A fragment-based in situ combinatorial approach to identify high-affinity ligands for unknown binding sites

Article abstract of DOI:10.1002/anie.200907254

[Figure Presented] In the lead: The title method for the identification of ligands is particularly useful for binding sites where little or no structural information is available. In a fragment-based approach, a suitable pair of first- and second-site ligands is identiled by NMR experiments. By applying a receptor-mediated in situ combinatorial approach, the two ligands are then linked to generate a new high-affinity lead structure (see picture).

Full text of DOI:10.1002/anie.200907254

Angewandte
Chemie
DOI: 10.1002/anie.200907254
Combinatorial Chemistry
A Fragment-Based In Situ Combinatorial Approach To Identify High-
Affinity Ligands for Unknown Binding Sites**
Sachin V. Shelke, Brian Cutting, Xiaohua Jiang, Hendrik Koliwer-Brandl, Daniel S. Strasser,
Oliver Schwardt, Soerge Kelm, and Beat Ernst*
Dedicated to Professor Robert E. Ireland
With his publication “On the attribution and additivity of
binding energies” in 1981, Jencks[1] launched a new area in
medicinal chemistry, that is, the so-called fragment-based
drug discovery (FBDD).[2] When fragments are linked, their
individual binding energies are additive. In addition, because
of the reduction of translational and rigid body rotational
degrees of freedom, the entropy barrier is markedly low-
ered.[3] Thus, by linking two low-affinity fragments, a new
ligand with a substantially improved affinity for the target can
be generated. However, this intriguing concept resulted in
only a few scattered applications[4,5] and had no immediate
impact on drug discovery. For a practical application of this
strategy two problems remained to be solved; firstly, how
suitable fragments that bind to proximal binding sites (so-
called first- and second-site fragments) can be identified, and
secondly, how these fragments can be linked without dis-
tortions of their individual binding modes.
example, Sharpless and co-workers,[9] who used the target
itself as an atomic-scale reaction vessel for creating its own
inhibitor or by applying a shape-modulating linker design.[10]
Herein we present a novel fragment-based approach that
does not require any spatial information on the binding site
and can be conducted with modest amounts of unlabeled
protein. Our target is the myelin-associated glycoprotein
(MAG, Siglec-4), a sialic acid binding immunoglobulin-like
lectin (Siglec),[11] which inhibits as one of several myelin
components axonal regrowth after injury.[12] The recently
reported use of monovalent glycosides[13] to reverse MAG-
mediated blocking of axonal regeneration encouraged the
search for high-affinity ligands. Oligo- and monosaccharide
derivatives based on the ganglioside GQ1ba,[14] which was the
hitherto best reported natural MAG antagonist, exhibit only
micromolar affinities.[15] Therefore, an alternative approach to
identify high-affinity ligands was required.
