Discovery of (1 H-Pyrazolo[3,4- c]pyridin-5-yl)sulfonamide Analogues as Hepatitis B Virus Capsid Assembly Modulators by Conformation Constraint

Article abstract of DOI:10.1021/acs.jmedchem.0c00292

Hepatitis B virus (HBV) capsid assembly modulators (CAMs) have been suggested to be effective anti-HBV agents in both preclinical and clinical studies. In addition to blocking HBV replication, CAMs could reduce the formation of covalently closed circular DNA (cccDNA), which accounts for the persistence of HBV infection. Here, we describe the discovery of (1H-indazole-5-yl)sulfonamides and (1H-pyrazolo[3,4-c]pyridin-5-yl)sulfonamides as new CAM chemotypes by constraining the conformation of the sulfamoylbenzamide derivatives. Lead optimization resulted in compound 56 with an EC50 value of 0.034 μM and good metabolic stability in mouse liver microsomes. To increase the solubility, the amino acid prodrug (65) and its citric acid salt (67) were prepared. Compound 67 dose dependently inhibited HBV replication in a hydrodynamic injection-based mouse model of HBV infection, while 56 did not show in vivo anti-HBV activity, likely owing to its suboptimal solubility. This class of compounds may serve as a starting point to develop novel anti-HBV drugs.

Full text of DOI:10.1021/acs.jmedchem.0c00292

Subscriber access provided by Uppsala universitetsbibliotek
Article
Discovery of (1H-Pyrazolo[3,4-c]pyridin-5-yl)sulfonamide Analogs as
Hepatitis B Virus Capsid Assembly Modulators by Conformation Constraint
Chunting Wang, Yameng Pei, Lin Wang, Shuo Li, Chao
Jiang, Xu Tan, Yi Dong, Ye Xiang, Yao Ma, and Gang Liu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00292 • Publication Date (Web): 18 May 2020
Downloaded from pubs.acs.org on May 18, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of (1H-Pyrazolo[3,4-c]pyridin-5-
yl)sulfonamide Analogs as Hepatitis B Virus
Capsid Assembly Modulators by Conformation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Constraint
Chunting Wang1,#, Yameng Pei1,3#, Lin Wang4, Shuo Li5, Chao Jiang5, Xu Tan5, Yi
Dong2, Ye Xiang4,*,Yao Ma2,*, Gang Liu1,*
1, School of Pharmaceutical Sciences, Tsinghua University, Beijing 100084, China.
2, Institute of Materia Medica, Chinese Academy of Medical Sciences & Peking Union
Medical College, Beijing 100050, China.
3, Tsinghua-Peking Center for Life Sciences, Tsinghua University, Beijing 100084,
China.
4, Beijing Advanced Innovation Center for Structural Biology, Collaborative
Innovation Center for Diagnosis and Treatment of Infectious Diseases, Center for
Global Health and Infectious Diseases, Department of Basic Medical Sciences, School
of Medicine, Tsinghua University, Beijing 100084, China.
5, Beijing Advanced Innovation Center for Structural Biology, Beijing Frontier
Research Center for Biological Structure, School of Pharmaceutical Sciences, Center
for Infectious Disease Research, School of Medicine, Tsinghua University, Beijing
100084, China.
Key words: Hepatitis B; capsid assembly; conformation constraint; pro-drug
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 86
1
2
3
4
5
6
7
8
9
Abstract: Hepatitis B virus (HBV) capsid assembly modulators (CAMs) have been
suggested to be effective anti-HBV agents in both preclinical and clinical studies. In
addition to blocking HBV replication, CAMs could reduce the formation of covalently
closed circular DNA (cccDNA), which accounts for the persistence of HBV infection.
Here, we describe the discovery of (1H-indazole-5-yl)sulfonamides and (1H-
pyrazolo[3,4-c]pyridin-5-yl)sulfonamides as new CAM chemotypes by constraining
the conformation of the sulfamoylbenzamide derivatives. Lead optimization resulted in
compound 56 with an EC50 value of 0.034 μM and good metabolic stability in mouse
liver microsomes. To increase the solubility, the amino acid prodrug (65) and its citric
acid salt (67) were prepared. Compound 67 dose-dependently inhibited HBV
replication in a hydrodynamic injection-based mouse model of HBV infection, while
56 did not show in vivo anti-HBV activity, likely owing to its sub-optimal solubility.
This class of compounds may serve as a starting point to develop novel anti-HBV drugs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Introduction
Chronic hepatitis B (CHB), caused by infection of the hepatitis B virus (HBV), is
one of the major infectious diseases in the world, leading to approximately 887,000
deaths in 2015.1 Although the worldwide application of efficient HBV vaccines greatly
decreased morbidity, approximately 257 million people still live with CHB. CHB
tremendously increases the risk of developing cirrhosis and hepatic carcinoma, which
are the leading causes of mortality for CHB patients.1 Two classes of drugs,
nucleos(t)ide analogs (NAs) and interferon (IFN), have been approved for the treatment
of CHB. For most patients, entecavir and two prodrugs of tenofovir, namely tenofovir
2
ACS Paragon Plus Environment

