Complete Print

CHEMGX 7 (1) 1176 (2012) ISSN 1860-7179 Vol. 7 No.
1 January, 2012
A Journal of
1/2012
Communication: Rofecoxib-Based Anti-in ammatory Agents (E. E. Knaus) Full Paper: TASK-3 Antagonists (C. A. Coburn)
www.chemmedchem.org
The inside cover picture shows a homology model of the catalytic domain of matriptase-2, a protease that plays a key role in human iron homeostasis. A substrate-analogue inhibitor was docked into the active site of a mutated variant of matriptase-2. A detailed view of the complex is presented on the right. The mutation is highlighted in green. For more details, please see the Communication by Michael Gtschow et al. on p. 68 ff.
01/2012
ChemMedChem, European in origin but international in scope, deals with all aspects of drug discovery. It is co-owned by Chemistry Publishing Society Europe (ChemPubSoc Europe) and is published by Wiley-VCH. Contributions in ChemMedChem cover medicinal and pharmaceutical sciences, drug design, drug development and delivery, molecular modeling, combinatorial chemistry, target validation, lead generation, and ADMET studies, that is, research from the overlapping areas between biology, chemistry, and medicine. ChemMedChem publishes Communications and Full Papers, as well as Reviews, Minireviews, Highlights, Concepts, Essays, Book Reviews, and occasionally Conference Reports. Authors can submit manuscripts to ChemMedChem online through our homepage (see over) by clicking on Submit an Article and following the simple instructions. Most of the articles in this issue have already appeared online on wileyonlinelibrary.com. See www.chemmedchem.org under Early View
Discover this journal online at:
wileyonlinelibrary.com
DOI: 10.1002/cmdc.201100462
Quality Research. Outstanding Publications.

Looking Forward
We continue to be diverse in our scope, our readership and our authorship, with manuscripts coming from all over the world (Figure 2). While there are countries that dominate, this
With the excitement of the International Year of Chemistry

and Angewandte Chemie International Editions 50th anniversary behind us, we are looking towards what lies ahead in 2012 and beyond. ChemMedChem is now in its seventh volume, and the training wheels have certainly come off! With a current impact factor (IF) of 3.306 (two-year IF taken from ISI Journal Citation Report 2010), we are established as a leading journal in the medicinal chemistry community, publishing primary research in the field. Our submissions are growing at a pace that exceeds the growth of the journal (Figure 1). This has two contrasting effects in real terms: we are becoming more selective in the manuscripts we accept, but as a consequence, the quality of the work we publish is improving. The quality of our contributions can be seen at a glance by looking at the top-ten cited articles of 2010 (Table 1) and the top-ten downloaded papers of 2011 (Table 2).
ChemMedChem 2012, 7, 3 6
Figure 1. Manuscript submissions from 2005 to 2011 (&): 2005 submissions are from JuneDecember, and 2011 submissions are extrapolated from November based on average monthly submission rate of 47 from January October. The number of articles published in ChemMedChem (&): contribution types are limited to Full Papers, Communications, Reviews, Minireviews, Highlights, Essays and Concepts.
2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
MED
Table 1. The top-ten most cited ChemMedChem manuscripts of 2010.[a] First Author S. Peukert E. Vergara S. Shukla D. D. Robinson R. K. Haynes J. P. Ribeiro D. M. Krger S. Amslinger A. Stary T. Tomasic Topic Hedgehog signaling pathway inhibitors Anticancer therapies that target selenoenzymes Chemical modifications for siRNA therapeutics Understanding kinase selectivity Mechanism of action of artemisinins Lectin-based drug design Structure- and ligand-based VS protocols Functionality of a,b-unsaturated carbonyls Consensus model of the hERG K + channel Bacterial Mur ligase inhibitors Times cited[b] 16 14 12 12 12 11 10 10 10 9 Type[c] R F R F F F F C F F Reference[d] 5(4), 5(1), 5(3), 5(4), 5(8), 5(3), 5(1), 5(3), 5(3), 5(2), 500 96 328 618 1282 415 148 351 455 286
EDITORIAL
DOI[e] 10.1002/cmdc.201000011 10.1002/cmdc.200900370 10.1002/cmdc.200900444 10.1002/cmdc.200900501 10.1002/cmdc.201000225 10.1002/cmdc.200900476 10.1002/cmdc.200900314 10.1002/cmdc.200900499 10.1002/cmdc.200900461 10.1002/cmdc.200900449
[a] JanuaryDecember 2010. [b] Data from Thompson ISI. [c] C = Concept, F = Full Paper, R = Review. [d] Volume(issue), first page. [e] Digital object identifier (DOI) numbers can be resolved at http://dx.doi.org.
Table 2. The top-ten ChemMedChem manuscript downloads for 2011.[a] First Author S. Peukert J. L. Yap T.-m. Ou A. Bcker K.-H. Chang N. Gavande W. C. Floyd T. W. Bell F. Kratz A. Spannhoff Topic Hedgehog signaling pathway inhibitors ERK signaling pathway inhibitors G-quadruplexes as anticancer drug targets Drug-like property rules for carboxylate-containing drugs Histone demethylase inhibitors Isoflavones as positive GABAA receptor modulators Chemotherapeutic evaluation of a tubulysin analoguedendrimer conjugate Drugs for hepatitis C Prodrug strategies in anticancer chemotherapy Histone methyltransferase and demethylase inhibitors Type[b] R M R F C C C H R R Reference[c] 5(4), 500 6(1), 38 3(5), 690 5(12), 2102 6(5), 759 6(8), 1340 6(1), 49 5(10), 1663 3(1), 20 4(10), 1568 DOI[d] 10.1002/cmdc.201000011 10.1002/cmdc.201000354 10.1002/cmdc.200700300 10.1002/cmdc.201000355 10.1002/cmdc.201100026 10.1002/cmdc.201100120 10.1002/cmdc.201000377 10.1002/cmdc.201000334 10.1002/cmdc.200700159 10.1002/cmdc.200900301
[a] JanuaryOctober 2011. [b] C = Communication, F = Full Paper, H = Highlight, M = Minireview, R = Review. [c] Volume(issue), first page. [d] Digital object identifier (DOI) numbers can be resolved at http://dx.doi.org.
speaks more of these countries as powerhouses of research and development, rather than any regional affiliations of the journal. We pride ourselves on being a truly global publication.
A Delicate Balance
In a competition-driven field, speed is of particular importance, but it should not be at the sacrifice of quality. ChemMedChem has responded well to the demands of our authors and readers alike by ensuring our publication times remain rapid, while maintaining the superior quality that has become synonymous with the journal. Typically, the whole process from receipt to online publication of a manuscript takes around 70 days. This includes a fast but thorough peer-review period, with initial decisions taking just 21 days on average, and a professional copyediting and production process that usually takes fewer than 35 days from receipt of the final version of the article to online publication Typically, the whole in EarlyView. process from receipt With these sorts of figures, to online publication we ensure your important reof a manuscript sults reach the community in a takes around timely fashion while also pre70 days. serving the overall quality of
Figure 2. Origin of manuscripts submitted to ChemMedChem in 2011; the region of origin is based on the corresponding author address. Values shown are percentages.
the forum, an all-important factor for readers who are faced with increasing amounts of information.
A More Discerning Readership
Medicinal chemistry research is progressing, changing and

developing. With improved computational, synthetic and anaChemMedChem 2012, 7, 3 6
www.chemmedchem.org
In a competitiondriven field, speed is of particular importance, but it should not be at the sacrifice of quality.
lytical methods along with more sophisticated technology and instrumentation comes greater assay sensitivity, improved understanding of mechanisms of action and drugdrug interactions and greater insight into biological processes at work. In short, the average medicinal chemistry research program is no longer average, but rather exceptional: serendipity no longer plays a critical role in drug discovery, but rather compounds are designed based on scientific data: crystal structures, homology models, HTS leads and so forth. Just as the field has developed, so have our readers become more discerning. Interesting compounds with some encouraging in vitro data are fine, but alone are often too preliminary. Our readers want to see the innovation behind the choices made by the authors: Why this derivative? Why that substituent? Why this assay? Why that biological target? These are also some of the questions we as Editors ask ourselves when evaluating a manuscript, and we encourage our authors to do the same when drafting their manuscript and cover letter.
If you are interested in submitting a manuscript for one of these issues, or perhaps you have an idea for another Special Issue, were always happy to hear from you.
A Changing Landscape
here is no doubt that 2011 saw some major changes that have had significant impacts on researchers in industry and academia. Funding cuts, open-access mandates from funding agencies, big-pharma mergersall of these have had a were doing our direct impact on pharmaceutibest to adapt to the cal and medicinal chemistry, tougher landscape often making it tougher for scifacing our authors, entists in the field to carry out and to accommotheir research, and even more date the changing so to publish their findings. needs of our Prof. Giorgio Tarzia (University readers. of Urbino, Italy) and Prof. Rainer Metternich (Hoffmann La Roche, Switzerland) have more to say on this topic in the section entitled A Very Severe Climb; their combined experience provides a unique insight into the challenges that we face in the coming years. From an Editors viewpoint, were doing our best to adapt to the tougher landscape facing our authors, and to accommodate the changing needs of our readers.
Special Issues
pecial Issues are a great way to highlight key topics that have caught the imagination of the field, or to bring areas of research into the limelight that might otherwise escape notice. Here at ChemMedChem, we try to do both! After three very successful Special Issues in 2011, which included one on Rare and Neglected Diseases (issue 02) and one celebrating Women in Medicinal Chemistry (issue 04), we are looking forward to equally exciting Special Issues in 2012! Our readers will see issues dedicated to medicinal chemistry research in fields such as chemical proteomics, neurological and CNS-related diseases, and ion channels as therapeutic targets.
Two New Family Members

hemMedChems co-owners, Wiley-VCH and ChemPubSoc Europe, a group of 16 European Chemical Societies, have recently added two new journals to their portfolio. ChemPlusChem is a genuinely multidisciplinary journal of chemistry and materials publishing collaborative research results in the field of chemistry plus one other scientific discipline. ChemistryOpen is the first society-owned chemistry journal. A gold-road, open-access journal, ChemistryOpen hopes to lead the way in offering a highquality platform for authors who want or are obliged to publish their primary research results in an open forum. A unique feature of this new journal is the ChemistryOpen Thesis Treasury, which will publish brief summaries of PhD theses, creating a fully citable and searchable document with a link to the full thesis. Full details for both ChemPlusChem and ChemistryOpen can be found on their websites at http://www.chempluschem.org and http://www.chemistryopen.org, respectively.
www.chemmedchem.org
MED
Here to Help!
EDITORIAL Finally, feel free to drop us an e-mail at the Editorial Office; our friendly staff will be happy to help in any way they can. We welcome your comments, your constructive criticisms, and of course, your next high-impact manuscript.
As Editors, we serve the community and its members, which

means you! If you have a question regarding the technicalities of the publication process, we refer you to our Notice to Authors available in full via our homepage, but also condensed and refined for your information on p. 1920 of this issue. If you are interested in the inner workings of the overall the process, we suggest you read our article From Submission to Publication: Demystifying the Publication Process, also available under our homepage via the Author Guidelines link, or the very interesting Editorial entitled Ten Tips for Authors recently published in Chemistryan Asian Journal (DOI: 10.1002/ asia.201100829).
Dr. Natalia Ortzar Editor-in-Chief ChemMedChem
Dr. Scott D. Williams Sr. Assoc. Editor ChemMedChem
A Very Severe Climb It is clear that the drug-discovery process is slowing down due to a mixture of political, economic, regulatory, managerial and scientific factors. An editorial by Rudy M. Baum entitled Changing Pharma Paradigms [Chem. Eng. News 2011, 89, 3] reported briefly on some of these factors, echoing other recently published comments on the current state of the pharmaceutical industry. Pharma paradigms are changing, and not always for the better, but is there something that we as medicinal chemists can do to help improve the present unhappy situation? Not much in the short term, but we, as a scientific community, should at least recognize that at the basis of the current revolution and involution of the pharmaceutical industry, there is also a shortage of fundamental knowledge that prevents the discovery of truly novel medicines. Regrettably, we do not know enough about the most serious conditions that plague our world, and in some cases we do not have reliable pharmacological tools to test apparently reasonable hypotheses regarding their cause and treatment. It is clear that while Pharma, which is both profit- and knowledge-driven, can retreat from research deemed too risky, Academia, which is entirely knowledge-driven, must face its institutional role more effectively and undergo a serious rethinking of its paradigms. Regardless of the field in which we are engaged, we must devote greater effort to increase our fundamental knowledge and to better understand the diseases for which we wish to design treatments. To do this, medicinal chemists must seek and stimulate stronger collaborations with related disciplines such as biochemistry, pharmacology, and translational medicine, as well as with any other areas of research that are able to offer synergy. Additionally, medicinal chemistry needs to focus more on new approaches to address the identification of diseaserelevant targets and off-targets of hit and/or lead compounds, e.g. through chemical proteomics, early in the drug-discovery process in order to increase the probability of success of drug-discovery projects. A recent article examined this matter thoroughly [What
Do Medicinal Chemists Actually Make? A 50-Year Retrospective, W. P. Walters, J. Green, J. R. Weiss, M. A. Murcko, J. Med. Chem. 2011, 54, 64056416]. In short, medicinal chemistry must become a more fundamental branch of science than it currently is. The recent Noble Prizes in Chemistry and Medicine are a good example of how research, initially thought to be without practical applications, can open new doors to exciting and important science. Individual scientists, the medicinal chemistry community, and scientific journals have a role in the realization of this shift that is a precondition to the revitalization of the drug-discovery process. The emergence of drug resistance and the existence of untreatable diseases are areas too important to be forgotten. Sooner or later we will be able to treat them, and private or institutional funders will again be forthcomingthis will all be made easier as our basic knowledge progresses. In the meantime, those who can act will have to do their best.
Prof. Rainer Metternich F. HoffmannLa Roche Ltd (Switzerland)
Prof. Giorgio Tarzia Universit degli Studi di Urbino (Italy)
www.chemmedchem.org
COVER PICTURE
The cover picture shows the class III anti-arrhythmic agent E-4031, which affects heart rhythm by interacting with the hERG K+ channel resulting in prolongation of the QT interval in the electrocardiogram. In this study, a series of E-4031 derivatives (indicated by arrows) was designed, synthesized and tested for their inhibition of the hERG K+ channel so as to establish structureactivity relationships (SARs). Knowledge gained by this study can be used in the early stages of drug discovery and development to avoid or circumvent hERG K+ channel blockade, thereby reducing the risk of cardiotoxicity, ventricular arrhythmias and sudden death. For more details, see the Full Paper by Adriaan P. IJzerman et al. on p. 107 ff.
NEWS
Spotlights on our sister journals 16 18
REVIEWS
For decades, quaternary ammonium compounds have been widely used as antiseptics and disinfectants. Further development of bisquaternary ammonium compounds has led to a large group of antibacterial and antimalarial agents. Most of these are characterized by a defined mode of action rather than by nonspecific interactions with the outer membranes of the microorganisms. M. Tischer, G. Pradel, K. Ohlsen, U. Holzgrabe* 22 31 Quaternary Ammonium Salts and Their Antimicrobial Potential: Targets or Nonspecific Interactions?
MINIREVIEWS
Microtubules and microprocessors: Computational approaches have provided useful insight for anticancer drug discovery efforts in recent years. We present an overview of the role played by computational methods in anti-tubulin research, particularly as it pertains to colchicine binding agent research. We discuss recent results, highlighting the challenges and opportunities faced by scientists in this field. A. Massarotti,* A. Coluccia, R. Silvestri, G. Sorba, A. Brancale 33 42 The Tubulin Colchicine Domain: a Molecular Modeling Perspective
MED
COMMUNICATIONS
X. Xia, C. F. Zhou, L. Ballell, A. E. Garcia-Bennett* 43 48 In vivo Enhancement in Bioavailability of Atazanavir in the Presence of Proton-Pump Inhibitors using Mesoporous Materials Matters of the HAART! The current treatment for human immunodeficiency virus (HIV), HAART, makes use of a combination of antiretroviral drugs, which are poorly soluble in aqueous media. Enhancing the solubility of such drugs through the use of mesoporous materials could lead to improved treatment efficiency and might provide a solution to the drugdrug interaction problems associated with these types of therapeutic regimes.
Y. Huang, S. Wolf, D. Koes, G. M. Popowicz, C. J. Camacho, T. A. Holak, A. Dmling* 49 52 Exhaustive Fluorine Scanning toward Potent p53Mdm2 Antagonists
Fluorine dance: We discovered potent p53Mdm2 antagonists by systematically varying the fluorine substitution pattern around a benzyl group that undergoes stacking interactions with His 96 of Mdm2. The potency of the optimized enantiomer (S)-7 e is > 50-fold better than the worst compound of the series. All compounds were efficiently synthesized by Ugi multicomponent reaction chemistry.
H. Park,* S. Hong, S. Hong* 53 56 Nocodazole is a High-Affinity Ligand for the Cancer-Related Kinases ABL, c-KIT, BRAF, and MEK
New from old: Nocodazole displays high affinity for a set of cancer-related kinases, including ABL, c-KIT, BRAF, and MEK. These experimental and computational findings may provide insight into the design of new and potent inhibitors of common kinases based on the nocodazole scaffold.
K. Splith, R. Bergmann, J. Pietzsch, I. Neundorf* 57 61 Specific Targeting of Hypoxic Tumor Tissue with NitroimidazolePeptide Conjugates Nitroimidazolepeptide conjugates: Attachment of a nitroimidazole unit to a cell-penetrating peptide was used to create conjugates able to target hypoxic tumor tissue. In vivo studies have demonstrated the successful delivery of these conjugates to even less well-perfused hypoxic regions in solid tumors. This approach provides a targeted delivery system with the option to use it as an imaging agent or to conjugate it with therapeutics.
www.chemmedchem.org
Expertly tailored! The development of safe anti-inflammatory (AI) agents that are highly selective COX-2 inhibitors has been challenging, as indicated by the clinical withdrawal of rofecoxib, valdecoxib, and lumiracoxib. We report novel rofecoxib analogues with a sulfohydroxamic acid nitric oxide (NO) donor COX-2 pharmacophore that exhibit potent AI activity. The release of NO is expected to circumvent the contraACHTUNGREindicated cardiovascular effects of ACHTUNGRErofecoxib.
A. Bhardwaj, Z. Huang, J. Kaur, E. E. Knaus* 62 67 Rofecoxib Analogues Possessing a Nitric Oxide Donor Sulfohydroxamic Acid (SO2NHOH) Cyclooxygenase-2 Pharmacophore: Synthesis, Molecular Modeling, and Biological Evaluation as Anti-inflammatory Agents
Matriptase-2 is a type II transmembrane serine protease that functions as key regulator of body iron homeostasis. To obtain insights into substrate/inhibitor enzyme interactions, site-directed mutagenesis, kinetic and molecular modeling studies were performed. Based on the active site structure of the related enzyme matriptase, amino acids that enhanced (Phe 665) or reduced (Asp 785, Tyr 712) the affinity of peptide ligands for matriptase-2 were identified.
E. Maurer, M. T. Sisay, M. Stirnberg, T. Steinmetzer, J. Bajorath, M. Gtschow* 68 72 Insights into Matriptase-2 Substrate Binding and Inhibition Mechanisms by Analyzing Active-Site-Mutated Variants
Stopping resistance in tuberculosis: The unusually regioversatile acetyltransferase Eis is a cause of resistance to ACHTUNGREkanamycin A in cases of tuberculosis. En route to new tuberculosis treatments, several inhibitors of the Eis enzyme were identified and characterized.
K. D. Green, W. Chen, S. Garneau-Tsodikova* 73 77 Identification and Characterization of Inhibitors of the Aminoglycoside Resistance Acetyltransferase Eis from Mycobacterium tuberculosis
FULL PAPERS
SELEX-tive uptake: The oligonucleotide KMF2-1a is one of a panel of aptamers that specifically recognize the MCF10AT1 breast cancer cell line, with Kd values in the nanomolar range. KMF2-1a binds specifically and very stably to target cells. Furthermore, this aptamer can bind to target cells at both 4 and 37 8C, extending its use for in vivo experimentation. K. Zhang, K. Sefah, L. Tang, Z. Zhao, G. Zhu, M. Ye, W. Sun,* S. Goodison, W. Tan* 79 84 A Novel Aptamer Developed for Breast Cancer Cell Internalization
www.chemmedchem.org
MED
K. J. Castor, J. Mancini, J. Fakhoury, N. Weill, R. Kieltyka, P. Englebienne, N. Avakyan, A. Mittermaier, C. Autexier,* N. Moitessier,* H. F. Sleiman* 85 94 Platinum(II) Phenanthroimidazoles for Targeting Telomeric G-Quadruplexes A rational progression of p-extended phenanthroimidazole PtII complexes were synthesized. Their interactions with and relative binding affinities for human telomere-derived quadruplex DNA were studied by CD, FID assays, molecular modeling, and a TRAP assay. Significant binding affinity and selectivity for quadruplex DNA over duplex DNA was observed; this led to promising preliminary cytotoxicity studies for [(CLIP)Pt(en)] in cancer cells.
D. Spinks, L. S. Torrie, S. Thompson, J. R. Harrison, J. A. Frearson, K. D. Read, A. H. Fairlamb, P. G. Wyatt, I. H. Gilbert* 95 106 Design, Synthesis and Biological Evaluation of Trypanosoma brucei Trypanothione Synthetase Inhibitors
Tipping the HAT: Thiazole methylsulfones and tetrahydroindazole methylamides were explored as potent inhibitors of the Trypanosoma brucei trypanothione synthetase (TbTryS). Scaffold optimisation of these series afforded nanomolar TbTryS inhibitors which act on target and provide useful leads to further explore the trypanothione pathway in kinetoplastids.
M. Vilums, J. Overman, E. Klaasse, O. Scheel, J. Brussee, A. P. IJzerman* 107 113 Understanding of Molecular Substructures that Contribute to hERG K + Channel Blockade: Synthesis and Biological Evaluation of E-4031 Analogues
Go-go dancing with hERG! Evaluation of derivatives of E-4031, a class III antiACHTUNGREarrhythmic agent, provides practical information on the molecular determinants for hERG K + channel blockade. Chemical features that are likely to increase affinity for the hERG K + channel can be omitted (!) during drug development; likewise, structural elements that reduce affinity (!) could be included to potentially circumvent cardiotoxicity related attrition at a later stage.
M. L. Huang, S. B. Y. Shin, M. A. Benson, V. J. Torres, K. Kirshenbaum* 114 122 A Comparison of Linear and Cyclic Peptoid Oligomers as Potent Antimicrobial Agents
Cycling builds endurance! A family of linear N-substituted glycine peptoid oligomers bearing cationic and hydrophobic side chains were subjected to macrocyclization, thereby enhancing antimicrobial potency. These compounds are active against Gram-negative and Gram-positive bacteria and are nonhemolytic toward human erythrocytes.
10
www.chemmedchem.org
A molecular TASKmaster! A 5,6,7,8tetrahydropyridoACHTUNGRE[4,3-d]pyrimidine highthroughput screening lead was optimized to give a novel series of potent and selective TWIK-related acid-sensitive K + (TASK-3) channel antagonists. Compound 23 demonstrated pharmacological target engagement as indicated by its ability to modulate sleep architecture in rodent electroencephalogram (EEG) telemetry models.
C. A. Coburn,* Y. Luo, M. Cui, J. Wang, R. Soll, J. Dong, B. Hu, M. A. Lyon, V. P. Santarelli, R. L. Kraus, Y. Gregan, Y. Wang, S. V. Fox, J. Binns, S. M. Doran, D. R. Reiss, P. L. Tannenbaum, A. L. Gotter, P. T. Meinke, J. J. Renger 123 133 Discovery of a Pharmacologically Active Antagonist of the Two-PoreDomain Potassium Channel K2P9.1 (TASK-3)
Sialic acid derivatives, already known to bind MAG (Siglec-4) with high affinity, were screened for their binding affinity to CD22 (Siglec-2) by SPR. The best compound identified was further modified with various hydrophobic substituents at the 2-, 5-, and 9-positions of the sialic acid scaffold, leading to nanomolar derivatives, of which ligand 17 b shows the most promising pharmacodynamic and pharmacokinetic profiles.
S. Mesch, K. Lemme, M. Wittwer, H. Koliwer-Brandl, O. Schwardt, S. Kelm, B. Ernst* 134 143 From a Library of MAG Antagonists to Nanomolar CD22 Ligands
Modified aspirins: Aspirin is a nonselective cyclooxygenase (COX) inhibitor that exhibits moderate anti-inflammatory (AI) activity. Replacement of the acetoxy group in aspirin by SO2NHOH provides compounds that are dual inhibitors of COX-2 and 5-lipoxygenase (5-LOX), release nitric oxide at pH 7.4, and which exhibit potent AI activity. Binding interactions were identified by docking into the binding sites of COX-2 and 5-LOX.
J. Kaur, A. Bhardwaj, Z. Huang, E. E. Knaus* 144 150 Aspirin Analogues as Dual Cyclooxygenase-2/5-Lipoxygenase Inhibitors: Synthesis, Nitric Oxide Release, Molecular Modeling, and Biological Evaluation as AntiInflammatory Agents
Antiprotozoals from herbicides: A library screen revealed a 1,3-diiminoisoindoline carbohydrazide derivative to be a potent inhibitor of the in vitro proliferation of P. falciparum in red blood cells. A variety of derivatives, including FeIII complexes, were synthesized and displayed antimalarial activity down to the double-digit nanomolar IC50 range.
P. Mombelli, M. C. Witschel,* A. W. van Zijl, J. G. Geist, M. Rottmann, C. Freymond, F. Rhl, M. Kaiser, V. Illarionova, M. Fischer, I. Siepe, W. B. Schweizer, R. Brun, F. Diederich* 151 158 Identification of 1,3-Diiminoisoindoline Carbohydrazides as Potential Antimalarial Candidates
www.chemmedchem.org
11
MED
P. M. Toth, S. Naruhn, V. F. S. Pape, S. M. A. Drr, G. Klebe, R. Mller, W. E. Diederich* 159 170 Development of Improved PPARb/d Inhibitors Ten-times better! Structural optimization of the PPARb/d-selective inhibitor GSK0660 (1) revealed that replacement of the 4-aminophenyl substituent by medium-length n-alkyl chains is most beneficial for an increased binding afACHTUNGREfinity. The compounds presented here show activity down to the one-digit nanomolar range representing up to a tenfold higher binding affinity compared with 1.
* Author to whom correspondence should be addressed. Supporting information on the WWW (see article for access details). This article is available online free of charge (Open Access). A video clip is available as Supporting Information on the WWW (see article for access details). Contributions labeled with this symbol have been judged as Very Important Papers by the referees.
BOOKS
Analogue-Based Drug Discovery II J. Fischer, C. R. Ganellin (Eds.) Antiviral Drugs: From Basic Discovery Through Clinical Trials W. M. Kazmierski (Ed.) H. E. Burks, S. Peukert . . . . . . . . . . . . . . . . . . . . 171 M. Rowley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
SERVICE
Preview ... .. .. ... .. .. .. ... .. .. ... .. .. ... .. .. .. ... .. .. ... .. .. ... .. 175 All the Tables of Contents may be found on the WWW under: http://www.chemmedchem.org
Issue 12, 2011, was published online on December 2, 2011.
12
www.chemmedchem.org
MED
Editor: Natalia Ortuzar Senior Associate Editor: Scott D. Williams
Editorial Board
Chairmen Giorgio Tarzia (Italy) Rainer Metternich (Switzerland) Members Karl-Heinz Altmann (Switzerland) Paul A. Bartlett (USA) Maurizio Botta (Italy) Gloria Cristalli (Italy) Harren Jhoti (UK) Gerhard Klebe (Germany)
Managing Editor: Lisa Abel Editorial Assistant: Jean-Pierre Parks Production: Stefanie Klein, Denis Ott
Martin Stahl (Switzerland) Hans Ulrich Stilz (Germany) Renxiao Wang (China) Yasuyoshi Watanabe (Japan) David A. Winkler (Australia) Steven Young (USA)
Secretary: Luisa Prieto Telephone: (+ 49) 6201-606-142 Fax: (+ 49) 6201-606-331 or -328 E-mail: chemmedchem@wiley.com Courier Services: Boschstrae 12 69469 Weinheim (Germany) Regular Mail: Postfach 10 11 61 69451 Weinheim (Germany) Homepage: http://www.chemmedchem.org Manuscript Submission & Personal Homepage
International Advisory Board

Christopher Abell (UK) Geo Adam (Switzerland) Ivano Bertini (Italy) William R. Bishai (USA) Maria Laura Bolognesi (Italy) Maria Jose Camarasa (Spain) Bernd Clement (Germany) Paul J. Coleman (USA) Manoj C. Desai (USA) Franois Diederich (Switzerland) Romano Di Fabio (Italy) Edmond Differding (Belgium) Beat Ernst (Switzerland) Gert Fricker (Germany) Arun K. Ghosh (USA) Athanassios Giannis (Germany) Sylviane Giorgi-Renault (France) Ernest Giralt (Spain) William A. Goddard (USA) Concepcion Gonzlez-Bello (Spain) Gilles Gosselin (France) Pilar Goya Laza (Spain) Stephen Hanessian (Canada) Michael M. Hann (UK) Adriaan P. IJzerman (The Netherlands) Hualiang Jiang (China) Kazunori Kataoka (Japan) Thomas Keller (Singapore) Alan P. Kozikowski (USA) Karen Lackey (USA) Dominique Lesuisse (France) Steven V. Ley (UK) Dawei Ma (China) Ulf Madsen (Denmark) Antonello Mai (Italy) ter Pe Matyus (Hungary) Lieven Meerpoel (Belgium) Marco Mor (Italy) Stefano Moro (Italy) Sylviane Muller (France) Klaus Mller (Switzerland) Rolf Mller (Germany) Peter Nussbaumer (Germany) Paul M. O'Neill (UK) Tudor I. Oprea (USA) Eckhard Ottow (Germany) Michael Rowley (Sweden) Thomas Ryckmans (UK) Jagarlapudi A. R. P. Sarma (India) Carlo Scolastico (Italy) Henning Steinhagen (Germany) Ralph Weissleder (USA) Hanno Wild (Germany) Paul Wyatt (UK)
http://www.manuscriptxpress.com Subscriptions: Customer Services (cs-germany@wiley.com) Marketing: Monika Silz (ext. 458, fax: ext. 100, e-mail: msilz@wiley.com) Advertising: Marion Schulz (ext. 565, fax: ext. 550, e-mail: mschulz@wiley.com, jspiess@wiley.com) Copyright Permissions: Bettina Loycke (ext. 280, fax: ext. 447, e-mail: rightsde@wiley.com) Reprints, e-prints, posters: Carmen Leitner (fax: ext. 226) e-mail: chem-reprints@wiley.com Publishers: WILEY-VCH Verlag GmbH & Co. KGaA, Postfach 10 11 61 69451 Weinheim (Germany) http://www.wiley-vch.de/ Founding Editor: Peter Glitz Coordinating Managing Editor: Karen J. Hindson Publication: 12 issues in 2012 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim. Further masthead information follows directly after the Table of Contents.
ChemMedChem succeeds IL FARMACO, ` a journal of the Societa Chimica Italiana
ChemPubSoc Europe
G
16 societies G 11 journals G 1 network
5.476
Impact Factor:
2.909
Impact Factor:
3.206
Impact Factor:
3.345
Impact Factor:
3.945
Impact Factor:
3.306
Impact Factor:
3.339
Impact Factor:
6.325
Impact Factor:
New!
New!
E-magazine for chemists
G G G G G
H High Impact Factor journals Short publication times S Excellent authors E Top editorial services T Supported by 73,000 society S members
C Closely associated with
12.730
Impact Factor:
PSJ-11-36067_gu
www.ChemistryViews.org
www.chempubsoc.eu
ChemMedChem is indexed in Biological Abstracts (Thomson ISI), BIOSIS Previews (Thomson ISI), Cambridge Structural Database (CSA/CIG), CAS: Chemical Abstracts Services (ACS), Chemical Abstracts Service/SciFinder (ACS), ChemInform (Wiley-VCH/FIZ Chemie), COMPENDEX (Elsevier), Current Contents/Life Sciences (Thomson ISI), EMBASE (Elsevier), EMNursing (Elsevier), GEOBASE/Geographical & Geological Abstracts (Elsevier), IBIDS: International Bibliographic Information on Dietary Supplements, Index Medicus/MEDLINE/ PubMed (NLM), International Pharmaceutical Abstracts (Thomson ISI), Journal Citation Reports/Science Edition (Thomson ISI), MEDLINE/PubMed (NLM), Mosby Yearbooks (Elsevier), Science Citation Index (Thomson ISI), Science Citation Index Expanded (aka SciSearch) (Thomson ISI), SCOPUS (Elsevier), VINITI (All-Russian Institute of Science & Technological Information) and Web of Science (Thomson ISI). When citing work presented herein, please ensure the abbreviation ChemMedChem is used.
MED
Subscription rates for 2012 In 2007 a new, tiered pricing model was introduced for new journals. The appropriate rate for an institutional subscription is based on the number of Full Time Equivalents (FTE) at the institution purchasing the journal. Customer Type* Institutional Rates for ChemMedChem in 2012 small medium large 10 000 40 000 40 000+ 1 000 10 000 10 000+ only) 4832 / 4201 7564 / 6577 6220 / 5409 FTE Academic Customer FTE Corporate/Government Customer Subscription Rate (print & electronic / print only or electronic Europe E 2621 / 2278 3487 / 3032 Switzerland SFr 4084 / 3551 5438 / 4728 Rest of the World US$ 3364 / 2925 4571 / 3974 *Rates for hospitals upon request.
ISSN 1860-7179 (Print) ISSN 1860-7187 (Online) registered with their local Reproduction Rights Organisation (RRO), e.g. Copyright Clearance Center (CCC), 222 Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), provided the appropriate fee is paid directly to the RRO. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale. Special requests should be addressed to: permissionsuk@wiley.com
Open Access: ChemMedChem accepts articles for Open Access publication. Please see www.chemmedchem. org and select Online Open. Access ChemMedChem from your desktop. The journal is available through Wiley Online Library at wileyonlinelibrary.com
Personal rates for members of a national chemical society, print only (including postage and handling charges): Europe: E 155.00. Individual subscription rates are available on request. All Wiley-VCH prices are exclusive of VAT. Postage and handling charges included. Prices are subject to change. Individual rate subscriptions may not be resold or used as library copies. Orders may be placed through your bookseller or directly at the publishers; for readers in Germany, Switzerland, and Austria: WILEY-VCH, Customer Service Journals, Postfach 101161, 69451 Weinheim (Germany); Fax: (+49) 6201-606-184; E-mail: service@wiley-vch.de for readers in the rest of the world: John Wiley and Sons Ltd., Journals Administration Dept., 1 Oldlands Way, Bognor Regis, West Sussex, PO22 9SA (England); Tel.: (+44) 1234-779-777; Fax: (+44) 1234-843-232; E-mail: cs-journals@wiley.co.uk Delivery: By direct mail (printed matter) or through book-sellers. * Change of address: Please notify your bookseller or the publishers immediately. * Cancellation of subscriptions: The publishers must be notified not later than three months before the end of the calendar year. Registered names, trademarks, etc., used in this journal, even when not marked as such, are not to be considered unprotected by law. Responsibility: This journal was carefully produced in all its parts. Nevertheless, the Publisher, Editors and Authors cannot be held responsible for errors or any consequences arising from the use of information contained in this journal; the views and opinions expressed do not necessarily reflect those of the Publisher, and Editors neither does the publication of advertisements constitute any endorsement by the Publisher, and Editors of the products advertised. Copyright and Photocopying Copyright 2012 Wiley-VCH Verlag GmbH & Co. KGaA. All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by any means without the prior permission in writing from the copyright holder. Authorization to photocopy items for internal and personal use is granted by the copyright holder for libraries and other users
Articles are published online days or even weeks before the Delivery Terms and Legal Title Where the subscription price includes printed issues are mailed, print issues and delivery is to the rethrough the wiley online cipients address, delivery terms are Delivered Duty Unpaid (DDU); the re- library service. A full-rate subcipient is responsible for paying any scription entitles users to import duty or taxes. Title to all issues online access to full text plus transfers FOB our shipping point, freight prepaid. We will endeavour to access to the tables of confulfil claims for missing or damaged tents and abstracts of huncopies within six months of publication, within our reasonable discretion dreds of other Wiley journals. and subject to availability. Speak to your librarian or visit For the USA and Canada: the site for full details.
ChemMedChem (ISSN 1860-7179) is published monthly by Wiley-VCH, PO Box 191161, 69451 Weinheim, Germany. Periodical postage pending at Jamaica, NY 11431. Air freight and mailing in the USA by Publications Expediting Services Inc., 200 Meacham Ave., Elmont, NY 11003. US POSTMASTER: Send address changes to ChemMedChem, c/o John Wiley & Sons Inc., 350 Main St., Malden, MA 02148-5020. Wileys Corporate Citizenship initiative seeks to address the environmental, social, economic, and ethical challenges faced in our business and which are important to our diverse stakeholder groups. We have made a long-term commitment to standardize and improve our efforts around the world to reduce our carbon footprint. Follow our progress at www.wiley.com/go/citizenship
Typesetting, Printing and Binding:
Johannes-Gutenberg-Strasse 13 97199 Ochsenfurt-Hohestadt www.triltsch.de
Paper: acid-free Printed in the Federal Republic of Germany
Authors who wish to publish in ChemMedChem should consult the current Notice to Authors, which appears on the ChemMedChem homepage at http://www.chemmedchem.org, before they draft their manuscript.
ChemMedChem 2012, 7, 15
15
SPOTLIGHTS ...
On these pages, we feature a selection of the excellent work that has recently been published in our sister journals. If you are reading these pages on a
computer, click on any of the items to read the full article. Otherwise please see the DOIs for easy online access through Wiley Online Library.
Biogenesis
A. Eschenmoser* Etiology of Potentially Primordial Biomolecular Structures: From Vitamin B12 to the Nucleic Acids and an Inquiry into the Chemistry of Lifes Origin A Retrospective The quest for the etiology of a biomolecular structure acquires special significance when the questions asked refer to molecules the existence of which are fundamental to life, and in particular to the origin of life. In a search for the chemistry of emergence of life, one needs to pay strict heed to molecular guidesnucleic acids, proteins, cofactorsthat all carry a message which it is our job to decipher.
Angew. Chem. Int. Ed. DOI: 10.1002/anie.201103672
Supramolecular Chemistry
B. J. Slater, E. S. Davies, S. P. Argent, H. Nowell, W. Lewis, A. J. Blake, N. R. Champness* A Perylene Diimide Rotaxane: Synthesis, Structure and ElectroACHTUNGREchemically Driven De-Threading Perylene diimides make excellent building blocks for the formation of [2]-rotaxanes. The rich electrochemistry of the perylene-based recognition site facilitates a pathway to different oxidation states and properties and allows a mechanism for electrochemically driven de-treading of the interlocked species (see figure).
Chem. Eur. J. DOI: 10.1002/chem.201103090
Nitrogen Heterocycles
T. M. Klaptke,* B. Krumm, F. A. Martin, J. Stierstorfer New Azidotetrazoles: Structurally Interesting and Extremely Sensitive Hypersensitivity: 1-Amino-5-azidotetrazole (1), 5-azido-1-diazidocarbamoyltetrazole (2), and 1-(amino-azidocarbamoyl)-5-azidotetrazole (3) were formed by the diazotation of triaminoguanidinium chloride and separated by short-column chromatography. Their high nitrogen content results in extremely high sensitivity, therefore handling and characterization were very challenging.
Chem. Asian J. DOI: 10.1002/asia.201100632

16
... ON OUR SISTER JOURNALS

Immunology
A. A. Khan, S. H. Chee, R. J. McLaughlin, J. L. Harper, F. Kamena, M. S. M. Timmer,* B. L. Stocker* Long-Chain Lipids Are Required for the Innate Immune Recognition of Trehalose Diesters by Macrophages Going to any length? Trehalose diesters of various chain lengths have been synthesised in order to determine the effect of lipid length on innate immune recognition, as determined by NO (see graph) and cytokine production by macrophages. In this work, we show that longer lipids (C20C26) are required for macrophage activation, with C22 giving optimal activity.
ChemBioChem DOI: 10.1002/cbic.201100451
Hyperpolarized Gases
N. Amor,* K. Hamilton,* M. Kppers, U. Steinseifer, S. Appelt, B. Blmich, T. Schmitz-Rode NMR and MRI of Blood-Dissolved Hyperpolarized Xe-129 in Different Hollow-Fiber Membranes In the blood: Hollow-fiber membranes for continuous dissolution of hyperpolarized xenon gas into blood are investigated via NMR spectroscopy and imaging with a xenonizer setup (see picture). Spatially resolved functionality is analyzed and a comparison of different fiber materials is presented.
ChemPhysChem DOI: 10.1002/cphc.201100446
Water oxidation
N. H. Chou, P. N. Ross, A. T. Bell,* T. D. Tilley* Comparison of Cobalt-based Nanoparticles as Electrocatalysts for Water Oxidation Water oxidation electrocatalysts: The controlled synthesis of e-Co, CoO, and Co3O4 nanoparticles with nearly identical particle size and shape allows comparisons of the inherent catalytic properties of these materials in the oxygen evolution reaction (OER). The nanoparticle electrodes exhibit relatively low overpotentials and very similar catalytic activities under basic conditions.
ChemSusChem DOI: 10.1002/cssc.201100075
Oxidation Catalyst
E. A. Eilertsen, S. Bordiga,* C. Lamberti, A. Damin, F. Bonino, B. Arstad, S. Svelle, U. Olsbye, K. P. Lillerud* Synthesis of Titanium Chabazite: A New Shape Selective Oxidation Catalyst with Small Pore Openings and Application in the Production of Methyl Formate from Methanol Take CHAnce on zeolites: A new titanium silicate oxidation catalyst (Ti-CHA) and a bifunctional titanium aluminum silicate catalyst (TiAl-CHA) with the chabazite (CHA) topology have been synthesized. The materials have a 3-dimensional channel system with small pore openings that enable shape selective oxidation catalysis. The new Ti-Al-CHA material facilitates a high conversion of methanol to methyl formate with a selectivity of 85 % at 60 8C, using H2O2 as the oxidation agent.
ChemCatChem DOI: 10.1002/cctc.201100281

www.chemmedchem.org
17
SPOTLIGHTS
MetalAlkyl Homolysis
P. H. M. Budzelaar* Radical Chemistry of Iminepyridine Ligands Ligands containing at least two imine or pyridine ligands in conjugation reduce the energy required for metalalkyl homolysis dramatically; ligand alkylation reactions are best explained via free alkyl radicals. Eur. J. Inorg. Chem. DOI: 10.1002/ejic.201100698
Mushroom Extraction Artefacts

F. von Nussbaum, M. Rth, P. Spiteller, T. Hbscher-Weissert, F. Lbermann, K. Polborn, W. Steglich* Coloured Artefacts Formed by Oxidation of Benzene-1,2,4-triol and b-Dopa During the Extraction of Cortinarius violaceus (Agaricales) with Alcohols Extraction of the mushroom Cortinarius violaceus with methanol leads to the formation of several colourful artefacts. They arise from benzene-1,2,4-triol, b-dopa, and ironACHTUNGRE(III) ions, all of which are present in the fungus.
Eur. J. Org. Chem. DOI: 10.1002/ejoc.201101215
Hybrid Cars
ChemViews magazine Hybrid Cars Mass produced hybrids have come a long way since 2001 when the Prius was first launched in Japan. The Industry Roundup gives an overview of hybrid car sales by region, hybrid type, and brand. ChemViews magazine DOI: 10.1002/chemv.201000140
18
www.chemmedchem.org
MED
Important Basics for Authors
Contributions in ChemMedChem cover a wide range of topics within the field of medicinal chemistry. Naturally, ChemMedChem does not publish manuscripts that have already appeared elsewhere in any format. Authors must inform the Editor of manuscripts submitted, soon to be submitted, or in press at other journals that have a bearing on the manuscript being submitted to ChemMedChem. The Ethical Guidelines for Publication in Journals and Reviews issued by the European Association for Chemical and Molecular Sciences (EuCheMS) are followed and applied by the ChemMedChem Editorial Office; these guidelines are available at our website. Peer review is paramount to this journal. Except for Book Reviews, Author Profiles, Conference Reports, and similar articles, all manuscripts that are suitable for consideration will be peer reviewed and, if accepted for publication, edited for clarity, brevity, and consistency.
Concise Notice to Authors

should be 98 % or higher. We understand that these requirements cannot always be met, and we are able to make exceptions, for example, in cases for which the nature of the chemical synthesis precludes certain aspects of analysis, or where access to proper instrumentation is limited. Please contact the Editorial Office if you have any questions.
Supporting information Any Supporting Information should be appended to the end of the main document being submitted. This way, the referees will be able to access your manuscript and Supporting Information in a single file.
Preparing Your Manuscript for Submission

Visit www.chemmedchem.org Below we briefly discuss some key issues to consider when getting ready to submit your next paper to ChemMedChem. All specific information can be found by clicking the Author Guidelines link on ChemMedChems homepage; we strongly urge you to visit there, even if youve already published with us before.
The cover letter The importance of a concise and well-written cover letter is often underappreciated. Occasionally even skilled and seasoned researchers overlook this rather simple opportunity to put their most recent body of work in the best possible light (but no hype, please). Rather than a mere summary of results, cover letters should instead focus on justifying why the given manuscript should appear in ChemMedChem. You should declare any conflicts of interest in your manuscripts cover letter; you are also welcome to include referee suggestions.
Accepted Manuscripts
Revisions & final production data upload Notice of a manuscripts acceptance is always good news, and this is the point in particular when authors thank themselves for having prepared their manuscript as carefully as possible from the very start because the news comes with a list of required changes, both from the referees and the Editor. Its clear that the faster one is able to carry out revisionseither in preparation for a second round of peer review, or in finalizing the materials for upload to the copyediting stagethe sooner the work will be published. Therefore, please carry out any revisions, as requested by the referees and/or the Editorial Office, as quickly as possible.
Iron out any kinks as early as possible Double check your manuscript for proper organization. Making sure that all the graphics are numbered uniquely and in sequence, for example, will help ensure that confusion is avoided during peer review. A little extra time spent weeding out duplicate reference citations before submission will save a lot more time down the road, both for you and our copyeditors, thus minimizing the time to publication.
Manuscript templates & layout The use of ChemMedChems various manuscript submission templates is highly encouraged, as is the insertion of Tables, Schemes and Figures where they are first mentioned in the text. This ensures that your manuscript is as reader- and printer-friendly as possible, significantly facilitating the work of the referees. The aim is to efficiently present your work to the peer-review panel in a condensed yet legible manner. The templates are available under the Author Guidelines link on our homepage.
Preparing your production data Once you have your final version ready, we ask that you provide the text as a single word document without the template and without any embedded graphics. Provide all graphics as individual files in their original format (ChemDraw, PowerPoint, etc.), and send us any Supporting Information as a single PDF file, remembering to delete any title/author/affiliation details, as a cover sheet is automatically generated.
Analytical data The characterization of new compounds reported in Communications and Full Papers is important, and should be detailed in the Experimental Section or Supporting Information. Our minimum requirements include optical rotation (for optically active compounds), 1H and 13C NMR, IR, and HRMS data for all compounds evaluated biologically; the purity for such compounds
Color production The production of color graphics is expensive, and we ask that authors carry part of the additional costs. However, if you are unable to meet the associated costs and color is essential to the scientific understanding, then we are happy to assist you; in short, the science comes first! We ask that you contact
19
MED
the Editorial Office upon acceptance to discuss this with the Editor. Please note that the online and print versions of ChemMedChem are identical; we are unable to publish graphics in color online for a given article while publishing the same graphics in greyscale in print.
Concise Notice to Authors end of the paper alerting readers that minor changes have been made since the initial online publication. Significant errors, noted either after a manuscripts publication in EarlyView or in an issue, should be corrected in a Corrigendum. Corrigenda are published directly after the Table of Contents and should be kept as short as possible. We request that authors submit the Corrigendum electronically like any other article, and that they cite the publication to be corrected along with its digital object identifier (DOI).
Wiley OnlineOpen The Wiley OnlineOpen option allows authors of primary research articles to make their work available to non-subscribers of the journal. With Wiley OnlineOpen, you, your funding agency, or your institution pays a one-off fee of E2500, and the article is made available to everyone via Wiley Online Library. Authors of OnlineOpen articles are also permitted to post the final, published PDF of their article on a website, institutional repository, or other free publicly accessible server.
A Few Other Things

Book reviews Researchers working in all fields of medicinal chemistry are occasionally invited by the Editorial Office to write reviews for books that have been recently published. Book Reviews should include a brief summary of the books contents, a constructive critique of any shortcomings, and a balanced highlight of positive features; a short general introduction to the topic and a discussion of its relevance is encouraged.
For the special few All papers published in ChemMedChem have been successful in passing a thorough and rigorous peer-review process, and therefore reflect top-notch medicinal chemistry research. Among these, there are a select few that stand out with our referees as being Very Important Papers (VIPs). Authors of VIPs are often invited to supply a short synopsis of their work in which the important findings are highlighted in terms that laypeople and non-chemists can understand; the synopses are posted to the News page at ChemMedChems website, and are usually featured in ChemistryViews, Wileys online chemistry portal. Moreover, certain VIPs are chosen to be the subject of a press release. Each issue of ChemMedChem has a Front and Inside Cover, and any author may submit proposals for consideration. The Editors decision is based on a variety of criteria, including a manuscripts peer-review results and topic. We do ask for an author contribution to the additional production costs of the covers, and further details are available upon request.
Author profiles Authors who have published on average one manuscript per year since the journal began in 2006 may be invited by the Editor to submit a brief biographical sketch to appear in the journal. Author Profiles are intended to give our readers a more personal view of researchers who are making a significant impact in the medicinal chemistry community. If you would like to nominate an individual for profiling, please contact the Editorial Office; self-nominations are certainly possible.
Full Details Are Available Online

As stated above, a complete body of information for authors can be found under the Author Guidelines link on the ChemMedChem homepage. The Notice to Authors is indexed by topic, and contains all the pertinent information you should need for submitting your next manuscript. If you should experience any difficulties in uploading manuscripts online, please contact the Editorial Office at the E-mail address given below.
Corrections: Galley Proofs & Beyond

Although our copyeditors do their best to ensure each manuscript is 100 % error-free, on rare occasions, mistakes are overlooked or even made. Therefore, please scrutinize the galley proofs as carefully as possible, ensuring that all content in your paper is true and accurate. Be sure to respond clearly to all queries, which are marked in the text with black squares (e.g., &&Rewording OK?&&). As with all steps in the publication process, we encourage a prompt return of proof corrections from our authors (within 48h), but this should not come at the expense of careful proofreading. Errors recognized only after an article has been published online in EarlyView can be corrected. How the correction is carried out depends on the severity of the error. Minor typographical errors and misspellings are left alone as long as reader comprehension of the content remains unaltered. Minor scientifically incorrect or incomplete information in published articles can be corrected for the articles eventual publication in an issue. In these cases, a short text is added toward the
Contact Us!
Please feel free to contact the Editorial Office with any queries you might have on a future or current submission. We are always happy to assist our authors and readers alike, and look forward to hearing from you. For regular mail: ChemMedChem Postfach 10 11 61 69451 Weinheim Germany Telephone: Fax: E-mail: Homepage: For courier services: ChemMedChem Boschstrae 12 69469 Weinheim Germany
(+49) 6201-606-142 (+49) 6201-606-331 or -328 chemmedchem@wiley.com http://www.chemmedchem.org
20
www.chemmedchem.org
You are invited

... TO READ AND BROWSE:
The Chemical Record (TCR), the interdisciplinary review journal from The Chemical Society of Japan.
FEATURES: Record of Records: Keep up with all the latest world records in chemistry, and contribute an entry yourself from the literature EarlyView: All articles are published online prior to print All issues are available at www.tcr.wiley-vch.de Dont miss the latest articles register for RSS feeds and E-Mail alerts From the publishers of Angewandte Chemie and Chemistry An Asian Journal
ISSN: 1527-8999 print ISSN: 1528-0691 online 2011. Volume 11. 6 Issues.
EDITOR-IN-CHIEF
Hisashi Yamamoto (University of Chicago, USA) in cooperation with Brian Johnson (Wiley-VCH, Managing Editor)
www.tcr.wiley-vch.de
Watch out for newly published Records and Challenges articles and the latest world records in chemistry in the Record of Records!
For further information and to subscribe please send an e-mail to:
cs-journals@wiley.com (Americas, Europe, Middle East, Africa, Asia Pacific) service@wiley-vch.de (Germany, Austria, Switzerland) cs-japan@wiley.com (Japan)
2733401102_bu
MED
DOI: 10.1002/cmdc.201100404
Quaternary Ammonium Salts and Their Antimicrobial Potential: Targets or Nonspecific Interactions?
Maximilian Tischer,[a] Gabriele Pradel,[b] Knut Ohlsen,[b] and Ulrike Holzgrabe*[a]
Dedicated to Prof. Dr. Dr. h.c. G. Bringmann on the occasion of his 60th birthday.
22
For more than 50 years dequalinium chloride has been used successfully as an antiseptic drug and disinfectant, particularly for clinical purposes. Given the success of dequalinium chloride, several series of mono- and bisquaternary ammonium compounds have been designed and reported to have im-
proved antimicrobial activity. Furthermore, many of them exhibit high activity against mycobacteria and protozoa, especially against plasmodia. This review discusses the structureactivity relationships and the modes of action of the various series of (bis)quaternary ammonium compounds.
Introduction
Since the discovery in 1876 by Robert Koch that microorganisms cause infectious diseases, physicians, microbiologists, and medicinal chemists have been searching for drugs to prevent infections. As early as the 1930s quaternary ammonium compounds (QACs) were found to have antimicrobial activities; patents were immediately filed for this purpose.[1] QACs are the most useful antiseptics and disinfectants, and have been used for a variety of clinical purposes such as preoperative disinfection of unbroken skin, application to mucous membranes, and disinfection of noncritical surfaces.[2] A systematic variation of bisisoquinolinium and bisquinolinium salts led to the first-in-class bisquaternary ammonium compound (BQAC) dequalinium chloride, i.e., 1,1-(1,10-decanediyl)bis(4-amino-2-methyl)quinolinium dichloride (CAS 6707-58-0, Figure 1).[3] For the past 55 years, this biscation has been widely used as an antiseptic and disinfectant. It has also been investigated for manifold activities (Table 1), although its mechanism of action is not yet clear.
Table 1. Activities of dequalinium toward various microorganisms. Bacteria Staphylococcus aureus Streptococcus agalactiae Streptococcus pyogenes Enterococcus faecalis Listeria monocytogenes Escherichia coli Gardnerella vaginalis Proteus mirabilis Proteus vulgaris Mycobacterium phlei Mycobacterium tuberculosis MIC [mg mL1] 1.28[4] 3[4] 1[4] 24[4] 8[4] 64[4] 128[4] > 1024[4] 63.0[3] 1.66[3] 1.2 (aerobic) 0.3 (anaerobic)[5] Activity ED50 = 0.043 mm[6] MIC = 57.6 mg mL1[4] IC50 = 0.8 mm[a] IC50 = 0.055 mm[b] IC50 [mm] 5.5[7] 11[7]
Parasites Trypanosoma brucei Trichomonas vaginalis Leishmania major Plasmodium falciparum Fungi Candida albicans Cryptococcus neoformans
[a] T. lschlger, unpublished results. [b] L. Sologub, G. Pradel, unpublished results.
Figure 1. Structure of dequalinium chloride.
With this review we demonstrate the anti-infective power of QACs and BQACs toward bacteria and protozoa, especially against plasmodia. The structureactivity relationships (SAR) of the various series of QACs and BQACs toward the different microorganisms as well as their modes of action are also discussed.
nels,[1824] and disintegration of amyloid fibrils.[25] In addition to these effects, dequaliniums toxicity has also been reported; it may cause skin necrosis if administered on intertriginous skin areas under occlusive conditions.[26, 27] Despite this, none of these effects are relevant for its current medical indications.
Possible mode of action of dequalinium chloride and related compounds In contrast to some other QACs, the mode of action of dequalinium seems to be more specific. For example, the low dose of 10 mg per vaginal tablet, 0.25 mg per lozenge, or 10 mg per 100 mL gurgling solution points to a specific mode of action. Moreover, SARs derived by Babbs et al.[3] indicate that dequalinium has one or more specific targets; otherwise, distinct differences in activity between closely related compounds would be difficult to explain. Figure 2 clearly shows the influence of
[a] M. Tischer, Prof. Dr. U. Holzgrabe Institute of Pharmacy and Food Chemistry University of Wrzburg, Am Hubland, 97074 Wrzburg (Germany) E-mail: u.holzgrabe@pharmazie.uni-wuerzburg.de [b] Dr. G. Pradel, Dr. K. Ohlsen Research Center for Infectious Diseases, University of Wrzburg Joseph-Schneider-Str. 2, Building D15, 97080 Wrzburg (Germany)
Dequalinium
Today dequalinium is licensed only as a topical medication. This includes pharyngeal sprays, throat lozenges, mouth washes and decongestant sprays, topical creams, gels, and ointments, as well as tablets or suppositories for vaginal application. Beside its antibacterial activity, dequalinium has been reported to exhibit an array of other biological effects such as potent antitumor activity through toxicity against mitochondria,[812] inhibition of protein kinase C,[13, 14] antimalarial,[1517] antitrypanosomal,[3, 6] and antifungal[7] activities, inhibition of mycothiol ligase in Mycobacterium tuberculosis,[5] selective blockade of small-conductance Ca2 + -activated K + chanChemMedChem 2012, 7, 22 31
www.chemmedchem.org
23
MED
Maximilian Tischer completed his graduate studies in pharmacy at the University of Wrzburg in 2006 and received his pharmaceutical license in 2007. He is now working toward his PhD in medicinal chemistry at the University of Wrzburg under the supervision of Professor Holzgrabe. His fields of research are organic synthesis, SAR studies, and structural development of anti-infective quaternary ammonium compounds. Gabriele Pradel studied biology at the University of Gieen and received her PhD from the University of Frankfurt am Main in 1999. She then worked as a postdoctoral fellow and instructor at the New York University Medical Center and the Weill Cornell Medical Center of New York, respectively. In 2005 she was awarded by the Emmy Noether Program of the Deutsche Forschungsgemeinschaft, and since then has headed a research group at the Wrzburg Research Center for Infectious Diseases. Her research focus lies in the gametocyte stages of malaria parasites as targets for transmission-blocking strategies. Knut Ohlsen studied biology at the University of Leipzig. In 1998 he received his PhD in microbiology from the University of Wrzburg. After a postdoctoral fellowship in the research group of Jrg Hacker at the University of Wrzburg, he became a group leader at the University of Wrzburgs Institute for Molecular Infection Biology. His research areas are hostpathogen interactions of Staphylococcus aureus, antibacterial drug development, and in vivo imaging techniques. Ulrike Holzgrabe studied chemistry and pharmacy at the Universities of Marburg and Kiel, respectively, receiving her PhD under the supervision of Prof. Dr. R. Haller. In 1989 she finished her habilitation, and in 1990 she accepted an associate professorship at the University of Bonn, where she served as a vice rector from 1997 to 1999. Since 1999 she has held a full professorship at the University of Wrzburg. She was president of the German Pharmaceutical Society from 2003 to 2007. Her main research interests include drug discovery in the fields of muscarinic ligands and anti-infectives, as well as quality analysis of active pharmaceutical ingredients.
U. Holzgrabe et al.
Figure 2. Dependence of anti-S. aureus activity on the length of the central linker of dequalinium derivatives. The negative logarithm of the MIC values (pMIC) is plotted as a function of the number of methylene groups (n); hence, an increase in pMIC value reflects an increase in anti-infective potency.
varying the length of the central linker on minimal inhibitory concentration (MIC) values against Staphylococcus aureus. Replacing the quinoline moiety with an isoquinoline group increases the activity threefold (10 methylene groups). With further modification of the structure toward dequalinium, the activity increases another 50-fold. Substances without a specific target would not show such subtle differences in activity due to slight alterations in the chemical structure. Similar findings were reported by Babbs et al. for other series of corresponding compounds (see Figure 3 and Figure 4).[3]
Figure 3. Dependence of anti-S. aureus activity on the length of the central linker of isoquinolinium derivatives.
Figure 4. Dependence of anti-S. aureus activity on the length of the central linker of quinolinium derivatives.
24
www.chemmedchem.org
Quaternary Ammonium Salts In 1961, Caldwell et al. investigated the antibacterial activity of a series of dequalinium derivatives differing in the alkyl linker chain length. It was demonstrated that the mode of action of dequalinium may be closely related to its interaction with proteins such as pepsin, casein, gelatin, and egg albumin.[28] Investigations into the site of action revealed a dequaliniums high affinity for membranes, no dissolution of the cell wall in S. aureus, and almost no lytic effect on B. megaterium protoplasts. Additionally, the substance is taken up by the cells, and the morphological abnormalities are mainly confined to intracellular changes.[29] A few years later, Hugo and Frier found that dequalinium is taken up by the cells and that the cell wall is not the primary site of absorption. Although Gramnegative E. coli were observed to be much less susceptible to dequalinium than S. aureus, the uptake under inhibitory conditions was similar, indicating a difference in affinity rather than in mode of action. The outer cell membrane of Gram-negative organisms seems to constitute a permeability barrier to dequalinium. Additional experiments with the aerobic and anaerobic metabolism of glucose indicated no specific attack on the cytochrome system. Taken together, it is rather likely that dequalinium exhibits its bactericidal action by access to the cytoplasm followed by precipitation of cytoplasmic material. Nucleic acids are particularly susceptible to precipitation by this compound.[30] Using fluorescence microscopy, Weiss et al. demonstrated that dequalinium accumulates in the mitochondria of several human and murine tumor cell lines, resulting in potent anticarcinoma activity.[9] Furthermore, it binds DNA and inhibits calmodulin as well as ATP production in isolated rat liver mitochondria. Through delocalization of the positive charge over the quinaldine moiety, it is a lipophilic biscation that can readily penetrate the lipid bilayers of mitochondria and bacteria, for example.[912, 31, 32] This observation lends support to the endosymbiotic theory, which proposes free-living bacteria as the evolutionary origin of mitochondria. In another study by Locher et al.,[38] the antimicrobial properties of nostocarbolines (see Figure 7 below) were investigated; these compounds were observed to exhibit activities only as dimers. There was a strong correlation between the length of the linker chain and antimicrobial activity. The most active compounds had linkers of 10 and 12 methylene units and were bactericidal against methicillin-resistant S. aureus (MRSA), vancomycin-resistant enterococci (VRE), and Escherichia coli. Killing activity correlated with membrane damage, as indicated by ATP leakage. Moreover, a SOS response following DNA damage was observed in E. coli. Clearly, these substances interact with bacterial membranes, given their chemical structures with positive charges and hydrophobic regions. However, the role of intracellular targets remains vague. The observed effect on the SOS response may suggest some specific interaction with DNA; however, this has never been proven. It is possible that the most potent agents exerted membrane damage, leading to disruption of the cell envelope, disturbing proton motive force, and arresting intracellular activity by binding targets in the cytoplasm.[38] Interestingly, Gram-positive bacteria are generally more sensitive than Gram-negative bacteria. This has been attributed to the outer membrane of Gram-negative bacteria, which is absent in Gram-positive strains (cf. dequalinium findings). As a consequence, QACs must first pass the outer membrane by lysis of this layer, for example, and then interact with the inner membrane and later with intracellular components. The Gramnegative bacterium Pseudomonas aeruginosa and bacteria that are embedded in biofilm matrices are less susceptible to QACs and other antiseptics.[39, 40] These observations indicate a crucial role played by particular surface structures of various bacterial species in sensitivity to QACs. Likewise, mycobacteria, which have a unique cell wall consisting of a hydrophobic mycolate layer and a peptidoglycan layer linked by the polysaccharide arabinogalactan, show high resistance to QACs.[41] Importantly, many antibiotic-resistant staphylococci (such as MRSA or methicillin-resistant S. epidermidis) have acquired plasmids such as qacA, qacB, qacC or qacD, which encode efflux pumps and thus confer resistance determinants to QACs.[4244] This suggests that intracellular accumulation of QACs is important for full activity and underscores the importance of the interaction between QACs and intracellular targets for antibacterial activity. Several bisquaternary bisnaphthalimides have been evaluated against a variety of human pathogens, leading to the discovery of compounds that show inhibitory effects against Gram-positive bacteria, including multi-resistance staphylococci. Originally, bistertiary bisnaphthalimides were described as putative anticancer drugs, but clinical trials were unsuccessful.[45, 46] In a reevaluation of bisquaternary bisnaphthalimides against bacteria, a highly active substance was identified, which has been designated as MT02 (Figure 5). This compound exerts high antimicrobial activity, with a MIC value of 0.31 mg mL1 against community-acquired MRSA lineage USA300. It is also active against other Gram-positive bacteria such as Streptococcus pneumonia, Listeria monocytogenes, and Bacillus subtilis. Gram-negative bacteria such as Citrobacter
Targets of QACs in Bacteria: More Than Nonspecific Interactions with Membranes

Many surface-active compounds are bactericidal, and some explanations have been proposed for this action. In 1951, Salton investigated the adsorption of cetyltrimethylammonium bromide (CTAB), a monoquaternary ammonium compound (MQAC), in S. aureus and some other bacteria.[33] It was observed in electron microscopy studies that the cytoplasmic membrane contracts away from the cell wall, indicating cytolytic damage and cell leakage.[34] CTAB-treated bacteria released their cell contents into their suspending media. This supports the hypothesis, previously raised by Baker et al., of cell membrane disorganization by detergent-like activity as the main mode of action of QACs.[35] More than 50 years later, Denyer also found that the main effects of QACs in bacteria are structural and functional changes in the cell wall, release of wall components and initiation of autolysis, inhibition of membrane ATPase, and electrostatic interactions with negatively charged polar head groups of phospholipids.[36, 37]
www.chemmedchem.org
25
MED
U. Holzgrabe et al. ic base sequence as reported for the bisnaphthalimide elinafide.[49] The direct binding of MT02 to DNA as a mode of action represents an example in which the molecular target of a BQAC has been evaluated in detail. So far, no specific target has been identified for most QACs; it is assumed that the effect is rather generalized than specific to one target. However, as already pointed out, there should be some target specificities, as shown for the bisquaternary bisnaphthalimide MT02, because the activity of QACs toward different bacterial species varies substantially and cannot be explained simply by the structure of cationic charge and hydrophobic portions. Interestingly, similar to QACs, membrane damage has long been assumed as the major mechanism of action of cationic antimicrobial peptides.[50] These peptides share chemical features with QACs, such as a cationic portion and a hydrophobic region. It became clear more recently that other target sites, including interaction with lipid II, an important precursor of cell wall synthesis, are major targets for some cationic antibacterial peptides.[51] Whether QACs also interact with lipid II has not yet been investigated, and remains an attractive hypothesis for further evaluation. With all these findings taken into account, the general mechanism of QAC action can be summarized as follows:
*
Figure 5. Structure of MT02.
koseri, Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Pseudomonas aeruginosa, Salmonella typhimurium, Serratia marcescens, and Shigella dysenteriae were not susceptible, possibly owing to their outer membrane, which acts as a barrier that prevents access of the compound into the cytoplasm.[47] Different lengths of the methylene linker showed no correlation with antimicrobial activity (Figure 6). However, derivatives of MT02 without nitro groups and with shorter or longer connecting chains also exerted activity against S. aureus, but only at higher concentrations. Importantly, studies into the mode of action based on DNA microarray global expression analysis and radiolabeling experiments generated evidence that MT02 is a DNA binding compound that leads to the inhibition of DNA replication. The transcriptional signature was characterized by a strong expression of genes involved in DNA metabolism, DNA replication, SOS response, and transport of positively charged compounds.[47] The direct binding of MT02 to DNA was demonstrated by surface plasmon resonance and gel retardation experiments. Importantly, the substance did not show cytotoxic activity in cell culture systems using four different eukaryotic cell lines. Furthermore, expression analysis suggested that MT02 indirectly impacts cell wall metabolism and cell propagation. In contrast to bistertiary bisnaphthalimides, which exhibit antitumor activity (see above),[46, 48] the two permanent positive charges of MT02 prevent penetration of the compound into eukaryotic cells and may therefore be the cause of its low cytotoxicity despite its binding to eukaryotic DNA. Binding of MT02 to double-stranded DNA is reversible, and is probably not restricted to a specif-
* *
adsorption and penetration through the cell wall (the cell wall has no penetration barrier); binding to components of the cell membrane, probably with both lipids and proteins; disorganization of cell membranes and, at higher concentrations, causing leakage with subsequent loss of low-molecular-weight material; intracellular degradation of proteins and nucleic acids; lysis of cell wall components by release of autolytic enzymes; complete loss of structural organization of the cell.
QACs in the Treatment of Malaria

Malaria is the most prevalent tropical disease in the world, with 225 million new cases reported each year and an annual death toll of 0.8 million people.[52] Many of these fatal cases are instances of drug-resistant malaria tropica, caused by the protozoan parasite Plasmodium falciparum. In the search for new drugs for chemotherapeutic intervention, dequalinium was tested for its effect on malaria parasites. In vitro parasite growth tests showed that dequalinium is active against the P. falciparum blood stages, with an IC50 value of 55 nm (see Table 1; references [16], [17], and [60]; L. Sologub and G. Pradel, unpublished observations). The mode of action included compound accumulation in mitochondria combined with an inhibition of calmodulin and of the membrane-bound ATPase of the parasitophorous vacuole.[17, 53] The in vitro results
Figure 6. Chain lengthactivity correlation of bisquaternary naphthalimides against S. aureus.
26
www.chemmedchem.org
Quaternary Ammonium Salts were confirmed in vivo by using Plasmodium berghei-infected mice. Dequalinium was able to cure 40 % of the infected animals, monitored 30 days post infection.[16] While dequalinium was tested in malaria parasites due to its overall antimicrobial effect, other QACs were investigated for their potential as antimalarials by a target-oriented approach that is linked to the de novo membrane biosynthesis of bloodstage parasites. The rapid, manifold, and recurrent replication of malaria parasites in red blood cells (RBCs) requires high turnover rates in membrane components, making parasite lipid metabolism an attractive weak spot for targeting drugs.[54] Phospholipid synthesis in malaria parasites After infection by Plasmodium, the total amount of phospholipids increases sixfold in parasitized human RBCs.[55] The membrane composition of malaria parasites differs significantly from that of uninfected erythrocytes, and is primarily composed of phospholipids such as phosphatidylcholine (PtdCho, 4050 %; see Figure 7) and phosphatidylethanolamine (PtdEtn, 3545 %). Malaria parasites synthesize phospholipids from polar head groups such as choline (Figure 7), ethanolamine, or serine.[54, 55] PtdEtn is synthesized by the parasite via a pathway following phosphorylation of ethanolamine. Ethanolamine is obtained in limited amounts from plasma and in larger quantities through decarboxylation of serine, which, in turn, is either imported or derived from hemoglobin degradation.[56] PtdCho is synthesized from choline by two routes in Plasmodium. During the first route, which is highly conserved in eukaryotes, PtdCho is derived from choline via an enzymatic cascade of the de novo cytidine diphosphate (CDP)choline pathway. This route involves three enzymes: choline kinase, CTP phosphocholine cytidylyltransferase, and choline/ethanolamine phosphotransferase (PfCEPT). The second route of PtdCho synthesis was only recently identified in P. falciparum and is typical for plants and nematodes, but is absent in mammals.[57] Through this route, which is termed the serine decarboxylasePtdEtn methyltransferase (SDPM) pathway, phosphoethanolamine is converted into phosphocholine, a rate-limiting step mediated by the enzyme phophoethanolamine methyltranserase PfPMT.[57] Both routes of PtdCho synthesis represent excellent targets for novel chemotherapeutics. Only the CDPcholine pathway, however, was hitherto reported to be sensitive to QACs (Figure 8). The effect of QACs on the plasmodial choline carrier
The CDPcholine pathway requires the transport of host choline from plasma into the infected erythrocyte, a process that involves the remnant erythrocytic choline carrier as well as an unknown transporter of the parasite-induced new permeation pathway (NPP).[58] Subsequently choline crosses the barriers of the parasitophorous vacuole membrane and the parasite plasma membrane by an as-yet unidentified organic cation transporter of Plasmodium,[54, 55, 59, 60] which is proposed to be located in the plasmalemma of the parasite (H. Vial, personal communication). The transport of choline from plasma is a rate-limiting step in membrane biosynthesis by malaria parasites, and thus the choline carriers of infected erythrocytes represent excellent target structures for the design of phospholipid polar head analogues as competitors of choline. In 1985, Vial and colleagues reported for the first time that compounds which mimic the structures of lipid precursors such as ethanolamine, serine, or choline are able to block blood-stage replication of the malaria parasite in the low micromolar range. The first three compounds tested as structural analogues of choline were the mono-QAC (MQAC) dodecyltrimethylammonium and the bis-QACs (BQACs) decamethonium and hemicholinium 3 (Figure 7), which exhibited antimalarial activities toward P. falciparum with half-maximal inhibitory concentration (IC50) values of 0.710 mm.[61] Treatment of parasites with these compounds resulted in a specific decrease in the biosynthesis of PtdCho, paving the way for a more rational approach in designing choline analogues.[62] In follow-up studies, Vial and co-workers designed extensive sets of primary, secondary, and tertiary amines as well as MQACs and BQACs as potential choline analogues, which were tested in various P. falciparum strains.[63, 64] SAR analyses indicated that the Figure 7. Structures of choline, phosphatidylcholine (PtdCho), hemicholinium 3, G25, 3m, shape, electronegativity, and lipophilicity of the comM34, M38, M40, M60, T3/T4, T16, TE3, MAP-412, and nostocarbolines.
ChemMedChem 2012, 7, 22 31 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
27
MED
U. Holzgrabe et al.
Figure 8. SDPM and CDP-choline pathways for PtdCho synthesis in P. falciparum.[54, 55] Cho, choline; CS, cytostome; NPP, new permeation pathway; E, erythrocyte; eChoT, erythrocyte choline transporter; EM, erythrocyte membrane; Etn, ethanolamine; FV, food vacuole; Hb, hemoglobin; OCT, organiccation transporter; P-Cho, phosphocholine; PfCCT, CTP-phosphocholine cytidylyltransferase; PfCEPT, choline/ethanolamine cytidylyltransferase; PfCK, choline kinase; PfEK, ethanolamine kinase, PfPMT, phosphoethanolamine methyltransferase; PLMT, phospholipid methyltransferase; PPM, parasite plasma membrane; PtdCho, phosphatidylcholine; PtdEtn, phosphatidylethanolamine; PtdEtnMT, phosphatidylethanolamine methyltransferase; PV, parasitophorous vacuole; PVM, parasitophorous vacuole membrane; Ser, serine; SD, serine decarboxylase.
Figure 9. Chain lengthactivity correlation of BQACs reported by Vial and colleagues.[6466]
pounds are related to in vitro antiplasmodial activity. For MQACs, increasing lipophilicity around the nitrogen atom was shown to be beneficial for activity, as was an increase in alkyl chain length up to 12 methylene groups. Beyond 12 methylene units, the antimalarial activity decreased slightly, probably due to curling up of the chain. In contrast, increasing alkyl chain length (n) between the two nitrogen atoms constantly and dramatically increased the antimalarial properties of BQACs (e.g., compound G19: IC50 = 3 pm at n = 21; see Figure 9).[64] Furthermore, a moderate increase in lipophilicity around the nitrogen atom is beneficial for the BQACs. While it remains to be determined which of the three above-mentioned choline carriers acts as the main target for the QACs, these new data point toward targeting the organic cation transporter within the parasite plasmalemma (H. Vial, personal communication). Rational drug design of antiplasmodial BQACs SAR analyses pointed to a lead compound, G25 (Figure 7), which has an alkyl chain length of n = 16 and IC50 values of 0.65.3 nm toward various drug-sensitive and drug-resistant P. falciparum strains.[64, 67] A radioactive G25 analogue accumulated in infected RBCs, with a 180-fold increase of substance in the membrane fraction of infected relative to uninfected RBCs.[67] G25 had no cytotoxic effects in cell culture, with a selectivity index (SI) ! 1000. During a 4 day suppression test in mice, which were infected with P. chabaudi, G25 successfully cleared parasites from the blood, but with a rather low therapeutic index (TI) of 18.[67, 68] At high doses, G25 provoked rapid and transient hypoxia in the rodents. G25 was subsequently
tested in Aotus monkeys, which were infected with P. falciparum, as well as in Rhesus monkeys infected with P. cynomolgi, and cured the infected monkeys at doses of 0.03 mg kg1 and a TI > 30. These promising results, however, experienced a drawback in the discovery that subcutaneously administered radioactive compounds had a short elimination half-life in mice (3.3 h) with low bioavailability (17.3 %). The effective dose of orally administered G25 was 100-fold higher than the intraperitoneally or subcutaneously administered compound.[68] G25 was thus limited as a lead compound owing to its weak oral absorption, which was assigned to the permanent positive charge of the quaternary ammonium moiety, which prevented the compound from crossing the gastrointestinal barrier. To remedy the problems of low bioavailability and toxicity, in follow-up studies the pyrrolidinium moiety of G25 was substituted by a thiazolium group that is present in vitamin B1.[69] Two highly active compounds were identified, T3 and T4 (n = 12 for both compounds; see Figure 7), which exhibited in vitro activities against P. falciparum with IC50 values of 2.6 and 0.7 nm and in vivo activities against P. vinckei in the mouse model with half-maximal effective dose (ED50) values of 0.2 and 0.14 mg kg1, respectively.[69] T3 and its neutral bio-precursor TE3 (Figure 7) were subsequently investigated for their pharmacological properties. Low doses administered intraperitoneally offer protection against malaria in the P. vinckei mouse model, with ED50 values of 0.2 mg kg1 for T3 and 0.25 mg kg1 for TE3, while oral administration led to protection, with ED50 values of 13 and 5 mg kg1, respectively.[70] The bioavailability of T3 was 72 % after intraperitoneal administration and 15 % after oral administration in rats. Furthermore, the T3 plasma concentrations of 8 nm at 24 h following oral administration of TE3 were higher than the IC50 values for most chloroquine-resistant strains of P. falciparum.[70] Based on these promising results, compound T3/SAR97276 is now being clinically investigated by SanofiAventis and has meanwhile entered phase II clinical trials.[54] Two recent studies by the Vial research group were aimed at designing more effective neutral prodrugs of T3 to enhance its
28
www.chemmedchem.org
Quaternary Ammonium Salts bioavailability.[71, 72] To circumvent the poor oral absorption of previous BQACs, the Vial group further replaced the quaternary ammonium heads by bioisosteric groups that are highly basic, charged at physiological pH, and capable of creating bonds with the target similar or stronger than those of conventional BQACs.[65] Because of the equilibrium between protonated and unprotonated forms, these compounds are expected to diffuse more easily through cell barriers. Out of 60 new compounds, 2-aminopyridium salts, amidines, and guanidines showed the best antiplasmodial activities in the low nanomolar range. Again, dimerization of the cationic head produced effectors with enhanced antimalarial activities, and the optimal alkyl chain length between the cationic heads was n = 12. Three lead compounds, the amidines M34, M38, and M40 (Figure 7), were identified, but the high flexibility of the linker was considered a drawback for membrane permeation.[65] In a subsequent study, hybrid biscationic salts were designed from efficient symmetrical derivatives (T3 or T4 and M34, 38, or M40).[66] However, the generation of hybrids did not result in any synergy between amidine and thiazolium cationic heads, and the oral absorption was still weak. Another lead compound of the study by Calas et al. was M64, a reversed N-alkylamidine, in which the alkyl chain is attached to the functional nitrogen atom (Figure 7).[65, 73] Because M64 turned out to be unstable in vivo, reverse benzamidine derivatives were designed. Several promising candidates were identified, among them compound 3m (Figure 7), which exhibited an IC50 value of 6.6 nm in P. falciparum in vitro and an ED50 value of 5 mg kg1 in the P. vinckei mouse model.[73] Further advances in the design of antiplasmodial QACs Other research groups joined the search for antimalarial BQACs. Sasaki and co-workers synthesized antimalarial compounds from isonicotinic acid as a starting material. The first lead compounds, PAM-86 (Figure 7), exhibited promising antiplasmodial activities in vitro (IC50 = 10 nm), but not in vivo.[74] The subsequently designed MAP series showed increased activities with increased alkyl chain linker length, whereas the cytotoxicity of the compounds increased with the number of side methylene groups (m).[75] A lead compound, MAP-412 (n = 12, m = 4; see Figure 7), exhibited an IC50 value of 100 nm with no cytotoxicity and an ED50 value of 8.2 mg kg1 in mice infected with P. berghei.[75] The inhibitory effect of the cyanobacterium-derived alkaloid nostocarboline (Figure 7) against protozoan parasites was described.[76] Biscationic dimeric derivatives exhibited activities against Leishmania donovani in the sub-micromolar range, against Trypanosoma brucei in the low-micromolar range, and against P. falciparum in the low-nanomolar range with weak cytotoxicity. When tested in the P. berghei mouse model, however, the dimers did not show significant activity, while nostocarboline exhibited an ED50 value of 50 mg kg1.[77] Furthermore, the above-mentioned bisquaternary naphthalimides were tested for their antimalarial properties.[78] SAR analyses showed that an alkyl chain length of n = 8 between the positively charged nitrogen atoms exhibited optimal antiplasChemMedChem 2012, 7, 22 31
modial activities, with IC50 values of 60 nm in P. falciparum in vitro (Figure 10), while n > 8 decreased the effect on the parasite. A dysfunction in membrane biosynthesis was shown on the ultrastructural level for two of the compounds.[78]
Figure 10. Chain lengthactivity correlation of bisquaternary naphthalimides reported by Tischer et al.[78]
Additional modes of action of antiplasmodial QACs Because PtdCho can be synthesized by the parasite via the SDPM pathway and thus plasma-derived choline is not essential for parasite blood-stage survival, it was postulated that the inhibition of choline uptake alone cannot account for the antimalarial activity of BQACs. In this context, Biagini et al. studied the uptake of the antimalarially active BQAC T16 (IC50 = 25 nm) and its radioactive analogue (Figure 7).[79] The compound was reported to enter infected erythrocytes by NPP-induced channels, as shown by using the NPP inhibitor furosemid. Approximately 40 % of total T16 accumulated in the food vacuole, where it bound to the hemoglobin-derived ferriprotoporphyrin IX as well as to hemozoin.[79, 80] The authors proposed that BQACs act on hemozoin, which would thus strongly contribute to their lethal effect on malaria parasites. In contrast to these assumptions, two new studies indicate that BQACs affect enzymatic activities downstream of choline entry. A recent proteomics analysis on T4-treated P. falciparum showed that T4 treatment decreases the expression of the PfCEPT in these parasites.[81] The enzyme is responsible for the final steps of PtdCho and PtdEtn synthesis (Figure 8). A follow-up study revealed that the BQACs hemicholinium 3 and T3 inhibit the phosphorylation activities of recombinantly expressed choline kinase (PfCK) and ethanolamine kinase (PfEK, Figure 8).[82] The latter two studies indicate that BQACs exhibit multiple modes of action during phospholipid biosynthesis in Plasmodium by targeting the choline carrier as well as enzymes of both the SDPM and the CDPcholine pathways, which would conclusively explain the lethal effect of BQACs on malaria parasites.
Summary and Outlook

Since the advent of dequalinium chloride as a potent disinfectant and antiseptic drug that is also active against mycobacteria and protozoa, a variety of new BQACs were found to have higher activity especially against Gram-positive bacteria and
www.chemmedchem.org
29
MED
plasmodia. On one hand, the antimicrobial activity seems to depend on the length of the chain between the two positive charges, indicating that a specific target, for example, related to the structure and biosynthesis of the cell wall components, is addressed. On the other hand, the lateral heterocyclic moieties are able to direct the BQACs to other targets, such as DNA, as was found with the bisnaphthalimides. Although nonspecific interactions with membranes was regarded for a long time to be the main mechanism of QAC and BQAC action, closer investigations into the mode of action revealed similar and different targets for this class of charged compounds. Particularly in the case of antimalarial BQACs, it was shown that members of this compound class are able to address very specific targets. None of the newly reported compounds are currently in clinical use; however, some of them are in, or are close to entering, clinical trials.
U. Holzgrabe et al.
[20] D. Galanakis, C. A. Davis, C. R. Ganellin, P. M. Dunn, J. Med. Chem. 1996, 39, 359 370. [21] D. Galanakis, C. A. Davis, B. D. R. Herrero, C. R. Ganellin, P. M. Dunn, D. H. Jenkinson, J. Med. Chem. 1995, 38, 595 606. [22] D. Galanakis, C. R. Ganellin, S. Malik, P. M. Dunn, J. Med. Chem. 1996, 39, 3592 3595. [23] C. R. Ganellin, D. Galanakis, J. C. Rosa, S. Malik, P. M. Dunn, D. H. Jenkinson, Book of Abstracts, 211th ACS National Meeting, New Orleans, LA (USA), March 24 28, 1996, MEDI-002. [24] J. C. Rosa, D. Galanakis, C. R. Ganellin, P. M. Dunn, J. Med. Chem. 1996, 39, 4247 4254. [25] J. W. Park, I. H. Lee, J. S. Hahn, J. Kim, K. C. Chung, S. R. Paik, Biochim. Biophys. Acta Gen. Subj. 2008, 1780, 1156 1161. [26] L. Spngberg, K. E. Safavi, A. Kaufman, E. A. Pascon, J. Endod. 1988, 14, 175 178. [27] D. S. Wilkinson, Hautarzt 1970, 21, 114 116. [28] D. Caldwell, W. A. Cox, P. F. DArcy, L. R. Rowe, J. Pharm. Pharmacol. 1961, 13, 554 564. [29] W. A. Cox, Appl. Microbiol. 1965, 13, 956 966. [30] W. B. Hugo, M. Frier, Appl. Microbiol. 1969, 17, 118 127. [31] W. L. Bodden, S. T. Palayoor, W. N. Hait, Biochem. Biophys. Res. Commun. 1986, 135, 574 582. [32] T. C. Rowe, V. Weissig, J. W. Lawrence, Adv. Drug Delivery Rev. 2001, 49, 175 187. [33] M. R. J. Salton, J. Gen. Microbiol. 1951, 5, 391 404. [34] M. R. J. Salton, R. W. Horne, V. E. Cosslett, J. Gen. Microbiol. 1951, 5, 405 407. [35] Z. Baker, R. W. Harrison, B. F. Miller, J. Exp. Med. 1941, 74, 611 620. [36] S. P. Denyer, Intl. Biodeterioration Biodegradation 1995, 36, 227 245. [37] M. R. Brown, E. Tomlinson, J. Pharm. Sci. 1979, 68, 146 149. [38] H. H. Locher, D. Ritz, P. Pfaff, M. Gaertner, A. Knezevic, D. Sabato, S. Schroeder, D. Barbaras, K. Gademann, Chemotherapy 2010, 56, 318 324. [39] A. D. Russell, G. W. Gould, J. Appl. Bacteriol. 1988, 65, S167 S195. [40] D. O. Kolawole, FEMS Microbiol. Lett. 1984, 25, 205 209. [41] S. J. Broadley, P. A. Jenkins, J. R. Furr, A. D. Russell, Lett. Appl. Microbiol. 1991, 13, 118 122. [42] A. Leelaporn, I. T. Paulsen, J. M. Tennent, T. G. Littlejohn, R. A. Skurray, J. Med. Microbiol. 1994, 40, 214 220. [43] T. G. Littlejohn, D. Diberardino, L. J. Messerotti, S. J. Spiers, R. A. Skurray, Gene 1991, 101, 59 66. [44] I. T. Paulsen, M. H. Brown, T. G. Littlejohn, B. A. Mitchell, R. A. Skurray, Proc. Natl. Acad. Sci. USA 1996, 93, 3630 3635. [45] M. F. Braa, J. M. Castellano, M. Moran, M. J. Perez de Vega, C. R. Romerdahl, X. D. Qian, P. Bousquet, F. Emling, E. Schlick, G. Keilhauer, AntiCancer Drug Des. 1993, 8, 257 268. [46] M. F. Braa, M. Cacho, M. A. Garci, B. de Pascual-Teresa, A. Ramos, M. T. Domnguez, J. M. Pozuelo, C. Abradelo, M. F. Rey-Stolle, M. Yuste, M. Bez-Coronel, J. C. Lacal, J. Med. Chem. 2004, 47, 1391 1399. [47] T. M. Menzel, M. Tischer, P. Francois, J. Nickel, J. Schrenzel, H. Bruhn, A. Albrecht, L. Lehmann, U. Holzgrabe, K. Ohlsen, Antimicrob. Agents Chemother. 2011, 55, 311 320. [48] J. Gallego, B. R. Reid, Biochemistry 1999, 38, 15104 15115. [49] L. Gonzlez-Bulnes, J. Gallego, J. Am. Chem. Soc. 2009, 131, 7781 7791. [50] H. Jenssen, P. Hamill, R. E. Hancock, Clin. Microbiol. Rev. 2006, 19, 491 511. [51] T. Schneider, T. Kruse, R. Wimmer, I. Wiedemann, V. Sass, U. Pag, A. Jansen, A. K. Nielsen, P. H. Mygind, D. S. Raventos, S. Neve, B. Ravn, A. M. Bonvin, L. De Maria, A. S. Andersen, L. K. Gammelgaard, H. G. Sahl, H. H. Kristensen, Science 2010, 328, 1168 1172. [52] World Malaria Report 2010, World Health Organization http:// www.who.int/malaria/world_malaria_report_2010/worldmalariareport2010.pdf (accessed November 8, 2011). [53] D. A. Bullough, E. A. Ceccarelli, J. G. Verburg, W. S. Allison, J. Biol. Chem. 1989, 264, 9155 9163. [54] C. Ben Mamoun, S. T. Prigge, H. Vial, Drug Dev. Res. 2010, 71, 44 55. [55] S. Dchamps, S. Shastri, K. Wengelnik, H. J. Vial, Int. J. Parasitol. 2010, 40, 1347 1365. [56] N. Elabbadi, M. L. Ancelin, H. J. Vial, Biochem. J. 1997, 324, 435 445. [57] G. Pessi, G. Kociubinski, C. B. Mamoun, Proc. Natl. Acad. Sci. USA 2004, 101, 6206 6211.
Acknowledgements
We thank Henri Vial (Universit Montpellier II, France) for helpful discussions. G.P. acknowledges funding by the Emmy Noether Program of the Deutsche Forschungsgemeinschaft; U.H., K.O., and M.T. acknowledge funding by the Collaborative Research Grant (SFB 630) of the Deutsche Forschungsgemeinschaft. Keywords: ammonium compounds cations dequalinium drug discovery malaria
[1] G. Domagk (Alba Pharmaceuticals Co.), US002108765, 1938. [2] G. McDonnell, A. D. Russell, Clin. Microbiol. Rev. 1999, 12, 147 179. [3] M. Babbs, H. O. J. Collier, W. C. Austin, M. D. Potter, E. P. Taylor, J. Pharm. Pharmacol. 1956, 8, 110 119. [4] V. Della Casa, H. Noll, S. Gonser, P. Grob, F. Graf, G. Pohlig, Arzneim.Forsch. 2002, 52, 699 705. [5] M. T. Gutierrez-Lugo, H. Baker, J. Shiloach, H. Boshoff, C. A. Bewley, J. Biomol. Screening 2009, 14, 643 652. [6] D. C. Jones, I. Hallyburton, L. Stojanovski, K. D. Read, J. A. Frearson, A. H. Fairlamb, Biochem. Pharmacol. 2010, 80, 1478 1486. [7] C. K. Ng, V. Singhal, F. Widmer, L. C. Wright, T. C. Sorrell, K. A. Jolliffe, Bioorg. Med. Chem. 2007, 15, 3422 3429. [8] G. Gamboa-Vujicic, D. A. Emma, S. Y. Liao, C. Fuchtner, A. Manetta, J. Pharm. Sci. 1993, 82, 231 235. [9] M. J. Weiss, J. R. Wong, C. S. Ha, R. Bleday, R. R. Salem, G. D. Steele, Jr., L. B. Chen, Proc. Natl. Acad. Sci. USA 1987, 84, 5444 5448. [10] A. I. Garca-Prez, E. Galeano, E. Nieto, P. Sancho, Leuk. Res. 2011, 35, 1395 1401. [11] P. Sancho, E. Galeano, E. Nieto, M. D. Delgado, A. I. Garca-Prez, Leuk. Res. 2007, 31, 969 978. [12] E. Galeano, E. Nieto, A. I. Garca-Prez, M. D. Delgado, M. Pinilla, P. Sancho, Leuk. Res. 2005, 29, 1201 1211. [13] C. Abeywickrama, S. A. Rotenberg, A. D. Baker, Bioorg. Med. Chem. 2006, 14, 7796 7803. [14] D. Qin, R. Sullivan, W. F. Berkowitz, R. Bittman, S. A. Rotenberg, J. Med. Chem. 2000, 43, 1413 1417. [15] J. R. Rodrigues, N. D. Gamboa, Parasitol. Res. 2009, 104, 1491 1496. [16] J. R. Rodrigues, N. Gamboa de Dominguez, Exp. Parasitol. 2007, 115, 19 24. [17] M. T. Makler, (Portland Veterans Administration Medical Center, Portland, OR, USA), US004946849, 1990. [18] D. Galanakis, J. A. D. Calder, C. R. Ganellin, C. S. Owen, P. M. Dunn, J. Med. Chem. 1995, 38, 3536 3546. [19] D. Galanakis, C. A. Davis, B. Del Rey Herrero, C. R. Ganellin, P. M. Dunn, D. H. Jenkinson, Bioorg. Med. Chem. Lett. 1995, 5, 559 562.
30
www.chemmedchem.org
Quaternary Ammonium Salts

[58] M. L. Ancelin, M. Parant, M. J. Thuet, J. R. Philippot, H. J. Vial, Biochem. J. 1991, 273, 701 709. [59] G. A. Biagini, E. M. Pasini, R. Hughes, H. P. De Koning, H. J. Vial, P. M. ONeill, S. A. Ward, P. G. Bray, Blood 2004, 104, 3372 3377. [60] A. M. Lehane, K. J. Saliba, R. J. Allen, K. Kirk, Biochem. Biophys. Res. Commun. 2004, 320, 311 317. [61] M. L. Ancelin, H. J. Vial, J. R. Philippot, Biochem. Pharmacol. 1985, 34, 4068 4071. [62] M. L. Ancelin, H. J. Vial, Antimicrob. Agents Chemother. 1986, 29, 814 820. [63] M. Calas, G. Cordina, J. Bompart, M. Ben Bari, T. Jei, M. L. Ancelin, H. Vial, J. Med. Chem. 1997, 40, 3557 3566. [64] M. Calas, M. L. Ancelin, G. Cordina, P. Portefaix, G. Piquet, V. Vidal-Sailhan, H. Vial, J. Med. Chem. 2000, 43, 505 516. [65] M. Calas, M. Ouattara, G. Piquet, Z. Ziora, Y. Bordat, M. L. Ancelin, R. Escale, H. Vial, J. Med. Chem. 2007, 50, 6307 6315. [66] S. Ortial, S. Denoyelle, S. Wein, O. Berger, T. Durand, R. Escale, A. Pellet, H. Vial, Y. Vo-Hoang, ChemMedChem 2010, 5, 52 55. [67] K. Wengelnik, V. Vidal, M. L. Ancelin, A. M. Cathiard, J. L. Morgat, C. H. Kocken, M. Calas, S. Herrera, A. W. Thomas, H. J. Vial, Science 2002, 295, 1311 1314. [68] M. L. Ancelin, M. Calas, A. Bonhoure, S. Herbute, H. J. Vial, Antimicrob. Agents Chemother. 2003, 47, 2598 2605. [69] A. Hamz, E. Rubi, P. Arnal, M. Boisbrun, C. Carcel, X. Salom-Roig, M. Maynadier, S. Wein, H. Vial, M. Calas, J. Med. Chem. 2005, 48, 3639 3643. [70] O. Nicolas, D. Margout, N. Taudon, S. Wein, M. Calas, H. J. Vial, F. M. Bressolle, Antimicrob. Agents Chemother. 2005, 49, 3631 3639. [71] S. A. Caldarelli, M. Boisbrun, K. Alarcon, A. Hamz, M. Ouattara, X. Salom-Roig, M. Maynadier, S. Wein, S. Peyrottes, A. Pellet, M. Calas, H. Vial, Bioorg. Med. Chem. Lett. 2010, 20, 3953 3956. [72] S. A. Caldarelli, J. F. Duckert, S. Wein, M. Calas, C. Perigaud, H. Vial, S. Peyrottes, ChemMedChem 2010, 5, 1102 1109. [73] O. Berger, S. Wein, J. F. Duckert, M. Maynadier, S. El Fangour, R. Escale, T. Durand, H. Vial, Y. Vo-Hoang, Bioorg. Med. Chem. Lett. 2010, 20, 5815 5817. [74] K. Fujimoto, D. Morisaki, M. Yoshida, T. Namba, K. Hye-Sook, Y. Wataya, H. Kourai, H. Kakuta, K. Sasaki, Bioorg. Med. Chem. Lett. 2006, 16, 2758 2760. [75] K. Motoshima, Y. Hiwasa, M. Yoshikawa, K. Fujimoto, A. Tai, H. Kakuta, K. Sasaki, ChemMedChem 2007, 2, 1527 1532. [76] D. Barbaras, M. Kaiser, R. Brun, K. Gademann, Bioorg. Med. Chem. Lett. 2008, 18, 4413 4415. [77] S. Bonazzi, D. Barbaras, L. Patiny, R. Scopelliti, P. Schneider, S. T. Cole, M. Kaiser, R. Brun, K. Gademann, Bioorg. Med. Chem. 2010, 18, 1464 1476. [78] M. Tischer, L. Sologub, G. Pradel, U. Holzgrabe, Bioorg. Med. Chem. 2010, 18, 2998 3003. [79] G. A. Biagini, E. Richier, P. G. Bray, M. Calas, H. Vial, S. A. Ward, Antimicrob. Agents Chemother. 2003, 47, 2584 2589. [80] E. Richier, G. A. Biagini, S. Wein, F. Boudou, P. G. Bray, S. A. Ward, E. Precigout, M. Calas, J. F. Dubremetz, H. J. Vial, Antimicrob. Agents Chemother. 2006, 50, 3381 3388. [81] K. G. Le Roch, J. R. Johnson, H. Ahiboh, D. W. Chung, J. Prudhomme, D. Plouffe, K. Henson, Y. Zhou, W. Witola, J. R. Yates, C. B. Mamoun, E. A. Winzeler, H. Vial, BMC Genomics 2008, 9, 513. [82] B. Alberge, L. Gannoun-Zaki, C. Bascunana, C. Tran van Ba, H. Vial, R. Cerdan, Biochem. J. 2010, 425, 149 158.
Received: August 23, 2011 Revised: October 28, 2011 Published online on November 24, 2011
www.chemmedchem.org
31
NEW Journal
Scope
Advanced Healthcare Materials is an international, interdisciplinary journal for peer-reviewed papers on materials science aimed at promoting human health.
Advanced Healthcare Materials covers all aspects of materials science in medicine and biotechnology: Design Synthesis Biomaterials Devices for medical applications Physical, chemical, and biological properties Clinical and biotechnological applications Advanced Healthcare Materials contains an attractive mix of Communications, Full Papers, and Review Articles, Progress Reports, Research News, and Essays. All articles will be covered by major indexing and abstracting services, including Web of Science and Medline.
2012. Volume 1, 6 issues. Print ISSN: 2192-2640 Online ISSN: 2192-2659
Dont miss the opportunity to receive COMPLIMENTARY ONLINE ACCESS through 2012 & 2013.
This special service is available to all institutions. Ask your librarian to register at
Readership
Materials scientists, chemists, physicists, engineers, and medical and biotech researchers in academia as well as industry.
wileyonlinelibrary.com/newjournals
www.advhealthmat.com
Editorial Ofce
Advanced Healthcare Materials is handled by the experienced Advanced Materials, Advanced Functional Materials, and Small editorial teams at Wiley-VCH.
3546211110_bu
DOI: 10.1002/cmdc.201100361
The Tubulin Colchicine Domain: a Molecular Modeling Perspective

Alberto Massarotti,*[a] Antonio Coluccia,[b] Romano Silvestri,[b] Giovanni Sorba,[a] and Andrea Brancale[c]
Computational approaches have been increasingly applied to drug design over the past three decades and have already provided some useful results in the discovery of anticancer drugs. Given the increased availability of crystal structures in recent years, a growing number of molecular modeling studies on tubulin have been reported. Herein we present a brief overview of the role played by computational methods in anti-tubulin research, specifically in the context of colchicine binding agent research. An overview of current structures is reported, along with a brief discussion on the issues associated with the various tubulin isotypes. Finally, a summary of the most recent and relevant results is presented, highlighting the challenges and opportunities faced by researchers in this field.
Introduction
Microtubules are important components of the cytoskeleton and play several crucial roles in a diverse array of cellular processes such as morphogenesis, motility, organelle and vesicle trafficking, and chromosome segregation during mitosis.[1] Furthermore, alteration of tubulin polymerization disrupts the formation of tumor vasculature, and it is not surprising that the microtubule cytoskeleton has emerged as an effective target for cancer chemotherapy.[25] Microtubules are highly dynamic polymers, and their essential element is the a/b-tubulin heterodimer.[6, 7] Under certain conditions, the dimers polymerize in a head-to-tail fashion to form linear protofilaments. The polymerization process is extremely fluid, with tubulin dimers added at one end, while dimers are removed from the other end. The transition between these two processes is known as dynamic instability.[810] The tubulin heterodimer contains at least three distinct drug binding sites: the paclitaxel, vinblastine, and colchicine binding sites (Figure 1).[12] For the first two of these sites, drugs are in current use in clinical oncology.[1315] Over the last decades, a large number of compounds able to interact with the colchicine binding site have been reported (Figure 2).[1622] However, no colchicine site inhibitor has found clinical application in anticancer therapy. Colchicine itself binds to tubulin very tightly, but its severe toxicity to normal tissues has hampered its use in the clinic.[23] At the moment, combretastatin A-4 (CA-4) is one of the most promising anti-tubulin agents that targets the colchicine site.[24] Its water-soluble phosphate prodrug form (CA-4P, also known as fosbretabulin, trade name Zybrestat) is in phase II/III clinical trials either alone or in combination with traditional chemotherapeutic agents or with radiotherapy.[2530] Ombrabulin and indibulin are two other colchicine-site-binding agents that are currently being evaluated in phase I/II clinical trials (Figure 3).[31]
A brief overview of the colchicine domain

Colchicine has played a pivotal role in the discovery of tubulin. This molecule was first isolated from the leaves of meadow saffron (Colchicum autumnale) and has been usedand continues to be usedin medicine since the 18th Century for the treat[a] Dr. A. Massarotti, Prof. G. Sorba Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, and Drug and Food Biotechnology Center Universit degli Studi del Piemonte Orientale A. Avogadro Via Bovio 6, 28100 Novara (Italy) E-mail: alberto.massarotti@pharm.unipmn.it Figure 1. Location of the tubulin binding sites: a- and b-tubulin are represented as yellow and green ribbon structures, respectively; vinblastine and DAMA-colchicine are shown as magenta and blue space-filling models; GTP and GDP are represented as cyan stick structures (PDB ID: 1Z2B). Paclitaxel is shown in orange space-filling representation (adapted from PDB ID: 1JFF). All images were prepared with PyMOL.[11] [b] Dr. A. Coluccia, Prof. R. Silvestri Istituto PasteurFondazione Cenci Bolognetti Dipartimento di Chimica e Tecnologie del Farmaco Sapienza Universit di Roma, Piazzale Aldo Moro 5, 00185 Roma (Italy) [c] Dr. A. Brancale Welsh School of Pharmacy, Cardiff University King Edward VII Avenue, Cardiff, CF10 3NB (UK)
33
MED
A. Massarotti et al. The crystal structures era Tubulin crystal structures have been reported for more than a decade. The first X-ray crystallographic structure of tubulin was reported in 1998,[38] but only recently have crystals useful for understanding the colchicine site been obtained (Table 1). In 2000 Gigant et al.[39] published a double tubulin heterodimer in complex with an RB3 protein stathmin-like domain. This crystal structure was obtained with-
Figure 2. Structures of the main compounds that are active at the colchicine site.
Table 1. Reported crystal structures of tubulin colchicine binding domain inhibitors. PDB 1FFX 1SA0 1SA1 1Z2B 3E22 3DU7 3HKB 3HKC 3HKD 3HKE 3N2G 3N2K Ligands 1 4 1, vinblastine 1, soblidotin 1, phomopsin A 12 13 14 15 16 Res. [] 3.95 3.58 4.20 4.10 3.80 4.10 3.65 3.80 3.70 3.60 4.00 4.00 Origin Bos taurus Bos taurus Bos taurus Bos taurus Year 2000 2004 2005 2008 Ref. [39] [37] [40] [41]
Ovis aries
2009
[42]
Figure 3. Examples of colchicine site inhibitors currently in clinical trials.
Ovis aries
2010
[43]
ment of gout. In the 1950s, the effects of colchicine in various cell and tissue types were studied,[32] and in 1968 tubulin was recognized as the biological target of colchicine by Taylor and out any ligands. Four years later Ravelli et al.[37] published the colleagues.[33] The name tubulin was coined by Mohri, who was the first to determine the amino acid composition of the first description of the interaction of colchicine and podophylheterodimer by using microtubules extracted from sea urchin lotoxin with tubulin, showing how these two compounds parsperm.[34] After this influential contribution, little structural intially overlap in their sites of interaction. Other X-ray structures became available between 2005 and formation regarding the ligandreceptor complex become 2008, all of which showed tubulin in complex with both colchiavailable owing to the chemical instability of tubulin.[35] cine and other compounds active at a different domain.[40, 41] A study published in 2000 by Hamel and co-workers[36] can Over the last three years, molecules structurally distinct from be considered the first structure-based approach for comcolchicine and podophyllotoxin have been crystallized in the pounds that interact with the colchicine site. The authors tried tubulin colchicine site (Figure 4). These structures show a new to identify the colchicinoids binding site on tubulin through biochemical and molecular modeling techniques. The colchicine binding site was finally identified in 2004 by Ravelli et al.[37] They reported the structure of tubulin in complex with N-deacetyl-N-(2mercaptoacetyl)colchicine (DAMA-colchicine) and with the RB3 protein stathmin-like domain (PDB ID: 1SA0). At this point, structure-based molecular modeling on the tubulin colchicine site began to emerge. Figure 4. Compounds co-crystallized in the colchicine domain of tubulin (structures of colchicine and podophyllotoxin are omitted).
34
www.chemmedchem.org
The Tubulin Colchicine Domain
Figure 5. Structural interaction fingerprint (SIFt) of the tubulin crystal structures (PDB IDs listed down the left side): each residue (listed across the top) is represented by a seven-bit-long bit string, in which each position is indicated as polar, nonpolar, or as a hydrogen bond interaction.
and interesting binding mode.[42, 43] It is important to stress the complete absence of human tubulin structures in the RCSB Protein Data Bank until now.
The crystallized colchicine domain

