Rational conversion of noncontinuous active region in proteins into a small orally bioavailable macrocyclic drug-like molecule: The HIV-1 CD4:gp120 paradigm

Article abstract of DOI:10.1016/j.bmc.2010.04.053

Rational conversion of noncontinuous active regions of proteins into a small orally bioavailable molecule is crucial for the discovery of new drugs based on inhibition of protein-protein interactions. We developed a method that utilizes backbone cyclization as an intermediate step for conversion of the CD4 noncontinuous active region into small macrocyclic molecules. We demonstrate that this method is feasible by preparing small inhibitor for human immunodeficiency virus infection. The lead compound, CG-1, proved orally available in the rat model.

Full text of DOI:10.1016/j.bmc.2010.04.053

Bioorganic & Medicinal Chemistry 18 (2010) 5754–5761
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Rational conversion of noncontinuous active region in proteins into a
small orally bioavailable macrocyclic drug-like molecule: The HIV-1
CD4:gp120 paradigm
Mattan Hurevich a, , , Avi Swed b, , Salim Joubran a, Shira Cohen a, Noam S. Freeman a,
*
Elena Britan-Rosich d, Laurence Briant-Longuet c, Martine Bardy c, Christian Devaux c,
Moshe Kotler d, Amnon Hoffman b, Chaim Gilon a,
*
a Institute of Chemistry, The Hebrew University of Jerusalem, Edmond Safra Campus, Givat Ram Campus, The Hebrew University, Jerusalem 91904, Israel
b Departments of Pharmaceutics, Scholl of Pharmacy, The Hebrew University of Jerusalem, Hadassa Ein Karem Campus, The Hebrew University, Jerusalem 91120, Israel
c Institut de Biologie CNRS UMR5121, Laboratoire Infections Retrovirales et Signalisation Cellulaire, CNRS UMR5121, Institut de Biologie, Montpellier 34960, France
d Departments of Pathology, Medical Scholl, Hadassa Ein Karem Campus, The Hebrew University, Jerusalem 91120, Israel
a r t i c l e i n f o
a b s t r a c t
Article history:
Rational conversion of noncontinuous active regions of proteins into a small orally bioavailable molecule
is crucial for the discovery of new drugs based on inhibition of protein–protein interactions. We devel-
oped a method that utilizes backbone cyclization as an intermediate step for conversion of the CD4 non-
continuous active region into small macrocyclic molecules. We demonstrate that this method is feasible
by preparing small inhibitor for human immunodeficiency virus infection. The lead compound, CG-1,
proved orally available in the rat model.
Received 8 March 2010
Revised 16 April 2010
Accepted 17 April 2010
Available online 21 April 2010
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Backbone cyclization
Drug design
HIV-1
Macrocycles
Proteomics
1. Introduction
cophores are in spatial proximity yet distant in the protein se-
quence. In these cases only large peptides are considered able to
Protein–protein interactions (PPI) are key players in many
intercellular interactions and are involved in many diseases. They
are considered as potential targets for drug intervention. However,
conversion of active region of proteins into a drug-like molecule is
challenging. It is even more problematic when the active pharma-
mimic the entire active region. We previously reported a procedure
to convert a noncontinuous active region in proteins, comprising
few pharmacophors, into a backbone cyclic peptide that mimics
the active region of the target protein.1 Backbone cyclization (BC)
is a general method for conversion of peptides into drugable pep-
tidomimetic. In this method conformational constraint is imposed
on peptides.2 The advantage of BC is achieving cyclization mainly
by using backbone atoms, thus, avoiding the use of side chains
and peptidic ends that are essential for the biological activity.
Moreover, backbone cyclization has been shown to impose meta-
bolic stability and intestinal permeability.3,4 Screening of backbone
cyclic libraries with conformational diversity (cycloscan) was used
here in order to reduce the size of the active region of CD4 into a
small molecule.
Human immunodeficiency virus (HIV-1) entry into the host cell
is initially mediated by the interaction of viral envelope proteins
with T-cell transmembrane proteins.
CD4 engagement by gp120 is a critical step in virus penetration
and its inhibition by drug intervention is the goal of several re-
search groups and pharmaceutical companies.5 Small molecules
Abbreviations: AcOH, acetic acid; Alloc, allyloxycarbonyl; Boc, t-butoxycarbonyl;
BTC, bis(trichloromethyl)carbonate; DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene;
DCM, methylene chloride; DIC, N,N0-diisopropylcarbodiimide; DIPEA, diisopropyl-
ethylamine; DMAP, dimethylaminopyridine; DMF, N,N0-dimethylformamide; Fmoc,
9-fluorenylmethyloxycarbonyl; HATU, 2-(7-aza-1H-benzotriazole-1-yl)-1,1,3,3-
tetramethyluronium hexafluorophosphate; HBTU, (2-(1H-benzotriazole-1-yl)-
1,1,3,3-tetramethyluronium hexafluorophosphate; HIV-1, human immunodefi-
ciency virus; HOAt, 1-hydroxy-7-aza-benzotriazole; IV, intravenous; MAGI, multi-
nuclear activation of a galactosidase indicator; MBHA, 4-methylbenzhydrylamine;
NMM, 4-methylmorpholine; PO, per os; PPI, protein–protein interaction; PyClock,
6-chloro-benzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium
hexafluorophos-
phate; TDW, triple distilled water; TFA, trifluoroacetic acid; TIPS, triisopropylsilane.
* Corresponding authors. Tel.: +972 (0)2 6586 181; fax: +972 (0)2 6416 358
(C.G.).
E-mail address: gilon@vms.huji.ac.il (C. Gilon).
