Sie sind auf Seite 1von 10

Journal of Neurochemistry, 2005, 95, 15751584

doi:10.1111/j.1471-4159.2005.03477.x

Differential phospholipase C activation by phenylalkylamine serotonin 5-HT2A receptor agonists


Jason C. Parrish, Michael R. Braden, Emily Gundy and David E. Nichols
Department of Medicinal Chemistry and Molecular Pharmacology, School of Pharmacy and Pharmaceutical Sciences, Purdue University, West Lafayette, Indiana, USA

Abstract Experiments compared a series of phenethylamine hallucinogens with their phenylisopropylamine analogues for binding afnity and ability to stimulate serotonin 5-HT2A receptor-mediated hydrolysis of phosphatidyl inositol in cells expressing cloned rat and human 5-HT2A receptors. The () phenylisopropylamine analogues had signicantly higher intrinsic activities for 5-HT2A receptor-mediated hydrolysis of phosphatidyl inositol compared to their phenethylamine analogues. With respect to the effects of the stereochemistry of the phenylisopropylamines, those with the (R) absolute conguration at the alpha carbon had higher intrinsic activities for hydrolysis of phosphatidyl inositol in a cell line expressing the human 5-HT2A receptor compared to those with the (S) absolute conguration. In virtual docking studies comparing

the (R)- and (S)-phenylisopropylamines with their phenethylamine analogues, there were distinct differences in the orientations of key ligand binding domain residues that have been identied as important by previous mutagenesis studies. In conclusion, our data support the hypothesis that phenylisopropylamines have higher hallucinogenic potency than their phenethylamine analogues primarily because they have higher intrinsic activities at 5-HT2A receptors. Although virtual ligand binding led to signicant perturbations of certain key residues, our results emphasize the conclusion reached by others that overall three-dimensional structural microdomains within the receptor must be considered. Keywords: 5-HT2A, hallucinogen, phenethylamine, phenylisopropylamine, phospholipase C, serotonin. J. Neurochem. (2005) 95, 15751584.

Since Arthur Heffters 1897 discovery that mescaline (3,4,5trimethoxyphenethylamine) was the active constituent of the hallucinogenic peyote cactus, Lophophora williamsii (Heffter 1898; Perrine 2001), the biochemical mechanism(s) responsible for altered states of consciousness (ASC) induced by hallucinogens is still incompletely understood. Although today hallucinogens (psychedelics) remain largely pharmacological curiosities, much has been learned about their pharmacology in the past 100 years (Nichols 2004). All hallucinogens have in common relatively high afnity for the serotonin 5-HT2A receptor subtype, and it is now widely thought that the effects of hallucinogens are dependent on activation of brain 5-HT2A receptors, particularly in frontal cortex and thalamus (Nichols 2004). The recent demonstration that the hallucinogenic effect of the indoleamine psilocybin in human subjects is blocked by ketanserin, a selective 5-HT2A receptor antagonist, provides compelling evidence for a 5-HT2A receptor agonist-mediated mechanism of action (Vollenweider et al. 1998). Although mescaline is the least potent of the hallucinogens, it has served as the template for the synthesis of

numerous phenethylamine (PEA) analogues with various efcacies for producing ASC. In the earliest such report, adding an alpha-methyl group to the structure of mescaline provided the phenylisopropylamine (PIA) analogue,

Received April 30, 2005; revised manuscript received July 13, 2005; accepted July 23, 2005. Address correspondence and reprint requests to Dr David E. Nichols, Department of Medicinal Chemistry and Molecular Pharmacology, School of Pharmacy and Pharmaceutical Sciences, 575 Stadium Mall Drive, Purdue University, West Lafayette, IN 479072091, USA E-mail: drdave@pharmacy.purdue.edu Abbreviations used: 2C-B, 4-bromo-2,5-dimethoxyphenethylamine; 2C-I, 4-iodo-2,5-dimethoxyphenethylamine; 2C-T2, 4-ethylthio-2,5-dimethoxyphenethylamine; 2C-TFM, 4-triuoromethyl-2,5-dimethoxyphenethylamine; 5-HT, 5-hydroxytryptamine, serotonin; Aleph-2, 4-ethylthio-2,5-dimethoxyphenylisopropylamine; ASC, altered states of consciousness; DOB, 4-bromo-2,5-dimethoxyphenylisopropylamine; DOI, 4-iodo-2,5-dimethoxyphenylisopropylamine; DOTFM, 4-triuoromethyl-2,5-dimethoxyphenylisopropylamine; PEA, phenethylamine; PIA, phenylisopropylamine; PLC, phospholipase C; TFMy, 4-triuoromethyl-benzoditetrahydrofuranylisopropylamine; TM, transmembrane; TMA, trimethoxyphenylisopropylamine.

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

1575

1576 J. C. Parish et al.

3,4,5-trimethoxyamphetamine (TMA) (Hey 1947). This analogue had roughly twice the potency of mescaline (Shulgin et al. 1961). Following the synthesis of TMA, numreous ring-substituted PEAs and PIAs were prepared and tested for biological activity. In general, PIA analogues were found to be more potent and have longer durations of action when compared to PEA analogues (Glennon et al. 1983), but the reason(s) for the difference in hallucinogenic potency has not been elucidated. The addition of the alpha-methyl group increases hydrophobicity and also will likely retard metabolic deamination, but this explanation is not adequate to explain the potency differences, and fails to account for differences observed using in vitro experiments. A recent study has systematically examined the difference in efcacy between PEAs and their corresponding PIA analogues. In that work, voltage clamp techniques were used to measure current generated in response to phenylalkylamine hallucinogens using Xenopus laevis oocyte membranes transiently expressing the rat 5-HT2A receptor (Acuna-Castillo et al. 2002). PEAs were found to be virtually non-efcacious whereas their corresponding PIA analogues were reported to be weak partial agonists. The authors proposed that an antagonist or weak partial agonist mechanism of action for hallucinogens might still be tenable. More recently, using the same oocyte assay system, these workers have reported that PEAs are antagonists at the transiently expressed rat 5-HT2A receptor (Villalobos et al. 2004). They concluded that their results cast doubt on the notion that phenylalkylamines hallucinogens are full or partial 5-HT2A agonists. The fact that a 5-HT2A receptor antagonist blocks the effects of hallucinogens in man, however (Vollenweider et al. 1998), indicates that the Xenopus oocyte model may not reect

