Heteroaromatic ester inhibitors of hepatitis A virus 3C proteinase: Evaluation of mode of action

Article abstract of DOI:10.1016/j.bmc.2008.03.059

The related 3C and 3C-like proteinase (3C^pro and 3CL^pro) of picornaviruses and coronaviruses, respectively, are good drug targets. As part of an effort to generate broad-spectrum inhibitors of these enzymes, we screened a library of inhibitors based on a halopyridinyl ester from a previous study of the severe acute respiratory syndrome (SARS) 3CL proteinase against Hepatitis A virus (HAV) 3C^pro. Three of the compounds, which also had furan rings, inhibited the cleavage activity of HAV 3C^pro with K_ics of 120-240 nM. HPLC-based assays revealed that the inhibitors were slowly hydrolyzed by both HAV 3C^pro and SARS 3CL^pro, confirming the identity of the expected products. Mass spectrometric analyses indicated that this hydrolysis proceeded via an acyl-enzyme intermediate. Modeling studies indicated that the halopyridinyl moiety of the inhibitor fits tightly into the S1-binding pocket, consistent with the lack of tolerance of the inhibitors to modification in this portion of the molecule. These compounds are among the most potent non-peptidic inhibitors reported to date against a 3C^pro.

Full text of DOI:10.1016/j.bmc.2008.03.059

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 5761–5777
Heteroaromatic ester inhibitors of hepatitis A virus 3C
proteinase: Evaluation of mode of action
Carly Huitema,a Jianmin Zhang,b Jiang Yin,c Michael N. G. James,c
John C. Vederasb and Lindsay D. Eltisa,*
aDepartments of Microbiology and Biochemistry, University of British Columbia, 2350 Health Science Mall, Vancouver,
British Columbia, Canada V6T 1Z3
bDepartment of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2
cDepartment of Biochemistry, University of Alberta, Edmonton, AB, Canada T6G 2H7
Received 28 January 2008; accepted 24 March 2008
Available online 27 March 2008
Abstract—The related 3C and 3C-like proteinase (3Cpro and 3CLpro) of picornaviruses and coronaviruses, respectively, are good
drug targets. As part of an effort to generate broad-spectrum inhibitors of these enzymes, we screened a library of inhibitors based
on a halopyridinyl ester from a previous study of the severe acute respiratory syndrome (SARS) 3CL proteinase against Hepatitis A
virus (HAV) 3Cpro. Three of the compounds, which also had furan rings, inhibited the cleavage activity of HAV 3Cpro with Kics of
120–240 nM. HPLC-based assays revealed that the inhibitors were slowly hydrolyzed by both HAV 3Cpro and SARS 3CLpro, con-
firming the identity of the expected products. Mass spectrometric analyses indicated that this hydrolysis proceeded via an acyl-
enzyme intermediate. Modeling studies indicated that the halopyridinyl moiety of the inhibitor fits tightly into the S1-binding pock-
et, consistent with the lack of tolerance of the inhibitors to modification in this portion of the molecule. These compounds are
among the most potent non-peptidic inhibitors reported to date against a 3Cpro
.
ꢀ 2008 Elsevier Ltd. All rights reserved.
1. Introduction
The crystal structure of HAV 3Cpro was the first solved
of the picornaviral proteinases.2 It has an overall fold
resembling the two-domain b-barrel fold of chymotryp-
sin but with a cysteine in the active site typical of the 3C
proteinases.3 The active site consists of a long groove be-
tween the two b-barrel domains and the enzyme’s spec-
ificity is determined by the residues that line this groove.
In the central portion of the groove between the two do-
mains, lie the conserved residues of a catalytic triad: the
nucleophilic Cys172; the general acid-base catalyst
His44; and Asp84, that forms a hydrogen bond to
His44.4 This catalytic triad is consistent with other 3Cpro
structures identified to date including those of PV,
HRV, and FMDV.5–7 The coronaviral 3CLpros have a
similar two-domain b-barrel fold with a third a-helical
domain that plays a role in dimerization and contributes
to the active site pocket.8 Coronaviruses have an active
site Cys–His catalytic dyad9 but for both types of pro-
teinases the catalytic mechanism is similar. Following
binding of a substrate, the latter’s carbonyl carbon atom
of the scissile peptide bond is subject to nucleophilic at-
tack by the sulfur atom of the active site Cys. Formation
of a tetrahedral intermediate is stabilized by residues of
Regulated proteolysis is a critical feature in the process-
ing of viral polyproteins and is essential for their replica-
tion. As such, proteinases constitute an important target
for novel antivirals. The main proteinases responsible
for this processing in picornaviruses and coronaviruses
are the 3C and 3C-like proteinases (3Cpro and 3CLpro),
respectively.1 Picornaviruses include important human
pathogens such as human rhinovirus (HRV), poliovirus
(PV), and hepatitis A virus (HAV), as well as significant
insect, plant and agricultural pathogens, such as foot-
and-mouth disease virus (FMDV). Coronaviruses in-
clude the etiological agent of the severe acute respiratory
syndrome (SARS) outbreak in 2002. Given the signifi-
cance of these viral pathogens, intense efforts are under-
way to develop novel antiviral compounds that target
3Cpros and 3CLpros.
Keywords: Cysteine proteinase; Antiviral; Picornavirus; Halopyridinyl
ester.
*
Corresponding author. Tel.: +1 604 822 0042; fax: +1 604 822
6041; e-mail: leltis@interchange.ubc.ca
0968-0896/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.03.059

5762
C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
the electrophilic oxyanion hole, and loss of the amide
leads to formation of an acyl enzyme intermediate.4
The catalysis is complete following deacylation and res-
toration of the active site.
enzyme) using a microplate assay. Of the 82 compounds,
49 completely inhibited the proteinase’s activity when
tested at 10 lM, 10 showed complete inhibition at
1 lM, and none showed complete inhibition at
0.25 lM (Table 1). For six compounds (Table 2) that
inhibited the proteinase activity to the greatest extent
at 1 lM, IC50 values were assessed. The IC50 experi-
ments were performed as described for the initial screen-
ing except that the substrate at a concentration of 20 lM
was used. As summarized in Table 2, the six compounds
had IC50s ranging from 50 nM to 1.2 lM. The lowest
value, for 24, is close to the theoretical minimal value
that can be determined in this experiment (i.e., 50% of
the concentration of the enzyme).
With some exceptions, such as peptidyl aldehydes10 and
peptidyl fluoromethyl ketones,11 many of the inhibitors
of HAV 3Cpro that have been developed to date are
not particularly potent in their initial binding to the en-
zyme. Keto-glutamine analogues have been well studied
as potential inhibitors. Such analogues with
a
phthalhydrazide moiety display initial competitive
(reversible) inhibition of HAV 3Cpro with the best hav-
ing a Ki of 9 lM.12 Over a longer period (hours), they
then alkylate the active site cysteine thiol with concom-
itant loss of the phthalhydrazide moiety.13 Such com-
pounds were improved through cyclization of the
glutamine side chain, resulting in inhibitors with IC50s
in the low micromolar range.14 Lall et al.15 also investi-
gated interaction of b-lactones with HAV 3C and found
a reversible inhibitor having a Ki of 9 lM. Interestingly,
2.2. Determination of kinetic parameters
In studies using sufficient pyridinyl ester to completely
inhibit the initial hydrolysis of the fluorogenic peptide
by HAV 3Cpro, peptide hydrolysis was sometimes ob-
served after incubating the reaction mixture for
ꢀ4 min. This suggested that HAV 3Cpro might catalyze
the slow hydrolysis of the ester inhibitor. To investigate
this possibility, we developed an HPLC-based assay to
monitor inhibitor hydrolysis by HAV 3Cpro and SARS
3CLpro. Reactions were performed using a concentra-
tion of pyridinyl ester high enough to ensure enzyme sat-
uration. HPLC analysis of the reaction mixture
demonstrated that the expected hydrolysis products
could be observed in a time-dependent manner concom-
itant with the disappearance of the ester (Fig. 1) and
that this hydrolysis was dependent on the enzyme. For
each of the tested esters, product identification was con-
firmed when possible by comparison to known stan-
dards. The assay was performed at several
concentrations of ester to confirm saturation of the pro-
teinase with the ester and enzymatic hydrolysis rates
were corrected for the non-enzymatic hydrolysis of the
inhibitor.
the enantiomeric compound was an irreversible inhibi-
tor with a kinact/Ki = 3800 MÀ1 minÀ1
.
4
Given the similarity of the active site architectures of
3Cpro and 3CLpro, it is expected that these proteinases
will be inhibited by similar classes of compounds. In-
deed, it may be possible to generate broad-spectrum
inhibitors of these enzymes. Screening of a 50,000-mem-
ber library of small molecules revealed a pyridinyl thio-
phene ester, 1 (Table 1), that inhibited SARS 3CLpro
with an IC50 of 0.5 lM.16 This compound also inhibited
HAV 3Cpro with an essentially identical IC50. More re-
cently, a library of 82 heteroaromatic esters based on 1
17
was created and screened against SARS 3CLpro
.
The
most potent ester 24 (Table 2) inhibited the proteinase
with an IC50 of 50 nM. Mass spectrometric analysis sug-
gested a mechanism of inhibition involving acylation of
the enzyme by the furoyl group. Slow reactivation of the
enzyme further suggested that the acyl-enzyme was sub-
ject to deacylation, and that the enzyme catalyzed the
slow hydrolysis of the ester.
The calculated and deduced steady-state kinetic param-
eters for three of the best inhibitors are summarized in
Table 2. The kcat values were calculated from the hydro-
lysis rates at saturating ester concentrations. The Km
values were taken to be the equivalent of the determined
Kic values, which is the case for substrates that are
slowly turned over (e.g., Seah et al.).18 Among the three
inhibitors tested with HAV 3Cpro, the two with the same
acyl group, 24 and 20, had very similar kcat values.
Equivalent studies using SARS 3CLpro and the most po-
tent inhibitor, 24, demonstrated that this proteinase also
catalyzed ester hydrolysis with parameters similar to
Herein, we investigate the ability of the heteroaromatic
esters to inhibit HAV 3Cpro. The inhibitors were
screened against HAV 3Cpro using a fluorescent peptide
substrate and the steady-state inhibition parameters of
the most potent inhibitors were evaluated. The mode
of action of the inhibitors was further investigated using
an HPLC-based kinetic assay and mass spectrometric
analyses. The results of these studies were compared to
the inhibition of SARS 3CLpro by the heteroaromatic es-
ters as well as to the previously described HAV 3Cpro
inhibitors.
those observed for HAV 3Cpro
.
2.3. MS analyses of inhibitor–proteinase complex
2. Results
2.1. Initial screen
The kinetic data indicate that HAV 3Cpro catalyzes the
slow hydrolysis of the pyridinyl ester inhibitors. To
investigate whether hydrolysis proceeds via an acyl-en-
zyme intermediate, samples of proteinase were incu-
bated for 10 min with 100 lM 24 and the resulting
sample was analyzed by mass spectrometry. As shown
in Figure 2, incubation of HAV 3Cpro resulted in a mass
The library of 82 pyridinyl esters (Scheme 1) that yielded
17
potent inhibitors of SARS 3CLpro (a coronoviral pro-
teinase) was screened against HAV 3Cpro (a picornaviral

