A novel, potent, and orally bioavailable thiazole HCV NS5A inhibitor for the treatment of hepatitis C virus

Article abstract of DOI:10.1016/j.ejmech.2019.02.016

A medicinal chemistry program based on the small-molecule HCV NS5A inhibitor daclatasvir has led to the discovery of dimeric phenylthiazole compound 8, a novel and potent HCV NS5A inhibitor. The subsequent SAR studies and optimization revealed that the cycloalkyl amide derivatives 27a-29a exhibited superior potency against GT1b with GT1b EC₅₀ values at picomolar concentration. Interestingly, high diastereospecificity for HCV inhibition was observed in this class with the (1R,2S,1′R,2′S) diastereomer displaying the highest GT1b inhibitory activity. The best inhibitor 27a was found to be 3-fold more potent (GT1b EC₅₀ = 0.003 nM) than daclatasvir (GT1b EC₅₀ = 0.009 nM) against GT1b, and no detectable in vitro cytotoxicity was observed (CC₅₀ > 50 μM). Pharmacokinetic studies demonstrated that compound 27a had an excellent pharmacokinetic profiles with a superior oral exposure and desired bioavailability after oral administration in both rats and dogs, and therefore it was selected as a developmental candidate for the treatment of HCV infection.

Full text of DOI:10.1016/j.ejmech.2019.02.016

Accepted Manuscript
A novel, potent, and orally bioavailable thiazole HCV NS5A inhibitor for the treatment
of hepatitis C virus
Teng-Kuang Yeh, Iou-Jiun Kang, Tsu-An Hsu, Yen-Chun Lee, Chung-Chi Lee,
Sheng-Ju Hsu, Ya-Wen Tian, Hui-Yun Yang, Chiung-Tong Chen, Yu-Sheng Chao,
Andrew Yueh, Jyh-Haur Chern
PII:
S0223-5234(19)30126-6
DOI:
https://doi.org/10.1016/j.ejmech.2019.02.016
EJMECH 11108
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 7 November 2018
Revised Date: 2 February 2019
Accepted Date: 4 February 2019
Please cite this article as: T.-K. Yeh, I.-J. Kang, T.-A. Hsu, Y.-C. Lee, C.-C. Lee, S.-J. Hsu, Y.-W. Tian,
H.-Y. Yang, C.-T. Chen, Y.-S. Chao, A. Yueh, J.-H. Chern, A novel, potent, and orally bioavailable
thiazole HCV NS5A inhibitor for the treatment of hepatitis C virus, European Journal of Medicinal
Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.02.016.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
A Novel, Potent, and Orally Bioavailable Thiazole HCV NS5A Inhibitor
for the Treatment of Hepatitis C Virus
#
#
Teng-Kuang Yeh, Iou-Jiun Kang, Tsu-An Hsu, Yen-Chun Lee, Chung-Chi Lee,
Sheng-Ju Hsu, Ya-Wen Tian, Hui-Yun Yang, Chiung-Tong Chen, Yu-Sheng Chao,
*
*
Andrew Yueh, Jyh-Haur Chern
Institute of Biotechnology and Pharmaceutical Research, National Health Research
Institutes, Miaoli County 350, Taiwan, ROC
*
To whom correspondence should be addressed. Dr. Jyh-Haur Chern, E-mail:
jhchen@nhri.org.tw; Tel.: 886-37-246-166 ext. 35716; Fax: 886-37-586-456.
Corresponding address: Institute of Biotechnology and Pharmaceutical Research,
National Health Research Institutes, No. 35, Keyan Road, Zhunan Town, Miaoli
County 350, Taiwan, ROC
#
Authors with equal contribution
1