The rapid development of this promising area[6] was
initiated in 1996, when a conclusive practical demonstration
of FBDD, called structure–affinity relationship by NMR
(SAR-by-NMR) was reported.[7] With this novel approach,
antagonists with nanomolar affinities were rapidly identified
by tethering two fragments that were individually optimized
by NMR spectroscopy. However, the implementation range
of this technique was limited by the requirement for labeled
proteins (13C and 15N) and for structural information on the
binding site in order to design the linker. Subsequently, a
broad array of innovative strategies for screening fragments
were reported, for example, the needle approach[5] or tether-
ing techniques detected by mass spectrometry.[8] Furthermore,
the problem of the linker design was addressed by, for
Because the crystal structure of MAG is not yet available,
a homology model[16] based on the crystal structure of
sialoadhesin (Siglec-1),[17] another member of the Siglec
family, was investigated. This model revealed a shallow,
unstructured binding site, which does not provide any obvious
additional interaction sites and therefore little prospect for
success by a structure-based approach. This result prompted
us to develop a novel, three-step fragment-based in situ
combinatorial approach, which is especially suited if little or
no structural information on the binding site is available.[18]
A
first-site ligand with a moderate, that is, micromolar affinity,
either based on a physiological ligand or identified by random
screening, serves as starting point. In order to search for
second-site ligands, members of a fragment library, which
bind to the target protein, are identified by an NMR
experiment that is based on the change of their transverse
magnetization decay upon binding (Figure 1a).[19]
[*] Dr. S. V. Shelke, Dr. B. Cutting, Dr. X. Jiang, Dr. D. S. Strasser,
Dr. O. Schwardt, Prof. Dr. B. Ernst
From these hits, fragments that bind to a second site
located adjacent to the first site are identified by their
enhanced paramagnetic relaxation caused by the spin-labeled
first-site ligand (Figure 1b). Successful applications of spin
labels for ligand screening[20a–c] as well as the characterization
of binding sites[20d] have already been reported. Because the
paramagnetic relaxation is distance-dependent, not only
fragments that bind at a proximal subsite, but also their
spatial orientation and hence the correct linking point can be
elucidated.[20e] Finally, the target protein is incubated with a
library of first- and second-site fragments functionalized with
azido or acetylene groups at the end of flexible methylene
chains of variable length (Figure 1c). Only in the case of an
Institute of Molecular Pharmacy, University of Basel
Klingelbergstrasse 50, 4056 Basel (Switzerland)
Fax: (+41)61-267-1552
E-mail: beat.ernst@unibas.ch
H. Koliwer-Brandl, Prof. Dr. S. Kelm
Department of Physiological Biochemistry, University of Bremen
28334 Bremen (Germany)
[**] We would like to thank the Volkswagen Foundation, the Swiss
National Science Foundation, the German Federal Ministry for
Education and Research (BMBF, project 031632A) and the Tꢀnjes-
Vagt Foundation (project XXI) for their support of this work. We are
indebted to Jonas Egger for the graphic artwork (Figure 1).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200907254.
Angew. Chem. Int. Ed. 2010, 49, 5721 –5725
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5721

Communications
Figure 1. Fragment-based in situ combinatorial approach for the iden-
tification of ligands for an unknown binding site. a) Identification of
second-site ligands based on their transverse magnetization decay
upon binding; b) selection of second-site ligands binding adjacent to
the spin-labeled first-site ligand; c) incubation of the target protein
with libraries of first- and second-site ligands; d) a high-affinity ligand
is generated by receptor-mediated triazole formation. The receptor is
shown in green.
optimal spatial orientation, which results from the binding of
two compatible fragments to MAG, is a high-affinity ligand
generated by a receptor-mediated triazole formation (Fig-
ure 1d).[9]
For our investigations, recombinant MAG, which consists
of the three N-terminal domains of MAG and the Fc fragment
of human antibody immunoglobulin G (IgG; MAGd1-3-Fc)
was expressed in chinese hamster ovary (CHO) cells and
affinity purified on protein A agarose.[21] As a first-site ligand,
the sialic acid derivative 1 with 137 mm MAG affinity[15c] was
selected (Figure 2).
Our fragment library for the identification of second-site
ligands was composed of 80 low-molecular-weight (Mw <
300), moderately lipophilic (clogP < 3; P = partition coeffi-
cient octanol/water) and highly soluble organic molecules.
Whereas the limitations of Mw and logP for fragments are well
documented (Rule of Three),[22] water solubility is a pre-
requisite for the success of the planned NMR spectroscopic
screens. The fragment library was divided into sublibraries,
each of which contained 6 to 10 compounds. The sublibraries
were composed according to two criteria: 1) the members of a
sublibrary do not interact with each other and 2) each
component shows at least one isolated resonance in the
1H NMR spectrum of the sublibrary.