The colchicine site is a deep pocket located at the ab interface of tubulin heterodimers. Notwithstanding its deep location, significant conformational changes in the protein are necessary for accommodating the inhibitors. Three zones of interaction can be identified (Figure 5). The main zone (zone 2) accommodates most of the structure of the ligands and is located on the b subunit. The two remaining are accessory zones, with one located at the a subunit interface (zone 1), and the other buried deeper in the b subunit (zone 3). A structural interaction fingerprint (SIFt) analysis[44] of all reported crystal structures reveals the partial overlap between the inhibitor binding sites (Figure 5). For this reason, it is convenient to refer to this region with the more extensive term colchicine binding domain instead of colchicine binding site. Empty colchicine domain The overlay of the apo form of tubulin (PDB ID: 3HKB) on the structure of tubulin complexed with colchicine (PDB ID: 1SA0) shows an important structural change. The T7 loop from Alab250 to Pheb244 moves far from S8S9 (Valb351Cysb356, Leub313Argb320), leading to an opening of the colchicine binding domain (Figure 6). Colchicine binding Colchicine (1) is known to bind the non-polymerized tubulin subunits in a two-step process. At the beginning, colchicine forms a reversible and low-affinity complex with the proteins; this is followed by a slow conformational change in tubulin which leads to the formation of a pseudo-irreversible complex.[45] The C ring of colchicine interacts in zone 1, establishing van der Waals contacts with Vala181, Sera178, and Valb315.
The carbonyl group behaves as a hydrogen bond acceptor, interacting with Val181a. The A ring is buried in a hydrophobic pocket (zone 2) delimited by Lysb352, Asnb350, Leub378, Alab316, Leub255, Lysb254, Alab250 and Leub242, and the methoxy group at position 3 is involved in a hydrogen bond interaction within the thiol group of Cysb241 (Figure 6 b). Podophyllotoxin binding Podophyllotoxin (4) prevents colchicine from binding to the colchicine binding domain by competitive inhibition. Its binding mode to tubulin is more rapid and reversible than that of colchicine.[19] Even though podophyllotoxin binds at the same binding site as colchicine, it adopts a slightly different orientation (Figure 6 c): the trimethoxyphenyl ring is embedded in the hydrophobic pocket occupied by the A ring of colchicine (van der Waals contact with Lysb352, Asnb350, Valb318, Alab317, Alab316, Leub255, Lysb254, Alab250 and Leub242), and the methoxy group at position 3 is involved in a hydrogen bond interaction with the thiol group of Cysb241. The A and B rings occupy the first hydrophobic region bounded by Valb315 and Metb259 (zone 1). E7010 binding E7010 (12) is a synthetic sulfonamide compound that has been evaluated as an anticancer drug.[46] As expected from binding competition assays,[47] it interacts with the colchicine site. The methoxy and pyridine groups of E7010 superimpose with the colchicine C and A rings, respectively; the sulfonamide overlaps with the B ring. Nevertheless, E7010 is more deeply buried than colchicine in b-tubulin. This has two consequences: first, E7010 interacts with strand S6 through a hydrogen bond between the phenolic group and Tyrb202; second, E7010 does not interact with the a-subunit (Figure 6 d). TN16 binding TN16 (13) has long been recognized as an inhibitor of microtubule assembly that competes with colchicine for tubulin bind-
www.chemmedchem.org
35
MED
A. Massarotti et al.
Figure 6. Structures of crystallized colchicine domains: a) PDB ID: 3HKB without ligand, b) 1SA0 with 1, c) 1SA1 with 4, d) 3HKC with 12, e) 3HKD with 13, f) 3HKE with 14, g) 3N2G with 15, and h) 3N2K with 16. Residues thought to interact with the compounds are represented as thin lines, compounds are shown in stick representation, and superimposition of colchicine [from panel b), PDB ID: 1SA0] is represented as a blue thin line in each case. Tubulin subunit ribbon structures are colored as in Figure 1; amino acid numbering is based on the 1SA0 crystal structure. Two views are shown for each panel: the right side in each case represents an approximate 908 upward rotation about the horizontal midline in the left-side view.
ing.[48] The overlap of the TN16 binding site with the colchicine site is limited to zone 2. TN16 is even more deeply buried in the b monomer than the E7010 ligand (Figure 6 e). The tubulin residues involved in binding belong to b strands S4, S5, and S6 (van der Waals contact with Thrb239, Valb238, Tyrb202, Glub200, and Pheb169). T13 binding T13 (14) has been reported to bind covalently and selectively to a subset of b-tubulin isotypes, disrupting microtubule polymerization.[49] Importantly, two different binding modes have been described in the same crystal. A T13 molecule occupies zone 2, largely overlapping with the binding mode of colchicines, while another T13 molecule is covalently linked to Cysb241. This is not surprising due to the high reactivity of the perfluorinated benzene ring. When T13 is not covalently linked to the protein, it interacts mainly with strand S9 (Leub313
Argb320) and loop T7 (Alab250Pheb244). The B ring is disordered and is directed toward the solvent (Figure 6 f). G2N and K2N binding G2N and K2N are enantiomers. They are mitotic inhibitors with in vivo and in vitro activity against various cell lines.[50] These two compounds are more deeply buried into b-tubulin than colchicine and they show no interactions with the a subunit (zone 1). There is just a partial overlap with colchicine that it is limited to the phenyl ring and to the A ring of colchicine (zone 2). Glub200 undergoes a hydrogen bonding interaction with the nitrogen atom of the aminopyridine group, while two additional hydrogen bonds are formed between Leub255 and Valb238 with the aniline group and the NH or the heterocyclic ring, respectively (Figure 6 g,h). The carbamate group is embedded in a hydrophobic pocket (van der Waals contact with Thrb239, Tyrb202, Asnb167, Glnb136, and Ileb4).
36
www.chemmedchem.org
Isotypes
Different isotypes for both a and b subunits are present in human cells. For the b subunit, eight isotypes are known and share a high degree of homology (90 % similarity; the major differences are related to the C-terminal region of the sequence, beyond residue 430) with different tissue distributions in normal cells (Table 2). For the a subunit at least six different isotypes are known. Microtubules incorporate all available isotypes.[51] In addition to a- and b-tubulin, other tubulin homologues have been identified (g, d, and e) with uncertain roles
show differences between the b-tubulin isoforms (Table 3, Figure 7). These mutations might play an important role in the stability of the tubulin-disrupting agents that bind at the colchicine domain. Two main mechanisms can be proposed: 1) These muTable 3. Tubulin isotype sequence variability in the colchicine domain. Isotype bIII bV bVI bVII bVIII V238I C241S C241S C241S Position V257M V315A V315A A317T A317T A317C T353V T353V T353V
Table 2. Tissue distributions of b tubulin isotypes in normal cells.
tations alter the direct interaction. This is related to the bI Constitutive All cells Serb241 mutation in place of bIIa/b Major neuronal isotype: Brain, nerves, muscle Restricted to particular cell type Cys (C241S). It is conceivable bIII Minor neuronal isotype: Brain Neurons only that the different orientations of Testis Sertoli cells Colon Epithelial cells only the side chains of Serb241 and bIVa Brain Neurons and glia Cysb241, as well as their differbIVb Constitutive High in ciliated cells, lower in others ent sizes (-OH versus -SH) could Minor isotype Almost all tissue except brain bV[52] partially explain the lower affinibVI Hematopoietic-specific bVII Brain Unknown ty of bIII for colchicine.[57] 2) These mutations induce a shape change in the binding in the life cycle of the cell.[52] Notably, different isotypes might site, limiting the selectivity of colchicine for the bII, bIII, and play a specific role in the stability of microtubules.[51] FurtherbIV isoforms.[58] This is related to the A317T and T353V mutamore, tubulin isoform expression is influenced not only by the tions,[51] which may disrupt the interactions with adjacent helitype of cells or tissue, but also by external conditions such as ces 9 and 10, resulting in a change in the binding site. anticancer therapy. In summary,[59] it is possible to suppose that alteration in the expression of the bI isoform, which increases the amount of For the purpose of this review, we focus mainly on the isobIII and bV, destabilizes the microtubules, conferring on them forms of b-tubulin, as it represents the main binding partner a significant degree of resistance to drugs that act as stabilizfor colchicine. The interest in b-tubulin isoforms began with ers, such as paclitaxel and colchicine. On the other hand, the the discovery that cells are capable of changing the expression overexpression of bI stabilizes the microtubules, imparting rerates of individual tubulin b isotypes in response to chemosistance to compounds that act as destabilizers, such as vincristherapy with paclitaxel or colchicine.[53] tine and vinblastine.[52] Several research groups have investigated the possibility of Currently there are no crystal structures available for the bdesigning a compound that can selectively bind to a specific tubulin isoforms. Regardless, the quantity of structural informaisotype.[55] The current microtubule-disrupting agents bind to all isotypes, having only a slight preference for one over anothtion about bI tubulin and the high degree of homology beer. The vinca alkaloids bind preferentially to bII, providing an tween different isotypes and bI is sufficient to allow homology explanation for their good activity against leukemia and Hodgmodeling in order to obtain 3D structures of each b-tubulin kins lymphoma, as these cancerous cells express high levels of isotype. Great efforts in this approach were carried out by the bII isotype.[56] Mane and co-workers.[51, 52, 54, 60] They performed homology modeling to obtain the various isoform structures and reportBased on sequence analysis of the b isotypes, it is apparent ed an in silico evaluation (free energy perturbation (FEP) analythat the major sequence variability occurs between bIII, bV, sis) of a colchicine analogue training set, identifying one candibVI, bVII, and bVIII. As mentioned above, the main differences date which should bind selectively to the abIII isoform.[55] are located at the C-terminal regions, which are rich in negatively charged residues. It is possible that the differences in dynamic instability between these b-tubulin isotypes is linked to differences in electrostatics and dipole moment.[52] Structure-based drug design By focusing on the colchicine domain and analyzing the colPharmacophore chicine binding mode proposed by Ravelli et al.,[37] it is possiAs discussed above, there are several chemical entities that are ble to identify 29 residues within a cutoff range of 6 that are able to bind tubulin, acting as disrupting agents. Most of them involved in colchicine binding.[37] Among these residues, six
Isotype[53, 54] Tissue/Organ expression Cellular expression
www.chemmedchem.org
37
MED
Figure 7. Representation of isotype differences in the colchicine domain: the amino acid variations between b isotypes within 6 of the bound colchicine molecule were selected and are shown in orange for the a) reference and b) mutated structures.
are not chemically related. With the aim of rationalizing their key common binding interactions at the colchicine domain, many pharmacophore models have been developed. Nguyen et al. proposed the most comprehensive one, identifying a common pharmacophore model by starting from a training set of 15 structurally unrelated derivatives.[61] It consists of seven pharmacophore points located in two different planes that intersect at an angle of ~ 458. There are three hydrogen bond acceptors: A1 in contact with Valb181, A2 in contact with Cysb241, and A3 establishing one contact mainly with Alab250, Aspb251, and Leub252. There is also a hydrogen bond donor, D1, which interacts with Thrb179, and finally two hydrophobic centers, H1 and H2, and a planar group R1 (Figure 8). Nguyen and colleagues established that one hydrogen bond acceptor, two hydrophobic centers, and a planar group were the minimum features required for drug activity.
This proposed pharmacophore model was taken into account by several other research groups[62, 63] as a reference or starting point for the development of new pharmacophore hypotheses. The relevance of this model was also confirmed in other publications that described the selected features for the authors pharmacophore models as being chemically and geometrically very similar to those proposed by Nguyen et al.[6466]
Docking studies Since the publication of the first tubulin structure with DAMAcolchicine co-crystallized into the active site, molecular docking has been used extensively to support the discussion of the biological data observed. Several research groups have used a variety of docking algorithms and scoring function combinations to obtain somewhat consistent results.[6777] Although a full review of all the simple docking studies reported is not
Figure 8. The seven-point pharmacophore model: a) red, magenta, yellow, purple, blue, cyan, and green spheres represent pharmacophore points A1, A2, A3, D1, H1, H2, and R1, respectively; b) points A1, A2, and A3 are hydrogen bond acceptors, D1 is a hydrogen bond donor, H1 and H2 are hydrophobic centers, and R1 is a planar group.
38
www.chemmedchem.org
Figure 9. Details of the colchicine domain: zones 13 are shown in stick representation and are colored respectively in orange, cyan, and magenta. The views in panels a) and b) are related analogously as those described for Figure 6.
practical here, it is possible to highlight some of the most common interactions observed. Figure 9 illustrates the residues that are in contact with the crystallized DAMA-colchicine compound. Following the same approach used above, it is possible to highlight three sub-pockets in the binding site, indicated by three different colors. The same color scheme is used in
Figure 10, in which the structural features of some of the most representative compounds reported (1730) are highlighted with the same color of the sub-pocket with which they interact. Upon examination of these images, it is apparent that all but one structure (compound 18, reported by Luo and coworkers)[76] present an aromatic group in contact with Cysb241
Figure 10. Compounds active on the colchicine domain, used in various docking studies. The structures are colored according to the colchicine domain zone interactions; see text for details.
www.chemmedchem.org
39
MED
(zone 2, cyan sub-pocket). This feature is often considered to be the most relevant, and indeed, the presence of this interaction is sometimes used as a filtering method in docking results.[68] Along with this interaction, another hydrophobic contact in the orange sub-pocket is commonly observed, often with an additional hydrogen bond with Valb181 (compounds 2025).[67, 69, 78] Furthermore, somewhat less common interactions are also reported in the more external part of this subpocket (compounds 19, 21, 2325, represented in red) as possible hydrogen bonds with Serb178 and Thrb179.[68, 69] Considering the studies reported, it appears that highly active colchicine-site-binding compounds occupy both the two sub-pockets discussed above (zone 12, cyan and orange). Indeed, there is a good correspondence between the pharmacophore points discussed above and the docking results covered in this section. These observations could provide an initial rational base in the design of novel tubulin inhibitors. It is also worth mentioning that Fortin et al. observed a further sub-pocket deep in the colchicine site (magenta sub-pocket),[70] although it seems that only relatively small groups can be accommodated at this position (compound 26). Finally, in a few cases (compounds 18, 25, 27, and 28) some structures are partially placed outside the colchicine domain, projecting toward the GTP site (black).[7476, 79] In this position no specific interaction are normally observed. Despite the good agreement that is often reported between structural information from docking results and biological data, good correlations between the docking score and tubulin polymerization inhibition data have rarely been observed.[80] This lack of scoring reliability has severely limited the use of molecular docking as a predictive tool in the design and discovery of novel tubulin binding compounds. To overcome this problem, several research groups have recently started to apply more computationally expensive techniques to evaluate ligand binding energies with tubulin. In particular, molecular dynamics (MD) simulations have been used after molecular docking to refine compound binding and to calculate interaction energies,[78] reporting a better correlation between in silico and in vitro results.[81] Interestingly, a docking/MD methodology has also been used to investigate the differences in binding energies of a select group of colchicine analogues in complex with various tubulin isoforms.[82] Indeed, this latter study represents the first step in the direction of designing novel tubulin binding agents that are selective for specific tubulin isoforms, with the support of computer-aided drug design. Virtual screening studies Few studies so far have reported the discovery of anti-tubulin agents by structure-based virtual screening. Kim et al. used the crystal structures of colchicine and podophyllotoxin in complex with tubulin to build a pharmacophore model and to screen a database of natural products; they discovered a chalcone derivative 31 with antimitotic activity (Figure 11).[83] Chiang et al. used a pharmacophore model based on an indole series of compounds to identify novel antimitotic agents (i.e., 32) with potent antiproliferative activity.[64] Knox and co-workers report-
Figure 11. Compounds active on the colchicine domain as selected by virtual screening studies; the color scheme is analogous to that used for Figure 10.
ed the first dual-targeting agent 33 for heat shock protein 90 (Hsp90) and tubulin. Their approach combined a docking/scoring protocol on Hsp90 with subsequent use of a receptorbased pharmacophore on tubulin.[84] Recently, Kim and co-workers,[66] using a pharmacophore model based on the above-mentioned study by Nguyen,[61] screened a focused library of 3,4,5-trimethoxyphenyl compounds selected from their in-house chemical database. Three compounds (e.g., 34) showed strong anti-proliferative activity against HL60 cell lines and were able to inhibit tubulin polymerization in vitro.
Outlook
Tubulin binding agents have played an important role in anticancer therapy since the approval of vincristine in 1963, and research in this field is still very active.[85] Among the classes of anti-tubulin agents, only the colchicine binding site agents have not yet reached the commercial phase. In this context, thanks to the amount of structural data that has recently become available, computer-aided drug design methodologies could play a pivotal role in the discovery of novel agents in this compound class. Indeed, from the publication of the first tubulinDAMA-colchicine 3D structure in 2004,[37] the use of molecular modeling applications in this field has increased dramatically. Although it is probably too early to make an assessment of the impact such applications have had on the discovery of novel therapeutic agents, it should be noted that the computer models reported in the literature are becoming increasingly accurate, especially when relatively rapid calculations such as pharmacophore queries and molecular docking are combined with more advanced methodologies like molecular dynamics. However, this research field still presents several challenges. The fact that all the available 3D structures of tubulin in complex with a colchicine analogue have poor resolution should not be overlooked (Table 1). This could explain the generally poor performance of the scoring functions and could also proChemMedChem 2012, 7, 33 42
40
www.chemmedchem.org
The Tubulin Colchicine Domain vide justification for why molecular dynamics is often required to improve model accuracy. A second point to keep in mind is that the tubulin structures available are derived from mammalian proteins, but none are from Homo sapiens. A final important element to consider is the various isoforms of tubulin and their specific roles. Indeed, this latter point represents a challenge as well as an exciting opportunity. The discovery of novel tubulin binding agents that could selectively target specific isoforms could have a significant impact in this research field, and molecular modeling could prove to be an essential tool in this scientific endeavor.
[25] A. Dowlati, K. Robertson, M. Cooney, W. P. Petros, M. Stratford, J. Jesberger, N. Rafie, B. Overmoyer, V. Makkar, B. Stambler, A. Taylor, J. Waas, J. S. Lewin, K. R. McCrae, S. C. Remick, Cancer Res. 2002, 62, 3408 3416. [26] G. J. Rustin, S. M. Galbraith, H. Anderson, M. Stratford, L. K. Folkes, L. Sena, L. Gumbrell, P. M. Price, J. Clin. Oncol. 2003, 21, 2815 2822. [27] J. P. Stevenson, M. Rosen, W. Sun, M. Gallagher, D. G. Haller, D. Vaughn, B. Giantonio, R. Zimmer, W. P. Petros, M. Stratford, D. Chaplin, S. L. Young, M. Schnall, P. J. ODwyer, J. Clin. Oncol. 2003, 21, 4428 4438. [28] C. J. Mooney, G. Nagaiah, P. Fu, J. K. Wasman, M. M. Cooney, P. S. Savvides, J. A. Bokar, A. Dowlati, D. Wang, S. S. Agarwala, S. M. Flick, P. H. Hartman, J. D. Ortiz, P. N. Lavertu, S. C. Remick, Thyroid 2009, 19, 233 240. [29] Study of Combretastatin and Paclitaxel/Carboplatin in the Treatment of Anaplastic Thyroid Cancer, US National Institutes of Health, http:// www.clinicaltrials.gov/ct2/show/NCT00507429?term = combretastatin (accessed September 20, 2011). [30] OXiGENE Clinical Trials Pipeline, http://www.oxigene.com/pipeline/ clinical_trials/ (accessed September 20, 2011). [31] C. Dumontet, M. A. Jordan, Nat. Rev. Drug Discovery 2010, 9, 790 803. [32] J. R. Peterson, T. J. Mitchison, Chem. Biol. 2002, 9, 1275 1285. [33] R. C. Weisenberg, G. G. Borisy, E. W. Taylor, Biochemistry 1968, 7, 4466 4479. [34] H. Mohri, Nature 1968, 217, 1053 1054. [35] B. Gigant, A. Cormier, A. Dorleans, R. B. Ravelli, M. Knossow, Top. Curr. Chem. 2009, 286, 259 278. [36] R. Bai, D. G. Covell, X. F. Pei, J. B. Ewell, N. Y. Nguyen, A. Brossi, E. Hamel, J. Biol. Chem. 2000, 275, 40443 40452. [37] R. B. Ravelli, B. Gigant, P. A. Curmi, I. Jourdain, S. Lachkar, A. Sobel, M. Knossow, Nature 2004, 428, 198 202. [38] E. Nogales, S. G. Wolf, K. H. Downing, Nature 1998, 391, 199 203. [39] B. Gigant, P. A. Curmi, C. Martin-Barbey, E. Charbaut, S. Lachkar, L. Lebeau, S. Siavoshian, A. Sobel, M. Knossow, Cell 2000, 102, 809 816. [40] B. Gigant, C. Wang, R. B. Ravelli, F. Roussi, M. O. Steinmetz, P. A. Curmi, A. Sobel, M. Knossow, Nature 2005, 435, 519 522. [41] A. Cormier, M. Marchand, R. B. Ravelli, M. Knossow, B. Gigant, EMBO Rep. 2008, 9, 1101 1106. [42] A. Dorleans, B. Gigant, R. B. Ravelli, P. Mailliet, V. Mikol, M. Knossow, Proc. Natl. Acad. Sci. USA 2009, 106, 13775 13779. [43] P. Barbier, A. Dorleans, F. Devred, L. Sanz, D. Allegro, C. Alfonso, M. Knossow, V. Peyrot, J. M. Andreu, J. Biol. Chem. 2010, 285, 31672 31681. [44] Z. Deng, C. Chuaqui, J. Singh, J. Med. Chem. 2004, 47, 337 344. [45] D. A. Skoufias, L. Wilson, Biochemistry 1992, 31, 738 746. [46] A. M. Mauer, E. E. Cohen, P. C. Ma, M. F. Kozloff, L. Schwartzberg, A. I. Coates, J. Qian, A. E. Hagey, G. B. Gordon, J. Thorac. Oncol. 2008, 3, 631 636. [47] K. Yoshimatsu, A. Yamaguchi, H. Yoshino, N. Koyanagi, K. Kitoh, Cancer Res. 1997, 57, 3208 3213. [48] T. Arai, FEBS Lett. 1983, 155, 273 276. [49] B. Shan, J. C. Medina, E. Santha, W. P. Frankmoelle, T. C. Chou, R. M. Learned, M. R. Narbut, D. Stott, P. Wu, J. C. Jaen, T. Rosen, P. B. Timmermans, H. Beckmann, Proc. Natl. Acad. Sci. USA 1999, 96, 5686 5691. [50] C. de Ines, D. Leynadier, I. Barasoain, V. Peyrot, P. Garcia, C. Briand, G. A. Rener, C. Temple, Jr., Cancer Res. 1994, 54, 75 84. [51] J. Y. Mane, M. Klobukowski, J. T. Huzil, J. Tuszynski, J. Chem. Inf. Model 2008, 48, 1824 1832. [52] R. Bhattacharya, F. Cabral, J. Biol. Chem. 2009, 284, 13023 13032. [53] H. K. Berrieman, M. J. Lind, L. Cawkwell, Lancet Oncol. 2004, 5, 158 164. [54] J. T. Huzil, R. F. Luduea, J. Tuszynski, Nanotechnology 2006, 17, S90S100. [55] C. Y. Tseng, J. Y. Mane, P. Winter, L. Johnson, T. Huzil, E. Izbicka, R. F. Luduea, J. A. Tuszynski, Mol. Cancer 2010, 9, 131. [56] I. A. Khan, R. F. Luduea, Invest. New Drugs 2003, 21, 3 13. [57] A. Banerjee, R. F. Luduea, J. Biol. Chem. 1992, 267, 13 335 13 339. [58] A. Banerjee, Y. Engelborghs, A. DHoore, T. J. Fitzgerald, Eur. J. Biochem. 1997, 246, 420 424. [59] C. Stengel, S. P. Newman, M. P. Leese, B. V. Potter, M. J. Reed, A. Purohit, Br. J. Cancer 2010, 102, 316 324. [60] E. J. Carpenter, J. T. Huzil, R. F. Luduea, J. A. Tuszynski, Eur. Biophys. J. 2006, 36, 35 43.
Acknowledgements
This work was supported by MIUR PRIN 2008 Italy (A.M., A.C., R.S. and G.S., grant 200879X9N9). A.C. gratefully acknowledges Istituto PasteurFondazione Cenci Bolognetti for his Borsa di Studio per Ricerche in Italia. Keywords: antitumor agents colchicine colchicine domain molecular modeling tubulin
[1] The Role of Microtubules in Cell Biology, Neurobiology, and Oncology (Ed.: T. Fojo), Humana Press, Totowa, NJ (USA), 2008. [2] A. Jordan, J. A. Hadfield, N. J. Lawrence, A. T. McGown, Med. Res. Rev. 1998, 18, 259 296. [3] P. Giannakakou, D. Sackett, T. Fojo, J. Natl. Cancer Inst. 2000, 92, 182 183. [4] P. M. Checchi, J. H. Nettles, J. Zhou, J. P. Snyder, H. C. Joshi, Trends Pharmacol. Sci. 2003, 24, 361 365. [5] M. A. Jordan, L. Wilson, Nat. Rev. Cancer 2004, 4, 253 265. [6] J. Bryan, L. Wilson, Proc. Natl. Acad. Sci. USA 1971, 68, 1762 1766. [7] R. F. Luduea, E. M. Shooter, L. Wilson, J. Biol. Chem. 1977, 252, 7006 7014. [8] G. van Loo, X. Saelens, M. van Gurp, M. MacFarlane, S. J. Martin, P. Vandenabeele, Cell Death Differ. 2002, 9, 1031 1042. [9] A. L. Risinger, F. J. Giles, S. L. Mooberry, Cancer Treat. Rev. 2009, 35, 255 261. [10] R. A. Stanton, K. M. Gernert, J. H. Nettles, R. Aneja, Med. Res. Rev. 2011, 31, 443 481. [11] The PyMOL Molecular Graphics System, version 1.3, Schrdinger LLC, 2010. [12] Molecular Modeling Approaches to Study the Binding Mode on Tubulin of Microtubule Destabilizing and Stabilizing Agents, M. Botta, S. Forli, M. Magnani, F. Manetti, T. Carlomagno in Tubulin-Binding Agents, Vol. 286, Springer, Berlin/Heidelberg, 2009, pp. 279 328. [13] R. Altaha, T. Fojo, E. Reed, J. Abraham, Curr. Pharm. Des. 2002, 8, 1707 1712. [14] A. Kruczynski, B. T. Hill, Crit. Rev. Oncol. Hematol. 2001, 40, 159 173. [15] E. K. Rowinsky, R. C. Donehower, N. Engl. J. Med. 1995, 332, 1004 1014. [16] M. J. De Brabander, R. M. Van de Veire, F. E. Aerts, M. Borgers, P. A. Janssen, Cancer Res. 1976, 36, 905 916. [17] E. Hamel, C. M. Lin, Biochemistry 1984, 23, 4173 4184. [18] M. A. Jordan, D. Thrower, L. Wilson, J. Cell Sci. 1992, 102, 401 416. [19] D. L. Sackett, Pharmacol. Ther. 1993, 59, 163 228. [20] N. J. Lakhani, M. A. Sarkar, J. Venitz, W. D. Figg, Pharmacotherapy 2003, 23, 165 172. [21] N. J. Lawrence, A. T. McGown, Curr. Pharm. Des. 2005, 11, 1679 1693. [22] G. R. Pettit, S. B. Singh, E. Hamel, C. M. Lin, D. S. Alberts, D. Garcia-Kendall, Experientia 1989, 45, 209 211. [23] J. Zhou, P. Giannakakou, Curr. Med. Chem. Anti-Cancer Agents 2005, 5, 65 71. [24] G. C. Tron, T. Pirali, G. Sorba, F. Pagliai, S. Busacca, A. A. Genazzani, J. Med. Chem. 2006, 49, 3033 3044.
www.chemmedchem.org
41
MED
[61] T. L. Nguyen, C. McGrath, A. R. Hermone, J. C. Burnett, D. W. Zaharevitz, B. W. Day, P. Wipf, E. Hamel, R. Gussio, J. Med. Chem. 2005, 48, 6107 6116. [62] B. Bhattacharyya, D. Panda, S. Gupta, M. Banerjee, Med. Res. Rev. 2008, 28, 155 183. [63] K. Nakagawa-Goto, P. C. Wu, C. Y. Lai, E. Hamel, H. Zhu, L. Zhang, T. Kozaka, E. Ohkoshi, M. Goto, K. F. Bastow, K. H. Lee, J. Med. Chem. 2011, 54, 1244 1255. [64] Y. K. Chiang, C. C. Kuo, Y. S. Wu, C. T. Chen, M. S. Coumar, J. S. Wu, H. P. Hsieh, C. Y. Chang, H. Y. Jseng, M. H. Wu, J. S. Leou, J. S. Song, J. Y. Chang, P. C. Lyu, Y. S. Chao, S. Y. Wu, J. Med. Chem. 2009, 52, 4221 4233. [65] F. M. Abdel Bar, M. A. Khanfar, A. Y. Elnagar, F. A. Badria, A. M. Zaghloul, K. F. Ahmad, P. W. Sylvester, K. A. El Sayed, Bioorg. Med. Chem. 2010, 18, 496 507. [66] N. D. Kim, E. S. Park, Y. H. Kim, S. K. Moon, S. S. Lee, S. K. Ahn, D. Y. Yu, K. T. No, K. H. Kim, Bioorg. Med. Chem. 2010, 18, 7092 7100. [67] C. Rappl, P. Barbier, V. Bourgarel-Rey, C. Gregoire, R. Gilli, M. Carre, S. Combes, J. P. Finet, V. Peyrot, Biochemistry 2006, 45, 9210 9218. [68] G. De Martino, M. C. Edler, G. La Regina, A. Coluccia, M. C. Barbera, D. Barrow, R. I. Nicholson, G. Chiosis, A. Brancale, E. Hamel, M. Artico, R. Silvestri, J. Med. Chem. 2006, 49, 947 954. [69] L. Cafici, T. Pirali, F. Condorelli, E. Del Grosso, A. Massarotti, G. Sorba, P. L. Canonico, G. C. Tron, A. A. Genazzani, J. Comb. Chem. 2008, 10, 732 740. [70] S. Fortin, L. Wei, E. Moreau, P. Labrie, E. Petitclerc, L. P. Kotra, C. G. R, Bioorg. Med. Chem. 2009, 17, 3690 3697. [71] R. Romagnoli, P. G. Baraldi, M. D. Carrion, C. L. Cara, O. Cruz-Lopez, M. Tolomeo, S. Grimaudo, A. Di Cristina, M. R. Pipitone, J. Balzarini, N. Zonta, A. Brancale, E. Hamel, Bioorg. Med. Chem. 2009, 17, 6862 6871. [72] R. lvarez, C. Alvarez, F. Mollinedo, B. G. Sierra, M. Medarde, R. Pelaez, Bioorg. Med. Chem. 2009, 17, 6422 6431. [73] A. S. Kiselyov, M. N. Semenova, N. B. Chernyshova, A. Leitao, A. V. Samet, K. A. Kislyi, M. M. Raihstat, T. Oprea, H. Lemcke, M. Lantow, D. G. Weiss, [74] [75]
N. N. Ikizalp, S. A. Kuznetsov, V. V. Semenov, Eur. J. Med. Chem. 2010, 45, 1683 1697. Y. Qian, G. Y. Ma, Y. Yang, K. Cheng, Q. Z. Zheng, W. J. Mao, L. Shi, J. Zhao, H. L. Zhu, Bioorg. Med. Chem. 2010, 18, 4310 4316. N. M. OBoyle, L. M. Greene, O. Bergin, J. B. Fichet, T. McCabe, D. G. Lloyd, D. M. Zisterer, M. J. Meegan, Bioorg. Med. Chem. 2011, 19, 2306 2325. Y. Luo, K. M. Qiu, X. Lu, K. Liu, J. Fu, H. L. Zhu, Bioorg. Med. Chem. 2011, 19, 4730 4738. M. A. Reddy, N. Jain, D. Yada, C. Kishore, J. R. Vangala, P. Surendra Reddy, A. Addlagatta, S. V. Kalivendi, B. Sreedhar, J. Med. Chem. 2011, DOI: 10.1021/jm200639r. F. Bellina, S. Cauteruccio, S. Monti, R. Rossi, Bioorg. Med. Chem. Lett. 2006, 16, 5757 5762. R. Romagnoli, P. G. Baraldi, M. G. Pavani, M. A. Tabrizi, D. Preti, F. Fruttarolo, L. Piccagli, M. K. Jung, E. Hamel, M. Borgatti, R. Gambari, J. Med. Chem. 2006, 49, 3906 3915. A. Tripathi, M. Fornabaio, G. E. Kellogg, J. T. Gupton, D. A. Gewirtz, W. A. Yeudall, N. E. Vega, S. L. Mooberry, Bioorg. Med. Chem. 2008, 16, 2235 2242. A. Coluccia, D. Sabbadin, A. Brancale, Eur. J. Med. Chem. 2011, 46, 3519 3525. J. T. Huzil, J. Mane, J. A. Tuszynski, Interdiscip. Sci. 2010, 2, 169 174. D. Y. Kim, K. H. Kim, N. D. Kim, K. Y. Lee, C. K. Han, J. H. Yoon, S. K. Moon, S. S. Lee, B. L. Seong, J. Med. Chem. 2006, 49, 5664 5670. A. J. Knox, T. Price, M. Pawlak, G. Golfis, C. T. Flood, D. Fayne, D. C. Williams, M. J. Meegan, D. G. Lloyd, J. Med. Chem. 2009, 52, 2177 2180. I. S. Johnson, J. G. Armstrong, M. Gorman, J. P. Burnett, Jr., Cancer Res. 1963, 23, 1390 1427.
[76] [77]
[78] [79]
[80]
[81] [82] [83] [84] [85]
Received: July 21, 2011 Revised: September 14, 2011 Published online on October 12, 2011
42
www.chemmedchem.org
DOI: 10.1002/cmdc.201100500
In vivo Enhancement in Bioavailability of Atazanavir in the Presence of Proton-Pump Inhibitors using Mesoporous Materials
Xin Xia,[b, c] Chunfang Zhou,[b] Llus Ballell,[d] and Alfonso E. Garcia-Bennett*[a]
Poor aqueous solubility is acknowledged as one of the major barriers in the development of novel pharmaceutical agents, inhibiting their progress into clinical studies even from an early toxicological assessment of the compound. More than 40 % of newly identified lead compounds are lipophilic, resulting in poor pharmacokinetic properties.[1] Insufficient dissolution caused by inherent crystallization behavior, molecular structure, or unfavorable drugdrug interactions, results in inferior drug absorption and bioavailability, which often leads to the need for high dosing in order to reach therapeutic levels in blood plasma. Organic and inorganic nanomaterials are currently being studied as drug delivery vehicles, to achieve supersaturation in the gastrointestinal environment, which could lead to enhanced bioavailability.[2] Traditional formulation strategies rely on administering the poorly water-soluble compound as a solution in the presence of surfactants, complexing agents, such as cyclodextrins, or as emulsions.[35] Increasing the surface area and curvature of the drug particles, delivery of the drug in an amorphous state, as well as the addition of polymeric crystallization inhibitors, are techniques that have been explored as a means to facilitate and prolong a supersaturation state of a compound in the gastrointestinal lumen.[69] Many of these methods, however, are costly and do not stabilize the supersaturation state sufficiently, nor do they resolve issues associated with drugdrug interactions. Mesoporous silica-based particles have emerged as suitable excipients with a complete set of desirable properties in comparison to organic-based nanostructured drug carriers, such as liposomes or dendrimers.[1012] The potential applications of mesoporous silica-based particles in pharmaceuticals and nanomedicine have recently been reviewed,[1315] and these materials have been shown to be adequate vehicles for the controlled and targeted release of a variety of drugs, both in vitro and in vivo.[16, 17] An itraconazole formulation with ordered mesoporous silica has been developed by Mellaerts et al.,[18] where the solubility of itraconazole, a weak base, was enhanced in simulated gastric fluid (SGF), and where the precipitation of the compound was decreased in simulated intestine fluid (SIF). Recently Van Speybroeck et al. compared the release rate of fenofibrate from mesoporous silica particles with different pore sizes, reporting that the dissolution rate of fenofibrate increased with decreasing pore size under supersaturating conditions.[19] Herein, we describe the results of our investigation into the relationship between mesoporous structure and solubility enhancement using mesoporous silica structures NFM-1 (cylindrical, 2d-hexagonal, small pore; Figure 1 a),[20] AMS-6 (cylindrical 3d-cubic and small pore; Figure 1 b),[21] and STA-11 (cylindrical 3d-cubic, large pore; Figure 1 b)[22] in the context of drugdrug interactions. We performed in vivo bioavailability studies in a rat model to observe how these improvements translate into actual oral bioavailability data. The compound chosen for our study was atazanavir (ATV, Figure 1), a antiretroviral protease inhibitor used for the treatment of infection by human immunodeficiency virus (HIV). Worldwide, over 40 million people are infected with HIV. The high activity antiretroviral therapy (HAART) combines at least three antiretroviral drugs and has been used to extend the lifespan of HIV-infected patients. Chronic use of HAART is needed to control HIV infection. The frequent administration of several drugs in relatively high doses is the main cause of patient noncompliance and a hurdle toward the efficacy of the pharmacotherapy.[23]
[a] Dr. A. E. Garcia-Bennett Nanotechnology and Functional Materials The ngstrom Laboratory, Uppsala University Lgerhyddsvgen 1, Polacksbacken, 75121 Uppsala (Sweden) E-mail: alfonso.garcia@angstrom.uu.se Homepage: www.angstrom.uu.se [b] X. Xia, Dr. C. F. Zhou Nanologica AB, Drottning Kristinas vg 45, 11428 Stockholm (Sweden) [c] X. Xia Department of Materials and Environmental Chemistry Arrhenius Laboratory, Stockholm University Svante Arrhenius vg 16C, 10691 Stockholm (Sweden) [d] Dr. L. Ballell Diseases of the Developing World, GlaxoSmithKline Severo Ochoa 2, 28769 Tres Cantos, Madrid (Spain) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100500. Figure 1. Unit cell model of mesoporous structures with a) cylindrical 2dhexagonal symmetry and b) cylindrical 3d-cubic symmetry of mesoporous particles used in this work. c) Antiretroviral atazanivir (ATV) with partition constant: log P = 5.20.
43
MED
therms, lies in a similar range (~ 800 m2 g1) whilst the pore ATV is a lipophilic drug with partition constant (log P) value of 5.20. Despite four violations of the Lipinskis Rule of Five, its volume is double for the cubic 3D-structures compared with bioavailability is between 6068 % and it has a half-life of 6.5 h that of the hexagonal particles. when administered orally.[24] However, the bioavailability of ATV ATV was loaded into the mesoporous materials via an evaporation protocol using methanol as a solvent, since the solubilis severely hampered, by as much as 78 % reduction in plasma concentration, when it is coadministered with proton-pump inTable 1. Textural properties of mesoporous silica used in the study. hibitors, leading to a significant decrease in its effectiveness.[25] Proton-pump inhibitors are administered to HIV patients to Mesoporous Surface Pore Total pore Particle treat the symptoms of heartburn and stomach pains that are silica area[a] width[b] volume[c] size[d] common secondary effects of HIV medication. This decrease in [m2 g1] [] [cm3 g1] [mm] bioavailability is caused by precipitation of the drug under the NFM-1 793 (322) 26 (24) 0.40 (0.16) 10.16 less acidic conditions caused by the proton-pump inhibitors. (2dhexagonal) AMS-6 847 (284) 43 (37) 0.80 (0.24) 9.44 Hence, development of a formulation capable of maintaining (3dcubic) high bioavailability when ATV is coadministered with protonSTA-11 847 (216) 79 (71) 0.85 (0.27) 9.85 pump inhibitors is likely to lead to a considerable improvement (3dcubic) in patient comfort, as well as effectiveness of the treatment. [a] BrunauerEmmettTeller (BET) surface area. [b] Value obtained by apCalcined mesoporous materials NFM-1, AMS-6 and STA-11 plying a density functional theory (DFT) model and assuming a cylindrical were chosen as model ATV carriers due to their varying texturpore geometry. [c] Measured at 0.98 in relative pressure. [d] Average al and structural characteristics (Table 1). NFM-1 possesses a value from dynamic light scattering (DLS) analysis performed in SIF. Values in brackets are properties of ATV-loaded samples. two-dimensional (2D) hexagonal network (space group: p6mm, Figure 2) of cylindrical pores with relatively small mesopore size (26 ). In contrast, both AMS-6 and STA-11 possess three-dimensional (3D) cylindrical cubic structures (space " group: Ia3d), hence will possess faster diffusion coefficients,[26] in comparison to the 2D-hexagonal structure. Low angle X-ray diffraction (XRD; see Supporting Information for details) and transmission electron microscopy (Figure 2) reveal highly ordered bicontinuous cubic structures for STA-11 and AMS-6, and a hexagonal structure for NFM-1. The pore size of the two cubic structured mesoporous materials, AMS-6 and STA-11, is 43 and 78 , respectively, as determined by nitrogen sorption isotherms. The particles vary in size with the following trend, AMS6 < STA-11 < NFM-1 (Figure 2 and Table 1), as determined by scanning electron microscopy (TEM); however, when the particles are in solution, agglomeration results in a larger average particle size for all materials, as measured by dynamic light scattering (DLS) in simulated intestine fluid (SIF). The Brunauer EmmettTeller (BET) surface area Figure 2. Transmission (left) and scanning (right) electron microscopy images of mesoporous materials: a) STA-11 for all three materials, as deter- along the [110] orientation and showing spherical morphologies, b) AMS-6 recorded along the [111] orientation mined by nitrogen sorption iso- and spherical particles, and c) NFM-1 along the [001] directions and showing rod-type morphologies.
44
www.chemmedchem.org
ity of ATV is low in aqueous media (3.27 mg L1 at pH 7), even under acidic condition. Themogravimetric (TG) analysis was used to determine the amount of ATV loading in the different mesoporous solids. Raw data TG curves show a loading capacity of 28.2, 31.5 and 32.8 wt % for STA-11, NFM-1 and AMS-6, respectively (for TG curves, see Supporting Information, ESI-2). The decomposition curves for all loaded mesoporous structures are consistent with those observed for the free (crystalline) drug. Nitrogen sorption isotherms, performed post-loading, are useful to determine the location of the drug. The BET surface area of loaded NFM-1-ATV decreases to 322 m2g1, while the total pore volume decreases to 0.166 cm3g1 (see Supporting Information, ESI-3). The results for all mesostructures are consistent with a loading of ATV inside the mesopores that are not fully filled at these loading amounts. As shown in Figure 3, no XRD peaks could be detected for any of the loaded-ATV mesoporous materials, a strong indication that the drug resides in its amorphous state, and within the pores. Figure 4 shows in vitro dissolution curves performed in SIF, for mesoporous loaded and free drug. Dissolution curves in SGF (pH 1.2, see Supporting Information, ESI-4) under sink conditions do not show a significant improvement in the solubility of ATV released from mesoporous particles in comparison to that of the free drug, and rapid solubilization is observed for all mesoporous materials. In SIF, the free drug shows a poor solubility profile due to the higher pH level (Figure 4 a). All mesoporous materials yield a significant enhancement in dissolution. ATV-loaded STA-11 and AMS-6 release approximately 45 % of the drug content after 40 min, while around 40 % is released from ATV-loaded NFM-1. STA-11-ATV reached the highest dissolution percentage after 4 h, while AMS-6-ATV and NFM-1-ATV took 7 h to reach the maximum dissolution value (87 %, Figure 4 b). The percentage of drug released from all of the mesoporous silica decreased after 8 h; the smallest decrease was observed for NFM-1-ATV compared with the other two mesoporous materials. This is presumably due to a recrystallization of the free drug soon after release. Analysis by 1 H nuclear magnetic resonance (NMR) performed on samples
of ATV released from NFM-1 does not show any indication of drug decomposition (see Supporting Information, ESI-5). To investigate further the extent of the solubility enhancement using mesoporous silica carriers, dissolution curves were obtained for NFM-1-ATV at different particles-to-SIF ratios: 20 mg L1, 40 mg L1, 80 mg L1 and 160 mg L1 (Figure 4 d). With increased particle concentration, the dissolution rate appears to decrease as a percentage of amount loaded. However, the actual solubility suggests that a higher concentration of NFM-1-ATV particles (160 mg L1) provides greater solubility, whereby the maximum solubility reached was 18 mg L1 after 4 h, in comparison to 0.254 mg L1 after a similar time for free ATV. In the case of the highest particle concentration, a supersaturation state is maintained for 4 h, after which time the amount of solubilized ATV decreases rapidly. Dissolution parameters are given in Table 2, and the curves can be fitted for both the power law and Higuchi model, supporting a diffusion-based mechanism of release of ATV from NFM-1 particles. The curves are consistent with variations in concentration, showing that mesoporous materials are versatile excipients for controlled release. Stabilization of the amorphous state of compounds in carbon nanotubes due to confinement has previously been discussed by Prasad et al. and later by Beiner et al. with respect to pharmaceutical compounds.[27, 28] Our in vitro results are consistent with their model of suppression of crystallization, where below a certain critical pore diameter (d*), the surface energy contributions are larger than the energetic gain upon crystallization. This thermodynamic effect increases with decreasing pore size, according to Equation (1):[2730] d * 4scl T m 1 =T m 1 TDHm 1c 1
Figure 3. High-angle, X-ray diffraction patterns of free atazanavir (ATV), STA-11-ATV, AMS-6-ATV and NFM-1-ATV.
where scl is the surface energy between crystal and melt, DHm is the heat of melting, Tm1 is the bulk melting temperature, and 1c is the crystal density. Further to thermodynamic concerns, nanoconfinement will decrease the crystallization kinetics of the compound through a dispersion of nucleation sites within the porous structure and an immobilization of crystallization nutrients within the pores. Considering a polar surface area of 125.06 2 for ATV, it is unlikely that crystallization can proceed within pores smaller than 1314 nm in diameter.[31] Furthermore, the crystal structure of ATV has unit cell dimensions of a = 9.86 , b = 29.245 and c = 8.327 , making it unlikely that the dimensions of the pore in NFM-1 would allow any crystal formation.[32] The stability of the amorphous state of loaded ATV within mesoporous materials shown here has been checked for periods of longer than one year (data not shown), and no additional peaks suggesting crystallization or structural rearrangement of the drug within the pores have been observed. An in vivo pharmacokinetic study was conducted in order to validate the enhancement in dissolution observed for ATV, in the context with coadministration with proton-pump inhibitors, for the best performing mesoporous silica NFM-1. A single oral dose of 10 mg kg1 ATV or 50 mg kg1 NFM-1-ATV (20 wt % loading of ATV) was given to female Sprague Dawley
www.chemmedchem.org
45
MED
Figure 4. a) Dissolution curves of free () and loaded ATV (STA-11-ATV: ^; AMS-6-ATV: &; NFM-1-ATV: ~) in simulated intestine fluid (SIF, 40 mg L1, sink). b) Maximum dissolution of ATV released from mesoporous silica in SIF. Dissolution curves for free (J) and loaded NFM-1-ATV (^: 20 mg L1; &: 40 mg L1; ~: 80 mg L1; : 160 mg L1) with varied particle/SIF ratio as c) a percentage of solubilized ATV, and d) compound concentration.
Table 2. Parameters of power law and Higuchi equations for the release curves of loaded ATV samples. Sample [mg L1] NFM-1-ATV (20) NFM-1-ATV (40) NFM-1-ATV (80) NFM-1-ATV (160) Power law: ln F = ln kp + n ln t n kp R2 0.47 0.52 0.55 0.58 0.44 0.34 0.24 0.17 0.98 0.99 0.99 0.98 Higuchi: F = kH t50 % kH R2 t50 % [h] 0.41 0.32 0.25 0.17 0.97 0.98 0.98 0.98 1.50 2.41 4.07 8.46
kp : kinetic constant; n: release exponent; R: coefficient of correlation; t50 %: time taken to release 50 % of ATV.
rats (n = 3), fasted overnight prior to the experiment, approximately 5 h after administration of a proton-pump inhibitor (omeprazole, 100 mg kg1; see the Experimental Section for full details). Blood samples were collected in ethylenediaminetetraacetic acid (EDTA) tubes at six different time points after administration: 0.5 h, 1 h, 2 h, 4 h, 8 h and 24 h, and blood
plasma concentrations of ATV were determined using HPLCMS. The pharmacokinetic profile of whole-blood plasma ATV concentration for rats treated with NFM-1-ATV in comparison to those treated with free ATV is shown in Figure 5. A very pronounced improvement in ATV absorption is observed during the first hour of the study. Overall, a statistically significant improvement in bioavailability was observed for NFM-1-ATV in comparison with the free drug. The maximum concentration achieved (Cmax) for free ATV in the present study was 18.35 ng mL1. In contrast, NFM-1-ATV results in a Cmax value of 85.25 ng mL1. The total area under the curve (AUC) for an eight-hour period shows a similar contrast in bioavailability, with values of 271.91 ng h mL1 and 65.84 ng h mL1 for NFM-1-ATV and ATV, respectively. Two absorption effects are apparent from Figure 5: within the first hour after oral administration, a burst effect can be seen owing to the rapid absorption of solubilized drug; next, between 18 h, the data presumably indicate absorption of solubilized ATV released from the core of the mesoporous parChemMedChem 2012, 7, 43 48
46
www.chemmedchem.org
sized silica materials were calcined at 550 8C for 6 h. After calcination, the materials were stored in airtight containers. Characterization of the texture and structure is summarized in Table 2 and in the Supporting Information; full details of the synthetic protocols are also given there. Characterization of porosity: A nitrogen adsorptiondesorption isotherm was applied to the calcined mesoporous silica powders for the characterization of porosity using a Micromeritics Tristar II 3020 apparatus (Norcross, GA, USA). All samples were degassed at 300 8C (40 8C for drug-loaded samples) for 6 h under a flow of nitrogen. The surface area was calculated using the Brunauer EmmettTeller (BET) equation in the relative pressure range between 0.05 and 0.2.[33] The pore-size distribution was calculated by density functional theory (DFT) assuming a cylindrical pore geometry.[34, 35] The total pore volume was calculate using the t-plot method.[36] Scanning electron microscopy (SEM): SEM images were obtained using a JSM-7401F scanning electron microscope (JEOL Ltd., Tokyo, Japan) operating at 12 kV with no gold coating. Transmission electron microscopy (TEM): TEM images of calcined samples were obtained using a JEOL-3010 transmission electron microscope (JEOL Ltd., Tokyo, Japan) operating at 300 kV (Cs = 0.6 mm, resolution = 1.7 ). Images were recorded using a chargecoupled device (CCD) camera with specialized imaging shadowgraph (SIS) analysis (size = 1024 1024 megapixels, pixel size = 23.5 23.5 mm; model: Keen View, Olympus Soft Imaging Solutions, Mnster, Germany) at 30000100000 magnification using lowdose conditions on calcined samples. Thermogravimetric analyses: The loading amount was determined using a thermogravimetric analysis instrument (model TGA7, PerkinElmer, MA, USA). Scanning was performed at 20!900 8C at a heating rate of 20 8C min1. The plug-in gas atmosphere was dry air (flow rate = 20 mL min1). The sample weights varied from 5 mg to 10 mg. The derivative weight loss was calculated using the PyrisInstrument Managing software version 10.1 (PerkinElmer, MA, USA). Particle size analysis by dynamic light scattering: The particle size in simulated intestine fluid (SIF) was measured on a Mastersizer 2000 particle size analyzer (Malvern instruments Ltd., Worcestershire, UK). Media for dissolution experiments: For use as a buffer, SIF was prepared by dissolving NaOH (0.896 g, 0.0224 mol) and KH2PO4 (13.61 g, 0.10 mol) in purified water (1 L; ) giving a pH of 6.8. GlycineHCl buffer was used as simulated gastric fluid (SGF) and contained NaCl (0.7 g, 0.012 mol), glycine (1 g, 0.013 mol) and aq HCl (63.2 mL, 1 m) in purified water (1 L) to give a pH of 1.2. All chemicals were purchased from SigmaAldrich. Purified water was obtained using a RiOsTM 5 water system (Millipore SAS, Molsheim, France). Drug loading: A rotary evaporating method was applied to incorporate atazanavir with the mesoporous silica. In brief, a concentrated drug solution in methanol was treated with a sample of mesoporous silica (200 mg), and the solven was removed by rotary evaporation at 40 8C. Samples were left to dry open to the air overnight. Drug release: The in vitro release performance of drug-loaded, ordered mesoporous silica was assessed in SIF (pH 6.8) and SGF (pH 1.2). All release experiments showed good reproducibility. The release profile was obtained by means of UV/Vis absorbance scan
Figure 5. Pharmacokinetic profile for ATV in whole-blood samples taken from Sprague Dawley rats treated with NFM-1-ATV (~) and ATV free-base (&). Data points represent the mean standard deviation (n = 3 animals per time point).
ticle. However, this explanation neglects other pharmacokinetic and absorption processes that might be taking place, as well as the pharmacokinetic profile of omeprazole. Further studies are required to fully understand the in vivo mechanism of improved bioavailability derived through the use of mesoporous silica particles. In conclusion, mesoporous materials with different pore structures and sizes have been evaluated as suitable drug solubility enhancers. All mesoporous materials studied here showed a considerable improvement in the dissolution of atazanavir, under sink conditions in SIF. Smaller pores sizes relative to the size of the drug appear to be more important than pore connectivity (3D or 2D) or drug diffusion within the pores, and materials with these smaller pores achieved slightly higher drug dissolution. Results suggest that the mechanism of dissolution enhancement is based on the nanoconfinement of the drug within the pores of the mesoporous material, leading to a suppression of crystallization. In our study, mesoporous silica with a 2D-hexagonal structure (NFM-1) exhibited a diffusioncontrolled mechanism of release, and led to an enhancement of solubility equivalent to 71 times higher than a similar dose of the unloaded, free drug. These values were reproduced in vivo, in a single administration study in rats, and the data obtained was consistent with a fourfold enhancement in AUC and Cmax in the presence of proton-pump inhibitors. These results demonstrate the versatility of mesoporous drug carriers in solving formulation and pharmacokinetic problems associated with drugdrug interactions, which not only affect the efficiency of therapeutic treatments but can also cause severe patient discomfort, as is the case in the treatment of HIV using antiretroviral drugs such as atazanavir.
Experimental Section
Chemistry
Synthesis of mesoporous silica: Three different types of mesoporous silica have been synthesized according to previously described procedures: STA-11,[22] AMS-6[21] and NFM-1.[20] All syntheChemMedChem 2012, 7, 43 48
www.chemmedchem.org
47
MED
using a UV/Vis 920 spectrometer (GBC Scientific Equipment Pty Ltd, Braeside, Australia).
[4] R. G. Strickley, R. Oliyai in Solvent Systems and their Selection in Pharmaceutics and Biopharmaceutics, (Eds.: P. Augustijns, M. E. Brewster), AAPS Press and Springer, New York, 2007, pp. 257 308. [5] L. Appel, W. Babcock, D. T. Friesen, R. J. Ray, S. Shamblin, R. Shanker, D. Smithey, (Pfizer Products Inc., Groton, CT, USA), Int. Pat. Appl. WO 2006/ 024944 A2, 2006. [6] M. E. Matteucci, B. K. Brettmann, T. L. Rogers, E. J. Elder, R. O. Williams III, K. P. Johnston, Mol. Pharm. 2007, 4, 782 793. [7] K. A. Overhoff, J. T. McConville, W. Yang, K. P. Johnston, J. I. Peters, R. O. Williams III, Pharm. Res. 2008, 25, 167 175. [8] A. T. Serajuddin, J. Pharm. Sci. 1999, 88, 1058 1066. [9] C. Leuner, J. Dressman, Eur. J. Pharm. Biopharm. 2000, 50, 47 60. [10] A. R. Mohammed, N. Weston, A. G. A. Coombes, M. Fitzgerald, Y. Perrie, Int. J. Pharm. 2004, 285, 23 34. [11] M. T. Morgan, Y. Nakanishi, D. J. Kroll, A. P. Griset, M. A. Carnahan, M. Wathier, N. H. Oberlies, G. Manikumar, M. C. Wani, M. W. Grinstaff, Cancer Res. 2006, 66, 11913 11921. [12] V. Swali, N. J. Wells, G. J. Langley, M. Bradley, J. Org. Chem. 1997, 62, 4902 4903. [13] J. Salonen, L. Laitinen, A. M. Kaukonen, J. Tuura, M. Bjrkqvist, T. Heikkil, K. Vh-Heikkil, J. Hir-vonen, V. Lehto, J. Controlled Release 2005, 108, 362 374. [14] A. M. Kaukonen, L. Laitinen, J. Salonen, J. Tuura, T. Heikkil, T. Limnell, J. Hirvonen, V. Lehto, Eur. J. Pharm. Biopharm. 2007, 66, 348 356. [15] R. Mellaerts, C. A. Aerts, J. Van Humbeeck, P. Augustijns, G. Van den Mooter, J. A. Martens, Chem. Commun. 2007, 1375 1377. [16] M. Vallet-Reg, F. Balas, D. Arcos, Angew. Chem. 2007, 119, 7692 7703; Angew. Chem. Int. Ed. 2007, 46, 7548 7558. [17] J. Lu, M. Liong, Z. Li, J. I. Zink, F. Tamanoi, Small 2010, 6, 1794 1805. [18] R. Mellaerts, R. Mols, P. Kayaert, P. Annaert, J. Van Humbeeck, G. Van den Mooter, J. A. Martens, P. Augustijns, Int. J. Pharm. 2008, 357, 169 179. [19] M. Van Speybroeck, R. Mellaerts, R. Mols, T. D. Thi, J. A. Martens, J. Van Humbeeck, Eur. J. Pharm. Sci. 2010, 41, 623 630. [20] R. Atluri, N. Hedin, A. E. Garcia-Bennett, J. Am. Chem. Soc. 2009, 131, 3189 3191. [21] A. E. Garcia-Bennett, S. Che, K. Miyasaka, O. Terasaki, Chem. Mater. 2004, 16, 3597 3605. [22] R. P. Hodgkins, A. E. Garcia-Bennett, P. A. Wright, Micro. Meso. Mater. 2005, 79, 241 252. [23] A. Sosnik, D. A. Chiappetta, A. M. Carcaboso, J. Controlled Release 2009, 138, 2 15. [24] C. A. Lipinski, F. Lombardo, B. W. Dominy, P. J. Feeney, Adv. Drug Delivery Rev. 2001, 46, 3 26. [25] Prescribing information for atazanavir sulfate (Reyataz) capsules. Bristol Myers Squibb (Princeton, NJ, USA), July 2004. [26] M. Strmme, U. Brohede, R. Atluri, A. E. Garcia-Bennett, Wiley Interdiscip. Rev.: Nanomed. Nanotechnol. 2009, 1, 140 148. [27] R. Prasad, S. Lele, Philos. Mag. Lett. 1994, 70, 357 361. [28] G. T. Rengarajan, D. Enke, M. Steinhart, M. Berner, J. Mater. Chem. 2008, 18, 2537 2539. [29] C. L. Jackson, G. B. McKenna, Chem. Mater. 1996, 8, 2128 2137. [30] M. Alcoutlabi, G. B. McKenna, J. Phys. Condens. Matter 2005, 17, R461 R524. [31] The structure and properties of atazanavir are summarized here: http:// www.chemspider.com/Chemical-Structure.130642.html (Last accessed: November 10, 2011). [32] S. Kim, B. T. Lotz, J. Z. Gougoutas, M. Davidovich, S. K. Srivastava, (Bristol Myers Squibb Co.), US Pat. Appl. US 2011/0124689 A1, 2011. [33] S. Brunauer, P. H. Emmett, E. Teller, J. Am. Chem. Soc. 1938, 60, 309 319. [34] P. Tarazona, Phys. Rev. A 1985, 31, 2672 2679. [35] P. Tarazona, U. Marini Bettolo Marconi, R. Evans, Mol. Phys. 1987, 60, 573 595. [36] B. C. Lippens, J. H. De Boer, J. Catal. 1965, 4, 319 323. [37] Y. Takehara, K. Sumii, A. Tari, M. Yoshihara, M. Sumii, K. Haruma, G. Kajiyama, S. V. Wu, J. H. Walsh, Am. Physiol. Soc. 1996, 271, G799 804.
In vivo bioavailability studies

All experiments were conducted by Vivotecnia Research (Tres Cantos, Spain), according to the good laboratory practice (GLP) requirements of European Directive 2004/10/EC and Spanish law RD 1369/2000. Animals: Sprague Dawley rats (67 weeks in age; Harlan laboratories) were used throughout this study. Rats were kept (and bred) at the animal facility at Vivotecnia Research. All animals were allowed to acclimatize for at least 5 days prior to the experiments. Dosing: All rats were dosed orally with 100 mg kg1 of omeprazole (proton-pump inhabitor; Maxplus Int. Ltd, Lianyungang, China) 5 h prior to the administration of atazanavir (10 mg kg1),[37] which was given as either a silica-based formulation or the free base (purchased from Legend Pharmaceutical & Chemical Co., Ltd., Nanjing, China). All powders were administrated by oral gavage as a suspension in phosphate buffered saline 1X (PBS1X; Invitrogen, Carlsbad, CA, USA). Blood sampling: The time points for blood extraction were 0.5 h, 1 h, 2 h, 4 h, 8 h and 24 h after NFM-1-ATV/ATV oral administration. All animals were bled once during the study period. The predose samples (0 h) were obtained from three out of the four spare animals. At each time point, at least 0.5 mL of whole blood were extracted from isofluorane-anesthetized animals by sublingual puncture and transferred directly to commercially available ethylenediaminetetraacetic acid (EDTA)-containing tubes (BD-Bioscience). Tubes containing EDTA/whole blood samples were kept at room temperature under rotation until each sample was divided into two different aliquots of ~ 250 mL. Thereafter, the EDTA/whole blood aliquots were immediately frozen in an upright position and stored at 80 8C 10. The analytical determination of atazanavir concentration in EDTA/whole blood samples from animals treated with NFM-1-ATV/ATV was performed by HPLC-MS/MS.
Acknowledgements
AEGB is supported by the Swedish Research Council (Vetenskaprdet). The authors are grateful for support from the Seventh Framework Programme of the European Commission: EC-FP-7ORCHID (grant no. 261378) and EC-FP-7-DENDREAMERS (grant no. 215884). The authors express their gratitude to Dr. Albert Mihranyan (Uppsala University, Sweden) for many helpful comments during the writing of this manuscript, and to Mr. Richard Lihammar (Stockholm University, Sweden) for performing NMR analysis. Keywords: drug delivery human immunodeficiency virus (HIV) mesoporous silica proton-pump inhibitors solubility
[1] Part 1: Oral Delivery of Poorly Soluble Drugs, M. Hite, S. Turner, C. Federici, Pharm. Manuf. Pack. Sourc. (PMPS) 2003, summer; available here: www.scolr.com/lit/PMPS_2003_1.pdf (Last accessed: November 10, 2011). [2] H. Hou, R. A. Siegel, J. Pharm. Sci. 2006, 95, 896 905. [3] R. G. Strickley, Pharm. Res. 2004, 21, 201 230.
Received: October 27, 2011 Published online on December 5, 2011
48
www.chemmedchem.org
DOI: 10.1002/cmdc.201100428
Exhaustive Fluorine Scanning toward Potent p53Mdm2 Antagonists

Yijun Huang,[a] Siglinde Wolf,[b] David Koes,[c] Grzegorz M. Popowicz,[b] Carlos J. Camacho,[c] Tad A. Holak,[b] and Alexander Dmling*[a]
Dedicated to George Olah, a father of modern fluorine chemistry, on the occasion of his 85th birthday. The unmatched properties of the element fluorine are reflected in organic fluorine compounds. The richness of modern fluorine chemistry allows for the regioselective introduction of this element at virtually any position of a given molecule.[1, 2] Thus the introduction of fluorine has been especially valuable in the process of drug discovery to fine-tune many different target- and off-target-related properties.[36] For example, fluorine has been used to increase the binding affinity of small molecules to their targets,[7] to tune pKB and log D values,[8] to improve target selectivity,[9] to improve oral absorption and exposure,[10] to prevent from metabolism,[11] or to increase the antibacterial spectrum.[12] Not surprisingly, an estimated 20 % of all pharmaceuticals contain fluorine.[4] Additionally, 18F is often used as a positron emission tomography (PET)-active element to perform time- and space-resolved distribution studies of drugs in animals and humans.[13, 14] Last but not least, 19F is very useful in the NMR spectroscopy of complex biological systems owing to its high sensitivity and zero natural background.[15] Herein we report the systematic F-scanning of a scaffold derived from the Ugi four-component reaction (Ugi4CR) for the synthesis, optimization, and biophysical characterization of potent p53Mdm2 antagonists. The proteinprotein interaction (PPI) between the transcription factor p53 and its negative regulator Mdm2 has recently attracted great interest as a novel therapeutic target to treat cancer.[16] The antagonistic compound RG7112 (a nutlin-3 derivative) is currently undergoing clinical trials, while several p53 Mdm2 antagonists are in preclinical development. Due to the wealth of structural information available for the p53Mdm2 interaction,[17, 18] a pharmacophore-based virtual screening platform ANCHOR.QUERY was recently introduced for the rational design of small-molecule antagonists.[19] Based on this approach we described a number of unprecedented multicomponent reaction (MCR) scaffolds that are able to efficiently disrupt the p53Mdm2 interaction.[20, 21] A structurally characterized p53Mdm2 antagonist 1 showed good alignment with the three-finger pharmacophore model for Mdm2 antagonists,[22] in which key moieties fit into the Trp 23, Phe 19, and Leu 26 binding pockets.[23] Notably, the benzyl substituent of the small molecule involves a stacking interaction with the imidazole ring of His 96 (Figure 1).[22] In an attempt to modulate the p-
Figure 1. A co-crystal structure of p53Mdm2 antagonist 1 reveals a parallel alignment between the benzyl group of the Ugi scaffold and the imidazole ring of His 96 (Mdm2), which seems suitable for optimization of the pp interaction.[22]
stacking interaction, all 19 possible isomers with different fluorine substitution patterns at the benzyl position were synthesized (Scheme 1). Fluorine-substituted benzylamines 2 bt were obtained from commercial sources or were prepared from the corresponding benzaldehydes by reductive amination.[24] The MCR chemistry used herein is efficient and straightforward, and compounds 6 at were synthesized by the Ugi-4CR in good to excellent yields (Supporting Information [SI]). The ester group of compounds 6 at was saponified to give the
[a] Y. Huang,+ Prof. A. Dmling Departments of Pharmaceutical Sciences and Chemistry University of Pittsburgh, Pittsburgh, PA 15261 (USA) E-mail: asd30@pitt.edu [b] S. Wolf,+ G. M. Popowicz, Prof. T. A. Holak Max Planck Institute of Biochemistry Am Klopferspitz 18, 82152 Martinsried (Germany) [c] Dr. D. Koes, Prof. C. J. Camacho Department of Computational and Systems Biology University of Pittsburgh, Pittsburgh, PA 15261 (USA) [+] These authors contributed equally to this work. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100428. Scheme 1. Synthesis of p53Mdm2 antagonists using the Ugi-4CR followed by saponification. Reagents and conditions: a) MeOH, RT; b) LiOH, EtOH/H2O (1:1), RT. (Detailed procedures are given in the Supporting Information.)
49
MED
corresponding acid compounds 7 at, as 2-carboxylic acid of the indole ring is known to improve the binding affinity with Mdm2.[20] By the nature of the Ugi reaction, all products were formed as racemic mixtures and screened as such. The binding constants (Ki) of compounds 6 and 7 to Mdm2 were measured by using our recently described fluorescence polarization assay (Table 1 and SI table 1).[25] As a reference
Table 1. Inhibition constants for compounds 6 and 7. No. a b c d e f g h i j k l m n o p q r s t nutlin-3a X 6[a] H 4-F 3-F 2-F 3,4-di-F 2,4-di-F 2,3-di-F 2,5-di-F 3,5-di-F 2,6-di-F 2,3,4-tri-F 2,4,5-tri-F 3,4,5-tri-F 2,3,6-tri-F 2,4,6-tri-F 2,3,5-tri-F 2,3,5,6-tetra-F 2,3,4,6-tetra-F 2,3,4,5-tetra-F 2,3,4,5,6-penta-F 1.5 2.2 1.3 3 0.5 6 3 10 2.4 4.5 2.1 5 0.4 4.3 6.8 2.5 6.7 5.8 7 5.8 Ki [mm] 7[a] 1.8 0.45 0.81 1.7 0.25 2.3 0.2 2.5 0.3 5.7 0.15 2.3 0.13 3.2 3.2 0.17 3 3 0.7 1.8 0.04 3.26 3.40 clog P (7)[b] Figure 2. Stereo image of compound (S)-7 e (pink sticks) in Mdm2 (surface with secondary structure and lines): His 96, which forms a p-stacking interaction with the 3,4-difluorobenzyl group of (S)-7 e, is highlighted as blue sticks. The indole fragment of (S)-7 e aligns well with the anchor fragment Trp 23 of p53 (green sticks, PDB ID: 1YCR; RMSD = 0.701 ). The indole NH of (S)-7 e also forms the characteristic conserved hydrogen bond with the carbonyl group of Leu 54 (yellow dotted line).
3.54
3.69
3.83 3.97 5.17
[a] Measured by fluorescence polarization assay. [b] Calculated using Instant JChem 2.5.3 from ChemAxon.
compound we used the well-known nutlin-3a, and the binding data compare well with previously published results.[21] The acid compounds 7 show overall improved potency compared with the corresponding parent ethyl ester compounds 6. Interestingly, the Ki values of compounds 7 with various fluorine substitution patterns varied by a factor of 44, between 5.7 mm and 130 nm. The best compound, 7 m (Ki = 130 nm, Mr = 495 Da), is among the most potent p53Mdm2 antagonists known, which has affinity similar to that of the reference compound nutlin-3a (Ki = 40 nm, Mr = 581.5 Da). Compound 7 m exhibits a binding efficiency index (BEI) of 13.9, which indicates a better ligand efficiency than nutlin-3a (BEI = 12.7).[26] The aqueous solubility and calculated lipophilicity of 7 m are 0.85 mg mL1 and clog P = 3.69 (SI), which are superior to the respective values for nutlin-3a (< 0.1 mg mL1, clog P = 5.17). To better understand the structural basis of the interaction, 7 e was co-crystallized with Mdm2, and the X-ray structure was resolved (PDB ID: 3TU1, Figure 2 and SI table 2). As expected, the small molecule binds into the p53 binding site of Mdm2. The 6-chloroindole moiety aligns well with the anchor residue Trp 23 of p53 and forms a hydrogen bond with Leu 54. The 3,4difluorobenzyl group mimics Leu 26, and the tert-butylamide substituent derived from the isocyanide component is deeply
buried in the Phe 19 pocket. The formyl substituent is directed into the solvent space. Two amide oxygen atoms of 7 e and the residues in the receptor (His 96 and Val 93) form hydrogen bond bridges with surrounding water molecules (SI figure 1). The overall interactions, however, are governed mostly by hydrophobic contacts of the indole, 3,4-difluorobenzyl, and tertbutyl substituents with the receptor amino acids (SI figures 2 6). Clearly, the 3,4-difluorobenzyl group is nicely aligned parallel with the imidazole ring of His 96, suggesting an attractive interaction. Short distances between the atoms of two aromatic rings are between 3.4 and 3.9 (SI figure 4). Compared with other crystallographically characterized small-molecule antagonists of the p53Mdm2 interaction,[17, 18] the Leu 26 binding site encloses the 3,4-difluorobenzyl group very tightly. The enantiomers of 6 e and 6 m were efficiently separated by preparative supercritical fluid chromatography (SFC), and transformed into the corresponding enantiomers of 7 e and 7 m, respectively (SI). The enantiomer (+)-7 e (Ki = 200 nm) is more potent than ()-7 e (Ki = 400 nm). Similarly, the enantiomer (+)-7 m (Ki = 100 nm) is more potent than ()-7 m (Ki = 280 nm). Figure 3 shows an NMR heteronuclear single quantum coherence (HSQC) experiment for (+)-7 e and Mdm2. Strong binding, with KD < 1 mm, (and a slow chemical exchange) of (+)-7 e to Mdm2 is indicated by the doubling of signals.[27] A key property of the CF bond is the reversed polarization relative to the CH bond. Thus, the pattern of fluorine substitution on the benzene ring should be able to modulate the aromatic interaction between the small molecule and its protein receptor.[28] We investigated this effect with docking models derived from the co-crystal structure (Figure 2) by performing all possible fluorine substitution patterns in the benzyl group of compounds 7 bt on a fixed receptor. All the refined docked small molecules show only minor differences relative to the cocrystal structure. The models reveal that the CH-to-CF substitution at the buried ortho position leads to a highly repulsive dipoledipole interaction (Figure 4). In fact, the straightforward change in Coulomb energy due to the charge swap indicated in Figure 4 is ~ 4 kcal mol1. Not surprisingly, the fluorine in the buried ortho position does not rank well either in our
50
www.chemmedchem.org
Figure 3. Superposition of NMR HSQC spectra of 15N-labeled Mdm2 titrated against (+)-7 e: The spectrum of free Mdm2 is shown in red. The spectrum of (+)-7 eMdm2 (intermediate ratio, 3:10, respectively) is shown in green, and the spectrum of (+)-7 eMdm2 (final ratio, 1:1) is shown in blue.
Figure 4. Dipole reversal by fluorine is the main determinant of binding affinity: Docking models (yellow sticks) of A) compound (S)-7 m and B) compound (S)-7 j are minimized structures of the superposition of the small molecules onto the ligand of the co-crystal shown in Figure 2 with a fixed receptor. Arrows depict the dipoledipole interaction between the His 96eVal 93O hydrogen bond (blue sticks, red arrow) and the CH (A, blue arrow) and CF (B, cyan arrow) dipoles that are attractive and repulsive, respectively. Also indicated are the Merck molecular force field (MMFF) partial charges.
models or in the experiments. Note that at least for the subset of compounds with symmetric fluorine patterns there is no ambiguity on the position of the fluorine atoms in the docked models based on the co-crystal structure. Given the strong electrostatic coupling between the two rings, we expected the intermolecular component of the electrostatic free energy to be a main determinant of the binding affinity. Indeed, as shown in Figure 5, the correlation coefficient between the computed change in electrostatics and the experimentally determined affinity (pKi) yielded a strikingly accurate R = 0.96 value for compounds with symmetric fluorine substitutions. The analysis correctly ranks compounds with the best and worst Ki values among the benzyl rings with a symmetric
fluorine pattern. By itself, this result is quite impressive, as most physically based scoring functions do not recapitulate thermodynamic data.[29] The correlation of electrostatics with pKi for the full set of compounds, including those with fluorine substitutions that form asymmetric patterns, sharply drops to R = 0.58. An even lower correlation (R = 0.22) is obtained by using the full binding affinities calculated by the software OpenEye Szybki, further emphasizing the key role of electrostatics in capturing changes in Ki for the different ligands. Refining the ligand protonation states by considering pKa and tautomer enumeration with the AM1-BCC charge model of OpenEye QuacPac version 1.3.1 does not change the results.[30] The conspicuous over-prediction of the electrostatic component of the binding free energy for asymmetric compounds is likely to reflect the specificities of the internal energies for each rotamer of the benzyl ring. This effect is offset for symmetric compounds, but it should significantly affect the free energy of the unbound asymmetric ligands in ways that are ignored by the fixed ligand structures imposed by the Poisson Boltzmann calculations. In summary, we have efficiently synthesized 40 derivatives of p53Mdm2 antagonist 1 via a one-pot Ugi-4CR including all possible fluorine substitutions on the benzyl group in order to optimize aromatic interactions. A co-crystal structure of (S)-7 e with Mdm2 reveals the structural basis of the potent interaction. The introduction of fluorine substitutions on the benzyl group can considerably improve the potency of p53Mdm2 antagonists, due to the electrostatic interaction between the small molecules and the receptor. Although limited in scope, the computational analysis of docked configurations that are presumed to bind via the same binding mode shows that the electrostatics of unique rigid structures can be reasonably ex-
www.chemmedchem.org
51
MED
[6] S. Purser, P. R. Moore, S. Swallow, V. Gouverneur, Chem. Soc. Rev. 2008, 37, 320 330. [7] B. J. Backes, K. Longenecker, G. L. Hamilton, K. Stewart, C. Lai, H. Kopecka, T. W. von Geldern, D. J. Madar, Z. Pei, T. H. Lubben, B. A. Zinker, Z. Tian, S. J. Ballaron, M. A. Stashko, A. K. Mika, D. W. A. Beno, A. J. Kempf-Grote, C. Black-Schaefer, H. L. Sham, J. M. Trevillyan, Bioorg. Med. Chem. Lett. 2007, 17, 2005 2012. [8] M. Morgenthaler, E. Schweizer, A. Hoffmann-Rder, F. Benini, R. E. Martin, G. Jaeschke, B. Wagner, H. Fischer, S. Bendels, D. Zimmerli, J. Schneider, F. Diederich, M. Kansy, K. Mller, ChemMedChem 2007, 2, 1100 1115. [9] H. Su, Y. Xie, W.-B. Liu, S.-L. You, Bioorg. Med. Chem. Lett. 2011, 21, 3578 3582. [10] A. D. Kerekes, S. J. Esposite, R. J. Doll, J. R. Tagat, T. Yu, Y. Xiao, Y. Zhang, D. B. Prelusky, S. Tevar, K. Gray, G. A. Terracina, S. Lee, J. Jones, M. Liu, A. D. Basso, E. B. Smith, J. Med. Chem. 2011, 54, 201 210. [11] B. K. Park, N. R. Kitteringham, P. M. ONeill, Annu. Rev. Pharmacol. Toxicol. 2001, 41, 443 470. [12] A. Percival, J. Antimicrob. Chemother. 1991, 28, 1 8. Figure 5. Correlation between the calculated change in electrostatics and binding affinity [13] D. Le Bars, J. Fluorine Chem. 2006, 127, 1488 1493. (pKi) for compounds (S)-7: Electrostatic calculations were performed with OpenEye Szybki [14] L. Cai, S. Lu, V. W. Pike, Eur. J. Org. Chem. 2008, 2008, 2853 1.3.1 with a PoissonBoltzmann solvation model and MMFF partial charges. Compounds 2873. with a symmetric arrangement of fluorine atoms in the benzyl group correlate closely [15] N. Yanamala, A. Dutta, B. Beck, B. Van Fleet, K. Hay, A. (R = 0.96) to the calculated electrostatics. The remaining asymmetric compounds are Yazbak, R. Ishima, A. Dmling, J. Klein-Seetharaman, Chem. more difficult to characterize and decrease the overall correlation (R = 0.58). Following Biol. Drug Des. 2010, 75, 237 256. Table 1, letters indicate the fluorine substitution pattern in the phenyl ring. [16] C. F. Cheok, C. S. Verma, J. Baselga, D. P. Lane, Nat. Rev. Clin. Oncol. 2011, 8, 25 37. [17] K. Khoury, G. M. Popowicz, T. A. Holak, A. Dmling, Med. Chem. Comm. 2011, 2, 246 260. plained by current methods, whereas assessing the free energy [18] G. M. Popowicz, A. Dmling, T. A. Holak, Angew. Chem. 2011, 123, 2732 of an unbound state with a variety of hard-to-evaluate internal 2741; Angew. Chem. Int. Ed. 2011, 50, 2680 2688. states is quite detrimental to empirical estimates of the free [19] D. Koes, K. Khoury, Y. Huang, W. Wang, M. Bista, G. M. Popowicz, S. Wolf, energy. Our findings underscore the principally known but T. A. Holak, A. Dmling, C. J. Camacho, PLoS Comput. Biol. in press. [20] a) A. Czarna, B. Beck, S. Srivastava, G. M. Popowicz, S. Wolf, Y. Huang, M. often surprising effects that fluorine can exert on the biological Bista, T. A. Holak, A. Dmling, Angew. Chem. 2010, 122, 5480 5484; activity of small molecules. Our findings may also lead mediciAngew. Chem. Int. Ed. 2010, 49, 5352 5356; b) Y. Huang, S. Wolf, M. nal chemists to success in modulating molecular interactions Bista, L. Meireles, C. Camacho, T. A. Holak, A. Dmling, Chem. Biol. Drug with other types of targets. Des. 2010, 76, 116 129; c) S. Srivastava, B. Beck, W. Wang, A. Czarna, T. A. Holak, A. Dmling, J. Comb. Chem. 2009, 11, 631 639. [21] G. M. Popowicz, A. Czarna, S. Wolf, K. Wang, W. Wang, A. Dmling, T. A. Holak, Cell Cycle 2010, 9, 1104 1111. Acknowledgements [22] S. Wolf, Y. Huang, G. M. Popowicz, S. Goda, T. A. Holak, A. Dmling, J. Am. Chem. Soc. unpublished results. This work was supported by NIH grants 1R21GM087617-01, [23] A. Dmling, Curr. Opin. Chem. Biol. 2008, 12, 281 291. [24] T. Ono, V. P. Kukhar, V. A. Soloshonok, J. Org. Chem. 1996, 61, 6563 1P41GM094055-01 and 1R01GM097082-01, the Qatar Foundation 6569. NPRP (grant 4-319-3-097 to A.D.), and the Deutsche Krebshilfe [25] G. M. Popowicz, A. Czarna, T. Holak, Cell Cycle 2008, 7, 2441 2443. (grant 108354 to T.H.); this work is part of a NCINExT program. [26] C. Abad-Zapatero, J. T. Metz, Drug Discovery Today 2005, 10, 464 469. [27] a) K. Wthrich, NMR of Proteins and Nucleic Acids, Wiley, New York, 1986; b) T. Rehm, R. Huber, T. A. Holak, Structure 2002, 10, 1613 1618. Keywords: drug discovery fluorine multicomponent [28] a) C. Y. Kim, J. S. Chang, J. B. Doyon, T. T. Baird, C. A. Fierke, A. Jain, D. W. reactions p53mdm2 proteinprotein interaction Christianson, J. Am. Chem. Soc. 2000, 122, 12125 12130; b) C. Y. Kim, P. P. Chandra, A. Jain, D. W. Christianson, J. Am. Chem. Soc. 2001, 123, 9620 9627. [1] G. A. Olah, R. D. Chambers, G. K. Surya Prakash, (Eds.) Synthetic Fluorine [29] G. L. Warren, C. W. Andrews, A. M. Capelli, B. Clarke, J. LaLonde, M. H. Chemistry, Wiley, New York, 1992. Lambert, M. Lindvall, N. Nevins, S. F. Semus, S. Senger, G. Tedesco, I. D. [2] M. Shimizu, T. Hiyama, Angew. Chem. 2005, 117, 218 234; Angew. Wall, J. M. Woolven, C. E. Peishoff, M. S. Head, J. Med. Chem. 2006, 49, Chem. Int. Ed. 2005, 44, 214 231. 5912 5931. [3] a) K. Mller, C. Faeh, F. Diederich, Science 2007, 317, 1881 1886; b) L. M. [30] S. Wlodek, A. G. Skillman, A. J. Nicholls, J. Chem. Theory Comput. 2010, Salonen, M. Ellermann, F. Diederich, Angew. Chem. 2011, 123, 4908 6, 2140 2152. 4944; Angew. Chem. Int. Ed. 2011, 50, 4808 4842. [4] H.-J. Bhm, D. Banner, S. Bendels, M. Kansy, B. Kuhn, K. Mller, U. ObstSander, M. Stahl, ChemBioChem 2004, 5, 637 643. [5] W. K. Hagmann, J. Med. Chem. 2008, 51, 4359 4369. Received: September 8, 2011 Published online on September 27, 2011
52
www.chemmedchem.org
DOI: 10.1002/cmdc.201100410
Nocodazole is a High-Affinity Ligand for the Cancer-Related Kinases ABL, c-KIT, BRAF, and MEK
Hwangseo Park,*[a] Seunghee Hong,[b] and Sungwoo Hong*[b]
Nocodazole (1, Janssen Pharmaceutica), originally identified in a chemical library screen for anticancer activity, has been used in clinical cancer chemotherapy for some time.[1] Nocodazole is a standard anti-microtubule agent that interferes with the polymerization of microtubules. A series of experimental studies has shown that nocodazole induces apoptosis in chronic lymphocytic leukemia cells,[2] inhibits insulin-stimulated glucose transport,[3] decreases apoptosis in some human colon carcinoma cells,[4] and impairs the morphology and directionality of migrating medial ganglionic eminence cells.[5] However, the molecular mechanisms underlying its anticancer activities remain unclear. It was recently demonstrated that nocodazole treatment induces significant phosphoproteomic changes in a variety of tumor cell lines,[6] suggesting that nocodazole modulates the activities of kinases. In addition, docking simulations indicate that the aminobenzimidazole moiety of nocodazole can fit into the ATP binding sites of kinases. These observations led us to hypothesize that the anticancer activity of nocodazole may be attributed in large part to the inhibition of kinases associated with cell-cycle signaling. Herein we report that nocodazole displays high affinity toward a set of cancer-related kinases, including ABL, c-KIT, BRAF, and MEK. Notably, nocodazole effectively binds to the most drug-resistant mutant, ABL(T315I). Docking models of nocodazole bound to these kinases were also studied to understand the structural properties of binding modes. The binding affinities of nocodazole[7] were tested at 10 mm in a high-throughput binding assay (KINOMEscan, Ambit Biosciences) against a panel of 96 kinases. The assay monitors the strength of binding by using phage display technology. Table 1 lists the percent of control (POC) values for nocodazole. Nocodazole appears to be a potent inhibitor of 16 kinases with high affinity (POC < 22). Full Kd values were determined for 10 kinases that are considered promising targets for cancer treatment. As shown in Table 1, the highest binding affinity was observed for ABL, with an associated Kd value of 0.091 mm. Nocodazole
Table 1. Binding affinities of nocodazole (1) for various kinases. Kinase ABL phosphorylated ABL(E255K) phosphorylated ABL(T315I) phosphorylated BRAF BRAF(V600E) CDK7 c-KIT DYRK1B JNK2 JNK3 MEK1 MEK2 MET MKNM2 PDGFRB PI3Kg ROCK2 SNARK ULK2 POC [10 mm][a] 0.25 0.8 2.4 21 11 13 22 1.1 21 20 8.4 7.3 8.9 11 9.7 8.2 18 16 7.1 Kd [mm][b] 0.091 0.12 0.17 1.8 1.1 ND 1.6 ND ND ND 1.7 1.6 1.7 ND ND 1.5 ND ND ND
[a] Kinases with percent of control (POC) values < 22 are shown. [b] Values are the average of two determinations; ND = not determined.
[a] Prof. H. Park Department of Bioscience and Biotechnology Sejong University, 98 Kunja-dong, Kwangjin-ku, Seoul 143-747 (Korea) E-mail: hspark@sejong.ac.kr [b] S. Hong, Prof. S. Hong Department of Chemistry Korea Advanced Institute of Science and Technology (KAIST) Guseong-dong, Yuseong-gu, Daejeon 305-701 (Korea) E-mail: hongorg@kaist.ac.kr Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100410.
also revealed a sub-micromolar binding affinity for the E255K (Kd = 0.12 mm) and T315I (Kd = 0.17 mm) mutant forms of ABL. In addition, nocodazole bound to the wild-type and V600E mutants of BRAF, c-KIT, MEK1, MEK2, and MET with Kd values in the range of 12 mm. The activity of nocodazole was further confirmed with radiometric kinase assays,[8] in which the compound exhibited good potency against ABL (IC50 = 0.21 mm), ABL(E255K) (IC50 = 0.53 mm), and ABL(T315I) (IC50 = 0.64 mm). The low molecular weight of nocodazole (Mr = 301 Da) suggests that the compound may serve as a good scaffold from which more potent inhibitors may be derivatized. ABL is ubiquitously expressed and is responsible for many cellular functions, including the regulation of cell growth and survival, DNA damage responses, actin dynamics, and cell migration in response to the stimulation of cell-surface receptors.[9] The fusion of the Bcr-Abl gene encodes a BCR-ABL chimeric kinase with constitutively activated ABL kinase activity. BCR-ABL kinase is considered to be one of the most attractive targets for treating leukemia because its overactivation is observed in > 90 % of chronic myelogenous leukemia (CML) patients. The inhibition of BCR-ABL by imatinib has been successfully used to treat patients with chronic-phase CML as an effective first-line therapy.[10] Despite the approved therapeutic options, CML cases that include the T315I gatekeeper binding site point mutation remain a major clinical challenge, because CML cells that harbor this mutation are insensitive to the ABLtargeted drugs. Thus, the identification of targeted ABL inhibitors capable of effectively impairing the activity of the T315I
53
MED
mutation would have important therapeutic implications.[11] In this regard, nocodazole may serve as a promising lead compound for the development of a new drug for the treatment of CML because of its sub-micromolar activities against both wild-type and mutant forms of ABL. Docking simulations were performed to obtain energetic and structural insight into the binding modes at the ATP binding sites of ABL and ABL(T315I). The AutoDock program was used with consideration for the motional flexibility of the amino acid side chains in the ATP binding site; 3D atomic coordinates of nocodazole were generated using the CORINA program, and GasteigerMarsilli atomic charges were assigned.[12] Noting that multiple binding modes may be generated in docking simulations, we carried out 20 independent docking runs for each target kinase to select the most plausible binding mode (tables 1S3S in the Supporting Information). Among the 20 conformations of nocodazole generated in docking simulations, those clustered together have similar binding modes differing by < 1.5 in positional root-meansquare deviation. In the calculated ABLnocodazole complex shown in Figure 1, nocodazole appears to be in close contact with residues Leu 248Tyr 253, Ile 313Gly 321, and Ala 269 Lys 271, which belong to the Gly loop, the gatekeeper site, and a b sheet in the N-terminal domain, respectively. It is noted that two nitrogen atoms on the aminobenzimidazole group of nocodazole form two hydrogen bonds with the backbone amidic nitrogen and the aminocarbonyl oxygen of Met 318. Nocodazole may be further stabilized in the ATP binding site via hydrophobic interactions among its nonpolar groups with the side chains of Leu 248, Tyr 253, Val 256, Ala 269, Phe 317, Leu 370, and Phe 382. Thus, the overall structural features derived from docking simulations indicate that the binding affinities of nocodazole for both wild-type and the T315I mutant of ABL stem from multiple hydrogen bonds and hydrophobic interactions established simultaneously in the ATP binding site. Despite the overall similarities in the binding mode of nocodazole toward the wild-type and T315I mutant, some differences were observed in how the inhibitor interacts with the gatekeeper residues. Two additional residues, Val 299 and Ile 315, were found at the interface formed by van der Waals contacts between nocodazole and the T315I mutant of ABL (Figure 1 b), indicating that the hydrophobic interactions established in the T315I mutant are stronger than those in the wild-type. However, the strengthening of hydrophobic interactions appear to be insufficient to compensate for the weakening of the hydrogen bonds. The hydrogen bonds between Met 318 and nocodazole weaken in going from the wild-type to the T315I mutant of BCR-ABL: the associated NO interatomic distance and NHO angle change from 2.66 and 171.38, respectively, in the wildtype to 2.98 and 158.78 in the mutant. These structural features may explain the slight decrease in activity toward the T315I mutant. Nocodazole exhibited good affinity toward c-KIT, with a Kd value of 1.6 mm. c-KIT mutations that cause a constitutive activation of c-KIT are implicated in highly malignant human cancers, including gastrointestinal stromal tumors (GIST), acute myeloid leukemia (AML), and mast cell leukemia (MCL), as well as in mastocytosis.[2, 13] Therefore, nocodazole is a good starting point for the de novo design and SAR studies to develop new drugs for the treatment of various cancers associated with c-KIT. The calculated binding mode of nocodazole in the ATP binding site of c-KIT is shown in Figure 2. This binding mode was found to be similar to that of ABL, in that one nitrogen atom on the aminobenzimidazole group donates a hydrogen bond
Figure 1. Calculated binding modes of nocodazole in the ATP binding sites of a) wild-type and b) T315I mutant ABL. Indicated in the dashed boxes are the sites of mutation. Each dotted line indicates a hydrogen bond.
Figure 2. Calculated binding mode of nocodazole in the ATP binding site of c-KIT. Each dotted line indicates a hydrogen bond.
54
www.chemmedchem.org
to the backbone aminocarbonyl oxygen and another receives a hydrogen bond from the amidic nitrogen of a residue at the center of the gatekeeper site (Cys 673 for c-KIT). These two hydrogen bonds seem to provide the most significant contribution toward stabilizing nocodazole in the ATP binding site of cKIT. Nocodazole may be further stabilized in the ATP binding site of c-KIT via hydrophobic interactions among the nonpolar groups with side chains Leu 595, Val 603, Ala 621, Val 654, Leu 799, and Phe 811. However, the side chain Tyr 672 is directed away from the benzimidazole ring of nocodazole at a distance > 5 , whereas Phe 317 points toward the benzimidazole ring within 4.3 in the ABLnocodazole complex. Nocodazole also displays good binding affinity toward the components of the mitogen-activated protein kinase (MAPK) pathway, such as BRAF (Kd = 1.8 mm), BRAF(V600E) (Kd = 1.1 mm), MEK1 (Kd = 1.7 mm), and MEK2 (Kd = 1.6 mm). Because more than half of all metastatic melanoma patients present BRAF(V600E) mutations,[13a] the targeted inhibition of signaling by the mutant BRAF provides a promising therapeutic strategy for melanoma patients. Figure 3 compares the calculated binding modes of nocodazole in the ATP binding sites of the wild-type and V600E mutant form of BRAF. In contrast to the structural
chain indole ring of Trp 531 is directed toward the benzimidazole ring of nocodazole at a distance of 3.34.5 . The hydrophobic groups of nocodazole are also stabilized by van der Waals interactions with the nonpolar side chains of Ile 463, Val 471, Ala 481, Leu 514, Trp 531, and Phe 583. In conclusion, nocodazole was submitted to enzymatic kinase assays to test the hypothesis that the inhibitory activity of nocodazole toward these kinases contributes to its anticancer activity. Nocodazole was found to be a common inhibitor of various cancer-related kinases, including ABL, c-KIT, BRAF, MEK1, MEK2, and MET. Based on these results, the kinase inhibitory activity of nocodazole needs to be taken into account when analyzing the results of mechanistic cell biology studies with nocodazole. Binding mode analyses and docking simulations indicate that nocodazole is stabilized in the ATP binding sites of the target kinases through formation of hydrogen bonds between its aminobenzimidazole moiety and the backbone groups at the center of the gatekeeper site. Hydrophobic interactions with nonpolar residues help stabilize the binding of nocodazole in the ATP binding sites of the target kinases. These experimental and computational findings may provide insight into the design of new potent common kinase inhibitors as anticancer drugs through de novo design and SAR studies based on the nocodazole scaffold.
The binding modes of nocodazole in the ATP binding site of various kinases were estimated by docking simulations with the AutoDock version 4.0. It combines a rapid energy evaluation through pre-calculated grids of affinity potentials with the Lamarckian genetic algorithm to find the suitable binding positions for a ligand on a protein receptor under consideration of the structural flexibilities of both protein and ligand. In these docking simulations, we used the empirical scoring function improved by the implementation of a new solvation model for a compound. The modified scoring function has the following form: DGaq WvdW bind Welec X X Aij
i1 j>i 12
rij
Bij 6 rij
Whbond
XX
i1 j>i
E t
X X qi qj X X r Wtor Ntor Wsol Si Occmax Vj e2s i e rij rij i1 j>i i1 j>i
Cij Dij 12 10 rij rij !

2 ij 2
1 in which WvdW, Whbond, Welec, Wtor, and Wsol are the respective weighting factors of van der Waals, hydrogen bond, electrostatic interactions, torsional term, and desolvation energy of nocodazole; rij represents the interatomic distance, and Aij, Bij, Cij, and Dij are related to the depth of the potential energy well and the equilibrium separations between the atoms of the target kinases and nocodazole. The hydrogen bond term has an additional weighting factor, E(t), which represents the angle-dependent directionality. In the entropic term, Ntor is the number of all rotatable bonds in the ligand. In the desolvation term, Si and Vi are the solvation parameter and the fragmental volume of atom i in nocodazole, respectively, while Occimax stands for the maximum atomic occupancy. In the calculation of the molecular solvation free energy term in Equation (1), we used the atomic parameters developed by Kang et al.,[14] because
Figure 3. Calculated binding modes of nocodazole in the ATP binding sites of a) wild-type and b) V600E mutant BRAF. Each dotted line indicates a hydrogen bond.
features in the ABLnocodazole and c-KITnocodazole complexes, only one hydrogen bond is present between the amidic nitrogen atom of nocodazole and the aminocarbonyl oxygen atom of Cys 532. However, hydrophobic interactions in the BRAFnocodazole complexes are significant, and the side
www.chemmedchem.org
55
MED
their solvation model proved successful in predicting the solvation free energies of a variety of organic molecules.
[7] Nocodazole was purchased from InterBioScreen Ltd. (Moscow, Russia). The purity was determined to be 97 % by HPLC. A 10 mm stock solution in DMSO was prepared. [8] IC50 value determinations were performed by Reaction Biology Corp. (Malvern, PA, USA). [9] A. M. Pendergast, Adv. Cancer Res. 2002, 85, 51 100. [10] H. Kantarjian, C. Sawyers, A. Hochhaus, F. Guilhot, C. Schiffer, C. Gambacorti-Passerini, D. Niederwieser, D. Resta, R. Capdeville, U. Zoellner, M. Talpaz, B. Druker, N. Engl. J. Med. 2002, 346, 645 652. [11] T. OHare, W. C. Shakespeare, X. Zhu, C. A. Eide, V. M. Rivera, F. Wang, L. T. Adrian, T. Zhou, W.-S. Huang, Q. Xu, C. A. Metcalf III, J. W. Tyner, M. M. Loriaux, A. S. Corbin, S. Wardwell, Y. Ning, J. A. Keats, Y. Wang, R. Sundaramoorthi, M. Thomas, D. Zhou, J. Snodgrass, L. Commodore, T. K. Sawyer, D. C. Dalgarno, M. W. Deininger, B. J. Druker, T. Clackson, Cancer Cell 2009, 16, 401 412. [12] a) G. M. Morris, D. S. Goodsell, R. S. Halliday, R. Huey, W. E. Hart, R. K. Belew, A. J. Olson, J. Comput. Chem. 1998, 19, 1639 1662; b) J. Gasteiger, M. Marsili, Tetrahedron 1980, 36, 3219 3228. [13] a) K. T. Flaherty, I. Puzanov, K. B. Kim, A. Ribas, G. A. McArthur, J. A. Sosman, P. J. ODwyer, R. J. Lee, J. F. Grippo, K. Nolop, P. B. Chapman, N. Engl. J. Med. 2010, 363, 809 819; b) J. A. Schumacher, K. S. ElenitobaJohnson, M. S. Lim, J. Clin. Pathol. 2008, 61, 109 114. [14] H. Kang, H. Choi, H. Park, J. Chem. Inf. Model. 2007, 47, 509 514.
Acknowledgements
This work was supported by the Korea Institute of Oriental Medicine (grant no. K11061) and the National Research Foundation of Korea (NRF-2011-0003609, 2011-0016436, and 2011-0020322). Keywords: ABL anticancer agents inhibitors kinases nocodazole
[1] M. J. de Brabander, R. M. van de Veire, F. E. Aerts, M. Borgers, P. A. Janssen, Cancer Res. 1976, 36, 905 916. [2] R. W. Beswick, H. E. Ambrose, S. D. Wagner, Leuk. Res. 2006, 30, 427 436. [3] J. C. Molero, J. P. Whitehead, T. Meerloo, D. E. James, J. Biol. Chem. 2001, 276, 43829 43835. [4] H. Zhang, X. Shi, Q.-J. Zhang, M. Hampong, H. Paddon, D. Wahyuningsih, S. Pelech, J. Biol. Chem. 2002, 277, 43648 43658. [5] J.-P. Baudoin, C. Alvarez, P. Gaspar, C. Metin, Dev. Neurosci. 2008, 30, 132 143. [6] K. Nagano, T. Shinkawa, H. Mutoh, O. Kondoh, S. Morimoto, N. Inomata, M. Ashihara, N. Ishii, Y. Aoki, M. Haramura, Proteomics 2009, 9, 2861 2874.
Received: August 26, 2011 Revised: October 1, 2011 Published online on October 14, 2011
56
www.chemmedchem.org
DOI: 10.1002/cmdc.201100401
Specific Targeting of Hypoxic Tumor Tissue with NitroimidazolePeptide Conjugates