Both authors contributed equally to this work.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.04.053

M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
5755
like BMS-8066 and NBD-5567 recently showed promising prelimin-
O
ary results while peptide inhibitors like T-208 have been commer-
cially available. Although the structure of the CD4:gp120 complex
has been previously resolved9 only few drug candidates were
developed in order to target this interaction. This shortcoming
can be accounted for the large conformational changes experi-
enced in the viral envelope structure during CD4 binding. These
changes influence the Phe43 pocket and the structure of the site
in the unbound state is still unclear. Another difficulty in rational
design of CD4 mimetic stems from the fact that HIV-1 gp120 bind-
ing site in CD4 is noncontinuous. Insertion of mutation into CD4
molecule pointed out that Lys29, Lys35, Phe43, Lys46, and Arg59
residues participate in gp120 binding.10–13 The CD4 Phe43 is posi-
tioned in a type II0 b-turn on the first domain of the extra-cellular
part of the protein and is the most significant residue for gp120
binding. The phenylic side chain is involved in numerous interac-
tions with large hydrophobic pocket in gp120. Arg59, which is
Arg
NH
n=2,3,4,6
m=2-5
(CH2)m
(CH2)n
C-n-m
Gln Gly Ser Phe
N
CH2
NH2
O
O
Scheme 1. Structure of the C-n-m backbone cyclic library. Nomenclature, in the C-
n-m library n represents the number of methylenes in the N-alkylated building unit
and m represents the number of methylenes in the dicarboxylic acid linker.
cophores of the CD4 active site (Phe43 and Arg59). To obtain an ac-
tive backbone cyclic CD4 mimetic, four amino acids comprising the
turn around the Phe43 were preserved. In addition, the Arg moiety
was introduced as part of the bridge (Scheme 1). Optimization of
the distance between the Arg and Phe residues was achieved by
gradual addition of one or two methylenes to the n or m side of
the bridge. These systematic changes enabled us to partially scan
the conformational space of the scaffold, thus leading to the de-
sired bioactive conformation.
positioned on an
a-helix in close spatial proximity to the above
mentioned b-turn also plays a pivotal role in the interaction. Calcu-
lations based on the X-ray structure of the CD4:gp120 complex
indicate that the mean distance between the two pharmacophores
is less than 10 Å. The short distance between the pharmacophores
suggests that this site can be mimicked by small molecule and not
only by large peptides14 (Fig. 1).
The effect of backbone cyclic peptides on HIV-1 propagation
was studied by using MAGI cells, which express the b-galactosi-
dase gene under transcription activator region regulation.15 Mock
or HIV-1 infected cultured cells show that 100 lM of the cyclic
A rational method for the conversion of a noncontinuous active
site in a protein into a small macrocyclic proteinomimetic mole-
cule is described. We applied a backbone cyclic peptide library
screening method as an intermediate step for the rational design
of CD4 mimetic small anti HIV-1 drug lead based on two important
residues. We prepared a macrocyclic molecule that inhibits HIV-1
infection in cells. This molecule has substantial oral bioavailability
in animals and shows good distribution in tissues.
peptide C-2-2 (for structure and nomenclature see Scheme 1 foot-
note) reduced the viral infectivity by over 80% (Fig. 2). Further-
more, the four peptides, C-2-2, C-2-3, C-2-4, and C-2-5 (all
bearing two methylenes in the n side) were the most potent inhib-
itors in the library and reduced infectivity by 84%, 80%, 76%, and
68%, respectively in 100 lM concentrations. Contrary, no distinc-
tive correlation between the inhibitory effect and the m size was
observed.
Results from the C-n-m library indicate that the proximity of the
Phe moiety to the Arg residue on the n side directly affects the effi-
cacy of the CD4 derived cyclic peptides on HIV-1 infectivity. These
results suggest that mimicking of the CD4 active site is possible
using small backbone cyclic peptides and that the size of the inhib-
itor can be further reduced.
2. Results and discussion
2.1. Backbone cyclic HIV-1 inhibitors 1st generation
The in vitro model of CD4 gp120 interaction is unfortunately
not well established. Other groups that developed anti HIV-1 entry
inhibitors found it hard to prove the exact inhibition mechanism of
these drug leads. These works (e.g., Si et al.) suggest that gp120
inhibitors can occupy the hydrophobic pocket without disturbing
the interaction with CD4 and can lead to inhibition by inducing
conformational changes to gp120. The cellular approach is consid-
ered as an efficient alternative to the somewhat indecisive binding
assay. We decided to use the direct cell inhibition assay in order
to establish the concept of minimization through backbone
cyclization.
2.2. Design and synthesis of small molecule macrocyclic HIV-1
infection inhibitors
The straightforward conformational screening of the C-n-m li-
brary drew a pattern which allowed us to design small molecule
inhibitors. Our design focused on creating small macrocyclic
gp120 inhibitors by reducing the distance between the important
pharmacophores namely, Phe and Arg on the m side of the most
potent backbone cyclic inhibitor C-2-2.
We synthesized backbone cyclic peptide library (named C-n-m,
for nomenclature Scheme 1) that preserve the two crucial pharma-
In order to reduce the size of the potential HIV-1 inhibitors,
three amino acids from the Phe region (Gln40, Gly41, and Ser42)
were replaced with a diamide linker (pimelic acid). Furthermore,
since the proximity of the Arg and Phe pharmacophores is critical
for the anti HIV-1 activity, we decided to shorten the distance be-
tween these two crucial pharmacophores. The distance between
the Arg guanidine and the Phe aromatic moiety in C-2-2 is 12
atoms. By inserting the guanidine or the phenylic moiety as part
of the bridge methylene linker the distance between the two func-
tional groups could be shorten to only nine atoms (Scheme 2).
Three small molecules, CG-1, SC-1, and MC-1, with molecular
weights of about 500 g/mol were synthesized on solid support.
The three molecules share a similar scaffold comprising fourteen
atoms in the ring and the same chirality of the side chains.
As described in Scheme 2, all macrocyclic scaffolds were con-
structed using similar synthetic steps. The most challenging step
was the on-resin reductive amination. Three aldehydes were syn-
Figure 1. CD4:gp120 complex crystal structure. Blue-Arg59, red-Phe43 two CD4
pharmacophors crucial for the gp120 binding interaction.

