An amphiphilic conjugate approach toward the design and synthesis of betulinic acid-polyphenol conjugates as inhibitors of the HIV-1 gp41 fusion core formation

Article abstract of DOI:10.1002/cmdc.201100149

Exploration of potent inhibitors of the HIV-1 gp41 fusion core formation is a promising strategy to discover small-molecule HIV-1 entry inhibitors for the treatment of HIV-1 infection. In this paper, a series of novel betulinic acid-polyphenol conjugates was designed, guided by molecular modeling of the binding of betulinic acid (BA) and phenolic galloyl/caffeoyl groups in the groove on the gp41 N-terminal heptad repeat (NHR) trimeric coiled coil. These conjugates were synthesized via conjugation of galloyl and caffeoyl groups with BA at the C-28 position. Their inhibitory activities of HIV gp41 six-helix bundle (6-HB) formation between the NHR peptide N36 and the C-terminal heptad repeat (CHR) peptide C34 were evaluated with size-exclusion HPLC. Conjugates bearing a galloyl group were found to exhibit four to sixfold higher inhibitory activities than that of parent compound BA, suggesting that they may be exploitable as HIV-1 fusion/entry inhibitors targeting gp41. The docking study on BA and its derivatives suggests that hydrophobic and hydrogen-bonding pockets exist in the groove of the gp41 NHR trimeric coiled coil and that a potent inhibitor should have amphiphilic structures to cooperatively interact with both pockets. This possibility was explored by incorporating both lipophilic and hydrophilic groups into the conjugates in a well-defined orientation to bind with both pockets in the gp41 NHR-trimer.

Full text of DOI:10.1002/cmdc.201100149

MED
DOI: 10.1002/cmdc.201100149
An Amphiphilic Conjugate Approach toward the Design
and Synthesis of Betulinic Acid–Polyphenol Conjugates as
Inhibitors of the HIV-1 gp41 Fusion Core Formation
Yan Liu,[a] Zhuofeng Ke,[b] Kwok Yiu Wu,[a] Shuwen Liu,[c] Wen-Hua Chen,[c] Shibo Jiang,[d] and
Zhi-Hong Jiang*[a, c]
Dedicated to Professor Isao Kouno on the occasion of his 65th birthday
Exploration of potent inhibitors of the HIV-1 gp41 fusion core
formation is a promising strategy to discover small-molecule
HIV-1 entry inhibitors for the treatment of HIV-1 infection. In
this paper, a series of novel betulinic acid–polyphenol conju-
gates was designed, guided by molecular modeling of the
binding of betulinic acid (BA) and phenolic galloyl/caffeoyl
groups in the groove on the gp41 N-terminal heptad repeat
(NHR) trimeric coiled coil. These conjugates were synthesized
via conjugation of galloyl and caffeoyl groups with BA at the
C-28 position. Their inhibitory activities of HIV gp41 six-helix
bundle (6-HB) formation between the NHR peptide N36 and
the C-terminal heptad repeat (CHR) peptide C34 were evaluat-
ed with size-exclusion HPLC. Conjugates bearing a galloyl
group were found to exhibit four to sixfold higher inhibitory
activities than that of parent compound BA, suggesting that
they may be exploitable as HIV-1 fusion/entry inhibitors target-
ing gp41. The docking study on BA and its derivatives sug-
gests that hydrophobic and hydrogen-bonding pockets exist
in the groove of the gp41 NHR trimeric coiled coil and that a
potent inhibitor should have amphiphilic structures to cooper-
atively interact with both pockets. This possibility was explored
by incorporating both lipophilic and hydrophilic groups into
the conjugates in a well-defined orientation to bind with both
pockets in the gp41 NHR-trimer.
Introduction
The rapid spread of human immunodeficiency virus (HIV) has
resulted in the infection of more than 60 million people world-
wide and the death of over one-third of those infected. There-
fore, finding effective therapeutic agents for this disease is one
of the major goals in modern medicinal chemistry. To date, 28
anti-HIV new chemical entities have been approved by the US
Food and Drug Administration (FDA), most of which target re-
verse transcriptase (RT) and protease.[1,2] The emergence of
drug-resistant HIV mutants and serious adverse effects has,
however, led to increasing response failure of HIV patients to
the current reverse transcriptase inhibitors (RTIs) and protease
inhibitors (PIs).[3–5] Therefore, major efforts are currently fo-
cused on the discovery of new drugs that act in the early
stages of HIV fusion and entry.[6–8]
Figure 1. Structure of the gp41 six-helix bundle.
From this perspective, gp41, a transmembrane subunit of
the HIV type 1 (HIV-1) envelope glycoprotein, can serve as a
promising target. X-ray crystallographic studies have shown
that the gp41 fusion core is a six-helix bundle (6-HB), in which
three N-terminal heptad repeat (NHR) helices exist as an inter-
nal trimeric coiled coil and three C-terminal heptad repeat
(CHR) peptide helices pack in an antiparallel orientation into
the hydrophobic grooves on the surface of the NHR coiled coil
(Figure 1).[9] The deep hydrophobic pockets in these grooves
are thought to be critical for the stabilization of the gp41 6-HB
core and for gp41-mediated membrane fusion;[10,11] as such,
they are attractive drug targets to prevent early viral fusion
events.[12] Up to now, several peptides derived from the gp41
[a] Dr. Y. Liu, K. Y. Wu, Prof. Z.-H. Jiang
School of Chinese Medicine, Hong Kong Baptist University
7 Baptist University Road, Kowloon Tong, Kowloon, Hong Kong (P.R. China)
Fax: (+852)341-12-461
zhjiang@ hkbu.edu.hk
[b] Dr. Z. Ke
Fukui Institute for Fundamental Chemistry, Kyoto University
Takano-Nishihiraki-cho 34-4, Sakyou-ku, Kyoto 606-8103 (Japan)
[c] Prof. S. Liu, Prof. W.-H. Chen, Prof. Z.-H. Jiang
School of Pharmaceutical Sciences, Southern Medical University
Tonghe, Guangzhou 510515 (P.R. China)
[d] Prof. S. Jiang
Lindsley F. Kimball Research Institute, New York Blood Center
310 E 67th Street, New York, NY 10021 (USA)
1654
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1654 – 1664

Betulinic Acid–Polyphenol Conjugates
CHR region, such as SJ-2176,[13] DP-178 (enfuvirtide, also
known as T-20)[14] and C34,[15] have been reported to inhibit
HIV-1 fusion/entry at nanomolar concentrations. Enfuvirtide
was approved by the US FDA in 2003 as the first member of a
new class of anti-HIV drugs—HIV fusion inhibitors.[16,17] These
CHR peptides are believed to interact with the HIV-1 gp41
NHR to form heterologous 6-HB and block viral gp41 6-HB
core formation, thereby inhibiting fusion between the viral and
target cell membranes.[18,19] However, these inhibitors have
some disadvantages, such as poor oral bioavailability and high
production costs,[20] both of which have largely restricted their
extensive application. Therefore, the development of orally
available nonpeptide, small-molecule HIV-1 fusion inhibitors
with a mechanism of action similar to the CHR peptides is ur-
gently needed.
trimeric coiled coil. The docking studies revealed that the gp41
NHR trimeric coiled coil contains both hydrophobic and hydro-
gen-bonding pockets. A potent HIV gp41 inhibitor could po-
tentially be obtained through cooperative binding of an agent
in these two kinds of pockets. Based on these findings, we de-
signed and synthesized a series of BA–polyphenol conjugates,
tested their inhibitory activities against HIV-1 gp41 6-HB forma-
tion, and studied their potential binding modes.
Results and Discussion
Molecular design
It is well known that the hydrophobic grooves of the gp41
NHR trimeric coiled coil play a crucial role in the formation and
stabilization of the 6-HB.[10,11] As such, these grooves are attrac-
tive targets for anti-HIV drugs to inhibit early fusion events.[9,12]
Some key interactions between C34 and the N36 trimeric
coiled coil are depicted in Figure 2 (PDB ID: 1AIK[9]). It can be
seen that three hydrophobic residues of C34, with large side
chains (Trp628, Trp631 and Ile635) and a charged residue,
Asp632, bind to the hydrophobic pocket of trimeric N36 pep-
tide through specific hydrophobic and salt-bridge interactions.
