Inhibitors of hepatitis C virus NS3·4A protease 1. Non-charged tetrapeptide variants

Article abstract of DOI:10.1016/j.bmcl.2003.08.050

Tetrapeptide-based peptidomimetic compounds have been shown to effectively inhibit the hepatitis C virus NS3·4A protease without the need of a charged functionality. An aldehyde is used as a prototype reversible electrophilic warhead. The SAR of the P₁ and P₂ inhibitor positions is discussed.

Full text of DOI:10.1016/j.bmcl.2003.08.050

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4059–4063
.
Inhibitors of Hepatitis C Virus NS3 4A Protease 1.
Non-Charged Tetrapeptide Variants
Robert B. Perni,* Shawn D. Britt, John C. Court, Lawrence F. Courtney,
David D. Deininger, Luc J. Farmer, Cynthia A. Gates, Scott L. Harbeson,
Joseph L. Kim, James A. Landro, Rhonda B. Levin, Yu-Ping Luong, Ethan T. O’Malley,
JanosPitlik, B. Govinda Rao, Wayne C. Schairer, John A. Thomson, Roger D. Tung,
John H. Van Drie and Yunyi Wei
Vertex Pharmaceuticals Inc., 130 Waverly Street, Cambridge, MA 02139, USA
Received 13 June 2003; accepted 18 August 2003
.
Abstract—Tetrapeptide-based peptidomimetic compounds have been shown to effectively inhibit the hepatitis C virus NS3 4A
protease without the need of a charged functionality. An aldehyde is used as a prototype reversible electrophilic warhead. The SAR
of the P1 and P2 inhibitor positions is discussed.
# 2003 Elsevier Ltd. All rights reserved.
HCV infection hasreached epidemic proportion world-
wide and to date therapy optionsare limited and clinical
results are often unsatisfactory. Prevalence of the dis-
ease and currently available treatments have been
recently reviewed.1 Inhibition of the hepatitisC virus
NS3 4A protease has been an intense area of research
since the mid-1990sand numerousgroupshave reported
progress in the field.2ꢀ4
thisapproach, but these compoundsrequire carboxylate
residues to attain reasonable potency.9 Deletion of the
acidic residues at P5 and P6 results in a dramatic
decrease in binding affinity. We herein report a series of
potent tetrapeptide inhibitorsthat do not require car-
boxyl groupsfor binding potency and therefore offer
improved potential for cell potency.
.
The shallow, hydrophobic, highly flexible binding
.
pocket of the NS3 4A protein represents a formidable
challenge to drug design and only recently has a com-
pound been reported to enter human testing.5 Inclusion
of terminal charged groups(e.g., carboxylates) on either
or both sides of the enzyme active site has proven to
provide substantial biochemical potency. Indeed, pro-
.
duct inhibition of the NS3 4A proteolytic reaction have
formed the basis of many design efforts.6 However,
charged groups often result in compromising cellular
penetration and in vivo pharmacokinetics.7 Reversible
covalent binding to the catalytic serine is an alternative
to electrostatic binding that has been utilized in other
serine protease inhibitor series.8 A series of peptide
aldehydeshave recently been reported which exploits
We and othershave recently decsribed the effectsof
truncating substrates and substrate-derived competitive
inhibitorsfrom the optimal recognition length of ten
*Corresponding author. Tel.: +1-617-444-6237; fax: +1-617-444-
6766; e-mail: robert_perni@vrtx.com
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.08.050

4060
R. B. Perni et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4059–4063
amino acids down to a four to six residue length.6b,10
Based on this work, we synthesized inhibitor 1 (Ki=0.89
mM).4 Removal of the P5 and P6 amino acidsand cap-
ping with a neutral heterocycle provided 2, which pos-
sesses reasonable inhibitory activity (Ki=12 mM) asa
prototype tetrapeptide inhibitor. Thislatter compound
served as the basis for our subsequent optimization
efforts.
and subsequently oxidized with DMP to give the final
products, 23–26. The P1 position was usually epimerized
during the oxidation step.
Compoundswere evaluated for enzyme-binding activity
(Ki) via colorimetric monitoring of the hydrolysis of an
HCV NS5A-pNA substrate.10 The P2 SAR (Table 1)
resists any simple interpretation. In a related series of
compounds(data not hsown)
l isomers were always
Compound 2 wasprepared from commercially available
4
more potent than d isomers consistent with the P1 geo-
metry of the natural substrate. Comparing compounds
2 and 8, and compounds 14 and 15, one may be tempted
to infer that increased steric bulk and/or p-density
improvesaffinity. However, the phenyl on compound 10
represents one of the smallest groups, yet yields potency
comparable to the naphthyl. The orientation of the P2
group clearly influencesaffinity, asevidenced by com-
pounds 11 (Ki=2.1 mM) and 12 (Ki=7.4 mM), and
compounds 15 (Ki=1.9 mM) and 16 (Ki=0.4 mM). It
appearsthat thisinfluence isdue to interactionswith the
side chains forming the S2 pocket, Arg181, Asp107, and
His83, the latter two part of the catalytic triad. Further
analysis of these interactions with structure-based
methods should allow us to refine our understanding of
them.
4-benzyloxyproline aswe have previously reported.
A
related method for the solid phase synthesis of peptidic
aldehydeshasbeen published by others. 11 P2 derivatives
were prepared via solid-phase methodology utilizing a
semicarbazone linker. P2 ethers, esters and carbamates
were prepared ashsown in Scheme 1 using standard
methodology followed by removal of the solid support
under acidic conditions.
Carbamate 19 and benzylic ethers 8 and 9 were pre-
pared by preassembly of the P2 group (5a and 5b) via
the chemistry shown in Scheme 2. The proline inter-
mediates, 5a and 5b, were then elaborated to the com-
plete tetrapeptide via standard peptide coupling
methodology.
P1 modified analoguesof compound 2 were prepared as
shown in Scheme 3. Tripeptide 7 wasprepared via
standard peptide coupling chemistry. The carboxylic
acid 7 wascoupled to the corresponding amino alcohol
The linkage of the P2 substituent to the proline also has
an effect. In general, esters provide more potency rela-
tive to ether linked substituents. Unfortunately, esters
were found to be too hydrolytically unstable to be use-
Scheme 1. (a) PPh3, DEAD, ROH; (b) AcOH, THF, HCHO or 50% TFA/DCM; (c) RCOCl, iPrNEt2, CH2Cl2 DCM or RCOOH, HOBt, HBTU,
NMP; (d) RNCO or RCOOH, (PhO)2P(O)N3.
Scheme 2. (a) NaH, RCH2Br or RCH2Cl, DMF/THF; (b) LiOH, H2O, THF; (c) carbonyldiimidazole, tetrahydroquinoline, CH2Cl2.

R. B. Perni et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4059–4063
4061
Scheme 3. (a) Fmoc-Val-NCA, iPr2NEt, CH2Cl2; (b) Et2NH, CH3CN; (c) Fmoc-Val-NCA, iPr2NEt, CH2Cl2; (d) Et2NH, CH3CN; (e) pyr-
azinecarboxylic acid, HOBt, EDC, iPr2NEt, CH2Cl2; (f) 4 N HCl/dioxane; (g) amino alcohol, EDC, HOBt, CH2Cl2; (h) DMP, CH2Cl2 or Swern.
ful beyond determining inhibition constants. Simple
primary carbamatesasan alternative to esters(e.g.,
.
Table 1. Inhibition of the HCV NS3 4A protease enzyme by tetra-
peptide P2 variants
17–
18 and 20–22) were prepared and were found to be less
potent relative to ester 16. However, carbamates 20–22
were essentially equipotent to esters 14 and 15. Electron
donating, or electron withdrawing groupsappear to
have little influence on binding affinity. A naphthyl car-
bamate, 18, however shows improved potency relative
to the phenyl carbamate 17.
To address the hydrolytic instability of the ester groups,
the tetrahydroisoquinolinyl (THIQ) carbamate 19 was
synthesized. The carbamate nitrogen is incorporated
into the ring system of 19 to mimic ester 16 in termsof
overall steric demand. Figure 1 shows the X-ray struc-
turesof 16 and 19 in identical orientations. While the
scaffold itself binds identically in both cases, including
the OC¼O link, the ringsof the naphthyl and THIQ do
not. The THIQ ringsof 19 are rotated approximately
180ꢁ. Additionally, the puckered aliphatic THIQ ring
tiltsthe ringsaway from the psace occupied by the
naphthyl group of 16. The low RMS deviation for the
molecule excluding the bicyclic rings(RMSD=0.3 A ,
maximum interatom variability 0.6 A) impliesthat the
S2 pocket accommodatesthe different orientationsbut
doesnot force the orientation on them. If the protein
required the change in conformation of the P2 ring sys-
tems it would likely disturb the scaffold backbone as
well. Inhibitor 19 did show improved binding relative to
primary carbamatesand similar potency to 16 (Ki=0.89
mM vs Ki=0.40 mM, respectively).
R2
P1 Stereochemistrya
Ki (mM)
2
8
9
CH2C6H5
CH2-1-Naphthyl
CH2-2-Naphthyl
C6H5
1-Naphthyl
2-Naphthyl
8-Quinolinyl
COC6H5
CO-1-Naphthyl
CO-2-Naphthyl
CONHC6H5
CONH-1-Naphthyl
l
12
2.9
1.7
3.9
2.1
7.4
>50
5.8
1.9
d,l
d,l
d,l
d,l
l
10
11
12
13
14
15
16
17
18
d,l
d,l
d,l
d,l
l
0.40
22.3
12.5
l
19
20
d,l
d,l
0.89
2.5
21
22
l
3.2
7.9
The S1 specificity pocket is defined by the Leu135,
Phe154, and Ala157 side chains and allows only for the
inclusion of relatively small, preferably hydrophobic P1
substituents. This is the specificity pocket9 that provides
excellent selectivity versus the clotting cascade enzymes
such as thrombin, kallikrein and factor Xa all of which
require basic substitution at P1. The consensus sequence
for all trans cleavage sites, NS4A-4B, NS4B-5A and
NS5A-5B incorporatesa cytseine at P 1. The incor-
poration of an electrophilic warhead in an inhibitor is
d,l
aWhere indicated, single isomers were separated from a d,l mixture by
HPLC.

4062
R. B. Perni et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4059–4063
Figure 1. X-ray comparison of 16 (left, resolution 2.9 A) and 19 (right, resolution 2.8 A).
single isomer. Given that the trifluoroethyl group pro-
.
Table 2. Inhibition of the HCV NS3 4A protease enzyme by tetra-
peptide P1variants
videsno significant advantage over the n-propyl group,
norvaline waschosen asthe P 1 substituent with which to
pursue optimization of the scaffold.12 Not surprisingly
the geminal substitution (25) at P1 significantly reduces
binding affinity while the introduction of an oxygen
atom (26) also is substantially detrimental to activity.
These data are consistent with those reported for
charged, non covalent P1 termini.13,14
Ra
Ki (mM)
Incorporation of a single charged P4 terminusaffor-
ded 2615 (Ki=0.11 mm), which demonstrated
improved potency relative to the bis-carboxylate (P5
and P6) 1 (Ki=0.89 mM). Lengthening or shortening
the chain length haslittle effect on potency (data not
shown).
2
12
3.1
23
24
25
26
3.8
>50
>50
aCompoundsare single P isomers except for 23 which wasprepared
asmixture of epimerswhich were not separated.
1
consequently incompatible with a cysteine residue at P1
due to intermolecular and possibly intramolecular reac-
tivity. Table 2 summarizes the P1 SAR for a series of
small substituents replacing the cysteine side chain with
non-reactive groups. Replacing the cysteine sulfur with a
carbon atom results in an ethyl (amino butyric acid deri-
vative, e.g, 2) P1 side chain, The ethyl P1 side-chain while
an effective cysteine surrogate is not the optimal group
for the S1 subsite. Increasing the hydrophobicity of the
side-chain in the form of trifluoroethyl (23 asan epimeric
P1 mixture) or propyl (norvaline, 24) substituents resulted
in more potent inhibitors. Though expensive, 4,4,4-tri-
fluoro-2-aminobutyric acid iscommercially available as
a racemate. Norvaline iscommercially available as
In summary, using an aldehyde as a prototype electro-
philic covalent warhead, a series of P1 and P2 analogues
were prepared asan initial step in the optimization of a
.
tetrapeptide HCV NS3 4A protease inhibitor scaffold.
Small non-polar substituents at P1 were found to be
optimal, consistent with results previously obtained
with non-covalent peptidic inhibitors.13,14 Large hydro-
phobic substituents appended to the 4-position of a
proline residue at P2 in an appropriate vector were
found to provide substantial binding recognition by the
protease, resulting in the discovery of a sub-micromolar
tetrapeptide inhibitor. We believe thisimportant classof
inhibitorsoffersgreat potential for the treatment of
HCV infection in humans.

R. B. Perni et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4059–4063
4063
References and Notes
T. J.; Schostarez, H. J.; Palazuk, B. J.; Bleasdale, J. E. Bioorg.
Med. Chem. Lett. 2003, 13, 971.
8. Leung, D.; Abbenante, G.; Fairlie, D. P. J. Med. Chem.
2000, 43, 305.
1. Tam, S.-L.; Pause, A.; Shi, Y.; Sonenberg, N. Nature Rev.
2002, 1, 867. McHutchison, J. G.; Patel, K. Hepatology 2002,
36, S245.
9. Attwood, M. R.; Bennet, J. M.; Campbell, A. D.; Canning,
G. G. M.; Carr, M. G.; Conway, E.; Dunsdon, R. M.;
Greening, J. R.; Jones, P. S.; Kay, P. B.; Handa, B. K.; Hurst,
D. N.; Jennings, N. S.; Jordon, S.; Keech, E.; O’Brien, M. A.;
Overton, H. A.; King-Underwood, J.; Raynham, T. M. Anti-
viral Chem. Chemother. 1999, 10, 259.
10. Landro, J. A.; Raybuck, S. A.; Luong, Y. P. C.; O’Malley,
E. T.; Harbeson, S. L.; Morgenstern, K. A.; Rao, B. G.;
Livingston, D. Biochemistry 1997, 36, 9340. Detailsof the
hexapeptide to tetrapeptide scaffold truncation are the subject
of a manuscript currently in preparation.
11. Ede, N. J.; Eagle, S. N.; Wickham, G.; Bray, A. M.;
Warnes, B.; Shoemaker, K.; Rosenberg, S. J. Pept. Sci. 2000,
6, 11.
12. Further optimization of the tetrapeptide scaffold descri-
bed in this paper is the subject of manuscripts currently in
preparation.
2. (a) Narjes, F.; Koch, U.; Steinkuhler, C. Exp. Op. Inves.
New Drugs 2003, 12, 153. Perni, R. B.; Kwong, A. D. In Pro-
gress in Medicinal Chemistry, King F. D., Oxford, A. W., Eds.
Elsevier Science B.V.: Amsterdam, 2002; Volume 39, p 215
3. Recent representative peptide inhibitor papers: (a) Llinas-
Brunet, M.; Bailey, M.; Fazal, G.; Ghiro, E.; Gorys, V.;
Goulet, S.; Halmos, T.; Maurice, R.; Poirier, M.; Poupart, M.-
A.; Rancourt, J.; Thibeault, D.; Wernic, D.; Lamarre, D.
Bioorg. Med. Chem. Lett. 2000, 10, 2267. (b) Han, W.; Hu, Z.;
Jiang, X.; Decicco, C. P. Bioorg. Med. Chem. Lett. 2000, 10,
711. (c) Johansson, A.; Poliakov, A.; Akerblom, E.; Linde-
berg, G.; Winiwarter, S.; Samuelsson, B.; Danielson, U. H.;
Hallberg, A. Bioorg. Med. Chem. 2002, 12, 3915. (d) Nizi, E.;
Koch, U.; Ponzi, S.; Matassa, V. G.; Gardelli, C. Bioorg. Med,
Chem. Lett. 2002, 12, 3325.
4. Tung, R. D.; Harbeson, S. L.; Deininger, D. D.; Murcko,
M. A.; Bhisetti, G. R.; Farmer, L. J. PCT WO 98/17679,
1998.
13. Llinas-Brunet, M.; Bailey, M.; Deziel, R.; Fazal, G.;
Gorys, V.; Goulet, S.; Halmos, T.; Maurice, R.; Poirier, M.;
Poupart, M.-A.; Rancourt, J.; Thibeault, D.; Wernic, D.;
Lamarre, D. Bioorg. Med. Chem. Lett. 1998, 8, 2719.
14. For recent reports of novel P1 substituents see: (a) Tsan-
trizos, Y.; Bolger, G.; Bonneau, P.; Cameron, D. R.; Gou-
dreau, N.; Kukolj, G.; LaPlante, S. R.; Llina-Brunet, M.; Nar,
H.; Lamarre, D. Angew. Chem. Int. Ed. 2003, 42, 1355. (b)
Poupart, M.-A.; Cameron, D. R.; Chabot, C.; Ghiro, E.;
Goudreau, N.; Goulet, S.; Poirier, M.; Tzantrizos, Y. J. Org.
Chem. 2001, 66, 4743. (c) Narjes, F.; Koehler, K. F.; Koch,
U.; Gerlach, B.; Colarusso, S.; Steinkuler, C.; Brunetti, M.;
Altamura, S.; DeFrancesco, R.; Matassa, V. G. Bioorg. Med.
Chem. Lett. 2002, 12, 701.
5. Lamarre, D.; Bailey, M.; Bolger, G.; Cameron, D.; Cartier,
M.; Faucher, A.-M.; Goudreau, N.; Kukolj, G.; Lagace, L.;
Pause, A.; Rancourt, J.; Thibeault, D.; Tsantrizos, Y.; Llinas-
Brunet, M. Hepatology 2002, 36, 92A (Abstract #464). Hin-
richsen, H.; Benhamou, Y.; Reiser, M.; Sentjens, R.; Wede-
meyer, H.; Calleja, J. L.; Forns, X.; Cronlein, J.; Nehmiz, G.;
Steinmann, G. Hepatology 2002, 36, 379A (Abstract #866).
6. Product based inhibitors: (a) Steinkuhler, C.; Basiol, G.;
Brunetti, M.; Urbani, A.; Koch, U.; Cortese, R.; Pessi, A.; De
Francesco, R. Biochemistry 1998, 37, 8899. (b) Ingallinella, P.;
Altamura, S.; Bianchi, E.; Taliani, M.; Ingenito, R.; Cortese,
R.; De Francesco, R.; Steinkuhler, C.; Pessi, A. Biochemistry
1998, 37, 8906.
15. Compound 26 wasprepared according to the route shown
in Scheme 1. The terminal carboxylate wasintroduced into the
scaffold by using the unprotected dicarboxylic acid in large
excess under standard coupling conditions.
7. For a recent example of this phenomenon see: Larsen, S. D.;
Stevens, F. C.; Lindberg, T. J.; Bodnar, P. M.; O’Sullivan,