Table 1. Percentage of inhibition of HAV 3Cpro for preliminary analysis of inhibitor library at various concentrations of inhibitors
Compound Structure
% 10 lM % 1 lM % 0.25 lM Compound Structure
% 10 lM % 1 lM % 0.25 lM
O
O
O
Cl
S
1
93
89
ND
ND
ND
43
44
>90
54
67
69
23
53
O
N
N
2
3
8
See Table 2
See Table 2
S
S
O
O
N
83
21
ND
ND
ND
ND
45
46
47
48
>90
>90
ND
57
50
O
4
5
6
<10
<10
<10
10
82
ND
<10
10
S
O
N
H
N
ND
ND
85
S
S
O
O
N
H
N
>90
59
N
H
N
7
8
<10
<10
ND
ND
ND
ND
49
50
>90
11
ND
ND
S
S
O
O
N
N
Cl
H
N
59
ND
(continued on next page)

Table 1 (continued)
Compound Structure
% 10 lM % 1 lM % 0.25 lM Compound Structure
% 10 lM % 1 lM % 0.25 lM
H
N
S
9
<10
<10
ND
ND
ND
ND
51
52
>90
26
ND
ND
O
N
N
Cl
H
N
S
S
10
11
12
13
14
82
<10
O
O
F
O
N
Cl
16
91
12
23
ND
<10
ND
ND
ND
ND
ND
ND
53
54
55
56
>90
NDa
64
>90
ND
57
20
O
Cl
O
O
N
Me
O
ND
<10
ND
S
S
O
O
O
O2N
N
N
Me
O
O
S
S
N
N
39
23
O
O
Me
Cl
15
25
ND
ND
57
29
50
ND

Cl
O
O
16
11
ND
ND
58
63
32
ND
S
N
O
O
Me
17
<10
<10
ND
59
<10
ND
ND
S
S
N
NO2
N
O
O
18
19
14
92
ND
<10
ND
<10
60
61
74
19
54
27
O
O
Me
Cl
N
H
ND
ND
N
20
See Table 2
>90
>90
49
62
<10
ND
<10
O
Cl
21
22
>90
>90
>90
ND
<10
ND
63
64
>90
>90
53
30
ND
ND
O
O
N
O
O
Cl
O
N
(continued on next page)

Table 1 (continued)
Compound Structure
% 10 lM % 1 lM % 0.25 lM Compound Structure
% 10 lM % 1 lM % 0.25 lM
O
Cl
23
>90
>90
<10
65
13
ND
ND
O
N
24
25
26
27
28
29
See Table 2
ND
>90
ND
ND
ND
ND
ND
87
66
67
68
69
70
71
82
75
32
N
67
ND
ND
ND
ND
ND
>90
ND
ND
O
O
O
Br
O
O
O
Cl
Cl
11
87
60
22
O
O
O
<10
>90
<10
>90
27
ND
ND
ND
N
O
S
75
ND
O
S
N
N
H
N
81
32
O
O

S
30
31
32
33
10
ND
ND
ND
ND
37
72
73
74
75
>90
45
ND
ND
ND
ND
O
N
O
Cl
Cl
80
58
77
70
N
H
O
O
N
N
O
>90
>90
47
85
ND
S
S
O
Cl
85
>90
24
N
O
N
O
Cl
34
>90
50
ND
76
64
ND
ND
MeO
Cl
O
N
O
Cl
O
35
36
>90
>90
ND
ND
36
54
77
78
40
76
ND
ND
ND
O
N
See Table 2
36
(continued on next page)

Table 1 (continued)
Compound Structure
% 10 lM % 1 lM % 0.25 lM Compound Structure
% 10 lM % 1 lM % 0.25 lM
37
See Table 2
83
>90
43
79
56
21
ND
38
39
40
41
42
<10
>90
37
ND
ND
ND
ND
47
ND
80
81
82
83
37
22
44
23
34
ND
ND
ND
ND
24
ND
ND
<10
34
19
>90
35
33
ND, not determined.
a Not soluble.

C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
5769
Table 2. Kinetic parameters of a series of pyridinyl analogue inhibitors of HAV 3Cpro and SARS 3CLpro and half-life of pyridinyl analogues in buffer
Compound Structure
IC50 (nM)
Km (nM) kcat (sÀ1) · 10À5 kcat/Km (MÀ1sÀ1) · 103 Half-life (sÀ1) · 10À3
HAV 3Cpro
20
37
24
45
36
44
338
2
240 30
180 30
120 20
ND
110
6
4.6 0.6
1.2 0.2
4.9 0.6
1.1 0.3
ND
140 20
25
8
1.4 0.5
53
2
119
ND
ND
8
10
2
470 60
ND
ND
1200 130 ND
ND
250 50
ND
26
ND
17
ND
ND
SARS 3CLpro
24
ND
7
6
7
3
ND, not determined.
Method A:
EDCl, HOBt, DIPEA, DMF
5-Chloro-3-pyridinol
R
O
Cl
N
RCO2H
RCO2H
O
N
O
Method B:
1) SOCl2, CH2Cl2
Cl
R
2) 5-Chloro-3-pyridinol,Pyr. CH2Cl2.
O
Scheme 1. Synthesis of a library of 3-chloropyridinyl esters by Method A or B.
increase of 100 2.5 Da. This is consistent with the
covalent attachment of the furoyl moiety (MW 95 Da)
and the departure of the 3-bromo-5-hydroxypyridine
as the leaving group. In samples that were incubated
for both 2 and 10 min, only the acylated form of the pro-
teinase was detected.
2.4. Docking
To gain insight into the interactions between the inhib-
itors and HAV 3Cpro prior to acylation, compound 24
was computer-modeled into the active site pocket of
the enzyme. From the 20 docked complexes obtained

5770
C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
Figure 1. Hydrolysis of 24 by HAV 3Cpro. The reaction mixture contained 50 lM inhibitor and 2 lM 3Cpro. It was subjected to HPLC analysis after
0, 17, 35, and 52 min. Traces were recorded at (A) 252 nm and (B) 287 nm. Peaks corresponding to each of 24, furoic acid, and 3-bromo-5-
hydroxypyridine (3B5H) are labeled. Over the time course of this experiment, non-enzymatic hydrolysis of the ester was negligible (data not shown).
Figure 2. (A) Mass spectrum of wild-type HAV 3Cpro (M+ 23875.08 Da). (B) Mass spectrum of the complex of 3Cpro and inhibitor 24 following
2 min of incubation (M+ 23974.38 Da).

