Design, asymmetric synthesis, and evaluation of pseudosymmetric sulfoximine inhibitors against HIV-1 protease

Article abstract of DOI:10.1016/j.bmc.2010.01.020

The HIV-1 protease is a validated drug target for the design of antiretroviral drugs to combat AIDS. We previously established the sulfoximine functionality as a valid transition state mimetic (TSM) in the HIV-1 protease inhibitors (PI) design and have id

Full text of DOI:10.1016/j.bmc.2010.01.020

Bioorganic & Medicinal Chemistry 18 (2010) 2037–2048
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Design, asymmetric synthesis, and evaluation of pseudosymmetric
sulfoximine inhibitors against HIV-1 protease
*
Ding Lu , Yuk Yin Sham, Robert Vince
Center for Drug Design, Academic Health Center, University of Minnesota, 516 Delaware St SE, Minneapolis, MN 55455, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The HIV-1 protease is a validated drug target for the design of antiretroviral drugs to combat AIDS. We
previously established the sulfoximine functionality as a valid transition state mimetic (TSM) in the
HIV-1 protease inhibitors (PI) design and have identified a lead pseudosymmetric compound with nano-
molar enzymatic inhibitory activity. Here, we report the asymmetric synthesis of this compound and its
application in the synthesis of sulfoximine-based peptidomimetic HIV-1 protease inhibitors. Molecular
modeling revealed the potential mode of binding of the sulfoximine inhibitor as a TSM. The predicted
absolute binding free energies suggested similar inhibitory effect as observed in our enzymatic inhibitory
studies.
Received 19 November 2009
Revised 6 January 2010
Accepted 7 January 2010
Available online 21 January 2010
Keywords:
HIV
HIV protease
Sulfoximine inhibitors
Transition state mimetic
Peptidomimetic
Asymmetric synthesis
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
the catalytic D25/D250 residues as a hydrogen bond donor and an
acceptor within the enzymatic active sites.8
Acquired Immunodeficiency Syndrome (AIDS) remains one of
the world’s greatest public health challenges with over 25 million
mortalities estimated since it was first recognized in 1981.1 The
highly active antiretroviral therapy (HAART) used in the treatment
of AIDS utilizes a combination of antiviral drugs with at least two
mechanisms of action to effectively inhibit the replication cycle
of HIV.2 Due to the persistence of HIV in the infected host cell res-
ervoir within the tissues, AIDS remains an incurable disease.3 With
the rapid emergence of drug resistance strains worldwide, the ur-
gency to expand the drug repertoire for maintaining the efficacy of
HAART continues to be a priority in antiviral research.
The HIV protease is a C2 symmetric homodimeric aspartyl pro-
tease4,5 that catalyzes the proteolysis of the polyprotein precursor,
gag-pol, and is responsible for producing functional and structural
proteins required for the assembly and maturation of infectious
HIV.6 The HIV-1 protease has been one of the primary targets for
antiviral drug development.7 Currently, there are nine HIV PI’s ap-
proved by the US Food and Drug Administration (FDA) for AIDS
treatment. These drugs feature a secondary alcohol as the transi-
tion state mimetic (TSM) of the tetrahedral intermediate during
amide bond hydrolysis. This secondary alcohol specifically targets
Previously, we have characterized the sulfoximine moiety as a
valid TSM in HIV-1 PI design and have identified a potent pseudo-
symmetric lead compound 1. This compound exhibits an IC50 of
2.5 nM against the HIV-1 protease and IC50 of 408 nM against the
virus.9 We are continuing our efforts towards the development of
a novel HIV PI featuring the sulfoximine moiety to exploit its dual
functionality as a hydrogen donor and acceptor. We have recently
demonstrated that the non-symmetric sulfoximine-based inhibi-
tors are far less potent than their pseudosymmetric analogues.10
This report presents a novel asymmetric synthetic method specif-
ically developed to construct the pseudosymmetric sulfoximine
0
core inhibitors. Conformationally relaxed analogues at the P2=P2
2 substituents were also synthesized to explore stereochemical
0
preferences of the S2=S2 subsites (Fig. 1). The application of this
methodology to construct sulfoximine-based peptidomimetic
HIV-1 PI’s was also demonstrated. To access the role of the sul-
foxime moiety at the active site of HIV-1 protease, an N-methyl
substituted sulfoxime compound 23 was synthesized to remove
the hydrogen bonding donor ability of 1. The resulting compound
displayed a decreased potency validating the importance of the
dual functionality of sulfoximine both as a hydrogen donor and
an acceptor. A molecular modeling study was carried out to evalu-
ate the mode of binding of the sulfoximine compounds. The ob-
served binding modes further support the rationality of the TSM
design. Additionally, the absolute binding free energies of these
inhibitors were evaluated and were consistent with the observed
enzymatic inhibitory activities.
* Corresponding author. Tel.: +1 612 624 9911; fax: +1 612 625 2633.
E-mail address: vince001@umn.edu (R. Vince).
Present address: Masonic Cancer Center, University of Minnesota, 425 East River
Road, Minneapolis, MN 55455, USA.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.01.020

2038
D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
I50
I50'
A
N
N
H
H
H
O
G48'
N
G48
C
G48'
C
G48'
N
H
S3
S1
P1
S2'
P2'
S4'
P4'
H
H
O
O
O
P3
O
O
O
H
N
H
N
H
H2N
N
O
N
H
N
H
N
H
N
H
C
O
P4
S4
O
P2
S2
O
P1'
O
P3'
S3'
O
O
S1'
O OH
O O
H
N
O O
O O
H
N
C
C
D29
D29
G27 D29
G27'
D29'
D29'
D29'
I50
N
I50'
B
N
H
H
O
S2
S2'
H
H
O
O
N
S
N
O
O
H
H
H
H
O
N
Bz
S1
Bz
H
O
C
O
H
O O
H
N
S1'
O O O O
H
N
O O
C
D29
D29
G27
D29 D29'
G27'
D29' D29'
Figure 1. General hydrogen bonding scheme for HIV-1 protease binding to (A) a putative peptide substrate and (B) the symmetric sulfoximine inhibitor, 1. While HIV
protease is a C2 symmetric homodimeric protein in the apo form, the binding of native peptide substrates involves asymmetric hydrogen bonding in each monomeric unit.30
Symmetric inhibitors involve the same hydrogen bonding network on both monomeric units.
2. Results and discussion
Ph
1'
Ph
1
O
NH
H
N
OH
H
N
HO
S
2.1. Chemistry
2
2'
3
3'
O
O
The identification of the sulfoximine-based compound 1 (2S,20S)
as a potent HIV-1 PI has prompted us to further explore the poten-
tial of this new molecular scaffold in the design of HIV PIs. The pre-
vious racemic synthesis generates several diastereomers for the
key sulfide precursor and requires excessive HPLC separations.9
Therefore, sufficient supply of the key intermediate requires a
more scalable synthetic route. We envision that an asymmetric
synthesis with enantiomerically enriched early intermediates
would significantly alleviate the complication of generating and
separating diastereomeric mixtures, hence, greatly increase the
overall synthetic efficiency.
1
Ph
Ph
OH
OH
H
N
H
N
LG
RS
+
O
O
LG = leaving group
The symmetry of the target sulfoximine molecule allows the
adoption of a divergent–convergent synthetic strategy, which re-
lies on an enzyme catalyzed desymmetrization reaction of prochi-
ral diols to construct key chiral building blocks (Scheme 1).
Creating stereogenic carbon centers with high enantioselectivity
has proven valuable in the study of reaction mechanisms and in
the practical syntheses of optically active compounds. Specifically,
the lipase catalyzed enantiotopic differentiation of two primary
hydroxyl groups of a prochiral diol molecule has been shown to
be an efficient method for the preparation of important chiral
building blocks with defined stereochemistry.11 Thus, we antici-
pated that the lipase catalyzed desymmetrization reaction would
provide an efficient synthetic approach to optically pure interme-
diate 2.
R = Ac, H
Ph
Ph
OAc
Ph
HO
OH
HO
HO
OH
O
2
Scheme 1. Retrosynthetic strategy for pseudosymmetric lead compound.
catalyzed desymmetrization reaction was carried out by treating
diol 4 with lipase PS (50% wt/wt) in the presence of excess vinyl
acetate as an acyl donor and diisopropyl ether and water
(1000:1) as a mixed solvent (Scheme 2). This reaction successfully
produced optically active monoester (R)-(+)-5 in excellent yield.
The execution of this strategy started with diethyl benzylmalo-
nate 3 which was reduced to 2-benzyl-1,3-propanediol 4 in high
yield with lithium aluminum hydride (LAH) in ether. A lipase PS

D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
2039
Ph
O
Ph
Ph
b
a
OEt
EtO
OH
HO
OAc
HO
O
5
4
3
8
c
Ph
Ph
Ph
d
e
OH
OH
BnO
BnO
OAc
BnO
O
7
6
Scheme 2. Reagents and conditions: (a) LAH in ether, quantitative; (b) lipase PS, vinyl acetate, 92%; (c) benzyl 2,2,2-trichloroacetimidate, CF3SO3H (cat.), 70%; (d) K2CO3,
MeOH, 87%; (e) Jones reagent, 99%.
The assignment of (R)-configuration to the predominant isomer
was based on literature precedent12 and later confirmed by NMR
and single crystal X-ray analysis of its derivatives.9 The enantio-
meric composition of 5 was measured by NMR analysis of the cor-
responding (+)-a-methoxy-a-trifluoromethyl-a-phenyl acetate
(Mosher’s ester) and the enantiomeric ratio was determined to
Coupling of 8 with (1S,2R)-(ꢀ)-cis-1-amino-2-indanol in the
presence of BOP reagent afforded amide (2R)-9 in good yield. A
small amount of (2S)-9 generated from (S)-8 was isolated by flash
chromatography. Therefore, (1S,2R)-(ꢀ)-cis-1-amino-2-indanol
also functions as a chiral resolving agent, which leads to an enan-
tiomeric enrichment of the starting material. We planned to carry
out a coupling reaction between a nucleophile and an electrophile
to construct the C–S bond in the target skeleton (Scheme 1), and
both parts could be derived from 9 after unmasking the benzyl
ether functional group. Consequently, the hydroxyl group of the
amino indanol moiety was protected to avoid potential interfer-
ence with the new primary hydroxyl group formed from debenzy-
lation at the next step. Therefore, 9 was reacted with excess 2,2-
dimethoxypropane using camphorsulfonic acid (CSA) as a catalyst
to afford acetonide 10, which was then subjected to hydrogenoly-
sis to remove benzyl protection and provide intermediate 11 with
a free hydroxyl functionality for the subsequent coupling reactions.
Compound 11 served as the precursor for the bromide compound
12, which is an important intermediate in our divergent–conver-
gent synthetic scheme. The hydroxyl group in 11 was successfully
converted to bromide 12 with N-bromosuccinimide (NBS) and tri-
phenylphosphine (PPh3) in good yield. The free hydroxyl group in
11 was converted to thioacetate in 13 via a Mitsunobu reaction
(Scheme 3).
be 96:4.
This lipase PS catalyzed monoacylation of prochiral diol was
found to be applicable to a variety of diols with both electron with-
drawing and electron donating groups tolerated on the aromatic
ring.13 Thus, this strategy could be applied to control electron den-
0
sity of P1=P1 substituents of the designed HIV-1 PIs. It is noticed
that chemical manipulation of a monoacetate such as 5 may be
complicated by the vulnerability of the acyl moiety to migrate to
the vicinal hydroxyl group under basic conditions.14 Therefore,
typical alkaline benzylation conditions (benzyl bromide, sodium
hydride in DMF) must be avoided. A solution was to use the
acid-catalyzed conditions15 in which the enantiomerically en-
riched monoester (R)-(+)-5 was treated with the commercially
available benzyl 2,2,2-trichloroacetimidate in the presence of a
catalytic amount of trifluoromethanesulfonic acid (CF3SO3H) to
furnish the benzyl ether 6 in 70% yield. The enantiomeric purity
was subsequently checked by derivatizing with Mosher’s reagent
and no racemization was observed. The acetate group in 6 was re-
moved by treatment with K2CO3 in MeOH16 to afford intermediate
7. The primary alcohol in 7 was subsequently oxidized by Jones’ re-
agent17 to afford the carboxylic acid compound 8 with the desired
stereochemistry (Scheme 2).
A phase-transfer catalysis (PTC) was applied to construct the
C–S coupling product. This strategy was successfully demonstrated
in a stereoselective synthesis of various glycosyl derivatives,
including S-aryl glycosides, from glycosyl halides with complete
Ph
O
Ph
O
Ph
O
OH
a
H
N
OH
BnO
BnO
N
S
O
O
9
8
13
b
e
Ph
O
Ph
Ph
O
O
O
N
BnO
O
d
N
HO
c
N
Br
O
10
11
12
Scheme 3. Reagents and conditions: (a) (1S,2R)-(ꢀ)-cis-1-amino-2-indanol, BOP reagents, TEA, 94%; (b) DMP, CSA, 92%; (c) H2, Pd/C, 88%; (d) NBS, PPh3, 88%; (e) DIAD, PPh3,
84%.