5-HT2A receptor-mediated signaling relevant to human psychopharmacology. Although the PEAs are achiral, addition of the methyl group to the alpha side chain carbon affords the PIAs, which have two optical isomers. It is a well-accepted tenet of pharmacology that the optical isomers (enantiomers) of biologically active molecules have different pharmacological properties. Most biological targets, often macromolecular proteins, are chiral, stereospecic and/or stereoselective in their actions. The difference in potency (the eudismic ratio) between the more active enantiomer (eutomer) and the less active enantiomer (distomer) can typically vary from several-fold to several hundred-fold, depending on the chemical structure. Of the PIA enantiomers, the (R)-isomers [(R)-()-PIAs, e.g. (R)-4-bromo2,5-dimethoxyphenylisopropylamine (DOB), Fig. 1] are more potent in vitro (Dyer et al. 1973; Cheng et al. 1974), in animal models (Harris et al. 1977; Young and Glennon 1996), and in their ability to induce ASC in humans (Shulgin 1973; Shulgin and Shulgin 1991). We reasoned that further study of the pharmacological differences between the enantiomers of selected PIAs and their corresponding PEAs might provide additional insight into the underlying mechanism of action for hallucinogenic phenethylamines. The present work focuses on the role of the alpha-methyl in PIAs in 5-HT2A receptor activation. We describe herein the pharmacological characterization of a small series of substituted phenethylamine 5-HT2A agonists (Fig. 1) in activating the phospholipase C (PLC) signaling pathway in both the cloned rat and human 5-HT2A receptors. We then docked and minimized one set of alpha-methyl enantiomers (R)-DOB and (S)-DOB, along with their achiral homologue 4bromo-2,5-dimethoxyphenethylamine (2C-B) (Fig. 1) into a homology model of the human 5-HT2A receptor that was

Fig. 1 Phenethylamine (PEA) and phenylisopropylamine (PIA) compounds used in this study. 4-Triuoromethyl-2,5-dimethoxyphenylisopropylamine (DOTFM) and 4ethylthio-2,5-dimethoxyphenylisopropylamine (Aleph-2) were not available as the pure enantiomers and were only tested as the racemic () materials.

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

Phenylalkylamines at the 5-HT2A receptor 1577

developed based on an in silico activated model of bovine rhodopsin (Chambers and Nichols 2002).

et al. 2003). [3H]Ketanserin and [125I]4-iodo-2,5-dimethoxyphenylisopropylamine (DOI) were used for saturation assays. Inositol phosphate accumulation After treatment with phenylalkylamine agonists, GF-6 or A20 cells were assayed for the accumulation of phosphoinositides as previously described in detail (Kurrasch-Orbaugh et al. 2003). These cells provided a comparison of both the rat and human 5HT2A receptor responses. Computational modeling/virtual docking Ligand structures were drawn and energy minimized (Powell method, 0.01 kcal/mol*A gradient termination, MMFF94s force eld, MMFF94 charges, 1000 maximum iterations) using the Sybyl modeling program (Tripos, St. Louis, MO, USA). The in silicoactivated homology model of the h5-HT2A receptor was prepared as previously described (Chambers and Nichols 2002). Virtual dockings of energy minimized ligands to the h5-HT2A receptor were performed using the program GOLD (Cambridge Crystallographic Data Center, Cambridge, UK) and scored using GOLDScore with default settings except for constraints. The GOLDScore tness algorithm was constrained to orientations containing ligandprotein interactions implicated by site-directed mutagenesis and previous modeling (Chambers and Nichols 2002) to possibly be essential for binding. Distance constraints of 0.20.3 nm (23 A) were set between the side chain carbonyl carbon of D155(3.32) and the amine nitrogen of the ligand (Wang et al. 1993; Kristiansen et al. 2000); from S159(3.36) and T160(3.37) to the ligand 2-position oxygen (Almaula et al. 1996); and between S239(5.43) and the ligand 5position oxygen (Johnson et al. 1997; Shapiro et al. 2000). The highest ranked docking output structures were merged with the h5HT2A receptor model and analyzed with Sybyl. Merged ligandreceptor structures were energy minimized as a subset based on the ligand molecule (aggregates set to monomers > 0.6 nm (6 A) radius from the ligand, monomers > 1.2 nm (12 A) radius ignored, Powell method, 0.1 kcal/mol*A gradient termination, MMFF94s force eld, MMFF94 charges, distance dielectric of 4, 1000 maximum iterations). Constraints for subsequent molecular dynamics simulations and minimizations in Sybyl were dened as above for GOLD; however, hydrogen bond constraints were dened as a distance range constraint of 0.10.2 nm (12 A) between each polar residues hydrogen and the appropriate oxygen atom on the ligand. Constrained molecular dynamics simulations were then run on the energy-minimized ligandreceptor structures (constraints, force eld, charges and dielectric as outlined above, aggregates as above plus backbone atoms, NTV ensemble at 300K, Boltzmann distribution of initial velocities, 5000 steps of 1 fs, and 5 fs snapshots). Structures with lowest potential energy after the rst 1000 fs equilibration period were then energy minimized as outlined above with dened constraints. Data analysis GraphPad Prism Software (GraphPad Prism version 4.00 for Windows, GraphPad Software, San Diego, California, USA, http:// www.graphpad.com) was used to calculate Ki values for radioligand displacement, EC50s, and intrinsic activities for PLC activation. The radioligand displacement data were best t to a one-site model and Ki values were calculated from EC50s using the concentration of

