A chemical adaptor system designed to link a tumor-targeting device with a prodrug and an enzymatic trigger

Article abstract of DOI:10.1002/anie.200390108

Targeting tumors can be achieved by using a novel chemical adaptor system (see picture) comprising, for example, the anticancer drug epotoside, an N-(2-hydroxypropyl) methacrylamide (HPMA) copolymer as the tumor-targeting device, and catalytic antibody 38C2 as the triggering enzyme in a single entity. When the enzyme cleaves the substrate, it triggers a spontaneous reaction that releases the active drug from the targeting moiety.

Full text of DOI:10.1002/anie.200390108

Angewandte
Chemie
residual electron density, 2.19/ꢀ2.18 eäꢀ3.Compound 1 has
Chemical Adaptor Systems
ꢀ
crystallographic Ci(1) site symmetry such that the asymmetric
part of the crystal structure consists of 1/2 of a neutral cluster and
one solvated acetone molecule;b) 2¥2Me2CO: monoclinic; P21/
n; a ¼ 17.517(1), b ¼ 24.471(2), c ¼ 24.216(2) ä, b ¼ 106.233(1)8,
V¼ 9966.5(13) ä3; Z ¼ 2, 1calcd ¼ 2.878 Mgmꢀ3. MoKa data col-
lected at 173(2) K with SMART CCD 1000 area detector
diffractometer by 0.3w scans over 2q range 2.9 56.68;empirical
absorption correction (SADABS) applied (m(MoKa) ¼
4.900 mmꢀ1;max./min. transmission, 0.758/0.374). Anisotropic
refinement (1044 parameters;74 restraints) on 24577 independ-
ent reflections converged at wR2(F2) ¼ 0.158 with R1(F) ¼ 0.052
for I > 2s(I);GOF (on F2) ¼ 0.99;max./min. residual electron
A Chemical Adaptor System Designed To Link a
Tumor-Targeting Device with a Prodrug and an
Enzymatic Trigger**
Anna Gopin, Neta Pessah, Marina Shamis,
Christoph Rader, and Doron Shabat*
Selective chemotherapy remains a key issue for successful
treatment in cancer therapy. Prolonged administration of
effective concentrations of chemotherapeutic agents is usu-
ally not possible because of dose-limiting systemic toxicities.
Furthermore, strong side effects involving nonmalignant
tissues are often observed. Therefore, much effort has been
devoted to the development of new drug delivery systems that
mediate drug release selectively at the tumor site. One way to
achieve such selectivity is to activate a prodrug specifically by
a confined enzymatic activity. In this concept, the enzyme is
either expressed by the tumor cells, or brought to the tumor
by a targeting moiety such as a monoclonal antibody.[1] The
prodrug is converted to an active drug by the local or localized
enzyme at the tumor site, thereby minimizing nonspecific
toxicity to other tissues.
Here we present a new concept that combines a tumor-
targeting device, a prodrug, and a prodrug activation trigger
in a single entity. We designed a generic module or chemical
adaptor that is based on three chemical functionalities as
shown in Figure 1. The first functionality is attached to an
active drug and, thereby, masks it to yield a prodrug. The
second functionality is linked to a targeting moiety, which is
responsible for guiding the prodrug to the tumor site, and the
third functionality is attached to an enzyme substrate. When
the corresponding enzyme cleaves the substrate, it triggers a
spontaneous reaction that releases the active drug from the
targeting moiety. As a result, prodrug activation will prefer-
entially occur at the tumor site.
density, 2.13/ꢀ2.56 eäꢀ3. Compound 2 has crystallographic Ci(1)
ꢀ
site symmetry such that the asymmetric part of the crystal
structure consists of 1/2 of a neutral molecule and one solvated
acetone molecule. CCDC-195160 (1) and -195161 (2) contain the
supplementary crystallographic data for this paper. These data
can be obtained free of charge via www.ccdc.cam.ac.uk/conts/
retrieving.html (or from the Cambridge Crystallographic Data
Centre, 12, Union Road, Cambridge CB21EZ, UK;fax: ( þ
44)1223-336-033;or deposit@ccdc.cam.ac.uk).