2040
D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
anomeric inversion18 (Scheme 4). The acetate group in 13 was re-
moved by NaOH under dithiothreitol (DTT) protection to afford the
free thiol compound, which was directly coupled with 12 in a two-
phase solvent system of ethyl acetate and saturated aqueous NaH-
CO3 solution in the presence of tetrabutylammonium hydrogen
sulfate (TBAHS) as the phase-transfer catalyst. The desired sym-
metric sulfide 14 was obtained in 70% yield. No racemization of
the product or elimination of the starting materials was observed.
Having obtained intermediate 14 in sufficient quantity, we
proceeded to synthesize the sulfoxide derivative 15 via direct
meta-chloroperoxybenzoic acid (mCPBA) oxidation. The subse-
quent deprotection of isopropylidene in 15 using para-toluenesul-
fonic acid (PTSA) as catalyst in methanol provided optically pure
intermediate 16, which was finally converted to the target sulfox-
imine derivative 1 by MSH (Scheme 4).
It has been shown that conformationally constrained P2=P2
0
substituents, such as (1S,2R)-(ꢀ)-cis-1-amino-2-indanol moiety
0
18, effectively occupy the lipophilic S2=S2 pockets of the HIV-1
protease and also form hydrogen bonds with D29/D290 in the back-
bone of the enzyme.19 This observation has led to the discovery of
numerous highly potent inhibitors, such as Merck L700,41720 and
the subsequent clinically approved drug, Indinavir. However, the
amino indanol group is susceptible to CYP450 metabolism,21 which
limits its oral bioavailability. To find a replacement for the meta-
bolically labile (aromatic and benzylic CYP450 oxidation) amino
indanol moiety and to investigate contributions of binding affinity
from hydrogen bond interactions with the HIV-1 protease, we
incorporated more conformationally relaxed amino indanol ana-
logs such as 19 and 20 into our pseudosymmetric sulfide core
structure 17. The
L
-valine N-methylamide has been previously re-
0
The development of an asymmetric synthesis method to make
stereochemically controlled core structure prompted us to investi-
gate its application. Previous results revealed the importance of the
sulfoximine functional group and the stereochemical preference of
ported to improve ligand binding to the S2=S2 pockets of the
HIV-1 protease.22 It has been featured in various peptidomimetic
HIV-1 PI’s with different TSMs to maintain a high degree of enzy-
matic inhibition activities.23 Therefore, we incorporated the
L-va-
the P1/P1 substituents in our HIV-1 PI design.9 Therefore, the sul-
line N-methylamide substituent into the pseudosymmetric
sulfide core structure 17 to construct the corresponding peptidom-
imetic inhibitor 21.
It is reasonable to believe that a sulfoximine group may func-
tion as a TSM, which presumably forms hydrogen bonds with the
two aspartic acid residues in the active site. To further demonstrate
the importance of the sulfoximine core hydrogen bonding ability,
0
foximine group and the (2S,20S)-benzyl substituents were retained
as a core structure. Based on the retrosynthetic analysis shown in
Scheme 5, the key intermediate is optically pure (2S,20S)-diacid
17. Compound 17 was subsequently obtained by refluxing diamide
intermediate 14 in 4 N HCl, and verified by NMR as a single
diastereomer.
Ph
O
Ph
O
Ph
O
Ph
O
Ph
O
O
S
O
O
O
O
O
AcS
N
a
N
S
N
b
N
N
13
14
15
c
Ph
O
Ph
Ph
Ph
O
O
HO
OH
O
HO
H
H
OH
H
H
N
S
N
N
S
N
d
NH
O
O
16
1
Scheme 4. Reagents and conditions: (a) (i) NaOH, DTT, MeOH/H2O; (ii) 12, NaHCO3, TBAHS, ethyl acetate/H2O, 70%; (b) mCPBA, 91%; (c) PTSA, 97%; (d) MSH, 60%.
P1'
P1
O
HO
S
OH
1
1'
O
RHN
P2
3 S
NHR
P2'
RHN
S
NHR
3'
2
2'
NH
17
O
O
O
O
O
optically pure diacid
symmetric amide
pseudosynmmetric sulfoximine
OH
H2N
OH
O
H2N
H2N
OH
NH2
N
H
R
=
21
,
,
,
18
20
19
Scheme 5. Retrosynthetic analysis of pseudosymmetric sulfoximine compounds and selections of P2/P20 substitution.

D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
2041
the N-methylated sulfoximine analogue 23 was synthesized and
evaluated to determine the biological effect of this chemical
modification.
inhibit the spread of viral infection in a cell based assay. The results
are presented as percent inhibition against the HIV-1 protease/
HIV-1 at 10 lM concentration. For those compounds that exhibited
Several methods have been developed for the preparation of N-
substituted sulfoximine from NH-sulfoximines, such as Eschweil-
er–Clark conditions24 and Meerwein’s salt (trimethyloxonium tet-
rafluroborate)25 enforced N-methylation, and base-promoted N-
alkylation.26 We chose the powerful Meerwein’s salt as the methy-
lating agent in our synthesis because it allows methylation of sen-
sitive or weakly nucleophilic functional groups under very mild
reaction conditions. However, direct N-methylation of 1 proved
to be challenging in the presence of nearby hydroxyl groups.
Therefore the amino indanol groups were protected as acetonides
to afford 22 (Scheme 6). On the other hand, it was observed that
direct imination of 15 by MSH failed to provide compound 22. In-
stead, deprotection of acetonide was observed, which is presum-
ably due to the mesitylsulfonic acid produced in the reaction
media. Non-nucleophilic base of N,N-diisopropyl-ethylamine (DI-
PEA)27 was employed in the N-methylation reaction using the
Meerwein’s salt to prevent the deprotection of the acid-labile ace-
tonide groups. As expected, the crude reaction mixture of the N-
methylation reaction was directly reacted with PTSA in methanol
to remove the isopropylidene and afforded the desired product
23 (Scheme 6).
significant inhibitory effects, IC50 values were also obtained.
The amino indanol residue in 1 was replaced with phenylglyci-
nols, which can be considered as acyclic unconstrained versions of
amino indanol, to afford a series of HIV-1 PIs in the sulfide, sulfox-
ide, and sulfoximine forms. With the intention to evaluate the ste-
reochemical preference of the hydroxyl amine, S- and R-
phenylglycinol derivatives were prepared. All compounds show
moderate activity against HIV-1 protease (Table 1), albeit less po-
tent than (1S,2R)-(ꢀ)-cis-1-amino-2-indanol substituted lead com-
pound 1. Apparently, the conformational constraint inherent in the
amino indanol system did not allow placement of even smaller,
flexible substituents at maximum binding position. The observed
trend of enzymatic activity of sulfoximine > sulfoxide > sulfide is
in agreement with previous conclusions. The stereochemical pref-
erence of phenylglycinol is S, as in the most potent compound 31,
which is similar to that of (1S,2R)-(ꢀ)-cis-1-amino-2-indanol group
in 1.
The antiviral activities of these compounds were evaluated in a
cell based assay. Surprisingly, these compounds with opposite con-
0
figurationally stereogenic centers at the P2=P2 sites are not very
active against virus and exhibit marginal differences in potency.
Thus, we conclude that the stereochemistry of phenylglycinols ap-
pears to have little effect on antiviral potency. It is observed that
sulfide compounds 24 and 25 have the best antiviral activity in
the two series suggesting that sulfide compounds have better cell
permeability than the sulfoxide compounds.28
2.2. Biological activities
The enzymatic inhibitory effects of synthesized compounds
were evaluated with purified HIV-1 protease. Enzymatic activity
was determined based on substrate conversion quantified by a
HPLC method. All compounds were also tested for their ability to
Replacement of amino indanol with
L-valine-N-methylamide
resulted in peptidomimetic analogues 26, 29, and 32 displaying
Ph
O
Ph
O
Ph
O
Ph
O
Ph
Ph
O
O
S
N
O
S
NH
O
S
NH
OH
OH
O
HO
HO
H
N
H
N
H
N
H
N
N
N
b
a
O
O
22
23
1
Scheme 6. Reagents and conditions: (a) 2,2-dimethoxypropane, CSA, 76%; (b) (i) Me3ꢁBF3, DIPEA; (ii) PTSA, MeOH, 56% in two steps.
Table 1
Biological activities against the HIV-1 protease and virus
Compound Structure
% Inhibition @ 10
against enzyme
lM
% Inhibition @
10 M against virus
Compound structure
% Inhibition @ 10
against enzyme
lM
% Inhibition @
10 lM against virus
l
Ph
Ph
Ph
Ph
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
N
H
OH
OH
OH
H
N
H
S
N
S
N
51
50
54
15.4
2.5
0
41
47
87
1.4
O
O
O
O
24
25
Ph
Ph
Ph
Ph
O
S
O
S
H
H
H
H
N
N
N
N
0
0
O
O
O
O
27
28
Ph
Ph
Ph
Ph
O
S
NH
O
S
NH
H
H
H
H
N
N
N
N
O
O
O
O
30
31