Experimental procedures
Materials All agonists used in this study were synthesized in our laboratory (e.g. see Nichols et al. 1973). Drugs were fully characterized by melting point, NMR, mass spectrometric, TLC, and elemental analysis. Enantiomers were analyzed for optical rotation and values matched those reported. Cell culture methods The GF-6 cell line stably expressing the rat 5-HT2A receptor (5500 fmol/mg) was the kind gift of Dr David Julius of the Department of Pharmacology, University of California, San Francisco, CA, USA. A549 cells, stably expressing the human 5-HT2A receptor (150 fmol/mg protein) (A20 cells) were the kind gift of Dr Ulrike Weyer-Czernilofsky of the Ernst Boehringer Institute, Bender and Co., Vienna, Austria. Both GF-6 and A20 cells were maintained in culture asks at 37C with 5% CO2 in Dulbeccos modied Eagles medium (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% (v/v) fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA, USA), 2 mM L-glutamine (Gibco, Grand Island, NY), 100 units/mL penicillin (Gibco, Grand Island, NY), 100 lg/mL streptomycin (Gibco), and 300 mg/L G-418 m (Sigma-Aldrich). Cells were passaged at 9095% conuency and discarded after approximately 30 passages or when they failed to respond to 5-HT. Establishing a stable high-expression human 5-HT2AR cell line The h5-HT2A receptor gene insert was excised from pcDNA3.1h5HT2AR (Guthrie cDNA resource center, Sayre, PA, USA; http:// www.cdna.org) with PmeI and XhoI and subcloned into the pBudCE4 vector (Invitrogen, Carlsbad, CA, USA), that was cut with NotI, blunted with T4 DNA polymerase in the presence of dNTPs, and then cut with XhoI (New England Biolabs, Beverly, MA, USA). Insert and vector were puried, quantied, ligated, and amplied utilizing standard methods. Orientation, identity, and sequence were veried by primer-directed sequencing (Retrogen, San Diego, CA, USA) utilizing custom synthesized EF-1a forward and BGH reverse primers (Integrated DNA Technologies, Coralville, IA). HEK-293 cells were transfected with 3.5 lg pBudCE4h5HT2AR utilizing the Fugene-6 lipofection reagent (Roche Biomolecules, Indianapolis, IN, USA) in a 3:1 (w : v) ratio, and stable clones were selected as per the package insert instructions in the presence of 60 lg/mL Zeocin (Invitrogen). Cells were maintained in culture asks at 37C with 5% CO2 in Dulbeccos modied Eagles medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 units/mL penicillin, 100 lg/mL streptomycin, and 30 lg/mL Zeocin. Colonies of stable Hh2A clones surviving selection and maintenance with Zeocin were then assayed for h5HT2AR expression by saturation isotherm binding. Radioligand assays Membrane preparations, saturation isotherm, and competition binding assays were performed as previously described (Chambers

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

1578 J. C. Parish et al.

[125I]DOI and a Kd value as determined by saturation isotherm. Ki and EC50 values were calculated as the mean SEM from separate determinations. A Students t-test was used to compare values for each analogue pair. Values were considered signicantly different if they generated p < 0.05. Visual representations of docking ensembles were generated using PyMol (DeLano 2002). Residues are numbered with their actual sequence location after the residue code, followed by their index number (in parentheses) based on the method introduced by Ballesteros and Weinstein (1995), in which the most conserved residue in each transmembrane TM segment is assigned the position index 50.

afnities of the analogue pairs proved to be statistically indistinguishable at the human 5-HT2A receptor (Table 1). (R)-phenylisopropylamine analogues have higher binding afnities for both the rat and human 5-HT2A receptor compared to the (S)-phenylisopropylamine analogues with identical 4-position substituents As an additional preliminary experiment, the binding afnities of enantiomeric PIA pairs were compared at the rat and human 5-HT2A receptors. In all cases the (R)-isomer exhibited higher afnity for the rat and human 5-HT2A receptor when compared to the (S)-isomer (Table 2). Phenylethylamines and ()-phenylisopropylamines have similar EC50s for phospholipase C activation in both A20 and GF-6 cells, although ()phenylisopropylamines have higher intrinsic activity than their phenylethylamine analogues Having shown that PEAs and their ()-PIA analogues have indistinguishable binding afnities for either the rat or human 5-HT2A receptor, we subsequently determined the functional characteristics of these ligands with respect to 5-HT2A receptor-mediated PLC activation. In A20 cells, with low expression levels of the human 5-HT2A receptor (150 fmol/ mg protein), 10 lM 5-HT gave a 10-fold over basal stimulation of inositol phosphate accumulation on average. The potencies of PEAs and their corresponding ()-PIAs for receptor-mediated PLC activation were indistinguishable (p < 0.05). In other words, the concentration to reach half maximal stimulation (EC50) was the same for the ()-PIAs

Results

Phenylethylamines and their ()-phenylisopropylamine analogues have indistinguishable binding afnities at both the rat and human 5-HT2A receptors Initial studies were designed to test the hypothesis that PEAs and their ()-PIA analogues did not differ in 5-HT2A receptor afnity. For these experiments, the binding afnity of four analogue pairs was determined by radioligand competition assay with [125I]DOI in GF-6 cells expressing the rat 5-HT2A receptor (5500 fmol/mg protein). The binding afnities for all analogue pairs were statistically indistinguishable in the rat 5-HT2A receptor (Table 1). In addition to the rat 5-HT2A receptor, the binding afnities of the analogue pairs at the human 5-HT2A receptor were determined. The Hh2A cell receptor line, expressing the human 5-HT2A (7890 610 fmol/mg protein) was developed for radioligand competition assays. As with the rat 5-HT2A receptor,

Table 1 Comparison of afnities and potencies for phosphatidyl inositol hydrolysis of racemic phenylisopropylamines and their corresponding phenylethylamines Hh2A (h5-HT2A) Ki [125I]DOI (nM) 0.83 1.13 0.60 0.70 0.46 0.63 0.70 0.78 (0.18) (0.13) (0.04) (0.05) (0.02) (0.07) (0.05) (0.02) A20 (h5-HT2A) EC50 PI hydrolysis (nM) 17.4 18.2 9.7 9.8 10.1 7.7 13.1 14.4 (3.0) (2.3) (1.8) (1.7) (1.6) (1.5) (1.6) (2.1) Intrinsic activity (% 5-HT) 70.2 40.0 49.7 30.3 50.1 25.7 57.0 43.5 (4.0) (4.5)* (2.2) (1.5)* (3.7) (1.6)* (4.4) (4.6) GF-6 (r5-HT2A) Ki [125I]DOI (nM) 0.66 0.66 0.65 0.65 0.61 0.65 1.78 1.81 (0.13) (0.11) (0.12) (0.07) (0.09) (0.10) (0.30) (0.23) EC50 PI hydrolysis (nM) 25.9 27.0 19.2 19.0 57.0 47.6 76.3 84.6 (2.1) (2.0) (2.6) (2.6) (2.0) (5.3) (4.3) (12.7) Intrinsic activity (% 5-HT) 87.2 64.8 77.1 59.4 77.6 41.7 84.2 78.7 (2.4) (4.6)* (2.8) (4.1)* (3.8) (3.4)* (1.7) (2.2)