ACCEPTED MANUSCRIPT
ABSTRACT
A medicinal chemistry program based on the small-molecule HCV NS5A inhibitor
daclatasvir has led to the discovery of dimeric phenylthiazole compound 8, a novel
and potent HCV NS5A inhibitor. The subsequent SAR studies and optimization
revealed that the cycloalkyl amide derivatives 27a-29a exhibited superior potency
against GT1b with GT1b EC values at picomolar concentration. Interestingly, high
5
0
diastereospecificity for HCV inhibition was observed in this class with the
1R,2S,1’R,2’S) diastereomer displaying the highest GT1b inhibitory activity. The
best inhibitor 27a was found to be 3-fold more potent (GT1b EC = 0.003 M) than
M) against GT1b, and no detectable in vitro
cytotoxicity was observed (CC > 50 µM). Pharmacokinetic studies demonstrated
(
n
5
0
daclatasvir (GT1b EC50 = 0.009
n
50
that compound 27a had an excellent pharmacokinetic profiles with a superior oral
exposure and desired bioavailability after oral administration in both rats and dogs,
and therefore it was selected as a developmental candidate for the treatment of HCV
infection.
Keywords: HCV; discovery; SAR; NS5A inhibitor; phenylthiazole; inhibitory
activity
2

ACCEPTED MANUSCRIPT
1
. Introduction
Hepatitis C virus (HCV), first discovered in 1989, is a positive sense single stranded
1
RNA virus that causes inflammation of the liver. At least six major genotypes,
including multiple subtypes within each genotype, of HCV have been described and
2
genotype 1 is the most common infection worldwide. HCV, chronically infected
more than 170 million people worldwide, can lead to liver fibrosis, liver cancer and
3
liver failure. Traditional dual therapy (pegylated interferon plus ribavirin) produces
relatively low rates of sustained virological response (SVR) for Genotype 1,
4
historically the most difficult-to-treat genotype. Telaprevir and boceprevir (Figure
1
), approved by US Food and Drug Administration (FDA) in 2011, were the first oral
NS3 protease inhibitors used with pegylated interferon and ribavirin to treat genotype
5
1
chronic hepatitis C. Treatment with a telaprevir- or boceprevir-based triple
therapy has resulted in an improved SVR rate in genotype 1 patients, when compared
5
c-e, 6
with pegylated interferon plus ribavirin therapy.
These triple therapies,
however, have significant limitations due to side effects and unfavorable tolerability,
5
c-e, 6a-e, 7
especially in patients with difficult-to-treat liver cirrhosis.
The selling of
telaprevir and boceprevir was stopped on 2014 and 2015, respectively, because highly
efficacious and better tolerated direct-acting antiviral drugs (DAAs) were approved by
3

ACCEPTED MANUSCRIPT
8
FDA since 2013.
Insert Figure 1
A great advance has been recently made in treating chronic hepatitis C (CHC) due to
newly approved DAAs (Figure 2), such as the NS3/4A protease inhibitors
9
10
11
12,13
simeprevir, paritaprevir, grazoprevir, and glecaprevir,
the NS5A inhibitors
1
4
15
16
11c-g,
17
18
ledipasvir,
ombitasvir,
daclatasvir,
elbasvir
velpatasvir,
and
1
3,19,20
18c-f, 21
22
pibrentasvir,
the NS5B polymerase inhibitors sofosbuvir
and dasabuvir .
A number of DAAs, with or without ribavirin, have been approved and those all-oral,
2
3
interferon-free regimens , which have higher antiviral efficacy and fewer
dose-related adverse effects as well as lower risk of resistant strains have been
2
4
marketed.
For the currently approved combination products, the NS5A inhibitors
are the most commonly used DAAs combined with an NS5B polymerase inhibitor
and/or an NS3/4A protease inhibitor for effectively treating treatment-naive and
treatment-experienced patients who have HCV genotype 1 to 6. Within a treatment
period of 12 weeks, new combination drugs have a cure rate of over 90% and no
significant effects, with most recent treatments only requiring 8 weeks of therapy in
multiple genotypes for treatment-naïve non-cirrhotic populations. Unfortunately, the
4

