Synthesis and antiviral activities of novel N-alkoxy-arylsulfonamide-based HIV protease inhibitors

Article abstract of DOI:10.1016/j.bmcl.2005.05.101

A series of novel N-alkoxy-arylsulfonamide HIV protease inhibitors with low picomolar enzyme activity and single digit nanomolar antiviral activity is disclosed.

Full text of DOI:10.1016/j.bmcl.2005.05.101

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3560–3564
Synthesis and antiviral activities of novel
N-alkoxy-arylsulfonamide-based HIV protease inhibitors
Ronald G. Sherrill,* Eric S. Furfine, Richard J. Hazen, John F. Miller, David J. Reynolds,
Douglas M. Sammond, Andrew Spaltenstein, Pat Wheelan and Lois L. Wright
GlaxoSmithKline, 5 Moore Drive, Research Triangle Park, NC 27709, USA
Received 2 May 2005; revised 13 May 2005; accepted 16 May 2005
Available online 21 June 2005
Abstract—A series of novel N-alkoxy-arylsulfonamide HIV protease inhibitors with low picomolar enzyme activity and single digit
nanomolar antiviral activity is disclosed.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
HIV protease inhibitors are the only antivirals that have
been employed in HAART (highly active anti-retroviral
therapy) that do not target the reverse transcriptase dur-
ing viral replication. These agents, which prevent pro-
duction of infectious virions, have become the
mainstay in an ever-increasing milieu of drug combina-
tions that successfully suppress viral replication in
infected patients. However, these agents are not without
side effects, owing in part to the relatively high doses
necessary to completely suppress viral replication. Pa-
tient compliance coupled with the development of viral
resistance can ultimately lead to virologic breakthrough
and antiretroviral failure.1 Therefore, the need for
agents with increased potency and activity against resis-
tant strains of HIV is well recognized. This in turn has to
be accommodated against a complex competitive land-
scape of extensive prior art directed against this aspartyl
protease.
Figure 1.
soned that incorporation of an N-alkoxysulfonamide to
occupy the P10/P20 pocket of the active site of the viral
protease might provide a novel pharmacophore from
which to generate agents with increased antiretroviral
potency (Fig. 1). The results of this investigation are de-
tailed below.
2. Chemistry
Amprenavir (Fig. 1) and its prodrug fosamprenavir are
prototypical HIV protease inhibitors (PI) which are gen-
erally well tolerated by patients and exhibit a favorable
clinical resistance profile.2 As such, the arylsulfonamide
scaffold coupled to an ethanolamine backbone, which is
found in most inhibitors from this class, might serve as a
logical starting point for further improvements. We rea-
The general synthetic strategy for construction of these
analogs involved either addition of suitably substituted
N-alkylhydroxylamines 1 or N-alkoxy-arylsulfonamides
2
into known phenyl alanine-derived epoxide 33
(Scheme 1). Hydroxylamines were prepared through
known methodology of either alkylation or Mitsonobu
coupling of N-hydroxyphthalimide, followed by hydra-
zine deprotection to yield the free hydroxylamine 1.4
Addition of these substituted hydroxylamines into epox-
ide 3 was often sluggish (Method A). However, addition
of lithium triflate to a THF solution provided reasonable
Keyword: HIV protease.
*
Corresponding author. Tel.: +1 9194836281; fax: +1 9194836053;
e-mail: ron.g.sherrill@gsk.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.101

R. G. Sherrill et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3560–3564
3561
All aminobenzenesulfonamides were derived from their
corresponding nitrophenylsulfonyl chloride. We did ob-
serve that nitro-substituted N-alkoxy-arylsulfonamides
were incompatible with P4-phosphazene and therefore
were reduced to the aniline prior to addition to epoxide
3. Otherwise, reduction of the nitro groups, if present in
final analogs, was performed at the end of the sequence.
Similarly, the O-benzyl-protecting group on the pheno-
lic sulfonamides n and t was removed at the last stage
via hydrogenolysis.