Drugs ()-DOB 2C-B ()-DOI 2C-I ()-DOTFM 2C-TFM ()-Aleph-2 2C-T2

Values are the mean and (SEM) from a minimum of three separate determinations. *p < 0.05 compared to the PIA analogue. 2C-B, 4-bromo-2,5-dimethoxyphenethylamine; 2C-I, 4-iodo-2,5-dimethoxyphenethylamine; 2C-T2, 4-ethylthio-2,5-dimethoxyphenethylamine; 2C-TFM, 4-triuoromethyl-2,5-dimethoxyphenethylamine; Aleph-2, 4-ethylthio-2,5-dimethoxyphenylisopropylamine; DOB, 4-bromo-2,5-dimethoxyphenylisopropylamine; DOI, 4-iodo-2,5-dimethoxyphenylisopropylamine; DOTFM, 4-triuoromethyl-2,5-dimethoxyphenylisopropylamine; PI, phosphatidyl inositol.

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

Phenylalkylamines at the 5-HT2A receptor 1579

Table 2 Comparison of phenylisopropylamines enantiomer afnities and potencies for phosphatidyl inositol hydrolysis Hh2A (h5-HT2A) Ki [125I]DOI (nM) 0.29 (0.04) 1.9 (0.2)* 0.27 (0.02) 0.90 (0.04)* 0.12 (0.01) 4.9 (0.3)* A20 (h5-HT2A) EC50 PI hydrolysis (nM) 6.5 25.3 4.8 14.7 5.0 8.3 (1.2) (1.0)* (0.3) (0.6)* (1.0) (1.2) Intrinsic activity (% 5-HT) 79.5 50.1 59.8 38.4 73.0 44.6 (4.5) (7.9)* (2.3) (1.3)* (4.6) (1.6)* GF-6 (r5-HT2A) Ki [125I]DOI nM) 0.27 1.49 0.31 0.98 0.15 0.34 (0.05) (0.27)* (0.03) (0.17)* (0.01) (0.05)* EC50 PI hydrolysis (nM) 15.3 74.4 17.2 54.2 9.6 8.1 (2.5) (10.5)* (2.9) (7.9)* (1.1) (0.4) Intrinsic activity (% 5-HT) 82.4 75.7 74.9 62.4 79.0 70.4 (1.2) (3.1) (4.4) (6.9) (3.7) (5.8)

Drugs (R)-DOB (S)-DOB (R)-DOI (S)-DOI (R)-TFMy (S)-TFMy

Values are the mean and (SEM) from a minimum of three separate determinations. *p < 0.05 compared to the R isomer. DOB, 4-bromo-2,5-dimethoxyphenylisopropylamine; DOI, 4-iodo-2,5-dimethoxyphenylisopropylamine; TFMy, 4-triuoromethylbenzoditetrahydrofuranylisopropylamine; PI, phosphatidyl inositol.

as for their non-chiral PEA analogues. Conversely, a signicant difference in intrinsic activity was observed between each analogue pair, excluding only the 4-ethylthio analogues (p 0.08) (Table 1). In the cases of the 4-bromo, 4-iodo, and 4-triuoromethyl analogue pairs, the ()-PIA analogues had higher intrinsic activity than their achiral PEA analogues. We examined the ability of the four analogue pairs to stimulate PLC via the rat 5-HT2A receptor expressed in GF-6 cells with a large receptor reserve for the PLC signal transduction pathway (Kurrasch-Orbaugh et al. 2003). In the GF-6 cell line, 10 lM 5-HT gave a 30-fold over basal stimulation of inositol phosphate accumulation. As with the human receptor studies, 5-HT2A receptor-mediated PLC activation occurred with no signicant difference in potency between each analogue pair. Additionally, there was a signicant difference between the relative intrinsic activity of each analogue pair, again excluding only the 4-ethylthio analogue pair (p 0.10) (Table 1). (R)-phenylisopropylamines generally were both more potent and had higher intrinsic activities for phospholipase C activation compared to their corresponding (S)-phenylisopropylamine analogues in A20 cells expressing the human 5-HT2A receptor Having observed that the (R)-PIA isomers had higher afnities for both the rat and human 5-HT2A receptors compared to the (S)-PIA isomers, additional studies were undertaken in order to determine whether there were differences in the functional properties of the isomers, measured by 5-HT2A receptor-mediated PLC activation. The ability of (R) and (S) enantiomeric pairs to stimulate phosphoinositide turnover was rst assessed in A20 cells. For the 4-bromo and 4-iodo enantiomeric PIA pairs, the EC50s for the (R)-enantiomers were signicantly different

from the (S)-enantiomers, with the (R)-enantiomers being more potent (Table 2). This result is consistent with in vivo data, where (R)-enantiomers are more potent than the (S)-enantiomers (Shulgin 1973; Shulgin and Shulgin 1991). Curiously, the EC50s for the (R)- and (S)-4-triuoromethylbenzoditetrahydrofuranylisopropylamine (TFMy analogue pair were statistically indistinguishable (p 0.08). The relative intrinsic activities of the enantiomeric pairs also were all signicantly different in A20 cells, with the (R)-isomers having higher intrinsic activity compared to the (S)-isomers. (R)-Phenylisopropylamines were generally more potent than their corresponding (S)-phenylisopropylamine analogues for phospholipase C activation in the GF-6 cell line expressing the rat 5-HT2A receptor, although the differences between intrinsic activities of the enantiomeric pairs were not signicantly different In GF-6 cells, the EC50s for 5-HT2A receptor-mediated PLC activation were signicantly different for the enantiomeric pairs of the 4-bromo and 4-iodo PIAs. As with the human receptor studies, the EC50s for the (R)- and (S)-TFMy analogue pair were statistically indistinguishable (p 0.32). In all cases, the intrinsic activities of both enantiomeric pairs were statistically indistinguishable (Table 2). This result regarding intrinsic activity differs from the human receptor studies, where the (R)-isomers had greater relative efcacy for PLC activation. (R)- and (S)-phenylisopropylamines share similar localization and orientation in the receptor, yet produce distinct perturbations in key residues upon virtual docking into the human 5-HT2A receptor model The comparison of (R)- vs. (S)-DOB shown in Fig. 2a illustrates marked differences in the conformations of