ACCEPTED MANUSCRIPT
high prices of the medications have caused a heavy barrier to most people to access
that care even in high-income countries. Therefore many research groups have been
working toward the development of a new effective, safe, and cheap treatment for
HCV infection. The primary object of our research effort was to identify chemically
novel, potent, and orally bioavailable HCV NS5A inhibitor with lower cost and
shorter duration of therapy.
Insert Figure 2
As mentioned above, several NS5A inhibitors have been successfully developed and
approved for the treatment of HCV infection, such as ledipasvir, ombitasvir,
daclatasvir, elbsvir, velpatasvir, and pibrentasvir (Figure 2).
In this paper,
daclatasvir was selected as a basis for further structural modification in our new drug
discovery program since it has been reported to exert strong potency against HCV in a
cell-based HCV replicon assay. It is important to note that the scaffold of daclatasvir
consisted of two symmetrical imidazole rings, which play a crucial role for the strong
potency of HCV.
In a preliminary study, we observed that the replacement of the
two imidazole rings of daclatasvir with the hydrophobic thiazole rings has led to the
discovery of the dimeric phenylthiazole compound 8 as our initial lead of new HCV
5

ACCEPTED MANUSCRIPT
NS5A inhibitor (Figure 3). Surprisingly, compound 8 was found to possess
moderate GT1b inhibitory activity with a GT1b EC of 56.3 nM, although it was
50
much less potent than daclatasvir (GT1b EC50 = 0.009 nM) against GT1b. This
interesting result prompted us to further investigate this new class of HCV NS5A
inhibitors.
Insert Figure 3
In an attempt to improve the GT1b potency of the lead compound 8, a series of
structurally related dimeric phenylthiazole derivatives 9-57 have been designed,
synthesized, and evaluated for their inhibitory activity against GT1b in our anti-HCV
screening program. The detailed SAR and pharmacokinetic studies were described.
Among these compounds, the best inhibitor 27a (Figure 4) was identified as a highly
potent (GT1b EC = 0.003 nM) and orally bioavailable HCV NS5A inhibitor. It
5
0
appeared to be the most promising developmental candidate for the treatment of HCV
infection. In this paper, we described the discovery and optimization of a novel class
of dimeric phenylthiazole derivatives as potent HCV NS5A inhibitors via a medicinal
chemistry program based on the scaffold of compound 8.
6

ACCEPTED MANUSCRIPT
Insert Figure 4
2
. Results and discussion
2
.1. Chemistry
The synthetic procedures for this class of dimeric phenylthiazole derivatives 8-57 are
presented in Schemes 1-4. The key intermediate 6 was prepared according to the
synthetic method shown in Scheme 1 beginning from the commercially available
2
-amino-4’-bromoacetophenone hydrochloride 1.
Neutralization of the
hydrochloride salt 1 with DIPEA in CH Cl at room temperature followed by
2
2
coupling reaction of its corresponding free base with N-Boc-L-proline in the presence
of HOBt·H O/EDC led to the formation of amide 2 in 90% yield. Reaction of 2 with
2
2
5
Lawesson’s reagent gave the phenylthiazole 3 in 78% yield, which was reacted with
bis(pinacolato)diboron in the presence of Pd(PPh ) and potassium acetate to give the
3
4
2
6
corresponding organoboron compound 4 in 99% yield.
The dimeric phenylthiazole
compound 5 was then prepared by reacting the organoboron compound 4 with 3 under
2
7
PdCl (dppf) catalysis using Suzuki coupling reaction in 47% yield.
Acidic
2
deprotection of 5 with trifluoroacetic acid in CH Cl at room temperature gave the key
2
2
intermediate 6 in 99% yield.
7