Members of the sublibraries that bind to MAGd1-3-Fc were
identified based on the change of their transverse magnet-
ization decay, which occurs upon complex formation.[23] The
magnetization decay is molecular-weight-dependent and
therefore allows a differentiation between small ligands,
that is, fragments in solution, and large molecules, that is,
fragments that are bound to the receptor. When the 1H NMR
spectra of the sublibraries were recorded at relaxation times
of 10 and 200 ms, the low-molecular-weight fragments of the
sublibrary experienced only small magnetization decays.
When MAGd1-3-Fc was added, those fragments that bind to
the protein, that is, those that formed a fragment–protein
complex, behaved as large molecules and therefore suffered
Figure 2. Principle of second-site screening. a) 1H NMR resonance of
H-C(4) of 5-nitroindole 4; b) decay of transverse magnetization caused
by binding of 4 to MAGd1À3-Fc; c) paramagnetic relaxation enhance-
ment on 4 caused by the spin-labeled first-site ligand 2*; d) recovery of
the paramagnetic relaxation enhancement by the addition of ascorbic
acid, which reduced 2* to 3.
from large magnetization decays. A number of fragments of
the various sublibraries were identified to bind on the surface
of the target protein. However, their exact binding site was
still not known.
For the identification of the hits that bound adjacent to
the binding site of 1, spin–spin relaxation NMR experiments
reported by Jahnke et al.[20a] were applied. Spin–spin relaxa-
tion rates are proportional to the product of the squares of the
gyromagnetic ratio g of the involved spin. Although it is the
second highest among all isotopes, the g value of protons and
hence also the spin–spin relaxation rate are small. However,
for the detection of ligand binding, the spin–spin relaxation
rates should be as large as possible. Because the g value of an
unpaired electron is 658 times larger than that of a proton, the
spin–spin relaxation rates caused by an unpaired electron on
protons (known as paramagnetic relaxation enhancement)
are dramatically larger than the effect of a nuclear–nuclear
interaction.[24] Therefore, the spin-labeled first-site ligand 2*,
5722
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5721 –5725

Angewandte
Chemie
which contains an unpaired electron, was synthesized
(Figure 2; for the synthesis see the Supporting Information).
Compound 2* is a conjugate of first-site ligand 1 and the
TEMPO (2,2,6,6-tetramethylpiperidin-1-oxyl) spin label.
Compound 1 and the N-hydroxy derivative 3, which is the
reduced form of TEMPO derivative 2*, exhibited similar
affinities (1: 137 mm; 3: 96 mm) in surface plasmon resonance
(SPR) experiments[15c] as well as comparable transfer of
magnetization from MAG in saturation transfer difference
NMR (STD NMR).[25] It was therefore assumed that the
TEMPO spin label does not substantially alter the binding
mode of the first-site ligand.
The sequence of NMR experiments required for the
identification of a second-site ligand binding adjacent to the
first-site ligand is shown for the second-site ligand 5-nitro-
indole 4 in Figure 2. When the 1H NMR spectrum was
recorded at two relaxation times, 10 and 200 ms, only a
minimal decay of magnetization—as exemplified for H-C(4)
of 4—was observed (Figure 2a). Addition of MAGd1-3-Fc led
to a pronounced decay of transverse magnetization, thus
indicating binding of 4 (Figure 2b). When the spin-labeled
first-site ligand 2* was added, a further enhancement of
paramagnetic relaxation was observed, therefore suggesting
simultaneous binding of 2* and 4 at neighboring binding sites
(Figure 2c).[20a] Finally, when the spin label was reduced by
adding ascorbic acid to the NMR sample, the effect was
cancelled, thus attributing the paramagnetic relaxation effect
unambiguously to the spin label (Figure 2d).
Because the paramagnetic relaxation enhancement
caused by the spin label is distance-dependent and is there-
fore potentially different for each proton of the second-site
ligand,[24] the spatial orientation of 4 could be determined.