Page 3 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
disoproxil fumarate and tenofovir alafenamide, are recommended as the first-line
treatment because of their high potency, good safety profile and very low probability of
developing resistance after long-term treatment.2 However, these NAs only inhibit the
replication of HBV and cannot disturb its persistence in hepatocytes. HBV surface
antigen (HBsAg) loss, which is considered as a functional cure, is only achieved in less
than 10% of patients after five years of treatment.3 The major reason for HBV
persistence is the formation of covalently closed circular DNA(cccDNA) in the nucleus
of host cells, which is very stable and acts as a template for HBV transcription.4, 5 IFN
has been reported to decrease the amount of cccDNA6 and result in HBsAg
seroclearance more frequently than NAs with a 58% cure rate in genotype A infection.7,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
However, the severe side effects and limited efficiency of other genotypes of HBV
infection have hindered its use in the clinic.2
Capsid assembly modulators (CAMs) have recently been proven to be effective for
blocking HBV replication in preclinical and clinical studies.8 The HBV capsid
comprises 120 core protein (Cp) dimers arranged with icosahedral T4 symmetry.9 Core
nucleation is the rate limiting step of the self-assembly,10 during which three core
protein dimers interact with HBV polymerase and pregenomic RNA (pgRNA) to form
a precursor complex.11, 12 Then, the Cp dimers are rapidly added to this precursor and
interact with each other through hydrophobic interactions to form the capsid. The weak
interaction between the subunits enables error correction in the capsid assembly, which
is important for correct assembly.13, 14 The HBV genome is synthesized only in the
properly formed capsid via reverse transcription using pgRNA as the template.15 In
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 86
1
2
3
4
5
6
7
8
9
addition to enabling reverse transcription, capsid assembly or core proteins regulate
almost every step in the HBV life cycle, e.g., transporting the HBV genome to the
nucleus and epigenetic regulation of HBV gene expression,16, 17 making capsid
assembly one of the most attractive targets in anti-HBV drug discovery.18
The weak interaction between Cps also provides opportunities for regulation of the
capsid assembly process by small molecules. CAMs, through kinetically and
thermodynamically modulating capsid assembly, could disturb the proper formation of
capsids and regulate the distribution and accessibility of Cps,19 which affords additional
therapeutic benefits besides blocking HBV replication. Several CAMs have been
reported to inhibit the de novo establishment of cccDNA by modulating the stability of
the capsid.20-24 Long-term treatment with CAMs could inhibit the expression of HBV e
antigen (HBeAg) and HBsAg, and the clearance of the latter has been considered a
functional cure of CHB.21 Thus, CAMs could potentially increase the functional cure
rate for CHB by modulating multiple steps in the HBV life cycle.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Several chemotypes of CAMs have been discovered, among which
phenylpropenamides (PPAs, e.g., AT130),25 sulfamoylbenzamide (SBA, e.g., DVR-
23)26 and heteroarylpyrimidines (HAPs, e.g., BAY41-4109)27 (Figure 1) are the most
studied. Although they each have different chemical structures, all these CAMs bind to
the same binding pocket at the Cp dimer-dimer interface.28-30 They act as molecular
glues to enhance the interaction between the Cps, thus accelerating the self-assembly
of Cps and decreasing the formation of functional capsids. However, DVR-23 and
AT130 promote the formation of morphologically normal but empty capsids (type II
4
ACS Paragon Plus Environment

Page 5 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CAMs, CAM-II), while BAY41-4109 misdirects core proteins to form aberrant
structures (type I CAMs, CAM-I). This indicates that the difference in the interaction
with Cp accounts for the varying effects of CAMs. Both CAM-II (JNJ-6379, Phase II
trial; AB-423, Phase I trial) and CAM-I (GLS4, Phase II trial) have been evaluated in
clinical trials.18. However, no compounds have been approved. Discovery of new
chemotypes of CAMs would provide additional candidates, promoting the development
of CAMs for new CHB therapeutics. In addition, different chemotypes could
differentially modulate the kinetics or thermodynamics of capsid assembly,31 providing
variable effects against HBV compared with present CAMs. Here, we describe the
discovery of (1H-indazole-5-yl)sulfonamide and (1H-pyrazolo[3,4-c]pyridin-5-
yl)sulfonamide as new chemotypes of CAMs by constraining the conformation of
SBAs. The (1H-pyrazolo[3,4-c]pyridin-5-yl)sulfonamide analogs 56 and 59 are less
cytotoxic and approximately 10-fold more potent than SBA analogs with good
metabolic stability in mouse liver microsomes. The citric acid salt (67) of N,N-
dimethylglysine prodrug of 56 has good solubility at various pHs, and dose dependently
inhibits in vivo HBV replication in a hydrodynamic injection-based mouse model.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Structures of CAMs.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 86
1
2
3
4
5
Results
6
7
1. Compound design and optimization
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Design of novel CAMs through conformation constraint. (A) Crystal
structure of Cps binding with SBA-R01 (PDB: 5T2P); (B) Rational design of CAMs;
(C) Merging of SBA-R01 and HAP-R01 (PDB: 5WRE) in the binding site.
Sulfamoylbenzamide (SBA) was recently discovered as a new CAM chemotype
from high-throughput screening.26 Different from heteroarylpyrimidines (HAPs), SBA
treatment induced the formation of morphologically normal capsids, which is
characteristic of type II CAMs.19 SBAs bind to the same pocket as HAP at the Cp dimer-
dimer interface (Figure 2C). The halogen-substituted aniline moiety occupies a
hydrophobic pocket. The amide forms hydrogen bonds with W102 and T128 (Figure
2A and 2B),30 which are essential for the inhibitory activity.32 The sulfonamide moiety
was exposed to the solvent and can tolerate various substituents.32, 33 In addition to SBA
and HAP, the PPA compound AT-13028 and several other recently discovered CAMs34,
35
have been reported to bind to the same pocket, indicating that various chemical
scaffolds can be accommodated in this pocket. In the active conformation of SBA, the
oxygen atom of the amide and carbon A in the central phenyl are close in space, as
6
ACS Paragon Plus Environment