[16] Observed electron counts of 372, 648, and 290 electrons were
obtained for 1, 2, and 3, respectively. Application of the Mingos
electron-counting model[17] for a close-packed metal cluster gives
rise to the following predicted electron counts: namely, 290 elec-
trons (i.e., 12 î 18(surface) þ 18(interior) þ 4 î 14(wingtip)) for
3 and 378 electrons (i.e., 12 î 24(surface) þ 34(two interior) þ
4 î 14(wingtip)) for 1. The highly condensed geometry of 2
prevents a reliable predicted electron count.
[17] a) D. M. P. Mingos, J. Chem. Soc. Chem. Commun. 1985, 1352;
b) D. M. P. Mingos, L. Zhenyang, J. Chem. Soc. Dalton Trans.
1988, 1657.
[18] J. Donohue, The Structures of the Elements, Wiley, New York,
1974, p. 216.
[19] In sharp contrast, corresponding mean radial Pd(i) Pd(s1)
separations found in face-condensed icosahedral-based Pd-core
geometries of high-nuclearity palladium carbonyl trimethyl-
phosphine clusters are approximately 0.14 ä shorter than the
mean tangential intrashell Pd(s1) Pd(s1) separations (that is, 2.73
(av) versus 2.87 ä (av)).[9c] This large difference is ascribed to
geometrically imposed radial compressions in icosahedral-based
polyhedra. The close agreement of the mean Pd Pd separations
determined for the radial and tangential (intrashell) connectiv-
ities in 1, 2, and 3 point to the absence of geometrical distortions
in the metal-core geometries of cuboctahedral-based systems.
The central core of our chemical adaptor (Scheme 1) is
based on 4-hydroxymandelic acid, which is commercially
available and has three functional groups suitable for linkage.
Group I is a carboxylic acid that is conjugated to a targeting
moiety through an amide bond. The drug is linked through
the benzyl alcohol group II, and the enzyme substrate is
attached through the phenol group III by a carbamate bond.
[*] Dr. D. Shabat, A. Gopin, N. Pessah, M. Shamis, Dr. C. Rader
Department of Organic Chemistry
School of chemistry, Faculty of Exact Sciences
Tel Aviv University
Tel Aviv 69978 (Israel)
Fax: (þ972)3-640-9293
E-mail: chdoron@post.tau.ac.il
[**] We thank Professor Richard A. Lerner and Professor Carlos F.
Barbas III of The Scripps Research Institute for providing antibo-
dy38C2 and TEVA Pharmaceutical Industries Ltd., Israel for a gift of
etoposide. We also thank Ms. Rajeswari Thayumanavan for early
contributions to this work.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2003, 42, No. 3
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0327 $ 20.00+.50/0
327

Communications
Targeting Ligand
Enzyme
Substrate
I
HOOC
III
OH
Enzyme Substrate
HO
II
Drug
Chemical
Adaptor
Scheme 1. General design of the chemical adaptor system. Cleavage of
the enzyme substrate generates an intermediate that spontaneously re-
arranges to release the drug from the targeting device.
Targeting
Device
Drug
Enzyme
Scheme 2 illustrates the release mechanism of the drug.
Cleaving the enzyme substrate generates a free amine group
(2) that spontaneously cyclizes to form a dimethyl urea
derivative and phenol 3. The latter undergoes spontaneous
rearrangement to give a quinone methide intermediate
(trapped by water to give benzyl alcohol 4), which releases
the drug in the form of carbamic acid, which then decarbox-
ylates spontaneously, yielding the active drug.
Chemical
Adaptor
Targeting
Device
Drug
Spontaneous
The design of our generic module allows us to potentially
link any targeting device to a variety of drugs and to release
them with any enzyme by using the corresponding substrate
as a trigger. As proof of the concept, we designed a pilot
system for which we chose catalytic antibody 38C2 as the
cleaving enzyme.[2] Antibody 38C2 catalyzes a sequence of
retro-aldol retro-Michael reactions, using substrates that are
not recognized by human enzymes, and thus, are ideal triggers
for selective prodrug activation. Antibody 38C2 has proven
its ability to catalyze the retro-aldol reactions with a broad
range of substrates[3] and recently was conjugated to an N-(2-
hydroxypropyl)methacrylamide (HPMA) copolymer for tu-
mor targeting.[4] Furthermore, antibody 38C2 has demon-
Chemical
Adaptor
Drug
Targeting
Device
Figure 1. 4-Hydroxymandelic acid, the central core of the chemical
adaptor system.
