Lysine sulfonamides as novel HIV-Protease inhibitors: Optimization of the Nε-acyl-phenyl spacer

Article abstract of DOI:10.1016/j.bmcl.2003.09.058

A series of Nα-isobutyl-Nα-arylsulfonamido-(Nε acyl) lysine and lysinol derivatives were prepared and evaluated as inhibitors of HIV protease and wild type virus. A simple original synthesis was devised to form Nα-(arylsulfonamide)-Nα-isobutyl lysine, whi

Full text of DOI:10.1016/j.bmcl.2003.09.058

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4289–4292
Lysine Sulfonamides as Novel HIV-Protease Inhibitors:
Optimization of the N"-Acyl-Phenyl Spacer
Brent R. Stranix,*,y Gilles Sauve,{ Abderrahim Bouzide, Alexandre Cote,
Guy Sevignyy and Jocelyn Yelley
Pharmacor Inc., 535 Cartier West, blvd., Laval, Quebec, Canada H7V 3S8
Received 18 July 2003; revised 24 September 2003; accepted 27 September 2003
Abstract—A series of Na-isobutyl-Na-arylsulfonamido-(Ne acyl) lysine and lysinol derivatives were prepared and evaluated as
inhibitors of HIV protease and wild type virus. A simple original synthesis was devised to form Nꢀ-(arylsulfonamide)-Nꢀ-isobutyl
lysine, which could be easily acylated with carboxylic acids at the N" position. A two-atom spacer was found to be optimal between
this acyl group and a phenyl yielding compounds of sub-nanomolar potency on purified enzyme.
# 2003 Elsevier Ltd. All rights reserved.
One of the main strategies applied to address the AIDS
pandemic consists of the inhibition of the virally enco-
ded enzyme human immunodeficiency virus (HIV)
aspartyl protease.1À3 In particular, much success has
been obtained through the use of small molecules,
which mimic the hydrolytic transition state of the
enzyme’s natural peptide substrate. The most commonly
encountered motif for these peptidomimetics is the
hydroxyethylene bioisostere.4 Although protease inhibi-
tors have radically improved the life of AIDS patients
and contributed in large part to the success of highly
active anti-retroviral therapy (HAART), new problems
have recently been identified. Most importantly, the
rapid emergence of several viral strains resistant to one
or more of the drugs currently available for the treat-
ment of AIDS has now become a major issue in the
treatment of HIV infection.5 This drawback, along with
long term toxicity, has generated a pressing need for a
next generation of protease inhibitors. Some of the
requirements include designs containing original struc-
tures and simple means of preparation. Very few HIV
protease inhibitors containing a non-peptidic amino
acid backbone are known.6À8 In the course of our studies
to design and develop inhibitors of HIV-protease,9,10 we
screened commercially available protected amino acids
and found two distinct series of amino acids, which
included lysine, that gave significant enzymatic inhibi-
tion.11 Further development of this lysine series per-
mitted us to identify several promising low molecular
weight leads with a general structure as shown in Figure
1. A synthetic route for preparing the enantiomerically
pure12 trisubstituted l-lysine derivatives was devised using
commercially available l-Lys("-Cbz)-OBn (Scheme 1).
Reductive alkylation with, in particular, isobutyr-
aldehyde, under mild reaction conditions gave good
yields of the monoalkylated compounds. Addition of
aryl sulfonyl chloride gave intermediate aryl sulfonamide
2a, which after flash chromatography, was easily
deprotected using Pd/C catalyzed hydrogenation to give
intermediates 3a. Acylation of the liberated e-amino
acid using Schotten–Baumann conditions was easily
accomplished with a series of activated carboxylic acids.
For those activated carboxylic acids that were not
already commercially available, activation of the car-
boxylic acid was done with 1,10-carbonyldiimidazole
(CDI), DCC/HOSu or EDAC/HOBt prior to the addi-
tion to the aqueous amino acid solution. The same
conditions were used to obtain the 4-aminobenzene-
sulfonyl derivatives, using the 4-nitrobenzenesulfonyl
*Corresponding author. Tel.: +1-514-496-1735; fax: +1-514-496-
4857; e-mail: bstranix@procyonbiopharm.com
yPresent address: Procyon Biopharma Inc., 6100 Royalmount, Mon-
treal, Qc, Canada H4P 2R2.
{Present address: INRS-Institut Armand-Frappier, 531, des Prairies
Blvd, Building 27, Office D-103, Laval, Quebec, Canada H7V 1B7.
Figure 1.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.09.058

4290
B. R. Stranix et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4289–4292
LC–MS for structure confirmation and enantiomeric
purity.13 The novel derivatives were evaluated as HIV
protease inhibitors. The activities were measured using
purified HIV protease14 in vitro, and are tabulated in
Table 1. The anti-protease activity of compounds 11a–
11d, suggest that the spacer length between the carbox-
amide and the aryl group is optimal at two methylenes.
Locking the conformation with rigid groups such as
trans-ethylene and trans-cyclopropane, such as in 11e–f,
appears to diminish the inhibitory properties. The
addition of two aryl groups also appears to be deleter-
ious (11g). By adding a bulky tert-butyl group in this
alpha position led to a complete loss of activity (11h).
We then proceeded to replace the dimethylene spacer
with isosteres containing N, O and S heteroatoms.
Insertion of a –O –CH2 –spacer between the carboxamide
and the phenyl group, as in 11i, abolishes the in vitro
activity whereas, conversely, the –CH2 –O –spacer of 11j
retains a significant amount of activity. The addition of
a hydroxyl group in the alpha position was obtained by
the CDI mediated condensation of l-phenyl lactic acid
did not significantly change the Ki and the addition of a
ketone, through a similar condensation of pyruvic acid
led to a slight loss of activity as shown for 11k and 11l.
Similarly, the –CH2 –S –spacer of 11m shows identical
activity as well as its oxidized analogue 11n. In the case
of the nitrogen analogues, a similar yet not as significant
pattern is seen. Substitution of the alpha carbon by
Scheme 1. (a) Isobutyraldehyde, NaCNBH3, MeOH, pH 4; (b)
ArSO2Cl, DIPEA, DCM; (c) H2 Pd/C, EtOH; (d) R-CO-X, THF/1 M
K2CO3.
(nosyl) group, in step b to give 2b, which was subse-
quently reduced in step c to the aniline 3b. The non-
desired acylation of this aniline during the coupling
reaction was rarely observed giving compounds of
general structure 4b exclusively. A slight modification to
the synthetic route was made in order to synthesize the
benzylic ureas.
l-Lys(Cbz)-OMe was used as the starting material,
which underwent identical chemical modifications as in
the Scheme 1 for the first three steps. Activation with
CDI gave a pendant isocyanate, which could be modified
with benzylic amines to give, after alkaline hydrolysis of
the methyl ester 8, compounds 11l, 11n and 11p (Scheme
2). N-Substituted-N-phenylglycines were synthesized by
reacting three equivalents of n-BuLi with N-phenyl glycine
in THF followed by an alkylating agent such as benzyl-
bromide (Scheme 3). Attempts at reductive aminations
of N-Ph-Gly were unsuccessful. Coupling of amino
acids was effected with the Boc protected l-Phe and l-
N-Me-Phe, followed by deprotection using TFA to yield
compounds 11q and 11r.
Table 1. HIV aspartyl protease inhibition constants for compounds
11a–v
Compd
Ki (nM) Compd
Ki (nM)
11a
>300
11l
14
The resulting compounds, purified to >95% by pre-
parative RP-HPLC, were characterized by NMR and
11b
11c
11d
11e
11f
>300
11m
11n
11o
11p
11q
18
20
7.8
32
12
6.9
>300
>300
>300
11g
171
11r
55
Scheme 2. (a)–(c) See Scheme 1; (d) CDI, amine DMF; (e) 1:1 1 M
NaOH/THF.
11h
11i
11j
>300
>300
18
11s
11t
11u
0.72
8.0
1.5
11k
6.2
11v
2.1
Scheme 3. (a) 3 equiv n-BuLi, R-Cl (Br).

B. R. Stranix et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4289–4292
4291
nitrogen leads to a compound of lesser activity than the
beta-substituted carbon as seen in 11o–p. Methylation
of the nitrogen atoms in 11q–r does not change sig-
nificantly the activity, whereas benzylation dramatically
improves the potency, especially in the case of 11s. The
addition of a pendant amino group in the alpha position
also leads to some improvement, as well as the methyl-
ated derivative as seen for 11u and 11v. Although
encouraged by the in vitro potency of 11s, this com-
pound did not show any activity when tested in
vivo.15,16 We hypothesized that removing the negative
charge, inherent to the carboxyl group, could improve
the potency in cell culture while retaining most of its
anti-protease activity. We esterified intermediate 3a and
reduced this ester to the corresponding alcohol with
LiAlH4 without loss of enantiomeric purity.17 This
transformation led to a compound of lesser potency in
vitro but which showed some modest anti-viral proper-
ties in cell culture. We carried out some slight modifi-
cations on 11o and its 4-aminobenzene sulfonamide
analogue by varying the R group and adding some
substituents on the N-benzyl moiety. Extension of the
spacer length in 13b by one methylene resulted in a sig-
nificant loss of activity. Deletion of this spacer gave
product 13d which retained the anti-protease and anti-
viral properties. Replacement of the benzyl group with
isopropyl in 13c conserved some activity and potency in
the whole-cell assay. Electron withdrawing substituents
on the N-benzyl moiety such as nitro also improved
potency on both the enzyme and viral culture as shown
in 13h–j and 14h–j. Notably, the para-nitro compound
14h showed a Ki of 0.6 nM. Also, both the meta and
ortho substituted compounds showed good enzyme
inhibition as well as good potency in the anti-viral
assay. Compounds 14k–m bearing fluoride substituents
showed some improved inhibition of the enzyme but
some loss of activity in cell culture. Addition of an
electron donating substituent such as para-methoxy (13e
and 14e) showed no improvement both in vitro and in
whole-cell assay. However, adding two methoxyls (13f–
g and 14f–g) yielded products of great potency, espe-
cially when the tosyl group was replaced by a more
water soluble para-aminobenzenesulfonyl group. The
meta-anilines derivatives 13n and 14n gave a slight loss
Table 2. HIV aspartyl protease inhibition constants and anti-viral potencies for compounds 13a–n and 14a–n
Compd
R
13
14
Compd
R
13
14
Ki (nM) EC50 (nM) Ki (nM) EC50 (nM)
Ki (nM) EC50 (nM) Ki (nM) EC50 (nM)
a
7.6
4200
2.4
ND
h
ND
0.59
540
b
c
52
>104
ND
ND
ND
ND
i
j
1.4
3.9
730
1.2
1.5
339
341
16
12
1100
3500
d
e
6700
ND
2.4
ND
ND
k
l
ND
ND
ND
ND
2.3
2.3
2200
1000
7.3
>104
f
2.4
3.5
760
1.6
2.9
340
366
m
ND
ND
1.1
6.9
1100
g
1300
n
7.6
3000
524
ND, not determined.

4292
B. R. Stranix et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4289–4292
acknowledge Dr. F. Dahani for his analysis of com-
pound purity.
References and Notes
Scheme 4. (a) MeOH, TMS–Cl; (b) LiAlH4 THF.
1. Tomasselli, A. G.; Heinrikson, R. L. Biochim. Biophys.
Acta 2000, 1477, 189.
of inhibitory properties on the protease in vitro but still
showed modest anti-viral properties. The substitution
pattern of the benzylic portion of this class of molecules
would seem to be crucial to the anti-viral activity but
retains very similar activities on the purified enzyme in
vitro. This observation could be explained by the better
accessibility of such substituted benzyl phenylglycines to
the viral enzyme when tested in whole-cell assays, due to
factors such as solubility and cell penetration (Table 2,
Scheme 4).
2. Vella, S. Handb. Exp. Pharmacol. 2000, 140, 23.
3. Eron, J. J., Jr. Clin. Infect. Dis. 2000, 30, S160.
4. Babine, R.; Bender, S. L. Chem. Rev. 1997, 97, 1359.
5. Swanstrom, R.; Erona, J. Pharmacol. Ther. 2000, 86, 145.
6. Marastoni, M.; Bortolotti, F.; Salvadori, S.; Tomatis, R.
Arzneim.-Forsch. 1998, 48, 709.
7. Marastoni, M.; Bazzaro, M.; Bortolotti, F.; Tomatis, R.
Arzneim.-Forsch. 2000, 50, 564.
8. Rutenber, E. E.; De Voss, J. J.; Hoffman, L.; Stroud, R. M.;
Lee, K. H.; Alvarez, J.; McPhee, F.; Craik, C.; Ortiz de Mon-
tellano, P. R. Bioorg. Med. Chem. 1997, 5, 1311.
9. Stranix, B. R.; Sauve, G.; Bouzide, A.; Cote, A.; Berube,
G.; Soucy, P.; Zhao, Y.; Yelle, J. US Patent No. 6,506,786;
Pharmacor Inc: US, 2002.
10. Bouzide, A.; Sauve, G.; Stranix, B. R.; Sevigny, G.; Yelle,
J. US Patent No. 6,455,587; Pharmacor Inc.: US, 2002.
11. Sauve, G.; Bouzide, A. unpublished results.
In summary, simple and potent anti-HIV protease
compounds were fashioned from the amino acid l-
lysine. A straightforward and original synthesis was
devised to form Na-(arylsulfonamide)-Na-isobutyl
lysine, which could be easily acylated with carboxylic
acids in the Ne position. A two-atom spacer was found
to be optimal between this acyl group and a phenyl with
or without aryl substituents. In particular, the N-ben-
zylated derivatives of N-phenyl glycinamides gave
highly potent derivatives in vitro. Reduction of the
pendant carboxylate to the corresponding alcohol gave
a novel class of compounds, which retained the in vitro
inhibition and showed potent inhibition in anti-viral
assays.
12. Hiebert, C. K.; Kopp, W. C.; Richerson, H. B.; Barf-
knecht, C. F. J. Med. Chem. 1988, 31, 2022.
13. The carboxylic acid 11a was derived with both R and S-1-
1
phenethylamine to form the diastereomeric amides. H NMR
showed a diastereomeric purity of greater than 95%.
14. Matayoshi, E. D.; Wang, G. T.; Krafft, G. A.; Erickson, J.
Science 1990, 247, 954.
15. Japour, A. J.; Mayers, D. L.; Johnson, V. A.; Kuritzkes,
D. R.; Beckett, L. A.; Arduino, J. M.; Lane, J.; Black,
R. J. P. S.R.; D’Aquila, R. T. Antimicrob. Agents Chemother.
1993, 37, 1095.
16. Pauwels, R.; Balzarini, J.; Baba, M.; Snoeck, R.; Schols,
D.; Herdewijn, P.; Desmyter, J.; Clercq, E. D. J. Virol. Meth-
ods 1988, 20, 309.
17. Compound 14f was analyzed by H NMR (acetone-d6) in
the presence of 0.9 equiv Eu(hfc)3 and showed an enantiomeric
purity of >95%.
Acknowledgements
1
We thank A. Dubois, P. Soucy Y. Zhao and B. Tian for
their participation in the biological analyses. We also