2042
D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
Table 3
very potent activity in the enzymatic assay, with the sulfoximine
derivative 32 exhibiting the highest enzyme activity (Table 2). This
suggests that pseudosymmetric sulfoximine is an efficient core
structure with higher activity than sulfide and sulfoxide cores.
Biological activity of lead compounds
Compound structure
IC50 against
IC50 against
enzyme (nM)
virus (lM)
However, it is noticed that 32, with
L
-valine N-methylamide as
Ph
O
Ph
0
O
S
NH
OH
OH
OH
OH
P2=P2 substitution, exhibits lower antiviral activity than 1, which
has an amino indanol group in the same position. The difference
in antiviral activities may be attributed to a comparably lower cell
permeability of 32 and higher peptide character, that is, four amide
bonds rather than two.
H
N
H
N
2.5
0.41
O
O
1
Ph
Ph
From the preliminary biological activity data of the new com-
pounds, we determined that incorporating the selected P2=P2 sub-
O
S
N
H
0
H
N
N
stituents into the central sulfoximine core structure resulted in
compounds with moderate to good enzymatic inhibitory activity.
However, this modification failed to improve enzymatic binding
460
>10
O
23
0
affinity. Apparently, different P2=P2 substituents are well tolerated
in the enzymatic binding pockets. It is also concluded that the
0
binding of phenyl residues at S1=S1 pockets and their stereochem-
0
sisting of the sidechains of L23, V80, P81, V82, and I84. The S2=S2
ical orientation within the enzyme binding sites are critical to the
enzymatic inhibition as the relative potencies of our newly synthe-
sites are mostly hydrophobic, consisting of the sidechains of A28,
V32, I47, I50, L76 as well as residues 28–32 of the opposing chain.
A conserved crystallographic water, which mediates ligand bind-
ing, is commonly found hydrogen bonded to the backbone amide
hydrogen of I50/I500 of the ‘flap’ regions at residues 45–55.
Replacement of this crystallographic water has been shown previ-
ously to enhance inhibition potency.31 Compounds which also take
advantage of the C2 symmetry of the homodimeric protein have
also been previously reported.32 All of the published X-ray struc-
tures of HIV-1 protease in complex with each of the known FDA ap-
proved HIV PI consist of the hydroxyl group positioned within
hydrogen bonding distance between the catalytic D25/D250 car-
boxyl groups.
0
sized compounds differ slightly for different P2=P2 substituents.
Nevertheless, the activities of these compounds are clearly depen-
dent on the sulfoximine-centered core structure.
By introducing the N-methyl group into the sulfoximine struc-
ture, we observed decreased enzymatic and antiviral activities
compared with the parent compound. N-Methylsulfoximine com-
pound 23 exhibits an IC50 value of 0.46 lM against the HIV-1 pro-
tease, which is about 180-fold less potent than the sulfoximine
compound 1 (IC50 of 2.5 nM). Additionally, this compound exhibits
poor activity against virus (Table 3). These results reveal that intro-
ducing an N-methyl group into the lead compound significantly re-
duces the hydrogen bonding ability of the sulfoximine group.
In the final docked models of 1, 30, 31, and 32, several con-
served intermolecular hydrogen bonds were observed. The main
chain amide oxygens were found to hydrogen bond to the con-
served crystallographic water (Fig. 2) while its amide donated
hydrogen to the carbonyl group of G27/G270 backbone. The hydro-
xyl groups of 1 and 31 acted as hydrogen acceptors to the D29/D290
amide hydrogens as well as hydrogen donors to the charged car-
2.3. Computational evaluation
The substrate binding site is characterized by the presence of
the catalytic D25/D250 residues with extended Sn/Sn binding sub-
0
0
sites for each of the corresponding Pn/Pn substituents on the ligand
(Schechter and Berger notation29) (Fig. 1). While binding of the
natural peptide substrates involves a network of asymmetric
hydrogen bonding on each of the two monomeric units,30 symmet-
ric inhibitor binding involves the same hydrogen bonding network
0
boxylate oxygens of D29/D290. The S1=S1 pockets were occupied
0
by the P1=P1 benzyl substituents, forming favorable hydrophobic
interactions. The hydrophobic rings of the indanol groups occupied
the S2=S20 binding sites. The central sulfoximine moiety of 1, 30, 31,
and 32 were found to be within the hydrogen bonding distances of
both D25/D250 carboxylate sidechain, similar to that conformation
found in the binding of the parent compound, L700,417. Previous
studies have demonstrated that the pKa in one of the two aspar-
tates is perturbed when the other is ionized.33 As a result, the cen-
tral sulfoximine group of the inhibitor, when hydrogen bonded to
the two catalytic aspartic groups, simultaneously functioned as a
hydrogen donor to the charged D25 and as a hydrogen acceptor
to the protonated D250. The replacement of the central hydroxyl
group to the sulfoximine moiety in L700,417 has minimum effect
on the mode of binding as observed in 1. With the exception of
the central sulfoximine, the hydrogen bonding network of the
0
on both monomeric units. The nearest S1=S1 subsites from the cat-
alytic D25/D250 residues are considered hydrophobic subsites con-
Table 2
Biological data of peptidomimetic inhibitors
Compound structure
% Inhibition @
10 M against
% Inhibition @
10 M against
virus
l
l
enzyme
Ph
Ph
O
H
O
H
N
S
N
87
12.6
N
H
N
H
O
O
26
0
main chain and P2=P2 substituents for all four sulfoximine inhibi-
Ph
Ph
tors to each of the HIV-1 protease were symmetric in each of the
two HIV protease monomeric units. Compound 30 showed a rela-
tive slight decrease in enzymatic inhibitory activity. While the
change in the pseudosymmetric chiral center did not affect the
O
O
O
S
O
O
H
H
N
N
90
99
2.1
5.5
N
H
N
H
O
O
29
0
placement of the indanol group into the S2=S2 pockets, it resulted
Ph
Ph
O
S
in the indanol hydroxyl bonding with the backbone carbonyl oxy-
gen of G48/G480. Overall, the incorporation of the more conforma-
tionally flexible indanol analogues did not have a major effect in
the overall binding affinity relative to compound 1. Replacement
H
H
N
N
N
H
N
H
NH
O
O
32
of the indanol group with the L-valine-N-methylamide group in

D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
2043
Figure 2. Binding of compounds 1, 30, 31, and 32 (a–d) to HIV-1 protease. Chains A and B are shown in ribbons and are color in pink and green, respectively. The consistently
observed modes of binding support the validity of the sulfoximine moiety as a TSM, participating both as a hydrogen donor and an acceptor. The incorporation of the less
restrained stereospecific P2/P20 substituents resulted in a change in the hydrogen bonding network from the carboxylate sidechain of D29/D290 for compound 31 to the
backbone carbonyl oxygen of G48/G480 for 30.
the pseudosymmetric peptidomimetic compound 32 resulted in
gies. Both compounds were able to bind in a similar mode of bind-
ing within the HIV protease binding site by re-establishing
alternative hydrogen bonding to D29/D290 and G48/G480, respec-
tively (Fig. 2), resulting in a relative difference of 0.4 kcal/mol
change in their binding free energy. The most potent sulfoxime
inhibitor observed by the enzymatic inhibition study was the
pseudosymmetric peptidomimetic compound 32, which was eval-
uated to have the lowest binding energy at ꢀ12.5 kcal/mol, similar
to that of L700,417.
0
the placement of the isopropyl group in the P2=P2 subsites as well
0
as the N-methyl groups into the P3/P3 subsites. The indanol hydro-
xyl groups that hydrogen bonded to D29/D290 carboxylates in 1
were replaced by the terminal amides in 32, which bonded directly
to D29/D290 backbone amide oxygen and the G48/G480 amide
hydrogen.
The calculated binding free energy of all the sulfoximine inhib-
itors are shown (see Table 4). As expected, the calculated binding
free energies of all synthesized sulfoximine inhibitors were higher
than all of the FDA approved inhibitors, as observed in the lower
inhibitory enzymatic activities. The replacement of the hydroxyl
in L700,419 to the sulfoximine group resulted in a net loss of
0.9 kcal/mol in binding free energy in 1. The major contribution
to the change was due to the overall desolvation effect during
binding with an insufficient compensation by the protein-ligand
coulombic energy in the bound state. The opening of the five-
member ring in the indanol resulted in the chiral species of 30
and 31 did not entropically affect their overall binding free ener-
3. Conclusions
HIV-1 PIs based on the pseudosymmetric sulfoximine-centered
core structure were synthesized. All prepared compounds exhibit
micromolar to nanomolar enzymatic inhibitory activities; how-
ever, most of the compounds show moderate or poor antiviral
activity against the HIV-1 strain in cells. Nevertheless, it is ob-
served that enzymatic activity is marginally affected by changing
0
the P2=P2 substituents, which argues for the importance of the
central sulfoximine core structure as well as the preferred stereo-
0
chemistry of the P1=P1 sites.
Table 4
Pseudosymmetric sulfoximine core structure was identified as a
valid TSM in the design of potent anti-HIV drugs with novel struc-
tures and properties. The asymmetric synthesis we developed will
allow the design of structurally diversified inhibitors featuring the
stereochemically controlled sulfoximine functionality. The diver-
gent–convergent synthesis of 1 could be readily modified to gener-
ate analogs with different P1=P1 and P2=P2 substituents which
may better match the enzyme’s binding pocket and lead to im-
proved antiviral activity.
The sulfoximine group may function as a TSM by providing NH
and O as potential hydrogen bond donor and acceptor to the cata-
lytic aspartates. N-Methylation of the sulfoximine derivative
exhibits decreased potency against both the enzyme and virus.
Docking studies further support the sulfoximine as a TSM and pro-
Calculated absolute binding free energies are in kcal/mol
exp
f!b
f !b
f!b
calc
Compound
D
G
bind
hDU
iq
DG
bind
hD
U
iq
hDU
iq
cav
elec
vdw
SQR
IDV
RTV
APV
LPV
NFV
DRV
L700,417
1
ꢀ11.9
ꢀ32.1
ꢀ38.5
ꢀ29.6
ꢀ43.6
ꢀ31.9
ꢀ32.8
ꢀ20.4
ꢀ13.4
ꢀ13.7
ꢀ19.7
ꢀ14.1
ꢀ82.4
ꢀ81.2
ꢀ82.9
ꢀ64.1
ꢀ76.7
ꢀ72.6
ꢀ66.2
ꢀ81.1
ꢀ91.9
ꢀ86.1
ꢀ82.4
ꢀ77.5
11.1
8.4
10.7
5.5
7.5
7.7
6.5
5.7
6.8
5.9
4.8
9.1
ꢀ13.0
ꢀ12.4
ꢀ13.7
ꢀ13.2
ꢀ15.1
ꢀ12.8
ꢀ15.2
ꢀ12.7
N/A
ꢀ12.5
ꢀ13.4
ꢀ13.9
ꢀ13.9
ꢀ14.1
ꢀ13.8
ꢀ14.1
ꢀ12.4
ꢀ11.5
ꢀ11.8
ꢀ12.2
ꢀ12.5
0
0
30
31
32
N/A
N/A
N/A