In this hydrophobic pocket, Lys574 of the N36 peptide (in
chain 3) is the key residue for the salt-bridge interaction (hy-
drogen bond and charged interactions) with Asp632 of the
C34 peptide, in an agreement with previous observations.[39,40]
Residues Leu565, Leu568, Trp571 of N36 (in chain 2) and
It has been reported that betulinic acid (BA) and its deriva-
tives exhibit anti-HIV-1 activities[21–24] via modes of action asso-
ciated with several steps in the virus life cycle, includ-
ing fusion, reverse transcription, and maturation.[25–31]
Further study of their structure–activity relationships
verified that the C-3 group is important for inhibition
of the later stages of the virus life cycle[32,33] and is
the pharmacophoric group for antimaturation activi-
ty, whereas the C-28 side chain is the pharmacophor-
ic moiety for HIV-1 entry inhibition.[22,34,35] Recent
studies have indicated that some plant polyphenols
with a hydrophilic galloyl group, such as tannins, cat-
echins and theaflavins, can significantly inhibit HIV-1
entry by blocking the formation of HIV-1 gp41 6-HB,
a fusion core conformation.[36–38] Based on these find-
ings, we hypothesized that 1) two different binding
sites may exist in gp41, and 2) incorporating phenolic
galloyl or caffeoyl groups into the C-28 position of
BA may lead to the discovery of a new class of HIV-1
entry inhibitors. Thus far, our preliminary studies
have led to the design and synthesis of novel inhibi-
tors using the natural product BA and plant-derived
polyphenols as building blocks.
Therefore, in the present study, we specifically
sought to rationally design BA–polyphenol conju-
gates, to evaluate their inhibitory activity on gp41 6-
HB formation, and to clarify their potential mecha-
nism of action against gp41 peptides at the molecu-
lar level. To accomplish this objective, we first carried
out a docking study to model the possible binding
Figure 2. Key residues of N36 (blue) and C34 (green) peptides for the formation and sta-
mode of BA and galloyl/caffeoyl groups in the N36 bilization of gp41 6-HB.
ChemMedChem 2011, 6, 1654 – 1664
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1655

Z.-H. Jiang et al.
MED
Val570, Ile573 of N36 (in chain 3) are key residues for the hy-
drophobic interactions with the C34 peptide. Based on these
interactions, a molecule that can bind to this hydrophobic
pocket is suggested to be capable of blocking the formation
of the gp41 6-HB core.
Coupling with excess hexane-1,6- and heptane-1,7-diamines
formed amides 4 and 5, respectively. Saponification of 4 and 5
with sodium hydroxide in a mixture of tetrahydrofuran and
methanol yielded 6 and 7, respectively.
Condensation of 4–7 with 3,4,5-triacetoxybenzoic acid or 3-
(3,4-diacetoxyphenyl)acrylic acid in the presence of N,N’-dicy-
clohexylcarbodiimide (DCC) in tetrahydrofuran at 08C afforded
compounds 8–11 and 18–21 in good to high yields (64–86%).
Subsequent treatment with hydrochloric acid in acetone fur-
nished the target BA–polyphenol conjugates 12–15 and 22–25
in moderate to good yields. Palladium-catalyzed hydrogena-
tion of the isopropylidene moiety in compounds 1, 12 and 13,
formed 26, 16 and 17 in high yield.
Therefore, we chose BA and plant polyphenols as the build-
ing blocks to evaluate how conjugates of these two compli-
mentary moieties would bind with the gp41 NHR trimeric
coiled coil. These two scaffolds are quite different in structure;
BA has a highly hydrophobic skeleton, whereas polyphenols
have multiple hydrophilic hydroxy groups. Their inherent struc-
tural dichotomy may correspond to different inhibitory types
of behavior against HIV-1 gp41 6-HB formation. To prove this
concept at the molecular level, we carried out docking studies
for BA and polyphenol groups with an N36 trimeric coiled coil
structure (PDB ID: 1AIK[9]) using AutoDock4.2.[41]
Biological assays
As shown in Figure 3, there were three possible binding
modes for BA (A, B and C) with similar calculated binding free
energies (ꢀ8.6, ꢀ8.4 and ꢀ8.3 kcalmolꢀ1, respectively). Howev-
er, binding mode A appeared to be the most favorable. In this
mode, we observed a key salt bridge between Lys574 and the
carboxylic group at C-28, as well as a hydrogen-bonding inter-
action between Arg579 and the hydroxy group at C-3. We also
saw a hydrophobic interaction between the skeleton of BA
and key residues Leu568, Val570, Trp571 and Ile573 in the hy-
drophobic pocket. Similar to the binding mode A, binding
modes B and C both have salt-bridge interaction with Lys574,
but the skeleton of BA has different orientations in the hydro-
phobic pocket. From the standpoint of new drug discovery,
the multiple binding modes of BA may raise the possibility of
modifying BA in a way that affords more potent inhibitors.
That is, with different functional side chains, BA derivatives
may act through the induced-fit process to interact with the
protein.
The in vitro inhibitory activity of BA–polyphenol conjugates 6–
26 against HIV-1 gp41 6-HB formation between the NHR pep-
tide N36 and the CHR peptide C34 was evaluated using size-
exclusion (SE)-HPLC. This assay is based on SE chromatography
separation of the 6-HB from the free peptides N36 and C34.
This 6-HB formation mimics the function of HIV-1 gp41 in the
biological system that is required for HIV-1 virus to enter into
the host cell. The experimental protocols used were similar to
those described in the literature.[42–44] The typical SE-HPLC ex-
perimental design and the results of the assay are presented in
Figure 5.
Specifically, under our assay condition, the Bio-Rad gel filtra-
tion standard (thyroglobulin, gamma globulin, ovalbumin, my-
oglobin and vitamin B12) could be effectively separated (Fig-
ure 5a). Isolated C34 peptide showed a single peak with a re-
tention time of 8.20 min (Figure 5b). The calculated molecular
weight was approximately 12.9 kDa, corresponding to a trimer
in phosphate-buffered saline. N36 peptide could not be de-
tected at 280 nm. These observations were consistent with
findings in the literature.[45] The mixture of N36 and C34 pep-
tides showed two peaks at 6.99 min and 8.20 min (C34) (Fig-
ure 5c). The former has a calculated molecular weight of about
25.7 kDa, corresponding to 6-HB. The peak area of 6-HB was
reduced when BA was added, suggesting that the process of
6-HB formation was, in part, inhibited by BA (Figure 5d). Sub-
sequently, the effect of synthesized BA–polyphenol conjugates
against the formation of the HIV-1 gp41 6-HB were determined
using the same method. Based on inhibition of 6-HB formation,
the percent inhibition was calculated according to Equa-
tion (1), where A0 is the 6-HB peak area measured without in-
hibitor and Ai is the 6-HB peak area when treating with inhibi-
tor.
To provide insight into the possible binding sites of galloyl
and caffeoyl groups, we used 1,2,3-trihydroxybenzene and 1,2-
dihydroxybenzene in our docking study. As shown in Figure 4a
and b, specific hydrogen-bonding sites exist for these two
compounds. In both binding sites, Arg579 and Gln575 are pre-
dicted to play important roles in the hydrogen-bonding inter-
actions with polyphenols. Interestingly, this hydrogen-bonding
site is in close proximity to the hydrophobic pocket. Thus, con-
jugation of BA with polyphenols may lead to a potent inhibitor
of 6-HB core formation, in which the hydrophobic interaction
of the BA subunit and the multiple hydrogen bonds formed by
the polyphenol group cooperatively enhance the binding affin-
ity and specificity of the conjugate.
Synthesis
Inhibition ¼ ðA0ꢀAiÞ=A0 ꢁ 100 %
ð1Þ
Based on the above-mentioned docking study, we designed
and synthesized BA–polyphenol conjugates 2–26 (Scheme 1).
BA 1 was conjugated with galloyl and caffeoyl groups at the
C-28 position of BA through a flexible spacer. First, the C-3 hy-
droxy group was protected with an acetyl group to afford
compound 2. Subsequent treatment of 2 with oxalyl chloride
in dichloromethane at room temperature gave acyl chloride 3.
As can be seen from Figure 6, BA showed moderate inhibi-
tion at 200 mgmLꢀ1 (56%), whereas gallic acid was inactive
under our assay condition. Interestingly, BA–polyphenol conju-
gates bearing the galloyl group exhibited significantly en-
hanced inhibitory activities. For example, compound 12 and
13 resulted in more than 90% loss of 6-HB, and compounds
1656
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1654 – 1664

Betulinic Acid–Polyphenol Conjugates
Figure 3. Three binding modes of BA with N36 trimeric coiled coil.
Table 1. The inhibitory activities of BA (1), gallic acid and compounds 6–
26 against the formation of HIV-1 gp41 6-HB in PBS (pH 7.2) at 378C.
Compd
IC50 [mm]
>5000
Compd
IC50 [mm]
16 and 17 completely inhibit the formation of the HIV-1 gp41
6-HB.
Gallic acid
1
6
7
8
9
12
13
14
15
16
17
22
23
24
25
26
375.3ꢂ20.8
66.4ꢂ2.5
408.4ꢂ23.4
427.5ꢂ21.1
359.3ꢂ29.1
511.6ꢂ61.7
503.2ꢂ42.1
85.3ꢂ5.6
To better understand the structure–activity relationships, the
IC50 value of BA was calculated from the subsequent dose–re-
sponse experiment. The inhibitory activities of all of the other
BA–polyphenol conjugates were evaluated in a similar way.
The obtained IC50 values, together with those of BA and gallic
acid for comparison, are summarized in Table 1.[46]
77.2ꢂ2.4
213.6ꢂ23.8
170.4ꢂ11.1
543.3ꢂ66.0
490.7ꢂ59.8
291.7ꢂ28.3
98.4ꢂ4.5
362.9ꢂ54.0
ChemMedChem 2011, 6, 1654 – 1664
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1657

Z.-H. Jiang et al.