Katrin Splith,[b] Ralf Bergmann,[c] Jens Pietzsch,[c] and Ines Neundorf*[a]
Hypoxia is one of the characteristics of solid tumors and originates in fast-growing tumors from an imbalance between arterial oxygen supply and consumption. Normally, hypoxic regions occur where tumor growth exceeds the capacity of the corresponding blood vessels. Hypoxic regions receive a lower drug dose than normoxic regions due to insufficient blood supply.[1] Furthermore, hypoxic cells are more resistant to commonly used radiotherapy and chemotherapy[1, 2] and show an increased metastatic potential.[3] However, the existence of hypoxic cells opens up the possibility to target hypoxic regions specifically and exploit hypoxia for diagnosis and therapy applications. In this respect, several radiopharmaceuticals have been developed that bear a 2-nitroimidazole moiety.[4] These nitroimidazole derivatives have high electron affinities and are hypoxic cell sensitizers. Following an enzyme-mediated one-electron reduction, nitroimidazoles accumulate within hypoxic regions of tumors. In contrast to normoxic tissue, where the resultant radical anion is reoxidized, in hypoxic regions, the resultant radical anion undergoes further reduction steps ultimately leading to final products that are linked to macromolecules. Thus, selective accumulation of nitroimidazole compounds is possible, and this mechanism is already used in positron emission tomography (PET) tumor imaging.[57] However, due to the insufficient blood supply in tumor cells and the low bioavailability of therapeutics in many cases, for successful imaging and/or treatment, a specific cell-permeable transporter is required that carries a sensitizer, label and/or therapeutic agent. Cell-penetrating peptides (CPPs) are delivery vectors for a great variety of different substances and can help to bypass the limited transport of bioactive molecules into their target cells.[8, 9] Using this method, we recently reported the successful and efficient delivery of nanocrystals,[10] oligonucleotides,[11] gallium complexes[12] and bioactive metal complexes[1316] into cancer, as well as primary cells. However, while CPPs provide a tool through which cell-impermeable compounds can reach intracellular targets, one of their major drawbacks is their lack of selectivity towards different cell lines.[17] The focus of this work was the design and characterization of peptide conjugates composed of a hypoxic radiosensitizer, a CPP and a radioactive label. We used the CPP sC18, which is a C-terminal fragment of the cationic antimicrobial peptide CAP18 and has been identified to deliver small organic substances like fluorophores and toxic peptide sequences efficiently into different cell lines.[18] We aimed at developing a targeted delivery system and imaging agent, with the option to eventually conjugate it with therapeutics. We generated peptide conjugates by coupling the 2-(2-nitroimidazol-1-yl)acetic acid (NIA) unit and a reporter group to the sC18 CPP. First, carboxyfluorescein (CF) was chosen as an internal label to study the influence of the NIA group on the CPP peptide structure and cellular uptake properties in vitro. As a control, peptides without the NIA group but with the CF label were tested also. Additionally, b-alanine was introduced as a linker between the CF label and peptide sequence. We aimed to avoid changes in uptake properties and to ensure the accessibility of the chelator, which was coupled as a second reporter group in subsequent studies. The peptides were N-terminally coupled with CF, whereas NIA was introduced at the side chain of Lys 109 in peptide 2. NIA was prepared as described elsewhere[19] and coupled directly to the peptide while still on solid support. After complete synthesis, all acid-labile protecting groups were removed, and the peptides were cleaved from the resin with trifluoroacetic acid (TFA). The peptide conjugates were purified by preparative reverse-phase HPLC and analyzed by MALDI-ToF and ESI mass spectrometry. Experimentally determined and calculated masses of the peptides were consistent (Table 1, see also Figure S1 in the Supporting Information).
Table 1. Names, sequences and molecular weights of synthesized peptides 14. Peptide Sequence[a]
106 121
MWcalcd[b] MWexptl
[a] Prof. Dr. I. Neundorf Department fr Chemie, Institut fr Biochemie Mathematisch-Naturwissenschaftliche Fakultt, Universitt zu Kln Zlpicher Str. 47, 50674 Kln (Germany) E-mail: ines.neundorf@uni-koeln.de [b] K. Splith Institut fr Biochemie, Fakultt fr Biowissenschaften Pharmazie und Psychologie, Universitt Leipzig Brderstr. 34, 04103 Leipzig (Germany) [c] Dr. R. Bergmann, Prof. Dr. J. Pietzsch Institut fr Radiopharmazie, Forschungszentrum Dresden-Rossendorf PF 510119, 01314 Dresden (Germany) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100401.
1 2 3 4
CF-sC18 CF-sC18(NIA) DOTA-sC18 DOTA-sC18(NIA)
CF- GLRKRLRKFRNKIKEK 2426.4 2427.5 CF-bA-GLRK(NIA)RLRKFRNKIKEK 2650.4 2651.4 DOTA-bA-GLRKRLRKFRNKIKEK 2525.4 2526.5 DOTA-bA-GLRK(NIA)RLRKFRNKIKEK 536.8[c] 536.9
[a] All peptides are N-terminally amidated. [b] [M + H] + . [c] [M + 5H]5 + .
Subsequently, compound 2 was characterized in terms of structural conformation, cellular uptake and toxicity compared to the parent compound 1. Primarily, we tested the influence of the NIA unit on the structure of the peptides by recording CD spectra of peptides 1 and 2 (Figure 1 A and B). For this pur-
57
MED
ing an endocytotic uptake. This finding is in good agreement with recent studies and supports the fact that endocytosis was highlighted as an important mode of cellular uptake for many CPPs.[14, 18, 21] Moreover, the uptake is concentration dependent, as was described previously for peptide 1 in HEK 293 cells.[18] Only a weak reduction of uptake was observed when NIA is coupled to the peptide; an effect that is more pronounced when lower concentrations are used. In addition, we tested peptides 1 and 2 with respect to their cytotoxicity in a fluorimetric resazurin-based cell viability assay in a concentration range of 1 to 100 mm (see Supporting Information). Recently, sC18 CPP was shown to be nontoxic when incubated with several different cell lines.[18] After treatment of subconfluent MCF-7 cells for 24 h, peptide 1 exhibited no influence on cell viability below 100 mm relative to untreated control cells. Notably, peptide 2, which is coupled with one moiety of NIA, did not affect cell viability when applied in concentrations below 100 mm. Even when the cells were incubated with the peptide solutions for 72 h, no toxicity could be observed (see Figure S2 in the Supporting Information). Inspired by these promising in vitro results we proceeded to investigate the biodistribution and pharmacokinetics of the NIA-modified peptide in vivo. Thus, we synthesized peptides 3 and 4 that are coupled N-terminally with the chelator DOTA (1,4,7,10-tetraaza-cyclododecane-1,4,7,10-tetraacetic acid). Peptide 4 was modified with the NIA group by the same procedure as peptide 2. 64Cu2 + labeling of derivatives 3 and 4 was performed as described previously.[22] The radiochemical yield was determined by radio-HPLC. After labeling with 64Cu2 + , the DOTA derivatives exhibited labeling yields of 91 12 % (n = 4) for 64Cu-3, and 90 6 % (n = 6) for 64Cu-4. However, after purification by radio-HPLC, 64Cu-3 and 64Cu-4 were obtained in radiochemical purities greater than 98 %, making the radiolabeled peptides useful for the following of in vivo assays. For this purpose, the radioactive copper-labeled peptides were injected into the tail veins of tumor-bearing Wistar rats. Whole-body PET scans were carried out over 120 min post-injection, and blood samples were taken at different time points. Qualitative analysis of the PET studies revealed that both with and without the NIA group, the peptides exhibit the same fast blood clearance in animals. However, a different accumulation pattern in the liver and the kidneys was observed for peptides 64 Cu-3 and 64Cu-4. Quantitative analysis of the PET data was performed to determine the in vivo biokinetics of the labeled peptides 64Cu-3 and 64Cu-4 within the regions of interest (ROI). For this reason, the standardized uptake values (SUV)activity concentration/injected activity*body weightin the blood (Figure 2 A), the liver and kidneys (Figure 2 B), and the tumor (Figure 2 C) were determined at different time points. Both radiolabeled compounds are immediately washed out, and after 5 min, nearly no radioactivity remained in the blood. Nevertheless, the peptides exhibited a very different accumulation pattern in the liver and kidneys. The NIApeptide 64Cu-4 accumulates in higher concentrations in both organs compared with 64 Cu-3, whereas accumulation in the kidneys was higher than in the liver. Control peptide 64Cu-3 accumulates to a higher extent in the kidneys, indicating different lipophilic properties,
Figure 1. Circular dichroism (CD) spectra of the investigated peptides 1 (c) and 2 (c). The spectra were measured at a peptide concentration of 20 mm in A) 10 mm aqueous phosphate buffer (pH 7.0) and B) 10 mm aqueous phosphate buffer (pH 7.0) with the addition of 50 % TFE. Uptake in MCF-7 cells of sC18 peptides C) 1 (&) and 2 (&) at different concentrations as determined by flow cytometry after 60 min. Also shown are representative fluorescence microscopy images of D) 25 mm 1 and E) 25 mm 2.
pose, the peptides were dissolved in 10 mm aqueous phosphate buffer (pH 7.0) as well as in phosphate buffer (10 mm, pH 7.0) supplemented with 50 % trifluoroethanol (TFE), a helixinducing solvent. Without TFE, the peptides displayed a spectrum typical of a random coil conformation with minima around 199 nm. With TFE, peptide 1 adopts an amphipathic ahelical secondary structure with two minima at 222 and 208 nm and a maximum at 190 nm, as previously observed.[18] This was also previously reported for the parent peptide fragment C18.[20] After introducing the NIA unit to the sC18 peptide (2), the a-helical properties in TFE-containing solvent are still present. Consequently, attachment of the NIA moiety has no influence on the backbone structure of the peptide, and thus, probably no influence on the uptake, even though the NIA group is coupled at an internal position of the sC18 structure in contrast to previous studies where modifications were simply introduced at the N-terminus.[14, 15] Based on this result, we assumed no change in uptake level or mechanism. Therefore, we examined the cellular uptake of peptides 1 and 2 in MCF-7 cells by flow cytometry (Figure 1 C) and fluorescence microscopy (Figure 1 D and E). After incubation for 60 min with 25 mm peptide solutions, the cells showed potent uptake of the CF-peptides, which were found to be located within cytoplasmic vesicles in all cell lines tested, indicat-
58
www.chemmedchem.org
peptides had been secreted into the urine. No relevant accumulation in other tissues was observed. Fluorine-18 fluoromisonidazole (FMISO), another hypoxic radiotracer that can enter the cells by passive diffusion, was even observed in the brain,[23] which seems not to be true for our conjugate indicating that 64Cu-4 is not able to cross the bloodbrain barrier. To determine the kinetics of radiolabeled peptides 64Cu-3 and 64Cu-4 and to obtain a more detailed distribution pattern, they were injected into the tail vein of tumor-bearing mice, and whole-body autoradiography experiments were conducted after 60 min (Figure 3). As expected, the control peptide only localized in well-perfused regions. Only small amounts of
Figure 3. Autoradiography images for peptides A) 64Cu-3 and B) 64Cu-4 in tumor-bearing mice. The black circles surround the region where the tumor is localized.
Figure 2. Pharmacokinetics of the injected peptides 64Cu-3 and 64Cu-4 presented as standardized uptake values (SUV) over time for the peptides in A) blood (64Cu-3: &; 64Cu-4: ~), B) in the kidneys (64Cu-3: ~; 64Cu-4: ! ) and liver (64Cu-3: ^; 64Cu-4: *), and C) in the tumor (64Cu-3: *; 64Cu-4: &). Curves of the control peptide 64Cu-3 are displayed in grey and those of the NIA peptide 64Cu-4 are shown in black.
probably attributable to the coupled NIA moiety. Remarkably, a higher accumulation of 64Cu-4 was detectable in tumor tissue, which is a first indication for the entrapment of the NIA conjugate inside tumor cells. The biodistribution of the radiolabeled NIApeptide 4 was studied in more detail by organ extraction following intravenous injection of the radiolabeled peptide (see Figure S3 in the Supporting Information). After 5 or 60 min, organs were harvested, weighed and radioactivity was measured. In agreement with the PET studies, analysis of the biodistribution of peptide 64 Cu-4 revealed that it is metabolized by the liver and excreted by the kidneys and bladder. After 60 min, the majority of the
the compound reached central hypoxic regions of the tumor, an observation that has been previously described for many therapeutics and markers.[1] In contrast, the NIA conjugate 64 Cu-4 showed a higher overall accumulation, mainly in lungs and kidneys, as was previously observed in the biodistribution studies. Again, for both peptides, no accumulation in the brain was visible. Most importantly, with the NIA unit attached, a higher accumulation of peptide 64Cu-4 was visible in the tumor, even in the central hypoxic area. Obviously, the attachment of NIA to the sC18 CPP seems to support trapping of the whole conjugate, similar to other nitroimidazole compounds.[2326] These results nicely confirm the transport and accumulation of peptide 64Cu-4 in the central hypoxic regions of the tumor. We also assume that the active mechanism of binding to bioreductive metabolites probably leads to the observed increased tumor uptake, dependent on the presence of hypoxia. In summary, we have demonstrated that attachment of 2-(2nitroimidazol-1-yl)acetic acid (NIA) to cell-penetrating peptides (CPPs) can be used to generate conjugates able to target hypoxic tumor tissues. We observed increased uptake of these conjugates even in less well-perfused hypoxic regions, indicating a successful capture of the conjugate due to the coupled NIA unit. With this new approach, we created a targeted delivery system that is applicable to act both as an imaging agent and as a platform to deliver therapeutics through conjugation. In our opinion, this system is a promising tool for applications in the emerging field of theragnostics, the combination of therapeutics and diagnostics.
www.chemmedchem.org
59
MED
Radiolabeling: 64Cu-labeling was performed by adding an aqueous solution of 64CuCl2 (50 MBq, 100 mL, 0.1 m NH4OAc) to peptide (100 mg) dissolved in MeCN/H2O (1:1 v/v, 100 mL). The reaction mixture was heated at 90 8C for 30 min and then analyzed by radioHPLC (Hewlett Packard Series 1100; detector: Ramona, Raytest Isotopenmessgerte GmbH, Straubenhardt, Germany). The compounds were analyzed using a Zorbax C18 300SB column (9.4 250 mm, 4 mm) and gradient H2O + 0.1 % TFA (A), MeCN + 0.1 % TFA (B) with 5 min 95 % A; through 10 min to 95 % B; and 5 min 95 % B with a flow of 3 mL min1. An identical system was used for the purification of the reaction mixture in the case of radiochemical yields lower than 95 %. Before application, the labeled peptides were filtered (Rotilabo-Spritzenfilter; polyvinylidene fluoride, nonsterile, pore size = 0.45 mm, = 33 mm) and diluted with electrolyte solution E153 (Serumwerk Bernburg AG, Bernburg, Germany). Animal experiments: All animal experiments were carried out in compliance with German law governing the conduct of animal experimentation and approved by the regional council for animal care in Dresden and the Landesdirektion Dresden. For in vivo studies, six-week-old male Wistar rats with an average weight of 164 22 g were used. The animals were housed under standard laboratory conditions with open access to standard food and tap water. Animals were anesthetized by inhalation of desfluran (8 % in a 30 % oxygen/air mixture). The experiments were performed using four female NMRI (nu/nu) mice from the specific pathogen-free breeding facility of the Experimental Centre of the Medical Faculty Carl Gustav Carus, University of Technology (Dresden, Germany). To further immunosuppress the nude mice, they were whole-body irradiated two days before tumor transplantation with 4 Gy (200 kV X-rays, 0.5 mm Cu filter, dose rate ~ 1 Gy min1). The established human hypopharyngeal squamous cell carcinoma line FaDu is kept in high passage by the American Type Culture Collection (Rockville, MD, USA). FaDu grows as an undifferentiated carcinoma in nude mice. For the experiments, source tumors were cut into small pieces and transplanted subcutaneously into the right hind-leg of anaesthetized mice. Quality assurance included the monitoring of the tumors by histology, LDH-isoenzyme pattern, volume doubling time (VDT) and microsatellite analysis.[27] Small-animal positron emission tomography (PET) imaging studies: Small-animal PET imaging studies were performed after injecting 1 mL min1 of the diluted 68Ga-DOTApeptide solution in isotonic NaCl solution with an activity between 15 and 25 MBq in the tail vein. Arterial blood samples were withdrawn using a catheter in the right femoral artery. Blood samples were taken after 1, 3, 5, 10, 20, 30, 60 min; plasma proteins were precipitated with MeCN/TFA/ H2O (50:5:45, v/v/v) and separated by centrifugation (13 000 rpm, 3 min). The supernatants were analyzed by radio-HPLC as described above. Dynamic PET studies (120 min duration) were obtained using a microPET Primate P4 Scanner (Siemens Medical Solutions, Erlangen, Germany) with a field of view (FOV) of axial 7.5 cm and transaxial 22 cm. The scanner had a computer-controlled bed and operated in three-dimensional list mode. The raw data were transformed into three-dimensional sinograms followed by Fourier rebinning and three-dimensional iterative image reconstitution combined with an MAP reconstruction (3D-OSEM + MAP). The intrinsic resolution at the center of FOV was 1.85 mm, and the spatial resolution obtained with this reconstitution ranged from 2.2 to 2.7 mm. Transmission correction was achieved by transmission scans of 18 min (57Co source). The activitytime curves in the interesting organs were derived from three-dimensional regions of interest (ROI) by the ROVER program (ABX GmbH, Radeberg, Germany) and analyzed using the program system R. The obtained values were normalized to the injected activity. Biodistribution: Male Wistar rats were intravenously injected with radiolabeled peptide solutions with an activity of 2 MBq in 0.5 mL isotonic NaCl solution. The animals were euthanized 5 and 60 min post-injection, respectively. Blood and organs were collected, wetweighed, and radioactivity was counted in a PerkinElmer Precisely Wizard 3 Automatic Gamma Counter (PerkinElmer, Waltham, USA). The percent dose per gram (% ID g1) was determined for each sample, and the radioactivity of the tissue samples was decay-corrected. Values are quoted as mean standard deviation (SD) for a group of four animals. Ex vivo autoradiography: The in vivo distribution of the compounds was studied 60 min after intravenous application. The FaDu tumorbearing mice were sacrificed with CO2 and directly frozen in EtOH/ dry ice at 60 8C. Further deep-frozen animals were embedded in a carboxymethyl cellulose gel matrix and frozen. The stage-mounted mice were cut in a Cryo-Polycut S (Reichert-Jung, Germany) cryomicrotome into 40 mm coronal sections, which were attached to Tesa tape and dried. For autoradiography, the whole-body cryostat sections were exposed to imaging plates (20 25 cm imaging plates for BAS-5000, Fujifilm) for 30 min and scanned for histologic autographic comparison.
Acknowledgements
We thank K. Lbner for help with cell culture, R. Herrlich for the help with biodistribution, R. Reppich-Sacher for help with MALDI and ESI mass spectrometry, and J. Hoyer for carefully reading this manuscript. Radiolabeling and in vivo studies were conducted at the FZ Dresden-Rossendorf. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) within the project FOR 630 Biological function of organometallic compounds. I.N. thanks Professor Dr. A.G. Beck-Sickinger (Universitt Leipzig, Germany) for generous access to all facilities of the institute. Keywords: antitumor agents cell-penetrating peptides drug delivery hypoxia tumor targeting
[1] J. M. Brown, A. J. Giaccia, Cancer Res. 1998, 58, 1408. [2] A. L. Harris, Nat. Rev. Cancer 2002, 2, 38. [3] S. D. Young, R. S. Marshall, R. P. Hill, Proc. Natl. Acad. Sci. USA 1988, 85, 9533. [4] J. D. Chapman, E. L. Engelhardt, C. C. Stobbe, R. F. Schneider, G. E. Hanks, Radiother. Oncol. 1998, 46, 229. [5] L. Hoigebazar, J. M. Jeong, S. Y. Choi, J. Y. Choi, D. Shetty, Y. S. Lee, D. S. Lee, J. K. Chung, M. C. Lee, Y. K. Chung, J. Med. Chem. 2010, 53, 6378. [6] G. Lucignani, Eur. J. Nucl. Med. Mol. Imaging 2008, 35, 838. [7] A. Nunn, K. Linder, H. W. Strauss, Eur. J. Nucl. Med. 1995, 22, 265. [8] A. Joliot, A. Prochiantz, Nat. Cell Biol. 2004, 6, 189. [9] S. Pujals, J. Fernandez-Carneado, C. Lopez-Iglesias, M. J. Kogan, E. Giralt, Biochim. Biophys. Acta Biomembr. 2006, 1758, 264. [10] C. Walther, K. Meyer, R. Rennert, I. Neundorf, Bioconjugate Chem. 2008, 19, 2346. [11] R. Rennert, I. Neundorf, H. G. Jahnke, P. Suchowerskyj, P. Dournaud, A. Robitzki, A. G. Beck-Sickinger, ChemMedChem 2008, 3, 241. [12] C. Walther, I. Ott, R. Gust, I. Neundorf, Biopolymers 2009, 92, 445. [13] I. Neundorf, J. Hoyer, K. Splith, R. Rennert, W. Peindy NDongo H, U. Schatzschneider, Chem. Commun. 2008, 5604.
60
www.chemmedchem.org
[14] K. Splith, W. Hu, U. Schatzschneider, R. Gust, I. Ott, L. A. Onambele, A. Prokop, I. Neundorf, Bioconjugate Chem. 2010, 21, 1288. [15] K. Splith, I. Neundorf, W. Hu, H. W. P. NDongo, V. Vasylyeva, K. Merz, U. Schatzschneider, Dalton Trans. 2010, 39, 2536. [16] I. Neundorf, R. Rennert, J. Franke, I. Kozle, R. Bergmann, Bioconjugate Chem. 2008, 19, 1596. [17] E. Vivs, J. Schmidt, A. Plegrin, Biochim. Biophys. Acta, Rev. Cancer 2008, 1786, 126. [18] I. Neundorf, R. Rennert, J. Hoyer, F. Schramm, K. Lbner, I. Kitanovic, S. Wlfl, Pharmaceuticals 2009, 2, 49. [19] Z. F. Su, X. Zhang, J. R. Ballinger, A. M. Rauth, A. Pollak, J. R. Thornback, Bioconjugate Chem. 1999, 10, 897. [20] A. Tossi, M. Scocchi, B. Skerlavaj, R. Gennaro, FEBS Lett. 1994, 339, 108. [21] S. Trabulo, A. L. Cardoso, M. Mano, M. C. P. de Lima, Pharmaceuticals 2010, 3, 961.
[22] G. Gasser, L. Tjioe, B. Graham, M. J. Belousoff, S. Juran, M. Walther, J. U. Kunstler, R. Bergmann, H. Stephan, L. Spiccia, Bioconjugate Chem. 2008, 19, 719. [23] S. T. Lee, A. M. Scott, Semin. Nucl. Med. 2007, 37, 451. [24] W. J. Koh, J. S. Rasey, M. L. Evans, J. R. Grierson, T. K. Lewellen, M. M. Graham, K. A. Krohn, T. W. Griffin, Int. J. Radiat. Oncol. Biol. Phys. 1992, 22, 199. [25] M. B. Parliament, J. D. Chapman, R. C. Urtasun, A. J. McEwan, L. Golberg, J. R. Mercer, R. H. Mannan, L. I. Wiebe, Br. J. Cancer 1992, 65, 90. [26] A. M. Rauth, Int. J. Radiat. Oncol., Biol., Phys. 1984, 10, 1293. [27] C. Schtze, R. Bergmann, A. Yaromina, F. Hessel, J. Kotzerke, J. Steinbach, M. Baumann, B. Beuthien-Baumann, Radiother. Oncol. 2007, 83, 311. Received: August 22, 2011 Published online on September 29, 2011
www.chemmedchem.org
61
MED
DOI: 10.1002/cmdc.201100393
Rofecoxib Analogues Possessing a Nitric Oxide Donor Sulfohydroxamic Acid (SO2NHOH) Cyclooxygenase-2 Pharmacophore: Synthesis, Molecular Modeling, and Biological Evaluation as Anti-inflammatory Agents
Atul Bhardwaj, Zhangjian Huang, Jatinder Kaur, and Edward E. Knaus*[a]
The design of selective cyclooxygenase-2 (COX-2) inhibitors as anti-inflammatory (AI) drugs that preferentially block the inducible COX-2 isozyme, which produces undesirable peripheral inflammation, over the constitutive COX-1 isozyme, which provides desirable gastroprotection and vascular homeostasis, represents an important milestone in the development of nonsteroidal anti-inflammatory drugs (NSAIDs). The discovery that celecoxib (1 a),[1] rofecoxib (2),[2] and valdecoxib (3)[3] show a low risk of gastrointestinal irritation provided validation for this original drug design concept (Figure 1). The subsequent obserand anti-aggregatory prostacyclin (PGI2) along with a contraindicated simultaneous increase in the level of the undesirable prothrombotic thromboxane A2 (TxA2), which induces vasoconstriction and platelet aggregation.[6] Nitric oxide (NO) induces vascular relaxation (vasodilation) and is an inhibitor of platelet aggregation and adhesion.[7] Benzenesulfohydroxamic acid [PhSO2NHOH, Pilotys acid (PA)] is a NO donor.[8] Accordingly, replacement of the methanesulfonylphenyl substituent in rofecoxib (2) with a para-benzenesulfohydroxamic acid moiety offers an attractive drug design concept to circumvent the adverse cardiovascular events associated with the chronic clinical use of highly selective COX-2 inhibitors, particularly rofecoxib. Herein we describe an investigation directed toward the synthesis, molecular modeling, and biological evaluation of NO donor sulfohydroxamic analogues of rofecoxib (compounds 1619) that, unlike rofecoxib, may be devoid of adverse cardiovascular effects. The sulfonyl chloride (SO2Cl) precursors 1315 (R1 = H, F, Cl) required for synthesis of the target sulfohydroxamic acids 16 18 and the sulfohydroxamic acid methyl ether 19 were prepared by synthetic methodologies reported previously[10] as illustrated in Scheme 1. Chorosulfonation occurred primarily at
Figure 1. Some selective COX-2 inhibitors (1 a, 1 c (SC-558),[9] 2, and 3 a), sulfohydroxamic acid analogue of celecoxib (1 b), sulfohydroxamic acid metabolite of valdecoxib (3 b), and the target sulfohydroxamic acid nitric oxide donor analogues of rofecoxib (1619).
vation of an increased risk of thrombosis and stroke in patients undergoing chronic treatment with some COX-2-selective inhibitors[4] ultimately culminated with the withdrawal of rofecoxib and valdecoxib.[5] A rational explanation for the elevation in blood pressure and increased incidence of adverse cardiovascular events was attributed to alterations of the balance in the COX pathway by highly selective COX-2 inhibitors; this leads to a decrease in the level of the beneficial vasodilatory
[a] Dr. A. Bhardwaj, Dr. Z. Huang, Dr. J. Kaur, Prof. E. E. Knaus Faculty of Pharmacy and Pharmaceutical Sciences University of Alberta, Edmonton, AB, T6G 2N8 (Canada) E-mail: ekanus@pharmacy.ualberta.ca Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100393.
Scheme 1. Reagents and conditions: a) Et3N, dry CH3CN, 25 8C, 1 h; b) NaH, dry DMSO, 25 8C, 1 h; c) ClSO3H, 25 8C, 1 h; d) HONH2HCl, K2CO3, dry THF, 25 8C, 5 h; e) compound 14, CH3ONH2HCl, NaHCO3, dry THF, 25 8C, 6 h.
62
the para position of the C4 phenyl ring as previously reported.[10] Subsequent reaction of the sulfonyl chlorides 1315 with hydroxyamine hydrochloride in the presence of potassium carbonate in tetrahydrofuran at 25 8C for 5 h afforded the respective sulfohydroxamic acid (SO2NHOH) products 1618 in 5963 % isolated yield. A similar reaction of the fluoro precursor 14 with methoxyamine hydrochloride furnished the sulfohydroxamic methyl ether (SO2NHOCH3) 19 in 56 % yield. The target sulfohydroxamic acids 1618 prepared in this study, unlike N-hydroxycelecoxib (1 b, Figure 1) which is unstable,[11] are stable products that are readily isolated and purified. The sulfohydroxamic acids 1618 were designed with the expectation that: 1) replacement of the SO2CH3 COX-2 pharmacophore present in rofecoxib (2) by a SO2NHOH moiety will provide an effective COX-2 pharmacophore that can insert into the secondary pocket of the COX-2 binding site,[12] 2) the SO2NHOH moiety will release NO at pH 7.4 to counteract the adverse cardiovascular effects shown by rofecoxib, 3) incorporating a halogen substituent (R1 = F, Cl) at the para position of the C3 phenyl ring will prevent hydroxylation by a process called metabolic halogen obstruction, and 4) the sulfohydroxamic acids 1618 will exhibit AI activity because N-hydroxyvaldecoxib (3 b) is an active metabolite resulting from N-hydroxylation of valdecoxib (3 a)[13] in humans.[14] Results of in vitro COX-1/COX-2 isozyme inhibition studies (Table 1) show that the sulfohydroxamic acids 1618 and the N-methoxylsulfonamide 19 are more potent inhibitors of COX2 (IC50 : 0.2851.4 mm) than of COX-1 (IC50 : 33.0 > 100 mm). Compounds 1618, having a SO2NHOH substituent, are more potent COX-2 inhibitors than the SO2NHOCH3 compound 19. The para substituent (R1) on the C3 phenyl ring is a determinant of both COX-2 inhibitory activity and the COX-2 selectivity index (SI), for which the relative potency profile and selectivity are 18 (Cl) > 17 (F) > 16 (H). The chloro compound 18 showed higher COX-2 inhibition (IC50 = 0.28 mm), and similar selectivity (SI = 304), relative to the parent reference drug rofecoxib (IC50 = 0.48 mm; SI > 209). The larger molecular volume (Table 1) of the SO2NHOH group or R1 substituent in com-
pounds 1618 relative to the smaller SO2CH3 substituent in rofecoxib (2) dictates that the molecular volume of compounds 1618 (269282 3) is slightly larger than that of rofecoxib (265 3). The SO2NHOH compounds 1618 (log P: 1.11.7) are more polar (less lipophilic) than rofecoxib (log P = 1.8), which has a SO2CH3 substituent (Table 1). These data show that SO2NHOH is an excellent COX-2 pharmacophore to replace the SO2CH3 substituent in rofecoxib. The percent NO released from the N-hydroxy (methoxy) sulfonamides 1619 and PA upon incubation in phosphate-buffered saline (PBS at pH 7.4) was measured by quantitation of nitrite using the Griess reaction (Table 1). NO release from the sulfohydroxamic acids 1618 varied over ranges of 1620 % at 1 h, 3035 % at 4 h, 3339 % at 16 h, 3443 % at 24 h, and 34 40 % at 48 h. Near maximal release of NO occurred within the 416 h incubation timeframe. In contrast, the N-methoxylsulfonamide 19 showed an expected low percent NO release (1 4 % range). In comparison, the percent NO release from the reference drug PA was 15, 26, 48, 53, and 60 % at 1, 4, 16, 24, and 48 h, respectively. These data provide credence for the belief that a SO2NHOH moiety may release NO at an acceptable rate to circumvent the adverse cardiovascular effects associated with the use of rofecoxib. Release of NO from compounds 16 18 is believed to occur by the mechanism proposed by Zamora et al.[15] wherein oxidation of PA at pH 78 occurs on the intact molecule to furnish a radical intermediate that subsequently decomposes to release benzenesulfinic acid (PhSO2H) and NO. AI ED50 values acquired for the sulfohydroxamic acids 1618 were determined using a carrageenan-induced rat paw edema model (Table 1). Compounds 1618 exhibited good AI activity (ED50 = 17.730.2 mg kg1), for which the relative potency profile with respect to the R1 substituent on the C3 phenyl ring is Cl (18) > F (17) > H (16). In contrast, the SO2NHOCH3 compound 19 was a much weaker AI agent (17.9 % inhibition of inflammation for a 50 mg kg1 p.o. dose). These data indicate that compounds 1618, having a SO2NHOH COX-2 NO donor pharmacophore, possess acceptable absorption, distribution,
Table 1. Evaluation of rofecoxib analogues: in vitro COX-1 and COX-2 inhibition, NO release, AI (ED50) activity, molecular volume, and log P data. Compd 16 17 18 19 2 1a PA R1 H F Cl F X OH OH OH OCH3 IC50 [mm][a] COX-1 COX-2 33.0 89.0 85.7 > 100 > 100 7.7 8.6 1.6 0.28 51.4 0.48 0.07 COX-2 SI[b] 1h 3.8 56.3 304 > 1.9 > 209 110 16 20 18 2 NO Release [%][c,d] 4h 16 h 24 h 30 34 35 4 33 37 39 3 34 38 43 1 ED50[e] [mg kg1] 30.2 26.8 17.7 moderate[h] 2.6 10.8 Vmolec [3][f] 269 274 282 291 265 292 log P[g] 1.1 1.3 1.7 2.2 1.8 4.3
48 h 34 40 40 1
[a] Test compound concentration required to produce 50 % inhibition of ovine COX-1 or human recombinant COX-2 in vitro; data are the mean of two determinations acquired using the enzyme immunoassay kit (see Experimental Section for details), and the deviation from the mean is < 10 % of the mean value. [b] In vitro COX-2 selectivity index: [(COX-1 IC50)/(COX-2 IC50)]. [c] Amount of NO released as a percentage of the theoretical maximum release of 1 mol NO per mol test compound, determined using the Griess reaction; data represent the mean value of n = 3 determinations; variation from the mean was 0.02 %. [d] Incubated for various time intervals in PBS (pH 7.4) at 37 8C. [e] Inhibitory activity in a carrageenan-induced rat paw edema assay; results are expressed as the ED50 value at 3 h after oral administration of the test compound. [f] After energy minimization with the molecular mechanics geometry optimization module, the molecular volume was calculated using the Alchemy 2000 program (Tripos Inc.). [g] The log P value was calculated using ChemDraw Ultra 6.0 (CambridgeSoft Co.). [h] For a 50 mg kg1 oral dose, 17.9 % inhibition of inflammation was observed.
www.chemmedchem.org
63
MED
metabolism, excretion (ADME) properties. The most potent compound 18 (R1 = Cl; ED50 = 17.7 mg kg1) showed respectable AI activity relative to the reference drugs celecoxib (ED50 = 10.8 mg kg1) and rofecoxib (ED50 = 2.6 mg kg1). Molecular modeling studies were carried out for compounds 17 and 18 to gain insight into the probable nature of their respective binding interactions with the COX-2 and COX-1 isozymes which may explain their selective inhibition of the COX2 isozyme. The SO2NHOH group of compound 18 (COX-2 IC50 = 0.28 mm and ED50 = 17.7 mg kg1) is oriented toward the secondary pocket region of the COX-2 active site (Figure 2). However, the SO2NHOH group in compound 18 is tilted more to that described previously for compound 18 (Figure 2). However, the SO2NHOH nitrogen atom in compound 17 forms hydrogen bonds with H90 (d = 2.80 ) and S353 (d = 2.94 ), and the hydroxy group that is located in the vicinity of the R513, A516, and H90 residues of the COX-2 active site shows hydrogen bonding (OHN d = 2.41 ) with H90. These hydrogen bonding interactions show that the SO2NHOH group interacts favorably within the COX-2 active site. The 4-fluorophenyl ring of compound 17 is positioned near W387 at the entrance of the hydrophobic pocket (Supporting Information), whereas the furanone ring carbonyl oxygen atom shows a hydrogen bonding interaction with the NH2 group of A527 (d = 2.58 ). Notably, the furanone carbonyl oxygen atom in compound 18 shows hydrogen bonding (Figure 3 b) with the hydroxy group of S530 (d = 2.87 ) in the COX-2 binding site. S530 is the acetylation site of aspirin in both COX-1 and COX-2. In contrast, when 18 was docked in the COX-1 active site, no interaction of the carbonyl oxygen atom with S530 was observed (Supporting Information). Accordingly, this selective and significant hydrogen bonding of the carbonyl oxygen atom in compound 18 with S530 in the COX-2, but not COX-1, active site may play a key role in determining the selective inhibition of
Figure 2. Molecular modeling (docking) of compound 18 (carbon atoms in pink) in the binding site of COX-2 (PDB ID: 6COX; Eintermolecular = 13.55 kcal mol1). Hydrogen atoms of amino acid residues have been removed for clarity.
toward the L352, S353, and Q192 active site residues. For compound 18, the hydroxy oxygen atom shows hydrogen bonding with the NH2 group of S353 (NH2O d = 2.92 ) and L352 (d = 2.11 ) residues in the COX-2 active site. Moreover, the SO2NHOH nitrogen atom shows a hydrogen bond with L352 (d = 1.92 ). The para-chloro substituent on the C3 phenyl ring of 18 has moved into the vicinity of R120 (guanidine moiety; d = 3.14 ) near the entrance to the COX-2 binding site (Figure 2). This positioning of the para-chlorophenyl ring is consistent with the more potent in vitro COX-2 inhibition shown by 18, as the guanidine moiety (R120) forms a salt bridge with the carboxylic group of arachidonic acid during COX-2-catalyzed metabolism of arachidonic acid. A similar molecular docking simulation for compound 17 within the COX-2 active site (Supporting Information) showed that compound 17 (COX-2 IC50 = 1.6 mm and ED50 = 26.8 mg kg1) assumes a position in a region that is very similar
Figure 3. Binding conformations found for a) compound 17 (pink), and b) compound 18 (pink) in comparison with rofecoxib (green) docked in the COX-2 binding site (Eintermolecular = 13.08 kcal mol1).
64
www.chemmedchem.org
COX-2 shown by compound 18 (COX-2 SI = 304). Compound 17, unlike compound 18, does not show any hydrogen bonding with the S530 residue of the COX-2 active site, as the interspatial distance of the carbonyl oxygen atom and the hydroxy group of S530 is 7.31 (Figure 3 a). Inspection of the molecular docking illustrations for compound 17 (data not shown) and 18 (Supporting Information) within the COX-1 active site clearly indicates that the key SO2NHOH pharmacophore, which likely contributes to potent and selective COX-2 inhibition, is not able to enter into the COX-1 active site for either compounds 17 or 18. This inability may be due, at least in part, to a steric conflict with side chains of amino acid residues in the COX-1 active site, as the molecular volume of the COX-1 active site is ~ 25 % smaller (316 3) than the primary COX-2 active site and its associated secondary pocket (394 3).[12] Because the COX-1 binding site does not have an associated secondary pocket into which the SO2NHOH moiety can insert, the SO2NHOH moiety is forced to position itself outside the COX-1 active site (Supporting Information). Furthermore, none of the active site residues in COX-1 showed hydrogen bonding or significant van der Waals interactions with any substituent on compounds 17 or 18. The partial entry of compounds 17 and 18 into, and their weak interactions within, the COX-1 active site is consistent with the weak in vitro inhibition of COX-1 determined experimentally. These docking studies indicate that the orientation of the SO2NHOH moiety within, or relative to, the respective COX-1 or COX-2 active site plays a crucial role in the ligandenzyme interactions (Eintermolecular) that occur. When compounds 17 and 18 were docked in the COX-2 active site, both compounds were positioned in a region that is very similar to that observed for rofecoxib and 1 c (SC-558). In this regard, the SO2NHOH group in compounds 17 and 18 is aligned in a similar manner to the SO2CH3 group of rofecoxib (Figure 3), and this orientation is also similar to that of the SO2NH2 group of 1 c crystallized within COX-2. This observation further illustrates why the SO2NHOH analogues of rofecoxib investigated in this study retain the COX-2 inhibitory performance of the parent drug rofecoxib. In conclusion, a new group of stable and readily isolated rofecoxib analogues were designed, wherein the methylsulfonyl (SO2CH3) COX-2 pharmacophore is replaced by a sulfohydroxamic acid (SO2NHOH) moiety, and a para substituent (16, R1 = H; 17, R1 = F; 18, R1 = Cl) is introduced on the C3 phenyl ring. Biological and molecular modeling studies showed that: 1) the SO2NHOH moiety is an effective NO donor, 2) the SO2NHOH moiety confers potent and selective COX-2 inhibitory activity (R1 = F, Cl), 3) the fluoro (17) and chloro (18) compounds assume a favorable orientation inside the COX-2 binding site which allows for multiple hydrogen bonding interactions, whereas 17 and 18 undergo partial insertion into the COX-1 binding site, and 4) compound 18 (R1 = Cl) is a potent and selective COX-2 inhibitor (IC50 = 0.28 mm; SI = 304) that exhibits appreciable in vivo AI activity (ED50 = 17.7 mg kg1). The results of this investigation provide strong evidence that replacement of the SO2CH3 moiety in rofecoxib by a NO donor SO2NHOH COX-2 pharmacophore offers a rational drug design approach
to circumvent adverse thrombotic and hypertensive effects associated with the chronic use of selective COX-2 inhibitors.
Chemistry
General: Melting points (mp) were measured in capillaries using a ThomasHoover capillary apparatus and are uncorrected. 1H and 13 C NMR spectra were measured on a Bruker AM 300 NMR spectrometer with [D6]DMSO as the solvent. Chemical shifts (d) are given in parts per million (ppm) with tetramethylsilane (TMS) as an internal reference. Infrared (IR) spectra were recorded as films on NaCl plates with a Nicolet 550 series II Magna FTIR spectrometer. Mass spectra (MS) were recorded on a Waters Micromass ZQ 4000 mass spectrometer using electrospray ionization (ESI). The purity of the compounds was assessed by elemental analysis, and these microanalyses were performed for C, H, N, and S by the Microanalytical Service Laboratory, Department of Chemistry, University of Alberta. Compounds 1619 showed a single spot on Macherey Nagel Polygram Sil G/UV254 silica gel plates (0.2 mm) using a low, medium, and highly polar solvent system, and no residue remained after combustion, indicating purity > 98 %. Column chromatography was performed on a Combiflash Rf system with a gold silica column. All other reagents, purchased from Aldrich Chemical Co. (Milwaukee, WI, USA), were used without further purification. Compounds 1315 were synthesized according to our previously reported method.[10] N-Hydroxy-4-(5-oxo-4-phenyl-2,5-dihydrofuran-3-yl)benzenesulfonamide (16): HONH2HCl (83 mg, 1.2 mmol) and K2CO3 (332 mg, 2.4 mmol) were added to a solution of 13 (200 mg, 0.6 mmol) in dry THF (15 mL) with stirring, and the reaction was allowed to proceed with vigorous stirring at 25 8C for ~ 5 h, after which time the reaction was complete as indicated by the disappearance of 13 (TLC; EtOAc/hexane, 1:1 v/v). The reaction mixture was filtered through a pad of Celite to furnish a clear filtrate, which was concentrated in vacuo to remove the solvent. EtOAc (30 mL) was added, and this mixture was washed with H2O (25 mL) and then brine (20 mL), and the organic fraction was dried (MgSO4) and filtered. Removal of EtOAc in vacuo gave a brown viscous oil that was purified by flash silica gel column chromatography (n-hexane/ EtOAc, 3:2 v/v) to afford 16 as a white solid (63.2 %); mp: 190 192 8C; 1H NMR (300 MHz, [D6]DMSO): d = 5.43 (s, 2 H), 7.367.44 (m, 5 H), 7.57 (d, J = 6 Hz, 2 H), 7.86 (d, J = 6 Hz, 2 H), 9.64 (s, 1 H), 9.68 ppm (s, 1 H); 13C NMR (75 MHz, [D6]DMSO): d = 71.0, 123.4, 127.6, 128.3, 129.0, 129.7, 130.1, 130.81, 135.3, 137.2, 159.4, 172.4 ppm; IR (film): n = 3367, 3225, 1732, 1175 cm1; MS (ESI +): m/z: 354 [M + Na] + ; Anal. calcd for C16H13NO5S: C 58.00, H 3.95, S 9.68, N 4.23, found: C 57.97, H 4.11, S 9.79, N 4.28. N-Hydroxy-4-[4-(4-fluorophenyl)-5-oxo-2,5-dihydrofuran-3-yl]benzenesulfonamide (17): Starting from 14 and using a similar procedure to that described for the preparation of 16, the title compound was obtained as a white solid (59.6 %); mp: 150151 8C; 1 H NMR (300 MHz, [D6]DMSO): d = 5.39 (s, 2 H), 7.227.31 (m, 2 H), 7.377.41 (m, 2 H), 7.58 (d, J = 8 Hz), 7.67 (d, J = 6 Hz, 2 H), 9.62 (s, 1 H), 9.70 ppm (s, 1 H); 13C NMR (75 MHz, [D6]DMSO): d = 70.8, 115.8 (d, JCCF = 20.7 Hz), 125.6, 128.2, 129.5, 130.0, 131.3 (d, JCCCF = 7.6 Hz), 135.1, 138.6, 156.2, 162.2 (d, JCF = 243.47 Hz), 172.4 ppm; IR (film): n = 3398, 3228, 1728, 1169 cm1; MS (ESI): m/z: 348 [MH] ; Anal. calcd for C16H12FNO5S: C 55.01, H 3.46, S 9.18, N 4.01, found: C 55.08, H 3.49, S 9.19, N 4.02.
www.chemmedchem.org
65
MED
N-Hydroxy4-[4-(4-chlorophenyl)-5-oxo-2,5-dihydrofuran-3-yl]benzenesulfonamide (18): Starting from 15 and using a similar procedure to that described for the preparation of 16, the title compound was obtained as a white solid (59 %); mp: 170172 8C; 1 H NMR (300 MHz, [D6]DMSO): d = 5.41 (s, 2 H), 7.357.38 (m, 2 H), 7.477.52 (m, 2 H), 7.59 (d, J = 6 Hz, 2 H), 7.67 (d, J = 6 Hz, 2 H), 9.64 (s, 1 H), 9.70 ppm (s, 1 H); 13C NMR (75 MHz, [D6]DMSO): d = 70.9, 125.4, 128.2, 128.5, 128.6, 128.8, 129.6, 130.8 (sulfonylphenyl C2, C6; C3, C5 and 4-chlorophenyl C2, C6; C3, C5), 130.8, 133.6, 134.9, 138.1 (sulfonylphenyl C1, C4), 156.7, 172.2 ppm; IR (film): n = 3392, 3237, 1740, 1183 cm1; MS (ESI): m/z: 364 [MH] , 366 [MH]1; Anal. calcd for C16H12ClNO5S: C 52.54, H 3.31, S 8.77, N 3.83, found: C 52.57, H 3.34, S 8.73, N 3.81. N-Methoxy-4-[4-(4-fluorophenyl)-5-oxo-2,5-dihydrofuran-3-yl]benzenesulfonamide (19): CH3ONH2HCl (100 mg, 1.2 mmol) and NaHCO3 (201 mg, 2.4 mmol) were added to a stirred solution of 14 (200 mg, 0.6 mmol) in dry THF (15 mL), and the reaction was allowed to proceed with vigorous stirring at 25 8C for ~ 6 h, until the reaction was complete as determined by TLC (EtOAc/hexane, 3:2 v/v). The reaction mixture was filtered through a Celite pad to give a clear filtrate, and the solvent was removed in vacuo. EtOAc (30 mL) was added and this mixture was washed with H2O (30 mL) and then brine (20 mL), and the organic fraction was dried (MgSO4) and filtered. Removal of EtOAc in vacuo gave a brown viscous oil that was purified by flash silica gel column chromatography (n-hexane/EtOAc, 3:2 v/v) to afford 19 as a white solid (56 %); mp: 122 8C; 1H NMR (300 MHz, [D6]DMSO): d = 3.64 (s, 3 H), 5.41 (s, 2 H), 7.247.30 (m, 2 H), 7.367.41 (m, 2 H), 7.69 (d, J = 6 Hz, 2 H), 7.87 (d, J = 6 Hz, 2 H), 10.6 ppm (s, 1 H); 13C NMR (75 MHz, [D6]DMSO): d = 64.8, 70.8, 115.8 (d, JCCF = 21.84 Hz), 125.3, 128.4, 129.3, 130.0, (d, JCCC = 3.3 Hz, 131.3 (d, JCCF = 10.9 Hz), 135.4, 138.4, 156.1, 162.2 (d, JCF = 243.5 Hz), 172.5 ppm; IR (film): n = 3184, 1734, 1340, 1172 cm1; MS (ESI): m/z: 362 [MH] ; Anal. calcd for C17H14FNO5S: C 56.19, H 3.88, S 8.82, N 3.85, found: C 55.64, H 3.81, S 8.72, N 3.81. Nitric oxide release assay: In vitro NO release, upon incubation of the test compound (2.4 mL of 5.0 102 mm) with PBS at pH 7.4 and 37 8C for 1, 4, 16, 24, and 48 h was determined by quantification of nitrite produced by the reaction of NO with oxygen and water using the Griess reaction. The amount of NO liberated at each time interval was acquired for test compounds 1619 using reported procedures.[18]
Molecular modeling (docking)

Coordinates from the X-ray crystal structure of COX-1 (ovine, 1EQG, ibuprofen bound in the active site) and COX-2 (murine, 6COX, 1 c bound in the active site) were taken from the RCSB Protein Data Bank. Compounds were constructed with the builder toolkit of the software package ArgusLab 4.0.1,[19] and energy minimized using the semiempirical quantum mechanical method PM3. The monomeric structure of the enzyme was chosen, and the active site was defined around the ligand. The molecule to be docked in the enzyme active site was inserted into the work space carrying the structure of the enzyme. The docking program implements an efficient grid-based docking algorithm, which approximates an exhaustive search within the free volume of the binding site cavity. The conformational space was surveyed by the geometry optimization of the flexible ligand (rings are treated as rigid) in combination with the incremental construction of the ligand torsions. Thus, docking occurred between the flexible ligand parts of the compound and enzyme. The ligand orientation was determined by a shape scoring function based on Ascore and the final positions were ranked by lowest interaction energy values. The Einteraction value is the sum of the energies involved in hydrogen bond interactions, hydrophobic interactions, and van der Waals interactions. Hydrogen bond and hydrophobic interactions between the compound and enzyme were explored by distance measurements.
Acknowledgements
We are grateful to the Canadian Institutes of Health Research (grant no. MOP-14712) for financial support of this research. Keywords: biological activity cyclooxygenases drug design inflammation inhibitors
[1] T. D. Penning, J. J. Talley, S. R. Bertenshaw, J. S. Carter, P. W. Collins, S. Docter, M. J. Graneto, L. F. Lee, J. W. Malecha, J. M. Miyashiro, R. S. Rogers, D. J. Rogier, S. S. Yu, G. D. Anderson, E. G. Burton, J. N. Cogburn, S. A. Gregory, C. M. Koboldt, W. E. Perkins, K. Seibert, A. W. Veenhuizen, Y. Y. Zhang, P. C. Isakson, J. Med. Chem. 1997, 40, 1347 1365. [2] P. Prasit, Z. Wang, C. Brideau, C.-C. Chan, S. Charleson, W. Cromlish, D. Ethier, J. F. Evans, A. W. Ford-Hutchinson, J. Y. Gauthier, R. Gordon, J. Guay, M. Gresser, S. Kargman, B. Kennedy, Y. Leblanc, S. Leger, J. Mancini, G. P. ONeill, M. Ouellet, M. D. Percival, H. Perrier, D. Riendeau, I. Rodger, P. Tagari, M. Therien, P. Vikers, E. Wong, L.-J. Xu, R. N. Young, R. Zamboni, S. Boyce, N. Rupniak, M. Forrest, D. Visco, D. Patrick, Bioorg. Med. Chem. Lett. 1999, 9, 1773 1778. [3] J. J. Talley, D. L. Brown, J. S. Carter, M. J. Graneto, C. M. Koboldt, J. L. Masferrer, W. E. Perkins, R. S. Rogers, A. F. Shaffer, Y. Y. Zhang, B. S. Zweifel, K. Seibert, J. Med. Chem. 2000, 43, 775 777. [4] D. Mukherjee, S. E. Nissen, E. J. Topol, JAMA 2001, 286, 954 959. [5] L. J. Marnett, Can. Prev. Res. 2009, 2, 288 291. [6] D. Mukherjee, Biochem. Pharmacol. 2002, 63, 817 821. [7] A. R. Butler, D. L. H. Williams, Chem. Soc. Rev. 1993, 22, 233 241. [8] S. B. King, H. T. Nagasawa in Methods in Enzymology (Ed.: L. Packer), New York, 1998, pp. 211 220.
Biology
In vivo anti-inflammatory assays were carried using a protocol approved by the Health Sciences Animal Welfare Committee at the University of Alberta (Edmonton, Canada). Cyclooxygenase inhibition assay: The ability of the test compounds 1619 and rofecoxib listed in Table 1 to inhibit ovine COX-1 and human recombinant COX-2 (IC50 values, mm) was determined using an enzyme immunoassay (EIA) kit (catalogue no. 560131, Cayman Chemical, Ann Arbor, MI, USA) according to a previously reported method.[16] Anti-inflammatory assay: In vivo anti-inflammatory activity was measured using a carrageenan-induced rat paw edema assay described by Winter et al.[17] In brief, three male SpragueDawley rats weighing 160180 g were used in each group. Test compounds 1619 suspended in water containing 1 % methyl cellulose were administered orally for a minimum of three different doses (10 50 mg kg1) 1 h prior to a 0.05 mL subcutaneous injection of 1 % carrageenan in 0.9 % NaCl solution under the planter skin of the right hind paw. Control experiments were identical, except that the vehicle did not contain a test compound. The volume of the injected paw was measured at 0 and 3 h using a UGO Basile 7141 Plethysmometer (series no. 43201). A doseresponse curve was constructed which was used to determine the ED50 value.
66
www.chemmedchem.org
[9] R. G. Kurumbail, A. M. Stevens, J. K. Gierse, J. J. McDonald, R. A. Stegeman, J. Y. Pak, D. Gildehaus, J. M. Myashiro, T. D. Penning, K. Seibert, P. C. Isakson, W. C. Stallings, Nature 1996, 384, 644 648. [10] M. J. Uddin, P. N. P. Rao, E. E. Knaus, J. Heterocycl. Chem. 2003, 40, 861 868. [11] G. Szab, J. Fischer, A. Kis-Varga, K. Gyires, J. Med. Chem. 2008, 51, 142 147. [12] C. Luong, A. Miller, J. Barnett, J. Chow, C. Ramesha, M. F. Browner, Nature Struct. Biol. 1996, 3, 927 933. [13] P. Erdlyi, T. Fodor, A. K. Varga, M. Czugler, A. Gere, L. Fischer, Bioorg. Med. Chem. 2008, 16, 5322 5330. [14] J. J. Yuan, D. C. Yang, J. Y. Zhang, R. Bible, A. Karim, J. W. A. Findlay, Drug Metab. Dispos. 2002, 30, 1013 1021. [15] R. Zamora, A. Grzesiok, H. Weber, M. Feelisch, Biochem. J. 1995, 312 (Pt. 2), 333 339.
[16] M. J. Uddin, P. N. P. Rao, R. McDonald, E. E. Knaus, J. Med. Chem. 2004, 47, 6108 6111. [17] C. A. Winter, E. A. Risley, G. W. Nuss, Proc. Soc. Exp. Biol. Med. 1962, 111, 544 547. [18] M. T. Cocco, C. Congiu, V. Onnis, M. Morelli, O. Cauli, Eur. J. Med. Chem. 2003, 38, 513 518. [19] A. Mark, ArgusLab version 4.0.1, Thompson Planaria Software LLC, Seattle, WA, (USA): http://www.arguslab.com (accessed September 14, 2011).
Received: August 15, 2011 Revised: September 12, 2011 Published online on October 11, 2011
www.chemmedchem.org
67
MED
DOI: 10.1002/cmdc.201100350
Insights into Matriptase-2 Substrate Binding and Inhibition Mechanisms by Analyzing Active-Site-Mutated Variants
Eva Maurer,[a] Mihiret T. Sisay,[a] Marit Stirnberg,[a] Torsten Steinmetzer,[b] Jrgen Bajorath,[c] and Michael Gtschow*[a]
Matriptase-2 (transmembrane protease, serine 6, TMPRSS6; MT2), predominantly expressed in the liver, was initially identified in human and mouse.[1, 2] MT2 belongs to the family of type II transmembrane serine proteases, representing an emerging class of cell surface proteolytic enzymes.[3] Recent findings have revealed a link between mutations in MT2 and iron-refractory iron deficiency anemia.[4, 5] MT2 suppresses the expression of hepcidin,[4, 6] the main regulator of systemic iron homeostasis, through cleavage of the bone morphogenetic protein co-receptor hemojuvelin.[79] Due to the important role of MT2 in iron homeostasis, the enzyme represents a novel target for the development of inhibitors, potentially useful in the treatment of hemochromatosis or beneficial as pharmacological tools. In a previous work, we identified two dipeptide amides as the first low-molecular weight inhibitors of MT2.[10] MT2 shares high sequence similarity with matriptase (membrane-type serine protease 1, MT-SP1, suppressor of tumorigenicity 14), particularly in the catalytic domain which is about 45 % identical.[1] Matriptase is expressed in epithelial cells and involved in tumor pathogenesis.[1115] Owing to its potential as a therapeutic target, synthetic inhibitors for matriptase have been described, such as mono/bisbenzamidines and derivatives of the sunflower trypsin inhibitor.[1618] In this study, we have investigated the interaction of MT2 with prototype low-molecular weight ligands using site-directed mutagenesis, kinetic analysis and molecular modeling. A peptidic substrate[1] and inhibitors from our previous study[10] were chosen to perform an active site scan of MT2. Our data reveal insights into substrate/inhibitorenzyme interactions and structural differences in the active sites of MT2 and matriptase. Furthermore, amino acids that enhanced (Phe 665) or reduced (Tyr 712, Asp 785) the affinity of peptide ligands were identified. For the design of potent and low-molecular weight inhibitors, an X-ray crystal structure of the target enzyme is beneficial. Although MT2 was confirmed as a key regulator of iron homeostasis, no crystal structure is currently available, in con[a] E. Maurer, Dr. M. T. Sisay, Dr. M. Stirnberg, Prof. Dr. M. Gtschow Pharmazeutisches Institut, Universitt Bonn An der Immenburg 4, 53121 Bonn (Germany) E-mail: guetschow@uni-bonn.de [b] Prof. Dr. T. Steinmetzer Institut fr Pharmazeutische Chemie, Universitt Marburg Marbacher Weg 6, 35032 Marburg (Germany) [c] Prof. Dr. J. Bajorath Bonn-Aachen International Center for Information Technology Universitt Bonn Dahlmannstr. 2, 53113 Bonn (Germany) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100350.
trast to the already well-characterized matriptase.[19] As such, the previously generated model of the MT2 active site, based on the crystal structure of the closely related matriptase,[10] was used in the present study. The catalytic domain of MT2 features the common fold of chymotrypsin-like serine proteases. In the active site, MT2 and matriptase vary in relevant amino acids participating in the formation of the S1 and S2 pockets and the S3/S4 binding region. To obtain insights into these structural differences between MT2 and matriptase and the resulting enzymeligand interactions, four important MT2 amino acids (His 665, Glu 712, Ala 757, Leu 785; corresponding to His 99, Glu 146, Ala 190, Leu 217 according to the chymotrypsinogen numbering) were mutated to the equivalent residues in matriptase (Figure 1).
Figure 1. Location of His 665, Glu 712, Ala 757 and Leu 785 in the active site of matriptase-2. The active site is shown in surface representation. Amino acids selected for mutation are displayed in stick representation.
These amino acids are involved in the formation of the active site pockets and are known to directly interact with both the substrate and inhibitor. We introduced phenylalanine at position His 665 in MT2 to investigate the role of hydrophobicity within the S2 and upper part of the S3/S4 region on substrate/inhibitor accommodation. Glu 712 was mutated to tyrosine in order to characterize the role of the small binding subsite above the S1 pocket. Both matriptases have almost identical residues inside the S1 pocket except for Ala 757 in MT2 instead of serine in matriptase at the bottom of the pocket. Therefore, the MT2A757S variant was generated. Furthermore, we mutated Leu 785 to aspartic acid at the lower part of the S3/S4 pocket to analyze whether the change in hydrophobicity at this position influences the binding mode of the sub strate or the potency of the inhibitors. Finally, variants bearing
68
simultaneous mutations were included: MT2A757S/L785D and MT2E712Y/A757S/L785D. These MT2 variants were obtained by site-directed mutagenesis. HEK cells were transfected with cDNA encoding wild-type or the mutated forms of MT2 with a Myc-tag at the C terminus of the protein. Anti-c-Myc antibody immunoblot analysis showed the presence of the catalytic domain (~ 30 kDa) in the conditioned media, and the full length form (~ 120 kDa) in membrane fractions of all mutants (Figure 2). These results are consistent with the detection of the enzyme in media and membrane fractions of transfected wild-type MT2 HEK cells,[20] indicating that the proteolytic processing of MT2 is not affected by these mutations.
Figure 2. Immunoblot analysis of active-site-mutated matriptase-2 (MT2) variants. HEK cells expressing MT2WT, MT2H665F, MT2E712Y, MT2A757S, MT2L785D, MT2A757S/L785D and MT2E712Y/A757S/L785D were cultured for two days with OptiMEM and then fractionated as described.[20] Total protein (30 mg) from conditioned media (CM) and membrane fractions (MF) were separated under reducing conditions (M: molecular mass marker). Samples were electroblotted onto nitrocellulose membranes and MT2 was visualized using monoclonal mouse anti-c-Myc antibody. The immunoblots shown are representative of three independent experiments.
Previously, we analyzed the kinetics of the MT2- and matriptase-catalyzed cleavage of chromogenic substrate Boc-Gln-AlaArg-para-nitroanilide.[10] In the present study, the Km values of the active-site-mutated variants were calculated via nonlinear regression and compared with wild-type MT2 (Table 1). Docking of the substrate into the active site of wild-type MT2 suggests that the basic group of the P1 arginine side chain makes a salt bridge with Asp 756 at the bottom of the S1 pocket (Figure 3). The NH and CO groups of the substrates P3 glutamine form a short antiparallel b sheet-like structure with the CO and NH groups of the enzymes Gly 784 that is critical for the proper binding of the substrate within the active site. In
this orientation, the P2 alanine side chain occupies the S2 pocket and the tert-butyloxycarbonyl (Boc) group forms lipophilic interactions with the side chains of Leu 785 and Trp 783. The scissile bond is positioned directly above the catalytic Ser 762 residue (Ser 195 according to the chymotrypsinogen numbering). Models of the variants with the substrate show a binding pattern similar to that of wild-type MT2. The P2P3 backbone is kept in position mainly by the presence of the aforementioned hydrogen-bonding network between glutamine and Gly 784, forming a short b sheet-like structure. Noteworthy, the Km value determined for MT2H665F (Km = 24 mm, Table 1) was 10-fold lower than that of wild-type MT2 (Km = 210 mm, Table 1). The introduction of phenylalanine instead of His 665 results in an overall increase in hydrophobicity of the S2 and the upper part of the S3/S4 pocket, which may have a positive effect on the binding of the P2 alanine side chain by forming more favorable lipophilic interactions with the mutated variant (Figure S1 in the Supporting Information). In the case of MT2A757S, the enzyme mutated in the S1 pocket, the Km value obtained (Km = 190 mm, Table 1) was close to that of wild-type MT2. Because the S1 pockets of MT2 and matriptase are very similar and both enzymes preferably accept substrates with arginine at the P1 position,[1, 21, 22] one might also expect similar Km values (Km = 380 mm for matriptase, Table 1). The kinetic analysis of MT2L785D showed a Km value of 620 mm, threefold higher compared with that of the wild-type enzyme (Table 1). The introduction of the acidic aspartic acid instead of Leu 785 in the lower part of the S3/S4 pocket very likely disrupts favorable lipophilic interactions with the Boc group. Inhibition of MT2 by dipeptide amides 14 has recently been reported.[10] Compounds 1 and 3 share the 4-amidino-
Table 1. Kinetic parameters of wild-type (WT) and mutated variants of matriptase-2. Variant WT H665F E712Y A757S L785D A757S/L785D E712Y/A757S/L785D Matriptase Km [mm][a] compd 1 210 10[c] 24 1 > 1000 190 10 620 30 830 50 > 1000 380 30[c] 0.19 0.01[c] 0.046 0.002 > 20 0.16 0.01 1.2 0.1 1.6 0.1 > 30 0.055 0.003[c] compd 2 30 3 2.6 0.1 > 200 69 11 54 0.2 620 280 > 200 0.22 0.01[c] Ki [mm][b] compd 3 0.29 0.02[c] 0.052 0.003 > 20 0.21 0.01 15 1 20 1 > 100 0.77 0.15[c] compd 4 37 2 3.0 0.3 > 200 75 9 220 30 280 30 > 200 2.1 0.3[c]
[a] Data represent the mean SEM from duplicate measurements at eight different substrate concentrations. [b] Data represent the mean SEM from duplicate measurements at five different inhibitors concentrations. [c] Data taken from Reference [10].
www.chemmedchem.org
69
MED
0.052 mm versus Ki = 0.29 mm for 3; Ki = 3.0 mm versus Ki = 37 mm for 4, Table 1). The cyclohexyl group is directed towards the upper part of the S3/S4 pocket forming hydrophobic interactions with Trp 783 and the introduced Phe at position 665 (Figure 4), leading to improved inhibitory activities of compounds 3 and 4. The replacement of Glu 712 by tyrosine resulted in an overall decrease in the inhibitory effect caused by 14 (Table 1). The Figure 3. Modeled complex of the substrate Boc-Gln-Ala-Arg-para-nitroanilide bound in the active site of wildmodel of this variant indicates type matriptase-2. The substrate is shown in stick representation and the carbon atoms are colored in cyan. The active site is shown in line (right) and surface (left) representations. Residues forming the active site are labeled that compound 1 binds in a simand the catalytic triad residues are displayed in stick representation. ilar manner as to wild-type MT2 (Figure S3 in the Supporting Information). However, in MT2E712Y benzylamine group, which is replaced by 2-aminomethylthe benzylsulfonyl group is located on top of the S1 pocket and is solvent exposed because the small binding subsite 5-chlorobenzylamine in compounds 2 and 4. This part of the above the S1 pocket is now occupied by the bulky side chain molecule is expected to bind to the S1 pocket, and the d-argiof Tyr 712. This orientation may have a negative impact on the nine (in 1 and 2) or d-cyclohexylalanine (d-Cha) moiety (in 3 and 4) is anticipated to bind in the S3/S4 pocket of MT2. In this study, the inhibitory potencies of 14 towards MT2 variants were determined and compared with their inhibitory activities against wild-type MT2 and matriptase (Table 1). The variant MT2H665F was fourto 12-fold more strongly inhibited by compounds 1 and 2 than the wild-type enzyme (Ki = 0.046 mm versus Ki = 0.19 mm for 1; Ki = 2.6 mm versus Ki = 30 mm for 2, Table 1). The region between the upper part of the S3/ S4 pocket and the S2 pocket is much more hydrophobic owing to the histidine!phenylalanine replacement. In the case of compounds 1 and 2, the formation of an additional cationp interaction between the phenylalanine side chain and the positively charged guanidino group of the inhibitor might be possible (Figure S2 in the Supporting Information). The inhibitory potency of compounds 3 and 4, both H665F and b) MT2WT. The inhibitor is of which contain a d-cyclohexy- Figure 4. Modeled complexes of compound 3 bound in the active site of a) MT2 shown in stick representation and the carbon atoms are colored in cyan. The active site is shown in line (right) lalanine, also increased when and surface (left) representations. Residues forming the active site are labeled and the catalytic triad residues are comparing the mutated variant displayed in stick representation. The carbon atoms of the mutated amino acid and the corresponding surface are with wild-type MT2 (Ki = colored in green.
70
www.chemmedchem.org
affinities of the inhibitors. In addition, this mutation would result in the loss of an important salt bridge formed with Arg 789, which could destabilize the active site geometry. Concerning MT2A757S, the potencies of compounds 1 and 3, both of which contain the benzamidine moiety, were slightly improved compared with those against wild-type MT2 (Table 1). The modeled complex of compound 1 inside MT2A757S shows a binding pattern similar to that of wild-type MT2. Peptides 2 and 4 exhibited stronger affinities for matriptase than for MT2 (Table 1). As amino acid 757 (chymotrypsinogen numbering 190) represents the main difference between the S1 pockets of these enzymes, one might assume an improvement in potency of 2 and 4. However, the affinities of compounds 2 and 4 decreased in the mutated variant relative to the wild-type enzyme. In agreement with these findings, similar compounds bearing the 2-aminomethyl-5-chlorobenzylamine moiety showed lower affinities for trypsin (with Ser 190) than thrombin (with Ala 190).[23] In MT2L785D variants, the inhibitory effects of compounds 3 and 4 decreased six- to 50-fold compared with those observed with wild-type MT2 (Table 1). In the wild-type enzyme, a favorable hydrophobic interaction between the cyclohexane ring and Leu 785 might account for the improved binding affinity, at least in the case of 3. In contrast, the MT2L785D model indicates that the cyclohexyl group is oriented towards the upper part of the S3/S4 pocket, the sulfonyl oxygen atoms are directed away from the introduced Asp 785, and the benzyl moiety is solvent exposed (Figure S4 in the Supporting Information). The simultaneous introduction of serine at position Ala 757 and aspartic acid at position Leu 785 caused similar results as discussed above for MT2A757S and MT2L785D (Table 1). The modeled complex of this variant and peptide 3 suggests an overall binding mode analogous to that of MT2L785D (Figure S4 in the Supporting Information). The additional replacement of Glu 712 for tyrosine in MT2A757S/L785D further decreased the affinity of all four compounds towards MT2E712Y/A757S/L785D (Table 1). As mentioned above, the single mutation at position 712 already resulted in a loss of affinity of the substrate and compounds 14, as was observed for the triple-mutated variant. We have explored the importance of several active site amino acids in the substrateenzyme and inhibitorenzyme interactions by mutational analysis of MT2 and demonstrated that a hydrophobic area in the S2 and the upper part of the S3/S4 pocket are important for a strong affinity of Boc-Gln-AlaArg-para-nitroanilide and compounds 14 for the enzyme. Only the mutation of His 665 to phenylalanine was sufficient to reduce the Km and Ki values of these peptides by four- to 12fold. In contrast, mutation of Glu 712 to tyrosine resulted in an overall decrease in affinities of the examined peptides, presumably because the small binding subsite above the S1 pocket is occupied by the side chain of Tyr 712. The introduction of a serine residue at position Ala 757 in the S1 pocket has only a marginal influence on the activities of the inhibitors. Leu 785 at the lower part of the S3/S4 pocket appears to be important for lipophilic interactions with the cyclohexyl group of compounds 3 and 4.
Our findings might elucidate slight differences in the active site architecture and ligand interactions of MT2 and matriptase, two closely related enzymes with different (patho)physiological importance. Moreover, this study may contribute to future improvement in potency and selectivity of MT2 inhibitors, accomplished by structural modifications such as those designed to further address remote binding sites. The resulting low-molecular weight compounds are promising candidates for use as pharmacological tools and potential therapeutic agents in the treatment of hemochromatosis.
Experimental details can be found in the Supporting Information.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft (SFB645). E.M. was supported by a fellowship award from Bayer HealthCare Pharmaceuticals. Keywords: matriptase-2 molecular modeling mutagenesis protease inhibitors type II transmembrane serine proteases
[1] G. Velasco, S. Cal, V. Quesada, L. M. Snchez, C. Lpez-Otn, J. Biol. Chem. 2002, 277, 37637 37646. [2] J. D. Hooper, L. Campagnolo, G. Goodarzi, T. N. Truong, H. Stuhlmann, J. P. Quigley, Biochem. J. 2003, 373, 689 702. [3] T. H. Bugge, T. M. Antalis, Q. Wu, J. Biol. Chem. 2009, 284, 23177 23181. [4] X. Du, E. She, T. Gelbart, J. Truksa, P. Lee, Y. Xia, K. Khovananth, S. Mudd, N. Mann, E. M. Moresco, E. Beutler, B. Beutler, Science 2008, 320, 1088 1092. [5] K. E. Finberg, M. M. Heeney, D. R. Campagna, Y. Aydinok, H. A. Pearson, K. R. Hartman, M. M. Mayo, S. M. Samuel, J. J. Strouse, K. Markianos, N. C. Andrews, M. D. Fleming, Nat. Genet. 2008, 40, 569 571. [6] A. R. Folgueras, F. M. de Lara, A. M. Pends, C. Garabaya, F. Rodrguez, A. Astudillo, T. Bernal. , R. Cabanillas, C. Lpez-Otn, G. Velasco, Blood 2008, 112, 2539 2545. [7] L. Silvestri, A. Pagani, A. Nai, I. de Domenico, J. Kaplan, C. Camaschella, Cell Metab. 2008, 8, 502 511. [8] S. Altamura, F. DAlessio, B. Selle, M. U. Muckenthaler, Biochem. J. 2010, 431, 363 371. [9] J. E. Maxson, J. Chen, C. A. Enns, A. S. Zhang, J. Biol. Chem. 2010, 285, 39021 39028. [10] M. T. Sisay, T. Steinmetzer, M. Stirnberg, E. Maurer, M. Hammami, J. Bajorath, M. Gtschow, J. Med. Chem. 2010, 53, 5523 5535. [11] Y. E. Shi, J. Torri, L. Yieh, A. Wellstein, M. E. Lippman, R. B. Dickson, Cancer Res. 1993, 53, 1409 1415. [12] C. Y. Lin, J. Anders, M. Johnson, Q. A. Sang, R. B. Dickson, J. Biol. Chem. 1999, 274, 18231 18236. [13] C. M. Overall, E. M. Tam, R. Kappelhoff, A. Connor, T. Ewart, C. J. Morrison, X. Puente, C. Lpez-Otn, A. Seth, Biol. Chem. 2004, 385, 493 504. [14] A. J. Sanders, C. Parr, G. Davies, T. A. Martin, J. Lane, M. D. Mason, W. G. Jiang, J. Exp. Ther. Oncol. 2006, 6, 39 48. [15] L. K. Vogel, M. Saeb, C. F. Skjelbred, K. Abell, E. D. Pedersen, U. Vogel, E. H. Kure, BMC Cancer 2006, 6, 176. [16] I. Enyedy, S.-L. Lee, A. H. Kuo, R. B. Dickson, C.-Y. Lin, S. Wang, J. Med. Chem. 2001, 44, 1349 1355. [17] P. Li, S. Jiang, S. L. Lee, C. Y. Lin, M. D. Johnson, R. B. Dickson, C. J. Michejda, P. P. Roller, J. Med. Chem. 2007, 50, 5976 5983. [18] T. Steinmetzer, A. Schweinitz, A. Strzebecher, D. Dnnecke, K. Uhland, O. Schuster, R. Friedrich, M. E. Than, W. Bode, J. Strzebecher, J. Med. Chem. 2006, 49, 4116 4126.
www.chemmedchem.org
71
MED
[19] R. Friedrich, P. Fuentes-Prior, E. Ong, G. Coombs, M. Hunter, R. Oehler, D. Pierson, R. Gonzalez, R. Huber, W. Bode, E. L. Madison, J. Biol. Chem. 2002, 277, 2160 2168. [20] M. Stirnberg, E. Maurer, A. Horstmeyer, S. Kolp, S. Frank, T. Bald, K. Arenz, A. Janzer, K. Prager, P. Wunderlich, J. Walter, M. Gtschow, Biochem. J. 2010, 430, 87 95. [21] S. L. Lee, R. B. Dickson, C. Y. Lin, J. Biol. Chem. 2000, 275, 36720 36725. [22] F. Bliveau, A. Dsilets, R. Leduc, FEBS J. 2009, 276, 2213 2226. [23] K. E. Rittle, J. C. Barrow, K. J. Cutrona, K. L. Glass, J. A. Krueger, L. C. Kuo, S. D. Lewis, B. J. Lucas, D. R. McMasters, M. M. Morrissette, P. G. Nantermet, C. L. Newton, W. M. Sanders, Y. Yan, J. P. Vacca, H. G. Selnick, Bioorg. Med. Chem. Lett. 2003, 13, 3477 3482.
Received: July 18, 2011 Published online on September 14, 2011
72
www.chemmedchem.org
DOI: 10.1002/cmdc.201100332
Identification and Characterization of Inhibitors of the Aminoglycoside Resistance Acetyltransferase Eis from Mycobacterium tuberculosis
Keith D. Green,[a] Wenjing Chen,[a, b] and Sylvie Garneau-Tsodikova*[a, b, c]
With an anticipated 9.8 million new cases this year,[1] the tuberculosis (TB) epidemic is one of the most serious health problems worldwide. The continuous emergence and global spread of multidrug-resistant (MDR) and extensively drug-resistant (XDR) strains of Mycobacterium tuberculosis (Mtb), the causative agent of TB, underscore the pressing clinical need for novel treatments of this deadly infectious disease and for new solutions to alleviate the resistance problem.[2, 3] Aminoglycoside (AG) antibiotics[4] such as kanamycin A (KAN) (1) and amikacin (AMK) (2) are currently used as a last resort for treatment of XDR-TB (Figure 1 A). However, resistance to KAN is constantly rising, and treatment options for patients affected with XDR-TB are becoming fewer.[5] In most bacterial strains, a major mechanism of resistance to AGs is the enzymatic modification of the drugs by AG-modifying enzymes such as AG acetyltransferases (AACs), AG phosphotransferases (APHs), and AG nucleotidyltransferases (ANTs).[6, 7] In Mtb, resistance to AGs results either from mutations of the ribosome that prevent the drugs from binding to it,[810] or from up-regulation of the chromosomal eis (enhanced intracellular survival) gene caused by mutations in its promoter.[11, 12] Other biological functions of the mycobacterial protein Eis have been the subject of numerous investigations.[1320] We recently demonstrated that Eis is a unique AAC that inactivates a broad set of AGs via a multi-acetylation mechanism.[21] Two main strategies to overcome the effect of Eis in Mtb can be envisioned: 1) the development of new AGs not susceptible to Eis and 2) the use of Eis inhibitors. We recently reported a chemoenzymatic methodology[22] and a complementary protecting-group-free chemical strategy[23] for the production of novel AG derivatives. However, as Eis is capable of multi-acetylation of a large variety of AG scaffolds, it is unlikely that novel AGs will provide a viable and/or sustainable solution to the resistance problem in Mtb. Blanchard and co-workers previously showed that, when used in conjunction, the b-lactamase inhibitor clavulanate and meropenem are effective against XDRTB.[24] The AG tobramycin and the macrolide antibiotic clari[a] Dr. K. D. Green, W. Chen, Dr. S. Garneau-Tsodikova Life Sciences Institute, University of Michigan 210 Washtenaw Ave., Ann Arbor, MI 48109 (USA) Fax: (+ 1) 734-615-5521 E-mail: sylviegt@umich.edu [b] W. Chen, Dr. S. Garneau-Tsodikova Chemical Biology Doctoral Program, University of Michigan 210 Washtenaw Ave., Ann Arbor, MI 48109 (USA) [c] Dr. S. Garneau-Tsodikova Department of Medicinal Chemistry College of Pharmacy, University of Michigan 210 Washtenaw Ave., Ann Arbor, MI 48109 (USA) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100332.
thromycin have also showed promising synergistic effects in Mtb clinical isolates.[25] Wright and colleagues also demonstrated that, in general, combinations of antibiotics and non-antibiotic drugs could result in enhancement of antimicrobial efficacy.[26] Similarly, an inhibitor of the resistance acetyltransferase Eis in combination with the currently used second-line antituberculosis drugs KAN or AMK may provide a potential solution to overcome the problem of XDR-TB. Herein, by using in vitro high-throughput screening (HTS), we identified and characterized the first series of potent inhibitors of Eis (Figure 1 B). To identify inhibitors of Mtb Eis, we used neomycin B (NEO) (3) owing to the robust activity of the enzyme with this AG. We screened a total of 23 000 compounds from three smallmolecule libraries: ChemDiv, BioFocus NCC, and MicroSource MS2000 spectrum. From the 23 000 molecules tested, 300 (1.3 %) showed a reasonable degree of inhibition (> 3 s from the mean negative control) against Eis, out of which 56 showed dose-dependent inhibition. The 25 compounds discussed herein (Figure 1 B) were found to have IC50 values in the low-micromolar range (Table 1 and Figure 2, as well as figures S1 and S2 in the Supporting Information). While most of these have not yet been biologically characterized, compounds 7, 14, 27, and 28 have found application as anti-HIV treatments (27[27, 28] and 28[2729]), molecules to prolong eukaryote longevity (7),[30] antibacterials (27 and 28),[31] anticancer agents (28),[32] and hypoglycemia therapeutics (14).[33] At first glance, the 25 identified compounds appear to have vastly different structures. However, upon closer inspection of their scaffolds, two structural features link these 25 Eis inhibitors: the presence of at least one aromatic ring and one amine functional group. In general, we observed that positively or potentially positively charged molecules, including chlorhexidine (6), displayed lower IC50 values than preferably negatively charged (27 and 28) or neutral compounds. The highly negatively charged AG-binding cavity of the Eis protein (PDB ID: 3R1K)[21] is consistent with this general trend. Seven of the 25 Eis inhibitors identified were divided into three groups for a preliminary and limited structureactivity relationship (SAR) analysis: 1) compounds 4 and 5, 2) 14, 15, and 16, as well as 3) 27 and 28 (Figure 1 B). Compounds 14, 15, and 16 differ in their imidazolium versus benzoimidazolium substitution on one side of the ketone and in their para substituents at the phenyl ring on the opposite side of the carbonyl. These differences had no effect on the IC50 values, indicating the importance of the imidazolium, but a secondary role of the additional features to the core structure for biological activity. In contrary, the differences in benzyl ring substitutions in compounds 27 and 28 (alternative placement (ortho versus meta) of the carboxylic acid and replacement of the para-chloro group with a para-hydroxy group) resulted in a
73
MED
Figure 1. A) Structures of AGs used in this study. B) Structures of the 25 inhibitors of Eis identified by high-throughput screening.
greater than fivefold increase in the inhibition of Eis. Similarly, replacement of the ethyl group of 5 adjacent to the cationic nitrogen with a phenyl moiety in 4 resulted in a 25-fold increase in the inhibitory activity of compound 4. Further kinetic analysis of compound 4 revealed a mixed mode of inhibition against NEO (Figure 2 B). The observed mixed mode of inhibition could be explained by the three substrates (NEO, acetylNEO, and diacetyl-NEO) that are produced during the reaction of NEO with Eis. Here, compound 4, may be competing differently with each possible substrate. Interestingly, in contrast to compound 4, the best inhibitor identified in this study with an IC50 value of 188 30 nm, chlo-
rhexidine (6), was found to behave as an AG-competitive inhibitor against NEO, KAN, and AMK (Figure 2 A). Chlorhexidine is an antibiotic used mainly as a topical antibacterial, as a mouthwash, and as a sterilizing agent for surgical equipment.[34] Because of its toxic effects on pulmonary tissues,[35] chlorhexidine cannot be pursued as a potential TB treatment, but will continue to serve as a positive control for future HTS experiments for the identification of additional Eis inhibitor scaffolds. With their structurally diverse scaffolds, the remaining compounds cannot be divided into distinct groups for SAR analyses. However, grouping the compounds by their IC50 values does reveal some trends. In comparison with compounds 4
74
www.chemmedchem.org
Table 1. Eis inhibition by hit compounds 428 for NEO acetylation. Compd[a] 4 4 4 5 6 6 6 7 8 9 10 11 12 13 14 IC50 [mm][b] 0.364 0.032 0.331 0.082 (AMK)[c] 0.585 0.113 (KAN)[c] 9.25 1.50 0.188 0.030 0.321 0.058 (AMK)[c] 0.666 0.193 (KAN)[c] 1.09 0.14 1.24 0.16 2.01 0.12 2.29 0.52 2.37 0.41 2.63 0.60 2.64 0.36 3.06 0.56 Compd[a] 15 16 17 18 19 20 21 22 23 24 25 26 27 28 IC50 [mm][b] 3.24 0.32 3.84 0.55 3.39 0.61 4.90 0.75 5.54 0.63 5.68 0.88 5.75 0.66 6.50 1.32 7.64 0.60 9.79 1.97 11.4 1.6 15.9 2.6 > 200 41 9
[a] See Figure 1 B for structures. [b] Determined from at least three trials; best-fit values were obtained by using KaleidaGraph 4.1. [c] IC50 values were also determined for compounds 4 and 6 using AMK and KAN (figure S2, Supporting Information).
and 68, the fewer hydrogen bonding sites of compounds 9 13 could explain the relatively higher IC50 values for these molecules. Likewise, the increased structural rigidity of compounds 1726 could limit the ability of these molecules to adopt an ideal conformation for binding, potentially explaining the higher inhibitory constants observed for these molecules. Because many AACs have a negatively charged AG-binding site that could be accessible for ligand binding,[3638] in order to confirm the specificity of the identified inhibitors for Eis, we tested whether the four best compounds (4, 6, 7, and 8) inhibit other AAC enzymes with negatively charged AG-binding sites from three different classes: AAC(2)-Ic from Mtb, AAC(3)IV from Escherichia coli, and AAC(6)/APH(2) from Staphylococcus aureus. With the exception of compound 4 against AAC(2)-Ic, which displayed an IC50 value of 367 129 mm (1000-fold worse than with Eis), no significant inhibition was observed for the combinations tested. This lack of cross-inhibition indicates that the inhibitors identified display high selectivity toward the Eis AG-binding site. Eis has been shown to multi-acetylate a large number of AGs[21] and is therefore po-
tentially able to accommodate various conformations of structurally diverse and/or similar molecules in contrast to the mono-acetylating AACs [AAC(2), AAC(3), and AAC(6)] for which substrates can only bind in a single conformation. The unique flexibility of the AG-binding site of Eis could therefore explain the intriguing selectivity of the inhibitors identified for this enzyme. For example, the selectivity of chlorhexidine (6) for Eis, normally non-selectively binding to negatively charged sites and therefore expected to inhibit AAC(2), AAC(3), and AAC(6), could be justified by the uniqueness of the Eis AGbinding site that could accommodate compound 6 in conformation(s) that the other AACs cannot. In summary, by using an in vitro HTS UV/Vis assay, we have identified 25 inhibitors of Eis from Mtb with 21 distinct scaffolds. The compounds display selective and potent inhibitory activity in vitro against the purified Mtb Eis and different modes of inhibition, with the known antibacterial chlorhexidine (6) competing with the AG for binding Eis. These findings provide the foundation for testing whether the Eis inhibitors will overcome KAN resistance in Mtb strains in which Eis is upregulated. This work also lays the groundwork for exploration of scaffold diversification and SAR studies of the identified biologically active compounds to be used in combination therapies with KAN or AMK against TB.
Reagents and small-molecule libraries: All reagents including DTNB, NEO, KAN, AMK, and acetyl-CoA were purchased from SigmaAldrich (St. Louis, MO, USA). Eis was screened against 23 000 compounds from three diverse libraries of small molecules: 1) the BioFocus NCC library, 2) the ChemDiv library (20 000 compounds), and 3) the MicroSource MS2000 library, composed of ~ 2000 bioactive compounds (343 molecules with reported biological activities, 629 natural products, 958 known therapeutics, and 70 compounds approved for agricultural use). The activity of promising compounds was confirmed by using repurchased samples from Sigma Aldrich (compound 6) and ChemDiv (San Diego, CA, USA) (compounds 4, 5, and 728). Expression and purification of Eis and other AAC proteins: The Eis and AAC(2)-Ic from Mtb,[21] as well as the AAC(3)-IV from
Figure 2. Representative examples of IC50 curves for A) chlorhexidine (6) and B) compound 4. Plots showing the competitive and mixed inhibition with respect to NEO for compounds 6 and 4, respectively, can be viewed as the inset in each panel.
www.chemmedchem.org
75
MED
E. coli[22, 39] and AAC(6)/APH(2)-Ia from S. aureus[22, 40] were overexpressed and purified as previously described. Eis chemical library screening: The inhibition of Eis activity was determined by a UV/Vis assay monitoring the increase in absorbance at 412 nm (e412 = 13 600 m1 cm1) resulting from the reaction of DTNB with the CoASH released upon acetylation of NEO. The final reaction mixtures (40 mL) contained Eis (0.25 mm), NEO (100 mm), Tris-HCl (50 mm, pH 8.0 adjusted at RT), AcCoA (40 mm), DTNB (0.5 mm), and the potential inhibitors (20 mm). Positive and negative control experiments were performed using chlorhexidine (6) (5 mm) and DMSO (0.5 % v/v), respectively, instead of the potential inhibitors. Briefly, a mixture (30 mL) containing Eis (0.33 mm) and NEO (133.33 mm) in Tris-HCl (50 mm, pH 8.0 adjusted at RT) was added to 384-well non-binding-surface plates (Thermo Fisher Scientific, Waltham, MA, USA) using a Multidrop dispenser (Thermo Fisher Scientific). The potential inhibitors (0.2 mL of a 4 mm stock), chlorhexidine (6) (0.2 mL of a 1 mm stock), or DMSO (0.2 mL) were then added to each well by Biomek HDR (Beckman, Fullerton, CA, USA). After 10 min at RT, reactions were initiated by the addition of a mixture (10 mL) containing AcCoA (160 mm), DTNB (2 mm), and Tris-HCl (50 mm, pH 8.0 adjusted at RT). After an additional 5 min of incubation at RT, the absorbance was measured at 412 nm using a PHERAstar plate reader (BMG Labtech, Cary, NC, USA). The average Z score for the entire HTS assay was 0.65. Hit validation: Using the above conditions, all compounds deemed a hit (> 3 s as a statistical hit threshold from the mean negative control) were tested in triplicate. Compounds that displayed inhibition at least in two of the three independent assays were then tested for a doseresponse using twofold dilutions from 20 mm to 78 nm. IC50 values were determined for all compounds displaying dose-dependent activity. Inhibition kinetics: IC50 values were determined on a multimode SpectraMax M5 plate reader using 96-well plates (Thermo Fisher Scientific) by monitoring absorbance at 412 nm taking measurements every 30 s for 20 min. Eis inhibitors were dissolved in TrisHCl (50 mm, pH 8.0 adjusted at RT containing 10 % v/v DMSO) (100 mL) and a two- or fivefold dilution was performed. To the solution of inhibitors, a mixture (50 mL) containing Eis (1 mm), NEO (400 mm), and Tris-HCl (50 mm, pH 8.0 adjusted at RT) was added. After 10 min, the reactions were initiated by addition of a mixture (50 mL) containing AcCoA (2 mm), DTNB (2 mm), and Tris-HCl (50 mm, pH 8.0 adjusted at RT). Overall, inhibitor concentrations ranged from 200 mm to 4 pm. Initial rates (first 25 min of reaction) were calculated and normalized to reactions containing DMSO only. All assays were performed at least in triplicate. IC50 values were calculated by using a Hill plot fit with KaleidaGraph 4.1 software. Two representative examples of IC50 curves are provided in Figure 2, while the other 23 IC50 curves are presented in figure S1 (Supporting Information). Determination of IC50 values of compounds 4 and 6 against AMK and KAN were also performed as described above (figure S2, Supporting Information). All IC50 values are listed in Table 1. Mode of inhibition: By using the conditions described for inhibition kinetics with varying concentrations of NEO (50, 75, 100, 125, 150, and 200 mm) and compounds 4 (1, 0.5, 0.25, and 0.125 mm) or 6 (5, 10, 20, and 40 nm), mixed inhibition was determined for compound 4, and compound 6 was found to be a competitive inhibitor of NEO. Resulting reaction rates were plotted as Lineweaver Burk plots (Figure 2 inserts of panels A and B). Using the same assay conditions, chlorhexidine (6) was also found to be a competitive inhibitor of KAN and AMK. Inhibitor selectivity for Eis: To determine if the identified inhibitors are selective for Eis, we tested the four best Eis inhibitors (4, 6, 7, and 8) with three other AACs: AAC(2)-Ic, AAC(3)-IV, and AAC(6)/ APH(2)-Ia. The conditions described for inhibition kinetics were used with varying concentrations of compounds 4, 6, 7, or 8 (200 0.2 mm, 10-fold serial dilution) and AAC(2)-Ic (0.125 mm), AAC(3)-IV (0.25 mm), or AAC(6)/APH(2)-Ia (0.25 mm) instead of Eis. For AAC(2)-Ic with compound 4, the concentration of inhibitor ranged from 1 mm to 500 pm and a fivefold serial dilution was used.
Acknowledgements
This work was supported by the Life Sciences Institute and the College of Pharmacy at the University of Michigan (S.G.-T.), by a grant from the Firland Foundation (S.G.-T.), by a Center for Chemical Genomics (CCG) Pilot Grant at the University of Michigan (S.G.-T.), and by a National Institutes of Health (NIH) Grant AI090048 (S.G.-T.). We thank Martha J. Larsen, Steve van de Roest, and Thomas J. McQuade (CCG, University of Michigan) for their help with HTS. We thank Oleg V. Tsodikov for critical reading of the manuscript. Keywords: aminoglycoside acetyltransferase antibiotics bacterial resistance Eis protein high-throughput screening inhibitors
[1] C. Dye, B. G. Williams, Science 2010, 328, 856. [2] A. Koul, E. Arnoult, N. Lounis, J. Guillemont, K. Andries, Nature 2011, 469, 483. [3] C. E. Barry III, J. S. Blanchard, Curr. Opin. Chem. Biol. 2010, 14, 456. [4] J. L. Houghton, K. D. Green, W. Chen, S. Garneau-Tsodikova, ChemBioChem 2010, 11, 880. [5] W. W. Yew, C. Lange, C. C. Leung, Eur. Respir. J. 2010, 37, 441. [6] K. D. Green, W. Chen, S. Garneau-Tsodikova, Antimicrob. Agents Chemother. 2011, 55, 3207. [7] M. S. Ramirez, M. E. Tolmasky, Drug Resist. Updates 2010, 13, 151. [8] D. Shcherbakov, R. Akbergenov, T. Matt, P. Sander, D. I. Andersson, E. C. Bttger, Mol. Microbiol. 2010, 77, 830. [9] L. Jugheli, N. Bzekalava, P. de Rijk, K. Fissette, F. Portaels, L. Rigouts, Antimicrob. Agents Chemother. 2009, 53, 5064. [10] Y. Zhang, W. W. Yew, Int. J. Tuberc. Lung Dis. 2009, 13, 1320. [11] P. J. Campbell, G. P. Morlock, R. D. Sikes, T. L. Dalton, B. Metchock, A. M. Starks, D. P. Hooks, L. S. Cowan, B. B. Plikaytis, J. E. Posey, Antimicrob. Agents Chemother. 2011, 55, 2032. [12] M. A. Zaunbrecher, R. D. Sikes, Jr., B. Metchock, T. M. Shinnick, J. E. Posey, Proc. Natl. Acad. Sci. USA 2009, 106, 20 004. [13] D. M. Shin, B. Y. Jeon, H. M. Lee, H. S. Jin, J. M. Yuk, C. H. Song, S. H. Lee, Z. W. Lee, S. N. Cho, J. M. Kim, R. L. Friedman, E. K. Jo, PLoS Pathog. 2010, 6, e1001230. [14] J. L. Dahl, J. Wei, J. W. Moulder, S. Laal, R. L. Friedman, Infect. Immun. 2001, 69, 4295. [15] J. L. Dahl, K. Arora, H. I. Boshoff, D. C. Whiteford, S. A. Pacheco, O. J. Walsh, D. Lau-Bonilla, W. B. Davis, A. G. Garza, J. Bacteriol. 2005, 187, 2439. [16] S. Wu, P. F. Barnes, B. Samten, X. Pang, S. Rodrigue, S. Ghanny, P. Soteropoulos, L. Gaudreau, S. T. Howard, Microbiology 2009, 155, 1272. [17] J. Wei, J. L. Dahl, J. W. Moulder, E. A. Roberts, P. OGaora, D. B. Young, R. L. Friedman, J. Bacteriol. 2000, 182, 377. [18] R. K. Lella, C. Sharma, J. Biol. Chem. 2007, 282, 18671. [19] E. A. Roberts, A. Clark, S. McBeth, R. L. Friedman, J. Bacteriol. 2004, 186, 5410. [20] L. P. Samuel, C. H. Song, J. Wei, E. A. Roberts, J. L. Dahl, C. E. Barry III, E. K. Jo, R. L. Friedman, Microbiology 2007, 153, 529.
76
www.chemmedchem.org
[21] W. Chen, T. Biswas, V. R. Porter, O. V. Tsodikov, S. Garneau-Tsodikova, Proc. Natl. Acad. Sci. USA 2011, 108, 9804. [22] K. D. Green, W. Chen, J. L. Houghton, M. Fridman, S. Garneau-Tsodikova, ChemBioChem 2010, 11, 119. [23] P. Shaul, K. D. Green, R. Rutenberg, M. Kramer, Y. Berkov-Zrihen, E. Breiner-Goldstein, S. Garneau-Tsodikova, M. Fridman, Org. Biomol. Chem. 2011, 9, 4057. [24] J. E. Hugonnet, L. W. Tremblay, H. I. Boshoff, C. E. Barry III, J. S. Blanchard, Science 2009, 323, 1215. [25] K. Stoffels, H. Traore, F. Vanderbist, M. Fauville-Dufaux, Int. J. Tuberc Lung Dis. 2009, 13, 1041. [26] L. Ejim, M. A. Farha, S. B. Falconer, J. Wildenhain, B. K. Coombes, M. Tyers, E. D. Brown, G. D. Wright, Nat. Chem. Biol. 2011, 7, 348. [27] K. Liu, H. Lu, L. Hou, Z. Qi, C. Teixeira, F. Barbault, B. T. Fan, S. Liu, S. Jiang, L. Xie, J. Med. Chem. 2008, 51, 7843. [28] C. Teixeira, F. Barbault, J. Rebehmed, K. Liu, L. Xie, H. Lu, S. Jiang, B. Fan, F. Maurel, Bioorg. Med. Chem. 2008, 16, 3039. [29] Y. Wang, H. Lu, Q. Zhu, S. Jiang, Y. Liao, Bioorg. Med. Chem. Lett. 2010, 20, 189. [30] D. S. Goldfarb, University of Rochester, NY (USA), 2009, US20090163545, 57 pp.
[31] I. A. Schepetkin, A. I. Khlebnikov, L. N. Kirpotina, M. T. Quinn, J. Med. Chem. 2006, 49, 5232. [32] H. Qin, J. Shi, R. Noberini, E. B. Pasquale, J. Song, J. Biol. Chem. 2008, 283, 29473. [33] S. J. Dominianni, T. T. Yen, J. Med. Chem. 1989, 32, 2301. [34] M. E. Saravia, P. Nelson-Filho, I. Y. Ito, L. A. da Silva, R. A. da Silva, C. G. Emilson, Microbiol. Res. 2010, 166, 63. [35] Y. Xue, S. Zhang, Y. Yang, M. Lu, Y. Wang, T. Zhang, M. Tang, H. Takeshita, Hum. Exp. Toxicol. 2011, DOI: 10.1177/0960327111400104. [36] M. W. Vetting, S. S. Hegde, F. Javid-Majd, J. S. Blanchard, S. L. Roderick, Nat. Struct. Biol. 2002, 9, 653. [37] E. Wolf, A. Vassilev, Y. Makino, A. Sali, Y. Nakatani, S. K. Burley, Cell 1998, 94, 439. [38] M. W. Vetting, C. H. Park, S. S. Hegde, G. A. Jacoby, D. C. Hooper, J. S. Blanchard, Biochemistry 2008, 47, 9825. [39] M. L. Magalhaes, J. S. Blanchard, Biochemistry 2005, 44, 16275. [40] D. D. Boehr, D. M. Daigle, G. D. Wright, Biochemistry 2004, 43, 9846. Received: July 1, 2011 Revised: August 18, 2011 Published online on September 5, 2011
www.chemmedchem.org
77
www
Online vacancies worldwide in physics, chemistry, chemical engineering, construction engineering, materials science and life sciences.
scitec-career.
the ultimate global JobMachine for scientists and engineers.
com
DOI: 10.1002/cmdc.201100457
A Novel Aptamer Developed for Breast Cancer Cell Internalization

Kejing Zhang,[a, b] Kwame Sefah,[a] Lili Tang,[b] Zilong Zhao,[a, c] Guizhi Zhu,[a] Mao Ye,[c] Weijia Sun,*[b] Steve Goodison,[d] and Weihong Tan*[a, c]
Breast cancer affects one in eight women in the United States, with a mortality rate that is second only to lung cancer. Although chemotherapy is widely used in breast cancer treatment, its side effects remain a challenge. One way to address this problem is through drug delivery by the internalization of cell-type-specific probes. Although nucleic acid aptamers are excellent probes for molecular recognition, only a few studies have demonstrated that aptamers can be internalized into living cells. Therefore, herein we report the development of a cancer-cell-specific DNA aptamer probe, KMF2-1a. By using the cell-SELEX method, this aptamer was selected against breast cancer cell line MCF-10AT1. Our results show that KMF2-1a is internalized efficiently and specifically to the endosome of target breast cancer cells. These results indicate that KMF2-1a is a promising agent for cell-type-specific intracellular delivery with both diagnostic and therapeutic implications.
Introduction
Breast cancer is by far the most frequently diagnosed cancer among women, with an estimated 1.38 million new cancer cases in 2008 (23 % of all cancers) and 458 503 deaths worldwide (13.7 % of cancer mortality in women). It is now the most common cancer, with an estimated 690 000 new cases in both developed and developing countries.[1] Among the newest tools in the fight against cancer, aptamers have gained increasing research interest. Aptamers are single-stranded oligonucleotides or peptides that bind to specific targets with high affinity and specificity. They are derived from an iterative process called systematic evolution of ligands by exponential enrichment (SELEX).[2, 3] Compared with the antibody approach, aptamers have high affinity, excellent specificity, and low toxicity. Moreover, they are stable and easy to synthesize, modify, and manipulate.[47] Although aptamers have been very effective in vitro, most aptamers cannot be directly taken up by cells without external assistance.[8] Up to now, successful cellular internalization has been reported for only a few aptamers, including anti-PSMA, which targets PSMA in prostate cancer cells, and Sgc8, which targets PTK7 in acute leukemia cells.[911] Clearly, internalization ability is very important for the application of aptamers in vivo, especially in targeted drug delivery. A key challenge confronting aptamer internalization is the limited number of aptamer candidates available for cell-specific recognition. Our research group has developed the cell-SELEX method to generate a panel of aptamers against several types of cancer cells, including lymphocytic leukemia, myeloid leukemia, liver cancer, small-cell lung cancer, ovarian cancer, and colon cancer.[1215] These aptamers were selected by using whole intact cells as targets, while a related cell line was used as a negative control. This method enables the generation of multiple aptamers for molecular recognition of the target cells, and these aptamers could be used as excellent molecular recognition probes. We recently used the cell-SELEX method to generate molecular probes for specific recognition of breast cancer cells. Two cell lines were used: the breast cancer cell line, MCF-10AT1, as target cells, and a non-cancer cell line, MCF-10TA1, as negative control.[16, 17] We identified a panel of aptamers (KMF2-1a, KMF3, and KMF9b) that specifically recognize the cancer cell line MCF-10AT1 with Kd values in the nanomolar range. One of these aptamers, KMF2-1a, binds specifically to target cells, and its binding capacity was shown to be very stable in our study. Furthermore, this aptamer can bind to target cells at both 4 and 37 8C, extending its use for in vivo experimentation. Herein we present the results of an internalization study of dye-labeled KMF2-1a by flow cytometry and confocal imaging. The results show that this aptamer can be specifically internalized by target breast cancer cells. Thus, it is anticipated that
[a] K. Zhang,+ Dr. K. Sefah,+ Dr. Z. Zhao, G. Zhu, Prof. W. Tan Departments of Chemistry, and Department of Physiology and Functional Genomics, Shands Cancer Center, UF Genetics Institute University of Florida, Gainesville, FL 32611-7200 (USA) E-mail: tan@chem.ufl.edu [b] K. Zhang,+ Prof. L. Tang, Prof. W. Sun Xiangya Hospital, Central South University P.O. Box 190, Changsha, Hunan 410008 (P.R. China) E-mail: sunweijia2009@126.com [c] Dr. Z. Zhao, M. Ye, Prof. W. Tan State Key Laboratory for Chemo/Biosensing and Chemometrics College of Chemistry and Chemical Engineering Hunan University, Changsha 410082 (P.R. China) [d] Prof. S. Goodison CRI, MD Anderson Cancer Center Orlando 6900 Lake Nona Blvd., Orlando, FL, 32827 (USA) [+] These authors contributed equally to this work.
79
MED
aptamer KMF2-1a will be a suitable candidate for further study in targeted drug delivery, both in vitro and in vivo.
W. Sun, W. Tan, et al. Therefore, aptamer KMF2-1a was incubated with the target cell line at 37 8C, and it showed binding ability similar to that observed at 4 8C (Figure 1 b), suggesting that KMF2-1a could be suitable for in vivo studies. Importantly, the negative control MCF-10TA1 cells did not bind KMF2-1a at 4 8C, indicating that KMF2-1a binding is indeed target-cell specific (Figure 1 c). To exclude interference signals from outside the cell, confirmation of aptamer internalization requires removal of aptamers bound to the cell surface. In previous studies with aptamer Sgc8, our research group found that trypsin treatment is an easy method for this.[18] Therefore, before flow cytometry and prior to the internalization study, trypsin was used to remove cell-surface-bound aptamers. We found that trypsin treatment could also successfully eliminate specific binding of KMF2-1a to target cells. Thus, at 4 8C, the fluorescence signal resulting from the binding of KMF2-1abiotin to the surfaces of MCF-
Results and Discussion

We first tested the binding capacity of aptamer KMF2-1a under various time and temperature conditions. After incubation with KMF2-1a for 30 min at 4 8C, MCF-10AT1 cells were washed twice with washing buffer to remove free aptamer, and were then analyzed by flow cytometry (Figure 1 a). After further incubation at 4 8C for 2 h in binding buffer without aptamer, flow cytometry confirmed identical results, indicating that the time elapsed after removal of the aptamer does not affect the binding ability. It is well known that the secondary structure of DNA aptamers can change under varying temperatures, explaining why some aptamers lose their binding ability at 37 8C.
Figure 1. Flow cytometry results of aptamer binding with MCF-10AT1 cells under various conditions; all aptamers were labeled with biotin. a) KMF2-1a shows stable binding to target cells at 4 8C: line 1, MCF-10AT1 cells; line 2, library aptamers (Lib) as negative control; line 3, 30 min after incubation of KMF2-1a with target cells at 4 8C; line 4, 2 h after incubation of KMF2-1a with target cells at 4 8C. b) KMF2-1a can still bind target cells at 37 8C: line 1, MCF-10AT1 cells; line 2, Lib as negative control; line 3, 30 min after incubation of KMF2-1a with target cells at 37 8C. c) Negative control MCF-10TA1 cells cannot bind KMF2-1a: line 1, MCF-10TA1 cells; line 2, Lib aptamers; line 3, MCF-10TA1 cells incubated with KMF2-1a at 4 8C for 30 min. d) Trypsin removes specific binding of KMF21a: line 1, MCF-10AT1 cells; line 2, Lib (negative control) incubated with target cells without trypsin treatment at 4 8C; line 3, KMF2-1a incubated with target cells without trypsin treatment at 4 8C; line 4, KMF2-1a incubation with target cells after trypsin treatment at 4 8C.
80
www.chemmedchem.org
An Aptamer Targeting Breast Cancer 10AT1 cells was entirely eliminated after trypsin treatment (Figure 1 d). Based on this finding, two conclusions can be drawn: first, the removal of specific binding between KMF2-1a and target cell surfaces by trypsin treatment indicates that the molecular target of this aptamer may be membrane protein(s); second, once surface-bound aptamers have been removed, the remaining fluorescence signal should arise only from aptamers that have been taken up (internalized) by the target cell. The internalization assay is shown schematically in Figure 2 a. MCF-10AT1 is an adherent cell line. Before starting the internalization study, cells were split into a six-well plate and cultured for another 24 h to ensure that cell membrane proteins had recovered to normal levels. The aptamer internalization study was carried out at 4 and 37 8C, and the flow cytometry results confirm aptamer internalization. Although the cells had been treated with trypsin after aptamer application at 37 8C, we could still detect the shift in signal from the 37 8C group, but not the 4 8C group, after incubation for 2 h with aptamer KMF2-1a. Because trypsin had removed all cell-surface-bound aptamers, the signal detected from the 37 8C group must have originated from aptamers inside the cell. Similar to KMF2-1a internalized by MCF-10AT1 assay, we used MCF-10TA1 cells as a negative control, and no signal was detected from the 37 8C group. Thus, we conclude that KMF2-1a is internalized specifically by the target MCF-10AT1 cells. We used confocal microscopy to further confirm aptamer internalization. As shown in Figure 3, after 2 h incubation with KMF2-1aTAMRA, we observed fluorescence signals from inside the MCF-10AT1 cells. On the other hand, no fluorescence was observed inside the cells incubated with TAMRA-conjugated negative control library aptamers (Lib), and no fluorescence shift was observed from Libbiotin in flow cytometry under the same conditions as KMF2-1aTAMRA/KMF2-1abiotin. These results provide strong evidence that KMF2-1a is internalized by MCF-10AT1 cells. We then studied the time dependence of internalization by flow cytometry. The intensity at each time interval was normalized to the fluorescence signal from the cell surface at 4 8C. As shown in Figure 2 c, cell internalization of KMF2-1a constantly increased over the course of the 90 min monitoring period. In our internalization study, streptavidincy5.5 was used to monitor aptamer behavior. Streptavidin is a 52.8 kDa tetrameric protein purified from the bacterium Streptomyces avidinii, and it binds biotin with extraordinarily high affinity.[19] Our results show that the entire complex of streptavidin and biotin-labeled KMF2-1a could be delivered into target breast cancer cells. This advantage could be further used for the targeted delivery of large molecules. Therefore, KMF2-1a may have the potential for aptamer-based targeted drug delivery in the combined therapy and diagnosis (theragnosis) of breast cancer. One aptamer that has already been used successfully for targeted delivery is the anti-PSMA aptamer, which specifically binds to prostate cancer cells. This aptamer has been used to deliver chemotherapeutic and other cytotoxic agents, including gelonin, a ribosomal toxin that causes cell death by disrupting protein synthesis,[20] and a quantum dotdoxorubicin (Dox) conjugate.[21] Dox is the most used drug in breast cancer chemotherapy, but it can cause many side effects. To overcome this liability, we are interested in exploring the conjugation of KMF2-1a with Dox for breast cancer targeted chemotherapy. In this study, we were also very interested in the fate of the KMF2-1a aptamer once inside the target cells. By using Alexa633-labeled transferrin as a tracker, we showed previously that aptamer Sgc8 is taken up by endosomes in cells.[11] We also used this method to investigate the fate of the KMF2-1a aptamer inside breast cancer cells. It has been reported that transferrin can be internalized via receptor-mediated endocytosis, and Alexa633-labeled transferrin is commonly used to localize in the endosome.[22] Thus, co-
Figure 2. Internalization study showing that KMF2-1a can be internalized specifically into target MCF-10AT1 cells. a) Schematic of internalization assay: MCF-10AT1 cells were split into a six-well plate. After binding with KMF2-1a biotin at 4 8C, streptavidincy5.5 dye was added. After washing away all the free aptamer and dye and adding medium, one group was incubated at 4 8C for 2 h and another group was incubated at 37 8C for 2 h. Trypsin was used to detach and treat cells. b) Internalization of KMF2-1a aptamers: lines 1, MCF-10AT1 cells; lines 2, Lib aptamers (negative control) were not internalized; lines 3, trypsin was applied to remove the membrane-bound aptamers, and a fluorescence shift was present with KMF2-1a-incubated cells at 37 8C, but not at 4 8C. However, KMF2-1a could not bind to negative control MCF10TA1 cells, and was not internalized. c) Quantitative analysis of KMF2-1a internalization over time: fluorescence intensities of cells that had internalized aptamers were normalized to those of cells at 4 8C. Flow cytometry data showed increasing fluorescence intensity inside target cells, indicating increasing internalization of aptamers.
www.chemmedchem.org
81
MED
W. Sun, W. Tan, et al.
Figure 3. Internalization study of KMF2-1aTAMRA into MCF-10AT1 cells by confocal microscopy (left: fluorescence signal; right: overlay of transmission and fluorescence signals). a) Fluorescence signals were observed from inside the cells after incubation of KMF2-1aTAMRA with cells for 2 h at 37 8C. b) Lib TAMRA could not be internalized into cells at 37 8C, even after incubation for 2 h.
localization of a given aptamerin this case KMF2-1awith transferrin would indicate that most internalized aptamers are taken up by endocytosis (Figure 4). In the context of drug delivery, the endosomal compartment is usually acidic, and many internalized extracellular carrier molecules alter their structures at low pH to release the bound substance. As such, internalized aptamers could potentially respond to these conditions to release their drug payloads. For example, one such acid-labile linkage was used by Yufen et al. for the Sgc8cDox conjugate. This aptamerdrug conjugate was cleaved inside the acidic endosomal environment, resulting in Dox release.[23]
line MCF-10AT1 by cell-SELEX, is specifically internalized into target cells without external assistance and delivered to the endosome. This aptamer was also shown to deliver large molecules (streptavidin) into cells. As a molecular recognition probe, KMF2-1a can be conjugated with other agents of interest, such as chemotherapeutic drugs or siRNA, for specific intracellular delivery. This aptamer is anticipated to be widely used in breast cancer research.
Cell lines and regents: The MCF-10AT1 and MCF-10A1 cell lines were maintained in DMEM-F12 medium (ATCC) supplemented with 10 mg mL1 insulin, 0.5 mg mL1 hydrocortisone, 20 ng mL1 epidermal growth factor (EGF), 100 ng mL1 cholera enterotoxin, 5 % horse serum, and 20 U mL1 penicillinstreptomycin. Cells were
Conclusions
In conclusion, we have shown that a cell-specific aptamer, KMF2-1a, which was generated against the breast cancer cell
82
www.chemmedchem.org
An Aptamer Targeting Breast Cancer
Figure 4. Co-localization of KMF2-1a and transferrin in endosomes: a) KMF2-1aTAMRA, b) transferrinAlexa633, and c) overlay of the fluorescence channels and bright-field channel.
washed after harvesting by using 1 non-enzymatic cell dissociation solution (Sigma) with the washing buffer [4.5 g L1 glucose and 5 mm MgCl2 in Dulbeccos phosphate-buffered saline with CaCl2 and MgCl2 (Sigma)]. The binding buffer used for aptamer selection was prepared by adding yeast tRNA (0.1 mg mL1; Sigma) and BSA (1 mg mL1; Fisher) to the washing buffer to minimize nonspecific binding. DNA synthesis and labeling: All DNA synthesis reagents were from Glen Research (Sterling, VA, USA). KMF2-1a: 5-AGG CGG CAG TGT CAG AGT GAA TAG GGG ATG TAC AGG TCT GCA CCC ACT CGA GGA GTG ACT GAG CGA CGA AGA CCC C-3. Library aptamers (Lib): 5-ATG AGA GCG TCG GTG TGG TA-N40-TAC TTC CGC ACC CTC CTA CA-3. All probes and DNAs were synthesized with an ABI3400 DNA/RNA synthesizer (Applied Biosystems, Foster City, CA, USA). TAMRA, biotin, and CPG phosphoramidites were used for modification on aptamer KMF2-1a in the DNA synthesizer. HPLC purification was performed on a ProStar HPLC station equipped with fluorescence and a photodiode array detector. A C18 reversed-phase column was used with CH3CN-TEAA as the mobile phase. Trypsin treatment of cells: Cells were first washed with washing buffer (500 mL) to remove the FBS in the binding buffer (which can quench the function of trypsin) and then incubated with trypsin (Fisher Biotech, 500 mL, 0.05 %)/EDTA (0.53 mm) in HBSS at 37 8C for 15 min. After incubation, 50 mL FBS was added, the cells were washed with washing buffer (500 mL) once again, and suspended in binding buffer (200 mL with 0.1 % NaN3) for detection. FACS-based internalization assay: Flow cytometry analysis was performed on a FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA) by counting 20 000 events, and the data were analyzed with WinMDI software. For the KMF2-1abiotin binding experiment, cells were detached by using non-enzymatic cell dissociation solution (Sigma). Cells were first washed with washing buffer (2 500 mL) and then suspended in binding buffer (200 mL, with 0.1 % NaN3). Next, aptamer conjugates KMF2-1a biotin and Libbiotin (200 nm) were added for 30 min incubation at 4 8C and 37 8C. After incubation, cells were washed with washing buffer (2 500 mL); the streptavidincy5.5 dye was then added in binding buffer (200 mL) for another 15 min incubation at 4 8C. Before the internalization study was carried out, 1 106 cells were split into a six-well plate for 24 h cell culture to ensure that cell membrane proteins recovered to normal levels. Cells were separated into two groups for experiments at 4 and 37 8C. After binding both groups of cells with KMF2-1a at 4 8C, free aptamers and dye were washed away, and then fresh medium was added. The 4 8C
cell group was incubated at 4 8C for 2 h in medium, and the 37 8C cell group was incubated for 10, 20, 30, 60, 90, and 120 min in medium. Internalization was stopped by placing cells on ice. In accordance with the trypsin (Fisher Biotech) treatment method described above, the two cell groups were analyzed by flow cytometry. All experiments were repeated three times. Confocal imaging: All cellular fluorescence images were collected on an FV500-IX81 confocal microscope (Olympus America Inc., Melville, NY, USA). Excitation wavelengths and emission filters were as follows: TAMRA, 543 nm laser line excitation, BP580 20 nm emission filter; Alexa633, 633 nm laser line excitation, LP650 emission filter. Cells were split into poly-d-lysine-coated 35 mm glassbottom dishes (Mat Tek Corp.) for 24 h cell culture. Cells were washed with washing buffer and then incubated with KMF2-1a TAMRA (200 nm) in medium at 37 8C for 2 h. For the co-localization experiment, transferrinAlexa633 (60 nm) was added 30 min before the termination of incubation. Experiments were repeated three times and analyzed by using Fluoview analysis software.
Abbreviations
ATCC: American Type Culture Collection, BSA: bovine serum albumin, DMEM: Dulbeccos modified Eagles medium, FBS: fetal bovine serum, HBSS: Hanks buffered salt solution, PSMA: prostatespecific membrane antigen, PTK7: protein tyrosine kinase 7, SELEX: systematic evolution of ligands by exponential enrichment, TAMRA: tetramethy1-6-carboxyrhodamine.
Authors contributions
K.Z. performed laboratory experiments, data analysis, and participated in the experimental design and drafting of the manuscript. K.S., Z.Z., and G.Z. participated in flow cytometry, confocal imaging, experiments and interpretation of the results. L.T. contributed to the interpretation of the data and critical revision of the manuscript. W.S., S.G., and W.T. participated in the design and coordination of the research, performed the statistical analysis, and helped to draft the manuscript. All authors read and approved the final manuscript.
www.chemmedchem.org
83
MED
Acknowledgements
Prof. Zhi Zhu, Da Han, Lu Peng, and Mingxu You (Department of Chemistry, University of Florida) are acknowledged for technical assistance. This work was supported by the National Key Scientific Program of China (2011CB911001, 2011CB911003) and by grants awarded by the National Institutes of Health (GM066137, GM079359, and CA133086). We also acknowledge the Interdisciplinary Center for Biotechnology Research (ICBR) at the University of Florida. Keywords: aptamers breast cancer drug delivery internalization MCF-10AT1
[1] GLOBOCAN 2008 Cancer Fact Sheet, Breast Cancer Incidence and Mortality Worldwide in 2008 Summary: [http://globocan.iarc.fr/factsheets/ cancers/breast.asp]. [2] A. D. Ellington, J. W. Szostak, Nature 1990, 346, 818 822. [3] C. Tuerk, L. Gold, Science 1990, 249, 505 510. [4] C. J. Yang, S. Jockusch, M. Vicens, N. J. Turro, W. Tan, Proc. Natl. Acad. Sci. USA 2005, 102, 17278 17283. [5] R. Nutiu, Y. Li, Angew. Chem. 2005, 117, 1085 1089; Angew. Chem. Int. Ed. 2005, 44, 1061 1065. [6] J. Liu, Y. Lu, Angew Chem. 2005, 118, 96 100; Angew Chem. Int. Ed. 2005, 45, 90 94. [7] D. Shangguan, Z. Tang, P. Mallikaratchy, Z. Xiao, W. Tan, ChemBioChem 2007, 8, 603 606. [8] M. Rimmele, ChemBioChem 2003, 4, 963 971.
W. Sun, W. Tan, et al.

