New series of potent, orally bioavailable, non-peptidic cyclic sulfones as HIV-1 protease inhibitors

Full text of DOI:10.1021/jm960340o

© Copyright 1996 by the American Chemical Society
Volume 39, Number 18
August 30, 1996
Communications to the Editor
New Ser ies of P oten t, Or a lly
Bioa va ila ble, Non -P ep tid ic Cyclic
Su lfon es a s HIV-1 P r otea se In h ibitor s
Choung U. Kim,*,† Lawrence R. McGee,†
Steven H. Krawczyk,† Erik Harwood,† Yoko Harada,†
S. Swaminathan,† Norbert Bischofberger,†
Ming S. Chen,† J ulie M. Cherrington,†
Shelly F. Xiong,† Linda Griffin,† Kenneth C. Cundy,†
Angela Lee,‡ Betty Yu,‡ Sergli Gulnik,‡ and
J ohn W. Erickson‡
F igu r e 1. Structures of protease inhibitors.
Gilead Sciences Inc., 353 Lakeside Drive, Foster City,
California 94404, and Structural Biochemistry Program,
SAIC Frederick, NCI-Frederick Cancer, Research and
Development Center, Frederick, Maryland 21702
Received May 7, 1996
Inhibitors of human immunodeficiency virus protease
(HIV PR) prevent assembly of viral proteins and matu-
ration of HIV and result in noninfective virions.1
Despite intensive research effort and a great number
of inhibitors reported in the literature, only a few have
reached clinical trials mainly due to metabolic instabil-
ity and low oral bioavailability.2 Although some clinical
results of HIV PR inhibitors for the treatment of AIDS
are encouraging, emergence of viral resistance presents
a major challenge for the clinical use of this class of
drugs.3 Therefore, there is still an urgent need for new
classes of protease inhibitors possessing different bind-
ing interactions with protease, which could potentially
lead to fewer and/or different resistance development.
HIV PR is an aspartic protease that functions as a
homodimer.4 Inhibitors incorporating C2-symmetric
diols as transition state analogues have been reported
to impart high affinity toward the enzyme.5 In addition
to the inhibitory potency, good pharmacological proper-
ties (oral bioavailability, metabolic stability, and
pharmacokinetics) are essential factors for protease
inhibitors to be developed as clinical candidates. For-
F igu r e 2. Hydrogen-bonding interaction between 1c and
HIV-1 protease.
mulating these goals, our design strategy had the
following requirements: (1) low molecular weight (<600),
† Gilead Sciences Inc.
‡ NCI-Frederick Cancer, Research and Development Center.
S0022-2623(96)00340-8 CCC: $12.00 © 1996 American Chemical Society

3432 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Communications to the Editor
F igu r e 3. Stereoview of the X-ray structure of 1c and HIV-1 protease complex.
(2) non-peptidic, (3) C2-symmetric cyclic structure with
1,2-diol functionality, and (4) some water solubility.6
Taking into account these factors, together with struc-
tural information available from X-ray crystallography
of C2-symmetric diol inhibitors complexed with HIV PR,
our molecular modeling studies led to the discovery of
the cyclic sulfones 1 as potent, non-peptidic inhibitors
of HIV PR.7 The absence of amide bonds in these
structures prevents these molecules from proteolytic
cleavage and should lead to pharmacologically superior
antiviral agents compared to classical peptide-based
inhibitors.
plexes with linear peptidomimetic inhibitors.9 Such
replacement and the hydrogen bond bridge may enhance
the binding of the inhibitor in terms of both enthalpy
and entropy. The replacement of the structural water
in the enzyme with the sulfonyl group is a novel
extension of DuPont Merck’s previously published C2-
symmetric cyclic urea-based HIV PR inhibitors,10 in
which the carbonyl of the urea also replaced water 301.
The OH groups of the C3 and C4 atoms straddle the
two active site aspartyl carboxylate groups symmetri-
cally. The C3 OH is 2.67 Å from OD2 of Asp125 and
2.80 Å from OD1 of Asp25. The C4 OH is positioned 2.74
and 3.04 Å from OD2 and OD1, respectively, of Asp25
and Asp125. This geometry provides an extensive
network of hydrogen bonds among all the oxygen atoms
of the carboxylategroups of the protease and the hy-
droxyl groups of the inhibitor.
The X-ray crystallographic structure of the inhibitor
1c complexed with HIV-1 PR at 2.2 Å resolution was
obtained.8 The inhibitor binds to the protease in a
highly symmetric fashion revealing many of the stabi-
lizing interactions (Figure 2). The sulfonyl and the two
hydroxy groups of the inhibitor provide anchoring
hydrogen bonds, while benzyl and thiazolyl side chains
gain substantial hydrophobic interactions from the
substrate pockets of the protease. The two oxygen
atoms of the sulfonyl group form a hydrogen-bonding
bridge to the backbone NH atoms (1.87 and 2.04 Å) of
the residues Ile50 and Ile150 located at the tip of the
flaps of the protease. These interactions replace the
structural water 301 which is normally found in com-
The side chain interactions of the inhibitor with the
protease are described in Figure 3 in a view which is
approximately perpendicular to the view in Figure 2.
The C1 and C6 benzyl groups fill the S2 and S2′ pockets,
while the C2 and C5 thiazoyl side chains occupy the S1
and S1′ pockets. All four rings make hydrophobic
interactions with the side chains of the protease in these
pockets. Ile50 and Ile184 line the S1 and S2′ pockets,