ACCEPTED MANUSCRIPT
Insert Scheme 1
Exploration of the substituents on the pyrrolidine nitrogen in this class of dimeric
phenylthiazole compounds has been done (Schemes 2-4). The synthetic procedures
for a series of alkyl carbamate derivatives (1S,2S,1’S,2’S) 8-17 are presented in
Scheme 2 (synthetic method A). The L form amino acid derivatives (III) that were
not commercially available were prepared by reacting the corresponding amino acid (I)
with the substituted chloroformate (II) at room temperature in the presence of sodium
28
carbonate and 1.0 M aqueous sodium hydroxide solution.
Subsequent peptide
coupling reaction of the pyrrolidine nitrogen of intermediate 6 with compound (III) in
the presence of HOBt·H O/EDC in DMF at room temperature gave the desired
2
coupling product (1S,2S,1’S,2’S) 8-17 as a single diastereomer in 54-86% yields
following flash column purification. This result indicated that no racemization
occurred during the coupling reaction.
Insert Scheme 2
The synthetic procedures for the carbamate derivatives 18a,b, sulfonamide derivatives
8

ACCEPTED MANUSCRIPT
1
9a,b, and a variety of substituted amide derivatives 20a,b - 49a,b are presented in
Scheme 3 (synthetic method B). The N-Boc protected phenylglycine derivative 7
was prepared by coupling the intermediate 6 with N-Boc-D-phenylglycine at room
temperature in the presence of HOBt·H O/EDC in 75% yield. Acidic deprotection
2
of 7 with trifluoroacetic acid at room temperature gave the amine intermediate, which
underwent substitution reaction with methyl chloroformate in the presence of
triethylamine to provide the methyl carbamate derivative (1R,2S,1’R,2’S) 18a and its
corresponding diastereomer (1R,2S,1’S,2’S) 18b. In addition, reaction of the amine
intermediate with methanesulfonyl chloride gave the methyl sulfonamide derivative
(
1R,2S,1’R,2’S) 19a and its corresponding diastereomer (1R,2S,1’S,2’S) 19b.
Similarly the substituted amide derivatives (1R,2S,1’R,2’S) 20a-49a and
1R,2S,1’S,2’S) 20b-49b were also prepared by deprotection of 7 with trifluoroacetic
(
acid followed by reactions with a variety of acyl chloride in the presence of
trimethylamine in 10-55% yield. These results may be due to the rapid racemization
at the C position of the (R)-phenylglycine moiety during the reaction process.
1
Insert Scheme 3
The cyclopropyl amide derivatives (1R,2S,1’R,2’S) 50-57 were prepared as shown in
9

ACCEPTED MANUSCRIPT
Scheme 4 (synthetic method A). The D form amino acids (IV) were acylated with
cyclopropyl carbonyl chloride at room temperature in the presence of sodium
carbonate and 1.0 M aqueous sodium hydroxide solution to give the corresponding
2
8
amide intermediate (V) in excellent yields.
Peptide coupling reaction of 6 with (V)
in the presence of HOBt·H O/EDC in DMF at room temperature gave the desired
2
cyclopropyl amide derivatives (1R,2S,1’R,2’S) 50-57 as a single diastereomer in
5
1-69% yields.
Insert Scheme 4
2
.2. Biological evaluation
The dimeric phenylthiazole compounds 8-57 described herein were evaluated for
GT1b inhibitory activity using an in vitro assay system that is suitable for monitoring
anti-HCV activities of compounds. This system is composed of a human
29
hepatocarcinoma cell line (Huh-7) supporting multiplication of a GT1b replicon.
While there are several genotypes in HCV, our assay employed genotype 1b replicon.
We examined compounds’ ability to inhibit the replication of subgenomic HCV RNA
in a GT1b replicon cell system. In addition, all target compounds were also tested
for their cytotoxicity toward human Huh-7 cells. Details of this SAR investigation
1
0