5756
M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
120
100
80
60
40
20
0
peptide
Figure 2. Inhibition of viral infectivity by C-n-m peptides.
thesized in solution (see Supplementary data). Aldehydes Fmoc-
Arg(Alloc)2-H and Alloc- -Tyr(t-Bu)-H were produced by oxidation
of the corresponding alcohols using Dess–Martin periodinane as
L
-
clic scaffolds synthesized here possess an active pharmacophore
as part of the cyclization bridge.
L
We report an extension of works previously published that
introduce functional linkers by on-resin reductive amination pro-
cedure.16,22,23 Utilizing these methods we were able to synthesize
the macrocycles on solid support. Adopting these methods mini-
mize the amount of synthetic steps involved in precursors’
preparation.
Previous studies demonstrated that the role of CD4 Phe43 is
crucial for gp120 binding and its replacement with other amino
acids including tyrosine resulted in dramatic decrease in binding
ability.12 To test whether CG-1 binds gp120 in the same manner
as CD4, we replaced the phenyl moiety in CG-1 with a phenol.
The resulting molecule, MC-1, has exactly the same scaffold, struc-
ture and pharmacophores topology as CG-1 but possesses a
hydroxyl on the para position of the phenyl moiety replacing
hydrogen.
described previously.16,17 Alloc-
L-Phe-H was prepared by reduction
of the corresponding Weinreb amide as described.18,19
The alpha nitrogen of Rink-amide 4-methylbenzhydrylamine
(MBHA) resin bound glycine was treated with the above aldehydes
and reduced with NaCNBH3 to give the corresponding secondary
amine. A protected amino acid was coupled to the obtained sec-
ondary amine using either bis(trichloromethyl)carbonate (BTC) or
2-(7-aza-1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexa-
fluorophosphate (HATU) as coupling reagent. The use of milder re-
agents proved insufficient for complete coupling. After coupling,
the temporary protecting groups (Fmoc on MC-1, Boc on SC-1
and allyloxycarbonyl (Alloc) on CG-1) were removed. Selective
Boc deprotection in the synthesis of SC-1 was a challenge. Boc is
commonly used as permanent protection in Fmoc based solid
phase synthesis and its removal was considered as nonorthogonal
to the appropriate resin. Sivanandaiah et al. reported iodotrichlo-
rosilane as an efficient agent for Boc deprotection.20 However, this
method proved too cumbersome to be used routinely and was
abandoned after few attempts. We have used a recently published
procedure described by Freeman and Gilon for the selective re-
moval of Boc from peptide-bound to Rink-amide MBHA resin. This
procedure proved to be mild enough to allow selective Boc removal
without major reduction in overall yield.21 Pimelic acid was at-
tached to the primary amine using N,N0-diisopropylcarbodiimide
(DIC) for activation. The second temporary protecting group was
removed according to standard procedures (Alloc for MC-1, Fmoc
for CG-1 and SC-1). However, incomplete Fmoc removal from the
Arg moiety in CG-1 synthesis using standard conditions occurred.
We have used the more reactive base, 1,8-diazabicyclo[5.4.0]
undec-7-ene (DBU), to ensure complete removal of Fmoc before
the cyclization step. Cyclization was performed using the new
coupling reagent 6-chloro-benzotriazole-1-yl-oxy-tris-pyrrolidi-
no-phosphonium hexafluorophosphate (PyClock). For MC-1 and
CG-1 a simple cleavage procedure was used for consecutive re-
moval of the remaining semi-permanent protecting groups and
the macrocycles from the resin. For SC-1, Alloc removal from the
guanidine moiety was performed prior to cleavage. The macrocy-
2.3. Biological activity of small molecule macrocyclic HIV-1
infection inhibitors
The effect of MC-1, SC-1, and CG-1 on HIV-1 propagation was
studied. We showed that MC-1 did not reduce the infectivity of
HIV-1 in cultured cells at the lM range (Fig. 3A). This indicates that
the CD4:gp120 interaction was not interrupted by MC-1. This
agrees with previous results, suggesting that the addition of the
hydroxyl group to Phe43 on CD4 disrupts low energy interactions
with the gp120 hydrophobic pocket. The mechanism for
CD4:gp120 interaction inhibition by small molecules is still un-
clear. While it is possible that the inhibitors occupy the Phe43
pocket, X-ray studies suggest a Water-filled channel as an alterna-
tive binding site for these inhibitors.24 Furthermore, inhibitors like
T-20 disrupt the interaction by imposing allosteric changes to the
envelope protein and not by occupation of the interaction site. Pre-
liminary docking experiments suggest that CG-1 fit the Phe43
pocket. However, the conformational changes gp120 experiences
during binding put in doubt any design and development based so-
lely on the X-ray structure of the complex.25 Although our results
imply that CG-1 binds gp120 at the Phe43 pocket this point will
be further investigated by in vitro studies.