Page 7 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
shown in Figure 2A. To constrain this conformation, we considered whether C=O and
carbon A could be cyclized to obtain a new scaffold in which the N atom in the aromatic
heterocycle was designed to interact with W102 by hydrogen bonding in place of the
C=O (Figure 2B). Various rings, such as phenyl or heterocycles, could be included in
the ring A position, which enables structural diversity in lead discovery and
optimization. To prove our idea, compounds with indazole or quinazoline scaffold were
synthesized. Quinazoline derivatives did not inhibit HBV replication. Compound 1 with
indazole scaffold (Table 1) inhibited extracellular HBV DNA in HepAD38 cells with
an EC50 value of 1.28 μM. We proposed that the 2-N atom at indazole scaffold could
form a proper hydrogen bond with W102 of HBV core protein. In addition, the 1-N
atom provided an additional site for structure optimization, which distinguishes
indazole from other scaffolds. The 1H-indazole-5-sulfonamide compounds were once
reported as anti-HBV agents in a patent,36 which increases our confidence in the design
concept. However, the SAR and mechanism of this class of compounds are yet to be
elucidated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The effect of substitution at the 1-position was first explored (Table 1). N-
methylation resulted in compound 2 with similar anti-HBV activity (EC50=1.38 μM).
However, the selectivity index (SI) of compound 2 increased compared with that of 1.
N-cyclopropanation greatly increased the potency compared to 1 by approximately 5-
fold (3, EC50=0.19 μM). The increase in activity may result from the occupation of a
small hydrophobic pocket, which is close to carbon A and occupied by the 2-substituent
of HAP derivatives (Figure 1C).30 N-methylation of the aniline moiety (4) and
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 86
1
2
3
4
5
6
7
replacement of the aniline with phenol (5), which prevents the hydrogen bond
interaction with T128, completely abolished the activity.
8
9
Table 1. Anti-HBV activity and cytotoxicity of 1H-indazole-5-sulfonamide analogs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compd.
R1
H
X
N
N
N
N
O
-
R2
H
H
H
CH3
-
EC50 (μM)a
1.28
CC50 (μM)b
24
SIc
18.75
>72
96
1
2
CH3
1.38
>100
3
4
0.19±0.10
>10
18.3
>100
-
5
>10
75
<7.5
129
3785
AB-423
GLS4
-
-
-
0.31±0.092
39.85±5.11d
-
-
0.013±0.0045 49.2±1.30d
a The concentration of 50% inhibition for HBV DNA in HepAD38 cell supernatant; b
The concentration that kills 50% HepAD38 cells; c selectivity index, SI=CC50/EC50; d
CC50 in HepG2 cells.
8
ACS Paragon Plus Environment

Page 9 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Compound 6 and its effect on capsid assembly. (A) Structure of compound 6;
(B) and (C) Compound 6 accelerated capsid assembly as detected by EM (B) and SEC
(C); (D) Calculation of the peak AUC and SEC peak AUC ratio in Figure 3C. Cp149
protein (5 μM) was incubated with 6 (10 μM) in a buffer containing 150 mM NaCl and
50 mM HEPES (pH 7.4) at 37 °C overnight prior to EM or SEC detection.
Although potent inhibition activity was achieved, the selective index of these
derivatives was relatively low (Table 1). Interestingly, replacement of 3-chloro-4-
fluoroaniline moiety with 3,4,5-trifluoroaniline greatly decreased cytotoxicity (6,
Figure 3A). To detect the effect of compound 6 on capsid assembly, electron
microscopy (EM) and size exclusion chromatography (SEC) analyses were performed.
The N-terminal domain of the core protein (aa 1-149, 5 μM) was incubated with 6 (10
μM) in a buffer containing 150 mM NaCl. EM images showed that more capsids were
formed after treatment with 6 compared with DMSO (Figure 3B). Two peaks of the
DMSO treated core protein appeared in the SEC analysis (Figure 3C), indicating the
co-existence of the capsids (left peak, elution volume: 1.08 mL, early fraction) and the
core protein dimers (right peak, elution volume: 1.97 mL, late fraction).The ratio
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 86
1
2
3
4
5
6
7
8
9
between the left and right peak area under the curve (AUC) (capsid peak AUC/core
protein dimers peak AUC) was used to evaluate the ability of compound 6 to accelerate
the capsid assembly. As shown in Figure 3D, the SEC peak AUC ratio of compound 6
was 1.43 while DMSO was 0.59, indicating that compound 6 promoted more core
protein dimers to assemble into capsids than DMSO did, consistent with the EM results.
The morphology and size of the capsid formed after treatment with 6 was the same as
that of the normal capsid, which was characteristic of type II CAMs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Anti-HBV activity and cytotoxicity of the compounds generated by scaffold
modification.
Compd.
Core
EC50 (μM)
2.0
CC50 (μM)
SI
>50
>769
-
7
8
7-F
6-F
4-F
>100
>100
>100
-
0.13
>20
9
X=Y=CH
Z=N
X=Z=CH
Y=N
Y=Z=CH
X=N
10
11
12
>20
-
0.24
>5
>100
-
>417
-
To explore the structural diversity of ring A (Figure 2B), we optimized the scaffold
by fluorine scan and N-walking strategy, which is widely applied in lead optimization.37,
38 In this article, fluorination and N-replacement were performed on the three positions
of the indazole scaffold (Table 2, 7-12). Compounds 8 and 11 were comparable in
10
ACS Paragon Plus Environment