The effect is largest for H-C(2), thus indicating that this
proton is facing the spin label. Therefore, the carbon atoms of
the pyrrole ring of 5-nitroindol 4 represent the optimal linking
positions (see the Supporting Information).
For the linking of an appropriate first- and second-site
ligand, the in situ click chemistry approach reported by
Sharpless and co-workers[9] was applied. In this approach,
the triazole formation does not result from copper catalysis
but from the receptor-mediated preorganization of azide and
acetylene to enable the [3+2] cycloaddition. For libraries of
first- (5a–d) and second-site ligands (6a–c; Scheme 1, for
synthetic details see the Supporting Information), two
requirements had to be met. Firstly, flexible spacer arms
should guarantee that the two terminal functionalities, that is,
the acetylene and the azido groups attached to the first- and
second-site ligand, can adopt the optimal spatial orientation
to undergo a [3+2] Huisgen cycloaddition reaction.[26] Sec-
ondly, the entropic penalties associated with rotatable bonds
should be kept low by the short spacers.[27] This requirement is
in agreement with the reach of the spin label of about 10 ꢀ.[24]
In previous in situ click chemistry studies,[9a–d] building
blocks with nano- to low micromolar affinity for the target
have been utilized. Because our ligands exhibit affinities in
the range of 100 mm to millimolar as estimated by SPR[15c] and
the fluorescent hapten binding assay,[21] their receptor-medi-
ated linking represents a borderline case similar to that
already discussed by Fokin et al.[9e]
Scheme 1. The library of first-site ligands consisted of four members,
5a–d, the library of second-site ligands of three, 6a–c. In total, 24
different triazoles, 12 anti-triazoles and 12 syn-triazoles could be
formed.
For the in situ click experiment, MAGd1À3-Fc[21] (6.47 mm)
was incubated with a mixture of the four first-site ligands
5a–d (each 330 mm) and the three second-site ligands 6a–c
(each 660 mm) in phosphate buffer (pH 7.4) at 378C for 3 days,
thus in principle permitting the formation of 12 syn-sub-
stituted and 12 anti-substituted triazoles. The differing con-
centrations of the first- and second-site ligands were used to
compensate for their different affinities for MAG (see above).
The analysis of the supernatant by LC–MS in selected ion
mode (SIM) after 3 days showed one major new molecular
ion of m/z 694.2, which fits for three syn- and three anti-
substituted triazoles, formed by [3+2]cycloaddition of 5a and
6c, 5b and 6b, and 5c and 6a, respectively. In addition, the
other possible molecular ions were also present in low, but
still detectable amounts. A control experiment carried out in
the absence of MAG showed that the cycloaddition products
were formed at comparable rates, which can be attributed to
uncatalyzed background reactions.
The differentiation between the three syn/anti pairs with
the same molecular weight but different spacer patterns was
possible by using mass spectrometry (Scheme 2 and the
Supporting Information). In the MS–MS mode with negative
ionization, only two major fragments, with m/z 300.1 and
m/z 393.1, could be detected beside the molecular ion
(m/z 694.2). These species are formed by fragmentation of
the glycosidic bond. Although the collision energy was varied
from 5 to 50 V, fragments originating from exocyclic a frag-
mentation of the triazole could not be identified. However, a
fragment m/z 492.2 was identified by SIM; this fragment can
only be obtained by a fragmentation of the syn- and anti-
triazoles 7 formed from 5a and 6c (see the Supporting
Angew. Chem. Int. Ed. 2010, 49, 5721 –5725
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5723

Communications
Scheme 2. As differentiation between anti-7 and syn-7 was not possible
by spectroscopic means, they were both synthesized (see the Support-
ing Information) and biologically evaluated. Differentiation can be
unambiguously made between the three possible syn/anti pairs with a
molecular weight of 694.2 according to the fragments obtained by
a fragmentation.
