See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270090572

Evolution of Macrocyclic Polyamines From Molecular Science to

Supramolecular Science
Article in Bulletin of Japan Society of Coordination Chemistry · January 2012
DOI: 10.4019/bjscc.59.26

CITATIONS READS
4 627

1 author:

Eiichi Kimura
Hiroshima University
298 PUBLICATIONS 12,249 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

supramolecular chemistry View project

Development of new photochemical reactions and their applications to biological and material sciences View project

All content following this page was uploaded by Eiichi Kimura on 20 January 2016.

The user has requested enhancement of the downloaded file.

 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science

Evolution of Macrocyclic Polyamines From Molecular

Science to Supramolecular Science
Department of Chemistry, Faculty of Sciences, Shizuoka University
Eiichi Kimura
Received March 19, 2012; E-mail: sekimur@ipc.shizuoka.ac.jp

Our chemical exploitation of saturated macrocyclic polyamines has led to discoveries of numerous new functions, and
synthesis of new functional molecules and supramolecular complexes. In this account, we focus on 1) anion complexes
with biologically relevant polyanions at neutral pH; 2) uptake of selective transition metal ions by new peptide-mimic
amide-containing macrocyclic polyamines; 3) specific complexation with noble metal ions; 4) the first fluorinated
macrocyclic polyamines; 5) macrocyclic polyamines with new functional pendants; 6) stabilization of unusually high
oxidation states of transition metal ions (e.g. Cu III, NiIII), and uptake and activation of O2 by the NiII-peptide-mimic
macrocyclic complexes;
zinc enzyme models by ZnII-macrocyclic polyamine complexes; 8) selective recognition of a nucleobase thymine
(and uracil) in nucleic acids by ZnII-12-membered macrocyclic tetraamine (“cyclen”) complexes; and 9) new
supramolecular assemblies from ZnII-cyclen building blocks in aqueous solution.

■■ 1. Introduction often appear in the following chapters. Figure 2 summarizes

our work in this account.
The history of saturated macrocyclic polyamines goes back to
1960s.1) In early days they had been used mostly as chelating
agents for transition metal ions for the study of basic macrocyclic
coordination chemistry.2) The discovery of macrocyclic polyethers
(“crown ethers”) and their novel host-guest properties by Pedersen
in late 1960s3) have given new inspiration and vision to
macrocyclic polyamines beyond conventional uses. The difference
in functionalities of macrocyclic polyamines from those of crown
ethers derives from the composite nitrogen donor ligand properties,
Fig. 1 Names and Structures of Typical Polyamines 1-9
their strong basicities and strong proton uptake that frequently
competes against metal chelation in aqueous solution. Polyamine
ligands are highly protonated at neutral pH or strongly metallated
with transition metals, noble metals, and heavy metals via
coordination bonds. Moreover, upon cyclization polyamine ligands
gain new properties beyond those anticipated for mere assemblies
of amines of linear polyamines.

Since early 1970s, we have synthesized dozens of new

macrocyclic poyamines to find new functions on the basis of
Fig. 2 New Functionalities of Macrocyclic Polyamine Ligands
new concept of host-guest interactions. Figure 1 shows name
and structure of typical linear and macrocyclic polyamines that
■■ 2. Recognition and Uptake of Specific Molecules
Corresponding Author: Eiichi Kimura
Address: 836 Ohya, Suruga-ku, Shizuoka 422-8529, Japan and Metal Ions
Keywords:macrocyclic polyamine, anion complex, selective transition metal ion
uptake, noble metal ion uptake, fluorinated macrocyclic polyamine, peptide- 2.1. Uptake of Biologically Relevant Polyanions by Highly
mimic macrocyclic dioxopolyamine, O2 uptake and activation, zinc enzyme
models, nucleobase recognition, supramolecular assembly Protonated Macrocyclic Polyamines.

Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 26

Accounts
It has been known that naturally occurring linear polyamines
such as spermidine 1 and spermine 2 (as protonated species) bind
with adenosine mono-(AMP2-), di-(ADP3-), and triphosphate
(ATP4-) anions in biological conditions. For example, the
apparent complexation constants, log Kapp, at pH 8 for the 1:1
complexation of spermine 1 with 5ʼ-AMP2-, 5ʼ-ADP3-, and
5ʼ-ATP4- are 2.6, 3.1, and 4.0, respectively. Multiprotonated
macrocyclic polyamines (hereafter MP) should have higher
positive charge density within the constrained macrocyclic
frames and thus were expected to have higher affinities to the
phosphate anions. Indeed, an 18-membered macrocyclic
hexamine 9 bearing 3H+ at pH 8 had higher log Kapp values of
3.2 with 5ʼ-AMP2-, 5.6 with 5ʼ-ADP3-, and 6.4 with 5ʼ-ATP4-
respectively.4) Likewise, 9 (as 3H+ form) yielded 1:1 anion
complexes with polycarboxylate anions (e.g. succinate2- or
citrate3-) but not with acetate1- in neutral pH aqueous solution.5)
The protonated 9-citrate3- complex formation was initially
discovered during identification of MP on paper
electrophoresis in citrate buffer.6)
The phosphate recognition by MP also took place at
membrane surfaces.7) A lipophilic macrocyclic pentaamine
bearing a long alkyl chain 10 was incorporated as a new
chemical sensor element into poly(vinyl chloride) (PVC)
matrix. A conventional anion exchange type electrode uses a
lipophilic quarternary ammonium cation, capriquat, but there
was no anion sensor based on the host-guest interaction. The
proton uptake is prerequisite for the potential response of
polyanions. The linearity of the log C vs. E curve ranges from
10ー7 to 10―3 M with a slope of 14.5 mV/decade in a HEPES
buffer (pH 6.7). The order of the potentiometric response with
5ʼ-AMP, 5ʼ-ADP, and 5ʼ-ATP followed the order of anion
complexation constants. The present ATP sensor shows a
higher sensitivity (~10-7 M) and wider dynamic range (10-7-10-3
M) than the previous electrochemical methods. Therefore, 10
or its homologue may serve a new ATP sensor prototype.
A new ditopic host molecule 11, composed of 9 and benzo-
15-crown-5 covalently linked, were synthesized, which formed
1:1 complexes (K ～102 M-1) with zwitterionic molecules such
as amino acids, peptides, or dopamine in aqueous solution at
neutral pH.8) The crown ether moiety and the protonated MP
concertedly interacted with guests at the ammonium cation
segment and anionic (or catechol) segment, respectively.
27 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
2.2 Selective Uptake and Transport of Metal Ions by
Dioxocyclam 6
Saturated macrocyclic polyamine ligands (L) can complex
with virtually all kinds of heavy metal ions and transition metal
ions. All of these complexes MLn+ are thermodynamically so
stable (log KML > ~15) that they cannot practically be used for
selective metal chelating agents. Another disadvantage with the
strong MP complexation is irreversibility; i.e. the metal
dissociation from MP complexes is so slow that recovery of
metal ions is very difficult. As a consequence, MP, unlike
macrocyclic polyethers, are unsuitable for membrane carriers
of metal ions.
We therefore investigated dioxocyclam 6 that we viewed as
a hybrid ligand of cyclam 5 and a tripeptide.9) The original idea
came from a CuII-carrier protein in human blood, whose N-
terminal Asp-Ala-His tripeptide part is responsible for the
function of CuII-uptake and -release by deprotonation or
protonation of its two amides (Fig. 3a). Indeed, 6 was found to
capture specifically CuII, NiII, CoII, PdII, and PtII that tend to take
an N4 square-planar configuration, accompanying concomitant
dispelling of the two amide protons at neutral pH (Fig. 3b).
These MII(H-2L) 12 complexes are very stable at neutral to
alkaline pH. Upon acidified, 12 quantitatively releases the
enclosing metal ion with full recovery of the free ligand 6
unharmed.9,10)
Fig. 3a) CuII-uptake and -release mechanism in blood
Fig. 3b) MII-uptake and release mechanism by our biomimic
macrocyclic ligand
The measured stability constants for log K(MH-2L) = [MH-
２
2L][H+] /[M][L] are 1.0 (M = Cu), -5.6 (Ni), and -11.4 (Co),
indicating appreciably different complex stability among the
divalent metal ions. By controlling pH, one can separate NiII
from CuII: 6 can sequester CuII at pH 5, but not NiII. Only at
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
pH > 7, NiII can form a stable complex 12. Together with well-
known fast complexation of CuII, CuII-selective transport liquid
membrane was devised using a lipophilic dioxocyclam 13 as a
membrane-carrier (Fig. 4).10) From aqueous layer I to aqueous
layer II, the selective CuII extraction was achieved. Another
feature of 6 is that no counteranion is needed for CuII transport,
but countertransport of protons is accompanied against the
increasing CuII-concentration uphill.
Fig. 4 CuII-selective transport liquid membrane via a lipophilic
dioxocyclam 13
The complexes 12 with PdII and PtII were more stable than with
CuII against acid.11) Since PtII is kinetically more inert, the
complexation into macrocyclic cavity of 6 hardly occurred.
However, the very sluggish 12 formation from K2PtCl4 was
accelerated by addition of a reducing agent Na 2S2O3 or ascorbic
acid. From cis-PtII(NH3)2Cl2, a cancer drug, PtII is transferred into 6
in a similar way. Later, an X ray crystal structure of 12 (with PtII)
confirmed a square-planar coordination geometry.12)
2.3. A Characteristic GoldIII Complex with Cyclam 5
The first 1:1 AuIII－MP complex was isolated with cyclam
5.13) The pH titration of AuIII-cyclam complex 14 with 0.1 M
NaOH showed dissociation of a proton with a pKa value of 5.4
(25 ˚C, I = 0.1), which was assigned to deprotonation from one
of the secondary amine of the coordinated cyclam. The
deprotonation of the cyclam NH at such a low pH has not been
observed with any other MII-cyclam complexes, indicating
extremely acidic nature (toward square-planar direction) of
AuIII. We found 14 to stay stable only in acidic solution. In
alkaline solution 14 decomposed. An X-ray crystal structure
showed a square-planar structure with the most stable “trans-
III” cyclam configuration and an average Au-N distance of 2.02
Å.14)
The AuIII-cyclam complex 14 was
found to possess interesting
electrochemical properties suitable for
a novel gold-plating agent.14) Its cyclic
voltammetry (CV) at pH 3 (citrate buffer) showed irreversible
reduction waves with peak potentials at -0.3 V (nickel disk
electrode), where no corresponding oxidation waves were
observed. On repetition of the CV scans, the reduction potential
gradually shifted to more positive potentials and the surface of the
nickel electrode turned to golden color with the metallic gold
plating. Controlled potential coulometry of 1.0 mM 14 at -0.45 V
indicated an electron flow of 3.0 e/Au. After the electroreduction
finished, unharmed cyclam ligand was quantitatively recovered, as
so analyzed by the formation of equimolar CuII-cyclam complex
(UV absorption at 506 nm,  80). Thus, we concluded that AuIII is
100% deposited on nickel metal surface as gold-plate. A
comparison of the reported reduction potentials of AuIII complexes
of linear polyamine ligands, ethylenediamine and
diethylenetriamine indicated that the AuIII state remains more
stabilized by the macrocyclic ligand 5 than by the linear polyamine
ligands. Probably because of the facile reduction of AuIII to Au0, the
electroreduction of the linear polyamine complexes did not yield
gold plating. Instead, black Au0 particles quickly precipitated. The
gold plate from 14 on nickel metal was as robust and of good
quality as currently adopted gold plate from reduction of AuI(CN)2
(in the presence of excess CN- and at -0.9 V). The advantage of our
gold plating method is 1) non-use of a pollutant CN- ion , 2)
quantified AuIII source from the stable crystalline AuIII complex 14,
and 3) 100% recoverable cyclam 5 ligand.
2.4. New Ligands Specific to Noble Metal Ions
New dioxomacrocyclic ligands containing sulfur donors
1515) and 1616) accommodate only noble metal ions PtII and PdII
to yield the corresponding complexes 17 and 18, but not other
common transition metal ions such as CuII, NiII, or CoII in
MeOH-H2O (at pH < 10). In these new macrocyclic ligands,
such discriminating functionalities come from a combination of
the unique S donors and amide groups in the macrocyclic
skeleton to concertedly work only for PtII and PdII. The divalent
noble metal ions possess mixed properties of hard and soft
acids. Their acidities to planar directions are greater than those
of CuII or NiII, so that PdII and PtII can displace the amide
protons at lower pH than CuII or NiII. To the softer S donor, the
noble metal ions act as soft metals so as to better fit than less
soft CuII or NiII.
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 28
Accounts
■■ 3. Fluorinated Cyclams and Dioxocyclams
Since cyclam 5 and dioxocyclam 6 were found very useful
and versatile tetraamine ligands in coordination chemistry,
bioinorganic and biomimetic chemistry, or catalysts, a series of
momo-, di-, tetrafluorinated dioxocyclams 19-21 and cyclams
22-24 have been synthesized for further functionalization
without changing the basic structures.17,18)
The effects of successive F substitution are demonstrated by
the drop in the basicities and the ligand field strengths in their
CuII and NiII complexes. These are ascribable to the strong
electron-withdrawing effect and lipophilic effect of the F atom.
Despite the weaker ligand field strength, the fluorinated
dioxocyclams 19-21 form more stable CuII complexes than
nonfluorinated dioxocyclam 6. In the complexation kinetics in
acetate buffer (4.7 < pH < 5.7) the fastest overall rate was
observed with the tetrafluorinated dioxocyclam 21, where the
cyclam Nʼs are least protonated and so most reactive to CuII.
NiII in the complex with tetrafluorinated cyclam 21 is almost
100% high-spin (purple-color), while with nonfluorinated
cyclam 6, NiII is a mixture of high- and low-spin states. The
NiII-fluorinated cyclam complexes were tested for an
electroreduction catalyst for CO2 → CO + H2 in H2O. The
reduction efficiency and CO/H2 selectivity depended upon the
number of F atom. The tetrafluorinated NiII complex worked
more efficiently and selectively at a lower energy of -1.1 V vs.
SCE than the nonfluorinated cyclam complex.
■■ 4. NiII Complexes of Cyclam appended with a
RuII-Tris-(2,2’-bipyridine) Complex 2619)
and RuII-Tris-(phen) Complex 2720)
NiII-cyclam complex acts as catalysts for reduction of CO2 →
CO, oxidation of alkenes to epoxides, or other transformations.
RuII(bpy)3 has facile photoexcitation properties. Calvin reported
photo-catalyzed reduction of CO2 to CO using a combination of
RuII(bpy)3 and NiII-cyclam.21) We therefore linked these two
functions intramolecularly to develop a novel hybrid photoredox
catalysis system (Fig. 5). Bipyridine-
29 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
pendant cyclam 25 was first synthesized. This new ligand was
heated with RuII(bpy)2Cl2 to yield a RuII(bpy)3 complex, to
which NiII was incorporated into the cyclam part to yield the
desired product 26. The emission intensity of 26 is drastically
reduced (1/160 of RuII(bpy)3) suggesting efficient quenching of
the excited state by the intramolecular attachment of NiII-
cyclam.
Fig. 5 Synthesis of a new CO2 photoreduction catalysts
Regrettably, 26 did not exhibit a good catalytic activity in
the photochemical CO2 reduction, probably due to unfavorable
configuration of the NiII-cyclam unit. The Ru(bpy)3 subunit
was found to dissociate readily on irradiation.
We substituted bipyridine for phenanthroline to make a
homologous RuNi complex 27,20) which turned out to be
photochemically much more stable than Ru(phen)3 upon
irradiation and exhibit a better photocatalytic activity for the
reduction of CO2 than 26 or the combination of RuII(phen)3/
NiII-cyclam. Further structural modification of 27 would make
a more efficient CO2 photoreduction catalyst. Since Ru(phen)3
binds intercalatively to a DNA major groove and NiII-cyclam
plays as a reactant to modify DNA, 27 or the like may find
another interesting biochemical applications.
■■ 5. The Uptake of O2 by NiII-Macrocyclic
Dioxopolymine Complexes22)
Our original dioxo[16]aneN5 8 formed a 1:1 5-coordinate
square-pyramidal complex 28 with a high-spin NiII, which
possesses two deprotonated amide coordination at equatorial
positions.23) This new complex showed an interesting redox
properties. Whereas a 4-coordinate square-planar NiII-
dioxocyclam complex 12 has a reversible redox potential E1/2
(for NiII-NiIII) of +0.57 V (vs. SCE) and a 5-coordinate square-
pyramidal NiII-[16]aneN5 7 complex shows a reversible E1/2 of
+0.66 V, the NiII-dioxo[16]aneN5 complex 28 that hybridizes
those two coordination functions showed a remarkably low E1/2
value of +0.24 V under the same conditions. This is the lowest
oxidation potential ever reported for NiII/III in neutral aqueous
solution. An X-ray crystal structure of 28 confirmed the 5-
coordinate, square-pyramid geometry,24,25) where the steric
strain for the relatively large sized high-spin NiII suffering from
the tight macrocyclic cavity is evident from (1) distorted
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
square pyramid structure, (2) the central NiII ion lying 0.22 Å
from the basal plane toward the apical nitrogen donor, and (3)
the apical Ni-N bond bent from perpendicular by 18.4˚.
Air bubbling of a 28- homologous NiII-benzdioxo[16] aneN5
complex 29 in aqueous solution (pink-color) turned brown-
colored solution to precipitate NiIII-benzdioxo[16] aneN5
complex 30 as crystals. Its X-ray analysis displayed a
tetragonally distorted octahedral structure with a H2O
occupying the sixth position, and reduced macrocyclic steric
strain of the high-spin NiII complex to a more ideal square-
pyramidal macrocyclic ligand geometry by the low-spin NiIII.25)
It was concluded that the equatorial electron-rich amide anion
coordination and reduction of steric stress contribute to the
unusually low E1/2 value of +0.27 V. It is of interest to see
almost no change in the E1/2 value for CuII/III with dioxocyclam
complex (+0.64 V) and dioxo[16]aneN5 complex (+0.68 V).
Most interestingly, during the air bubbling of the
aqueous solution at room temperature, 28 binds with O2
stoichiometrically to yield a 1:1 O2 adduct 31. Although we
failed to isolate and unequivocally prove the structure 31, all of
the circumstantial evidence pointed to the formation of 31.26) It
activates O2 to convert a mixing benzene into phenol (Fig. 6).
The produced phenol oxygen was found to be 100% derived
from O2 and not from H2 O.27)
Although this discovery attracted broad interest as the first
example of O2 activation by NiII to convert benzene into phenol
under mild conditions, i.e. atmospheric pressure and room
temperature, 28 or 29 failed to be developed into a catalytic
system, because the activated O2 oxygenates also the
macrocyclic ligand itself intramolecularly to yielded 32,28)
which can no longer activate O2.
In search for other O2-activating NiII complexes, a potentially
Fig. 6 O2 Activation by 28 and the resulting reaction products
Fig. 7 NiII complex with pyridyl-pendant dioxocyclam and its
reaction with O2
pentadentate dioxocyclam appended with a pyridine 33 was
synthesized.29) The reaction with Ni(OH)2 yielded a doubly
deprotonated Ni complex 34, where the pendant pyridine
remains uncoordinated, as shown by X-ray crystal structure.
The NMR study in D2O solution showed the four-coordinate
complex 34 being in equilibrium with a 5-coordinate, square-
pyramid complex 35 with an apical coordination of the pendant
pyridine (Fig. 7). A voltammetric study disclosed an unusually
low NiII/III quasi-reversible redox potential of +0.29 V vs. SCE
at pH 9.5 and 25 ˚C, a value near to +0.24 V for 28. The NiII
complex 34 absorbed almost equimolar O2 at pH 9.4 and 25 ˚C.
After 60 min of exposure to O2, more than 80% of the Ni
complex 34 was converted to a macrocyclic ligand-oxygenated
complex 37. The ligand-inserted oxygen atom was proven to
derive entirely from O2 but not from the solvent H2O, allowing
us to conclude a transient formation of a 1:1 O2 adduct 36
before O2 attacked intramolecularly the close ligand site.
Furthermore, thus activated O2 on 36 promoted cleavage of
supercoiled DNA (form I) to nicked DNA (form II), like 28 or
other O2 activating metal complexes. However, 36, when
mixed with benzene, did not yield phenol.
■■ 6. A New Series of Macrocyclic Polyamines
with C-Pivot Donor Pendants
We have developed a new and versatile synthetic method
that led to a new class of macrocyclic polyamines having
functional side arms attached at a ring carbon atoms. The
reaction in general uses ,-unsaturated carboxylic acid esters
and linear polyamines for a one-step annulation that
successively involves a Michael addition followed by an
intramolecular lactam formation.30) This method has been
applied to the synthesis of macrocycles having a phenol
pendant by “recycling” coumarin in refluxing methanol with
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 30
Accounts
1,9-diamino-3,7-diazanonane (2,3,2-tet) for two weeks, which
afforded a recycled 14-membered monooxocyclam 38 in 20%
yield after purification by silica gel chromatography. Reduction
of 38 with B2H6 yielded a phenol-pendant cyclam 3930a) (Fig.
8). This method was applied to synthesis of a 13-membered
tetraamine 4031) and a 12-membered triamine 41.32) In a similar
fashion, a catechol- (42),33) hydroquinone- (43),30b) pyridyl-
(44),34,35) and imidazole-pendant cyclam derivatives (45)36,37)
were synthesized from each correspondingly substituted ,-
unsaturated carboxyesters.
The facile coordination to the metal ions held in the
macrocycles by the phenol side-arms in 39,34) 40,31,39), and 4140)
is established by the solution studies and X-ray crystal
analyses.41) The apical phenolate coordination makes a square-
pyramidal complex with high-spin NiII, where the phenolate
oxygen is nearly at the apex of the pyramid with a very short