2044
D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
vide insight into the symmetric hydrogen bonding network during
binding. Further confirmation of this hypothesis by X-ray crystal-
lography study is currently underway.
room temperature for 3 h. The reaction mixture was then filtered
through Celite. After evaporation of methanol, the remaining solid
was taken up in H2O, extracted with DCM, washed with water and
brine. The organic layer was dried over Na2SO4, evaporated and
purified by flash chromatography on silica gel using 15% ethyl ace-
tate in hexane. Compound 7 was isolated as an oil (0.148 g, 87%).
1H NMR (CDCl3, 600 MHz) d 7.35–7.13 (m, 10H), 4.34–4.29 (m,
2H), 3.70 (dd, J = 4.2, 11.4 Hz, 1H), 3.61 (dd, J = 6.6, 11.4 Hz, 1H),
3.56 (dd, J = 4.2, 9 Hz, 1H), 3.46 (dd, J = 6.6, 9 Hz, 1H), 2.68–2.57
(m, 3H), 2.15–2.09 (m, 1H); 13C NMR (CDCl3, 75 MHz) d 140.3,
138.2, 129.3, 128.7, 128.0, 127.9, 126.3, 73.8, 72.9, 65.4, 42.9,
4. Experimental
4.1. Chemistry
Lipase PS and chemicals were purchased from Sigma Chemicals
(St. Louis, MO). All commercial reagents were used as provided un-
less otherwise indicated. An anhydrous solvent dispensing system
(J. C. Meyer) using two columns packed with neutral alumina was
utilized for drying THF, Et2O and CH2Cl2, while two columns
packed with molecular sieves were used to dry DMF. For all com-
pounds, 1H NMR and 13C NMR spectra were acquired using a Mer-
cury 300 or Varian 600 spectrometer in deuterium solvent with
Me4Si as internal standard unless otherwise stated. Optical rota-
tions were recorded on Rudolph autopol III polarimeter. Mass spec-
tra were recorded using Bruker BioTOF II or Agilent LC-TOF 1100
mass spectrometers equipped with either an ESI or APCI source.
All melting points were uncorrected and obtained from Mel-Temp
II melting point apparatus. All moisture and/or air sensitive exper-
iments were performed under a positive pressure of nitrogen or ar-
gon atmosphere in oven dried glassware. Solvent and liquid
reagents were transferred by a syringe or cannula. Solvents were
removed in vacuum using a rotary evaporator equipped with a
condenser. Yields were not optimized.
34.8; ½a 2D0
ꢀ34.0° (c 1.0 CHCl3); ESI MS C17H20O2 256.1, found
ꢂ
279.1 (M+Na)+.
4.1.4. 2R-Benzyl 3-benzyloxy propanoic acid (8)
Compound 7 (0.049 g, 0.19 mmol) was dissolved in 2 mL of ace-
tone and cooled to 0 °C. Jones reagent (0.38 mmol, 2.67 M, pre-
pared as follows: mix 2.67 g CrO3 with 23 mL of concentrated
H2SO4, slowly diluted with distilled H2O to a total volume of
10 mL) was added slowly and the mixture was stirred at 0 °C for
2 h. 2-Propanol was added until the solution turned greenish blue
and the solvent was then evaporated. Ethyl acetate and H2O were
added to the residue and the separated organic layer was washed
with brine and concentrated under reduced pressure. Flash chro-
matography eluting with 30% ethyl acetate in hexane afforded
compound 8 as an oil (0.051 g, 99%). 1H NMR (CDCl3, 600 MHz) d
7.34–7.15 (m, 10H), 4.54–4.48 (m, 2H), 3.64–3.58 (m, 2H), 3.05–
2.97 (m, 2H), 2.90–2.86 (m, 1H); 13C NMR (CDCl3, 75 MHz) d
179.4, 138.6, 137.9, 129.2, 128.7, 128.6, 128.4, 127.9, 126.7, 73.5,
4.1.1. 2R-Hydroxymethyl 3-phenylpropyl acetate (5)
A suspension of 4 (1.000 g, 6.0 mmol) and lipase PS (0.500 g) in
vinyl acetate (830 lL) and H2O (30 lL) was stirred at room temper-
69.6, 47.8, 24.6; ½a D20
ꢀ90.0° (c 1.0 CHCl3); ESI MS C17H18O3
ꢂ
270.1, found 293.1 (M+Na)+, 541.1 (2 M+H)+.
ature. After being stirred for 5 h, the mixture was filtered through a
pad of Celite. The filtrate was evaporated and purified by flash
chromatography on silica gel using 30% ethyl acetate in hexane.
Compound 5 was isolated as an oil (1.153 g, 92%). 1H NMR (CDCl3,
600 MHz) d 7.29–7.17 (m, 5H), 4.15 (dd, J = 4.8, 11.4 Hz, 1H), 4.07
(dd, J = 6, 10.8 Hz, 1H), 3.58 (dd, J = 4.8, 11.4 Hz, 1H), 3.50 (dd,
J = 6.6, 11.4 Hz, 1H), 2.68 (dd, J = 7.2, 13.8 Hz, 1H), 2.61 (dd,
J = 7.8, 13.8 Hz, 1H), 2.54 (br, 1H), 2.14–2.09 (m, 1H), 2.05 (s,
3H); 13C NMR (CDCl3, 150 MHz) d 171.2, 138.6, 129.3, 128.6,
4.1.5. N-[(1S,2R)-cis-Aminoindan-2-ol] 2R-benzyl-3-benzyloxy
propionamide (9)
A solution of 8 (0.050 g, 0.19 mmol) and (1S,2R)-(ꢀ)-cis-1-ami-
no-2-indanol (0.033 g, 0.22 mmol) in DCM (3 mL) was cooled in an
ice bath and treated with triethyl amine (51.5 lL, 0.37 mmol) fol-
lowed by adding (benzotrizol-1-yloxy) tris (dimethylamino) phos-
phonium hexafluorophosphate (BOP reagent, 0.098 g, 0.22 mmol).
The mixture was stirred at 0 °C for 1 h and then at room tempera-
ture overnight. The solvent was removed under reduced pressure
and the residue was dissolved in ethyl acetate and washed succes-
sively with 0.5 N HCl, H2O, saturated NaHCO3 and brine. The organ-
ic layer was dried over Na2SO4 and concentrated under reduced
pressure. The crude product was purified by flash chromatography
on silica gel using 30% ethyl acetate in hexane. Compound 9 was
isolated as a white solid (0.070 g, 94%). Mp: 128–130 °C; 1H NMR
(CDCl3, 600 MHz) d 7.32–7.09 (m, 14H), 6.14 (d, J = 7.8 Hz, 1H),
5.32 (dd, J = 4.8, 7.8 Hz, 1H), 4.56–4.51 (m, 2H), 4.35 (br, 1H),
3.77 (t, J = 8.4 Hz, 1H), 3.68–3.66 (m, 1H), 3.07–2.95 (m, 2H),
2.91–2.77 (m, 6H); 13C NMR (CDCl3, 150 MHz) d 174.0, 140.6,
140.4, 139.6, 138.0, 128.8, 128.6, 128.3, 128.0, 127.2, 126.8,
126.4, 63.4, 62.1, 42.3, 34.1, 21.2; ½a D20
ꢂ
+26.3° (c 1.0 CHCl3); ESI
MS C12H16O3 208.1, found 209.1 (M+H)+.
4.1.2. 2R-Phenyl 3-benzyloxy propyl acetate (6)
To a stirred solution of 5 (0.200 g, 0.96 mmol) and benzyl 2,2,2-
trichloroacetimidate (0.290 g, 1.15 mmol) in a mixed solvent of
DCM (3 mL) and cyclohexane (6 mL) was added trifluoromethane-
sulfonic acid (10 lL) under argon. The reaction mixture was stirred
at room temperature overnight. The crystalline trichloroacetamide
was removed by filtration and the filtrate was washed with aque-
ous saturated solution of NaHCO3 (15 mL) and H2O (15 mL). The
organic layer was dried over Na2SO4, evaporated and purified by
flash chromatography on silica gel using 5% ethyl acetate in hex-
ane. Compound 6 was isolated as an oil (0.201 g, 70%). 1H NMR
(CDCl3, 600 MHz) d 7.33–7.13 (m, 10H), 4.49–4.44 (m, 2H), 4.12–
4.06 (m, 2H), 3.43–3.37 (m, 2H), (dd, J = 6.6, 9 Hz, 1H), 2.74 (dd,
J = 7.8, 13.8 Hz, 1H), 2.67 (dd, J = 7.2, 13.8 Hz, 2H), 2.27–2.23 (m,
1H), 1.99 (s, 3H); 13C NMR (CDCl3, 150 MHz) d 171.3, 140.4,
137.5, 129.8, 128.5, 127.8, 126.2, 125.7, 75.2, 72.3, 67.8, 45.5,
125.4, 124.5, 73.7, 73.5, 7.09, 57.7, 50.6, 39.5, 35.6; ½a D20
ꢂ
+29.2° (c
1.0 CHCl3); ESI MS C26H27NO3 401.2, found 424.2 (M+Na)+.
4.1.6. 1-[2,2-Dimethyl-8,8a-dihydro-3aH-indeneo[(1S,2R)-1,2]-
oxazol-3-yl] (2R)-benzyl-3-benzyloxy propanone (10)
To a suspension of 9 (0.062 g, 0.16 mmol) in toluene (4 mL) was
added 2,2-dimethoxypropane (0.59 mL, 6.4 mmol) followed by ( )-
camphor-10-sulfonic acid (0.008 g, 0.03 mmol). The mixture was
stirred at room temperature for 4 h and the reaction was quenched
by adding ethyl acetate and 10% NaHCO3. The organic layer was
separated and dried over Na2SO4 and concentrated under reduced
pressure. Flash chromatography eluting with 15% ethyl acetate in
hexane afforded 10 as a solid (0.063 g, 92%). Mp: 109–110 °C; 1H
NMR (CDCl3, 600 MHz) d 7.38–7.10 (m, 12H), 6.86–6.80 (m, 1H),
40.3, 23.9; ½a 2D0
ꢀ2.6° (c 1.0 CHCl3); ESI MS C19H22O3 298.2, found
ꢂ
321.2 (M+Na)+.
4.1.3. 2R-Phenyl 3-benzyloxy propan-1-ol (7)
To a solution of 6 (0.200 g, 0.67 mmol) in methanol (40 mL) was
added K2CO3 (0.185 g, 1.34 mmol) and the mixture was stirred at