MED
Several interesting structure–activity relationships can be
elucidated from analysis of the IC50 values of these BA–poly-
phenol conjugates. First, among the starting compounds (BA,
gallic acid), intermediate compounds (6 and 7) and BA–poly-
phenol conjugates (12, 13, 16 and 17), those bearing the gal-
loyl group exhibited significantly enhanced inhibitory activities.
For example, compounds 12
(IC50 =85.3ꢂ5.6 mm) and 13
(IC50 =98.4ꢂ4.5 mm) showed up
to fivefold higher inhibitory ac-
tivity than BA. However, com-
pounds
14
(IC50 =362.9ꢂ
54.0 mm) and 15 (IC50 =375.3ꢂ
20.8 mm) exhibited comparable
activities to BA, suggesting that
a caffeoyl group at the C-28 side
chain is not favorable for im-
proving
inhibitory
activity.
Second, acetylation of the galloyl
and caffeoyl groups resulted in a
dramatic decrease in the inhibi-
tory activities, suggesting that
free phenolic hydroxy groups
are essential to the biological ac-
tivities. Third, compounds 6 and
7 exhibited comparable 6-HB in-
hibitory activities to BA, indicat-
ing that flexible diamine spacers
have little effect on activity. All
these results strongly suggest
that the galloyl group in these
conjugates make substantial
contributions to the inhibitory
activities against gp41 6-HB core
formation.
It has been reported that the
C-3 group in BA is important for
activity against the later stages
of the virus life cycle. However,
the influence of the C-3 group
on gp41 6-HB core formation is
not clear. Under our assay condi-
tion, acetylation of the C-3 hy-
droxy group in conjugates 12–
15 (compounds 22–25) led to a
pronounced decrease in activity.
This result indicates that a free
hydroxy group in the C-3 posi-
tion plays a role that cannot be
ignored in the inhibition of gp41
6-HB core formation. In addition,
reduction of the double bond of
compounds 1, 12 and 13 (com-
pounds 26, 16 and 17, respec-
tively) did not show much effect
on the activity.
Scheme 1. Synthetic route for BA–polyphenol conjugates and their derivatives. Reagents and conditions: a) H2, Pd/
C, RT, 2 h; b) Ac2O, Py, RT, overnight; c) (CO)2Cl2, CH2Cl2, RT, 0.5 h; d) H2N(CH2)nNH2, CH2Cl2, RT, overnight; e) NaOH,
CH3OH, THF, RT, overnight; f) DCC, THF, 08C!RT, 8 h (64–86%); g) HCl, acetone, reflux, 3 h (34–53%); h) H2, PD/C,
RT, 2 h (91–94%).
1658
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1654 – 1664

Betulinic Acid–Polyphenol Conjugates
Figure 5. SE-HPLC chromatograms. a) Standards for calibration (1: thyroglo-
bulin; 2: gamma globulin; 3: ovalbumin; 4: myoglobin; 5: vitamin B12);
b) C34 peptide; c) mixture of C34 and N36 peptides; and d) mixture of C34,
BA and N36.
Figure 4. Binding modes of a) 1,2-dihydroxybenzene and b) 1,2,3-trihydroxy-
benzene with N36 trimeric coiled coil.
Binding modes
To clarify how these BA–polyphenol conjugates interact with
the gp41 NHR domains, as well as to provide some guidance
for the future rational design of gp41 6-HB inhibitors, docking
studies were performed for two typical inhibitors, 12 and 14.
As shown in Figure 7a, in the most favorable binding mode
of compound 12, the BA moiety adopts an orientation that is
very similar to the binding mode C of BA (Figure 3c), although
mode A was the most favorable for BA binding (Figure 3a).
The binding free energy of 12 was calculated to be ꢀ10.8 kcal
molꢀ1, which is much lower than that of BA itself. In this bind-
ing mode (Figure 7a), the BA moiety of 12 bound to the NHR
Figure 6. Inhibitory activities of BA (1), gallic acid and compounds 6–26
(200 mgmLꢀ1) against HIV-1 gp41 6-HB formation.
sequence through a hydrogen-bonding interaction of the C-28
carbonyl group with Lys574, and hydrophobic interactions
with residues Leu565, Leu568, Val570 and Ile573, which are
key residues for the formation and stabilization of gp41 6-HB
(Figure 2). Meanwhile, as expected, the galloyl moiety of com-
ChemMedChem 2011, 6, 1654 – 1664
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1659

Z.-H. Jiang et al.
MED
less potent inhibitors than those bearing a galloyl group. Thus,
proper size/topology of the side chain in BA conjugates may
be one crucial structural factor that modulates the cooperative
interactions of the conjugate with both the hydrophobic and
hydrogen-bonding pockets. This consideration should weigh
heavily in optimizing the design of BA-based anti-HIV drugs
that target gp41.
Although it is still a challenge to quantitatively calculate the
binding free energy by docking procedures, our simulations
qualitatively correlate the predicted binding free energies and
the inhibitory activities of the BA–polyphenol conjugates (see
Figure 8a). Compounds 12, 13, 16 and 17, exhibiting higher in-
Figure 7. Binding modes of compounds a) 12 and b) 14 in the N36 trimeric
coiled coil.
pound 12 is predicted to form a hydrogen-bonding network
with residues Arg579, Gln577 and Gln575. Exactly consistent
with our design, these docking studies predict cooperative in-
teractions of the BA and galloyl moieties of compound 12 with
the two pockets, explaining the potent inhibitory activity.
It should be noted that introduction of a caffeoyl group at
C-28 of BA gave no significant enhancement in inhibitory activ-
ity. To better understand the interaction with the hydrophobic
and hydrogen-bonding pockets, a docking study was carried
out for compound 14. Its most favorable binding mode is de-
picted in Figure 7b. Interestingly, although the BA moiety of
14 still interacts well with the hydrophobic pocket, the caffeoyl
group has no significant interactions with the hydrogen-bond-
ing pocket. The binding free energy is calculated to be only
ꢀ9.7 kcalmolꢀ1, which is much higher than that of compound
12. This result is fully consistent with the aforementioned ob-
servations that the conjugates bearing a caffeoyl group are
Figure 8. a) Correlation of binding free energies with inhibitory activities for
BA–polyphenol conjugates. b) Binding free energy contribution of the BA
moiety (&), the polyphenol moiety (&) and the total binding free energy
(&) for the BA–polyphenol conjugates.
hibitory activities, are predicted to have binding free energies
lower than ꢀ10.5 kcalmolꢀ1, while conjugates without signifi-
cant improvements in inhibitory activity are predicted to have
weaker binding interactions with the gp41 NHR trimeric coiled
coil (ꢀ9.6 to ꢀ9.8 kcalmolꢀ1). All these conjugates have very
similar predicted binding modes, where BA binds to the hydro-
phobic pocket and the polyphenol moiety binds to the hydro-
gen-bonding pocket. This further supports our hypothesis that
1660
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1654 – 1664

Betulinic Acid–Polyphenol Conjugates
Brandstead Electrothermal IA9100 melting apparatus and are un-
corrected. HPLC analyses were performed using an Agilent 1100
series instrument (Santa Clara, CA, USA), and a Bio-Sil SEC 125–5
column (300 mmꢁ7.8 mm, Bio-Rad, Hercules, CA, USA) was used
for size-exclusion (SE)-HPLC to determine the 6-HB formation.
Docking calculations were performed using AutoDock v4.2.
BA–polyphenol conjugates interact cooperatively with the hy-
drophobic and hydrogen-bonding pockets through the BA and
polyphenol motifs, respectively, to enhance the overall inhibi-
tory activity of the conjugate. The binding contributions from
the BA and the polyphenol moieties for each conjugate are
shown in Figure 8b. It can be found that conjugates with
stronger activity usually have better binding contributions
from both BA and polyphenol parts. Less active conjugates
have decreased binding contribution from the BA moiety due
to adapting the conformational change of the polyphenol
moiety, and vice versa. Therefore, it remains an opportunity,
and also a challenge, to further design BA–polyphenol-type
conjugates that are able to perfectly fit these crucial binding
pockets to obtain more potent inhibitors targeting the gp41
NHR trimeric coiled coil.
Peptides N36 (SGIVQQQNNLLRAIEAQQHLLQLTVWGIKQL QARIL)
and C34 (WMEWDREINNYTSLIHSLIEESQNQQEK NEQELL) were
purchased from Bachem AG (Bubendorf, Switzerland). Betulinic
acid (BA) was purchased from Aladdin (Shanghai, P.R. China). Com-
pounds 2–7 and 26 were prepared according to the reported pro-
cedures.[31,47–49] All other reagents were of analytical quality and
used as received.