Page 11 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potency and cytotoxicity to 6, with EC50 values of 0.13 and 0.24 µM, respectively, while
modification at the other positions was not tolerable. Compounds 8 and 11 were thus
considered a proper starting point for structure optimization for the potential
improvement in drug-like property. Due to its facile synthesis compared to 8 and 11,
indazole analog 6 was used as a template for the SAR study. The SAR information was
then applied to 8 and 11, as we proposed that modification of the scaffold should not
influence the interaction between the other parts of the molecule and Cp.
R4 (Table 3) was supposed to bind to the hydrophobic pocket at Cp dimer interface
as the halogen-substituted phenyl in SBA analogs does. Thus, the steric match, which
is determined by the substituents on the phenyl group, is vital to the interaction between
R4 and this hydrophobic pocket.39 Removal of the fluorine at the meta position (13,
EC50=0.22 μM, Table 3) was tolerable and maintained potency compared with 6. The
4-fluorophenyl group is a common fragment in several chemotypes of CAMs.26, 32-34 In
this case, replacement of the p-fluorine with a larger group, such as chlorine (15) and
methyl (14), led to a ten-fold decrease or even a loss in activity, probably due to steric
hindrance. Removal of the para fluorine (16) also resulted in diminished activity
compared with 13. These data suggested that the para fluorine-substituted aniline is
essential for ligand-target interactions. Introducing chlorine and methyl groups at the
2-position reduced the activity by more than five-fold (17 and 18). The addition of
another fluorine or chlorine at the 3-position, which was expected to enhance
hydrophobic interactions, did not significantly influence the anti-HBV activity (19 in
Table 3, 3 in Table 1), suggesting that a 3-fluorine or 3-chlorine group did not cause
11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 86
1
2
3
4
5
6
7
8
9
steric hindrance and was not perfectly matched with the target. Inspired by the above
result, we replaced the 3-substituents with a larger group, such as methyl, leading to 20,
which was approximately ten-fold more potent than 13 with an EC50 value of 0.032 µM.
Although Cl and CH3 are isosteres with a similar size, the Van der Waals radius of CH3
(2 Å) is slightly larger than Cl (1.75 Å). As F, an electron withdrawing atom like Cl,
significantly contributed to the increase in potency, we proposed that the lower potency
of 3 compared to 20 is, if not all, partly due to the smaller volume of Cl than CH3 instead
of the electron withdrawing effect of Cl. Replacing the phenyl with a pyridinyl (22, 23
and 24) or introducing a polar group (CN in 21) resulted in a loss of activity, consistent
with the hydrophobic nature of the pocket accommodating the aniline moiety. To
increase the flexibility of the compound, a methylene was inserted between the phenyl
and the NH (25 and 26). The benzyl amine fragment was also present in other SBA
analogs.26 However, benzyl amine substituents were not tolerable with the indazole
scaffold (Table 3). Aliphatic substituents also resulted in a great decrease in potency
(27 and 28). Although 20 was the most potent compound obtained from the
optimization of R4, it was somewhat more cytotoxic than the others.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12
ACS Paragon Plus Environment

Page 13 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
Table 3. Effect of aniline substituents on anti-HBV activity and cytotoxicity.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R4
R5
-
CC50
(μM)
18.3
>100
ND
Compd
EC50 (μM)
SI
R6
R7
3
0.19±0.10
0.22
>5
96
>455
-
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
F
H
H
H
H
CH3
Cl
H
F
H
H
3.14
0.74
1.15
4
6.9
2.2
44
>87
>25
625
343
-
H
H
32.2
>100
>100
75
H
CH3
Cl
H
F
H
F
F
0.12
0.032
>10
>20
>10
>10
>10
>10
>10
>10
F
CH3
CN
H
11.0
ND
F
H
46.6
ND
<23
-
---
---
H
H
CH3
---
---
ND
-
---
---
ND
-
F
---
ND
-
100.2
>100
<10
-
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 86
1
2
3
4
5
6
7
8
9
To reduce the LogP value of the compounds, the cyclopropane was replaced with
trimethylene oxide (29, Table 4). This led to a loss in activity, suggesting the
requirement of hydrophobic substituents at R1. Replacement of the cyclopropane with
a smaller (ethyl, 30) or larger (isopropyl in 31 or cyclobutyl in 32) group significantly
diminished anti-HBV activity with EC50 values >1 µM. As the sulfonamide moiety was
supposed to bind in the solvent exposed region and interact with the target through a
water-mediated network,30 fragments containing hydrophilic groups, which enable the
interaction with water, were included. Similar anti-HBV activity but varied cytotoxicity
was achieved with various substitutions at the R3 position. Replacement of the 4-
hydroxyl piperidine with morpholine led to compound 33 with similar potency
(EC50=0.43 µM). Moving the hydroxyl group from the 4-position to the 3-position of
piperidine did not affect the activity (EC50=0.25 µM) but greatly increased the
cytotoxicity (34). Carboxylic acid at R3 was detrimental to anti-HBV activity (35).
Replacement of the piperidine with pyrrolidine (36) or insertion of one carbon atom
between the hydroxyl and piperidine (37) reduced the potency by 10-fold and increased
the cytotoxicity. Other groups containing oxygen at the 4-position of piperidine (38 and
39) were well-tolerated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14
ACS Paragon Plus Environment