D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
2045
6.12 (d, J = 15 Hz, 1H), 5.54 (d, J = 9 Hz, 1H), 4.51–4.44 (m, 3H),
79.4, 65.8, 48.4, 37.6, 36.4, 33.1, 30.9, 26.7, 24.3; ½a D20
ꢂ
+50.5° (c
3.80–3.66 (m, 2H), 3.52–3.43 (m, 1H), 3.38–3.31 (m, 1H), 3.03–
2.92 (m, 2H), 2.74–2.69 (m, 1H), 1.62 (s, 3H), 1.32 (s, 3H); 13C
NMR (CDCl3, 150 MHz) d 170.7, 141.1, 140.5, 139.9, 138.3, 129.7,
128.9, 128.6, 128.2, 127.9, 127.7, 127.2, 126.9, 125.7, 124.2, 96.9,
1.0 CHCl3); ESI MS C24H27NO3S 409.2, found 410.2 (M+H)+.
4.1.10. Bis-[N-(2,2-dimethyl-8,8a-dihydro-3aH-indeno[(1S,2R)-
1,2]-oxazol-3-yl) (2R)-benzyl-3-yl propianamide] sulfide (14)
DL-Dithiothreitol (0.083 g, 0.54 mmol) was added to a solution
of 13 (1.100 g, 2.6 mmol) in methanol (45 mL). The solution was
sparged with N2 for 15 min, cooled to 0 °C and kept under N2 atmo-
sphere. In a separate flask, 1 N NaOH (10 mL, 10.4 mmol) was
sparged with N2 for 20 min and then added to the reaction mixture
slowly. After 1 h, the ice bath was removed to allow the mixture to
warm to room temperature. The reaction was quenched by adding
saturated NH4Cl and then diluted with CHCl3. The organic layer
was washed by brine, dried over Na2SO4. All the solvent was evap-
orated to afford 1.2 g oil, which was used directly for the next step.
To ethyl acetate (70 mL) and saturated NaHCO3 (60 mL) solution,
which were sparged with N2 for 20 min, of 12 (1.100 g, 2.6 mmol)
and the above crude product was added tetrabutylammonium
hydrogen sulfate (TBAHS, 4.5 g, 5 equiv). The two-phase reaction
mixture was vigorously stirred at room temperature under N2
atmosphere. The reaction was quenched after seven days and the
organic layer was separated, washed with brine and dried over
Na2SO4. The crude product was purified by flash chromatography
on silica gel using 10% ethyl acetate in hexane to afford compound
14 as a white solid (1.200 g, 70%). Mp: 80–81 °C; 1H NMR (CDCl3,
600 MHz) d 7.37–7.01 (m, 14H), 6.92–6.89 (m, 2H), 6.21 (d,
J = 7.8 Hz, 2H), 5.49 (d, J = 4.2 Hz, 2H), 4.87–4.85 (m, 2H), 3.38
(dd, J = 7.8, 12.6 Hz, 2H), 3.28 –3.23 (m, 2H), 3.08–3.02 (m, 4H),
2.97–2.94 (m, 2H), 2.86 (dd, J = 5.4, 13.2 Hz, 2H), 2.72 (dd, J = 6.0,
12.6 Hz, 2H), 1.70 (s, 6H), 1.35 (s, 6H); 13C NMR (CDCl3,
150 MHz) d 170.4, 140.8, 140.6, 139.4, 129.7, 129.0, 128.5, 127.4,
79.1, 73.9, 73.7, 65.8, 48.6, 36.3, 35.3, 26.7, 24.1; ½a D20
ꢂ
+120.2° (c
1.0 CHCl3); ESI MS C29H31NO3 441.2, found 442.2 (M+H)+.
4.1.7. 1-[2,2-Dimethyl-8,8a-dihydro-3aH-indeneo[(1S,2R)-1,2]-
oxazol-3-yl] (2R)-benzyl-3-hydroxyl propanone (11)
A mixture of ethanol (4 mL) solution of 10 (0.062 g, 0.14 mmol)
and Pd/C (10 wt%, 0.012 g) was hydrogenated at room temperature
under hydrogen atmosphere for 12 h. The catalyst was filtered
through Celite, the cake was rinsed with ethanol and ethyl acetate.
The filtrate was evaporated under reduced pressure and the resi-
due was purified by flash chromatography on silica gel using 50%
ethyl acetate in hexane to afford compound 11 as a white solid
(0.043 g, 88%). Mp: 136–138 °C; 1H NMR (CDCl3, 600 MHz) d
7.34–7.16 (m, 7H), 6.90 (t, J = 7.2 Hz, 1H), 6.28 (d, J = 7.8 Hz, 1H),
5.48 (d, J = 4.2 Hz, 1H), 4.85–4.83 (m, 1H), 3.88–3.83 (m, 2H),
3.39–3.16 (m, 2H), 3.14–3.02 (m, 3H), 2.87–2.82 (m, 1H), 1.68 (s,
3H), 1.35 (s, 3H); 13C NMR (CDCl3, 150 MHz) d 171.1, 140.7,
139.7, 129.7, 128.9, 128.9, 128.4, 127.3, 126.9, 125.8, 124.2, 97.1,
79.3, 65.9, 65.2, 49.7, 36.3, 35.0, 26.6, 24.3; ½a D20
ꢀ21.4° (c 1.0
ꢂ
CHCl3); ESI MS C22H25NO3 351.2, found 374.5 (M+Na)+.
4.1.8. 1-[2,2-Dimethyl-8,8a-dihydro-3aH-indeneo[(1S,2R)-1,2]-
oxazol-3-yl] (2R)-benzyl-3-bromo propanone (12)
To a solution of 11 (0.810 g, 2.3 mmol) and triphenylphosphine
(0.725 g, 2.9 mmol) in CH2Cl2 (40 mL) in an ice bath, was added
crystalline N-bromosuccinimide (0.492 g, 2.8 mmol, recrystallized
from hot H2O) slowly. After addition was complete, the mixture
was stirred at room temperature for 12 h. The reaction was
quenched by adding saturated NaHCO3, the organic layer was then
washed by brine, dried over Na2SO4 and concentrated under vac-
uum. The crude product was purified by flash chromatography
on silica gel using 15% ethyl acetate in hexane to afford compound
12 as a white solid (0.844 g, 88%). Mp: 139–141 °C; 1H NMR (CDCl3,
600 MHz) d 7.36–7.18 (m, 7H), 6.92 (t, J = 6.6 Hz, 1H), 6.24 (d,
J = 7.2 Hz, 1H), 5.59 (d, J = 4.8 Hz, 1H), 4.87 (m, 1H), 3.68 (m, 1H),
3.53–3.49 (m, 2H). 3.45–3.41 (m, 1H), 3.08 (m, 2H), 2.89 (dd,
J = 5.4, 13.2 Hz, 1H), 1.72 (s, 3H), 1.37 (s, 3H); 13C NMR (CDCl3,
150 MHz) d 169.4, 140.7, 140.6, 138.5, 139.7, 129.2, 128.5, 127.4,
125.9, 124.1, 97.5, 79.2, 66.4, 49.3, 38.7, 36.3, 34.6, 26.7, 24.1;
125.9, 97.2, 90.3, 66.3, 47.9, 38.6, 37.5, 36.3, 26.7, 24.4; ½a D20
ꢂ
+90.0° (c 1.0 CHCl3); HRMS (ESI) C44H48N2O4S 700.3335, found
723.3235 (M+Na)+ 1.1 ppm error.
4.1.11. Bis-[N-(2,2-dimethyl-8,8a-dihydro-3aH-indeno[(1S,2R)-
1,2]-oxazol-3-yl) (2R)-benzyl-3-yl propianamide] sulfoxide (15)
To a solution of 14 (0.586 g, 0.84 mmol) in DCM (15 mL) at
ꢀ78 °C under N2 was added
a solution of mCPBA (0.187 g,
0.84 mmol) in DCM (3 mL). The mixture was stirred for 1 h then
warmed to room temperature for 1 h. The reaction was quenched
by diluting with DCM and extracted with saturated NaHCO3. The
organic layer was washed by brine, dried over Na2SO4 and concen-
trated under vacuum. The crude product was purified by flash
chromatography on silica gel using 50% ethyl acetate in hexane
to afford compound 15 as a white solid (0.545 g, 91%). Mp: 198–
199 °C; 1H NMR (acetone-d6, 600 MHz) d 7.48–7.09 (m, 14H),
6.85 (t, J = 7.8 Hz, 1H), 6.75 (t, J = 7.2 Hz, 1H), 6.11 (d, J = 7.2 Hz,
1H), 5.94 (d, J = 7.8 Hz, 1H), 5.83 (br, 1H), 5.44 (br, 1H), 4.84–4.79
(m, 2H), 3.63–3.55 (m, 2H), 3.52–3.47 (m, 2H), 3.23–2.88 (m,
10H), 1.64 (s, 3H), 1.59 (s, 3H), 1.26 (s, 3H), 1.21 (s, 3H); 13C
NMR (acetone-d6, 150 MHz) d 169.4, 168.8, 141.2, 141.1, 141.0,
139.9, 139.4, 129.9, 129.9, 129.0, 128.9, 128.1, 128.0, 127.2,
127.1, 127.1, 127.0, 125.6, 125.5, 124.1, 96.6, 96.2, 79.4, 79.3,
66.1, 65.2, 57.4, 56.9, 43.0, 42.9, 38.9, 36.7, 35.9, 26.3, 26.2, 23.5,
½
a 2D0
+100.2° (c 1.0 CHCl3); ESI MS C22H24BrNO2 413.1, found
ꢂ
414.1 (M+H)+.
4.1.9. 1-[2,2-Dimethyl-8,8a-dihydro-3aH-indeneo[(1S,2R)-1,2]-
oxazol-3-yl] (2R)-benzyl-3-acetylthio propanone (13)
A mixture of 11 (0.090 g, 0.26 mmol) and thioacetic acid
(28.6 lL, 0.38 mmol) in THF (1 mL) was added dropwise at
ꢀ10 °C to a stirred suspension of preformed adduct of triphenyl-
phosphine (0.101 g, 0.26 mmol) and diisopropyl azocarboxylate
(97
l
L, 0.38 mmol). The mixture was stirred at ꢀ10 °C for 1 h then
room temperature for 2 h. The reaction was quenched by addition
of ether (5 mL) and washed twice with saturated NaHCO3. The or-
ganic layer was dried over Na2SO4 and concentrated under reduced
pressure. The crude product was purified by flash chromatography
on silica gel using 15% ethyl acetate in hexane. Compound 13 was
isolated as a pale yellow oil (0.087 g, 84%); 1H NMR (CDCl3,
600 MHz) d 7.34–7.11 (m, 6H), 6.74 (t, J = 7.8 Hz, 1H), 5.80 (d,
J = 7.2 Hz, 1H), 5.51 (d, J = 3.6 Hz, 1H), 4.89–4.87 (m, 1H), 3.44
(m, 1H), 3.28 (dd, J = 5.4, 13.2 Hz, 1H), 3.18–3.16 (m, 1H), 3.12–
3.00 (m, 3H), 2.86 (dd, J = 3.6, 13.2 Hz, 1H), 2.34 (s, 3H), 1.67 (s,
3H), 1.31 (s, 3H); 13C NMR (CDCl3, 150 MHz) d 195.8, 169.6,
140.6, 139.7, 129.9, 129.0, 128.3, 127.3, 127.1, 125.7, 124.2, 97.0,
23.3;
½
a 2D0
+103.4° (c 1.0 CHCl3); HRMS (ESI) C44H48N2O5S
ꢂ
716.3284, found 717.3339 (M+H)+ 2.5 ppm error.
4.1.12. Bis-[[N-(1S,2R)-2,3-dihydro-2-hydroxy-1H-inden-1-yl]-
2S-phenyl-3-yl-propionamide] sulfoxide (16)
To a stirred and cooled solution of sulfide 15 (1 equiv) in DCM
(2 mL) and MeOH (1 mL) was added mCPBA (1.1 equiv in 1 mL of
DCM) slowly. The reaction was kept stirring under N2 for 3 h,
and the solution was concentrated and diluted with CHCl3. Organic
layer was washed by saturated NaHCO3, brine and dried over
Na2SO4. The solvent was evaporated to afford 16 as white solids