General procedure for the synthesis of 8–11 and 18–21: N,N’-Di-
cyclohexylcarbodiimide (DCC; 206 mg, 1 mmol) was added to a so-
lution of 3,4,5-triacetoxybenzoic acid (triacetyl gallic acid; 1 mmol)
or 3-(3,4-diacetoxyphenyl)acrylic acid (diacetyl caffeic acid; 1 mmol)
in anhydrous THF (10 mL). The resulting mixture was stirred at 08C
for 0.5 h. Then a solution of compound 4, 5, 6 or 7 (1 mmol) in
THF (5 mL) was added dropwise, and stirring was continued for an-
other 8 h at RT. The reaction mixture was poured onto ice water
(100 mL) and extracted with EtOAc (50 mL). The organic layer was
washed with brine and water, dried (MgSO4), filtered and concen-
trated in vacuo. The crude products were purified by flash silica gel
column chromatography to afford compounds 8–11 and 18–21 as
white powders.
Conclusions
In summary, this paper describes the design of betulinic acid
(BA)–polyphenol conjugates as novel inhibitors of gp41 fusion
core formation. Molecular modeling suggested that hydropho-
bic and polyphenol motifs are capable of cooperatively inter-
acting with the hydrophobic and hydrogen-bonding pockets
in the binding groove of the gp41 NHR trimeric coiled coil.
The hydrophobic pocket mainly consists of residues Lys574,
Leu565, Leu568, Trp571, Val570 and Ile573, whereas the hy-
drogen-bonding pocket is mainly composed of Arg579,
Gln577 and Gln575. The hydrophobic skeleton of the BA motif
is predicted to bind to the hydrophobic pocket, and the poly-
phenol motif is suggested to form a hydrogen-bonding net-
work within the hydrogen-bonding pocket.
3,4,5-Triacetoxy-N-[6-[[3b-hydroxylup-20(29)-en-28-oyl]amino]-
hexyl]benzamide (8): From compound 6 and 3,4,5-triacetoxyben-
zoic acid (64%); mp: 175–1778C; 1H NMR (300 MHz, CDCl3): d=
7.54 (s, 2H, galloyl-H), 6.44 (s, 1H, CONH), 5.72 (s, 1H, CONH), 4.71
(s, 1H, =CH), 4.58 (s, 1H, =CH), 3.48–3.42 (m, 2H, CONHCH2), 3.32–
3.27 (m, 1H, H-3), 3.19–3.09 (m, 2H, CONHCH2), 2.30 (s, 9H,
CH3COOAr), 1.94 (s, 1H, C-3OH), 1.68 (s, 3H, H-30), 0.96 (s, 9H, CH3),
0.81 (s, 3H, CH3), 0.75 (s, 3H, CH3), 0.67 ppm (s, 1H, H-5); IR (KBr):
To validate this concept, a series of novel BA–polyphenol
conjugates was designed and synthesized in moderate to high
yields. The results of size-exclusion (SE)-HPLC assays showed
that BA–polyphenol conjugates block HIV-1 gp41 six-helix
bundle (6-HB) core formation in a dose-dependent manner.
Among them, conjugates 12, 13, 16 and 17, bearing a galloyl
group at the C-28 side chain, exhibit four to sixfold higher in-
hibitory activity than that of the parent compound BA. This
result not only demonstrates the advantage of conjugate
chemistry in drug design, but also verifies the applicability and
reliability of computational docking studies on BA derivatives
binding to HIV-1 gp41 6-HB. Moreover, by considering both
the hydrophobic and hydrogen-bonding pockets of the gp41
NHR trimeric coiled coil in our molecular design, the BA–poly-
phenol conjugates can be further modified to discover novel
molecules able to perfectly fit these crucial binding pockets.
n˜ =3329, 2934, 2857, 1783, 1628, 1536, 1371, 1325, 1188, 1052,
892, 546 cmꢀ1; HRMS (ESI): m/z [M+H] calcd for C49H73N2O9
:
+
833.5316, found: 833.5289.
3,4,5-Triacetoxy-N-[7-[[3b-hydroxylup-20(29)-en-28-oyl]amino]-
heptyl]benzamide (9): From compound 7 and 3,4,5-triacetoxyben-
zoic acid (74%); mp: 184–1868C; 1H NMR (300 MHz, CDCl3): d=
7.51 (s, 2H, galloyl-H), 6.25 (t, J=5.4 Hz, 1H, CONH), 5.64 (t, J=
5.4 Hz, 1H, CONH), 4.72 (s, 1H, =CH), 4.58 (s, 1H, =CH), 3.44–3.37
(m, 2H, CONH-CH2), 3.31–3.25 (m, 1H, H-3), 3.22–3.11 (m, 2H,
CONH-CH2), 2.30 (s, 9H, CH3-COO-Ar), 2.05 (s, 1H, C3-OH), 1.68 (s,
3H, H-30), 0.96 (s, 6H, CH3), 0.94 (s, 3H, CH3), 0.82 (s, 3H, CH3), 0.75
(s, 3H, CH3), 0.66 ppm (br d, J=9.0 Hz, 1H, H-5); IR (KBr): n˜ =3328,
2934, 2853, 1781, 1627, 1536, 1371, 1188, 1051, 892, 641 cmꢀ1
;
HRMS (ESI): m/z [M+H]+ calcd for C50H75N2O9: 847.5473, found:
847.5465.
(E)-3-(3,4-Diacetoxyphenyl)-N-[6-[[3b-hydroxylup-20(29)-en-28-
oyl]amino]hexyl] acrylamide (10): From compound 6 and 3-(3,4-
diacetoxyphenyl)acrylic acid (82%); mp: 171–1738C; 1H NMR
(300 MHz, CDCl3): d=7.52 (d, J=15.6 Hz, 1H, caffeoyl-H-8), 7.37 (d,
J=8.1 Hz, 1H, caffeoyl-H-6), 7.32 (s, 1H, caffeoyl-H-2), 7.18 (d, J=
8.1 Hz, 1H, caffeoyl-H-5), 6.37 (d, J=15.6 Hz, 1H, caffeoyl-H-7), 6.11
(s, 1H, CONH), 5.74 (s, 1H, CONH), 4.72 (s, 1H, =CH), 4.58 (s, 1H, =
CH), 3.38 (m, 2H, CONH-CH2), 3.29 (m, 1H, H-3), 3.20–3.15 (m, 2H,
CONH-CH2), 2.30 (s, 6H, CH3-COO-Ar), 2.05 (s, 1H, C3-OH), 1.68 (s,
3H, H-30), 0.97 (s, 3H, CH3), 0.96 (s, 3H, CH3), 0.94 (s, 3H, CH3), 0.81
(s, 3H, CH3), 0.74 (s, 3H, CH3), 0.68 ppm (br d, J=9.0 Hz,1H, H-5); IR
(KBr): n˜ =3379, 2943, 1735, 1714, 1637, 1606, 1520, 1446, 1371,
Experimental Section
Chemistry
NMR spectra were recorded on a Varian INOVA 300 MHz NMR spec-
trometer in either CDCl3 or [D6]DMSO, and tetramethylsilane (TMS)
was used as an internal standard. High-resolution mass spectra
(HRMS) were measured on a Bruker MicrOTOFQ mass spectrometer.
Infrared (IR) spectra were obtained on a Perkin–Elmer Spectrum
One FT-IR spectrometer. Melting points (mp) were determined on a
ChemMedChem 2011, 6, 1654 – 1664
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1661

Z.-H. Jiang et al.
MED
1288, 1195, 1031, 982, 841 cmꢀ1; HRMS (ESI): m/z [M+H]+ calcd for
C49H73N2O7: 801.5418, found: 801.5406.
CH), 3.38 (m, 2H, CONH-CH2), 3.27–3.25 (m, 1H, H-3), 3.20–3.10 (m,
2H, CONH-CH2), 2.31 (s, 6H, CH3-COO-Ar), 2.04 (s, 3H, C3-COOCH3),
1.68 (s, 3H, H-30), 0.96 (s, 3H, CH3), 0.94 (s, 3H, CH3), 0.84 (s, 9H,
CH3), 0.77 ppm (br d, J=9.0 Hz, 1H, H-5); IR (KBr): n˜ =1735, 1713,
(E)-3-(3,4-Diacetoxyphenyl)-N-[7-[[3b-hydroxylup-20(29)-en-28-
oyl]amino]heptyl] acrylamide (11): From compound 7 and 3-(3,4-
diacetoxyphenyl)acrylic acid (86%); mp: 179–1818C; 1H NMR
(300 MHz, CDCl3): d=7.55 (d, J=15.6 Hz, 1H, caffeoyl-H-8), 7.35 (d,
J=8.1 Hz, 1H, caffeoyl-H-6), 7.32 (s, 1H, caffeoyl-H-2), 7.18 (d, J=
8.1 Hz, 1H, caffeoyl-H-5), 6.31 (d, J=15.6 Hz, 1H, caffeoyl-H-7), 5.71
(t, J=5.1 Hz, 1H, CONH), 5.61 (t, J=5.1 Hz, 1H, CONH), 4.73 (s, 1H,
=CH), 4.58 (s, 1H, =CH), 3.40–3.34 (m, 2H, CONH-CH2), 3.32–3.23
(m, 1H, H-3), 3.21–3.10 (m, 2H, CONH-CH2), 2.30 (s, 6H, CH3-COO-
Ar), 2.05 (s, 1H, C3-OH), 1.68 (s, 3H, H-30), 0.97 (s, 6H, CH3), 0.94 (s,
3H, CH3), 0.81 (s, 3H, CH3), 0.75 (s, 3H, CH3), 0.68 ppm (br d, J=
9.0 Hz,1H, H-5); IR (KBr): n˜ =3381, 2943, 1734, 1716, 1638, 1607,
1520, 1445, 1370, 1288, 1194, 1032, 983, 843 cmꢀ1; HRMS (ESI): m/z
[M+H]+ calcd for C50H75N2O7: 815.5574, found: 815.5586.