[9] H. Liu, A. K. Rajasekaran, P. Moy, Y. Xia, S. Kim, V. Navarro, R. Rahmati, N. H. Bander, Cancer Res. 1998, 58, 4055 4060. [10] S. E. Lupold, B. J. Hicke, Y. Lin, D. S. Coffey, Cancer Res. 2002, 62, 4029 4033. [11] Z. Xiao, D. Shangguan, Z. Cao, X. Fang, W. Tan, Chem. Eur. J. 2008, 14, 1769 1775. [12] X. Fang, W. Tan, Acc Chem. Res. 2010, 43, 48 57. [13] K. Sefah, Z. W. Tang, D. H. Shangguan, H. Chen, D. Lpez-Coln, Y. Li, P. Parekh, J. Martin, J. A. Phillips, Y. M. Kim, W. H. Tan, Leukemia 2009, 23, 235 244. [14] H. W. Chen, C. D. Medley, K. Sefah, D. Shangguan, Z. Tang, L. Meng, J. E. Smith, W. Tan, ChemMedChem 2008, 3, 991 1001. [15] Z. Tang, D. Shangguan, K. Wang, H. Shi, K. Sefah, P. Mallikratchy, H. W. Chen, Y. Li, W. Tan, Anal Chem. 2007, 79, 4900 4907. [16] G. H. Heppner, S. R. Wolman, Breast J. 1999, 5, 122 129. [17] P. J. Dawson, S. R. Wolman, L. Tait, G. H. Heppner, F. R. Miller, Am. J. Pathol. 1996, 148, 313 319. [18] D. Van Simaeys, D. Lpez-Coln, K. Sefah, R. Sutphen, E. Jimenez, W. Tan, PLoS One 2010, 5, e13770. [19] N. M. Green, Adv. Protein Chem. 1975, 29, 85 133. [20] T. C. Chu, J. W. Marks III, L. A. Lavery, S. Faulkner, M. G. Rosenblum, A. D. Ellington, M. Levy, Cancer Res. 2006, 66, 5989 5992. [21] V. Bagalkot, L. Zhang, E. Levy-Nissenbaum, S. Jon, P. W. Kantoff, R. Langer, O. C. Farokhzad, Nano Lett. 2007, 7, 3065 3070. [22] A. Ciechanover, A. L. Schwartz, A. Dautry-Varsat, H. F. Lodish, J. Biol. Chem. 1983, 258, 9681 9689. [23] Y. H. Huang, D. Shangguan, H. Liu, J. A. Phillips, X. Zhang, Y. Chen, W. Tan, ChemBioChem 2009, 10, 862 868. Received: September 30, 2011 Revised: October 31, 2011 Published online on December 14, 2011
84
www.chemmedchem.org
DOI: 10.1002/cmdc.201100453
Platinum(II) Phenanthroimidazoles for Targeting Telomeric G-Quadruplexes

Katherine J. Castor,[a] Johanna Mancini,[b] Johans Fakhoury,[a] Nathanael Weill,[a] Roxanne Kieltyka,[a] Pablo Englebienne,[a] Nicole Avakyan,[a] Anthony Mittermaier,[a] Chantal Autexier,*[b] Nicolas Moitessier,*[a] and Hanadi F. Sleiman*[a]
A rationally designed progression of phenanthroimidazole platinum(II) complexes were examined for their ability to target telomere-derived intramolecular G-quadruplex DNA. Through the use of circular dichroism, fluorescence displacement assays, and molecular modeling we show that these complexes template and stabilize G-quadruplexes from sequences based on the human telomeric repeat (TTAGGG)n. The greatest stabilization was observed for the p-chlorophenyl derivative 6 (G4DC50 = 0.31 mm). We also show that the G-quadruplex binding complexes are able to inhibit telomerase activity through a modified telomerase repeat amplification protocol (TRAP-LIG assay). Preliminary cell studies show that complex 6 is preferentially cytotoxic toward cancer over normal cell lines, indicating its potential use in cancer therapy.
Introduction
Guanine quadruplexes have gained recognition as important targets for chemotherapeutic drug design as a consequence of their potential to interfere with cancer cell proliferation.[1, 2] These higher-order DNA structures, held together by Hoogsteen hydrogen bonds, result from the folding of a guanine (G)rich DNA sequence in the presence of potassium or sodium cations.[3] G-quadruplex-forming sequences have been identified throughout the human genome in telomeres, promoter regions of oncogenes, nuclease hypersensitivity regions, and untranslated regions of RNA.[4, 5] From this list, one highly investigated G-quadruplex (G4) target for small-molecule drugs is the telomere. At its 3-terminus, the human telomere is a single DNA strand composed mainly of (TTAGGG)n repeats.[6] Biologically relevant phenomena are reported to occur upon the folding of this sequence into a G4 structure, such as loss of telomere integrity through disruption of the protective shelterin complex of proteins.[7, 8] This results in activation of DNA damage response mechanisms that follow recognition of double-strand breaks,[9] in the activation of PARP and ATM signaling,[10] and in the prevention of telomere elongation by the reverse transcriptase enzyme telomerase.[11] Interestingly, stabilization of G-quadruplexes present in promoter regions of oncogenes was found to result in transcriptional repression of the oncogene.[12, 13] This, coupled with their structural diversity in various promoter regions, makes them attractive drug targets for selective therapy. Thus, the stabilization of G-quadruplexes is predicted to have profound effects on cancer cell proliferation through a number of mechanisms, thereby accomplishing one of the main goals of chemotherapy: to halt tumor growth. The use of small-molecule binders to stabilize the G4 DNA motif is a possible approach to achieve this therapeutic goal. To date, many compounds with varying degrees of efficacy in
their ability to target this higher-order DNA structure have been reported.[14, 15] Based on this wealth of empirical data obtained not only from binding studies, but also from molecular modeling and crystal structure determination of G4ligand complexes,[16, 17] several clues about the rational design of G4 binders have been uncovered, including the need for an electron-poor aromatic p-surface, positive charges in the scaffold, and the presence of positively charged side arms.[14a] Most classical G4 binders are purely organic heteroaromatic compounds.[14a] A particularly attractive strategy for designing G4 binders is the incorporation of metal centers, as they offer a wide variety of structural and functional features that can be exploited for optimal binding. Georgiades et al. recently reviewed metal complexes and their abilities to bind G4 DNA.[15] In particular, platinum(II) complexes can introduce a square planar geometry unavailable to carbon compounds and can act as electron-withdrawing groups to further increase positive character on the ligand. They can also simplify synthetic access to G4 binders and allow the modular generation of compound libraries in a minimal number of steps. Platinum(II) complexes
[a] K. J. Castor, Dr. J. Fakhoury, Dr. N. Weill, Dr. R. Kieltyka, Dr. P. Englebienne, N. Avakyan, Prof. A. Mittermaier, Prof. N. Moitessier, Prof. H. F. Sleiman Department of Chemistry, McGill University 801 Sherbrooke West, Montreal, QC, H3A 2K6 (Canada) Fax: (+ 1) 514-398-3797 E-mail: hanadi.sleiman@mcgill.ca nicolas.moitessier@mcgill.ca [b] J. Mancini, Prof. C. Autexier Departments of Anatomy and Cell Biology and Experimental Medicine McGill University, Lady Davis Institute of Medical Research 3755 Cte Ste-Catherine Road, Montreal, QC, H3T 1E2 (Canada) E-mail: chantal.autexier@mcgill.ca Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100453.
85
MED
such as cisplatin are commonly used antitumor therapeutics. These compounds bind directly to two adjacent DNA bases upon loss of their labile chloride ligands.[18] In contrast, platinum diamine and phenanthroline complexes possess stable ligand shells and, as a result, can display a more selective antitumor profile. Herein we examine the construction of a new class of platinum(II) phenanthroimidazole complexes as guanine quadruplex binders for telomeric DNA sequences. Through circular dichroism, fluorescence intercalator displacement (FID) assays, and a modified telomeric repeat amplification protocol (TRAPLIG) assay, we structurally evolve these highly modular and synthetically accessible constructs toward optimal G4 binding. In particular, we examine the effect of increasing the p-surface area of the ligand through aromatic ring extension and by introducing different hydrogen bonding interactions. We also probe the effect of introducing an electron-withdrawing substituent on binding affinity, selectivity, and telomerase inhibition. Lastly, we perform preliminary cell studies which confirm the ability of the optimized structure to target cancer cells.
C. Autexier, N. Moitessier, H. F. Sleiman, et al.

Design and synthesis of phenanthroimidazole platinum(II) complexes We have previously shown that two platinum-based complexes with extended p-surfaces afford greater binding affinity to the tetrameric, intermolecular G-quadruplex than to duplex B-form DNA (B-DNA).[19] We were interested in examining the binding affinity of these complexes toward more biologically relevant, telomere-derived intramolecular G-quadruplexes. Moreover, we wished to rationally evolve the structure of these complexes toward optimal affinity and selectivity for Gquadruplexes over B-DNA. Given the structural and electronic requirements of transition metals, the criteria for optimal G4 binding with our complexes may be distinct from those developed with purely organic molecules. Complexes 16 (Figure 1) possess phenanthroimidazole ligands that are bound to platinum(II) ethylenediamine metal centers. Their phenanthroimidazole units were synthesized by condensation of 1,10-phenanthroline-5,6-dione with an aromatic (phenyl, naphthyl, indolyl, quinolinyl, salicyl, or p-chlorophenyl) aldehyde. These ligands were then treated with potassium tetrachloroplatinate, resulting in [Pt(phenanthroimidazole)Cl2], followed by treatment with ethylenediamine and precipitation with ammonium hexafluorophosphate, resulting in complexes 1 (62 % yield),[19] 2 (65 % yield),[19] 3 (42 % yield), 4 (39 % yield), 5 (54 % yield), and 6 (53 % yield). The complexes were characterized by 1H and 13C NMR spectroscopy, as well as HRMS. Complexes 3 and 4 are light-sensitive and required handling in the dark to avoid decomposition. Binding affinity and selectivity of the phenyl complex 1 We first examined the ability of the parent platinum phenyl phenanthroimidazole complex 1 to template the folding of a
Figure 1. Phenylphenanthroimidazole ethylenediamine platinum(II) 1, naphthylphenanthroimidazole ethylenediamine platinum(II) 2, indolylphenanthroimidazole ethylenediamine platinum(II) 3, quinolinylphenanthroimidazole ethylenediamine platinum(II) 4, salicylphenanthroimidazole ethylenediamine platinum(II) 5, p-chlorophenylphenanthroimidazole ethylenediamine platinum(II) 6, bipyridine ethylenediamine platinum(II) 7. Hashed lines in complexes 3, 4, and 5 represent intramolecular hydrogen bonds.
G4 structure from single-stranded DNA (ssDNA). This was carried out by titration of the human telomeric sequence 22AG (see Experimental Section) with complex 1 in the absence of additional Na + or K + salts. In this assay, if templation occurs, a decrease in the circular dichroic peaks associated with the single-stranded oligonucleotide (slight negative peak at 238 nm and strong positive peak at 257 nm) and in the concomitant appearance of peaks associated with G4 structures are expected (Figure 2). An antiparallel, basket-type G-quadruplex, such as that found in buffer solutions containing Na + cations for this sequence, would result in a characteristic positive peak at 292 nm and a negative peak at 262 nm (Figure 2 a). A mixed hybrid structure, such as that found in buffer solutions
Figure 2. Schematics of the G4 polymorphs: a) the basket-type antiparallel structure found in 10 mm sodium cacodylate buffer (pH 7.2) with 100 mm added NaCl, and b) form 1 and c) form 2 of the (3+1) hybrid-type structure found in 10 mm sodium cacodylate buffer (pH 7.2) with 100 mm added KCl. The gray and white parallelograms represent syn and anti orientations, respectively, of the guanine bases contained within the tetrads.
86
www.chemmedchem.org
PtII Phenanthroimidazoles containing K + cations, is expected to give two positive maxima at 295 and 260 nm (Figure 2 b,c). To ensure the 22AG sequence is single-stranded at the start of titration, the DNA was dialyzed against Milli-Q water for 1 3 days to remove any excess salts remaining from the purification process. Upon gradual addition of complex 1, a strong positive peak at 295 nm and negative peak at 260 nm were observed, consistent with templation of a basket-type antiparallel G-quadruplex from this strand (Figure 3 a) in the absence of added Na + or K + . In contrast, the less p-extended platinum bipyridine complex 7 showed no distinct differences upon titration of the single-stranded 22AG, consistent with weaker binding to the G-quadruplex (Figure 3 b). been reported for other strong G4 binders, such as telomestatin[21] and a macrocyclic oligoamide.[21, 22] We next examined the binding affinity of complex 1 to the G-quadruplex using a fluorescence intercalator displacement (FID) assay. This experiment is based on previous work by Teulade-Fichou and co-workers.[23] It relies on competitive displacement of the dye thiazole orange from the DNA target using platinum complex 1, with a concomitant decrease in its fluorescence intensity. The binding affinity was evaluated from the concentration of platinum complex required to give a 50 % decrease in dye fluorescence (DC50). Due to the polymorphic nature of human telomeric quadruplexes in the presence of different metal cations,[16a,g] we investigated binding affinities to both the basket-type quadruplex found in 10 mm sodium cacodylate/100 mm NaCl buffer and to the hybrid-type structure, found in 10 mm sodium cacodylate/100 mm KCl (pH 7.2) (Figure 2). We compared these DC50 values to those obtained using two duplex DNA targets (a 17mer duplex and a self-complementary 26mer duplex mixed sequence) in order to evaluate the selectivity of our complexes to G4 versus B-DNA. The observed G4DC50 value for complex 1 of ~ 0.6 mm for both G4 polymorphs is consistent with good binding affinity, and indicates stronger binding to these quadruplex structures than to B-DNA (dsDC50/G4DC50 ~ 1.21.7). These studies also show that phenyl phenanthroimidazole complex 1 binds significantly better than bipyridine complex 7, a compound with a smaller p-surface area, emphasizing the importance of large, extended p-surfaces for G4 binding to these platinum complexes (Table 1).
Table 1. DC50 values determined from FID assays. Complex 1 2 3 4 5 6 7

G4 DC50 [mm] Na + K+ ds DC50 [mm] 17mer 26mer
Selectivity[b] 1.21.7 0.921.6 0.662.6 1.32.4 1.63.0 1.84.7
0.66 1.30 0.53 0.70 0.53 0.31 > 2.5
0.68 1.71 1.55 1.09 0.70 0.66 ND[a]
1.11 2.12 1.37 1.65 1.60 1.47 > 2.5
0.84 1.58 1.03 1.37 1.10 1.20 > 2.5
Figure 3. CD spectra of the G-quadruplex templation with a) 1 and b) 7. Arrows in panel a) show the direction of change in CD signal with increasing additions of platinum complex 1. The inset in panel a) shows the decrease in the circular dichroic peak between 250 and 260 nm with added complex 1. The saturation point is at ~ 2:1 ratio of complex 1:G4 DNA strand and indicates a 2:1 binding stoichiometry.[20]
[a] Not determined: given the inability of complex 7 to bind strongly to G4 DNA, no thiazole orange displacement was observed, and therefore no DC50 value could be determined with the G4 structures formed in potassium-containing buffer. [b] Selectivity range determined from dividing the lowest dsDC50 by the highest G4DC50 and the highest dsDC50 by the lowest G4DC50 for each complex.
The binding stoichiometry of complex 1 to quadruplex DNA can be determined from the changes in molar ellipticity values of the CD bands at 250260 nm or 285295 nm (Figure 3 a inset). Upon plotting these values as a function of molar equivalents of platinum(II) complex added, the curve saturates at a complex 1:G4 DNA strand ratio of ~ 2:1. This is consistent with a binding stoichiometry of 2:1, a value that has previously
We initially attempted to obtain binding data by using two other standard methods, FRET melting[24] and UV/Vis titration. The FRET melting experiment relies on heating a labeled nucleic acid structure 5-FAM-G3(T2AG3)3-TAMRA-3 in the presence of transition metal G4 binders, and monitoring G-quadruplex denaturation from the onset of fluorescence. However, while there were significant melting temperature differences with and without our platinum complexes (DT1/2 % 20 8C), the con-
www.chemmedchem.org
87
MED
centration of complexes required to effect these changes was higher than for other potent G4 binders (5 mm versus 1 mm). Careful control experiments showed that platinum complex 1 can quench the fluorescence of dyes attached to the DNA strand, thus interfering with the fluorescence measurement. Therefore, caution should be taken when interpreting FRET melting temperature results, which can be complicated by changes in dye fluorescence brought about by the typically extended aromatic G4 binders.[25] We also attempted to measure the binding affinity by monitoring the UV/Vis spectrum of the platinum complex while titrating in the G-quadruplex. For these extended aromatic complexes, the experiment was complicated by aggregation of the platinum complexes in solution at the start of the experiment and the consequent difficulty in obtaining the initial absorbance value for the free, unstacked complex. This is a common problem for this technique,[26] as most G4 binders possess extended aromatic rings. In light of these observations, the fluorescence displacement assay was the most reliable method to compare binding affinities and selectivities for these platinum complexes. Increasing the p-surface area of the platinum complexes The parent phenyl phenanthroimidazole complex 1 displays good affinity and selectivity toward G-quadruplexes, whereas the bipyridine complex 7 shows significantly lower binding strength toward this target. This encouraged us to examine increasing the p-surface area of these complexes with the aim of optimizing binding interactions. As such, we introduced a naphthyl rather than a phenyl substituent (complex 2) into the phenanthroimidazole ligand on the platinum center. Titration of ssDNA with complex 2 resulted in templation of the antiparallel G-quadruplex (Supporting Information figure S1), similarly to complex 1. However, FID assays showed significantly lower binding affinity to the intramolecular telomeric G-quadruplex, both in the presence of Na + (G4DC50 = 1.30 mm) and K + (G4DC50 = 1.71 mm), which was similar to the affinity for duplex DNA. This observation contrasts with our earlier studies, which showed higher affinity of this naphthyl complex toward the all-parallel intermolecular G-quadruplex.[19] This result prompted us to examine the structure of complex 2 more closely. While increasing the p-surface area of the aromatic group may result in greater stacking of this group with the G-quadruplex, it would also be expected to impose a larger twist angle between the phenanthroimidazole and naphthyl moieties. This twisted structure may prevent favorable overlap of the more extended p-surface ligand with the intramolecular telomeric G-quadruplex. The effective twist angles between the phenanthroimidazole ligand and the substituent rings were analyzed to further understand the extent of planar p-surface each ligand can provide. Density functional theory (DFT) calculations at the B3LYP/LACV3P** level of theory were performed by scanning the NCCC torsion angle between the phenanthroimidazole ligand and the substituents for the phenyl and naphthyl ligands at 58 increments. The phenyl ligand in complex 1 shows a shallow potential energy dependence on the torsion angle between the two aromatic units and
C. Autexier, N. Moitessier, H. F. Sleiman, et al. an energy minimum at 158. However, the naphthyl ligand in complex 2 shows an energy minimum at a higher torsion angle of 308, with the energy more sharply increasing at 08 angle, or when the two aryl groups are aligned (Figure 4).
Figure 4. Energy diagram from the torsional scan around the NCCC bond in the phenanthroimidazole ligands of complexes 16; the inset shows the bond around which the torsion angles were measured. Complex 1 (c), 2 (c), 3 (c), 4 (b), 5 (b), 6 (b).
This encouraged us to examine a new class of ligands in which the attached aryl substituent is p-extended, but can be locked in an aligned, 08 twist angle through an intramolecular hydrogen bond (Figure 1). Complex 3 has an indole unit that is expected to engage in hydrogen bonding between the indole nitrogen (hydrogen bond donor) and the imidazole nitrogen atom. Complex 4 has a quinoline unit, the nitrogen atom of which can act as a hydrogen bond acceptor with the imidazole nitrogen acting as donor. Complex 5 has a less p-extended phenol group that can act as a hydrogen bond donor to the imidazole nitrogen (Figure 1). DFT calculations show that the effectively locked ligands exhibit a minimum in the potential energy surface at a torsion angle value of 08, with a sharp increase in energy for deviations of more than 10158 (Figure 4). By removing the internal strain within the ligands and locking the biphenyl torsion angle, the entropic penalty can be decreased upon binding. Intramolecularly hydrogen bonded complexes Titration of ssDNA with complex 4 was monitored by CD (Figure 5). Similar to complexes 1 and 2, complex 4 resulted in exclusive templation of the antiparallel quadruplex form and exhibited a binding ratio of ~ 2:1 complex 4:G4 DNA strand (Supporting Information figure S2). Complex 3 also templates an antiparallel structure, or a mixture of structures that have antiparallel characteristics (Supporting Information figure S3).[27] FID assays showed increased binding affinity of complex 3 to the G4 structure (G4DC50 = 0.53 mm) relative to the naphthyl complex 2. This is consistent with our torsion angle studies,
88
www.chemmedchem.org
PtII Phenanthroimidazoles for this structure over duplex DNA. In addition, selectivity toward a specific G4 polymorph (the antiparallel basket structure) is enhanced for these structures, outlining new design features that can be used to target specific G4 structures.
Introduction of a halogen substituent Locking the aryl substituent of complexes 3 and 4 with an intramolecular hydrogen bond resulted in an increase in affinity and selectivity for the antiparallel G4 structure. We were also interested in investigating whether electronic effects could further improve the binding affinity of this class of molecules. In particular, given the electron-rich nature of the G-quartet, we examined the effect of increasing the electron deficiency of our complexes by adding a chlorine substituent at the para position of the phenyl phenanthroimidazole (complex 6).[28] DFT calculations showed a significant impact of the chlorine atom on the highest occupied molecular orbital (HOMO) of the complex (and no impact on the lowest unoccupied molecular orbital (LUMO)), with a decrease in electron density and aromatic character of the imidazole unit in complex 6, relative to the unsubstituted complex 1 (Figure 6). Thus, the intervening
Figure 5. CD spectra of the G-quadruplex templation with 4. Arrows show the direction of the change in CD signal as a result of increasing additions of platinum complex 4.
which predict a significantly decreased twist angle for this complex. The quinoline complex 4 showed similar results with a slightly decreased binding affinity for the antiparallel G-quadruplex (G4DC50 = 0.70 mm). Selectivity for G4 versus B-DNA was increased for both complexes (complex 3: dsDC50/G4DC50 ~ 0.7 2.6, complex 4: dsDC50/G4DC50 ~ 1.32.4; Table 1). Interestingly, complexes 3 and 4 bind more strongly to the G4 antiparallel basket than to the hybrid structure (e.g., G4 DC50 = 0.53 mm versus G4DC50 = 1.55 mm for complex 3). Thus, unlike the phenyl complex 1, these complexes display greater affinity and selectivity for a specific G4 polymorph. This binding selectivity is a valuable feature, as numerous G4 targets of different structures have recently been described in the promoter regions of oncogenes.[10] A current challenge is to find small molecules that preferentially target one of these polymorphic quadruplex structures preferentially. Molecules such as complexes 3 and 4 display such a selectivity. We are currently examining the binding kinetics of these complexes to the various G4 polymorphs using surface plasmon resonance to better understand this selectivity. CD titration studies show that phenol derivative 5 templates clean formation of the antiparallel basket G4 polymorph (Supporting Information figure S4) with a 2:1 complex 5:G4 DNA strand binding stoichiometry, much like the closely related phenyl derivative complex 1 (Supporting Information figure S5). Introduction of the ortho-hydroxy substituent was found to increase the binding affinity of this complex for the antiparallel G4 form, as well as to increase the selectivity for G4 versus duplex DNA (G4DC50 = 0.53 mm, dsDC50/G4DC50 ~ 1.6 3.0). However, complex 5 displays less selectivity toward the antiparallel versus the hybrid G4 forms than do complexes 3 and 4. Thus, this complex shows similar binding features to those of the closely related phenyl complex 1, albeit with improved affinity and selectivity. The fact that binding affinity did not significantly improve in the case of complex 5 may be a result of the electron-rich nature of the phenol ring, which can play a role in decreasing binding to the G-quartet structure. In summary, the introduction of hydrogen bonding substituents on the aromatic ring increases the binding affinity of complexes 35 to the G-quadruplex, and also increases selectivity
Figure 6. Comparison of the HOMOs of a) complex 1 and b) complex 6. The dark isosurfaces represent the HOMO of the complexes computed by Jaguar 7.7 using an isovalue of 0.05.
phenyl ring efficiently mediates a shift in electron density away from the imidazole ring toward the chlorine atom. Moreover, the chlorine substituent results in a calculated dipole moment for complex 6 that is significantly greater than that of the unsubstituted complex 1 (17.8 versus 10.7 Debye). Finally, scanning the energy as a function of torsion angle between the pchlorophenyl and imidazole groups for complex 6 shows an energy minimum at 058 (Figure 4), that is, the two aromatic rings are predicted to be aligned. This is likely the result of increased conjugation between the imidazole and the electrondeficient p-chlorophenyl group. Thus, complex 6 is not only predicted to display a more electron deficient p-surface, it is assumed to be planar with minimal twisting between its aromatic rings despite its lack of internal hydrogen bond. These modeling results encouraged us to synthesize this complex and assess its binding affinity for the G4 structures.
www.chemmedchem.org
89
MED
CD titration of single-stranded telomeric DNA with chlorophenyl complex 6 resulted in the exclusive formation of the antiparallel G4 form (Supporting Information figure S6), in a similar manner to phenyl 1, quinoline 4, and phenol 5 complexes with the same binding stoichiometry of 2:1 (Supporting Information figure S7). FID assays showed significantly increased binding affinity of complex 6 to the antiparallel G4 form in Na + buffer (G4DC50 = 0.31 mm), which is similar to some of the more potent G4 binders reported.[28] Selectivity for G4 versus duplex DNA is increased (dsDC50/G4DC50 ~ 1.84.7), and preference for the antiparallel G4 form over the hybrid form is maintained. Thus, by creating a conjugated donoracceptor system (imidazole and chlorophenyl rings in complex 6) the two aromatic rings are kept in a parallel conformation. With this method, binding affinity and selectivity to G-quadruplexes can be significantly increased, without the need for fused polycyclic aromatic rings. TRAP-LIG assay To analyze the telomerase inhibition potential of platinum(II) complexes, an adapted version of the telomeric repeat amplification protocol (TRAP) assay[29] was used. Standard TRAP reactions contain telomeric substrate (TS) primer, NT primer, and TSNT internal control primer. These latter primers ensure that the telomerase inhibition observed is not due to inhibition of the PCR reaction step. Elongation of the telomeric substrate by telomerase generates DNA products that are then PCR amplified in the presence of a reverse primer termed ACX. DNA products generated by telomerase and amplified by PCR are then resolved on a non-denaturing polyacrylamide gel and analyzed. The resulting ladder of DNA products is normalized to the internal control (IC, generated by the PCR amplification of TSNT by both the TS and NT primers), and the ratio obtained for the inhibited telomerase reactions is then compared against the ratio of uninhibited telomerase (positive control containing protein only). Although the standard TRAP assay has been used successfully for assessing the ability of G4 motifs to inhibit telomerase, a recent study indicates that the efficiency of inhibition can be overestimated due to the inhibition of PCR amplification on quadruplex-prone motifs, even though the internal PCR control (IC) is not affected.[30] The modified TRAP assay, the TRAP-LIG protocol, removes bound ligand prior to the PCR step via a spin column, eliminating any inhibition that may be due to the presence of the ligand.[29] In the presence of increasing concentrations of platinum(II) complex (complexes 16) a decrease in TRAP product is observed, indicating that telomerase activity is inhibited by these G4 binders (Figure 7 a). The IC50 values range from 0.46 to 11.6 mm (Figure 7 b), which are on the same order as those of telomestatin (IC50 = 0.6 mm) and BRACO-19 (IC50 = 6.3 mm) as measured with the TRAP-LIG protocol.[29] Moreover, the result obtained for 5,10,15,20-tetrakis-(1-methyl-4-pyridino)porphyrin was on the same order of magnitude as those previously reported,[29] demonstrating the reliability of the TRAP-LIG assay (data not shown). In order to confirm the complete removal of the complex by the spin column, we added an excess of com-
Figure 7. a) Non-denaturing polyacrylamide gel from the TRAP-LIG assay for complex 1 [Pt(PIP)(en)]. The first two lanes at left represent negative (no protein extract) and positive (no added ligand) controls. Increasing concentration of complex 1 shows a decrease in the length of elongation products, consistent with the inhibition of telomerase activity. The lane at far right (bp) is a DNA ladder used for size comparison with the elongation products. Bands appearing at 35 bp are the internal control (IC) of the TRAP-LIG. b) Inhibition curves from TRAP-LIG assays for complexes 16: complex 1 (^), 2 (*), 3 (*), 4 (&), 5 (~), 6 ( ! ).
plex 2 after the extension reaction and subsequently purified the extended product with the spin column. After the PCR step we observed that the signal of the amplified product for the positive control (no complex added) was the same as the reaction in which the complex was added after the extension reaction (data not shown). This observation confirms that all platinum(II) complexes added before the extension step are completely removed by the spin column, and that telomerase inhibition is indeed due to the G4 binders. From the above observations we conclude that complexes 16 are efficient and specific telomerase inhibitors. A closer examination of the inhibition curves (Figure 7 b) reveals additional detail that cannot be obtained merely from the IC50 values. At a concentration of 32 nm, there is already significant inhibition of telomerase elongation products for some of the complexes (e.g., up to ~ 40 % for complex 4). Interestingly, only complexes 5 and 6, which are the strongest G4 binders, result in 100 % inhibition of telomerase elongation at the end of the experiment. Complex 6 in particular has the largest inhibition window (from ~ 100 % enzyme activity at 32 nm to ~ 0 % at 50 mm). The inhibition of telomerase elongation in the TRAP-LIG assay is a complex process. It can result
90
www.chemmedchem.org
PtII Phenanthroimidazoles from the ability of these complexes to bind and sequester the telomerase substrates in their G4 form, or possibly from their direct inhibition of the enzyme activity or processivity.[30] Thus, although the inhibition of telomerase elongation by complexes 16 correlates with G-quadruplex binding affinity, it is clear that each complex gives a unique profile with different initial and final activities and different inhibition slopes. Preliminary cytotoxicity studies Preliminary studies using the MTS assay[31] investigated the cytotoxicity of complex 6, the complex with the best selectivity for G4 versus duplex DNA, in telomerase-expressing cancer cell lines A549 and MCF-7, as well as in a normal fibroblast cell line, MRC-5. Complex 6 was found to be threefold more cytotoxic to the cancer cell lines than to the normal fibroblast cell line (IC50 values of ~ 20 mm versus 60 mm upon treatment for 72 h). To ensure G-quadruplex specificity, we tested the cytotoxicity of bipyridine complex 7, a platinum complex that showed no binding to G-quadruplexes (see above). We were consistently unable to obtain any IC50 values for complex 7 (up to 100 mm) using the same assay. These preliminary data indicate that complex 6 causes preferential cell cytotoxicity in cancer cell lines while being less toxic toward normal cells, a selectivity that compares well with previously reported strong G4 binders.[17] These complexes warrant further investigation for use as therapies for stabilizing G4 structures at telomeres and/or G-rich oncogene promoter regions. ic rings. Experimentally, this complex 6 was found to bind to antiparallel G-quadruplexes with the highest affinity and selectivity, which is similar to some of the most potent G-quadruplex binders. 5. Complexes 16 are effective telomerase inhibitors, but differ significantly in their inhibition profiles. 6. Complex 6, a selective G-quadruplex binder, is more cytotoxic to cancer cells than to normal cells. These studies have illustrated a number of design principles for guanine quadruplex binders. An extended p-aromatic surface had been found to result in high binding affinity in most previous reports; however, extended polycyclic aromatic molecules may not be sufficiently drug-like. These studies have shown that 1) introducing an intramolecular hydrogen bond between two aromatic rings (in complexes 35), or b) creating a conjugated donoracceptor system (imidazole and chlorophenyl rings in complex 6) can effectively lock the two aromatic rings in a parallel conformation, thus significantly increasing binding affinity and selectivity. These strategies may circumvent the need for including fused polycyclic aromatic motifs in G-quadruplex binders. Because of their ease of synthesis and modification, we are currently examining strategies to further simplify the design of these complexes, and are carrying out more detailed examinations of the mechanism by which they target cancer cells.
Chemistry
NMR spectra were recorded at 300 or 400 MHz on a Varian Mercury operated with VNMRJ 2.2D software under LINUX Red Hat 5 using TMS as internal standard. The following abbreviations are used: singlet (s), doublet (d), triplet (t), multiplet (m), and broad (br). Mass spectrometry was performed on a Kratos MS25RFA highresolution mass spectrometer equipped with gas chromatograph and El-Cl source. UV/Vis experiments were carried out on a CaryBio 300 spectrophotometer using the Cary WinUV Scan application version 3.00. Fluorescence measurements were performed on a FluoroMax-2 spectrofluorimeter (JobinYvon). IR spectra were collected on a PerkinElmer 400 instrument. Circular dichroism studies were performed at 25 8C on a JASCO J-810 spectropolarimeter with the Spectra Manager II software. All chemicals were used as received from SigmaAldrich without further purification for synthesis of the ligands and their respective complexes. Quinoline-8-carboxaldehyde was received from AlfaAesar and used without further purification. 1,10-Phenanthroline-5,6-dione, bipyridine 7, phenylphenanthroimidazole 1, and naphthylphenanthroimidazole 2 salicylphenanthroimidazole (SIP) platinum(II) complexes,[19] ligand,[32] and p-chlorophenylphenanthroimidazole (CLIP) ligand[33] were synthesized as previously reported. 2-Indolylimidazo[4,5-f]1,10-phenanthroline (PII): 1,10-Phenanthroline-5,6-dione (0.1500 g, 0.721 mmol), and ammonium acetate (1.1115 g, 14.42 mmol) were dissolved in glacial acetic acid (50 mL) in the absence of light and under inert atmosphere. Once mostly dissolved, indole-7-carboxaldehyde (0.1255 g, 0.865 mmol) was added, and this reaction mixture was held at reflux for 1 h. The mixture was cooled on ice and then neutralized using NH4OH to pH 7. Crude product precipitated as a light-brown solid. The solid was collected by vacuum filtration before being dissolved in
Conclusions
We have shown that phenanthroimidazole platinum(II) complexes template and strongly bind G4-forming sequences based on the human telomeric repeat, (TTAGGG)n, inhibit telomerase, and are cytotoxic to cancer cells. The specific findings are summarized below: 1. Phenanthroimidazole platinum(II) complex 1 can cleanly template the formation of an antiparallel G-quadruplex from single-stranded telomeric DNA, in a 2:1 stoichiometry, without the need for the addition of Na + or K + ions. It binds to Gquadruplexes with good affinity and selectivity over duplex DNA. 2. Through modeling studies, it was found that increasing the p-surface area of this complex by moving from a phenyl to a naphthyl substituent in complex 2 also increases the torsion angle between the two aromatic rings. Experimentally, it results in decreasing the binding affinity for G-quadruplexes. 3. By locking the aromatic rings together using an intramolecular hydrogen bond in complexes 35, higher binding affinities and selectivities for G-quadruplexes are observed. Interestingly, the selectivity of these complexes to the antiparallel over the hybrid G-quadruplex structure is also significantly increased. 4. Most notably, by introducing a chloro substituent at the para position, DFT studies predicted a significant decrease in electron density of the aromatic ring, an increase in dipole moment, and a near-zero twist angle between the two aromatChemMedChem 2012, 7, 85 94
www.chemmedchem.org
91
MED
MeOH/CH2Cl2 and evaporated onto silica gel under reduced pressure for purification by column chromatography. The reaction mixture was separated first with 100 % EtOAc followed by a gradient of CH2Cl2/MeOH from 0!8 % MeOH. The final product was evaporated to dryness as a yellow solid. It is important to exclude light at all times. Yield: 62.8 mg, 26 %; 1H NMR (300 MHz, [D6]DMSO): d = 13.81 (s, 1 H), 11.75 (s, 1 H), 9.43 (dd, 1 H), 9.06 (m, 2 H), 8.99 (dd, 1 H), 8.07 (d, 1 H), 7.88 (m, 2 H), 7.75 (d, 1 H), 7.58 (t, 1 H), 7.27 (t, 1 H), 6.62 ppm (t, 1 H); 13C NMR (500 MHz, [D6]DMSO): d = 172.3, 150.4, 147.9, 143.6, 132.7, 130.2, 129.1, 126.8, 123.2, 122.2, 118.9, 118.5, 113.0, 101.6, 21.3 ppm; HRMS (ESI, 50:50 MeOH/H2O) m/z (%): calcd 336.12437, found 336.12423. 2-Quinolinylimidazo[4,5-f]1,10-phenanthroline (PIQ): This compound was prepared similarly to PII, except 1,10-phenanthroline5,6-dione (0.1500 g, 0.714 mmol), ammonium acetate (0.2751 g, 3.57 mmol), and quinoline-8-carboxaldehyde (0.1570 g, 0.999 mmol) were used, and the reaction mixture was stirred at room temperature for ~ 1 h. The final product was isolated as a brown solid. It is important to exclude light at all times. Yield: 99.1 mg, 40 %; 1H NMR (300 MHz, [D6]DMSO): d = 14.10 (s, 1 H), 9.24 (m, 1 H), 9.14 (m, 1 H), 9.05 (m, 2 H), 8.94 (m, 1 H), 8.91 (d, 1 H), 8.55 (dd, 1 H), 8.13 (dd, 1 H), 7.85 (m, 3 H), 7.76 ppm (dd, 1 H); 13 C NMR (500 MHz, [D6]DMSO): 151.5, 150.0, 148.4, 148.3, 144.8, 144.2, 144.0, 138.0, 135.6, 130.8, 130.4, 130.4, 130.2, 128.9, 127.3, 126.9, 126.4, 124.2, 123.9, 123.5, 122.5, 119.8 ppm; HRMS (ESI, MeOH/formic acid 90:10) m/z (%): calcd 346.1218, found 348.1244. [Pt(PII)(en)][PF6]2 (3): K2PtCl4 (0.0689 g, 0.166 mmol) was first dissolved in distilled water (10 mL) and heated at ~ 80 8C. Indolylphenanthroimidazole (PII) (0.0548 g, 0.164 mmol) was dissolved in DMSO (10 mL) at ~ 80 8C in the absence of light. The K2PtCl4 solution was then added to the PII solution, and the reaction mixture was allowed to stir overnight at room temperature before being vacuum filtered. The intermediate (PII)PtCl2 was washed with warm water before drying with Et2O to afford (PII)PtCl2 as a light-brown solid. Ethylenediamine (16.2 mL, 0.242 mmol) was then added to a suspension of (PII)PtCl2 (0.0160 g, 0.027 mmol) in EtOH (10 mL) in an amber vial to exclude light. The reaction mixture was stirred at room temperature overnight before vacuum filtration was used to remove unreacted material and rinsing with distilled water. Ammonium hexafluorophosphate was then added to the filtrate to precipitate the light-brown salt. The product was vacuum filtered on a glass frit and washed with a small amount of distilled water before drying with Et2O. The resulting solid was light brown. Yield: 9.98 mg, 42 %; 1H NMR (500 MHz, [D6]DMSO): d = 11.81 (s, 1 H), 9.96 (m, 1 H), 9.45 (m, 1 H), 9.04 (m, 2 H), 8.33 (m, 2 H), 8.07 (d, 1 H), 7.78 (m, 1 H), 7.58 (s, 1 H), 7.28 (t, 1 H), 6.91 (br s, 4 H), 6.64 (s, 1 H), 2.77 ppm (s, 4 H); 13C NMR (300 MHz, [D6]DMSO): d = 148.7, 143.8, 134.9, 133.2, 129.2, 126.8, 125.9, 124.1, 122.5, 119.3, 118.9, 102.1, 47.7 ppm; IR (ATR) n = 3645, 3398, 3322, 3143, 1599, 1414, 1341, 1170, 1055, 832, 555 cm1; HRMS (ESI, 50:50 MeOH/H2O) m/z (%): calcd [M2 PF6]2 + 294.57371, found 294.57337 with 194Pt. [Pt(PIQ)(en)][PF6]2 (4): 4 was prepared similarly to 3. After precipitation with NH4PF6, a dark-brown product was isolated. Yield: 39 %; 1 H NMR (500 MHz, [D6]DMSO): d = 9.84 (m, 1 H), 9.50 (m, 1 H), 9.24 (m, 1 H), 9.08 (m, 2 H), 8.90 (d, 1 H), 8.66 (d, 1 H), 8.278.33 (m, 3 H), 7.91 (t, 1 H), 7.82 (t, 1 H), 6.95 (m, 4 H), 2.78 (s, 4 H) ppm; 13C NMR (500 MHz, [D6]DMSO): d = 152.5, 151.7, 144.8, 138.1, 131.5, 131.3, 128.9, 127.4, 126.4, 122.8, 105.0, 47.7 ppm; IR (ATR) n = 3665, 3323, 2989, 2961, 1596, 1411, 1057, 834, 556 cm1; HRMS (ESI, 50:50 MeOH/H2O) m/z (%): calcd [M2 PF6]2 + 300.57371, found 300.57383 with 194Pt.

[Pt(SIP)(en)][PF6]2 (5): 5 was prepared similarly to 3. After precipitation with NH4PF6, a redorange product was isolated. There is no need to exclude light during this reaction. Yield: 54 %; 1H NMR (500 MHz, [D6]DMSO): d = 14.07 (br s, 1 H), 9.23 (d, 2 H), 8.87 (d, 2 H), 8.32 (d, 1 H), 8.10 (t, 2 H), 7.20 (t, 1 H), 6.946.86 (m, 6 H), 2.75 ppm (s, 4 H); 13C NMR (300 MHz, [D6]DMSO): d = 152.9, 147.6, 143.0, 134.4, 127.0, 125.5, 125.4, 118.9, 116.8, 47.6 ppm; IR (ATR) n = 3321, 1612, 1516, 1485, 1418, 1373, 1306, 1169, 1055, 830, 755, 714, 555 cm1; HRMS (ESI, 50:50 MeOH/H2O) m/z (%): calcd [M2 PF6]2 + 283.06571, found 283.06540 with 194Pt. [Pt(CLIP)(en)][PF6]2 (6): 6 was prepared similarly to 3. After precipitation with NH4PF6, a yellow product was isolated. There is no need to exclude light during this reaction. Yield: 53 %; 1H NMR (500 MHz, [D6]DMSO): d = 9.23 (dd, 2 H), 8.79 (dd, 2 H), 8.38 (d, 2 H), 8.05 (dd, 2 H), 7.47 (d, 2 H), 6.83 (br 3, 4 H), 2.74 ppm (s, 4 H); 13 C NMR (300 MHz, [D6]DMSO): d = 147.1, 143.2, 34.3, 128.8, 128.3, 126.2, 125.2, 47.6 ppm; IR (ATR) n = 3645, 3323, 2989, 2959, 1594, 1461, 1360, 1056, 838, 715, 557 cm1; HRMS (ESI, 50:50 MeOH/H2O) m/z (%): calcd [M2 PF6]2 + 292.04877, found 292.04871 with 194Pt.
Biophysical assays
All platinum(II) complexes 16 were prepared as 1.5 mm stock solutions in DMSO and stored at 4 8C. Complexes 3 and 4 were wrapped in aluminum foil and their exposure to light was minimized in all experiments. 22AG DNA oligonucleotides were obtained from SigmaGenosys Canada. 17mer and 26mer oligonucleotides were synthesized in house on a MerMade 6 DNA synthesizer from BioAutomation Corporation. 2-Cyanoethyl diisopropylchlorophosphoramidites, nucleoside (dC and dG) derivatized 500 and 1000 LCAA-CPG supports with loading densities between 25 and 40 mmol g1, 5-ethylthiotetrazole, and phosphoramidite reagents used for automated DNA synthesis were purchased from ChemGenes Inc. Oligonucleotides were dissolved in deionized water prior to use and quantified spectrophotometrically at room temperature (with the exception of the 22AG strand, which was heated at 95 8C) using their respective molar extinction coefficients[34] at l 260 nm: 5-AG3(T2AG3)3-3 (22AG) = 228 500 m1 cm1, 5-CAA TCG GAT CGA ATT CGA TCC GAT TG-3 (26mer) = 253 200 m1 cm1, and the 17mer complements were quantified separately with 5-CCA GTT CGT AGT AAC CC-3 (17mer A) = 160 900 m1 cm1 and 5-GGG TTA CTA CGA ACT GG-3 (17mer B) = 167 400 m1 cm1. FID assay: Dilutions of the nucleic acid stock solutions in Milli-Q water were done in 10 mm sodium cacodylate buffer (pH 7.4) with 100 mm added NaCl or KCl where appropriate. 22AG, 17mer, and 26mer DNA were diluted from the original stock solutions in water to 100, 30, and 10 mm, respectively, before being annealed. For the 22AG sequence, solutions were heated at 95 8C for 5 min before being allowed to cool to room temperature over 12 h. For the 17mer and 26mer duplexes, concentrated solutions were heated at 95 8C for 5 min before being allowed to slowly cool to room temperature over 67 h. For each platinum(II) complex, a solution of 175 mm complex in DMSO was prepared. Final dilutions of DNA solutions were made into the fluorescence cuvettes such that the DNA concentration was 0.25 mm. Thiazole orange was prepared as a 1 mm stock solution in DMSO. New solutions were used for each titration to minimize aggregation effects from freezethaw cycles. After dilution of the DNA, the fluorescence spectra were recorded (lex = 501 nm scanned from 510750 nm with 3 mm slits). These spectra were used as blanks. Then 2 molar equivalents (0.50 mm) thiazole orange was added (or 3 molar equivalents in the case of
92
www.chemmedchem.org
PtII Phenanthroimidazoles
the 26mer). Spectra were recorded after 5 min equilibration time. Then 0.510 molar equivalents of the appropriate platinum samples were pipetted into each cuvette, followed by 5 min equilibration time before recording the spectra. The fluorescence area (FA) is converted into percentage displacement (PD) by: PD = 100[(FA/FA0) 100], in which FA0 is FA before addition of platinum complexes. The concentration of added ligand is then plotted against the PD, and a DC50 value is determined by the concentration of platinum complex needed to achieve a 50 % displacement of thiazole orange. Each DC50 value is the average of two or three replicates. Circular dichroism studies: CD studies were performed at 25 8C on a JASCO J-810 spectropolarimeter using a 1 mm path length cuvette. Temperature was kept constant using the Peltier unit within the instrument. Spectra were recorded from 400230 nm at a scan rate of 50 nm min1 and a response time of 4.0 s with two acquisitions recorded for each spectrum. Data were smoothed using the meansmovement function within the JASCO graphing software. To observe the ability of these complexes to fold the human telomeric DNA sequence, 22AG was dialyzed with Pierce Slide-a-Lyzer mini-dialysis kits (Mr cutoff: 3500 Da; Fisher Scientific) against deionized Milli-Q water for 13 days before being diluted to 15 mm in 50 mm Tris buffer (pH 7.6) for a total volume of 200 mL. Platinum(II) complexes were titrated in 0.5 mL aliquots from 1.5 mm solutions in DMSO to the 22AG sequence. With each aliquot added, 5 min elapsed prior to recording the CD spectrum. Spectra were collected with the aforementioned parameters, until 3 or 4 molar equivalents were reached in platinum(II) complex. dNTPs (2.5 mm), 1 mL TS primer (20 pmol mL1), 1 mL NT primer (20 pmol mL1), 1 mL TSNT primer (1013 m), 1 mL ACX primer (20 pmol mL1), 0.4 mL Taq polymerase (4 U mL1, Invitrogen). The PCR conditions used: (Step 1) 30 s at 94 8C, (Step 2) 60 s at 60 8C, (Step 3) 60 s at 72 8C, repeated for 34 cycles. PCR product (20 mL) was resolved by non-denaturing PAGE. To visualize extension products, the gel was stained with SYBR Safe (Invitrogen) and visualized using a Storm 840 instrument (GE Healthcare). Quantification was performed using ImageQuantTL software by measuring the intensity of the extension products normalized to the intensity of the internal PCR control and then compared with the positive control. IC50 values were calculated using GraphPad Prism 5 software. Cytotoxicity studies: The cytotoxicity of complexes 6 and 7 was assessed using the CellTiter96 kit from Promega according to the manufacturers instructions. Briefly, A549 cells (human lung adenocarcinoma), MCF-7 cells (human breast adenocarcinoma), and MRC5 cells (normal human fibroblast) were seeded at a density of 5000 cells per well in a 96-well plate. Platinum complex stock solutions were initially diluted in growth medium to yield final concentrations ranging from 1.6 to 100 mm. Cells were incubated for 72 h in 5 % CO2 at 37 8C. After the incubation period, the MTS reagent was added to each well and further incubated for 3 h in 5 % CO2 at 37 8C. Subsequently, 96-well plates were allowed to equilibrate at room temperature and the absorbance was read at 490 nm using a BioTek Epoch micro-plate reader. All quantifications were done using GraphPad Prism 5 software.
Acknowledgements
This work was supported by a Canadian Institute for Health Research (CIHR) Grant MOP-89978 to H.F.S., N.M. and C.A. K.J.C. was supported by a Drug Development Training Grant from the McGill University Faculty of Medicine and a Chemical Biology Scholarship from the McGill University Department of Biochemistry. J.F. was supported by a CIHR Cancer Consortium Training Grant Award from the McGill Cancer Centre and a McGill University Faculty of Medicine Internal Studentship. P.E. was supported by a J.W. McConnell Memorial Fellowship. C.A. is a Chercheur National of the Fonds de la Recherche en Sant du Qubec. H.F.S. is a Cottrell scholar of the Research Corporation. Keywords: antitumor agents G-quadruplexes human telomeres phenanthroimidazole platinum
[1] A. De Cian, L. Lacroix, C. Douarre, N. Temime-Smaali, C. Trentesaux, J. F. Riou, J.-L. Mergny, Biochimie 2008, 90, 131 155. [2] Y. Qin, L. H. Hurley, Biochimie 2008, 90, 1149 1171. [3] A. N. Lane, J. B. Chaires, R. D. Gray, J. O. Trent, Nucleic Acids Res. 2008, 36, 5482 5515. [4] S. Balasubramanian, S. Neidle, Curr. Opin. Chem. Biol. 2009, 13, 345 353. [5] J. L. Huppert, Biochimie 2008, 90, 1140 1148. [6] W. Palm, T. De Lange, Ann. Rev. Genetics 2008, 42, 301 344. [7] H. Tahara, K. Shin-ya, H. Seimiya, H. Yamada, T Tsuruo, T. Ide, Oncogene 2006, 25, 1955 1966. [8] R. Rodriguez, S. Mller, J. A. Yeoman, C. Trentsaux, J. F. Riou, S. Balasubramanian, J. Am. Chem. Soc. 2008, 130, 15758 15759. [9] S. Neidle, FEBS J. 2010, 277, 1118 1125. [10] S. Balasubramanian, L. Hurley, S. Neidle, Nat. Rev. Drug Discovery 2011, 10, 261 275. [11] T. Tauchi, K. Shin-ya, G. Sashida, M. Sumi, S. Okabe, J. H. Ohyashiki, K. Ohyashiki, Oncogene 2006, 25, 5719 5725.
Molecular modeling
Torsional barrier analysis: To assess the stabilization of the biphenyl torsional barrier due to the intramolecular hydrogen bond, models of the ligands were built in the Maestro 9.1 interface (Schrdinger Inc.) and optimized at the B3LYP/LACV3P** level of theory using Jaguar 7.7 (Schrdinger). The value for the NCCC torsion (with the biphenyl junction between the imidazole and the aryl substituents as a central bond) was varied between 08 and 458 at 58 intervals, and the geometry optimized at each step. Dipole moment and molecular orbital characterization: Starting from the conformation with the lowest energy, molecular orbitals and dipole moments were computed at the B3LYP/LACV3P** level of theory using Jaguar 7.7 (Schrdinger).
Biological studies
Telomeric repeat amplification protocol (TRAP-LIG) assay: For the evaluation of telomerase inhibition by the platinum(II) complexes, an adapted version of the TRAP assay[29] was used. Briefly, 5 mL 10 stock solution was added for each concentration of ligand to the extension reaction (Vt = 50 mL). Each extension reaction contained 5 mL 10 TRAP buffer (200 mm Tris pH 8.3, 15 mm MgCl2, 630 mm KCl, 10 mm EGTA pH 8.0, 0.05 % Tween-20, 1 mg mL1 BSA), 1 mL dNTPs (2.5 mm), 1 mL TS primer (20 pmol mL1), 37 mL ddH2O, and 1 mL HeLa extract (1 mg mL1). The extension reaction was incubated for 30 min at 30 8C. The reaction was then stopped, and the ligands were removed using the Nucleotide Removal Kit (Qiagen) according to the manufacturers recommendations. Bound product was eluted with 50 mL ddH2O and then vacuum evaporated for 30 min with medium heat (ThermoFisher). Product was then resuspended in 39 mL ddH2O and subsequently amplified by adding the following (Vt = 50 mL): 5 mL 10 TRAP buffer, 1 mL
www.chemmedchem.org
93
MED
[12] P. Wang, C.-H. Leung, D.-L. Ma, S.-C. Yan, C.-M. Che, Chem. Eur. J. 2010, 16, 6900 6911. [13] K. I. E. McLuckie, Z. A. E. Waller, D. A. Sanders, D. Alves, R. Rodriguez, J. Dash, G. J. McKenzie, A. R. Venkitaraman, S. Balasubramanian, J. Am. Chem. Soc. 2011, 133, 2658 2663. [14] a) D. Monchaud, M.-P. Teulade-Fichou, Org. Biomol. Chem. 2008, 6, 627 636; b) A. Arola, R. Vilar, Curr. Top. Med. Chem. 2008, 8, 1405 1415; c) T.-M. Ou, Y.-J. Lu, J.-H. Tan, Z.-S. Huang, K.-Y. Wong, L.-Q. Gu, ChemMedChem 2008, 3, 690 713. [15] S. N. Georgiades, N. H. Abd Karim, K. Suntharalingam, R. Vilar, Angew. Chem. 2010, 122, 4114 4128; Angew. Chem. Int. Ed. 2010, 49, 4020 4034. [16] a) S. Neidle, Curr. Opin. Struct. Biol. 2009, 19, 239 250; b) D.-L. Ma, C.-M. Che, S.-C. Yan, J. Am. Chem. Soc. 2009, 131, 1835 1846; c) M. A. Read, A. A. Wood, J. Harrison, S. M. Gowan, L. R. Kelland, H. S. Dosanjh, S. Neidle, J. Med. Chem. 1999, 42, 4538 4546; d) G. Parkinson, R. Ghosh, S. Neidle, Biochemistry 2007, 46, 2390 2397; e) S.-B. Chen, J.-H. Tan, T.-M. Ou, S.-L. Huang, L.-K. An, H.-B. Luo, D. Li, L.-Q. Gu, Z.-S. Huang, Bioorg. Med. Chem. Lett. 2011, 21, 1004 1009; f) G. Clark, P. Pytel, C. Squire, S. Neidle, J. Am. Chem. Soc. 2003, 125, 4066 4067; g) A. T. Phan, The FEBS J. 2010, 277, 1107 1117; h) S. Alcaro, A. Artese, G. Costa, S. Distinto, F. Ortuso, L. Parrotta, Biochimie 2011, 93, 1267 1274. [17] S. Sparapani, S. M. Haider, F. Doria, M. Gunaratnam, S. Neidle, J. Am. Chem. Soc. 2010, 132, 12263 12272. [18] D. P. Bancroft, C. A. Lepre, S. J. Lippard, J. Am. Chem. Soc. 1990, 112, 6860 6871. [19] R. Kieltyka, J. Fakhoury, N. Moitessier, H. F. Sleiman, Chem. Eur. J. 2008, 14, 1145 1154. [20] The inflection point appears to be between 1.5 and 2:1 complex 1:G4 DNA, possibly suggesting a mixture of structures due to the dynamic nature of quadruplex folding. [21] E. M. Rezler, J. Seenisamy, S. Bashyam, M. Y. Kim, E. White, W. D. Wilson, L. H. Hurley, J. Am. Chem. Soc. 2005, 127, 9439 9447.

[22] P. S. Shirude, E. R. Gillies, S. Ladame, F. Godde, K. Shin-ya, I. Huc, S. Balasubramanian, J. Am. Chem. Soc. 2007, 129, 11890 11891. [23] D. Monchaud, C. Allain, M.-P. Teulade-Fichou, Bioorg. Med. Chem. Lett. 2006, 16, 4842 4845. [24] A. De Cian, L. Guittat, M. Kaiser, B. Sacc, S. Amrane, A. Bourdoncle, P, Alberti, M.-P. Teulade-Fichou, L. Lacroix, J.-L. Mergny, Methods 2007, 42, 183 195. [25] P. A. Rachwal, K. R. Fox, Methods 2007, 43, 291 301. [26] M. Cusumano, M. L. Di Pietro, A. Giannetto, Inorg. Chem. 2006, 45, 230 235. [27] For complex 3, we noted light-sensitivity and some decomposition during the experiment, which likely resulted in atypical results. [28] D. Monchaud, C. Allain, H. Bertrand, N. Smargiasso, F. Rosu, V. Gabelica, A. De Cian, J.-L. Mergny, M.-P. Teulade-Fichou, Biochimie 2008, 90, 1207 1223. [29] J. Reed, M. Gunaratnam, M. Beltran, A. P. Reszka, R. Vilar, S. Neidle, Anal. Biochem. 2008, 380, 99 105. [30] A. De Cian, G. Cristofari, P. Reichenbach, E. De Lemos, D. Monchaud, M.P. Teulade-Fichou, K. Shin-ya, L. Lacroix, J. Lingner, J.-L. Mergny, Proc. Natl. Acad. Sci. USA 2007, 104, 17347 17352. [31] Y. Mikami-Terao, M. Akiyama, Y. Yuza, T. Yanagisawa, O. Yamada, T. Kawano, M. Agawa, H. Ida, H. Yamada, Exp. Eye Res. 2009, 89, 200 208. [32] Y. Xiong, X. Zou, W. Jianzhong, Y. Huiying, J. Liangnian, Synth. React. Inorg. Met. Org. Chem. 1998, 28, 1445 1454. [33] W. Mei, F. Sun, G. Wu, L. He, J. Li, X. Yao, X. Zhou, S. Cai, Guang Yaoxueyuan Xuebao 2004, 20, 451 452. [34] IDT SciTools OligoAnalyzer ver. 3.1: http://www.idtdna.com/analyzer/ Applications/OligoAnalyzer/ (accessed December 12, 2010).
Received: September 29, 2011 Published online on November 3, 2011
94
www.chemmedchem.org
DOI: 10.1002/cmdc.201100420
Design, Synthesis and Biological Evaluation of Trypanosoma brucei Trypanothione Synthetase Inhibitors
Daniel Spinks, Leah S. Torrie, Stephen Thompson, Justin R. Harrison, Julie A. Frearson, Kevin D. Read, Alan H. Fairlamb, Paul G. Wyatt, and Ian H. Gilbert*[a]
Trypanothione synthetase (TryS) is essential for the survival of the protozoan parasite Trypanosoma brucei, which causes human African trypanosomiasis. It is one of only a handful of chemically validated targets for T. brucei in vivo. To identify novel inhibitors of TbTryS we screened our in-house diverse compound library that contains 62 000 compounds. This resulted in the identification of six novel hit series of TbTryS inhibitors. Herein we describe the SAR exploration of these hit series, which gave rise to one common series with potency against the enzyme target. Cellular studies on these inhibitors confirmed on-target activity, and the compounds have proven to be very useful tools for further study of the trypanothione pathway in kinetoplastids.
Introduction
Human African trypanosomiasis (HAT), or sleeping sickness, is endemic in sub-Saharan Africa, claiming the lives of about 30 000 people every year and putting approximately 60 million people at risk of infection.[1] HAT is a progressive and fatal disease caused by the protozoan parasites Trypanosoma brucei gambiense and T. b. rhodesiense, which are transmitted to the human host by the bite of the tsetse fly. If left untreated the disease progresses to the central nervous system and is ultimately fatal. There is a clinical need for more effective drug therapies. Current therapies are toxic and have inappropriate treatment regimens for a rural African setting. There are also problems with treatment failures.[13] Differences in metabolic pathways have been discovered between parasite and host, which may be exploited for drug discovery programmes. An example of such a difference is found in thiol metabolism and the response of T. brucei to oxidative stress.[48] Studies have shown that trypanosomatid parasites are uniquely dependent on trypanothione (N1,N8-bis(glutathionyl)spermidine) as their principal thiol, in contrast to most other organisms (including their mammalian hosts) that use glutathione (g-l-glutamyl-l-cysteinylglycine, GSH).[9] In T. brucei trypanothione is synthesised from GSH and spermidine (Spd) by an ATP-dependent CN ligase, trypanothione synthetase (TryS; EC 6.3.1.9), with N1- and N8-glutathionylspermidine as intermediates.[10, 11] Selective inhibition of the trypanothione pathway with chemical agents (targeting trypanothione reductase, tryparedoxin, and tryparedoxin peroxidise) or classical gene knockout studies have shown a clear trypanocidal effect.[12, 13] TbTryS has also been genetically validated as a drug target, with RNAi and gene knockout studies confirming that TbTryS is essential for T. brucei growth in both bloodstream and procyclic forms, and that there is no alternative bypass mechanism available to the parasite.[14, 15] Before commencing a drug discovery programme, TbTryS was assessed for its suitability as a drug target using the traffic light scoring system that we have developed in house.[16] The assessment indicated TbTryS is an attractive target for drug development, especially as it is unlikely to have resistance or toxicity issues, as there is no obvious bypass metabolism or equivalent enzyme in humans.[10] The main concern was the potential druggability of the target. Because the active site of TbTryS is large enough to accommodate trypanothione and precursors, this may be an issue if the active site is a large featureless pocket, as is observed in T. brucei trypanothione reductase (TbTryR).[17] However, the structure of TryS from Leishmania major suggests this is not the case,[18] and the potential to cocrystallise ligands with the protein to inform a chemistry programme was a distinct advantage. Importantly, TbTryS is a bifunctional enzyme, which catalyzes the biosynthesis and hydrolysis of the GSHSpd adduct trypanothione. The two catalytic domains are separate in Leishmania. The N-terminal domain is a cysteine-containing amidohydrolase/peptidase amidase site, with the C-terminal ATP grasp domain responsible for the synthetase activity of the enzyme.[18] Figure 1 shows the only previously disclosed inhibitor of TbTryS, compound 1.[19] Whilst a valuable tool molecule, the optimisation and development of this phosphinate inhibitor into a potential clinical candidate is limited due to the peptidic nature of such a compound, with a high polar surface area
[a] D. Spinks, Dr. L. S. Torrie, Dr. S. Thompson, Dr. J. R. Harrison, Prof. J. A. Frearson, Dr. K. D. Read, Prof. A. H. Fairlamb, Prof. P. G. Wyatt, Prof. I. H. Gilbert Drug Discovery Unit, Division of Biological Chemistry & Drug Discovery College of Life Sciences, University of Dundee, Dundee DD1 5EH (UK) E-mail: i.h.gilbert@dundee.ac.uk Re-use of this article is permitted in accordance with the Terms and Conditions set out at http://onlinelibrary.wiley.com/journal/ 10.1002/(ISSN)1860-7187/homepage/2452_onlineopen.html.
95
MED
Table 1. Hit series clusters identified from HTS. Series Structure
I. H. Gilbert et al.
LE[a]
1a Figure 1. Previously published TbTryS inhibitor: compound 1, K 1200 nm.

app i
0.44
: 500
(PSA), and charges at physiological pH (which are detrimental for cellular penetration, metabolic stability, bioavailability, and bloodbrain barrier permeability). Herein we describe a medicinal chemistry programme to develop inhibitors of TbTryS, which gave rise to some potent compounds. We recently reported the biological experiments to chemically validate TbTryS using five of these compounds: 9, 20, 71, 84, and 89.[15, 20]
1b
0.45
1c
0.33
2a
0.33
Results and Discussions

High-throughput screening As previously described, a high-throughput screening assay for TbTryS was developed and validated.[20] TbTryS was then screened against a 62 000 compound diversity set at singlepoint concentration (30 mm). This gave rise to > 720 hits (compounds with TbTryS percentage inhibition > 33 % at 30 mm). These hits were clustered and filtered down to 174 compounds that underwent potency testing (10-point, half-log dilution doseresponse curves from which an accurate IC50 could be calculated). This gave rise to the six putative hit series, plus a number of singletons. Appropriate hits were then re-purchased to confirm identity and activity. A round of purchasing (where available) and synthesis of analogues was initiated to validate the series and to investigate the SAR around these potential hit series to assess optimisation towards lead series. Table 1 shows the hit series identified from the HTS screen, and the ligand efficiencies of compounds from these series.[21] Although there are six distinct chemical series shown in Table 1, the series can be clustered into two distinct groups that share common pharmacophoric features. Attempts were made to co-crystallise hit ligands in the protein to obtain an Xray crystal structure showing ligands in the binding domain, but unfortunately these have been unsuccessful so far. Initial hit exploration around series/group 1 Group 1 series were identified as having a common pharmacophore with two hydrogen bond acceptors (HBA) in 1,5 relationship, with one of the HBAs from a heterocyclic ring. The pharmacophore also includes regions of hydrophobicity, indicating a potential lipophilic pocket. Hit series 1 a: thiazole methylene sulfone The synthetic route to prepare this series involved the condensation of a thioamide with an appropriate a-bromoketone
2b 0.30
2c
0.29
[a] Ligand efficiency, calculated as 0.6 ln(IC50)/(heavy atom count).[21]
Table 2. SAR around hit series 1 a.
Compd 3 4 5 6 7 8
R H 2-CH3 2-F 3-OCH3 3-CF3 3-Cl
IC50 [mm][b] 1.2 3.2 1.1 12.0 1.1 0.4
Compd 9 10 11 12 13
R 3-F 4-CH3 4-CF3 3,4-di-Cl 3,4-di-F
IC50 [mm][b] 0.4 > 100 > 100 2.5 0.4
[a] Yields: 6087 %; the synthesis of compound 9 was previously described by Torrie et al.[20] [b] IC50 values against TbTryS.
(Table 2), based on the methodology of Dunn et al.[22] The corresponding a-bromoketones could be bought or synthesised from the corresponding acetyl compound (see Experimental Section below for more details). Activities of hit series 1 a compounds against TbTryS are listed in Table 2. Exploration around the scaffold of 1 a (Table 2) shows that 2substitution with a small substituent, methyl (compound 4) and fluorine (compound 5), is tolerated without loss of potency. An electron-donating substituent at the 3-position, as in
96
www.chemmedchem.org
Trypanothione Synthetase Inhibitors compound 6, gave a 10-fold loss in potency relative to unsubstituted compound 3; however, all potency was regained if the 3-position group was changed to an electron-withdrawing group, such as trifluoromethyl (compound 7). In addition, a slight improvement in potency was observed for the 3-F (9) and 3-Cl (8) derivatives (IC50 : 0.4 mm). These analogues showed sub-micromolar potency is achievable within this hit series. Substitution at the 4-position was not tolerated with either an electron-withdrawing or -donating group. Both methyl (10) and trifluoromethyl (11) were found to be inactive. However, 3,4-disubstitution with an electron-withdrawing substituent at the 3-position restored potency (compounds 12 and 13), although ligand efficiency was decreased. Hit series 1 b: tetrazole methylene carbonyl In this series, the thiazole subunit is replaced by a tetrazole, and the sulfone by a carbonyl group. The tetrazoles were alkylated with a-bromoketones (Table 3). Amide analogue 23 was synthesised via an amide coupling of the corresponding acid, which in turn was obtained from saponification of the ethyl ester.
Table 3. SAR around hit series 1 b.
bulk in this area is required. In contrast, the phenyl ketone 24 shows a marked loss in potency, with an IC50 value of 45 mm. Hit series 1 c: tetrahydroindazole methylene amide The pyrazole core for series 1 c was made by condensation of ethyl hydrazinoacetate with the commercially available dione. The ester was then hydrolysed, allowing amide couplings using standard methodology (Table 4).[23, 24]
Table 4. SAR around hit series 1 c.
Compd 25 26 27 28 29 30 31 32 33 34 35
R1 H H H H H H H H H CH2Ph CH2CH3
R (+/)-CH(CH3)Ph CH2-4-chlorophenyl CH2-2-thiophene CH2CH2Ph (+/)-CH(CH3)CH2CH3 CH2CH2CH3 CH2CH2OCH3 Ph cHex CH2Ph CH2CH3
TbTryS IC50 [mm] 5.6 3.0 3.9 5.6 12 13 30 > 100 > 100 > 100 15
Compd 14 15 16 17 18 19 20 21 22 23 24
R H 4-CF3 3-Cl, 4-CH3 3-F 3-CF3 3,5-di-F 3,5-di-Cl 3,5-di-Cl 3,5-di-Cl 3,5-di-Cl 3,5-di-F
R1 tBu tBu tBu tBu tBu tBu tBu CH3 CH2CH3 N(CH3)2 Ph
TbTryS IC50 [mm] 8.8 > 100 7.2 1.5 3.5 1.9 0.3 6.5 2.7 0.6 45
[a] Conditions: 1) ethyl 2-hydrazinylacetate, Et3N, EtOH, reflux, 61 % yield; 2) 2 m NaOH, THF/H2O (2:1), 89 % yield; 3) RR1NH, EDCHCl, HOBt, DIPEA, DMF, 50 8C; yields: 7980 %.
[a] Conditions: NaI, K2CO3, DMF, 90 8C; yields: 4486 %; the synthesis of compound 20 was previously described by Torrie et al.[20]
Small changes to optimise the amide group in this series generated a flat SAR plateau (2528; Table 4). Aliphatic amides (2931) were tolerated, albeit at weaker potency, but bulky rigid substitution with the phenyl 32 and cyclohexyl 33 moieties was not tolerated, with complete loss in potency observed. Tertiary amide 34 with two bulky benzyl groups also lost all potency, but the less bulky diethyl amide 35 retained some inhibitory activity. Initial hit exploration around series/group 2 Following the HTS campaign we also identified three hit series based around a second pharmacophore. This second group of hit series described a slightly different pharmacophore, with the heteroatom of the heterocycle in a 1,6 relationship with the HBA motif of the sulfonamide. These hit series had higher molecular weight and generally lower potency, with lower ligand efficiency (see Table 1). To explore the SAR around these series, compounds were purchased and screened. Data for these are listed in Tables 5, 6, and 7. None of the group 2 series offered any benefits in terms of potency over the compound series from group 1. In addition, the DMPK data indicated less favourable properties of group 2
Subtle trends in SAR similar to series 1 a were observed in series 1 b (Table 3). In particular, there was a 25-fold gain in potency for the 3,5-dichloro analogue 20, relative to the unsubstituted scaffold 14. This could be due to the greater lipophilicity of this compound causing an increase in nonspecific binding, although the ligand efficiency improved (LE from 0.39 to 0.45), suggesting a favourable specific interaction. It was possible to modify the ketone moiety to an amide without loss of potency (compare amide 23 (IC50 = 0.6 mm), with ketone 20 (IC50 = 0.3 mm)). In going from methyl (21) to ethyl (22) to tertbutyl ketone (20), an improvement in potency was observed (IC50 : 8.8 to 6.5 to 0.3 mm, respectively), indicating lipophilic
www.chemmedchem.org
97
MED
Table 5. SAR around hit series 2 a.
I. H. Gilbert et al. gen atoms becomes the HBA, and the other is used for attachment of the other HBA. This indazole series is also predicted to have reasonable physicochemical properties: low molecular weight, clogP < 5, and low PSA. For example, compound 60 (Table 8) has Mr = 292 Da, clogP = 4.0, and PSA = 35 2.
Compd 36 37 38 39 40 41 42
R 2-thiophene 5-(2,4-dimethylthiazole) 4-(1,3,5-trimethylpyrazole) 1-naphthyl 2,5-dimethoxyphenyl 3-(2,5-dimethylthiophene) tBu
TbTryS IC50 [mm] 8.5 2.2 2.6 > 100 > 100 24 37
Figure 2. Group 1 common pharmacophore: R and R1 are hydrophobic binding domains; N and O (boxed) are the hydrogen bond acceptor motifs (1,5 relationship); X, Y, and Z are C, N, O, or S; Q is C or SO (sulfone).
Table 6. SAR around hit series 2 b. Figure 3. Generation of the new scaffold.
Validation of the indazole series

Compd 43 44 45 46 47 48 49 50 51 52 R2 H H H H H H H H H CH3 R1 H 2-CH3 2-Cl H H H 2-CH3 2-CH3 2-Cl H R H H H 2,5-di-Cl 4-NHCOCH3 4-OCH3 4-OCH3 4-NHCOCH3 4-NHCOCH3 4-NHCOCH3 TbTryS IC50 [mm] > 50 15 25 16 5.9 11 4.0 1.7 3.7 > 50
The indazole scaffold was prepared as shown in Scheme 1. Indazole was first iodinated (at the 3-position) using standard conditions.[25] A Suzuki reaction afforded the 3-aryl intermedi-
Table 8. SAR around lead series 3.
Compd Table 7. SAR around hit series 2 c. 59 60 61 62 63 64 65 66 67 68 69 70 71[b]
R Ph Ph 4-chlorophenyl 4-indole 4-(N-methyl)indole 2-naphthyl 4-pyridyl 2-furanyl 3-pyridyl 3-thiophenyl 4-(N-isobutyl)pyrazole 3-chlorophenyl 3-fluorophenyl
R1 OCH3 tBu tBu tBu tBu tBu tBu tBu tBu tBu tBu tBu tBu
TbTryS IC50 [mm] 20 0.15 > 100 3.7 > 100 > 100 0.3 1.9 3.0 0.4 > 100 0.35 0.09
LE[a] 0.32 0.43 0.30
Compd 53 54 55 56 57 58
R H H H H H Ph
R1 2,5-difluorophenyl phenyl-4-COCH3 2,5-dichlorophenyl 3,4-dichlorophenyl 2-naphthyl 4-methylphenyl
TbTryS IC50 [mm] 6.1 8.7 6.3 5.8 2.4 > 100
0.41 0.38 0.35 0.42 0.39 0.42
[a] Ligand efficiency, calculated as 0.6 ln(IC50)/(heavy atom count).[21] [b] Compound 71 reported previously.[20]
series compounds, relative to data for compounds from group 1 (see Table 12 below). Therefore, further optimisation work was focussed on the group 1 series. Hit to lead optimisation strategy Hit series 1 ac were successfully validated and shared an overlapping pharmacophore for TbTryS activity (Figure 2), demonstrating clear SAR and a visible potential for further optimisation. The pharmacophoric features of the three group 1 series scaffolds were hybridised into one new core scaffold, which was based on an indazole (Figure 3). One of the indazole nitro-
ate. Alkylation of the N1 nitrogen atom with the required achloroketone (or arylmethylchloride) gave final products 60, 71, and 7579. Alternatively for compounds 6170, the alkylation of the 3-iodoindazole was completed first and the Suzuki reaction, to add the aryl substituent, was employed as a final step. 3-Indazole substitution Table 8 shows the data for key compounds from the initial SAR study, describing changes to the aromatic group at position 3
98
www.chemmedchem.org
Trypanothione Synthetase Inhibitors tion from which the pendant HBA is attached to the core scaffold was also investigated. In compounds 72 and 73 the sulfone HBA is appended from the 3-position of the indazole core (Table 9). This modification had little effect on activity, which is probably due to the symmetry of the molecule. For ease of synthesis, further HBA modifications were investigated at the N1 position of the indazole. Approaches were made to use a heterocycle as the second HBA motif (Table 10). Pyridine and oxazole were investigated, but showed lower activity than the ketone or sulfone groups.
Scheme 1. General synthesis of indazole analogues. Conditions: 1) KOH, I2, DMF, as in Edwards et al.,[25] 89 % yield; 2) ArB(OH)2, Na2CO3, DME, EtOH, H2O, yields: 5078 %; 3) NaH, DMF, a-haloketone (or other), yields: 3762 %. Synthesis of compounds 71 and 89 were described previously.[20] Table 10. SAR around pendant heterocyclic side chain indazoles.
of the indazole (depicted R in the structure). The initial compound 60 (R = phenyl), had an IC50 value of 150 nm in the TbTryS enzyme assay (LE calculated as 0.43) and was the most potent compound to date. Para substituents led to loss in activity (e.g. 61, 63, 64, and 69). This is possible evidence for the presence of a tight binding pocket into which the aromatic subunit sits. A variety of heterocycles were tolerated, although some gave a 10-fold loss in activity, such as the 2-furanyl compound 66 and 3-pyridyl compound 67. The 3-fluorophenyl motif (71, IC50 = 90 nm, LE = 0.42) was more potent than 3chlorophenyl (70, IC50 = 354 nm), and was used as the standard template for exploration of substituent SAR around the R1 position (see Tables 911 below). Having established the indazole as a potent heterocyclic core with good ligand efficiency, a number of SAR studies were carried out to explore chemical optimisation around this scaffold.
Compd 74 75 76 77 78 79
R SO2tBu 2-pyridyl 3-pyridyl 2,5-dimethyl-4-oxazole 5-phenyl-3-isoxazole 5-methyl-3-isoxazole
TbTryS IC50 [mm] 0.12 2.9 4.5 28 > 100 > 100
LE[a] 0.40 0.33 0.32 0.26
Amide HBA analogues Alkylation of the 3-arylindazole (synthesis shown in Scheme 2) with ethylbromoacetate, followed by saponification, yielded the indazole-N-acetic acid intermediate. This was coupled under standard amide coupling conditions to the appropriate amine to give access to amide analogues.
Investigation of alternative HBA moieties We were concerned that the ketone group could react with nucleophiles, so a number of alternative HBAs were investigated. The initial sulfone analogue 74 (TbTryS IC50 = 120 nm) was found to be equipotent to the ketone 60 (Table 9). The posi-
Scheme 2. Synthesis of amide analogues. Conditions: 1) NaH, DMF, ethyl bromoacetate, as in Fujimura et al.,[29] yields: 3762 %; 2) NaOH, THF/H2O, 95 % yield; 3) HOBt, EDC, DMF, DIPEA, R1R2NH; yields: 4048 %.
Table 9. SAR around indazole sulfone analogues.
Compd 72 73 74
R H 4-fluoro
TbTryS IC50 [mm] 0.22 0.08 0.12
LE[a] 0.40 0.41 0.40
Encouragingly, the ketone subunit could be replaced by a simple amide without loss of potency or ligand efficiency (Table 11). Thus the diethyl and dimethyl amides (87 and 88) had IC50 values similar to that of the ketone 71, with compound 88 having a ligand efficiency of 0.44. Potency was retained when fusing the dialkyl amide into an N-piperidine amide (89 TbTryS IC50 = 135 nm), but activity was completely abolished with a larger alkyl aromatic substituent (e.g. 86 IC50 > 100 mm). Addition of an appended basic amine, to potentially pick up the TbTryS endogenous substrate Spd binding domain interactions and improve ligand potency, was investigated. This basic group could also improve the aqueous solubility of our compounds and lower the lipophilicity (logP). Although the pipera-
www.chemmedchem.org
99
MED
Table 11. Amide side chain indazoles.
I. H. Gilbert et al. Cell potency and DMPK parameters for key compounds
Following the discovery of promising novel TryS inhibitors, compounds were progressed into a trypanosomal cell proliferation assay and a human cell counter-screen. Selected compounds were also screened in an Compd R1 R2 TbTryS IC50 [mm] LE[a] in vitro metabolic clearance 80 (CH2)3-N-methylpiperazine H 0.83 0.28 assay (Table 12), to assess suitaH 0.35 0.32 81 (CH2)3-N1-imidazole H 2.9 0.22 82 (CH2)2-N-Boc-piperazine bility for series progression. Met83 (CH2)2N(CH3)2 H 0.145 0.38 abolic stability studies using (CH2)3N(CH3)2 H 0.045 0.39 84[b] pooled human liver microsomes CH3 0.17 0.35 85 (CH2)3N(CH3)2 indicate a range of stabilities. H > 100 86 CH2-4-chlorophenyl CH2CH3 0.20 0.39 87 CH2CH3 Compounds from the group 2 CH3 0.11 0.44 88 CH3 series are highly metabolically fused into piperidine ring 0.14 0.38 89[b] unstable. However, compounds H 0.86 0.29 90 (CH2)2-N-methylpiperazine 8, 26, and 71 are reasonably [21] [a] Ligand efficiency, calculated as 0.6 ln(IC50)/(heavy atom count). [b] Compounds 84 and 89 reported metabolically stable, suggesting [15, 20] previously. nothing fundamentally problematic with these scaffolds in terms of cytochrome P450-driven metabolism. zine-containing compounds 90 and 80 lost potency (ninefold Table 12 also shows cell data for key compounds from severrelative to 71), and were less efficient binders (ligand efficienal of the various series of TbTryS inhibitors discovered. Alcies of 0.29 and 0.28 respectively), they were still sub-micromothough these compounds were not toxic to the MRC5 mamlar inhibitors of the TbTryS enzyme, with TbTryS IC50 values of malian cell line, there was up to a 100-fold decrease in going 0.86 and 0.83 mm, respectively. If the second basic amine from enzyme to trypanosomal cell efficacy, even with the lead centre was removed and the compounds were truncated to compounds 71 and 84 (TbTryS IC50 : 90 and 45 nm, respectivemake the C2- and C3-linked dimethyl amine compounds 83 and 84, a significant improvement in potency (six- and 18-fold, ly). While these cell potencies are equivalent to the drugs eflorrespectively) over the analogous piperazines was observed. nithine (22 mm) and nifurtimox (2 mm) currently in clinical use These compounds were also significantly more efficient bindfor late-stage human African trypanosomiasis, they are much ers than compounds 80 and 90, with respective ligand efficienless potent than the alternative arsenic-containing drug melarcies of 0.38 and 0.39. The C3-linked dimethylamine compound soprol (8 nm), although arsenic-based compounds do show 84 was observed to be the most potent compound to date, significant toxicity in the clinic.[26] with a TbTryS IC50 value of 45 nm. Compound 84 shows imAs this large potency shift between the enzyme IC50 values proved physicochemical properties over compound 60, espeand parasite EC50 values was unexpected, further experiments cially in decreased lipophilicity, with a clogP value of 2.8 were carried out to confirm whether hit compounds were en(1.3 units lower than that of 60), and Mr = 354 Da and PSA = tering the cell and acting on-target. As described fully elsewhere, exposing T. brucei parasites to the model TbTryS inhibi50 2. As compound 84 shows similar potency to an analogue not tors 89 and 84 (2 EC50 for 72 h) resulted in trypanothione levels dropping to < 10 % of wild-type levels.[15, 20] In addition, containing an appended amine (71, IC50 90 nm) it is unlikely that compound 84 has picked up the Spd binding domain. there was a corresponding increase in the TbTryS substrate This conclusion is supported by competition binding studies which revealed that the comTable 12. Trypanosome cell, human (MRC5) cell, and pooled human microsomal intrinsic clearance (Cli) data pounds displayed mixed inhibifor key (representative) compounds from series 13. tion with respect to Spd, and did not show classical competiTb EC50 [mm] MRC5 EC50 [mm] Cli [mL min1 g1] Compd Series TbTryS IC50 [mm] tive binding kinetics.[20] Finally, 8 1a 0.41 22 > 50 0.9 capping the NH group of the 26 1c 14 14 > 50 3.0 amide of 84 with a methyl 36 2a 8.5 32 > 50 9.1 57 2c 2.4 19 > 50 17 group (compound 85) resulted 71 3 0.09 7.9 > 50 1.6 in a fourfold decrease in poten84 3 0.045 5.4 20 ND[a] cy, and the C3-linked imidazole 89 3 0.14 5.1 > 50 ND[a] 81 was eightfold less potent [a] Not determined. than the dimethylamine.
100
www.chemmedchem.org
Trypanothione Synthetase Inhibitors GSH, providing strong evidence that these compounds were acting on-target. As previously reported, the on-target effects of these hit compounds were further confirmed by generating TbTryS single knockout (SKO) and TbTryS overexpressing (OE) cell lines. Western blot analysis and densitometry demonstrated that TbTryS protein levels were decreased in the SKO cells and elevated in the OE cell line, relative to wild-type cells, and these cell lines showed the expected changes in potency to 89 (EC50 values: 20.4, 6.9, and 44.5 mm for wild-type, SKO, and OE cell lines, respectively) and 84 (EC50 values: 7.1, 1.2, and 23.3 mm for wild-type, SKO, and OE cell lines, respectively), confirming that TbTryS is the specific target of these compounds.[15, 20]
(s), broad singlet (bs), doublet (d), triplet (t), quartet (q), multiplet (m) or combination thereof. LCMS analyses were performed with either an Agilent HPLC 1100 series instrument connected to a Bruker Daltonics MicrOTOF, or an Agilent Technologies 1200 series HPLC connected to an Agilent Technologies 6130 quadrupole LC MS; both instruments were connected to an Agilent diode-array detector. LCMS chromatographic separations were conducted with a Phenomenex Gemini C18 column, 50 3.0 mm, 5 mm particle size; mobile phase, H2O/CH3CN + 0.1 % HCOOH 80:20!5:95 over 3.5 min, and then held for 1.5 min; flow rate: 0.5 mL min1. Highresolution electrospray MS measurements were performed on a Bruker Daltonics MicrOTOF mass spectrometer. Thin-layer chromatography (TLC) was carried out on Merck silica gel 60 F254 plates using UV light and/or KMnO4 for visualisation. TLC data are given as the Rf value with the corresponding eluent system specified in brackets. Column chromatography was performed using RediSep 4 or 12 g silica pre-packed columns. LCMS chromatographic separations were conducted with a Waters Xbridge C18 column, 50 mm 2.1 mm, 3.5 mm particle size; Method A: mobile phase, H2O/ CH3CN + 0.1 % NH3 ; linear gradient 80:20!5:95 over 3.5 min, and then held for 1.5 min; flow rate 0.5 mL min1. All reactions were carried out under dry and inert conditions, unless otherwise stated.
Conclusions
In this work we successfully took HTS hits, clustered them into putative hit series, and rationalised their activities based on common pharmacophores. Initial investigation of SAR around the hit series confirmed an overlapping pharmacophore, and the optimisation potential of group 1 hit series in particular. Following the SAR on group 1 series, a hybridisation strategy and scaffold-hopping approach led us to discover the indazole lead series. Optimisation of this series for potency and improved DMPK properties led to compounds 71 and 84, which displayed in vitro enzyme potencies > 10-fold improved over the best HTS hits. Attempts so far to co-crystallise our inhibitors with the TbTryS enzyme have failed to produce robust data. Although these indazoles inhibit TbTryS with IC50 values of < 100 nm, they failed to show sub-micromolar potency in a trypanosome proliferation assay. This can be rationalised by the observation that parasites can survive with low levels of trypanothione beyond the timeframe of the standard whole-parasite proliferation assay. The extension of the time-course in screening assay format is prohibited by the need for repeated dilutions of samples to remain in log-phase growth, leading to unacceptable variability. The lead compounds do, however, show a robust biochemical effect in T. brucei, and are proven to act on-target, inhibiting TbTryS in cells.[15, 20] The current lead compounds could also prove very useful in combination therapy with known trypanocides (such as melarsoprol), as studies have revealed TryS-depleted T. brucei procyclics are significantly more susceptible to trypanocides.[27] Our compounds are the most advanced, potent, and drug-like (as predicted by physicochemical and in vitro DMPK properties) inhibitors of TbTryS reported to date, and are extremely useful leads to further explore the trypanothione pathway in kinetoplastids.
Compounds in series 1 a:
2-(tert-Butylsulfonylmethyl)-4-(3-fluorophenyl)thiazole (9): Synthesis previously described.[20] To a stirred solution of 3-fluoroacetophenone (250 mg, 1.81 mmol) in THF (6 mL), was added trimethylphenylammonium tribromide (681 mg, 1.81 mmol) solution in THF (4 mL). The reaction was stirred at room temperature for 18 h; the resulting white precipitate was filtered off, and the filtrate was added to petroleum ether (PE; 20 mL). The PE solution containing the product was washed with H2O (30 mL) and then dried (MgSO4). The solvent was then removed in vacuo to give intermediate 2bromo-1-(3-fluorophenyl)ethanone (390 mg, 99 %) as a paleyellow oil; [M + H] + = 217/219. To a stirred solution of 2-bromo-1(3-fluorophenyl)ethanone (326 mg, 1.8 mmol) in EtOH (20 mL) was added 2-(tert-butylsulfonyl)ethanethioamide (386 mg, 1.98 mmol), and the reaction was heated at reflux and stirred for 2 h. The solvent was then removed in vacuo to give a crude residue which was purified by column chromatography (CH2Cl2 !10 % MeOH/ CH2Cl2 eluent) to give the title compound 9 (380 mg, 81 %) as a white solid: 1H NMR ([D6]DMSO, 500 MHz): d = 8.40 (1 H, bs), 8.31 (2 H, m), 7.70 (2 H, m), 5.15 (2 H, bs), 1.4 (9 H, bs); [M + H] + = 314; HRMS C14H16FNO2S2 calcd: 314.0679, obsd: 314.0690. Compounds 38 were prepared according to the same procedures as described above for the preparation of compound 9, using the corresponding acetophenone.
1 H NMR 2-((tert-Butylsulfonyl)methyl)-4-(phenyl)thiazole (3): ([D6]DMSO, 500 MHz): d = 8.23 (1 H, s), 7.98 (2 H, dd, J = 1.4, 8.5 Hz), 7.46 (2 H, t, J = 7.5 Hz), 7.37 (1 H, t, J = 12.5 Hz), 5.07 (2 H, bs), 1.39 ppm (9 H, bs); [M + H] + = 296; HRMS C14H17NO2S2 calcd: 296.0767, obsd: 296.0773.
Experimental section
Chemistry
H NMR spectra were recorded on either Bruker Avance DPX 500 or Bruker Avance 300 spectrometers. Chemical shifts (d) are expressed in ppm. Signal splitting patterns are described as singlet
1
2-((tert-Butylsulfonyl)methyl)-4-(2-methylphenyl)thiazole (4): H NMR ([D6]DMSO, 500 MHz): d = 7.88 (1 H, m), 7.59 (1 H, m), 7.31 (3 H, m), 5.06 (2 H, bs), 2.42 (3 H, s), 1.38 ppm (3 H, s); [M + H] + = 310; HRMS C15H19NO2S2 calcd: 310.0930, obsd: 310.0943.
2-((tert-Butylsulfonyl)methyl)-4-(2-fluorophenyl)thiazole (5): H NMR ([D6]DMSO, 500 MHz): d = 8.10 (2 H, m), 7.44 (1 H, m), 7.35 (2 H, m), 5.10 (2 H, m), 1.39 ppm (9 H, s); [M + H] + = 314; HRMS C14H16FNO2S2 calcd: 314.0679, obsd: 314.0690.
www.chemmedchem.org
101
MED
2-((tert-Butylsulfonyl)methyl)-4-(3-methoxyphenyl)thiazole (6): 1 H NMR ([D6]DMSO, 500 MHz): d = 8.26 (1 H, s), 7.55 (1 H, m), 7.52 (1 H, m), 7.38 (1 H, t, J = 8 Hz), 6.95 (1 H, m), 5.08 (2 H, s), 3.82 (3 H, s), 1.39 ppm (9 H, s); [M + H] + = 326; HRMS C15H19NO3S2 calcd: 326.0879, obsd: 326.0888. 2-(tert-Butylsulfonylmethyl)-4-(3-trifluoromethylphenyl)thiazole (7): 1H NMR ([D6]DMSO, 500 MHz): d = 8.49 (1 H, bs), 8.30 (2 H, m), 7.74 (2 H, m), 5.12 (2 H, bs), 1.4 ppm (9 H, bs); [M + H] + = 364; HRMS C15H16F3NO2S2 calcd: 364.0647, obsd: 364.0639. 2-((tert-Butylsulfonyl)methyl)-4-(3-chlorophenyl)thiazole (8): 1 H NMR ([D6]DMSO, 500 MHz): d = 8.38 (1 H, bs), 8.03 (1 H, m), 7.95 (1 H, m), 7.51 (1 H, m), 7.44 (1 H, m), 5.09 (2 H, bs), 1.39 (9 H, s); 13 C NMR (125 MHz, CDCl3): d = 23.8, 52.3, 61.6, 116.9, 124.4, 126.6, 128.4, 130.1, 134.9, 135.6, 154.0, 156.9 ppm; [M + H] + = 330/332; HRMS C14H16ClNO2S2 calcd: 330.0384, obsd: 330.0384. Compounds 913 were obtained from Maybridge.
1-(5-(3,5-Difluorophenyl)-2H-tetrazol-2-yl)-3,3-dimethylbutan-2one (19): 1H NMR (CDCl3, 300 K): d = 7.72 (2 H, m), 6.94 (1 H, tt, J = 8.8 and 2.4 Hz), 5.71 (2 H, s), 1.36 ppm (9 H, s); [M + H] + = 281; HRMS C13H14F2N4O calcd: 281.1208, obsd: 281.1210. 1-(5-(3,5-Dichlorophenyl)-2H-tetrazol-2-yl)propan-2-one (21): H NMR (CDCl3, 300 K): d = 8.08 (2 H, d, 1.9 Hz), 7.48 (1 H, t, 1.9 Hz), 5.51 (2 H, s), 2.30 ppm (3 H, s); [M + H] + = 271/273; HRMS C10H8Cl2N4O calcd: 271.0148, obsd: 271.0141.
1
1-(5-(3,5-Dichlorophenyl)-2H-tetrazol-2-yl)butan-2-one (22): H NMR (CDCl3, 300 K): d = 8.09 (2 H, d, 1.9 Hz), 7.50 (1 H, t, 1.9 Hz), 5.52 (2 H, s), 2.57 (2 H, q, J = 7.2 Hz), 1.19 ppm (3 H, t, J = 7.2 Hz); [M + H] + = 285/287; HRMS C11H10Cl2N4O calcd: 285.0304, obsd: 285.0292.
1
2-(5-(3,5-Difluorophenyl)-2H-tetrazol-2-yl)-1-phenylethanone (24): 1H NMR (CDCl3, 300 K): d = 8.02 (2 H, m), 7.72 (3 H, m), 7.58 (2 H, t, J = 7.8 Hz), 6.93 (1 H, tt, J = 8.8 and 2.4 Hz), 6.16 ppm (2 H, s); [M + H] + = 301; HRMS C11H6F2N4O calcd: 301.0895, obsd: 301.0901. Compounds 15, 16 and 18 were obtained from Maybridge.
Compounds in series 1 b:
1-(5-(3,5-Dichlorophenyl)-2H-tetrazol-2-yl)-3,3-dimethylbutan-2one (20): Synthesis previously described.[20] A mixture of commercially available 5-(3,5-dichlorophenyl)tetrazole (0.215 g, 1.0 mmol), 1-bromopinacolone (0.14 mL, 1.0 mmol) and diisopropylethylamine (0.19 mL, 1.1 mmol) in CH3CN (5 mL) was heated at 75 8C for 3 days. The mixture was cooled, diluted with EtOAc (15 mL), and filtered through a plug of silica. The filtrate was concentrated and purified by chromatography on silica (eluent: EtOAc/hexane 0! 50 %) to give the title compound 20 (0.270 g, 86 %) as a white solid: 1H NMR (CDCl3, 300 K): d = 8.08 (2 H, d, J = 1.9 Hz), 7.48 (1 H, t, J = 1.9 Hz), 5.71 (2 H, s), 1.36 ppm (9 H, s); 13C NMR (125 MHz, CDCl3): d = 26.2, 43.7, 56.6, 125.3, 130.0, 130.3, 135.7, 163.4, 204.2 ppm; [M + H] + = 313; HRMS C13H14Cl2N4O calcd: 313.0586, obsd: 313.0601. 2-(5-(3,5-Dichlorophenyl)-2H-tetrazol-2-yl)-N,N-dimethylacetamide (23): A mixture of 5-(3,5-dichlorophenyl)tetrazole (0.215 g, 1.0 mmol), NaI (0.150 g, 1.0 mmol) and K2CO3 (0.210 g, 1.5 mmol) in DMF (5 mL) was heated at 90 8C until effervescence was nearly complete. 2-Chloro-N,N-dimethylacetamide (0.11 mL, 1.1 mmol) was added, and the mixture was heated at 90 8C overnight. The mixture was diluted with H2O (15 mL) and allowed to stand for several hours. The precipitate was collected by filtration and washed with H2O and Et2O to give the title compound 23 (0.150 g, 50 %) as a white solid: 1H NMR (CDCl3, 300 K): d = 8.09 (2 H, d, J = 1.9 Hz), 7.47 (1 H, t, J = 1.9 Hz), 5.58 (2 H, s), 3.17 (3 H, s), 3.07 ppm (3 H, s); [M + H] + = 300; HRMS C11H11Cl2N5O calcd: 300.0413, obsd: 300.0411. Compounds below were prepared according to the same procedures as described above for the preparation of compound 23, using the corresponding aryl tetrazole and a-haloketone. 3,3-Dimethyl-1-(5-phenyl-2H-tetrazol-2-yl)butan-2-one (14): H NMR (CDCl3, 300 K): d = 8.17 (2 H, m), 7.51 (3 H, m), 5.70 (2 H, s), 1.35 ppm (9 H, s); [M + H] + = 245; HRMS C13H16N4O calcd: 245.1397, obsd: 245.1387.
1
Compounds in series 1 c:
N-(4-Chlorobenzyl)-2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1Hindazol-1-yl)acetamide (26): To a stirred suspension of 2-(trifluoroacetyl)cyclohexanone (1 g, 5.151 mmol) and ethylhydrazinoacetate hydrochloride (2.39 g, 15.5 mmol) in EtOH (30 mL) was added Et3N (2.17 mL, 15.5 mmol) and the reaction heated at reflux for 24 h. The reaction mixture was cooled to room temperature and the solvent removed in vacuo. The resultant crude residue was taken up in CH2Cl2, washed (H2O, brine), dried (MgSO4) and concentrated in vacuo. The resultant crude residue was purified by column chromatography, eluting with 050 % EtOAc/hexane, to give ethyl 2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1H-indazol-1-yl)acetate (864 mg, 61 %) as a white solid; [M + H] + = 277. To a stirred solution of ethyl 2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1H-indazol-1-yl)acetate (691 mg, 2.5 mmol) in 1:1 THF/H2O (30 mL) was added 2 m (aq) NaOH (5 mL, 10 mmol), and the reaction stirred for 3 h. The solvent was then removed in vacuo to give a crude residue which was suspended in H2O, acidified (1 m aq. HCl), stirred for 5 min, filtered and the precipitate dried in vacuo to give 2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1H-indazol-1-yl)acetic acid (550 mg, 89 %) as a white solid; [M + H] + = 249. To a stirred solution of 2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1Hindazol-1-yl)acetic acid (124 mg, 0.5 mmol), 4-chlorobenzylamine (0.061 mL, 0.5 mmol) and 1-hydroxybenzotriazole hydrate (68 mg, 0.5 mmol) in N,N-dimethylformamide (DMF; 3 mL) at 50 8C was added N-(3-dimethylaminopropyl)-N-ethylcarbodiimide hydrochloride (96 mg, 0.5 mmol) and N,N-diisopropylethylamine (0.18 mL, 1 mmol) and the reaction stirred for 16 h. The reaction mixture was cooled to room temperature, taken up in EtOAc, washed (H2O, brine), dried (MgSO4) and concentrated in vacuo. The resultant crude residue was purified by reversed-phase HPLC (method A), to give title compound 26 (120 mg, 79 %) as a white solid: 1H NMR ([D6]DMSO, 500 MHz): d = 8.78 (1 H, t, J = 5.9 Hz), 7.41 (2 H, m), 7.30 (2 H, m), 4.84 (2 H, s), 4.30 (2 H, d, J = 3.0 Hz), 2.56 (2 H, m), 2.51 (2 H, m), 1.74 (2 H, m), 1.67 ppm (2 H, m); 13C NMR (125 MHz, CDCl3): d = 19.9, 21.2, 21.9, 22.2, 42.8, 52.3, 116.1, 120.6, 122.7, 128.8, 128.9, 133.5, 136.0, 140.1, 166.3 ppm; [M + H] + = 372; HRMS C17H17ClF3N3O calcd: 372.1085, obsd: 372.1103.
1-(5-(3-Fluorophenyl)-2H-tetrazol-2-yl)-3,3-dimethylbutan-2-one (17): 1H NMR (CDCl3, 300 K): d = 7.98 (1 H, d, J = 7.7 Hz), 7.88 (1 H, d, J = 9.3 Hz), 7.48 (1 H, m), 7.19 (1 H, td, J = 8.3 and 2.1 Hz), 5.71 (2 H, s), 1.32 ppm (9 H, s); [M + H] + = 263; HRMS C13H15FN4O calcd: 263.1303, obsd: 263.1291.
102
www.chemmedchem.org
Trypanothione Synthetase Inhibitors

N,N-Diethyl-2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1H-indazol1-yl)acetamide (35): By proceeding in a similar manner to compound 26, using 2-(3-(trifluoromethyl)-4,5,6,7-tetrahydro-1H-indazol-1-yl)acetic acid and diethylamine the title compound 35 (100 mg, 80 %) was obtained as a white solid. 1H NMR ([D6]DMSO, 500 MHz): d = 5.10 (2 H, s), 3.37 (2 H, q, J = 7.2, 7.5 Hz), 3.27 (2 H, q, J = 7.1 Hz), 1.71 (4 H, m), 1.17 (3 H, t, J = 7.1 Hz), 1.03 ppm (3 H, t, J = 7.1 Hz); [M + H] + = 304; HRMS C14H20F3N3O calcd: 304.1631, obsd: 304.1628. Compounds 25, 27, 28, 31, 32, 33 and 34 were obtained from Asinex. Compounds 29 and 30 were obtained from ChemDiv. with argon and irradiated (Biotage Initiator) at 150 8C for 20 min. 3Fluorophenylboronic acid (126 mg, 0.9 mmol) and 2 m (aq) Na2CO3 (0.56 mL, 1.12 mmol) were added, the system degassed, flooded with argon and irradiated (Biotage Initiator) at 150 8C for 20 min. The reaction mixture was taken up in EtOAc, washed (H2O, 1 m (aq) NaOH, brine), dried (MgSO4) and concentrated in vacuo. The resultant crude residue was purified by column chromatography, eluting with 050 % EtOAc/hexane, to give 3-(3-fluorophenyl)-1H-indazole (196 mg, 62 %) as a white solid. To a stirred suspension of 60 % NaH (72 mg, 1.79 mmol) in DMF (4 mL) was added 3-(3-fluorophenyl)-1H-indazole (95 mg, 0.448 mmol) and the reaction stirred for 10 min. To the reaction mixture 1-bromopinacolone (0.24 mL, 1.79 mmol) was added, the reaction heated at 75 8C and stirring continued for 66 h. The reaction mixture was cooled to room temperature, diluted with EtOAc, washed (H2O, brine), dried (MgSO4) and concentrated in vacuo. The crude residue was purified by column chromatography, eluting with 050 % EtOAc/hexane to give title compound 71 (80 mg, 58 %) was obtained as a white solid: 1H NMR ([D6]DMSO, 500 MHz): d = 8.12 (1 H, d, J = 4.1 Hz), 7.84 (1 H, m), 7.71 (1 H, m), 7.59 (1 H, m), 7.55 (1 H, m), 7.45 (1 H, m), 7.27 (2 H, m), 5.75 (2 H, s), 1.38 ppm (9 H, s); [M + H] + = 311; HRMS C19H19FN2O calcd: 311.1554, obsd: 311.1555. By proceeding in a similar manner to compound 71, using 3-(3-fluorophenyl)-1H-indazole (or 3-phenyl-1H-indazole) and the corresponding alkylating agent, the following compounds were obtained. 1-(3-Phenyl-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (60): H NMR ([D6]DMSO, 500 MHz): d = 8.09 (1 H, d, J = 8.2 Hz), 7.97 (2 H, m), 7.54 (3 H, m), 7.43 (2 H, m), 7.25 (1 H, m), 5.73 (2 H, s), 1.28 ppm (9 H, s); 13C NMR (125 MHz, CDCl3): d = 26.4, 43.5, 53.4, 109.0, 121.2, 121.6, 122.1, 126.7, 127.6, 128.0, 128.8, 133.5, 141.8, 144.9, 207.9 ppm; [M + H] + = 293; HRMS C19H20N2O calcd: 293.1648, obsd: 293.1658.
Compounds in series 2 a:
2-Methyl-N-((5-methyl-2-phenyl-2H-1,2,3-triazol-4-yl)methyl)propane-2-sulfonamide (42): tert-Butylsulfinyl chloride (0.18 mL, 1.5 mmol) was added to a solution of (5-methyl-2-phenyl-2H-1,2,3triazol-4-yl)methanamine (0.188 g, 1.0 mmol) and Et3N (1.4 mL, 10 mmol) in CH2Cl2 (10 mL) at 0 8C. The mixture was stirred at 0 8C for 2 h and then quenched with aqueous NaHCO3. The phases were separated and the aqueous phase was extracted with CH2Cl2. The combined organic phases were dried (Na2SO4) and concentrated. Chromatography on silica (50100 % EtOAc/PE 40:60) gave 2methyl-N-((5-methyl-2-phenyl-2H-1,2,3-triazol-4-yl)methyl)propane2-sulfinamide as an orange solid (0.240 g, 82 %): 1H NMR (CDCl3, 300 K): d = 8.01 (2 H, d, J = 8.4 Hz), 7.47 (2 H, t, J = 7.9 Hz), 7.33 (1 H, t, J = 7.3 Hz), 4.50 (1 H, dd, J = 14.1 and 4.7 Hz), 4.36 (1 H, dd, J = 14.1 and 7.5 Hz), 3.60 (1 H, m), 2.42 (3 H, s), 1.27 ppm (9 H, s); [M + H] + = 293. A solution of 2-methyl-N-((5-methyl-2-phenyl-2H-1,2,3-triazol-4-yl)methyl)propane-2-sulfinamide (0.146 g, 0.5 mmol) in CH2Cl2 (7.5 mL) was treated with mCPBA (Aldrich, 77 %; 0.225 g, 1.0 mmol) and stirred at room temperature for 45 min. The reaction was quenched with 2 m aqueous NaHSO3 (10 mL) and saturated aqueous NaHCO3 (10 mL). The organic phase was applied to a plug of silica and eluted with CH2Cl2/EtOAc (0!100 %) to give title compound 42 as a white solid (0.107 g, 69 %): 1H NMR (CDCl3, 300 K): d = 8.01 (2 H, m), 7.49 (2 H, m), 7.34 (1 H, tt, J = 7.4 and 1.1 Hz), 4.51 (2 H, d, J = 5.7 Hz), 4.43 (1 H, m), 2.44 (3 H, s), 1.47 ppm (9 H, s); [M + H] + = 309; HRMS C14H20N4O2S calcd: 309.1307, obsd: 309.1312.
3-(3-Fluorophenyl)-1-(pyridin-2-ylmethyl)-1H-indazole (75): H NMR ([D6]DMSO, 500 MHz): d = 8.63 (1 H, d, J = 4.5 Hz), 8.06 (1 H, d, J = 8.5 Hz), 7.83 (1 H, m), 7.75 (1 H, m), 7.58 (1 H, m), 7.47 (2 H, m), 7.42 (1 H, m), 7.27 (1 H, m), 7.21 (1 H, m), 7.12 (1 H, m), 6.95 (1 H, d, J = 8.0 Hz), 5.83 ppm (2 H, s); [M + H] + = 304; HRMS C19H14FN3 calcd: 304.1245, obsd: 304.1258.
1
Compounds in series 2 a, 2 b and 2 c:

Compounds 3641 were obtained from Maybridge. Compounds 43, 44, 45, 53, and 54 were obtained from ChemDiv. Compound 46 was obtained from ChemBridge. Compounds 4752 were obtained from Asinex. Compounds 5557 were obtained from Life Chemicals. Compound 58 was obtained from Enamine.
3-(3-Fluorophenyl)-1-(pyridin-3-ylmethyl)-1H-indazole (76): 1 H NMR ([D6]DMSO, 500 MHz): d = 8.63 (1 H, d, J = 2.0 Hz), 8.49 (1H dd, J = 1.5, 5.0 Hz), 8.13 (1 H, d, J = 8.5 Hz), 7.87 (2 H, t, J = 8.5 Hz), 7.72 (1 H, dt, J = 2.0, 9.5 Hz), 7.69 (1 H, dt, J = 2.0, 9.5 Hz), 7.58 (1 H, m), 7.50 (1 H, m), 7.35 (1 H, dd, J = 4.5, 8.0 Hz), 7.28 (2 H, m), 5.81 ppm (2 H, s); [M + H] + = 304; HRMS C19H14FN3 calcd: 304.1245, obsd: 304.1235. 4-((3-(3-Fluorophenyl)-1H-indazol-1-yl)methyl)-2,5-dimethyloxazole (77): 1H NMR ([D6]DMSO, 500 MHz): d = 8.10 (1 H, d, J = 8.5 Hz), 7.83 (1 H, m), 7.77 (1 H, m), 7.71 (1 H, m), 7.58 (1 H, m), 7.48 (1 H, m), 7.27 (2 H, m), 5.52 (2 H, bs), 3.35 (3 H, bs), 2.36 ppm (3 H, bs); [M + H] + = 322; HRMS C19H16FN3O calcd: 322.1350, obsd: 322.1345. 3-((3-(3-Fluorophenyl)-1H-indazol-1-yl)methyl)-5-phenylisoxazole (78): 1H NMR ([D6]DMSO, 500 MHz): d = 8.15 (1 H, d, J = 8.0 Hz), 7.86 (4 H, m), 7.77 (1 H, m), 7.59 (1 H, m), 7.51 (5 H, m), 7.30 (2 H, m), 6.95 ppm (2 H, s); [M + H] + = 370; HRMS C23H16FN3O calcd: 370.1350, obsd: 370.1345.
Compounds in series 3:
1-(3-(3-Fluorophenyl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (71): The synthesis was previously described elsewhere.[20] 3-Iodo-1H-indazole was prepared as reported by Edwards et al;[25] [M + H] + = 245. A capped process vial containing 3-iodo-1H-indazole (366 mg, 1.5 mmol), 3-fluorophenylboronic acid (252 mg, 1.8 mmol), 2 m (aq) Na2CO3 (1.12 mL, 2.25 mmol) and tetrakis-(triphenylphosphine)palladium(0) in 7:3:2 DME/H2O/EtOH (3 mL) was degassed, flooded
www.chemmedchem.org
103
MED
3-((3-(3-Fluorophenyl)-1H-indazol-1-yl)methyl)-5-methylisoxazole (79): 1H NMR ([D6]DMSO, 500 MHz): d = 8.13 (1 H, d, J = 8.50), 7.86 (1 H, m), 7.80 (1 H, m), 7.74 (1 H, m), 7.59 (1 H, m), 7.51 (1 H, m), 7.28 (2 H, m), 6.08 (1 H, s), 5.80 (2 H, s), 2.32 ppm (3 H, s); [M + H] + = 308; HRMS C18H14FN3O calcd: 308.1194, obsd: 308.1179. By proceeding in a similar manner to compound 71, using the corresponding boronic acid and 1-(3-iodo-1H-indazol-1-yl)-3,3-dimethylbutan-2-one, the following compounds were obtained: 1-(3-(4-Chlorophenyl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (61): 1H NMR ([D6]DMSO, 500 MHz): d = 8.09 (1 H, d, J = 8.3 Hz), 8.00 (2 H, m), 7.59 (2 H, m), 7.55 (1 H, d, J = 8.5 Hz), 7.45 (1 H, m), 7.27 (1 H, m), 5.74 (2 H, s), 1.28 ppm (9 H, s); [M + H] + = 327/329; HRMS C19H19ClN2O calcd: 327.1259, obsd: 327.1265. 1-(3-(1H-Indol-4-yl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (62): 1H NMR ([D6]DMSO, 500 MHz): d = 11.21 (1 H, s), 8.00 (1 H, d, J = 8.2 Hz), 7.55 (2 H, m), 7.44 (3 H, m), 7.24 (2 H, m), 6.88 (1 H, m), 5.75 (2 H, s), 1.30 ppm (9 H, s); [M + H] + = 332; HRMS C21H21N3O calcd: 332.1757, obsd: 332.1754. 3,3-Dimethyl-1-(3-(1-methyl-1H-indol-4-yl)-1H-indazol-1-yl)butan-2-one (63): 1H NMR ([D6]DMSO, 500 MHz): d = 8.01 (1 H, d, J = 8.4 Hz), 7.57 (3 H, m), 7.43 (1 H, t, J = 7.5 Hz), 7.39 (1 H, d, J = 3.2 Hz), 7.33 (1 H, t, J = 7.8 Hz), 7.23 (1 H, t, J = 7.5 Hz), 6.87 (1 H, d, J = 3.0 Hz), 5.75 (2 H, s), 3.86 (3 H, s), 1.30 ppm (9 H, s); [M + H] + = 346; HRMS C22H23N3O calcd: 346.1914, obsd: 346.1913. 3,3-Dimethyl-1-(3-(naphthalen-2-yl)-1H-indazol-1-yl)butan-2-one (64): 1H NMR ([D6]DMSO, 500 MHz): d = 8.56 (1 H, s), 8.30 (1 H, d, J = 8.4 Hz), 8.14 (2 H, m), 8.06 (1 H, d, J = 8.6 Hz), 7.97 (1 H, d, 7.7 Hz), 7.57 (3 H, m), 7.47 (1 H, m), 7.31 (1 H, t, J = 7.5 Hz), 5.78 (2 H, s), 1.29 ppm (9 H, s); [M + H] + = 343; HRMS C23H22N2O calcd: 343.1805, obsd: 343.1806. 3,3-Dimethyl-1-(3-(pyridin-4-yl)-1H-indazol-1-yl)butan-2-one (65): H NMR ([D6]DMSO, 500 MHz): d = 8.69 (2 H, dd, J = 1.6, 6.1 Hz), 8.22 (1 H, d, J = 8.3 Hz), 7.97 (2 H, dd, J = 1.6, 6.2 Hz), 7.59 (1 H, d, J = 8.5 Hz), 7.48 (1 H, t, J = 7.5 Hz), 7.32 (1 H, t, J = 7.5 Hz), 5.82 (2 H, s), 1.28 ppm (9 H, s); [M + H] + = 294; HRMS C18H19N3O calcd: 294.1601, obsd: 294.1614.
1 1
1-(3-(3-Chlorophenyl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (70): 1H NMR ([D6]DMSO, 500 MHz): d = 8.11 (1 H, d, J = 8.2 Hz), 7.97 (1 H, m), 7.95 (1 H, m), 7.56 (2 H, m), 7.47 (2 H, m), 7.28 (1 H, t, J = 7.5 Hz), 5.77 (2 H, s), 1.30 ppm (9 H, s); [M + H] + = 327/329; HRMS C19H19ClN2O calcd: 327.1259, obsd: 327.1262. Compound 59 was obtained from ChemDiv. 3-(tert-Butylsulfonylmethyl)-1-(3-fluorophenyl)-1H-indazole (73): To a stirred suspension of indazole-3-carboxylic acid (5 g, 30.8 mmol) in MeOH (125 mL) was added concd H2SO4 (0.5 mL) and the reaction heated at reflux for 16 h. The reaction mixture was cooled to room temperature and the solvent removed in vacuo. The resultant crude residue was taken up in CH2Cl2, washed (saturated NaHCO3 solution), dried (MgSO4) and concentrated in vacuo to give methyl 1H-indazole-3-carboxylate (4.95 g, 91 %) as a white solid; [M + H] + = 177. To a stirred solution of methyl 1H-indazole-3-carboxylate (881 mg, 5 mmol) in CH2Cl2 (125 mL), was added 3-fluorophenylboronic acid (1.40 g, 10 mmol), pyridine (0.81 mL, 10 mmol), copper(II) acetate (1.36 g, 7.5 mmol) and molecular sieves (4 , 3.82 g) and the reaction stirred (open to air) for 46 h. The reaction mixture was then filtered through a pad of Celite and concentrated in vacuo to give a crude residue which was purified by column chromatography, eluting with 050 % Et2O/hexane, to give methyl 1-(3-fluorophenyl)-1Hindazole-3-carboxylate (332 mg, 25 %) as a white solid; [M + H] + = 271. To a stirred solution of methyl 1-(3-fluorophenyl)-1H-indazole-3-carboxylate (329 mg, 1.2 mmol) in THF (10 mL) at 0 8C, under an inert atmosphere, was added LiAlH4 (185 mg, 4.9 mmol), the reaction warmed to room temperature and stirred for 16 h. The reaction mixture was quenched (Na2SO410 H2O), filtered through a pad of Celite and concentrated in vacuo. The resultant crude residue was taken up in 3:1 THF/CH2Cl2 (12 mL), PPh3 (479 mg, 1.826 mmol) and N-chlorosuccinimide (NCS; 244 mg, 1.8 mmol) added and the reaction stirred for 24 h. Extra PPh3 (479 mg, 1.8 mmol) and NCS (244 mg, 1.8 mmol) were then added and stirring continued for 24 h. The reaction mixture was then concentrated in vacuo to give a crude residue which was taken up in EtOAc, washed (saturated NaHCO3 solution, H2O, and brine), dried (MgSO4) and concentrated in vacuo. The resultant crude residue was purified by column chromatography, eluting with 050 % Et2O/hexane, to give 3-(chloromethyl)-1-(3-fluorophenyl)-1H-indazole (143 mg, 45 %) as a strawcoloured oil; [M + H] + = 261/263. To a stirred solution of 3-(chloromethyl)-1-(3-fluorophenyl)-1H-indazole (46 mg, 0.18 mmol) in CH3CN (3 mL) was added Cs2CO3 (63 mg, 0.19 mmol) and tert-butylthiol (0.022 mL, 0.19 mmol) and the reaction stirred for 16 h. Extra Cs2CO3 (32 mg, 0.097 mmol) and tert-butylthiol (0.011 mL, 0.097 mmol) were then added and stirring continued for 6 h. The reaction mixture was taken up in CH2Cl2, washed (H2O, brine), dried (MgSO4) and concentrated in vacuo. The resultant crude residue was purified by column chromatography, eluting with 050 % Et2O/hexane, to give 3-(tert-butylthiomethyl)-1(3-fluorophenyl)-1H-indazole (45 mg, 81 %) as a colourless oil; [M + H] + = 315. To a stirred solution of 3-(tert-butylthiomethyl)-1-(3-fluorophenyl)1H-indazole (45 mg, 0.14 mmol) in CH2Cl2 (5 mL) was added 70 % mCPBA (141 mg, 0.57 mmol) and the reaction stirred for 2 h. NaHSO3 solution (2 m aq) was added, the layers separated, the organic layer washed (saturated NaHCO3 solution) and filtered through a pad of silica, eluting with CH2Cl2 and EtOAc. The filtrate was concentrated in vacuo to give a crude residue which was puriChemMedChem 2012, 7, 95 106
1-(3-(Furan-2-yl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (66): H NMR ([D6]DMSO, 500 MHz): d = 8.10 (1 H, d, J = 8.2 Hz), 7.87 (1 H, d, J = 1.7 Hz), 7.52 (1 H, d, J = 8.5 Hz), 7.56 (2 H, m), 7.44 (1 H, m), 7.26 (1 H, m), 5.72 (2 H, s), 1.27 ppm (9 H, s); [M + H] + = 283; HRMS C17H18N2O2 calcd: 283.1441, obsd: 283.1452. 1-(3-(3-Pyridyl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (67): H NMR ([D6]DMSO, 500 MHz): d = 9.17 (1 H, m), 8.63 (1 H, dd, J = 1.6, 4.8 Hz), 8.35 (1 H, dt, J = 1.9 7.9 Hz), 8.12 (1 H, d, J = 8.3 Hz), 7.56 (2 H, m), 7.47 (1 H, m), 7.29 (1 H, m), 5.77 (2 H, s), 1.28 ppm (9 H, s); [M + H] + = 294; HRMS C18H19N3O calcd: 294.1601, obsd: 294.1616.
3,3-Dimethyl-1-(3-(thiophen-3-yl)-1H-indazol-1-yl)butan-2-one (68): 1H NMR ([D6]DMSO, 500 MHz): d = 8.13 (2 H, m), 7.69 (2 H, m), 7.49 (1 H, d, J = 8.5 Hz), 7.42 (1 H, m), 7.25 (1 H, m), 5.70 (2 H, s), 1.27 ppm (9 H, s); [M + H] + = 299; HRMS C17H18N2OS calcd: 299.1213, obsd: 299.1219. 1-(3-(1-Isobutyl-1H-pyrazol-4-yl)-1H-indazol-1-yl)-3,3-dimethylbutan-2-one (69): 1H NMR ([D6]DMSO, 500 MHz): d = 8.39 (1 H, s), 8.04 (1 H, d, J = 8.1 Hz), 7.99 (1 H, s), 7.45 (1 H, d, J = 8.4 Hz), 7.39 (1 H, m), 7.19 (1 H, t, J = 7.1 Hz), 5.66 (2 H, s), 4.01 (2 H, d, J = 7.2 Hz), 2.20 (1 H, m), 1.26 (9 H, s), 0.88 ppm (6 H, d, J = 6.6 Hz); [M + H] + = 339; HRMS C20H26N4O calcd: 339.2179, obsd: 339.2183.
104
www.chemmedchem.org
Trypanothione Synthetase Inhibitors

fied by reversed-phase HPLC (method A), to give title compound 73 (35 mg, 71 %) as a white solid: 1H NMR ([D6]DMSO, 500 MHz): d = 7.97 (1 H, dt, J = 0.9, 8.1 Hz), 7.95 (1 H, dt, J = 1.0, 8.5 Hz), 7.67 (3 H, m), 7.57 (1 H, m), 7.36 (1 H, m), 7.30 (1 H, m), 5.00 (2 H, s), 1.44 ppm (9 H, s); [M + H] + = 347; HRMS C18H19FN2O2S calcd: 347.1224, obsd: 347.1236. 3-(tert-Butylsulfonylmethyl)-1-phenyl-1H-indazole (72): By proceeding in a similar manner to 3-(tert-butylsulfonylmethyl)-1-(3-fluorophenyl)-1H-indazole (see compound 73), except using phenylboronic acid, title compound 72 (35 mg, 81 %) was obtained as a white solid. 1H NMR ([D6]DMSO, 500 MHz): d = 7.96 (1 H, dt, J = 0.9, 8.1 Hz), 7.87 (1 H, dt, J = 0.9, 8.6 Hz), 7.79 (2 H, m), 7.63 (2 H, m), 7.54 (1 H, m), 7.45 (1 H, m), 7.33 (1 H, m), 5.00 (2 H, s), 1.44 ppm (9 H, s); 13 C NMR (125 MHz, CDCl3): d = 23.9, 48.0, 61.1, 110.5, 121.9, 122.3, 123.0, 124.8, 127.1, 127.7, 129.5, 135.8, 139.7, 140.1 ppm; [M + H] + = 329; HRMS C18H20N2O2S calcd: 329.1318, obsd: 329.1302. 1-(tert-Butylsulfonylmethyl)-3-(3-fluorophenyl)-1H-indazole (74): 3-(3-Fluorophenyl)-1H-indazole (see compound 71) (0.106 g, 0.5 mmol) was alkylated with tert-butyl(chloromethyl)sulfane (0.104 g, 0.75 mmol), as in accordance with reported procedures[28] to give the intermediate 1-(tert-butylthiomethyl)-3-(3-fluorophenyl)1H-indazole (0.055 g, 35 %). 1H NMR (CDCl3, 300 K) d = 8.03 (1 H, dt, J = 7.5 and 0.8 Hz), 7.79 (1 H, dt, J = 7.7 and 1.2 Hz), 7.71 (1 H, ddd, J = 10.0, 2.5 and 1.0 Hz), 7.65 (1 H, dt, J = 8.5 and 0.8 Hz), 7.51-7.47 (2 H, m), 7.29 (1 H, m), 7.12 (1 H, tdd, J = 8.5, 2.5 and 0.8 Hz), 5.68 (2 H, s), 1.34 ppm (9 H, s); [M + H] + = 315. A solution of this sulfide in CH2Cl2 (2 mL) was treated with mCPBA (Aldrich, 77 %; 0.11 g, 0.5 mmol) and stirred at room temperature for 90 min. The reaction was quenched with 2 m aqueous NaHSO3 (10 mL) and saturated aqueous NaHCO3 (10 mL). The organic phase was applied to a plug of silica and eluted with CH2Cl2/ EtOAc (0!100 %) to give the title compound 74 as a white solid (0.054 g, 89 %): 1H NMR (CDCl3, 300 K): d = 8.01 (1 H, dt, J = 8.2 and 0.9 Hz), 7.78 (1 H, dt, J = 8.5 and 0.8 Hz), 7.76 (1 H, ddd, J = 11.7, 1.4 and 1.0 Hz), 7.67 (1 H, ddd, J = 9.9, 2.5 and 1.5 Hz), 7.55 (1 H, ddd, J = 8.4, 7.0 and 1.0 Hz) 7.50 (1 H, td, J = 8.1 and 5.9 Hz), 7.34 (1 H, ddd, J = 8.2, 6.8 and 0.8 Hz), 7.15 (1 H, tdd, J = 8.4, 2.6 and 0.9 Hz), 5.76 (2 H, s), 1.43 ppm (9 H, s); 13C NMR (125 MHz, CDCl3): d = 23.4, 61.2, 64.6, 110.0, 114.4, 115.5, 121.1, 122.3, 123.2, 127.9, 130.5, 134.8, 142.2, 145.1, 161.2, 164.1 ppm; [M + H] + = 347; HRMS C18H19FN2O2S calcd: 347.1224, obsd: 347.1214. 2-(3-(3-Fluorophenyl)-1H-indazol-1-yl)-N-(3-(4-methylpiperazin-1yl)propyl)acetamide (80): To a stirred suspension of 60 % NaH (754 mg, 18.9 mmol) in DMF (20 mL) under an inert atmosphere, was added 3-(3-fluorophenyl)-1H-indazole (1.0 g, 4.7 mmol) in DMF (10 mL) and the reaction was stirred for 15 min. To the reaction mixture was added ethyl bromoacetate (2.09 mL, 18.9 mmol), the reaction was warmed to 75 8C and stirring continued for 66 h. The reaction mixture was cooled to room temperature, taken up in EtOAc, washed (H2O, brine), dried (MgSO4) and concentrated in vacuo. The crude residue was taken up in 1:1 THF/H2O (30 mL) and 2 m (aq) NaOH (9.43 mL, 18.9 mmol), the reaction was heated at 50 8C and stirred for 3 h. The reaction was cooled to room temperature, diluted with H2O, extracted (EtOAc), the aqueous layer acidified (1 m (aq) HCl, pH 2), extracted (EtOAc). The organic layer was dried (MgSO4) and concentrated in vacuo to give 2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetic acid (980 mg, 77 %) as a white foam; [M + H] + = 471. By proceeding in a similar manner to 26, except using 2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetic acid and 1-(2-aminopropyl)-4methylpiperazine, the title compound 80 (65 mg, 43 %) was obChemMedChem 2012, 7, 95 106
tained as a white solid: 1H NMR ([D6]DMSO, 500 MHz): d = 8.25 (1 H, t, J = 5.4 Hz), 8.11 (1 H, d, J = 8.4 Hz), 7.85 (1 H, d, J = 7.9 Hz), 7.72 (1 H, m), 7.68 (1 H, d, J = 8.9 Hz), 7.58 (1 H, m), 7.47 (1 H, t, J = 7.7 Hz), 7.27 (2 H, m), 5.16 (2 H, s), 3.12 (2 H, m), 2.51 (8 H, m), 2.27 (2 H, m), 2.12 (3 H, s), 1.58 ppm (2 H, m); [M + H] + = 410; HRMS C23H28FN5O calcd: 410.2351, obsd: 410.2363. By proceeding in a similar manner to 26, except using 2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetic acid and the appropriate amine, the following compounds were prepared: N-(3-(1H-Imidazol-1-yl)propyl)-2-(3-(3-fluorophenyl)-1H-indazol1-yl)acetamide (81): 1H NMR ([D6]DMSO, 500 MHz): d = 8.41 (1 H, t, J = 5.4 Hz), 8.25 (1 H, s), 8.12 (1 H, d, J = 8.3 Hz), 7.84 (1 H, d, J = 7.9 Hz), 7.72 (2 H, m), 7.58 (1 H, m), 7.48 (1 H, t, J = 7.7 Hz), 7.44 (1 H, m), 7.27 (2 H, m), 7.23 (1 H, m), 5.20 (2 H, s), 4.10 (2 H, m), 3.09 (2 H, m), 1.93 ppm (2 H, m); [M + H] + = 378; HRMS C21H20FN5O calcd: 378.1725, obsd: 378.1739. tert-Butyl-4-(2-(2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetamido)ethyl)piperazine-1-carboxylate (82): 1H NMR ([D6]DMSO, 500 MHz) d mixture of rotamers d = 8.12 (2 H, m), 7.86 (1 H, d, J = 7.6 Hz), 7.72 (2 H, m), 7.59 (1 H, m), 7.48 (1 H, m), 7.28 (2 H, m), 5.22 (2 H, s), 3.20 (2 H, m), 2.51 (8 H, m), 2.37 (2 H, m), 1.40(5 H, s), 1.39 ppm (4 H, s); [M + H] + = 482; HRMS C26H32FN5O3 calcd: 482.2562, obsd: 482.2570. N-(2-(Dimethylamino)ethyl)-2-(3-(3-fluorophenyl)-1H-indazol-1yl)acetamide (83): NMR 1H NMR ([D6]DMSO, 500 MHz): d = 8.22 (1 H, t, J = 5.5 Hz), 8.12 (1 H, d, J = 8.2 Hz), 7.85 (1 H, d, J = 7.9 Hz), 7.72 (1 H, m), 7.69 (1 H, d, J = 8.5 Hz), 7.59 (1 H, m), 7.47 (1 H, m), 7.26 (2 H, m), 5.19 (2 H, s), 3.20 (2 H, m), 2.31 (2 H, m), 2.15 ppm (6 H, s); [M + H] + = 341; HRMS C19H21FN4O calcd: 341.1772, obsd: 341.1767. N-(3-(Dimethylamino)propyl)-2-(3-(3-fluorophenyl)-1H-indazol-1yl)acetamide (84): 1H NMR ([D6]DMSO, 500 MHz): d = 8.26 (1 H, t, J = 5.3 Hz), 8.13 (1 H, d, J = 8.1 Hz), 7.85 (1 H, d, J = 7.8 Hz), 7.71 (1 H, m), 7.68 (1 H, d, J = 8.6 Hz), 7.59 (1 H, m), 7.48 (1 H, m), 7.26 (2 H, m), 5.16 (2 H, s), 3.13 (2 H, m), 2.20 (2 H, m), 2.01 (6 H, s), 1.54 ppm (2 H, m); 13C NMR (125 MHz, CDCl3): d = 24.4, 40.7, 44.5, 52.6, 59.4, 109.5, 114.2, 115.1, 121.2, 122.1, 123.0, 127.3, 130.5, 135.3, 141.7, 144.0, 162.2, 164.2, 167.3 ppm; [M + H] + = 355; HRMS C20H23FN4O calcd: 355.1856, obsd: 355.1862. N-Methyl-(3-(dimethylamino)propyl)-2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetamide (85): 1H NMR ([D6]DMSO, 500 MHz) mixture of rotamers; d = 8.12 (1 H, d, J = 8.4 Hz), 7.85 (1 H, d, J = 7.9 Hz), 7.71 (1 H, m), 7.60(2 H, m), 7.44 (1 H, m), 7.25 (2 H, m), 5.60 (1 H, s), 5.50 (1 H, s), 3.48 (1 H, m), 3.30 (1 H, m), 3.14 (1.5 H, s), 2.84 (1.5 H, s), 2.27 (1 H, m), 2.20 (3 H, s), 2.16 (1 H, m), 2.09 (3 H, s), 1.77 (1 H, m), 1.60 ppm (1 H, m); [M + H] + = 369; HRMS C21H25FN4O calcd: 369.2085, obsd: 369.2097. N-(4-Chlorobenzyl)-2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetamide (86): 1H NMR ([D6]DMSO, 500 MHz): d = 8.78 (1 H, t, J = 5.7 Hz), 8.12 (1 H, d, J = 8.1 Hz), 7.85 (1 H, d, J = 7.8 Hz), 7.71 (2 H, m), 7.59 (1 H, m), 7.48 (1 H, t, J = 7.4 Hz), 7.38 (2 H, d, J = 8.2 Hz), 7.32 (2 H, d, J = 8.3 Hz), 7.28 (2 H, m), 5.26 (2 H, s), 4.32 ppm (2 H, d, J = 6.0 Hz); [M + H] + = 394; HRMS C22H17ClFN3O calcd: 394.1117, obsd: 394.1111. 2-(3-(3-Fluorophenyl)-1H-indazol-1-yl)-N-(2-(4-methylpiperazin-1yl)ethyl)acetamide (90): 1H NMR ([D6]DMSO, 500 MHz): d = 8.12 (1 H, d, J = 8.1 Hz), 8.02 (1 H, t, J = 5.4 Hz), 7.86 (1 H, d, J = 8.1 Hz), 7.72 (1 H, m), 7.71 (1 H, d, J = 8.6 Hz), 7.59 (1 H, m), 7.47 (1 H, t, J = 7.8 Hz), 7.27 (2 H, m), 5.19 (2 H, s), 3.20 (2 H, m), 2.51 (8 H, m), 2.34
www.chemmedchem.org
105
MED
(2 H, m), 2.10 ppm (3 H, s); [M + H] + = 396; HRMS C22H26FN5O calcd: 396.2194, obsd: 396.2212. N,N-Diethyl-2-(3-(3-fluorophenyl)-1H-indazol-1-yl)acetamide (87): 3-(3-Fluorophenyl)-1H-indazole (0.106 g, 0.5 mmol), prepared as in 71, was alkylated with 2-chloro-N,N-diethylacetamide (0.102 mL, 0.74 mmol) to give the title compound 87 (0.073 g, 45 %) as a white solid: 1H NMR (CDCl3, 300 K): d = 8.02 (1 H, d, J = 8.2 Hz), 7.78 (1 H, d, J = 7.8 Hz), 7.70 (1 H, ddd, J = 10.1 2.2 and 1.5 Hz), 7.53 (1 H, d, J = 8.5 Hz), 7.50-7.45 (2 H, m), 7.27 (1 H, m), 7.11 (1 H, td, J = 7.5 and 2.0 Hz), 5.30 (2 H, s), 3.54 (2 H, q, J = 7.1 Hz), 3.43 (2 H, q, J = 7.1 Hz), 1.17 ppm (6 H, m); [M + H] + = 326; HRMS C19H20FN3O calcd: 326.1663, obsd: 326.1663. 2-(3-(3-Fluorophenyl)-1H-indazol-1-yl)-N,N-dimethylacetamide (88): 3-(3-Fluorophenyl)-1H-indazole (0.106 g, 0.5 mmol), prepared as in 71, was alkylated with 2-chloro-N,N-dimethylacetamide (0.191 g, 1.6 mmol) to give the title compound 88 (0.114 g, 77 %) as a white solid. 1H NMR (CDCl3, 300 K): d = 1 H, d, J = 8.2 Hz), 7.78 (1 H, d, J = 7.7 Hz), 7.70 (1 H, ddd, J = 10.1, 2.4 and 1.6 Hz), 7.52-7.45 (3 H, m), 7.27 (1 H, m), 7.10 (1 H, ddt, J = 8.5, 2.6 and 0.8 Hz), 5.32 (2 H, s), 3.18 (3 H, s), 3.01 ppm (3 H, s); [M + H] + = 298; HRMS C17H16FN3O calcd: 298.1350, obsd: 298.1348. 2-(3-(3-Fluorophenyl)-1H-indazol-1-yl)-1-(piperidin-1-yl)ethanone (89): Synthesis previously described.[20] 3-(3-Fluorophenyl)-1H-indazole (0.106 g, 0.50 mmol), prepared as in 71, was alkylated with 2chloro-1-piperidin-1-ylethanone (0.350 g, 2.2 mmol) to give the title compound (0.05 g, 30 %) as a white solid: 1H NMR (CDCl3, 300 K): d = 8.03 (1 H, d, J = 8.2 Hz), 7.77 (1 H, d, J = 7.8 Hz), 7.69 (1 H, dt, J = 10.0 and 1.7 Hz), 7.55 (1 H, d, J = 8.5 Hz), 7.50-7.45 (2 H, m), 7.27 (1 H, m), 7.11 (1 H, td, J = 8.4 and 1.4 Hz), 5.32 (2 H, s), 3.60 (4 H, m), 1.64 (2 H, m), 1.54 (2 H, m), 1.48 ppm (2 H, m); [M + H] + = 338; HRMS C20H20FN3O calcd: 338.1663, obsd: 338.1660.
[1] K. D. Stuart, R. Brun, S. L. Croft, A. H. Fairlamb, R. E. Gurtler, J. H. McKerrow, S. Reed, R. L. Tarleton, J. Clin. Invest. 2008, 118, 1301 1310. [2] A. H. Fairlamb, Trends Parasitol. 2003, 19, 488 494. [3] M. P. Barrett, A. H. Fairlamb, Parasitol. Today 1999, 15, 136 140. [4] N. M. El-Sayed, P. J. Myler, G. Blandin, M. Berriman, J. Crabtree, G. Aggarwal, E. Caler, H. Renauld, E. A. Worthey, C. Hertz-Fowler, E. Ghedin, C. Peacock, D. C. Bartholomeu, B. J. Haas, A. N. Tran, J. R. Wortman, U. C. Alsmark, S. Angiuoli, A. Anupama, J. Badger, F. Bringaud, E. Cadag, J. M. Carlton, G. C. Cerqueira, T. Creasy, A. L. Delcher, A. Djikeng, T. M. Embley, C. Hauser, A. C. Ivens, S. K. Kummerfeld, J. B. Pereira-Leal, D. Nilsson, J. Peterson, S. L. Salzberg, J. Shallom, J. C. Silva, J. Sundaram, S. Westenberger, O. White, S. E. Melville, J. E. Donelson, B. Andersson, K. D. Stuart, N. Hall, Science 2005, 309, 404 409. [5] A. H. Fairlamb, A. Cerami, Annu. Rev. Microbiol. 1992, 46, 695 729. [6] R. L. Krauth-Siegel, S. K. Meiering, H. Schmidt, Biol. Chem. 2003, 384, 539 549. [7] S. Mller, E. Liebau, R. D. Walter, R. L. Krauth-Siegel, Trends Parasitol. 2003, 19, 320 328. [8] K. Augustyns, K. Amssoms, A. Yamani, P. K. Rajan, A. Haemers, Curr. Pharm. Des. 2001, 7, 1117 1141. [9] A. H. Fairlamb, P. Blackburn, P. Ulrich, B. T. Chait, A. Cerami, Science 1985, 227, 1485 1487. [10] S. L. Oza, M. R. Ariyanayagam, N. Aitcheson, A. H. Fairlamb, Mol. Biochem. Parasitol. 2003, 131, 25 33. [11] M. Comini, U. Menge, L. Floh, Biol. Chem. 2003, 384, 653 656. [12] S. Krieger, W. Schwarz, M. R. Ariyanayagam, A. H. Fairlamb, R. L. KrauthSiegel, C. Clayton, Mol. Microbiol. 2000, 35, 542 552. [13] S. R. Wilkinson, D. Horn, S. R. Prathalingam, J. M. Kelly, J. Biol. Chem. 2003, 278, 31640 31646. [14] M. A. Comini, S. A. Guerrero, S. Haile, U. Menge, H. Lunsdorf, L. Floh, Free Radical Biol. Med. 2004, 36, 1289 1302. [15] S. Wyllie, S. L. Oza, S. Patterson, D. Spinks, S. Thompson, A. H. Fairlamb, Mol. Microbiol. 2009, 74, 529 540. [16] J. A. Frearson, P. A. Wyatt, I. H. Gilbert, A. H. Fairlamb, Trends Parasitol. 2007, 23, 589 595. [17] D. Spinks, E. J. Shanks, L. A. T. Cleghorn, S. McElroy, D. Jones, D. James, A. H. Fairlamb, J. A. Frearson, P. G. Wyatt, I. H. Gilbert, ChemMedChem 2009, 4, 2060 2069. [18] P. K. Fyfe, S. L. Oza, A. H. Fairlamb, W. N. Hunter, J. Biol. Chem. 2008, 283, 17672 17680. [19] S. L. Oza, S. Chen, S. Wyllie, J. K. Coward, A. H. Fairlamb, FEBS J. 2008, 275, 5408 5421. [20] L. S. Torrie, S. Wyllie, D. Spinks, S. L. Oza, S. Thompson, J. R. Harrison, I. H. Gilbert, P. G. Wyatt, A. H. Fairlamb, J. A. Frearson, J. Biol. Chem. 2009, 284, 36137 36145. [21] A. L. Hopkins, C. R. Groom, A. Alex, Drug Discovery Today 2004, 9, 430 431. [22] D. Dunn, J. Husten, M. A. Ator, S. Chatterjee, Bioorg. Med. Chem. Lett. 2007, 17, 542 545. [23] M. Mochizuki, S. Miura (Takeda Pharmaceutical Co., Ltd.), WO 2009119088, 2009. [24] D. S. Goldfarb (University of Rochester, NY, USA), US 20090163545, 2009. [25] M. L. Edwards, P. J. Cox, S. Amendola, S. D. Deprets, T. A. Gillespy, C. D. Edlin, A. D. Morley, C. J. Gardner, B. Pedgrift, H. Bouchard, D. G. L. Babin, A. Le Brun, T. N. Majid, J. C. Reader, L. J. Payne, N. M. Khan, M. Cherry (Aventis Pharmaceuticals Inc.), WO 2003035065, 2003. [26] D. C. Jones, I. Hallyburton, L. Stojanovski, K. D. Read, J. A. Frearson, A. H. Fairlamb, Biochem Pharmacol. 2010, 80, 1478 1486. [27] M. R. Ariyanayagam, S. L. Oza, M. L. Guther, A. H. Fairlamb, Biochem. J. 2005, 391, 425 432. [28] D. W. Beight, S. Mehdi, J. R. Koehl, G. A. Flynn, Bioorg. Med. Chem. Lett. 1996, 6, 2053 2058. [29] Y. Fujimura, H. Nagano, I. Matcunga, M. Shindo, Chem. Pharm. Bull. 1984, 32, 3252 3254. Received: September 5, 2011 Revised: November 4, 2011 Published online on December 8, 2011
Biological assays
Trypanothione synthetase assay: The TryS enzyme assays were conducted as reported.[20] Cell-based assays: Proliferation assays using bloodstream form T. brucei and human MRC5 fibroblasts were conducted as reported.[17]
Abbreviations
Glutathione (GSH); human African trypanosomiasis (HAT); spermidine (Spd); single knockout (SKO); Trypanothione synthetase (TryS).
Acknowledgements
We thank Irene Hallyburton and Bhavya Rao for technical support in performing the cell potency assays, Daniel James for data management support, and Suzanne Norval for technical support in performing the metabolic stability experiments. This work was funded by the Wellcome Trust (WT 077705, WT 079838, and WT 083481). Keywords: antiprotozoal agents drug design Trypanosoma brucei trypanothione synthetase
106
www.chemmedchem.org
DOI: 10.1002/cmdc.201100366
Understanding of Molecular Substructures that Contribute to hERG K + Channel Blockade: Synthesis and Biological Evaluation of E-4031 Analogues
Maris Vilums,[a] Jeroen Overman,[a] Elisabeth Klaasse,[a] Olaf Scheel,[b] Johannes Brussee,[a] and Adriaan P. IJzerman*[a]
Cardiotoxicity is a common side effect of a large variety of drugs that is often caused by off-target human ether--go-gorelated gene (hERG) potassium channel blockade. In this study, we designed and synthesized a series of derivatives of the class III antiarrhythmic agent E-4031. These compounds where evaluated in a radioligand binding assay and automated patch clamp assay to establish structureactivity relationships (SAR) for their inhibition of the hERG K + channel. Structural modifications of E-4031 were made by altering the peripheral aromatic moieties with a series of distinct substituents. Additionally, we synthesized several derivatives with a quaternary nitrogen and modified the center of the molecule by introduction of an additional nitrogen and deletion of the carbonyl oxygen. Some modifications caused a great increase in affinity for the hERG K + channel, while other seemingly minor changes led to a strongly diminished affinity. Structures with quaternary amines carrying an additional aromatic moiety were found to be highly active in radioligand binding assay. A decrease in affinity was achieved by introducing an amide functionality in the central scaffold without directly interfering with the pKa of the essential basic amine. The knowledge gained from this study could be used in early stages of drug discovery and drug development to avoid or circumvent hERG K + channel blockade, thereby reducing the risk of cardiotoxicity, related arrhythmias and sudden death.
Introduction
Potassium (K + ) channels constitute a large group of transmembrane proteins, which are highly diverse and ubiquitously spread throughout the body. Amidst this vast family of ion channels are the voltage-gated potassium channels (Kv), including the human ether--go-go-related gene (hERG) K + channel. Class III antiarrhythmic drugs affect the heart rhythm by their interaction with the hERG K + channel, which results in prolongation of the QT interval in the electrocardiogram. However, non-antiarrhythmic drugs and their metabolites may also cause prolongation of the QT interval, thereby increasing the risk of ventricular arrhythmia and fibrillation that could lead to torsades de pointes and sudden death.[13] This has become a major safety concern for the pharmaceutical industry.[4] Astemizole, cisapride, sertindole and terfenadine are a few examples of the plethora of drugs that have been, at least partially, withdrawn from the market due to their unwanted ability to prolong the QT interval and bring about associated cardiac complications.[5, 6] As a result, the screening of lead compounds in a hERG channel assay is an indispensable step in the drug development process. The patch clamp technique can be applied to analyze ion channel function and modulation directly. As its throughput is relatively low even for automated patch clamp platforms, radioligand binding protocols have been developed to provide a higher throughput screen at lower costs.[7, 8] In the present study we aimed to investigate compounds primarily targeting the hERG channel. We reasoned that more precise SAR data for such compounds can facilitate the discovery of chemical alerts for hERG K + channel-related toxicity.
ChemMedChem 2012, 7, 107 113
More specifically, we designed and synthesized 26 derivatives of the experimental class III antiarrhythmic agent E-4031 and evaluated these compounds in a [3H]astemizole radioligand binding assay, using HEK293 cell membranes expressing the hERG K + channel. A selection of these compounds was subse-
quently studied in an automated patch clamp assay. E-4031 is a member of a methane sulfonanilide class of antiarrhythmic drugs with high affinity for the hERG channel.[9] Relatively small changes to the structure of E-4031 were introduced to allow for a precise study of the hERG channel binding site. Quite a
[a] M. Vilums, J. Overman, Dr. E. Klaasse, Dr. J. Brussee, Prof. A. P. IJzerman Division of Medicinal Chemistry Leiden/Amsterdam Center for Drug Research, Leiden University P.O. Box 9502, 2300 RA Leiden (The Netherlands) E-mail: ijzerman@lacdr.leidenuniv.nl [b] Dr. O. Scheel Cytocentrics AG, Joachim-Jungius-Strae 9, 18059 Rostock (Germany) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100366.
107
MED
few changes actually led to higher affinity compounds in both assays, when compared to the parent molecule. However, some changes led to a decrease in affinity. This approach finally enabled us to suggest a number of chemical and structural features that should be avoided if one wants to reduce the interaction of an agent with the hERG channel.
Compd E-4031 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 R 4-NHSO2Me H H 4-Me 4-Me 4-OMe 4-OMe 4-Cl 4-Cl 3,4-diCl 3,4-diCl 4-NHSO2Me 4-NHSO2Me 4-NHSO2Me 4-NHSO2Me 4-Cl 4-Cl 4-Cl 4-Cl 4-Cl R 6-Me H H H H H H H H H H H H 3-Me H 4-Me 4-OMe 4-Cl 3,4-diCl 1-naphthyl[b] Z N CH CH CH CH CH CH CH CH CH CH CH CH CH N CH CH CH CH CH
A. P. IJzerman et al.
Table 1. Binding affinities of benzoylpiperidines 1533. n 2 2 3 2 3 2 3 2 3 2 3 2 3 2 2 2 2 2 2 2 IC50[a] [nm] 249 60 209 60 199 59 28 1 102 32 319 106 157 27 75 13 55 20 55 15 333 65 46 10 96 13 155 36 39 12 72 40 37 8 35 4 136 23 71 21

Chemistry The main scaffold for the E-4031 derivatives that were designed and synthesized consists of an aliphatic six-membered heterocycle flanked by two aromatic rings, where substituent R is in the 4-, or 3,4-position on the left-hand aromatic ring and R is an aromatic ring system; X is either a carbonyl or methylene unit, and Y represents either nitrogen or carbon. The aromatic rings were substituted based on the Topliss scheme.[10]
[a] High-affinity IC50 SEM (n = 3) of specific [3H]astemizole binding to membranes of HEK293 cells stably expressing the hERG K + channel. [b] For derivative 33, phenyl group was replaced by 1-naphthyl.
This synthetic strategy afforded several benzoylpiperazines, benzylpiperidines and benzoylpiperidines. In addition to these compounds that all share a tertiary amine, several quaternary ammonium analogues were prepared. Synthesis of the benzoylpiperidine derivatives was achieved following the synthetic approach reported by Oinuma et al.[11] for the preparation of E-4031 (Scheme 1). The preparation of phenyl-ring-substituted analogues 1526 (substituents given in Table 1) involved a four-step synthesis. First, the amine of 4-piperidine carboxylic acid (1) was protected with an acetyl group by reacting it with acetic anhydride (Ac2O) to afford 1-acetyl-4-piperidine carboxylic acid (2). The carboxylic acid was then converted to its corresponding acid chloride by chlorination with thionyl chloride in 1,2-dichloroethane. In a FriedelCrafts acylation reaction, using aluminum
chloride as the catalyst, the acid chloride was reacted with substituted benzenes to give benzoyl intermediates 38. These amides were then deprotected by cleavage of the acetyl group from the nitrogen using aqueous hydrochloric acid to afford the amines as hydrochloride salts 914. In the last step, the amines were subsequently alkylated with bromoalkyl benzenes to yield compounds 1526 and converted to their hydrochloride salts using hydrochloric acid in ethanol to obtain crystalline solids, if necessary. In addition to these compounds, which only have a substituted aromatic system on the left-hand side of the central scaffold, several other derivatives were prepared with a substituted aromatic moiety on the right hand also (2733; substituents given in Table 1). These derivatives were synthesized by alkylation of 4-(4-chlorobenzoyl)piperidinium chloride (12) and 4-(4(methylsulfonamido)benzoyl)piperidinium chloride (14) with 2bromoethyl aromatic compounds. Compounds 27 and 28 have a notable close resemblance to E-4031. In order to afford compound 32, the corresponding alkylating agent (2-bromoethyl)-3,4-dichlorobenzene was first prepared from the commercially available 3,4-dichlorophenethyl alcohol using standard methods. The synthesis of compounds 3536 and 3839 was achieved by a single, straightforward alkylation of the secondary nitrogen of commercially available benzylpiperidine (34) and benzoylpiperazine (37; Scheme 2). Piperidine 34 and piperazine 37 were reacted with either phenethyl bromide or phenylpropyl bromide (substituents given in Table 2). These comScheme 1. Reagents and conditions: a) Ac2O, reflux; b) 1. SOCl2, 1,2-dichloroethane, heat; pounds were converted to their hydrochloride salt to 2. substituted benzene, AlCl3, reflux; c) aq HCl, reflux; d) 1. aq NaOH, alkylating agent, obtain a crystalline product, if necessary. CH3CN, reflux; 2. ethanolic HCl. For designation of R; R; Z and n, see Table 1.
108
www.chemmedchem.org
ChemMedChem 2012, 7, 107 113
hERG Inhibitors
Scheme 2. Reagents and conditions: a) 1. bromoalkyl benzene, aq NaOH, CH3CN, microwave; 2. ethanolic HCl. For designation of n, see Table 2.
Table 2. Binding affinities of benzylpiperidines 3536 and benzoylpiperazines 3839.[a] Compd 35 36 38 39 n 2 3 2 3 IC50 [nm] 24 12 20 10 Displacement [%] 16 44
Figure 1. Representative displacement curves of specific [3H]astemizole binding to HEK293 cell membranes stably expressing the hERG K + channel by E-4031 (*), compound 17 (&) and quaternary derivative 42 (~). All curves are best described by a two-site model, representing a high- and low-affinity binding site.
[a] High affinity IC50 SEM or displacement at 10 mm (n = 3) of [3H]astemizole binding to membranes of HEK293 cells stably expressing the hERG K + channel.
The quaternary derivatives 4042 were prepared from phenethyl piperidine 21 by alkylation with the corresponding alkylating agents in a microwave synthesizer (Scheme 3; substituents given in Table 3). Biology Displacement of [3H]astemizole with increasing concentrations of cold ligand afforded concentrationeffect curves that showed two affinity sites for all the assessed compounds, including E-4031. This is most notable for the quaternary derivatives, such as 42, as there is a large separation between the high- and low-affinity sites (Figure 1).
The affinity (IC50) values of the parent compound E-4031 for these two sites, derived from three independent experiments, were 249 60 nm for the high- and 52 4 mm for the low-affinity site. This biphasic behavior in radioligand binding studies has been demonstrated by us before,[12] and may have to do with the different conformational states the channel can attain. It was decided to test all other newly synthesized E-4031 analogues at 12 concentrations at least, allowing us to reliably determine the high-affinity site IC50 values for each compound. In whole-cell voltage clamp recordings, performed on an automated patch clamp platform, E-4031 displayed sub-micromolar affinity as a functional blocker of the hERG current (Figure 2). The concentration-dependent response of E-4031 yielded an IC50 value of 124 nm (95 % current inhibition (CI) at 96161 nm), which is comparable to the IC50 value found for the high-affinity site (249 nm) determined in the radioligand binding assay. This is entirely in agreement with our previous findings with dofetilide and a potent derivative thereof for which patch clamp data were correlated with the high-affinity, and not the low-affinity, binding site in the radioligand binding experiments.[12]
Scheme 3. Reagents and conditions: a) 1. alkylating agent, CH3CN, microwave. For designation of R and X, see Table 3.
Table 3. Binding affinities of quaternary derivatives 4042. Compd 40 41 42 R Me Benzyl Phenethyl X I Br Br IC50[a] [nm] 93 17 10 4 11 2 Figure 2. Concentrationresponse curve for the hERG tail current by E-4031 (&).
[a] High affinity IC50 SEM (n = 3) of specific [3H]astemizole binding to membranes of HEK293 cells stably expressing the hERG K + channel.
ChemMedChem 2012, 7, 107 113
www.chemmedchem.org
109
MED
Structureactivity relationships Substituents on the benzoyl moiety As the methane sulfonanilide group on E-4031 is thought to interact with Tyr 652 and Phe 656 of the hERG channel through p-stacking,[9] introducing p-electron-withdrawing groups or removing electron-donating groups was hypothesized to intervene with this interaction. In fact, the methane sulfonamide substituent of E-4031 may already influence that interaction. Therefore, we prepared both a derivative with an unsubstituted phenyl moiety and four differently substituted analogues. We replaced the methylpyridine ring in E-4031 with a phenyl ring and varied the spacer length. These new compounds (15 26) where evaluated in our radioligand binding assay (Table 1). Compounds 15 and 16 with phenyl groups on both sides of the molecule showed virtually the same affinity as E-4031: IC50 values of 209 nm for 15; 199 nm for 16 and 249 nm for E-4031. However, reintroduction of the 4-NHSO2Me group in analogues 25 and 26 significantly increased the affinity of these agents for hERG, with IC50 values of 46 nm and 96 nm, respectively, indicating that a simple phenyl ring instead of the original 2methylpyridine moiety favors hERG channel affinity. Derivatives 17 and 2123 also showed increased binding affinity. The 3,4dichloro-substituted derivatives 23 and 24 yielded no improvement compared to the 4-chloro-substituted derivatives. Further increase in lipophilicity caused by the extra chloro substituent is apparently of little importance. The derivatives bearing methoxy groups (1920) had lower affinity. The carbon spacer length also influenced binding, albeit in a rather small and subtle way. The biggest difference was a four- to sixfold decrease in binding affinity for the longer chain analogues (17 vs 18 and 23 vs 24). Substitution of the phenethyl moiety Apart from the methane sulfonamide substituent that is characteristic of many antiarrhythmic drugs, the 2-methylpyridine is a notable structural feature of E-4031 and raises questions concerning its relevance for binding. Therefore, we tested two additional compounds that either have a 3-methyl substituted phenyl (27) or an unsubstituted pyridine (28). Both compounds 27 and 28 had increased affinity (IC50 = 155 nm and 39 nm, respectively) in comparison to E-4031 (IC50 = 249 nm; Table 1). However, when we compare 27 to compound 25 with a methane sulfonamide group and an unsubstituted phenyl (IC50 = 46 nm), the methyl group on the 3-position appears to reduce binding. In contrast, pyridine derivative 28 exhibits a similar affinity for hERG, which indicates that the nitrogen per se has little influence on binding affinity and is probably not involved in specific molecular interactions with the channel. Considering affinity and synthetic feasibility, we synthesized and tested another series of benzoyl derivatives with different phenethyl substituents, bearing a 4-chloro substituent on the benzoyl moiety. Compounds 2933 all have this feature and only differ in the R substituent on the phenethyl moiety, which allowed us to explore this side of the molecule further (Table 1). In fact, all compounds appeared more potent than
A. P. IJzerman et al. E-4031 on the hERG channel. Compounds 29 and 33 both had similar affinity for the channel as reference compound 21 (IC50 = 72 nm, 71 nm and 75 nm, respectively). Compound 32 displayed three- to fourfold reduced affinity in comparison with the congeneric 4-chloro-substituted derivative 31, which was among the more potent derivatives, like compound 30 (IC50 = 37 nm). The high affinity of the bulky derivative 33 with a 1-naphthyl group suggests there is ample space in the hERG channel pore. Removal of carbonyl oxygen and piperazine derivatives Working on the central part of the molecule, we discovered that deletion of the carbonyl oxygen had an activating effect for compounds 35 and 36. In comparison with the unsubstituted benzoylpiperidines 15 and 16 (IC50 = 209 nm and 199 nm, respectively), the unsubstituted benzylpiperidines 35 and 36 displayed a 10-fold higher affinity (IC50 = 24 nm and 20 nm, respectively). In fact, these derivatives were among the most active compounds tested, even rivaling the methane sulfonamide- and chloro-substituted derivatives. There is thus no evidence of the carbonyl oxygen being involved in interactions with the hERG K + channel. By introducing a piperazine ring instead of a piperidine, we learned that resulting benzoylpiperazines 38 and 39 recognized the binding site very poorly and were only able to displace 16 % and 44 % of the radioligand, respectively, at a concentration of 10 mm (Table 2). This is in agreement with findings for vesnarinone, a known cardiotonic agent.[13] The introduction of piperazine amide in the center of the molecule disables cationp interactions with Tyr 652, decreasing affinity to the micromolar range. This cationp interaction is one of the key determinants for hERG K + channel binding and is facilitated by the central basic amine on E-4031 or other blockers such as dofetilide.[12] At the same time, this finding is an opportunity to reduce hERG channel affinity, as it may be expected that for other drug targets, this change from benzoylpiperidine to benzoylpiperazine is not necessarily influential. Finally, the carbon spacer length in this series affected the binding affinity quite marginally, as an increase in length from ethyl to propyl yielded only a slight increase in affinity, if at all. Quaternary amine derivatives Quaternary amine antiarrhythmic agents such as clofilium have high affinity for the hERG channel,[14] and thus we decided to examine quaternary ammonium derivatives of E-4031 as well. For synthetic feasibility and considering biological activity, we developed a small series from compound 21 that bears a 4chloro group on the R position and has an unsubstituted phenethyl group on the right hand side of the molecule. Compound 40 (IC50 = 93 nm) had approximately the same affinity as parent compound 21 (IC50 = 75 nm), which suggests that a quaternary and a tertiary nitrogen are accommodated equally well by the binding pocket on the hERG channel. As the tertiary amine is readily protonated under physiological conditions, the cationp interactions apparently do not discriminate beChemMedChem 2012, 7, 107 113
110
www.chemmedchem.org
hERG Inhibitors tween a protonated and alkylated cationic nitrogen. Compounds 41 and 42 had largely increased affinities with IC50 values of 10 nm and 11 nm, respectively (Table 3). These compounds distinguish themselves from 40 by having an additional aromatic moiety on the central nitrogen. The radioligand binding assay on cell membranes that express the hERG channel is an in vitro biochemical assay that does not account for factors such as compound diffusion through the membrane and the accessibility of the compound binding site in dependence of the conformational state of the ion channel. This is, however, an important determinant of hERG K + channel blockade, since most inhibitors access the pore cavity from the inside of the cell and there is no active transporter known. So, the quaternary ammonium compounds may have difficulty in passing the cell membrane. In order to analyze that we performed patch clamp studies on intact cells, as described in the next section. Patch-clamp studies Whole-cell voltage clamp measurements were performed on an automated patch clamp platform (see Experimental Section). As mentioned above, the inhibitory potency of E-4031 on the hERG tail current was first evaluated. In this protocol, the compound displayed an IC50 value of 124 nm, which is in fair agreement with the IC50 value of 249 nm determined in the radioligand binding studies. Its interaction kinetics with the channel are quite slow, as the blockade (at 312 nm of E-4031, the concentration that causes a 70 % block of the current) gradually developed over time with an estimated wash-in time of approximately 10 min (2000 s1400 s = 600 s, Figure 3 a). We tested three potent and structurally different compounds in the same assay (compounds 31, 35 and 42). All these compounds showed a very rapid wash-in time (< 2 min), exemplified in Figure 3 b for compound 35. Figure 4 shows the concentrationeffect curves for these derivatives, with IC50 values determined to be 19 nm for 31, 79 nm for 35 and 830 nm for 42. The potencies of the tertiary amine derivatives 31 and 35 determined in the patch clamp assay are quite comparable to the IC50 values determined in the radioligand binding assays (35 nm and 24 nm, respectively). However, for compound 42, this is not the case; a 75-fold difference in potency between the freely accessible channels in the membrane preparation (IC50 = 11 nm) and the intact cells in the patch clamp
Figure 4. Concentrationeffect curves of compounds 31 (^), 35 (~) and 42 (&) in the patch clamp experiments.
experiments (IC50 = 830 nm) was observed. As mentioned above, it has been demonstrated that most blockers enter the channel from the cytoplasm.[15] Within this concept, it is entirely feasible that the quaternary amine derivative is compromised in crossing the cell membrane, yielding a lower overall potency.
Conclusions
The hERG K + channel is an off-target for many drugs because of its ability to bind a wide variety of chemically diverse structures. This was again proven in this study by designing and evaluating structurally diverse derivatives of E-4031. We showed that by choosing specific groups on specific spots of the E-4031 scaffold, it is possible to either reduce or increase the affinity for the hERG channel. Changing the methane sulfonamide group in E-4031 to other substituents (compounds 1726) did not affect their affinities. However, complete removal of the substituents (compounds 15 and 16) caused the affinity to decrease by fivefold. Exploring the phenyl moiety on the right-hand side of the scaffold provided further insights. There is sufficient space in the cavity to allow bulky aromatic groups (compounds 33 and 4142). A great increase in affinity was observed with the shift from a tertiary to a quaternary amine, allowing further (aromatic) substitution of the molecule, as in derivative 42. Apart from changing or deleting the substituents on the peripheral aromatic rings, the most successful strategy to reduce hERG K + channel affinity was the introduction of an amide in the centre of the molecule (compounds 38 and 39). This virtually abolished the binding affinity of the unsubstituted derivatives and might have the same effect in similar structures. The carbonyl moiety itself is unfavorable for binding, and the combination with a neighboring nitrogen atom is a particularly useful structural motif for preventing hERG K + Figure 3. Wash-in kinetics of a) E-4031 and b) 35. Arrows indicate the time point of compound addition. channel blockade.
ChemMedChem 2012, 7, 107 113 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
111
MED
In conclusion, the series of E-4031 derivatives evaluated in this study provide practical information on the molecular determinants for hERG K + channel blockade. Chemical features that are likely to increase the affinity of a compound for the hERG K + channel can be omitted; likewise, structural elements that reduce the affinity of a drug for hERG could be applied to a scaffold, by adapting the strategies represented in Figure 5. Pharmaceutical industries and academic groups could use these considerations when designing new drug entities based on similar scaffolds in order to reduce cardiotoxicity in the early stages of drug development, eventually increasing drug safety and reducing development costs.
A. P. IJzerman et al.
in vacuo. Purification by column chromatography followed by conversion to the hydrochloride salt form with ethanolic HCl afforded the desired product 1526. (1-Phenethylpiperidin-4-yl)(phenyl)methanone (15): Prepared from 9, crystalline solid (36 %): 1H NMR (400 MHz, CDCl3): d = 7.94 7.17 (m, 10 H), 3.28 (m, 1 H), 3.10 (m, 2 H), 2.84 (m, 2 H), 2.65 (m, 2 H), 2.23 (m, 2 H), 1.89 ppm (m, 4 H); 13C NMR (101 MHz, CDCl3): d = 202.6, 140.3, 136.1, 133.0, 128.8, 128.7, 128.4, 128.3, 126.1, 60.8, 53.2, 43.5, 33.6, 28.6 ppm; HRMS (ESI): m/z [M + H] + calcd for C20H23NO: 294.18524, found: 294.18527; HPLC: tR = 41.3 min. General procedure for the synthesis of compound 35, 36: Compounds 3536 were prepared from 1-benzylpiperazine monohydrochloride 34 using the same procedure as described for compounds 1526, but rather than being refluxed, the reaction mixture was irradiated in the microwave at 110 8C for 1.52 h. 4-Benzyl-1-phenethylpiperidine hydrochloride (35): White crystalline solid (80 %): 1H NMR (400 MHz, CD3OD): d = 7.417.15 (m, 10 H), 3.54 (d, J = 12.0 Hz, 2 H), 3.233.19 (m, 2 H), 3.063.01 (m, 2 H), 2.88 (t, J = 12.0 Hz, 2 H), 2.65 (d, J = 6.4 Hz, 2 H), 1.92 (s, 1 H), 1.89 (s, 2 H), 1.651.48 ppm (m, 2 H); 13C NMR (101 MHz, CD3OD): d = 139.3, 136.8, 128.8, 128.5, 128.4, 128.1, 126.7, 125.9, 58.0, 52.6, 41.5, 35.6, 30.4, 29.2 ppm; HRMS (ESI): m/z [M + H] + calcd for C20H26ClN: 280.20595, found: 280.25098; HPLC: tR = 41.7 min. General procedure for the synthesis of compound 3839: Compounds 3839 were prepared from 1-benzoylpiperazine monohydrochloride 37 using the same procedure as described for compounds 3536. (4-Phenethylpiperazin-1-yl)(phenyl)methanone (38): Yellow crystalline solid (50 %): 1H NMR (400 MHz, CDCl3): d = 7.407.17 (m, 10 H), 3.82 (br s, 2 H), 3.45 (br s, 2 H), 2.80 (dd, J = 9.8, 6.2 Hz, 2 H), 2.63 (dd, J = 9.8, 6.2 Hz, 2 H), 2.61 (br s, 2 H), 2.45 ppm (br s, 2 H); 13 C NMR (101 MHz, CDCl3): d = 170.3, 140.0, 135.8, 129.7, 128.7, 128.5, 128.5, 127.1, 126.2, 60.3, 53.6, 52.8, 47.8, 42.2, 33.5 ppm; HRMS (ESI): m/z [M + H] + calcd for C19H22N2O: 295.18054, found: 295.18049; HPLC: tR = 43.9 min. General procedure for the synthesis of compound 4042: The alkyl bromide/iodide (0.5 mmol, 1 eq) was added to a solution of compound 21 (0.5 mmol, 1 eq) and aq NaOH (0.5 mmol, 1 eq) in CH3CN (10 mL). This mixture was heated with microwave irradiation at 110 8C for 4 h, then concentrated in vacuo. Purification by column chromatography on silica gel (EtOAc) afforded the desired product. 4-(4-Chlorobenzoyl)-1-methyl-1-phenethylpiperidinium iodide (40): Yellow crystalline solid (8 %): 1H NMR (400 MHz, CD3OD): d = 8.09 (d, J = 8.3 Hz, 2 H), 7.59 (d, J = 8.3 Hz, 2 H), 7.457.25 (m, 5 H), 3.983.86 (m, 1 H), 3.813.65 (m, 6 H), 3.25 (s, 3 H), 3.263.22 (m, 2 H), 2.322.09 ppm (m, 4 H); 13C NMR (101 MHz, CD3OD): d = 200.1, 141.5, 132.0, 130.0, 128.9, 128.8, 128.7, 127.0, 125.7, 59.7,
Chemistry
General: All solvents and reagents were analytical grade and purchased from commercial sources. E-4031 was purchased from Tocris Bioscience. Microwave reactions were carried out on a Biotage AB Emrys Optimizer, using Emrys Optimizer software v.2.6.0.4. Wattage was automatically adjusted to maintain the desired temperature. 1H and 13C NMR spectra were recorded on a Bruker AV400 spectrometer (1H NMR, 400 MHz; 13C NMR, 100 MHz) at ambient temperature and analyzed with MestReNova v.6.1.16384. Chemical shifts (d) are given in parts per million (ppm); multiplicities are indicated by s (singlet), d (doublet), dd (doublet of doublet), t (triplet), tt (triplet of triplets), q (quartet), m (multiplet), br (broad). High-resolution mass spectrometry (HRMS) of the final compounds was performed by the Leiden Institute of Chemistry (The Netherlands) using a Thermo Finnigan LTQ Orbitrap mass spectrometer. Calculated values are within 2 ppm of the theoretical values. Analytical purity of the final compounds was determined by HPLC with a Merck Discovery C18 column (12.5 cm 4.6 mm, 5 mm), measuring UV absorbance at 254 nm. Unless stated otherwise, sample preparation and HPLC method were as follows: compound (~ 1 mg) was dissolved in 1.5 mL of CH3CN/H2O (40:60) and eluted from the column within 60 min at flow rate 0.6 mL min1, with a two component system of MeOH/H2O, decreasing polarity of the solvent mixture over time. All compounds showed a single peak at the designated residence time and were at least 95 % pure. Thin-layer chromatography (TLC) was routinely consulted to monitor the progress of reactions, using aluminum coated Merck silica gel F254 plates. General procedure for the synthesis of compound 1526: The appropriate bromoalkyl benzene (0.5 mmol, 1 equiv) was added to a solution of compound 914 (0.5 mmol, 1 equiv) and 1 m NaOHaq (1.0 mmol, 2 equiv) in CH3CN (10 mL). This mixture was refluxed for 14 h, concentrated in vacuo, extracted with CH2Cl2 and washed with brine, dried (MgSO4), filtered by glass filter, and concentrated
Figure 5. Overview of proposed strategies to reduce hERG K + channel affinity.
112
www.chemmedchem.org
ChemMedChem 2012, 7, 107 113
hERG Inhibitors
53.1, 43.4, 39.6, 28.0, 22.6 ppm; HRMS (ESI): m/z [M + H] + calcd for C21H25ClNO: 342.16221, found: 342.16192; HPLC: tR = 41.6 min. form CytoPatch Instrument (Cytocentrics).[17] To determine the inhibitory effect of the compounds, hERG outward tail currents were measured by executing the following pulse protocol every 10 s: from a holding potential of 70 mV, cells were voltage-clamped for 100 ms to 50 mV and then for 2 s to + 40 mV. In order to evoke outward tail currents, a 2 s step to 50 mV followed. The peak tail current was corrected for the leak current determined during the first short voltage step to 50 mV. Cells with a wholecell membrane resistance of more than 500 MOhm and a hERG tail current amplitude of at least 200 pA were then continuously perfused with EC buffer for 10 min to establish stable predrug recordings. After this control period, test compounds were continuously applied to the cell via the transport channel for 12 min. Up to two increasing concentrations of the same compound were applied to one cell. hERG tail currents were averaged over 50 s at the end of the control phase and the end of each application phase. If the whole-cell membrane resistance decreased to less than 500 MOhm during compound application, the hERG tail current was only averaged if a clear steady state developed upon drug wash-in before cell loss. Current inhibition was calculated by dividing the mean tail current in the presence of the drug by the mean tail current of the control phase. All measurements were performed at room temperature.
Biology
Radioligand binding studies: The [3H]astemizole radioligand displacement assay to determine the IC50 values of E-4031 derivatives for the hERG K + channel was performed as described in the literature,[7] with some minor adjustments. A 400 mg mL1 solution of hERG/HEK293 membranes was prepared by diluting aliquots with buffer (10 mm 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 130 mm NaCl, 60 mm KCl, 0.8 mm MgCl2, 1 mm NaEGTA, 10 mm glucose and 0.1 % bovine serum albumin (BSA), pH 7.4). This sample (total volume = 100 mL) was vortexed and incubated at 25 8C for 1 h in the presence of the cold ligand and 1.52 nm of radioligand. Initially, only single-point measurements were taken at 10 mm of cold ligand to increase throughput. Full displacement curves were obtained from at least 12 concentrations in the range 1010104 m with intervals of 0.33, 0.5 or 1.0 log units depending on the activity of the evaluated compound. Nonspecific binding was determined in the presence of 10 mm astemizole. After incubation, the samples were diluted with an ice-cold wash buffer (25 mm Tris-HCl, 130 mm NaCl, 60 mm KCl, 0.8 mm MgCl2, 0.05 mm CaCl2 and 0.05 % BSA, pH 7.4). Unbound radioligand was separated from bound radioligand by rapid filtration through UniFilter GF/B 96-well plates using a FilterMate harvester (Perkin Elmer). The filters were then washed with wash buffer. Filter-bound radioactivity was subsequently determined by addition of 25 mL MicroScint 20 and measuring scintillation using a 1450 MicroBeta Trilux microplate scintillation counter. The data was analyzed using GraphPad Prism (v.5.00) by comparison of curve fitting for one site Fit log IC50 versus two sites Fit log IC50 with an extra sum of square F-test. All curves preferred a biphasic Fit with a P value < 0.05. Patch clamp experiments: Extracellular (EC) buffer used for automated patch clamp recordings, compound dilutions and for storage of cells was composed of: 140 mm NaCl, 2.5 mm KCl, 2 mm MgCl2, 2 mm CaCl2, 10 mm HEPES, 10 mm glucose, 15 mm sucrose (pH 7.4 0.1); osmolality: 320 5 mOsm kg1. The buffer was stored at 4 8C, but degassed and warmed to RT prior to use. Intracellular (IC) buffer was composed of: 100 mm K-gluconate, 20 mm KCl, 1 mm CaCl2, 1 mm MgCl2, 10 mm HEPES, 11 mm EGTA-KOH, 4 mm MgATP, 3 mm phosphocreatine-Na2-H2O, 9 mm sucrose (pH 7.2 0.1); osmolality: 295 5 mOsm kg1. Aliquots were stored at 20 8C, thawed prior to use and used for a maximum of 4 h. Compounds: Stock solutions of the compounds were prepared in DMSO and stored at 20 8C. For working dilutions, compounds were further diluted in DMSO and finally diluted in a 1:1000 step in EC buffer, yielding the intended compound concentration with a 0.1 % final DMSO content. Working dilutions were prepared freshly before use. Cells: CHO-K1 cells stably transfected with cDNA encoding the hERG ion channel.[16] Cells were cultured in Dulbeccos modified Eagle medium (DMEM) with 10 % fetal bovine serum (FBS), 100 mm sodium pyruvate, 200 mm l-glutamine (all from PAA Laboratories GmbH, Pasching, Germany) with 500 mg mL1 Geneticin (Carl Roth GmbH, Karlsruhe, Germany) at 37 8C in 8.5 % CO2. For patch clamp experiments, the cells were harvested using Accutase (PAA Laboratories GmbH) and stored in EC buffer in the Cytocentrics CellReservoir. Automated patch clamp recordings: Whole-cell voltage clamp experiments were performed using the automated patch clamp platChemMedChem 2012, 7, 107 113
Acknowledgements
This study was financially supported by the Dutch Top Institute Pharma, project number D2201. Keywords: E-4031 hERG ion channels patch clamp assay radioligand binding structureactivity relationships
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] V. Ruta, J. Chen, R. MacKinnon, Cell 2005, 123, 463 475. Z. Lu, A. M. Klem, Y. Ramu, J. Gen. Physiol. 2002, 120, 663 676. M. C. Sanguinetti, M. Tristani-Firouzi, Nature 2006, 440, 463 469. J. P. Valentin, T. Hammond, J. Pharmacol. Toxicol. Methods 2008, 58, 77 87. R. L. Woosley, Y. Chen, J. P. Freiman, R. A. Gillis, JAMA 1993, 269, 1532 1536. A. M. Brown, Cell Calcium 2004, 35, 543 547. P. J. Chiu, K. F. Marcoe, S. E. Bounds, C. H. Lin, J. J. Feng, A. Lin, F. C. Cheng, W. J. Crumb, R. Mitchell, J. Pharmacol. Sci. 2004, 95, 311 319. K. Finlayson, L. Turnbull, C. T. January, J. Sharkey, J. S. Kelly, Eur. J. Pharmacol. 2001, 430, 147 148. K. Kamiya, R. Niwa, J. S. Mitcheson, M. C. Sanguinetti, Mol. Pharmacol. 2006, 69, 1709 1716. J. G. Topliss, J. Med. Chem. 1972, 15, 1006 1011. H. Oinuma, K. Miyake, M. Yamanaka, K. Nomoto, H. Katoh, K. Sawada, M. Shino, S. Hamano, J. Med. Chem. 1990, 33, 903 905. Shagufta, D. Guo, E. Klaasse, H. de Vries, J. Brussee, L. Nalos, M. B. Rook, M. A. Vos, M. A. van der Heyden, A. P. Ijzerman, ChemMedChem 2009, 4, 1722 1732. K. Kamiya, J. S. Mitcheson, K. Yasui, I. Kodama, M. C. Sanguinetti, Mol. Pharmacol. 2001, 60, 244 253. H. Suessbrich, R. Schonherr, S. H. Heinemann, F. Lang, A. E. Busch, FEBS Lett. 1997, 414, 435 438. J. S. Mitcheson, Chem. Res. Toxicol. 2008, 21, 1005 1010. Z. Zhou, Q. Gong, B. Ye, Z. Fan, J. C. Makielski, G. A. Robertson, Biophys. J. 1998, 74, 230 241. O. Scheel, H. Himmel, G. Rascher-Eggstein, T. Knott, ASSAY Drug Dev. Technol. 2011; DOI: 10.1089/adt.2010.0352.
[13] [14] [15] [16] [17]
www.chemmedchem.org
113
MED
DOI: 10.1002/cmdc.201100358
A Comparison of Linear and Cyclic Peptoid Oligomers as Potent Antimicrobial Agents

Mia Lace Huang,[a] Sung Bin Y. Shin,[a] Meredith A. Benson,[b] Victor J. Torres,[b] and Kent Kirshenbaum*[a]
We investigated the antimicrobial activities of N-substituted glycine peptoid oligomers incorporating cationic and hydrophobic side chains. Head-to-tail macrocyclization was employed to enhance antimicrobial activity. Both linear and cyclic peptoids, ranging from six to ten residues, demonstrate potent antimicrobial activity against Gram-positive and Gram-negative bacteria. These peptoids do not cause significant lysis of human erythrocytes, indicating selective antimicrobial activity. Conformational ordering established upon macrocyclization is generally associated with an enhanced capacity to inhibit bacterial cell growth. Moreover, increased hydrophobic surface area also plays a role in improving antimicrobial activity. We demonstrate the potency of a cyclic peptoid in exerting antimicrobial activity against clinical strains of S. aureus while deterring the emergence of antimicrobial resistance.
Introduction
The emergence of bacterial pathogens resistant to conventional antibiotics is a growing public health threat. The current set of antibiotic agents is becoming incapable of combating many pathogenic bacteria, creating an unfortunate synergy of increased drug resistance combined with a diminishing rate of new antibiotic development.[1] Thus, not only must new antibiotic agents be discovered, but in order to properly address antimicrobial resistance, new antibiotics must possess novel mechanisms of action.[2] The proper selection of the antibiotic target is an essential consideration. In particular, targets that exhibit a low propensity for rapid resistance selection may lead to therapeutic regimens with sustained clinical efficacy.[3] Previous studies have explored the use of host-defense antimicrobial peptides (AMPs) that primarily target the bacterial membrane as novel antibiotic agents.[4] AMPs have been studied as potential therapeutic agents owing to the low frequency of bacterial resistance observed for this compound class.[3, 5] AMPs interact with the bacterial cytoplasmic membrane. Accordingly, a common feature found among AMPs is their amphiphilic structure, wherein residues are segregated into hydrophobic and cationic regions. The amphiphilic structure of AMPs enables both the initial electrostatic interaction between the AMPs and the anionic bacterial phospholipid head groups, accompanied by hydrophobic interactions with the apolar lipid acyl chains. The cationic nature of AMPs also serves to establish AMP selectivity for the negatively charged bacterial membranes over zwitterionic eukaryotic membranes. However, the entry of AMPs into clinical use has been hampered by poor selectivity and by in vivo degradation that can impair bioavailability. Extensive research efforts have sought to harness and enhance the desirable features present in naturally occurring AMPs.[6] Peptidomimetic compounds can recapitulate the structure and function of natural AMPs while circumventing their susceptibility to protease degradation. Peptidomimetics such as b-peptides[7] and arylamides[8] have generated significant interest as potential therapeutic agents,[9] especially those with a propensity to fold into stable secondary structures. Such molecules are anticipated to perturb the bacterial cell membrane similar to the action of AMPs, although the exact mechanisms may vary. Our investigations are focused on a class of peptidomimetic oligomers composed of N-substituted glycine monomer units.[10] These oligomers, termed peptoids, are an important class of sequence-specific peptidomimetics known to generate diverse biological activities.[11] Peptoids are well suited for AMP mimicry because of their ability to display side chains similar to bioactive peptides.[6c] In addition, the abiotic backbone in peptoids confers resistance from proteolytic degradation.[11b] There is ongoing debate about the importance of enforcing conformational order in peptides and other amphiphilic oligomers to obtain potent antimicrobial activity. Whereas some studies indicate a strong correlation between the stability of secondary structures and antimicrobial activity,[4, 8b] there have been several examples that offer an alternate view.[12] Moreover, it has been reported that improvements in activity can be attained by mitigating the extent of secondary structure formation.[13] It is still unclear whether conformational ordering
[a] M. L. Huang, Dr. S. B. Y. Shin, Prof. K. Kirshenbaum Department of Chemistry, New York University 100 Washington Square East, New York, NY 10003-6688 (USA) E-mail: kent@nyu.edu [b] M. A. Benson, Prof. V. J. Torres Department of Microbiology, New York University School of Medicine 522 First Avenue, Smilow 1010, New York, NY 10016 (USA) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100358.
114
ChemMedChem 2012, 7, 114 122
Peptoid Oligomers is generally advantageous or deleterious for antimicrobial activity in peptidomimetic systems. A crucial challenge to the design of peptoids with stable secondary structures is the substantial conformational heterogeneity present primarily in the form of cistrans amide bond isomerization.[14] A number of strategies have been developed to establish defined conformations in peptoids.[15] The incorporation of bulky a-chiral side chains has been reported to result in peptoids that adopt a polyproline type I helical structure.[16] The Barron research group has shown that peptoid sequences that exhibit such stable helical secondary structures possess potent antimicrobial activity and are good functional mimics of a well-known helical antimicrobial peptide, magainin-2amide.[11g, 17] Another strategy in enforcing rigidity in peptoid structures is through the introduction of covalent constraints by head-totail macrocyclization.[18] We have demonstrated by X-ray diffraction studies that macrocyclization can organize peptoid side chain groups onto opposing faces of the planar macrocycle. This suggests that the proper placement of cationic and hydrophobic side chains within cyclic peptoid oligomer sequences would result in a globally amphiphilic structure. The resulting well-defined cationic and hydrophobic surfaces would lead to potent mimics of AMPs capable of combating bacterial pathogens. Initial studies have provided proof of this principle by demonstrating modest antimicrobial activity for cyclic peptoid oligomers against fungi and bacteria.[19] Here we aim to carefully evaluate the effect of head-to-tail macrocyclization on the antimicrobial activity of short peptoid sequences that contain cationic and hydrophobic side chains. We anticipate that this strategy will generate globally amphiphilic peptoid macrocycles with enhanced antimicrobial activity relative to linear peptoid oligomers. In addition, these shortchained oligomers may exhibit improved pharmacological attributes over other oligomer antimicrobials, which are typically 1225 residues in length.[4a, 11g, 14a] We investigate a library of linear and cyclic peptoids to study the effects of macrocyclization, cationic and hydrophobic group variations, and oligomer sizes on antimicrobial activity.
Figure 1. Peptoid cationic and hydrophobic monomer structures and their designators used in this study. Nap: N-(3-aminopropyl)glycine; Nab: N-(4aminobutyl)glycine; Nah: N-(6-aminohexyl)glycine; Ngb: N-(4-guanidinobutyl)glycine; Npm: N-(phenylmethyl)glycine; Nnm: N-(1-naphthylmethyl)glycine; Ndp: N-(2,2-diphenylethyl)glycine; Nip: N-(isopropyl)glycine; Nib: N(isobutyl)glycine.
Sequence-specific N-substituted glycine oligomers can be efficiently synthesized via sub-monomer chemistry methods (Scheme 1) to incorporate a large number of diverse side chain functionalities.[20] This approach iterates sequential steps of bromoacylation and nucleophilic displacement to construct

Design and synthesis of peptoid sequences Using a structurally diverse set of cationic and hydrophobic monomers (Figure 1), we designed a library of short amphiphilic linear and cyclic peptoid oligomers composed of six to ten residues. This library contains a variety of structures to explore the effects of macrocyclization, hydrophobic surface area, cationic charge composition, and chain length upon antimicrobial activity. Based on crystal structures of cyclic peptoids,[18a] we anticipated that the incorporation of alternating cationic and hydrophobic monomers into peptoid macrocycles would result in a highly amphiphilic molecule in which the cationic polar and hydrophobic side chains segregate onto the distinct opposing faces of the macrocycle.[18a]
ChemMedChem 2012, 7, 114 122
Scheme 1. Solid-phase synthesis of A) N-acetylated linear and B) cyclic peptoid oligomers. Reagents and conditions: a) bromoacetic acid and DIC in DMF; b) primary amine (R-NH2) sub-monomer in DMF; c) Ac2O and DIEA in DMF, then 95 % TFA in H2O; d) HFIP in CH2Cl2, then PyBOP in DMF. Rink amide resin was used to generate N-acetylated linear oligomers, and 2-chlorotrityl chloride resin was used to generate linear precursors for cyclic peptoid oligomers; n = 6, 8, or 10.
each monomer unit. The side chain moieties are introduced upon displacement of bromide by primary amine sub-monomer synthons. N-acetylated linear oligomers are synthesized on Rink amide resin to afford C-terminal amides, for which the N termini are acetylated with acetic anhydride prior to TFA cleavage. Cyclic peptoid oligomers are generated from linear precursors and cyclized in high yield in PyBOP/DIEA/DMF[18a] (figure S1, Supporting Information). For the cyclic peptoids, linear precursors are synthesized on 2-chlorotrityl chloride resin to generate free N-terminal amino and C-terminal carboxylic acid groups, and are then cleaved with HFIP/CH2Cl2 (20 % v/v, Scheme 1). All peptoids are then purified to > 95 % by RP-
www.chemmedchem.org
115
MED
K. Kirshenbaum et al.
Figure 2. Structures of selected linear and cyclic peptide and peptoid sequences.
HPLC and characterized by ESI-MS (table S1, figure S2, Supporting Information). Synthesis of gramicidin S The cyclic decapeptide gramicidin S (CGS, Figure 2) was used as a positive control in comparing antimicrobial activity. CGS is composed of a cyclo(VOLdFP)2 sequence, in which O is an ornithine residue, and dF is a d-phenylalanine residue.[21] CGS and its N-acetylated linear counterpart were synthesized using standard Fmoc solid-phase peptide synthesis protocols. Antimicrobial and hemolytic activities Compounds were screened for antimicrobial and hemolytic activities (Table 1). Antimicrobial activity was tested against Gram-negative Escherichia coli ATCC 25922 (E. coli), Gram-positive Staphylococcus aureus ATCC 25923 (S. aureus), and Bacillus subtilis ATCC 6633 (B. subtilis) according to the guidelines given in the document M07-A7 of the Clinical and Laboratory Standards Institute (CLSI).[22] Hemolytic activity was tested with fresh human erythrocytes by measuring the absorbance of free hemoglobin as a marker of erythrocyte lysis. These experiments were conducted in three replicates of three independent trials. As evident in Table 1, CGS exhibits potent antimicrobial activity, with minimum inhibitory concentration (MIC) values of 2 mg mL1 for B. subtilis, and 15.6 mg mL1 for E. coli and S. aureus. These values are in agreement with previous reports.[23] Some peptoid sequences proved to be inactive as antimicrobials (L5, L13, L15, C5, C13, and C15), with MIC values > 500 mg mL1, whereas some sequences showed antimicrobial activity, with MIC values < 100 mg mL1 (L3, L8, L11, C3, C4, C7,
and C8; Table 1). The most potent antimicrobial activities were observed with the cyclic peptoid decamer C8 (Figure 2), which showed MIC values of 0.5 mg mL1 for B. subtilis, and 7.8 mg mL1 for E. coli and S. aureus, demonstrating that peptoid sequences can be potent antimicrobial agents. It was also observed that many active peptoid sequences did not readily induce hemolysis of human erythrocytes. Typically, hemolytic activity is evaluated by measuring the concentration at which 50 % of the erythrocytes in solution are lysed (hemolytic concentration value HC50). Significant hemolysis as evaluated by HC50 was detected only for the cyclic peptide CGS at 40 mg mL1. Peptoid oligomers were found to be nonhemolytic using this measurement (HC50 > 250 mg mL1 for all peptoids). It has been reported that antimicrobials exhibiting acceptable HC50 values may still exhibit insufficient selectivity.[12a] Thus, to evaluate the selectivity of peptoid antimicrobials rigorously, a more stringent parameter was used. Instead of HC50, the concentration at which 10 % hemolysis occurs was determined (HC10). HC10 values were detectable only for four peptoid oligomers in the library: L3, L12, C3, and C11. All four peptoid sequences contain Ndp residues, suggesting a relationship between hydrophobicity and hemolytic activity. Other peptoid sequences did not cause 10 % hemolysis even at the highest concentrations used in this assay. A selectivity ratio (SR), which we define here as HC10/MICE. coli, measures the ratio of hemolytic activity to antimicrobial activity. High SR values indicate that a compound possesses potent antimicrobial activity and low hemolytic activity. Gramicidin S (CGS) is strongly lytic toward erythrocytes, with HC50 and HC10 values of 40 and 15.6 mg mL1, respectively. Thus, although CGS possesses antimicrobial activity, it exhibits a low SR value of 1 because of its poor selectivity for bacterial over mammalian cells. The SR values obtained in this study range from 0.5 to
ChemMedChem 2012, 7, 114 122
116
www.chemmedchem.org
Peptoid Oligomers positive (B. subtilis and S. aureus) bacteria, we chose to use the MIC values against E. coli for comparing antimicrobial activity. Analyzing the antimicrobial activity of six pairs of linear and cyclic peptoid sequences shows that cyclization generally improves antimicrobial activity, as apparent by the decrease in MIC values for cyclic relative to linear sequences (Figure 3). For example, cyclic sequences of (NapNpm)4 (C7), and (NdpNibNapNipPro)2 (C14) are about eightfold more active than their linear counterparts, L7 and L14. The enhancement in antimicrobial activity upon macrocyclization is consistent with similar effects obtained by enforcing secondary structure in helical peptoid antimicrobial oligomers.[17b] Interestingly, both compounds are also non-hemolytic up to 250 mg mL1, providing selectivity ratio values > 8 (Table 1).
Table 1. Antimicrobial and hemolytic activities of linear and cyclic peptoids. MIC [mg mL1][b] E. coli 250 > 500 15.6 62.5 > 500 125 250 31.3 > 500 > 500 31.3 > 500 > 500 250 > 500 125 250 500 15.6 31.3 > 500 > 500 31.3 7.8 500 250 250 > 500 > 500 31.3 > 500 15.6 HC [mg mL1][c] HC50 HC10 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 40 > 250 > 250 > 62.5 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 250 250 > 250 > 250 > 250 > 250 > 250 > 250 31.3 > 250 > 250 > 250 > 250 > 250 > 250 > 250 > 62.5 > 250 > 250 > 250 > 250 > 15.6
Sequence[a] L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L12 L13 L14 L15 LGS C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 CGS Ac(NgbNpm)3 Ac(NahNpm)3 Ac(NapNdp)3 Ac(NapNnm)3 Ac(NapNpm)2Nap2 Ac(NapNpm)3 Ac(NapNpm)4 Ac(NapNpm)5 Ac(NpmNap)2Npm2 Ac(NabNpm)3 Ac(NgbNdp)5 Ac(NdpNgb)2Ndp2 Ac(NipNapNibNpmPro)2 Ac(NdpNibNgbNipPro)2 Ac(NpmNibNapNipPro)2 Ac(VOLdFP)2 C(NgbNpm)3 C(NahNpm)3 C(NapNdp)3 C(NapNnm)3 C(NapNpm)2Nap2 C(NapNpm)3 C(NapNpm)4 C(NapNpm)5 C(NpmNap)2Npm2 C(NabNpm)3 C(NgbNdp)5 C(NdpNgb)2Ndp2 C(NipNapNibNpmPro)2 C(NdpNibNgbNipPro)2 C(NpmNibNapNipPro)2 C(VOLdFP)2
B. sub. 62.5 > 500 1 2 > 500 125 31.3 15.6 31.3 > 500 62.5 > 500 > 500 500 > 500 7.8 125 500 1 0.5 > 500 500 2 0.5 15.6 250 250 > 500 > 500 62.5 > 500 2
S. aur. 125 > 500 15.6 125 > 500 250 250 62.5 > 500 > 500 62.5 62.5 > 500 500 > 500 125 250 > 500 7.8 31.3 > 500 500 31.3 7.8 500 500 250 > 500 > 500 62.5 > 500 15.6
SR[d] >1 NA >4 >4 NA >2 >1 >8 NA NA >8 NA NA >1 NA >2 >1 > 0.5 >2 >8 NA NA >8 > 32 > 0.5 NA >1 NA NA >8 NA 1
[a] Prefixes Ac and C refer to N-acetylated linear and cyclic sequences, respectively. [b] Minimum inhibitory concentrations against Bacillus subtilis, Escherichia coli, and Staphylococcus aureus. [c] Hemolytic concentrations: HC50, concentration at which 50 % hemolysis is observed; HC10, concentration at which 10 % hemolysis is observed. [d] Selectivity ratio: SR = HC10/MICE. coli ; NA: not applicable due to insufficient antimicrobial activity; experiments were conducted in three replicates of three independent trials; data reported in mm units are listed in table S2 of the Supporting Information.
Effect of hydrophobic surface area on antimicrobial activity The effect of hydrophobic surface area on antimicrobial activity was tested by comparing sequences that incorporate different residues of varying hydro-
32 for active peptoid compounds. Among cyclic peptoid sequences, four (C4, C7, C8, C14) were found to be potent and selective, with SR > 8, whereas only two linear peptoidsL8 and L11showed selectivity ratios greater than this value (Table 1). The most potent antimicrobial compound, C8 (Figure 2), also showed the highest selectivity ratio (SR > 32). The high selectivity ratios observed for this library of short peptoid oligomers are indicative of the potential of peptoids to be potent and selective antimicrobial agents. Because the HC10 values of many of the compounds in this library are above the assay limit of detection, the actual SR values could possibly be even higher than the calculated values. These values indicate that the compounds are able to kill bacterial cells without exerting significant lytic activity toward mammalian erythrocytes. Effect of cyclization on antimicrobial activity Because individual peptoids show generally similar trends in antimicrobial activity against Gram-negative (E. coli) and GramChemMedChem 2012, 7, 114 122
Figure 3. Antimicrobial activity against E. coli for six pairs of linear and cyclic compounds tested in this study shows that macrocyclization enhances antimicrobial activity.
www.chemmedchem.org
117
MED
phobic surface area (Figure 1: Npm, Nnm, and Ndp). N-Acetylated linear and cyclic peptoid hexamer sequences of (NapNpm)3 (L6 and C6), (NapNnm)3 (L4 and C4), and (NapNdp)3 (L3 and C3) were used to represent three sets of oligomers with varying degrees of hydrophobic surface area. L6 and C6 contain three phenyl groups per molecule as the hydrophobic functional groups, and possess the least hydrophobic surface area in the group. L4 and C4 contain three naphthyl groups per molecule as the hydrophobic functional groups, and these compounds possess greater hydrophobic surface area than L6 and C6. L3 and C3 contain six phenyl groups per molecule, and possess the most hydrophobic surface area in the group. Antimicrobial data shown in Figure 4 demonstrate that the sequences containing Ndp residues have greater antimicrobial activity than the sequences containing Npm or Nnm residues, suggesting a strong relationship between hydrophobic surface area and antimicrobial activity. These results are in agreement with previous reports.[11g, 17b] Interestingly, the enhancement of antimicrobial activity with increasing hydrophobic surface area is more pronounced in cyclic than in linear peptoids.
K. Kirshenbaum et al. Effect of various cationic side chains on antimicrobial activity The effect of various cationic side chains on antimicrobial activity was tested by comparing sequences that incorporate different alkylamino and guanidino functional groups as cationic residues (Figure 1: Nap, Nab, Nah, and Ngb). N-Acetylated linear and cyclic peptoid hexamer sequences of (NapNpm)3 (L6 and C6), (NabNpm)3 (L10 and C10), (NahNpm)3 (L2 and C2), and (NgbNpm)3 (L1 and C1) were used to represent a set of peptoid oligomers with two different types of cationic functional groups (amino and guanidino) and three different alkyl linker lengths. Results shown in Figure 5 and Table 1 indicate that N-acetylated linear hexamers that contain Nab or Nah are inactive (MIC > 500 mg mL1) against all bacteria tested, while Nap- and Ngb-containing hexamers showed moderate antimicrobial activity (MIC: 125250 mg mL1). From the set of various cationic molecules tested herein, the variations in the length of the amino alkyl side chain or the nature of the cationic functional group (amino versus guanidino) do not show a consistent correlation to antimicrobial activity.
Figure 4. Antimicrobial activity against E. coli for three linear and three cyclic peptoid oligomers containing different hydrophobic residues. Results show the importance of increasing hydrophobic surface area to enhancing antimicrobial activity.
Because peptoids that incorporate Ndp residues as hydrophobic monomers (such as L3 and C3) are associated with both enhanced antimicrobial activity (Figure 4) and hemolytic activity (Table 1), this suggests that an increased hydrophobic surface area also causes greater hemolytic activity. Thus, a balance in hydrophobic surface area must be considered in the molecular design of antimicrobial agents that exhibit potency but are also nontoxic toward mammalian cells. This observation is consistent with results from studies on antimicrobial polymers and foldamers that mimic the action of AMPs.[24]
Figure 5. Antimicrobial activity against E. coli for four pairs of linear and cyclic peptoid oligomers containing different cationic residues. No consistent relationship was observed between the length of the cationic side chain or the type of cationic functional group (amino versus guanidino) on antimicrobial activity.
Effect of molecular size on antimicrobial activity The effect of molecular size on the antimicrobial activity was tested by comparing hexameric, octameric, and decameric sequences. N-Acetylated linear and cyclic compounds of (NapNpm)3 (L6 and C6), (NapNpm)4 (L7 and C7), and (NapNpm)5 (L8 and C8) sequences were used to represent three sets of oligomers with different lengths.
ChemMedChem 2012, 7, 114 122
118
www.chemmedchem.org
Peptoid Oligomers We observed a correlation between oligomer length and antimicrobial activity for both linear and cyclic oligomers. Increasing the chain length generally enhances antimicrobial activity, as evident by the large decrease in MIC values for both the linear and cyclic series (Figure 6). The linear decamer L8 is the
Figure 7. Evaluation of antimicrobial resistance against cyclic peptoid C3 in clinical strains of a) methicillin-resistant (MRSA) and b) methicillin-sensitive S. aureus (MSSA). No change in MIC was observed over 21 days, indicating the absence of the development of antimicrobial resistance.
Figure 6. Antimicrobial activity against E. coli for three pairs of linear and cyclic peptoid oligomers containing different chain lengths and sizes. Results show that increasing chain length also enhances antimicrobial activity.
was repeated for 21 days. Results in Figure 7 show that the MIC values for parent and succeeding cultures remained constant over 21 days, indicating the inability of clinical S. aureus strains to rapidly evolve resistance in vitro against peptoid C3.
largest among the linear sequences and possesses the most potent antimicrobial activity (MICE. coli = 31.3 mg mL1). Similarly, the cyclic decamer C8 is the largest macrocycle among cyclic sequences and possesses the most potent antimicrobial activity (MICE. coli = 7.8 mg mL1); its activity is followed by the cyclic octamer C7 (MICE. coli = 31.3 mg mL1), and the inactive cyclic hexamer C6 (MICE. coli > 500 mg mL1). Resistance study The efficacy of conventional antibiotics in treating infections caused by multidrug-resistant pathogens has been diminished by the ability of bacteria to overcome diverse pharmacological regimens. Studies have indicated that the development of resistance is less frequent for some membrane-active agents than for other classes of antibiotics.[3] We evaluated the potential for clinical strains of S. aureus to develop antimicrobial resistance in vitro using serial passages of the bacteria[8b] with a representative cyclic peptoid chosen from this study. Serial dilutions of the peptoid C3 (Figure 2) were incubated with S. aureus LAC and MW2, which are clinical strains of methicillinresistant S. aureus (MRSA), as well as with S. aureus Newman and RN6734, which are methicillin-sensitive strains of S. aureus (MSSA).[25] Peptoid C3 was chosen because of its potent antimicrobial activity and relatively small size (~ 1000 Da, see table S1, Supporting Information), which may be important for pharmacological considerations. The MIC values were determined each day after daily propagation of surviving isolates with fresh media and compound dilutions. The experiment
ChemMedChem 2012, 7, 114 122
Conclusions
Macrocycles are a promising but perhaps underexploited class of potential therapeutic compounds.[26] We report that peptoid macrocycles are potent and selective antimicrobials. These compounds are active against Gram-negative and Gram-positive bacteria and are non-hemolytic toward human erythrocytes. We observed a general trend of enhanced antimicrobial activity with macrocyclization. These results suggest that for this class of oligomer compounds, enhancing conformational order is beneficial in attaining antimicrobial activity. In addition, we report that potent antimicrobial activity of a cyclic peptoid hexamer against clinical strains of S. aureus can be established, while impeding the emergence of antimicrobial resistance in vitro. Further studies are being conducted to improve the potency and selectivity of these molecules, as well as studies to determine their mechanism of action.
General
Solvents and reagents purchased from commercial sources were used without further purification. Abbreviations for reagents are as follows: 9-fluorenylmethoxycarbonyl (Fmoc); tert-butoxycarbonyl (Boc); benzotriazole-1-yloxytrispyrrolidinophosphonium hexafluorophosphate (PyBOP); trifluoroacetic acid (TFA); 1,1,1,3,3,3-hexafluoroisopropanol (HFIP); N,N-dimethylformamide (DMF); N,N-diisopropylcarbodiimide (DIC); diisopropylethylamine (DIEA); N-methylmorpholine (NMM); O-benzotriazole-N,N,N,N-tetramethyluronium hex-
www.chemmedchem.org
119
MED
afluorophosphate (HBTU). Rink amide resin, 2-chlorotrityl chloride resin, PyBOP, HBTU, and amino acids used for the synthesis of peptides were purchased from Novabiochem (San Diego, CA, USA). Typical resin loading levels used for solid-phase peptoid and peptide syntheses are 0.49 and 1.1 mmol g1 for Rink amide and 2chlorotrityl chloride resins, respectively. Bromoacetic acid was purchased from SigmaAldrich. DIC and DIEA were purchased from Chem Impex International. TFA was purchased from Acros Organics.
DMF (3 1.5 mL) and CH2Cl2 (3 1.5 mL); N-termini-capped peptoid oligomers were cleaved with a solution of TFA/H2O (3 mL, 95 % v/ v) at room temperature. The cleavage was conducted in a glass vial with constant agitation for 120 min. The cleavage cocktail was removed under a stream of N2 gas. The final product was dissolved in 5 mL 50 % CH3CN/H2O and filtered with a 0.5 mm stainless steel fritted syringe tip filter (Upchurch Scientific). Peptoid oligomers were analyzed on a C18 reversed-phase analytical RP-HPLC column at room temperature (Peeke Scientific, 5 mm, 120 , 2.0 50 mm) using a Beckman Coulter System Gold instrument. A linear gradient of 5!95 % CH3CN/H2O (0.1 % TFA) over 20 min was used at a flow rate of 0.7 mL min1. Preparative RP-HPLC was then performed on a Delta-Pak C18 column (Waters, 15 mm, 100 , 25 100 mm) with a linear gradient of 5!95 % CH3CN/H2O (0.1 % TFA) over 60 min at a flow rate of 5 mL min1. Mass spectrometry was performed on an Agilent 100 Series LCMSD Trap XCT (Agilent Technologies). Purified HPLC fractions were then frozen and lyophilized. For the generation of cyclic peptoid oligomers, 2-chlorotrityl chloride resin (200 mg) was washed in CH2Cl2 (2 2 mL), followed by swelling in DMF (2 mL). The first monomer was added manually by allowing bromoacetic acid (37 mg, 0.27 mmol) to react with DIEA (189 mL, 1.08 mmol) in CH2Cl2 (2 mL) on a shaker platform for 30 min at room temperature, followed by extensive washes with CH2Cl2 (5 2 mL) and DMF (5 2 mL). After the initial manual loading of bromoacetic acid, the first sub-monomer displacement step and all subsequent bromoacetylation and amine displacement steps were performed on a robotic synthesizer until desired oligomer lengths were obtained. The automated bromoacetylation steps were performed by adding bromoacetic acid (1660 mL of a 1.2 m solution in DMF) and DIC (400 mL). Next, a solution of sub-monomer (2 mL of a 1 m solution in DMF, 2 mmol) was added to introduce the side chain by nucleophilic displacement of bromide. The mixture was agitated for 20 min, drained, washed with DMF (3 2 mL), and washed with CH2Cl2 (3 2 mL). The peptoid-bound resin was cleaved in HFIP/ CH2Cl2 (2 mL, 20 % v/v) at room temperature for 30 min. The cleavage cocktail was evaporated under a stream of N2 gas, and the final product was dissolved in 50 % CH3CN/H2O (5 mL), frozen, and lyophilized overnight to remove any traces of HFIP in the sample. Linear precursors were subjected to cyclization without purification. Cyclization reactions were conducted with the crude linear precursors in dry deoxygenated DMF. Typically, the linear oligomer (12 mmol) was suspended in DMF (5.25 mL) with freshly prepared solutions of PyBOP (375 mL of a 96 mm solution in DMF) and DIEA (375 mL of a 192 mm solution in DMF). The reaction vessel was flushed with N2 and sealed with Parafilm to exclude air. The reaction was allowed to proceed for 5 min at room temperature, and the reaction mixture (10 mL) was diluted with 50 % CH3CN/H2O (140 mL) to quench the reaction. The diluted sample was then analyzed by RP-HPLC and MS as described above. Preparative RP-HPLC was then conducted as previously described. Purified relevant fractions were then frozen and lyophilized overnight. Boc protecting groups were subsequently removed with TFA/H2O (5 mL, 95 % v/v) over 30 min at room temperature. A stream of N2 gas was used to decrease the volume of TFA solution to 10 % of its original volume, and the concentrated sample was dissolved in 50 % TFA/H2O (15 mL), frozen, and lyophilized.
Generation of peptoid sub-monomers

For the generation of N-(3-aminopropyl)glycine (Nap), N-(4-aminobutyl)glycine (Nab), N-(6-aminohexyl)glycine (Nah), and N-(4-guanidinobutyl)glycine (Ngb) residues, N-Boc-1,3-diaminopropane, NBoc-1,4-diaminobutane, N-Boc-1,6-diaminohexane, and N,N-bisBoc-agmatine were used as the sub-monomer reagents, respectively. N-Boc-1,3-diaminopropane, N-Boc-1,4-diaminobutane, and N-Boc1,6-diaminohexane were synthesized from di-tert-butyldicarbonate (Alfa Aesar) and corresponding diamino alkanes [1,3-diaminopropane (SigmaAldrich), 1,4-diaminobutane (Alfa Aesar), and 1,6-diaminohexane (Acros)] following a previously reported protocol.[27] N,N-bis-Boc-agmatine was synthesized from 1,4-diaminobutane and N,N-bis-Boc-methylisothiourea (SigmaAldrich) following previously reported procedures.[28] For the generation of N-(phenylmethyl)glycine (Npm), N-(naphthylmethyl)glycine (Nnm), N-(2,2-diphenylethyl)glycine (Ndp), N-isopropylglycine (Nip), and N-isobutylglycine (Nib) residues, benzylamine (Alfa Aesar), 1-naphthylmethylamine (SigmaAldrich), 2,2-diphenylethylamine (Acros), isopropylamine (Alfa Aesar), and isobutylamine (SigmaAldrich) were used as the sub-monomer reagents, respectively.
Synthesis and purification of peptoid oligomers

Previously reported solid-phase peptoid sub-monomer synthesis protocols were used with modifications in reaction times and washing conditions.[10a] Peptoid synthesis was performed with an automated synthesizer (Charybdis Technologies Inc., San Diego, CA, USA) at room temperature. For the generation of linear peptoid oligomers, Rink amide resin (100 mg) was washed in CH2Cl2 (2 1.5 mL), followed by swelling in DMF (1.5 mL) for 20 min. The swelling step was performed twice. The Fmoc protecting group was removed from the resin by the addition of piperidine/DMF (30 % v/v). The mixture was agitated for 10 min, drained, and the piperidine treatment was repeated, followed by extensive washes with DMF (5 1.5 mL). After removal of the Fmoc group from the Rink amide resin, the amino-functionalized resin was bromoacetylated by adding bromoacetic acid (830 mL of a 1.2 m solution in DMF) and DIC (200 mL). The mixture was agitated for 20 min, drained, and washed with DMF (3 1.5 mL). Next, a solution of a desired amine sub-monomer reagent (1 mL of a 1 m solution in DMF, 1 mmol) was introduced. The mixture was agitated for 20 min, drained, washed with DMF (3 1.5 mL) and CH2Cl2 (3 1.5 mL). These bromoacetylation and amine displacement steps were repeated until the desired oligomer lengths and sequences were obtained. After the last coupling step, the acetylated N termini were generated using Ac2O (1.5 mL of a 1 m solution in DMF) and DIEA (50 mL). The mixture was agitated for 20 min, drained, and washed with
120
www.chemmedchem.org
ChemMedChem 2012, 7, 114 122
Peptoid Oligomers Synthesis and purification of peptides