3. Results and discussion
Scheme 1. Reagents. Method A: LiSO3CF3, THF, epoxide 3. Method
B: ArSO2Cl,5 DIEA, cat. DMAP, THF. Method C: ArSO2Cl,5 DIEA,
CH2Cl2. Method D: (1) P4-phosphazene, epoxide 3, THF; or (2)
LiHMDS, epoxide 3, THF; or (3) (a) H2, Pd on BaSO4, EtOH; (b) P4-
phosphazene, epoxide 3, THF. Method E: (a) TFA; (b) DIEA,
(3R,3aS,6aR)-hexahydrofuro[2,3-b]furan-3-yl 4-nitrophenyl carbon-
ate,6 THF. Method F: (1) (a) TFA; (b) (3R,3aS,6aR)-hexahydrof-
uro[2,3-b]furan-3-yl 4-nitrophenyl carbonate,6 DIEA, CH2Cl2 or
CH3CN; or (2) HATU, DIEA, 3-methyl-N-[(methyloxy)carbonyl]-L-
valine,7 DMF; or (3) EDC, HOBT, DIEA, N2-(2-quinolinylcarbonyl)-
L-asparagine;7 or (4) (a) TFA; (b) (3R,3aS,6aR)-hexahydrofuro[2,3-
b]furan-3-yl 4-nitrophenyl carbonate,6 DIEA, CH2Cl2; (c) H2, 10% Pd
on C, 2 M NH3 in MeOH. Method G: (1) (a) ArSO2Cl,5 DIEA, cat.
DMAP, THF; (b) H2, 10%Pd on C, EtOH; or (2) ArSO2Cl,5 DIEA,
cat. DMAP, THF.
Our SAR strategy was to initially optimize the hydrox-
ylamine at P10 since this functionality was unprecedent-
ed in HIV PIs. We felt that with this accomplished, we
could then complement these results with known SAR
at P2 from published1 and internal data and finally ob-
tain completely optimal antiviral activity through sub-
stituents on the arylsulfonamide group. We elected to
keep the benzyl group at P1 fixed on the ethanolamine
backbone due to the arduous synthetic investment nec-
essary for these analogs.
Optimization of the P10 position of our inhibitors was
focused on lipophilic groups due to known elements of
the active site for this position of the protease.8 Table 1
summarizes a subset of prepared analogs that address
the intrinsic enzyme inhibition9 activity of the core
hydroxylamine substituent. The O-isopropyl analog a,
isosteric to the isobutyl group in amprenavir, was sur-
prisingly a relatively weak inhibitor with a Ki of
yields of intermediate 4 without requiring large excesses
of the hydroxylamine. Treatment of 4 with an arylsulfo-
nyl chloride (Method B) then provided inhibitors 5 with
all of the key functionalities in place at P10, P20, and P1,
albeit with the suboptimal BOC-amine group occupying
the P2 pocket. Arylsulfonyl chlorides were either avail-
able commercially or prepared through literature meth-
ods.5 This approach allowed rapid optimization of the
hydroxylamine at P1 and provided a useful intermediate
for further elaboration. Alternatively, intermediate 5
could also be generated by initial sulfonylation of
hydroxylamine 1 (Method C), followed by addition of
the N-alkoxysulfonamide anion, generated with a suit-
ably strong base (e.g., lithium bis(trimethylsilyl)amide
or P4-phosphazene), into epoxide 3 (Method D).
Table 1. Optimization of P10 N-alkoxysulfonamide substituent
Entrya
R
Ki (nM)b
Synthetic sequence (yields)c
C, D1 (69%, 60%)
a
14.0
Analogs with optimized substituents at P2 could be pre-
pared in two separate ways. Boc-deprotection of 5 with
TFA followed by acylation of the free amine provided
analogs 6, which allowed diversification at the P2 posi-
tion subsequent to fixing the P10 substituent (Method
F). Acylation protocols involved either reaction with
known activated carbonates6 or coupling of known car-
boxylic acids7 with EDC or HATU in DMF.
b
c
d
e
3.2
4.0
2.8
1.6
A, B (51%, 70%)
C, D2 (92%, 100%)
C, D2 (83%, 22%)
C, D1 (79%, 89%)
A final synthetic strategy involved Boc-deprotection of
intermediate 4, providing a diamine intermediate where
the primary amine could be selectively acylated over the
less reactive N-alkoxyamine (Method E) to provide
intermediates 7. Final analogs 6 were then provided
through sulfonylation of 7 with the appropriate aryl-
sulfonyl chloride (Method G). This sequence provided
our strategy for optimizing P20 after fixing substituents
at P2.
f
12.0
22.0
C, D2 (80%, 99%)
C, D1 (90%, 99%)
g
a All compounds were >95% pure by 1H NMR and HPLC.
b Ki, enzyme inhibition constant.9
c See Figure 1 for details.