distance of 2.015 Å. The strong -donation leads to
stabilization of NiIII with NIII/II redox potential of +0.35 V to be
compared with +0.50 V for the cyclam complex.
The intramolecular, axial phenolate coordination makes 39 a
rare type of FeIII-sequestering macrocyclic polyamine ligand
forming a stable red-colored 1:1 complex in neutral aqueous
solution.30) Owing to the kinetic inertness of macrocyclic ligand
dissociation, the ligand exchange reaction of FeIII from 39 to
EDTA practically did not occur. Cyclam without the phenol
side arm cannot dissolve solid Fe(OH)3 in aqueous solution.
The FeIII-39 complex was reversibly reduced to the yellow-
colored FeII-39 complex with a quasi-reversible redox potential
of -0.16 V vs.SCE at 7 < pH < 9.
Fig. 8 Synthesis of phenol-pendant cyclam39 by recycling
coumarin
The catechol-pendant cyclam 42 was designed to see a
possible catechol monodentate coordination that might
influence the redox properties of itself, coupled with a
31 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
coordinating metal ion. The study might be helpful in
elucidating the mechanism of FeIII-containing catechol 1,2-
dioxygenase that catalyzes oxidative cleavage of catechol. For
stability reasons, we synthesized a cyclam appended with o-
methoxyphenol (an electrochemical precursor of catechol)
46.33) Under argon atmosphere, with NiII and FeII, 46 formed 5-
coordinate square-pyramidal 1:1 complexes in situ with apical
phenolate coordination . The most unusual synergistic
oxidation behavior was revealed by the definite simultaneous
3e oxidation of the FeII -46 complex at -0.30 V (pH 7.3 Tris
buffer) on RED. The potential of -0.3 V is too low for the 2e
catechol oxidation (+0.14 V) of the free ligand 46, and NiII-
bound o-methoxyphenolate (+0.33 V). The 1e oxidation
potential for FeII/III in 39 was -0.16 V. We thus concluded that
FeII in 46 was initially oxidized to FeIII at a lower potential of -
0.30 V, whereupon FeIII catalytically drains 2e out of the ligand
to a possible quinone-pendant 47. Separately, mild aeration of
the FeII-46 initially oxidized FeII to FeIII, which underwent
further 2e oxidation to 47 at -0.3 V. Although 47 failed to be
isolated for full characterization, the present FeII-46 complex
may serve the first model of synergistic intramolecular redox
coupling between monodentate catecholate and metal ions to
render the catechol unusually vulnerable to oxidation.
■■ 7. ZnII-Macrocyclic Polyamine Complexes for Zinc
Enzyme Models
Innumerable ZnII-containing enzymes are known. Metal
enzymes that are involved in hydrolysis of carboxylic esters,
amides, peptides, phosphates, or hydration of carbonyls almost
exclusively contain ZnII ions at the enzyme active centers.
Questions may arise why nature picks up Zn among dozens of
metal ions? or what are its special properties specific to its
biological functions? On the basis of his number of studies on
Zn enzymes, Prof. Bertini summarized special characters of
ZnII for enzymes as follows: 1) Zn can take tetra-, penta-, or
hexa-coordination without special preference
thermodynamically or catalytically. Zinc usually has
coordination number fewer than six to leave the catalytic sites
open; 2) the Zn-bound H2O in catalytic sites has pKa value of
near 7 much lower than free H2O (15.5), so that significant
concentration of Zn-OH species exists at physiological pH,
which may function as a nucleophile for electrophilic sites of
substrates such as carbonyl group of esters, amides or CO2; and
3) the pKa values seem to be controlled by surrounding charged
groups of amino acids or anions. 42)
Direct spectroscopic or magnetic investigation of dynamics
of Zn enzymes poses difficulty because ZnII (d10) is colorless
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
and diamagnetic. Zn enzymes were mostly studied by
substituting Zn for spectroscopically visible metal ions such as
MnII, or CoII. Another approach to elucidate the role of Zn was
construction of model complexes that have structure and
functions similar to enzyme active center.
However, few models had been successful in disclosing
underlying properties of ZnII ion. Because of scarce basic
knowledge about the inherent chemical properties of ZnII, some
of the feasibilities of the propositions derived from
enzymological studies about the functions of Zn remain to be
verified by chemical model studies.
In 1975, a ZnII- macrocyclic tetraamine complex 48 was
introduced as a carbonic anhydrase (CA) model by Woolley.43)
This model presented intriguing pictures about some kinetic
functions of ZnII in CA. A drawback of this model was a higher
pKa value of 8.7 than 7.5 of CA for the ZnII-bound H2O.
Moreover, it was insufficient in elucidating thermodynamic
functions of Zn, such as mechanism of CA inhibition by anions
or aromatic sulfonamides.
In 1990,44) we have discovered that a 12-membered
macrocyclic triamine [12]aneN3 (3) complex of ZnII 49
reproduces the simplest and yet the nearest environment of the
active Zn center of CA, which we believe is the best structural
and functional model. A number of scattering biochemical
information about CA have been well-correlated with the
revealed ZnII properties of our first model complex 49.45) Our
second model was ZnII-[12]aneN4 (or cyclen, 4) complex 50,
which is far more stable with a rigid 4-coordinate configuration
and easier to study the intrinsic ZnII properties well mimicking
a variety of zinc enzyme functions.
In the course of the basic study of macrocyclic polyamine
metal complexes , we have obtained colorless crystals of 49
from ZnSO4 and an equimolar amount of [12]aneN3 3 in
aqueous solution adjusted to pH 8.44) From its X-ray
crystallographic study we found an interesting trimeric
structure, where three ([12]aneN3-Zn-OH-)+ lie around a
crystallographic threefold axis and one HClO4 resides above on
this axis passing through Cl and O (Fig. 9). The three ZnII-
bound O make an extremely small equilateral triangle with O---
O separation of 2.25 Å, which is much shorter than the range
for regular hydrogen bond distances of 2.4-2.8 Å. This fact is
accounted for by extremely strong hydrogen bonds between the
three ZnII-bound OH-, a picture reminding us of a hydrogen-
bond network that plays a critical role in the CA active center.
The amphoteric nature of the ZnII ion may render the solid state
trimeric Zn-OH- hydrogen bonds so stable that even a proton
bound to ClO4- could not break them up. The ZnII has a slightly
distorted tetrahedral coordination with three
Nʼs and one OH-. The Zn-O bond is extremely short 1.944 Å,
shorter than the Zn-N bond distances of average 2.02 Å. In the
tetrahedral Zn coordinate environment of CA, the Zn-OH－and
Zn-N(imidazole) distances are estimated to be 1.96 and 2.1 Å,
respectively, which are close to those of our model complex.
Fig. 9 Hydrogen bind network in ([12]aneN3-ZnII-OH)3(ClO4)3·
HClO4
The trimeric structure in the crystals is broken down to two
basic Zn-OHー species and one neutral Zn-OH2 species in
aqueous solution, as established by pH titration. The 1H NMR
of this crystalline complex in D2O shows distinct spin
decouplings assigned to the axial and equatorial protons of
NCH2C and CCH2C, implying that the Zn-OH- complex is
inert (or very slow Zn dissociation on NMR timescale) and the
macrocyclic configuration is solid (not flipping over as in the
free ligand) with the three propylene groups adopting the same
chair configuration. By the pH titration measurement, the pKa
value for Zn-H2O ⇋ Zn-OH- was determined to be 7.3 (25 ˚C
and I = 0.1), which is very close to the reported pKa value of
CA. Since Zn in CA is similarly tetrahedrally coordinated to
three N (imidazole), we postulated that this macrocyclic
triamine complex would structurally and functionally make an
appropriate model of CA.
We found that 49 indeed catalyzes the hydration of CO2 and
the reverse dehydration of HCO3- around neutral pH in aqueous
solution.46) The mechanism and dynamic aspect of reactivities
of 49 well mimic CA.
Later, another type of tetrahedral Zn-OH- complex 51 was
reported as a new mimic of CA.47) This model well mimics the
CA structure and some functions such as CO2 hydration
(though in organic solvents). However, fundamental Zn
properties, such as pKa values for Zn-OH2 or reactivities of Zn
in aqueous environment, were unmeasurable because of the
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 32
Accounts
water insolubility of 51.
7.1. Anion Affinities of ZnII-[12]aneN3 49 vs Carbonic
Anhydrase (CA)
Carbonic anhydrase (and other Zn enzymes such as alkaline
phosphatases) are subject to competitive or uncompetitive
inhibition by anions (e.g. I-, SCN-, CH3COO-) or organic
molecules (e.g. aromatic sulfonamides). Moreover, substrates
are often anions (e.g. HCO3-, phosphate). Questions would
arise how Zn enzymes selectively and efficiently interact with
the desired anionic or neutral substrates? Also why affinities of
anions are significantly different for Zn in CA and for naked
(hydrated) Znaq ?, as seen in the opposite affinity orders for
halogen ions; I- > Br- > Cl- >F- for CA 48)vs. I- < Br- <Cl- < F-
for Znaq.49) Are there any other mechansim in CA that modify
or reverse the intrinsic anion affinities of Znaq? Here, we see a
good example of the gap between enzymological findings and
conventional chemical knowledge, which should be accounted
for by appropriate model studies.50)
Since we already discovered similar pKa values for Zn-OH2
ー
Zn-OH-, i.e. similar OH affinities of ZnII in CA (log K = 6.5)
and in our model 49 (log K = 6.4), we were tempted to further
examine the affinities of other anions with our model.
Indeed, from the pH-metric titrations of 1:1 ZnII-[12]aneN3 in
the presence of excess anions (CH3COO-, SCN-, I-, Br-, Cl-, or
F-) or HCO3- inhibition kinetic study (see below), the 1:1 49-A-
equilibrium (or anion affinity) constants K(A-) were obtained.51)
It is instructive that the order and magnitude of log K(A-) values
are similar for our model (HCO3- 4.0 > SCN- 2.4 > I- 1.6 > Br-
1.5 > Cl- 1.3 > F- 0.8) and CA (SCN- 3.2 > HCO3- 1.6 > I- 1.2 >
Br- 1.1 > Cl- 0.7 > F- -0.1), which lends a chemical support to
the biochemical conclusion that the inhibitor anions
are intimately binding (i.e. coordinating) to Zn at the active
center of CA. It is also instructive that HCO3- has the highest
1:1 affinity constant (log K = 3.1 at pH 8.4) second only to
OH- (log K = 6.4).51) Among common anions (except for OH-)
found in blood, the strongest binding to Zn is a substrate HCO3-,
which is biologically rational; because of the strong affinity,
HCO3- binding to CA is unhindered by more abundant other
anions.
One could reinterpret the same pH titration data in terms
of the raised pKa values for the Zn-OH2 ⇋ Zn-OH- in the
presence of anion inhibitors; e.g. pKa 7.6 at 100 mM F-, 8.0 at
100 mM I-, or 8.5 at 20 mM SCN-. The decreased concentration
of the Zn-OH- species at physiological pH would account for
the decreased catalytic efficiency of CA in the presence of
anion inhibitors. It is now clear that OH- has the highest affinity
to ZnII among the anions including the HCO3- substrate
33 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
and inhibitors. Together with a now accepted postulate that Zn-
OH- is a catalytically active nucleophilic reactant, the following
chemical interpretation of CO2 hydration and HCO3
dehydration on CA would be possible: (1) the competitive
inhibition of HCO3- dehydration (HCO3- → CO2 + OH-) by
anion inhibitors is accounted for as competition of the substrate
HCO3- and inhibitor anions for Zn at acidic pH, (2) HCO3-
dehydration accelerates in acidic pH, but diminish in alkaline
pH by the preferred binding of OH-, (3) the product HCO3-
resulting from the nucleophilic attack of Zn-OH- to CO2 will be
readily replaced by the stronger binding OH-, (4) thus the
catalytic cycle proceeds in either way depending on pH.
7.2. Elucidation of Aromatic Sulfonamide Inhibition of CA
by Our Model Study51)
The strong inhibition of CA by aromatic sulfonamides (e.g.
acetazolamide which is therapeutically prescribed as a diuretic
drug) had been extensively studied from an enzymological
point of view and now is accounted for by their direct binding
to Zn in the active center. An X-ray crystallographic analysis (2
Å resolution) of the CA－acetazolamide complex supports this
explanation. However, this is somewhat puzzling by
conventional chemical knowledge, because transition metal
ions (e.g. NiII) bind to the thiazole N rather than to the
sulfonamide N. There was no report of Zn complex with
acetazolamide. Chemical study has been awaited. In this
context, one may ask whether ZnII is capable of displacing the
sulfonamide hydrogen to coordinate with the resulting
sulfonamide anion at physiological pH, as we saw in the above
chapter for H2O. Note that aromatic sulfonamides are weak
acids, having pKa values of 8-10. Unless Zn especially favors
ArSO2NH- as a coordinate partner, the dissociation of
ArSO2NH2 would not occur near pH 7.
The binding constants of various ArSO2NH2 with Zn-[12] aneN3
complex 49 could not be determined by the pH titration method in
aqueous solution due to ArSO2NH2 insolubility problem. So we
resorted to an inhibition kinetic method.51) The Zn-OH- species in
49 catalyzes 4-nitrophenolacetate (NA) hydrolysis, which was
found to be subjected to inhibition by (trace amount of ) aromatic
sulfonamides. The kinetics of the initial NA hydrolysis in the
presence of various sulfonamides at pH 8.4 at 25 ˚C in 10% (v/v)
CH3CN aqueous solution was followed spectrophotometrically for
the release of nitrophenol. From the order of the inhibition, the 1:1
binding constants Ki were determined using Dickson plot (1/v = Ki
[I]/vmax + 1/vmax). Our model results of log Ki are parallel to the
reported log Ki for CA, implying that our chemical model 49 and
CA would have a common inhibition mechanism, i.e. the aromatic
sulfonamide
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
anions binding to the catalytic site of ZnII. Separately, we have
isolated a 1:1 49-deprotonated acetazolamide complex.44)
A p-toluensulfonamide was intramoleculaly attached to [12]
aneN3 to synthesize tosylamidopropyl-[12]aneN3 52, which was
tested if the pendant sulfonamide would be deprotonated in neutral
pH solution to bind to ZnII in the macrocyclic complex.51) In the
absence of ZnII, the pKa value of the pendant tosylamide was as
high as 11.2, i.e. an extremely weak acid. However, in the presence
of ZnII the dissociation took place at pH 7 and the deprotonated
tosylamide-coordinating tetrahedral complex 53 was isolated. Its
X-ray crystal structure was determined.51) This model proves that
ZnII indeed can deprotonate aromatic sulfonamides to bind to the
resulting amide N anions at neutral pH. As anticipated, 53 showed
no NA hydrolysis activity.
7.3. Development from ZnII-Sulfonamide Binding to New
Zinc Fluorophores
As an application of unique propensity of aromatic
sulfonamide anion binding to ZnII-macrocyclic polyamine
complexes, a fluorescent dansylamide was attached to cyclen to
54, which formed a dansylamide-deprotonated ZnII complex
53, whereupon the dansylamide fluorescence remarkably
increased , 4.9-fold at 540 nm and 10-fold at 490 nm on
excitation at 330 nm. Thus, 54 was expected to be a sensitive
indicator of ZnII at neutral pH in aqueous solution.52) The Zn-
dependent fluorescence with 5 M 54 is linearly responsive
to 0.01-5 M concentration of ZnII, which is unaffected by the
presence of mM concentration of other biological metal ions
such as Na+, K+, Ca2+, Mg2+, FeII, or FeIII. This new Zn
fluorophore 54 forms a far more stable 1:1 complex (Kd = 6 x
10―13 M at pH 7.8) than any previous Zn fluorophores and can
be useful as a new prototype of zinc fluorophore.53)
The dansylamidoethylcyclen 54 is partially lipophilic and
cell-permeable to emit weak fluorescence (emission at 550 nm)
by itself in cells. When ZnII is incorporated into the cells using
a zinc carrier pyrithione, the strong fluorescence appears due to
the 1:1 Zn complex 55 formation (emission at 530 nm). Thus
formed Zn complex is stable and cell-impermeable to remain
intact and keep emitting fluorescence over a few hours.54)
When apoptosis of cancer cells HeLa or HL60 are artificially
induced (by an anticancer agent etoposide or H2O2), 54
selectively and strongly stains only apoptoic cells at their early
stages with emission of strong fluorescence.54) Thus, 54 serves
also as a fluorescent sensor of apoptosis.
The Zn fluorophore studies have been continued to create
new cyclen series 56,55) 57,56) and 58.57-59)
7.4. [12]aneN3-ZnII-OH- Complex as a New Alcohol
Dehydrogenase Model
ZnII-containing alcohol dehydrogenase catalyzes “hydride
transfer” from alcohol to nicotinamide adenosine dinucleotide
NAD+ (oxidized form) and the reverse “hydride donation” from
NADH (reduced form) to carbonyls. The X-ray structure of
horse liver alcohol dehydrogenase revealed an active site
containing a Zn tetrahedrally coordinated with two cystein
sulfur atoms (cys 46 & cys174), one histidine nitrogen (his67),
and a water molecule.60) Enzymatic studies suggested that the
zinc ion generates alkoxide anions from alcohols with
simultaneous coordination, whereupon the “hydride transfer“
occurs from -C-H of the Zn-bound alkoxide to NAD+. To
facilitate the alcohol deprotonation (pKa ~16), a general-base
assistance from a proximate His61 is invoked. In classic organic
reactions, the Meerwine-Pondorf-Verley reaction is well
known that involves a similar hydride transfer from alcohols to
carbonyls on an extremely acidic metal ion AlIII in aprotic
solvents. On amphoteric ZnII, however, there was no chemical
precedence for such alcohol deprotonation to generate alkoxide
ion for the hydride transfer.
In one of our series of Zn enzyme model studies,61) 4-
nitrobenzaldehyde (0.125 mmol) was heated in 1.0 ml of a
refluxing alcohol (2-propanol, 2-propanol-d8, ethanol,
methanol, and 2,2,2-trifluoroethanol) containing 0.8 mol% of
ZnII complex (1 mol); ([12]aneN3-Zn-OH-)3(TfO)3·TfOH
(TfO = CF3SO3-), ([12]aneN4-Zn-OH2)(ClO4)2, ([12]aneN4-
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 34
Accounts