Communications to the Editor
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3433
Ta ble 1. Biological Activity of Cyclic Sulfones 1a -c against
duration that the plasma drug concentration remained
40 times above EC90 exceeded 12 h. Conversely, com-
pounds 1a ,c, which had almost no water solubility (< 1
µg/mL), showed virtually no oral absorption, confirming
the importance of balancing water/lipid solubility for
oral bioavailability.
In conclusion, we have demonstrated that the seven-
membered ring sulfone (thiepane dioxide) served as a
conformationally constrained scaffold for the rational
drug design of potent HIV PR inhibitors. Certain
members of this class represent some of the simplest,
most potent HIV PR inhibitors reported to date. Studies
are in progress to assess the potential of this novel class
of compounds as chemotherapeutic agents for the treat-
ment of AIDS.
HIV-1 Virus
compd IC50 (nM)a EC50 (nM)b/EC90 (nM) CC50 (µM)c
SI
1a
1b
1c
0.6
1.0
0.3
40/70
9/20
6.5/40
6
30
40
150
3333
6153
a
Concentration needed to inhibit HIV-1 protease activity by
b
50%. For the detailed assay, see ref 11. Concentration needed
to inhibit HIV-1 virus replication in MT2 cells by 50% as
determined by an XTT assay. For the detailed assay, see ref 12.
c Concentration needed to produce a 50% reduction in the number
of viable cells as assayed by metabolism of a tetrazolium dye.
Ack n ow led gm en t. We would like to thank Dr. J ohn
C. Martin for his many contributions and continuing
support of this project.
Su p p or tin g In for m a tion Ava ila ble: Physical and spec-
tral data for compounds 1a -c (2 pages). Ordering information
is found on any current masthead page.
Refer en ces
(1) J ohnston, M. I.; Hoth, D. F. Present Status and Future Prospects
for HIV Therapies. Science 1993, 260, 1286-1293.
(2) West, M. L.; Fairlie, D. P. Targeting HIV-1 Protease: A Test of
Drug-Design Methodologies. Trends Pharmacol. Sci. 1995, 16,
67-75.
F igu r e 4. Mean ( SD concentrations of 1b in plasma
following intravenous or oral administration to male beagle
dogs at 10 mg/kg.13
(3) Condra, J . H.; Schief, W. A.; Blahy, O. M.; Gabryelski, L. J .;
Graham, D. J .; Quintero, J . C.; Rhodes, A.; Robbins, H. L.; Roth,
E.; Shivaprakash, M.; Titus, D.; Yang, H.; Teppler, K. E.; Squire,
K. E.; Deutsch, P. J .; Emini, E. A. In Vivo Emergence of HIV-1
Variants Resistant to Multiple Protease Inhibitors. Nature
1995, 374, 569-571.
and Ile150 and Ile84 line the S1′ and S2 pockets. The
other major hydrophobic interactions involve Val82 and
Pro81 as well as Ile47 and Val32 along with the same
groups from the other protease monomer. In addition,
there are long range interactions (4.60 and 4.12 Å)
between the N atoms of the thiazole rings and Arg8 and
Arg108 of the protease. The absolute configuration of
the C1 and C2 side chains (1R,2R) presented in structure
1 is critically important for potent antiviral activity.
Thus, the 1R,2S diastereomer of 1a was 20-fold less
active compared to 1a , and the 1S,2S isomer exhibited
500-fold reduced activity in the enzymatic assay. Mod-
eling analysis of the 1S,2S isomer revealed that the
(1S)-benzyl group is not well-accommodated in the
active site where it collides with the flap.
(4) Pearl, L.; Taylor, W. A Structural Model for the Retroviral
Proteases. Nature 1987, 329, 351-354.
(5) Erickson, J . W. Design and Structure of Symmetry-Based
Inhibitors of HIV-1 Protease. Prospect. Drug Discovery Des.
1993, 1, 109-128.
(6) Kim, E. E.; Baker, C. T.; Dwyer, M. D.; Murcko, M. A.; Rao, B.
G.; Tung, R. D.; Navia, M. A. Crystal Structure of HIV-1 Protease
in Complex with VX-478, A Potent and Orally Bioavailable
Inhibitor of the Enzyme. J . Am. Chem. Soc. 1995, 117, 1181-
1182.
(7) The structure-activity relationships of a series of compounds
related to 1 will be published separately.
(8) Lee, A.; Erickson, J . W. Unpublished data.
(9) Wlodawer, A.; Erickson, J . W. Structure-Based Inhibitors of
HIV-1 Protease. Annu. Rev. Biochem. 1993, 62, 543-585.
(10) Lam, P. Y. S.; J adhav, P. K.; Evermann, C. J .; Hodge, C. N.;
Ru, Y.; Bacheler, L. T.; Meek, J . L.; Otto, M. J .; Rayner, M. L.;
Wong, N. Y.; Chang, C. H.; Weber, P. C.; J ackson, D. A.; Sharpe,
T. R.; Erickson-Viitanen, S. Rational Design of Potent, Bioavail-
able, Nonpeptide Cyclic Ureas as HIV Protease Inhibitors.
Science 1994, 263, 380-383.
(11) Matayoshi, E. D.; Wang, G. T.; Kraft, G. A.; Erickson, J . Novel
Fluorogenic Substrates for Assaying Retroviral Proteases by
Resonance Energy Transfer. Science 1990, 247, 954-958.
(12) Gu, Z.; Salomon, H.; Cherrington, J . M.; Mulato, A. S.; Chen,
M. S.; Yachoan, R.; Foli, A.; Sogocto, K. M.; Weinberg, M. A.
K65R Mutation of Human Immunodeficiency Virus Type 1
Reverse Transcriptase Encodes Cross-Resistance to 9-(2-Phos-
phonylmethoxyethyl)adenine. Antimicrob. Agents Chemother.
1995, 39, 1888-1891.
In addition to the benzyl analogue 1a , the meta
aminobenzyl and thiazolyl analogues 1b,c were also
prepared to provide variations in lipophilic, electronic,
and hydrogen-bonding properties. Both the meta ami-
nobenzyl group in P2 (P2′) positions16 and the thiazolyl
group in P2′ position17 in HIV PR inhibitors have been
reported to impart high antiviral activity. Cyclic sul-
fones 1a -c exhibited high antiviral potency against
HIV-1 as assessed by inhibition of protease in an
enzyme assay and viral replication in tissue culture
(Table 1). The high level of antiviral activity (EC50) and
low cytotoxicity (CC50) provided high selective indices
(SI) for compounds 1b,c. In our design goals, aqueous
solubility was considered to be important in order to
achieve good oral bioavailability. Without reducing
antiviral activity, the amino group can be added on to
the C2 (C5) benzyl phenyls at the meta position. The
resulting 1b had a water solubility of 20 µg/mL at pH
7.0. At a dose of 10 mg/kg in dogs (n ) 5), compound
1b had an oral bioavailability of 74% (Figure 4). The
(13) Dog PK study and sample analysis of 1b: The pharmacokinetics
of 1b were examined in five male beagle dogs (8-12 kg) following
intravenous or oral administration. 1b was formulated as an
aqueous solution at pH 2.5 and administered at 10 mg/kg. Each
dog received a single intravenous injection via a cephalic vein
and a single oral administration by gavage with
a 1 week
washout period. Blood samples were obtained from a peripheral
vein at intervals over 24 h and processed for plasma. Concen-
trations of 1b in dog plasma were determined by reverse phase
HPLC with UV detection. Plasma samples (100 µL) were mixed
with 100 µL of acvetonitrile to precipitate proteins and centri-
fuged at 2500g and 4 °C for 10 min. Supernatant (170 µL) was
added to 90 µL of 20 mM potassium phosphate buffer, pH 6.9,
and injected onto the HPLC. Detection was by UV absorption
at 250 nM. The limit of quantitation was 0.1 µg/mL.