Gramicidin S (CGS) was synthesized using standard Fmoc solidphase synthesis protocols. 2-Chlorotrityl chloride resin (200 mg) was washed in CH2Cl2 (2 2 mL), followed by swelling for 20 min in CH2Cl2 (2 mL). The first amino acid was added manually by allowing Fmoc-Phe (0.27 mmol) and DIEA (142 mL) to react in CH2Cl2 (2 mL) on a shaker platform for 30 min at room temperature, followed by extensive washes with CH2Cl2 (5 2 mL) and DMF (5 2 mL). Resin loaded with Fmoc-Phe was incubated twice with piperidine/DMF (2 mL, 20 % v/v) on a shaker platform for 15 min at room temperature, followed by extensive washes with DMF (5 2 mL). After manual loading of Phe, all subsequent amino acid addition and Fmoc deprotection steps were performed on a robotic synthesizer until the desired peptide length was obtained. The automated amino acid addition step was performed by adding Fmoc-amino acid (1 mL, 0.5 m in DMF), HBTU (1 mL, 0.5 m in DMF) and NMM (1 mL, 1.5 m in DMF, Alfa Aesar). The mixture was agitated for 30 min, drained, and washed with DMF (3 2 mL). The resin was then incubated twice with piperidine/DMF (2 mL, 20 % v/v) for 15 min. The reaction was drained and washed with DMF (3 2 mL) and with CH2Cl2 (3 2 mL) at room temperature. The cleavage was conducted with HFIP/CH2Cl2 (20 % v/v) with constant agitation for 30 min. The cleavage cocktail was evaporated under a stream of N2 gas, and the final product was dissolved in CH3CN/H2O (70 %, 5 mL) prior to freezing and lyophilization. Cyclization of gramicidin S was performed as described above for the cyclization of peptoid oligomers. The linear version of gramicidin S (LGS) was prepared by following the same procedure, except for the use of Rink amide resin (100 mg) and treatment with piperidine in DMF (20 % v/v) prior to the addition of Fmoc-Phe amino acid. The peptide N terminus was also acetylated using Ac2O (1.5 mL of a 1 m solution in DMF) and DIEA (50 mL), and cleaved from the resin using TFA/H2O (95 % v/v). Peptide oligomers were purified and analyzed by as described above. All products were brought to > 95 % purity by RP-HPLC and lyophilized to dry powders prior to biological testing.
[12] [7] [8] Chem. 2002, 9, 811 822; c) J. A. Patch, A. E. Barron, Curr. Opin. Chem. Biol. 2002, 6, 872 877; d) J. A. Robinson, Curr. Opin. Chem. Biol. 2011, 15, 379 386. T. L. Raguse, E. A. Porter, B. Weisblum, S. H. Gellman, J. Am. Chem. Soc. 2002, 124, 12774 12785. a) G. N. Tew, D. H. Liu, B. Chen, R. J. Doerksen, J. Kaplan, P. J. Carroll, M. L. Klein, W. F. DeGrado, Proc. Natl. Acad. Sci. USA 2002, 99, 5110 5114; b) S. Choi, A. Isaacs, D. Clements, D. Liu, H. Kim, R. W. Scott, J. D. Winkler, W. F. DeGrado, Proc. Natl. Acad. Sci. USA 2009, 106, 6968 6973. a) N. Srinivas, P. Jetter, B. J. Ueberbacher, M. Werneburg, K. Zerbe, J. Steinmann, B. Van der Meijden, F. Bernardini, A. Lederer, R. L. Dias, P. E. Misson, H. Henze, J. Zumbrunn, F. O. Gombert, D. Obrecht, P. Hunziker, S. Schauer, U. Ziegler, A. Kach, L. Eberl, K. Riedel, S. J. DeMarco, J. A. Robinson, Science 2010, 327, 1010 1013; b) A. Bragonzi, Sci. Transl. Med. 2010, 2, 21ps9. a) G. M. Figliozzi, R. Goldsmith, S. C. Ng, S. C. Banville, R. N. Zuckermann, Method Enzymol. 1996, 267, 437 447; b) R. J. Simon, R. S. Kania, R. N. Zuckermann, V. D. Huebner, D. A. Jewell, S. Banville, S. Ng, L. Wang, S. Rosenberg, C. K. Marlowe, D. C. Spellmeyer, R. Y. Tan, A. D. Frankel, D. V. Santi, F. E. Cohen, P. A. Bartlett, Proc. Natl. Acad. Sci. USA 1992, 89, 9367 9371. a) J. M. Holub, M. J. Garabedian, K. Kirshenbaum, Mol. BioSyst. 2011, 7, 337 345; b) S. M. Miller, R. J. Simon, S. Ng, R. N. Zuckermann, J. M. Kerr, W. H. Moos, Drug Dev. Res. 1995, 35, 20 32; c) J. E. Murphy, T. Uno, J. D. Hamer, F. E. Cohen, V. Dwarki, R. N. Zuckermann, Proc. Natl. Acad. Sci. USA 1998, 95, 1517 1522; d) Y. Utku, E. Dehan, O. Ouerfelli, F. Piano, R. N. Zuckermann, M. Pagano, K. Kirshenbaum, Mol. BioSyst. 2006, 2, 312 317; e) J. T. Nguyen, C. W. Turck, F. E. Cohen, R. N. Zuckermann, W. A. Lim, Science 1998, 282, 2088 2092; f) S. Ng, B. Goodson, A. Ehrhardt, W. H. Moos, M. Siani, J. Winter, Bioorg. Med. Chem. 1999, 7, 1781 1785; g) N. P. Chongsiriwatana, J. A. Patch, A. M. Czyzewski, M. T. Dohm, A. Ivankin, D. Gidalevitz, R. N. Zuckermann, A. E. Barron, Proc. Natl. Acad. Sci. USA 2008, 105, 2794 2799; h) P. A. Wender, D. J. Mitchell, K. Pattabiraman, E. T. Pelkey, L. Steinman, J. B. Rothbard, Proc. Natl. Acad. Sci. USA 2000, 97, 13003 13008; i) C. W. Wu, S. L. Seurynck, K. Y. C. Lee, A. E. Barron, Chem. Biol. 2003, 10, 1057 1063; j) M. M. Reddy, T. Kodadek, Proc. Natl. Acad. Sci. USA 2005, 102, 12672 12677; k) B. Yoo, K. Kirshenbaum, Curr. Opin. Chem. Biol. 2008, 12, 714 721; l) R. N. Zuckermann, T. Kodadek, Curr. Opin. Mol. Ther. 2009, 11, 299 307. a) B. P. Mowery, A. H. Lindner, B. Weisblum, S. S. Stahl, S. H. Gellman, J. Am. Chem. Soc. 2009, 131, 9735 9745; b) E. A. Porter, B. Weisblum, S. H. Gellman, J. Am. Chem. Soc. 2005, 127, 11516 11529; c) J. R. Lai, R. F. Epand, B. Weisblum, R. M. Epand, S. H. Gellman, Biochemistry 2006, 45, 15718 15730. I. S. Radzishevsky, S. Rotem, D. Bourdetsky, S. Navon-Venezia, Y. Carmeli, A. Mor, Nat. Biotechnol. 2007, 25, 657 659. a) C. W. Wu, K. Kirshenbaum, T. J. Sanborn, J. A. Patch, K. Huang, K. A. Dill, R. N. Zuckermann, A. E. Barron, J. Am. Chem. Soc. 2003, 125, 13525 13530; b) G. L. Butterfoss, P. D. Renfrew, B. Kuhlman, K. Kirshenbaum, R. Bonneau, J. Am. Chem. Soc. 2009, 131, 16798 16807. a) N. H. Shah, G. L. Butterfoss, K. Nguyen, B. Yoo, R. Bonneau, D. L. Rabenstein, K. Kirshenbaum, J. Am. Chem. Soc. 2008, 130, 16622 16632; b) K. Kirshenbaum, A. E. Barron, R. A. Goldsmith, P. Armand, E. K. Bradley, K. T. V. Truong, K. A. Dill, F. E. Cohen, R. N. Zuckermann, Proc. Natl. Acad. Sci. USA 1998, 95, 4303 4308; c) B. Paul, G. L. Butterfoss, M. G. Boswell, P. D. Renfrew, F. G. Yeung, N. H. Shah, C. Wolf, R. Bonneau, K. Kirshenbaum, J. Am. Chem. Soc. 2011, 133, 10910 10919. a) P. Armand, K. Kirshenbaum, R. A. Goldsmith, S. Farr-Jones, A. E. Barron, K. T. V. Truong, K. A. Dill, D. F. Mierke, F. E. Cohen, R. N. Zuckermann, E. K. Bradley, Proc. Natl. Acad. Sci. USA 1998, 95, 4309 4314; b) P. Armand, K. Kirshenbaum, A. Falicov, J. R. L. Dunbrack, K. A. Dill, R. N. Zuckermann, F. E. Cohen, Fold Des. 1997, 2, 369 375. a) R. Kapoor, M. W. Wadman, M. T. Dohm, A. M. Czyzewski, A. M. Spormann, A. E. Barron, Antimicrob. Agents Chemother. 2011, 55, 3054 3057; b) J. A. Patch, A. E. Barron, J. Am. Chem. Soc. 2003, 125, 12092 12093. a) S. B. Y. Shin, B. Yoo, L. J. Todaro, K. Kirshenbaum, J. Am. Chem. Soc. 2007, 129, 3218 3225; b) B. Yoo, S. B. Shin, M. L. Huang, K. Kirshenbaum, Chem. Eur. J. 2010, 16, 5528 5537. a) K. Kirshenbaum, S. B. Shin, Peptoid Oligomers, Their Preparation, Pharmaceutical Compositions, and Their Use in Treatment and Prevention of Bacterial, Viral, and Fungal Infections, PCT Intl. Appl. WO
[9]
[10]
[11]
Acknowledgements
M.A.B. is supported by an American Heart Association predoctoral fellowship (10PRE3420022). Research in the Torres Laboratory is enabled by the New York University School of Medicine Development Funds (V.J.T.). The authors gratefully acknowledge support from the NSF (CAREER Award CHE-0645361 to K.K.) and the NIH (Research Facilities Improvement Grant C060RR-165720). Keywords: antibiotic resistance antimicrobial peptides foldamers macrocycles peptidomimetics
[1] a) I. M. Gould, Int. J. Antimicrob. Agents 2009, 34, S2 S5; b) H. W. Boucher, G. H. Talbot, J. S. Bradley, J. E. Edwards, D. Gilbert, L. B. Rice, M. Scheld, B. Spellberg, J. Bartlett, Clin. Infect. Dis. 2009, 48, 1 12. [2] D. J. Payne, M. N. Gwynn, D. J. Holmes, D. L. Pompliano, Nat. Rev. Drug Discovery 2007, 6, 29 40. [3] J. G. Hurdle, A. J. ONeill, I. Chopra, R. E. Lee, Nat. Rev. Microbiol. 2011, 9, 62 75. [4] a) K. A. Brogden, Nat. Rev. Microbiol. 2005, 3, 238 250; b) R. E. Hancock, H. G. Sahl, Nat. Biotechnol. 2006, 24, 1551 1557. [5] H. G. Boman, Annu. Rev. Immunol. 1995, 13, 61 92. [6] a) A. S. Ripka, D. H. Rich, Curr. Opin. Chem. Biol. 1998, 2, 441 452; b) D. L. Steer, R. A. Lew, P. Perlmutter, A. I. Smith, M. I. Aguilar, Curr. Med.
[13] [14]
[15]
[16]
[17]
[18]
[19]
ChemMedChem 2012, 7, 114 122
www.chemmedchem.org
121
MED
[20] [21] [22] 2010098843 A2 20100902, 2010; b) D. Comegna, M. Benincasa, R. Gennaro, I. Izzo, F. De Riccardis, Bioorg. Med. Chem. 2010, 18, 2010 2018. R. N. Zuckermann, J. M. Kerr, S. B. H. Kent, W. H. Moos, J. Am. Chem. Soc. 1992, 114, 10646 10647. a) G. F. Gause, M. G. Brazhnikova, Nature 1944, 154, 703 703; b) A. M. Liquori, F. Conti, Nature 1968, 217, 635. Clinical and Laboratory Standards Institute, Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically, 7th Ed., Approved Standard CLSI document M07-A7, CLSI, Wayne PA (USA), 2006. P. Wadhwani, S. Afonin, M. Ieromino, J. Buerck, A. S. Ulrich, J. Org. Chem. 2006, 71, 55 61. a) K. Kuroda, G. A. Caputo, W. F. DeGrado, Chem. Eur. J. 2009, 15, 1123 1133; b) E. F. Palermo, K. Kuroda, Appl. Microbiol. Biotechnol. 2010, 87, 1605 1615; c) G. N. Tew, R. W. Scott, M. L. Klein, W. F. Degrado, Acc. Chem. Res. 2010, 43, 30 39.
[25] a) A. L. Dumont, T. K. Nygaard, R. L. Watkins, A. Smith, L. Kozhaya, B. N. Kreiswirth, B. Shopsin, D. Unutmaz, J. M. Voyich, V. J. Torres, Mol. Microbiol. 2011, 79, 814 825; b) H. F. Chambers, F. R. Deleo, Nat. Rev. Microbiol. 2009, 7, 629 641. [26] E. M. Driggers, S. P. Hale, J. Lee, N. K. Terrett, Nat. Rev. Drug Discovery 2008, 7, 608 624. [27] C. Dardonville, C. Fernandez-Fernandez, S. L. Gibbons, G. J. Ryan, N. Jagerovic, A. M. Gabilondo, J. J. Meana, L. F. Callado, Bioorg. Med. Chem. 2006, 14, 6570 6580. [28] F. Sarabia, A. Sanchez-Ruiz, S. Chammaa, Bioorg. Med. Chem. 2005, 13, 1691 1705.
[23] [24]
Received: July 21, 2011 Revised: September 8, 2011 Published online on September 23, 2011
122
www.chemmedchem.org
ChemMedChem 2012, 7, 114 122
DOI: 10.1002/cmdc.201100351
Discovery of a Pharmacologically Active Antagonist of the Two-Pore-Domain Potassium Channel K2P9.1 (TASK-3)
Craig A. Coburn,*[a] Yunfu Luo,[b] Mingxiang Cui,[b] Jiabing Wang,[a] Richard Soll,[b] Jingchao Dong,[b] Bin Hu,[b] Michael A. Lyon,[c] Vincent P. Santarelli,[d] Richard L. Kraus,[d] Yun Gregan,[d] Yi Wang,[d] Steven V. Fox,[e] Jacquelyn Binns,[e] Scott M. Doran,[e] Duane R. Reiss,[f] Pamela L. Tannenbaum,[f] Anthony L. Gotter,[f] Peter T. Meinke,[g] and John J. Renger[f]
TWIK-related acid-sensitive K + (K2P9.1, TASK-3) ion channels have the capacity to regulate the activity of neuronal pathways by influencing the resting membrane potential of neurons on which they are expressed. The central nervous system (CNS) expression of these channels suggests potential roles in neurologic disorders, and it is believed that the development of TASK-3 antagonists could lead to the therapeutic treatment of a number of neurological conditions. While a therapeutic potential for TASK-3 channel modulation exists, there are only a few documented examples of potent and selective small-molecule channel blockers. Herein, we describe the discovery and lead optimization efforts for a novel series of TASK-3 channel antagonists based on a 5,6,7,8-tetrahydropyrido[4,3-d]pyrimidine high-throughput screening lead from which a subseries of potent and selective inhibitors were identified. One compound was profiled in detail with respect to its physical properties and demonstrated pharmacological target engagement as indicated by its ability to modulate sleep architecture in rodent electroencephalogram (EEG) telemetry models.
Introduction
Potassium channels are a diverse class of proteins that play an essential role in many physiological processes by allowing passive permeability of potassium ions (K + ) across biological membranes. There are three major families of potassium channel proteins each with structural and functional distinctions that are characterized by the number of transmembrane (TM) and pore (P) domains in each subunit. The first and second families of potassium channels are the voltage-dependent potassium channels (Kv) and the inwardly rectifying potassium channels (Kir). The members of the third family are known as the KCNK or K2P (two-pore-domain potassium) proteins, which generate well-modulated background K + currents that contribute to the resting membrane potential in a wide variety of cells.[1] Currently, the human K2P channel super family has 15 members that can be divided into six main subfamilies based on their structure and functional properties. The TWIK-related acid-sensitive K + (TASK) ion channels are pH-sensitive subunits of the KCNK family that are closely related and coexpressed in many brain regions.[2] The TASK channels are generally referred to as TASK-1 (KCNK3 or K2P3.1), TASK-3 (KCNK9 or K2P9.1), and the nonfunctional TASK-5 (KCNK15 or K2P15.1).[3] TASK-1 and TASK-3 channels share the closest homology within this family with greater than 50 % amino acid identity. TASK channels are distributed throughout the central nervous system (CNS) as well as the periphery and are thought to play key roles in a number of neuronal processes. TASK-3 is particularly abundant in the hippocampus, cerebellum and cortex, and in specific nuclei including the locus coeruleus, parChemMedChem 2012, 7, 123 133
aventricular nuclei of thalamus and the dorsal raphe.[4] TASK-3 channel activity has been shown to regulate both neurotransmitter release as well as mediating the effects of neurotransmitter activation including the activity of 5-HT-releasing neurons of the dorsal raphe.[5] 5-HT-mediated gamma-aminobutyric acid (GABA) release by entorhinal cortical neurons also occurs through the inhibition of TASK-3 and contributes to thalamocortical neuron activity, thus having the potential to regulate sleep-stage progression and cognition.[6] It is therefore believed that the development of TASK-3 inhibitors could lead to
[a] Dr. C. A. Coburn, Dr. J. Wang External Medicinal Chemistry, Merck & Co. 770 Sumneytown Pike, West Point, PA 19486 (USA) E-mail: craig_coburn@merck.com [b] Y. Luo, Dr. M. Cui, Dr. R. Soll, Dr. J. Dong, Dr. B. Hu Department of Medicinal Chemistry, WuXi AppTec 288 Fute Zhong Road, Shanghai, 200131 (China) [c] Dr. M. A. Lyon Array BioPharma, Inc., 3200 Walnut Street, Boulder, CO 80301 (USA) [d] Dr. V. P. Santarelli, Dr. R. L. Kraus, Y. Gregan, Y. Wang Department of In Vitro Pharmacology, Discovery and Preclinical Sciences Merck & Co., 770 Sumneytown Pike, West Point, PA 19486 (USA) [e] S. V. Fox, J. Binns, Dr. S. M. Doran Department of Central Pharmacology, Discovery and Preclinical Sciences Merck & Co., 770 Sumneytown Pike, West Point, PA 19486 (USA) [f] D. R. Reiss, Dr. P. L. Tannenbaum, Dr. A. L. Gotter, Dr. J. J. Renger Department of Neuroscience, Discovery and Preclinical Sciences Merck & Co., 770 Sumneytown Pike, West Point, PA 19486 (USA) [g] Dr. P. T. Meinke External Medicinal Chemistry, Merck & Co. 126 E. Lincoln Ave, Rahway, NJ 07065 (USA)
123
MED
therapeutic agents against neurological conditions including sleep disorders, neurodegeneration, cognitive impairment, Parkinsons disease, Huntingtons disease, or major depressive disorder. In fact, TASK-3 knockout (KO) mice exhibit augmented circadian amplitude, suppressed REM sleep, and exhibit resistance to despair behavior in animal models, indicating the potential of TASK-3 as an antidepressant target.[7] While a therapeutic potential for TASK-3 channel modulation exists, there are only a few documented examples of smallmolecule inhibitors.[8] All members of the TASK family are nonselectively inhibited by the cannabinoids and local anesthetics such as bupivacaine (IC50 ~ 40 mm) and lidocaine (IC50 ~ 220 mm). TASK-3 is selectively inhibited by the cationic dye ruthenium red (IC50 = 0.70 mm).[9] In 2007 SanofiAventis claimed that a series of known Kv1.5 channel blockers have sub-micromolar TASK-1 activity.[10] A recent report from the Decher lab described A-1899 (1) as a potent and selective small-molecule TASK-1 antagonist.[11] The IC50 value for compound 1 versus TASK-1 expressed in CHO cells was reported to be 7 nm, while TASK-3 channels were blocked at an IC50 = 70 nm.
C. A. Coburn et al. corporate compound collection (~ 1.8 MM compounds) was screened, and initial hits from this FLIPR assay were identified and further assessed in a single-concentration exposure electrophysiological assay using the IonWorks Quattro platform. A select set of these hits were then titrated for potency in eightpoint doseresponse curves also using the IonWorks Quattro system.[12] From this subset of hits, we identified the 5,6,7,8tetrahydropyrido[4,3-d]pyrimidine (THPP) analogue 2 as a moderately active TASK-3 antagonist (IC50 = 0.65 mm). Compound 2 showed no activity against other K + channels such as hERG (IC50 > 15 mm) or the voltage-gated potassium channel Kv1.5 (IC50 > 10 mm). Compound 2 was originally synthesized as part of a three-component, 15 5 40 compound library for our sample repository acquisition committee (SRAC) initiative. As such, compounds from this library provided us with an initial observation of the structureTASK-3 activity relationships, which resulted in prioritization of this series toward an optimization candidate (Figure 1).
Figure 1. Summary of the structureTASK-3 activity relationships elucidated from HTS of the 3000-membered 15 (A) 5 (BC) 40 (D) THPP compound library.
Given the vast range of biologically relevant functions of TASK-3, the development of more potent and selective TASK-3 inhibitors would help facilitate the evaluation of its functional significance and clinical utility. Here, we describe the discovery of a novel class of potent and selective TASK-3 antagonists, along with preliminary structureactivity relationships (SAR), developed from the high-throughput screening (HTS) screening hit 2. This work ultimately led to the first positive pharmacological proof of concept study of small-molecule TASK-3 antagonism by demonstrating significant effects in electroencephalogram (EEG) patterns in rodents.
SAR gleaned from the initial library showed a preference for the 6,6-tetrahydropyridopyridine scaffold over a wide variety of homologues and isosteres (BC). Substitution on the saturated C ring was not represented in the initial library. Substitution at the C-4 position (A) tolerated a wide variety of substituents, while the C-2 position showed a clear preference for hydrogen over alkyl, aryl, heteroaryl, and dialkylamino groups. Region D showed a stronger preference for amides and sulfonamides over alkyl and urea groups, while analogues having an aromatic ring proximal to the scaffold were more potent than the corresponding alkyl or cycloalkyl compounds. Chemistry and SAR Efforts aimed at exploring the SAR and improving the potency of the lead compound (2) focused on each of the three subunits described above (Figure 1) and a modular synthetic scheme was designed to provide easy access for analogue production. Modification of the Eastern D region: A variety of conditions for the cyclocondensation of R1-substituted amidines with appropriately X- and Y-substituted N-Boc-3-carbethoxy-4-piperidones were explored to identify the optimal base and solvent for the requisite condensation reaction (Scheme 1, step b). Ethanol or THF with an aqueous solution of potassium carbonate at room temperature emerged as the most appropriate and
ChemMedChem 2012, 7, 123 133

High-throughput screen Our efforts began with an HTS campaign to discover structurally unique TASK-3 antagonists to serve as a springboard for a lead optimization medicinal chemistry program. As such, we developed an HTS assay using a human TASK-3 expressing cell line (HEK293) in which activity was measured by a voltage sensitive fluorescent dye approach (FLIPR assay). The 2008 Merck
124
www.chemmedchem.org
TASK-3 Antagonists
Scheme 1. General scheme for the synthesis of THPP analogues: a) 1. H2, Boc2O, 20 % Pd(OH)2/C, RT, o/n; 2. Et3N, 0 8C, 2 h; b) R1C(=NH)NH2HX, K2CO3, EtOH, water, RT, o/n; c) PPh3, NCS, dioxane, 50 8C, 18 h; d) ethyl isonipecotate, RT, 16 d; e) HCl (4 n), dioxane, RT, 118 h.
general protocol for the formation of the THPP scaffolds 5. The resulting pyrimidinone functional group necessitated conversion to the corresponding chloropyrimidines 6 in order to facilitate the SNAr displacement reactions with the various amines. It was therefore critical to find a chlorination method that was mild enough to effect the halogenation reaction while not adversely affecting the tert-butoxycarbonyl protecting group. Sugimoto and co-workers accomplished a requisitely mild and facile halogenation of a series of hydroxy heterocycles using triphenylphosphine and N-halosuccinimides.[13] Specifically, by heating a suspension of pyrimidinones 5 with N-chlorosuccinimide (NCS) and triphenylphosphine in dioxane at 50 8C for 18 h, we reproducibly obtained 5060 % isolated yields of the desired chloropyrimidines 6. With the syntheses of the key intermediates for the SNAr reactions complete, we turned our attention to the preparation of the penultimate scaffold 8. Standard conditions for the introduction of the isonipecotic ester required stirring a solution of the chloropyrimidines in tetrahydrofuran (THF) with two equivalents of the amine at room temperature. Intermediates 7 were then deprotected with 4 n hydrochloric acid in dioxane and neutralized to afford free amines 8 in good overall yields (Scheme 1). The approach to final product synthesis involved cleavage of a resin-bound activated ester for amide (12) formation, solution phase urea (10) and sulfonamide (11) synthesis with solidsupported scavenging agents and alkylations (9), which were effected through a reductive alkylation protocol using a solidsupported reducing agent as shown in Scheme 2. This modular approach allowed us to rapidly define the SAR at the 6-position of the THPP scaffold, and the results are shown in Table 1. Small alkyl groups, such as methyl (9 a) or n-butyl (9 b), were not tolerated while the introduction of a benzylic residue showed activity comparable to the HTS lead compound (9 c; IC50 = 0.7 mm). Urea analogues 10 showed a similar SAR profile with small alkyl groups showing weak TASK-3 activity while benzyl and p-biphenyl compounds displaying low micromolar IC50 values.
ChemMedChem 2012, 7, 123 133
Scheme 2. End-game synthesis for library production of THPP analogues: a) PS-DCC, CH2Cl2, RT, 8 d; b) ZSO2Cl, PS-DIPEA, CH2Cl2, RT, o/n; c) ZCHO, MP(CN)BH3, AcOH/MeOH (10 %), RT, o/n; d) ZNCO, PS-DIPEA, CH2Cl2, RT, o/n.
Table 1. Structureactivity relationship data for 5,6,7,8-tetrahydropyrido[4,3-d]pyrimidine (THPP) analogues with modifications to the Eastern moiety.
Compd 9a 9b 9c 10 a 10 b 10 c 11 a 11 b 11 c 11 d 12 a 12 b 12 c 12 d 12 e 12 f
Linker CH2 CH2 CH2 CONH CONH CONH SO2 SO2 SO2 SO2 CO CO CO CO CO CO
Z H CH2CH2CH3 Ph CH3 CH2Ph p-biphenyl CH2Ph 4-Cl-Ph 3-Cl-Ph 2-Cl-Ph Ph 4-MeO-Ph 4-Br-Ph 4-Me-Ph 4-cHx-Ph 4-PhPh
IC50 [mm][a] TASK-3 > 100 > 100 0.71 0.08 > 100 3.4 0.4 9.6 0.5 3.6 0.4 3.9 0.5 0.7 0.07 0.9 0.1 3.6 0.77 1.6 0.18 0.43 0.06 0.31 0.08 0.12 0.01 0.074 0.009
[a] Values represent the mean SD of three independent experiments.
Benzylic sulfonamides 11 a and para-substituted phenyl sulfonamides 11 b showed weak TASK-3 activity, while the orthoand meta-substituted benzenesulfonamides (11 c, d) exhibited an enhanced profile with activity in the sub-micromolar range.
www.chemmedchem.org
125
MED
Incorporation of a carbonyl linker (12) afforded the most promising results as these analogues showed good activity in the TASK-3 IonWorks Quattro electrophysiology assay. An in-depth examination of the SAR within the benzamide analogues revealed that substitution with lipophilic groups para to the amide carbonyl group was preferred with the potency trend favoring Ph < Chx < Me % Br < Cl < F < OMe < H (Table 1). Modification of the core BC scaffold: Several analogues were synthesized to probe the tolerability of substitution on the tetrahydropyridine scaffold. The synthetic route employed for the preparation of analogues 13 was similar to that described in Scheme 1 after construction of the appropriately substituted N-Boc piperidinones. For the 2-substituted THPP analogues 14, we also followed the chemistry described in Scheme 1 but employed acetamidine or benzamidine in the cyclocondensation reaction with N-Boc-3-carbethoxy-4-piperidone. It should be noted that gentle warming at 45 8C was required to drive these two reactions to completion. Substitutions of alkyl groups on the saturated C ring were not tolerated in any examples as evidenced by the slightly weaker potencies of 13 a, b when compared to the unsubstituted parent compound 12 f. Similarly, methyl and phenyl substituents at the R1-position of the aromatic B ring resulted in a loss in potency (Table 2).
C. A. Coburn et al.
Scheme 3. General scheme for the synthesis of northern analogues: a) HCl (4 n), dioxane, 0 8C!RT, 5 h; b) BOP reagent, DIPEA, RT, 4 h; c) R1R2NH, DMSO, K2CO3, 908C, 4 h.
Table 2. Structureactivity relationship data for 5,6,7,8-tetrahydropyrido[4,3-d]pyrimidine (THPP) analogues with modifications to the central ring.
Compd 12 f 13 a 13 b 14 a 14 b
R1 H H H Ph Me
X H, H Me, Me H, H H, H H, H
Y H, H H, H Me, Me H, H H, H
IC50 [mm][a] TASK-3 0.074 0.009 0.57 0.03 0.65 0.03 92 4 0.26 0.05
shown). The six-membered cyclic amines were favored over pyrrolidines and azetidines (Table 3), and those analogues that possessed an hydrogen-bond acceptor (HBA) at the piperidine 4-position were more potent than the unsubstituted piperidine analogue 17 c. To further delineate this finding, we explored the SAR at the 4-position of the piperidine ring by incorporating a variety of HBA groups including methyl sulfone, methyl ether and fused cyclic ethers (Table 4). These modifications consistently resulted in a 1050-fold increase in TASK-3 potency versus the unsubstituted analogue 17 c. Carboxyl surrogates such as the tertiary alcohol group (21) and the oxadiazole functional group (22) gave compounds with slightly better IC50 values compared with the parent ethyl ester 12 f, while the n-propyl ketone (23) proved to be the optimal group at this position and resulted in the most potent compound realized in this series (IC50 = 35 nm). The affinity of compounds 2124 for TASK-1 (also expressed in HEK-293 cells and recorded on the QPatch HT platform) was tested, and each analogue was found to show modest selectivity for TASK-3. Selectivity values ranged from 2- to 20-fold depending on the nature of the piperidine C-4 substituent. The IC50 values determined for several potassium channels suggest that compound 23 is a relatively specific TASK-3 channel inhibitor, exhibiting good selectivity over other K2P chan-
[a] Values represent the mean SD of three independent experiments. Table 3. Structureactivity relationship data for 5,6,7,8-tetrahydropyrido[4,3-d]pyrimidine (THPP) analogues with modifications to the Northern moiety.
Modification of the Northern A sector: To address the SAR at the C-4 position of the core scaffold, we held the optimized biphenylamide D ligand constant. For these studies the synthetic scheme was modified slightly in order to prepare a penultimate intermediate that would allow for rapid assessment of the SAR (Scheme 3). As such, deprotection and acylation of N-Boc intermediate 6 provided compound 16, which was subjected to SNAr reactions with a variety of amines. Final compound preparation was achieved by heating chloropyrimidine 16 with the appropriately substituted amine. SAR from this effort revealed a preference for cyclic amines over alkyl or dialkylamines (results not
Compd 17 a 17 b 17 c 17 d 17 e
R1R2 CH2CH(OCH3)CH2 CH2CH2CH2CH2 CH2CH2CH2CH2CH2 CH2CH2OCH2CH2 CH2CH2SO2CH2CH2
IC50 [mm][a] TASK-3 15 6 27 2 4 1.1 4.6 1.1 0.57 0.06
[a] Values represent the mean SD of three independent experiments.
126
www.chemmedchem.org
ChemMedChem 2012, 7, 123 133
TASK-3 Antagonists progression from their sleep-to-wake states, and, during their sleeping period, their sleep episodes as well as their REM theta oscillations are more fragmented.[13] We have also shown that the sleep architecture of animals lacking TASK-3 channels exhibit far greater day/night variations in time spent in active wake, delta sleep, as well as substantial oscillations in the number of entries into all sleep stages.[7a] Additionally, REM sleep time showed reductions across most of the circadian period. Therefore, we reasoned that electroencephalogram (EEG) measurements in various animal models should serve as a relevant pharmacodynamic readout of both brain penetration and TASK-3 channel blockade. As such, compound 23 was advanced into our rodent sleep assays to assess its effects on sleep architecture. Compound 23 was dosed subcutaneously (s.c., 100 mg kg1) to telemetry implanted mice at the onset of their inactive phase (lights on), and the amount of time spent in each state of arousal was measured, combined into 30 min intervals, and compared to that of vehicle-treated animals (Figure 2). Concurrently, we conducted a satellite pharmacokinetics (PK) experiment in C57/BL6 mice (n = 8) to measure drug exposure in cerebral spinal fluids (csf) at 60 min post-subcutaneous dose and found that the average drug concentration was 150 nm (average plasma drug levels = 4.6 mm). Effects of TASK-3 antagonism on sleep parameters were assessed using telemetric recording of electrocorticogram (ECoG) and electromyogram (EMG) signals to measure time spent in various states of arousal, including active wake, light sleep, REM sleep, and delta sleep. We were pleased to find that compound 23 produced a significant increase in active wake with a concurrent decrease in both REM and delta sleep immediately following administration to wild-type (WT) mice. The sleep/ wake effects lasted for 34 h across the natural wake to sleep transition with some values reaching significance as much as 5 h after dosing (Figure 2 a). In addition, after administration of compound 23 to the TASK-3 deficient (KO) mice at the same dose (100 mg kg1) and PEG 400 vehicle, the effects on REM suppression were completely absent, and the magnitude of effects on wake and slow-wave sleep were substantially attenuated relative to wild-type controls (Figure 2 b). Coupled with the results from the study in WT mice, these results are consistent with TASK-3 target engagement thereby producing the desired pharmacological effect. To explore the pharmacological effect in more detail, we treated male SpragueDawley rats in a doseresponse manner with both 15 mg kg1 (n = 14) and 50 mg kg1 (n = 16) of compound 23. Radiotelemetry implanted animals were treated with vehicle (PEG 400, s.c.), 15 mg kg1 or 50 mg kg1 of compound 23 (PEG 400, s.c.) in a three-day crossover paradigm. Dose-dependent effects across all sleep stages were marked by more consistent sleep reductions and greater next-day homeostatic rebound of sleep/wake patterns. Stronger effects were observed in mid- to high-frequency bands with increasing dose, while low-frequency qEEG effects were more consistent at the low dose (Figure 3). As evidenced in the mouse study, there was an immediate post-dose increase in light sleep with concomitant decreases
Table 4. Structureactivity relationship data for 5,6,7,8-tetrahydropyrido[4,3-d]pyrimidine (THPP) analogues with modifications to the piperidine ring system.
Compd 18 19 20 a 20 b 21 22 23 24
R TASK-3 SO2Me OMe cis trans C(OH)Me2
IC50 [mm][a] TASK-1 nd nd nd nd 0.96 0.22 0.42 0.11 0.30 0.02 0.19 0.02 0.082 0.005 0.135 0.037 0.45 0.02 0.07 0.007 0.05 0.006 0.07 0.01
C(O)CH2CH2CH3 C(O)CH2CH3
0.035 0.005[b] 0.08 0.009
[a] Values represent the mean SD of three independent experiments; [b] Values represent the mean SD of twelve independent experiments; nd = not determined.
nels. For example, the potency of compound 23 against the closest relative of TASK-3 (TASK-1; IC50 = 303 nm) is approximately nine-times weaker, while it is inactive against the K2P channels TREK-1 (IC50 > 10 000 nm), as measured in our planar Qpatch HT assays. Compound 23 is weakly active against the voltage-gated Kv1.5 (IC50 ~ 5 mm; 140-times less potent), insensitive to hERG at the highest test concentration (> 15 mm), and does not inhibit KATP (< 10% inhibition at 10 mm). There was no activity at 10 mm when assayed in a panel of off-target CNS assays, such as the 5-HT serotonin receptor, and a Panlabs screen resulted in no activity being observed against more than 100 different receptors, ion channels and enzymes (10 mm exposure) further supporting compound 23 as a selective TASK-3 inhibitor. To be effective at inhibiting TASK-3 channels in vivo, compounds subjected to the efflux-mediated by P-glycoprotein transporter (P-gp) should be avoided as this transporter can often limit brain penetration. Gratifyingly, compound 23 showed a favorable profile for a CNS-penetrant drug as it displayed low directional transport ratios in both human (BA/ AB = 0.3) and mouse (BA/AB = 0.4) P-gp expressing cell lines while maintaining excellent cell permeability (Papp = 36 106 cm s1). Pharmacological activity The availability of knockout (KO) mice has aided the evaluation of the physiological function of TASK-3. A recent report from Pang et al. has shown that TASK-3 KO animals exhibit a slower
ChemMedChem 2012, 7, 123 133
www.chemmedchem.org
127
MED
C. A. Coburn et al. alpha, and sigma). The magnitude and duration of these effects were once again dose responsive with changes in lowand high-frequency powers lasting for up to 4 and 9 h, respectively, in response to the highest dose. These results, particularly the suppression of qEEG theta power and REM sleep, are consistent with previous observations in TASK-3 KO mice in which type II theta oscillations were observed to be eliminated by this mutant model. While acute inhibition of TASK-3 channels by compound 23 did not increase the number of REM bouts indicative of REM fragmentation (results not shown), the overall wake-promoting effects of the compound are in line with findings in TASK-3 KO animals showing increased time in active wake at the expense of REM and non-REM at inactive phase onset.[13] Further, these findings suggest the therapeutic potential of TASK-3 antagonists not only for sleep disorders such as excessive daytime sleepiness, but also as antidepressant agents. The REM-suppressing effect of compound 23 is reminiscent of that observed with antidepressant treatment, and animals lacking the TASK-3 channel have recently been shown to display resistance to despair behavior.[7a, 15]
Conclusions
K2P channels are expressed throughout the CNS and play an important role in controlling the neuronal resting membrane potential as well as its excitability. The development of compounds that can specifically block these channels is a field of active interest and holds potential therapeutic promise. By developing a high-throughput screen and using a rational drug design approach, we have discovered a novel series of TASK-3 antagonists and elucidated the preliminary SAR. One of the more potent and selective compounds was profiled in detail, and target engagement was shown to occur through the alteration of the sleep/wake patterns in rodents. Additional pharmacological studies with compound 23 have been conducted and will be reported in subsequent communications.
Figure 2. Effects of a 100 mg kg1 s.c. dose of 23 on mouse sleep. The duration of time spent in active wake, delta sleep and REM sleep in vehicle (*) versus drug (*) for a) wild-type (WT) animals and b) TASK-3 KO animals. Radiotelemetry implanted wild-type (n = 6; panel a) or TASK-3 KO mice (n = 6; panel b) were treated with either vehicle (PEG 400, s.c.) or 100 mg kg1 of compound 23 in a 5 day crossover paradigm. Mean times in sleep stage from all 5 days are plotted over a single 24 h period, dose time occurring at 14:00, inactive phase onset. Intervals at which significant differences exist between control and experimental values are indicated by grey vertical lines with short, medium and long tick marks representing p values of < 0.05, 0.01 or 0.001, respectively (linear mixed effects model for repeated measures).
Biology
All animal studies were performed according to the US National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals, and experimental protocols were reviewed by the Merck Animal Care and Use Committee. Sleep architecture and qEEG: Rodent electrocorticogram/electroencephalogram (ECoG/EEG) and electromyogram (EMG) measurements for sleep architecture and quantitative EEG (qEEG) analysis were performed as previously described.[14] Briefly, male C57/BL6 background or TASK-3 mutant mice and male SpragueDawley rats implanted with radiotelemetric monitors (Data Sciences International, Arden Hills, MN, USA) were alternatively treated with compound 23 and vehicle (PEG 400, s.c.) daily for either three (rat) or five (mice) consecutive days in a counterbalanced cross-over design. Rats: two baseline days (no dosing), a two-day vehicle-only run-in, a three day arm of drug or vehicle, followed by three days of conditional crossover. Mice: a five day arm of drug or vehicle,
ChemMedChem 2012, 7, 123 133
in delta and REM sleep, with sleep/wake effects 24 h after dosing and the magnitude of these effects increasing at the higher dosage. After 4 h, there was a rebound in REM sleep shortly after the onset of the subsequent active period in animals treated with 50 mg kg1 of compound 23. Consistent with these wake-promoting effects were significant quantitative EEG effects marked by increases in high-frequency gamma power and decreases in low- to mid-frequencies (delta, theta,
128
www.chemmedchem.org
TASK-3 Antagonists
Figure 3. Effects of lower doses of 23 on SpragueDawley rat sleep. Animals were dosed with compound 23 (*) at 15 mg kg1 s.c. (panels a and c) or 50 mg kg1 s.c. (panels b and d) or vehicle (PEG 400, s.c.; *). Data are group averages ( SEM) of vehicle versus compound at each 0.5 h time point for 23 h after dosing. Data were compared using a mixed-model ANOVA at each time point. Significant conditional differences are indicated as tick marks at the top of the graphs (short = p 0.05; medium = p 0.01; long = p 0.001) and gray bars through significantly different data points.
two baseline days, followed by five days of conditional crossover. Mean time and the mean number of entries into sleep stages in 30 min intervals were scored in an automated routine using Somnologica (Embla Systems, Denver, CO, USA); measures for vehicle
and compound conditions for all treatment days were averaged and plotted on a 24-hour timescale, and significance for each 30 min interval was evaluated using a linear mixed effects model for repeated measures.
ChemMedChem 2012, 7, 123 133
www.chemmedchem.org
129
MED
Chemistry
Solvents, reagents, and intermediates that are commercially available were used as received. Reagents and intermediates that were not commercially available were prepared in the manner as described below. 1H NMR spectra were obtained on a Varian VNMR System 400 (400 MHz), and chemical shifts (d) are reported in ppm downfield from tetramethylsilane (TMS) with the number of protons, multiplicities, and coupling constants (J) in Hertz (Hz) indicated parenthetically. Where LC/MS data are presented, analyses was performed using an Agilent 6110A MSD or an Applied Biosystems API-100 mass spectrometer; the parent ion is given. Flash column chromatography was performed using prepacked normal-phase silica (Biotage, Inc.) or bulk silica (Fisher Scientific). Unless otherwise indicated, column chromatography was performed using a gradient elution of hexane/EtOAc (10:0!0:10). Representative procedure for the synthesis of intermediate compounds 8 (R1=X=Y=H) 1-tert-Butyl 3-ethyl 4-oxopiperidine-1,3-dicarboxylate (4): A solution of 1-benzyl-3-carboxyethyl-4-piperidone hydrochloride (140 g, 470.1 mmol) and (Boc)2O (113 g, 517 mmol) in MeOH (600 mL) was treated with 20 % Pd(OH)2/C (16.5 g, 23.5 mmol). The flask was then charged with H2 on a Parr apparatus, and the mixture agitated at RT at 20 psi of H2 overnight. The mixture was then cooled in an ice bath, treated with Et3N (72 mL, 517 mmol), and stirred for 2 h. The reaction mixture was filtered through a pad of Celite to remove Pd/C, and the filtrate was concentrated in vacuo to give an orange oil. The crude was diluted with EtOAc (600 mL) and washed with 1 n HCl (60 mL), brine (60 mL) and water (60 mL), then dried (MgSO4), filtered and concentrated to a colorless oil. Purification by flash chromatography (5 % EtOAc/hexane) gave the desired product as a colorless oil that slowly solidified to a white solid (117 g, 91 %): 1H NMR (400 MHz, CDCl3): d = 4.12 (q, J = 7.8 Hz, 2 H), 3.923.85 (m, 2 H), 3.41 (t, J = 7.1 Hz, 2 H), 2.63 (dt, J = 14.8, 5.8 Hz, 1 H), 2.482.41 (m, 1 H), 1.781.70 (m, 1 H), 1.49 (s, 9 H), 1.25 ppm (t, J = 7.1 Hz, 3 H); MS (ESI): m/z: 214 [MtBu] . tert-Butyl 4-oxo-3,5,7,8-tetrahydropyrido[4,3-d]pyrimidine-6(4H)carboxylate (5): A solution of 4 (86 g, 317 mmol) in THF (250 mL) was treated with formamidine acetate (50 g, 476 mmol), followed by a solution of K2CO3 (66 g, 476 mmol) in water (100 mL), and the mixture was stirred at RT overnight. The solution was concentrated in vacuo and taken up in CH2Cl2 (300 mL), and then stirred overnight. The mixture was filtered and the filtrate was dried (MgSO4), filtered and concentrated in vacuo to give a cream solid. Purification by flash chromatography (50 % EtOAc/hexane then 5 % MeOH/CH2Cl2) gave the desired pyrimidinone as a white solid (51 g, 64 %); MS (ESI): m/z: 252.0 [M + H] + . tert-Butyl 4-chloro-7,8-dihydropyrido[4,3-d]pyrimidine-6(5H)-carboxylate (6): A mixture of PPh3 (81 g, 310 mmol) and N-chlorosuccinimide (NCS; 41.5 g, 310 mmol) in dioxane (850 mL) was stirred at RT for 30 min. Pyrimidinone 5 (26 g, 103 mmol) was added the suspension, and the mixture was stirred at 50 8C for 18 h. The brown suspension was subsequently treated with Et3N (25 mL), and the resulting brown solution concentrated in vacuo. The brown-black oil was dry packed on silica gel and purified by flash chromatography (1520 % EtOAc/hexane) to give an oil that slowly solidified to a light brown solid (25 g, 89 %); MS (ESI): m/z: 270.0 and 271.9 [M + H] + . tert-Butyl 4-[4-(ethoxycarbonyl)piperidin-1-yl]-7,8-dihydropyrido[4,3-d]pyrimidine-6(5H)-carboxylate (7): A solution of chloropyrimdine 6 (7.3 g, 27 mmol) in THF (50 mL) was treated with ethyl
C. A. Coburn et al.
isonipecotate (8.5 g, 54 mmol), and the mixture stirred at RT for 1 6 days as indicated by loss of starting material using LC/MS. The mixture was then filtered to remove the salts, and the filtrate concentrated in vacuo to an oil. The oil was taken up in EtOAc (100 mL) and washed with saturated aq NH4Cl (25 mL) and water (25 mL), then dried (MgSO4), filtered and concentrated to a yellow or brown gum, which was purified by chromatography to give the desired 4-aminopyridopyrimidine (8.9 g, 85 %): 1H NMR (400 MHz, CD3OD): d = 8.49 (s, 1 H), 4.52 (s, 2 H), 4.144.22 (m, 2 H), 4.14 (q, J = 7.2 Hz, 2 H), 3.703.80 (br s, 2 H), 2.89 (t, J = 6.0 Hz, 2 H), 2.702.80 (m, 1 H), 2.002.10 (m, 2 H), 1.701.85 (m, 2 H), 1.45 (s, 9 H), 1.24 ppm (t, J = 7.2 Hz, 3 H); 13C NMR (500 MHz, CD3OD): d = 174.3, 161.9, 156.4, 155.4, 116.0, 80.8, 60.4, 47.0, 40.5, 40.2, 39.9, 28.1, 27.9, 27.1, 13.1 ppm; HRMS (ESI): m/z [M + H] + calcd for C20H30N4O4 : 391.2340, found: 391.2324. Ethyl 1-(5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl) piperidine-4-carboxylate (8): A suspension of carbamate 7 (8.9 g, 23 mmol) was stirred in 4 m HCl in dioxane (100 mL) at RT for 1 18 h, as indicated by loss of the carbamate molecular ion using LC/ MS. The suspension was then diluted with Et2O (400 mL) and stirred for 1 h. The hydrochloride salt of the amine was collected by filtration, washed with Et2O and dried in vacuo. The salt was dissolved in 20 % MeOH/CH2Cl2 (100 mL) and treated with NaHCO3 (40 g), and this suspension stirred overnight. The mixture was filtered, the filtrate was concentrated in vacuo and suspended in MeCN and filtered. The filtrate was concentrated in vacuo to yield the free amine (quant): 1H NMR (400 MHz, CD3OD): d = 8.61 (s, 1 H), 4.404.45 (m, 2 H), 4.224.32 (m, 2 H), 4.074.18 (m, 2 H), 3.583.67 (m, 2 H), 3.48 (t, J = 7.6 Hz, 2 H), 3.173.26 (m, 2 H), 2.712.85 (m, 1 H), 2.042.15 (m, 2 H), 1.751.89 (m, 2 H), 1.24 ppm (t, J = 7.2 Hz, 3 H); 13C NMR (500 MHz, CD3OD): d = 176.2, 165.1, 163.4, 156.5, 116.6, 61.7, 49.6, 45.6, 43.2, 37.8, 30.9, 29.2, 14.5 ppm; HRMS (ESI): m/z [M + H] + calcd for C15H22N4O2 : 291.1816, found: 291.1802. General procedure for the synthesis of analogues 9: A solution of intermediate 8 (0.40 mmol) in 10 % AcOH/MeOH (3 mL) was treated with the appropriate aldehyde (0.52 mmol), followed by MP-(CN)BH3 (0.52 mmol), and the suspensions were agitated at RT overnight. The resin was then filtered and washed with MeOH (3 mL), and the filtrates were concentrated in vacuo to yield the desired alkylamine (0.40 mmol, quant). General procedure for the synthesis of ureas 10: A suspension of polymer-supported N,N-diisopropylethylamine (PS-DIPEA; 0.440 mmol) in CH2Cl2 (3.0 mL) was treated with a solution of intermediate 8 (0.40 mmol) in CH2Cl2 (1.0 mL), followed by the appropriate isocyanate (0.44 mmol), and the mixture was agitated at RT for 5 h. Excess isocyanate was scavenged by addition of PS-trisamine or silica amine (0.440 mmol), and the suspension agitated at RT overnight. The resin was then removed by filtration and washed with CH2Cl2/THF (1:1, 3 mL), and the combined filtrates concentrated in vacuo to yield the desired urea (0.40 mmol, quant). General procedure for the synthesis of sulfonamides 11: A suspension of PS-DIPEA (0.52 mmol) in CH2Cl2 (3.0 mL) was treated with a solution of intermediate 8 (0.400 mmol) in CH2Cl2 (1.0 mL), followed by the appropriate sulfonyl chloride, and the mixture agitated at RT overnight. Excess sulfonyl chloride was scavenged by the addition of PS-trisamine or silica amine (0.52 mmol), and the suspension agitated at RT overnight. The resin was then removed by filtration and washed with CH2Cl2/THF (1:1, 3 mL), and the combined filtrates concentrated in vacuo to yield the desired sulfonamide (0.4 mmol, quant).
ChemMedChem 2012, 7, 123 133
130
www.chemmedchem.org
TASK-3 Antagonists
General procedure for the synthesis of amides 12: A suspension of polymer-supported N,N-dicyclohexylcarbodiimide (PS-DCC; 0.900 mmol) in CH2Cl2 (6 mL) was treated with the appropriate carboxylic acid (0.900 mmol), followed by a solution of intermediate 8 (0.300 mmol) in CH2Cl2 (1 mL), and the mixture was agitated at RT for 8 days. The reaction was filtered and the resin washed with CH2Cl2/MeOH (3:1, 3 mL), and the combined filtrates were concentrated in vacuo to yield the desired amide (0.30 mmol, quant). Ethyl 1-[6-(biphenyl-4-ylcarbonyl)-8,8-dimethyl-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]piperidine-4-carboxylate (13 a): A solution of amine 8 (R1=H, X=CH3, Y=H; 285 mg, 0.49 mmol) in CH2Cl2 (10 mL) was treated with biphenyl-4-carboxylic acid (89 mg, 0.448 mmol), benzotriazol-l-yl-oxy-tris-(dimethylamino)phosphonium hexafluorophosphate (BOP reagent; 218 mg, 0.493 mmol) and DIPEA (231 mg, 1.79 mmol). The mixture was stirred at RT for 3 h then concentrated in vacuo. The crude product was purified by reverse-phase (RP)-HPLC to give the desired compound (175 mg, 78 %): 1H NMR (400 MHz, CD3OD): d = 8.48 (s, 1 H), 7.66 (d, J = 8.4 Hz, 2 H), 7.57 (d, J = 7.2 Hz, 2 H), 7.46 (d, J = 8.4 Hz, 2 H), 7.35 7.39 (m, 2 H), 7.277.30 (m, 1 H), 4.514.75 (m, 2 H), 3.684.32 (m, 6 H), 3.273.46 (m, 1 H), 2.893.15 (m, 1 H), 2.532.75 (m, 1 H), 1.78 2.03 (m, 2 H), 0.901.61 ppm (m, 11 H); 13C NMR (500 MHz, CD3OD): d = 175.9, 171.8, 163.4, 162.5, 151.5, 144.9, 141.2, 134.5, 130.0, 129.2, 129.0, 128.5, 128.3, 113.0, 61.7, 57.0, 52.0, 43.5, 41.4, 38.6, 29.6, 25.9, 14.4 ppm; HRMS (ESI): m/z [M + H] + calcd for C30H34N4O3 : 499.2704, found: 499.2698. Ethyl 1-[6-(biphenyl-4-ylcarbonyl)-7,7-dimethyl-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]piperidine-4-carboxylate (13 b): A solution of amine 8 (R1=H, X=H, Y=CH3 ; 100 mg, 0.314 mmol) in CH2Cl2 (5 mL) was treated with biphenyl-4-carboxylic acid (62 mg, 0.314 mmol), BOP reagent (153 mg, 0.345 mmol) and DIPEA (162 mg, 1.26 mmol). The mixture was stirred at RT for 4 h then concentrated in vacuo. The crude product was purified by RPHPLC to give the desired compound (175 mg, 78 %): 1H NMR (400 MHz, CD3OD): d = 8.56 (s,1 H), 7.547.75 (m, 4 H), 7.457.54 (m, 4 H), 7.367.47 (m, 1 H), 4.114.27 (m, 2 H), 3.803.96 (m, 3 H), 3.35 3.45 (m, 1 H), 3.123.25 (m, 4 H), 2.602.73 (m, 1 H), 1.982.06 (m, 4 H), 1.221.68 ppm (m, 9 H); MS (ESI): m/z: 499 [M + H] + . Ethyl 1-[6-(biphenyl-4-ylcarbonyl)-2-phenyl-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]piperidine-4-carboxylate (14 a): 1 H NMR (400 MHz, CD3OD): d = 8.35 (d, J = 3.6 Hz, 2 H), 7.76 (d, J = 8.0 Hz, 2 H), 7.72 (d, J = 7.2 Hz, 2 H), 7.447.60 (m, 7 H), 7.407.44 (m, 1 H), 4.504.80 (br s, 2 H), 3.604.20 (m, 6 H), 2.903.20 (br s, 4 H), 2.652.75 (br s, 1 H), 1.702.10 (br s, 3 H), 1.001.40 ppm (br s, 4 H); 13C NMR (500 MHz, CD3OD): d = 174.1, 169.1, 163.2, 162.6, 160.1, 141.5, 139.3, 137.5, 134.4, 130.3, 129.0, 128.4, 127.9, 127.6, 127.5, 126.8, 126.7,113.2, 59.9, 47.2, 44.1, 41.0, 40.0, 32.2, 27.9, 14.0 ppm; HRMS (ESI): m/z [M + H] + calcd for C34H34N4O3 : 547.6751, found: 547.2683. Ethyl 1-[6-(biphenyl-4-ylcarbonyl)-2-methyl-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]piperidine-4-carboxylate (14 b): 1 H NMR (400 MHz, CD3OD): d = 7.707.80 (m, 2 H), 7.657.68 (m, 2 H), 7.467.60 (m, 4 H), 7.377.44 (m, 1 H), 4.704.80 (br s, 1 H), 4.504.60 (br s, 1 H), 3.604.25 (m, 6 H), 2.853.20 (m, 4 H), 2.48 (s, 3 H), 1.802.10 (br s, 3 H), 1.101.50 ppm (m, 5 H); 13C NMR (500 MHz, CD3OD): d = 177.6, 172.6, 166.2, 164.5, 163.3, 144.6, 141.1, 135.1, 130.0, 128.7, 128.2, 128.1, 126.1, 113.3, 61.6, 47.8, 42.1, 40.9, 31.3, 29.8, 25.2, 14.5 ppm; HRMS (ESI): m/z [M + H] + calcd for C29H32N4O3 : 485.2547, found: 485.2527. Biphenyl-4-yl(4-chloro-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)yl)methanone (16): A solution of compound 6 (5.38 g, 20.0 mmol)
ChemMedChem 2012, 7, 123 133
in EtOAc (100 mL) was cooled to 0 8C and treated with 4 m HCl in dioxane (100 mL). The resulting mixture was stirred to RT for 5 h, then concentrated in vacuo, using toluene as an azeotrope. Crude compound 15 was redissolved in CH2Cl2 (100 mL) and treated with biphenyl-4-carboxylic acid (3.96 g, 20.0 mmol), BOP reagent (9.77 g, 22.0 mmol) and DIPEA (13.9 mL, 80.0 mmol). The mixture was stirred at RT for 4 h, then quenched with a saturated aq NaHCO3 (20 mL). The organic phase was washed with water (20 mL) and brine (20 mL), dried (Na2SO4), filtered, and concentrated in vacuo. Purification by flash chromatography on an ISCO flash system (MeOH/CH2Cl2, 0!100) gave the desired compound as a brown solid (3.67 g, 52 %): MS (ESI): m/z: 350 [M + H] + . General procedure for the synthesis of 17 ae: A solution of intermediate 16 (350 mg, 1.0 mmol) in anhyd DMSO (5 mL) was treated with excess amine (2.0 mmol) and K2CO3 (276 mg, 2.0 mmol). The mixture was heated to 90 8C for 4 h, then cooled and filtered. The filtrate was purified by RP-HPLC to give the desired compound. Biphenyl-4-yl[4-(3-methoxycyclobutyl)-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]methanone (17 a): Yield = 235 mg, 58 %: 1 H NMR (400 MHz, CDCl3): d = 8.44 (br s, 1 H), 7.407.69 m, 9 H), 4.81(br s, 3 H, 4.36 (br s, 4 H), 3.80 (s, 2 H), 3.35 (s, 3 H), 2.72 ppm (br s, 2 H); MS (ESI): m/z: 401.2 [M + H] + . Biphenyl-4-yl[4-(pyrrolidin-1-yl)-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]methanone (17 b): Yield = 195 mg, 51 %: 1H NMR (400 MHz, CD3OD): d = 8.48 (s, 1 H), 7.75 (d, J = 8.0 Hz, 2 H), 7.65 (d, J = 7.2 Hz, 2 H), 7.58 (d, J = 8.4 Hz, 2 H), 7.417.50 (m, 2 H), 7.367.39 (m, 1 H), 5.005.10 (m, 2 H), 3.704.20 (m, 6 H), 2.903.00 (s, 2 H), 1.802.15 ppm (m, 4 H); HRMS (ESI): m/z [M + H] + calcd for C24H24N4O: 385.2023, found: 385.2010. Biphenyl-4-yl[4-(piperidin-1-yl)-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]methanone (17 c): Yield = 225 mg, 56 %: 1H NMR (400 MHz, CD3OD): d = 8.42 (s, 1 H), 7.73 (d, J = 8.0 Hz, 2 H), 7.64 (d, J = 7.6 Hz, 2 H), 7.407.55 (m, 4 H), 7.307.38 (m, 1 H), 4.704.80 (br s, 1 H), 4.354.55 (br s, 1 H), 4.004.10 (br s, 1 H), 3.753.85 (br s, 1 H), 3.403.50 (br s, 2 H), 3.143.25 (br s, 2 H), 2.903.10 (br s, 2 H), 1.301.85 ppm (m, 6 H); HRMS (ESI): m/z [M + H] + calcd for C25H26N4O: 399.2179, found: 399.2163. Biphenyl-4-yl[4-(morpholin-4-yl)-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]methanone (17 d): Yield = 245 mg, 61 %: 1H NMR (400 MHz, CD3OD): d = 8.53 (s, 1 H), 7.76 (d, J = 8.0 Hz, 2 H), 7.68 (d, J = 7.2 Hz, 2 H), 7.497.60 (m, 4 H), 7.387.48 (m, 1 H), 4.454.80 (br s, 2 H), 4.10 (br s, 1 H), 3.783.86 (br s, 4 H), 3.483.57 (br s, 4 H), 3.243.34 (m, 1 H), 2.903.10 ppm (br s, 2 H); 13C NMR (500 MHz, CD3OD): d = 172.7, 164.8, 163.4, 156.4, 144.6, 141.3, 135.0, 130.0, 129.8, 129.1, 128.3, 128.1, 116.5, 67.7, 49.9, 48.0, 40.8, 31.4 ppm; HRMS (ESI): m/z [M + H] + calcd for C24H24N4O2 : 401.1972, found: 401.1956. Biphenyl-4-yl[4-(1,1-dioxidotetrahydro-2H-thiopyran-4-yl)-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl]methanone (17 e): Yield = 251 mg, 56 %: 1H NMR (400 MHz, CD3OD): d = 8.56 (s, 1 H), 7.68 7.34 (m, 9 H), 4.413.63 (m, 8 H), 3.633.39 (m, 3 H), 2.862.45 (m, 5 H), 2.451.63 ppm (m, 3 H); MS (ESI): m/z: 449 [M + H] + . Biphenyl-4-yl{4-[4-(methylsulfonyl)piperidin-1-yl]-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl}methanone (18): A solution of intermediate 16 (200 mg, 0.57 mmol) in dioxane (6 mL) was treated with 4-(methanesulfonyl)piperidine (188 mg, 1.44 mmol) and K2CO3 (158 mg 1.44 mmol). The mixture was heated to 70 8C and stirred at this temperature overnight. The reaction mixture was cooled and filtered, and the filtrate was then concentrated in vacuo. The
www.chemmedchem.org
131
MED
residue was purified by RP-HPLC to give the desired product (150 mg, 55 %): 1H NMR (400 MHz, CD3OD): d = 8.45 (s, 1 H), 7.71 (d, J = 7.2 Hz, 2 H), 7.64 (d, J = 6.8 Hz, 2 H), 7.51 (s, 2 H), 7.43 (t, J = 6.4 Hz, 2 H), 7.34 (t, J = 7.2 Hz, 1 H), 4.75 (s, 1 H), 4.61 (s, 1 H), 4.08 (s, 2 H), 3.82 (s, 2 H), 3.33 (d, J = 15.2 Hz, 1 H), 3.152.72 (m, 7 H), 2.20 1.80 (m, 3 H), 1.55 ppm (s, 1 H); MS (ESI): m/z: 477 [M + H] + . Biphenyl-4-yl[4-(4-methoxypiperidin-1-yl)-7,8-dihydropyrido [4,3-d]pyrimidin-6(5H)-yl]methanone (19): A solution of intermediate 16 (100 mg, 0.286 mmol) in dioxane (3 mL) was treated with 4-methoxypiperidine (66 mg, 0.572 mmol), K2CO3 (79 mg 0.572 mmol) and MgSO4 (86 mg, 0.716 mmol). The resulting mixture was heated to 70 8C and stirred overnight. The mixture was cooled to RT and filtered, and the filtrate was concentrated in vacuo. The residue was purified by RP-HPLC to give the desired compound (64 mg, 53 %): 1H NMR (400 MHz, CD3OD): d = 8.46 (s, 1 H), 7.66 (d, J = 8.4 Hz, 2 H), 7.57 (d, J = 8.4 Hz, 2 H), 7.46 (d, J = 8.0 Hz, 2 H), 7.37 (t, J = 7.2 Hz, 2 H), 7.207.30 (m, 1 H), 4.604.70 (m, 1 H), 3.504.15 (m, 1 H), 3.203.38 (m, 3 H), 2.803.02 (m, 2 H), 1.50 2.10 ppm (m, 7 H); HRMS (ESI): m/z [M + H] + calcd for C26H28N4O2 : 429.2285, found: 429.2268. Biphenyl-4-yl{4-[(8aR)-hexahydro-2H-pyrano[3,2-c]pyridin-6(5H)yl]-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl}methanone (20 a and 20 b): A mixture of octahydro-2H-pyrano[3,2-c]pyridine[16] (141 mg, 1.0 mmol), intermediate 16 (175 mg, 0.5 mmol) and K2CO3 (414 mg, 3.0 mmol) in anhyd DMSO (6 mL) was stirred at 120 8C for 5 h. After cooling to RT, the reaction mixture was filtered. The filtrate was purified by preparative HPLC to yield trans isomer 20 b (125 mg, 55 %) and cis isomer 20 a (82 mg, 36 %). trans: 1H NMR (400 MHz, CDCl3): d = 8.50 (s, 1 H), 7.61 (d, J = 8.0 Hz, 2 H), 7.54 (d, J = 7.2 Hz, 2 H), 7.337.44 (m, 5 H), 4.70 (s, 2 H), 4.49 (d, J = 7.6 Hz, 1 H), 4.28 (d, J = 8.0 Hz, 1 H), 3.96 (d, J = 10.0 Hz, 1 H), 3.81 (s, 2 H), 3.45 (t, J = 10.8 Hz, 1 H), 3.123.26 (m, 4 H), 2.84 (t, J = 12.0 Hz, 1 H), 2.012.05 (m, 1 H), 1.461.77 ppm (m, 5 H); HRMS (ESI): m/z [M + H] + calcd for C28H30N4O2 : 455.2442, found: 455.2423. cis: 1H NMR (400 MHz, CDCl3): d = 8.50 (s, 1 H), 7.60 (d, J = 8.0 Hz, 2 H), 7.53 (d, J = 7.6 Hz, 2 H), 7.327.43 (m, 5 H), 4.69 (s, 2 H), 3.99 4.13 (m, 3 H), 3.673.79 (m, 4 H), 3.47 (t, J = 10.8 Hz, 2 H), 3.14 (s, 2 H), 1.671.94 (m, 6 H), 1.161.18 ppm (m, 1 H); HRMS (ESI): m/z [M + H] + calcd for C28H30N4O2 : 455.2442, found: 455.2420. Biphenyl-4-yl{4-[4-(2-hydroxypropan-2-yl)piperidin-1-yl]-7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl}methanone (21): A solution of carboxylic ester 8 (R1=X=Y=H; 580 mg, 2 mmol) in anhyd THF (20 mL) under N2 was cooled to 78 8C and treated dropwise with CH3MgBr (1.3 mL, 3.0 m in ether, 4.0 mmol). After the addition was finished, the mixture was stirred at 78 8C for 1 h and then slowly warmed to 0 8C. When TLC showed complete starting material consumption, the mixture was quenched carefully with saturated aq NH4Cl (5 mL). The aqueous layer was extracted with CH2Cl2 (2 5 mL), and the combined organic extracts were washed with brine (5 mL), dried (Na2SO4), filtered and concentrated in vacuo to give the crude tertiary alcohol as a brown oil (550 mg). The crude was redissolved in CH2Cl2 (10 mL) and treated with 4-biphenylcarboxylic acid (2.4 mmol), N-hydroxybenzotriazole (HOBt; 331 mg, 2.4 mmol) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC; 5 mmol) at RT. The reaction mixture was stirred overnight, and then the mixture was diluted with CH2Cl2 (20 mL) and washed with water (5 mL) and brine (5 mL), dried (Na2SO4), filtered and concentrated in vacuo. The crude residue was purified by column chromatography on silica gel (CH2Cl2/MeOH, 100:1!30:1) to give the desired compound (410 mg, 45 %): 1H NMR (400 MHz, CD3OD):
C. A. Coburn et al.
d = 8.56 (s, 1 H), 7.73 (d, J = 7.2 Hz, 2 H), 7.69 (d, J = 6.8 Hz, 2 H), 7.59 (d, J = 7.6 Hz, 2 H), 7.417.49 (m, 3 H), 4.42 (br s, 2 H), 3.934.13 (m, 3 H), 2.693.19 (m, 5 H), 1.902.03 (m, 3 H), 1.491.71 (m, 2 H), 0.94 1.19 ppm (m, 8 H); MS (ESI): m/z: 457 [M + H] + . Biphenyl-4-yl{4-[4-(3-ethyl-1,2,4-oxadiazol-5-yl)piperidin-1-yl]7,8-dihydropyrido[4,3-d]pyrimidin-6(5H)-yl}methanone (22): A solution of 4-(3-ethyl-1,2,4-oxadiazol-5-yl)piperidine[16] (50.0 mg, 0.14 mmol) in anhyd dioxane (3 mL) was treated with intermediate 16 (50 mg, 0.28 mmol), K2CO3 (58.00 mg, 0.42 mmol) and MgSO4 (14 mg, 0.14 mmol). The reaction was heated at 100 8C for 6 h. The reaction mixture was cooled to RT and filtered, and the filtrate was concentrated in vacuo. Purification by preparative HPLC yielded the title compound (29 mg, 42 %): 1H NMR (400 MHz, CD3OD): d = 8.51 (s, 1 H), 7.647.66 (d, J = 7.2 Hz, 2 H), 7.467.53 (m, 4 H), 7.37 (t, J = 7.2 Hz, 2 H), 7.277.29 (m, 1 H), 5.41 (s, 1 H), 4.204.37 (m, 2 H), 3.813.99 (m, 2 H), 3.483.58 (m, 2 H), 2.903.00 (m, 2 H), 2.592.65 (m, 4 H), 1.912.20 (m, 4 H), 1.121.21 ppm (m, 3 H); MS (ESI): m/z: 495 [M + H] + . 1-{1-[6-(biphenyl-4-ylcarbonyl)-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]piperidin-4-yl}propan-1-one (23): A solution of 1(4-piperidinyl)-1-butanone (831 mg, 5.3 mmol) in EtOH (50 mL) was treated with intermediate 16 (1.7 g, 4.87 mmol) and DIPEA (1.25 g, 9.74 mmol). The mixture was stirred at 80 8C for 3 h. After cooling to RT, the reaction mixture was concentrated in vacuo, and the residue was purified by silica gel chromatography (petroleum ether/ EtOAc, 10:1) to give the desired product (1.35 g, 59 %): 1H NMR (400 MHz, CD3OD): d = 8.56 (s, 1 H), 7.727.75 (m, 2 H), 7.677.69 (m, 2 H), 7.567.63 (m, 2 H), 7.437.49 (m, 2 H), 7.367.38 (m, 1 H), 4.464.83 (m, 2 H), 3.904.07 (m, 2 H), 3.213.72 (m, 2 H), 2.953.15 (m, 2 H), 2.212.94 (m, 3 H), 1.842.19 (m, 2 H), 1.631.84 (m, 1 H), 1.131.64 (m, 3 H), 0.581.13 ppm (m, 3 H); 13C NMR (500 MHz, CD3OD): d = 213.6, 172.4, 162.5, 152.8, 149.0, 146.9, 141.2, 134.4, 130.3, 129.9, 129.2, 128.3, 128.0, 113.1, 49.6, 44.0, 43.3, 38.9, 29.1, 27.4, 17.9, 14.0 ppm; HRMS (ESI): m/z [M + H] + calcd for C29H32N4O2 : 469.2598, found: 469.2621. 1-{1-[6-(Biphenyl-4-ylcarbonyl)-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]piperidin-4-yl}ethan-1-one (24): A solution of intermediate 16 (1.1 g, 3.1 mmol) in EtOH (15 mL) was treated with 1(4-piperidinyl)-1-propanone (437 mg, 3.1 mmol) and DIPEA (1.55 g, 12 mmol). The mixture was stirred at 80 8C for 3 h then concentrated in vacuo to dryness. Purification by silica gel chromatography (petroleum ether/EtOAc, 10:1) gave the desired product (950 mg, 67 %): 1H NMR (400 MHz, CD3OD): d = 8.55 (s, 1 H), 7.657.74 (m, 2 H), 7.567.63 (m, 2 H), 7.467.54 (m, 2 H), 7.427.44 (m, 2 H), 7.36 7.38 (m, 1 H), 4.504.80 (m, 2 H), 4.234.60 (m, 2 H), 3.644.12 (m, 2 H), 3.363.60 (m, 1 H), 2.703.12 (m, 3 H), 2.232.69 (m, 2 H), 1.64 2.17 (m, 2 H), 1.161.56 (m, 1 H), 0.641.15 ppm (m, 3 H); MS (ESI): m/z: 454 [M + H] + .
Keywords: antagonists electroencephalogram (EEG) neurochemistry potassium ion channels structureactivity relationships TASK-3
[1] P. Enyedi, G. Czirjak, Physiol. Rev. 2010, 90, 559 605. [2] W. J. Coetzee, Y. Amarillo, J. Chiu, A. Chow, T. McCormack, H. Moreno, M. Nadal, A. Ozaita, D. Pountney, E. Vega-Saenz de Miera, B. Rudy, N. Y. Acad. Sci. 1999, 868, 233 285. [3] S. Bittner, T. Budde, H. Wiendl, S. Meuth, G. Sven, Brain Pathol. 2010, 20, 999 1009. [4] E. M. Talley, G. Solorzano, Q. Lei, D. Kim, D. A Bayliss, J. Neurosci. 2001, 21, 7491 7505.
132
www.chemmedchem.org
ChemMedChem 2012, 7, 123 133
TASK-3 Antagonists
[5] S. G. Meuth, T. Budde, T. Kanyshkova, T. Broicher, T. Munsch, H.-C. Pape, J. Neurosci. 2003, 23, 6460 6469. [6] C. P. Washburn, J. E. Sirois, E. M. Talley, P. G Guyenet, D. A. Bayliss, J. Neurosci. 2002, 22, 1256 1265. [7] a) A. L. Gotter, V. P. Santarelli, S. M. Doran, P. L. Tannenbaum, R. L. Kraus, T. W. Rosahl, H. Meziane, M. Montial, D. R. Reiss, K. Wessner, A. McCampbell, J. Stevens, J. Brunner, S. V. Fox, V. N. Uebele, D. A. Bayliss, C. J. Winrow, J. J. Renger, Brain Res. 2011, in press; DOI: 10.1016/ j.brainres.2011.08.021; b) D. S. Pang, C. J. Robleo, D. R. Carr, T. C. Gent, A. L. Vyssotski, A. Caley, A. Y. Zecharia, W. Wisden, S. G. Brickley, N. P. Franks, Proc. Natl. Acad. Sci. USA 2009, 106, 17546 17551; c) A. M. Linden, C. Sandu, M. I. Aller, O. Y. Vekovischeva, P. H. Rosenberg, W. Wisden, E. R. Korpi, J. Pharmacol. Exp. Ther. 2007, 323, 924 934. [8] a) A. Mathie, E. L. Veale, Curr. Opin. Invest. Drugs 2007, 8, 555 562; b) D. A. Bayliss, P. Q. Barrett, Trends Pharmacol. Sci. 2008, 29, 566 575. [9] G. Czirjak, P. Enyedi, Mol. Pharmacol. 2003, 63, 646 652. [10] J. Brendel, H. Goegelein, K. Wirth, W. Kamm, (Sanofi-Aventis Deutschland GMBh, Frankfurt-am-Main, Germany), Int. Pat. Appl. WO 2007/ 124849, 2007; Compound 1 in this patent was reported to have a tenfold weaker activity at blocking the TASK-3 channel (IC50 = 1.0 mm) versus TASK-1 (IC50 = 0.10 mm). [11] A. Streit, M. Netter; F. Kempf, M. Walecki, S. Rinne, M. Bollepalli, R. Preisig-Muller, V. Renigunta, J. Daut, T. Baukrowitz, M. Sansom, P. Stansfeld, N. Decher, J. Biol. Chem. 2011, 286, 13977 13984; F. Kempf, M. Walecki, S. Rinne, M. Bollepalli, R. Preisig-Muller, V. Renigunta, J. Daut, T. Baukrowitz, M. Sansom, P. Stansfeld, N. Decher, J. Biol. Chem. 2011, 286, 13 977 13 984. The TASK-3 antagonist assay was developed using the IonWorks Quattro system (Molecular Devices LLC, Sunnyvale, CA, USA). In this electrophysiology platform, cells are sealed on Population Patch Plate (PPC) technology. Electrical access is obtained using nystatin (SigmaAldrich, cat. no.: N6261). Currents were recorded using a 200 ms depolarization to + 50 mV followed by a 200 ms ramp from 110 to + 70 mV for both the pre- and post-compound recordings. O. Sugimoto, M. Mori, K-I Tanji, Tetrahedron Lett. 1999, 40, 7477 7480. C. J. Winrow, A. L. Gotter, C. D. Cox, S. M. Doran, P. L. Tannenbaum, M. J. Breslin, S. L. Garson, S. V. Fox, C. M. Harrell, J. Stevens, D. R. Reiss, D. Cui, P. J. Coleman, J. J. Renger, J. Neurogenet. 2011, 25, 52 61. a) R. H. Pastel, J. D. Fernstrom, Brain Res. 1987, 436, 92 102; b) I. H. Slater, G. T. Jones, R. A. Moore, Neuropharmacology 1978, 17, 383 389. C. A. Coburn, J. Wang, V. P. Santarelli, S. Hu, M. Cui, B. Hu, J. Dong, Y. Luo, R. Soll, (Merck & Co., West Point, USA), Int. Pat. Appl. WO 2011/ 106276, September 1, 2011.
[12]
[13] [14]
[15] [16]
ChemMedChem 2012, 7, 123 133
www.chemmedchem.org
133
MED
DOI: 10.1002/cmdc.201100407
From a Library of MAG Antagonists to Nanomolar CD22 Ligands

Stefanie Mesch,[a] Katrin Lemme,[a] Matthias Wittwer,[a] Hendrik Koliwer-Brandl,[b] Oliver Schwardt,[a] Srge Kelm,[b] and Beat Ernst*[a]
Siglec-2, also known as CD22, is involved in the regulation and survival of B-cells and has been successfully targeted in cell depletion therapies with antibody-based approaches. Sialic acid derivatives, already known to bind with high affinity to myelinassociated glycoprotein (MAG, Siglec-4), were screened for their binding affinity for CD22 by surface plasmon resonance. The best compound identified was further modified with various hydrophobic substituents at the 2-, 5-, and 9-positions of the sialic acid scaffold, leading to nanomolar derivatives, of which ligand 17 b shows the most promising pharmacodynamic and pharmacokinetic profiles. Isothermal titration calorimetry measurements demonstrate that the binding is enthalpy driven. Interestingly, the thermodynamic fingerprints reveal an excellent correlation between gains in enthalpy and compensation by increased entropy costs. Moreover, 17 b exhibits a residence time in the range of a few seconds, clearly prolonged relative to residence times typically observed for carbohydratelectin interactions. Finally, initial tests regarding druglike properties of 17 b demonstrate the required high plasma protein binding yet a lack of oral availability, although its distribution coefficient (log D) is in the required range.
Introduction
Patients diagnosed with B-cell lymphomas can be effectively treated with the anti-CD20 antibody rituximab.[1] However, this therapy is not a cure, and especially for patients with indolent lymphoid malignancies novel treatments with alternative mechanisms of B-cell killing are required.[2] Numerous antibodies for B-cell depletion therapy are therefore under clinical development.[3] Two of them, the immunotoxins BL22[4] and CMC544,[5] target CD22, a B-lymphocyte-specific receptor. CD22 (Siglec-2) is a member of the sialic acid binding immunoglobulin-like lectin (Siglec) family.[6] It is an inhibitory co-receptor for the B-cell receptor (BCR) and plays a crucial role in the regulation of activity,[7] homeostasis,[8] and survival of B-cells.[9] Upon BCRantigen binding, tyrosine phosphorylation is induced,[10] which triggers further phosphorylation processes and finally leads to a dampening of the BCR-induced signal.[6, 11] Apart from a few exceptions,[12] sialic acid binding sites of CD22 were shown to be effectively occupied by cis ligands, that is, ligands located on the same cell surface.[13] This interaction is important for the regulation of CD22 activity. Furthermore, it was shown that CD22 can also interact with ligands in trans,[14] that is, ligands located on other cells, to enable cellcell communication. As an alternative to antibodies, sialosides that were shown to directly influence CD22 activity[15] initiated an intense search for high-affinity ligands. The physiological ligands of Siglecs are gangliosides and sialidated glycoproteins.[16, 17] CD22 recognizes sialic acid a2,6linked to d-Gal, d-GalNAc (!1), or d-GlcNAc. A decrease in the structural complexity yielded sialic acid derivatives 24[15, 18, 19] (Figure 1), indicating that a biphenyl carboxamide moiety at the 9-position can improve affinity substantially. When this extension in the 9-position was applied to the disaccharide epitope, ligands such as 5 and 6 exhibiting sub-micromolar affinities were identified.[20] Finally, benzyl substituents at the reducing end of the disaccharide epitope led to a further improvement in affinity.[19, 21] When testing in B-cell based assays, a further aspect, the socalled density-dependent binding, must be considered.[22] In tests of monomeric sialosides (e.g., 7[23] or 8,[24] Figure 1) coupled to various supports such as a polyacrylamide backbone,[23] a glutamate cluster,[18] or a polymer obtained by ring-opening metathesis,[25] a substantial increase in binding affinity relative to monovalent sialosides was observed.[18, 26] Interestingly, with a multivalent presentation, cis interactions could be abolished without precedent treatment with sialidase to remove masking by cis ligands.[13] Additionally, OReilly et al. showed that bifunctional ligands comprising a ligand of CD22 linked to an antigen can self-assemble by antibody triggering on B-cell surfaces and are able to overcome cis interactions as well.[24] Based on the current state of knowledge, an oligovalent presentation of glycan ligands is required for a successful competition with cis ligands.[27] However, further investigation is required to clarify whether an oligomeric display is mandatory to compete with cis ligands, or if monomeric high-affinity ligands could success[a] Dr. S. Mesch, K. Lemme, Dr. M. Wittwer, Dr. O. Schwardt, Prof. Dr. B. Ernst Institute of Molecular Pharmacy, Pharmacenter, University of Basel Klingelbergstr. 50, 4056 Basel (Switzerland) E-mail: beat.ernst@unibas.ch [b] Dr. H. Koliwer-Brandl, Prof. Dr. S. Kelm Centre for Biomolecular Interactions Bremen, Glycobiochemistry P.O. Box 330440, 28334 Bremen (Germany) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100407.
134
ChemMedChem 2012, 7, 134 143
From MAG Antagonists to CD22 Ligands to be the superior substituent, leading to a ligand with nanomolar affinity for CD22 and a modest selectivity regarding MAG (~ 12-fold).
Synthesis Based on the screening hits 9 c and 9 k as well as published data on CD22 ligands,[20] compound 13 a,[28] with biphenylcarboxamide at the 9-position and a benzyl aglycone, was considered as the starting point for our optimization campaign. Our goal was not only to elucidate the kinetic and thermodynamic profile
Figure 1. Overview of sialic acid based ligands of CD22: Ligands of CD22 showing affinities from the low-micromolar (24, 7, and 8) to sub-micromolar range (5 and 6).
fully be applied as well, as it has been shown that the monovalent ligand 3 is able to elicit a rapid modulation of B-cell signaling in the presence of fully sialylated cis ligands.[15] Herein we present a series of CD22 ligands based on a screening hit identified from a small library of sialic acid derivatives, which were originally synthesized as antagonists of the myelin-associated glycoprotein (MAG, Siglec-4).[28, 29] They were tested for pharmacodynamic (affinity and selectivity) and pharmacokinetic (lipophilicity, plasma protein binding, oral availability) properties.
Table 1. Affinities of compounds 9 ak for hCD22 and mMAG as determined by SPR experiments.
Entry 1 2
Compd 9a 9b
R1
R2 Ac Ac
KD [mm] CD22 MAG 3.0 11 15 13

In the intensive search for high-affinity CD22 ligands, most efforts have focused on modifications at the 9-position of sialic acid.[20] Moreover, Kiso and co-workers have reported ligands with phenyl, benzyl, and biphenylmethyl substituents at the reducing end.[21] Regarding the 5-position, van Rossenberg et al. showed that sterically demanding acyl groups, such as benzoyl and long aliphatic chains such as octanoyl, decrease affinity,[18] whereas hydroxyacetyl is well accepted.[26, 30] Screening To further elaborate on optimal substitutions at the 2- and 5positions, a stepwise approach was conducted, assuming that the individual contributions to binding are additive. With a small library of sialosides from our MAG project, an initial screen regarding the optimal aglycone (R1 group) was performed (Entries 16, Table 1).[28, 29] The affinities were determined by surface plasmon resonance (SPR, see below for details). 2,3-Dichlorobenzyl (!9 c) led to the highest affinity and was therefore selected as aglycone for the new series of CD22 ligands. The aim of the second screenagain based on available MAG antagonistswas the identification of the optimal modification at the 5-position (R2, Entries 711, Table 1).[29] As a result, o-nosyl (ortho-nitrophenylsulfonyl, !9 k) was identified
ChemMedChem 2012, 7, 134 143
9c
Ac
0.4
4.3
9d
Ac
3.9
2.4
9e
Ac
2.7
6.1
9f
Ac
1.8
11.6
9g
0.36
0.5
9h
0.48
4.1
9i
1.40
5.8
10
9j
1.17
17
11
9k
0.11
1.4
www.chemmedchem.org
135
MED
B. Ernst et al. tion of the amino group by reductive amination, reduction of the nitro group could not be avoided. However, alkylation of the primary amine with 1-(bromomethyl)-4-phenylbenzene in the presence of potassium carbonate furnished sialoside 20, although only in a modest yield.
Surface plasmon resonance (SPR) The compound library 9 ak (Table 1) and the newly synthesized CD22 ligands 13 a,b, 17 ac, and 20 were evaluated by SPR (Table 2). Here, hCD22d13Fc[34] was immobilized on a protein A surface, which had been covalently attached to the chip by amino coupling. In the reference cell, only protein A was immobilized to compensate for unspecific binding to the matrix. Dilution series of the compounds were prepared either in pure HBS-EP buffer (9 ak, 13 a,b) or in the same buffer additionally containing 5 % DMSO (17 a c, 20). Finally, the sensorgrams were fitted according to a 1:1 binding model. Ligand 13 a showed a binding affinity in the submicromolar range (Entry 13). With the 2,3-dichlorobenzyl aglycone (!13 b), a gain in affinity by a factor of eight was observed (Entry 14). Modifications at the 9-position caused a gain in affinity for 4-(4-hydroxyphenyl)benzamide 17 b (Entry 16), equal affinity for 4-(pyridin-4-yl)benzamide 17 c (Entry 17), and a loss in affinity for amine 20 (Entry 18). Surprisingly, 17 a (Entry 15) showed no affinity, neither in the SPR experiments, nor in the isothermal titration calorimetry (ITC) assay (Table 2), although the best substituents Scheme 1. Reagents and conditions: a) 4-phenylbenzoyl chloride, PPh3, DCE (77 %); identified so far for the 2- (2,3-dichlorobenzyl) and b) ROH, NIS, TfOH, CH3CN [12 a: 70 %, 12 b: 50 %, 16: 76 %]; c) 10 % aq NaOH [13 a: 51 %, the 5-positions (o-nosyl) were present. Regarding se13 b: 25 %]; d) o-nosyl-Cl, NEt3, DMAP, CH2Cl2 (78 %); e) RCOCl, PPh3, DCE; f) LiOH, H2O, lectivity, 17 b exhibits a 50-fold higher affinity for THF [17 ac: 619 % for two steps, !18: 52 %]; g) PPh3, H2O, NEt3, 50 8C (13 %); h) 1-(broCD22 than for MAG (KD = 3.0 mm). This might be momethyl)-4-phenylbenzene, K2CO3, THF (12 %). largely due to the biphenyl moiety at the 9-position, which has previously been shown to decrease the afof the ligandCD22 interaction, but also the pharmacokinetic finity for MAG.[20, 28, 35] properties of the best representatives. For the ligands 13 b, 17 b, and 17 c a clear prolongation of Starting from sialyl donor 10,[31] test compound 13 b was obthe residence time of the proteinligand complex was obtained according to published procedures.[28, 29] The o-nosyl deserved (Table 3, Figure 2). Potential advantages of a prolonged rivatives were synthesized according to a different approach (Scheme 1). After cleavage of the N-acetate (!14),[32] the amine was allowed to react with o-nosyl chloride (!15) followed by glycosylation with 2,3-dichlorobenzyl alcohol to yield intermediate 16. Acylation by modified Staudinger conditions[33] and deprotection yielded the acyl derivatives 17 ac. The low yields in this particular reaction can be explained by the formation of side products due to acetate migration. In addition, because of the small scale, purification, especially removal of PPh3=O, was extremely difficult. Interestingly, lithium hydroxide in water/tetrahydrofuran had to be used for deprotection, because treatment with 10 % aqueous sodium hydroxide resulted in loss of the nitro substituent. To avoid acetate migration, the 9-amino derivative 19 was obtained by deprotection of 16 (!18) followed by azide reduction. Upon alkylaFigure 2. Sensorgram of 17 b at 25 8C measured in SPR; RU = response units.
136
www.chemmedchem.org
ChemMedChem 2012, 7, 134 143
From MAG Antagonists to CD22 Ligands with each other. The thermodynamic fingerprint of a ligand describes enthalpy and entropy contributions of the interaction with its target and therefore provides characteristic insight into KD [nm] the binding process. ITC measSPR ITC urements for the ligands 9 k, 13 a,b, and 17 ac were per110 131 formed at 25 8C using hCD22d1 [34] (Table 4) and confirmed 3Fc 800 751 224 the behavior typically observed for carbohydratelectin interac100 82 19 tions,[4043] namely an enthalpydriven binding. Less common are entropy-driven carbohy[a] [a] NB NB dratelectin interactions, such as heparin binding to the agrin-G3 domain[42] or the interaction of 60 80 9 di- and trisaccharides with calreticulin.[43] 110 152 43 The interaction of the parent compound 9 k is dominated by a large enthalpic contribution, 500 ND[b] which is, however, decreased by substantial entropy costs (Entry 23). The approximate sixfold drop in affinity observed for 13 a results from a large increase in entropy costs, which is only partially compensated by an improved enthalpy contribution (Entry 24). A possible explanation for the increased entropy term is a substantial conformational change caused by the spatial demands of the biphenyl substituent. Replacement of the benzyl aglycone in 13 a with the 2,3-dichlorobenzyl group (!13 b, Entry 25) gives a further increase in enthalpy, whereas the entropy term remains stable. Surprisingly, 17 a (Entry 26) does not bind to CD22, although the o-nosyl substituent does not prohibit binding of 9 k. As reported by Kiso and colleagues,[19] intramolecular attraction between benzyl at the 2-position and biphenyl at the 9-position can lead to a hydrophobic collapse. The introduction of an onosyl group at the 5-position may further support this process, abolishing the binding of 17 a. This hypothesis is also supported by the pharmacokinetic properties of 17 a because it is the only compound that was retained in an artificial membrane
Table 2. Affinity data for CD22 ligands as determined by SPR and ITC experiments.
Entry
Compd
R1
R2
R3
12
9k
13
13 a
Bn
Ac
14
13 b
Ac
15
17 a
16
17 b
17
17 c
18
20
[a] No binding. [b] Not determined.
Table 3. Kinetic properties of 9 k, 13 b, and 17 b,c. Entry 19 20 21 22 Compd 9k 13 b 17 b 17 c KD [nm] 130 100 60 110 kon [106 m1 s1] 2.1 1.1 1.1 1.3 koff [s1] 0.27 0.13 0.08 0.16 t1/2 [s][a] 2.6 5.3 8.7 4.3
[a] t1/2 = ln2/koff.
residence time are extended duration of the pharmacological effect and target selectivity.[36] Additionally, a therapeutic effect can be reached with a lower dose. As the half-life (t1/2) of carbohydrateprotein complexes is typically very short (< 1 s),[37, 38] this is clearly an improved property of these new CD22 ligands, although there is still a need for further improvement. The extended t1/2 of the ligandCD22 complex[39] might be due to the improved lipophilicity of the liTable 4. Thermodynamic parameters as determined at 25 8C by ITC. gands (see log D values in Table 5 below). DG8 [kJ mol1] Entry Compd N KD [nm] Isothermal titration calorimetry (ITC) The binding affinities measured by SPR and ITC experiments (Table 2) are in good agreement
ChemMedChem 2012, 7, 134 143
23 24 25 26 27 28 9k 13 a 13 b 17 a 17 b 17 c 0.96 0.96 0.04 0.94 0.02 NB[a] 1.07 0.01 0.95 0.02 131 751 224 82 19 NB[a] 80 9 152 43 39.3 35.1 0.8 40.5 0.6 NB[a] 40.6 0.4 39.0 0.7
DH8 [kJ mol1] 54.8 73.4 0.8 80.2 3.4 NB[a] 61.6 1.1 83.4 2.8
TDS8 [kJ mol1] + 15.5 + 38.3 1.6 + 39.7 3.6 NB[a] + 21.0 0.7 + 44.4 3.5
[a] No binding.
www.chemmedchem.org
137
MED
Table 5. Pharmacokinetic properties of CD22 ligands. Entry 29 30 31 Compd 13 a 13 b 17 a log D7.4 0.91 2.24 3.25 PAMPA log Pe NP[a] 9.8 8.7 PAMPA [% Mm] NR[b] NR[b] pH 5.0: 24 % pH 6.2: 14 % pH 7.4: NR[b] NR[b] NR[b] PPB [%] 97 98 > 99
B. Ernst et al. Pharmacokinetics As CD22 is found exclusively on B-cells, oral or intravenous applications must be considered. To elucidate the potential for oral availability, log D values were determined, as they are generally > 99 considered a rough indicator of > 99 a given compounds membrane permeation behavior,[46, 47] although restrictions apply.[48] The log D values of compounds 13 a,b and 17 ac were determined, and are in the range of 0.913.25 (Table 5), which indicates possible membrane permeability. However, determination of log Pe with the parallel artificial membrane permeability assay (PAMPA)[44] showed that 13 a and 17 b do not permeate at all, and 13 b and 17 a,c only to a very limited extent. Furthermore, retention within the membrane (PAMPA % Mm) could be excluded, except for 17 a. Here, at pH 5, 24 % of the total amount of compound was retained in the membrane. The different results in the PAMPA for other ligands despite similar log D values might be caused by other molecular descriptors. For example, the log D values of 17 a and 17 b are in the same range, but their nonpolar surface areas and hydrogen bonding capacity differ. The former are generally observed to favor partition, whereas the latter seems to hinder it.[49] Due to these findings, no oral absorption based on passive diffusion can be expected for the presented CD22 ligands, as log Pe values below 6.7 indicate a very poor permeation through membranes. Therefore, if no active transport processes are involved, oral administration seems unrealistic. Finally, plasma protein binding (PPB) of our CD22 ligands turned out to be very high. As CD22 is located on the surface of B-cells which circulate in the blood, high PPB might be beneficial, as suggested by Urien et al. for cetirizine, as an example.[50] Moreover, side effects due to distribution in various tissues of the body might be decreased as a consequence of both the high extent of PPB and the compounds inability to cross membranes. An additional positive aspect might be prolongation of plasma t1/2, resulting from the fact that only free ligands can be metabolized and excreted. Nevertheless, the amount of unbound ligand, and consequently its distribution and clearance, depends not only on the extent of binding at equilibrium but also on the rates of association and dissociation (kon and koff)[51] as well as tissue distribution and binding. The latter seems negligible, as PAMPA results do not indicate membrane permeation. The kinetic behavior of the ligands with respect to plasma proteins could be assessed by SPR and could provide valuable additional information.[52]
32 33
17 b 17 c
3.12 2.08
NP[a] 8.9
[a] No permeation. [b] No retention.
(PAMPA,[44] see Table 5, Entry 31). Interestingly, compound 20, in which the amide linker of 17 a is replaced by an amine linker, again exhibited nanomolar binding affinity in SPR (Table 2, Entry 18). Introduction of 4-(4-hydroxyphenyl)benzamide (!17 b, Entry 27) resulted in a ligand with an affinity similar to that of compound 13 b, although the DH8 and TDS8 terms substantially change; they both differ by 19 kJ mol1, but with opposite signs. Introduction of a para-hydroxy substituent caused a favorable enthalpy change similar to an observation reported by Kiso et al.[20] Interestingly, when the terminal phenol (!17 b) was replaced by pyridine (!17 c, Entry 28), a substantially improved DH8 value (22 kJ mol1) was observed, which, however, is overcompensated by a loss of entropy leading to an overall twofold decrease in affinity. Clearly, the o-nosyl substituent elicits a reorientation of the ligand in the binding site, or leads to a remarkable change in desolvation energies. As a consequence, a dramatic change in the thermodynamic fingerprint results. An enthalpyentropy plot (Figure 3) reveals a linear re-
Figure 3. Enthalpyentropy compensation plot: correlation of the change in enthalpy (DH8) and the change in entropy (TDS8) of CD22 ligands interacting with hCD22d13Fc.[34]
lationship with a high correlation coefficient (R2 = 0.97), indicating enthalpyentropy compensation, a property often reported for carbohydratelectin interactions.[40, 45] The slope of 1 denotes that the enthalpic gain is completely compensated by an entropic penalty.
Conclusions
A series of high-affinity CD22 ligands was synthesized and investigated with regard to their biological and pharmacokinetic properties. The best ligand, sialoside 17 b, contains a dichlorobenzyl substituent at the anomeric position, an ortho-nitrobenChemMedChem 2012, 7, 134 143
138
www.chemmedchem.org
From MAG Antagonists to CD22 Ligands zylsulfonamide at the 5-position, and a 4-hydroxy-4-biphenylcarboxamide at the 9-position. In addition to nanomolar affinity, ligand 17 b exhibits a prolonged residence time of the proteinligand complex, which is atypical for carbohydratelectin interactions.[38] With respect to its pharmacokinetic properties, 17 b is not expected to be orally available based on PAMPA data, and therefore intravenous administration would need to be considered for an in vivo validation. However, the low permeation may be improved by a prodrug approach[47] or by replacing the carboxylate with a bioisostere.[53] Finally, 17 b shows extended plasma protein binding, which might be beneficial concerning various aspects such as pharmacological effect, plasma half-life, or side effects.
7.43 (m, 1 H, CHar), 7.47 (dd, J = 4.8, 10.3 Hz, 2 H, CHar), 7.547.63 (m, 2 H, CHar), 7.647.72 (m, 2 H, CHar), 7.90 ppm (d, J = 8.5 Hz, 2 H, CHar); 13C NMR (125 MHz, CDCl3): d = 11.4 (SMe), 21.0, 21.3 (3C, 3 OAc), 23.3 (NHAc), 37.3 (C3), 38.1 (C9), 49.9 (C5), 53.1 (OMe), 68.4 (C7), 69.4 (C4), 70.7 (C8), 71.0 (C6), 84.9 (C2), 127.3, 127.4, 127.7, 128.2, 129.1, 133.0, 140.1, 144.5 (12C, Car), 167.3, 168.2, 170.4, 170.5, 171.2, 172.2 ppm (6 CO); MS (ESI) m/z calcd for C32H38N2O11S [M + Na] + : 681.20, found: 681.28. Methyl [2,3-dichlorobenzyl 5-acetamido-4,7,8-tri-O-acetyl-3,5,9trideoxy-9-(4-phenylbenzamido)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (12 b): Compound 11 (50 mg, 7.5 mmol, 1.0 equiv) was dissolved in dry CH3CN (2.0 mL). 2,3-Dichlorobenzyl alcohol (40 mg, 23 mmol, 3.0 equiv) and powdered MS (3 ) were added. The mixture was stirred at room temperature for 1.5 h. The suspension was then cooled to 40 8C and subsequently treated with N-iodosuccinimide (NIS; 27 mg, 12 mmol, 1.6 equiv) and triflic acid (5.3 mL, 6 mmol in 0.2 mL CH3CN, 0.8 equiv). After 30 min, the reaction mixture was warmed to 30 8C and stirring was continued for 16 h. The mixture was allowed to warm to room temperature, stirred for another 2 h and filtered through a pad of Celite. The Celite was washed with CH2Cl2 (5 mL) and the filtrate was subsequently washed with 20 % aq Na2S2O3 (1 mL) and satd aq NaHCO3 (3 2 mL). The organic phase was dried over Na2SO4, filtered and concentrated under reduced pressure. The crude product was purified by chromatography on silica gel (0.5 % gradient of iPrOH in petroleum ether (PE)/CH2Cl2 8:4) to yield 12 b (30 mg, 50 %) as an oil. 1H NMR (500 MHz, CDCl3): d = 1.89 (s, 6 H, NHAc, OAc), 2.01 2.06 (m, 4 H, H-3a, OAc), 2.15 (s, 3 H, OAc), 2.57 (dd, J = 4.1, 13.4 Hz, 1 H, H-3b), 3.49 (dd, J = 5.6, 12.2 Hz, 1 H, H-9a), 3.79 (s, 3 H, OMe), 3.83 (d, J = 12.4 Hz, 1 H, H-9b), 4.09 (t, J = 10.5 Hz, 1 H, H-5), 4.15 (m, 1 H, H-6), 4.58 (A of AB, J = 13.7 Hz, 1 H, CH2Ar), 4.834.98 (m, 2 H, H-4, CH2Ar), 5.335.40 (m, 2 H, H-7, H-8), 7.21 (t, J = 7.5 Hz, 1 H, CHar), 7.237.31 (m, 4 H, CHar), 7.347.50 (m, 6 H, CHar), 7.56 7.69 ppm (m, 1 H, CHar); 13C NMR (125 MHz, CDCl3): d = 20.8, 20.9, 21.1 (3 OAc), 23.3 (NHAc), 37.8 (C3), 43.2 (C9), 49.5 (C5), 53.0 (OMe), 64.3 (CH2Ar), 68.3 (C7), 68.9 (C4), 70.7 (C8), 73.0 (C6), 98.8 (C2), 126.2, 127.2, 127.3, 127.5, 129.4, 129.6, 137.3 (18C, Car), 168.1, 170.2, 170.4, 170.7, 171.0, 171.2 ppm (6 CO); MS (ESI) m/z calcd for C38H40Cl2N2O12 [M + Na] + : 809.18, found: 809.28. Sodium [2,3-dichlorobenzyl 5-acetamido-3,5,9-trideoxy-9-(4-phenylbenzamido)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (13 b): Compound 12 b (30 mg, 38 mmol) was treated with 10 % aq NaOH (0.4 mL) in CH3OH (2 mL). The crude product was purified by LCMS to yield 13 b as a white solid (6.3 mg, 25 %). [a]D20 21.8 (c = 0.33, CH3OH); 1H NMR (500 MHz, CD3OD): d = 1.73 (t, J = 11.0 Hz, 1 H, H-3a), 2.00 (s, 3 H, NHAc), 2.93 (d, J = 11.4 Hz, 1 H, H3b), 3.46 (d, J = 9.0 Hz, 1 H, H-7), 3.53 (m, 1 H, H-9a), 3.68 (d, J = 9.0 Hz, 1 H, H-6), 3.713.85 (m, 3 H, H-4, H-5, H-9b), 4.04 (t, J = 7.7 Hz, 1 H, H-8), 4.73 (A, B of AB, J = 13.4 Hz, 2 H, CH2Ar), 7.26 (t, J = 7.9 Hz, 1 H, CHar), 7.39 (t, J = 6.6 Hz, 2 H, CHar), 7.46 (t, J = 7.4 Hz, 2 H, CHar), 7.57 (d, J = 7.7 Hz, 1 H, CHar), 7.67 (d, J = 7.4 Hz, 2 H, CHar), 7.72 (d, J = 7.8 Hz, 2 H, CHar), 7.91 (d, J = 7.9 Hz, 2 H, CHar), 8.22 (d, J = 6.5 Hz, 1 H, 5-NH), 8.30 ppm (s, 1 H, 9-NH); 13C NMR (125 MHz, CD3OD): d = 22.6 (NHAc), 42.5 (C3), 44.5 (C9), 54.1 (C5), 64.6 (CH2Ar), 69.6 (C4), 71.2 (C8), 72.5 (C7), 74.5 (C6), 101.4 (C2), 127.9, 128.0, 128.2, 128.5, 128.6, 129.0, 129.1, 130.0, 130.1, 140.3, 145.6 (18C, Car), 175.5 ppm (3C, 3 CO); HRMS (ESI) m/z calcd for C31H32Cl2N2O9 [M + Na] + : 669.1384, found: 669.1384. Methyl [methyl 4,7,8-tri-O-acetyl-9-azido-3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-2-thio-d-glycero-a-d-galacto-2-nonulopyranosid]onate (15): Nosyl chloride (105 mg, 0.47 mmol), NEt3 (34.0 mL, 48.0 mg, 0.47 mmol), and DMAP (10.0 mg, 0.08 mmol)
Chemistry
NMR spectra were recorded on a Bruker Avance DMX-500 spectrometer (500 MHz). Assignment of 1H and 13C NMR spectra was performed by using 2D methods (COSY, HSQC, TOCSY). Chemical shifts are expressed in ppm using residual CHCl3, CHD2OD and HDO as references. Optical rotations were measured with PerkinElmer Polarimeters 241 and 341. MS analyses were carried out with a Waters Micromass ZQ Detector system. The spectra were recorded in positive or negative ESI mode. HPLCHRMS analyses were carried out with an Agilent 1100 instrument equipped with a photodiode array detector and a Micromass QTOF I equipped with a 4 GHz digital time converter. All target compounds exhibited a purity of ! 95 %. Reactions were monitored by TLC using glass plates coated with silica gel 60 F254 (Merck) and visualized by using UV light and/or by charring with a molybdate solution (0.02 m solution of ammonium cerium sulfate dihydrate and ammonium molybdate tetrahydrate in 10 % aq H2SO4). Column chromatography was performed on silica gel (Uetikon, 4060 mesh). CH3OH was dried by holding at reflux with NaOMe and distilled immediately before use. THF was distilled over Na immediately before use. CH2Cl2, dichloroethane (DCE), CH3CN, toluene, and benzene were dried by filtration over Al2O3 (Fluka, type 5016 A basic). Molecular sieves (3 ) were activated under vacuum at 500 8C for 2 h immediately before use. Synthetic procedures for compounds 10,[31] 12 a, 13 a,[28] 14, and 15[29] were reported earlier. Methyl [methyl 5-acetamido-4,7,8-tri-O-acetyl-3,5,9-trideoxy-9(4-phenylbenzamido)-2-thio-d-glycero-a-d-galacto-2-nonulopyranosid]onate (11): Compound 10 (250 mg, 0.50 mmol, 1.0 equiv) was dissolved in dry DCE (10 mL). 4-Phenylbenzoyl chloride (432 mg, 2.00 mmol, 4.0 equiv) and PPh3 (290 mg, 1.10 mmol, 2.2 equiv) were added, and the reaction mixture was stirred at room temperature for 24 h. After dilution with CH2Cl2 (10 mL), the organic layer was washed with satd aq NaHCO3 (3 5 mL) and H2O (5 mL), dried over Na2SO4 and filtered. Afterward, the solvents were removed under reduced pressure and the crude product was purified by chromatography on silica gel (0.5 % gradient of CH3OH in CH2Cl2) to yield 11 (251 mg, 77 %) as an oil. 1H NMR (500 MHz, CDCl3): d = 1.90 (s, 3 H, OAc), 1.97 (s, 3 H, SMe), 2.03, 2.07 (2 s, 6 H, 2 OAc), 2.20 (dd, J = 11.8, 13.8 Hz, 1 H, H-3a), 2.24 (s, 3 H, NHAc), 2.56 (dd, J = 4.9, 13.8 Hz, 1 H, H-3b), 3.05 (ddd, J = 0.5, 3.5, 15.3 Hz, 1 H, H-9a), 3.82 (s, 3 H, OMe), 4.16 (q, J = 10.3 Hz, 1 H, H-5), 4.28 (dd, J = 1.9, 10.6 Hz, 1 H, H-6), 4.50 (ddd, J = 3.9, 8.6, 15.0 Hz, 1 H, H-9b), 5.16 (dt, J = 3.5, 7.3 Hz, 1 H, H-8), 5.205.33 (m, 2 H, H-4, H-7), 5.43 (d, J = 10.1 Hz, 1 H, 5-NH), 7.22 (dd, J = 4.3, 8.5 Hz, 1 H, 9-NH), 7.36
ChemMedChem 2012, 7, 134 143
www.chemmedchem.org
139
MED
were added successively to a solution of 14 (73.0 mg, 0.16 mmol) in dry CH2Cl2 (3 mL) at 0 8C. The reaction mixture was stirred at room temperature overnight and then washed with satd aq NaHCO3 (2 5 mL) and H2O (5 mL). The organic phase was dried over Na2SO4, filtered and concentrated under reduced pressure. The residue was purified by chromatography on silica gel (PE/ EtOAc 1:1!1:2) to yield 15 (81 mg, 78 %) as an oil. 1H NMR (500 MHz, CDCl3): d = 1.85 (m, 1 H, H-3a), 2.02 (s, 3 H, SMe), 2.10, 2.13, 2.21 (3 s, 9 H, 3 OAc), 2.80 (m, 1 H, H-3b), 3.32 (dd, J = 6.2, 13.4 Hz, 1 H, H-9a), 3.57 (dd, J = 3.2, 13.4 Hz, 1 H, H-9b), 3.80 (m, 1 H, H-5), 3.82 (s, 3 H, OMe), 3.91 (d, J = 10.5 Hz, 1 H, H-6), 4.97 (td, J = 4.7, 11.4 Hz, 1 H, H-4), 5.30 (m, 2 H, H-7, H-8), 5.75 (d, J = 9.4 Hz, 1 H, NH), 7.70 (m, 2 H, CHar), 7.90 (d, J = 7.9, 1 H, CHar), 8.10 ppm (d, J = 6.5, 1 H, CHar); 13C NMR (125 MHz, CDCl3): d = 12.1 (SMe), 20.5, 21.1, 21.1 (3 OAc), 38.1 (C3), 50.8 (C9), 53.2, 53.5 (C5, OMe), 68.8 (C7), 69.7 (C4), 70.4 (C8), 74.7 (C6), 82.8 (C2), 125.5, 130.4, 133.5, 135.5, 147.5 (6C, Car), 167.8, 170.4 ppm (4C, 4 CO); MS (ESI) m/z calcd for C23H29N5O13S2 [MH] : 646.12, found: 646.56. Methyl [2,3-dichlorobenzyl 4,7,8-tri-O-acetyl-9-azido-3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-d-glycero-a-d-galacto-2nonulopyranosid]onate (16): Compound 15 (370 mg, 0.57 mmol, 1.0 equiv) was dissolved in dry CH3CN (10 mL). Then, 2,3-dichlorobenzyl alcohol (302 mg, 1.71 mmol, 3.0 equiv) and powdered MS (3 ) were added. The mixture was stirred at room temperature for 1.5 h. Afterward, the suspension was cooled to 40 8C and subsequently treated with NIS (205 mg, 0.92 mmol, 1.6 equiv) and triflic acid (40 mL, 0.4 mmol, 0.8 equiv). After 30 min, the reaction mixture was warmed to 30 8C, and stirring was continued for 16 h. The mixture was then warmed to room temperature, stirred for another 2 h and filtered through a pad of Celite. The Celite was washed with CH2Cl2 (15 mL), and the filtrate was subsequently washed with 20 % aq Na2S2O3 (5 mL), satd aq NaHCO3 (3 5 mL), and H2O (5 mL). The organic phase was dried over Na2SO4, filtered and concentrated under reduced pressure. The crude product was purified by chromatography on silica gel (0.5 % gradient of iPrOH in PE/CH2Cl2 2:1) to yield 16 as a yellow solid (337 mg, 76 %). [a]D20 43.7 (c = 0.46, CH3OH); 1H NMR (500 MHz, CDCl3): d = 1.96 (m, 1 H, H-3a), 2.02, 2.14, 2.22 (3 s, 9 H, 3 OAc), 2.75 (m, 1 H, H-3b), 3.31 (m, 1 H, H9a), 3.50 (m, 1 H, H-9b), 3.78 (s, 3 H, OMe), 3.88 (m, 1 H, H-5), 4.15 (d, J = 10.6 Hz, 1 H, H-6), 4.58, 4.66 (A, B of AB, J = 12.7, 2 H, CH2Ar), 4.96 (m, 1 H, H-4), 5.245.32 (m, 2 H, H-7, H-8), 5.60 (d, J = 9.3 Hz, 1 H, 5-NH), 7.25 (m, 1 H, CHar), 7.357.48 (m, 2 H, CHar), 7.74 (dd, J = 4.5, 10.7 Hz, 1 H, CHar), 7.777.86 (m, 2 H, CHar), 8.19 ppm (m, 1 H, CHar); 13C NMR (125 MHz, CDCl3): d = 20.5, 20.7, 21.0 (3 OAc), 38.2 (C3), 50.5 (C9), 51.0 (OMe), 52.9 (C5), 64.4 (CH2Ar), 68.6 (C4), 69.7 (C7), 71.1 (C8), 76.7 (C6), 98.1 (C2), 123.7, 125.4, 126.4, 127.5, 127.7, 129.4, 132.4, 132.7, 134.1, 135.4, 149.5 (12C, Car), 168.6, 170.7, 170.9 ppm (4C, 4 CO); MS (ESI) m/z calcd for C29H31Cl2N5O14S [M + Na] + : 798.08, found: 797.95. General procedure A for the preparation of acid chlorides: The acid (0.15 mmol) was suspended in dry CH2Cl2 (12 mL) and cooled to 0 8C. Chloroenamine (0.17 mmol) was added dropwise, and the reaction was allowed to warm to room temperature. After 2 h, the initial precipitate was dissolved, and the acid chlorides were directly used without work-up and purification. General procedure B for the synthesis of compounds 17 ac: Compound 16 (75 mmol) was added to the corresponding acid chloride (0.15 mmol), dissolved in dry CH2Cl2 (1 mL). PPh3 (0.18 mmol) in dry CH2Cl2 (1.5 mL) was added after 5 min, and the solution was stirred at room temperature for 24 h. Then, the reaction mixture was diluted with CH2Cl2 (5 mL) and washed with satd aq NaHCO3 (3 3 mL) and H2O (3 mL). The organic phase was dried
B. Ernst et al.
over Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by chromatography on silica gel. Afterward, the acetylated intermediate was dissolved in THF/H2O (2.5 mL, 4:1) and treated with LiOH (0.28 mmol) at room temperature for 4 h. After completion of the reaction, the crude product was purified by LCMS to yield the final compound. Sodium [2,3-dichlorobenzyl 3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-9-(4-phenylbenzamido)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (17 a): Prepared from 16 (50 mg, 0.06 mmol) and 4-phenyl benzoic acid according to general procedures A and B. Yield: 10 mg (19 %) as a white solid. [a]D20 1.8 (c = 0.04, CH3OH); 1 H NMR (500 MHz, CD3OD): d = 1.64 (t, J = 12.1 Hz, 1 H, H-3a), 2.40 (dd, J = 3.7, 13.0 Hz, 1 H, H-3b), 3.51 (m, 1 H, H-9a), 3.59 (t, J = 10.0 Hz, 1 H, H-5), 3.693.80 (m, 2 H, H-7, H-9b), 3.98 (m, 1 H, H-8), 4.05 (m, 1 H, H-4), 4.12 (d, J = 10.5 Hz, 1 H, H-6), 4.53, 4.85 (A, B of AB, J = 12.3 Hz, 2 H, CH2Ar), 7.29 (t, J = 8.3 Hz, 1 H, CHar), 7.38 (t, J = 7.2 Hz, 1 H, CHar), 7.427.49 (m, 3 H, CHar), 7.64 (d, J = 7.7 Hz, 1 H, CHar), 7.68 (d, J = 7.1 Hz, 3 H, CHar), 7.73 (d, J = 7.6 Hz, 4 H, CHar), 7.91 (d, J = 7.6 Hz, 2 H, CHar), 8.18 (d, J = 7.6 Hz, 1 H, CHar), 8.35 ppm (s, 1 H, NH); 13C NMR (125 MHz, CD3OD): d = 42.3 (C3), 45.7 (C9), 58.1 (C5), 63.9 (CH2Ar), 68.4 (C4), 71.0 (C7), 71.2 (C8), 72.6 (C6), 99.3 (C2), 125.6, 125.9, 128.1, 128.2, 128.7, 129.0, 129.1, 129.2, 130.0, 130.3, 132.4, 133.5, 134.3, 136.5, 139.6, 141.3, 145.7, 149.0 (24C, Car), 171.0 ppm (2C, 2 CO); HRMS (ESI) m/z calcd for C35H32Cl2N3NaO12S [M + Na] + : 834.0880, found: 834.0893. Sodium [2,3-dichlorobenzyl 3,5,9-trideoxy-9-[4-(4-hydroxyphenyl)benzamido]-5-(2-nitrophenylsulfonamido)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (17 b): Prepared from 16 (30 mg, 40 mmol) and 4-(4-hydroxyphenyl)benzoic acid according to general procedures A and B. Yield: 2.0 mg (6 %) as a pale-yellow solid. [a]D20 88.0 (c = 0.01, CH3OH); 1H NMR (500 MHz, CD3OD): d = 1.65 (t, J = 12.0 Hz, 1 H, H-3a), 2.78 (dd, J = 3.6, 12.0 Hz, 1 H, H-3b), 3.41 3.58 (m, 2 H, H-5, H-9a), 3.67 (m, 1 H, H-4), 3.703.79 (m, 2 H, H-7, H9b), 3.88 (d, J = 10.3 Hz, 1 H, H-6), 3.97 (m, 1 H, H-8), 4.64, 4.87 (A, B of AB, J = 12.6 Hz, 2 H, CH2Ar), 6.91 (d, J = 7.8 Hz, 2 H, CHar), 7.26 (t, J = 7.7 Hz, 1 H, CHar), 7.39 (d, J = 8.0 Hz, 1 H, CHar), 7.477.60 (m, 3 H, CHar), 7.67 (d, J = 7.8 Hz, 2 H, CHar), 7.707.80 (m, 4 H, CHar), 7.85 (d, J = 7.8 Hz, 2 H, CHar), 8.16 (d, J = 7.3 Hz, 1 H, NH), 8.23 ppm (m, 1 H, NH); 13C NMR (125 MHz, CD3OD): d = 42.6 (C3), 43.8 (C9), 49.5 (C5), 64.7 (CH2Ar), 70.2, 71.2, 71.9 (C4, C7, C8), 74.0 (C6), 116.8, 125.8, 127.3, 128.6, 128.7, 128.9, 129.2, 130.1, 132.1, 132.5, 133.3, 133.7, 134.6, 135.8 (24C, Car), 170.5 ppm (2C, 2 CO); HRMS (ESI) m/z calcd for C35H33Cl2N3O13S [M + Na] + : 828.1009, found: 828.1009. Sodium [2,3-dichlorobenzyl 3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-9-[4-(pyridin-4-yl)benzamido]-d-glycero-a-d-galacto-2nonulopyranosid]onate (17 c): Prepared from 16 (58 mg, 70 mmol) and 4-(pyridin-4-yl)benzoic acid according to general procedures A and B. Yield: 8.0 mg (13 %) as a pale-yellow solid. [a]D20 45.0 (c = 0.32, CH3OH); 1H NMR (500 MHz, CD3OD): d = 1.71 (t, J = 12.1 Hz, 1 H, H-3a), 2.72 (dd, J = 4.3, 12.3 Hz, 1 H, H-3b), 3.51 (t, J = 10.0 Hz, 1 H, H-5), 3.57 (dd, J = 6.3, 13.7 Hz, 1 H, H-9a), 3.69 (td, J = 4.3, 11.3 Hz, 1 H, H-4), 3.763.86 (m, 2 H, H-7, H-9b), 3.91 (d, J = 10.3 Hz, 1 H, H-6), 4.02 (t, J = 7.4 Hz, 1 H, H-8), 4.69, 4.92 (A, B of AB, J = 13.2 Hz, 2 H, CH2Ar), 7.26 (t, J = 7.8 Hz, 1 H, CHar), 7.40 (d, J = 8.0 Hz, 1 H, CHar), 7.49 (d, J = 7.7 Hz, 1 H, CHar), 7.76 (t, J = 6.2 Hz, 2 H, CHar), 7.81 (m, 1 H, CHar), 7.867.93 (m, 5 H, CHar), 7.99 (d, J = 7.8 Hz, 2 H, CHar), 8.17 (d, J = 7.0 Hz, 1 H, NH), 8.67 ppm (d, J = 5.1 Hz, 2 H, CHar); 13 C NMR (125 MHz, CD3OD): d = 42.3 (C3), 44.4 (C9), 57.9 (C5), 64.7 (CH2Ar), 69.8 (C4), 71.2 (C7), 71.7 (C8), 74.5 (C6), 100.9 (C2), 123.9, 125.9, 128.5, 128.6, 128.7, 129.5, 130.4, 132.3, 133.5, 133.7, 134.6, 136.2, 136.9, 139.6, 141.2, 148.9, 151.4 (23C, Car), 170.0, 172.6 ppm
ChemMedChem 2012, 7, 134 143
140
www.chemmedchem.org
From MAG Antagonists to CD22 Ligands