Zn-OH-)2(ClO4)2·HClO4, ([14]aneN4-Zn)(TfO)2 (all the Zn Interestingly, the reverse “hydride donation” reaction was not

complexes were isolated and characterized), or Zn(TfO)2. The catalyzed by the Zn-[12]aneN3 complex.
reactions gave only two products, A (4-nitrobenzyl alcohol)
and B (4-nitrobanzaldehyde dialkyl acetal) (Fig. 10), which
were identified by high-performance chromatography and 1H
NMR. Comparison of the product distribution and turnover
number with zinc catalysts (Table 1) is highly instructive
concerning the catalytic role of the Zn complexes for two types
of nucleophilic reactions; 1. hydride transfer to product A
(corresponding to the hydride transfer to NAD+) and 2.
alkoxide transfer to product B (corresponding to the hydration
of carbonyls in CA).

In a separate study,62) we synthesized [12]aneN3 with an

Fig. 10 Alcohol dehydrogenase reaction by ZnII-macrocyclic
complexes
intramolecularly attached alcohol 59 and its Zn complex 60.
We saw facile deprotonation of the pendant-alcohol in
equilibrium with an alkoxide-coordinated complex 61 with a
pKa value of 7.4. For more details, see below.
7.5. Activation of Alcohols by ZnII to Attack Carboxy- and
Phosphoesters
Activation of proximal (intramolecular type) serine-OH via
metal coordination is illustrated by dinuclear ZnII alkaline
phosphatases for phosphate hydrolysis. On the basis of X-ray

Table 1 Products distribution and TON (turnover number) (24 h crystal structure and NMR analysis, it was proposed that a
at reflux) substrate phosphate first bound to Zn1 and Zn2 is attacked by
the Zn2-bound serine-OH(or serine-O-) to yield a transient
Most interestingly, the reaction in 2-propanol with [12] phosphoseryl intermediate 62, which is then attacked by the
aneN3-Zn-OH-complex gave the product A in TON 7820 based emptied Zn1-bound OH- to complete the phosphate hydrolysis
on [ZnII] (corresponding to 62.4% oxidation of the starting and go back to the original start state (Fig. 11).
alcohol) after 24 h. Other ZnII complexes were virtually
noncatalytic for this ADH-like reaction. The reaction catalyzed
by [12]aneN3-Zn-OH- in 2-propanol-d6 was followed up to 40
h by 1H NMR , which unequivocally proved D-transfer to yield
4-nitrobenzyl alcohol monodeuterated at the benzylic position.
The outstanding hydride transfer TON by [12] aneN3-Zn-OH
complex implies facile generation of Zn-OR to labilize its -C-
H. The low barrier for the 4- to 5-coordinate interconversion on
the [12]aneN3 complex may contribute to the catalytic
turnover. A proposed mechanism is the initial Zn-OH acting as Fig. 11 A proposed mechanism of alkaline phosphatase
a general base to generate an alkoxide complex.
An NAD+ model compound, N-benzylnicotinamide also An above mentioned alcohol-pendant [12]aneN3-Zn complex 60
undergoes the hydride transfer from 2-propanol to yield 1,4- was tested whether the Zn-bound alcohol is activated for
dihydronicotinamide as a major product, although the reaction nucleophilic attack at electrophilic esters, in analogy to the Zn-
proceeded slower.61) When the reaction was followed by 1H activated serine-OH in alkaline phosphatases. The reaction of 60,
NMR spectroscopy, we observed almost exclusive 1,4- however, was too slow with phosphomonoester substrates.
dihydronicotinamide formation. No product other than a minor Therefore, we resorted to hydrolysis of a more reactive
1,6-dihydronicotinamide adduct was detected. carboxyester, 4-nitrophenylacetate (NA).60) The

35 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)

 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science

initial hydrolysis rate was followed by the appearance of 4- OH-coordinating at an apical position. The potentiometric
nitrophenolate anion in aqueous solution of 10% (v/v) CH3CN. titration of 66 showed dissociation of a proton with a pKa value
The second-order rate constant at 25 ˚C was 1.4 x 10-1 M-1s-1. of 7.6. The structure of the deprotonated species was assigned
Since this initial rate continued beyond one catalytic cycle, this to ZnL-OH- complex 67, which is an active reactant catalyzing
hydrolysis was concluded as a catalytic reaction. This rate did NA hydrolysis. The detailed kinetic study revealed that the Zn-
not change when D2O solvent was used, suggesting that the OH- of 67 played dual roles: as a general base to activate the
ZnII-bound alkoxide anion did not act as a general base (to proximate alcohol OH for the first acyl-transfer reaction and
produce a nucleophilic OH- from water), but reacted directly as subsequently as a nucleophile to attack the acyl intermediate in
a nucleophile. The initial product was confirmed to be a the final hydrolysis step (Fig. 13).
transient “acetyl intermediate” 63, which was isolable upon
addition of excess EDTA to the early reaction mixture. The
acetyl intermediate complex with ZnII 64 was not stable enough
to allow isolation. Hydrolysis of the acetate group in the
isolated 63 immediately took place by mixing it with ZnII,
apparently by the attack of a spontaneously generated Zn-OH-.
The rate-determining step in the overall reaction was thus the
first “acetyl transfer” process (Fig. 12).

Fig. 13 Mechanism of catalysis of p-nitrophenolacetate

hydrolysis by 66

Later, a new benzylalcohol-pendant cyclen complex with

ZnII was synthesized.62) The Zn-bound alcohol deprotonation
took place with a pKa value of 7.3 (25 ˚C). The pendant alcohol
undeprotonated complex 68 and the deprotonated complex 69
were both isolated. The Zn-bound alkoxide ion in 69 served an
Fig. 12 Mechansim of catalysis of p-nitrophenol acetate instructive model for Zn alkaline phosphatase (Fig. 14). It
hydrolysis by 61 reacted with bis(4-nitrophenyl)phosphate to yield 4-nitrophenol
and a “phosphoryl-transfer” intermediate 70, which was isolated
The second-order rate constant for NA hydrolysis by [12] and fully characterized. The Zn-bound alkoxide in 69 is more
aneN3-Zn-OH－ under the identical conditions was 3.6 x 10-2 reactive than N-methylcyclen-Zn-OH-
M-1s-1. Comparing the kinetic data for the acetate hydrolysis
catalyzed by the Zn-OR- and Zn-OH-, we conclude that Zn-OR-
is 4 times stronger nucleophile than Zn-OH-, despite almost the
same basicity (i.e. pKa values) for both anions. This is the first
model of Zn hydrolases, where Zn-activated serine or threonine
residue preferentially acts as a nucleophile to yield a transient
intermediates, which are then rapidly hydrolyzed by the Zn-
OH- attack. It is of interest to note that chymotrypsin (a serine
enzyme) also involves “acyl transfer” to the His57-activated
Ser195-OH in the rate-determining step, which is then rapidly
hydrolyzed by the same imidazole-activated H2O to complete
the substrate hydrolysis
In the case of an analogous alcohol-pendant cyclen 65,61) an
isolated ZnII complex from pH 6 aqueous solution had a 5-
Fig. 14 Mechanism of bis(nitrophenol)phosphate hydrolysis by 68
coordinate structure 66 with the (undeprotonated) alcohol

Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 36

Accounts
71. The phosphoryl transfer reaction by 69 was 1,700 times
faster in anhydrous DMF solution than in aqueous solution,
which is explained by the enhanced nucleophilicity by less
polar or poorer solvation of DMF. This fact is of interest in
view of the hydrophobic environments for most of the active
center of Zn enzymes.
Subsequently, the phosphoryl transfer intermediate 70 was
subjected to the intramolecular hydrolysis by the spontaneously
generated Zn-bound OH- 72 at a more alkaline pH to yield a
phosphomonoester product 73. The pKa value for 68 ⇋ 72 was
9.1, as determined potentiometrically and kinetically. The first
order rate constant for the reaction 70 → 73 was 3.5 x 10-5 s-1 at
35 ˚C. As a reference to this intramoelcular phosphodiester
hydrolysis, the intermolecular hydrolysis of ethyl (4-
nitrophenyl) phosphodiester by N-methylcyclen-Zn-OH 71
gave a second-order rate constant of 7.9 x 10-7 M-1s-1 at 35 ˚C.
Thus, the intramolecular hydrolysis is about 45,000 times faster
than the intermolecular hydrolysis with 1 mM 71. This model
finding of the phosphate hydrolysis, along with the above
carboxyester hydrolysis model, supports the significance of the
proximate serine in Zn hydrolases for the indirect two-step
reactions, involving the initial Zn-alkoxide attack at substrate
esters to yield ester-transfer intermediates in the rate limiting
steps, followed by the fast intramolecular Zn-OH- attack at the
intermediates to complete the catalytic hydrolysis. The indirect
two-step reactions are kinetically more favorable than the direct
one-step attack by Zn-OH-.
The chemical principle disclosed by our models would
prevail to the reaction schemes in Zn hydrolases containing
serine or threonine at the active centers. In our last phosphatase
model, regrettably, the second Zn-bound phosphomonoester
intermediate 73 did not release the second phosphate ion,
because the phosphomonoester remained strongly coordinated,
which prevented recurrence of the Zn-OH- species. By contrast,
alkaline phosphatases being dinuclear Zn2 enzymes, a Zn-OH
spontaneously generates from the Zn1 attacks the
phosphoserine intermediate at Zn2 to complete the hydrolysis
for catalytic turnover (Fig. 11).
7.6. A Dinuclear ZnII-Phosphomonoesterase Model
Mononuclear ZnII-[12]aneN3 and ZnII-cyclen comlexes,
though hydrolyze phosphotri- and diesters, do not react at all
with phosphomonoesters. Rather, they bind strongly with
monoesters to form a fairly stable 1:1 complexes(log K 3~4).
This is due to the -2 charge and strong bidentate ligand nature
of the monoesters.63)
A new alkoxide-bridged, dizinc cryptate 74, which
we synthesized incidentally, was found to be a rare
37 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
phosphomonoesterase model.64) Initially, 74 did not seem
reactive toward phosphates, because the two zinc ions appeared
to be coordinatively saturated, both in a 5-coordinate, trigonal-
bipyramidal environment. To our surprise, 74 actually reacted
exclusively with a dianionic 4-nitrophenyl phosphomonoester
(NPP2-), but not with a monoanionic bis(4-nitrophenyl)
phosphodiester (BNP-), a neutral tris(4-nitrophenyl)
phosphotriester (TNP), or 4-nitrophenyl acetate. We already
have learned that the easiness of phosphate hydrolysis is TNP >
BNP- > > NPP2- by mononuclear [12]ane N3- and [12] aneN4
(cyclen)-ZnII complexes.63) Thus, 74 is an exceptional Zn
complex that selectively reacts with phosphomonoester
dianions (and ATP4-) to cleave the P-O ester bond, thus serving
an interesting model for dinuclear Zn2 phosphomonoesterases.
An X-ray structure analysis confirmed two identical Zn
centers having a trigonal bipyramidal coordination sphere. The
reaction of 74 with NPP2- in pH 5-7 aqueous solutions at 55 ˚C
yielded a phosphoryl amide complex as a sole product 75,
whose structure was determined by an x-ray structural study.
The reaction of 74 with ATP4- also gave 75 (and ADP3-),
although the reaction rate was much slower. The second-order
dependence of the rate (first order in [74] and [NPP2-]) fit to the
kinetics. The rate-pH profile showed a bell-shaped relationship
with pK1 of 5.2 (assigned to NPP2- + H+ ⇋ HNPP-) and pK2 of
6.3 (apical NH + H+ ⇋ HNH+), having the maximum rate at pH
5.9. The overall reaction picture of the P-O ester bond-cleavage
was unambiguously drawn on the basis of the mechanistic
study (Fig. 15).
Fig. 15 Mechansim of phosphomonoester bond cleavage by 74
In the beginning appropriately separated two Zn ions (3.42 Å)
allowed the two anionic O- donors of NPP2- to bind to the sixth
coordination sites of each Zn to neutralize and activate NPP 2- to
become susceptible to the subsequent reaction. Concomitantly, the
two axially coordinated Zn-N bonds loosened and one of the
loosened Nʼs attacked the activated NPP2- electrophilic site to
displace 4-nitrophenolateto yield 75. The weakly acidic conditions
(at 5.9 < pH < 7) would assist the dissociation of the axial NH by
protonation to facilitate the association of NPP2-, which accounts
for the increased rate as pH is lowered from 7 to 5.9. Further
acidification (below pH 5.9), on the other hand, would convert the
reactant NPP2- to an unreactive protonated
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
species HNPP-, which has weaker affinity to 74, resulting in the
slower reaction.
The present model findings may be relevant to the reaction
mechanism of dinuclear phosphomonoesterase: first, the strong
bidentate association of the substrate phosphate triggers
dissociation of the labile ligand: and secondly, the outgoing
ligand is in a favorable position to attack at the electrophilic
site of the metal-interacting (and thus activated) substrate.
7.7. Development of Cooperative Phosphate Receptors
with Multi ( ZnII-cyclen) Units
A new cyclic tris(ZnII-cyclen) 76, which has three ZnII-
cyclen complexes connected through a 1,3,5-trimetylbenzene
spacer, has been synthesized as a new receptor for organic
phosphate dianions in aqueous solution.68) The design of 76 was
based on X-ray crystal structure of a 1:3 complex of NPP2- with
ZnII-cyclen complex. The P NMR titration of 76 with
31
phenylphosphate (PP2-) dianion in D2O confirmed the
formation of a 1:1 76-PP2-complex 78. By potentiometric pH
titrations, the 1:1 complex affinity constants, log Kaff, were
determined to be 5.8 with NPP2- and 6.6 with PP2- in aqueous
solution. The new cyclic tris(ZnII-cyclen) 76 is a much better
host for phosphates than the parent ZnII-cyclen (log Kaff = 3.3
with NPP2-).
A new supramolecular complex 80 was synthesized as a
luminescence sensor for inositol 1,4,5-trisphosphate (IP3), an
important second messenger in intracellular signal transduction,
and its achiral model compound, cis,cis-1,3,5-cyclohexanetriol
triphosphate (CTP3). The new sensor 80 is Ru(II)-templated
assembly of three molecules of a bis(ZnII-cyclen) complex having a
2,2-bipyridyl linker 79 (Fig. 16).69) Single-crystal X-ray diffraction
analysis of a racemic mixture of 80 showed that three of the six
ZnII-cyclen units are orientated to face the opposite side of the
molecule with three apical ligands (ZnII-bound OH-) of each of the
three Zn2+ located on the same
face. 1H NMR and UV titrations of 80 with CTP3 indicated that
80 forms a 1:2 complex with CTP 3 in aqueous solution at
neutral pH. In the absence of guest molecules, 80 (10 M) has
an emission maximum at 610 nm at pH 7.4 and 25 °C
(excitation at 300 nm). An addition of 2 equiv of CTP 3 induced
a 4.2-fold enhancement in the emission of 80 at 584 nm. It
should be noted that 80 is the first chemical sensor that directly
responds to CTP3 and IP3 and discriminates these triphosphates
from monophosphates (PP2- and D-glucose-6-phosphate) and
diphosphates (cis-1,3-cyclohexanediol diphosphate and D--
fructose-1,5-diphosphate).
Fig. 16 Synthesis of ruthenium complex of bis(ZnII-cyclen) as the
luminescence sensor of IP3
7.8. Discovery that Guanidine is a ZnII-Binding Ligand in
Neutral pH Aqueous Solution70)
Guanidines, arginine residues, having pKa of 12.5 in Zn
enzyme centers are perceived to function only as guanidium
cations to bind to anions such as phosphates. Their coordinate
bonding to ZnII has not been conceived.
We have synthesized a new guanidine-pendant cyclen 81,
which yielded a 1:1 ZnII complex 82 in neutral pH aqueous
solution. X-ray crystal structure analysis revealed a distinct
apical coordination of the pendant guanidine through an imine
nitrogen (Fig. 17).
We then studied how the guanidine in 82 prefers ZnII over
protons at neutral pH. Potentiometric pH titration analysis
disclosed the pendant guanidine equilibrating between
coordination (to 82) and protonation (to 83) with a
deprotonation constant pKa value of very low 5.9. The
speciation diagram for a mixture of 1 mM 81 and 1mM ZnII as
a function of pH at 25 ˚C and I = 0.1 indicates that the initially
formed ZnII complex 83 is populated most abundantly (75 %) at
pH 5.5. Above pH 5.9, 82 is predominant.
In the presence of dianionic phosphomonoesters, which are
good ligands to ZnII as described above and may displace the
guanidine for ZnII, we performed 31P NMR and potentiometric
titrations. The results indicate that a 1:1 NPP-83 complex 84
predominates at lower pH (< 7) and 82 becomes predominant
at pH > 7. The 84-homologue (phenyl phosphate in place of
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 38
Accounts
NPP) complex was isolated and its X-ray crystal analysis
confirmed the structure.
Fig. 17 Zinc complex of guanidyl cyclen 81 and its properties
■■ 8. ZnIIMacrocyclic Polyamine Complexes for
Selective Nucleobase Recognition71)
The fundamental knowledge about ZnII acquired from the
preceding studies of zinc enzyme models has been evolved to
new molecular recognition of nucleobases. In view of the
special characters of strengthened acidity and interaction with
aromatic sulfonamides, ZnII-cyclen complexes were next
examined if ZnII interacts with organic compounds containing a
similarly acidic imide group that may generate stable imide
anions like sulfonamides. In 1993, we were fortunate to
initially choose azidothymidine (AZT, a nucleoside derivative)
and discover that ZnII-cyclen complex indeed interacted with it
in pH 8.5 aqueous solution to yield a crystalline 1:1 complex
85.72) Its X-ray crystal analysis revealed a 5-coordinate, square-
pyramidal structure, where a new coordinate bonding occurs
between Zn and thymine N(3)- (viewed as a deprotonated imide
anion), which is reinforced by two complementary hydrogen
bonds between the two imide oxygens and two NH groups of
cyclen. The 1:1 complex 85 is thermodynamically fairly
stable (Kapp = 3.2 x 103 M-1 at pH 8 and 25 ˚C), but kinetically
labile, as so concluded from 1H NMR study. The following
potentiometric titrations of deoxyribonucleosides (dA, dG, dC,
and dT) and related compounds in the presence of Zn-cyclen
disclosed selective interactions of Zn-cyclen to dT, AZT, and
39 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
uridine (U). Most remarkably, the present selective recognition
occurs in aqueous media, whereas naturally occurring dA-dT
(or U) base recognition takes place only in the hydrophobic
domain of DNA (or RNA) polymers.
To further strengthen this nucleoside recognition in
aqueous solution, cyclen was appended with aromatic ligands
capable of stacking with nucleobases. First synthesized was
Zn-acridinylmethylcyclen complex 86,73) which yielded a
crystalline 1:1 complex with 1-methylthymine. An X-ray
crystal structure confirmed an additional - stacking
interaction between acridine and thymine, which was also
supported by 1H NMR study. Its thymine recognition was
enhanced thermodynamically 20 times (Kapp = 6.3 x 104 M-1 at
pH 8), and also became kinetically more inert. It is also noted
that Zn-cyclen complexes were found to preferentially interact
with neutral thymine or uracil bases over potentially
competitive phosphate monoanions in DNA or RNA backbone.
Another type of cyclen derivative was 2,4-dinitrophenylcyclen
87.74) Its ZnII complex 88, where cyclen acts as a tridentate
ligand (like [12]aneN3), interacted with 1-methylthymine to
yield an analogous 1:1 complex, which is one order more stable
than the (Zn-cyclen)-1-methylthymine complex. The X-ray
crystal structure showed4-coordination of ZnII with three
secondary nitrogens of cyclen and a deprotonated imide of 1-
methyltymine. The guest stands cofacial to the 2,4-
dinitrophenyl ring (indicating - stacking interaction) with the
two imide carbonyl oxygen hydrogen bonding with two NH
groups of the cyclen.
For application of our nucleobase recognition, lipophilic Zn II-
cyclen complexes 89a,b,c were synthesized to make carriers for dT
or U nucleosides or nucleotides for selective transport from an
aqueous phase to a CHCl3 phase.75) The transport mechanism that
involves the formation of lipophilic 1:1 dT-(or U-)-ZnL complexes
under slightly alkaline conditions is different from previously
reported transport mechanisms. The CHCl3-extracted lipophilic 1:1
complexes were dissociated to release nucleosides or nucleotides
into pH 6 aqueous solution.
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
Thus, membrane transport of dT, U, or relevant drugs (e.g.
AZT) becomes possible from weak alkaline pH to acidic pH
aqueous solution, which is coupled with proton antiport.
8.1. Recognition of Thymine (or Uracil) Base in Single
and Double-Stranded Nucleic Acids
A central goal of designing DNA- or RNA-targeting agents
is to manipulate gene activities with agents that recognize and
modify specific nucleic acid sequence. The Zn-cyclen
recognition of thymine (or uracil) base in nucleic acids was
first examined with a single-stranded poly(U) and a double-
stranded poly(U)-poly(A).76) The affinity constant of Zn-cyclen
with each imide-deprotonated uracil in poly(U) was K = 105.1
M-1, which is close to 105.2 M-1 for the interaction of Zn-cyclen
with a deprotonated uracil. This fact implies that Zn-cyclen has
negligible interaction with monoanionic phosphodiester
backbone of poly(U). Moreover, Zn-cyclen complex disrupts
U-A hydrogen bonds to dissociate(or melt) the duplex of
poly(U)-poly(A), as demonstrated by the decreasing melting
temperature (Tm) of the duplex in pH 7.6 with an increase in
the concentration of Zn-cyclen. At r = [Zn-cyclen]added/ [uracil]
poly(U) = 2, the poly(U)-poly(A) duplex completely disrupted at
room temperature. Additional evidence for the inhibition of
poly(A)-poly(U) hybridization by Zn-cyclen was obtained by a
P NMR study.
31
A linear bis(Zn-cyclen) 90 and a linear tris(Zn-cyclen) 91
matched a dinucleotide thymidylthymidine (TpT) and a
trinucleotide thymidylthymidylthymidine (TpTpT) for a stable
1:1 complex 92 and 93, respectively, at pH 7.4 in aqueous
solution (Fig. 18) .77) The stoichiometric 1:1 complexation of
TpT with 90 was established by 1H NMR titration, and UV-
spectrophotometric titration. In FAB mass spectroscopic
measurement of 1:1 TpT/90 mixture in pH 8 aqueous solution,
a major peak at m/z 1119 was assigned to the formation of 92
(m/z = 1120.84). The much stronger 1:1 complexation of
TpTpT with the linear tris(Zn-cyclen) 91 was evident by the
steeper slope and the sharper break in the UV titration. The
FAB mass data revealed a major peak at m/z at 1763 with Zn
isotopic peaks for 93 (m/z = 1762.81) prepared in situ from 1:1
TpTpT/91 mixture in pH 8 aqueous solution.
Fig. 18 Stable complexes of bis(ZnII-cyclen) and tris(ZnII-cyclen)
with TpT and TpTpT
Potentiometric pH titration for the interaction of 90 (1 mM)
with TpT (1 mM) determined the 1:1 dissociation constant Kd
of 6.3 x 10-7 M at pH 7.4 and 25 ˚C, which implies more than
95% of 91 and TpT are in the form of the 1:1 complex 92 in the
pH range of 7 to 10. Likewise, the potentiometric pH titration
for the interaction of 91 (1 mM) with TpTpT (1 mM) gave the
dissociation constant Kd of further10-3 times smaller value of
8.0 x 10-10 M at pH 7.4 and 25 ˚C. In conclusion, a bis(Zn-
cyclen) 90 and a tris(Zn-cyclen) 91 are a new type of DNA base
sequence-selective ligands that very efficiently bind to TpT and
TpTpT, respectively, at physiological pH, where the
recognition occurs at the concentrations of μM and nM order,
respectively.
The selective and efficient nucleobase recognition by Zn II-
acridinylmethylcyclen complex 86 was firmly established for
the thymine in dinucleotides, single-stranded d(GTGACGCC)
and a double-stranded d(CGCTAGCG)2 by UV
spectrophotometric titration, potentiometric titration, 1H NMR,
FAB-MS measurements and molecular modeling.78) 86 was
shown to bind to all the deprotonated thymine T- in TpT, GpT,
CpT, ApT, and single-stranded d(GTGACGCC) with similar
1:1 binding constants, log Kapp of ca. 5 at pH 8. The NMR
study demonstrated destabilization of the double-stranded
d(CGCTAGCG)2 due to the binding of 86 to T-.
The linear bis(Zn-cyclen) 90 has been displayed to be the
first highly selective ditopic receptors (for T－and phosphate
dianion) for various derivatives of dT (and U) nucleotides
including AZTMP (monophosphate) and AZTDP (diphosphate)
at physiological pH in aqueous solution.79) The resulting 1:1
ternary complexes 94 are kinetically inert and
thermodynamically much more stable than a 1:1 T-- (Zn-
benzylcyclen) complex. This is due to the simultaneous
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 40
Accounts