C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
5771
Figure 3. Possible prehydrolysis-binding modes of 24 to (a) HAV 3Cpro and (b) SARS 3CLpro. Protein surfaces are colored according to electrostatic
potentials with red and blue indicating negative and positive potentials, respectively. The carbon atoms of 24 and the amino acid residues are gray
and purple, respectively. The nitrogen, oxygen and sulfur atoms are colored blue, red, and orange, respectively. Hydrogen bonds are displayed by
dashed lines.
using the program AUTODOCK, the best binding
mode as judged from binding energies and allowable
bond lengths and angles is one in which the halopyridine
ring occupies the S1-binding pocket (Fig. 3a). More spe-
cifically, the nitrogen of the halopyridinyl ring is hydro-
than for good peptide substrates (2.1 0.5 mM for Ac-
ELRTQSFS-NH2). By contrast, the enzyme turns over
peptide
substrates
much
more
efficiently
(kcat = 1.8 0.1 sÀ1 for Ac-ELRTQSFS-NH2 versus
25 8 · 10À5 sÀ1 for 24). Importantly, the specificity
of the viral proteinase for its natural polypeptide sub-
strate has yet to be determined. The presence of an ester
linkage is a critical feature for effective inhibition of
3Cpro by the tested small molecules. For example, com-
pounds lacking the ester bond, such as 29, 30, and 61,
are poor inhibitors suggesting that these compounds
do not interact with the proteinase particularly strongly.
20
gen bonded to the Ne2 of His191 (3.0 A) and the halogen
˚
atom points out toward solvent. The carbonyl oxygen of
the ester is located in the oxyanion hole of the enzyme,
forming two hydrogen bonds each with the N of Gly170
˚
and Cys172 at distances of 3.1 and 2.7 A, respectively.
3. Discussion
Comparison of the inhibition characteristics of the esters
reveals that there is more flexibility in the cyclic acid
moiety of the ester than in the pyridinyl alcohol. For
example, 60, 66, 68 and 78 moderately inhibited HAV
3Cpro at concentrations of 1 lM despite the different
sizes of the cyclic acid moiety. While the cyclic acid moi-
ety is subject to some constraints, as illustrated by com-
pounds 65, 69, and 78, these constraints are difficult to
deduce from the current data set. The spacing between
the ester and alcohol or acid moiety was also more crit-
ical for the alcohol than the acid. Compound 37 was a
good inhibitor despite possessing two additional carbon
atoms (–CH@CH–) on the acid side of the ester with re-
spect to 20. By contrast, the insertion of a single addi-
tional carbon atom (–CH2–) on the alcohol side of the
ester of the potent inhibitor 24 yielded 25, a poor inhib-
itor. Finally, the presence and position of the halogen on
the pyridine ring were important determinants of inhibi-
tion as was the position of the ring nitrogen with respect
to the ester bond. Thus, compounds containing 2- or 4-
pyridinyls (2, 4, 11 and 13) were poor inhibitors. Simi-
larly, compounds lacking the halogen substituent (3 vs
45) or in which this substituent is ortho or para to the es-
In this study, a targeted library of compounds was
screened to identify several potent inhibitors of the
picornaviral HAV 3Cpro. The six most potent inhibitors
all contain an ester bond connecting a halopyridine ring
to a 5-membered ring (Table 2). The most potent of
these, 24, competitively inhibited the hydrolysis of a
peptidyl substrate with a Kic (120 nM) that is among
the lowest for an inhibitor of a 3Cpro enzyme reported
to date. HPLC and mass spectrometric analysis demon-
strated that the esters are slow substrates of HAV 3Cpro
,
with hydrolysis proceeding via an acyl-enzyme interme-
diate. Consistent with a mechanism involving rate-limit-
ing deacylation, the two compounds possessing the same
acylating group, 20 and 24, had very similar kcat values.
Interestingly, the halopyridinyl esters are better sub-
strates for HAV 3Cpro than the peptide substrates char-
acterized to date. For example, the kcat/Km value for 37
was 1400 500 MÀ1 sÀ1 versus 800 100 MÀ1 sÀ1 for
Ac-ELRTQSFS-NH2.19 This is due to the enzyme’s
exceptionally low Km values for the halopyridinyl esters
(e.g., 120 20 nM for 24), that were considerably lower

5772
C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
ter group (15 and 16 vs 45) were also poor inhibitors.
Overall, these results indicate that the halopyridine ring
is a particularly important determinant of inhibition.
gests that these compounds could be used to inhibit the
related proteinases of other coronaviruses and picornav-
iruses such as FMDV and HRV. These inhibitors might
be improved by increasing their specificity, presumably
by targeting the S2 pocket better, and further slowing
their enzymatic hydrolysis. Indeed, given the correlation
between turnover of the esters (kcat) and non-enzymatic
stability (half-life), decreasing enzymatic hydrolysis
might improve their stability in aqueous solution. Final-
ly, given the slightly different substrate-binding pockets
of the different 3Cpros, a comparison of the inhibition
of related proteinases by a series of these halopyridinyl
esters might provide insight into the mode of substrate
binding and how the physicochemical properties of the
substrate-binding pocket determine the excellent speci-
ficity of these enzymes.
The model of 24 docked to HAV 3Cpro is consistent with
the structure–activity relationship of the inhibitors. In
particular, the model indicates that the halopyridinyl
ring of 24 essentially fills the S1 pocket of the enzyme
with the pyridine nitrogen atom forming a hydrogen
bond to Ne2 of His191 in this pocket (Fig. 3a) in contrast
to the considerable space in the enzyme’s active site
around the furan ring. Although the modeled HAV
3Cpro-24 complex does not clarify the role of the halogen
group on the pyridine ring as this group, it is possible
that the halogen influences the electronic characteristics
and/or the solvation of the pyridinyl ring so as to en-
hance inhibition.
The modeled HAV 3Cpro-24 complex is also consistent
with the hydrolysis of these ester compounds via an
acyl-enzyme intermediate in two aspects. First, the dis-
tance between the Sc of Cys172 and the carbonyl carbon
5. Experimental
5.1. Chemicals and reagents
˚
in 24 is 3.4 A. Second, these two atoms are almost co-
Factor Xa was purchased from Hematologic Technolo-
gies Inc. (Essex Junction, USA). Peptides were synthe-
sized at the Nucleic Acid Protein Service Unit of the
University of British Columbia and confirmed by mass
spectrometric analysis. All other chemicals were of ana-
lytical grade and used without further purification.
planar with the imidazole ring of His44. Therefore,
His44 is in an ideal position to have its Ne2 acting as a
general base. During normal enzymatic hydrolysis of
peptidyl substrates, the scissile peptide bond is expected
to be in line with the spatial orientation between the Cys
and His catalytic pair in the active site. However, the
modeled HAV 3Cpro-24 complex shows a rather differ-
ent direction for the central scissile ester bond of com-
pound 24, primarily due to the predicted strong
propensity of the bromopyridinyl ring to bind inside
the S1 pocket. Interestingly, a similar binding mode
was predicted for the complexes between SARS 3CLpro
and a group of pyridinyl ester-based inhibitors.21
5.2. Inhibitor synthesis
Two general procedures were used to prepare pyridinyl
esters. In method A, the following compounds were
added to a solution of carboxylic acid (0.5 mmol, 1.0
equiv) in DMF (2 mL) at rt: EDCI (97 mg, 0.5 mmol,
1.0 equiv), HOBt (68 mg, 0.5 mmol, 1.0 equiv), DIPEA
(90 lL, 0.5 mmol, 1.0 equiv), and 5-chloro pyridinol
(65 mg, 0.5 mmol, 1.0 equiv). After 24 h of stirring, the
solvent was removed in vacuo to afford the crude
mixture.
Considering the similarities of picornaviral 3Cpro and
coronaviral 3CLpro, it is not surprising that the library
of pyridinyl esters yielded similar potent inhibitors of
the HAV and SARS proteinases.17 In particular, both
enzymes are catalytically driven by a Cys–His pair and
have a high specificity of Gln as the P1 residue. This
specificity reflects their similar S1 pockets, especially
the histidine at its base (His163 and His191 of SARS
3CLpro and HAV 3Cpro, respectively) which is an impor-
tant determinant of P1 specificity in both enzymes.4,22
Consistent with these similarities, the predicted binding
In method B, the following compounds were added to a
solution of carboxylic acid (1 mmol, 1.0 equiv) in DCM
(5 mL) at rt: thionyl chloride (0.4 mL, 2.6 equiv) and a
catalytic amount of DMF (2 drops). After 20 h of stir-
ring, the solvent was removed in vacuo to afford the acyl
chloride product. A solution of the acyl chloride in
DCM (5 mL) was added dropwise to a solution of 5-
chloro pyridinol (130 mg, 1 mmol, 1.0 equiv) and pyri-
dine (0.09 mL, 1.1 equiv) in DCM (5 mL) at 0 ꢁC. After
3 h of stirring, the solvent was removed in vacuo to af-
ford the crude mixture.
mode of halopyridinyl esters to SARS 3CLpro is very
17
similar to that predicted for HAV 3Cpro, with the
halopyridinyl moiety occupying the S1 pocket and the
ring of the acid moiety extending into the S2 pocket.
Crude mixtures were purified using an 1100 HPLC cou-
pled with an ES-MSD Agilent 1956B with positive ion
detection. The HPLC was fitted with a semi-preparative
column, Zorbax RX-C8 (9.4 · 250 mm, 5 lM) equipped
with a guard column. The column was operated at a
flow rate of 3 mL/min. Compounds were eluted using
a linear gradient of 35–100% acetonitrile in 0.05% for-
mic acid/H2O over 20 min, followed by 100% acetoni-
trile in 0.05% formic acid/H2O (2 min) and a final
return to 35% acetonitrile in 0.05% formic acid/H2O in
4. Conclusion
We have screened a library of 82 non-peptidic inhibitors,
including 49 not previously described, against HAV
3Cpro and have characterized in more detail several of
the most potent of these inhibitors. Some of these inhib-
itors had been previously shown to potently inhibit the
related SARS 3CLpro. The ability of the halopyridinyl
esters to inhibit two such divergent 3Cpros strongly sug-