Figure 3. The binding epitope of anti-7 was mapped through
STD NMR experiments. The percentages shown in the structure
indicate the level of transfer of magnetization to a particular hydrogen
atom relative to transfer of magnetization received by the N-acetyl
methyl group. The large percentages on the benzamide hydrogens are
consistent with previously reported SAR studies.[15d] In addition, large
percentages were identified on the nitroindole, thus rationalizing the
enhanced affinity of anti-7.
Information). As a further differentiation between the syn
and anti product was not possible, syn-7 and anti-7 were
synthesized under thermal conditions (see the Supporting
Information) and biologically tested. In an SPR experiment
with MAGd1-3-Fc immobilized on a protein A modified chip
surface,[15c] KD values of 190 nm for anti-7 and 2 mm for syn-7
were obtained (Scheme 2). This result leads to the assumption
that anti-7 was formed in the in situ click experiment,
although the formation of both, anti-7 and syn-7 can not yet
be excluded. An MAG antagonist that exhibits a 1000-fold
improved affinity relative to the tetrasaccharide epitope of
GQ1ba (KD: 180 mm[15b]) was identified for anti-7.
successful, for example, for the trimannoside-DC-SIGN
interaction.[15a] When applied to these proteins, the presented
fragment-based in situ combinatorial approach could be
valuable in supporting the identification of high-affinity
glycomimetics and could develop into a useful tool in drug
discovery.
STD NMR[25] experiments showed large contributions to
binding from both terminal aromatic groups, the benzamide
and the 5-nitroindole moiety (Figure 3). In addition, the
triazole linking the first-site ligand 5a and the second-site
ligand 6c also contributes to the interaction with MAG. This
observation is consistent with previous experiments where the
replacement of the methyl aglycon in sialoside 1 by a benzyl
substituent led to an affinity enhancement that was approx-
imately 10-fold.[15h]
Received: December 23, 2009
Revised: May 3, 2010
Published online: July 7, 2010
Keywords: in situ click chemistry ·
.
fragment-based in situ combinatorial approach · drug discovery ·
glycoproteins · NMR spectroscopy
In summary, the identification of a second-site ligand by
NMR screening using a spin-labeled first-site ligand[20a]
followed by
a receptor-mediated linking of first- and
second-site ligand[9] yielded a nanomolar MAG antagonist,
which is currently under further biological evaluation. The
synthetic effort is substantially reduced compared to a
conventional approach. Beside the composition of the
second-site library and the synthesis of the first-site ligand
substituted with TEMPO, only seven compounds had to be
synthesized to successfully identify a high-affinity antagonist.
A particular strength of this approach is that no structural
information of the target protein is required.
[1] W. P. Jencks, Proc. Natl. Acad. Sci. USA 1981, 78, 4046 – 4050.
[2] For recent reviews, see: a) P. J. Hajduk, J. Greer, Nat. Rev. Drug
Discovery 2007, 6, 211 – 219; b) G. Chessari, A. J. Woodhead,
Drug Discovery Today 2009, 14, 668 – 675; c) C. W. Murray, D. C.
Rees, Nat. Chem. 2009, 1, 187 – 192.
[3] A. V. Finkelstein, J. Janin, Protein Eng. 1989, 3, 1 – 3.
[4] C. L. M. J. Verlinde, G. Rudenko, W. G. J. Hol, J. Comput.-Aided
Mol. Des. 1992, 6, 131 – 147.
[5] H. J. Boehm, M. Boehringer, D. Bur, H. Gmuender, W. Huber,
W. Klaus, D. Kostrewa, H. Kuehne, T. Luebbers, N. Meunier-
Keller, F. Mueller, J. Med. Chem. 2000, 43, 2664 – 2674.
[6] Fragment-based Approaches in Drug Discovery (Eds: W. Jahnke,
D. A. Erlanson), Wiley-VCH, Weinheim, 2006.
The identification of a nanomolar mimetic of the physio-
logical ligand is extremely tedious, especially for the shallow
binding sites of various lectins such as the sLex-E-selectin
interaction. In many cases the approach is only partially
[7] S. B. Shuker, P. J. Hajduc, R. P. Meadows, S. W. Fesik, Science
1996, 274, 1531 – 1534.