bindings of the T moieties and dianionic phosphate moieties to
the two Zn-cyclen units of 90.
In our experiment, the presence of HgII reduced 34% of the
initial photodimerization rate. After subsequent experiments,
we could summarize the inhibitory effect on the photo[2 + 2]
cycloaddition of TpT by the monomeric and dimeric Zn-
cyclens as in Fig. 19.

8.2. Inhibition of Photo[2 + 2] Cycloaddition of TpT

and Promotion of Photosplitting of the Cis-syn-
cyclobutane Thymine Dimer by bis(Zn-cyclen) 90 80)
Exposure of cellular nucleic acids to UV irradiation leads to Fig. 19 Effect of bis(Zn-cyclen) 90 on photodimerization of TpT

a variety of lesion, which, unless repaired, are potentially

carcinogenic, mutagenic, or cytotoxic. Among the known
photoproducts of DNA, the major ones are cis-syn-(T[c,s]T),
and trans-syn(T[t,s]T)-cyclobutane thymine dimers, which
result from the photo[2+2]cycloaddition of two adjacent
thymidilyl(3ʼ-5ʼ)thymidine site. In the aforementioned
(thermodynamically and kinetically stable) TpT-bis(Zn-cyclen)
90 binding, the two thymine rings separate far more remotely
(ca. 10 Å according to energy-minimization calculation) than
the ordinary 3―4 Å in DNA. This finding led us to suspect that
90 might prohibit the photo[2 +2] cycloaddition and/or
promote the reverse photosplitting of the thymine dimers for
protection of TpT sites of DNA against UV light.
The irradiation of TpT (0.2 mM in 10 mM Tris buffer at pH 7.6,
I = 0.1, 3-5 ˚C) was simultaneously carried out in the absence and
presence of Zn-cyclen or bis(Zn-cyclen) 90 in the same UV
reaction vessel. An equimolar amount of 90 more effectively
inhibited the rate of photodimerization (85% inhibition of the
control reaction) than two equimolar Zn-cyclen during the UV
exposure for 20 min (where possible secondary reverse
photosplitting is insignificant). Even after the photodimerization
and photosplitting reached an equilibrium at prolonged irradiation
for 3 h, 90% of TpT remained. When two equivalent of 90 were
used, even stronger inhibition of the photodimerization was seen.
The presence of 1 mM ZnSO4 without the cyclen ligands did not
affect the reaction rate. From the dissociation constant of ca. 1 μＭ
at pH 7.6 (obtained from isothermal calorimetric titration method),
95% of the 1:1 complex 92 is present at [TpT] = [91] = 0.2 mM at
pH 7.6 and 25 ˚C. Under the same conditions, TpT (0.2 mM) and
Zn-cyclen (0.4 mM) forms only 4% of the 1:2 complex. Mercury
(II) is known to inhibit pyrimidine photodimerization of DNA.
8.3. Biochemical Identification of the Zn-Cyclen Binding
Sites in Natural DNA
In order to identify the interaction sites of Zn-cyclen
complexes in natural DNA, a biochemical assay was employed
with a DNA footprinting technique.81) Very useful information
was obtained about the sequence-selective binding on 5ʼ-32P-
labeled DNA fragments (arbitrary 150 base pairs) from plasmid
pUC19 by means of a combination of DNase I and micrococcal
nuclease footprinting methods. For references, typical AT
binders(distamycin A and DAPI )and a GC binder
(echinomycin) were tested side by side on the same gel plate.
It was immediately apparent that most of the DNaseI-protected
regions by the Zn-acridinylmethylcyclen complex 86 overlap with
those by distamycin A and DAPI, but not with those by
echinomycin. All of the protected sequences by various Zn-cyclen
complexes were associated with homopolymeric AT. The IC50 (=
concentration required for 50% inhibition of the DNase I digestion
at position 45-50, TpTpTpTpTpTpT in the upperstrand) were in the
range of 8-30 μM, which may not appear exciting values in
comparison with 0.5 μM distamycin A and 2 μM DAPI. However,
we believe that these values are significant enough for a starter
prototype. Free ZnII ion or cyclen ligands alone, did not protect
DNA from the DNase I hydrolysis. CuII or NiII-cyclen complexes,
which do not interact with thymine, were not effective at all in
protecting DNA.
For more insightful mechanism of the AT protection by the
Zn-cyclen complexes, another footprinting assay was
conducted with micrococcal nuclease that recognizes single-
stranded sites in DNA without much influence from the partner
strand and cut pA and pT sites more effectively than pG and

41 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)

 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
pC bonds. This enzyme turned out be be useful in comparing
the patterns of AT recognition by the Zn-cyclen complexes and
by distamycin A. In the micrococcal nuclease footprinting,
distamycin A and DAPI were shown to protect both pT and pA
pairs together in the homopolymeric AT regions, indicating
distamycine A and DAPI bind to both A and T in the AT minor
groove to stabilize the double helix. On the other hand, Zn-
cyclen complexes protected only pT, but not the pairing pA in
the same homopolymeric AT regions, indicating that Zn-cyclen
complexes selectively bind to T to melt the A-T base pairs and
that the separated pA partners were exposed to micrococcal
nuclease and become susceptible to the enzyme hydrolysis.
In order to investigate whether distamycin A and Zn-cyclen
complexes target the common AT-rich regions, a competitive
binding study was performed. To the DNA preincubated with
60 μM ZnII-quinolyl-cyclen complex 95 was added distamycin
(0-10 μM), which was then digested with micrococcal
nuclease. It was found that the initial 95-controlled footprinting
pattern at the T and A regions were dose-dependently changed
to the distamycin-controlled pattern. The reverse was also true.
We conclude that distamycine A and 95, despite the different
binding mechanisms, reversibly compete for the common AT
regions.
8.4. Selective Interaction of Zn-cyclen Complexes with
TATA Box
The AT-rich DNA sequence located at 25 to 30 base (“TATA
box”) upstream from the transcriptional start sites is an essential
element of the promoter for eukaryotic RNA polymerase II. The
TATA box plays a key role in regulating the overall level of
transcription and participates in the selection of transcriptional start
sites. TATA binding protein, TBP (a transcriptional factor) must
bind to it for initiation of the transcription. Distamycin A and DAPI
bind to TATA box and thus inhibit the binding of TBP. Likewise,
Zn-cyclen complexes were demonstrated to selectively bind to
TATA box region by DNase I footprinting assay of a 230 bp SV40
early promoter sequence containing an AT-rich TATA box
(TATTAT) and GC-rich GC box.82)
Further investigation by gel mobility assay using a 25-bp
TATA consensus DNA fragment demonstrated that Zn-cyclen
complexes inhibit the binding of human TBP to the DNA: the
IC50 values were 15 μM (for Zn-acridinylmethylcyclen 86),
and 2.5 μM (for Zn-1,7-bis(4-quinolylmethyl)cyclen 96) to be
compared with 0.4 μM for distamycin A or 0.8 μM for DAPI.
Thus, Zn-cyclen complexes are a new prototype as a small
molecular drug for genetic transcriptional factor.
The chemical interaction of Zn-cyclen complexes with
AT-rich region of DNA inhibited the subsequent biochemical
processes.83) The inhibition of calf thymus DNA-directed
transcription was assayed by measuring the incorporation of
[3H]-UTP into the transcribed RNA in the presence of E. coli
RNA polymerase. The experiment indeed showed the
inhibition of transcription by Zn-cyclen complexes. The Zn-
cyclen complexes having higher binding affinities to calf
thymus DNA, showed the higher inhibition. The most effective
inhibitor was Zn-bis(naphthylmethyl)cyclen 97 (IC50 = 55
μM).
8.5. Inhibition of HIV-1 TAR RNA-TAT Peptide Binding by
a Linear Tris(Zn-cyclen) 9184)
The search for small molecules that interact with RNA is
attracting interest for drug discovery. One typical example is
AIDS therapeutics which target viral RNA sites to prevent the
formation of key RNA-protein complexes. Transcription of
HIV-1 genome is facilitated by a HIV-1 regulatory protein TAT
which activates synthesis of full-length HIV-1 mRNA by its
binding to TAR RNA. The TAR element adopts a hairpin
structure with a U-rich bulge (UUU or UCU) located four base
pairs below a six-nucleotide loop. The bulge is the TAT
binding site.
We found that the linear tris(Zn-cyclen) 91 is the most potent
inhibitor compound ever reported for the TAR-TAT binding,
due to its extremely strong binding to the UUU bulge.
Footprinting analysis using micrococcal nuclease, RNase A and
phosphodiesterase I has revealed the UUU bulge to be strongly
protected by 91 in the TAR model RNA sequence. Then, we
studied inhibition of the complexation of TAR with a TAT-
consensus peptide by a gel mobility shift assay. Among the Zn-
cyclen complexes tested, 91 was the most potent inhibitor with
an extremely low IC50 value of 20 nM (Fig. 20). By
comparison, the currently most potent inhibitor
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 42
Accounts

is aminoglycoside antibiotics, neomycin (IC50 = 1 μM). Our supercomplex 100, isolated as crystals in a quantitative
yield
present findings show that Zn-cyclen complexes efficiently (Fig. 21).
and selectively recognize a U-rich single-strand element in a
biologically important RNA and may serve as a new type of
HIV-1 RNA targeting small molecules.