1635, 1601, 1515, 1443, 1370, 1286, 1193, 1032, 979, 839 cmꢀ1
;
HRMS (ESI): m/z [M+H]+ calcd for C52H77N2O8: 857.5680, found:
857.5684.
General procedure for the synthesis of 12–15 and 22–25: A solu-
tion of compound 8–11 or 18–21 (0.5 mmol) in acetone (10 mL)
was treated with HCl (3n, 2 mL). The reaction mixture was refluxed
for 3 h. After cooling to RT, the reaction mixture was diluted with
EtOAc (50 mL), washed with brine and water, dried (MgSO4), fil-
tered and concentrated in vacuo, and purified by flash column
chromatography to give 12–15 or 22–25 as white powders.
3,4,5-Trihydroxy-N-[6-[[3b-hydroxylup-20(29)-en-28-oyl]amino]-
hexyl]benzamide (12): From compound 8 (38%); mp: 151–1538C;
1H NMR (300 MHz, [D6]DMSO): d=8.93 (s, 2H, galloyl-C3,5-OH),
8.56 (s, 1H, galloyl-C4-OH), 7.98 (t, J=5.1 Hz, 1H, CONH), 7.51 (t,
J=5.1 Hz, 1H, CONH), 6.78 (s, 2H, ArH), 4.63 (s, 1H, =CH), 4.51 (s,
1H, =CH), 3.17–3.11 (m, 2H, CONH-CH2), 3.07–3.00 (m, 1H, H-3),
2.96–2.87 (m, 2H, CONH-CH2), 1.98 (s, 1H, C3-OH), 1.61 (s, 3H, H-
30), 0.90 (s, 3H, CH3), 0.85 (s, 3H, CH3), 0.83 (s, 3H, CH3), 0.75 (s, 3H,
CH3), 0.63 (s, 3H, CH3), 0.59 ppm (s, 1H, H-5); IR (KBr): n˜ =3376,
2938, 2864, 1637, 1599, 1520, 1452, 1374, 1347, 1196, 1036, 982,
881, 765, 648 cmꢀ1; HRMS (ESI): m/z [MꢀH]ꢀ calcd for C43H65N2O6:
705.4843, found: 705.4842.
3,4,5-Triacetoxy-N-[6-[[3b-acetoxylup-20(29)-en-28-oyl]amino]-
hexyl]benzamide (18): From compound 4 and 3,4,5-triacetoxyben-
zoic acid (70%); mp: 205–2078C; 1H NMR (300 MHz, CDCl3): d=
7.53 (s, 2H, galloyl-H), 6.37 (t, J=6.0 Hz, 1H, CONH), 5.71 (t, J=
6.0 Hz, 1H, CONH), 4.72 (s, 1H, =CH), 4.58 (s, 1H, =CH), 3.45–3.38
(m, 2H, CONH-CH2), 3.35–3.26 (m, 1H, H-3), 3.20–3.09 (m, 2H,
CONH-CH2), 2.30 (s, 9H, CH3-COO-Ar), 2.04 (s, 3H, C3-COOCH3), 1.68
(s, 3H, H-30), 0.96 (s, 3H, CH3), 0.94 (s, 3H, CH3), 0.84 (s, 6H, CH3),
0.83 (s, 3H, CH3), 0.78 ppm (br d, J=9.0 Hz, 1H, H-5); IR (KBr): n˜ =
1781, 1627, 1575, 1536, 1496, 1449, 1372, 1325, 1271, 1187, 1050,
891, 545 cmꢀ1; HRMS (ESI): m/z [M+H]+ calcd for C51H75N2O10:
875.5422, found: 875.5402.
3,4,5-Trihydroxy-N-[7-[[3b-hydroxylup-20(29)-en-28-oyl]amino]-
heptyl]benzamide (13): From compound 9 (46%); mp: 163–
1
3,4,5-Triacetoxy-N-[7-[[3b-acetoxylup-20(29)-en-28-oyl]amino]-
heptyl]benzamide (19): From compound 5 and 3,4,5-triacetoxy-
benzoic acid (76%); mp: 213–2158C; 1H NMR (300 MHz, CDCl3):
d=7.51 (s, 2H, galloyl-H), 6.17 (t, J=5.4 Hz, 1H, CONH), 5.62 (t, J=
5.4 Hz, 1H, CONH), 4.73 (s, 1H, =CH), 4.59 (s, 1H, =CH), 3.44–3.38
(m, 2H, CONH-CH2), 3.29–3.25 (m, 1H, H-3), 3.20–3.13 (m, 2H,
CONH-CH2), 2.30 (s, 9H, CH3-COO-Ar), 2.04 (s, 3H, C3-COOCH3), 1.68
(s, 3H, H-30), 0.96 (s, 3H, CH3), 0.94 (s, 3H, CH3), 0.84 (s, 9H, CH3),
0.78 ppm (br d, J=9.0 Hz,1H, H-5); IR (KBr): n˜ =1784, 1629, 1576,
1538, 1495, 1450, 1327, 1188, 1052, 892, 547 cmꢀ1; HRMS (ESI): m/z
[M+H]+ calcd for C52H77N2O10: 889.5578, found: 889.5508.
1658C; H NMR (300 MHz, [D6]DMSO): d=8.94 (s, 2H, galloyl-C3,5-
OH), 8.56 (s, 1H, galloyl-C4-OH), 7.98 (t, J=5.1 Hz, 1H, CONH), 7.50
(t, J=5.1 Hz, 1H, CONH), 6.78 (s, 2H, ArH), 4.63 (s, 1H, =CH), 4.51
(s, 1H, =CH), 3.17–3.08 (m, 2H, CONH-CH2), 3.01–3.06 (m, 1H, H-3),
2.96–2.88 (m, 2H, CONH-CH2), 1.98 (s, 1H, C3-OH), 1.61 (s, 3H, H-
30), 0.90 (s, 3H, CH3), 0.85 (s, 3H, CH3), 0.84 (s, 3H, CH3), 0.75 (s, 3H,
CH3), 0.63 (s, 3H, CH3), 0.59 ppm (s, 1H, H-5); IR (KBr): n˜ =3354,
2936, 2853, 1697, 1637, 1528, 1450, 1375, 1318, 1195, 1042, 882,
639, 603 cmꢀ1; HRMS (ESI): m/z [MꢀH]ꢀ calcd for C44H67N2O6:
719.4999, found: 719.4973.
(E)-3-(3,4-Dihydroxyphenyl)-N-[6-[[3b-hydroxylup-20(29)-en-28-
oyl]amino]hexyl] acrylamide (14): From compound 10 (50%); mp:
(E)-3-(3,4-Diacetoxyphenyl)-N-[6-[[3b-acetoxylup-20(29)-en-28-
oyl]amino]hexyl]acryl amide (20): From compound 4 and 3-(3,4-
diacetoxyphenyl)acrylic acid (78%); mp: 199–2018C; 1H NMR
(300 MHz, CDCl3): d=7.55 (d, J=15.6 Hz, 1H, caffeoyl-H-8), 7.36
(dd, J=8.1, 2.1 Hz, 1H, caffeoyl-H-6), 7.31 (d, J=2.1 Hz, 1H, caffeo-
yl-H-2), 7.18 (d, J=8.1 Hz, 1H, caffeoyl-H-5), 6.36 (d, J=15.6 Hz, 1H,
caffeoyl-H-7), 5.95 (t, J=5.7 Hz, 1H, CONH), 5.70 (t, J=5.7 Hz, 1H,
CONH), 4.72 (s, 1H, =CH), 4.58 (s, 1H, =CH), 3.41–3.35 (m, 2H,
CONH-CH2), 3.31–3.26 (m, 1H, H-3), 3.19–3.09 (m, 2H, CONH-CH2),
2.30 (s, 6H, CH3-COO-Ar), 2.04 (s, 3H, C3-COOCH3), 1.68 (s, 3H, H-
30), 0.97 (s, 3H, CH3), 0.94 (s, 3H, CH3), 0.84 (s, 6H, CH3), 0.83 (s, 3H,
CH3), 0.77 ppm (br d, J=9.0 Hz, 1H, H-5); IR (KBr): n˜ =1736, 1714,
1
136–1388C; H NMR (300 MHz, [D6]DMSO): d=9.19 (s, 2H, ArOH),
7.90 (t, J=5.1 Hz, 1H, CONH), 7.52 (t, J=5.1 Hz, 1H, CONH), 7.19
(d, J=15.9 Hz, 1H, caffeoyl-H-8), 6.91 (s, 1H, caffeoyl-H-2), 6.80 (d,
J=8.1 Hz, 1H, caffeoyl-H-6), 6.71 (d, J=8.1 Hz, 1H, caffeoyl-H-5),
6.29 (d, J=15.9 Hz, 1H, caffeoyl-H-7), 4.63 (s, 1H, =CH), 4.51 (s, 1H,
=CH), 3.13–3.10 (m, 2H, CONH-CH2), 3.05–3.01 (m, 1H, H-3), 2.97–
2.92 (m, 2H, CONH-CH2), 2.08 (s, 1H, C3-OH), 1.61 (s, 3H, H-30),
0.90 (s, 3H, CH3), 0.83 (s, 6H, CH3), 0.75 (s, 3H, CH3), 0.62 (s, 3H,
CH3), 0.62 ppm (s, 1H, H-5); IR (KBr): n˜ =3367, 2938, 2854, 1638,
1603, 1515, 1446, 1375, 1282, 1195, 1112, 1043, 981, 882, 811 cmꢀ1
;
HRMS (ESI): m/z [M+H]+ calcd for C45H69N2O5: 717.5206, found:
1639, 1608, 1521, 1445, 1371, 1289, 1196, 1033, 985, 840 cmꢀ1
;
717.5215.