M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
5757
O
=Rink Amide MBHA-resin
Fmoc-Arg(Alloc)2-H
a,b
Method C
NH
Alloc-Phe-H
Method A
Fmoc
C
H2
a,b
Method B
Alloc
NH
Fmoc
Alloc-Tyr(tBu)-H
O
O
O
a,b
Alloc
NH
NH
NH
NH
C
CH2
C
CH2
(CH2)3
H
N
NH
H2
Alloc
Alloc NH
H2
N
CH2
C
H2
O
Boc-Phe-OH
t-Bu
Fmoc-Arg(pbf)-OH
c
Fmoc-Arg(pbf)-OH
c
c
Fmoc
NH
Boc
NH
Fmoc
NH
(CH2)3
NH
HN C
(CH2)3
NH
pbf
N
HN C
pbf
H
N
O
H
O
O
Alloc
NH
O
Alloc
NH
O
O
N
N
Fmoc NH
N
CH2
C
C
H2
CH2
CH2
C
Alloc N
H
N
H2
H2
(CH2)3
O
f,e
t-Bu
Alloc
NH
d,e
a,e
O
Fmoc
COOH NH
O
NH
(CH2)3 NH
(CH2)3 NH
H
NH
(CH2)5
H
N
(CH2)5
COOH
HN
C N pbf
HN
C
pbf
(CH2)5
COOH
O
O
O
O
O
O
NH
N
O
Alloc NH
N
N
Fmoc NH
CH2
i,g,h
C
CH2
C
H2
H2
CH2
C
Alloc
Alloc
N
H
N
O
t-Bu
H2
(CH2)3
NH
f,g,h
a,g,f,h
NH
O
O
O
NH
NH
C NH2
H
H
N
(CH2)3 N C NH2
(CH2)3
NH
NH
(CH2)5
(CH2)5
(CH2)5
NH
SC-1
MC-1
O
O
O
O
O
O
CG-1
O
O
O
N
NH
NH
N
N
CH CH2
C
CH CH2
C
C
CH CH2
NH2
NH2
H2
NH2
H2
NH
H2
H
H2N
N (CH2)3
HO
Scheme 2. Synthesis of the macrocyclic CD4 mimetics. Reagents and conditions: (a) piperidine; (b) Alloc-
L-Tyr(t-Bu)-H or Boc-L-Phe-H or Fmoc-L-Arg(Alloc)2-H, NaCNBH3; (c)
Boc- -Phe-OH or Fmoc-L-Arg(pbf)-OH, HATU, HOAt, DIPEA; (d) 0.05 M SnCl4; (e) pimelic acid, DIC, DMAP; (f) Pd(PPh3)4, AcOH, N-methylmorpholine; (g) PyClock, DIPEA; (h)
L
TFA, TDW, TIPS; (i) 10% DBU.
CG-1 and SC-1 are structural regioisomers having the same
scaffold that preserve the original CD4 active pharmacophores
but differ in the position of the functional groups (Scheme 2).
While SC-1 shows only weak inhibitory effect on HIV-1 infection
is 10–15 Å. CG-1 size falls well within this criteria and has drug-
like structure and characteristics.
We further evaluated its potential as drug candidate by study-
ing its oral bioavailability properties.
(inhibits virus infection in the high
then 80% of viral infection in the low
l
M range), CG-1 inhibits more
M range. The difference be-
l
2.4. Pharmacokinetic parameters following IV and PO
administration
tween the CG-1 and SC-1 isomers in HIV-1 infection inhibition
indicates that the potency is dictated not only by the nature of
the pharmacophores but also influenced by the correct orientation
of the active moieties. The effect of CG-1 on virus infection was
studied (Fig. 3B) by three independent assays that reaffirmed in a
The pharmacokinetic profile of CG-1 was studied by following
intravenous (IV) and per os (PO) administration to rats (presented
in Fig. 4). The pharmacokinetic parameters derived from adminis-
tration of CG-1 to rats are depicted in Table 1.
concentration dependent manner that CG-1 has low lM activity.
In 2004, Dahno et al. evaluated the reasonable size for drug can-
didates that aim to block transmembrane proteins interaction.14
The work was based on several commercially available drugs and
suggests that the optimal distance between the important residues
CG-1 half-life was about 73 min. This is a relatively long half-
life in comparison to native peptides. This finding reaffirm previous
reports that indicate that chemical modification (arise from cycli-
zation) resulted in a metabolically stable macrocyclic molecule.

5758
M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
3. Conclusion
Backbone cyclization and macrocyclization is presented here as
rational methods to mimic noncontinuous active regions in trans-
membrane proteins aimed to inhibit protein–protein interaction.
Screening a small set of backbone cyclic peptides enabled a detec-
tion of a structure activity relationship pattern that was used to de-
sign a small macrocyclic inhibitor. We applied this concept to the
HIV-1 CD4:gp120 model, and showed that the active site in CD4
can be minimized. Furthermore, the screening of backbone cyclic
peptide library proved useful in the design of macrocyclic HIV-1
inhibitors that can mimic the CD4 active site.
The unique protecting group manipulation and, in particular, the
novel orthogonal Boc deprotection procedure used here was essen-
tial for the synthesis of the scaffolds and can be adopted as strategy
for the preparation of other molecules on solid support. CG-1 inhib-
its viral infection in cells in the low micromolar range and has good
pharmacological properties, for example, oral bioavailability. We
claim that CG-1, having the combined properties of inhibitory ef-
fect, high stability and superior oral bioavailability, is an attractive
lead for further drug development. We also claim that by adapting
the approach presented here active regions in other protein binding
sites can be minimized to drug-like molecules.
4. Experimental section
4.1. Chemistry general
All starting materials were purchased from commercial sources
and were used without further purification. Nuclear magnetic res-
onance (NMR) spectra during synthesis were recorded on a Bruker
AMX 300, Bruker 400 or Bruker 500 MHz spectrometer. Chemical
shifts are reported downfield, relative to internal solvent peaks.
Coupling constants J are reported in Hz. High Resolution Mass spec-
trometry (HRMS) spectra were recorded on nanospray ionization
LTQ orbitrap. Matrix assisted laser desorption ionization (MALDI)-
time of flight (TOF) (MALDI-TOF) Mass spectra were recorded on
Figure 3. CG-1 blocks HIV-1 infection in MAGI HeLa cells. (A) Inhibition of HIV-1
infection was evaluated following treatment with 10
lM (grey) and 1
lM (dots) of
MC-1, SC-1, and CG-1. (B) CG-1 inhibits HIV-1 infection in
a concentration
dependent manner.
a PerSeptive Biosystems MALDI-TOF MS, using a-cyano-4-hydroxy-
cinnamic acid as matrix. Thin layer chromatography (TLC) was per-
formed on Merck aluminum sheets silica gel 60 F254. Column
chromatography was performed on Merck silica gel 60 (230–400
mesh). Peptides purification: Peptides purity was determined by
analytical HPLC, peptides below 95% purity were excluded from fur-
ther examination (see Supplementary data). Analytical HPLC was
performed on Vydac analytical columns (C18, 5
lm, 4.6 mm ꢀ
250 mm (218TP54)) using Merck–Hitachi system: Model LaChrom
with a L-7100 pump, L-7200 autosampler, L-7400 UV–Vis detector
and a D-7000 interface. Products were assayed at 215 and 220 nm.