3562
R. G. Sherrill et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3560–3564
14 nM. However, the analogous carbon analog has been
reported in the literature as having a Ki of 34 nM.10
Extension with a single methylene or two methyl groups
(analogs b and c, respectively) increased enzyme activity
3-fold. Direct substitution of cyclohexyl (d) and a cyclo-
pentyl (e) on the oxygen provided an almost 10-fold in-
crease in inhibitory potency relative to isopropyl.
However, activity was attenuated by either extending
the cycloalkyl moiety (f) or by substituting a phenyl
(g) for the cycloalkyl group. From these data, we were
able to conclude that O-cyclopentyl and O-cyclohexyl
alkoxysulfonamides would serve as preferred P10
substituents.
enzyme. An impressive antiviral activity of 24 nM was
recorded for the saquinavir-derived analog, whereas
the fused furan carbamate provided even further
improvements, with a single digit antiviral potency of
6 nM.
Finally, we directed our efforts toward optimization of the
P20 group. Given that the hexahydrofuro[2,3-b]furan-3-ol
carbamate at P2 displayed the optimal antiviral potency,
we elected to focus on this preferred substituent at this po-
sition. Similarly, the O-cyclopentyl and the O-cyclohexyl
substituents at P10 displayed the most favorable enzyme
potency in our study. We therefore focused on these two
substituents at P10 in our P20 study (Table 3).
Next, we addressed a small group of substituents at P2
that had historically given potent enzyme binding inhibi-
tion and antiviral potency, as measured in an MT4 cell
line (Table 2).11 The hydroxy tetrahydrofuran carba-
mate found in amprenavir led to subnanomolar enzyme
inhibitor h, although with an observed enzyme binding
constant roughly five times higher than that of the par-
ent drug. However, we were encouraged that the enzyme
potency translated into similar cellular antiviral activity
between the two scaffolds. The antiviral potency of h
was also roughly 5-fold less active than that of the mar-
keted drug. t-Butylglycine methylcarbamate analog i,
derived from atazanavir, displayed enzyme and antiviral
activities roughly comparable to those of amprenavir.
Without exception, the substituents that we chose at P20
displayed exceptional enzyme and antiviral activities
when combined with optimized P2 and P10 groups
(Table 3). Notably, only the combination of the P10
cyclohexyl group and the P20 4-aminophenylsulfonyl
group, analog s, displayed slightly attenuated enzyme
potency. However, a corresponding decrease in antiviral
potency was not observed. Each of the analogs in this
series displayed 620 nM antiviral activity in our
cell-based assay, which corresponded to a 7.5- to 50-fold
increase in antiviral potency relative to amprenavir.
Given the exceptional antiviral activity displayed by
these analogs, we screened a subset of compounds
against two multi-PI resistant viral strains12 (EP13 and
D545701) in MT4 cell lines in order to assess their po-
tential against resistant viral strains (Table 4). Our assay
cutoff of >240 nM reflected our estimation of a clinically
relevant acceptable antiviral activity. Notably, analogs p
and t displayed acceptable antiviral potency against
both mutant strains of virus.
Given sufficient enzyme potency, it was clear that we
could surpass the antiviral activity displayed by proto-
typical HIV PIs. Accordingly, the asparagine quinoline
carboxamide found in saquinavir and hexahydro-
furo[2,3-b]furan-3-ol carbamate P2 substituents generat-
ed inhibitor analogs j and k, respectively, with an
inhibitory concentration of less than 5 pM against the
Table 2. Optimization of P2
Entrya
R
Ki (nM)b
IC50 (nM)c
738
Synthetic sequence (yields)d
C, D1, F1 (69%, 60%, 87%)
h
0.19
i
0.090
<0.005
<0.005
0.04
203
24
C, D1, F2 (69%, 60%, 35%)
C, D1, F3 (69%, 60%, 41%)
C, D1, F1 (69%, 60%, 82%)
j
k
6
Amprenavir
150
a All compounds were >95% pure by 1H NMR and HPLC.
b Ki, enzyme inhibition constant.9
c IC50, antiviral inhibition against HIV-1 in MT-4 cell culture.11
d See Figure 1 for details.