C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
5773
0.5 min. The quality of selected purified samples was
confirmed by re-injection of the samples on an analytic
column (:Zorbax RX-C18, 4.6 · 150 mm, 5 lM) oper-
ated at a flow rate of 0.7 mL/min using the above-de-
scribed linear gradient.
5.2.8. 5-Chloropyridin-3-yl 2-chlorobenzoate (43). Meth-
od A. A white solid. H NMR (CDCl3, 300 MHz) d
8.53 (d, 1H, J = 2.1 Hz), 8.50 (d, 1H, J = 2.3 Hz), 8.06
(ddd, 1H, J = 7.9, 1.9, 1.0 Hz), 7.72 (dd, 1H, J = 2.2,
2.2 Hz), 7.57–7.54 (m, 2H), 7.46–7.40 (m, 1H); HRMS
[EI] Calcd for C12H7NO2Cl2 (M+) 266.9854, found
266.9857.
1
5.2.1. 5-Chloropyridin-3-yl 5-bromofuran-2-carboxylate
(36). Method B. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.52 (d, 1H, J = 2.1 Hz), 8.47 (d, 1H,
J = 2.3 Hz), 7.68 (dd, 1H, J = 2.3, 2.1 Hz), 7.38 (d, 1H,
J = 3.7 Hz), 6.59 (d, 1H, J = 3.6 Hz); HRMS [EI] Calcd
for C10H5ClNO3Br (M+) 302.9112, found 302.9114.
5.2.9. 5-Chloropyridin-3-yl 5-nitro-1H-pyrazole-3-car-
boxylate (44). Method B. A white solid. 1H NMR
(CDCl3, 500 MHz) d 8.60 (dd, 1H, J = 2.1, 0.5 Hz),
8.52 (dd, 1H, J = 2.4, 0.5 Hz), 7.72 (t, 1H, J = 2.1 Hz),
7.64 (s, 1H); HRMS [EI] Calcd for C9H5ClN4O4 (M+)
267.9999, found 267.9992.
5.2.2. (E)-5-Chloropyridin-3-yl 3-(furan-2-yl)acrylate
(37). Method A. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.47 (dd, 1H, J = 2.1, 0.4 Hz), 8.41 (dd,
1H, J = 2.3, 0.4 Hz), 7.63 (d, 1H, J = 15.5 Hz), 7.60 (t,
1H, J = 2.2 Hz), 7.57 (ddd, 1H, J = 2.3, 1.2, 0.5 Hz),
6.72 (d, 1H, J = 3.4 Hz), 6.51 (dd, 1H, J = 3.4, 1.8 Hz),
6.46 (d, 1H, J = 15.6 Hz); HRMS [EI] Calcd for
C12H8ClNO3 (M+) 249.0193, found 249.0194.
5.2.10. 5-Chloropyridin-3-yl thiophene-2-carboxylate
(45). Method A. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.51 (d, 1H, J = 2.1 Hz), 8.48 (d, 1H,
J = 2.1 Hz), 8.02 (dd, 1H, J = 3.9, 1.3 Hz), 7.74 (dd,
1H, J = 5.0, 1.2 Hz), 7.69 (dd, 1H, J = 2.1, 2.1 Hz),
7.22 (dd, 1H, J = 5.0, 3.9 Hz); HRMS [EI] Calcd for
C10H6O2NSCl (M+) 238.9808, found 238.9809.
5.2.3. 5-Chloropyridin-3-yl 3-methylthiophene-2-carbox-
ylate (38). Method A. A white solid. H NMR (CDCl3,
1
5.2.11. 5-Chloropyridin-3-yl 2-methylbenzoate (46). Meth-
od A. A white solid. 1H NMR (CDCl3, 300 MHz) d 8.56–
8.42 (m, 2H), 8.16 (dd, 1H, J = 8.3, 1.5 Hz), 7.68 (dd, 1H,
J = 2.0, 2.0 Hz), 7.53 (dt, 1H, J = 7.5, 1.4 Hz), 7.39–7.32
(m, 2H), 2.68 (s, 3H); HRMS [EI] Calcd for C13H10NO2Cl
(M+) 247.0400, found 247.0396.
500 MHz) d 8.50 (d, 1H, J = 2.3, 0.5 Hz), 8.47 (dd, 1H,
J = 2.3, 0.5 Hz), 7.69 (dd, 1H, J = 2.3, 2.3 Hz), 7.57 (d,
1H, J = 5.0 Hz), 7.03 (d, 1H, J = 5.0 Hz), 2.63 (s, 3H);
HRMS [EI] Calcd for C11H8ClNO2S (M+) 252.9964,
found 252.9962.
5.2.4. 5-Chloropyridin-3-yl 5-methylthiophene-2-carbox-
ylate (39). Method A. A white solid. H NMR (CDCl3,
5.2.12. 5-Chloropyridin-3-yl 2-methoxybenzoate (47).
Method A. A white solid. H NMR (CDCl3, 300 MHz)
1
1
500 MHz) d 8.49 (d, 1H, J = 2.0 Hz), 8.46 (d, 1H,
J = 2.4 Hz), 7.83 (d, 1H, J = 3.8 Hz), 7.68 (d, 1H,
J = 2.3, 2.0 Hz), 6.88 (dd, 1H, J = 3.9, 1.0 Hz), 2.60 (s,
3H); HRMS [EI] Calcd for C11H8ClNO2S (M+)
252.9964, found 252.9954.
d 8.50 (d, 1H, J = 2.0 Hz), 8.48 (d, 1H, J = 2.3 Hz), 8.03
(dd, 1H, J = 8.0, 1.8 Hz), 7.75 (dd, 1H, J = 2.2, 2.2 Hz),
7.60 (ddd, 1H, J = 8.8, 7.5, 1.8 Hz), 7.11–7.04 (m, 2H),
3.96 (s, 3H); HRMS [EI] Calcd. for C13H10NO3Cl
(M+) 263.0349, found 263.0350.
5.2.5. 5-Chloropyridin-3-yl 5-(2-nitrophenyl)furan-2-car-
boxylate (40). Method A. A slightly brown solid. H
5.2.13. 5-Chloropyridin-3-yl 4-chloro-2-hydroxybenzoate
(48). Method B. A white solid. 1H NMR (CDCl3,
300 MHz) d 10.3 (s, 1H), 8.56 (dd, 1H, J = 2.1,
0.4 Hz), 8.48 (dd, 1H, J = 2.4, 0.4 Hz), 7.98 (dd, 1H,
J = 8.6, 0.3 Hz), 7.67 (dd, 1H, J = 2.4, 2.1 Hz), 7.10
(dd, 1H, J = 2.0, 0.3 Hz), 6.99 (dd, 1H, J = 8.6,
2.0 Hz); HRMS [EI] Calcd for C12H7Cl2NO3 (M+)
282.9803, found 282.9805.
1
NMR (CDCl3, 500 MHz) d 8.52 (d, 1H, J = 2.2 Hz),
8.49 (dd, 1H, J = 2.3, 0.3 Hz), 7.87 (t, 1H, J = 1.2 Hz),
7.86 (t, 1H, J = 1.2 Hz), 7.72–7.69 (m, 2H), 7.59 (ddd,
1H, J = 8.2, 7.6, 1.5 Hz), 7.51 (d, 1H, J = 3.8 Hz), 6.82
(d, 1H, J = 3.7 Hz); HRMS [EI] Calcd for C16H9ClN2O5
(M+) 344.0200, found 344.0197.
5.2.6. 5-Chloropyridin-3-yl 5-(2-(trifluoromethyl)phenyl)
furan-2-carboxylate (41). Method A. A white solid. H
5.2.14. 5-Chloropyridin-3-yl benzo[d][1,3]dioxole-5-car-
boxylate (49). Method A. A white solid. 1H NMR
(CDCl3, 300 MHz) d 8.50 (d, 1H, J = 2.1 Hz), 8.45 (d,
1H, J = 2.3 Hz), 7.82 (dd, 1H, J = 8.2, 1.7 Hz), 7.67
(dd, 1H, J = 2.2, 2.2 Hz), 7.59 (d, 1H, J = 1.7 Hz),
6.93 (d, 1H, J = 8.2 Hz), 6.11 (s, 2H); HRMS [EI] Calcd
for C13H8NO4Cl (M+) 277.0142, found 277.0140.
1
NMR (CDCl3, 500 MHz) d 8.53 (d, 1H, J = 2.3 Hz),
8.51 (d, 1H, J = 2.3 Hz), 8.08–8.06 (m, 1H), 8.04–8.01
(m, 1H), 7.72 (dd, 1H, J = 2.2, 2.2 Hz), 7.68–7.58 (m,
2H), 7.54 (d, 1H, J = 3.7 Hz), 6.96 (d, 1H, J = 3.7 Hz);
HRMS [EI] Calcd for C17H9ClNO3F3 (M+) 367.0223,
found 367.0223.
5.2.15. 5-Chloropyridin-3-yl 3,4-dimethoxybenzoate (50).
Method A. A white solid. H NMR (CDCl3, 300 MHz)
1
5.2.7. 5-Chloropyridin-3-yl 1H-imidazole-4-carboxylate
(42). Method A. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.51 (d, 1H, J = 2.2 Hz), 8.48 (d, 1H,
J = 2.3 Hz), 7.99–7.98 (m, 1H), 7.85–7.85 (m, 1H), 7.69
(t, 1H, J = 2.2 Hz); HRMS [EI] Calcd for C9H6ClN3O2
(M+) 223.0149, found 223.0150.
d 8.51 (d, 1H, J = 2.1 Hz), 8.47 (d, 1H, J = 2.1 Hz), 7.86
(dd, 1H, J = 8.4, 2.1 Hz), 7.67 (dd, 1H, J = 2.3, 2.1 Hz),
7.64 (d, 1H, J = 2.1 Hz), 6.98 (d, 1H, J = 8.4 Hz), 3.99
(s, 3H), 3.997 (s, 3H); HRMS [EI] Calcd for
C14H12NO4Cl (M+) 293.0455, found 293.0454.