ACCEPTED MANUSCRIPT
has been done on GT1b and will be described herein.
Daclatasvir is a potent HCV NS5A inhibitor, which showed excellent GT1b inhibitory
activity (GT1b EC = 0.009 nM) in a HCV genotype 1b subgenomic replicon assay.
5
0
In our efforts to identify a new generation inhibitor possessing greater GT1b potency
relative to daclatasvir, we first examined the role of the two imidazole rings in
daclatasvir (Table 1). Surprisingly, the replacement of the two imidazole rings in
daclatasvir with the thiazole rings led to the discovery of a dimeric phenylthiazole
compound 8 as our initial lead of new HCV NS5A inhibitor, which showed moderate
GT1b inhibitory activity (GT1b EC = 56.3 nM) in a HCV genotype 1b subgenomic
5
0
replicon assay. Although compound 8 was much less potent than daclatasvir against
GT1b, it was found to possess low cytotoxicity (CC > 50 µM) and high therapeutic
50
index (CC /GT1b EC ), and worthy of further optimization. Having shown that the
50
50
imidazole rings can be replaced by the thiazole rings, a series of structurally related
dimeric phenylthiazole derivatives have been synthesized and evaluated for their
GT1b inhibitory activities (Table 1). Modifications of the two methyl groups on the
carbamate moiety (R ) and the isopropyl groups at the C-1 and C-1’ position (R ) of
1
2
the compound 8 led to the following general conclusions. The replacement of the
two methyl groups on the carbamates of 8 with either the ethyl (9) or phenyl group
11

ACCEPTED MANUSCRIPT
(
10) significantly reduced GT1b inhibitory activity (GT1b EC = 327.1 and 2815.6
5
0
nM, respectively). This effect might be due to their strict steric requirement at this
position in the binding site. Since it appeared that retention of the methyl carbamate
group was nevertheless beneficial, we turned our attention to the exploration of the
SAR around the alkyl substituents (R ) at the C-1 and C-1’ position of the compound.
2
Replacing the two isopropyl groups at the C-1 and C-1’ position of 8 with the methyl
groups (11) moderately reduced activity against GT1b (GT1b EC50 = 158.5 nM);
however, the ethyl derivative 12 had GT1b inhibitory activity (GT1b EC = 56.6 nM)
5
0
comparable to that of 8. In an effort to further improve the GT1b inhibitory activity,
we extended the alkyl substituents at the C-1 and C-1’ position of 8. Replacement of
the two isopropyl groups with either a n-propyl (13), n-butyl (14), i-butyl (15), or
t-butyl (16) resulted in a general reduction in activity (GT1b EC = 123-471 nM)
5
0
against GT1b. Importantly, the phenyl derivative 17, when compared to its
corresponding isopropyl derivative 8, provided a significant improvement in GT1b
inhibitory activity (GT1b EC = 40.9 nM), and demonstrated the best GT1b potency
5
0
in this series of dimeric phenylthiazole compounds.
Insert Table 1
1
2

ACCEPTED MANUSCRIPT
We next investigated the stereochemical requirements for GT1b inhibitory activity in
the methyl carbamate series (compounds 17, 18a, and 18b). Results are shown in
Table 2.
Interestingly, the (1R,2S,1’R,2’S) diastereomer 18a (GT1b EC = 0.19
50
nM) was found to be significantly more potent than the corresponding diastereomers
(
1S,2S,1’S,2’S) 17 (GT1b EC = 40.9 nM) and (1R,2S,1’S,2’S) 18b (GT1b EC =
50 50
1
07.9 nM) against GT1b, demonstrating the critical stereochemical requirement at
C-1 and C-1’ position for a R absolute configuration to display strong GT1b
inhibitory activity in this series. It seems that the stereochemistry (1R,2S,1’R,2’S) of
this class of dimeric phenylthiazole compounds optimally fits to a binding pocket of
NS5A. With these results, the methyl carbamate derivative (1R,2S,1’R,2’S) 18a was
chosen for further SAR study based on its strong GT1b potency.
Insert Table 2
In an attempt to improve the GT1b potency of the lead compound 18a, we explored
the possibility of replacing the carbamate group at the C-1 and C-1’ position of 18a
with other functional groups. The results are shown in Table 3 and are compared to
1
8a. Replacement of the carbamate group with the sulfonamide, compound 19a,
resulted in a significant reduction in GT1b inhibitory activity (GT1b EC = 3.4 nM).
5
0
1
3

