A conformationally frozen peptoid boosts CXCR4 affinity and anti-HIV activity

Article abstract of DOI:10.1002/anie.201202090

There can be only one: Using a peptoid motif obtained by shifting the arginine side chain of a pentapeptide previously developed by Fujii et al. to the neighboring nitrogen atom restricts the conformational freedom and yields a conformationally homogeneou

Full text of DOI:10.1002/anie.201202090

Angewandte
Chemie
DOI: 10.1002/anie.201202090
CXCR4 Ligands
A Conformationally Frozen Peptoid Boosts CXCR4 Affinity and
Anti-HIV Activity**
Oliver Demmer, Andreas O. Frank, Franz Hagn, Margret Schottelius, Luciana Marinelli,
Sandro Cosconati, Ruth Brack-Werner, Stephan Kremb, Hans-Jꢀrgen Wester, and Horst Kessler*
The chemokine receptor subtype CXCR4 belongs to the G-
protein coupled receptors (GPCRs) and is, together with its
natural ligand CXCL12 (or SDF-1), a central part of the
signaling system in the human body. Its functions range from
stem-cell trafficking during embryogenesis, through cardio-
vascular, hematopoietic, and brain development, to signaling
in the nervous and immune system.[1–6] Furthermore, CXCR4
is one of two major coreceptors used by the human
immunodeficiency virus (HIV) for cell entry and CXCR4-
using viruses are critical for the pathogenesis of AIDS.[7,8]
Hence, CXCR4 represents a valuable therapeutic option for
multiple diseases, such as inflammation, cancer,[9] and HIV/
AIDS.[10] The recent approval of the CXCR4 antagonist
AMD3100 (Mozobil) as a drug for stem-cell mobilization
paves the way for development of further compounds that
target CXCR4 related diseases.[1,11] However at present, none
of the FDA-approved anti-HIV drugs target the CXCR4
receptor.
One of the most interesting classes of CXCR4 antagonists
is derived from the naturally occurring bicyclic peptide
polyphemusin II which was modified in a stepwise fashion
into monocyclic T140.[12] The monocyclic T140 and its
derivatives are inverse agonists and therefore additionally
offer the advantage of selectivity towards CXCR4 over other
CXCR4 antagonists which function as partial agonists, for
example, AMD3100.[13,14] In our efforts to develop high-
affinity CXCR4 ligands as suitable probes for molecular
imaging[15–17] we extended our studies through ligand-based
design using conformational considerations and structure–
activity relationships (SAR). Herein we describe CXCR4
antagonists with picomolar affinity, their binding mode, and
their capacity to inhibit the HIV infection of cells.
The pioneering work of Fujii et al. demonstrated that the
peptidic CXCR4 antagonist T140 can be downsized from
14 amino acids into a head-to-tail cyclized pentapeptide with
a binding affinity (IC50) of 8 nm (FC131; 1a).[18] This peptide
was further modified into the N-methylated analogue cyclo(-
d-Tyr1-d-[NMe]Arg2-Arg3-Nal4-Gly5-)
(Nal =l-3-(2-naph-
thyl)alanine) 1b resulting in even higher affinity (IC50
=
3 nm, Figure 1).[19] This is the highest CXCR4 binding affinity
known to date for compounds with the cyclopentapeptide
scaffold.
[*] Dr. O. Demmer, Dr. A. O. Frank, Dr. F. Hagn, Prof. Dr. H. Kessler
Institute for Advanced Study at the Department Chemie
Technische Universitꢀt Mꢁnchen
Lichtenbergstrasse 4, 85748 Garching (Germany)
and
Chemistry Department, Faculty of Science
King Abdulaziz University
P.O. Box 80203, Jeddah 21589 (Saudi Arabia)
E-mail: kessler@tum.de
Homepage: http://www.org.chemie.tu-muenchen.de
Dr. M. Schottelius, Prof. Dr. H.-J. Wester
Lehrstuhl fꢁr Pharmazeutische Radiochemie
Garching (Germany)
Prof. Dr. L. Marinelli
Dipartimento di Chimica Farmaceutica e Tossicologica
Universitꢂ di Napoli “Federico II”, Napoli (Italy)
Dr. S. Cosconati
Dipartimento di Scienze Ambientali
Seconda Universitꢂ di Napoli, Caserta (Italy)
Figure 1. Constitutionally similar cyclic pentapeptides exhibiting
different conformational behavior because of the introduction of an
N-methyl group and different chirality in position 2.