2046
D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
(29 mg, 97%). Mp: 200–201 °C; 1H NMR (DMSO-d6, 600 MHz) d
8.15 (d, J = 9.0 Hz, 1H), 8.09 (d, J = 9.0 Hz, 1H), 7.28–7.05 (m,
18H), 5.18 (t, J = 4.8 Hz), 5.12 (t, J = 4.8 Hz, 1H), 4.76 (d, J = 3.0 Hz,
1H), 4.60 (d, J = 3.0 Hz, 1H), 4.34 (d, J = 3.6 Hz, 1H), 4.29 (d,
J = 3.6 Hz, 1H), 3.18–2.96 (m, 7H), 2.79–2.66 (m, 5H), 2.41 (d,
J = 10.8 Hz, 1H); 13C NMR (DMSO-d6, 150 MHz) d 173.1, 173.1,
142.7, 142.6, 141.4, 141.4, 139.3, 139.2, 129.9, 129.8, 128.9,
127.9, 127.8, 127.0, 126.9, 126.8, 125.5, 125.3, 125.0, 124.9, 72.9,
4.1.16. N-Methyl bis-[N-(2,2-dimethyl-8,8a-dihydro-3aH-
indeno[(1S,2R)-1,2]-oxazol-3-yl) 2R-benzyl-3-yl propianamide]
sulfoximine (23)
To a solution of 22 (20 mg, 0.027 mmol) in 2 mL of DCM was
added DIPEA (3.5 mg, 0.03 mmol) followed by Meerwein’s salt
(4.0 mg, 0.03 mmol) in an ice bath. The mixture was stirred under
N2 till no starting material left. The reaction was quenched by adding
H2O and MeOH. To this crude reaction mixture was added PTSA
(5 mg, 0.03 mmol) and the reaction was stirred for additional 12 h.
All the solvent was removed under reduced pressure and the residue
was partitioned between CHCl3 and saturated NaHCO3. The com-
bined organic layers were washed by brine and dried over Na2SO4.
The crude product was purified by flash chromatography using 1%
MeOH in CHCl3 to afford 23 as a solid (10 mg, 56%). Mp: 132–
133 °C; 1H NMR (CDCl3/MeOH-d4, 600 MHz) d 7.64 (br, 2H), 7.38–
6.94 (m, 18H), 5.24 (d, J = 4.2 Hz, 1H), 5.22 (d, J = 4.2 Hz, 1H), 4.35
(t, J = 4.8 Hz, 1H), 4.32 (t, J = 4.8 Hz, 1H), 3.89 (dd, J = 10.8, 15 Hz,
1H), 3.69 (dd, J = 10.8, 14.4 Hz, 1H), 3.28–3.22 (m, 2H), 3.08–3.05
(m, 4H), 3.02–2.98 (m, 2H), 2.93–2.85 (m, 2H), 2.82–2.77 (m, 2H),
2.63 (s, 3H); 13C NMR (CDCl3/MeOH-d4, 150 MHz) d 172.0, 171.8,
140.4, 140.4, 136.7, 136.6, 129.4, 129.4, 129.0, 129.0, 128.6, 128.1,
128.0, 127.6, 127.6, 127.0, 126.7, 125.2, 125.1, 124.3, 124.2, 73.0,
57.7, 57.6, 54.6, 54.2, 42.6, 42.2, 39.8, 39.2, 38.6; ½a D20
ꢀ4.4° (c
ꢂ
0.26 MeOH/CHCl3 = 1:1); HRMS (ESI) C38H40N2O5S 636.2658, found
637.2715 (M+H)+, 2.5 ppm error.
4.1.13. Bis-{[N-(1S,2R)-2,3-dihydro-2-hydroxy-1H-inden-1-yl]-
2S-phenyl-3-yl-propionamide} sulfoximine (1)
Sulfoxide 16 (1 equiv) was dissolved in DMF (2.5 mL) in an ice
bath. MSH (4 equiv) was added to the flask all in once. The reac-
tion was stirred at room temperature for 20 h and quenched by
adding pre-cooled NaHCO3 solution and kept stirring for addi-
tional 10 min. The reaction mixture was concentrated and parti-
tioned between CHCl3 and brine. Combined organic layers were
dried over Na2SO4 and purified by HPLC to afford 1 as white sol-
ids (7 mg, 45%). Mp: 208–209 °C; 1H NMR (CDCl3 + MeOH-d4,
300 MHz) d 7.38–6.93 (m, 18H), 5.24 (br, 2H), 4.31 (dd, J = 2.7,
2.7 Hz, 2H), 3.77 (dd, J = 5.4, 5.4 Hz, 2H), 3.33–3.28 (m, 2H),
3.14–2.97 (m, 6H), 2.88–2.80 (m, 4H); 13C NMR (CDCl3 + MeOH-
d4, 75 MHz) d 177.8, 177.7, 144.3, 144.2, 144.1, 141.9, 141.7,
133.3, 133.3, 133.8, 132.7, 131.9, 131.8, 131.1, 130.9, 130.8,
128.9, 128.3, 128.6, 77.1, 77.0, 61.8, 61.7, 60.4, 59.9, 58.7, 47.5,
46.7, 43.3, 43.3, 43.2, 43.0; ESI m/e C38H41N3O5S 651.2767, found
72.8, 57.8, 57.8, 52.7, 51.8, 42.3, 42.3, 39.6, 38.7, 38.7, 26.2; ½a D20
ꢂ
+14.0° (c 0.3 MeOH/CHCl3 = 1:1); HRMS (ESI) C39H44N3O5S
665.2923, found 666.2969 (M+H)+ 4.0 ppm error.
4.2. General procedure for the preparation of diamide sulfide
(24, 25, 26)
652.2849 (M+H)+, 0.61 ppm error; ½a 2D0
ꢂ
= +60.4° (c 0.25 MeOH/
A solution of 17 (1 equiv) and 19, 20, or 21 (3 equiv) in DCM
(4 mL) was cooled in an ice bath and treated with triethylamine
(6 equiv) followed by adding (benzotrizol-1-yloxy) tris (dimethyl-
amino) phosphonium hexafluorophosphate (BOP reagent, 3 equiv).
The mixture was stirred at 0 °C for 1 h and then at room tempera-
ture overnight. The solvent was removed under reduced pressure
and the residue was dissolved in ethyl acetate and washed succes-
sively with 0.5 N HCl, H2O, saturated NaHCO3 and brine. The organ-
ic layer was dried over Na2SO4 and concentrated under reduced
pressure. The crude product was purified by flash chromatography
on silica gel using 60% ethyl acetate in hexane.
CHCl3 = 1:1).
4.1.14. Bis-(2S-benzyl-3-yl propanoic acid) sulfide (17)
Compound 14 (100 mg, 0.14 mmol) was suspended in 4 N HCl
(20 mL, H2O/dioxane = 1:1). The reaction was refluxed in oil bath
for 8 h and quenched by diluting with DCM followed by washing
with brine. The combined organic layers, after being dried over
Na2SO4, were evaporated to give the crude product which was
purified by flash chromatography on silica gel using 2% MeOH in
DCM with addition of 0.3% acetic acid. Compound 17 was isolated
as an oil (39 mg, 77%). 1H NMR (CDCl3, 600 MHz) d 7.27–7.13 (m,
10H), 2.93 (dd, J = 6.6, 13.2 Hz, 2H), 2.82–2.73 (m, 4H), 2.64 (dd,
J = 8.4, 13.2 Hz, 2H), 2.51 (dd, J = 4.8, 13.2 Hz, 2H); 13C NMR (CDCl3,
4.2.1. Bis-[[N-(2R)-phenylglycinol]-2-phenyl-3-yl-propiona-
mide] sulfide (24)
Compound 24 was isolated as a white solid (145 mg, 62%). Mp:
207–208 °C; 1H NMR (CDCl3, 600 MHz) d 7.18–6.81 (m, 20H), 4.85
(dd, J = 4.2, 7.2 Hz, 2H), 3.66 (dd, J = 4.8, 11.4 Hz, 2H), 3.54 (dd,
J = 4.8, 11.4 Hz, 2H), 2.87 (dd, J = 6.6, 12.0 Hz, 2H), 2.78–2.72 (m,
6H), 2.56–2.53 (m, 2H); 13C NMR (CDCl3, 150 MHz) d 174.9,
139.0, 139.0, 129.0, 128.5, 128.4, 127.2, 126.6, 126.4, 65.2, 55.5,
150 MHz) d 180.5, 138.1, 129.1, 128.8, 127.0, 47.4, 38.0, 32.7; ½a D20
ꢂ
ꢀ14.8° (c 1.0 CHCl3); HRMS (ESI) C22H26O4S 358.1239, found
358.1299, 3.7 ppm error.
4.1.15. Bis-[N-(2,2-dimethyl-8,8a-dihydro-3aH-indeno[(1S,2R)-
1,2]-oxazol-3-yl) 2R-benzyl-3-yl propianamide] sulfoximine (22)
To a suspension of 1 (50 mg, 0.07 mmol) in toluene (4 mL)
was added 2,2-dimethoxypropane (4 mL) followed by 10-cam-
phorsulfonic acid (17.6 mg, 0.07 mmol). The reaction was stirred
at 50 °C for 12 h. The reaction was quenched by adding ethyl
acetate and 10% NaHCO3. Organic layer was separated, dried
over Na2SO4 and concentrated under reduced pressure. Flash
chromatography eluting with 20% ethyl acetate in hexane affor-
ded 22 as an oil (42 mg, 76%). 1H NMR (CDCl3, 600 MHz) d 7.34–
6.70 (m, 16H), 6.76 (d, J = 5.4 Hz, 1H), 6.70 (d, J = 5.4 Hz, 1H),
5.84–5.73 (m, 2H), 4.79–4.74 (m, 2H), 3.76–3.68 (m, 2H), 3.57–
3.46 (m, 4H), 3.06–2.87 (m, 8H), 1.35 (s, 6H), 1.26 (s, 6H); 13C
49.2, 38.5, 34.0; ½a D20
ꢀ52.7° (c 0.22 MeOH/CHCl3 = 1:1); ESI MS
ꢂ
C36H40N2O4S 596.3, found 597.3 (M+H)+.
4.2.2. Bis-[[N-(2S)-phenylglycinol]-2-phenyl-3-yl-propionamide]
sulfide (25)
Compound 25 was isolated as a white solid (160 mg, 64%). Mp:
188–189 °C; 1H NMR (DMSO-d6, 600 MHz) d 8.13 (d, J = 8.4 Hz, 2H),
7.26–7.12 (m, 20H), 4.74 (dd, J = 6.6, 13.8 Hz, 2H), 4.61 (t, J = 5.4 Hz,
2H), 3.35–3.32 (m, 4H), 2.77–2.57 (m, 8H), 2.40–2.38 (m, 2H); 13C
NMR (DMSO-d6, 150 MHz) d 173.0, 141.7, 140.0, 129.5, 128.8,
128.5, 127.7, 127.2, 126.7, 65.1, 55.3, 48.0, 38.5, 34.5; ½a D20
ꢂ
+52.0°
(c 0.2 MeOH/CHCl3 = 1:1); ESI MS C36H40N2O4S 596.3, found
597.2 (M+H)+.
NMR (CDCl3, 150 MHz)
d 169.9, 140.9, 140.8, 140.8, 137.4,
137.4, 129.3, 129.3, 129.1, 129.1, 128.3, 128.3, 127.4, 127.2,
127.2, 126.0, 125.9, 123.9, 123.9, 97.0, 79.2, 66.6, 66.4, 58.8,
4.2.3. Bis-[(L-valine-N-methylamide)-2-phenyl-3-yl-propiona-
mide] sulfide (26)
58.3, 41.3, 40.5, 39.1, 39.0, 36.0, 29.7, 26.5; ½a D20
ꢂ
+102.0° (c 0.5
The desired product was crystallized from ethyl acetate as a
CHCl3); HRMS (ESI) C44H49N3O5S 731.3393, found 732.3430
white solid (198 mg, 81%). Mp: 267–268 °C; 1H NMR (DMSO-d6,
(M+H)+ 4.9 ppm error.