ACCEPTED MANUSCRIPT
Encouragingly, the methyl amide derivative 20a (GT1b EC = 0.02 nM) was found
5
0
to be 10-fold more potent than the corresponding methyl carbamate derivative 18a
GT1b EC50 = 0.19 nM) against GT1b. We also investigated the stereochemical
comparison for GT1b inhibitory activity (Table 3). As expected, the
(
stereochemically unfavorable (1R,2S,1’S,2’S) diastereomers 18b-20b were found to
be significantly less potent than their corresponding (1R,2S,1’R,2’S) diastereomers
1
8a-20a against GT1b, respectively.
Insert Table 3
Having achieved excellent GT1b potency with amide derivatives, we investigated the
SAR around the alkyl substituents on the amide moiety of this class of dimeric
phenylthiazole compounds (Table 4). A variety of substituted amide derivatives
(1R,2S,1’R,2’S) 21a-49a and their corresponding (1R,2S,1’S,2’S) diastereomers
2
1b-49b were prepared in an attempt to increase potency against GT1b. The results
are shown in Table 4 and are compared to the methyl amide derivatives 20a and 20b.
In (1R,2S,1’R,2’S) isomer series, increasing the chain length of the alkyl substituent
on the amide from methyl (20a) to ethyl (21a) and n-propyl (22a) generally retained
the GT1b inhibitory activity. The n-butyl amide derivative 23a showed a slight
1
4

ACCEPTED MANUSCRIPT
reduction in potency against GT1b. Further extension of the side chain to n-pentyl
(24a) resulted in a further reduction in GT1b inhibitory activity. More importantly,
the isopropyl amide derivative 25a exhibited superior GT1b potency (GT1b EC =
50
0
.009 nM), which was comparable to that of daclatasvir. Surprisingly, the t-butyl
amide derivative 26a was considerably less active than 20a against GT1b.
We further investigated varying the substituents on the amide by changing from alkyl
to cycloalkyl group in this series in an effort to increase GT1b potency (Table 4).
Interestingly, the cycloalkyl amide derivatives 27a-29a were found to be significantly
more potent (GT1b EC < 0.01 nM) than the corresponding alkyl amide derivatives
5
0
2
2a-24a against GT1b. This is possibly due to their drastically conformational
change and steric requirement at this position in the binding site. It is worth noting
that the cyclopropyl amide derivative 27a was found to be 3-fold more potent (GT1b
EC50 = 0.003 nM) than daclatasvir (GT1b EC50 = 0.009 nM) in GT1b inhibitory
activity, and no apparent cellular cytotoxicity was observed (CC50 > 50 µM).
Therefore compound 27a was chosen for further pharmacokinetic study based on its
GT1b potency. However, the bulky cyclohexyl amide derivative 30a displayed a
significant reduction in GT1b inhibitory activity (GT1b EC = 0.25 nM).
5
0
1
5

ACCEPTED MANUSCRIPT
Although compound 30a showed weak GT1b inhibitory activity, we also investigated
replacing the cyclohexyl rings of 30a with aryl or heteroaryl rings on the amide
(Table 4). Intriguingly, aromatization of the cyclohexyl ring of 30a (31a) induced a
drastic decrease in GT1b inhibitory activity; however, the corresponding benzyl
amide derivative 32a had GT1b inhibitory activity comparable to 30a. More
interestingly, when the phenyl ring on the amide of 31a was replaced with a
4
-pyridinyl (33a), a dramatic improvement in GT1b inhibitory activity was observed.
Moving the nitrogen atom to the 3-position (3-pyridinyl amide derivative 34a) caused
a loss in GT1b inhibitory activity, indicating that this nitrogen atom is not tolerated at
this position, and the 2-pyridinyl amide derivative 35a exhibited a further reduction in
GT1b inhibitory activity. In addition, the 2- and 3-furanyl amide derivatives, 36a
and 37a, exhibited excellent GT1b inhibitory activity (GT1b EC = 0.07 and 0.08
5
0
nM, respectively), which were almost comparable to that of 4-pyridinyl amide
derivative 33a (GT1b EC = 0.011 nM). Surprisingly, the 2- and 3-thienyl amide
50
derivatives, 38a and 39a, were considerately less potent than the corresponding 2- and
-furanyl amide derivatives, 36a and 37a against GT1b.
3
We next tried to seek a more attractive structure for optimization instead of the aryl
and heteroaryl amides, thus several different types of heterocyclic and
1
6