Prof. Dr. R. Brack-Werner, Dr. S. Kremb
Helmholtz Zentrum Mꢁnchen
Institute of Virology Neuherberg/Mꢁnchen (Germany)
Despite its high affinity, the NMR spectrum of this N-
methylated peptide (1b) exhibits two conformations in slow
equilibrium. The major one exhibits a conformation similar to
1a and was assumed by Ueda et al.[19] to be the bioactive
conformation. We determined the two conformations of 1c
(IC50 = 6 Æ 1 nm) in water by NMR spectroscopy to obtain
detailed insight into the different binding modes with addi-
tional help of molecular modeling. For these experiments we
used the ornithine2 (Orn2) derivative instead of Arg2 as it was
[**] We thank M. Wolff and B. Cordes for technical assistance and E.
Gourni for the help in the cell binding assay. This work was
supported by the Center of Integrated Protein Science Munich
(CIPS). Financial support by the DFG (SFB824, Subproject B5) is
acknowledged. The representation of the HI virus in the Table of
Contents graphic was originally created by the US National Institute
of Health (Permission: PD-USGov-HHS-NIH)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201202090.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
previously modified for our molecular imaging study.[15–17] It
can be assumed that this substitution has no influence on the
backbone conformation; subsequently, we constructed
ligands with even greater affinity by using a peptoid motif
that exhibits one single conformation.
The structure of the major 1c conformer, as investigated
in water, is virtually the same as the one elucidated earlier for
1b in DMSO (Figure 2).[19] The only major difference
fluorophosphate (HATU). The linear pentapeptides were
then cyclized in dilute solution using diphenylphosphoryl
azide (DPPA) and NaHCO3. Deprotection with trifluoro-
acetic acid (TFA) and subsequent HPLC separation yielded
the pure peptides (Table 1). Aminoalcohols were tert-butoxy-
carbonyl (Boc) protected prior to alkylation and 7-amino-
heptan-1-ole was synthesized according to procedures de-
scribed in literature.[25–27]
Table 1: Values for binding affinity to CXCR4 (IC50) and anti-HIV activity
(EC50) of synthesized compounds.
Comp
n
R1
R2
IC50 [nm][a]
EC50 [nm][b]
1b
2
3
4
5
6
–
4
5
6
7
6
6
6
N-Me-d-Arg
3[c]
66Æ5.7
n.d.
n.d.
Me
Me
Me
Me
H
H
H
H
H
0.9Æ0.1
1.6Æ0.3
0.3Æ0.2
2.4Æ0.8
100Æ9
53Æ3.7
n.d.
Figure 2. The major (trans amide bond between d-Tyr1 and d-Orn2,
orange) and minor (cis, blue) conformations of 1c exhibit a similar
overall conformation of the peptide backbone except for the orienta-
tion at the N-methyl amide between d-Tyr1 and d-Orn2 which is flipped
from trans to cis. Both structures are the most representative con-
formations taken from their unrestrained MD trajectories in explicit
water (see Supporting Information for more details). All aliphatic
protons were omitted for clarity.
H
n.d.
7
H
Me
Me
H
94Æ38
979Æ84
29Æ1.5
365Æ35.0
8[d]
0.04Æ0.02
AMD 3100
651Æ37[e]
[a] Measured against [125I]FC131. Mean value of at least three experi-
ments. [b] Mean values Æstandard deviations of 3–5 infection assays.