Targeting Device
Targeting Device
NH
NH
O
O
NH
Enzyme Substrate
N
O
O
Enzyme
O
N
O
N
O
O
O
HN
DRUG
O
HN
O
1
2
DRUG
Targeting Device
NH
O
NH2
Targeting Device
Spontaneous
Spontaneous
O
OH
NH
DRUG
O
HN
O
O
OH
H2O
CO2
4
3
HO
N
N
DRUG
Scheme 2. General mechanism of drug release starting with specific enzymatic cleavage.
328
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0328 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 3

Angewandte
Chemie
strated its efficacy in activating several prodrugs in vitro and
in vivo.[5] A dramatic 75% decrease in subcutaneous (s.c.)
tumor size has been observed in mice that received an
intratumoral injection of antibody 38C2 and systemic treat-
ments with an etoposide prodrug.[6]
As a targeting device we chose a polymer molecule for our
pilot system. Water-soluble synthetic polymers such as
HPMA copolymers are biocompatible, nonimmunogenic,[7]
and nontoxic. Moreover, their in vivo distribution is well
characterized[8] and they are known to accumulate selectively
at tumor sites due to the enhanced permeability and retention
(EPR) effect.[9] This effect occurs because of the difference
between the vasculature physiology of solid tumors and
normal tissues. Compared with the regular ordered vascula-
ture of normal tissues, blood vessels in tumors are often highly
abnormal. The growth of the tumor creates a constant need
for the continuous supply of new blood vessels. This process,
termed angiogenesis, often results in the construction of
vessels with leaky walls, which allows enhanced permeability
of macromolecules within the tumor. In addition, poor
O
O
O
N
EDC, HOBT, 49%
HO
N
p-nitrophenyl chloroformate
O
OH
OH
HO
HO
67%
Boc-dimethyl ethylendiamine
5
O
O
N
Et3N, 90%
O
N
O
O
O
NO2
HO
O
HO
O
6
H2N
CF3CO2
N
O
O
O
N
p-nitrophenyl chloroformate
O
O
HO
N
O
O
N
Et3N, 72%
N
O
HO
O
7
O
O
N
O
O
HO
N
O
O
N
N
O
O
O
O
O
DRUG-NH2
71%
8
O2N
O
O
N
O
O
HO
N
O
O
N
N
O
O
O
1. TFA, 2. POLYMER
Et3N, 88%
HN
O
9
DRUG
O
POLYMER
N
O
O
HO
N
O
O
N
N
O
O
O
HN
O
DRUG
10
Scheme 3. Synthesis of the chemical adaptor system using catalytic antibody 38C2 as the activating enzyme, an HPMA copolymer as a targeting
device, and anydrug with a functional amine group. EDC ¼1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide, HOBT¼1-hydroxybenzotriazole,
TFA¼trifluoroacetic acid.
Angew. Chem. Int. Ed. 2003, 42, No. 3
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0329 $ 20.00+.50/0
329

Communications
lymphatic drainage at the tumor site promotes accumulation
clinical use by the FDA in the late 1970s, after it demonstrated
efficacy in chemotherapy. Etoposide exhibits its antitumor
activity by inhibition of enzymes that cut and pass double-
stranded DNA called topoisomerase inhibitors.[11] The phenol
group of etoposide was used to link it to the complex through
a carbamate bond (see Scheme 4).