5774
C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
5.2.16. 5-Chloropyridin-3-yl 4-aminobenzoate (51). Meth-
od B. A white solid. 1H NMR (CDCl3, 300 MHz) d 8.48
(d, 1H, J = 2.1 Hz), 8.44 (d, 1H, J = 2.3 Hz), 8.00 (d,
2H, J = 8.7 Hz), 7.67 (dd, 1H, J = 2.2, 2.2 Hz), 6.71 (d,
2H, J = 8.7 Hz), 4.50–3.60 (br s, 2H); HRMS [EI] Calcd
for C12H9N2O2Cl (M+) 248.0353, found 248.0352.
0.6 Hz), 2.47 (s, 3H); HRMS [EI] Calcd for
C13H10NO2Cl (M+) 247.0400, found 247.0398.
5.2.24. 5-Chloropyridin-3-yl 4-hydroxy-7-(trifluoromethyl)
quinoline-3-carboxylate (59). Method B. A yellow solid.
1H NMR (CD3OD, 300 MHz) d 9.03 (s, 1H), 8.55 (d,
1H, J = 2.1 Hz), 8.50 (d, 1H, J = 2.4 Hz), 8.36 (d, 1H,
J = 8.4 Hz), 8.06–8.04 (m, 1H), 8.01 (dd, 1H, J = 2.3,
2.2 Hz), 7.74 (dd, 1H, J = 8.5, 1.4 Hz), 4.00 (s, 1H);
HRMS [EI] Calcd for C16H8ClN2O3F3 (M+) 368.0176,
found 368.0176.
5.2.17. 5-Chloropyridin-3-yl 4-(methylamino)benzoate
(52). Method B. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.47 (dd, 1H, J = 2.1, 0.5 Hz), 8.45 (dd,
1H, J = 2.4, 0.5 Hz), 8.01 (d, 2H, J = 9.0 Hz), 7.67 (dd,
1H, J = 2.2, 2.2 Hz), 6.62 (d, 2H, J = 9.0 Hz), 2.95 (s,
3H); HRMS [EI] Calcd for C13H11N2O2Cl (M+)
262.0509, found 262.0513.
5.2.25. 5-Chloropyridin-3-yl benzo[d]thiazole-6-carboxyl-
ate (60). Method B. A white solid. H NMR (CDCl3,
1
300 MHz) d 9.24 (s, 1H), 8.88 (dd, 1H, J = 1.6,
0.6 Hz), 8.54 (dd, 1H, J = 2.1, 0.4 Hz), 8.52 (dd, 1H,
J = 2.4, 0.4 Hz), 8.35 (dd, 1H, J = 8.6, 1.7 Hz), 8.28
(dd, 1H, J = 8.7, 0.6 Hz), 7.73 (t, 1H, J = 2.2 Hz);
HRMS [EI] Calcd for C13H7N2O2ClS (M+) 289.9917,
found 289.9907.
5.2.18. 5-Chloropyridin-3-yl 4-(dimethylamino)benzoate
(53). A slightly yellow solid. 1H NMR (CDCl3,
500 MHz) d 8.47 (d, 1H, J = 2.0 Hz), 8.45 (d, 1H,
J = 2.3 Hz), 8.04 (d, 2H, J = 9.2 Hz), 7.67 (dd, 1H,
J = 2.2, 2.2 Hz), 6.71 (d, 2H, J = 9.2 Hz), 3.10 (s, 6H);
HRMS [EI] Calcd for C14H13N2O2Cl (M+) 276.0666,
found 276.0670.
5.2.26. N-(6-Chloropyridin-3-yl)thiophene-2-carboxamide
(61). Method B. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.47 (dd, 1H, J = 2.8, 0.6 Hz), 8.26 (ddd,
1H, J = 8.6, 2.8, 0.3 Hz), 7.71–7.68 (br s, 1H), 7.68
(dd, 1H, J = 3.8, 1.2 Hz), 7.62 (dd, 1H, J = 5.0,
1.2 Hz), 7.36 (dt, 1H, J = 8.7, 0.5 Hz), 7.17 (dd, 1H,
J = 5.0, 3.8 Hz); HRMS [EI] Calcd for C10H7N2OSCl
(M+) 237.9968, found 237.9966.
5.2.19. 5-Chloropyridin-3-yl 5-(3-nitrophenyl)furan-2-car-
boxylate (54). Method B. A white solid. 1H NMR (CDCl3,
500 MHz) d 8.65 (ddd, 1H, J = 2.2, 1.7, 0.5 Hz), 8.56–8.50
(m, 2H), 8.26 (ddd, 1H, J = 8.2, 2.2, 1.0 Hz), 8.17 (ddd,
1H, J = 7.8, 1.7, 1.0 Hz), 7.72 (t, 1H, J = 2.2 Hz), 7.68
(ddd, 1H, J = 8.3, 7.9, 0.5 Hz), 7.55 (d, 1H, J = 3.8 Hz),
7.03 (d, 1H, J = 3.7 Hz); HRMS [EI] Calcd for
C16H9ClN2O5 (M+) 344.0200, found 344.0200.
5.2.27. 5-Chloropyridin-3-yl 5-methyl-3-phenylisoxazole-
4-carboxylate (62). Method A. A white solid. H NMR
1
5.2.20. 5-Chloropyridin-3-yl 4-sulfamoylbenzoate (55).
Method A.
(CDCl3, 500 MHz) d 8.47 (d, 1H, J = 2.0 Hz), 8.28 (d,
1H, J = 2.4 Hz), 7.67- 7.64(m, 2H), 7.58 (dd, 1H,
J = 2.4, 2.0 Hz), 7.54–7.46 (m, 3H), 2.87 (s, 3H); HRMS
[EI] Calcd for C16H11ClN2O3 (M+) 314.0458, found
314.0458.
A
white solid. 1H NMR (CDCl3,
300 MHz) d 8.55 (dd, 1H, J = 2.1, 0.5 Hz), 8.50 (dd,
1H, J = 2.4, 0.5 Hz), 8.35 (d, 2H, J = 8.8 Hz), 8.10 (d,
2H, J = 8.9 Hz), 7.71 (dd, 1H, J = 2.4, 2.1 Hz), 4.92–
4.87 (br, 2H); HRMS [EI] Calcd for C12H9N2O4ClS
(M+) 311.9972, found 311.9978.
5.2.28. 5-Chloropyridin-3-yl cinnamate (63). Method A.
A white solid. 1H NMR (CDCl3, 300 MHz) d 8.49
(dd, 1H, J = 2.1, 0.4 Hz), 8.43 (dd, 1H, J = 2.3,
0.4 Hz), 7.92 (d, 1H, J = 16.1 Hz), 7.65 (t, 1H,
J = 2.2 Hz), 7.63–7.59 (m, 2H), 7.47–7.44 (m, 3H,),
6.62 (d, 1H, J = 16.0 Hz); HRMS [EI] Calcd for
C14H10NO2Cl (M+) 259.0400, found 259.0406.
5.2.21. 5-Chloropyridin-3-yl biphenyl-4-carboxylate (56).
Method A. A white solid. 1H NMR (CDCl3, 500 MHz) d
8.53 (dd, 1H, J = 2.1, 0.5 Hz), 8.51 (dd, 1H, J = 2.3,
0.5 Hz), 8.27 (d, 2H, J = 8.7 Hz), 7.77 (d, 2H,
J = 8.7 Hz), 7.72 (dd, 1H, J = 2.2, 2.2 Hz), 7.67 (dd, 2H,
J = 8.4, 1.4 Hz), 7.54–7.43 (m, 3H); HRMS [EI] Calcd.
for C18H12NO2Cl (M+) 309.0557, found 309.0556.
5.2.29. 5-Chloropyridin-3-yl 1-naphthoate (64). Method
1
A. A white solid. H NMR (CDCl3, 500 MHz) d 9.02
5.2.22. 5-Chloropyridin-3-yl 3-chlorobenzoate (57). Meth-
od A. A white solid. 1H NMR (CDCl3, 500 MHz) d 8.53
(dd, 1H, J = 2.0, 0.5 Hz), 8.48 (dd, 1H, J = 2.4, 0.5 Hz),
8.18 (ddd, 1H, J = 2.1, 1.7, 0.5 Hz), 8.08 (ddd, 1H,
J = 7.8, 1.7, 1.1 Hz), 7.69 (dd, 1H, J = 2.4, 2.1 Hz), 7.66
(ddd, 1H, J = 8.0, 2.1, 1.1 Hz), 7.50 (ddd, 1H, J = 8.0,
7.8, 0.5 Hz); HRMS [EI] Calcd for C12H7NO2Cl2 (M+)
266.9854, found 266.9858.
(ddd, 1H, J = 8.7, 1.8, 0.8 Hz), 8.55 (dd, 1H, J = 4.0,
0.5 Hz), 8.55 (t, 1H, J = 0.5 Hz), 8.51 (d, 1H,
J = 1.3 Hz), 8.50 (d, 1H, J = 1.3 Hz), 8.18–8.15 (m,
1H), 7.96 (ddt, 1H, J = 8.2, 1.5, 0.7 Hz), 7.76 (dd, 1H,
J = 2.3, 2.1 Hz), 7.69 (ddd, 1H, J = 8.4, 6.8, 1.5 Hz),
7.63–7.59 (m, 1H); HRMS [EI] Calcd for C16H10NO2Cl
(M+) 283.0400, found 283.0402.
5.2.30. 5-Chloropyridin-3-yl 1H-indole-3-carboxylate
(65). Method A. A white solid. 1H NMR (CDCl3,
300 MHz) d 8.82–8.66 (br s, 1H), 8.51 (dd, 1H,
J = 2.3, 0.4 Hz), 8.50 (dd, 1H, J = 2.1, 0.4 Hz), 8.22
(dddd, 1H, J = 6.1, 3.2, 0.8, 0.8 Hz), 8.13 (d, 1H,
J = 3.1 Hz), 7.74 (t, 1H, J = 2.2 Hz), 7.49 (ddd, 1H,
J = 6.0, 3.3, 0.8 Hz), 7.35 (dd, 2H, J = 6.1, 3.2 Hz);
5.2.23. 5-Chloropyridin-3-yl 3-methylbenzoate (58).
Method A.
500 MHz) d 8.51 (dd, 1H, J = 2.0, 0.5 Hz), 8.47 (dd,
1H, J = 2.3, 0.5 Hz), 8.02–8.00 (m, 1H), 8.00–7.98 (m,
1H), 7.68 (dd, 1H, J = 2.3, 2.1 Hz), 7.50 (dddd, 1H,
J = 7.6, 2.4, 1.3, 0.6 Hz), 7.43 (ddd, 1H, J = 8.0, 7.6,
A
white solid. 1H NMR (CDCl3,