[c] Taken from Ref. [19]. [d] The amine group in the N-alkyl chain is
exchanged for a guanidine group. [e] Measured against [125I]SDF1-a.[35]
between these two structures is the preferred orientation of
the peptide bond between Nal4 and Gly5. The minor
conformation exhibits a cis peptide bond at the N-methyl-
ation site, which can be already concluded from the strong
ROE intensity (ROE = rotating frame nuclear Overhauser
effect) between the two Ha atoms of Tyr1 to Orn2 (see
Supporting Information). This result demonstrates that the
allylic strain induced by the introduction of the N-methyl
group is strong enough to cause a flip from a trans to a cis
peptide bond.[20]
Determination of the CXCR4 binding affinities of pep-
toids 2–8 (see Supporting Information) revealed that they all
have at least twofold higher affinity towards CXCR4 than the
parent peptide 1c (Figure 3 and Table 1) regardless of their N-
alkyl chain length, n in Table 1. d-Ala was exchanged with l-
Ala (7) and Gly (6) in the backbone of the most active ligand
4. Both derivatives exhibited two orders of magnitude lower
affinity, indicating that the methyl group and its chirality in
the d-Ala side chain is important for keeping the conforma-
tion close to the bioactive one.
As shown in Figure 2, the side chains of Tyr1 and Orn2 lie
next to each other in solution and are inclined towards each
other. A similar behavior had been observed in Somatostatin
peptides where hydrophobic clustering is the main driving
force for the preferred side-chain orientation.[21] These
preliminary conformational studies prompted the design of
new analogues in which the Orn (or Arg) side chain was
shifted from the a-carbon to the adjacent nitrogen atom
assuming that this change into a peptoid[22] can be assumed to
result in a structure, where the peptide bond (cis-trans) would
be frozen in its trans conformation. Such a conformational
impact of the peptoid motif in combination with enhanced
biological properties has been observed previously.[23]
Figure 3. Peptoids with varying N-alkyl length and different chirality.
See Table 1 for additional information.
For the synthesis of the peptoids, d-Orn was replaced by
d-Ala and aminoalkyl residues were introduced to its Na atom
by the Fukuyama–Mitsunobu reaction on solid support in
combination with standard Fmoc chemistry (Fmoc = 9H-
fluoren-9-ylmethoxy)carbonyl).[24] Specifically, an o-nitroben-
zensulfonyl (o-nosyl; Ns) protected amine is alkylated with an
alcohol under Mitsunobu conditions. After deprotection of
the Ns group the secondary amine is acylated using 2-(7-aza-
Compound 4 has a subnanomolar affinity, which is nearly
20-fold higher than 1c. On the NMR time scale 4 has a single
conformation that corresponds to the all-trans form of 1c; the
alkylated amide bonds of 4 and 1c have a similar orientation
(Figure 4). This result confirms the assumption that the all-
trans conformation of 1c is the bioactive one,[19] and the Tyr1
side chain and the N-alkyl group are arranged in a stacked
conformation. The stacking leads to an upfield shift of the
1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium
hexa-
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Figure 4. Compound 4 exhibits only one conformation in contrast to
its N-methylated parent peptide 1c. The Tyr1 side chain and the N-alkyl
residue form a hydrophobic interaction stabilizing the trans conforma-
tion. All aliphatic protons are omitted for clarity .
Figure 5. Binding conformation of peptoid 8 in the CXCR4 crystal
structure as calculated by using the software Glide. The receptor is
represented as light-green sticks and ribbons. The ligand is repre-
sented as orange sticks. Hydrogen bonds are represented as yellow
dashed lines. For clarity, only interacting residues are shown.
proton signals on the second carbon from the backbone amide
to d = 0.58 and 1.17 ppm compared to the signals of g protons
in an average arginine at d = 1.56 ppm. Furthermore, the
peptide bond Nal–Gly in 4 is flipped so that the carbonyl
oxygen atom points in the opposite direction to that of 1c
(Figure 4).
In a final optimization step the guanidine group was
reintroduced in the peptoid aminoalkyl chain. This resulted in
compound 8 having an additional almost 10-fold increase in
affinity compared to 4 making it the most affine CXCR4
ligand reported to date.
while Nal4 is embedded in the inner portion of the receptor
making a well-oriented cation–p interaction with R188.
Noteworthy, such a contact is further reinforced by the
establishment of an H-bond between the Nal4 carbonyl
oxygen atom and the R188 guanidine group of the receptor.