HRMS (ESI): m/z [M+HCOOHꢀH]ꢀ calcd for C52H75N2O10: 887.5422,
(E)-3-(3,4-Didydroxyphenyl)-N-[7-[[3b-hydroxylup-20(29)-en-28-
oyl]amino]heptyl] acrylamide (15): From compound 11 (53%);
found: 887.5420.
1
(E)-3-(3,4-Diacetoxyphenyl)-N-[7-[[3b-acetoxylup-20(29)-en-28-
oyl]amino]heptyl] acrylamide (21): From compound 5 and 3-(3,4-
diacetoxyphenyl)acrylic acid (81%); mp: 207–2098C; 1H NMR
(300 MHz, CDCl3): d=7.56 (d, J=15.6 Hz, 1H, caffeoyl-H-8), 7.36 (d,
J=8.1 Hz, 1H, caffeoyl-H-6), 7.32 (s, 1H, caffeoyl-H-2), 7.18 (d, J=
8.1 Hz, 1H, caffeoyl-H-5), 6.34 (d, J=15.6 Hz, 1H, caffeoyl-H-7), 5.89
(s, 1H, CONH), 5.64 (s, 1H, CONH), 4.73 (s, 1H, =CH), 4.58 (s, 1H, =
mp: 144–1468C; H NMR (300 MHz, [D6]DMSO): d=9.30 (s, 1H, caf-
feoyl-OH), 9.07 (s, 1H, caffeoyl-OH), 7.91 (t, J=5.1 Hz, 1H, CONH),
7.51 (t, J=5.1 Hz, 1H, CONH), 7.19 (d, J=15.9 Hz, 1H, caffeoyl-H-8),
6.91 (d, J=2.1 Hz, 1H, caffeoyl-H-2), 6.79 (dd, J=8.1, 2.1 Hz, 1H,
caffeoyl-H-6), 6.71 (d, J=8.1 Hz, 1H, caffeoyl-H-5), 6.30 (d, J=
15.9 Hz, 1H, caffeoyl-H-7), 4.63 (s, 1H, =CH), 4.51 (s, 1H, =CH),
3.15–3.09 (m, 2H, CONH-CH2), 3.05–2.98 (m, 1H, H-3), 2.96–2.88 (m,
1662
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1654 – 1664

Betulinic Acid–Polyphenol Conjugates
2H, CONH-CH2), 1.98 (s, 1H, C3-OH), 1.61 (s, 3H, H-30), 0.90 (s, 3H,
CH3), 0.85 (s, 3H, CH3), 0.84 (s, 3H, CH3), 0.75 (s, 3H, CH3), 0.64 (s,
3H, CH3), 0.68 ppm (s, 1H, H-5); IR (KBr): n˜ =3360, 2939, 2866,
1637, 1602, 1516, 1447, 1357, 1282, 1195, 1112, 1031, 981,
882 cmꢀ1; HRMS (ESI): m/z [MꢀH]ꢀ calcd for C46H69N2O5: 729.5206,
found: 729.5209.
fied by column chromatography to yield compound 16 or 17 as a
white powder.
3,4,5-Trihydroxy-N-[6-[(3b-hydroxylupan-28-oyl)amino]hexyl]-
benzamide (16): From compound 12 (91%); mp: 179–1818C;
1H NMR (300 MHz, [D6]DMSO): d=8.97 (s, 2H, galloyl-C3,5-OH),
8.27 (s, 1H, galloyl-C4-OH), 7.96 (t, J=5.4 Hz, 1H, CONH), 7.41 (t,
J=5.4 Hz, 1H, CONH), 6.77 (s, 2H, galloyl-H), 3.16–3.10 (m, 2H,
CONH-CH2), 3.06–3.04 (m, 1H, H-3), 2.98–2.86 (m, 2H, CONH-CH2),
1.89 (s, 1H, C3-OH), 0.87 (s, 3H, CH3), 0.85 (s, 3H, CH3), 0.83 (s, 3H,
CH3), 0.79 (d, J=6.9 Hz, 1H, C29 or H-30), 0.76 (s, 3H, CH3), 0.69 (d,
J=6.9 Hz, 1H, H-29 or H-30), 0.63 (s, 3H, CH3), 0.60 ppm (s, 1H, H-
5); IR (KBr): n˜ =3365, 2941, 2883, 1653, 1597, 1517, 1449, 1369,
1345, 1192, 1035, 977, 869, 765 cmꢀ1; HRMS (ESI): m/z [MꢀH]ꢀ
calcd for C43H65N2O6: 707.4999, found: 707.4955.
3,4,5-Trihydroxy-N-[6-[[3b-acetoxylup-20(29)-en-28-oyl]amino]-
hexyl]benzamide (22): From compound 18 (37%); mp: 159–
1
1618C; H NMR (300 MHz, CDCl3): d=7.66 (s, 3H, galloyl-OH), 6.95
(s, 2H, galloyl-H), 6.47 (s, 1H, CONH), 5.94 (s, 1H, CONH), 4.70 (s,
1H, =CH), 4.58 (s, 1H, =CH), 3.46 (m, 1H, H-3), 3.33 (m, 2H, CONH-
CH2), 3.20–3.16 (m, 2H, CONH-CH2), 2.03 (s, 3H, C3-COOCH3), 1.67
(s, 3H, H-30), 0.95 (s, 3H, CH3), 0.90 (s, 3H, CH3), 0.81 (s, 9H, CH3),
0.78 ppm (br d, J=9.0 Hz, 1H, H-5); IR (KBr): n˜ =3352, 2948, 1736,
1715, 1658, 1525, 1447, 1378, 1268, 1199, 1110, 1026, 979, 871,
758 cmꢀ1; HRMS (ESI): m/z [M+H]+ calcd for C45H69N2O7: 749.5104,
found: 749.5094.
3,4,5-Trihydroxy-N-[7-[(3b-hydroxylupan-28-oyl)amino]heptyl]-
benzamide (17): From compound 13 (94%); mp: 193–1958C;
1H NMR (300 MHz, [D6]DMSO): d=8.96 (s, 2H, galloyl-C3,5-OH),
8.27 (s, 1H, galloyl-C4-OH), 7.96 (t, J=5.4 Hz, 1H, CONH), 7.41 (t,
J=5.4 Hz, 1H, CONH), 6.77 (s, 2H, ArH), 3.16–3.10 (m, 2H, CONH-
CH2), 3.05–3.03 (m, 1H, H-3), 2.98–2.86 (m, 2H, CONH-CH2), 1.89 (s,
1H, C3-OH), 0.87 (s, 3H, CH3), 0.85 (s, 3H, CH3), 0.83 (s, 3H, CH3),
0.79 (d, J=6.6 Hz, 1H, H-29 or H-30), 0.76 (s, 3H, CH3), 0.69 (d, J=
6.6 Hz, 1H, H-29 or H-30), 0.63 (s, 3H, CH3), 0.59 ppm (s, 1H, H-5);
IR (KBr): n˜ =3359, 2939, 2867, 1677, 1599, 1524, 1450, 1370, 1311,
1189, 1042, 892, 739 cmꢀ1; HRMS (ESI): m/z [MꢀH]ꢀ calcd for
C44H69N2O6: 721.5156, found: 721.5171.