C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
5775
HRMS [EI] Calcd for C14H9N2O2Cl (M+) 272.0353,
found 272.0349.
J = 8.4 Hz), 7.68 (dd, 1H, J = 2.2, 2.2 Hz), 7.34 (d, 2H,
J = 8.3 Hz), 2.47 (s, 3H); HRMS [EI] Calcd for
C13H10NO2Cl (M+) 247.0400, found 247.0403.
5.2.31. 5-Chloropyridin-3-yl 1H-benzo[d]imidazole-5-car-
boxylate (66). Method A. A white solid. 1H NMR
(CDCl3, 300 MHz) d 8.60–8.58 (m, 1H), 8.52 (d, 1H,
J = 2.1 Hz), 8.51 (d, 1H, J = 2.3 Hz), 8.25 (s, 1H), 8.17
(dd, 1H, J = 8.5, 1.7 Hz), 7.77 (d, 1H, J = 8.5 Hz), 7.73
(t, 1H, J = 2.0 Hz); HRMS [EI] Calcd for C13H8N3O2Cl
(M+) 273.0305, found 273.0303.
5.2.39. 5-Chloropyridin-3-yl 4-methoxybenzoate (74).
Method A.
A
white solid. 1H NMR (CDCl3,
300 MHz) d 8.50 (d, 1H, J = 2.0 Hz), 8.46 (d, 1H,
J = 2.3 Hz), 8.15 (d, 2H, J = 8.7 Hz), 7.68 (dd, 1H,
J = 2.3, 2.0 Hz), 7.01 (d, 2H, J = 8.8 Hz), 3.92 (s, 3H);
HRMS [EI] Calcd for C13H10NO3Cl (M+) 263.0349,
found 263.0351.
5.2.32. 5-Chloropyridin-3-yl 1H-benzo[d][1,2,3]triazole-5-
carboxylate (67). Method A. A white solid. H NMR
1
5.2.40. 5-Chloropyridin-3-yl 2-nitrobenzoate (75). Meth-
od A. A white solid. H NMR (CDCl3, 300 MHz) d
8.68–8.45 (m, 2H), 8.10 (dd, 1H, J = 7.6, 1.7 Hz),
7.91–7.76 (m, 3H), 7.75–7.72 (m, 1H); HRMS [EI] Calcd
for C12H7N2O4Cl (M+) 278.0094, found 278.0095.
1
(CD3OD, 300 MHz) d 8.85 (dd, 1H, J = 1.4, 0.7 Hz),
8.55 (d, 1H, J = 2.3 Hz), 8.53 (d, 1H, J = 2.1 Hz), 8.29
(ddd, 1H, J = 8.7, 1.5, 0.6 Hz), 7.99 (s, 1H), 7.98 (ddd,
1H, J = 4.9, 1.4, 0.6 Hz); HRMS [EI] Calcd for
C12H7N4O2Cl (M+) 274.0258, found 274.0260.
5.2.41. 5-Chloropyridin-3-yl 4-(diethylamino)benzoate
(76). Method B. A yellow solid. 1H NMR (CDCl3,
500 MHz) d 8.46 (d, 1H, J = 2.1 Hz), 8.44 (d, 1H,
J = 2.3 Hz), 8.01 (d, 2H, J = 9.2 Hz), 7.66 (dd, 1H,
J = 2.2, 2.2 Hz), 6.68 (d, 2H, J = 9.2 Hz), 3.46 (q, 4H,
J = 7.2 Hz), 1.24 (t, 6H, J = 7.1 Hz); HRMS [EI] Calcd
for C16H17N2O2Cl (M+) 304.0979, found 304.0978.
5.2.33. 5-Chloropyridin-3-yl 1H-indole-5-carboxylate
(68). Method A. A white solid. 1H NMR (CDCl3,
300 MHz)
d
8.60–8.58 (m, 1H), 8.50 (d, 2H,
J = 2.2 Hz), 8.48–8.40 (br s, 1H), 8.04 (dd, 1H, J = 8.7,
1.7 Hz), 7.51 (t, 1H, J = 2.2 Hz), 7.50 (dt, 1H, J = 8.6,
0.7 Hz), 7.34 (dd, 1H, J = 3.4, 2.5 Hz), 6.73 (ddd, 1H,
J = 3.3, 2.1, 1.0 Hz); HRMS [EI] Calcd for
C14H9N2O2Cl (M+) 272.0353, found 272.0354.
5.2.42. 5-Chloropyridin-3-yl 5-(4-chloro-2-nitrophenyl)
furan-2-carboxylate (77). Method A. A slightly yellow
solid. 1H NMR (CDCl3, 500 MHz) d 8.53 (d, 1H,
J = 2.0 Hz), 8.49 (d, 1H, J = 2.4 Hz), 7.85 (d, 1H,
J = 2.1 Hz), 7.83 (d, 1H, J = 8.6 Hz), 7.70 (dd, 1H,
J = 2.4, 2.1 Hz), 7.67 (dd, 1H, J = 8.4, 2.1 Hz), 7.50 (d,
1H, J = 3.8 Hz), 6.82 (d, 1H, J = 3.7 Hz); HRMS [EI]
Calcd for C16H8Cl2N2O5 (M+) 377.9810, found
377.9807.
5.2.34. 5-Chloropyridin-3-yl 3-acetoxybenzoate (69).
A
white solid. 1H NMR (CDCl3,
Method B.
300 MHz) d 8.52 (d, 1H, J = 2.1 Hz), 8.43 (d, 1H,
J = 2.4 Hz), 8.21 (dd, 1H, J = 7.9, 1.8 Hz), 7.69 (ddd,
1H, J = 8.0, 7.5, 1.7 Hz), 7.64 (t, 1H, J = 2.2 Hz), 7.42
(dt, 1H, J = 7.8, 1.2 Hz), 7.21 (dd, 1H, J = 8.1,
1.2 Hz), 2.33 (s, 3H); HRMS [ES] Calcd for
C14H11ClNO4 (MH+) 292.0371, found 292.0369.
5.2.43. 5-Chloropyridin-3-yl 5-(2-chloro-5-(trifluoro-
methyl)phenyl)furan-2-carboxylate (78). Method A. A
yellow solid. H NMR (CDCl3, 500 MHz) d 8.54 (dd,
1H, J = 2.0, 0.5 Hz), 8.52 (dd, 1H, J = 2.3, 0.5 Hz),
8.28–8.26 (m, 1H), 7.72 (t, 1H, J = 2.2 Hz), 7.67–7.64
(m, 1H), 7.61–7.58 (m, 1H), 7.56 (d, 1H, J = 3.8 Hz),
7.40 (d, 1H, J = 3.8 Hz); HRMS [EI] Calcd for
C17H8Cl2NO3F3 (M+) 402.9833, found 400.9823.
5.2.35. 5-Chloropyridin-3-yl 2-oxo-2H-chromene-3-car-
boxylate (70). Method B. A white solid. 1H NMR
(CDCl3, 300 MHz) d 8.76 (s, 1H), 8.53 (d, 1H,
J = 2.0 Hz), 8.50 (d, 1H, J = 2.2 Hz), 7.78–7.74 (m,
1H), 7.72 (t, 1H, J = 2.2 Hz), 7.70 (dd, 1H, J = 7.7,
1.4 Hz), 7.46–7.38 (m, 2H); HRMS [EI] Calcd for
C15H8NO4Cl (M+) 301.0142, found 301.0141.
1
5.2.36. 5-Chloropyridin-3-yl cyclohexanecarboxylate (71).
Method A. A white solid. 1H NMR (CDCl3, 300 MHz) d
8.45 (d, 1H, J = 2.1 Hz), 8.31 (d, 1H, J = 2.3 Hz), 7.53
(dd, 1H, J = 2.3, 2.1 Hz), 2.59 (tt, 1H, J = 11.0, 3.6 Hz),
2.12–2.02 (m, 2H), 1.90–1.75 (m, 2H), 1.75–1.50 (m,
3H), 1.44–1.26 (m, 3H); HRMS [EI] Calcd for
C12H14NO2Cl (M+) 239.0713, found 239.0717.
5.2.44. 5-Chloropyridin-3-yl 1H-pyrazole-4-carboxylate
(79). Method A. A white solid. 1H NMR (CDCl3,
500 MHz) d 10.6–10.4 (br s, 1H), 8.51 (d, 1H,
J = 2.1 Hz), 8.46 (d, 1H, J = 2.4 Hz), 8.24 (s, 2H), 7.67
(t, 1H, J = 2.2 Hz); HRMS [EI] Calcd for C9H6ClN3O2
(M+) 223.0149, found 223.0149.
5.2.45. 5-Chloropyridin-3-yl isonicotinate (80). Method
1
5.2.37. 5-Chloropyridin-3-yl 4-fluorobenzoate (72). Meth-
od A. A white solid. 1H NMR (CDCl3, 300 MHz) d 8.52
(d, 1H, J = 2.0 Hz), 8.47 (d, 1H, J = 2.3 Hz), 8.22 (dd,
2H, J = 9.1, 5.3 Hz), 7.68 (dd, 1H, J = 2.3, 2.0 Hz),
7.22 (dd, 2H, J = 9.0, 8.3 Hz)); HRMS [EI] Calcd for
C12H7NO2ClF (M+) 251.0149, found 251.0145.
B. A white solid. H NMR (CD3OD, 300 MHz) d 8.84
(d, 2H, J = 6.2 Hz), 8.54 (d, 2H, J = 2.2Hz), 8.11 (d,
2H, J = 6.2 Hz), 7.99 (t, 1H, J = 2.2 Hz); HRMS
[EI] Calcd for C11H7N2O2Cl (M+) 234.0196, found
234.0191.
5.2.46. 5-Chloropyridin-3-yl pyrazine-2-carboxylate (81).
Method B. A white solid. H NMR (CDCl3, 300 MHz)
d 9.48 (d, 1H, J = 1.5 Hz), 8.90 (d, 1H, J = 2.5 Hz), 8.84
(dd, 1H, J = 2.4, 1.5 Hz), 8.57 (d, 1H, J = 2.0 Hz), 8.55
1
5.2.38. 5-Chloropyridin-3-yl 4-methylbenzoate (73).
Method A.
300 MHz)
A
white solid. 1H NMR (CDCl3,
8.54–8.45 (m, 2H), 8.08 (d, 2H,
d