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

1580 J. C. Parish et al.

In the agonist (R)-DOB, shown in Fig. 3(a), F340(6.52) associates closely with the aromatic ring of the ligand through an edge to face interaction. The alpha-methyl group of the ligand also appears to undergo Van der Waals interactions with F340(6.52) in such a way as to form a tightly associated complex between the ligand and F340. In (S)DOB, however (Fig. 3b), F340(6.52) is clearly displaced from the aromatic ring of the ligand. The achiral ligand 2C-B (Fig. 3c) also appears to associate with F340(6.52) through an edge to face interaction, but there is less intimate association between F340(6.52) and the aromatic ring of the ligand than occurs with (R)-DOB (Fig. 3a). Thus, residues that are believed to be involved in ligand binding, S239(5.43), F243(5.47), and F340(6.52) show major perturbations when the (S)-enantiomer binds, as compared to the (R)-enantiomer. By contrast, F339(6.51) shows virtually no perturbation with virtual binding of any of the ligands, consistent with mutagenesis studies indicating it does not directly interact with small molecule agonists (Choudhary et al. 1993).
Discussion

Fig. 2 Comparison of binding poses for: (a) (R)-4-bromo-2,5-dimethoxyphenylisopropylamine (DOB) (red) and (S)-DOB (blue); (b) (R)-DOB (red) and 4-bromo-2,5-dimethoxyphenethylamine (2C-B) (yellow). The ligands in each panel are superimposed as grey space-lling spheres. The carbon framework for each ligand and its associated receptor complement is similarly colored. The lines are the Ca traces of the helical backbones, with the extracellular side of the receptor toward the top of the gure, and TM6 of the receptor in the foreground.

F240(5.44), F243(5.47), F244(5.48), and F340(6.52), as well as S239(5.43). The two enantiomers produce signicantly altered orientations of S239(5.43), along with a portion of the backbone of transmembrane helix 5 (TM5) within the nonaggregate region (and thus allowed to move within the simulations). The docking studies show that F243(5.47) contacts the bottom of the ligand, and is markedly displaced when the less potent (S)-DOB binds, relative to (R)-DOB. The conformations of F339(6.51) within these two docked ensembles are virtually superimposable. In comparing the binding of (R)-DOB and the partial agonist 2C-B (Fig. 2b), the displacements of F340(6.52), S239(5.43), and the backbone of TM5 all show signicant differences. F243(5.47), however, is not displaced to the same extent as occurs with (S)-DOB (Fig. 2a). Similar to what is seen when docking either (R)- or (S)-DOB, F339(6.51) shows very little difference in conformation following docking of the achiral 2C-B.

Consistent with previous reports (Johnson et al. 1987; Glennon et al. 1992; Nash et al. 1994), we observed no differences in binding afnity between PEA and ()-PIA analogue pairs. By contrast, they have marked differences in ability to activate second messenger systems, an effect rst reported in a comparison of a pair of 4-triuoromethyl PEA and PIA analogues (Nichols et al. 1994). The present results with a larger series, including PIA enantiomers, show that PIAs are more efcacious than their PEA analogues in stimulating both rat and human 5-HT2A receptor-mediated PLC activation. There was no difference in EC50 between any of the PEA/()-PIA pairs. Curiously, only the 4-ethylthio PEA/()-PIA pair lacked a signicant intrinsic activity difference at either the rat or human receptor. Perhaps the more polarizable hydrophobic sulfur substituent allows a unique binding pose that confers higher intrinsic activity on PEA partial agonists, independent of the presence of the side chain alpha-methyl. Consistent with this result, 4-ethylthio-2,5-dimethoxyphenylisopropylamine (Aleph-2) is only two-fold more potent in man than 4-ethylthio-2,5-dimethoxyphenethylamine (2C-T2), whereas both ()-DOB and ()-DOI are approximately 10 times more potent than their PEA analogues (Shulgin and Shulgin 1991). Since hallucinogenic PIA enantiomers rst became available (Nichols et al. 1973), subsequent biological assays have determined that (R)-isomers have higher 5-HT2 afnity than their corresponding (S)-isomers (e.g. Johnson et al. 1990). We report here that the (R)-isomers of PIAs also are more potent and more efcacious at stimulating human 5-HT2A receptor-mediated PLC activation. There was no difference,

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

Phenylalkylamines at the 5-HT2A receptor 1581

Fig. 3 View of the three ligands virtually docked looking down onto the ligands from the extracellular space. The ligand is the left-most structure in each panel. With the helical backbones aligned, distinct differences can be seen in the orientations of aromatic residue F340(6.52). In the agonist (R)-4-bromo-2,5-dimethoxyphenylisopropylamine (DOB), shown in (a), F340(6.52) associates closely with the aromatic ring of the ligand through an edge to face interaction and to the alpha-methyl group through Van der Waals interactions to form a tightly associated complex. In (S)-DOB (b), F340(6.52) is displaced from the aromatic ring of the ligand. The achiral ligand 4-bromo-2,5dimethoxyphenethylamine (2C-B) (c) also appears to associate with F340(6.52) through an edge to face interaction, but there appears to be a looser association between this residue and the ligand, perhaps at least partially attributable to the lack of Van der Waals interaction with an alpha-methyl.

however, between intrinsic activity of (R)- and (S)-PIAs in the GF-6 cell line expressing the rat receptor. This phenomenon may simply reect different signaling properties of the human vs. rat receptors. The TFMy analogues did not follow the general trend seen with the other phenylalkylamines. The lack of substantial differences in 5-HT2A receptor-mediated PLC activation