of large molecules. Based on the tumor-targeting properties
of molecules derivatized with HPMA copolymers, polymer-
directed enzyme prodrug therapy (PDEPT) was developed by
Duncan et al. as a novel approach toward selective chemo-
therapy.[10]
Etoposide was selected as the drug in our pilot system. It is
a well-established anticancer drug, which was approved for
Starting from 4-hydroxymandelic acid, we built the
complex as described in Scheme 3. In brief, 4-hydroxyman-
CH3
C
CH3
C
CH2
CH2
CO
CO
NH
CH2
CO
NH
CH2
NH
x
y
CH2
CH2OH
CH3
x = 95 mol%
y = 5 mol%
O
OC
N
O
O
HO
N
N
O
N
N
O
O
O
Antibody 38C2
O
11
N
O
O
O
H3CO
OCH3
O
O
N
N
CO2
O
O
O
O
OH
O
HO
O
O
O
H3C
CH3
C
CH3
CH3
C
CH3
C
H2O
CH2
CH2
C
CH2
CH2
CO
CO
CO
NH
CH2
CO
NH
CH2
CO
NH
CH2
CO
NH
CH2
NH
x
y
NH
x
y
CH2
CH2OH
CH3
CH2
CH2OH
CH3
O
CO2
N
N
x = 95 mol%
y = 5 mol%
x = 95 mol%
y = 5 mol%
O
O
O
OC
N
OC
N
13
N
N
N
OH
OH
HO
O
N
12
O
O
OH
H3CO
OCH3
H3CO
OCH3
O
O
O
O
O
O
O
O
OH
O
HO
OH
O
O
HO
O
H3C
O
O
O
H3C
O
Etoposide
Scheme 4. Mechanism of etoposide drug release from the HPMA-copolymer, using catalytic antibody 38C2 as the triggering enzyme.
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 1433-7851/03/4203-0330 $ 20.00+.50/0 Angew. Chem. Int. Ed. 2003, 42, No. 3
330

Angewandte
Chemie
delic acid was coupled with monoprotected Boc-dimethyle-
thylenediamine in the presence of EDC and HOBT to give
amide 5. Selective acylation of the phenol group with p-
nitrophenyl chloroformate afforded carbonate 6, which was
treated with antibody 38C2- substrate to give compound 7.
Acylation of the benzylic alcohol with p-nitrophenyl chlor-
oformate gave carbonate 8, which was treated with an amine
drug to give compound 9. The Boc protecting group was
removed by TFA to give the corresponding amine salt, which
was immediately coupled with the p-nitrophenyl ester of the
HPMA copolymer to give the final complex 10. The drug
polymer conjugate was purified by dialysis against water. By
measuring its UV spectrum, the number of drug molecules
attached to the HPMA copolymer was determined. Based on
the etoposide chromophore, which has a lmax value at 280 nm,
we found that on average three molecules of the drug were
linked to one molecule of the HPMA copolymer.
Next we tested whether the etoposide drug can be
released from complex 11 by the catalytic activity of
antibody 38C2. According to our design, the drug should be
spontaneously released after the generation of phenol 12 as
illustrated in Scheme 4. We incubated complex 11 with
catalytic antibody 38C2 in phosphate-buffered saline (PBS)
(pH 7.4) at 378C and monitored the appearance of etoposide
using an HPLC assay. As a positive control, we used a
previously described etoposide prodrug[6] that is activated by
antibody 38C2. As Figure 2 shows, etoposide was released by
the catalytic activity of antibody 38C2 to form compound 13
and the free drug. The rate of drug release was similar to the
activation rate of the known etoposide prodrug. No sponta-
neous etoposide release was observed in the absence of the
antibody.
a carbamate linkage. This strategy is convenient when the
drug has an available amine group such as in the case of
doxorubicin. Drugs like etoposide, which have free hydroxy
groups that need to be masked, are linked through an N,N-
dimethylethylenediamine spacer, which spontaneously im-
molates to form a dimethyl urea derivative (see Scheme 4).
A similar and elegant approach of enzymatic cleavage
leading to a small-molecule release, was recently published by
Waldmann et al.[12] An enzyme-labile linker group was
attached to a solid support and a third functionality was used
for solid-phase synthesis of a target molecule. The small
molecule was detached from the polymeric support, starting
by enzymatic cleavage, leading to generation of an inter-
mediate that spontaneously released the product. In one
example a quinone methide rearrangement was used to
release the target molecule, similarly to our approach.[13] The
intermediate was generated by enzymatic cleavage of an ester
group, which would result in a high background if used for in
vivo prodrug activation. Whereas ester functionalities are
stable in organic solvents, they undergo rapid hydrolysis in
aqueous solutions at physiological pH. In addition, endoge-
nous enzymes with esterase activity further contribute to
nonspecific prodrug activation. In view of these facts, we
designed a chemical adaptor system that does not contain any
ester linkages but utilizes rather stable amide and carbamate
bonds.