Page 15 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
Table 4. Effects of R1 and R3 on anti-HBV activity and cytotoxicity.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compd.
14
R1
R3
EC50 (μM)
0.22
>10
CC50 (μM)
>100
ND
SI
>455
-
cyclopropanyl 4-hydroxypiperidinyl
4-hydroxypiperidinyl
29
ethyl
4-hydroxypiperidinyl
4-hydroxypiperidinyl
4-hydroxypiperidinyl
morpholinyl
30
1.26
>10
>100
ND
>79
-
isopropanyl
cyclobutanyl
cyclopropanyl
31
32
≈8
10.79
>100
19.06
>100
≈60
≈1.3
>232
76.24
-
33
0.43
0.25
>10
cyclopropanyl 3-hydroxypiperidinyl
cyclopropanyl
34
35
cyclopropanyl 3-hydroxypyrrolidinyl
cyclopropanyl
36
2.1
≈28
23
37
1.1
24.77
>100
14.25
cyclopropanyl
38
0.25
0.33
>400
43
cyclopropanyl
39
Considering both HBV inhibition activity and cytotoxicity, we combined the optimal
substituents from Tables 3 and 4. As described above, the introduction of a 3-methyl
group on the aniline moiety (20, Table 3) greatly improved the antiviral activity. We
also found that different substitutes on the sulfonamide moiety resulted in variations in
cytotoxicity. Compound 40 (Table 5), designed by combining 3-methyl-4-fluorophenyl
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 86
1
2
3
4
5
6
7
8
9
at the R4 position with morpholinyl at the R3 position, inhibited HBV replication with
an EC50 value of 0.057 µM without cytotoxicity (CC50>100 μM). Introduction of these
two substitutes into the scaffold of 11 but not 8 improved the antiviral activity by 5-fold
without increasing cytotoxicity (42 and 41, Table 5). Compound 42 inhibited HBV
replication with an EC50 value of 0.042 μM, without cytotoxicity (CC50>100 μM).
Therefore, compounds 40 and 42 were selected for further optimization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. Anti-HBV activity and cytotoxicity of compounds generated by combining the
optimal substituents with different scaffolds.
Compd.
20
EC50 (μM) CC50 (μM)
HLM/MLM Clint (µL/min/mg)a
NDb
41.0/26.9
ND
11.0
0.032
40
0.057
0.21
>100
>100
>100
41
42
0.042
38.0/53.5
a The intrinsic clearance of compounds by human or mouse liver microsome
(HLM/MLM). b ND, not determined.
16
ACS Paragon Plus Environment

Page 17 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
Table 6. Structural modifications to improve the metabolic stability.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CC50
(μM)
>100
8.42
SI
Compd. Class
R1
R3
R4
cLogPa
EC50 (μM)
43
46
47
49
morpholinyl
morpholinyl
F
3.51
0.22±0.17
0.29±0.16
6.00
>454
29
CF3
CF3
-
-
4-hydroxypiperidinyl
4-hydroxypiperidinyl
≈7.52
≈5.86
1.3
3.83
0.68
8.6
422
54
CH3
3.06
0.045±0.021
19.0
57
58
50
51
52
53
55
44
45
48
56
59
60
4-hydroxypiperidinyl CF2H
3.22
3.55
-
0.29
0.12
13.55
50
47
A
morpholinyl
morpholinyl
CF2H
CH3
CH3
CH3
CH3
CH3
F
417
CF3CH2
CF2HCH2
CFH2CH2
CF2H
0.24
>100
>100
>100
>100
50.0
>417
>435
>244
≈10
morpholinyl
3.40
3.15
3.27
2.84
2.93
3.23
-
0.23
morpholinyl
0.41
morpholinyl
>10
H
morpholinyl
0.29±0.14
0.25±0.11
0.22
172
4-hydroxypiperidinyl
morpholinyl
>100
>100
>100
>100
>100
>100
>400
>454
>294
2941
1408
435
F
morpholinyl
CF3
CH3
0.34
B
4-hydroxypiperidinyl
2.78
3.04
3.12
0.034±0.012
0.071±0.035
0.23
4-hydroxypiperidinyl CF2H
morpholinyl CF2H
a Calculated by the ACD/Percepta platform.
The metabolic stability in liver microsomes of 40 and 42 was low (Table 5). To
improve the metabolic stability, we replaced the methyl group at R4 in 40 and 42, which
is prone to be oxidized by CYP450,39 with a F (43-45, Table 6), CF3 (46-48) or
cyclopropanyl group (49). These modifications resulted in a 5- to 100-fold decrease in
potency compared with 40 and 42 (Table 6). This was most likely because of the
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 86
1
2
3
4
5
6
7
8
9
improper steric match of these groups with Cp. The cyclopropanyl group at the 1-
position is another site prone to metabolism through N-dealkylation mediated by
CYP450.39 We replaced cyclopropanyl with fluorine-substituted methyl or ethyl groups,
which are less liable to CYP450-mediated oxidation due to the electron-withdrawing
effects of fluorine and the higher strength of the C-F bond.37 Although less potent than
40, compounds 50-52 inhibited HBV replication with submicromolar EC50 values.
However, difluoromethyl substitution (53) significantly decreased the anti-HBV
activity, probably because the strong electron-withdrawing effect of the difluoromethyl
group reduced the hydrogen bond acceptor ability of the N atom at the 2-position. All
these attempts to block metabolism sites were discontinued due to the accompanying
decreases in potency. Another popular approach to improve metabolic stability is to
reduce the lipophilicity of the compound.40, 41 Replacing the morpholine with its one-
carbon bridged analog (54), which has been reported to have lower lipophilicity,42
reduced the cLogP by 0.85 units compared to 40. Compound 54 inhibited HBV
replication with an EC50 value of 0.045 μM, similar to 40. However, the cytotoxicity
greatly increased compared to 40 (CC50 19.0 vs. >100 μM). The 3-methyl group at R4
in compounds 40 and 42 was found to significantly contribute to the increased anti-
HBV activity (Table 3). We intended to investigate the contribution of cyclopropane in
40. The removal of cyclopropane resulted in a 1.07-unit reduction in cLogP (55, Table
6). Compound 55 inhibited HBV replication at an EC50 of 0.25 μM, which is less potent
than 40, indicating that both the cyclopropane and 3-methyl groups significantly
contribute to the anti-HBV activity. Optimization of the solvent-exposed sulfonamide
18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 19 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
group was another approach to reduce the lipophilicity without significantly decreasing
potency. Unfortunately, most analogs of 40 with various substituents at the R3 position
were cytotoxic with CC50 values < 30 μM (Supplementary Table S1). Compound 56,
designed by replacing the morpholine in compound 42 with the more hydrophilic 4-
hydroxypiperidine, inhibited HBV replication with an EC50 value of 0.042 μM, without
cytotoxicity at 100 μM in HepAD38 cells. Interestingly, the metabolic stability of 56 in
mouse liver microsome (MLM)s was significantly improved compared to 42 (Table 7).
To block metabolism at the benzyl site, we replaced methyl with a difluoromethyl (57-
60), a less lipophilic bioisostere of the methyl group.43 The antiviral activity of these
compounds decreased slightly. Indazole derivatives 57 and 58 were found to be more
cytotoxic than the 1H-pyrazolo[3,4-c]pyridine derivatives 59 and 60 (Table 6),
consistent with the above result of the matched pair compounds 20 and 56. Although
replacement of methyl with difluoromethyl did not reduce the lipophilicity, we
observed that the metabolic stability of 59 increased compared with 56 (Table 7).
Table 7. Mouse liver microsome metabolism of 56 and 59.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compd.
56
EC50 (µM)
0.034±0.012
0.071±0.035
CC50 (µM)
>100
MLM Clint (µL/min/mg)a
10.3
6.90
59
>100
a The intrinsic clearance of compounds by mouse liver microsome (MLM).
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 86
1
2
3
4
5
2. In vitro effects of compounds 56 and 59 in HepAD38 and HepG2.2.15 cells
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. In vitro effects of compounds 56 and 59 in HepAD38 and HepG2.2.15 cells.
(A) EM analysis of the effects on capsid assembly. (B) and (C) Changes of HBV
extracellular and intracellular parameters in HepAD38 and HepG2.2.15 cells after
treatment with AT130, 3TC, GLS4, 56 and 59 at 2 µM. Cells were treated with
compounds or DMSO for 4 days. The medium was replaced every 2 days. DNA and
RNA were quantified with qPCR or qRT-PCR. HBeAg and HBsAg were quantified by
ELISA. VC, vehicle control. The results are presented as the means ± SD. Statistics
were determined by two-tailed t-test. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
As shown in Figure 4A, compounds 56 and 59 accelerated the capsid assembly and
20
ACS Paragon Plus Environment