ACCEPTED MANUSCRIPT
amino-substituted alkyl amide derivatives 40a-49a were also synthesized and
examined. Interestingly, the morpholyl, pyrrolidyl, piperidyl, and pyrrolyl amide
derivatives 40a-43a were found to possess excellent GT1b inhibitory activity with
GT1b EC values ranging from 0.016 - 0.035 nM. However, the amino-substituted
5
0
alkyl amide derivatives 44a-49a were about 100-fold less potent than 40a-43a.
Further investigation will be required to explore this unexpected result.
On the other hand, we also investigated the stereochemical comparison for GT1b
inhibitory activity in the amide series (Table 4).
Interestingly, high
diastereospecificity for HCV inhibition was observed with the (1R,2S,1’R,2’S)
diastereomer displaying the highest GT1b inhibitory activity. However, the
stereochemically unfavorable (1R,2S,1’S,2’S) diastereomers 21b-49b were found to
be significantly less potent than their corresponding (1R,2S,1’R,2’S) diastereomers
2
1a-49a against GT1b in the amide series.
Insert Table 4
Modifications of the alkyl substituents (R) at both the C-1 and C-1’ position of the
dimeric phenylthiazole derivatives are shown in Table 5 and are compared to the best
1
7

ACCEPTED MANUSCRIPT
inhibitor 27a. Replacement of the phenyl groups at both the C-1 and C-1’ position
of 27a with either a methyl (50), ethyl (51), n-propyl (52), n-butyl (53), i-propyl (54),
i-butyl (55), t-butyl (56), cyclohexyl (57) significantly reduced GT1b inhibitory
activity. This is probably due to the fact that the strongest hydrophobic interaction
could be provided by the phenyl rings at both the C-1 and C-1’ position in the binding
site. With these results, we observed that the isobutyl analogue of the amide (55)
bestowed the best GT1b potency (GT1b EC = 27.1 nM) in this series. However,
5
0
the bulky cyclohexyl analogue of the amide (57) was found to be devoid of GT1b
inhibitory activity (GT1b EC > 50 µM). The underlying cause of this effect is still
5
0
unknown and deserves further research.
Insert Table 5
To evaluate the cytotoxicity of this class of dimeric phenylthiazole derivatives, all
target compounds 8-57 were tested for their cytotoxic effect toward human Huh-7 cell
lines. Very encouragingly, compounds 8-57 were found to display no detectable
cytotoxicity at the highest concentration of 50 µM (CC > 50 µM).
50
2
.3. Pharmacokinetic studies
1
8