between the (R)- and (S)-TFMy analogues suggests the possibility of a unique binding orientation for these benzodifuranyl ligands. It may well be that the increased hydrophobicity of the dihydrofuran rings, coupled with the highly electronegative 4-triuoromethyl group, allows the ligand to interact with a different subset of receptor residues, but virtual docking experiments did not identify such a binding pose. Indeed, the binding poses of R-DOB and R-TFMy were essentially superimposable (not shown). Our docking and minimization experiments gave results compatible with previous mutagenesis studies. Comparisons of the nal energy-minimized structures show distinct differences in the positions and orientations of several residues lining the binding pocket, including S239(5.43), F240(5.44), F243(5.47), F244(5.48), and F340(6.52). These differences must ultimately derive from the presence or absence of the side chain alpha-methyl, and the stereochemistry at the alpha-carbon atom. Previous mutagenesis studies had shown three of these residues to be particularly important determinants of ligand afnity and intrinsic activity: S239(5.43), F243(5.47), and F340(6.52). The cognate residues in bovine rhodopsin for the latter two, F212(5.47) and A269(6.52), have been shown to have highly coupled evolution as part of a physically connected network that links distant functional sites in the tertiary structure of the G-protein coupled receptors (GPCRs) (Suel et al. 2003). Mutational analysis of the rat 5-HT2A receptor had previously found that F340(6.52) was essential both for high afnity and high potency of phenethylamine agonists (Choudhary et al. 1993; Roth et al. 1997). The cognate residue in several other GPCRs has been shown to have similar importance (Bluml et al. 1994; Heitz et al. 1999; Huang et al. 1999; Ward et al. 1999; Wieland et al. 1999). F340(6.52) is also essential for agonist activation of phosphatidyl inositol hydrolysis; the potency of DOI was nearly abolished, from an EC50 of 12 nM in the wild type to 28 000 nM in the F340(6.52)L mutant (Roth et al. 1997). Residues on TM6 are of particular signicance because rotation of helix 6 is believed to be a major event during G

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

1582 J. C. Parish et al.

protein-coupled receptor activation (Farrens et al. 1996). Movement of TM6 directly alters the conformation of intracellular loop 3 (i3), which in turn affects coupling and activation of Gq/11, and thus the activation of the PLC pathway studied here (Oksenberg et al. 1995; Hill-Eubanks et al. 1996; Gelber et al. 1999). The interaction of F340(6.52) with the three ligands differs considerably, as also illustrated in Fig. 3, but simple explanations for these differences are not evident. The most potent agonist, (R)-DOB, appears to form the tightest complex with F340(6.52). Although the alpha-methyl may help to anchor this association through Van der Waals forces, the overall docking pose is likely the result of numerous local, as well as long-range, conformational perturbations that are not attributable to a single interaction. Nevertheless, assuming that F340(6.52) is a key residue involved in agonistreceptor activation, its movement may be a reporter for perturbations in the overall ligandreceptor conformation. F243(5.47) appears somewhat less critical for 5-HT2A receptor function, although Shapiro et al. (2000) report that the F243(5.47)A mutant had 40-fold increased afnity for ()-DOI, an EC50 for phosphatidyl inositol hydrolysis that was decreased 200-fold, and intrinsic activity reduced to about half of that in the wild type. The similar EC50 values for (R)-DOB and 2C-B, and (R)-DOI vs. 4-iodo-2,5-dimethoxyphenethylamine (2C-I) (Table 2) suggest that the displacement of F243(5.47) by the ligand may somehow be correlated with functional potency. In our virtual docking studies, we observed dramatic differences between (R)-DOB vs. (S)-DOB or (R)-DOB vs. 2C-B in orientations of S239(5.43) and the segment of TM5 that was allowed to move (although this could be an artifact of the articial constraints employed to emphasize hydrogen bonds that might not otherwise be found using conventional force eld calculations.) In the rat 5-HT2A receptor, the S239(5.43)A mutant had ve-fold increased afnity for DOI (Shapiro et al. 2000), with the potency to activate phosphatidyl inositol hydrolysis decreased about three to four-fold, and similar intrinsic activity. We previously hypothesized that S239(5.43) forms a hydrogen bond with the 5-methoxy of DOI (Chambers and Nichols 2002). In the S239A mutant one might speculate that the large lipophilic iodine atom of DOI could form favorable Van der Waals interactions with A239(5.43). Our modeling suggests that F240(5.44) and F244(5.48) are projected outward from TM5 into the surrounding membrane. Consistent with this observation, F240(5.44)A and F244(5.48)A mutations had no signicant effect on DOI potency or intrinsic activity and little effect on afnity, leading to the conclusion that these residues probably do not directly interact with the ligand (Shapiro et al. 2000). F339(6.51) shows essentially no perturbation with virtual binding of any of the ligands. This observation also is

consistent with mutagenesis studies by Choudhary et al. (1993), who found that afnity of DOI for the F339(6.51)L mutant was reduced only by about ve-fold (Ki 0.92 vs. 3.5 nM). Those workers did not examine the intrinsic activity of DOI, but that of four other agonists remained unchanged. Interestingly, the only sequence difference in the putative binding domain between the rat and human 5-HT2A receptor is the substitution of a serine (human) for an alanine (rat) at location 5.46 in TM5. Although this divergence affects the binding of N(1)-substituted ergolines and tryptamines (Johnson et al. 1993), in our experiments we found no signicant difference in the afnity of most of the phenethylamines at the two receptors (Table 1). The exceptions are the 4-sulfur substituted molecules, Aleph-2 and 2C-T-2, where the serine in the human receptor could hydrogen bond to the sulfur atom of the ligand. This nding suggests that the phenethylamines examined here do not engage this residue, and it was gratifying to note that the virtual docking studies failed to place the ligand close enough to residue S242(5.46) for any interaction. Although there are signicant differences between some of the EC50 and intrinsic activity values for ligands at the two receptors, the relevance of virtual docking studies to functional effects is unclear. Thus, other than D155(3.32), the requisite aspartate in TM3 (Wang et al. 1993; Kristiansen et al. 2000) residues shown by previous mutagenesis studies to be most important for ligand binding and intrinsic activity in the rat 5-HT2A receptor were, in order of importance, F340(6.52) and F243(5.47). S239(5.43) was of somewhat lesser importance, with F240(5.44), F244(5.48), and F339(6.51) appearing to be relatively unimportant for receptor interactions with DOI. Because of the assumptions we were forced to make, and the relatively crude nature of the computational methods presently available, we do not propose that these docked ensembles accurately represent the receptor conformational effects produced by changes in chirality in these small exible agonist ligands. What they do show, however, is that small alterations in ligand structure can theoretically produce striking differences in the overall conformations of residues in the binding domain, which qualitatively appear compatible with previous mutagenesis results. We have previously noted that the region around the alpha carbon is very sensitive to steric bulk (Jacob and Nichols 1982), and that conclusion is consistent with these docking results. Although the enantiomers of large complex organic molecules obviously have dramatically different abilities to complement the threedimensional shape of the binding site in a macromolecular target, the basis for the difference in the biological activity of small exible enantiomeric molecules is less apparent. Our results are consistent with the suggestions of Shapiro et al. (2000), who concluded that relatively minor alterations in either receptor or ligand structure can produce dramatic and unpredictable changes in both binding interactions and 5-HT2A receptor activation.