Page 21 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
behaved as type II CAMs. To clarify their effect on the HBV life cycle, we treated
HepAD38 and HepG2.2.15 cells with 56, 59, positive control CAMs (AT130 and GLS4)
and Lamivudine (3TC) at 2 µM for four days. At the extracellular level, all these
compounds inhibited HBV replication in both cell lines, reducing extracellular HBV
DNA by approximately 90% at 2 µM (Figure 4B). None of the compounds, including
the positive controls, could inhibit the secretion of HBsAg in either cell line. The
inhibition of HBsAg secretion by CAMs was only observed in the de novo infection
models after long-term treatment,20, 21 resulting from the inhibition of de novo synthesis
of cccDNA. The secretion of HBeAg by HepAD38 cells was reduced by approximately
50% after treating with 4 CAMs. However, only GLS4 could weakly decrease HBeAg
in HepG2.2.15 cells (Figure 4B). Inhibition of HBeAg was most probably due to the
decrease in intracellular RNA accumulation (Figure 4C) or a direct effect on HBeAg
by CAMs, which could inhibit HBeAg secretion by inducing precore protein self-
assembly,44 as CAMs did not inhibit cccDNA in such cell lines after short-term
treatment.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CAMs inhibit reverse transcription by reducing the encapsulation of pgRNA into the
capsid. Thus, intracellular DNA levels in both cell lines were significantly reduced after
treatment with CAMs (Figure 4C). AT130, 56 and 59 decreased intracellular total RNA
in HepAD38 cells by 50%. GLS4 also induced a weak but significant decrease in the
total intracellular RNA in HepAD38 cells. We proposed that the reduction of
intracellular RNA may be due to the increased exposure to the cytoplasm as reduced
encapsulation into the capsid. However, more details about which RNA decreased are
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 86
1
2
3
4
5
6
7
8
9
needed to fully understand this phenomenon. The reverse transcriptase (RT) inhibitor
3TC slightly induced intracellular RNA accumulation due to the inhibition of reverse
transcription. In HepG2.2.15 cells, only 56 and 59 could significantly reduce
intracellular RNA, probably due to the different properties of each cell type.
3. Compounds 56 and 59 suppressed viral replication of both lamivudine-resistant
and multidrug-resistant HBV strains
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 8. Anti-HBV activity in wild type (HepAD38 and HepG2.2.15 cells),
lamivudine-resistant (HepG2.A64 cells) and multidrug-resistant (HepG2.1403F cells)
HBV strains.
HepG2.2.15
HepAD38
HepG2.A64
EC50 (μM)
HepG2.1403F
HepG2
CC50 (μM)
3TC
TDF
ETV
56
0.0038±0.0033
0.070±0.019
>10
>10
2.52±1.04
>100
15.04±3.88
42.07±2.92
>100
6.4E-05±7.62E-06 0.010±0.0017
0.84±0.36
2.2E-06±3.71E-07
0.0035±0.0012
0.042±0.010
0.011±0.011
0.034±0.012
0.071±0.035
0.31±0.092
>10
>10
0.00025±3.06E-05
0.0022±0.00087
0.048±0.028
0.0063±0.0017
0.00033±1.5E-04
0.013±0.0032
0.074±0.011
0.042±0.012
59
>100
AB-423
GLS4
0.28±0.074
39.85±5.11
49.2±1.30
0.0066±0.0020
0.013±0.0045
As CAMs have different action mechanisms with NAs, they should be effective
against HBV that is resistant to RT inhibitors. We then tested the effects of the
compounds on HBV DNA in a lamivudine/entecavir (3TC/ETV)-resistant cell line
(HepG2.A64) containing rtL180M+rtT184L+rtM204V mutations, and a multidrug-
resistant (MDR) cell line (HepG2.1403F) containing rtL180M+rtM204V+rtS202G+
rtN236T mutations.45 As shown in Table 8, compounds 56 and 59 inhibited HBV
replication in the HepG2.A64, HepG2.1403F, HepG2.2.15, and HepAD38 cell lines
with EC50 values well below 0.1 µM. Both compounds did not influence the viability
22
ACS Paragon Plus Environment