The mobile phase consisted of a gradient system, with solvent A
corresponding to TDW with 0.1% TFA and solvent B corresponding
to acetonitrile (ACN) with 0.1% TFA. The mobile phase started with
95% A from 0 to 5 min followed by a linear gradient from 5% B to
95% B from 5 to 55 min. The gradient remained at 95% B for an addi-
tional 5 min and then was reduced to 95% A and 5% B from 60 to
65 min. The gradient remained at 95% A for additional 5 min to
achieve column equilibration. The flow rate of the mobile phase
was 1 mL/min. Peptide purification was performed by reversed
Figure 4. Plasma concentrations of CG-1 were plotted against time scale after IV
bolus and PO administration to conscious Wistar rats, n = 5 in each group, values are
average SEM.
The calculated oral bioavailability (10%) further strengthened our
preliminary findings of good ex-vivo permeability. The relatively
high volume of distribution (Vss: 0.6 L/kg) provides an indirect
indication of the restricted ability to cross biological membranes.
Table 1
Pharmacokinetic properties of CG-1 after administration to Wistar rats
PK parameters
Cmax (ng/mL)
Tmax (min)
AUC (min ꢀ ng/mL)
21770 63
Cl (mL/min/kg)
14.53 0.33
Vss (L/kg)
0.6
T1/2 (min)a
Oral bioavailabilitya
PO
IV
866 76
102 19
18
5
73
ꢁ10%
a
Noncompartmental pharmacokinetic analysis was performed using WinNonlin software, standard.

M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
5759
phase semi-preparative HPLC on a Merck–Hitachi 665A model
equipped with a preparative pump (30 mL/min) and a high flow
UV–Vis detector using semi-preparative Vydac column (C18,
Fmoc-Gly-OH was coupled using HBTU activation followed by
washing with NMP (3 ꢀ 5 min). The Fmoc group was removed,
and the resin was washed with NMP (3 ꢀ 5 min). Fmoc-
L-Arg(Al-
5
l
m, 10 ꢀ 250 (208TP510)) flow rate of the mobile phase was
loc)2-CHO (4 equiv), dissolved in NMP/MeOH solution containing
1% AcOH, was added to the resin followed by the addition of
NaBH3CN (4 equiv) and left to stir for 4 h. Resin was washed with
NMP/MeOH (1 ꢀ 5 min), MeOH (1 ꢀ 5 min), 1% AcOH/water
(1 ꢀ 5 min), 10% water in MeOH (1 ꢀ 5 min), MeOH (1 ꢀ 5 min),
4.5 mL/min. All semi-preparative HPLC runs were carried out using
a gradient system similar to the one used in for the analytical HPLC.
Macrocycles purification: Analytical RP-HPLC were recorded at
220 nm at a flow of 1 mL/min on Merck–Hitachi system (LaChrom
with a L-7100 pump, L-7200 autosampler, L-7400 UV–Vis detector
NMP (1 ꢀ 5 min), DCM (1 ꢀ 5 min), and NMP (3 ꢀ 5 min). Boc-
L-
and a D-7000 interface) on Phenomenex RP-18 column (C18, 5
l,
Phenylalanine-OH was coupled using HATU activation overnight
followed by washing with NMP (3 ꢀ 5 min). Boc was removed
using the recently published procedure.21 Resin was treated with
0.05 M SiCl4 in DCM (dry) (2 ꢀ 15 min) followed by washing with
DCM (1 ꢀ 5 min), DMF (1 ꢀ 5 min), 20% MeOH/DMF (1 ꢀ 5 min),
1% DIPEA/DMF (2 ꢀ 5 min), and NMP (3 ꢀ 5 min). Pimelic acid
(10 equiv) was pre-activated with DIC (10 equiv) in NMP and
was added to the resin followed by addition of 4-(Dimethyl-
amino)pyridine (DMAP) (1 equiv) left for 3 h then washed with
NMP (3 ꢀ 5 min). The Fmoc group was removed, resin was washed
with NMP (3 ꢀ 5 min) then treated with a solution of PyClock
(6 equiv) and DIPEA (14 equiv) in NMP for 4 h (ꢀ2). Resin was
washed with NMP (2 ꢀ 5 min), DCM (2 ꢀ 5 min), and MeOH
(2 ꢀ 5 min) then dried under vacuum. Resin was treated with
Pd(PPh3)4(0) (1:1 weight equiv) in NMM/AcOH/DCM(dry)(2.5/2.5/
95%) solution for 2 h in dark then washed with 0.5% DIPEA in
NMP (3 ꢀ 5 min), 0.5% sodium diethyldithiocarbamate trihydrate
in NMP (5 ꢀ 2 min), NMP (2 ꢀ 2 min), DCM (2 ꢀ 2 min), MeOH
(2 ꢀ 2 min), and dried in vacuum. Crude was cleaved from the re-
sin by treatment with TFA/triisopropylsilane (TIPS)/TDW (92.5/5/
2.5%) solution for 2.5 h. The solution was separated by filtration
and the resin was rinsed with neat TFA. The TFA was evaporated
to give crude oil that was dissolved in ACN:TDW 1:1 solution
and lyophilized. Crude was purified on semi-preparative HPLC as
described above, collected peaks were analyzed using analytical
HPLC and pure compounds (over 95% purity were used for biolog-
ical screening).
4.6 ꢀ 75 mm (Luna)). Using the same solvent system previously de-
scribed, the mobile phase started with 95% A from 0 to 5 min fol-
lowed by a linear gradient from 5% B to 95% B from 5 to 17 min.
The gradient remained at 95% B for an additional 4 min and then
was reduced to 95% A from 21 to 25 min. The gradient remained
at 95% A for additional 5 min to achieve column equilibration.