R. G. Sherrill et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3560–3564
3563
Table 3. Optimization of P20
Entrya
R
R0
Ki (nM)b
IC50 (nM)c
17
Synthetic sequence (yields)d
C, D3, F1 (87%, 77%, 63%)
l
<0.005
m
n
<0.005
<0.005
<0.005
9
A, B, F4 (86%, 88%, 41%)
A, B, F4 (86%, 50%, 77%)
A, B, F1 (86%, 79%, 47%)
12
7
o
p
<0.005
7
A, B, F1 (86%, 89%, 68%)
q
r
s
<0.005
0.005
0.018
20
11
8
C, D2, F1 (83%, 22%, 27%)
A, E, G1 (36%, 72%, 82%)
A, E, G1 (36%, 72%, 74%)
t
<0.005
<0.005
<0.005
5
A, E, G1 (36%, 72%, 79%)
A, E, G1 (36%, 72%, 87%)
A, E, G1 (36%, 72%, 86%)
u
v
13
3
Amprenavir
0.04
150
a All compounds were >95% pure by 1H NMR and HPLC.
b Ki, enzyme inhibition constant.9
c IC50, antiviral inhibition against HIV-1 in MT-4 cell culture.11
d See Figure 1 for details.
key. In this case, bioavailability was enhanced in all spe-
cies via coadministration with the CYP 3A4 inhibitor,
ritonavir.13
Table 4. Resistance profile of selected analogs (IC50 nM in MT4 cells)
Entry
EP13
(fold change)
D545701
(fold change)
HIV-1
m
212 (24)
184 (26)
44 (9)
>240 (>27)
194 (28)
117 (23)
9
7
In summary, we have disclosed a novel series of N-alk-
oxy-arylsulfonamide-based HIV protease inhibitors that
display both exceptional enzyme inhibitory activity and
antiviral potency. Members of this series also demon-
strate antiviral activity against mutant viral strains. In
addition, one analog also demonstrated modest bio-
availability in both rat and dog without the need for
the added pharmacoenhancing properties of ritonavir.
p
t
5
u
107 (8)
100 (33)
440 (3)
>240 (>18)
>240 (>80)
>1000 (>7)
13
3
v
Amprenavir
150
Our final challenge remained achieving acceptable phar-
macokinetic profiles. As has been observed with most
HIV PIs, bioavailability of these analogs was uniformly
poor. However, one notable exception was found: ana-
log l demonstrated 11% oral bioavailability after dosing
intravenously (iv) and by gavage at 5 mg/kg in HW rats.
Encouraged by this modest level of plasma exposure, we
examined an identical oral dose and an iv dose of 1 mg/
kg in beagle dogs which produced a 16% bioavailability.
We were especially encouraged by these results in the
context that the second-generation HIV protease inhib-
itor lopinavir provides a modest 25% bioavailability in
rat and no intrinsic bioavailability in either dog or mon-
References and notes
1. Randolph, J. T.; DeGoey, D. A. Curr. Top. Med. Chem.
2004, 4, 1079.
2. Gatell, J. M. J. HIV Ther. 2001, 6, 95.
3. Chen, P.; Cheng, P. T. W.; Spegel, S. H.; Zahler, R.;
Wang, X.; Thottathil, J.; Barrish, J. C.; Polniaszek, R. P.
Tetrahedron Lett. 1997, 38, 3175.
4. (a) Grochowski, E.; Jurczak, J. Synthesis 1976, 10, 682; (b)
Kim, J. N.; Kim, K. M.; Ryu, E. K. Synth. Commun. 1992,
22, 1427.

3564
R. G. Sherrill et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3560–3564
5. 1,3-Benzodioxole-5-sulfonyl chloride and 2,3-dihydro-1,4-
E.; Homnick, C. F.; Nunberg, J.; Springer, J. P.; Huff, J.
R. J. Med. Chem. 1992, 35, 1685.
9. Toth, M. V.; Marshall, G. R. Int. J. Pept. Protein Res.
1990, 36, 544.
benzodioxin-6-sulfonyl chloride: Aikins, J. A.; Tao, E. V.
P. Eur. Pat. Appl. EP 254577 A1, 1988.4-[(Phenylmeth-
yl)oxy]benzenesulfonyl chloride: Toja, E.; Gorini, C.;
Zirotti, C.; Barzaghi, F.; Galliani, G. Eur. J. Med. Chem.
1991, 26, 403.
10. Ghosh, A. K.; Fidanze, S. J. Org. Chem. 1998, 63, 6146.
11. Daluge, S. M.; Purifoy, D. J. M.; Savina, P. M.; St. Clair, M.
H.; Parry, N. R.; Dev, I. K.; Novak, P.; Ayers, K. M.;
Reardon, J. E. Antimicrob. Agents Chemother. 1994, 38,
1590.
12. EP13 and D545701 viruses are multi-protease inhibitor-
resistant viruses that contain the following mutations
relative to the consensus sequence of wild type virus.
EP13: protease—M46I, L63P, A71V, V82F/L, and I84V;
no reverse transcriptase mutations. D545701: protease—
L10I, L19Q, K20R, E35D, M36I, S37N, M46I, I50V,
I54V, I62V, L63P, A71V, V82A, and L90M; reverse
transcriptase—E28K, K32E, V35I, T39S/T, E40D/V/Y/F,
K43E, and Y181Y/C.
13. Sham, H. L.; Kempf, D. J.; Molla, A.; Marsh, K. C.;
Kumar, G. N.; Chen, C. M.; Kati, W.; Stewart, K.;
Lal, R.; Hsu, A.; Betebenner, D.; Korneyeva, M.;
Vasavanonda, S.; McDonald, E.; Saldivar, A.; Wide-
burg, N.; Chen, X.; Niu, P.; Park, C.; Jayanti, V.;
Grabowski, B.; Granneman, G. R.; Sun, E.; Japour, A.
J.; Norbeck, D. W. Antimicrob. Agents Chemther. 1998,
42, 3218.
6. (3R,3aS,6aR)-Hexahydrofuro[2,3-b]furan-3-yl
4-nitro-
phenyl carbonate: Miller, J. F.; Furfine, E. S.; Hanlon,
M. H.; Hazen, R. J.; Ray, J. A.; Robinson, L.; Samano,
V.; Spaltenstein, A. Bioorg. Med. Chem. Lett. 2004, 14,
959; , 1-({[(3S)-Tetrahydro-3-furanyloxy]carbonyl}oxy)-
2,5-pyrrolidinedione:Kim, B. M.; Bae, S. J.; So, S. M.;
Yoo, H. T.; Chang, S. K.; Lee, J. H.; Kang, J. Org. Lett.
2001, 3, 2349.
7. 3-Methyl-N-[(methyloxy)carbonyl]-L-valine:
Xu,
Z.;
Singh, J.; Schwinden, M. D.; Zheng, B.; Kissick, T. P.;
Patel, B.; Humora, M. J.; Quiroz, F.; Dong, L.; Hsieh, D.;
Heikes, J. E.; Pudipeddi, M.; Lindrud, M. D.; Srivastava,
S. K.; Kronenthal, D. R.; Mueller, R. H. Org. Proc. Res.
Dev. 2002, 6, 323; , N2-(2-quinolinylcarbonyl)-L-aspara-
gine:Girijavallabhan, V. M.; Bennett, F.; Patel, N. M.;
Ganguly, A. K.; Dasmahapatra, B.; Butkiewicz, N.; Hart,
A. Bioorg. Med. Chem. 1994, 2, 1075.
8. Thompson, W. J.; Fitzgerald, P. M. D.; Holloway, M. K.;
Emini, E. A.; Darke, P. L.; McKeever, B. M.; Schleif, W.
A.; Quintero, J. C.; Zugay, J.; Tucker, T. J.; Schwering, J.