Page 23 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of HepG2 cells with CC50 values >100 µM. Although 3TC and ETV potently inhibited
HBV replication in HepG2.2.15 and HepAD38 cells, they did not show a significant
inhibitory effect even at 10 µM in HepG2.A64 and HepG2.1403F cells. The activity of
tenofovir (TDF) on the resistant strains was also more than 100-fold weaker than that
on HepG2.2.15 and HepAD38 cells. These results further proved the potential of CAMs
in CHB therapy.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4. Design of amino acid prodrugs to improve anti-HBV activity in vivo
The rigid and planar structure together with a high fraction of sp2 carbons led to the
poor solubility of compound 56 and its analogs (as shown in Table 10). Salt formation
is an efficient method to increase compound solubility.46 However, 56 and the related
analogs are weak bases, which are not ideal for salt formation. Thus, amino acid
prodrugs were designed, whose amino group was utilized for salt formation (Table 9).
Several amino acids, including N,N-dimethylglycine, L-valine, L-alanine and glycine,
were attached to the hydroxyl group of 56 through an ester bond. Different alpha
substituents were selected to modulate the stability of the prodrug. To evaluate the
stability of these prodrugs, the percent of remaining compound after incubation in
plasma and different buffers was determined by HPLC. Except for 62, all the other
prodrugs were rapidly hydrolyzed in plasma (Table 9). The steric hindrance of the i-Pr
in 62 may account for the resistance to hydrolysis. However, a significant portion of
compounds 63 and 64 was hydrolyzed in the aqueous buffers, indicating that they are
unstable in the digestive tract. Compound 61 was chosen for salt screening to increase
solubility.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 86
1
2
3
4
5
Table 9. Stability of the prodrugs in plasma and buffers.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Percent remaining
Half-life
Compound R’
R”
in plasma
pH 7.4 pH 6.8
pH 5.8
--
0.1 M HCl
61
62
63
64
H
i-Pr
CH3
H
CH3 47.2 min stablea
--b
--
stable
H
H
H
18 h
stable
--
--
--
--
31.2 min 55.6%c 80.2%
77.5 min 74.2% 67.6%
85.7%
88.3%
* Compounds were incubated in buffer at 37 °C; a No hydrolysis after 2 h at 37 °C; b
Not detected; c Percent remaining of the compound after 3 h.
Introduction of the amino acid moiety indeed improved the solubility in water (65,
Table 10), probably by increasing the fraction of sp3 carbons and the basicity. The TFA
salt further increased the solubility (61, Table 10). However, the solubility of 61 in pH
6.8 and 7.4 buffer were very low, which indicates poor solubility in the digestive tract.
To increase the solubility at various pH values, other salts with acetic acid, citric acid,
maleic acid, methanesulfonic acid and hydrochloric acid were prepared (66-70, Table
10). As excess salt in the detection samples would change the pH of the dissolvent, 1
mg of compound in 100 μL of buffer was used to determine the solubility (Table 10).
The more acidic acids used in the salt led to increased solubility (66 vs. 68 and 69).
Hydrophilic groups on the acid also increased the solubility (67 vs. 70). The solubility
24
ACS Paragon Plus Environment

Page 25 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
decreased with increasing pH (Table 10). The citric acid salt 66 was selected as the
acceptable solubility in various buffers and a low degree of hygroscopicity.
Table 10. Solubility of the prodrugs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Solubility (mg/mL)a
Compd.
Acid
water
ND
pH 5.8
ND
pH 6.8
ND
pH 7.4
ND
-
56
61
65
66
67
68
69
70
-
TFA
1.51
0.71
1.11
8.13
7.90
6.90
4.20
0.29
-
ND
-
-
-
AcOH
ND
ND
ND
0.031
0.23
0.016
ND
Citric Acid
(MeSO3H)2
HCl
7.60
7.08
6.75
7.93
6.65
7.69
0.040
ND
Maleic acid
a Solubility in water or 0.067 M phosphate buffer; ND Not detectable, below the
detection limit.
5. Anti-HBV activity of 56 and 67 in the hydrodynamic injection (HDI) HBV
mouse model
To evaluate the anti-HBV activity of 56 and 67, a hydrodynamic injection-based
mouse model of HBV infection was used. In this mouse model, the HBV genome was
transiently transduced into the hepatocytes of C57BL/6 mice under high pressure
conditions, and inhibition of HBV replication could be evaluated during the transient
time of about 7 days.47 Although this animal model cannot mimic the full cycle of HBV
infection, it was commonly used to evaluate the inhibition of HBV replication and liver
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 86
1
2
3
4
5
6
7
8
9
exposure of the compounds. Compounds 56 and 67 were chosen for in vivo anti-HBV
evaluation in HDI model as the higher potency and metabolic stability of 56 and the
increased solubility of 67. Oral administration once a day at doses of 50 mg/kg and 100
mg/kg were performed. Once daily oral dose of 100 mg/kg AB-423 and 0.03 mg/kg
ETV were used as positive controls. However, neither 56 nor 67 as well as AB-423
significantly inhibited HBV replication in mice (Supplementary Figure S1). In the
preclinical study, 100 mg/kg AB-423 was orally administrated twice daily for in vivo
anti-HBV evaluation. Thus, we adopt the administration frequency of twice a day, to
elucidate the therapeutic potential of this class of CAMs of AB-423. After 7 days
treatment, 100 mg/kg 67 achieved significant inhibition of both serum (Figure 5A) and
liver (Figure 5B) HBV DNA levels compared to vehicle control, whereas 50 mg/kg 67
only achieved weak inhibition of liver HBV DNA levels. Both doses of 56 did not lead
to a significant anti-HBV effect (Figure 5), which probably resulted from poor
absorption due to poor solubility.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Anti-HBV activity in an HDI HBV mouse model. (A) Serum HBV DNA
copies after hydrodynamic injection; (B) Liver HBV DNA on the 7th day after
hydrodynamic injection. N=5; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P <
0.0001.
The dose-dependent inhibition of in vivo HBV replication inspired us to further evaluate
26
ACS Paragon Plus Environment