Although the peptoid structures can be considered close
structural analogues of compound 1a–c, the introduction of
a well-defined (trans) conformation of the Tyr1-N-alkylated
Ala2 amide bond allows the ligand to make contact with
a receptor region that was previously unexplored by the
parent compounds (see Figure 5 and 2b in Supporting
Information). This contact most probably accounts for the
enhanced affinity of 8.
The extremely high binding affinity motivated us to
explore the capacity of this compound to inhibit HIV
infection. For this purpose we used a full HIV-replication
assay validated for the discovery and the analysis of the
efficacy of HIV inhibitory molecules.[30] This assay is based on
HIV-permissive cells that express the HIV receptor CD4 and
the CXCR4 co-receptor and contain a reporter gene activated
by HIV infection to express a red fluorescent protein.
All tested compounds inhibited infection of the reporter
cells by HIV-1(LAI) (Table 1), a prototypical virus strain that
uses CXCR4 for entry.[31] Compounds 4 and 8, which have the
highest affinity to CXCR4 also showed strongest anti-HIV
activities, with EC50 values of (53 Æ 3.7) and (29 Æ 1.5) nm,
respectively. Compound 7, which has the lowest affinity to
CXCR4 also showed the lowest anti-HIV activity, which was
over 30-limes lower than the anti-HIVactivity of compound 8.
Compounds 4, 8, and 1b all exhibited higher anti-HIVactivity
than AMD3100. All compounds failed to inhibit infection of
CCR5-expressing reporter cells by HIV-1 (AD8), an HIV-
1 isolate that enters cells through the CCR5 coreceptor,[32]
when tested in HIV permissive cells that express the CCR5
To rationalize the ultra-high affinity displayed by the
synthesized peptoids, docking studies (Figure 5) were carried
out on the most active ligand 8 using the recently published X-
ray structure of CXCR4 bound to the CVX15 peptide
(protein data bank (PDB) code: 3OE0).[28] The best binding
mode obtained with the docking program Glide places 8 in
the receptor region occupied by CVX15 (See Figure 2a in
Supporting Information for a comparison) with its Tyr1
residue establishing a well oriented p–p interaction with
F189 and a H-bond with the Y190 backbone CO (Figure 2 in
Supporting Information). The adjacent N-alkylated Ala2 is
optimally oriented to place the long guanidinoalkyl chain
towards a negatively charged region of the receptor (see
Figure 3 in Supporting Information). This results in strong
ionic interactions of the terminal guanidine group with D262
and E277. Interestingly, the same interaction pattern is
established by Arg14 of the crystallized CVX15 peptide
(see Figure 2a in Supporting Information). Consistent with
the SAR data, the CXCR4 affinity of these peptoids depends
on the length of the N-alkyl chain and on the nature of the
terminal positively charged group. The almost 10-fold higher
affinity of 8 in comparison with 4 should be ascribed to the
ability of the guanidine group to establish a double salt bridge
with the E277 and D262 residues. For the D262 residue,
mutagenesis data outlined that its mutation abrogates the
antagonistic activity of bicyclam compounds.[29] The hydrogen
bonding and electrostatic interactions are also established by
Arg3 of 8 which makes polar contacts with D97 and D187
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[16] O. Demmer, I. Dijkgraaf, U. Schumacher, L. Marinelli, S.
Cosconati, E. Gourni, H.-J. Wester, H. Kessler, J. Med. Chem.
2011, 54, 7648 – 7662.
coreceptor (see Figure 1 on page 6 of the Supporting Infor-
mation). This result confirms specificity of the tested com-
pounds for HIV-1 CXCR4-using HIV-1 variants.
[17] E. Gourni, O. Demmer, M. Schottelius, C. D’Alessandria, S.
Schulz, I. Dijkgraaf, U. Schumacher, M. Schwaiger, H. Kessler,
H.-J. Wester, J. Nucl. Med. 2011, 52, 1803 – 1810.