D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
2047
600 MHz) d 7.87 (d, J = 9.0 Hz, 2H), 7.40 (d, J = 4.2 Hz, 2H), 7.21–
7.10 (m, 10H), 4.01 (t, J = 8.4 Hz, 2H), 2.88–2.51 (m, 10H), 2.47 (s,
6H), 1.88 (dd, J = 6.6, 13.2 Hz, 2H), 0.77 (t, J = 6.0 Hz, 12H); 13C
NMR (DMSO-d6, 150 MHz) d 173.4, 171.8, 139.9, 129.5, 126.7,
ture for 12 h and was quenched by adding saturated NaHCO3 solu-
tion. The mixture was stirred for 30 min in an ice bath and the
crude product was purified by flash chromatography using 4%
MeOH in DCM to afford compound 30, 31, or 32.
47.5, 38.5, 34.2, 30.9, 26.1, 19.9, 19.1; ½a D20
ꢀ46.7° (c 0.24 MeOH/
ꢂ
CHCl3 = 1:1); ESI MS C32H46N4O4S 582.3, found 583.3 (M+H)+.
4.4.1. Bis-[[N-(2R)-phenylglycinol]-2-phenyl-3-yl-propionamide]
sulfoximine (30)
4.3. General procedure for the preparation of diamide sulfoxide
(27, 28, 29)
Compound 30 was obtained as a white solid (60 mg, 74%). Mp:
218–219 °C; 1H NMR (DMSO-d6, 600 MHz) d 8.38 (d, J = 8.4 Hz, 1H),
8.34 (d, J = 7.8 Hz, 1H), 7.19–6.95 (m, 20H), 4.81–4.78 (m, 2H),
4.70–4.60 (m, 2H), 3.53–3.47 (m, 4H), 3.41–3.38 (m, 2H), 3.29–
3.23 (m, 2H), 2.96–2.92 (m, 2H), 2.79–2.68 (m, 4H); 13C NMR
(DMSO-d6, 150 MHz) d 172.4, 172.3, 141.1, 141.0, 138.9, 138.8,
129.8, 129.7, 128.9, 128.8, 128.7, 128.5, 127.8, 127.4, 127.3,
127.2, 127.1, 127.0, 126.9, 65.4, 65.2, 57.1, 56.6, 55.4, 55.3, 42.4,
To a solution of 24, 25, or 26 (1 equiv) in DCM (12 mL) and
methanol (12 mL) at ꢀ78 °C under N2 was added mCPBA (1 equiv).
The mixture was stirred at ꢀ78 °C for 1 h then at room tempera-
ture for another 1 h. The reaction was quenched by diluting with
DCM and extracted with saturated NaHCO3. The organic layer
was then washed by brine, dried over Na2SO4, concentrated under
reduced pressure. The crude product was purified by flash chroma-
tography on silica gel using 4% MeOH in CHCl3 to afford compound
27, 28, or 29.
42.2, 38.9, 38.8; ½a D20
ꢀ23.8° (c 0.26 MeOH/CHCl3 = 1:1); ESI MS
ꢂ
C36H41N3O5S 627.3, found 628.3 (M+H)+.
4.4.2. Bis-[[N-(2S)-phenylglycinol]-2-phenyl-3-yl-propionamide]
sulfoximine (31)
4.3.1. Bis-[[N-(2R)-phenylglycinol]-2-phenyl-3-yl-propionamide]
sulfoxide (27)
Compound 31 was obtained as a white solid (50 mg, 54%). Mp:
185–186 °C; 1H NMR (DMSO-d6, 600 MHz) d 8.25 (d, J = 7.8 Hz, 1H),
8.22 (d, J = 7.8 Hz, 1H), 7.28–7.15 (m, 20H), 4.70–4.63 (m, 2H),
3.38–3.12 (m, 10H), 2.78–2.64 (m, 6H), 2.63–2.57 (m, 2H); 13C
NMR (DMSO-d6, 150 MHz) d 175.1, 141.4, 141.3, 129.9, 129.8,
128.8, 128.7, 128.6, 128.6, 128.5, 127.9, 127.8, 127.3, 127.3,
127.3, 127.0, 127.0, 64.9, 64.9, 56.6, 56.4, 55.5, 55.4, 42.3, 42.2,
Compound 27 was isolated as a white solid (95 mg, 92%). Mp:
208–209 °C; 1H NMR (DMSO-d6, 600 MHz) d 8.53 (d, J = 7.8 Hz,
1H), 8.34 (d, J = 8.4 Hz, 1H), 7.26–6.93 (m, 20H), 4.79–4.75 (m,
2H), 3.53–3.44 (m, 4H), 3.20–3.12 (m, 2H), 3.08–3.04 (m, 1H),
2.93 (dd, J = 7.2, 13.2 Hz, 1H), 2.88 (dd, J = 7.8, 13.8 Hz, 1H), 2.83–
2.75 (m, 3H), 2.68 (dd, J = 7.8, 13.2 Hz, 1H), 2.57–2.55 (m, 1H);
13C NMR (DMSO-d6, 150 MHz) d 172.4, 172.3, 141.6, 141.1, 139.2,
139.0, 129.8, 129.7, 128.9, 128.8, 128.6, 128.5, 127.5, 127.3,
127.2, 127.2, 127.0, 126.9, 65.3, 55.6, 55.4, 55.3, 54.5, 42.5, 39.2,
39.0, 38.9;
½
a 2D0
+82.0° (c 0.2 MeOH/CHCl3 = 1:1); ESI MS
ꢂ
C36H41N3O5S 627.3, found 628.2 (M+H)+.
38.2; ½a 2D0
ꢂ
ꢀ60.0° (c 0.13 MeOH/CHCl3 = 1:1); ESI MS C36H40N2O5S
4.4.3. Bis-[(L-valine-N-methylamide)-2-phenyl-3-yl-propiona-
mide] sulfoximine (32)
612.3, found 613.2 (M+H)+.
Compound 32 was obtained as a white solid (19 mg, 41%). Mp:
195–196 °C; 1H NMR (DMSO-d6, 600 MHz) d 7.97 (d, J = 6.6 Hz, 2H),
7.20 (br, 2H), 7.22–7.14 (m, 10H), 3.99–3.91 (m, 2H), 3.29–2.55 (m,
10H), 2.49 (s, 6H), 1.53–1.48 (m, 2H), 0.77 (s, 12H); 13C NMR
(DMSO-d6, 150 MHz) d 172.6, 172.4, 171.7, 163.6, 138.9, 138.8,
129.8, 129.7, 128.8, 127.0, 73.2, 58.8, 56.7, 56.2, 41.9, 41.7, 38.9,
4.3.2. Bis-[[N-(2S)-phenylglycinol]-2-phenyl-3-yl-propionamide]
sulfoxide (28)
Compound 28 was isolated as a white solid (147 mg, quantita-
tive). Mp: 205–206 °C; 1H NMR (DMSO-d6, 600 MHz) d 8.39 (d,
J = 7.8 Hz, 1H), 8.28 (d, J = 7.8 Hz, 1H), 7.28–7.12 (m, 20H), 4.79–
4.77 (m, 1H), 4.63 (t, J = 6.0 Hz, 1H), 3.39 (t, J = 6.0 Hz, 2H), 3.35–
3.29 (m, 2H), 3.13–3.12 (m, 1H), 3.02 (t, J = 6.6 Hz, 1H), 2.96–2.88
(m, 2H), 2.84–2.72 (m, 3H), 2.67 (dd, J = 7.2, 13.2 Hz, 1H), 2.62
(dd, J = 7.2, 13.2 Hz, 1H), 2.44–2.41 (m, 1H); 13C NMR (DMSO-d6,
150 MHz) d 172.2, 172.2, 141.6, 141.5, 139.2, 139.1, 129.8, 129.7,
128.9, 128.8, 128.6, 128.6, 127.7, 127.3, 127.3, 127.0, 126.9, 65.1,
38.7, 31.1, 26.1, 19.9, 19.1, 19.0; ½a D20
ꢀ11.7° (c 0.24 MeOH/
ꢂ
CHCl3 = 1:1); HRMS (ESI) C32H47N5O5S 613.3298, found 636.3158
(M+Na)+ 5.0 ppm error.
4.5. Biology
64.9, 55.5, 55.5, 54.9, 54.4, 42.4, 41.8, 39.3, 38.0; ½a D20
ꢂ
+15.5° (c
All the synthesized compounds were tested against recombined
HIV-1 protease (Bachem Biosciences, Philadelphia PA) by a HPLC
assay using HIV-1 Protease Substrate III (Bachem Biosciences, Phil-
adelphia PA) or a fluorescence assay using EnzoLyte 520 HIV-1 Pro-
tease Assay Kit (AnaSpec, San Jose CA).
0.22 MeOH/CHCl3 = 1:1); ESI MS C36H40N2O5S 612.3, found 613.2
(M+H)+.
4.3.3. Bis-[(L-valine-N-methylamide)-2-phenyl-3-yl-propionamide]
sulfoxide (29)
4.6. HPLC assay
Compound 29 was isolated as a white solid (85 mg, 92%). Mp:
228–229 °C; 1H NMR (DMSO-d6, 600 MHz) d 8.06 (d, J = 8.4 Hz,
1H), 8.00 (d, J = 8.4 Hz, 1H), 7.46 (br, 2H), 7.23–7.11 (m, 10H),
4.01–3.96 (m, 2H), 3.17–2.63 (m, 10H), 2.49 (s, 6H), 1.88–1.86
(m, 2H), 0.77 (s, 6H), 0.75 (s, 6H); 13C NMR (DMSO-d6, 150 MHz)
d 172.6, 172.6, 171.7, 171.7, 139.0, 129.8, 129.7, 128.9, 128.8,
127.0, 126.9, 58.7, 58.7, 42.0, 41.7, 39.0, 38.2, 31.0, 31.0, 26.1,
The reactions were carried out in a buffer containing 50 mM
sodium acetate (pH 4.9), 200 mM NaCl, 5 mM DTT, and 10% glyc-
erol. The HIV-1 Substrate III was added to the buffer followed by
inhibitors. The inhibitors were dissolved in DMSO, with a final
concentration of 5% (v/v). Baseline values were determined from
enzymatic reactions containing 5% DMSO in the absence of inhib-
itor. The reactions were initiated by the addition of the HIV-1
protease enzyme for 3.5 min at 37 °C. The reactions were
quenched with 1/10 volume of 10% TFA and stored at 4 °C until
analyzed. The samples were analyzed on a Beckman Coulter
HPLC system, using a reverse-phase C18 column with a gradient
of 5–50% acetonitrile in the presence of 0.1% TFA at a rate of
1 mL/min.
26.1, 19.9, 19.1, 18.9; ½a D20
ꢀ79.2° (c 0.24 MeOH/CHCl3 = 1:1); ESI
ꢂ
MS C32H46N4O5S 598.3, found 599.3 (M+H)+.
4.4. General procedure for the preparation of diamide sulfoximine
(30, 31, 32)
To a solution of 27, 28, or 29 (1 equiv) in DMF (10 mL) was
added MSH (5 equiv). The mixture was stirred at room tempera-