5776
C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
(d, 1H, J = 2.3 Hz), 7.74 (t, 1H, J = 2.2 Hz); HRMS [EI]
(0.002:1, w:w) overnight at 23 ꢁC. The concentrated pro-
tein was loaded onto a MonoQ 10/10 column pre-equil-
ibrated with 20 mM Tris, pH 7.5. The protein was eluted
with a gradient of 0–150 mM NaCl over 100 mL of the
equilibration buffer. The fractions containing proteinase
were combined, exchanged into 20 mM Tris, pH 7.5,
concentrated to 21 mg/mL, and then frozen as beads
in liquid nitrogen. A yield of 50 mg of 3CLpro was ob-
tained from 4 L of cell culture.
Calcd for C10H6N3O2Cl (M+) 235.0149, found 235.0149.
5.2.47. 5-Chloropyridin-3-yl 2,6-dichloro-5-fluoronicoti-
nate (82). Method B. A white solid. H NMR (CDCl3,
1
300 MHz) d 8.57 (dd, 1H, J = 2.0, 0.4 Hz), 8.50 (dd,
1H, J = 2.4, 0.4 Hz), 8.21 (d, 1H, JH–F = 7.1 Hz), 7.71
(t, 1H, J = 2.2 Hz); HRMS [EI] Calcd for
C11H4Cl3N2O2F (M+) 319.9323, found 319.9327.
5.2.48. 5-Chloropyridin-3-yl 2-naphthoate (83). Method
A. A white solid. H NMR (CDCl3, 300 MHz) d 8.80–
5.4. Fluorogenic peptide assay
1
8.78 (m, 1H), 8.54 (d, 2H, J = 2.2 Hz), 8.17 (dd, 1H,
J = 8.6, 1.8 Hz), 8.04–7.92 (m, 3H), 7.75 (t, 1H,
J = 2.2 Hz), 7.70–7.58 (m, 2H); HRMS [EI] Calcd for
C16H10NO2Cl (M+) 283.0400, found 283.0405.
The steady-state proteolytic activities of HAV 3Cpro and
SARS 3CLpro were measured using the fluorogenic pep-
tide substrates Dabcyl-GLRTQSND(edans)G and Abz-
SVTLQSGY(NO2)R, respectively. Standard assays
were performed using 0.1 M potassium phosphate, pH
7.5, 2 mM EDTA containing 0.1 lM proteinase and
10–75 lM peptide at 37 ꢁC. Initial inhibitor screening
and IC50 determination for HAV 3Cpro were done using
a 96-well plate and a Victor2 fluorescence plate reader
(PerkinElmer, Woodbridge, ON, Canada). Substrate
(10 lM) and inhibitor (0.25–10 lM) were prewarmed
for three minutes in black 96-well microplates in a total
volume of 200 ll before the reaction was started with
the addition of enzyme (kex = 355 nm, kem = 460 nm).
Determination of Ki was done using a Varian Eclipse
Fluorescence Spectrophotometer (Varian, Cary, NC,
USA). The assay was performed using a 100 lL quartz
5.3. Protein expression and purification
The HAV 3Cpro (C24S) was produced heterologously
using Escherichia coli BL21(DE3) pLysS containing
pHAV-3CEX.20 Substitution of the non-essential sur-
face cysteine residue in the C24S variant prevents inter-
molecular
disulfide
bond
formation.
Freshly
transformed cells were grown overnight at 30 ꢁC in LB
broth supplemented with 100 lg/mL ampicillin and
25 lg/mL chloramphenicol, and used to inoculate
(1:200) one litre of the same medium. The 1-L culture
was grown at 37 ꢁC to an optical density at 600 nm of
approximately 0.6 whereupon heterologous gene expres-
sion was induced by adding 0.25 mM IPTG. The cells
were incubated for a further 6 h at 30 ꢁC, harvested by
centrifugation, washed using 20 mM potassium phos-
phate, pH 6.5, containing 1 mM EDTA and 2 mM
DTT, and then frozen at À80 ꢁC until further use.
cuvette (kex = 340 nm, kem = 490 nm for SARS 3CLpro
;
k
ex = 320 nm, kem = 420 nm for SARS 3CLpro). Initial
rates were calculated using the first 3 min of the pro-
gress curve and were corrected for the inner filter effect
as described by Liu et al.,23 using SND(edans)G or Abz-
SVTLQ as appropriate. Steady-state equations were fit-
ted to the initial-rate data using the least-squares and
dynamic weighting options of LEONORA.24
To purify HAV 3Cpro, the frozen cells were resuspended
in 20 mL of 20 mM potassium phosphate, 1 mM EDTA,
2 mM DTT, pH 6.5 and disrupted using a french press
operated at 20,000 psi. Cell debris was removed by cen-
trifugation (37,000g for 30 min) and the supernatant was
passed through a 45 lm filter. The filtered cell extracted
was loaded onto a MonoS 10/10 column pre-equili-
brated with 20 mM potassium phosphate, 0.5 mM
5.5. Pyridinyl ester hydrolysis assay
Rates of pyridinyl ester hydrolysis were determined by
monitoring product formation using an HPLC. Assays
were performed at 37 ꢁC in HPLC vials containing
400 lL of 0.1 M potassium phosphate, pH 7.5, 2 mM
EDTA, 2 lM proteinase, and 10–75 lM pyridinyl ester.
Reactions were initiated with the addition of the ester.
Aliquots of 50 lL were withdrawn and analyzed using
a Waters 2695 separation module equipped with Phe-
nomenex prodigy 10 l OD-prep column and a Waters
2996 photodiode array detector. The column was oper-
ated at a flow rate of 1 mL/min and developed using the
following three-step profile: 1 mL 5% acetonitrile in
0.05% formic acid/H2O; 10 mL linear gradient of 5–
100% acetonitrile in 0.05% formic acid/H2O; and 1 mL
100% acetonitrile. Rates were calculated over the first
hour of the reaction. Enzymatic rates were corrected
for non-enzymatic hydrolysis.
¨
EDTA, pH 6.5 and operated at 3.5 mL/min an AKTA
Explorer (GE Healthcare). The proteinase was eluted
using a gradient of 80–280 mM NaCl in 96 mL of the
equilibration buffer. Eight milliliters of fractions were
collected. Those containing the proteinase, as judged
from SDS–PAGE, were combined and concentrated
using a stirred cell concentrator equipped with a
YM10 membrane (Amicon, Etobicoke, ON, Canada).
The protein solution was concentrated to 3.0 mg/mL
and frozen as beads in liquid nitrogen. Sixty milligrams
of 3Cpro was typically obtained from one liter of cell
culture.
The His-tagged SARS 3CLpro was produced and puri-
fied as described previously16 except that the His-tag
was removed. To remove the His-tag, the protein was
exchanged three times into 50 mM Tris pH 8.0 and then
amended with 100 mM NaCl and 1 mM Ca2+. The His-
tag was removed proteolytically using Factor Xa:3CL
5.6. Mass spectrometry of enzyme-inhibitor complexes
Samples of HAV 3Cpro were prepared by reacting
50 lM enzyme with 100 lM pyridinyl ester in 100 mM
potassium phosphate buffer, pH 7.5, 2 mM EDTA at