(2 CO); HRMS (ESI) m/z calcd for C34H32Cl2N4O12S [M + H] + : 791.1193, found: 791.1194. Sodium [2,3-dichlorobenzyl 9-azido-3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (18): A solution of 16 (310 mg, 0.4 mmol) in THF/H2O (6 mL:2.5 mL) was allowed to react with LiOH (78 mg, 3.2 mmol) at room temperature for 4 h. The crude product was purified by LCMS to yield 18 (107 mg, 52 %) as an oil. 1H NMR (500 MHz, CD3OD): d = 1.68 (t, J = 12.2 Hz, 1 H, H-3a), 2.71 (dd, J = 4.8, 12.5 Hz, 1 H, H-3b), 3.31 (m, 1 H, H-9a), 3.45 (m, 1 H, H-5), 3.53 (dd, J = 2.4, 12.8 Hz, 1 H, H-9b), 3.67 (ddd, J = 4.8, 9.9, 11.8 Hz, 1 H, H-4), 3.81 (dd, J = 1.5, 8.9 Hz, 1 H, H7), 3.86 (dd, J = 1.5, 10.5 Hz, 1 H, H-6), 3.95 (ddd, J = 2.4, 6.6, 8.9 Hz, 1 H, H-8), 4.70, 4.92 (A, B of AB, J = 13.3 Hz, 2 H, CH2Ar), 7.27 (t, J = 7.9 Hz, 1 H, CHar), 7.44 (dd, J = 1.4, 8.0 Hz, 1 H, CHar), 7.50 (d, J = 7.7 Hz, 1 H, CHar), 7.707.81 (m, 2 H, CHar), 7.88 (m, 1 H, CHar), 8.16 ppm (dd, J = 3.3, 6.0 Hz, 1 H, CHar); 13C NMR (125 MHz, CD3OD): d = 42.4 (C3), 55.0 (C9), 57.9 (C5), 64.6 (CH2Ar), 69.6 (C4), 70.5 (C7), 72.4 (C8), 74.5 (C6), 125.8, 128.5, 128.7, 130.3, 132.3, 133.5, 134.4 (12C, Car), 172.5 ppm (CO); MS (ESI) m/z calcd for C22H23Cl2N5O11S [MH] : 634.04, found: 634.16. Sodium [2,3-dichlorobenzyl 9-amino-3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (19): PPh3 (33 mg, 0.13 mmol, 1.6 equiv) and NEt3 (1.2 mL, 12 mmol, 1.5 equiv) were successively added to a solution of 18 (50 mg, 8.0 mmol, 1.0 equiv) in dry THF (5 mL) at 0 8C. After 1 h, PPh3 (66 mg, 0.26 mmol, 3.2 equiv) and H2O (1.5 mL) were added, and stirring was continued at 50 8C for 4 h. After cooling to room temperature, the solvent was removed under reduced pressure, and the residue was purified by LCMS to yield 19 (6.3 mg, 13 %) as an oil. [a]D20 43.7 (c = 0.46, CH3OH); 1H NMR (500 MHz, CD3OD): d = 1.60 (t, J = 12.0 Hz, 1 H, H-3a), 2.79 (dd, J = 4.9, 12.3 Hz, 1 H, H-3b), 2.93 (dd, J = 9.3, 12.7 Hz, 1 H, H-9a), 3.34 (m, 1 H, H-9b), 3.38 (t, J = 10.1 Hz, 1 H, H-5), 3.66 (m, 1 H, H-4), 3.803.89 (m, 2 H, H6, H-7), 4.05 (m, 1 H, H-8), 4.69, 4.91 (A, B of AB, J = 13.6 Hz, 2 H, CH2Ar), 7.27 (t, J = 7.9 Hz, 1 H, CHar), 7.42 (d, J = 8.0 Hz, 1 H, CHar), 7.54 (d, J = 7.8 Hz, 1 H, CHar), 7.717.81 (m, 2 H, CHar), 7.89 (dd, J = 3.3, 6.0 Hz, 1 H, CHar), 8.17 ppm (dt, J = 3.7, 7.5 Hz, 1 H, CHar); 13 C NMR (125 MHz, CD3OD): d = 43.0 (C3), 43.9 (C9), 57.9 (C5), 64.5 (CH2Ar), 69.5 (C8), 70.0 (C4), 72.0 (C7), 74.1 (C6), 125.9, 128.3, 128.6, 130.0, 132.3, 133.4, 133.6, 134.5, 136.0, 140.2, 149.1 (12C, Car), 173.9 ppm (CO); MS (ESI) m/z calcd for C22H25Cl2N3O11S [M + H] + : 610.07, found: 610.12. Sodium [2,3-dichlorobenzyl 3,5,9-trideoxy-5-(2-nitrophenylsulfonamido)-9-(4-phenylbenzylamine)-d-glycero-a-d-galacto-2-nonulopyranosid]onate (20): Compound 19 (2.0 mg, 3.3 mmol, 1.0 equiv) was dissolved in THF (0.6 mL). 1-(Bromomethyl)-4-phenylbenzene (1.1 mg, 6.6 mmol, 2.0 equiv) and K2CO3 (1.1 mg, 8.0 mmol, 2.4 equiv) were added in portions. The reaction mixture was stirred at room temperature for 4 days. The crude product was purified by LCMS to yield 20 as an oil (0.3 mg, 12 %). [a]D20 102.3 (c = 0.01, CH3OH); 1H NMR (500 MHz, CD3OD): d = 1.60 (t, J = 11.8 Hz, 1 H, H-3a), 2.77 (m, 1 H, H-3b), 2.98 (t, J = 10.9 Hz, 1 H, H-9a), 3.393.41 (m, 2 H, H-5, H-9b), 3.66 (m, 1 H, H-4), 3.763.90 (m, 2 H, H-6, H-7), 4.10 (t, J = 8.7 Hz, 1 H, H-8), 4.19 (s, 2 H, NCH2), 4.72, 4.95 (A, B of AB, 2 H, CH2Ar), 7.26 (t, J = 7.5 Hz, 1 H, CHar), 7.36 (m, 1 H, CHar), 7.387.48 (m, 3 H, CHar), 7.54 (d, J = 7.4 Hz, 2 H, CHar), 7.63 (d, J = 7.3 Hz, 2 H, CHar), 7.69 (d, J = 7.3 Hz, 2 H, CHar), 7.76 (s, 2 H, CHar), 7.87 (s, 1 H, CHar), 8.17 (s, 1 H, CHar), 8.55 ppm (s, 1 H, CHar); 13 C NMR (125 MHz, CD3OD): d = 41.6 (C3), 47.3 (C9), 51.2 (NCH2), 57.0 (C5), 63.6 (CH2Ar), 68.5 (C8), 69.1 (C4), 70.5 (C7), 72.3 (C6), 123.9, 126.6, 126.7, 127.1, 127.5, 127.6, 128.3, 128.4, 129.7, 130.6,
ChemMedChem 2012, 7, 134 143
132.3 ppm (24C, Car); HRMS (ESI) m/z calcd for C35H35Cl2N3O11S [M + Na] + : 798.1267, found: 798.1262.
Biological assays
SiglecFc proteins: Human CD22d13Fc from CHO-Lec1 and murine MAGd13Fc from CHO-Lec3.2.8.1 cell culture supernatants were affinity purified as described previously,[34] dialyzed against 10 mm phosphate buffer (pH 7.4), sterile filtered, and stored at 4 8C. The proteins were analyzed by an ELISA and binding assay with immobilized fetuin.[34] Surface plasmon resonance (SPR): The SPR measurements were performed on a Biacore 3000 SPR-based optical biosensor (Biacore, GE Healthcare, Uppsala, Sweden). Sensor chips (CM5), immobilization kits, maintenance supply, and HBS-EP buffer (10 mm HEPES pH 7.4, 150 mm NaCl, 3 mm EDTA, 0.005 % v/v surfactant P20) were purchased from GE Healthcare. The carboxy groups on the CM5 chip were activated for 10 min with a 1:1 mixture of 0.1 m N-hydroxysuccinimide (NHS) and 0.1 m 3-(N,N-dimethylamino)propyl-N-ethylcarbodiimide (EDC) at a flow rate of 10 mL min1. For immobilizing protein A (P6031, Sigma), a solution of 30 mg mL1 in acetate buffer was injected over the activated surface for 10 min at a flow rate of 10 mL min1. Protein A densities of ~ 4000 RU were achieved. Flow cells were blocked with a 10 min injection of 1 m ethanolamine, pH 8.0. For capturing, hCD22d13Fc or mMAGd13Fc solutions of 3040 mg mL1 concentration in NaOAc (pH 5.0) were injected at a flow rate of 5 mL min1 for 3 min. The surface was equilibrated with HBS-EP buffer overnight at a flow rate of 5 mL min1, achieving densities of ~ 30004000 RU. Tenfold dilution series of antagonists were freshly prepared in HBS-EP. All binding experiments were conducted at 25 8C at a flow rate of 20 mL min1. The samples were injected for 1 min followed by 1 min dissociation. Each sample concentration was measured in duplicate. Double referencing was applied to correct for bulk effects and other systematic artifacts (subtraction of reference surface and blank injections). Data processing and equilibrium binding constant determinations were accomplished with Scrubber (BioLogic Software, Version 1.1g or 2.0c). Kinetic data were simultaneously fit using Scrubber 2.0c. For compounds with low solubility (!17 ac, 20), 5 % DMSO (for molecular biology, > 99.9 %, Fluka) was used in the buffer. The running buffer was 5 % DMSO in HBS-EP. The surface was equilibrated at a flow of 5 mL min1 until the baseline was stable. To eliminate the influence of DMSO on the signals, a calibration curve was recorded with DMSO concentrations of 4.75.1 %. The DMSO calibration was accomplished directly in Scrubber 2.0c. Isothermal titration calorimetry (ITC): ITC experiments were performed using a VP-ITC instrument from MicroCal Inc. (GE Healthcare, Northampton, UK). The measurements were performed at 25 8C. Injections of 510 mL ligand solutions were added from a computer-controlled syringe at an interval of 5 min into the sample cell solution containing hCD22d13Fc[34] (cell volume: 1.4523 mL) with stirring (307 rpm). For control experiments, identical ligand solutions were injected into buffer without protein. The heat released in this control experiment was subtracted from the experimental data. The assay buffer was HBS-E (10 mm HEPES pH 7.4, 150 mm NaCl, 3 mm EDTA, 5 % DMSO). The concentration of hCD22d13Fc was 1.811.1 mm, calculated in terms of monomer, determined by HPLCUV,[54] and the ligand concentration was 60 200 mm. The experimental data were fitted to a theoretical titration curve (one-site binding model) using Origin version 7 (GE Healthcare, Northampton, UK), with DH8 (enthalpy change in kJ mol1), KA (association constant in m1), and N (number of binding sites). The
www.chemmedchem.org
141
MED
quantity c = KAMt0, where Mt0 is the initial macromolecule concentration, is important in titration microcalorimetry. The experiment was performed at c values between 2 and 85. Thermodynamic parameters were calculated from Equation (1): DG DH TDS RT lnK A RT lnK D 1
B. Ernst et al.
partment and 90 mL of blank buffer were added. After protein precipitation with 300 mL CH3CN (4 8C) the solutions were mixed, centrifuged (3600 rpm (1666 g), 4 8C, 11 min) and 50 mL of the supernatant were withdrawn. Analyte concentrations were determined by LCMS. The fraction unbound was calculated by dividing the concentration in the buffer compartment by the concentration in the plasma compartment (both concentrations adjusted for dilutions prior to analysis). The fraction bound was calculated by subtracting the fraction bound from 1. Values were accepted if the recovery of analyte was between 80 and 120 %. LCMS measurements: Separation was performed on an Agilent 1100 Series HPLC instrument with a 1200 series autosampler, connected to an Agilent 6400 Series Triple Quadrupole mass spectrometer for quantification (Agilent Technologies, Santa Clara, CA, USA). Double-distilled H2O with 0.1 % formic acid (A) and CH3CN with 0.1 % formic acid (B) were used as solvents. The gradient was a follows: 0.1 min 95 % A to 5 % B; 1 min 5 % A to 95 % B; 1.2 min 95 % A to 5 % B. The total method duration was 4 min. For the separation, a Waters Atlantis T3 column (3 mm, 2.1 50 mm) was used (Waters, Milford, MA, USA). The column was heated at 60 8C, and the flow rate was set to 0.6 mL min1; 5 mL of analyte were injected per run. Quantification was performed with the MassHunter software (Agilent Technologies, Version B.01.04).
for which DG8, DH8, and DS8 are the changes in free energy, enthalpy, and entropy of binding, respectively. T is the absolute temperature, and R = 8.314 J mol1 K1.[55] log D7.4 determination: Triplicate measurements were performed for every compound at two ratios of octanol to buffer according to the expected log D7.4 value. Equal amounts of phosphate buffer (0.1 m, pH 7.4) and octanol were mixed and shaken vigorously for 5 min. Upon separation of the two phases, the buffer phase was withdrawn and analytes were dissolved therein (105 m). Predefined volumes of octanol and analyte solutions were transferred to a PCR plate, which was thermo-sealed with aluminum foil and shaken (2 h, 1350 rpm, 25 8C on a Heidolph Titramax 100 plate shaker (Heidolph, Schwabach, Germany). The plate was then centrifuged at 2000 rpm (657 g) and 25 8C for 5 min, and buffer samples were withdrawn from each well. The analyte concentrations were determined by LCMS, and log D7.4 values were calculated. Values were accepted if the mean values of the two ratios did not differ by > 0.1 units. Parallel artificial membrane permeation assay (PAMPA): log Pe was determined in a 96-well format. For each compound, measurements were performed in quadruplicate at three pH values: 5.0, 6.2, and 7.4. Each well of a deep-well plate was filled with 650 mL System Solution (pIon, Woburn MA, USA, P/N 110151) at the according pH value. Samples (150 mL) were withdrawn from each well to determine the blank spectra by UV/Vis spectroscopy. Analyte was then added to the remaining System Solution to yield 50 mm solutions. To exclude precipitation, the optical density was measured at l 650 nm, with 0.01 being the threshold value. Again, samples of 150 mL were withdrawn to determine the reference spectra. A further 200 mL were transferred to each well of the donor plate of the PAMPA sandwich. The filter membranes at the bottom of the acceptor plate were impregnated with 5 mL GIT-0 Lipid Solution (pIon, P/N 110669), and 200 mL Acceptor Sink Buffer (pIon, P/N 110139) were filled into each acceptor well. The sandwich was assembled, then placed in the GutBox, left undisturbed for 16 h, and then disassembled again followed by the transfer of 150 mL from each donor and acceptor well to UV plates. Quantification was performed by both UV/Vis spectroscopy (SpectraMax 190, Molecular Devices, Sunnyvale, CA, USA) and LCMS; log Pe values were calculated based on the LCMS results and with the aid of the PAMPA Explorer Software (pIon, version 3.5). Plasma protein binding (PPB): The dialysis membranes (HTDialysis LCC, Gales Ferry, USA; MWCO 1214 kDa) were prepared according to the manufacturers instructions. Human plasma (Biopredic, Rennes, France) was centrifuged (5800 rpm (4325 g), 25 8C, 10 min), the centrifugate (without floating plasma lipids) was adjusted to pH 7.5, and analyte was added to yield 10 mm solutions. Equal volumes (150 mL each) of phosphate buffer (0.1 m, pH 7.5) and analyte-containing plasma were transferred to the separated compartments of the assembled 96-well high-throughput dialysis block (HTDialysis). Measurements were performed in triplicate. The plate was covered with a sealing film and incubated (5 h, 37 8C). Buffer and plasma compartment were processed separately: 90 mL were withdrawn from the buffer compartment and 10 mL blank plasma were added, while 10 mL were withdrawn from the plasma com-
Acknowledgements
We thank the Swiss National Science Foundation (project 200020-120628) for their support of this project. Keywords: carbohydratelectin interactions CD22 MAG pharmacokinetics surface plasmon resonance thermodynamic fingerprint
[1] A. Molina, Annu. Rev. Med. 2008, 59, 237 250. [2] a) C. Bello, E. M. Sotomayor, Hematology Am. Soc. Hematol. Educ. Program 2007, 233 242; b) J. Castillo, E. Winer, P. Quesenberry, Exp. Hematol. 2008, 36, 755 768. [3] M. K. OReilly, J. C. Paulson, Trends Pharmacol. Sci. 2009, 30, 240 248. [4] R. J. Kreitman, I. Margulies, M. Stetler-Stevenson, Q. C. Wang, D. J. FitzGerald, I. Pastan, Clin. Cancer Res. 2000, 6, 1476 1487. [5] J. F. DiJoseph, M. M. Dougher, D. C. Armellino, D. Y. Evans, N. K. Damle, Leukemia 2007, 21, 2240 2245. [6] L. Nitschke, Immunol. Rev. 2009, 230, 128 143. [7] P. R. Crocker, A. Varki, Immunology 2001, 103, 137 145. [8] P. R. Crocker, Curr. Opin. Pharmacol. 2005, 5, 431 437. [9] T. F. Tedder, J. C. Poe, K. M. Haas, Adv. Immunol. 2005, 88, 1 50. [10] K. G. Smith, D. M. Tarlinton, G. M. Doody, M. L. Hibbs, D. T. Fearon, J. Exp. Med. 1998, 187, 807 811. [11] G. M. Doody, L. B. Justement, C. C. Delibrias, R. J. Matthews, J. Lin, M. L. Thomas, D. T. Fearon, Science 1995, 269, 242 244. [12] H. Floyd, L. Nitschke, P. R. Crocker, Immunology 2000, 101, 342 347. [13] N. Razi, A. Varki, Proc. Natl. Acad. Sci. USA 1998, 95, 7469 7474. [14] B. E. Collins, O. Blixt, A. R. DeSieno, N. Bovin, J. D. Marth, J. C. Paulson, Proc. Natl. Acad. Sci. USA 2004, 101, 6104 6109. [15] S. Kelm, J. Gerlach, R. Brossmer, C. P. Danzer, L. Nitschke, J. Exp. Med. 2002, 195, 1207 1213. [16] L. D. Powell, D. Sgroi, E. R. Sjoberg, I. Stamenkovic, A. Varki, J. Biol. Chem. 1993, 268, 7019 7027. [17] a) S. Kelm, A. Pelz, R. Schauer, M. T. Filbin, S. Tang, M. E. de Bellard, R. L. Schnaar, J. A. Mahoney, A. Hartnell, P. Bradfield, Curr. Biol. 1994, 4, 965 972; b) S. Kelm, R. Schauer, J. C. Manuguerra, H. J. Gross, P. R. Crocker, Glycoconjugate J. 1994, 11, 576 585.
142
www.chemmedchem.org
ChemMedChem 2012, 7, 134 143
From MAG Antagonists to CD22 Ligands

[18] S. M. van Rossenberg, L. A. Sliedregt, R. Autar, C. Piperi, A. P. van der Merwe, T. J. van Berkel, J. Kuiper, E. A. Biessen, J. Biol. Chem. 2001, 276, 12967 12973. [19] H. H. Abdu-Allah, K. Watanabe, G. C. Completo, M. Sadagopan, K. Hayashizaki, C. Takaku, T. Tamanaka, H. Takematsu, Y. Kozutsumi, J. C. Paulson, T. Tsubata, H. Ando, H. Ishida, M. Kiso, Bioorg. Med. Chem. 2011, 19, 1966 1971. [20] H. H. Abdu-Allah, T. Tamanaka, J. Yu, L. Zhuoyuan, M. Sadagopan, T. Adachi, T. Tsubata, S. Kelm, H. Ishida, M. Kiso, J. Med. Chem. 2008, 51, 6665 6681. [21] H. H. Abdu-Allah, K. Watanabe, K. Hayashizaki, C. Takaku, T. Tamanaka, H. Takematsu, Y. Kozutsumi, T. Tsubata, H. Ishida, M. Kiso, Bioorg. Med. Chem. Lett. 2009, 19, 5573 5575. [22] T. K. Dam, C. F. Brewer, Glycobiology 2010, 20, 270 279. [23] B. E. Collins, O. Blixt, S. Han, B. Duong, H. Li, J. K. Nathan, N. Bovin, J. C. Paulson, J. Immunol. 2006, 177, 2994 3003. [24] M. K. OReilly, B. E. Collins, S. Han, L. Liao, C. Rillahan, P. I. Kitov, D. R. Bundle, J. C. Paulson, J. Am. Chem. Soc. 2008, 130, 7736 7745. [25] Z. Q. Yang, E. B. Puffer, J. K. Pontrello, L. L. Kiessling, Carbohydr. Res. 2002, 337, 1605 1613. [26] S. Kelm, R. Brossmer, R. Isecke, H. J. Gross, K. Strenge, R. Schauer, Eur. J. Biochem. 1998, 255, 663 672. [27] W. C. Chen, G. C. Completo, D. S. Sigal, P. R. Crocker, A. Saven, J. C. Paulson, Blood 2010, 115, 4778 4786. [28] S. V. Shelke, G. P. Gao, S. Mesch, H. Gthje, S. Kelm, O. Schwardt, B. Ernst, Bioorg. Med. Chem. 2007, 15, 4951 4965. [29] S. Mesch, D. Moser, D. S. Strasser, A. Kelm, B. Cutting, G. Rossato, A. Vedani, H. Koliwer-Brandl, M. Wittwer, S. Rabbani, O. Schwardt, S. Kelm, B. Ernst, J. Med. Chem. 2010, 53, 1597 1615. [30] E. R. Sjoberg, L. D. Powell, A. Klein, A. Varki, J. Cell. Biol. 1994, 126, 549 562. [31] A. Hasegawa, H. Ohki, T. Nagahama, H. Ishida, M. Kiso, Carbohydr. Res. 1991, 212, 277 281. [32] E. Kaiser, J. P. Tam, T. M. Kubiak, R. B. Merrifield, Tetrahedron Lett. 1988, 29, 303 306. [33] P. Boullanger, V. Maunier, D. Lafont, Carbohydr. Res. 2000, 324, 97 106. [34] H. Koliwer-Brandl, N. Siegert, K. Umnus, A. Kelm, A. Tolkach, U. Kulozik, J. Kuballa, S. Cartellieri, S. Kelm, Int. Dairy J. 2011, 21, 413 420. [35] N. R. Zaccai, K. Maenaka, T. Maenaka, P. R. Crocker, R. Brossmer, S. Kelm, E. Y. Jones, Structure 2003, 11, 557 567. [36] R. A. Copeland, D. L. Pompliano, T. D. Meek, Nat. Rev. Drug Discovery 2006, 5, 730 739; D. C. Swinney, J. Pharm. Med. 2008, 22, 23 34. [37] a) M. K. Wild, M. C. Huang, U. Schulze-Horsel, P. A. van der Merwe, D. Vestweber, J. Biol. Chem. 2001, 276, 31602 31612; b) O. Schwardt, H. Gthje, A. Vedani, S. Mesch, G. P. Gao, M. Spreafico, J. von Orelli, S. Kelm, B. Ernst, J. Med. Chem. 2009, 52, 989 1004. [38] T. R. Bakker, C. Piperi, E. A. Davies, P. A. Merwe, Eur. J. Immunol. 2002, 32, 1924 1932. [39] a) P. O. Markgren, M. Hmlinen, U. H. Danielson, Anal. Biochem. 2000, 279, 71 78; b) P. O. Markgren, M. T. Lindgren, K. Gertow, R. Karlsson, M. Hmlinen, U. H. Danielson, Anal. Biochem. 2001, 291, 207 218. [40] a) B. A. Williams, M. C. Chervenak, E. J. Toone, J. Biol. Chem. 1992, 267, 22 907 22 911; b) E. J. Toone, Curr. Opin. Struc. Biol. 1994, 4, 719 728; c) T. K. Dam, C. F. Brewer, Chem. Rev. 2002, 102, 387 429; d) M. Ambrosi, N. R. Cameron, B. G. Davis, Org. Biomol. Chem. 2005, 3, 1593 1608. a) N. Ahmad, H. J. Gabius, S. Sabesan, S. Oscarson, C. F. Brewer, Glycobiology 2004, 14, 817 825; b) Y. Ito, S. Hagihara, M. A. Arai, I. Matsuo, M. Takatani, Glycoconjugate J. 2004, 21, 257 266; c) M. A. Arai, I. Matsuo, S. Hagihara, K. Totani, J. Maruyama, K. Kitamoto, Y. Ito, ChemBioChem 2005, 6, 2281 2289; d) T. K. Dam, H. J. Gabius, S. Andr, H. Kaltner, M. Lensch, C. F. Brewer, Biochemistry 2005, 44, 12564 12571; e) C. F. Brewer, Glycoconjugate J. 2004, 19, 459 465; f) M. S. Quesenberry, R. T. Lee, Y. C. Lee, Biochemistry 1997, 36, 2724 2732; g) P. Srme, P. Arnoux, B. Kahl-Knutsson, H. Leffler, J. M. Rini, U. J. Nilsson, J. Am. Chem. Soc. 2005, 127, 1737 1743; h) K. Bachhawat-Sikder, C. J. Thomas, A. Surolia, FEBS Lett. 2001, 500, 75 79; i) P. Szabo, T. K. Dam, K. Smetana, B. Dvornkov, D. Kbler, C. F. Brewer, H. J. Gabius, Anat. Histol. Embryol. 2009, 38, 68 75. C. O. Sallum, R. A. Kammerer, A. T. Alexandrescu, Biochemistry 2007, 46, 9541 9550. M. Kapoor, H. Srinivas, E. Kandiah, E. Gemma, L. Ellgaard, S. Oscarson, A. Helenius, A. Surolia, J. Biol. Chem. 2003, 278, 6194 6200. M. Kansy, F. Senner, K. Gubernator, J. Med. Chem. 1998, 41, 1007 1010. E. Garcia-Hernandez, R. A. Zubillaga, E. A. Chavelas-Adame, E. VazquezContreras, A. Rojo-Dominguez, M. Costas, Protein Sci. 2003, 12, 135 142. a) C. Hansch, A. Leo, Substituent Constant for Correlation Analysis in Chemistry and Biology, Wiley, New York, 1979; b) M. J. Waring, Exp. Opin. Drug Discov. 2010, 5, 235 248. E. H. Kerns, L. Di, Drug-like Properties: Concepts, Structure Design and Methods, Academic Press, Amsterdam, 2008. G. Corti, F. Maestrelli, M. Cirri, N. Zerrouk, P. Mura, Eur. J. Pharm. Sci. 2006, 27, 354 362. a) P. S. Burton, J. T. Goodwin, T. J. Vidmar, B. M. Amore, J. Pharmacol. Exp. Ther. 2002, 303, 889 895; b) W. J. Egan, G. Lauri, Adv. Drug Delivery Rev. 2002, 54, 273 289. S. Urien, J.-P. Tillement, J. Barre in Pharmacokinetic Optimization in Drug Research: Biological, Physicochemical, and Computational Strategies (Eds.: B. Testa, H. van de Waterbeemed, G. Folkers, R. Guy), Helvetica Chimica Acta, Zrich, 2001, pp. 189 197. A. M. Talbert, G. E. Tranter, E. Holmes, P. L. Francis, Anal. Chem. 2002, 74, 446 452. R. L. Rich, Y. S. Day, T. A. Morton, D. G. Myszka, Anal. Biochem. 2001, 296, 197 207. N. A. Meanwell, J. Med. Chem. 2011, 54, 2529 2591. a) S. Mesch, K. Lemme, H. Koliwer-Brandl, D. S. Strasser, O. Schwardt, S. Kelm, B. Ernst, Carbohydr. Res. 2010, 345, 1348 1359; b) F. Bitsch, R. Aichholz, J. Kallen, S. Geisse, B. Fournier, J. M. Schlaeppi, Anal. Biochem. 2003, 323, 139 149. T. Wiseman, S. Williston, J. F. Brandts, L. N. Lin, Anal. Biochem. 1989, 179, 131 137.
[41]
[42] [43] [44] [45]
[46]
[47] [48] [49]
[50]
[51] [52] [53] [54]
[55]
Received: August 25, 2011 Published online on October 11, 2011
ChemMedChem 2012, 7, 134 143
www.chemmedchem.org
143
MED
DOI: 10.1002/cmdc.201100460
Aspirin Analogues as Dual Cyclooxygenase-2/5Lipoxygenase Inhibitors: Synthesis, Nitric Oxide Release, Molecular Modeling, and Biological Evaluation as AntiInflammatory Agents
Jatinder Kaur, Atul Bhardwaj, Zhangjian Huang, and Edward E. Knaus*[a]
Analogues of aspirin were synthesized through an efficient one-step reaction in which the carboxyl group was replaced by an ethyl ester, and/or the acetoxy group was replaced by an N-substituted sulfonamide (SO2NHOR2 : R2 = H, Me, CH2Ph) pharmacophore. These analogues were designed for evaluation as dual cyclooxygenase-2 (COX-2) and 5-lipoxygenase (5LOX) inhibitors. In vitro COX-1/COX-2 isozyme inhibition studies identified compounds 11 (CO2H, SO2NHOH), 12 (CO2H, SO2NHOCH2Ph), and 16 (CO2Et, SO2NHOH) as highly potent and selective COX-2 inhibitors (IC50 range: 0.070.7 mm), which exhibited appreciable in vivo anti-inflammatory activity (ED50 range: 23.131.4 mg kg1). Moreover, compounds 11 (IC50 = 0.2 mm) and 16 (IC50 = 0.3 mm), with a sulfohydroxamic acid (SO2NHOH) moiety showed potent 5-LOX inhibitory activity. Furthermore, the SO2NHOH moiety present in compounds 11 and 16 was found to be a good nitric oxide (NO) donor upon incubation in phosphate buffer at pH 7.4. Molecular docking studies in the active binding site of COX-2 and 5-LOX provided complementary theoretical support for the experimental biological structureactivity data acquired.
Introduction
Cyclooxygenase isozymes (COX-1 and COX-2) catalyze the metabolic transformation of arachidonic acid (AA) into prostaglandins (PGs), prostacyclin (PGI2), and thromboxane A2 (TxA2).[13] Nonsteroidal anti-inflammatory drugs (NSAIDs) commonly prescribed for the treatment of pain, fever, and inflammation, exert their desirable therapeutic and undesirable side effects through selective/nonselective inhibition of the COX-1 and/or COX-2 isozymes.[49] Acetylsalicylic acid (aspirin, compound 9, Figure 1) is a moderately potent NSAID and non-narcotic analgesic agent frequently present in pain relief and antipyretic medications. Aspirin, which inhibits both COX isozymes via irreversible acetylation of the Ser 530 residue present in the COX binding site, is 10- to 100-fold more potent toward inhibition of COX-1 over the COX-2 isozyme.[10] However, contraindicated side effects that include gastrointestinal (GI), ulcerogenic, hepatic, and renal toxicity are often observed; these are attributed to inhibition of the cytoprotective constitutively expressed COX-1 isozyme that is responsible for the biosynthesis of beneficial PGs involved in the maintenance and homeostatic control of normal physiological functions.[1121] The anti-inflammatory (AI) effects of aspirin are due to inhibition of the inducible COX-2 isoform. The development of selective COX-2 inhibitors (COXIBs), devoid of adverse GI and ulcerogenic effects often associated with the use of traditional nonselective COX-1/COX2 inhibitory NSAIDs, constituted an important milestone in medicinal chemistry drug design.[2225] Unfortunately, some highly selective COX-2 inhibitors such as rofecoxib and valdecoxib were subsequently found to cause adverse cardiovascular thrombotic events (myocardial infarctions and strokes) that resulted in their withdrawal from the market. Highly selective COX-2 inhibitors are believed to cause an adverse biochemical imbalance in the COX pathway where the biosynthesis of the beneficial PGI2 (a potent vasodilator and inhibitor of platelet aggregation) is decreased in conjunction with an increased biosynthesis of prothrombotic TxA2 (a potent vasoconstrictor and platelet aggregator).[2630] AA, in addition to the COX-mediated pathway, is also metabolized to leukotrienes (LTs) via lipoxygenase (LOX) isozymes. There is now evidence which suggests that inhibition of the COX-1/COX-2 isozymes by NSAIDs may result in more extensive metabolism of AA by the 5-LOX-mediated pathway, resulting in elevated biosynthesis of inflammatory LTs. LTs such as LTB4, formed via the 5-LOX pathway, are also associated with various side effects that include a possible contribution to the development of atherosclerosis, myocardial infarction, and asthmatic actions.[31] Accordingly, a dual inhibitor of the LOX/ COX enzymatic pathways[32] represents a rational approach for the design of more efficacious AI agents with a superior safety profile over those of selective COX-2 inhibitors that increase the incidence of adverse thrombotic events and elevate the levels of inflammatory LTs.[33, 34] The concept to replace the acetoxy (OAc) moiety of aspirin with a sulfohydroxamic (SO2NHOR: R = H, CH2Ph, Me) group
[a] Dr. J. Kaur, Dr. A. Bhardwaj, Dr. Z. Huang, Dr. E. E. Knaus Faculty of Pharmacy and Pharmaceutical Sciences University of Alberta, Edmonton, AB, T6G 2N8 (Canada) E-mail: ekanus@pharmacy.ualberta.ca Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100460.
144
ChemMedChem 2012, 7, 144 150
Aspirin Analogues as Dual Inhibitors bonate and para-toluenesulfonic acid in dry THF at 25 8C for 1 h furnished the sulfohydroxamic acid (SO2NHOH) product 11 (64 % isolated yield). A similar reaction of 10 with N-benzyloxyamineHCl or methoxyamineHCl gave the respective products 12 (80 %) and 13 (60 %), as illustrated in Scheme 1. The main
Figure 1. Chemical structures of some representative COX-1/COX-2 inhibitors 13 and 9, a selective COX-2 inhibitor 4, 5-LOX inhibitors 5 and 6, dual COX/5-LOX inhibitors 7 and 8, and the target sulfohydroxamic acid/N-substituted sulfonamide analogues of aspirin 1113 and 1618.
was based on the expectation that: 1) the SO2NHOR substituent would act as a COX-2 pharmacophore, like the SO2NH2 (celecoxib, 4) COX-2 pharmacophore; 2) the SO2NHOH group would act as a 5-LOX pharmacophore similar to hydroxamic acid (CONHOH), N-hydroxyurea (zileuton, 5), or N-hydroxyamide (tepoxalin, 6) based 5-LOX inhibitors[35, 36] (see structures in Figure 1) that act by chelation of iron present in the 5-LOX enzyme; and 3) the SO2NHOH group would act as nitric oxide (NO) donor to induce vascular smooth-muscle relaxation, decrease in blood pressure, and inhibition of platelet aggregation, providing chemoprotection against thrombotic events.[37, 38] As part of our ongoing program to develop new COX-2 and 5-LOX pharmacophores and NO donor moieties, herein we describe the synthesis of a group of CO2H (compounds 1113) and CO2Et (compounds 1618) analogues of aspirin (see structural comparisons shown in Figure 1). In vitro enzyme inhibition (COX-1, COX-2, 5-LOX), NO release, molecular modeling, and AI studies were carried out to acquire a set of structureactivity relationship data for this new type of NOreleasing dual COX-2/5-LOX inhibitory AI agent.
Scheme 1. Reagents and conditions: a) HONH2HCl, anhyd K2CO3, p-toluenesulfonic acid, dry THF, 25 8C, 1 h for compound 11; PhCH2ONH2HCl, anhyd K2CO3, p-toluenesulfonic acid, dry THF, 25 8C, 3 h for compound 12; b) MeONH2HCl, NaHCO3, p-toluenesulfonic acid, dry THF, 25 8C, 3 h; c) EtOH, NH3 ; d) SOCl2, dry DMF, reflux; e) HONH2HCl, NaHCO3, THF, 25 8C, 2 h; f) PhCH2ONH2HCl, NaOH, THF, 1 h for compound 17; MeONH2HCl, NaOH, THF, 3 h for compound 18.
advantage of this strategy is that it is a one-step reaction, which conveniently gave the target products formed by the required cleavage of the sulfonyloxygen bond, rather than the undesired carbonyloxygen bond present in 10. Furthermore, the products can be readily isolated by recrystallization from ethanol and ethyl acetate. In addition, compounds 1618 bearing an ethyl ester group in place of the carboxyl substituent present in compounds 1113 can be synthesized by using a reported procedure.[39] Thus, reaction of 10 with ammonia affords the ammonium salt 14, which is readily elaborated to the sulfonyl chloride 15. Subsequent reaction of the sulfonyl chloride 15 with either HONH2, PhCH2ONH2, or MeONH2 afforded the respective target products 1618 (Scheme 1). COX-1 and COX-2 inhibition In vitro COX-1/COX-2 isozyme inhibition studies (Table 1) showed that the compounds 1113 and 1618 are more potent inhibitors of COX-2 (IC50 range: 0.0746.6 mm) than COX-1 (IC50 range: 86.9 to > 100 mm). In this regard, replacement of the OAc group in aspirin, which is a more potent and selective inhibitor of COX-1 (COX-2 SI = 0.14), with a sulfohy-

Chemistry Reaction of 2-sulfobenzoic acid cyclic anhydride (10) with hydroxyamineHCl in the presence of anhydrous potassium carChemMedChem 2012, 7, 144 150
www.chemmedchem.org
145
MED
Table 1. In vitro inhibition, anti-inflammatory and NO-release data for COX-1, COX-2, and 5-LOX. Compd 11 12 13 16 17 18 4 9 5 R1 R2 IC50 [mm][a] COX-1 COX-2 86.9 > 100 91.6 > 100 > 100 96.6 7.7 0.35 0.73 0.07 35.2 0.09 20.1 46.6 0.07 2.4 SI[b] COX-2 119 > 1428 2.6 > 1111 > 4.9 2.1 110 0.14 IC50 [mm][c] 5-LOX 0.2 1.4 0.8 0.4 2.21 0.98 1.1 13 ED50 [mg kg1][d] AI 23.1 31.4 NT[g] 24.5 NT NT 10.8 128 1h 22 5.2 9.0 21 4 8.5 15
E. E. Knaus et al.
NO release [%][e,f] 4h 16 h 28 5.8 9.2 29 4.2 8.6 26 33 6 9.8 32 4.8 9.4 48
24 h 34 6.1 10.2 24 4.9 9.7 53
H H H CH2Ph H CH3 H C2H5 CH2Ph C2H5 CH3 C2H5 celecoxib aspirin zileuton NDGA[h] PA[i]
[a] In vitro test compound concentration required to produce 50 % inhibition of ovine COX-1 or human recombinant COX-2; values represent the mean of two determinations, and the deviation from the mean is < 10 % of the mean value. [b] In vitro COX-2 selectivity index (COX-1 IC50/COX-2 IC50). [c] The in vitro test compound concentration required to produce 50 % inhibition of potato 5-LOX; values represent the mean of two determinations, and the deviation from the mean is < 10 % of the mean value. [d] Inhibitory activity in a carrageenan-induced rat paw edema assay (mean value, n = 3); data collected at 3 h after oral administration of the test compound. [e] Percent NO released (mean value, n = 3) relative to a theoretical maximum release of 1 mol NO per mol test compound, as determined by using the Griess reaction; variation from the mean value was 0.02 %. [f] Incubated for various time intervals in PBS, pH 7.4, at 37 8C. [g] NT: not tested. [h] NDGA: nordihydroguaiaretic acid. [i] PA, benzenesulfohydroxamic acid (PhSO2NHOH; Pilotys acid).
droxamic acid moiety inverts the COX isozyme inhibitory and selectivity profile to furnish compounds that become more potent and selective COX-2 inhibitors. Within this group of compounds, compound 12 bearing a SO2NHOCH2Ph substituent (COX-2 IC50 = 0.07 mm and SI > 1428) and compound 16 with a SO2NHOH substituent (COX-2 IC50 = 0.09 mm and SI > 1111) emerged as the most potent and selective COX-2 inhibitors, being equipotent to the reference drug celecoxib (COX-2 IC50 = 0.07 mm and SI = 110). The SO2NHOR2 substituent is a determinant of COX-2 inhibitory activity, in which the relative potency order with respect to the R2 substituent was observed to be CH2Ph > H > Me. Replacement of the CO2H group in compounds 1113 by a CO2Et substituent furnished compounds 1618 that resulted in variable effects on COX-2 potency, which was increased ~ 8-fold for the SO2NHOH compound (16 > 11), decreased ~ 287-fold for the SO2NHOCH2Ph compound (12 > 17), and slightly decreased ~ 1.3-fold for the SO2NHOMe compound (13 > 18). In vitro studies showed that all compounds (1113 and 16 18) exhibited appreciable 5-LOX inhibition (IC50 range: 0.2 2.21 mm) relative to the reference drugs zileuton (IC50 = 1.1 mm) and NDGA (IC50 = 13 mm). The SO2NHOH compounds 11 (IC50 = 0.2 mm) and 16 (IC50 = 0.4 mm) showed almost equipotent inhibition of 5-LOX. Structureactivity data acquired show that: 1) the sulfohydroxamic acid pharmacophore is a determinant of 5-LOX inhibition, for which the relative potency sequence is SO2NHOH > SO2NHOMe > SO2NHOCH2Ph, and 2) the carboxylate moiety is not a major determinant of potency, as replacement of the CO2H group in compounds 1116 by a CO2Et substituent in compounds 1618 results in only a small decrease in 5-LOX inhibitory activity.
Anti-inflammatory activity The oral AI activities exhibited by the N-hydroxy/benzyloxysulfonamides 11, 12, and 16, which exhibited the best combination of COX-2 and 5-LOX inhibitory activities, were determined by using a carrageenan-induced rat foot paw edema model (see data in Table 1). Compounds 11, 12, and 16 exhibited appreciable AI activities (ED50 range: 23.131.4 mg kg1 p.o.) that were intermediate between those of the reference drugs celecoxib (ED50 = 10.8 mg kg1 p.o.) and aspirin (ED50 = 128 mg kg1 p.o.). One plausible explanation for the observation that the CO2H compound 11 and the CO2Et compound 16 exhibit equipotent AI activity may be due, at least in part, to esterasemediated cleavage of the CO2Et group to a CO2H substituent following oral (p.o.) administration of the ester 16. Nitric oxide release The percent NO released from the N-substituted sulfonamides 1113, 1618, and the well-known NO donor reference compound Pilotys acid (PA), upon incubation in phosphate-buffered saline (PBS) at pH 7.4 was measured by quantitation of nitrite using the Griess reaction (see data in Table 1) for incubation times of 1, 4, 16, and 24 h. Within the group of compounds 1113 and 1618, compounds 11 and 16, with a free SO2NHOH group, released nearly equal amounts of NO at all four times, increasing from 2122 % at 1 h to a maximal release of 34 % at 24 h. The sulfohydroxamic acid moiety was observed to be a determinant of the amount of NO released, where the relative release profile is SO2NHOH > SO2NHOMe > SO2NHOCH2Ph. The release of NO was expected to be lower for compounds that possess a SO2NHOMe or SO2NHOCH2Ph substituent, as O-demethylation or O-debenzylation must occur before NO can be released from compounds 12 and 13 or 17 and 18. These NO-release data validate the concept that
ChemMedChem 2012, 7, 144 150
146
www.chemmedchem.org
Aspirin Analogues as Dual Inhibitors aspirin analogues bearing a SO2NHOH moiety, like PA (see Table 1), can release NO, which has the potential to induce desirable cardiovascular effects including vasodilation and inhibition of platelet aggregation. Release of NO from compounds 11 and 16 is believed to occur by the mechanism proposed by Zamora et al.,[40] wherein oxidation of PA at pH 78 occurs on the intact molecule to furnish a radical intermediate that subsequently decomposes to release benzenesulfinic acid (PhSO2H) and NO. Molecular modeling (docking) studies To explore, at the molecular level, the enzymeligand interactions that probably govern the recognition and binding of compounds described herein to COX-1, COX-2, and 5-LOX, molecular docking experiments were carried out using enzyme crystal structure data obtained from the RCSB Protein Data Bank.[4143] In contrast to the orientation of aspirin,[44] molecular docking of compound 12 (R1 = H, R2 = CH2Ph) within the COX-2 active site shows that the benzoic acid component of compound 12 (COX-2 IC50 = 0.07 mm and AI ED50 = 31.4 mg kg1 p.o.) is positioned in the vicinity of the secondary pocket of the COX-2 isozyme, surrounded by residues His 90, Glu 192, Ala 516, Val 523, and Ile 517 (Figure 2 and Figure 3 a). The carsimilarly positioned to the phenyl ring of aspirin (Figure 3 a) in the hydrophobic pocket of the COX-2 isozyme (lined by Trp 387, Met 522, Phe 518, Leu 385, Tyr 385, Tyr 348, and Glu 526). Thus, the increased COX-2 inhibitory activity (COX-2 IC50 = 0.07 mm) observed upon replacement of the OAc group in aspirin by a SO2NHOCH2Ph moiety may be attributed to this peculiar orientation of compound 12, in which part of the compound (PhCH2O) overlaps the parent drug aspirin, and the other half (phenyl ring with carboxyl group) favorably moves toward the secondary pocket (Figure 3 a) and shows appreciable electrostatic interactions with COX-2 active site residues.
Figure 3. A comparative view of binding conformations found for a) 12 (pink) and b) 16 (pink) in comparison with aspirin (green) docked in the COX-2 active site.
Figure 2. Molecular docking of compound 12 in the binding site of COX-2 (PDB ID: 6COX; Eintermolecular = 15.57 kcal mol1). Hydrogen atoms of amino acid residues are omitted for clarity.
boxyl group is completely inserted into the secondary pocket region, where it is oriented toward Arg 513, His 90, and Ser 353. The hydroxy group of the CO2H substituent shows hydrogen bonds to a nitrogen atom of the imidazole ring of His 90 (OHN d = 2.14 ) and to the oxygen atom of the amide carbonyl group of Ser 353 (OHO d = 2.90 ). Interestingly, comparison of the orientation of compound 12 with that of the parent drug aspirin (Figure 3 a) shows that the phenyl ring present in the PhCH2ONHSO2 substituent of compound 12 is
ChemMedChem 2012, 7, 144 150
Similar to compound 12, a molecular docking study showed the best-scoring binding position (repeatedly obtained) for compound 16 (R1 = C2H5, R2 = H) is located near the secondary pocket region of the COX-2 active site (see Supporting Information figure S1), where the oxygen atom of the hydroxy group (SO2NHOH) shows hydrogen bonding to a nitrogen atom of the His 90 residue (OHN d = 2.92 ) and to one of the terminal amino groups of Arg 513 (OHNH2 d = 2.83 ). A comparison of the position of aspirin and compound 16 within the COX-2 active site as illustrated in Figure 3 b shows that in contrast to aspirin, compound 16 resides at the entrance to the secondary pocket region. This favorable orientation and electrostatic interactions with COX-2 active site residues may be a plausible explanation for the superior activity profile of compound 16 (COX-2 IC50 = 0.09 mm; AI ED50 = 24.5 mg kg1 p.o.) relative to that of aspirin. Notably, docking of compounds 12 or 16 with the COX-1 enzyme did not indicate any significant
www.chemmedchem.org
147
MED
interactions with COX-1 residues (see Supporting Information figures S2 and S3). The larger SO2NHOCH2Ph moiety of compound 12 may hinder the suitable orientation of compound 12 within the COX-1 binding site. Probable ligandenzyme modes of binding of compounds 11 and 16 with the 5-LOX active site were examined by molecular docking studies using the recently reported X-ray crystal structure of 5-LOX.[43] Docking compound 11 (5-LOX IC50 = 0.2 mm) in the active site shows that the sulfohydroxamic group assumes a position in the vicinity of catalytic iron (His 367, His 372, and Ile 673; Figure 4). The calculated distance
E. E. Knaus et al. pharmacophore present in compounds 11 and 16 to the catalytic iron of 5-LOX, along with significant electrostatic interactions with active site residues, provides complementary theoretical support for the appreciable 5-LOX inhibitory and in vivo AI activities observed.
Conclusions
A group of aspirin analogues were developed by incorporating a NO-releasing sulfohydroxamic pharmacophore that makes these compounds dual COX-2/5-LOX inhibitors. The target carboxylic acid compounds 1113 were synthesized by using an efficient one-step reaction. In vitro COX-1/COX-2 inhibition studies showed that this new class of sulfohydroxamic acids inhibit the COX-2 isozyme (IC50 values in the 0.0746.6 mm range). In particular, compounds 11, 12, and 16 were found to be potent (COX-2 IC50 range: 0.070.7 mm) and selective (COX2 SI range: 119 to > 1428) COX-2 inhibitors that exhibited appreciable in vivo anti-inflammatory activity (ED50 range: 23.1 31.4 mg kg1). Molecular docking analyses for compounds 12 and 16 on the COX-2 isozyme, and compounds 11 and 16 on the 5-LOX isozyme, complement the experimental results. The appreciable anti-inflammatory activity exhibited by the sulfohydroxamic compounds 11 and 16, in conjunction with the timedependent slow release of NO, makes them good lead compounds with potentially good gastrointestinal and cardiovascular safety profiles for the treatment of arthritic conditions.
Figure 4. Molecular modeling (docking) of compound 11 (carbon atoms in green) in the binding site of 5-LOX (PDB ID: 308Y; Eintermolecular = 13.22 kcal mol1). Iron is represented as an orange sphere.
General
Melting points were measured in capillaries using a Thomas Hoover capillary apparatus and are uncorrected. 1H and 13C NMR spectra were recorded on a Bruker AM 300 NMR spectrometer using CDCl3/[D6]DMSO as solvent. Chemical shifts are given in parts per million with tetramethylsilane (TMS) as an internal reference. IR spectra were recorded as films on NaCl plates using a Nicolet 550 series II Magna FTIR spectrometer. MS data were recorded on a Waters Micromass ZQ 4000 mass spectrometer using ESI mode. Compound purity was established by elemental analysis, which was performed for C, H, N, S by the Microanalytical Service Laboratory, Department of Chemistry, University of Alberta. Compounds showed a single spot on MachereyNagel Polygram Sil G/UV254 silica gel plates (0.2 mm) using a low, medium, and highly polar solvent system, and no residue remained after combustion, indicating a purity > 98 %. Column chromatography was performed on a Combiflash Rf system using a gold silica column. All other reagents, purchased from Aldrich Chemical (Milwaukee, WI, USA), were used without further purification. In vivo anti-inflammatory assays were carried using a protocol approved by the Health Sciences Animal Welfare Committee at the University of Alberta. Compounds 1618 were synthesized according to a previously reported procedure.[39] 2-Hydroxysulfamoylbenzoic acid (11): HydroxyamineHCl (101 mg, 1 mmol) and anhydrous K2CO3 (202 mg, 2 mmol) in dry THF (5 mL) was stirred at 25 8C for 10 min to neutralize the hydroxyamineHCl salt. The reaction mixture was cooled to 0 8C, 2-sulfobenzoic acid cyclic anhydride (6, 135 mg, 1 mmol) in dry THF (5 mL) was added dropwise, and finally a catalytic amount of p-toluenesulfonic acid was added. The reaction was allowed to proceed with vigorous
ChemMedChem 2012, 7, 144 150
between the NH group of the SO2NHOH pharmacophore and iron is 2.72 . Moreover, within this region of the 5-LOX active site, the nitrogen atom of the SO2NHOH group shows a hydrogen bond interaction with His 367 (NHN d = 1.94 ), and one of the oxygen atoms attached to sulfur (SO2NH) shows hydrogen bond with an NH2 group (ONH2 d = 2.54 ) of the His 372 residue. The terminal hydroxy group of the sulfohydroxamic acid (SO2NHOH) moiety is tilted toward the carboxyl group of His 550 (d = 2.68 ), which makes hydrogen bonding possible (Figure 4). Similar to the docking interactions observed for compound 11, a docking study with compound 16 shows that the sulfohydroxamic pharmacophore is tilted toward Val 671, His 372, and His 550 of the 5-LOX binding site (see Supporting Information figure S4), and it adopts a position near the catalytic iron center (NH to Fe distance: 3.01 ). Furthermore, the hydroxy group (SO2NHOH) undergoes hydrogen bonding with a nitrogen atom (OHNH d = 2.91 ) of the heterocyclic imidazole ring of His 372, and both of the oxygen atoms attached to sulfur (SO2NH) show hydrogen bonding with one of the nitrogen atoms of His 372. The ester group (COOC2H5) present in compound 16 is oriented in the vicinity of Leu 607, and the oxygen atom of the carbonyl group shows hydrogen bonding with one of the nitrogen atoms of His 550 (see Supporting Information figure S4). The close proximity of the SO2NHOH
148
www.chemmedchem.org
Aspirin Analogues as Dual Inhibitors

stirring at 25 8C for ~ 1 h during which time the disappearance of compound 6 was monitored (TLC; CH2Cl2/MeOH 1:3, v/v). After completion, the reaction mixture was filtered through a pad of Celite to furnish a clear filtrate, which was concentrated in vacuo. Removal of THF in vacuo gave a white semisolid, which was washed with Et2O to furnish a white solid that was recrystallized from the mixture of EtOH and EtOAc (1:3 v/v) to yield the product 11 (64 %) as a white solid; mp: 9394 8C (lit.[45] mp: 9193 8C); 1 H NMR (300 MHz, [D6]DMSO): d = 7.407.61 (m, 2 H, phenyl H-4, H-5), 7.727.84 (m, 2 H, phenyl H-3, H-6), 9.87 and 10.06 (two peaks, 1 H each, HO-NH), 11.51 ppm (br s, 1 H, COOH); 13C NMR (75 MHz, [D6]DMSO): d = 125.5, 129.2, 129.4, 130.5, 131.1, 141.7, 167.5 ppm; IR (film) n = 3275, 3175, 3032, 1732, 1147 cm1; ESIMS: 216 [MH] ; Anal. calcd for C7H7NO5S: C 38.71, H 3.25, S 14.76, N 6.45, found: C 38.77, H 3.28, S 14.70, N 6.49. 2-Benzyloxysulfamoylbenzoic acid (12): BenzyloxyamineHCl (172 mg, 1 mmol) and anhydrous K2CO3 (86 mg, 2 mmol) in dry THF (5 mL) was stirred at 25 8C for 15 min to neutralize the benzyloxyamineHCl, then the reaction mixture was cooled to 0 8C, and a solution of 2-sulfobenzoic acid cyclic anhydride (6, 100 mg, 1 mmol) in dry THF (5 mL) was added dropwise with stirring. A catalytic amount of p-toluenesulfonic acid was added, and the reaction was allowed to proceed with vigorous stirring at 35 8C for ~ 3 h, during which time the reaction progress was monitored for the disappearance of compound 6 (TLC; CH2Cl2/MeOH 1:2, v/v). After completion, the reaction mixture was filtered through a pad of Celite to furnish a clear filtrate. Removal of the solvent in vacuo gave a white solid, which was recrystallized from a mixture of EtOH/EtOAc (1:3, v/v) + Et2O (a few drops) to furnish the product 12 (85 %) as a white solid; mp 149151 8C; 1H NMR (300 MHz, [D6]DMSO): d = 4.92 (s, 2 H, OCH2), 7.297.56 (m, 8 H, phenyl hydrogen), 7.767.81 (m, 1 H, phenyl hydrogen), 12.0 (s, 1 H, NH), 12.0 ppm (br, 1 H, COOH); 13C NMR (75 MHz, [D6]DMSO): d = 76.5, 126.3, 128.0, 128.2, 128.4, 128.5, 128.7, 129.1, 129.7, 130.0, 131.4, 136.2, 144.2, 167.1 ppm; IR (film) n = 3286, 2919, 1729, 1191 cm1; ESIMS: 306 [MH] ; Anal. calcd for C14H13NO5S: C 54.71, H 4.26, S 10.43, N 4.56, found: C 54.69, H 4.31, S 10.37, N 4.59. 2-Methoxysulfamoylbenzoic acid (13): MethoxyamineHCl (90 mg, 1 mmol) was neutralized with Na2CO3 (100 mg, 2 mmol) in dry THF (5 mL), and upon reaction completion, the impure product 13 was isolated as a transparent viscous oil using the same methodology previously described for the preparation of 11 and 12. This transparent viscous oil, which, upon washing with Et2O gave a white solid, which was recrystallized from a mixture of EtOH/EtOAc (1:3, v/v) to afford the product 13 (60 %) as a white solid; mp 130 8C; 1 H NMR (300 MHz, [D6]DMSO): d = 3.72 (s, 3 H, OCH3), 7.417.55 (m, 3 H, phenyl hydrogen), 7.777.81 (m, 1 H, phenyl hydrogen), 10.45 (br s, 1 H, NH), 11.77 ppm (br, 1 H, COOH); 13C NMR (75 MHz, [D6]DMSO): d = 62.0, 126.4, 129.2, 129.7, 131.0, 131.4, 144.2, 166.7 ppm; IR (film) n = 3392, 3237, 1740, 1183 cm1; ESIMS: 230 [MH] ; Anal. calcd for C8H9NO5S: C 41.55, H 3.92, S 13.87, N 6.06, found: C 41.63, H 3.95, S 13.93, N 6.09.
5-LOX inhibition assay

The ability of the test compounds 1113 and 1618 listed in Table 1 to inhibit potato 5-LOX (cat. no. 60401, Cayman Chemical) was determined by using an enzyme immunoassay (EIA) kit (cat. no. 760700, Cayman Chemical) according to our previously reported method.[47]
Anti-inflammatory assay
In vivo anti-inflammatory activity was determined by using a carrageenan-induced rat paw edema assay described by Winter et al.[48] In brief, three male SpragueDawley rats weighing 160180 g were used in each group. Test compounds 11, 12, and 16 suspended in water containing 1 % methyl cellulose were administered orally for a minimum of three different doses (range: 1050 mg kg1) 1 h prior to a 0.05 mL subcutaneous injection of 1 % carrageenan in 0.9 % NaCl solution under the planter skin of the right hind paw. Control experiments were identical, with vehicle containing no test compound. The volume of the injected paw was measured at 0 and 3 h using a UGO Basile 7141 Plethysmometer (series no. 43201). A doseresponse curve was constructed which was used to determine the ED50 value.
Nitric oxide release assay

In vitro NO release, upon incubation of the test compound (2.4 mL of 5.0 102 mm) with PBS at pH 7.4 and 37 8C for 1, 4, 16, or 24 h was determined by quantification of nitrite produced by the reaction of NO with O2 and H2O using the Griess reaction. The amount of NO liberated at each time interval was acquired for test compounds 1113 and 1618 by using reported procedures.[49]
Molecular modeling (docking) procedure

The molecular docking experiments were performed using resolved coordinates for the X-ray crystal structures of COX-1 (ovine, PDB ID: 1EQG, ibuprofen bound in the active site), COX-2 (murine, PDB ID: 6COX, SC558 bound in the active site), and 5-LOX (PDB ID: 308Y, iron bound in the enzyme) that are available from the PDB. Compound structures were built using the toolkit in the software package ArgusLab 4.0.1[50] and energy minimized using the semiempirical quantum mechanical method PM3. The monomeric structure of the target enzyme was selected, and the active site was defined around the ligand. The molecule to be docked in the active site of the enzyme was introduced in the work space carrying the structure of the enzyme. The docking program implements an efficient grid-based docking algorithm, which approximates an exhaustive search within the free volume of the binding site cavity. The conformational space was surveyed by geometry optimization of the flexible ligand (rings are treated as rigid) in combination with the incremental construction of the ligand torsions. Thus, during the experiment, docking occurred between the flexible ligand parts of the compound and the enzyme. The ligand orientation was determined by a shape-scoring function based on Ascore, and the final positions were ranked by lowest interaction energy values. The calculated Einteraction value is the sum of the energies involved in hydrogen bonds, hydrophobic interactions, and van der Waals interactions. Distance measurements of hydrogen bonds and hydrophobic interactions between the compound and enzyme were acquired. Visual inspection for 5-LOX docking interactions was carried using Discovery studio 3.0.[51]
COX inhibition assays

The ability of the test compounds 1113 and 1618 listed in Table 1 to inhibit ovine COX-1 and human recombinant COX-2 was determined by using an enzyme immunoassay (EIA) kit (cat. no. 560131, Cayman Chemical, Ann Arbor, MI, USA) according to a previously reported method.[46]
ChemMedChem 2012, 7, 144 150
www.chemmedchem.org
149
MED
Acknowledgements
We are grateful to the Canadian Institutes of Health Research (grant no. MOP-14712) for financial support of this research. Keywords: anti-inflammatory activity inhibition lipoxygenase nitric oxide cyclooxygenase
E. E. Knaus et al.
[24] G. Szab, J. Fischer, . Kis-Varga, K. Gyires, J. Med. Chem. 2008, 51, 142 147. [25] J. J. Talley, D. L. Brown, J. S. Carter, M. J. Graneto, C. M. Koboldt, J. L. Masferrer, W. E. Perkins, R. S. Rogers, A. F. Shaffer, Y. Y. Zhang, B. S. Zweifel, K. Seibert, J. Med. Chem. 2000, 43, 775 777. [26] C. Bombardier, L. Laine, A. Reicin, D. Shapiro, R. Burgos-Vargas, B. Davis, R. Day, M. B. Ferraz, C. J. Hawkey, M. C. Hochberg, T. K. Kvien, T. J. Schnitzer, N. Engl. J. Med. 2000, 343, 1520 1528. [27] R. S. Bresalier, R. S. Sandler, H. Quan, J. A. Bolognese, B. Oxenius, K. Horgan, C. Lines, R. Riddell, D. Morton, A. Lanas, M. A. Konstam, J. A. Baron, N. Engl. J. Med. 2005, 352, 1092 1102. [28] T. Grosser, S. Fries, G. A. FitzGerald, J. Clin. Invest. 2006, 116, 4 1. [29] G. A. FitzGerald, N. Engl. J. Med. 2004, 351, 1709 1711. [30] J. A. Baron, R. S. Sandler, R. S. Bresalier, A. Lanas, D. G. Morton, R. Riddell, E. R. Iverson, D. L. DeMets, Lancet 2008, 372, 1756 1764. [31] A. Moreau, P. N. Praveen Rao, E. E. Knaus, Bioorg. Med. Chem. 2006, 14, 5340 5350 and references cited therein. [32] P. N. Praveen Rao, Q.-H. Chen, E. E. Knaus, J. Med. Chem. 2006, 49, 1668 1683. [33] A. J. Scheen, Rev. Med. Liege 2004, 59, 565 569. [34] J.-M. Dogn, C. T. Supuran, D. Pratico, J. Med. Chem. 2005, 48, 2251 2257. [35] E. M. F. Muri, M. J. Nieto, R. D. Sindelar, J. S. Williamson, Curr. Med. Chem. 2002, 9, 1631 1653. [36] G. W. Carter, P. R. Young, D. H. Albert, J. Bouska, R. Dyer, R. L. Bell, J. B. Summers, D. W. Brooks, J. Pharmacol. Exp. Ther. 1991, 256, 929 937. [37] A. R. Butler, D. L. H. Williams, Chem. Soc. Rev. 1993, 22, 233 241. [38] J. E. Keeble, P. K. Moore, Br. J. Pharmacol. 2002, 137, 295 310. [39] H. T. Nagasawa, S. P. Kawle, J. A. Elbeding, E. G. DeMaster, J. M. Fukuto, J. Med. Chem. 1995, 38, 1865 1871. [40] R. Zamora, A. Grzesiok, H. Weber, M. Feelisch, Biochem. J. 1995, 312, 333 339. [41] B. S. Selinsky, K. Gupta, C. T. Sharkey, P. J. Loll, Biochemistry 2001, 40, 5172 5180. [42] R. G. Kurumbail, A. M. Stevens, J. K. Gierse, J. J. McDonald, R. A. Stegeman, J. Y. Pak, D. Gildehaus, J. M. Miyashiro, T. D. Penning, K. Seibert, P. C. Isakson, W. C. Stallings, Nature 1996, 384, 644 648. [43] N. C. Gilbert, S. G. Bartlett, M. T. Waight, D. B. Neau, W. E. Boeglin, A. R. Brash, M. E. Newcomer, Science 2011, 331, 217 219. [44] P. J. Loll, D. Picot, R. M. Garavito, Nat. Struct. Biol. 1995, 2, 637 643. [45] F. Mincione, L. Menabuoni, F. Briganti, G. Mincione, A. Scozzafava, C. T. Supuran, J. Enzyme Inhib. 1998, 13, 267 284. [46] M. J. Uddin, P. N. P. Rao, R. McDonald, E. E. Knaus, J. Med. Chem. 2004, 47, 6108 6111. [47] M. A. Chowdhury, K. R. A. Abdellatif, Y. Dong, D. Das, M. R. Suresh, E. E. Knaus, Bioorg. Med. Chem. Lett. 2008, 18, 6138 6141. [48] C. A. Winter, E. A. Risley, G. W. Nuss, Proc. Soc. Exp. Biol. Med. 1962, 111, 544 547. [49] M. T. Cocco, C. Congiu, V. Onnis, M. Morelli, O. Cauli, Eur. J. Med. Chem. 2003, 38, 513 518. [50] A. Mark, ArgusLab version 4.0.1, Thompson Planaria Software LLC, Seattle, WA (USA), http://www.arguslab.com (accessed October 28, 2011). [51] Discovery Studio 3.0, Molecular Modeling Software, Accelrys Inc., http://accelrys.com/products/discovery-studio/ (accessed October 28, 2011). Received: October 4, 2011 Published online on November 18, 2011
[1] M. Hamberg, B. Samuelsson, J. Biol. Chem. 1967, 242, 5336 5343. [2] T. Miyamoto, N. Ogino, S. Yamamoto, O. Hayaishi, J. Biol. Chem. 1976, 251, 2629 2636. [3] B. S. Fletcher, D. A. Kujubu, D. M. Perrin, H. R. Herschman, J. Biol. Chem. 1992, 267, 4338 4344. [4] K. Brune, B. Hinz, Arthritis Rheum. 2004, 50, 2391 2399. [5] L. J. Marnett, Annu. Rev. Pharmacol. Toxicol. 2009, 49, 265 290. [6] A. K. Chakraborti, S. K. Garg, R. Kumar, H. F. Motiwala, P. S. Jadhavar, Curr. Med. Chem. 2010, 17, 1563 1593. [7] Y. Harrak, G. Casula, J. Basset, G. Rosell, S. Plescia, D. Raffa, M. G. Cusimano, R. Pouplana, M. D. Pujol, J. Med. Chem. 2010, 53, 6560 6571. [8] R. R. Ranatunge, M. Augustyniak, U. K. Bandarage, R. A. Earl, J. L. Ellis, D. S. Garvey, D. R. Janero, L. G. Letts, A. M. Martino, M. G. Murty, S. K. Richardson, J. D. Schroeder, M. J. Shumway, S. W. Tam, A. M. Trocha, D. V. Young, J. Med. Chem. 2004, 47, 2180 2193. [9] M. Anzini, M. Rovini, A. Cappelli, S. Vomero, F. Manetti, M. Botta, L. Sautebin, A. Rossi, C. Pergola, C. Ghelardini, M. Norcini, A. Giordani, F. Makovec, P, Anzellotti, P. Patrignani, M. Biava, J. Med. Chem. 2008, 51, 4476 4481. [10] E. A. Meade, W. L. Smith, D. L. DeWitt, J. Biol. Chem. 1993, 268, 6610 6614. [11] B. Cryer, Am. J. Gastroenterol. 2005, 100, 1694 1695. [12] M. F. Go, Gastrointest. Endosc. Clin. N. Am. 2006, 16, 83 97. [13] M. W. James, C. J. Hawkey, Br. J. Clin. Pharmacol. 2003, 56, 146 155. [14] M. Lazzaroni, G. B. Porro, Aliment. Pharmacol. Ther. 2004, 20, 48 58. [15] J. Nsdal, K. Brown, Drug Saf. 2006, 29, 119 132. [16] G. Mounier, C. Guy, F. Berthoux, M. N. Beyens, M. Ratrema, M. Ollagnier, Therapie 2006, 61, 255 266. [17] V. Schneider, L. E. Lvesque, B. Zhang, T. Hutchinson, J. M. Brophy, Am. J. Epidemiol. 2006, 164, 881 889. [18] J. Zadrazil, Vnitr. Lek. 2006, 52, 686 690. [19] D. Adebayo, I. Bjarnason, Postgrad. Med. J. 2006, 82, 186 191. [20] V. Chiroli, F. Benedini, E. Ongini, P. Del Soldato, Eur. J. Med. Chem. 2003, 38, 441 446. [21] E. Fosslien, Ann. Clin. Lab. Sci. 1998, 28, 67 81. [22] P. Prasit, Z. Wang, C. Brideau, C.-C. Chan, S. Charleson, W. Cromlish, D. Ethier, J. F. Evans, A. W. Ford-Hutchinson, J. Y. Gauthier, R. Gordon, J. Guay, M. Gresser, S. Kargman, B. Kennedy, Y. Leblanc, S. Leger, J. Mancini, G. P. ONeill, M. Ouellet, M. D. Percival, H. Perrier, D. Riendeau, I. Rodger, P. Tagari, M. Therien, P. Vikers, E. Wong, L.-J. Xu, R. N. Young, R. Zamboni, S. Boyce, N. Rupniak, M. Forrest, D. Visco, D. Patrick, Bioorg. Med. Chem. Lett. 1999, 9, 1773 1778. [23] T. D. Penning, J. J. Talley, S. R. Bertenshaw, J. S. Carter, P. W. Collins, S. Docter, M. J. Graneto, L. F. Lee, J. W. Malecha, J. M. Miyashiro, R. S. Rogers, D. J. Rogier, S. S. Yu, G. D. Anderson, E. G. Burton, J. N. Cogburn, S. A. Gregory, C. M. Koboldt, W. E. Perkins, K. Seibert, A. W. Veenhuizen, Y. Y. Zhang, P. C. Isakson, J. Med. Chem. 1997, 40, 1347 1365.
150
www.chemmedchem.org
ChemMedChem 2012, 7, 144 150
DOI: 10.1002/cmdc.201100441
Identification of 1,3-Diiminoisoindoline Carbohydrazides as Potential Antimalarial Candidates

Paolo Mombelli,[a] Matthias C. Witschel,*[b] Anthoni W. van Zijl,[a] Julie G. Geist,[a] Matthias Rottmann,[c, d] Cline Freymond,[c, d] Franz Rhl,[b] Marcel Kaiser,[c, d] Victoria Illarionova,[e] Markus Fischer,[e] Isabella Siepe,[b] W. Bernd Schweizer,[a] Reto Brun,[c, d] and Franois Diederich*[a]
A series of inhibitors of plant enzymes of the non-mevalonate pathway from herbicide research efforts at BASF were screened for antimalarial activity in a cell-based assay. A 1,3-diiminoisoindoline carbohydrazide was found to inhibit the growth of Plasmodium falciparum with an IC50 value < 100 nm. Synthesis of a variety of derivatives allowed an improvement of the initial antimalarial activity down to IC50 = 18 nm for the most potent compound, the establishment of a structureactivity relationship, and the evaluation of the cytotoxic profile of the diiminoisoindolines. Furthermore, interesting configurational and conformational aspects for this class of compounds were studied by computational and X-ray crystal structure analysis. Some of the compounds can act as tridentate ligands, forming 2:1 ligandiron(III) complexes, which also display antimalarial activity in the nanomolar IC50 range, paired with low cytotoxicity.
Introduction
Malaria is one of the most devastating infectious diseases worldwide, particularly affecting tropical and subtropical regions. With more than 40 % of the global population living under the threat of contracting the disease, Plasmodium falciparum malaria leads to nearly one million deaths and 250 million clinical cases annually.[1, 2] Over the last decade, the malaria burden was decreased by relying on both transmission control and highly efficient artemisinin-based combination therapies (ACTs).[3] With the recent emergence of delayed clinical response in P. falciparum-infected patients,[4] however, the need to identify novel lead structures with distinct mechanisms of action has become urgent. As an alternative to the targetbased approach, hits can be identified through high-throughput screening in whole-cell assays, followed by optimization of in vitro activity. The validity of this approach was proven,[5] as spiroindolones were recently discovered as a new class of potent antimalarials; the activity of the initial hit was optimized within a very short timeframe to a promising drug candidate, which is currently undergoing clinical trials.[6] Moreover, pharmaceutical companies and academic research groups have made their screening data against P. falciparum public, thus providing the antimalarial community with valuable information and new starting points for further research.[7] At BASF, the non-mevalonate pathway to isoprenoids[8] has been examined in detail in the course of herbicide research.[9] Most of the enzymes in this pathway have been proven to be essential in plants either through the action of herbicidal inhibitors such as ketoclomazone[10] and fosmidomycin,[11] or by antisense experiments.[12] As the enzymes of this pathway also seem to be essential in Plasmodium species,[11] all 461 inhibitors of these enzymes identified in the BASF herbicide projects were transferred to the Swiss Tropical and Public Health InstiChemMedChem 2012, 7, 151 158
tute for examination of their whole-cell in vitro activity toward P. falciparum. One of the most active hits in this assay is 1,3-diiminoisoindoline carbohydrazide 1 (Figure 1), which displays a promising half-maximal (50 %) inhibitory concentration (IC50) value of 58 nm. This compound was originally selectFigure 1. Structure of the lead compound 1 ed as a potential inhibitor of 4-diphosfound by screening of phocytidyl-2C-methyl-d-erythritol kinase a library of 461 com(IspE, EC 2.7.1.148), which was successpounds. fully addressed by our research group for the development of potent inhibitors before.[13] Compound 1 inhibits Lycopersicon esculentum IspE with an IC50 value of 76 mm, as determined by using the
[a] P. Mombelli, Dr. A. W. van Zijl, J. G. Geist, Dr. W. B. Schweizer, Prof. Dr. F. Diederich Laboratorium fr Organische Chemie ETH Zrich, Hnggerberg, HCI, 8093 Zrich (Switzerland) E-mail: diederich@org.chem.ethz.ch [b] Dr. M. C. Witschel, Dr. F. Rhl, Dr. I. Siepe BASF SE, GVA/HC-B009, 67056 Ludwigshafen (Germany) E-mail: matthias.witschel@basf.com [c] Dr. M. Rottmann, C. Freymond, M. Kaiser, Prof. Dr. R. Brun Swiss Tropical and Public Health Institute Socinstrasse 57, 4002 Basel (Switzerland) [d] Dr. M. Rottmann, C. Freymond, M. Kaiser, Prof. Dr. R. Brun University of Basel, Petersplatz 1, 4003 Basel (Switzerland) [e] Dr. V. Illarionova, Prof. Dr. M. Fischer Institute of Food Chemistry University of Hamburg, Grindelallee 117, 20146 Hamburg (Germany) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100441.
151
MED
available assay.[14] During the course of our optimization process, the P. falciparum IspE isoform became available; however, neither the lead compound 1 nor any of the prepared analogues showed activity toward P. falciparum IspE below 100 mm. Therefore, it seemed unlikely that the good P. falciparum activity observed in the cell-based assay was due to the inhibition of the IspE enzyme. Nevertheless, optimization of lead structure 1 was worthwhile and promising, due to its high antimalarial activity in the cell-based assay. The one-step procedure for the preparation of compound 1 was published recently in a study of the synthesis and structural properties of 1,3-diiminoisoindoline carbohydrazides.[15] However, very little is known about the biological activity of the diiminoisoindoline structural motif. Some in vivo activities in a carragenan-induced assay in mice[16] as well as potential applications as a C3a antagonist[17] have been reported for 3imino-1-oxoisoindolines and 1,3-disubstituted diiminoisoindolines, respectively. Encouraged by the results of the screening and the straightforward synthetic access that allows various structural modifications of the lead, a variety of derivatives were synthesized in order to optimize the in vitro antiplasmodial activity of the initial hit and to establish a structureactivity relationship (SAR) for this compound class.
M. C. Witschel, F. Diederich, et al. Scheme 1 for compound 4. Ester 26 was obtained from the corresponding carboxylic acid by following a published procedure,[18] and was subsequently converted into hydrazide 27 by treatment with hydrazine monohydrate. Reaction of the hydrazide with 1,3-diiminoisoindoline (28) in ethanol at reflux provided compound 4 in excellent yield and high purity by simple precipitation. The alternative route leading to compounds bearing a substituent on the isoindoline moiety relies on the cyclization of a substituted dinitrile with the desired hydrazide. As an example, the synthesis of compounds 9 and 17 is shown in Scheme 1. Aromatic nucleophilic substitution of 3-nitrophthalonitrile (29) with 4-fluorobenzenethiol gave 30 in good yield after recrystallization from ethanol. Treatment of dinitrile 30 with picolinyl hydrazide (31) in the presence of a catalytic amount of base in ethanol at reflux gave a mixture of 9 and 17. The major limitation in the preparation of the target compounds was the formation of two regioisomeric products in the reaction of asymmetrically substituted phthalonitriles, such as 30, with hydrazides. Because the isoindolines were obtained by precipitation from the reaction mixture, their isolated yields and isomeric ratios depended mainly on the solubility of the two isomers. However, in the reaction of 1,2,3-trisubstituted phthalonitriles, the C4-substituted isoindolines were obtained in excess, as expected based on both steric and electronic considerations. Repeated recrystallization of the mixture gave only moderate yields, but afforded enough regioisomerically pure material for biological testing, as shown in Scheme 1 for compounds 9 and 17.

Synthesis Two synthetic routes were followed to access isoindolines 2 25 (see Table 1 below for structures), depending on the desired substitution pattern on the pyridine ring or the isoindoline core. Both sequences consist of a maximum of three steps, with no chromatographic purification necessary in most cases. The synthetic route used for the compounds bearing an unsubstituted isoindoline moiety is shown as an example in
Tautomerism, configurational, and conformational analysis Diiminoisoindoline carbohydrazides are crystalline compounds with high melting temperatures (> 200 8C) and low solubility in common organic solvents and water. Although low water solubility can be problematic in the optimization of drug candi-
Scheme 1. Synthesis of compounds 4 (top), 9, and 17 (bottom): a) N2H4H2O, EtOH, reflux, 1 h, 43 %; b) EtOH, reflux, 1 h, 91 %; c) Na2CO3, 4-fluorobenzenethiol, DMF, 25 8C, 2 h, 64 %; d) NaOMe/MeOH, EtOH, reflux, 6 h, 9: 17 %, 17: 6 %.
152
www.chemmedchem.org
ChemMedChem 2012, 7, 151 158
Antimalarial 1,3-Diiminoisoindoline Carbohydrazides dates, this limitation could be partially overcome by forming the corresponding hydrochloride salts, which showed slightly improved water solubilities (0.11 mg mL1). This class of compounds also features interesting structural properties, such as tautomerism, diastereoisomerism around the exocyclic C=N bond, and E/Z amide-type isomerism.[15, 19] Quantum-mechanical calculations performed by Hordiyenko et al.[15] on the tautomeric forms of the acetic acid hydrazide derivative showed hydrazone A to be energetically preferred over B by 1.7 kcal mol1 (Scheme 2, R = CH3). In contrast, hydrazide C is clearly disfavored, being 15.6 kcal mol1 higher in energy than B. tomeric form A. All X-ray crystal structures confirm the conformation suggested by the calculation, with the molecule nearly planar and the hydrazide NH group engaged in intramolecular hydrogen bonds with the pyridine and the endocyclic isoindoline nitrogen atoms acting as hydrogen bond acceptors. In line with a strong preference for the discussed most stable conformer (1 a in figure 1SI) in solution as well, no isomerism was observed in the 1H NMR spectra (400 MHz, [D6]DMSO) at 25 8C.
Formation of iron(III) complexes Stable metal complexes have gathered increasing interest in the field of medicinal chemistry, in particular as tools to modulate and improve target specificity.[20] In their preferred tautomeric form and conformation, our compounds appear to be prone to act as tridentate ligands for the complexation of metal cations upon deprotonation. To investigate the biological profile of the metal complexes, we turned our attention to the ability of screening hit 1 to form a complex with iron(III). Indeed, by heating a mixture of ligand 1 and Fe(ClO4)3x H2O at a 2:1 ratio in ethanol in the presence of triethylamine, complex 32 was readily formed (Scheme 3). The structure of complex 32 was confirmed by X-ray crystallography, showing a slightly distorted octahedral geometry for FeIII, coordinated to the endocyclic nitrogen atom of the isoindoline moiety, the deprotonated hydrazide, and the pyridine nitrogen atom in each of the two ligands. Similarly to the aforementioned procedure, complexes 3339 were synthesized from the respective ligands (Table 2). The kinetic stability of the complexes was investigated in ligand-exchange experiments monitored by LCMS. For this purpose, a 1:1 mixture of complexes 32 and 35 was analyzed
Scheme 2. Tautomeric forms A, B, and C of 1,3-diiminoisoindoline carbohydrazides: DDG8 values refer to the acetic acid hydrazide derivative (R = CH3) as reported by Hordiyenko et al.[15]
We performed a conformational analysis of lead compound 1 in the tautomeric form A using density functional theory in the gas phase (1 bar, 298 K) at the B3LYP/6-311G(d,p) level of theory (see Supporting Information figure 1SI). The conformer with Z orientation around the C=N and NCO bonds, as depicted in Figure 1, was found to be by far the most energetically preferred. The results obtained by computation are supported by X-ray crystallography. Single crystals suitable for X-ray analysis were obtained for compounds 7 and 22 (Figure 2, and Supporting Information figures 2SI and 3SI), 5, 14, and 18 (Supporting Information figures 46SI), with all compounds in the same tau-
Figure 2. ORTEP plot of X-ray crystal structures of compounds 7 (top) and 22 (bottom) at 123 K. Atomic displacement parameters are shown at the 50 % probability level. Intramolecular hydrogen bonding distances in 7: d(NN(H)) = 2.69 and 2.68 ; in 22: d(NN(H)) = 2.66 and 2.64 .
Scheme 3. Synthesis of metal complex 32 (top) and ORTEP plot of X-ray crystal structure of 32 at 100 K (bottom): a) Fe(ClO4)3x H2O, Et3N, EtOH, reflux, 20 min, 87 %. Atomic displacement parameters are shown at the 50 % probability level; hydrogen atoms and ClO4 anion are omitted for clarity.
ChemMedChem 2012, 7, 151 158
www.chemmedchem.org
153
MED
at various temperatures over longer time periods. In both ethanol and in a phosphate-buffered solution (K2HPO4/KH2PO4, pH 7.4), only traces of ligand-exchange reactions were detected after stirring the mixture for 12 h at 40 8C, whereas the formation of the mixed [FeIIIAB]-type complex was observed after 12 h at 60 8C (see Supporting Information figures 7SI and 8SI). Similar results were obtained when the experiment was performed with a 1:2 mixture of complex 32 and ligand 8 (Supporting Information figures 9SI and 10SI), suggesting that the complexes are quite stable under physiological conditions.
M. C. Witschel, F. Diederich, et al. range as for 1, with IC50 values of 47, 61, and 51 nm, respectively (Table 1). Moving the methoxy group from pyridine C4 to C6 led to a decrease in activity by a factor of 100 (table 1SI, entry 25). Compounds bearing strong electron-withdrawing groups, such as nitrile or nitro groups, were consistently less active and inhibited P. falciparum growth in the single-digit micromolar range (table 1SI, entries 2628). This dependence of the activity on the electron donor/acceptor character of the pyridine substituent hints at the H-bond-accepting role of its nitrogen atom. It presumably undergoes intramolecular hydrogen bonding, as suggested by the structural data (Figure 2); however, a reorientation with intermolecular hydrogen bonding to a protein cannot be excluded at this stage. As for the pyridine core, the choice of substituents and the substitution pattern on the isoindoline moiety are essential. Because substituted isoindolines were obtained as mixtures of two regioisomers, they were first tested against P. falciparum as mixtures; then for the most potent compounds, the isomers were separated and tested independently. The introduction of various substituents at positions C5 or C6 of the isoindoline moiety led to a loss in potency by factors varying from three for a methyl group to 45 for a methoxyethoxy group (Supporting Information table 2SI, entries 616). Electron-donating substituents at positions C4 or C7, in contrast, were tolerated or even beneficial in terms of antimalarial activity. In particular, phenyl ether 6 and aryl thioethers 710 displayed activities in the same range as 1, with IC50 values between 25 and 50 nm. Similar results were obtained with alkyl ethers 11 and 12 or alkyl thioether 13 (Table 1). The values of the C7-substituted isoindolines 1417 were slightly higher than those determined for the corresponding isomers 69 substituted at position C4. Electron-withdrawing functional groups such as N,N-dialkylsulfonamido in compounds 18 and 19 (IC50 = 82 and 94 nm, respectively) did not improve the potency relative to the screening hit 1. Accordingly, compounds bearing nitro groups inhibited parasitic growth in the micromolar range (table 3SI, entries 19 and 20). Benzoisoindolines 22 and 23 showed IC50 values of 22 and 39 nm, respectively, whereas substitution of the pyridine at C5 with an n-butyl chain did not substantially affect the activity, as observed for 24, with an IC50 value of 51 nm. The introduction of a strong electron-donating group, such as N,Ndimethylamino at position C4, provided 25, which was revealed to be the most potent compound of the library (IC50 = 18 nm). Metal complex 32, with screening hit 1 as ligands, inhibited the growth of P. falciparum with an IC50 value of 76 nm (Table 2), in a similar range to that of the free ligand 1. Exchanging the counter-anion from perchlorate to chloride in complex 33 resulted in a minimal loss of activity (IC50 = 171 nm). In general, metal complexes 3238 were found to have similar potency to their respective free ligands, except for compound 39, for which the complexation was associated with a relevant decrease in antimalarial activity (IC50 = 454 nm).
Biological Results
In vitro antimalarial activity The synthesized compounds were evaluated for their in vitro antimalarial activity using erythrocytes infected by the chloroquine-sensitive P. falciparum strain NF54, with chloroquine and artesunate serving as positive controls. Initially, the importance of the pyridine and isoindoline moieties was investigated. Omitting the pyridine ring or replacing it with other aromatic, heteroaromatic (except for isoquinoline, see 20 in Table 1), or alicyclic residues led to a substantial loss in potency, with IC50 values > 1 mm (see Supporting Information table 1SI, entries 1 11). Furthermore, the position of the pyridine nitrogen atom was found to be crucial: 4-picolinyl and 3-picolinyl derivatives were dramatically less potent than 1, with IC50 values > 1 mm (table 1SI, entries 12 and 13). In addition, the inhibitors were also found to be very sensitive to changes at the isoindoline moiety, suggesting that the hydrogen bonding motif and a bicyclic aromatic system are essential for maintaining low-nanomolar activity (table 2SI, entries 15). Consequently, we focused on the synthesis of derivatives featuring both the isoindoline and pyridine moieties with different substituents. The short synthetic route allowed preparation of a larger library of compounds with activities in the lower nanomolar range, as observed for lead 1. However, the nature and position of the substituent was found to be crucial for maintaining activity. The introduction of functional groups on the pyridine ring was tolerated at positions C4 and C5. Small residues such as methyl, halogen atoms, or hydroxy groups did not greatly affect the activity, as IC50 values varied only from 54 to 363 nm (Supporting Information table 1SI, entries 1419). In contrast, the activity was lost if a methyl or a trifluoromethyl group was introduced at position C6 (IC50 values > 3 mm, table 1SI, entries 20 and 21). Isoquinoline derivative 20 was equipotent to 1 (Table 1), whereas the quinoline analogue inhibited parasite growth with an IC50 value of 2.7 mm (table 1SI, entry 22). The activities of the phenyl- and n-butyl-substituted pyridines 2 and 21 (IC50 : 91 and 78 nm, respectively, Table 1) are on the same order of magnitude as that of 1, suggesting that larger groups are tolerated on pyridine C4, and to a lesser extent on C5 as well. C6-Phenyl-substituted pyridine was not active at 10 mm (table 1SI, entry 23). The best results were obtained with compounds 3, 4, and 5, which bear electron-donating alkoxy substituents at position C4. Their activities are in the same
154
www.chemmedchem.org
ChemMedChem 2012, 7, 151 158
Antimalarial 1,3-Diiminoisoindoline Carbohydrazides
Table 1. Structures, in vitro activities against P. falciparum, cytotoxicities, and selectivities of compounds 125.[a]
Compd
IC50 [nm]
SI[c]
R1 1 2 3[d] 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 H Ph OMe OEt OiPr H H H H H H H Cl H H H H H H
R2 H H H H H OPh SPh S(p-Cl-C6H4) S(p-F-C6H4) S(o-F-C6H4) OCH2CF3 O(CH2)2OMe S(CH2)2NMe2 H H H H SO2NMe2 SO2NEt2
R3 H H H H H H H H H H H H H OPh SPh S(p-Cl-C6H4) S(p-F-C6H4) H H
P. fal. NF54 58 91 47 61 51 28 39 55 31 63 32 31 34 91 61 145 71 94 82
Cytotox. L-6[b] 1368 ND ND 847 699 22 7 7 6 11 120 588 10 1289 237 114 102 297 31 24 14 14 0.8 0.2 0.1 0.2 0.2 4 19 0.3 14 4 0.8 1.4 3 0.4
20
63
ND
21
78
ND
22
22
54
23
39
68
24
51
34
0.7
25
18
36
chloroquine artesunate podophyllotoxin
17.1 5.8
12.1
[a] All values represent the mean of at least two independent measurements, each performed in duplicate. [b] Rat skeletal myoblasts were used to assess cytotoxicity; ND: not determined. [c] Selectivity index calculated as the ratio IC50 (L-6)/IC50 (NF54). [d] Also tested as HCl salt: NF54 IC50 = 24 nm, L-6 IC50 = 597 nm, SI = 25.
ChemMedChem 2012, 7, 151 158
www.chemmedchem.org
155
MED
Table 2. Structures, in vitro activities against P. falciparum, cytotoxicities, and selectivities of FeIII complexes 3239.[a]
M. C. Witschel, F. Diederich, et al.
Compd
Ligand
Counterion
IC50 [nm]
SI[c]
R1 32 33 34 35 36 37 38 39 chloroquine artesunate podophyllotoxin 1 = HL1 1 = HL1 4 = HL2 8 = HL3 11 = HL4 18 = HL5 6 = HL6 21 = HL7 H H OEt H H H H H
R2 H H H H H H H nBu
R3 H H H S(p-Cl-C6H4) OCH2CF3 SO2NMe2 OPh H
X ClO4 Cl ClO4 ClO4 ClO4 ClO4 ClO4 ClO4
P. fal. NF54 76 171 105 69 34 95 70 454 17.1 5.8
Cytotox. L-6[b] 36 722 21 142 194 58 8 124 13 276 12.1 483 124 1.8 0.8 0.2 1.3 0.2 0.6
[a] All values represent the mean of at least two independent measurements, each performed in duplicate. [b] Rat skeletal myoblasts were used to assess cytotoxicity. [c] Selectivity index calculated as the ratio IC50 (L-6)/IC50 (NF54).
In vivo antimalarial activity In vivo studies were performed with four compounds among the most potent of the library. To ensure diversity, we selected derivatives featuring different substitution patterns. Therefore, the screening hit 1, 4-methoxypyridine 3, 4-substituted isoindoline 11, and benzoisoindoline 25 were converted into the corresponding hydrochloride salts and tested with mice infected by P. berghei. The inhibitors were administered to three mice orally and to three mice subcutaneously with 100 mg kg1 single daily doses on days 1, 2, and 3 after infection. On day 4, only mice that had orally received the hydrochloride salt of compound 25 showed a notable decrease in parasitemia relative to untreated control mice, with an activity of 46 %, but without any significant increase in the mean survival time (7 days). In another independent experiment, the hydrochloride of 25 (100 mg kg1) was administered orally on four consecutive days, showing a decrease in parasitemia of 60 %. The relatively low in vivo activity, considering the doubledigit nanomolar activity in the cell-based assay, may be due to low solubility and bioavailability issues of the isoindolines. Cytotoxicity In addition to investigating antimalarial activity, cytotoxicity assays were carried out with compounds 22 and 24 on human cervical cancer cells (HeLa B), histiocytic lymphoma cells (U937), and liver carcinoma cells (HepG2). No acute cytotoxicity against the human cells was observed at concentrations of 60 mm after 24 h. Independently, cell proliferation inhibition over 72 h was assessed for the most relevant compounds using rat skeletal myoblast (L-6) cells. Podophyllotoxin served as positive control in the assay. The selectivity index (SI) was determined as the ratio of the IC50 value against L-6 cells and the antiplasmodial IC50 value. Lead compound 1 showed an IC50 value of 1368 nm against L-6 cells, which corresponds to an SI of 24 (Table 1). Remarkably, the corresponding iron(III) complexes 32 and 33 showed SIs > 100 (Table 2), which could be explained by a different target specificity, but also by physicochemical properties of the complexes which differ from those of the free ligands. While the introduction of substituents on the pyridine moiety did not dramatically affect cell proliferation, modification of the isoindoline core was found to decrease the selectivity for P. falciparum considerably. In particular, the introduction of additional phenyl rings, either annulated to the isoindoline as in the case of benzoisoindolines 2225, or as phenyl ether or thiophenyl ether in 610, 16, and 17, lowered the selectivity by factors varying from 10 to 200. Alkyl-ether-substituted isoindolines 11 and 12 displayed IC50 values against L-6 cells in the triple-digit nanomolar range. The SIs for the C7-substituted isoindolines 1417 were found to be slightly higher than those determined for the corresponding C4 isomers 69. To further assess the cytotoxic/proliferation inhibition potential of these compounds, iron(III) complex 32, which displayed the highest SI in the rat L-6 cell proliferation assay of the tested compounds, was also tested on HepG2 cells. After 24 h, no inhibition of proliferation was observed at 50 mm, but after 72 h proliferation inhibition with an IC50 value of 0.75 mm was observed (mean of two independent measurements). The cell division inhibitor podophyllotoxin, used as comparison, showed very low activity (IC50 > 50 mm) in this assay after 24 h, but strong inhibition (IC50 < 1 mm) after 72 h. Although high doses (3 100 mg kg1) of 25 (SI = 2 in the L6 assay) were administered both orally and subcutaneously to the mice in the in vivo studies, no acute toxicity was observed in these experiments.
ChemMedChem 2012, 7, 151 158
156
www.chemmedchem.org
Antimalarial 1,3-Diiminoisoindoline Carbohydrazides Activity on other organisms The screening hit 1 also showed in vitro activity against Trypanosoma brucei rhodesiense (IC50 = 0.48 mm), T. cruzi (IC50 = 0.23 mm), and Leishmania donovani (IC50 = 1.36 mm). Additionally, isoindolines were subjected to a herbicidal assay performed on duckweed (Lemna minor, Supporting Information table 4SI). Several compounds showed interesting activities, with some correlation to the activities observed in the P. falciparum assay. This might indicate a related mode of action for this organism.
solution was treated with hydrazide 27 (72 mg, 0.40 mmol), heated at reflux for 1 h, and cooled to 25 8C. The resulting precipitate was filtered off, washed with Et2O, and dried in vacuo. Compound 4 (121 mg, 91 %) was obtained as a light-yellow solid; mp: 271 272 8C; 1H NMR (300 MHz, [D6]DMSO): d = 1.38 (t, J = 6.9 Hz, 3 H), 4.24 (q, J = 6.9 Hz, 2 H), 7.19 (dd, J = 5.7, 2.6 Hz, 1 H), 7.537.60 (m, 2 H), 7.63 (d, J = 2.1 Hz, 1 H), 7.797.82 (m, 1 H), 7.927.95 (m, 1 H), 8.48 (d, J = 5.7 Hz, 1 H), 8.60 (br s, 1 H), 8.81 (br s, 1 H), 11.81 ppm (s, 1 H); 13C NMR (100 MHz, [D6]DMSO): d = 14.3, 64.1, 108.3, 113.3, 120.7, 121.6, 129.6, 131.1, 134.9, 138.1, 150.1, 151.4, 158.5, 158.8, 166.0, 167.9 ppm; IR (ATR): ~ = 3436, 3296, 2959, 2930, 2859, 1667, n 1644, 1591, 1565, 1503, 1469, 1428, 1393, 1360, 1335, 1295, 1240, 1136, 1048, 965, 927, 852, 820 cm1; HRMS-ESI: calcd for C16H16N5O2 + [M + H] + : 310.1299, found: 310.1304 (100 %). N-{(1Z)-3-Amino-4-[(4-fluorophenyl)sulfanyl]-1H-isoindol-1-ylidene}pyridine-2-carbohydrazide (9) and N-{(1Z)-3-Amino-7-[(4-fluorophenyl)sulfanyl]-1H-isoindol-1-ylidene}pyridine-2-carbohydrazide (17): A suspension of hydrazide 31 (135 mg, 0.98 mmol) and phthalonitrile 30 (250 mg, 0.98 mmol) in EtOH (8.3 mL) was heated until complete dissolution of the substrates. The mixture was cooled to 25 8C, treated with NaOMe (0.5 m in MeOH, 0.20 mL, 0.10 mmol), and heated at reflux for 6 h. After cooling to 25 8C, the resulting precipitate was filtered off and washed with Et2O. Recrystallization in EtOH gave 17 (22 mg, 6 %) as a pale-green solid. Crystallization upon standing from the reaction solution gave 9 (67 mg, 17 %) as a dark-red crystalline solid. Compound 9: mp: 236238 8C; 1 H NMR (400 MHz, [D6]DMSO): d = 7.247.30 (m, 2 H), 7.39 (dd, J = 7.8, 0.9 Hz, 1 H), 7.457.50 (m, 2 H), 7.59 (t, J = 7.6 Hz, 1 H), 7.657.75 (br s, 1 H), 7.68 (ddd, J = 7.6, 4.8, 1.3 Hz, 1 H), 7.82 (dd, J = 7.5, 0.9 Hz, 1 H), 8.08 (td, J = 7.7, 1.7 Hz, 1 H), 8.17 (dt, J = 7.6, 1.1 Hz, 1 H), 8.70 (ddd, J = 4.8, 1.7, 0.9 Hz, 1 H), 9.21 (br s, 1 H), 11.80 ppm (s, 1 H); 13C NMR (100 MHz, [D6]DMSO): d = 116.9 (d, 2J(C,F) = 22.5 Hz), 120.6, 122.3, 127.1, 129.0 (d, 4J(C,F) = 3.3 Hz), 129.8, 132.3, 133.3, 133.4 (d, 3J(C,F) = 8.6 Hz), 134.4, 138.3, 139.9, 148.6, 149.2, 156.9, 158.6, 161.9 (d, 1J(C,F) = 246.0 Hz), 166.9 ppm; 19F NMR (376 MHz, n [D6]DMSO): d = 113.5 113.4 ppm (m, 1 F); IR (ATR): ~ = 3427, 3291, 3155, 1673, 1648, 1587, 1504, 1485, 1463, 1433, 1397, 1327, 1278, 1222, 1152, 1073, 1009, 997, 955, 834, 815 cm1; HRMSMALDI: calcd for C20H14FN5NaOS + [M + Na] + : 414.0795, found: 414.0795 (17 %), calcd for C20H15FN5OS + [M + H] + : 392.0976, found: 392.0973 (100 %). Compound 17: mp: 297299 8C; 1H NMR (400 MHz, [D6]DMSO): d = 6.696.71 (m, 1 H), 7.347.42 (m, 3 H), 7.667.73 (m, 4 H), 8.08 (td, J = 7.7, 1.7 Hz, 1 H), 8.18 (dt, J = 7.7, 1.0 Hz, 1 H), 8.55 (br s, 1 H), 8.71 (ddd, J = 4.8, 1.6, 0.9 Hz, 1 H), 8.76 (br s, 1 H), 11.90 ppm (s, 1 H); 13C NMR (100 MHz, [D6]DMSO): d = 117.3 (d, 2J(C,F) = 22.0 Hz), 118.0, 122.2, 125.7 (d, 4J(C,F) = 3.2 Hz), 127.0 ( 2), 129.8, 132.0, 134.8, 135.7, 138.1 (d, 3J(C,F) = 8.6 Hz), 138.2, 148.6, 149.4, 158.3, 158.9, 163.0 (d, 1J(C,F) = 247.9 Hz), 166.9 ppm; 19 F NMR (376 MHz, [D6]DMSO): d = 111.4 113.3 ppm (m, 1 F); IR (ATR): ~ = 3409, 3311, 3160, 1651, 1604, 1587, 1533, 1504, 1488, n 1467, 1434, 1412, 1335, 1247, 1216, 1137, 1121, 1089, 997, 944, 836, 801 cm1; HRMS-MALDI: calcd for C20H14FN5NaOS + [M + Na] + : 414.0795, found: 414.0792 (100 %), calcd for C20H15FN5OS + [M + H] + : 392.0976, found: 392.0976 (22 %). 4-Ethoxypyridine-2-carbohydrazide (27): A solution of ester 26[18] (1.95 g, 10.0 mmol) and N2H4H2O (1.0 mL, 20.0 mmol) in EtOH (2.5 mL) was heated at reflux for 50 min, then cooled to 5 8C. The resulting precipitate was isolated by filtration, re-dissolved in hot EtOH together with activated carbon, and filtered again. The filtrate was cooled to 15 8C, and 27 (190 mg, 11 %) was isolated as an off-white solid by filtration. The filtrate was concentrated under reduced pressure. Recrystallization of the residue in H2O/EtOH
Conclusions
The screening of a rather small set of compounds for antimalarial activity led to the discovery of 1,3-diiminoisoindoline 1 as a potent inhibitor of P. falciparum proliferation in red blood cells. The synthesis of various derivatives of 1 revealed that both the pyridine and the isoindoline moieties are essential to maintain the antimalarial activity in the low-nanomolar range. However, different substitution patterns on the pyridine and isoindoline are tolerated, and in some cases even beneficial, allowing the preparation of a large number of compounds that inhibit the growth of P. falciparum with IC50 values well below 100 nm. Moreover, a small series of iron(III) complexes was prepared using lead compound 1 and its derivatives as tridentate ligands. They are kinetically stable under physiological conditions and showed remarkable activity against P. falciparum in the nanomolar range. Some of the active compounds in the P. falciparum assay also displayed strong growth inhibition in the L-6 cell assay, indicating that these effects might be related. This should be taken into account for further optimization studies of the isoindoline compound class.
Supporting Information
See the Supporting Information for general details and procedures, cytotoxicity experiments, and in vivo antimalarial activity.
In vitro antimalarial activity

In vitro antimalarial activity was measured by using the [3H]hypoxanthine incorporation assay with the P. falciparum strain NF54. Results are expressed as the concentration effecting 50 % inhibition (IC50). See the Supporting Information for further details.
Chemistry
In the following, the experimental details for the syntheses of all compounds shown in Schemes 1 and 3 are described. Compounds 1[15] and 26[18] were prepared according to published procedures; all other synthetic details and experimental data are given in the Supporting Information. N-[(1Z)-3-Amino-1H-isoindol-1-ylidene]-4-ethoxypyridine-2-carbohydrazide (4): 1-Imino-1H-isoindol-3-amine (28) (73 mg, 0.50 mmol) was dissolved in EtOH (2.0 mL) by gentle heating. The
ChemMedChem 2012, 7, 151 158
www.chemmedchem.org
157
MED
gave additional 27 (585 mg, 32 %) as a yellow crystalline solid; mp: 119122 8C; 1H NMR (300 MHz, [D6]DMSO): d = 1.36 (t, J = 7.0 Hz, 3 H), 4.19 (q, J = 7.0 Hz, 2 H), 4.55 (s, 2 H), 7.10 (dd, J = 5.7, 2.7 Hz, 1 H), 7.46 (d, J = 2.6 Hz, 1 H), 8.40 (d, J = 5.6 Hz, 1 H), 9.80 ppm (s, 1 H); 13C NMR (100 MHz, [D6]DMSO): d = 14.7, 64.3, 108.3, 112.9, 150.4, 152.2, 162.9, 166.0 ppm; IR (ATR): ~ = 3673, 3277, 2988, 2901, n 2346, 1671, 1598, 1567, 1506, 1395, 1309, 1245, 1160, 1108, 1066, 1040, 994, 982, 929, 983, 882, 843, 815 cm1; HRMS-EI: calcd for C8H11N3O2 + [M] + : 181.0846, found: 181.0846 (100 %). 3-[(4-Fluorophenyl)sulfanyl]phthalonitrile (30): A suspension of 3nitrophthalonitrile (29) (692 mg, 4.0 mmol) and Na2CO3 (1.27 g, 12.0 mmol) in DMF (20 mL) was treated with 4-fluorobenzenethiol (0.43 mL, 4.0 mmol) at 25 8C. The mixture was stirred at 25 8C for 5 h and poured into a solution of half-satd aq NaCl (40 mL). The resulting precipitate was filtered off, washed with H2O, and dried over CaCl2. Recrystallization in EtOH gave 30 (655 mg, 64 %) as colorless needles; mp: 143 8C; 1H NMR (300 MHz, CDCl3): d = 7.11 (dd, J = 8.1, 1.3 Hz, 1 H), 7.14 (d, J = 4.8 Hz, 2 H), 7.47 (t, J = 8.1 Hz, 1 H), 7.527.60 ppm (m, 3 H); 13C NMR (100 MHz, CDCl3): d = 113.4, 113.5, 115.1, 117.3, 117.7 (d, 2J(C,F) = 22.5 Hz), 124.2 (d, 4J(C,F) = 3.6 Hz), 130.1, 131.3, 132.7, 137.5 (d, 3J(C,F) = 8.8 Hz), 146.8, 164.0 ppm (d, 1 J(C,F) = 252.8 Hz); 19F NMR (282 MHz, CDCl3): d = 108.5 ppm (tt, 3 J(H,F) = 8.4, 4J(H,F) = 5.3 Hz, 1 F); IR (ATR): ~ = 3145, 3078, 2988, 2233, n 1914, 1846, 1783, 1725, 1589, 1572, 1492, 1449, 1425, 1403, 1298, 1280, 1219, 1195, 1182, 1162, 1149, 1096, 1074, 1045, 1016, 994, 927, 839, 800 cm1; HRMS-EI: calcd for C14H7FN2S [M] + : 254.0309, found: 254.0306 (100 %); elemental analysis calcd (%) for C14H7FN2S (254.3): C 66.13, H 2.77, N 11.02; found: C 66.09, H 2.81, N 10.91. [FeL12]ClO4 (32): A suspension of 1 (178 mg, 0.67 mmol) and Et3N (93 mL, 0.67 mmol) in EtOH (5 mL) was treated with a solution of Fe(ClO4)3x H2O (159 mg, 0.34 mmol) in EtOH (1.5 mL). The mixture was heated at reflux for 20 min and cooled to 25 8C. The resulting precipitate was filtered off, washed with EtOH and Et2O, and dried in vacuo. Complex 32 (202 mg, 87 %) was obtained as a darkbrown solid; IR (ATR): ~ = 3326, 3112, 2981, 1641, 1597, 1558, 1474, n 1424, 1325, 1285, 1237, 1180, 1099, 1045, 1018, 957, 845, 809 cm1; HRMS-MALDI: calcd for C28H20FeN10O2 + [MClO4] + : 584.1115, found: 584.1114 (100 %), calcd for C14H12N5O + [L1 + 2 H] + : 266.1036, found: 266.1039 (53 %); X-ray: see Figure 2.
M. C. Witschel, F. Diederich, et al.