3434 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Communications to the Editor
(14) Compounds 1a -c were prepared according to the synthetic
sequence shown below. The detailed synthesis including alter-
native routes will be published separately.
(15) Kuszmann, J .; Sohar, P. 1,6-Anhydro-1(6)-thio-L-iditol and -D-
Mannitol, and Some Derivatives Thereof. Carbohydr. Res. 1977,
56, 105-115.
(16) Wong, Y. N.; Bucham, D. L.; Saxton, P. L.; Erickson-Viitanen,
S.; Grubb, M. F.; Quon, C. Y.; Huang, S.-M. A Pharmacokinetic
Evaluation of HIV Protease Inhibitors, Cyclic Ureas, in Rats and
Dogs. Biopharm. Drug Dispos. 1994, 15, 535-544.
(17) Kemp, D. J .; Marsh, K. C.; Denissen, J . F.; McDonald, E.;
Vasavanonda, S.; Flentge, C. A.; Green, B. E.; Fino, L.; Park, C.
H.; Kong, X.-P.; Wideburg, N. E.; Saldivar, A.; Ruiz, L.; Kati,
W. M.; Sham, H. L.; Robins, T.; Stewart, K. D.; Hsu, A.; Plattner,
J . J .; Leonard, J . M.; Norbeck, D. W. ABT-538 is a Potent
Inhibitor of Human Immunodeficiency Virus Protease and Has
High Oral Bioavailability in Humans. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 2484-2488.
J M960340O