Page 27 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the in vivo anti-HBV activity of 67. As the molecular weight of 67 is nearly twice that
of AB-423, we increased the dose to 200 mg/kg. However, no significant improvement
in the inhibition of HBV replication was observed (Supplementary Figure S2). This
may result from a limited absorption rate, induction of the expression of metabolic
enzymes etc. Blood and liver concentration of 56 after oral administration of 56 and its
pro-drug 67 were investigated. It was indicated that 56 was quickly eliminated in mice
with a clearance of 0.14 L/h and a short half-life time of 0.63 hour (Figure 6A, Table
11). Oral bioavailability of 56 was calculated as 12%. Blood concentration (1.19 μM)
of 56 reached maximum (Cmax) at 30 min after administration of 100 mg/kg parent
compound 56 (Table 11). At the dose of 100 mg/kg, oral bioavailability of 67 reached
87.6% (Table 11), much higher than that of parent compound 56 (12%). At the dose of
200 mg/kg, the oral bioavailability of 67 was increased to 132%, which might be
resulted from a phenomenon of saturated metabolic enzyme.48 Blood concentration of
56 reached Cmax of 3.73 μM and 8.84 μM at 1h after administrating 67 with the dose
of 100mg/kg and 200 mg/kg, respectively (Table 11). It was noted that very low level
of the pro-drug 67 in blood of mice was observed (data not shown). Considering the
fact that the half-life time of 67 in mice plasma was 47 min, we speculated that 67 could
be hydrolyzed by additional enzymes in small intestine and liver besides plasma.
Parent compound 56 distributed fast to liver after either intravascular or oral
administration of 56 or pro-drug 67 (Figure 6B). More than 10 nmol/g in liver was
detectable at 5 min after oral administration of 56 or 67. The liver Cmax of 56 reached
32 nmol/g at 0.5 h (Tmax) after oral administration of 56, corresponding with the Tmax
27
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 86
1
2
3
4
5
6
7
8
9
of 56 in blood. Fortunately, liver Cmax of 56 after oral administration of pro-drug 67
with the dose of 100 mg/kg and 200 mg/kg reached 40 nmol/g at 2 h and 59 nmol/g at
4 h, respectively, much higher than that after oral administration of parent compound
56 (Table 11). Correspondingly, liver exposure of 56 achieved by oral administrating
100 mg/kg 67 was two-fold higher than that achieved by oral administrating 100 mg/kg
56. Unexpectedly, 1.5-fold increase in liver exposure of 56 was observed at the dose of
200 mg/kg of 67 comparing with that at 100 mg/kg (Table 11). One possibility was that
high concentration of compound induced the metabolic enzyme overexpression.49
These results proved the advantage of amino acid pro-drug in improving the exposure
and in vivo anti-HBV efficacy of (1H-Pyrazolo[3,4-c]pyridin-5-yl)sulfonamide
derivative 56.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Pharmacokinetics of compound 56 and 67 in male C57BL/6 mice. (A) Plasma
concentration of 56 after administration of 56 and its prodrug 67; (B) Liver
concentration of 56 and its pro-drug. N=3.
28
ACS Paragon Plus Environment

Page 29 of 86
Journal of Medicinal Chemistry
1
2
3
4
5
Table 11. Pharmacokinetics of 56 and 67 in mice.
6
7
8
9
Liver
Cmax
t1/2
(h)
CL
AUCplasma
AUCLiver
Blood
Blood
Liver
F
Dose
(L/h) (μg*h/L) (nmol*h/g) Tmax (h)
Cmax (μM)
Tmax (h)
(%)
(nmol/g)
56 i.v.
20mg/kg
56 p.o.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.63 0.14
2597
1580
62
-
-
-
0.5
2
-
-
-
-
-
-
-
-
142
306
482
0.5
1.0
1.0
1.19
3.73
8.84
32
40
59
12
100mg/kg
67 p.o.
7013.2
21106
87.6
132
100mg/kg
67 p.o.
4
200mg/kg
6. Molecule modeling
Figure 7. Predicted binding mode of 56. (A) Docking of 56 with HBV core protein
crystal structure (PDB code: 5WRE); (B) Merging of 56 and NVR 3-778 at the binding
site; (C) Merging of 56 and HAP-R01 at the binding site.
To gain insight of the interaction between 56 and HBV core protein, molecule
modeling was performed. We carried out docking based on the co-crystal structure of
HAP-R01 and core protein (PDB code: 5WRE). The results showed that compound 56
binds to the core protein dimer-dimer interface with a similar interaction mode as that
of SBAs and HAPs (Figure 7). The modeling results show the formation of a hydrogen
bond between 2-N and W102. The aniline moiety pointed to a hydrophobic pocket
formed by the two core protein dimers, which is occupied by a halogen substituted
29
ACS Paragon Plus Environment