Fig. 20 Gel mobility shift of the complex of TAR33 and Tat46-86 in

the presence of increasing concentration of 91 (0-1.25 M) or
ZnSO4 (10 M)

■■ 9. Three-Dimensional Supramolecular Assembly

from Zn-cyclen Complexes and Cyanuric
Acid

Spontaneous self-assembly of natural polymers such as Fig. 21 Supramolecular assemblies from cyanuric acid and 76 in
nucleic acids or proteins through noncovalent bonds in aqueous solution

solution, gel, or solid states are ubiquitous in nature. Artificial

supramolecular architectures have been often constructed from
imide-containing organic molecules such as nucleic acid bases
(e.g. thymine) or barbiturates. Driving forces for such self-
assemblies are hydrogen-bond, aromatic - stacking,
hydrophobic interaction, or ionic interaction.
In a preceding Chapter, we have described recognition of
thymine and uracil (both mono-imide compounds) by Zn-
cyclen complexes in neutral aqueous solution to yield relatively
stable 1:1 complexes (ZnL-dT- and ZnL-U-), where the
deprotonated imide N- coordinates to Zn and the two carbonyl
oxygens hydrogen bond with the two complementary
cyclen NHʼs. The apparent affinity constant, log Kapp was ca. 3
at pH 7.6 and 25 ˚C. For recognition of a diimide compound,
barbital, we earlier found a linear bis(Zn-cyclen) 90 being an
appropriate host to yield a stable 1:1 complex 98 in equilibrium
with a 2:2 complex.85)
For recognition of a centrosymmetric tri-imide compound,
cyanuric acid (CA), we now have examined a cyclic tris(Zn-
cyclen) 76 as a host in anticipation of a 1:1 76-CA3- complex
99 formation.86) Instead, to our big surprise, totally unexpected
molecular recognition took place with a novel supramolecule,
4:4 76-CA3- complex 101 formation, which was isolated as
crystals in 77% yield from strong alkaline (pH > 11.5) aqueous
solution. Meanwhile, in neutral aqueous solution, a different
type of molecular recognition occurred to yield a 2:3 76-CA2-
Fig. 22 a) Cuboctahedral cage structure of 101
Binding mode between Zn-cyclen and cyanuric acid
The X-ray crystal analysis of 101 revealed a highly
symmetric 4:4 cuboctahedral-like cage structure; an assembly
of four triangular tris(Zn-cyclen) 76 linked with four smaller
triangular CA3- with 12 Znʼs at 12 vertices, together with six
open rectangular plane (Fig. 22a). This is the first example of
the trianionic CA form in aqueous solution, apparently
stabilized by the ZnII coordination. As shown in Fig. 22 b, the
stability of the superstructure was reinforced by the hydrogen-
bond network between Zn-N-H---O=C of CA3- (24 hydrogen
bonds in total). The cage 101 has a well-defined inner space,
which is separated from the outer space by four phenyl groups
and CA3-. The inner cavity is visualized as a three-dimensional
truncated tetrahedron with approximate volume of 140 Å3,
suggesting potential encapsulation space for lipophilic guest
molecules. Another 2:3 supercomplex 100 was stable at neutral

43 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)

 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science
pH and fully characterized by potentiometric pH titration, 1H
NMR, and X-ray crystal analysis.
The 4:4 assembly 101 is stable only in solid or in DMSO
solution and tends to disproportionate to the 2:3 complex 100
in neutral H2O, as was shown by the 1H NMR study. The 2:3
and 4:4 supercomplexes are equivalent in terms of acid-base
equilibrium and the direction of the disproportionation to the
4:4 complex was probably favored due to its crystal stability
that may be contributed from the intramolecular 24 hydrogen
bond network.
For construction of more robust and practical supramolecular
cages in neutral pH aqueous solution, we have replaced CA
with trithiocyanuric acid (TCA).87) An expected advantage of
TCA over CA was that three imide deprotonation constants
(pKa) of TCA are ca. 2 orders lower than those for CA, so that
the trianionic TCA3- might be stabilized near neutral pH.
Another advantage may be that TCA3- tends to aromatize to a
1,3,5-triazine structure and the anions would localize more on
the exocyclic sulfur atoms, which would favor stronger Zn-S-
interactions. In fact, slow evaporation of a mixture of tris(Zn-
cyclen) 76, TCA, and NaNO3 in pH 8 aqueous solution yielded
a 1:1 stoichiometric (76-TCA3-) (NO3-)3 complex as crystals.
The X-ray structure analysis, indeed, revealed a 101-like 4:4
assembly structure 102. The exterior frame is visualized as a
twisted (to right or left) cuboctahedron. ach triangular TCA3-
links to three Zn-cyclen moieties through Zn-S- coordination
bonds and cyclenN-H---N hydrogen bonds (Fig. 23). Since the
three C-S-Zn bonds around all of the four TCA3- units
Fig. 23 A 4:4 supramolecular assembly 102 from 76 and
trithiocyanuric acid
Fig. 24 X-ray crystal structure of adamantane-inclusion complex
of 102. Other hydrophobic guest compounds are also included.
are bent by 100 ˚in a clockwise or anticlockwise fashion, the
new 4:4 supercomplex 102 is chiral. As expected, 102 is
thermodynamically and kinetically (on the NMR timescale)
stable, and is formed quantitatively from a 1:1 76-TCA mixing
aqueous solution.
Most outstanding property of 102 is the encapsulation of
lipophilic organic compounds. An adamantane (ADM)-
including 102 was crystallized out from a mixed solution at pH
Its X ray structure analysis has proven the encapsulation of
ADM in the matching size of the inner cavity (Fig. 24). Of the
guest ADM the four composite cyclohexane rings squarely face
the tetrahedrally surrounding phenyl rings. The encapsulation
of guest molecules was confirmed by their 1H upfield shifts due
to the shielded magnetic field by the surrounding aromatic
rings. The guest molecules encapsulated were generally size-
matching hydrophobic molecules either neutral, anionic, or
cations, such as ADM, nitrophenols, ibuprofen, diadamantane,
or tetraethylammonium cation. The guest-including 102 is
kinetically stable on NMR timescale. However, the guest
molecules can be slowly displaced with other guests at room
temperature in aqueous solution.
Recently, molecular recognition of hydrocarbon guests by
102 was examined in detail.88) On the basis of 1H NMR data,
GC measurements, and crystal structure analysis, it was
confirmed that one molecule of a hydrocarbon such
as CnH(2n+2) (n = 1~8), adamantane, cis- and trans-decaline,
cyclododecane and (R,R)-BMBA can be incorporated as guests.
Computational simulations of 102–guest complexes using
Amber* of the MacroModel suggested that 102 has a flexible
structure originating from rotation about the C-S- bonds of the
(ZnII)3-(TCA3–) unit. The SASA values (= (SASAG + SASAH)
– (SASAHG)) were calculated as parameters to evaluate the
surface complementarity between guest molecules and the
inner cavity of the Zn-cage, in order to explain the order of
stability determined by the guest replacement experiments. In
addition, the storage and release of volatile molecules by 102 in
aqueous solution and solid state was reported. These data may
aid in the design, synthesis, and understanding of new artificial
supramolecular systems for molecular recognition, sensing, and
storage of various guest molecules in aqueous solutions and
solid states.
The above principle of molecular recognition and self-
assembly has been applied to synthesis of prismatic molecular
assembly.89) Upon replacing the cyclic tris(Zn-cyclen) 76
for the linear tris(Zn-cyclen) 91 in pH 8 aqueous solution,
the reaction with TCA quantitatively yielded a (91)3-(TCA3-)3
supercomplex 103. The 3:3 trigonal prism structure was
confirmed by X-ray crystal analysis. Its exterior may be
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 44
Accounts
represented schematically as a trigonal prism (Fig. 25); its
prism length is 2.9 Å, the distance between two adjacent TCA3-
is 1.2 Å. Of the three TCA3- units, the two terminal TCA3- units
bind to Zn ions all through Zn-S- coordination bonds, where
TCA3- takes an aromatic triazine form. Interestingly, the central
TCA3- has less aromatic structure with the anion localizing on
the imide Nʼs to bind to Zn and Zn-S coordination bonds are
supplementary. Three phenyl groups from three 91 are
assembled closely to each other. The potentiometric titrations,
UV spectrophotometric titrations, 1H NMR titrations, and ESI
MS all supported quantitative formation of the
thermodynamically and kinetically stable (on NMR timescale)
3:3 supramolecular complex 103 at M concentrations in
aqueous solution at neutral pH.
Recently, a tetrakis(Zn-cyclen) 104 was synthesized. It was
found that 104 forms a 4:4 complex 105 with TCA3-, which is
very stable at nM concentrations in neutral pH aqueous
solution, as evidenced by 1H NMR titration, potentiometric pH
and UV titrations, and MS measurements. 90) The supercomplex
105 does not dissociate into the starting building blocks in the
presence of strong ZnII-binding anions such as phosphates or
double-stranded DNA. The results of the competitive binding
assays with ethidium bromide (EB) and calf-thymus DNA
(ctDNA), thermal melting (Tm) experiments, gel mobility shift
assays (GMSA), and dynamic light scattering data (DLS)
strongly indicated that the trigonal prism functions as a
polycationic template to induce aggregation of double-stranded
DNA.
Fig. 25 Trigonal prism structure of supermolecules 103 and 105
Recently, a supramolecular catalyst 106 was constructed by
a 4:4:4 self-assembly of bis(ZnII-cyclen) complex 79 (Zn2L),
CA2– and Cu2+, as confirmed by potentiometric pH, UV/ Vis
spectrophotometric titrations and X-ray crystallographic
analyses (Fig. 26).91) Interestingly, 106 contains -Cu2(OH)2 at
center cores and selectively accelerates the hydrolysis of a
phosphate monoester, NPP, at neutral pH, which mimics the
active centers of dimetallic phosphatases. To our knowledge,
106 represents the first synthetic supramolecular
phosphomonoesterase that is formed by self-assembly and
45 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)
selectively hydrolyzes NPP catalytically under physiological
conditions, although catalytic turnover number (CTN) is not so
dramatic, about 2~4. Nevertheless, these results would be
useful for the future design of stable, biologically active, and
biocompatible supramolecular catalysts.
Fig. 26 A new supramolecular assembly 106 from 79, cyanuric
acid, and CuII ,which selectively hydrolyzes a phosphomonester
NPP
■■ Acknowledgement
Author wishes to thank to his laboratory staffs, coworkers,
students, and collaborators for their great contributions to make
the present account possible. With all these people he is
pleased to share this Award honor. The most of the works have
been done at Pharmaceutical Chemistry Lab, Hiroshima
University School of Medicine, and Coordination Chemsitry
Lab, Institute for Molecular Science at Okazaki. Author also
thanks to Prof. Shin Aoki (Tokyo University of Science) and
Dr. Naoko Nishina (Shizuoka University) for this manuscript
preparation.
■■ References
N. F. Curtis, Coord. Chem. Rev., 3, 3 (1968).
D. H. Busch, Acc. Chem. Res., 11, 392 (1978).
C. J. Pedersen, J. Am. Chem. Soc., 89, 2495 and 7017 (1967).
E. Kimura, M. Kodama, T. Yatsunami, J. Am, Chem. Soc., 104, 3182
(1982).
E. Kimura, A. Sakonaka, T. Yatsunami, M. Kodama, J. Am. Chem.
Soc., 103, 3041 (1981).
T. Yatsunami, A. Sakonaka, E. Kimura, Anal. Chem., 53, 477 (1981).
a) Y. Umezawa, M. Kataoka, W. Takami, E. Kimura, T. Koike,
Nada, Anal. Chem., 60, 2392 (1988). b) K. Odashima, R.
Naganawa, E. Kimura, T. Koike, J. Sessler, Supramol. Chem., 4,
101 (1994).
E. Kimura, H. Fujioka, M. Kodama, J. Chem. Soc. Chem. Commun.,
1158 (1986).
a) M. Kodama, E. Kimura, J. Chem. Soc. Dalton Trans., 325 (1079). b)
ibid, 694 (1981). c) K. Ishizu, J. Hirai, M. Kodama,
Kimura, Chem. Let., 1945 (1979). d) E. Kimura, Y. Lin, R. Machida, H.
Zenda, J. Chem. Soc. Chem. Commun., 1020 (1986).
 Evolution of Macrocyclic Polyamines From Molecular Science to Supramolecular
Science