C. Huitema et al. / Bioorg. Med. Chem. 16 (2008) 5761–5777
5777
4. Yin, J.; Bergmann, E. M.; Cherney, M. M.; Lall, M. S.;
Jain, R. P.; Vederas, J. C.; James, M. N. J. Mol. Biol.
2005, 354, 854.
5. Mosimann, S. C.; Cherney, M. M.; Sia, S.; Plotch, S.;
James, M. N. J. Mol. Biol. 1997, 273, 1032.
37 ꢁC for 2 or 10 min. Samples were quenched with 25%
acetic acid and submitted to NAPS for MALDI-TOF
analysis.
5.7. Molecular docking
6. Matthews, D. A.; Smith, W. W.; Ferre, R. A.; Condon, B.;
Budahazi, G.; Sisson, W.; Villafranca, J. E.; Janson, C. A.;
McElroy, H. E.; Gribskov, C. L., et al. Cell 1994, 77, 761.
7. Birtley, J. R.; Knox, S. R.; Jaulent, A. M.; Brick, P.;
Leatherbarrow, R. J.; Curry, S. J. Biol. Chem. 2005, 280,
11520.
8. Shi, J.; Wei, Z.; Song, J. J. Biol. Chem. 2004, 279, 24765.
9. Yang, H.; Yang, M.; Ding, Y.; Liu, Y.; Lou, Z.; Zhou, Z.; Sun,
L.; Mo, L.; Ye, S.; Pang, H.; Gao, G. F.; Anand, K.; Bartlam,
M.; Hilgenfeld, R.; Rao, Z. PNAS 2003, 100, 13190.
10. Malcolm, B. A.; Lowe, C.; Shechosky, S.; McKay, R. T.;
Yang, C. C.; Shah, V. J.; Simon, R. J.; Vederas, J. C.;
Santi, D. V. Biochemistry 1995, 34, 8172.
11. Morris, T. S.; Frormann, S.; Shechosky, S.; Lowe, C.;
Lall, M. S.; Gauss-Muller, V.; Purcell, R. H.; Emerson, S.
U.; Vederas, J. C.; Malcolm, B. A. Bioorg. Med. Chem.
1997, 5, 797.
12. Ramtohul, Y. K.; James, M. N. G.; Vederas, J. C. J. Org.
Chem. 2002, 67, 3169.
Automated molecular docking was performed using the
program AutoDock 3.0.5 and a recent structural model
of HAV 3Cpro (PDB: 2CXV)4 as the target. The 3D
structures of ester compounds were constructed using
Sybyl 7.1 (Tripos Inc., St. Louis, USA). Inhibitor-en-
zyme interactions were evaluated using the Lamarckian
genetic algorithm (LGA). Binding energies between the
compounds and the protein were evaluated using a grid
˚
map with 60 · 60 · 60 points spaced at 0.375 A gener-
ated using the AUTOGRID program. Default values
of the docking parameters were used except for the fol-
˚
lowing: step sizes for translation (0.2 A), orientation
(5ꢁ), and torsion (5ꢁ); and the number of generations
(37,000); energy evaluations (1,500,000), and docking
runs (20). The docked inhibitor–enzyme complexes were
ranked according to the lowest predicted binding ener-
gies and the conformity to ideal geometry of the docked
structures.
13. Yin, J.; Cherney, M. M.; Bergmann, E. M.; Zhang, J.;
Huitema, C.; Pettersson, H.; Eltis, L. D.; Vederas, J. C.;
James, M. N. J. Mol. Biol. 2006, 361, 673.
14. Jain, R. P.; Vederas, J. C. Bioorg. Med. Chem. Lett. 2004,
14, 3655.
15. Lall, M. S.; Ramtohul, Y. K.; James, M. N. G.; Vederas,
J. C. J. Org. Chem. 2002, 67, 1536.
Acknowledgments
16. Blanchard, J. E.; Elowe, N. H.; Huitema, C.; Fortin, P.
D.; Cechetto, J. D.; Eltis, L. D.; Brown, E. D. Chem. Biol.
2004, 11, 1445.
17. Zhang, J.; Pettersson, H. I.; Huitema, C.; Niu, C.; Yin, J.;
James, M. N.; Eltis, L. D.; Vederas, J. C. J. Med. Chem.
2007, 50, 1850.
18. Seah, S. Y.; Labbe, G.; Nerdinger, S.; Johnson, M. R.;
Snieckus, V.; Eltis, L. D. J. Biol. Chem. 2000, 275, 15701.
19. Jewell, D. A.; Swietnicki, W.; Dunn, B. M.; Malcolm, B.
A. Biochemistry 1992, 31, 7862.
20. Malcolm, B. A.; Chin, S. M.; Jewell, D. A.; Stratton-
Thomas, J. R.; Thudium, K. B.; Ralston, R.; Rosenberg,
S. Biochemistry 1992, 31, 3358.
Financial support from the Natural Sciences and Engi-
neering Research Council of Canada (Discovery grants
to L.D.E. and J.C.V.), the Canada Research Chairs Pro-
gram (CRC in Protein Structure & Function and CRC
in Bioorganic & Medicinal Chemistry), and the Alberta
Heritage Foundation for Medical Research (AHFMR)
is gratefully acknowledged. C.H. is the recipient of an
NSERC fellowship. J.Y. is the recipient of AHFMR
and Killam Foundation fellowships. We thank Suzanne
Perry for help with mass spectrometry and Geoff Hors-
man for helpful discussions.
21. Niu, C.; Yin, J.; Zhang, J.; Vederas, J. C.; James, M. N.
Bioorg. Med. Chem. 2007, 16, 293.
22. Lee, T. W.; Cherney, M. M.; Huitema, C.; Liu, J.; James,
K. E.; Powers, J. C.; Eltis, L. D.; James, M. N. J. Mol.
Biol. 2005, 353, 1137.
23. Liu, Y.; Kati, W.; Chen, C. M.; Tripathi, R.; Molla, A.;
Kohlbrenner, W. Anal. Biochem. 1999, 267, 331.
24. Cornish-Bowden, A. Analysis of Enzyme Kinetic Data;
Oxford University Press: Oxford, New York, 1995.
References and notes
1. Tong, L. Chem. Rev. 2002, 102, 4609.
2. Allaire, M.; Chernaia, M. M.; Malcolm, B. A.; James, M.
N. Nature 1994, 369, 72.
3. Bergmann, E. M.; Mosimann, S. C.; Chernaia, M. M.;
Malcolm, B. A.; James, M. N. J. Virol. 1997, 71, 2436.