Semi-preparative HPLC were recorded at 220 nm on Phenomenex
RP-18 column (C18, 10
l
250 ꢀ 10 mm, 110 Å (Gemini)). Using
the same solvent system previously described, the mobile phase
started with 95% A from 0 to 5 min followed by a linear gradient
from 5% B to 35% B from 5 to 30 min, then to 95% B in 15 min, the
gradient remained at 95% B for an additional 5 min and then was re-
duced to 95% A in 10 min. The gradient remained at 95% A for addi-
tional 5 min to achieve column equilibration.
4.2. Peptide synthesis
Backbone cyclic peptides were named systematically C-n-m
where n describe the number of methylenes in the building unit
and
m represents the number of methylenes in the linker
(Scheme 1). The Peptides were synthesized on methylbenzhydryl-
amine (MBHA) resin using the combinatorial ‘tea bag’ method10
using Fmoc and Boc chemistries as described before.1,11 Final
cleavage was performed using HF/anisole mixture or by trimethyl-
silyl trifluoromethane sulfonate/anisole/trifluoroacetic acid mix-
ture. Out of the designed peptide library thirteen were prepared
in sufficient yield and purity to allow biological screening (Table 1).
SC-1: Prepared from 200 mg of Fmoc-Rink MBHA resin. Yield:
0.8 mg. HPLC purity >95%. Rt 9.73. HRMS (Orbitrap-ESI): exact
mass calcd for C24H38N7O4 488.2980 (MH+). Found 488.2968.
4.3. General methods for solid phase synthesis of
macromolecules C
4.3.2. Solid-phase synthesis MC-1: method B
Swelling: The resin was swelled for at least 2 h in DCM. Fmoc
removal: The resin was treated with a solution of 20% piperidine
in NMP (2 ꢀ 20 min), and then washed with NMP (5 ꢀ 2 min).
HBTU coupling: Protected amino acids (1.5 equiv) were dissolved
in NMP. N,N-Diisopropylethylamine (DIPEA) (1.5 equiv) and 1-
hydroxybenzotriazole (HOBt) (1.5 equiv) were added and the mix-
ture was cooled to 0 °C. (2-(1H-Benzotriazole-1-yl)-1,1,3,3-tetram-
ethyluronium hexafluorophosphate) (HBTU) (1.5 equiv) was added
and the mixture was pre-activated by mixing for 10 min, added to
the resin, and shaken for 1 h. The resin was washed with NMP
(3 ꢀ 2 min). Capping: The resin was treated with a solution of
AC2O (10 equiv) and DIPEA (7.15 equiv) in DMF for 20 min and
washed with NMP (3 ꢀ 2 min). HATU coupling: Fmoc protected
amino acids (1.5 equiv) were dissolved in NMP, DIPEA (1.5 equiv)
and 1-hydroxy-7-aza-benzotriazole (HOAt) (1.5 equiv) were added
and the mixture was cooled to 0 °C. (2-(7-Aza-1H-benzotriazole-1-
yl)-1,1,3,3-tetramethyluronium hexafluorophosphate) (HATU)
(1.5 equiv) was added and the mixture was pre-activated by mix-
ing for 10 min, added to the resin, and shaken overnight at rt.
The resin was washed with NMP (3 ꢀ 2 min).
Fmoc-Rink-amide MBHA resin was washed with NMP and left
2 h for swelling. Fmoc group was removed and resin was washed
with NMP (3 ꢀ 5 min). Fmoc-Gly-OH was coupled using HBTU acti-
vation followed by washing with NMP (3 ꢀ 5 min). The Fmoc group
was removed and resin was washed with NMP (3 ꢀ 5 min). Alloc-
L-
Tyr(t-Bu)-CHO (4 equiv) in 1% AcOH in NMP/MeOH was added to
the resin followed by the addition of NaBH3CN (4 equiv) and left
to stir for 4 h. Resin was washed with NMP/MeOH (1 ꢀ 5 min),
MeOH (1 ꢀ 5 min), 1% AcOH/water (1 ꢀ 5 min),10% water in MeOH
(1 ꢀ 5 min), MeOH (1 ꢀ 5 min), NMP (1 ꢀ 5 min), DCM (1 ꢀ 5 min),
and NMP (3 ꢀ 5 min). Fmoc-L-Arg(2,2,4,6,7-pentamethyldihydro-
benzofuran-5-sulfonyl (pbf))-OH was coupled using HATU activa-
tion overnight followed by washing with NMP (3 ꢀ 5 min). Fmoc
group was removed and resin was washed with NMP (3 ꢀ 5 min).
Pimelic acid (10 equiv) was pre-activated with DIC (10 equiv) in
NMP and was added to the resin followed by addition of DMAP
(1 equiv) left for 3 h. Resin was washed with NMP (2 ꢀ 5 min),
DCM (2 ꢀ 5 min), and MeOH (2 ꢀ 5 min) then dried under vacuum.
Alloc was removed by treatment with Pd(PPh3)4(0) (0.5 equiv) in
NMM/AcOH/DCM(dry)(2.5/2.5/95%) solution for 2 h in dark, then
washed with 0.5% DIPEA in NMP (3 ꢀ 5 min), 0.5% sodium diethyl-
dithiocarbamate trihydrate in NMP (5 ꢀ 2 min), NMP (2 ꢀ 2 min),
DCM (2 ꢀ 2 min), MeOH (2 ꢀ 2 min), and dried in vacuum. The
Fmoc group was removed, resin was washed with NMP
(3 ꢀ 5 min) then treated with a solution of PyClock (6 equiv) and
4.3.1. Solid-phase synthesis SC-1: method A
Fmoc-Rink-amide MBHA resin was washed with NMP and left
2 h for swelling. Fmoc group was removed using 20% piperidine
solution and then resin was washed with NMP (3 ꢀ 5 min).