5724
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5721 –5725

Angewandte
Chemie
[8] D. A. Erlanson, A. C. Braisted, D. R. Raphael, M. Randal, R. M.
Stroud, E. M. Gordon, J. A. Wells, Proc. Natl. Acad. Sci. USA
2000, 97, 9367 – 9372.
Schwardt, B. Ernst, Bioorg. Med. Chem. 2006, 14, 4944 – 4957;
g) O. Blixt, K. Allin, L. Pereira, A. Datta, J. C. Paulson, J. Am.
Chem. Soc. 2002, 124, 5739 – 5746; h) S. Kelm, R. Brossmer, R.
Isecke, H.-J. Gross, K. Strenge, R. Schauer, Eur. J. Biochem.
1998, 255, 663 – 672.
[9] a) W. G. Lewis, L. G. Green, F. Grynszpan, Z. Radic, P. R.
Carlier, P. Taylor, M. G. Finn, K. B. Sharpless, Angew. Chem.
2002, 114, 1095 – 1099; Angew. Chem. Int. Ed. 2002, 41, 1053 –
1057; b) R. Manetsch, A. Krasinski, Z. Radic, J. Raushel, P.
Taylor, K. B. Sharpless, H. C. Kolb, J. Am. Chem. Soc. 2004, 126,
12809 – 12818; c) V. P. Mocharla, B. Colasson, L. V. Lee, S.
Roeper, K. B. Sharpless, C.-H. Wong, H. C. Kolb, Angew. Chem.
2005, 117, 118 – 122; Angew. Chem. Int. Ed. 2005, 44, 116 – 120;
d) A. Krasinski, Z. Radic, R. Manetsch, J. Raushel, P. Taylor,
K. B. Sharpless, H. C. Kolb, J. Am. Chem. Soc. 2005, 127, 6686 –
6692; e) M. Whiting, J. Muldoon, Y.-C. Lin, S. M. Silverman, W.
Lindstrom, A. J. Olson, H. C. Kolb, M. G. Finn, K. B. Sharpless,
J. H. Elder, V. V. Fokin, Angew. Chem. 2006, 118, 1463 – 1467;
Angew. Chem. Int. Ed. 2006, 45, 1435 – 1439.
[10] C. Palomo, J. M. Aizpurua, E. Balentovꢁ, I. Azecune, J. I. Santos,
J. Jimꢂnez-Barbero, J. C. Canada, J. I. Miranda, Org. Lett. 2008,
10, 2227 – 2300.
[11] a) G. Mukhopadhyay, P. Doherty, F. S. Walsh, P. R. Crocker,
M. T. Filbin, Neuron 1994, 13, 757 – 767; b) R. H. Quarles,
J. Neurochem. 2007, 100, 1431 – 1448.
[16] A. Bhunia, O. Schwardt, H. Gꢃthje, G.-P. Gao, S. Kelm, A. J.
Benie, M. Hricovini, T. Peters, B. Ernst, ChemBioChem 2008, 9,
2941 – 2945.
[17] A. P. May, R. C. Robinson, M. Vinson, P. R. Crocker, E. Y.
Jones, Mol. Cell 1998, 1, 719 – 728; Brookhaven protein data
bank acquisition code 1QFO.
[18] B. Ernst, B. Cutting, S. V. Shelke, PCT/WO2007/105094A1, 2006.
[19] G. Otting, Curr. Opin. Struct. Biol. 1993, 3, 760 – 768.
[20] a) W. Jahnke, L. B. Perez, C. G. Paris, A. Strauss, G. Fendrich,
C. M. Nalin, J. Am. Chem. Soc. 2000, 122, 7394 – 7395; b) W.
Jahnke, S. Rꢄdisser, M. Zurini, J. Am. Chem. Soc. 2001, 123,
3149 – 3150; c) J. Vazquez, L. Tautz, J. J. Ryan, K. Vuori, T.