[2] R. W. Snow, C. A. Guerra, A. M. Noor, H. Y. Myint, S. I. Hay, Nature 2005, 434, 214 217. [3] R. T. Eastman, D. A. Fidock, Nat. Rev. Microbiol. 2009, 7, 864 874. [4] A. M. Dondorp, F. Nosten, P. Yi, D. Das, A. P. Phyo, J. Tarning, K. M. Lwin, F. Ariey, W. Hanpithakpong, S. J. Lee, P. Ringwald, K. Silamut, M. Imwong, K. Chotivanich, P. Lim, T. Herdman, S. S. An, S. Yeung, P. Singhasivanon, N. P. J. Day, N. Lindegardh, D. Socheat, N. J. White, N. Engl. J. Med. 2009, 361, 455 467. [5] a) T. N. C. Wells, Science 2010, 329, 1153 1154; b) M. Schlitzer, R. Ortmann, ChemMedChem 2010, 5, 1837 1840. [6] M. Rottmann, C. McNamara, B. K. S. Yeung, M. C. S. Lee, B. Zou, B. Russell, P. Seitz, D. M. Plouffe, N. V. Dharia, J. Tan, S. B. Cohen, K. R. Spencer, G. E. Gonzlez-Pez, S. B. Lakshminarayana, A. Goh, R. Suwanarusk, T. Jegla, E. K. Schmitt, H.-P. Beck, R. Brun, F. Nosten, L. Renia, V. Dartois, T. H. Keller, D. A. Fidock, E. A. Winzeler, T. T. Diagana, Science 2010, 329, 1175 1180. [7] a) D. Plouffe, A. Brinker, C. McNamara, K. Henson, N. Kato, K. Kuhen, A. Nagle, F. Adrin, J. T. Matzen, P. Anderson, T.-g. Nam, N. S. Gray, A. Chatterjee, J. Janes, S. F. Yan, R. Trager, J. S. Caldwell, P. G. Schultz, Y. Zhou, E. A. Winzeler, Proc. Natl. Acad. Sci. USA 2008, 105, 9059 9064; b) F.-J. Gamo, L. M. Sanz, J. Vidal, C. de Cozar, E. Alvarez, J.-L. Lavandera, D. E. Vanderwall, D. V. S. Green, V. Kumar, S. Hasan, J. R. Brown, C. E. Peishoff, L. R. Cardon, J. F. Garcia-Bustos, Nature 2010, 465, 305 310; c) W. A. Guiguemde, A. A. Shelat, D. Bouck, S. Duffy, G. J. Crowther, P. H. Davis, D. C. Smithson, M. Connelly, J. Clark, F. Zhu, M. B. Jimnez-Daz, M. S. Martinez, E. B. Wilson, A. K. Tripathi, J. Gut, E. R. Sharlow, I. Bathurst, F. El Mazouni, J. W. Fowble, I. Forquer, P. L. McGinley, S. Castro, I. Angulo-Barturen, S. Ferrer, P. J. Rosenthal, J. L. DeRisi, D. J. Sullivan, Jr., J. S. Lazo, D. S. Roos, M. K. Riscoe, M. A. Phillips, P. K. Rathod, W. C. Van Voorhis, V. M. Avery, R. K. Guy, Nature 2010, 465, 311 315. [8] a) M. Rohmer, M. Knani, P. Simonin, B. Sutter, H. Sahm, Biochem. J. 1993, 295, 517 524; b) M. K. Schwarz, PhD Thesis, ETH Zrich, No. 10951, 1994; c) S. T. J. Broers, PhD Thesis, ETH Zrich, No. 10978, 1994; d) W. Eisenreich, B. Menhard, P. J. Hylands, M. H. Zenk, A. Bacher, Proc. Natl. Acad. Sci. USA 1996, 93, 6431 6436; e) W. Eisenreich, M. Schwarz, A. Cartayrade, D. Arigoni, M. H. Zenk, A. Bacher, Chem. Biol. 1998, 5, R221 R233. [9] M. C. Witschel, H. W. Hffken, M. Seet, L. Parra, T. Mietzner, F. Thater, R. Niggeweg, F. Rhl, B. Illarionov, F. Rohdich, J. Kaiser, M. Fischer, A. Bacher, F. Diederich, Angew. Chem. 2011, 123, 8077 8081; Angew. Chem. Int. Ed. 2011, 50, 7931 7935. [10] J. Zeidler, J. Schwender, C. Mueller, H. K. Lichtenthaler, Biochem. Soc. Trans. 2000, 28, 796 798. [11] H. Jomaa, J. Wiesner, S. Sanderbrand, B. Altincicek, C. Weidemeyer, M. Hintz, I. Trbachova, M. Eberl, J. Zeidler, H. K. Lichtenthaler, D. Soldati, E. Beck, Science 1999, 285, 1573 1576. [12] a) J. M. Estvez, A. Cantero, A. Reindl, S. Reichler, P. Len, J. Biol. Chem. 2001, 276, 22 901 22 909; b) C. S. Ahn, H.-S. Pai, Plant Mol. Biol. 2008, 66, 503 517; c) M.-H. Hsieh, H. M. Goodman, Planta 2006, 223, 779 784. [13] A. K. H. Hirsch, M. S. Alphey, S. Lauw, M. Seet, L. Barandun, W. Eisenreich, F. Rohdich, W. N. Hunter, A. Bacher, F. Diederich, Org. Biomol. Chem. 2008, 6, 2719 2730. [14] F. Rohdich, J. Wungsintaweekul, H. Lttgen, M. Fischer, W. Eisenreich, C. A. Schuhr, M. Fellermeier, N. Schramek, M. H. Zenk, A. Bacher, Proc. Natl. Acad. Sci. USA 2000, 97, 8251 8256. [15] O. V. Hordiyenko, A. V. Biitseva, M. Y. Kornilov, N. Brosse, A. Hocquet, B. Jamart-Grgoire, O. V. Shishkin, R. I. Zubatyuk, Eur. J. Org. Chem. 2006, 2833 2842. [16] L. Butner, Y. Huang, E. Tse, I. H. Hall, Biomed. Pharmacother. 1996, 50, 290 296. [17] E. B. Grant, D. Guiadeen, M. Singer, D. Argentieri, D. J. Hlasta, M. Wachter, Bioorg. Med. Chem. Lett. 2001, 11, 2817 2820. [18] R. J. Sundberg, S. Jiang, Org. Prep. Proced. Int. 1997, 29, 117 122. [19] E. Licandro, D. Perdicchia, Eur. J. Org. Chem. 2004, 665 675. [20] E. Meggers, Angew. Chem. 2011, 123, 2490 2497; Angew. Chem. Int. Ed. 2011, 50, 2442 2448.
Acknowledgements
This research was supported by the Swiss National Science Foundation and BASF SE. A.W.v.Z. is grateful to the Rubicon program of the Netherlands Organization for Scientific Research (NWO) for a postdoctoral fellowship and to the Novartis Jubilee Foundation for financial support; J.G.G. thanks Novartis for a graduate student fellowship. We acknowledge Carola Gunkel for performing biological assays, as well as Frank Stelzer, Jan Meyer, and Michaela Mller (all BASF) for the synthesis of building blocks. We also thank Pablo Rivera-Fuentes for computational studies, Laura M. Salonen, and Veronika Ehmke (all ETH) for proofreading the manuscript. Keywords: antiprotozoal agents herbicides isoindolines malaria biological activity
[1] World Health Organization (WHO), World Malaria Report 2010, Geneva (Switzerland), http://www.who.int/malaria/world_malaria_report_2010 (accessed November 1, 2011).
Received: September 21, 2011 Revised: October 25, 2011 Published online on November 16, 2011
158
www.chemmedchem.org
ChemMedChem 2012, 7, 151 158
DOI: 10.1002/cmdc.201100408
Development of Improved PPARb/d Inhibitors

Philipp M. Toth,[a] Simone Naruhn,[b] Veronika F. S. Pape,[a] Stefanie M. A. Drr,[a] Gerhard Klebe,[a] Rolf Mller,[b] and Wibke E. Diederich*[a]
GSK0660 (1) is the first peroxisome proliferator-activated receptor (PPAR) b/d-selective inhibitory ligand described in the literature. Based on its structure, we designed and synthesized a series of modified compounds to establish preliminary structureactivity relationships. Most beneficial for increased binding affinity towards the PPARb/d ligand binding domain was the replacement of the 4-aminophenyl substituent by medium-length n-alkyl chains, such as n-butyl or iso-pentyl. These compounds show activity down to the one-digit nanomolar range, thus possessing up to a tenfold higher binding affinity compared with GSK0660. Additionally, the subtype-specific inhibition of PPARb/d was confirmed in a cell-based assay making these compounds invaluable tools for the further exploration of the functions of PPARb/d.
Introduction
Nuclear receptors are transcription factors that are regulated by small-molecule ligands, such as steroid hormones, lipophilic vitamins, or fatty acid derivatives. Upon ligand binding, nuclear receptors modulate the transcription of their respective target genes. Their functions are therefore associated with vital biological processes, such as cell differentiation, proliferation, immune regulation, and metabolism. Accordingly, they represent highly relevant pharmacological targets for various diseases, including hyperlipidemia, diabetes and inflammatory disorders.[1] The peroxisome proliferator-activated receptors (PPARs) are members of the nuclear receptor superfamily and can be divided into the three subtypes: PPARa (NR1C1), PPARb/d (NR1C2), and PPARg (NR1C3).[2] PPARa is predominantly found in brown adipose tissue, liver, heart, kidney, and intestine, all of which are organs involved in the catabolism of fatty acids.[35] Thus, activation of PPARa leads to increased fatty acid catabolism, gluconeogenesis, and ketone body synthesis.[3, 6] It is the target of the fibrate class of hypolipidemic drugs that lead to lower plasma triglyceride levels,[7] higher plasma levels of apolipoprotein A-II,[8] and lower plasma concentrations of inflammation-associated substances, such as fibrinogen, interleukin-6, and C-reactive protein (CRP).[9] PPARg is mainly expressed in adipocytes, but is also found in other cell types including epithelial cells of the intestine and specific kinds of immune and inflammatory cells.[3, 4, 10] It is the target of the thiazolidinedione drugs used to treat type II diabetes, such as rosiglitazone and pioglitazone, which act as insulin sensitizers.[11] However, their clinical use is under debate due to possible serious side effects,[12] which have already led to the withdrawal of rosiglitazone-containing drugs in the European Union.[13] PPARb/d has the broadest expression pattern of all three PPAR subtypes.[3, 4] It is involved in a large number of biological processes, including lipid and glucose metabolism,[14] cell differentiation,[15, 16] proliferation,[17, 18] apoptosis,[19] and immune
ChemMedChem 2012, 7, 159 170
regulation.[20] It has also been associated with related pathophysiological processes, such as inflammation, obesity, dyslipidemia, diabetes, cancer, and cardiovascular diseases. Therefore, PPARb/d has gained tremendous significance and is currently believed to be a highly relevant drug target for the treatment of various human diseases.[21, 22] All three PPAR subtypes form heterodimers with the retinoid X receptors (RXRs) of the NR2B subfamily.[23] These heterodimers bind to peroxisome proliferator response elements (PPREs) in their target genes.[3] Binding of an agonistic PPAR ligand causes a conformational change in the protein resulting in the dissociation of attached corepressors and/or the binding of coactivators, which finally leads to increased transcription.[24, 25] The agonist-induced structural changes are well understood, and a number of crystal structures exist.[25] Several high-affinity, subtype-specific PPARb/d agonists have been developed, and their effect on different biological processes has been intensively investigated.[26, 27] These studies suggest that such agonistic ligands have the potential for further development, but definitive conclusions, including safety issues, are difficult to draw in view of the incomplete knowledge regarding the biological functions of PPARb/d. As an example, apparently conflicting data exist regarding the role of PPARb/d in cell proliferation, tumorigenesis and inflammatory processes.[16, 17, 21, 28] It is thus possible that the inhibition of
[a] P. M. Toth, V. F. S. Pape, S. M. A. Drr, Prof. Dr. G. Klebe, Prof. Dr. W. E. Diederich Institut fr Pharmazeutische Chemie Philipps-Universitt Marburg Marbacher Weg 6, 35032 Marburg (Germany) E-mail: wibke.diederich@staff.uni-marburg.de [b] S. Naruhn, Prof. Dr. R. Mller Institute of Molecular Biology and Tumor Research Philipps-Universitt Marburg Emil-Mannkopff-Str. 2, 35032 Marburg (Germany) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cmdc.201100408.
159
MED
PPARb/d rather than its activation may be beneficial under certain pathophysiological conditions. To address these issues, the development of subtype-specific, high-affinity PPARb/d inhibitors is of utmost importance, since none of the few compounds described to date fulfill these criteria. SR13904[29] and GSK3787[30] are not specific for PPARb/d, while GSK0660 (1),[31] the first published inhibitor, is subtype-specific but requires the application of relatively high concentrations, thus increasing the risk of off-target effects. Here, we describe the first systematic variation of GSK0660 (1) as a lead structure, in order to improve the binding affinity towards human PPARb/d and to gain a first insight into structureactivity relationships (SAR).
W. E. Diederich et al. of GSK0660 (1): the methoxy-substituted 1,4-diaminobenzene moiety. By varying the position of the methoxy substituent or by replacing it with a hydrogen atom (entries 13, Table 1), we sought to estimate its contribution to the binding affinity and thus determine its structural importance.
Table 1. Compounds used for the elucidation of the initial structureactivity relationships.
Entry
Compd GSK0660 (1) 2 3 4 5 6 7
R1 H OMe H H H OMe H
R2 OMe H H H H H H
X NH NH NH O CH2 NH NH
Y S S S S S CH=CH CH=CH

Initial structureactivity relationships At first, we conducted an initial SAR study by systematically varying the structural features of GSK0660 (1)[31] to assess the contribution of the various residues to the binding affinity (Figure 1). We first focused our attention on the central portion
1 2 3 4 5 6 7
Figure 1. Structural features of compound 1 feasible for variation.
In the competitive time-resolved fluorescence resonance energy transfer (TR-FRET) assay (Figure 2 b), a significant decrease in affinity to the human PPARb/d ligand binding domain (LBD) can be seen when the methoxy group is moved from the 2- (1) to the 3-position (2) or when it is replaced by a hydrogen atom (3), suggesting that the optimal position for the methoxy substituent is the 2-position, the same substitution pattern as present in the parent compound. In the following step, we investigated the importance of the bridging atom between the terminal phenyl group and the phenyl ring of the central moiety in order to probe the exis-
Figure 2. Competitive in vitro PPAR ligand binding assays. Results of the Lanthascreen TR-FRET PPARb/d competitive ligand binding assays at a ligand concentration of 100 nm. Relative FRET ratios were normalized for GSK0660 (1) to 100 %. All data points represent averages of triplicates ( S.D). Statistically significant differences to the GSK0660-treated sample are denoted: *** P < 0.001, ** P < 0.01, * P < 0.05, as determined by a t-test.
160
www.chemmedchem.org
ChemMedChem 2012, 7, 159 170
Improved PPARb/d Inhibitors tence of potential hydrogen bonds or other polar interactions at this position. Therefore, we replaced the bridging NH group of the parent compound 1 by either an oxygen atom or a methylene group. Due to their commercial availability, we employed the corresponding aromatic amines lacking the methoxy group in the central moiety for this investigation, thus leading to compounds 4 and 5 (entries 4 and 5, Table 1). Again, the competitive TR-FRET assay showed that replacement of the bridging atom and the 2-methoxy group significantly reduces the affinity of the compound towards the PPARb/d LBD in comparison to parent compound 1. At a concentration of 100 nm, compounds 3, 4, and 5 are nearly equally potent (Figure 2 a). Furthermore, we substituted the methyl thiophene carboxylate with the bioisosteric methyl benzoate (entries 6 and 7, Table 1). As we encountered difficulties concerning the nucleophilic aromatic substitution (SNAr) employing aniline as the nucleophile, in which primarily the methoxy substituent of the 4fluoro-2-methoxy-nitrobenzene moiety was replaced instead of the envisioned fluorine atom, we limited the compounds tested to those bearing either a methoxy group (6) or a hydrogen atom (7) in the 3-position. Switching from a thiophene to a phenyl ring leads to a considerable drop in affinity when comparing compound 2 with 6. In case of the related compounds 3 and 7, the affinity difference is not significantly pronounced, however, all of these compounds exhibit a lower binding affinity for PPARb/d than GSK0660 (Figure 2 b).
Scheme 1. General synthetic scheme. Reagents and conditions: a) R1R2NH, MeCN, mW irradiation, 1 h, 120 8C, 2598 %; b) Pd/C, H2, EtOAc, 24 h, RT, dark, no isolation; c) DMAP, pyridine, methyl 3-(chlorosulfonyl)thiophene-2-carboxylate, 48 h, RT, dark, 3488 % (over two steps).
Variation of the amino substituent As the final step of our initial SAR study, we substituted the phenyl moiety attached to the bridging NH group. As natural PPAR ligands are fatty acids and derivatives thereof, we chose medium-length n-alkyl chains as a starting point to create a small series of homologous n-alkyl GSK0660 derivatives (Table 2).
Table 2. Linear alkyl chain GSK0660 (1) derivatives 10 ad.
Entry 1 2 3 4
Compd 10 a 10 b 10 c 10 d
R1 n-propyl n-butyl n-pentyl n-hexyl
Structure
ed in a SNAr reaction using the corresponding alkylamines under microwave irradiation. This led to an impressive shortening of the reaction time from about 48 h to roughly 1 h compared to conventional heating. However, in the case of 9 e, although the reaction was stirred for over 8 h under microwave irradiation instead of 1 h, complete conversion of the starting material could not be achieved. Overall, the resulting 4-nitroaniline derivatives (9 ao) were obtained in isolated yields ranging from 25 % (9 e) to 98 % (9 n), providing an average yield of 73 % for the first synthetic step. Although our synthetic approach of introducing the amino substituent in the first reaction step seemed less appealing, all our attempts to establish a sequence allowing the desired variation in the last step (e.g., BuchwaldHartwig amination, etc.) were unsuccessful (data not shown). Subsequently, these 4-nitroaniline derivatives (9 ao) were reduced to the corresponding 1,4-diamines via hydrogenation employing palladium on charcoal as the catalyst and ethyl acetate as the solvent. Apart from that, the N-benzyl-substituted derivative (9 l) was reduced with tin(II)chloride as we expected the N-benzyl group to be cleaved under the reaction conditions mentioned above. Showing structural instability, in particular if exposed to light, the 1,4-diamines were directly used in the next reaction step without further purification. Conversion to the final sulfonamides (10 ap) was accomplished in the presence of 4-dimethylaminopyridine (DMAP) as a nucleophilic catalyst and pyridine as a basic solvent following a published procedure.[32] For the reduction of the nitro group and the following sulfonamidation, an average yield of 66 % over two steps could be achieved, while the individual yields for those two steps lay in the range of 34 % (10 l) to 88 % (10 g). Biological evaluation The results of the competitive TR-FRET assay showed that, with the exception of 10 a, all synthesized n-alkyl derivatives exhibit significantly higher binding affinities compared with GSK0660 (Figure 2 c). As the binding affinity markedly increased by elongating the side chain from three to four carbon atoms and no further improvement was seen with additional methylene
Chemistry All subsequent compounds were synthesized using a threestep procedure (Scheme 1) starting from commercially available 4-fluoro-2-methoxy-1-nitrobenzene (8), which was aminatChemMedChem 2012, 7, 159 170
www.chemmedchem.org
161
MED
groups (10 c and 10 d), we refrained from synthesizing derivatives of 1 with side chains longer than n-hexyl.
W. E. Diederich et al. ring consists of only one methylene group. Replacing the Nphenyl substituent in 1 with a cyclohexylamino moiety, representing the fully saturated analogue of the phenyl ring system, results in compound 10 k, which also exhibits a higher binding affinity than GSK0660 (1).
Further variations of the amino substituent Encouraged by these results, we systematically investigated the effect of these N-substituents on the binding affinity. As the n-alkyl chains attached to the nitrogen led to a significant increase in affinity, it was obvious to use diversely branched alkyl chains of various lengths to scrutinize the binding pocket. Hence, we synthesized a further series of GSK0660 derivatives with branched alkyl chains (10 ej) and alkyl chain-linked aromatics (10 l, 10 m) as substituents (Table 3).
Tertiary amines In order to investigate the significance of the NH group as a potential hydrogen-bond donor, we synthesized a small series of GSK0660 derivatives possessing a tertiary amine functionality (10 np; Table 4). The binding affinity of the methylated n-
Table 4. Tertiary amines 10 np. Table 3. Branched (10 ej) and cyclic (10 km) GSK0660 (1) derivatives.
Entry Entry 1 2 3 4 5 6 7 Compd 10 e 10 f 10 g 10 h 10 i 10 j 10 k R1 tert-butyl iso-propyl 3 sec-butyl iso-pentyl iso-hexyl iso-heptyl cyclohexyl Structure 1 2
Compd 10 n 10 o 10 p
X dimethylamine n-hexyl-methylamine morpholine
Structure
10 l
benzyl
hexyl derivative (10 o) is significantly improved in comparison to GSK0660 (1), while the dimethylamine (10 n) and morpholine derivatives (10 p) are both less active (Figure 2 e). Considering compound 10 n, the loss of affinity can most probably be attributed to the missing lipophilic interactions of the shorter alkyl chain. In the case of 10 p, steric repulsion could be an explanation, but without further investigation this remains speculative.
10 m
phenethyl
Review of the initial structureactivity relationships The competitive TR-FRET assay for the branched alkyl chain derivatives (10 ej) showed that all newly synthesized compounds possess an improved binding affinity to PPARb/d in comparison to GSK0660 (Figure 2 d). The compounds with the most bulky (10 e) as well as the longest (10 j) branched alkyl substituents possess the lowest binding affinities, while all other compounds in this series (10 fi) are nearly equally potent. Homologation of GSK0660 by insertion of a methylene group between the bridging NH function and the phenyl ring leads to compound 10 l, which exhibits an increased binding affinity while further homologation (10 m) results in a slight decrease in binding. However, compound 10 m is still more potent than GSK0660 (Figure 2 c). Therefore, it appears that the optimal linker length between the NH group and the aromatic As the majority of the compounds equipped with an alkyl chain exhibited a higher binding affinity than GSK0660 (1), we varied the substitution pattern of the central phenyl moiety (Table 5) to confirm the results of our initial SAR. The competitive TR-FRET assay (Figure 2 f) clearly demonstrated that the impact of the methoxy group position on the binding affinity is strongly influenced by the choice of the corresponding N-substituent. Whereas for GSK0660, 2 (3-methoxy), and 3 (2-hydrogen), the affinity towards the PPARb/d LBD significantly decreased, the corresponding n-hexyl derivatives 10 d and 11 are similarly potent. Compound 12, however, is markedly less active than 10 d and 11, nevertheless, even derivative 12 is more potent than GSK0660 (1), thus indicating that the effect the different substitutions have on the binding affinity towards the PPARb/d LBD is not merely additive but rather more complex.
ChemMedChem 2012, 7, 159 170
162
www.chemmedchem.org
Improved PPARb/d Inhibitors
Table 5. Revision of the initial structureactivity relationships.
Entry 1 2 3 4 5 6
Compd GSK0660 (1) 2 3 10 d 11 12
R1 Ph Ph Ph C6H13 C6H13 C6H13
R2 H OMe H H OMe H
R3 OMe H H OMe H H
Figure 3. Results of the competitive in vitro PPAR ligand binding assay.
Subtype-specific inhibition of PPARb/d in a cell-based assay Comparison of the most active compounds of the different series Next, we determined the Ki values for the best compounds of each series, that is, compounds 10 h and 10 l (branched alkyl series), 10 b and 10 d (linear alkyl series), and 10 o (tertiary amine series), with GSK0660 (1) given as a reference (Table 6 Finally, we analyzed the effect of 10 h and 11 on the agonistinduced transcriptional activity of the three PPAR subtypes in cultured human cells (Figure 4). Toward this end, WPMY-1 myofibroblasts were cotransfected with two plasmids: 1) a vector expressing a transcriptional activator consisting of a PPAR LBD fused to a LexA DNA binding domain; and 2) a luciferase reporter construct harboring multiple LexA binding sites upstream of a basal promoter. Treatment with agonists specific for either PPARa, PPARb/d or PPARg resulted in a 2- to 2.5-fold activation in all three cases. The two test compounds efficiently antagonized the L165,041-mediated transcriptional activation of PPARb/d, whereas they had no significant effect on the ligand activation of PPARa or PPARg. These data confirm that the compounds developed in the present study have retained the functionality in a cellular context and their specificity for PPARb/d. Studies to further elucidate the potential of the present compounds are currently ongoing.[33]
Table 6. Comparison of the inhibitory activity of selected derivatives from the different compound series.
Entry Compd 1 2 3 4 5 6
R1
Structure
Ki[a] [nm] 98.15 8.59 31.32 12.1 12.34 26.06
GSK0660 (1) phenyl 10 b 10 d 10 h 10 l 10 o n-butyl n-hexyl iso-pentyl benzyl n-hexyl-methylamine
Conclusions
A series of high-affinity ligands derived from GSK0660 (1) as selective inhibitors of PPARb/d has been synthesized and a preliminary SAR has been determined. Variation of the methoxysubstituted 1,4-diaminobenzene moiety of GSK0660, as well as substitution of the methyl thiophene carboxylate by a methyl benzoate led to a significantly reduced affinity for the PPARb/d LBD. Replacement of the 4-aminophenyl substituent by medium-length n-alkyl chains resulted in a clear gain in binding affinity. By systematically varying the 4-amino substituent, we could demonstrate that the introduction of a n-butyl, an iso-pentyl, or a benzyl moiety is most beneficial, leading to compounds exhibiting an up to tenfold higher binding affinity compared with GSK0660, while retaining druglike properties in agreement with Lipinskis rule of five.[34] Interestingly, the effect of the substitution pattern of the central 1,4-diaminobenzene moiety on the binding affinity is strongly dependent on the respective 4-amino substituent. This clearly indicates that the effect the different substitutions have on the binding affinity towards the PPARb/d LBD is not merely additive. Overall, these compounds represent invaluable tools to further elucidate the molecular and biological functions of PPARb/d.
[a] Ki values were determined by nonlinear regression analysis using GraphPad Prism version 5.0 software.
and Figure 3). Replacement of the phenyl moiety of GSK0660 by medium-length alkyl chain residues has a tremendous effect on binding affinity. Introducing an n-hexyl (10 d) or an n-hexyl-methylamine (10 o) leads to an increase in the affinity towards the PPARb/d LBD by a factor of three. The introduction of an n-butyl or an iso-pentyl moiety is even more beneficial resulting in an improvement in the binding affinity by a factor of eight. Remarkably, introducing an additional methylene group when switching from phenyl (1) to benzyl (10 l) also leads to a considerably more potent compound, the affinity of which is comparable to those of the best alkyl derivatives.
ChemMedChem 2012, 7, 159 170
www.chemmedchem.org
163
MED
W. E. Diederich et al.
with instrument settings as described in the manufacturers instructions for LanthaScreen assays on a VICTOR3V Multilabel Counter (WALLAC 1420; PerkinElmer, Waltham, MA). Cell culture: WPMY-1 cells (ATCC, CRL-2854) were cultured in Dulbeccos modified Eagle medium (DMEM) supplemented with 10 % fetal bovine serum, 100 U mL1 penicillin and 100 mg mL1 streptomycin in a humidified incubator at 37 8C and 5 % CO2. Transfection and luciferase reporter assay: LexA-mPPARb/d, LexAmPPARa, LexA-mPPARg, and 7L-TATA initiator module (TATAi) have been described previously.[36] Transfections were performed with polyethyleneimine (average MW = 25 000; SigmaAldrich). Cells were transfected at 7080 % confluence on 12-well plates in DMEM plus 2 % fetal calf serum (FCS) with 2.5 mg of plasmid DNA and 2.5 mL of PEI (1:1000 dilution, adjusted to pH 7.0 and preincubated for 15 min in 100 mL phosphate-buffered saline for complex formation). After transfection (4 h), the medium was changed, and cells were incubated in DMEM plus 10 % FCS for 48 h. Luciferase assays were performed as described,[37] except that a Renilla luciferase plasmid (pRL-SV40; Promega, Mannheim, Germany) was cotransfected for standardization.
Chemistry
General: All reactions requiring inert or dry conditions were conducted in heat-dried glass-flasks under an atmosphere of Argon (Ar N50 purchased from Air Liquide). During the course of the reaction a constant overpressure of Ar was maintained. Unless stated otherwise, all commercially available starting materials and solvents were purchased and used without further purification. 4-Fluoro-2-methoxy-1-nitrobenzene and 3-(chlorosulfonyl)thiophene-2-carboxylate were purchased from Fluorochem. GSK0660 (1) was purchased from SigmaAldrich and later re-synthesized in our laboratory. Thin-layer chromatography (TLC) was performed on precoated TLC plates (silica gel 60 F254 Merck). Flash column chromatography was performed manually using silica gel 60 (MachereyNagel) or on prepacked flash chromatography columns (PF 30-SIHP-JP/12G) purchased from Interchim using a Bchi separation system. Hexanes were purchased either in technical quality and distilled prior to use or purchased in p.a. quality from Grssing (cyclohexane). EtOAc was purchased from SigmaAldrich or Acros in p.a. quality. MeCN was purchased from VWR in HPLC quality. Pyridine was stored over KOH for 48 h, distilled and stored over molecular sieves (4 ) under an Ar atmosphere.
1
Figure 4. Inhibition of the agonist-induced transcriptional activity of a) LexAPPARa, b) LexA-PPARb/d and c) LexA-PPARg by 10 h and 11. WPMY-1 cells were transiently transfected with a luciferase reporter construct containing multiple LexA binding sites along with a plasmid expressing a transcriptional activator consisting of the PPARa, PPARb/d or PPARg ligand binding domain fused with the DNA binding domain of LexA. Post-transfection (4 h), the cells were treated with the test compounds (500 nm) for 48 h, followed by 300 nm of the PPARa agonist GW7647, 500 nm of the PPARb/d agonist L165,041, or 300 nm of the PPARg agonist GW1929 or solvent only. Induction values represent luciferase activities of agonist-treated cells relative to cells treated with solvent only. All data points represent averages of triplicate experiments ( SD). Statistically significant differences to the untreated control are denoted: ** P < 0.01, * P < 0.05, as determined by a t-test.
Biological data
Time-resolved fluorescence resonance energy transfer (TR-FRET) assays in vitro: Ligand binding to PPARb/d was determined by using the respective LanthaScreen TR-FRET PPAR competitive binding assay (Invitrogen, Karlsruhe, Germany) as previously described.[16] All assays were incubated in the dark at room temperature for 60 min and carried out in a buffer containing 100 mm KCl, 20 mm Tris (pH 7.9), 0.01 % Triton X100 and 1 mg mL1 bovine serum albumin. The assays were validated for their robustness by determining the respective Z-factors.[35] Measurements were performed
H NMR and 13C NMR spectra were recorded on a Jeol ECX-400 or a Jeol ECA-500 instrument. Chemical shifts (d) are given in ppm with the residual solvent signal used as reference (CDCl3 : s, d = 7.26 ppm (1H) and t, d = 77.1 ppm (13C); [D6]DMSO: quint, d = 2.50 ppm (1H) and quint, d = 40.1 ppm (13C)). Spectra measured in CDCl3 were recorded at room temperature, while spectra taken in [D6]DMSO were recorded at 30.0 8C. Coupling constants (J) are reported in hertz (Hz). Mass spectra (MS) were recorded on a doublefocusing sector field spectrometer type 7070H (Vacuum Generators) or a double-focusing sector field spectrometer type AutoSpec (Micromass). Elemental combustion analyses were performed on a vario MICRO cube (Elementar Analysensysteme GmbH, Hanau, Germany). Microwave irradiation was applied via a Discover BenchMate Plus (CEM GmbH, Kamp-Lintfort, Germany). Melting points (mp) were determined using a melting point meter KSP1N (A. Krss Optronic GmbH, Hamburg, Germany) and are uncorrected. Unless stated otherwise, all tested compounds were at least 98 % pure as determined by combustion analysis. The relevant data is
ChemMedChem 2012, 7, 159 170
164
www.chemmedchem.org

shown together with other analytical data directly after the corresponding synthetic procedure. Abbreviations: br = broad; d = doublet; DMAP = 4-(dimethylamino)pyridine; DIAD = diisopropyl azodicarboxylate; DMSO = dimethyl sulfoxide; EI = electron ionization; ESI = electrospray ionization; HRMS = high-resolution mass spectrometry; J = coupling constant; m = multiplet; PBS = phosphate buffered saline; PCR = polymerase chain reaction; quint = quintet; RT = room temperature; s = singlet; sep = septet; t = triplet. (1C, C-5), 118.8 (2C, C-2, C-6), 115.5 (1C, C-4), 113.9 (1C, C-6), 107.4 (1C, C-2), 55.8 (1C, OCH3), 53.5 ppm (CO2CH3); MS (EI): m/z (%): 419 (12), 418 (47) [M] + , 386 (7), 214 (26), 213 (100); HRMS-EI: m/z [M] + calcd for C19H18N2O5S2 : 418.065716, found: 418.063862; Anal. calcd for C19H18N2O5S2 : C 54.53 %, H 4.34 %, N 6.69 %, found: C 54.07 %, H 4.41 %, N 6.60 %. Methyl 3-(N-(4-(phenylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (3): Compound 3 was prepared according to general procedure A from N1-phenylbenzene-1,4-diamine (371 mg, 2.00 mmol) as a purple solid (623 mg, 80 %); mp: 119121 8C; 1H NMR (400 MHz, CDCl3): d = 8.14 (br s, 1 H, SO2NH), 7.42 (d, 3JH,H = 5.2 Hz, 1 H, 5-H), 7.39 (d, 3JH,H = 5.2 Hz, 1 H, 4-H), 7.247.19 (m, 2 H, 3-H, 5H), 6.98 (d, 3JH,H = 8.6 Hz, 4 H, 2-H, 3-H, 5-H, 6-H), 6.936.87 (m, 3 H, 2-H, 4-H, 6-H), 5.71 (br s, 1 H, NH), 3.98 ppm (s, 3 H, CO2CH3); 13 C NMR (100 MHz, CDCl3): d = 161.5 (1C, CO2CH3), 144.4 (1C, C-1), 142.6 (1C, C-3), 141.8 (1C, C-4), 131.7 (1C, C-5), 130.9 (1C, C-2), 130.5 (C-4), 129.4 (2C, C-3, C-5), 128.8 (1C, C-1), 124.9 (2C, C-3, C-5), 121.4 (1C, C-4), 118.1 (2C, C-2, C-6), 117.9 (2C, C-2, C-6), 53.5 ppm (CO2CH3); MS (EI): m/z (%): 389 (10), 388 (40) [M] + , 356 (6), 184 (24), 183 (100); HRMS-EI: m/z [M] + calcd for C18H16N2O4S2 : 388.055151, found: 388.052722; Anal. calcd for C18H16N2O4S2 : C 55.65 %, H 4.15 %, N 7.21 %, found: C 55.66 %, H 4.29 %, N 7.08 %. Methyl 3-(N-(4-phenoxyphenyl)sulfamoyl)thiophene-2-carboxylate (4): Compound 4 was prepared according to general procedure A from 4-phenoxyaniline (371 mg, 2.00 mmol) as a light brown solid (570 mg, 73 %, 94 % pure); mp: 136137 8C; 1H NMR (400 MHz, CDCl3): d = 8.23 (br s, 1 H, SO2NH), 7.47 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.43 (d, 3JH,H = 5.0 Hz, 1 H, 4-H), 7.347.29 (m, 2 H, 3-H, 5H), 7.127.06 (m, 3 H, 2-H, 6-H, 4-H), 6.976.92 (m, 2 H, 2-H, 6H), 6.896.83 (m, 2 H, 3-H, 5-H), 4.01 ppm (s, 3 H, CO2CH3); 13C NMR (100 MHz, CDCl3): d = 161.5 (1C, CO2CH3), 156.9 (1C, C1), 155.6 (1C, C-4), 144.2 (1C, C1), 131.7 (1C, C-5), 131.4 (1C, C-3), 131.1 (1C, C-2), 130.5 (1C, C-4), 129.9 (2C, C-3, C-5), 124.6 (2C, C-3, C-5), 123.7 (1C, C-4), 119.4 (2C, C-2, C-6), 119.1 (2C, C-2, C-6), 53.5 ppm (CO2CH3); MS (EI): m/z (%): 390 (19), 389 (83) [M] + , 358 (11), 357 (43), 185 (18), 184 (100); HRMS-EI: m/z [M] + calcd for C18H15NO5S2 : 389.0392, found: 389.0382. Methyl 3-(N-(4-benzylphenyl)sulfamoyl)thiophene-2-carboxylate (5): Compound 5 was prepared according to general procedure A from 4-benzylaniline (367 mg, 2.00 mmol) as a white solid (526 mg, 68 %); mp: 144146 8C; 1H NMR (400 MHz, CDCl3): d = 8.22 (br s, 1 H, SO2NH), 7.44 (s, 2 H, 4-H, 5-H), 7.297.23 (m, 2 H, 4-H, 6-H), 7.217.16 (m, 1 H, 5-H), 7.137.09 (m, 2 H, 3-H, 7-H), 7.04 (s, 4 H, 2-H, 3-H, 5-H, 6-H), 3.99 (s, 3 H, CO2CH3), 3.88 ppm (s, 2 H, 1-H2); 13 C NMR (100 MHz, CDCl3): d = 161.4(1C, CO2CH3), 144.4 (1C, C-2), 140.7 (1C, C-1), 138.9 (1C, C4), 134.5 (1C, C-3), 131.7 (1C, C-5), 131.0 (1C, C-2), 130.4 (1C, C-4), 129.8 (2C, C-3, C-5), 129.0 (2C, C4, C-6), 128.6 (2C, C-3, C-7), 126.3 (1C, C-5), 122.6 (2C, C-2, C6), 53.5 (CO2CH3), 41.4 ppm (1C, C-1); MS (EI): m/z (%): 388 (42), 387 (100) [M] + , 355 (26), 290 (30), 182 (49); HRMS-EI: m/z [M] + calcd for C18H15NO5S2 : 387.0599, found: 387.0590; Anal. calcd for C18H15NO5S2 : C 58.90 %, H 4.42 %, N 3.61 %, found: C 58.85 %, H 4.59 %, N 3.75 %. Methyl 2-(N-(3-methoxy-4-(phenylamino)phenyl)sulfamoyl)benzoate (6): Compound 6 was prepared according to general procedure A from 2-methoxy-N1-phenylbenzene-1,4-diamine (431 mg, 2.00 mmol) as a purple solid (640 mg, 78 %); mp: 135137 8C; 1 H NMR (500 MHz, CDCl3): d = 7.83 (br s, 1 H, SO2NH), 7.81 (dd, 3 JH,H = 7.5 Hz, 4JH,H = 1.2 Hz, 1 H, 6-H), 7.78 (dd, 3JH,H = 8.0 Hz, 4JH,H = 1.2 Hz, 1 H, 3-H), 7.57 (dt, 3JH,H = 7.7 Hz, 4JH,H = 1.4 Hz, 1 H, 4-H), 7.47 (dt, 3JH,H = 7.7 Hz, 4JH,H = 1.4 Hz, 1 H, 5-H), 7.267.21 (m, 2 H, 3-H, 5-
General procedures for the synthesis of sulfonamides

Procedure A: A solution of arylamine (1.0 equiv) in CH2Cl2 (3 mL mmol1) was cooled to 0 8C, and Et3N (1.0 equiv) and methyl 3-(chlorosulfonyl)thiophene-2-carboxylate (1.0 equiv) were added. The mixture was allowed to warm to RT and was then stirred for 16 h. Saturated aq NH4Cl was added and the mixture was extracted with CH2Cl2. The combined organic layers were washed with brine, dried with Na2SO4, filtered, and concentrated in vacuo. Flash chromatography (hexanes/EtOAc) gave the title compound. Procedure B: The corresponding arylamine (1.0 equiv) was dissolved in anhyd pyridine (10 mL mmol1) with stirring under an atmosphere of Ar. DMAP (0.5 equiv) and methyl 3-(chlorosulfonyl)thiophene-2-carboxylate (1.1 equiv) were added, and the reaction mixture was stirred at RT for 48 h in the dark. The solvent was evaporated in vacuo, and the residue was dissolved in 5 % aq HCl and then extracted with CH2Cl2. The combined organic layers were washed with brine, dried with Na2SO4, filtered and concentrated in vacuo. Flash chromatography (hexanes/EtOAc) gave the title compound. Methyl 3-(N-(2-methoxy-4-(phenylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (1): Compound 1 was prepared according to general procedure B from 3-methoxy-N1-phenylbenzene-1,4-diamine (109 mg, 0.509 mmol) as a yellow solid (173 mg, 81 %); mp: 151.1152.3 8C; 1H NMR (400 MHz, [D6]DMSO): d = 8.60 (s, 1 H, SO2NH), 8.16 (s, 1 H, NH), 7.90 (d, 3JH,H = 5.0 Hz, 1 H, 5-H), 7.32 (d, 3 JH,H = 5.0 Hz, 1 H, 4-H), 7.247.19 (m, 2 H, 3-H, 5-H), 7.12 (d, 3JH,H = 8.5 Hz, 1 H, 6-H), 7.067.02 (m, 2 H, 2-H, 6-H), 6.83 (m, 1 H, 4-H), 6.59 (dd, 3JH,H = 8.6 Hz, 4JH,H = 2.4 Hz, 1 H, 5-H), 6.54 (d, 4JH,H = 2.3 Hz, 1 H, 3-H), 3.94 (s, 3 H, CO2CH3), 3.47 ppm (s, 3 H, OCH3); 13C NMR (125 MHz, [D6]DMSO): d = 160.9 (1C, CO2CH3), 154.0 (1C, C-2), 144.7 (1C, C-1), 143.8 (1C, C-4), 143.4 (1C, C-3), 132.3 (1C, C-2), 132.1 (1C, C-5), 131.1 (1C, C-4), 129.7 (2C, C-3, C-5), 128.0 (1C, C4), 120.7 (1C, C-6), 117.8 (2C, C-2, C-6), 116.6 (1C, C-1), 108.4 (1C, C-5), 100.5 (1C, C-3), 55.5 (1C, OCH3), 53.7 ppm (1C, CO2CH3); MS (ESI): m/z (%): 214 (100) [MC6H4O4S2] + , 419 (61) [M + H] + , 441 (90) [M + Na] + ; HRMS-ESI: m/z [M + Na] + calcd for C19H18N2O5S2Na: 441.055485, found: 441.059223; Anal. calcd for C19H18N2O5S2 : C 54.53 %, H 4.34 %, N 6.69 %, found: C 54.30 %, H 4.41 %, N 6.67 %. Methyl 3-(N-(4-(hexylamino)-3-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (2): Compound 2 was prepared according to general procedure A from 2-methoxy-N1-phenylbenzene-1,4-diamine (431 mg, 2.00 mmol) as a light green solid (682 mg, 82 %); mp: 125126 8C; 1H NMR (400 MHz, CDCl3): d = 8.11 (s, 1 H, SO2NH), 7.43 (d, 3JH,H = 5.4 Hz, 1 H, 5-H), 7.41 (d, 3JH,H = 5.2 Hz, 1 H, 4-H), 7.26 7.21 (m, 2 H, 3-H, 5-H), 7.087.02 (m, 3 H, 5-H, 2-H, 6-H), 6.92 (tt, 3JH,H = 7.5 Hz, 4JH,H = 1.2 Hz, 1 H, 4-H), 6.83 (d, 4JH,H = 2.29 Hz, 1 H, 2-H), 6.42 (dd, 3JH,H = 8.3 Hz, 4JH,H = 2.3 Hz, 1 H, 6-H), 6.04 (br s, 1 H, NH), 3.99 (s, 3 H, CO2CH3), 3.82 ppm (s, 3 H, OCH3); 13C NMR (100 MHz, CDCl3): d = 161.5 (1C, CO2CH3), 148.4 (1C, C-3), 144.4 (1C, C-1), 142.2 (1C, C-3), 131.8 (1C, C-5), 131.7 (1C, C-2), 130.9 (1C, C-1), 130.5 (1C, C-4), 129.4 (2C, C-3, C-5), 128.1 (1C, C-4), 121.6
ChemMedChem 2012, 7, 159 170
www.chemmedchem.org
165
MED
H), 7.077.01 (m, 3 H, 5-H, 2-H, 6-H), 6.92 (tt, 3JH,H = 7.2 Hz, 4JH,H = 1.2 Hz, 1 H, 4-H), 6.79 (d, 4JH,H = 2.3 Hz, 1 H, 2-H), 6.47 (dd, 3JH,H = 8.3 Hz, 4JH,H = 2.3 Hz, 1 H, 6-H), 6.04 (br s, 1 H, NH), 4.03 (s, 3 H, CO2CH3), 3.80 ppm (s, 3 H, OCH3); 13C NMR (125 MHz, CDCl3): d = 168.5 (1C, CO2CH3), 148.3 (1C, C-3), 142.3 (1C, C-2), 138.2 (1C, C-1), 132.6 (1C, C-4), 131.8 (1C, C-1), 131.5 (1C, C-5), 130.7 (1C, C-1), 130.7 (1C, C-6), 130.5 (1C, C-3), 129.4 (2C, C-3, C-5), 128.4 (1C, C4), 121.5 (1C, C-5), 118.8 (2C, C-2, C-6), 116.4 (1C, C-4), 113.9 (1C, C-6), 107.9 (1C, C-2), 55.8 (1C, OCH3), 53.7 ppm (1C, CO2CH3); MS (EI): m/z (%): 412 (24) [M] + , 381 (27), 380 (100), 316 (13), 214 (14), 213 (65); HRMS-EI: m/z [M] + calcd for C18H15NO5S2 : 412.109294, found: 412.108092; Anal. calcd for C18H15NO5S2 : C 61.15 %, H 4.89 %, N 6.79 %, found: C 61.21 %, H 5.03 %, N 6.90 %. Methyl 2-(N-(4-(phenylamino)phenyl)sulfamoyl)benzoate (7): Compound 7 was prepared according to general procedure A from N1-phenylbenzene-1,4-diamine (371 mg, 2.00 mmol) as a purple solid (596 mg, 80 %); mp: 142144 8C; 1H NMR (500 MHz, CDCl3): d = 7.88 (br s, 1 H, SO2NH), 7.82 (dd, 3JH,H = 7.7 Hz, 4JH,H = 1.2 Hz, 1 H, 6-H), 7.77 (dd, 3JH,H = 7.7 Hz, 4JH,H = 1.2 Hz, 1 H, 3-H), 7.57 (dt, 3JH,H = 7.5 Hz, 4JH,H = 1.2 Hz, 1 H, 4-H), 7.50 (dt, 3JH,H = 7.7 Hz, 4JH,H = 1.4 Hz, 1 H, 5-H), 7.257.20 (m, 2 H, 3-H, 5-H), 7.016.96 (m, 4 H, 2-H, 3H, 5-H, 6-H), 6.91 (tt, 3JH,H = 7.5 Hz, 4JH,H = 1.2 Hz, 1 H, 4-H), 6.90 6.86 (m, 2 H, 2-H, 6-H), 5.70 (br s, 1 H, NH), 4.02 ppm (s, 3 H, CO2CH3); 13C NMR (125 MHz, CDCl3): d = 168.5 (1C, CO2CH3), 142.7 (1C, C-2), 141.8 (1C, C-1), 138.2 (1C, C-1), 132.6 (1C, C-4), 131.5 (1C, C-1), 130.6 (1C, C-6), 130.6 (1C, C-3), 129.4 (2C, C-3, C-5), 129.2 (1C, C-4), 129.1 (1C, C-1), 125.5 (2C, C-3, C-5), 121.4 (1C, C-4), 118.1 (2C, C-2, C-6), 117.9 (2C, C-2, C-6), 53.6 ppm (1C, CO2CH3); MS (EI): m/z (%): 382 (20) [M] + , 351 (56), 350 (100), 286 (67), 285 (54), 183 (51); HRMS-EI: m/z [M] + calcd for C18H15NO5S2 : 382.098729, found: 382.095154; Anal. calcd for C18H15NO5S2 : C 62.81 %, H 4.74 %, N 7.32 %, found: C 62.89 %, H 5.10 %, N 7.33 %. Methyl 3-(N-(2-methoxy-4-(propylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (10 a): Compound 10 a was prepared according to general procedure B from 3-methoxy-N1-propylbenzene-1,4diamine as a yellow solid (124 mg, 65 % over two steps); mp: 178.4179.7 8C; 1H NMR (400 MHz, CDCl3): d = 8.09 (br s, 1 H, SO2NH), 7.38 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.35 (d, 3JH,H = 5.3 Hz, 1 H, 4H), 7.28 (d, 3JH,H = 8.7 Hz, 1 H, 6-H), 6.12 (dd, 3JH,H = 8.7 Hz, 4JH,H = 2.5 Hz, 1 H, 5-H), 5.92 (d, 4JH,H = 2.5 Hz, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.61 (br s, 1 H, NH), 3.48 (s, 3 H, OCH3), 3.00 (t, 3JH,H = 7.1 Hz, 2 H, 1-H2), 1.60 (tq, 3JH,H = 7.1 Hz, 3JH,H = 7.1 Hz, 2 H, 2-H2), 0.97 ppm (t, 3JH,H = 7.3 Hz, 3 H, 3-H3); 13C NMR (125 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.5 (1C, C-2), 148.3 (1C, C-4), 145.7 (1C, C-3), 131.6 (1C, C-5), 129.3 (1C, C-4), 128.3 (1C, C-6), 114.4 (1C, C1), 104.6 (1C, C-5), 95.7 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 45.9 (1C, C-1), 22.7 (1C, C-2), 11.7 ppm (1C, C-3); MS (EI): m/z (%): 384 (19) [M] + , 179 (100) [MC6H5O4S2] + ; MS (ESI): m/z (%): 385 (17) [M + H] + , 769 (100) [2M + H] + , 791 (33) [2M + Na] + , 1153 (28) [3M + H] + , 1175 (13) [3M + Na] + ; HRMS-EI: m/z [M] + calcd for C16H20N2O5S2 : 384.081366, found: 384.083550; Anal. calcd for C16H20N2O5S2 : C 49.98 %, H 5.24 %, N 7.29 %, found: C 49.93 %, H 5.27 %, N 7.32 %. Methyl 3-(N-(4-(butylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 b): Compound 10 b was prepared according to general procedure B from N1-butyl-3-methoxybenzene-1,4diamine as a yellow solid (135 mg, 68 % over two steps); mp: 129.2131.8 8C; 1H NMR (400 MHz, CDCl3): d = 8.09 (s, 1 H, SO2NH), 7.38 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.35 (d, 3JH,H = 5.0 Hz, 1 H, 4-H), 7.28 (d, 3JH,H = 8.5 Hz, 1 H, 6-H), 6.12 (dd, 3JH,H = 8.7 Hz, 4JH,H = 2.5 Hz, 1 H, 5-H), 5.92 (d, 3JH,H = 2.5 Hz, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.59 (br s, 1 H, NH), 3.48 (s, 3 H, OCH3), 3.03 (t, 3JH,H = 7.1 Hz, 2 H, 1-H2),
1.591.52 (m, 2 H, 2-H2), 1.441.35 (m, 2 H, 3-H2), 0.94 ppm (t, 3 JH,H = 7.3 Hz, 3 H, 4-H3); 13C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.5 (1C, C-2), 148.3 (1C, C-4), 145.8 (1C, C-3), 131.6 (1C, C-5), 129.2 (1C, C-4), 128.3 (1C, C-6), 114.5 (1C, C-2), 106.8 (1C, C1), 104.7 (1C, C-5), 95.7 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 43.8 (1C, C-1), 31.6 (1C, C-2), 20.3 (1C, C-3), 14.0 ppm (1C, C-4); MS (ESI): m/z (%): 399 (6) [M + H] + , 421 (50) [M + Na] + , 819 (100) [2M + Na] + , 1217 (26) [3M + Na] + ; HRMS-ESI: m/z [M+Na] + calcd for C17H22N2O5S2Na: 421.086785, found: 421.083904; Anal. calcd for C17H22N2O5S2 : C 51.24 %, H 5.56 %, N 7.03 %, found: C 51.33 %, H 5.58 %, N 6.91 %. Methyl 3-(N-(2-methoxy-4-(pentylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (10 c): Compound 10 c was prepared according to general procedure B from 3-methoxy-N1-pentylbenzene-1,4diamine as a yellow solid (113 mg, 55 % over two steps); mp: 114.5115.3 8C; 1H NMR (400 MHz, CDCl3): d = 8.09 (br s, 1 H, SO2NH), 7.38 (d, 3JH,H = 5.0 Hz, 1 H, 5-H), 7.35 (d, 3JH,H = 5.3 Hz, 1 H, 4H), 7.28 (d, 3JH,H = 8.7 Hz, 1 H, 6-H), 6.12 (dd, 3JH,H = 8.6 Hz, 4JH,H = 2.5 Hz, 1 H, 5-H), 5.92 (d, 4JH,H = 2.5 Hz, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.59 (br s, 1 H, NH), 3.48 (s, 3 H, OCH3), 3.02 (t, 3JH,H = 7.1 Hz, 2 H, 1-H2), 1.611.54 (m, 2 H, 2-H2), 1.371.31 (m, 4 H, 3H2, 4-H2), 0.90 ppm (t, 3JH,H = 7.1 Hz, 3 H, 5-H3); 13C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.5 (1C, C-2), 148.3 (1C, C-4), 145.7 (1C, C-3), 131.6 (1C, C-5), 129.3 (1C, C-4), 128.3 (1C, C-6), 114.3 (1C, C-1), 104.6 (1C, C-5), 95.7 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 44.1 (1C, C-1), 29.4 (1C, C-2), 29.3 (1C, C-3), 22.6 (1C, C-4), 14.1 ppm (1C, C-5); MS (ESI): m/z (%): 413 (100) [M + H] + , 430 (70) [M + NH4] + , 435 (20) [M + Na] + , 825 (42) [2M + H] + ; HRMS-EI: m/z [M] + calcd for C18H24N2O5S2 : 412.112666, found: 412.111911; Anal. calcd for C18H24N2O5S2 : C 52.41 %, H 5.86 %, N 6.79 %, found: C 52.43 %, H 6.20 %, N 7.14 %. Methyl 3-(N-(4-(hexylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 d): Compound 10 d was prepared according to general procedure B from N1-hexyl-3-methoxybenzene-1,4diamine as a yellow solid (175 mg, 71 % over two steps); mp: 99.1 101.3 8C; 1H NMR (400 MHz, [D6]DMSO): d = 8.36 (s, 1 H, SO2NH), 7.86 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.26 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 6.94 (m, 1 H, 6-H), 6.056.02 (m, 2 H, 3-H, 5-H), 5.58 (t, 3JH,H = 5.4 Hz, 1 H, NH), 3.93 (s, 3 H, CO2CH3), 3.41 (s, 3 H, OCH3), 2.92 (m, 2 H, 1H2), 1.49 (tt, 3JH,H = 7.3 Hz, 3JH,H = 7.3 Hz, 2 H, 2-H2), 1.361.24 (m, 6 H, 3-H2, 4-H2, 5-H2), 0.86 ppm (t, 3JH,H = 6.9 Hz, 3 H, 6-H3); 13 C NMR (100 MHz, [D6]DMSO): d = 160.9 (1C, CO2CH3), 154.5 (1C, C-2), 149.7 (1C, C-4), 145.1 (1C, C-3), 132.1 (1C, C-2), 131.9 (1C, C5), 131.2 (1C, C-4), 128.9 (1C, C-6), 112.6 (1C, C-1), 103.8 (1C, C-5), 95.9 (1C, C-3), 55.5 (1C, OCH3), 53.7 (1C, CO2CH3), 43.4 (1C, C-1), 31.6 (1C, C-4), 29.2 (1C, C-2), 26.9 (1C, C-3), 22.6 (1C, C-5), 14.4 ppm (1C, C-6); MS (ESI): m/z (%): 427 (55) [M + H] + , 449 (100) [M + Na] + ; HRMS-ESI: m/z [M + Na] + calcd for C19H26N2O5S2Na: 449.118086, found: 449.118829; Anal. calcd for C19H26N2O5S2 : C 53.50 %, H 6.14 %, N 6.57 %, found: C 53.71 %, H 6.21 %, N 6.65 %. Methyl 3-(N-(4-(tert-butylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 e): Compound 10 e was prepared according to general procedure B from N1-tert-butyl-3-methoxybenzene-1,4-diamine as a yellow solid (231 mg, 80 % over two steps); mp: 167.5169.3 8C; 1H NMR (400 MHz, [D6]DMSO): d = 8.39 (s, 1 H, SO2NH), 7.87 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.29 (d, 3JH,H = 5.3 Hz, 1 H, 4H), 6.93 (d, 3JH,H = 8.5 Hz, 1 H, 6-H), 6.21 (dd, 3JH,H = 8.7, 4JH,H = 2.3 Hz, 1 H, 5-H), 6.17 (d, 4JH,H = 2.3 Hz, 1 H, 3-H), 5.23 (br s, 1 H, NH), 3.93 (s, 3 H, CO2CH3), 3.40 (s, 3 H, OCH3), 1.25 ppm (s, 9 H, C(CH3)3); 13 C NMR (100 MHz, [D6]DMSO): d = 160.9 (1C, CO2CH3), 154.1 (1C, C-2), 148.2 (1C, C-4), 145.1 (1C, C-3), 132.1 (1C, C-2), 132.0 (1C, C5), 131.1 (1C, C-4), 128.4 (1C, C-6), 113.0 (1C, C-1), 106.6 (1C, C-5),
ChemMedChem 2012, 7, 159 170
166
www.chemmedchem.org

99.2 (1C, C-3), 55.5 (1C, OCH3), 53.7 (1C, CO2CH3), 50.7 (1C, C(CH3)3), 30.0 ppm (3C, C(CH3)3); MS (ESI): m/z (%): 399 (100) [M + H] + , 421 (8) [M + Na] + , 797 (22) [2M + H] + ; HRMS-EI: m/z [M] + calcd for C17H22N2O5S2 : 398.097016, found: 398.100247; Anal. calcd for C17H22N2O5S2 : C 51.24 %, H 5.56 %, N 7.03 %, found: C 51.47 %, H 5.60 %, N 6.94 %. Methyl 3-(N-(4-(isopropylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 f): Compound 10 f was prepared according to general procedure B from N1-isopropyl-3-methoxybenzene-1,4-diamine as a yellow solid (320 mg, 78 % over two steps); mp: 135.1136.9 8C; 1H NMR (400 MHz, CDCl3): d = 8.09 (s, 1 H, SO2NH), 7.38 (d, 3JH,H = 5.0 Hz, 1 H, 5-H), 7.35 (d, 3JH,H = 5.3 Hz, 1 H, 4H), 7.27 (d, 3JH,H = 8.2 Hz, 1 H, 6-H), 6.10 (dd, 3JH,H = 8.6 Hz, 4JH,H = 2.4 Hz, 1 H, 5-H), 5.90 (d, 4JH,H = 2.3 Hz, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.53 (sep, 3JH,H = 6.2 Hz, 1 H, 1-H), 3.47 (s, 3 H, OCH3), 3.46 (br s,1 H, NH), 1.16 ppm (d, 3JH,H = 6.4 Hz, 6 H, 2-H3, 3-H3); 13C NMR (125 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.4 (1C, C-2), 147.2 (1C, C-4), 145.7 (1C, C-3), 131.6 (1C, C-5), 131.6 (1C, C-2), 129.2 (1C, C-4), 128.3 (1C, C-6), 114.3 (1C, C-1), 105.0 (1C, C-5), 96.3 (1C, C3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 44.4 (1C, C-1), 23.1 ppm (2C, C-2, C-3); MS (ESI): m/z (%): 385 (21) [M + H] + , 769 (100) [2M + H] + ; HRMS-ESI: m/z [M + H] + calcd for C16H21N2O5S2 : 385.089191, found: 385.092388; Anal. calcd for C16H20N2O5S2 : C 49.98 %, H 5.24 %, N 7.29 %, found: C 50.20 %, H 5.33 %, N 7.19 %. Methyl 3-(N-(4-(isobutylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 g): Compound 10 g was prepared according to general procedure B from N1-isobutyl-3-methoxybenzene1,4-diamine as a yellow solid (175 mg, 88 % over two steps); mp: 158.0159.8 8C; 1H NMR (400 MHz, [D6]DMSO): d = 8.35 (s, 1 H, SO2NH), 7.86 (d, 3JH,H = 5.0 Hz, 1 H, 5-H), 7.27 (d, 3JH,H = 5.0 Hz, 1 H, 4H), 6.94 (m, 1 H, 6-H), 6.066.03 (m, 2 H, 3-H, 5-H), 5.65 (t, 3JH,H = 5.6 Hz, 1 H, NH), 3.93 (s, 3 H, CO2CH3), 3.41 (s, 3 H, OCH3), 2.75 (dd, 3 JH,H = 6.2 Hz, 3JH,H = 6.2 Hz, 2 H, 1-H2), 1.77 (sep, 3JH,H = 6.6 Hz, 1 H, 2-H), 0.90 ppm (d, 3JH,H = 6.6 Hz, 6 H, 3-H3, 4-H3); 13C NMR (100 MHz, [D6]DMSO): d = 160.9 (1C, CO2CH3), 154.5 (1C, C-2), 149.8 (1C, C-4), 145.1 (1C, C-3), 132.1 (1C, C-2), 131.9 (1C, C-5), 131.2 (1C, C-4), 128.9 (1C, C-6), 112.5 (1C, C-1), 103.7 (1C, C-5), 95.9 (1C, C-3), 55.5 (1C, OCH3), 53.7 (1C, CO2CH3), 51.3 (1C, C-1), 28.0 (1C, C-2), 21.0 ppm (2C, C-3, C-4); MS (ESI): m/z (%): 194 (84) [MC6H4O4S2] + , 399 (100) [M + H] + , 421 (63) [M + Na] + , 797 (30) [2M + H] + , 819 (40) [2M + Na] + ; HRMS-EI: m/z [M] + calcd for C17H22N2O5S2 : 398.097016, found: 398.100771; Anal. calcd for C17H22N2O5S2 : C 51.24 %, H 5.56 %, N 7.03 %, found: C 51.00 %, H 5.72 %, N 6.88 %. Methyl 3-(N-(4-(isopentylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 h): Compound 10 h was prepared according to general procedure B from N1-isopentyl-3-methoxybenzene-1,4-diamine as a yellow solid (173 mg, 42 % over two steps); mp: 125.8129.1 8C; 1H NMR (400 MHz, CDCl3): d = 8.09 (s, 1 H, SO2NH), 7.37 (d, 3JH,H = 5.0 Hz, 1 H, 5-H), 7.35 (d, 3JH,H = 5.3 Hz, 1 H, 4H), 7.28 (d, 3JH,H = 8.7 Hz, 1 H, 6-H), 6.12 (dd, 3JH,H = 8.6 Hz, 4JH,H = 2.4 Hz, 1 H, 5-H), 5.92 (d, 4JH,H = 2.3 Hz, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.55 (br s, 1 H, NH), 3.48 (s, 3 H, OCH3), 3.03 (t, 3JH,H = 7.3 Hz, 2 H, 1-H2), 1.67 (sep, 3JH,H = 6.6 Hz, 1 H, 3-H), 1.46 (dt, 3 JH,H = 7.2 Hz, 3JH,H = 7.2 Hz, 2 H, 2-H2), 0.92 ppm (d, 3JH,H = 6.4 Hz, 6 H, 4-H3, 5-H3); 13C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.5 (1C, C-2), 148.4 (1C, C-4), 145.7 (1C, C-3), 131.6 (1C, C-2), 131.6 (1C, C-5), 129.3 (1C, C-4), 128.3 (1C, C-6), 114.3 (1C, C1), 104.6 (1C, C-5), 95.7 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 42.2 (1C, C-1), 38.5 (1C, C-2), 26.0 (1C, C-3), 22.7 ppm (2C, C-4, C-5); MS (ESI): m/z (%): 208 (68) [MC6H4O4S2] + , 413 (100) [M + H] + , 435 (41) [M + Na] + , 825 (31) [2M + H] + , 847 (32)
ChemMedChem 2012, 7, 159 170
[2M + Na] + ; HRMS-ESI: m/z [M + H] + calcd for C18H25N2O5S2 : 413.120491, found: 413.123613; Anal. calcd for C18H24N2O5S2 : C 52.41 %, H 5.86 %, N 6.79 %, found: C 52.48 %, H 5.99 %, N 6.61 %. Methyl 3-(N-(2-methoxy-4-(4-methylpentylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (10 i): Compound 10 i was prepared according to general procedure B from 3-methoxy-N1-(4-methylpentyl)benzene-1,4-diamine as a yellow solid (52 mg, 68 % over two steps); mp: 156.3158.3 8C; 1H NMR (400 MHz, CDCl3): d = 8.09 (s, 1 H, SO2NH), 7.37 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.34 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 7.27 (d, 3JH,H = 8.6 Hz, 1 H, 6-H), 6.11 (dd, 3JH,H = 8.6 Hz, 4 JH,H = 2.4 Hz, 1 H, 5-H), 5.91 (d, 4JH,H = 2.3 Hz, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.60 (br s, 1 H, NH), 3.48 (s, 3 H, OCH3), 3.00 (t, 3JH,H = 7.2 Hz, 2 H, 1-H2), 1.601.50 (m, 3 H, 4-H, 2-H2), 1.271.21 (m, 2 H, 3-H2), 0.88 ppm (d, 3JH,H = 6.6 Hz, 6 H, 5-H3, 6-H3); 1H NMR (400 MHz, [D6]DMSO): d = 8.36 (s, 1 H, SO2NH), 7.86 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.26 (d, 3JH,H = 5.0 Hz, 1 H, 4-H), 6.95 (m, 1 H, 6-H), 6.056.02 (m, 2 H, 5-H, 3-H), 5.59 (t, 3JH,H = 5.4 Hz, 1 H, NH), 3.93 (s, 3 H, CO2CH3), 3.41 (s, 3 H, OCH3), 2.91 (m, 2 H, 1-H2), 1.581.45 (m, 3 H, 4-H, 2-H2), 1.261.20 (m, 2 H, 3-H2), 0.85 ppm (d, 3JH,H = 6.6 Hz, 6 H, 5-H3, 6-H3); 13C NMR (125 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.5 (1C, C-2), 148.3 (1C, C-4), 145.7 (1C, C-3), 131.6 (1C, C-5), 129.2 (1C, C-4), 128.3 (1C, C-6), 114.4 (1C, C-1), 104.6 (1C, C5), 95.7 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 44.4 (1C, C1), 36.4 (1C, C-3), 27.9 (1C, C-4), 27.4 (1C, C-2), 22.6 ppm (2C, C5, C-6); MS (ESI): m/z (%): 222 (71) [MC6H4O4S2] + , 427 (100) [M + H] + , 449 (58) [M + Na] + , 853 (29) [2M + H] + , 875 (36) [2M+Na] + ; HRMS-EI: m/z [M] + calcd for C19H26N2O5S2 : 426.128316, found: 426.129707; Anal. calcd for C19H26N2O5S2 : C 53.50 %, H 6.14 %, N 6.57 %, found: C 53.55 %, H 6.19 %, N 6.48 %. Methyl 3-(N-(2-methoxy-4-(5-methylhexylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (10 j): Compound 10 j was prepared according to general procedure B from 3-methoxy-N1-(5methylhexyl)benzene-1,4-diamine as a yellow solid (53 mg, 73 % over two steps); mp: 110.2112.8 8C; 1H NMR (400 MHz, [D6]DMSO): d = 8.35 (br s, 1 H, SO2NH), 7.86 (d, 3JH,H = 5.3 Hz, 1 H, 5H), 7.26 (d, 3JH,H = 5.0 Hz, 1 H, 4-H), 6.94 (m, 1 H, 6-H), 6.056.02 (m, 2 H, 5-H, 3-H), 5.59 (t, 3JH,H = 5.3 Hz, 1 H, NH), 3.93 (s, 3 H, CO2CH3), 3.41 (s, 3 H, OCH3), 2.92 (m, 2 H, 1-H2), 1.561.44 (m, 3 H, 5-H, 2H2), 1.361.28 (m, 2 H, 3-H2), 1.191.13 (m, 2 H, 4-H2), 0.85 ppm (d, 3 JH,H = 6.6 Hz, 6 H, 6-H3, 7-H3); 13C NMR (100 MHz, [D6]DMSO): d = 160.9 (1C, CO2CH3), 154.5 (1C, C-2), 149.7 (1C, C-4), 145.1 (1C, C-3), 132.1 (1C, C-2), 131.9 (1C, C-5), 131.2 (1C, C-4), 128.9 (1C, C-6), 112.6 (1C, C-1), 103.8 (1C, C-5), 95.9 (1C, C-3), 55.5 (1C, OCH3), 53.7 (1C, CO2CH3), 43.4 (1C, C-1), 38.8 (1C, C-4), 29.5 (1C, C-2), 28.0 (1C, C-5), 25.0 (1C, C-3), 23.0 ppm (2C, C-6, C-7); MS (ESI): m/z (%): 236 (93) [MC6H4O4S2] + , 441 (100) [M + H] + , 463 (61) [M+Na] + , 903 (33) [2M + Na] + ; HRMS-EI: m/z [M] + calcd for C20H28N2O5S2 : 440.143966, found: 440.140167; Anal. calcd for C20H28N2O5S2 : C 54.52 %, H 6.41 %, N 6.36 %, found: C 54.60 %, H 6.42 %, N 6.28 %. Methyl 3-(N-(4-(cyclohexylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 k): Compound 10 k was prepared according to general procedure B from N1-cyclohexyl-3-methoxybenzene-1,4-diamine as a yellow solid (259 mg, 61 % over two steps); mp: 167.3168.5 8C; 1H NMR (400 MHz, CDCl3): d = 8.10 (br s, 1 H, SO2NH), 7.39 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.36 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 7.27 (d, 3JH,H = 9.2 Hz, 1 H, 6-H), 6.15 (dd, 3JH,H = 8.5 Hz, 4 JH,H = 1.8 Hz, 1 H, 5-H), 5.95 (br s, 1 H, 3-H), 4.00 (s, 3 H, CO2CH3), 3.48 (s, 3 H, OCH3), 3.15 (m, 1 H, 1-H), 1.99 (m, 2 H, cyclohexyl-H), 1.73 (m, 2 H, cyclohexyl-H), 1.63 (m, 1 H, cyclohexyl-H), 1.07 1.38 ppm (m, 5 H, cyclohexyl-H); 13C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.6 (1C, C-2), 147.2 (1C, C-4), 145.7 (1C, C-3),
www.chemmedchem.org
167
MED
131.6 (1C, C-5), 131.6 (1C, C-2), 129.3 (1C, C-4), 128.3 (1C, C-6), 114.1 (1C, C-2), 104.9 (1C, C-5), 96.2 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 51.9 (1C, C-1), 33.5 (2C, C-2, C-6), 25.9 (1C, C-4), 25.0 ppm (2C, C-3, C-5); MS (EI): m/z (%): 424 (13) [M] + , 220 (97) [M-C13H18NO] + , 219 (100) [MC6H5O4S2] + , 111 (83) [C5H3OS] + ; HRMS-EI: m/z [M] + calcd for C19H24N2O5S2 : 424.1127, found: 424.1138; Anal. calcd for C19H24N2O5S2 : C 53.75 %, H 5.70 %, N 6.60 %, found: C 53.53 %, H 5.69 %, N 6.43 %. Methyl 3-(N-(4-(benzylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 l): Compound 10 l was prepared according to general procedure B from N1-benzyl-3-methoxybenzene-1,4diamine as a yellow solid (145 mg, 34 % over two steps); mp: 141.6142.8 8C; 1H NMR (400 MHz, CDCl3): d = 8.06 (s, 1 H, SO2NH), 7.32 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.30 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 7.23 (d, 3JH,H = 8.7 Hz, 1 H, 6-H), 7.277.19 (m, 5 H, 3-H, 4-H, 5-H, 6-H, 7-H), 6.11 (dd, 3JH,H = 8.6 Hz, 4JH,H = 2.4 Hz, 1 H, 5-H), 5.91 (d, 4JH,H = 2.5 Hz, 1 H, 3-H), 4.19 (s, 2 H, 1-H2), 3.96 (br s, 1 H, NH), 3.94 (s, 3 H, CO2CH3), 3.39 ppm (s, 3 H, OCH3); 13C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.4 (1C, C-2), 147.9 (1C, C-4), 145.7 (1C, C-3), 139.0 (1C, C-2), 131.6 (1C, C-5), 129.3 (1C, C-4), 128.8 (2C, C-4, C6), 128.2 (1C, C-6), 127.6 (2C, C-3, C-7), 127.5 (1C, C-5), 114.9 (1C, C-1), 104.8 (1C, C-5), 96.0 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 48.5 ppm (1C, C-1); MS (ESI): m/z (%): 455 (82) [M + Na] + , 887 (100) [2M + Na] + , 1319 (22) [3M + Na] + ; HRMS-ESI: m/z [M+Na] + calcd for C20H20N2O5S2Na: 455.071135, found: 455.067108; Anal. calcd for C20H20N2O5S2 : C 55.54 %, H 4.66 %, N 6.48 %, found: C 55.59 %, H 4.80 %, N 6.39 %. Methyl 3-(N-(2-methoxy-4-(phenethylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (10 m): Compound 10 m was prepared according to general procedure B from 3-methoxy-N1-phenethylbenzene-1,4-diamine as a yellow solid (79 mg, 35 % over two steps); mp: 150.9152.5 8C; 1H NMR (400 MHz, CDCl3): d = 8.11 (br s, 1 H, SO2NH), 7.38 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.35 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 7.337.28 (m, 3 H, Ar-H), 7.267.18 (m, 3 H, Ar-H), 6.14 (dd, 3 JH,H = 8.7 Hz, 4JH,H = 2.5 Hz, 1 H, 5-H), 5.91 (d, 4JH,H = 2.3 Hz, 1 H, 3H), 4.00 (s, 3 H, CO2CH3), 3.66 (br s, 1 H, NH), 3.47 (s, 3 H, OCH3), 3.32 (t, 3JH,H = 6.9 Hz, 2 H, 1-H2), 2.88 ppm (t, 3JH,H = 7.0 Hz, 2 H, 2-H2); 13 C NMR (125 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.5 (1C, C-2), 147.8 (1C, C-4), 145.7 (1C, C-3), 139.1 (1C, C-3), 131.6 (1C, C-2), 131.6 (1C, C-5), 129.3 (1C, C-4), 128.8 (2C, C-5, C-7), 128.7 (2C, C4, C-8), 128.2 (1C, C-6), 126.6 (1C, C-6), 114.8 (1C, C-1), 104.9 (1C, C-5), 96.0 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 45.1 (1C, C-1), 35.6 ppm (1C, C-2); MS (ESI): m/z (%): 447 (47) [M + H] + , 469 (100) [M + Na] + , 915 (11) [2M + Na] + ; HRMS-ESI: m/z [M + H] + calcd for C21H23N2O5S2 : 447.104841, found: 447.104703; Anal. calcd for C21H22N2O5S2 : C 56.48 %, H 4.97 %, N 6.27 %, found: C 56.50 %, H 5.07 %, N 6.24 %. Methyl 3-(N-(4-(dimethylamino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 n): Compound 10 n was prepared according to general procedure B from 3-methoxy-N1,N1-dimethylbenzene-1,4-diamine as a yellow solid (227 mg, 80 % over two steps); mp: 161.3162.9 8C; 1H NMR (400 MHz, [D6]DMSO): d = 8.45 (br s, 1 H, SO2NH), 7.87 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.27 (d, 3JH,H = 5.0 Hz, 1 H, 4-H), 7.06 (d, 3JH,H = 8.7 Hz, 1 H, 6-H), 6.20 (dd, 3JH,H = 8.8 Hz, 4JH,H = 2.6 Hz, 1 H, 5-H), 6.14 (d, 4JH,H = 2.5 Hz, 1 H, 2-H), 3.94 (s, 3 H, CO2CH3), 3.49 (s, 3 H, OCH3), 2.85 ppm (s, 6 H, N(CH3)2); 13 C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.2 (1C, C-2), 150.3 (1C, C-4), 145.7 (1C, C-3), 131.6 (1C, C-2), 131.6 (1C, C-5), 129.3 (1C, C-4), 127.8 (1C, C-6), 114.0 (1C, C-1), 104.7 (1C, C-5), 95.5 (1C, C-3), 55.2 (1C, OCH3), 53.1 (1C, CO2CH3), 40.7 ppm (2C, N(CH3)2); MS (ESI): m/z (%): 371 (100) [M + H] + , 393 (26) [M + Na] + ; HRMS-EI: m/z [M] + calcd for C15H18N2O5S2 : 370.065716, found:
370.064960; Anal. calcd for C15H18N2O5S2 : C 48.63 %, H 4.90 %, N 7.56 %, found: C 48.68 %, H 4.92 %, N 7.63 %. Methyl 3-(N-(4-(hexyl(methyl)amino)-2-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (10 o): Compound 10 o was prepared according to general procedure A from N1-hexyl-3-methoxyN1-methylbenzene-1,4-diamine (172 mg, 0.728 mmol) as a yellow solid (141 mg, 44 %); mp: 110.1111.7 8C; 1H NMR (400 MHz, CDCl3): d = 8.11 (s, 1 H, SO2NH), 7.39 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.36 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 7.31 (d, 3JH,H = 8.9 Hz, 1 H, 6-H), 6.20 (dd, 3 JH,H = 8.8 Hz, 4JH,H = 2.7 Hz, 1 H, 5-H), 5.98 (d, 4JH,H = 2.5 Hz, 1 H, 3H), 4.01 (s, 3 H, CO2CH3), 3.51 (s, 3 H, OCH3), 3.22 (t, 3JH,H = 7.6 Hz, 2 H, 1-H2), 2.87 (s, 3 H, NCH3), 1.551.46 (m, 2 H, 2-H2), 1.321.25 (m, 6 H, 3-H2, 4-H2, 5-H2), 0.88 ppm (t, 3JH,H = 6.8 Hz, 3 H, 6-H3); 13 C NMR (100 MHz, CDCl3): d = 160.8 (1C, CO2CH3), 153.4 (1C, C-2), 149.2 (1C, C-4), 145.9 (1C, C-3), 131.6 (1C, C-5), 131.5 (1C, C-2), 129.2 (1C, C-4), 128.1 (1C, C-6), 113.3 (1C, C-1), 104.4 (1C, C-5), 95.1 (1C, C-3), 55.1 (1C, OCH3), 53.1 (1C, CO2CH3), 53.0 (1C, C-1), 38.5 (1C, NCH3), 31.8 (1C, C-4), 26.9 (1C, C-2), 26.8 (1C, C-3), 22.7 (1C, C-5), 14.1 ppm (1C, C-6); MS (ESI): m/z (%): 441 (29) [M + H] + , 463 (93) [M + Na] + , 904 (100) [2M + Na + H] + , 1344 (33) [3M + Na + H] + ; HRMS-ESI: m/z [M + H] + calcd for C20H29N2O5S2 : 441.151791, found: 441.148997; Anal. calcd for C20H28N2O5S2 : C 54.52 %, H 6.41 %, N 6.36 %, found: C 54.64 %, H 6.51 %, N 6.35 %. Methyl 3-(N-(2-methoxy-4-morpholinophenyl)sulfamoyl)thiophene-2-carboxylate (10 p): Compound 10 p was prepared according to general procedure B from 2-methoxy-4-morpholinoaniline as a yellow solid (164 mg, 80 % over two steps); mp: 144.1145.4 8C; 1 H NMR (400 MHz, [D6]DMSO): d = 8.56 (br s, 1 H, SO2NH), 7.88 (d, 3 JH,H = 5.3 Hz, 1 H, 5-H), 7.30 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 7.12 (m, 1 H, 6-H), 6.446.41 (m, 2 H, 3-H, 5-H), 3.93 (s, 3 H, CO2CH3), 3.69 (t, 3 JH,H = 4.8 Hz, 4 H, 3-H2, 5-H2), 3.52 (s, 3 H, OCH3), 3.06 ppm (t, 3 JH,H = 4.8 Hz, 4 H, 2-H2, 6-H2); 13C NMR (100 MHz, [D6]DMSO): d = 160.9 (1C, CO2CH3), 153.7 (1C, C-2), 151.1 (1C, C-4), 144.7 (1C, C-3), 132.2 (1C, C-2), 132.2 (1C, C-5), 131.1 (1C, C-4), 127.4 (1C, C-6), 116.4 (1C, C-1), 107.1 (1C, C-5), 99.6 (1C, C-3), 66.6 (2C, C-3, C-5), 56.0 (1C, OCH3), 53.7 (1C, CO2CH3), 48.8 ppm (2C, C-2, C-6); MS (ESI): m/z (%): 208 (18) [MC6H4O4S2] + , 413 (100) [M + H] + , 435 (85) [M + Na] + , 825 (31) [2M + H] + , 847 (48) [2M + Na] + ; HRMS-EI: m/z [M] + calcd for C17H20N2O6S2 : 412.076280, found: 412.074783; Anal. calcd for C17H20N2O6S2 : C 49.50 %, H 4.89 %, N 6.79 %, found: C 49.63 %, H 4.90 %, N 6.66 %. Methyl 3-(N-(4-(hexylamino)-3-methoxyphenyl)sulfamoyl)thiophene-2-carboxylate (11): Compound 11 was prepared according to general procedure A from N1-hexyl-2-methoxybenzene-1,4-diamine (500 mg, 2.25 mmol) as a brown oil (220 mg, 23 %); 1H NMR (400 MHz, CDCl3): d = 7.96 (s, 1 H, SO2NH), 7.42 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.38 (d, 3JH,H = 5.0 Hz, 1 H, 4-H), 6.68 (d, 4JH,H = 2.1 Hz, 1 H, 2-H), 6.42 (dd, 3JH,H = 8.3 Hz, 4JH,H = 2.3 Hz, 1 H, 6-H), 6.33 (d, 3JH,H = 8.3 Hz, 1 H, 5-H), 4.08 (br s, 1 H, NH), 4.01 (s, 3 H, CO2CH3), 3.77 (s, 3 H, OCH3), 3.01 (t, 3JH,H = 7.1 Hz, 2 H, 1-H2), 1.59 (quint, 3JH,H = 7.1 Hz, 2 H, 2-H2), 1.421.33 (m, 2 H, 5-H2), 1.321.23 (m, 4 H, 3H2, 4-H2), 0.910.86 ppm (m, 3 H, 6-H3); 13C NMR (100 MHz, CDCl3): d = 161.5 (1C, CO2CH3), 146.8 (1C, C-3), 144.7 (1C, C-3), 137.4 (1C, C-2), 132.0 (1C, C-5), 130.8 (1C, C-4), 130.2 (1C, C-4), 124.7 (1C, C-1), 116.8 (1C, C-5), 109.0 (1C, C-6), 107.2 (1C, C-2), 55.6 (1C, OCH3), 53.4 (1C, CO2CH3), 43.7 (1C, C-1), 31.7 (1C, C-2), 29.5 (1C, C-4), 27.0 (1C, C-3), 22.7 (1C, C-5), 14.1 ppm (1C, C-6); MS (EI): m/z (%): 427 (8), 426 (24) [M] + , 394 (5), 222 (22), 221 (100), 137 (6); HRMS-EI: m/z [M] + calcd for C19H26N2O5S2 : 426.1283, found: 426.1290; Anal. calcd for C19H26N2O5S2 : C 53.50 %, H 6.14 %, N 6.57 %, found: C 52.41 %, H 6.19 %, N 6.60 %.
ChemMedChem 2012, 7, 159 170
168
www.chemmedchem.org

Methyl 3-(N-(4-(hexylamino)phenyl)sulfamoyl)thiophene-2-carboxylate (12): Compound 12 was prepared according to general procedure A from N1-hexylbenzene-1,4-diamine (192 mg, 1.00 mmol) as a dark green solid (124 mg, 31 %); mp: 100101 8C; 1H NMR (400 MHz, CDCl3): d = 7.90 (s, 1 H, SO2NH), 7.35 (d, 3JH,H = 5.3 Hz, 1 H, 5-H), 7.30 (d, 3JH,H = 5.3 Hz, 1 H, 4-H), 6.856.79 (m, 2 H, 2-H, 6-H), 6.386.33 (m, 2 H, 3-H, 5-H), 3.94 (s, 3 H, CO2CH3), 2.94 (t, 3JH,H = 7.1 Hz, 2 H, 1-H2), 1.49 (quint, 3JH,H = 7.1 Hz, 2 H, 2-H2), 1.331.25 (m, 2 H, 5-H2), 1.251.18 (m, 4 H, 3-H2, 4-H2), 0.81 ppm (t, 3JH,H = 6.6 Hz, 3 H, 6-H2); 13C NMR (100 MHz, CDCl3): d = 161.5 (1C, CO2CH3), 147.4 (1C, C-4), 144.7 (1C, C-3), 131.9 (1C, C-4), 130.7 (1C, C-2), 130.3 (1C, C-5), 125.8 (2C, C-2, C-6), 125.3 (1C, C-1), 112.9 (2C, C-3, C-5), 53.4 (1C, CO2CH3), 44.1 (1C, C-1), 31.7 (1C, C2), 29.5 (1C, C-4), 26.9 (1C, C-3), 22.7 (1C, C-5), 14.1 ppm (1C, C6); MS (EI): m/z (%): 396 (30) [M] + , 192 (33), 191 (100), 121 (21), 111 (12); HRMS-EI: m/z [M] + calcd for C18H24N2O4S2 : 396.1178, found: 396.1155; Anal. calcd for C18H24N2O4S2 : C 54.52 %, H 6.10 %, N 7.06 %, found: C 54.56 %, H 6.65 %, N 7.04 %.
europa.eu/docs/en_GB/document_library/Press_release/2010/09/ WC500096996.pdf (Last accessed: September 29, 2011). a) Y.-X. Wang, C.-H. Lee, S. Tiep, R. T. Yu, J. Ham, H. Kang, R. M. Evans, Cell 2003, 113, 159 170; b) E. Bedu, W. Wahli, B. a. Desvergne, Expert Opin. Ther. Targets 2005, 9, 861 873; c) B. Desvergne, L. Michalik, W. Wahli, Physiol. Rev. 2006, 86, 465 514. a) C. Bastie, D. Holst, D. Gaillard, C. Jehl-Pietri, P. A. Grimaldi, J. Biol. Chem. 1999, 274, 21920 21925; b) A. D. Burdick, D. J. Kim, M. A. Peraza, F. J. Gonzalez, J. M. Peters, Cell. Signalling 2006, 18, 9 20. R. Mller, M. Rieck, S. Mller-Brsselbach, PPAR Res. 2008, 614852. a) L. Michalik, B. Desvergne, N. S. Tan, S. Basu-Modak, P. Escher, J. Rieusset, J. M. Peters, G. Kaya, F. J. Gonzalez, J. Zakany, D. Metzger, P. Chambon, D. Duboule, W. Wahli, J. Cell Biol. 2001, 154, 799 814; b) M. Romanowska, L. Reilly, C. N. Palmer, M. C. Gustafsson, J. Foerster, PLoS One 2010, 5, e9701. M. Romanowska, N. al Yacoub, H. Seidel, S. Donandt, H. Gerken, S. Phillip, N. Haritonova, M. Artuc, S. Schweiger, W. Sterry, J. Foerster, J. Invest. Dermatol. 2008, 128, 110 124. N. Di-Po, N. S. Tan, L. Michalik, W. Wahli, B. Desvergne, Mol. Cell 2002, 10, 721 733. a) F. Varnat, B. B.-T. Heggeler, P. Grisel, N. Boucard, I. Corthsy-Theulaz, W. Wahli, B. Desvergne, Gastroenterology 2006, 131, 538 553; b) S. Kanakasabai, W. Chearwae, C. C. Walline, W. Iams, S. M. Adams, J. J. Bright, Immunology 2010, 130, 572 588. a) K. S. Kilgore, A. N. Billin, Curr. Opin. Invest. Drugs 2008, 9, 463 469; b) S. Mller-Brsselbach, M. Kmhoff, M. Rieck, W. Meissner, K. Kaddatz, J. Adamkiewicz, B. Keil, K. J. Klose, R. Moll, A. D. Burdick, J. M. Peters, R. Mller, EMBO J. 2007, 26, 3686 3698; c) A. Abdollahi, C. Schwager, J. Kleeff, I. Esposito, S. Domhan, P. Peschke, K. Hauser, P. Hahnfeldt, L. Hlatky, J. Debus, J. M. Peters, H. Friess, J. Folkman, P. E. Huber, Proc. Natl. Acad. Sci. USA 2007, 104, 12890 12895. a) K. Kang, S. M. Reilly, V. Karabacak, M. R. Gangl, K. Fitzgerald, B. Hatano, C.-H. Lee, Cell Metab. 2008, 7, 485 495; b) J. I. Odegaard, R. R. Ricardo-Gonzalez, A. Red Eagle, D. Vats, C. R. Morel, M. H. Goforth, V. Subramanian, L. Mukundan, A. W. Ferrante, A. Chawla, Cell Metab. 2008, 7, 496 507; c) Y. Barak, D. Liao, W. He, E. S. Ong, M. C. Nelson, J. M. Olefsky, R. Boland, R. M. Evans, Proc. Natl. Acad. Sci. USA 2002, 99, 303 308; d) R. A. Gupta, D. Wang, S. Katkuri, H. Wang, S. K. Dey, R. N. DuBois, Nat. Med. 2004, 10, 245 247; e) D. Wang, H. Wang, Q. Shi, S. Katkuri, W. Walhi, B. Desvergne, S. K. Das, S. K. Dey, R. N. DuBois, Cancer Cell 2004, 6, 285 295; f) D. J. Kim, T. E. Akiyama, F. S. Harman, A. M. Burns, W. Shan, J. M. Ward, M. J. Kennett, F. J. Gonzalez, J. M. Peters, J. Biol. Chem. 2004, 279, 23719 23727; g) M. T. Bility, M. K. Devlin-Durante, N. Blazanin, A. B. Glick, J. M. Ward, B. H. Kang, M. J. Kennett, F. J. Gonzalez, J. M. Peters, Carcinogenesis 2008, 29, 2406 2414; h) A. Tesse, R. Andriantsitohaina, T. Ragot, PPAR Res. 2009, 745821. P. Germain, P. Chambon, G. Eichele, R. M. Evans, M. A. Lazar, M. Leid, A. R. De Lera, R. Lotan, D. J. Mangelsdorf, H. Gronemeyer, Pharmacol. Rev. 2006, 58, 760 772. S. Yu, J. K. Reddy, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2007, 1771, 936 951. V. Zoete, A. Grosdidier, O. Michielin, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2007, 1771, 915 925. a) A. Stefanski, L. Majkowska, Expert Opin. Ther. Pat. 2006, 16, 1713 1733; b) M. L. Sznaidman, C. D. Haffner, P. R. Maloney, A. Fivush, E. Chao, D. Goreham, M. L. Sierra, C. LeGrumelec, H. E. Xu, V. G. Montana, M. H. Lambert, T. M. Willson, W. R. Oliver, D. D. Sternbach, Bioorg. Med. Chem. Lett. 2003, 13, 1517 1521; c) J. Berger, M. D. Leibowitz, T. W. Doebber, A. Elbrecht, B. Zhang, G. Zhou, C. Biswas, C. A. Cullinan, N. S. Hayes, Y. Li, M. Tanen, J. Ventre, M. S. Wu, G. D. Berger, R. Mosley, R. Marquis, C. Santini, S. P. Sahoo, R. L. Tolman, R. G. Smith, D. E. Moller, J. Biol. Chem. 1999, 274, 6718 6725; d) R. A. Ngala, C. J. Stocker, A. G. Roy, D. Hislop, E. Wargent, R. Bell, D. G. Hassall, J. D. Harling, A. N. Billin, T. M. Willson, J. R. S. Arch, M. A. Cawthorne, Diabetes Obes. Metab. 2011, 13, 455 464; e) J.-M. Ye, J. Tid-Ang, N. Turner, X.-Y. Zeng, H.-Y. Li, G. J. Cooney, E. M. Wulff, P. Sauerberg, E. W. Kraegen, br J. Pharmacol. 2011, 163, 556 566. M. Rieck, W. Meissner, S. Ries, S. Mller-Brsselbach, R. Mller, Mol. Pharmacol. 2008, 74, 1269 1277. J. M. Peters, F. J. Gonzalez, Biochim. Biophys. Acta, Rev. Cancer 2009, 1796, 230 241.
[14]
[15]
[16] [17]
[18]
[19] [20]
Acknowledgements
[21]
This work is supported by the Deutsche Forschungsgemeinschaft (SFBTR17/A3) and the LOEWE-Schwerpunkt Tumor and Inflammation of the state of Hesse (Germany). Keywords: GSK0660 inhibitors NR1C2 nuclear receptors PPARb/d structureactivity relationships
[1] a) H. Gronemeyer, J.-A. Gustafsson, V. Laudet, Nat. Rev. Drug Discovery 2004, 3, 950 964; b) J. T. Moore, J. L. Collins, K. H. Pearce, ChemMedChem 2006, 1, 504 523. [2] Nuclear Receptors Nomenclature Committee, Cell 1999, 97, 161 163. [3] a) L. Michalik, J. Auwerx, J. P. Berger, V. K. Chatterjee, C. K. Glass, F. J. Gonzalez, P. A. Grimaldi, T. Kadowaki, M. A. Lazar, S. ORahilly, C. N. A. Palmer, J. Plutzky, J. K. Reddy, B. M. Spiegelman, B. Staels, W. Wahli, Pharmacol. Rev. 2006, 58, 726 741; b) T. M. Willson, P. J. Brown, D. D. Sternbach, B. R. Henke, J. Med. Chem. 2000, 43, 527 550. [4] a) C. Dreyer, G. Krey, H. Keller, F. Givel, G. Helftenbein, W. Wahli, Cell 1992, 68, 879 887; b) O. Braissant, F. Foufelle, C. Scotto, M. Daua, W. Wahli, Endocrinology 1996, 137, 354 366. [5] S. Mandard, M. Mller, S. Kersten, Cell. Mol. Life Sci. 2004, 61, 393 416. [6] a) S. Kersten, J. Seydoux, J. M. Peters, F. J. Gonzalez, B. a. Desvergne, W. Wahli, J. Clin. Invest. 1999, 103, 1489 1498; b) J. K. Reddy, T. Hashimoto, Annu. Rev. Nutr. 2001, 21, 193 230. [7] B. Staels, N. Vu-Dac, V. A. Kosykh, R. Saladin, J. C. Fruchart, J. Dallongeville, J. Auwerx, J. Clin. Invest. 1995, 95, 705 712. [8] N. Vu-Dac, K. Schoonjans, V. Kosykh, J. Dallongeville, J. C. Fruchart, B. Staels, J. Auwerx, J. Clin. Invest. 1995, 96, 741 750. [9] B. Staels, W. Koenig, A. Habib, R. Merval, M. Lebret, I. P. Torra, P. Delerive, A. Fadel, G. Chinetti, J.-C. Fruchart, J. Najib, J. Maclouf, A. Tedgui, Nature 1998, 393, 790 793. [10] a) A. Chawla, E. J. Schwarz, D. D. Dimaculangan, M. A. Lazar, Endocrinology 1994, 135, 798 800; b) P. Tontonoz, E. Hu, R. A. Graves, A. I. Budavari, B. M. Spiegelman, Genes Dev. 1994, 8, 1224 1234; c) P. Tontonoz, R. A. Graves, A. I. Budavari, H. Erdjument-Bromage, M. Lui, E. Hu, P. Tempst, B. M. Spiegelman, Nucleic Acids Res. 1994, 22, 5628 5634. [11] a) M. Stumvoll, H.-U. Hring, Ann. Med. 2002, 34, 217 224; b) F. Chiarelli, D. Di Marzio, Vasc. Health Risk Manage. 2008, 4, 297 304. [12] a) V. Ratziu, S. Caldwell, B. A. Neuschwander-Tetri, Hepatology 2010, 52, 2206 2215; b) R. Simo, A. Rodriguez, E. Caveda, Curr. Drug Saf. 2010, 5, 234 244. [13] European Medicines Agency (EMA) press release: European Medicines Agency recommends suspension of Avandia, Avandamet and Avaglim (September 23, 2010), Ref: EMA/585784/2010; http://www.ema.
[22]
[23]
[24] [25] [26]
[27] [28]
ChemMedChem 2012, 7, 159 170
www.chemmedchem.org
169
MED
[29] N. T. Zaveri, B. G. Sato, F. Jiang, J. Calaoagan, K. Laderoute, B. J. Murphy, Cancer Biol. Ther. 2009, 8, 1252 1261. [30] a) B. G. Shearer, R. W. Wiethe, A. Ashe, A. N. Billin, J. M. Way, T. B. Stanley, C. D. Wagner, R. X. Xu, L. M. Leesnitzer, R. V. Merrihew, T. W. Shearer, M. R. Jeune, J. C. Ulrich, T. M. Willson, J. Med. Chem. 2010, 53, 1857 1861; b) P. S. Palkar, M. G. Borland, S. Naruhn, C. H. Ferry, C. Lee, U. H. Sk, A. K. Sharma, S. Amin, I. A. Murray, C. R. Anderson, G. H. Perdew, F. J. Gonzalez, R. Mller, J. M. Peters, Mol. Pharmacol. 2010, 78, 419 430. [31] B. G. Shearer, D. J. Steger, J. M. Way, T. B. Stanley, D. C. Lobe, D. A. Grillot, M. A. Iannone, M. A. Lazar, T. M. Willson, A. N. Billin, Mol. Endocrinol. 2008, 22, 523 529. [32] A. Maderna, J.-M. Vernier, D. Barawkar, V. Chamakura, H. El Abdellaoui, Z. Hong, (Valeant Research & Development, Costa Mesa, CA, USA), Int. Pat. Appl. WO 2007/014011 A2, 2007. [33] S. Naruhn, P. M. Toth, T. Adhikary, K. Kaddatz, V. Pape, S. Drr, G. Klebe, S. Mller-Brsselbach, W. E. Diederich, R. Mller, Mol. Pharmacol. 2011, in press; DOI: 10.1124/mol.111.074039.
[34] C. A. Lipinski, F. Lombardo, B. W. Dominy, P. J. Feeney, Adv. Drug Delivery Rev. 2001, 46, 3 26. [35] J.-H. Zhang, T. D. Y. Chung, K. R. Oldenburg, J. Biomol. Screening 1999, 4, 67 73. [36] a) T. Fauti, S. Mller-Brsselbach, M. Kreutzer, M. Rieck, W. Meissner, U. Rapp, H. Schweer, M. Kmhoff, R. Mller, FEBS J. 2006, 273, 170 179; b) V. Jrme, R. Mller, Hum. Gene Ther. 1998, 9, 2653 2659; c) S. Naruhn, W. Meissner, T. Adhikary, K. Kaddatz, T. Klein, B. Watzer, S. Mller-Brsselbach, R. Mller, Mol. Pharmacol. 2010, 77, 171 184. [37] S. Gehrke, V. Jrme, R. Mller, Gene 2003, 322, 137 143.
Received: August 25, 2011 Published online on October 24, 2011
170
www.chemmedchem.org
ChemMedChem 2012, 7, 159 170
Eureka!
Analogue-Based Drug Discovery II

Edited by Jnos Fischer and C. Robin Ganellin.
Wiley-VCH, Weinheim 2010. 564 pp., hardcover $ 230.00.ISBN 978-3-52732-549-8
Drug discovery has undergone a significant paradigm shift in the latter part of the 20th Century, continuing to the present. Early medicinal chemistry was driven by organic synthesis, whereby chemists prepared structural analogues as a means to optimize a drug candidate, especially when the mechanism of action was not clear. As pharmacology evolved and the understanding of biological mechanisms improved, drug candidates were increasingly identified through the use of in vitro assays and optimized on ADME properties in addition to potency. At the end of the 20th Century, in an effort to find new chemical entities (NCEs) in short order, the field saw the rise and fall of combinatorial chemistry. At present, researchers continue to find new avenues for identifying NCEs, but continue to learn from experience, namely analoguebased drug discovery. This approach was highlighted in the first volume of Analogue-Based Drug Discovery, published by Wiley-VCH in 2006. This book focused on improving a lead compound by preparing and testing analogues to find a drug with improved potency and properties. The first volume described several analogue classes, and also included a chapter in which privileged structures of the pharmaceutical industry were discussed. Analogue-Based Drug Discovery II is a companion book to the first volume. The
ChemMedChem 2012, 7, 171 172
second volume includes examples of full (structural) analogues, and also contains examples of pharmacological analogues: structurally distinct compounds with the same pharmacology. Like the first volume, this book is organized into three sections, beginning with the general aspects of analogue-based drug discovery (Chapters 14), moving into analogue classes (Chapters 513), and cumulating in case studies (Chapters 14 21). For the novice researcher, the first section of the book provides a general overview of the principles of drug discovery that are not necessarily regulated specifically to the analogue-based approach. This section also highlights standalone drugs, which are drugs that have no analogues. While this section is very informative, its utility in the context of a book on analogue-based drug discovery is misleading. The second section of the book details various analogue classes, including DPP IV inhibitors, PDE5 inhibitors, rifamycins, monoterpenoid indole alkaloids, anthracyclines, paclitaxel and epothilone analogues, selective serotonin reuptake inhibitors, muscarinic receptor antagonists, and b-adrenoceptor agonists. These chapters delve into the biology associated with each analogue class, discussing the mechanism of action, drug discovery chemistry, and pharmacological/clinical data. The third section of the book is quite similar to the second in that each chapter has parallel content; however, the chapters describe case studies of a single compound rather than an analogue class. While the sections are separate, a streamlined approach to the presentation of the wealth of information encompassed in these chapters would greatly benefit the reader. The chapters in these sections are heterogeneous with regards to the depth of information on each analogue class or compound. For example, the authors of the DPP IV inhibitor chapter pro-
vide a concise discussion of general diabetes biology, in vitro assays, and animal models. The chapter develops the history of the analogue-based drug discovery effort of DPP IV inhibitors from early inhibitors to Sitagliptin and Vildagliptin. The authors also include a synopsis of the pharmacological data comparing Vildagliptin, Sitagliptin, and Alogliptin. The application of the format the authors chose for this chapter to the remainder of the chapters in this book would greatly enhance the readers understanding of analogue-based drug discovery as it pertains to analogue classes and case studies. In the latter section of the book, in addition to discussing the drug discovery program on a particular case study, the synthesis of several analogues is also highlighted. While some readers may be interested in organic synthesis, the broader audience may find this section misplaced within the context of the book. This volume provides an excellent overview of several biological targets, compound classes, and illustrations of problem solving in medicinal chemistry for chemists in drug discovery who are beginning to establish their careers. Analogue-Based Drug Discovery II would also be a resource for more established researchers who require a comprehensive analysis of case studies or compound classes. Additionally, the book is easily navigated by the table of contents or index. In addition to streamlining the content of the last two sections of the book, consistent formatting with regards to the chemical structures would be beneficial for the reader. Dr. Heather E. Burks, Dr. Stefan Peukert Novartis Institutes for BioMedical Research, Cambridge, MA (USA) DOI: 10.1002/cmdc.201100479
171
MED
Antiviral Drugs: From Basic Discovery Through Clinical Trials
Edited by Wieslaw M. Kazmierski.
John Wiley & Sons, Hoboken 2011. 428 pp., hardcover $ 149.95.ISBN 978-0-47045-563-0
There has been a major focus in many pharmaceutical companies over the past three decades on antivirals, spurred by the advent of human immunodeficiency virus (HIV) with its major medical and social consequences. Many viral infections can be prevented by vaccination, but there remain serious infections including HIV, hepatitis C virus (HCV) and others, that yet cannot. Efforts to make new antivirals for persistent infections are vital due to the inevitable emergence of resistance. This book covers the discovery and development of small molecules for several viruses, with the number of chapters on each virus approximating to the importance of the area. So for HIV, there are a total of 14 chapters covering protease, non-nucleoside reverse transcriptase (RT), nucleoside RT, entry, and integrase inhibitors; while for HCV, nine chapters on protease, polymerase, and other mechanisms are given. There are three chapters covering respiratory syncytial virus, and finally one each on influenza, hepatitis B, and cytomegalovirus. Each chapter mainly tells the story of the
discovery of a single molecule that has gone into development and in many cases onto the market. The remit of each chapter to cover the discovery of a drug through its clinical trials is a large one, and to do it comprehensively would be a major undertaking. Instead, the authors of each chapter provide their own balance of what they consider to be important in the whole process, depending, perhaps, on their areas of expertise. So, for example, the chapter on apricitabine has little to say on its discovery, but a quite detailed description of the process chemistry is given before describing the clinical trials. The chapter on atazanavir goes into considerable detail on its preclinical toxicology, with little on its discovery. The story on boceprevir is largely concerned with its discovery. There are chapters, for example that on entecavir, which provide a balanced story covering the whole process including discovery, synthetic route, toxicology, clinical trials, and resistance. All of the chapters cover or touch on resistance, as expected in a therapeutic area where this is such a major concern. This comment about the emphasis on particular areas for some of the chapters is not a criticismfor a person newly coming into the area of antiviral research, the book serves to highlight the challenges to be overcome going from a target to a drug. It is possible to read each chapter in isolation as all are selfcontained stories, and virtually all are well written, interesting, and will educate. Having said this, the most benefit to the reader working in the area or interested in it would come from reading the book in its entirety.
Since each chapter mainly tells the story of a single drug or target, what is lacking for a reader not working in the area is a separate overview of the epidemiology and treatment options for the diseases. Each chapter does try to put its story into context in that way, but tends to come at it from its own point of view. What would have added to the book would be a separate chapter (or perhaps more than one) on the extent of the diseases covered in the book, the available treatments, and the challenges that remain to be addressed. I very much enjoyed this book and learned greatly from it. The vital importance of addressing resistanceeither by novel mechanisms or engineering high barriers, or preferably bothcomes across strongly. The challenges of putting together potency, resistance profile, appropriate pharmacokinetics and safety in a single molecule are made clear. Some of the pitfalls in toxicology and clinical trials are pointed out, and some examples are given of the need for a scalable process for manufacture. What also comes across is the need for continued efforts in the area of antivirals as new treatments for existing diseases become required, and perhaps new viruses emerge in the human population. The book is a valuable resource for researchers working in the discovery and early development of antiviral agents, and also for those simply interested in this area Dr. Michael Rowley AstraZeneca (Sweden) DOI: 10.1002/cmdc.201100421
172
www.chemmedchem.org
ChemMedChem 2012, 7, 171 172
> Weekly > About 15.000 abstracts per year > Covering some 100 international journals
Current Awareness
Abstracts of the most important reactions in organic chemistry > New and improved methods and syntheses > New catalysts and reagents > Syntheses of natural products of general importance > Syntheses of novel organoelement compounds > Fundamental key steps > Stereochemical outcome Linked references Detailed reaction schemes
NEW
ALL IN ONE
> Complete ChemInform issue in a single pdf > Easy e-browsing by subject bookmarks > Integrated links to original literature
EDITORS CHOICE
> The Editors personal research highlight of the week.
www.cheminform.wiley-vch.de
2788431102_bu
Acids in Chemistry, Life Sciences, and Biotechnology

Hans Brckner and Noriko Fujii (Eds.) ISBN: 978-3-90639-065-9 Hardcover Publication date: December 2010
D-Amino
The increasing importance of D-Amino acids resulted in the foundation of the D-Amino Acid Research Society in Japan, followed by the First International Conference on D-Amino Acid Research This new title presents a collection of state-of-the art research in a eld that has signicant implications in relation to health care, research in age-related and neuronal diseases, pharmaceutical chemistry, and food science
The majority of the contributions published in this new volume result from the rst international conference and reect the longstanding interest of the Guest Editors, and their co-workers and collaborators in the eld The nal artefact provides a wealth of knowledge for organic, pharmaceutical, medicinal, and protein chemists as well as food and medical scientists
For further information visit www.wiley.com
VERLAG HELVETICA CHIMICA ACTA
2/2012
REVIEWS
A Decade of the Human Genome SequenceHow Does the Medicinal Chemist Benefit? A. Brunschweiger, J. Hall*
COMMUNICATIONS
Remarkable Stability and Cytostatic Effect of a Quercetin Conjugate, 3,7-Bis-O-Pivaloxymethyl (POM) Quercetin Immunosuppressive Effects of Subglutinol Derivatives A Synthetic Lipid A Mimetic Modulates Human TLR4 Activity Application of Barluenga Boronic Coupling (BBC) to the Parallel Synthesis of Drug-like and Drug Fragment-like Molecules M. K. Kim, K.-S. Park, Y. Chong* W.-G. Lee, W.-S. Kim, S.-G. Park, H. Kim, J. Hong, H. Ko, Y.-C. Kim* M. Piazza, V. Calabrese, G. Damore, R. Cighetti, T. Gioannini,* J. Weiss,* F. Peri* S. Nakagawa,* K. A. Bainbridge, K. Butcher, D. Ellis, W. Klute, T. Ryckmans*
FULL PAPERS
Galloflavin (CAS 568-80-9): A Novel Inhibitor of Lactate Dehydrogenase M. Manerba, M. Vettraino, L. Fiume, G. Di Stefano,* A. Sartini, E. Giacomini, R. Buonfiglio, M. Roberti, M. Recanatini L. Rocamora-Reverte, E. Carrasco-Garca, J. Ceballos-Torres, S. Prashar, G. N. Kaluderovic, J. A. Ferragut,* S. Gmez-Ruiz* M. Bhm, T. Khl, K. Hardes, R. Coch, C. Arkona, B. Schlott, T. Steinmetzer, D. Imhof* C. C. Williams, S. H. Thang, T. Hantke, U. Vogel, P. H. Seeberger, J. Tsanaktsidis,* B. Lepenies* A. Murza, A. Parent, E. Besserer-Offroy, H. Tremblay, F. Karadereye, N. Beaudet, R. Leduc, P. Sarret, . Marsault*
Study of the Anticancer Properties of Tin(IV) Carboxylate Complexes on a Panel of Human Tumor Cell Lines Synthesis and Functional Characterization of Tridegin and Its Analogues: Inhibitors and Substrates of Factor XIIIa RAFT-Derived PolymerDrug Conjugates: Poly(hydroxypropyl methacrylamide) (HPMA)7-Ethyl-10-hydroxycamptothecin (SN-38) Conjugates Elucidation of the StructureActivity Relationships of Apelin: Influence of Unnatural Amino Acids on Binding, Signaling, and Plasma Stability
Dont risk missing out! Subscribe to ChemMedChem. Contact cs-germany@wiley.com
ChemMedChem 2012, 7, 175
175
The back cover picture shows Eis, the enzyme that converts aminoglycoside antibiotics from their active (green) to their inactive (red) form, a process that causes drug resistance in tuberculosis (TB). The compounds shown are the first examples of Eis inhibitors, which could find application in combination therapies with aminoglycosides to treat TB. For more details, please see the Communication by Sylvie GarneauTsodikova et al. on p. 73 ff.

Complete Print

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Complete Print

Hochgeladen von

Copyright:

Verfügbare Formate

CHEMGX 7 (1) 1176 (2012) ISSN 1860-7179 Vol. 7 No.

Discover this journal online at:

Quality Research. Outstanding Publications.

With the excitement of the International Year of Chemistry

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

A More Discerning Readership

Medicinal chemistry research is progressing, changing and

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Two New Family Members

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

As Editors, we serve the community and its members, which

Dr. Natalia Ortzar Editor-in-Chief ChemMedChem

Dr. Scott D. Williams Sr. Assoc. Editor ChemMedChem

Prof. Rainer Metternich F. HoffmannLa Roche Ltd (Switzerland)

Prof. Giorgio Tarzia Universit degli Studi di Urbino (Italy)

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Issue 12, 2011, was published online on December 2, 2011.

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

International Advisory Board

ChemMedChem succeeds IL FARMACO, ` a journal of the Societa Chimica Italiana

16 societies G 11 journals G 1 network

E-magazine for chemists

C Closely associated with

Typesetting, Printing and Binding:

Johannes-Gutenberg-Strasse 13 97199 Ochsenfurt-Hohestadt www.triltsch.de

Paper: acid-free Printed in the Federal Republic of Germany

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Angew. Chem. Int. Ed. DOI: 10.1002/anie.201103672

Chem. Eur. J. DOI: 10.1002/chem.201103090

Chem. Asian J. DOI: 10.1002/asia.201100632

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

... ON OUR SISTER JOURNALS

ChemBioChem DOI: 10.1002/cbic.201100451

ChemPhysChem DOI: 10.1002/cphc.201100446

ChemSusChem DOI: 10.1002/cssc.201100075

ChemCatChem DOI: 10.1002/cctc.201100281

Mushroom Extraction Artefacts

Eur. J. Org. Chem. DOI: 10.1002/ejoc.201101215

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Concise Notice to Authors

Preparing Your Manuscript for Submission

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

A Few Other Things

Full Details Are Available Online

Corrections: Galley Proofs & Beyond

(+49) 6201-606-142 (+49) 6201-606-331 or -328 chemmedchem@wiley.com http://www.chemmedchem.org

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

You are invited

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

[a] T. lschlger, unpublished results. [b] L. Sologub, G. Pradel, unpublished results.

Figure 1. Structure of dequalinium chloride.

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Targets of QACs in Bacteria: More Than Nonspecific Interactions with Membranes

 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 5. Structure of MT02.

QACs in the Treatment of Malaria

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim