Design, synthesis, and evaluation of inhibitors of Norwalk virus 3C protease

Article abstract of DOI:10.1016/j.bmcl.2011.07.016

The first series of peptidyl aldehyde inhibitors that incorporate in their structure a glutamine surrogate has been designed and synthesized based on the known substrate specificity of Norwalk virus 3C protease. The inhibitory activity of the compounds wi

Full text of DOI:10.1016/j.bmcl.2011.07.016

Bioorganic & Medicinal Chemistry Letters 21 (2011) 5315–5319
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, and evaluation of inhibitors of Norwalk virus 3C protease
Kok-Chuan Tiew a, Guijia He a, Sridhar Aravapalli a, Sivakoteswara Rao Mandadapu a,
Mallikarjuna Reddy Gunnam a, Kevin R. Alliston a, Gerald H. Lushington b, Yunjeong Kim c,
Kyeong-Ok Chang c, William C. Groutas a,
⇑
a Department of Chemistry, Wichita State University, Wichita, KS 67260, USA
b Molecular Graphics and Modeling Laboratory, The University of Kansas, Lawrence, KS 66045, USA
c Department of Diagnostic Medicine/Pathobiology, College of Veterinary Medicine, Kansas State University, Manhattan, KS 66506, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The first series of peptidyl aldehyde inhibitors that incorporate in their structure a glutamine surrogate
has been designed and synthesized based on the known substrate specificity of Norwalk virus 3C prote-
ase. The inhibitory activity of the compounds with the protease and with a norovirus cell-based replicon
system was investigated. Members of this class of compounds exhibited noteworthy activity both in vitro
and in a cell-based replicon system.
Received 27 May 2011
Revised 28 June 2011
Accepted 6 July 2011
Available online 14 July 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Inhibitors
Peptidyl aldehydes
Norwalk virus
3C protease
Transition state inhibitors
Noroviruses are a leading cause of food-borne and water-borne
non-bacterial acute gastroenteritis.1 Norovirus infections consti-
tute an important health problem with an estimated 23 million
cases of gastroenteritis occurring annually in the US, causing
50,000 hospitalizations and 300 deaths.2 There are currently no
effective vaccines or antiviral therapeutics for the treatment of
norovirus infection.
alone8,9 or covalently-bound to an inhibitor, a peptidyl Michael
acceptor, have been reported.7
Norovirus 3C protease plays an essential role in virus replica-
tion, consequently, orally-bioavailable drug-like agents that inhibit
the 3C protease are of value as potential antiviral therapeutics. We
describe herein the results of preliminary studies related to the
inhibition of Norwalk virus 3C protease by a series of peptidyl alde-
hyde inhibitors (Fig. 1).
Initial design considerations included the use of a glutamine
surrogate10 for optimal synthetic tractability and design flexibility
(vide infra). Furthermore, our overarching goal was to identify a
suitably-functionalized di-peptide or tri-peptide inhibitor that
Noroviruses are small enveloped viruses of the Caliciviridae fam-
ily.3 The genome of the Norwalk virus, a prototype of noroviruses, is
comprised of a single-stranded, positive sense RNA molecule of ꢀ7.7
Kilo bases that consists of three open reading frames (ORFs) that en-
code a 200 kDa polyprotein (ORF1), a major capsid protein VP1
(ORF2), and a small basic protein VP2 (ORF3). The mature polypro-
tein is co- and post-translationally processed by a virus-encoded
protease to generate mature non-structural proteins.4 Processing
of the mature polyprotein is mediated by this 3C protease, a (chy-
mo)trypsin-like cysteine protease having a Cys-His-Glu catalytic
triad and an extended binding site. The substrate specificity of noro-
virus 3C protease has been determined using in vitro transcription/
translation studies, and peptidyl chromogenic and fluorogenic sub-
strates.5–7 The protease shows a strong preference for a –D/E–F/Y–
X–L–Q–G–P– sequence (where X is H, E or Q) corresponding to the
recognition element
OH
O
Cys SH
E
Z
X
X
Z
S
Cys
E
0
0
0
subsites S5–S4–S3–S2–S1–S1 –S2 –. Cleavage is at the P1–P1 (Q–G)
scissile bond. X-ray crystal structures of norovirus 3C protease
warhead
(Z = H, COOR, CONHR, etc)
⇑
Corresponding author. Tel.: +1 316 978 7374; fax: +1 316 978 3431.
E-mail address: bill.groutas@wichita.edu (W.C. Groutas).
Figure 1. General representation of the interaction between a cysteine protease
and a transition state inhibitor.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.07.016

5316
K.-C. Tiew et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5315–5319
Scheme 1. Synthesis of inhibitors 1–10. Reagents: (a) TFA/CH2Cl2; (b) Z-AA or Z-Gly-AA/EDCI/HOBt/NMM, AA = amino acid; (c) LiBH4; (d) Dess–Martin periodinane; (e)
(CH3)2CHNC/HOAc then K2CO3/aq CH3OH.
could be further transformed into a molecule possessing molecular
properties that are important for oral bioavailability and favorable
ADME/Tox characteristics.11–13 The design of the inhibitors was
further augmented by insights gained via the use of computer
graphics and modeling and the X-ray crystal structure of the en-
zyme.7 The synthesis of inhibitors 1–10 was carried out as shown
in Scheme 1.14 The glutamine surrogate starting material was syn-
thesized using literature procedures.15
Table 1
Inhibitory activity of compounds (1–10)
Compound
Structure
IC50 (lM)
O
NH
O
O
H
1
7.2
N
N
N
CHO
H
O
O
NH
O
O
H
2
3
4
Inactivea
Inactivea
1.82
N
N
N
CHO
H
O
O
NH
O
H
O
N
N
H
O
OH
O
NH
O
H
N
O
N
H
CHO
O
O
NH
O
O
H
5
1.45
N
O
N
N
CHO
H
H
O

K.-C. Tiew et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5315–5319
5317
Table 1 (continued)
Compound
Structure
IC50 (lM)
O
NH
O
O
H
N
6
Inactivea
O
N
H
N
H
O
OH
O
NH
O
O
H
N
7
0.87
O
N
H
N
H
CHO
O
O
NH
O
O
H
8
Inactivea
N
O
N
H
N
H
O
OH
O
NH
O
O
H
9
7.5
N
O
N
N
CHO
H
H
O
O
NH
O
H
Inactivea
O
10
O
N
N
H
NH
O
O
a
Compounds were designated as inactive if the percent inhibition was <25 when incubated with the enzyme for 30 min at
an [I]/[S] ratio of 25.
Deblocking with TFA, followed by coupling with an appropriate
Cbz-protected amino acid ester, yielded a product which was
subsequently reduced to the alcohol with lithium borohydride.
Dess–Martin oxidation yielded the desired aldehydes. Alpha-ketoa-
mide 10 was synthesized by reacting the corresponding peptidyl
aldehyde with isopropyl isonitrile in the presence of acetic acid, fol-
lowed by mild hydrolysis of the diastereomeric acetate ester to
is of paramount importance (compare, e.g., compounds 1 and 4,
Table 1). In order to gain a better insight and understanding into
the binding of Inhibitor 4 to the active site of the enzyme,
computer modeling was used to demonstrate that 4 is capable of
adopting a low-energy conformation that closely resembles the
conformer of the co-crystallized peptide (Fig. 3).7,18 Thus, in addi-
tion to covalent bond formation between the active site cysteine
residue (Cys139) and the inhibitor aldehyde carbonyl (see general
illustration in Fig. 1), inhibitor 4 engages in multiple favorable
binding interactions with the enzyme, including lipophilic interac-
tions involving the –CH2–CH2– segment of the ligand lactam with
the –CH2–CH2– segment of Pro136, the leucine side chain in the
inhibitor with His30, Ile109 and Val 114, and interactions of the
phenyl ring in the Cbz cap—partially occupying the S4 pocket—
with Ile109. In addition, a network of hydrogen bonds involving
Thr134 (backbone carbonyl), Ala158 (backbone carbonyl), Gln110
yield the a
-hydroxyamide, and then Dess–Martin oxidation.16 The
interaction of compounds 1–10 with Norwalk virus 3C protease17
was investigated and the results are summarized in Table 1.
Incubation of compound 4 with Norwalk virus 3C protease lead
to dose-dependent inhibition of the enzyme (Fig. 2). It is evident
from Table 1 that the presence of the aldehyde warhead is essential
for inhibitory activity since the precursor alcohols were either
inactive or had minimal activity (compare, e.g., compounds 3 and
4, 5 and 6, 7 and 8, Table 1). Furthermore, the nature of the cap

5318
K.-C. Tiew et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5315–5319
pounds 4 and 5 were found to be active against the virus with
effective doses that inhibit 50% of norovirus replication, ED50s, of
2.1 and 7.8 M, respectively. The median toxic dose, TD50, for both
4 and 5 was found to be >320 M. Compounds 4 and 5 also inhibit
the replication of murine norovirus (MNV) in RAW267.4 cells with
ED50s of 5.5 and 20.3
M, respectively.24 The TD50s for both 4 and 5
with RAW267.4 were found to be >320
M.24 The results of an
l
l
l
l
ongoing hit-to-lead optimization campaign will be reported in
due course.
In conclusion, the first series of transition state inhibitors of
norovirus protease has been reported. Members of this series of
compounds exhibited noteworthy activity in a cell-based replicon
system of norovirus infection.
Acknowledgment
The generous financial support of this work by the National
Institutes of Health (U01AI081891) is gratefully acknowledged.
References and notes
Figure 2. Log dose–response curve for the inhibition of NV 3C protease by inhibitor
4.
1. (a) Glass, R. I.; Parashar, U. D.; Estes, M. K. N. Eng. J. Med. 2009, 361, 1776; (b)
Tan, M.; Jiang, X. Curr. Opin. Invest. Drugs 2008, 9, 146.
2. Mead, P. S.; Slutsker, L.; Dietz, V.; McCaig, L. F.; Breese, J. S.; Shapiro, C.; Griffin,
P. M.; Tauxe, R. V. Emerg. Infect. Dis. 1999, 5, 607.
3. Green, Y. K. In Fields Virology; Knipe, D. M., Howley, P. M., Eds.; Lippincott,
Williams & Wilkins: Philadelphia, 2007; Vol. 1, pp 949–979.
4. Blakeney, S. J.; Cahill, A.; Reilly, P. A. Virology 2003, 308, 216.
5. Hardy, M. E.; Crone, T. J.; Brower, J. E.; Ettayebi, K. Virus Res. 2002, 89, 29.
6. Someya, Y.; Takeda, N.; Miyamura, T. Virus Res. 2005, 110, 91.
7. Hussey, R. J.; Coates, L.; Gill, R. S.; Erskine, P. T.; Coker, S.-F.; Mitchell, E.;
Cooper, J. B.; Wood, S.; Broadbridge, R.; Clarke, I. N.; Lambden, P. R.; Shoolingin-
Jordan, P. M. Biochemistry 2011, 50, 240.
8. Zeitler, C. E.; Estes, M. K.; Prasad Venkatarman, B. V. J. Virol. 2006, 80, 5050.
9. Nakamura, K.; Someya, Y.; Kumasaka, T.; Ueno, G.; Yamamoto, M.; Sato, T.;
Takeda, N.; Miyamura, T.; Tanaka, N. J. Virol. 2005, 79, 13685.
10. (a) Webber, S. E.; Okano, K.; Little, T. L.; Reich, S. H.; Xin, Y.; Fuhrman, S. A.;
Matthews, D. A.; Love, R. A.; Hendrickson, T. F.; Patick, A. K.; Meador, J. W.;
Ferre, R. A.; Brown, E. L.; Ford, C. E.; Binford, S. L.; Worland, S. T. J. Med. Chem.
1998, 41, 2786; (b) Dragovich, P. S.; Prins, T. J.; Zhou, R.; Webber, S. E.;
Marakovits, J. T.; Fuhrman, S. A.; Patick, A. K.; Matthews, D. A.; Lee, C. A.; Ford,
C. E.; Burke, B. J.; Rejto, P. A.; Hendrickson, T. F.; Tuntland, T.; Brown, E. L.;
Meador, J. W.; Ferre, R. A.; Harr, J. E.; Kosa, M. B.; Worland, S. T. J. Med. Chem.
1999, 42, 1213.
11. Lipinski, C. A. J. Pharmacol. Toxicol. Methods 2000, 44, 235.
12. Veber, D. F. J. Med. Chem. 2002, 45, 2615.
13. Ritchie, T. J.; Ertl, P.; Lewis, R. Drug Discovery Today 2011, 16, 65.
14. All compounds were characterized by 1H NMR and HRMS and had a >95%
purity.
Figure 3. Predicted covalently-bound conformer21 for NV 3C protease inhibitor 4
15. Mou, K.; Xu, B.; Ma, C.; Yang, X.; Zou, X.; Lu, Y.; Xu, P. Bioorg. Med. Chem. Lett.
2008, 18, 2198.
16. Bogen, S. L.; Arasappan, A.; Velazquez, F.; Blackman, M.; Huelgas, R.; Pan, W.;
Siegel, E.; Nair, L. G.; Venkatraman, S.; Guo, Z.; Dolle, R.; Shi, N.-Y.; Njoroge, F.
G. Bioorg. Med. Chem. 2010, 18, 1854.
(stick structure with green carbon atoms and CPK-colored
N and O atoms)
contrasted with the peptidic inhibitor acetyl-Glu-Phe-Gln-Leu-Gln-CH@CH–COOꢁ
(stick structure with gray carbon atoms and CPK-colored N and O atoms) resolved
in the 1IPH crystal structure.7 The NV 3C protease binding site is shown as a
Connolly surface colored as follows: yellow = non-polar groups, white = partially
polar C, H atoms, red = polar O, blue = polar N, cyan = polar H. Key pharmacophore
residues are labeled according to the positions on the receptor surface from which
they interact with the ligand (except for Q110 whose approximate position is
marked but whose surface is not shown because the residue is above the plane of
the molecule).
17. Recombinant NV protease was assayed as follows: 10
was added to a thermostatted cuvette at 30 °C containing 180
NaH2PO4 buffer, pH 8.0, containing 120 mM NaCl and 6 mM DTT, and 5
l
L of 50
l
M NV protease
l
L of 50 mM
lL
DMSO. Five microliter of 12 mM Edans-EPDFHLQGPEDLAK-Dabcyl substrate in
DMSO was then added and the increase in fluorescence was monitored for
30 min at an excitation and emission wavelength of 360 and 460 nm,
respectively, using
curves were linear. The final enzyme and substrate concentrations were
2.5 M and 300 nM, respectively. In a typical inhibition run, 10 L of 50 M NV
protease was added to a thermostatted cuvette at 30 °C containing 180 L of
50 mM NaH2PO4 buffer, pH 8.0, containing 120 mM NaCl and 6 mM DTT, and
L of inhibitor in DMSO. After a 30 min incubation period, 5 L of 12
Edans-EPDFHLQGPEDLAK-Dabcyl substrate in DMSO was then added and the
increase in fluorescence was monitored for 30 min at an excitation and
emission wavelength of 360 and 460 nm, respectively.
a HORIBA FluoroMax 4 spectrofluorometer. Hydrolysis
(side chain carbonyl), and Ala160 (backbone amide proton) is
clearly evident. Extending the inhibitor by an additional amino
acid (as in compound 5) improved potency, albeit not dramatically
(compare compounds 4 and 5, Table 1). Modeling studies sug-
gested that replacement of Leu by other hydrophobic amino acids
might result in an optimal fit of the amino acid side chain in the S2
pocket, improving potency. Indeed, compound 7 with a P2 Nle was
found to be a sub-micromolar inhibitor of the enzyme, however,
replacement of Leu with Ile (compound 9, Table 1) was detrimental
l
l
l
l
5
l
l
lM
18.
A prospective bound conformer for NV 3C protease inhibitor 4 was
determined via a genetic algorithms conformational optimization using the
SYBYL program (SYBYL 8.0, The Tripos Associates, St. Louis, MO, 2008) The
covalently-bound ligand–receptor complex was prepared from the PDB 1IPH
crystal structure by deleting the co-crystallized ligand and adding the new
ligand (one atom at a time) in an analogous conformation via the ‘Add Atom’
utility so as to have conformational control during construction of the ligand
and ensure automatically specification of low-energy bond lengths and bond
angles. Hydrogens were added to the entire complex according to the
automatic SYBYL algorithm (assuming cationic Lys and Arg residues, and
anionic Asp and Glu) and were positionally optimized via molecular
to inhibitory activity.
activity, suggesting that steric congestion in the vicinity of the S1
a
-Ketoamide 10 was devoid of inhibitory
0
subsite is severe.
The activity of inhibitors 4–5 against the Norwalk norovirus
was investigated using a cell-based replicon system.19–23 Com-

K.-C. Tiew et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5315–5319
5319
mechanics with all heavy atoms held rigid and default convergence criteria
20. Chang, K. O.; Sosnovtsev, S. V.; Belliot, G.; Green, K. Y. Virology 2006, 2, 463.
21. Chang, K. O.; George, D. W. J. Virol. 2007, 22, 12111.
22. Chang, K. O. J. Virol. 2009, 83, 8587.
23. The effects of compounds 4 and 5 were also examined in murine norovirus-1
(MNV-1). MNV-1 can be cultured in the murine macrophage-like cell line
RAW267.4, thus MNV-1 can be a surrogate system to examine the effects of
antiviral compounds on norovirus replication in cells. Confluent RAW267.4
cells in 6-well plates were inoculated with MNV-1 at a multiplicity of infection
using the Tripos Molecular Force Field24 and Gasteiger-Marsili charges.25 The
resulting complex was then subjected to a genetics algorithm conformational
search implemented in SYBYL, requesting identification of the top twenty
most favorable conformations. The search yielded only one plausible low-
energy conformation (depicted in Fig. 3).
19. The effects of compounds 4 and 5 were examined in NV replicon-harboring
cells (HG23 cells).20 The detailed procedures for studying the antiviral effects
using HG23 cells have been reported elsewhere.20–22 Briefly, 1-day old, 80–90%
confluent HG23 cells were treated with varying concentrations of compound 4
(MOI) of 2 with varying concentrations (0–320 lM) of compounds 4 and 5.
Virus infected cells were then incubated for an additional 12 and 24 h. After
freezing and thawing plates three times, the replication of MNV-1 in the
presence of the compound was measured by the 50% tissue culture infective
dose (TCID50) assay. The nonspecific cytotoxic effects in RAW267.4 cells by
ribavirin were monitored by the method described above.
or 5 (0 [mock-DMSO]–320 lM) to examine its effects on the replication of NV.
At 24 or 48 h of treatment, the NV protein or genome were analyzed with
Western blot analysis, or qRT-PCR, respectively. The ED50s of compounds 4 or 5
for NV genome levels were determined at 24 h post-treatment. The cytotoxic
effects of compounds 4 or 5 on HG23 cells were determined using a cell
cytotoxicity assay kit (Promega, Madison, WI) to calculate the median toxic
dose (TD50) at 48 h of treatment.
24. Clark, M.; Cramer, R. D., III; Van Opdenbosch, N. J. Comput. Chem. 1989, 10, 982.
25. Gasteiger, J.; Marsili, M. Tetrahedron Lett. 1978, 3181.