3,4,5-Trihydroxy-N-[7-[[3b-acetoxylup-20(29)-en-28-oyl]amino]-
heptyl]benzamide (23): From compound 19 (34%); mp: 164–
1
1668C; H NMR (300 MHz, CDCl3): d=7.67 (s, 3H, galloyl-OH), 6.96
(s, 2H, galloyl-H), 6.46 (s, 1H, CONH), 5.96 (s, 1H, CONH), 4.71 (s,
1H, =CH), 4.59 (s, 1H, =CH), 3.50–3.43 (m, 1H, H-3), 3.35–3.33 (m,
2H, CONH-CH2), 3.19–3.17 (m, 2H, CONH-CH2), 2.04 (s, 3H, C3-
COOCH3), 1.68 (s, 3H, H-30), 0.96 (s, 3H, CH3), 0.91 (s, 3H, CH3), 0.82
(s, 9H, CH3), 0.76 ppm (br d, J=9.0 Hz, 1H, H-5); IR (KBr): n˜ =3361,
2941, 1733, 1714, 1653, 1523, 1444, 1371, 1263, 1198, 1110, 1025,
974, 873, 756 cmꢀ1; HRMS (ESI): m/z [MꢀH]ꢀ calcd for C46H69N2O7:
761.5105, found: 761.5101.
Biological assay
(E)-3-(3,4-Dihydroxyphenyl)-N-[6-[[3b-acetoxylup-20(29)-en-28-
oyl]amino]hexyl] acrylamide (24): From compound 20 (49%); mp:
148–1508C; H NMR (300 MHz, [D6]DMSO): d=9.31 (s, 1H, caffeo-
The inhibitory activity of gallic acid, BA, and all the BA derivatives
against HIV gp41 6-HB formation was evaluated using methods
similar to those described previously.[42–44] Typically, the peptide
N36 was incubated with phosphate-buffered saline (PBS) or BA de-
rivatives at 378C for 30 min, followed by addition of the peptide
C34. The final concentrations of both N36 and C34 were 40 mm in
PBS containing 4% (v/v) dimethyl sulfoxide (DMSO). After incuba-
tion at 378C for another 30 min, 20 mL of sample was injected into
the size-exclusion chromatography column connected to the HPLC
system. PBS (pH 7.2) was used as a mobile phase with a flow rate
of 1.0 mLminꢀ1. The UV absorbance at 280 nm was recorded using
a photodiode array detector. Bio-Rad gel filtration standard (Bio-
Rad, Hercules, CA, USA) consisting of thyroglobulin (670 kDa),
gamma globulin (158 kDa), ovalbumin (44 kDa), myoglobin
(17 kDa) and vitamin B12 (1.35 kDa) was used to determine the
molecular weights of the peptides.
1
yl-OH), 9.07 (s, 1H, caffeoyl-OH), 7.91 (t, J=5.1 Hz, 1H, CONH), 7.51
(t, J=5.1 Hz, 1H, CONH), 7.19 (d, J=15.9 Hz, 1H, caffeoyl-H-8), 6.91
(s, 1H, caffeoyl-H-2), 6.80 (dd, J=8.1, 1.8 Hz, 1H, caffeoyl-H-6), 6.72
(d, J=8.1 Hz, 1H, caffeoyl-H-5), 6.29 (d, J=15.9 Hz, 1H, caffeoyl-H-
7), 4.63 (s, 1H, =CH), 4.51 (s, 1H, =CH), 3.15–3.10 (m, 2H, CONH-
CH2), 3.02 (m, 1H, H-3), 3.00–2.94 (m, 2H, CONH-CH2), 2.05 (s, 3H,
C3-COOCH3), 1.68 (s, 3H, H-30), 0.96 (s, 3H, CH3), 0.92 (s, 3H, CH3),
0.83 (s, 3H, CH3), 0.82 (s, 6H, CH3), 0.82 ppm (s, 1H, H-5); IR (KBr):
n˜ =3363, 2940, 1734, 1716, 1656, 1603, 1515, 1446, 1374, 1266,
1195, 1111, 1027, 978, 881, 812 cmꢀ1; HRMS (ESI): m/z [M+H]+
calcd for C47H71N2O6: 759.5312, found: 759.5323.
(E)-3-(3,4-Dihydroxyphenyl)-N-[7-[[3b-acetoxylup-20(29)-en-28-
oyl]amino]heptyl] acrylamide (25): From compound 21 (51%);
1
mp: 156–1588C; H NMR (300 MHz, [D6]DMSO): d=9.34 (s, 1H, caf-
feoyl-OH), 9.10 (s, 1H, caffeoyl-OH), 7.94 (t, J=5.1 Hz, 1H, CONH),
7.54 (t, J=5.1 Hz, 1H, CONH), 7.22 (d, J=15.9 Hz, 1H, caffeoyl-H-8),
6.94 (s, 1H, caffeoyl-H-2), 6.83 (dd, J=8.1, 1.8 Hz, 1H, caffeoyl-H-6),
6.75 (d, J=8.1 Hz, 1H, caffeoyl-H-5), 6.33 (d, J=15.9 Hz, 1H, caffeo-
yl-H-7), 4.66 (s, 1H, =CH), 4.54 (s, 1H, =CH), 3.16–3.13 (m, 2H,
CONH-CH2), 3.25–3.18 (m, 1H, H-3), 3.16–3.08 (m, 2H, CONH-CH2),
2.04 (s, 3H, C3-COOCH3), 1.68 (s, 3H, H-30), 0.96 (s, 3H, CH3), 0.94
(s, 3H, CH3), 0.83 (s, 3H, CH3), 0.82 (s, 6H, CH3), 0.76 ppm (s, 1H, H-
5); IR (KBr): n˜ =3380, 2942, 1733, 1715, 1635, 1603, 1516, 1446,
1374, 1266, 1195, 1112, 1027, 979, 883, 812 cmꢀ1; HRMS (ESI): m/z
[M+H]+ calcd for C48H73N2O6: 773.5469, found: 773.5464.
Docking simulations
The 6-HB gp41 structure was retrieved from the Protein Data Bank
(PDB ID: 1AIK).[9] The N36 trimeric coiled coil was selected as a re-
ceptor for the docking simulations of inhibitors. AutoDock tools
were used to assign Gasteiger partial charges and merge nonpolar
hydrogen atoms. The grip maps were calculated in a grip box of
60ꢁ70ꢁ110 points with a grid-point spacing of 0.375 ꢂ. Flexible
docking calculations were performed with AutoDock v4.2[41] defin-
ing the side chain of key residue Lys574 as a flexible part where
two sigma bonds adjacent to the ammonium group were rotata-
ble. The docking simulations were performed with the Lamarckian
genetic algorithm for at least 50 docking runs for each ligand,
using a population size of 250 individuals with a maximum
number of energy evaluations at 2.5ꢁ109 and a maximum number
of generations at 2.7ꢁ104. An extended model was used to esti-
General procedure for the synthesis of 16 and 17: A solution of
12 or 13 (0.2 mmol) in CH3OH (20 mL) was hydrogenated over Pd/
C (30% by wt) at RT for 2 h. The crude was filtered to remove the
catalyst, and the filtrate was concentrated in vacuo and then puri-
ChemMedChem 2011, 6, 1654 – 1664
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1663

Z.-H. Jiang et al.
MED
[28] I. Baglin, A. C. Mitaine-Offer, M. Nour, K. Tan, C. Cave, M. A. Lacaille-
Dubois, Mini-Rev. Med. Chem. 2003, 3, 525–539.
[29] I. C. Sun, C. H. Chen, Y. Kashiwada, J. H. Wu, H. K. Wang, K. H. Lee, J.
Med. Chem. 2002, 45, 4271–4275.
mate the energetics of the unbound states of the receptor and the
ligands when evaluating their binding free energies (DG).
[30] M. A. Walker, Drug Discovery Today 2001, 6, 859–861.
[31] Y. Kashiwada, F. Hashimoto, L. M. Cosentino, C. H. Chen, P. E. Garrett,
K. H. Lee, J. Med. Chem. 1996, 39, 1016–1017.
[32] T. Kanamoto, Y. Kashiwada, K. Kanbara, K. Gotoh, M. Yoshimori, T. Goto,
K. Sano, H. Nakashima, Antimicrob. Agents Chemother. 2001, 45, 1225–
1230.
[33] J. Zhou, C. H. Chen, C. Aiken, Retrovirology 2004, 1, 15.
[34] B. Labrosse, O. Pleskoff, N. Sol, C. Jones, Y. Hꢃnin, M. Alizon, J. Virol.
1997, 71, 8230–8236.
[35] S. L. Holz-Smith, I. C. Sun, L. Jin, T. J. Matthews, K. H. Lee, C. H. Chen, An-
timicrob. Agents Chemother. 2001, 45, 60–66.
Acknowledgements
This project was supported by the National Natural Science Foun-
dation of China (30729001 and U0832001) and the Southern
Medical University (Guangzhou, P.R. China).
Keywords: conjugates · drug design · fusion/entry inhibitors ·
HIV-1 gp41 · molecular modeling
[36] S. W. Liu, H. Lu, Q. Zhao, Y. X. He, J. K. Niu, A. K. Debnath, S. G. Wu, S. B.