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

Phenylalkylamines at the 5-HT2A receptor 1583

Finally, we found no evidence to support the conclusions of Acuna-Castillo et al. (2002), or Villalobos et al. (2004), derived from Xenopus oocyte studies, that the PEA analogues are antagonists. All PEA compounds were partial agonists in our systems, albeit with lower intrinsic activity than PIAs, but they are not antagonists, at least in the PLC signaling system. In view of the differences we see in these experiments between different mammalian cell lines expressing different levels of rat and human 5-HT2A receptors, it seems likely that conclusions drawn from rat 5-HT2A receptors transiently expressed in frog oocytes may be even less reliable.
Acknowledgements
This research was supported by NIH grant DA02189 from NIDA, and the work was conducted in a facility constructed with support from Research Facilities Improvement Program Grant Number C0614499 from the National Center for Research Resources of the National Institutes of Health.

References
Acuna-Castillo C., Villalobos C., Moya P. R., Saez P., Cassels B. K. and Huidobro-Toro J. P. (2002) Differences in potency and efcacy of a series of phenylisopropylamine/phenylethylamine pairs at 5-HT (2A) and 5-HT (2C) receptors. Br. J. Pharmacol. 136, 510519. Almaula N., Ebersole B. J., Zhang D., Weinstein H. and Sealfon S. C. (1996) Mapping the binding site pocket of the serotonin 5-Hydroxytryptamine2A receptor. Ser3.36 (159) provides a second interaction site for the protonated amine of serotonin but not of lysergic acid diethylamide or bufotenin. J. Biol. Chem. 271, 14 67214 675. Ballesteros J. A. and Weinstein H. (1995) Integrated methods for the construction of three-dimensional models and computational probing of structurefunction relations in G protein-coupled receptor. Meth. Neurosci. 25, 366428. Bluml K., Mutschler E. and Wess J. (1994) Functional role in ligand binding and receptor activation of an asparagine residue present in the sixth transmembrane domain of all muscarinic acetylcholine receptors. J. Biol. Chem. 269, 18 87018 876. Chambers J. J. and Nichols D. E. (2002) A homology-based model of the human 5-HT2A receptor derived from an in silico activated Gprotein coupled receptor. J. Comput. Aided Mol. Des. 16, 511520. Chambers J. J., Parrish J. C., Jensen N. H., Kurrasch-Orbaugh D. M., Marona-Lewicka D. and Nichols D. E. (2003) Synthesis and pharmacological characterization of a series of geometrically constrained 5-HT2A/2C receptor ligands. J. Med. Chem. 46, 3526 3535. Cheng H. C., Long J. P., Nichols D. E. and Barfknecht C. F. (1974) Effects of psychotomimetics on vascular strips: studies of methoxylated amphetamines and optical isomers of 2,5-dimethoxy-4methylamphetamine and 2,5-dimethoxy-4-bromoamphetamine. J. Pharmacol. Exp. Ther. 188, 114123. Choudhary M. S., Craigo S. and Roth B. L. (1993) A single point mutation (Phe340 Leu340) of a conserved phenylalanine abolishes 4-[125I]iodo-(2,5-dimethoxy) phenylisopropylamine and [3H]mesulergine but not [3H]ketanserin binding to 5-hydroxytryptamine2 receptors. Mol. Pharmacol. 43, 755761.

DeLano W. L. (2002) The PyMOL Molecular Graphics System. http:// www.pymol.org. Ref Type: Electronic Citation. Dyer D. C., Nichols D. E., Rusterholz D. B. and Barfknecht C. F. (1973) Comparative effects of stereoisomers of psychotomimetic phenylisopropylamines. Life Sci. 13, 885896. Farrens D. L., Altenbach C., Yang K., Hubbell W. L. and Khorana H. G. (1996) Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science 274, 768770. Gelber E. I., Kroeze W. K., Willins D. L., Gray J. A., Sinar C. A., Hyde E. G., Gurevich V., Benovic J. and Roth B. L. (1999) Structure and function of the third intracellular loop of the 5-hydroxytryptamine2A receptor: the third intracellular loop is alpha-helical and binds puried arrestins. J. Neurochem. 72, 22062214. Glennon R. A., Young R. and Jacyno J. M. (1983) Indolealkylamine and phenalkylamine hallucinogens. Effect of alpha-methyl and N-methyl substituents on behavioral activity. Biochem. Pharmacol. 32, 12671273. Glennon R. A., Teitler M. and Sanders-Bush E. (1992) Hallucinogens and serotonergic mechanisms. NIDA Res. Monogr. 119, 131135. Harris R. A., Snell D. and Loh H. H. (1977) Stereoselective effects of 1-(2,5-dimethoxy-4-methylphenyl)-2-aminopropane (DOM) on schedule-controlled behavior. Pharmacol. Biochem. Behav. 7, 307 310. Heffter A. (1898) Ueber pellote. Beitrag zur chemischen und pharmakologischen kenntnis der cacteen. Naunyn Schmiedebergs Arch. Exp. Path. Pharmacol. 40, 385429. Heitz F., Holzwarth J. A., Gies J. P., Pruss R. M., Trumpp-Kallmeyer S., Hibert M. F. and Guenet C. (1999) Site-directed mutagenesis of the putative human muscarinic M2 receptor binding site. Eur. J. Pharmacol. 380, 183195. Hey P. (1947) The synthesis of a new homolog of mescaline. Quart. J. Pharm. Pharmacol. 20, 129134. Hill-Eubanks D., Burstein E. S., Spalding T. A., Brauner-Osborne H. and Brann M. R. (1996) Structure of a G-protein-coupling domain of a muscarinic receptor predicted by random saturation mutagenesis. J. Biol. Chem. 271, 30583065. Huang X. P., Nagy P. I., Williams F. E., Peseckis S. M. and Messer W. S. Jr (1999) Roles of threonine 192 and asparagine 382 in agonist and antagonist interactions with M1 muscarinic receptors. Br. J. Pharmacol. 126, 735745. Jacob J. N. and Nichols D. E. (1982) Isomeric cyclopropyl ringmethylated homologues of trans-2-(2,5-dimethoxy-4-methylphenyl) cyclopropylamine, an hallucinogen analogue. J. Med. Chem. 25, 526530. Johnson M. P., Hoffman A. J., Nichols D. E. and Mathis C. A. (1987) Binding to the serotonin 5-HT2 receptor by the enantiomers of 125I-DOI. Neuropharmacology 26, 18031806. Johnson M. P., Mathis C. A., Shulgin A. T., Hoffman A. J. and Nichols D. E. (1990) [125I]-2-(2,5-dimethoxy-4-iodophenyl) aminoethane ([125I]-2C-I) as a label for the 5-HT2 receptor in rat frontal cortex. Pharmacol. Biochem. Behav. 35, 211217. Johnson M. P., Audia J. E., Nissen J. S. and Nelson D. L. (1993) N(1)-substituted ergolines and tryptamines show species differences for the agonist-labeled 5-HT2 receptor. Eur. J. Pharmacol. 239, 111118. Johnson M. P., Wainscott D. B., Lucaites V. L., Baez M. and Nelson D. L. (1997) Mutations of transmembrane IV and V serines indicate that all tryptamines do not bind to the rat 5-HT2A receptor in the same manner. Brain Res. Mol. Brain Res. 49, 16. Kristiansen K., Kroeze W. K., Willins D. L., Gelber E. I., Savage J. E., Glennon R. A. and Roth B. L. (2000) A highly conserved aspartic acid (Asp-155) anchors the terminal amine moiety of tryptamines and is involved in membrane targeting of the 5-HT (2A) serotonin receptor but does not participate in activation via