5760
M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
DIPEA (14 equiv) in NMP for 4 h (ꢀ2). Resin was washed with NMP
(2 ꢀ 5 min), DCM (2 ꢀ 5 min), and MeOH (2 ꢀ 5 min) then dried
under vacuum. Crude was cleaved from the resin by treatment with
TFA/TIPS/TDW (92.5/5/2.5%) solution for 2.5 h. The solution was
separated by filtration and the resin was rinsed with neat TFA.
The TFA was evaporated to give crude oil that was dissolved in
ACN:TDW 1:1 solution and lyophilized. Crude was purified on
semi-preparative HPLC as described above, collected peaks were
analyzed using analytical HPLC and pure compounds (over 95% pur-
ity were used for biological screening).
rocyanide, 4 mM potassium ferricyanide, 2 mM MgCl2 and 0.4 mg/
mL of X-Gal (Ornat, Israel). Blue cells were counted under a light
microscope.
4.3.5. Animals
Male Wistar rats (275 20 g) were purchased from Harlan lab-
oratories (Rehobot, Israel). Rats were kept in a light-controlled
room (light from 7:00 to 19:00) and were maintained on labora-
tory chow and water ad libitum.
All surgical and experimental procedures were reviewed and
approved by the Animal Experimentation Ethics Committee of
the Hebrew University Hadassah Medical Center, Jerusalem.
MC-1: Prepared from 200 mg of Fmoc-Rink MBHA resin. Yield:
1.4 mg. HPLC purity >95%. Rt 8.79. HRMS (Orbitrap-ESI): exact
mass calcd for C24H38N7O5 504.2929 (MH+). Found 504.2919.
4.3.6. Pharmacokinetic study
4.3.3. Solid-phase synthesis CG-1: method C
Studies were performed in conscious Wistar male rats. An
indwelling cannula was implanted into the left jugular vein 24 h
before the pharmacokinetic experiment to allow full recovery of
the animals from the surgical procedure. Animals (n = 5) received
an intravenous (IV) bolus dose of 1 mg/kg of CG-1 or 10 mg/kg
by oral gavage (n = 5), CG-1 was dissolved in water. Blood samples
(with heparin, 15 U/mL) were collected at several time points up to
24 h after CG-1 administration. Plasma was separated by centrifu-
gation (4000 g, 5 min, 4 °C) and stored at ꢂ70 °C pending analysis.
Noncompartmental pharmacokinetic analysis was performed
using WinNonlin software, standard
Fmoc-Rink-amide MBHA resin was washed with NMP and left
2 h for swelling. The Fmoc group was removed and resin was
washed with NMP (3 ꢀ 5 min). Fmoc-Gly-OH was coupled using
HBTU activation followed by washing with NMP (3 ꢀ 5 min). The
Fmoc group was removed and resin was washed with NMP
(3 ꢀ 5 min). Alloc-
L-Phe-CHO (4 equiv) in 1% AcOH in NMP/MeOH
was added to the resin followed by the addition of NaBH3CN
(4 equiv) and left to stir for 4 h. Resin was washed with NMP/MeOH
(1 ꢀ 5 min), MeOH (1 ꢀ 5 min), 1% AcOH/water (1 ꢀ 5 min), 10%
water in MeOH (1 ꢀ 5 min), MeOH (1 ꢀ 5 min), NMP (1 ꢀ 5 min),
DCM (1 ꢀ 5 min), and NMP (3 ꢀ 5 min). Fmoc-L-Arg(pbf)-OH was
coupled using HATU activation overnight followed by washing with
NMP (3 ꢀ 5 min), DCM (2 ꢀ 5 min), and MeOH (2 ꢀ 5 min) then
dried under vacuum. Alloc was removed by treatment with
Pd(PPh3)4(0) (0.5 equiv) in NMM/AcOH/DCM(dry)(2.5/2.5/95%)
solution for 2 h in dark, then washed with 0.5% DIPEA in NMP
(3 ꢀ 5 min), 0.5% sodium diethyldithiocarbamate trihydrate in
NMP (5 ꢀ 2 min), NMP (2 ꢀ 2 min), DCM (2 ꢀ 2 min), MeOH
(2 ꢀ 2 min), and NMP (2 ꢀ 3 min). Pimelic acid (10 equiv) was
pre-activated with DIC (10 equiv) in NMP and was added to the re-
sin followed by addition of DMAP (1 equiv) left for 3 h, then washed
with NMP (3 ꢀ 5 min) and the Fmoc group was removed using a
solution of 10% 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) in NMP
(2 ꢀ 1/2 h). The resin was washed with NMP (3 ꢀ 5 min) then trea-
ted with a solution of PyClock (6 equiv) and DIPEA (14 equiv) in
NMP for (2 ꢀ 4 h). The resin was washed with NMP (2 ꢀ 5 min),
DCM (2 ꢀ 5 min), and MeOH (2 ꢀ 5 min) then dried under vacuum.
Crude was cleaved from the resin by treatment with TFA/TIPS/TDW
(92.5/5/2.5%) solution for 2.5 h. The resin was filtered and rinsed
with neat TFA. The TFA solution was evaporated to give a crude
oil that was dissolved in acetonitrile:TDW 1:1 solution and lyophi-
lized. The crude product was purified on semi-preparative HPLC as
described above, collected peaks were analyzed using analytical
HPLC and pure compounds (over 95% purity) were used for biolog-
ical screening.
Acknowledgment
The authors like to thank Luxembourg Industries Ltd. for sup-
plying PyClock that was used throughout this work.
Supplementary data
Supplementary data (characterization of C-n-m library, syn-
thetic procedure of aldehyde preparation, and NMR results) associ-
ated with this article can be found, in the online version, at
doi:10.1016/j.bmc.2010.04.053.
References and notes
1. Kasher, R.; Oren, D. A.; Barda, Y.; Gilon, C. J. Mol. Biol. 1999, 292, 421.
2. Gilon, C.; Halle, D.; Chorev, M.; Selinger, Z.; Byk, G. Biopolymers 1991, 31, 745.
3. Hess, S.; Linde, Y.; Ovadia, O.; Safrai, E.; Shalev, D. E.; Swed, A.; Halbfinger, E.;
Lapidot, T.; Winkler, I.; Gabinet, Y.; Faier, A.; Yarden, D.; Xiang, Z.; Portitlo, F. P.;
Haskell-Luevano, C.; Gilon, C.; Hoffman, A. J. Med. Chem. 2008, 51, 1026.