Jiang, Biochim. Biophys. Acta Gen. Subj. 2005, 1723, 270–281.
[37] S. W. Liu, S. B. Jiang, Z. H. Wu, L. Lv, J. J. Zhang, Z. G. Zhu, S. G. Wu, Life
Sci. 2002, 71, 1779–1791.
[38] L. Lꢄ, S. W. Liu, S. B. Jiang, S. G. Wu, Acta Pharmacol. Sin. 2004, 25, 213–
218.
[39] Y. X. He, S. W. Liu, W. G. Jing, H. Lu, D. M. Cai, D. J. Chin, A. K. Debnath, F.
Kirchhoff, S. B. Jiang, J. Biol. Chem. 2007, 282, 25631–25639.
[40] Y. X. He, S. W. Liu, J. J. Li, H. Lu, Z. Qi, Z. H. Liu, A. K. Debnath, S. B. Jiang,
J. Virol. 2008, 82, 11129–11139.
[41] G. M. Morris, R. Huey, W. Lindstrom, M. F. Sanner, R. K. Belew, D. S. Good-
sell, A. J. Olson, J. Comput. Chem. 2009, 30, 2785–2791.
[42] S. W. Liu, Q. Zhao, S. B. Jiang, Peptides 2003, 24, 1303–1313.
[43] S. W. Liu, W. G. Jing, B. Cheung, H. Lu, J. Sun, X. X. Yan, J. K. Niu, J.
Farmar, S. G. Wu, S. B. Jiang, J. Biol. Chem. 2007, 282, 9612–9620.
[44] B. Liu, R. W. Joseph, B. D. Dorsey, R. A. Schiksnis, K. Northrop, M. Bukh-
tiyarova, E. B. Springman, Bioorg. Med. Chem. Lett. 2009, 19, 5693–5697.
[45] S. W. Liu, G. F. Xiao, Y. B. Chen, Y. X. He, J. K. Niu, C. R. Escalante, H. B.
Xiong, J. Farmar, A. K. Debnath, P. Tien, S. B. Jiang, Lancet 2004, 363,
938–947.
[46] We used size exclusion (SE)-HPLC to study the inhibition of gp41 6-HB
core formation by BA derivatives. This method has the following advan-
tages: i) The gp41 peptides can interact directly with inhibitors in this
assay, avoiding the influence of other complicated factors in cell culture
assay; ii) This method is reliable and practical for activities screening; iii)
With BA as a positive control, the determined IC50 values can make the
structure–activity relationship clearer. As far as we know, this is the first
time that the direct interaction of BA derivatives with gp41 peptides
has been studied by SE-HPLC. The SE-HPLC assay was performed ac-
cording to the literature.[42,43] We used 40 mm peptides (both N36 and
C34), which are higher than the peptides concentrations (only 2 mm) in
other protocols described in the literature.[42,43] The higher peptides
concentrations consequently lead to numerically “higher” IC50 values;
this explained why BA was reported in the literature to have an IC50
value (anti-HIV-1 activity) of 13.0 mm in H9 lymphocyte cells,[31] but ex-
hibits an IC50 value 408.4ꢂ23.4 mm under our assay condition. The
potent HIV-1 entry/fusion inhibitory activity of BA and its derivatives in
cell culture has previously been reported.[23]
[1] C. Finnegan, R. Blumenthal, Infect. Disord.: Drug Targets 2006, 6, 355–
357.
[2] A. K. Debnath, Expert Opin. Invest. Drugs 2006, 15, 465–478.
[3] V. A. Johnson, F. Brun-Vꢃzinet, B. Clotet, B. Conway, R. T. D’Aquila, L. M.
Demeter, D. R. Kuritzkes, D. Pillay, J. M. Schapiro, A. Telenti, D. D. Rich-
man, Top. HIV Med. 2003, 11, 215–221.
[4] D. D. Richman, S. C. Morton, T. Wrin, N. Hellmann, S. Berry, M. F. Shapiro,
S. A. Bozzette, AIDS 2004, 18, 1393–1401.
[5] A. Carr, D. A. Cooper, Lancet 2000, 356, 1423–1430.
[6] A. J. Hessell, E. G. Rakasz, D. M. Tehrani, M. Huber, K. L. Weisgrau, G.
Landucci, D. N. Forthal, W. C. Koff, P. Poignard, D. I. Watkins, D. R.
Burton, J. Virol. 2010, 84, 1302–1313.
[7] I. de La Arada, J. P. Julien, B. G. de La Torre, N. Huarte, D. Andreu, E. F.
Pai, J. L. R. Arrondo, J. L. Nieva, J. Phys. Chem. B 2009, 113, 13626–
13637.
[8] J. Mꢄnch, L. Stꢅndker, K. Adermann, A. Schuz, M. Schindler, R. Chinna-
durai, S. Pçhlmann, C. Chaipan, T. Biet, T. Peters, B. Meyer, D. Wilhelm,
H. Lu, W. G. Jing, S. B. Jiang, W. G. Forssmann, F. Kirchhoff, Cell 2007,
129, 263–275.
[9] D. C. Chan, D. Fass, J. M. Berger, P. S. Kim, Cell 1997, 89, 263–273.
[10] D. C. Chan, C. T. Chutkowski, P. S. Kim, Proc. Natl. Acad. Sci. USA 1998,
95, 15613–15617.
[11] H. Ji, W. Shu, F. T. Burling, S. B. Jiang, M. Lu, J. Virol. 1999, 73, 8578–
8586.
[12] B. D. Welch, A. P. VanDemark, A. Heroux, C. P. Hill, M. S. Kay, Proc. Natl.
Acad. Sci. USA 2007, 104, 16828–16833.
[13] S. B. Jiang, K. Lin, N. Strick, A. R. Neurath, Nature 1993, 365, 113–113.
[14] C. T. Wild, D. C. Shugars, T. K. Greenwell, C. B. Mcdanal, T. J. Matthews,
Proc. Natl. Acad. Sci. USA 1994, 91, 9770–9774.
[15] M. Lu, P. S. Kim, J. Biomol. Struct. Dyn. 1997, 15, 465–471.
[16] J. M. Kilby, J. J. Eron, N. Engl. J. Med. 2003, 348, 2228–2238.
[17] J. P. Lalezari, K. Henry, M. O’Hearn, J. S. Montaner, P. J. Piliero, B. Trottier,
S. Walmsley, C. Cohen, D. R. Kuritzkes, J. J. Eron, Jr., J. Chung, R. DeMasi,
L. Donatacci, C. Drobnes, J. Delehanty, M. Salgo, N. Engl. J. Med. 2003,
348, 2175–2185.
[18] D. C. Chan, P. S. Kim, Cell 1998, 93, 681–684.
[19] S. B. Jiang, A. K. Debnath, Biochem. Biophys. Res. Commun. 2000, 269,
641–646.
[20] S. W. Liu, S. G. Wu, S. B. Jiang, Curr. Pharm. Des. 2007, 13, 143–162.
[21] K. H. Lee, J. Nat. Prod. 2010, 73, 500–516.
[22] J. F. Mayaux, A. Bousseau, R. Pauwels, T. Huet, Y. Hꢃnin, N. Dereu, M.
Evers, F. Soler, C. Poujade, E. De Clercq, J. B. Le Pecq, Proc. Natl. Acad.
Sci. USA 1994, 91, 3564–3568.
[47] M. Evers, C. Poujade, F. Soler, Y. Ribeill, C. James, Y. Lelievre, J. C. Gue-
guen, D. Reisdorf, I. Morize, R. Pauwels, E. De Clercq, Y. Hꢃnin, A. Bous-
seau, J. F. Mayaux, J. B. Le Pecq, N. Dereu, J. Med. Chem. 1996, 39,
1056–1068.
[48] F. Soler, C. Poujade, M. Evers, J. C. Carry, Y. Hꢃnin, A. Bousseau, T. Huet,
R. Pauwels, E. De Clercq, J. F. Mayaux, J. B. Le Pecq, N. Dereu, J. Med.
Chem. 1996, 39, 1069–1083.
[23] C. Aiken, C. H. Chen, Trends Mol. Med. 2005, 11, 31–36.
[24] P. Yogeeswari, D. Sriram, Curr. Med. Chem. 2005, 12, 657–666.
[25] K. D. Qian, D. L. Yu, C. H. Chen, L. Huang, S. L. Morris-Natschke, T. J. Nitz,
K. Salzwedel, M. Reddick, G. P. Allaway, K. H. Lee, J. Med. Chem. 2009,
52, 3248–3258.
[49] L. Huang, P. Ho, K. H. Lee, C. H. Chen, Bioorg. Med. Chem. 2006, 14,
2279–2289.
[26] Z. Dang, W. H. Lai, K. D. Qian, P. Ho, K. H. Lee, C. H. Chen, L. Huang, J.
Med. Chem. 2009, 52, 7887–7891.
Received: March 20, 2011
[27] R. H. Cichewicz, S. A. Kouzi, Med. Res. Rev. 2004, 24, 90–114.
Published online on June 17, 2011
1664
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1654 – 1664