In conclusion, we have developed a drug delivery system
based on a chemical adaptor that provides a generic linkage of
a drug with a targeting device in a manner suitable for
triggering by defined enzymatic activity. The system is generic
and can be applied to a variety of drugs, targeting devices, and
enzymes by introducing the corresponding substrate as a
trigger for drug release in the chemical adaptor. The chemical
adaptor system was designed with stable chemical linkages to
avoid nonspecific drug release in vivo. Proof of the concept
was demonstrated by using etoposide as the drug, an HPMA
copolymer as the targeting device, and catalytic antibo-
dy 38C2 as the triggering enzyme. Currently we are synthesiz-
ing a variety of chemical adaptor systems using different
tumor-targeting devices, prodrugs, and enzymatic triggers.
Two additional examples, comprising a penicillin amidase
substrate as the enzymatic trigger and camptothecin as the
anticancer drug, respectively, are described in the Supporting
Information. In vivo studies aimed at testing our chemical
adaptor concept for the selective targeting of tumors have
been initiated.
The general synthetic approach of introducing the drug in
our chemical adaptor system as outlined in Scheme 3,
involves a reaction between an amine group of the drug and
a p-nitrophenyl carbonate of a hydroxybenzyl alcohol to form
Received: May 21, 2002
Revised: October 22, 2002 [Z19328]
[1] K. D. Bagshawe, S. K. Sharma, C. J. Springer, G. T. Rogers, Ann.
Oncol. 1994, 5, 879 891.
Figure 2. Determination of etoposide release from the HPMA copoly-
mer versus drug formation from a known etoposide prodrug bycatalyt-
[2] J. Wagner, R. A. Lerner, C. F. Barbas III, Science 1995, 270,
1797 1800.
^
ic antibody38C2.
Complex 11 (500 mm) with antibody38C2 (66 mm)
[3] T. Hoffmann, G. Zhong, B. List, D. Shabat, J. Anderson, S.
Gramatikova, R. A. Lerner, C. F. Barbas III, J. Am. Chem. Soc.
1998, 120, 2768 2779.
[4] R. Satchi-Fainaro, W. Wrasidlo, H. N. Lode, D. Shabat, Bio. Org.
Med. Lett. 2002, 10, 3023 3029.
&
in PBS-7.4. Etoposide prodrug (500 mm) with antibody38C2 (66 mm)
in PBS (pH 7.4). Reactions were incubated at 378C for the indicated
time. The drug release concentration c was monitored byHPLC analy-
sis.
Angew. Chem. Int. Ed. 2003, 42, No. 3
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0331 $ 20.00+.50/0
331

Communications
[5] D. Shabat, C. Rader, B. List, R. A. Lerner, C. F. Barbas III, Proc.
by the supramolecular ordering and spatial relationship of the
donor and acceptor chromophores. Molecular architectures
such as hydrogen-bonded systems,[1] dendrimers,[2] chromo-
phore-linked polymers,[3] Langmuir Blodgett films,[4] and
self-assembled monolayers[5] are of extremely important in
this context. Among a plethora of donor acceptor systems
investigated, energy and electron transfer from oligo(phenyl-
enevinylene)s (OPVs) and poly(phenylenevinylene)s (PPVs)
to acceptors, such as C60,[6] phenanthroline,[7] and doped
organic dyes[8] have generated enormous interest because of
their potential use in photovoltaic and light-emitting devices.
From this view point, the recent reports by Meijer and co-
workers on energy and electron transfer from OPV-function-
alized dendrimers and supramolecular assemblies to various
acceptors are of particular interest.[9] This situation has
prompted us to investigate on the potential of OPV based
organogels for the purpose of energy transfer and light
harvesting.[10]
Natl. Acad. Sci. USA 1999, 96, 6925 6930.
[6] D. Shabat, H. Lode, U. Pertl, R. A. reisfeld, C. Rader, R. A.