Mustelin, M. Pellecchia, J. Med. Chem. 2007, 50, 2137 – 2143;
d) N. U. Jain, A. Venot, K. Umemoto, H. Leffler, J. H. Preste-
gard, Protein Sci. 2001, 10, 2393 – 2400; e) I. Bertini, M. Fragai,
Y.-M. Lee, C. Luchinat, B. Terni, Angew. Chem. 2004, 116, 2304 –
2306; Angew. Chem. Int. Ed. 2004, 43, 2254 – 2256.
[21] N. Bock, S. Kelm in Methods in Molecular Biology, Vol. 347 (Ed.:
I. Brockhausen), Humana Press, Totowa, NJ, 2006, pp. 359 – 376.
[22] M. Congreve, R. Carr, C. Murray, H. Jhoti, Drug Discovery
Today 2003, 8, 876 – 877.
[12] R. H. Quarles, Neurochem. Res. 2009, 34, 79 – 86.
[13] A. A. Vyas, O. Blixt, J. C. Paulson, R. L. Schnaar, J. Biol. Chem.
2005, 280, 16305 – 16310.
[14] L. J.-S. Yang, C. B. Zeller, N. L. Shaper, M. Kiso, A. Hasegawa,
R. E. Shapiro, R. L. Schnaar, Proc. Natl. Acad. Sci. USA 1996,
93, 814 – 818.
[23] P. J. Hajduk, E. T. Olejniczak, S. W. Fesik, J. Am. Chem. Soc.
1997, 119, 12257 – 12261.
[24] I. Bertini, C. Luchinat, G. Parigi, R. Pierattelli, ChemBioChem
2005, 6, 1536 – 1549.
[25] a) M. Mayer, B. Meyer, Angew. Chem. 1999, 111, 1902 – 1906;
Angew. Chem. Int. Ed. 1999, 38, 1784 – 1788; b) M. Mayer, B.
Meyer, J. Am. Chem. Soc. 2001, 123, 6108 – 6117.
[26] a) R. Huisgen, G. Szeimies, L. Moebius, Chem. Ber. 1967, 100,
2494 – 2507; b) R. Huisgen in 1,3-Dipolar Cycloaddition Chemis-
try, Vol. 1 (Ed.: A. Padwa), Wiley, New York, 1984, pp. 1 – 176.
[27] F. Mammen, E. I. Shakhnovich, G. M. Whitesides, J. Org. Chem.
1998, 63, 3168 – 3175.
[15] a) B. Ernst, J. L. Magnani, Nat. Rev. Drug Discovery 2009, 8,
661 – 677; b) S.-Y. Shin, H. Gꢃthje, O. Schwardt, G.-P. Gao, B.
Ernst, S. Kelm, B. Meyer, ChemBioChem 2008, 9, 2946 – 2949;
c) O. Schwardt, H. Gꢃthje, G.-P. Gao, S. Mesch, J. von Orelli, M.
Spreafico, A. Vedani, S. Kelm, B. Ernst, J. Med. Chem. 2009, 52,
989 – 1004; d) S. V. Shelke, G.-P. Gao, S. Mesch, H. Gꢃthje, S.
Kelm, O. Schwardt, B. Ernst, Bioorg. Med. Chem. 2007, 15,
4951 – 4965; e) G.-P. Gao, M. Smiesko, O. Schwardt, H. Gꢃthje,
S. Kelm, A. Vedani, B. Ernst, Bioorg. Med. Chem. 2007, 15,
7459 – 7469; f) D. Schwizer, H. Gꢃthje, S. Kelm, M. Porro, O.
Angew. Chem. Int. Ed. 2010, 49, 5721 –5725
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5725