[18] N. Fujii, S. Oishi, K. Hiramatsu, T. Araki, S. Ueda, H. Tamamura,
A. Otaka, S. Kusano, S. Terakubo, H. Nakashima, J. A. Broach,
J. O. Trent, Z. X. Wang, S. C. Peiper, Angew. Chem. 2003, 115,
3373 – 3375; Angew. Chem. Int. Ed. 2003, 42, 3251 – 3253.
[19] S. Ueda, S. Oishi, Z.-X. Wang, T. Araki, H. Tamamura, J.
Cluzeau, H. Ohno, S. Kusano, H. Nakashima, J. O. Trent, S. C.
Peiper, N. Fujii, J. Med. Chem. 2007, 50, 192 – 198.
[20] J. Chatterjee, D. Mierke, H. Kessler, J. Am. Chem. Soc. 2006, 128,
15164 – 15172.
In conclusion, we succeeded in restricting the conforma-
tion of a known CXCR4 ligand into a single active con-
formation by using a peptoid motif. In combination with
structure–activity relationship studies the binding affinity was
optimized to be 400 to 1500-fold higher than compounds
currently under clinical development, such as KRH-1636 or
AMD3100. AMD3100 is approved by the FDA for treatment
of non-Hodgkin lymphoma and multiple myeloma.[33,34]
Together with the advantages of modified cyclic peptides as
drugs (e.g. biocompatible metabolism, high stability against
enzymatic degradation, selectivity profile[35,36]) this molecule
has the potential to be a promising candidate for future
medical applications in CXCR4 related diseases.[29]
[21] B. H. Arison, R. Hirschmann, D. F. Veber, Bioorg. Chem. 1978,
7, 447 – 451.
[22] H. Kessler, Angew. Chem. 1993, 105, 572 – 573; Angew. Chem.
Int. Ed. Engl. 1993, 32, 543 – 544.
[23] S. C. Shankaramma, K. Moehle, S. James, J. W. Vrijbloed, D.
Obrecht, J. A. Robinson, Chem. Commun. 2003, 1842 – 1843.
[24] O. Demmer, I. Dijkgraaf, M. Schottelius, H.-J. Wester, H.
Kessler, Org. Lett. 2008, 10, 2015 – 2018.
Received: March 16, 2012
Revised: May 2, 2012
Published online: && &&, &&&&
[25] J. M. Chong, M. A. Heuft, P. Rabbat, J. Org. Chem. 2000, 65,
5837 – 5838.
[26] C. C. Forbes, K. M. DiVittorio, B. D. Smith, J. Am. Chem. Soc.
2006, 128, 9211 – 9218.
[27] A. Nilsen, P. M. England, J. Am. Chem. Soc. 2007, 129, 4902 –
4903.
[28] B. Wu, E. Y. Chien, C. D. Mol, G. Fenalti, W. Liu, V. Katritch, R.
Abagyan, A. Brooun, P. Wells, F. C. Bi, D. J. Hamel, P. Kuhn,
T. M. Handel, V. Cherezov, R. C. Stevens, Science 2010, 330,
1066 – 1071.
Keywords: biological activity · drug design ·
medicinal chemistry · peptides · peptidomimetics
.
[1] W.-T. Choi, S. Duggineni, Y. Xu, Z. Huang, J. An, J. Med. Chem.
2012, 55, 977 – 994.
[2] C. C. Bleul, M. Farzan, H. Choe, C. Parolin, I. Clark-Lewis, J.
Sodroski, T. A. Springer, Nature 1996, 382, 829 – 833.
[3] E. Oberlin, A. Amara, F. Bachelerie, C. Bessia, J.-L. Virelizier, F.
Arenzana-Seisdedos, O. Schwartz, J.-M. Heard, I. Clark-Lewis,
D. F. Legler, M. Loetscher, M. Baggiolini, B. Moser, Nature 1996,
382, 833 – 835.
[29] S. Hatse, K. Princen, L.-O. Gerlach, G. Bridger, G. Henson, E.
De Clercq, T. W. Schwartz, D. Schols, Mol. Pharmacol. 2001, 60,
164 – 173.
[30] S. Kremb, M. Helfer, W. Heller, D. Hoffmann, H. Wolff, A.