4. Hess, S.; Ovadia, O.; Shalev, D. E.; Senderovich, H.; Qadri, B.; Yehezkel, T.;
Salitra, Y.; Sheynis, T.; Jelinek, R.; Gilon, C.; Hoffman, A. J. Med. Chem. 2007, 50,
6201.
5. Este, J. A.; Telenti, A. Lancet 2007, 370, 81.
6. Si, Z.; Madani, N.; Cox, J. M.; Chruma, J. J.; Klein, J. C.; Schon, A.; Phan, N.; Wang,
L.; Biorn, A. C.; Cocklin, S.; Chaiken, I.; Freire, E.; Smith, A. B., 3rd; Sodroski, J. G.
Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5036.
7. Zhao, Q.; Ma, L.; Jiang, S.; Lu, H.; Liu, S.; He, Y.; Strick, N.; Neamati, N.; Debnath,
A. K. Virology 2005, 339, 213.
8. Kilby, J. M.; Hopkins, S.; Venetta, T. M.; DiMassimo, B.; Cloud, G. A.; Lee, J. Y.;
Alldredge, L.; Hunter, E.; Lambert, D.; Bolognesi, D.; Matthews, T.; Johnson, M.
R.; Nowak, M. A.; Shaw, G. M.; Saag, M. S. Nat. Med. 1998, 4, 1302.
9. Kwong, P. D.; Wyatt, R.; Robinson, J.; Sweet, R. W.; Sodroski, J.; Hendrickson, W.
A. Nature 1998, 393, 648.
10. Brodsky, M. H.; Warton, M.; Myers, R. M.; Littman, D. R. J. Immunol. 1990, 144,
3078.
11. Kwong, P. D.; Wyatt, R.; Majeed, S.; Robinson, J.; Sweet, R. W.; Sodroski, J.;
Hendrickson, W. A. Structure 2000, 8, 1329.
12. Moebius, U.; Clayton, L. K.; Abraham, S.; Harrison, S. C.; Reinherz, E. L. J. Exp.
Med. 1992, 176, 507.
13. Myszka, D. G.; Sweet, R. W.; Hensley, P.; Brigham-Burke, M.; Kwong, P. D.;
Hendrickson, W. A.; Wyatt, R.; Sodroski, J.; Doyle, M. L. Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 9026.
CG-1: Prepared from 200 mg of Fmoc-Rink MBHA resin. Yield:
1.2 mg (2.1%). Analytical HPLC purity >95%. Rt 9.71. HRMS (Orbi-
trap-ESI): exact mass calcd for C24H38N7O4 488.2980 (MH+).
Found 488.2976.
4.3.4. HIV-1 titration
Titration of HIV-1 strand HXB2 in the absence or presence of the
inhibitor was carried out by the multinuclear activation of a galac-
tosidase indicator (MAGI) assay, as described by Kimpton and
Emerman.15 Briefly, HeLaꢂCD4+beta-gal cells were transferred
into 96-well plates at 15 ꢀ 103 cells per well. On the following
day, the cells were infected with 50 lL of serially diluted virus in
14. Dahno, W.; Swistok, J.; Khan, W.; Fry, D.; Sun, H.; Falcioni, F.; Goodnow, R., Jr.;
Truitt, T.; Hoffman, A.; Garippa, R. In Third International and Twenty-Eight
European Peptide Symposium; Martin, F., Mati, F., Chaim, G., Jirina, S., Eds.;
Kenes International: Prague, Czech Republic, 2004.
the presence of 20 mg/mL of DEAE-dextran (Pharmacia, Sweden).
Two days post-infection, cultured cells were fixed with 1% formal-
dehyde and 0.2% glutaraldehyde in PBS. Following intensive wash
with PBS, cells were stained with a solution of 4 mM potassium fer-
15. Kimpton, J.; Emerman, M. J. Virol. 1992, 66, 2232.

M. Hurevich et al. / Bioorg. Med. Chem. 18 (2010) 5754–5761
5761
16. Bondebjerg, J.; Xiang, Z. M.; Bauzo, R. M.; Haskell-Luevano, C.; Meldal, M. J. Am.
21. Freeman, N. S.; Gilon, C. Synlett 2009, 2097.
Chem. Soc. 2002, 124, 11046.
22. Hurevich, M.; Barda, Y.; Gilon, C. Heterocycles 2007, 73, 617.
17. Myers, A. G.; Zhong, B. Y.; Movassaghi, M.; Kung, D. W.; Lanman, B. A.; Kwon, S.
Tetrahedron Lett. 2000, 41, 1359.
18. Ede, N. J.; Eagle, S. N.; Wickham, G.; Bray, A. M.; Warne, B.; Shoemaker, K.;
Rosenberg, S. J. Pept. Sci. 2000, 6, 11.
19. Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
20. Sivanandaiah, K. M.; Sureshbabu, V. V.; Shankaramma, S. C. Int. J. Pept. Protein
Res. 1995, 45, 377.
23. Qvit, N.; Reuveni, H.; Gazal, S.; Zundelevich, A.; Blum, G.; Niv, M. Y.; Feldstein,
A.; Meushar, S.; Shalev, D. E.; Friedler, A.; Gilon, C. J. Comb. Chem. 2008, 10, 256.
24. Madani, N.; Perdigoto, A. L.; Srinivasan, K.; Cox, J. M.; Chruma, J. J.; LaLonde, J.;
Head, M.; Smith, A. B., 3rd; Sodroski, J. G. J. Virol. 2004, 78, 3742.
25. Caporuscio, F.; Tafi, A.; Gonzalez, E.; Manetti, F.; Este, J. A.; Botta, M. Bioorg.
Med. Chem. Lett. 2009, 19, 6087.