Lerner, C. F. Barbas III, Proc. Natl. Acad. Sci. USA 2001, 98,
7528 7533.
[7] B. Rihova, M. Bilej, V. Vetvicka, K. Ulbrich, J. Strohalm, J.
Kopecek, R. Duncan, Biomaterials 1989, 10, 335 342.
[8] L. W. Seymour, K. Ulbrich, J. Strohalm, J. Kopecek, R. Duncan,
Biochem. Pharmacol. 1990, 39, 1125 1131.
[9] H. Maeda, J. Wu, T. Sawa, Y. Matsumura, K. Hori, J. Controlled
Release 2000, 65, 271 284.
[10] R. Satchi, T. A. Connors, R. Duncan, Br. J. Cancer 2001, 1070
1076.
[11] K. R. Hande, Eur. J. Cancer 1998, 34, 1514 1521.
[12] U. Grether, H. Waldmann, Chem. Eur. J. 2001, 7, 959 971.
[13] B. Sauerbrei, V. Jungmann, H. Waldmann, Angew. Chem. 1998,
110, 1187 1190; Angew. Chem. Int. Ed. 1998, 37, 1143 1146.
Small-molecule-based organogels have attracted much
attention in recent years because of their interesting physical
properties and architectural elegance.[11] However, organo-
gels based on p-conjugated systems are relatively very few.[12]
Recently, we have reported hydrogen-bond- and p-stack-
induced supramolecular assembly of the OPV derivatives
1a,b, which leads to the formation of entangled nanostruc-
tures, which induce gelation of hydrocarbon solvents.[13] The
absorption and emission properties of 1a,b showed dramatic
changes during gelation, which is an indication of strong
intermolecular p-electronic coupling of the ordered self-
assembled OPV gel. Excitation of 1a,b in cyclohexane at 380
and 470 nm revealed the emission corresponding to the
monomeric species (lem ¼ 455 and 483 nm, Ff ¼ 0.40 ꢁ 0.01
Light Harvesting with Gels
Gelation-Assisted Light Harvesting by Selective
Energy Transfer from an Oligo(p-
phenylenevinylene)-Based Self-Assembly to an
Organic Dye**
Ayyappanpillai Ajayaghosh,* Subi J. George, and
Vakayil K. Praveen
Dedicated to the memory of Darshan Ranganathan
Excitation energy transfer within and between
molecular systems plays an important role in
natural processes, such as photosynthesis. In this
context, there has been widespread interest in
mimicking the mechanism of solar-energy har-
vesting of natural photosynthesis with the aid of
synthetic molecular systems. Apart from this,
photoinduced energy transfer has become sig-
nificant in the area of photovoltaics, organic
light-emitting diodes, fluorescent labeling, and
in a variety of photonic devices. In most of these
cases, energy transfer is considerably influenced
[*] Dr. A. Ajayaghosh, S. J. George, V. K. Praveen
PhotochemistryResearch Unit
related to quinine sulfate, t ¼ 1.46 ns;Figure 1, spectrum b)
and self-assembled species (lem ¼ 525 and 565 nm, Ff ¼ 0.40 ꢁ
0.01 related to Rhodamine 6G, t ¼ 1.62 ns;Figure 1, spec-
trum c), which showed strong dependency on solvent polarity
and temperature.
Herein we describe an interesting case of a thermorever-
sible fluorescence-resonance energy transfer (FRET) and
light harvesting, exclusively from OPV based supramolecular
gel nanostructures of 1a,b to an organic dye. To study the
feasibility of such an energy transfer, we chose Rhodamine B
as the acceptor, the absorption (lmax ¼ 555 nm) of which
Regional Research Laboratory, CSIR
Trivandrum-695019 (India)
Fax: (þ91)471-490-186
E-mail: aajayaghosh@rediffmail.com
[**] This work was supported bythe Department of Science and
Technology, Government of India under the Swarnajayanti scheme
(DST/SF/C6/99-2000). We thank Mr. P. Prabhakar Rao for SEM
analysis and CSIR for a research fellowship to S.J.G. This is
contribution number RRLT-PRU-155.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
332
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0332 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 3