Kleinschmidt, S. Cepok, B. Hemmer, J. Durner, R. Brack-
Werner, Antimicrob. Agents Chemother. 2010, 54, 5257 – 5268.
[31] K. Peden, M. Emerman, L. Montagnier, Virology 1991, 185,
661 – 672.
[4] C. Murdoch, Immunol. Rev. 2000, 177, 175 – 184.
[5] R. S. Klein, J. B. Rubin, Trends Immunol. 2004, 25, 306 – 314.
[6] M. Burger, T. Hartmann, M. Krome, J. Rawluk, H. Tamamura,
N. Fujii, T. J. Kipps, J. A. Burger, Blood 2005, 106, 1824 – 1830.
[7] P. Lusso, EMBO J. 2006, 25, 447 – 456.
[32] E. O. Freed, G. Englund, M. A. Martin, J. Virol. 1995, 69, 3949 –
3954.
[33] K. Ichiyama, S. Yokoyama-Kumakura, Y. Tanaka, R. Tanaka, K.
Hirose, K. Bannai, T. Edamatsu, M. Yanaka, Y. Niitani, N.
Miyano-Kurosaki, H. Takaku, Y. Koyanagi, N. Yamamoto, Proc.
Natl. Acad. Sci. USA 2003, 100, 4185 – 4190.
[8] P. R. Gorry, P. Ancuta, Curr. HIV/AIDS Rep. 2011, 8, 45 – 53.
[9] J. A. Burger, T. J. Kipps, Blood 2006, 107, 1761 – 1767.
[10] H. Tamamura, N. Fujii, Expert Opin. Ther. Targets 2005, 9, 1267 –
1282.
[11] J.-P. Levesque, I. G. Winkler, Curr. Opin. Organ Transplant.
2008, 13, 53 – 58.
[34] S. P. Fricker, V. Anastassov, J. Cox, M. C. Darkes, O. Grujic, S. R.
Idzan, J. Labrecque, G. Lau, R. M. Mosi, K. L. Nelson, L. Qin, Z.
Santucci, R. S. Y. Wong, Biochem. Pharmacol. 2006, 72, 588 –
596.
[12] H. Tamamura, Y. Xu, T. Hattori, X. Zhang, R. Arakaki, K.
Kanbara, A. Omagari, A. Otaka, T. Ibuka, N. Yamamoto, H.
Nakashima, N. Fujii, Biochem. Biophys. Res. Commun. 1998,
253, 877 – 882.
[35] P. Vlieghe, V. Lisowski, J. Martinez, M. Khrestchatisky, Drug
Discovery Today 2010, 15, 40 – 56.
[36] J. Chatterjee, B. Laufer, H. Kessler, Nat. Protoc. 2012, 7, 432 –
444.
[13] M. M. Rosenkilde, L.-O. Gerlach, J. S. Jakobsen, R. T. Skerlj,
G. J. Bridger, T. W. Schwartz, J. Biol. Chem. 2004, 279, 3033 –
3041.
[14] E. De Clercq, Biochem. Pharmacol. 2009, 77, 1655 – 1664.
[15] O. Demmer, E. Gourni, U. Schumacher, H. Kessler, H.-J.
Wester, ChemMedChem 2011, 6, 1789 – 1791.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
CXCR4 Ligands
There can be only one: Using a peptoid
motif obtained by shifting the arginine
side chain of a pentapeptide previously
developed by Fujii et al. to the neighbor-
ing nitrogen atom restricts the confor-
mational freedom and yields a conforma-
tionally homogeneous peptide (see pic-
ture) with a 100-fold higher binding
affinity to the chemokine receptor CXCR4
in the picomolar range. Its efficiency to
inhibit HIV-1 infections is also demon-
strated.
O. Demmer, A. O. Frank, F. Hagn,
M. Schottelius, L. Marinelli, S. Cosconati,
R. Brack-Werner, S. Kremb, H.-J. Wester,
H. Kessler*
&&&& — &&&&
A Conformationally Frozen Peptoid
Boosts CXCR4 Affinity and Anti-HIV
Activity
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!