2048
D. Lu et al. / Bioorg. Med. Chem. 18 (2010) 2037–2048
4.7. Fluorescence assay
for performing the HIV protease assay and Southern Research Insti-
tute for the antiviral testing data. We also thank Dr. William M.
Shannon for his advice as a consultant for the antiviral work on this
project. The authors thank Dr. Zhengqiang Wang for critical read-
ing of the manuscript and helpful scientific discussions. The Uni-
versity of Minnesota Supercomputing Institute provided all the
necessary computational resources.
These reactions were incubated in Costar 3915 plates (96 well,
non-treated, flat bottom, black) and followed the manufacturer
recommended protocol for screening HIV-1 protease inhibitors.
The fluorescent signal was measured every 2–3 min (Ex/
Em = 490/520) in a BioTek Synergy HT plate reader. The data were
collected using KC4 and further analyzed for IC50 values with Prism
Graph 4.
References and notes
1. UNAIDS/WHO AIDS Epidemic Update: Dec 2007 (http://www.unaids.org).
2. Yeni, P. J. Hepatol. 2006, 44, S100.
4.8. Molecular modeling
3. Kedzierska, K.; Maslin, C. L. V.; Crowe, S. M. Med. Chem. Rev. 2004, 1, 351.
4. Miller, M.; Schneider, J.; Sathyanarayana, B. K.; Toth, M. V.; Marshall, G. R.;
Clawson, L.; Selk, L.; Kent, S. B. H.; Wlodawer, A. Science 1989, 246, 1149.
5. Swain, A. L.; Miller, M. M.; Green, J.; Rich, D. H.; Schneider, J.; Kent, S. B. H.;
Wlodawer, A. Proc. Natl. Acad. Sci. 1990, 87, 8805.
6. Kohl, N. E.; Emini, E. A.; Schleif, W. A.; Davis, L. J.; Heimbach, J. C.; Dixon, R. A.;
Scolnick, E. M.; Sigal, I. S. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 4686.
7. Huff, J. R. J. Med. Chem. 1991, 34, 2305.
8. Umezawa, H. A. T.; Morishima, H.; Matsuzaki, M.; Hamada, M.; Takeuchi, T. J.
Antibiot. 1970, 23, 259.
9. Lu, D.; Vince, R. Bioorg. Med. Chem. Lett. 2007, 17, 5614.
10. Raza, A.; Sham, Y. Y.; Vince, R. Bioorg. Med. Chem. Lett. 2008, 18, 5406.
11. Santaniello, E.; Ferraboschi, P.; Grisenti, P.; Manzocchi, A. Chem. Rev. 1992, 92,
1071.
12. Itoh, T.; Chika, J.; Takagi, Y.; Nishiyama, S. J. Org. Chem. 2002, 58, 5717.
13. Campbell, R. F.; Fitzpatrick, K.; Inghardt, T.; Karlsson, O.; Nilsson, K.; Reilly, J. E.;
Yet, L. Tetrahedron Lett. 2003, 44, 5477.
14. Liu, K. K. C.; Nozaki, K.; Wong, C. H. Biocatalysis 1990, 3, 169.
15. Iversen, T.; Bundle, D. R. J. Chem. Soc., Chem. Commun. 1981, 23, 1240.
16. Carr, J. A.; Bisht, K. S. Organic Lett. 2004, 6, 3297.
17. Spino, C.; Tremblay, M. C.; Godbout, C. Org. Lett. 2004, 6, 2801.
18. Zhu, X.; Schmidt, R. R. Chem.-A Eur. J. 2004, 10, 875.
19. Dorsey, B. D.; Levin, R. B.; McDaniel, S. L.; Vacca, J. P.; Guare, J. P.; Darke, P. L.;
Zugay, J. A.; Emini, E. A.; Schleif, W. A., et al J. Med. Chem. 1994, 37, 3443.
20. Askin, D.; Wallace, M. A.; Vacca, J. P.; Reamer, R. A.; Volante, R. P.; Shinkai, I. J.
Org. Chem. 1992, 57, 2771.
All modeling was carried out using the Schrödinger modeling
suite package.34 Compounds 1, 30, 31, and 32 were built and en-
ergy minimized using the OPLS-AA force field35 with implicit gen-
eralized born solvent. Two previously reported X-ray
crystallographic structures of HIV-1 protease complexes were used
for the docking of the pseudosymmetric sulfoximine inhibitors of
1, 30, and 31 (PDB: 4PHV)36 and the sulfoximine peptidomimetic
32 (PDB: 1EBW).37 The conserved crystallographic water, com-
monly observed to hydrogen bond to the amide hydrogen atom
of G49 and G490, was kept as part of the catalytic binding site for
the protein for the docking calculation. To account for the previ-
ously observed ionization state of D25/D250, each of the four oxy-
gen atoms in D25/D250 carboxylate groups was systematically
protonated and allowed to be energy minimized in implicit solvent
while all heavy atoms were held fixed. D250 OD1 which was found
with the optimal hydrogen bonding network was selected as our
reference protonated state for docking of all our sulfoximine ana-
logues. The docking calculations were carried using GLIDE v5.5 with
the extra precision protocol. The van der Waals radii of non-polar
atoms for each of the ligands were scaled by a factor of 0.8. To ac-
count for the protein flexibility, each of the final docked complexes
was further refined by 100 ps equilibration followed by 1 ns
molecular dynamics simulation using spherical boundary condi-
tion with explicit TIP3P solvent. A two layer spherical shell systems
were employed with the nearest 18 Å sphere from the center of the
ligand restrained by force constant of 0.3 kcal/mol Å2 while fixing
the residues in the outer 4 Å region shell. A constant dielectric of
1.0 with a 14 Å non-bond cutoff was used. All TIP3P water, except
for the conserved crystallographic water bound to the ligand and
G49/G490, was removed from the final structure.
21. Lin, J. H.; Chiba, M.; Balani, S. K.; Chen, I. W.; Kwei, G. Y. S.; Vastag, K. J.;
Nishime, J. A. Drug Metab. Dispos. 1996, 24, 1111.
22. Thaisrivongs, S.; Turner, S. R.; Strohbach, J. W.; TenBrink, R. E.; Tarpley, W. G.;
McQuade, T. J.; Heinrikson, R. L.; Tomasselli, A. G.; Hui, J. O.; Howe, W. J. J. Med.
Chem. 1993, 36, 941.
23. Muhlman, A.; Lindberg, J.; Classon, B.; Unge, T.; Hallberg, A.; Samuelsson, B. J.
Med. Chem. 2001, 44, 3407.
24. Johnson, C. R.; Schroeck, C. W.; Shanklin, J. R. J. Am. Chem. Soc. 1973, 95, 7424.
25. Johnson, C. R.; Haake, M.; Schroeck, C. W. J. Am. Chem. Soc. 1970, 92, 6594.
26. Johnson, C. R.; Lavergne, O. M. J. Org. Chem. 1993, 58, 1922.
27. Diem, M. J.; Burow, D. F.; Fry, J. L. J. Org. Chem. 1977, 42, 1801.
28. Fehrentz, J. A.; Chomier, B.; Bignon, E.; Venaud, S.; Chermann, J. C.; Nisato, D.
Biochem. Biophys. Res. Commun. 1992, 188, 865.
29. Schechter, I.; Berger, A. Biochem. Biophys. Res. Commun. 1967, 27, 157.
30. Wlodawer, A.; Erickson, J. W. Annu. Rev. Biochem. 1993, 62, 543.
31. Lam, P. Y.; Jadhav, P. K.; Eyermann, C. J.; Hodge, C. N.; Ru, Y.; Bacheler, L. T.;
Meek, J. L.; Otto, M. J.; Rayner, M. M.; Wong, Y. N., et al Science 1994, 263,
380.
32. Erickson, J.; Neidhart, D. J.; VanDrie, J.; Kempf, D. J.; Wang, X. C.; Norbeck, D.
W.; Plattner, J. J.; Rittenhouse, J. W.; Turon, M.; Wideburg, N., et al Science 1990,
249, 527.
33. Hyland, L. J.; Tomaszek, T. A., Jr.; Meek, T. D. Biochemistry 1991, 30, 8454.
34. Schrodinger, LLC, New York, NY (http://www.schrodinger.com).
35. Jorgensen, W. L.; Maxwell, D. S.; Tirado-Rives, J. J. Am. Chem. Soc. 1996, 118,
11225.
The absolute binding energy,
DGbind, was evaluated based on
the linear response method38–41 implemented within Liason.34
f!b
f!b
f!b
D
Gbind
¼
a
h
D
U
iq þ bh
D
U
iq þ
c
h
D
U
iq
ð1Þ
cav
elec
vdw
f!b
f!b
f!b
where h
D
U
iq, h
DU
iq, h
DU
iq is the change in the van der
cav
elec
vdw
Waals, electrostatic and cavity energy from the free (f) to bound (b)
state during binding at the average charged state, q, as evaluated
by the Surface Generalized Born (SGB) continuum solvent model.
36. Bone, R.; Vacca, J. P.; Anderson, P. S.; Holloway, M. K. J. Am. Chem. Soc. 1991,
113, 9382.
37. Andersson, H. O.; Fridborg, K.; Lowgren, S.; Alterman, M.; Muhlman, A.;
Bjorsne, M.; Garg, N.; Kvarnstrom, I.; Schaal, W.; Classon, B.; Karlen, A.;
Danielsson, U. H.; Ahlsen, G.; Nillroth, U.; Vrang, L.; Oberg, B.; Samuelsson, B.;
Hallberg, A.; Unge, T. Eur. J. Biochem. 2003, 270, 1746.
a, b, and c are empirical fitting coefficients evaluated by multivar-
iable regression analysis. The absolute binding energies of the HIV
PI used for the training set was based on the previous studies42 and
the fitted equation was used to predict the absolute binding affin-
ities of the sulfoximine inhibitors.
38. Hansson, T.; Aqvist, J. Protein Eng. 1995, 8, 1137.
39. Sham, Y. Y.; Chu, Z. T.; Tao, H.; Warshel, A. Proteins 2000, 39, 393.
40. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong, L.; Lee, M.; Lee,
T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D. A.;
Cheatham, T. E., 3rd. Acc. Chem. Res. 2000, 33, 889.
41. Jones-Hertzog, D. K.; Jorgensen, W. L. J. Med. Chem. 1997, 40, 1539.
42. Hou, T.; McLaughlin, W. A.; Wang, W. Proteins 2008, 71, 1163.
Acknowledgments
This research work was supported by the Center for Drug De-
sign at the University of Minnesota. We thank Ms. Christine Dreis