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

1584 J. C. Parish et al.

a salt-bridge disruption mechanism. J. Pharmacol. Exp. Ther 293, 735746. Kurrasch-Orbaugh D. M., Watts V. J., Barker E. L. and Nichols D. E. (2003) Serotonin 5-hydroxytryptamine 2A receptor-coupled phospholipase C and phospholipase A2 signaling pathways have different receptor reserves. J. Pharmacol. Exp. Ther. 304, 229237. Nash J. F., Roth B. L., Brodkin J. D., Nichols D. E. and Gudelsky G. A. (1994) Effect of the R() and S(+) isomers of MDA and MDMA on phosphatidyl inositol turnover in cultured cells expressing 5-HT2A or 5-HT2C receptors. Neurosci. Lett. 177, 111115. Nichols D. E. (2004) Hallucinogens. Pharmacol. Ther. 101, 131181. Nichols D. E., Barfknecht C. F., Rusterholz D. B., Benington F. and Morin R. D. (1973) Asymmetric synthesis of psychotomimetic phenylisopropylamines. J. Med. Chem. 16, 480483. Nichols D. E., Frescas S., Marona-Lewicka D., Huang X., Roth B. L., Gudelsky G. A. and Nash J. F. (1994) 1-(2,5-Dimethoxy-4(triuoromethyl)phenyl)-2-aminopropane: a potent serotonin 5-HT2A/2C agonist. J. Med. Chem. 37, 43464351. Oksenberg D., Havlik S., Peroutka S. J. and Ashkenazi A. (1995) The third intracellular loop of the 5-hydroxytryptamine2A receptor determines effector coupling specicity. J. Neurochem. 64, 1440 1447. Perrine D. M. (2001) Visions of the night: Western medicine meets peyote 18871899. Heffter Rev. 2, 652. Roth B. L., Shoham M., Choudhary M. S. and Khan N. (1997) Identication of conserved aromatic residues essential for agonist binding and second messenger production at 5-hydroxytryptamine2A receptors. Mol. Pharmacol. 52, 259266. Shapiro D. A., Kristiansen K., Kroeze W. K. and Roth B. L. (2000) Differential modes of agonist binding to 5-hydroxytryptamine (2A) serotonin receptors revealed by mutation and molecular modeling of conserved residues in transmembrane region 5. Mol. Pharmacol. 58, 877886.

Shulgin A. T. (1973) Mescaline: the chemistry and pharmacology of its analogs. Lloydia 36, 4658. Shulgin A. and Shulgin A. (1991) PIHKAL. A Chemical Love Story. Transform Press, Berkeley, CA. Shulgin A. T., Bunnell S. and Sargent T. (1961) The psychotomimetic properties of 3,4,5-trimethoxyamphetamine. Nature 189, 1011 1012. Suel G. M., Lockless S. W., Wall M. A. and Ranganathan R. (2003) Evolutionarily conserved networks of residues mediate allosteric communication in proteins. Nat. Struct. Biol. 10, 5969. Villalobos C. A., Bull. P., Saez P., Cassels B. K. and Huidobro-Toro J. P. (2004) 4-Bromo-2,5-dimethoxyphenethylamine (2C-B) and structurally related phenylethylamines are potent 5-HT2A receptor antagonists in Xenopus laevis oocytes. Br. J. Pharmacol. 141, 11671174. Vollenweider F. X., Vollenweider-Scherpenhuyzen M. F., Babler A., Vogel H. and Hell D. (1998) Psilocybin induces schizophrenia-like psychosis in humans via a serotonin-2 agonist action. Neuroreport 9, 38973902. Wang C. D., Gallaher T. K. and Shih J. C. (1993) Site-directed mutagenesis of the serotonin 5-hydroxytrypamine2 receptor: identication of amino acids necessary for ligand binding and receptor activation. Mol. Pharmacol. 43, 931940. Ward S. D., Curtis C. A. and Hulme E. C. (1999) Alanine-scanning mutagenesis of transmembrane domain 6 of the M(1) muscarinic acetylcholine receptor suggests that Tyr381 plays key roles in receptor function. Mol. Pharmacol. 56, 10311041. Wieland K., Laak A. M., Smit M. J., Kuhne R., Timmerman H. and Leurs R. (1999) Mutational analysis of the antagonist-binding site of the histamine H(1) receptor. J. Biol. Chem. 274, 29 99430 000. Young R. and Glennon R. A. (1996) A three-lever operant procedure differentiates the stimulus effects of R. ()-MDA from S (+)-MDA. J. Pharmacol. Exp. Ther 276, 594601.

2005 International Society for Neurochemistry, J. Neurochem. (2005) 95, 15751584

Das könnte Ihnen auch gefallen