10) E. Kimura, C. A. Dalimunte, A. Yamashita, R. Machida, J. 41) E. Kimura, T. Koike, K. Uenishi, M. Hediger, M. Kuramoto,
Chem.
Soc. Chem. Commun., 1041 (1985). S. Joko, Y. Arai, M. Kodama, Y. Iitaka, Inorg. Chem., 26,
2975
11) E. Kimura, Y. Lin, R. Machida, H. Zenda, J. Chem. Soc., Chem. (1987).
Commun., 1020 (1986). 42) For review, E. Kimura, Tetrahdron, 48, 6175 (1992).
12) E. Kimura, S. Korenari, M. Shionoya, M. Shiro, J. Chem. Soc. 43) P. Woolly, Nature, 258, 677 (1975).
Chem. Comm. 1166 (1988). 44) E. Kimura, T. Shiota, T. Koike, M. Shiro, M. Kodama, J. Am.
13) E. Kimura, Y. Kurogi, T. Takahashi, Inorg. Chem., 30, 4117 Chem. Soc., 112, 5805 (1990).
(1991). 45) For review, a) E, Kimura, Progress in Inorg. Chem., Ed. By
K.
14) E. Kimura, Y. Kurogi, T. Koike, M. Shionoya, Y. Iitaka, J. D. Karlin, 41, 443 (1993). b) E. Kimura, Acc. Chem. Res., 34,
Coord. 171
Chem., 28, 33 (1993). (2001).
15) E. Kimura, Y. Kurogi, S. Wada, M. Shionoya, J. Chem. Soc., 46) X. Zhang, R. Eldik, T. Koike, E. Kimura, Inorg. Chem., 32,
5749
Chem.Commun., 78 (1989). (1993).
16) E. Kimura, Y. Kurogi, T. Tojo, M. Shionoya, M. Shiro, J. Am. 47) R. Alsfasser, S. Trofimenko, A. Looney, G. Parkin, H.
Chem. Soc., 113, 4857 (1991). Vahrenkamp, Inorg. Chem., 30, 4098 (1991).
17) E. Kimura, M. Shionoya, M. Okamoto, H. Nada, J. Am. Chem. 48) Y. Pocker, J. T. Stone, Biochemistry, 7, 2936 (1968).
Soc., 110, 3679 (1988). 49) R. M. Smith, A. E. Martell, Critical Stability Constants, Vol.
4,
18) M. Shionoya, E. Kimura, Y. Iitaka, J. Am. Chem. Soc., 112, 9237 Plenum, New York, 1976.
(1990). 50) E. Kimura, Comments on Inorg. Chem., 11, 285 (1991).
19) E. Kimura, S. Wada, M. Shionoya, T. Takahashi, Y. Iitaka, J. 51) T. Koike, E. Kimura, I. Nakamura, Y. Hashimoto, M. Shiro, J.
Am.
Chem. Soc. Chem. Commun., 397 (1990). Chem. Soc., 114, 7338 (1992).
20) ) E. Kimura, X. Bu, M. Shionoya, S. Wada, S. Maruyama, Inorg. 52) T. Koike, T. Watanabe, S. Aoki, E. Kimura, M. Shiro, J. Am.
Chem., 31, 4542 (1992). Chem. Soc., 118, 12696 (1996).
21) J. L. Grant, K. Goswami, L.O. Sprear, W. W. Otvos, M. Calvin, 53) For review, a) E. Kimura, T. Koike, Chem. Soc. Rev., 27, 179
J.
Chem.Soc., Dalton Trans., 2105 (1987). (1998). b) E. Kimura, S. Aoki, Bio Metals, 14, 191 (2001).
22) Review for macrocyclic oxopolyamines, E. Kimura, J. Coord. 54) E. Kimura, S. Aoki, E. Kikuta, T. Koike, Proc. Natl. Acad.
Sci.
Chem., 15, 1 (1986). 100, 3731 (2003).
23) E. Kimura, A. Sakonaka, R. Machida, M. Kodama, J. Am. Chem. 55) S. Aoki, S. Kaido, H. Fujioka, E. Kimura, Inorg. Chem., 42,
1023
Soc., 104, 4255 (1982). (2003).
24) Y. Kushi, R. Machida, E. Kimura, J. Chem. Soc., Chem. 56) S. Aoki, D. Kagata, M. Shiro, K. Takeda, E. Kimura, J. Am.
Commun., Chem.
216 (1985). Soc., 126, 13377 (2004).
25) R. Machida, E. Kimura, Y. Kushi, Inorg. Chem., 25, 3461 (1986). 57) S. Aoki, K. Sakurama, N. Matsuo, Y. Yamada, R. Takasawa,
S.
26) E. Kimura, R. Machida, M. Kodama, J. Am. Chem. Soc., 106, Tanuma, K. Takeda, E. Kimura, Chem. Eur. J. 12, 9066
(2006)
5497 (1984). 58) S. Aoki, K. Sakurama, N. R. Ohshima, N. Matsuo, Y.
Yamada,
27) E. Kimura, R. Machida, J. Chem. Soc., Chem. Commun., 499 R. Takawasa, S. Tanuma, K. Takeda, E. Kimura, Inorg. Chem.
47,
(1984). 2747 (2008).
28) D. Chen, A. E. Martell, J. Am. Chem. Soc., 112, 9411 (1990). 59) R. Ohshima, M. Kitamura, K. A. Morita, M. Shiro, Y. Yamada,
M.
29) E. Kimura, M. Sasada, M. Shionoya, T. Koike, H. Kurosaki, M. Ikekita, E. Kimura, S. Aoki, Inorg. Chem., 49, 888 (2010).
Shiro, J. Biol. Inorg. Chem., 2, 74 (1997). 60) H. Eklund, A. Jones, G. Schneider, p. 377, Zinc Enzymes, ed. by
I.
30) a) E. Kimura, T. Koike,M. Takahashi, J. Chem. Soc., Chem. Bertini, C. Luchinat, W. Maret, M. Zeooe, Birkhausen,
Boston.
Commun., 385 (1985). b) For review; E. Kimura, Pure & Appl. 61) E. Kimura, M. Shionoya, A. Hoshino, T. Ikeda, Y. Yamada, J.
Am.
Chem., 58, 1461 (1986). Chem. Soc., 114, 10134 (1992).
31) E. Kimura, T. Koike, K. Uenishi, R. B. Davidson, J. Chem. Soc., 62) E. Kimura, I. Nakamura, T. Koike, M. Shionoya, Y. Kodama,
T.
Chem. Commun., 1110 (1986). Ikeda, M. Shiro, J. Am. Chem. Soc., 116, 4764 (1994).
32) E. Kimura, M. Yamaoka, M. Morioka, T. Koike, Inorg. Chem., 63) T. Koike, S. Kajitani, I. Nakamura, E. Kimura, M. Shiro, J.
25, Am.
3883 (1986). Chem. Soc., 117, 1210 (1995).
33) E. Kimura, S. Joko, T. Koike, M. Kodama, J. Am. Chem. Soc., 64) E. Kimura, Y. Kodama, T. Koike, M. Shiro, J. Am. Chem.
Soc.,
109, 5528 (1987). 117, 8304 (1995).
34) E. Kimura, T. Koike, H. Nada, Y. Iitaka, J. Chem. Soc., Chem. 65) T. Koike, E. Kimura, J. Am. Chem. Soc., 113, 8935
(1991).
Commun., 1322 (1986). 66) T. Koike, M. Inoue, E. Kimura, M. Shiro, J. Am. Chem. Soc.,
118,
35) E. Kimura, T. Koike, H. Nada, Y. Iitaka, Inorg. Chem., 27, 1036 3091 (1996).
(1988). 67) T. Koike, E. Kimura, J. Am. Chem. Soc., 113, 8935
 (1991).
36) E. Kimura, M. Shionoya, T. Mita, Y. Iitaka, J. Chem. Soc., 68) E. Kimura, S. Aoki, T. Koike, M. Shiro, J. Am. Chem. Soc.,
Chem. 119,
Commun., 1712 (1987). 3068 (1997).
37) E. Kimura, M. Shionoya, T. Yamauchi, M. Shiro, Chem. Lett., 69) S. Aoki, M. Zulkefeli, M. Shiro, M. Kohsaka, K. Takeda, E.
1217 (1991). Kimura, J. Am. Chem. Soc., 127, 9129 (2005).
38) Y. Iitaka, T. Koike, E. Kimura, Inorg. Chem., 25, 402 (1986). 70) S. Aoki, K. Iwaida, N. Hanamoto, M. Shiro, E. Kimura, J.
Am.
39) a) E. Kimura, K. Uenishi, T. Koike, Y. Iitaka, Chem. Lett., 1137 Chem. Soc., 124, 5256 (2002).
(1986). b) E. Kimura, T. Koike, K. Uenishi, R. B. Davidson, J. 71) For review, a) E. Kimura, E. Kikuta, Progress in Reaction
Kinetics
Chem. Soc., Chem. Commun., 1110 (1986). and Mechanism, 25, 1 (2000). b) S. Aoki, E. Kimura, Chem.
Rev.,
40) E. Kimura, T. Koike, K. Toriumi, Inorg. Chem., 27, 3687 (1988). 104, 769 (2004).
Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012) 46
Accounts

72) M. Shionoya, E. Kimura, M. Shiro, J. Am. Chem. Soc., 115, 83) E. Kikuta, N. Katsube, E. Kimura, J. Biol. Inorg. Chem., 4, 431
6730
(1993). (1999).
73) M. Shionoya, T. Ikeda, E. Kimura, M. Shiro, J. Am. Chem. Soc., 84) E. Kikuta, S. Aoki, E. Kimura, J. Am. Chem. Soc., 123, 7911
116, 3848 (1994). (2001).
74) T. Koike, T. Gotoh, S. Aoki, E. Kimura, M. Shiro, Inorg. Chim. 85) a) T. Koike, M. Takashige, E. Kimura, H. Fujioka, M. Shiro,
Acta, 270, 424 (1998). Chem. Eur. J., 2, 617 (1996). b) H. Fujioka, N. Yamada, E.
75) S. Aoki, Y. Honda, E. Kimura, J. Am. Chem.Soc., 120, 10018 Kimura, Heterocycles, 42, 775 (1996).
(1998). 86) S. Aoki, M. Shiro, T. Koike, E. Kimura, J. Am. Chem. Soc.,
122,
76) M. Shionoya, M. Sugiyama, E. Kimura, J. Chem. Soc. Chem. 576 (2000).
Commun., 1747 (1994). 87) S. Aoki, M. Shiro, E. Kimura, Chem. Eur. J., 8, 929
(2002).
77) E. Kimura, M. Kikuchi, H. Kitamura, T. Koike, Chem. Eur. J., 88) S. Aoki, S. Suzuki, M. Kitamura, M. Shiro, E. Kimura, Chem.
5,
3113 (1999). Asian J., in press (2012)
78) E. Kimura, H. Kitamura, K. Ohtani, T. Koike, J. Am. Chem. 89) S. Aoki, M. Zulkefeli, M. Shiro, E. Kimura, Proc. Natl. Acad.
Soc.,
122, 4668 (2000). Sci., 99, 4894 (2002).
79) S. Aoki, E. Kimura, J. Am. Chem. Soc., 122, 4542 (2000). 90) M. Zulkefeli, T. Sogon, K. Takeda, E. Kimura, S. Aoki, Inorg.
80) S. Aoki, C. Sugimura, E. Kimura, J. Am. Chem. Soc., 120, Chem., 48, 9567 (2009).
10094
(1998). 91) M. Zulkefeli, A. Suzuki, Y. Hisamatsu, M. Shiro, E. Kimura, S.
81) E. Kikuta, M. Murata, N. Katsube, T. Koike, E. Kimura, J. Am. Aoki, Inorg. Chem., 50, 10113 (2011).
Chem. Soc., 121, 5426 (1999).
E. Kikuta, T. Koike, E. Kimura, J. Inorg. Biochem., 79, 253
(2000).

Profile
Eiichi Kimura was born in 1938 in Shizuoka city, Japan. He received his Bachelor (1961) and
Master degree (1963) in Pharmaceutical Sciences both from Tokyo University.He then as a
Fulbright student moved to Department of Chemistry, Graduate School, University of North
Carolina at Chapel Hill, where he obtained Ph.D (1967) under the supervision of Professor James P.
Collman. He worked as a postdoctoral research fellow in 1967-1968 at Syntex Pharmaceutical
Research Institute at Palo Alto, and in 1969-1970 at the University of Chicago with Professor Jack
Halpern. In 1970 he became an associate professor and in 1978 was promoted to professor at
Medical School, Hiroshima University. In 1988―1990 he served as professor of Institute for
Molecular Science. In 1990 he came back to Medical School, Hiroshima University, where he
continued to be professor until he retired in 2002. In August 1992, he was given The 2nd Izatt-
Christensen International Macrocyclic Chemistry Award. In April 1996, he received The
Pharmaceutical Society of Japan Award. In 2003 he became president of Pharmaceutical Society of
Japan. In May, 2004 he received Purple Ribon National Award (Shiju-Housho). Since 2005 till now
he has been a visiting professor at Faculty of Science, Shizuoka University.
 47 Bull. Jpn. Soc. Coord. Chem. Vol. 59 (2012)

View publication stats