ACCEPTED MANUSCRIPT
Following extensive SAR studies, the cyclopropyl amide derivative 27a was then
selected for animal pharmacokinetic evaluation since it appeared to demonstrate the
most potent inhibitory activity (GT1b EC = 0.003 nM) against GT1b in this novel
5
0
class of compounds. The pharmacokinetic parameters of 27a were determined in
rats and dogs and the data are summarized in Table 6 and 7. There were no deaths
and no treatment-related effects on body weight or clinical observations. After
intravenous (1.0 mg/kg) and oral (5.0 mg/kg) administration in rats, the compound
was characterized by a low plasma clearance (CL) of 1.1 mL/min/kg, a moderate
volume of distribution (Vss) of 1.3 L/kg, a long oral half-life (t ) of 20.4 h, and an
1
/2
acceptable bioavailability of 22% which resulted in a good oral exposure (AUC) of
4539 ng/mL h. In addition, compound 27a was dosed both intravenously (0.1
1
×
mg/kg) and orally (0.5 mg/kg) in dogs, similarly it had a low plasma clearance of 1.3
mL/min/kg, a moderate volume of distribution of 1.6 L/kg, a long oral half-life (t )
1
/2
of 21.4 h, and a reasonable bioavailability of 23.3%, leading to a good oral exposure
of 1622 ng/mL h. These significant results demonstrated that compound 27a had an
×
excellent GT1b potency and favorable pharmacokinetic properties with acceptable
bioavailability after oral dosing in both rats and dogs. Therefore, it appeared to be
the most promising preclinical candidate for further development as anti-HCV agent
1
9

ACCEPTED MANUSCRIPT
Insert Table 6
Insert Table 7
Furthermore, drug resistance studies have been carried out to elucidate how
3
0
compound 27a works in the HCV RNA replication.
The resistance profile showed
that the N terminus of NS5A is the region responsible for the 27a-mediated inhibition
3
0
of HCV replicon activity.
It may be possible that 27a exerts GT1b inhibitory
activity by directly binding to the N terminus of NS5A due to similar resistance
profile to daclatasvir, a NS5A inhibitor, resulting in the drug resistance mutations
3
1
within the N terminus of NS5A.
Of course, there is still a need to clarify the
binding site and molecular mechanism of action of this class of NS5A inhibitors.
Importantly, these findings from this study provide very useful information to develop
novel anti-HCV agents.
3
. Conclusions
In summary, a novel class of dimeric phenylthiazole derivatives was designed, synthesized,
and identified as potent HCV NS5A inhibitors based on daclatasvir scaffold. We have
developed the SAR for the substituent on the (S)-pyrrolidine and have shown that the
2
0

ACCEPTED MANUSCRIPT
(R)-phenylglycine moiety is sufficient for good potency against GT1b, and the potency
increases in the order SO < CO < CO for the linker from the methyl substituent to the
2
2
(R)-phenylglycine. In these studies, it was demonstrated that the stereochemistry at the
C-1 and C-1’ position of the phenylglycine moieties plays a very important role in affecting
their GT1b inhibitory activities of this class of dimeric phenylthiazole compounds. The
stereochemically preferred (1R,2S,1’R,2’S) diastereomers were found to be significantly
more potent than the corresponding (1R,2S,1’S,2’S) and (1S,2S,1’S,2’S) diastereomers
against GT1b. Furthermore, modifications of the substituent at the amide group on the
(R)-phenylglycine moiety resulted in an interesting pattern of GT1b inhibitory activity.
Several compounds described in this work exhibited superior potency against GT1b with
GT1b EC values at picomolar concentration. The cyclopropyl amide derivative 27a was
50
found to be the best inhibitor of GT1b replicon (GT1b EC = 0.003 nM), which showed
5
0
promising pharmacokinetic properties and bioavailability after oral administration in both
rats and dogs. Previous drug resistance studies demonstrated that compound 27a is likely
to inhibit HCV replication by directly binding to HCV NS5A. On the basis of these
encouraging results, compound 27a was selected as a preclinical candidate for the
treatment of HCV infection. Further SAR and mechanistic studies on this class of
compounds are currently under active investigation and will be reported in due course.
2
1

ACCEPTED MANUSCRIPT
4
. Experimental
4
.1. General procedures
All commercial chemicals and solvents are reagent grade and were used without
1
further treatment unless otherwise noted. H NMR spectra were obtained with a
Varian Mercury-300 or a Varian Mercury-400 spectrometer. Chemical shifts were
recorded in parts per million (ppm, δ) and were reported relative to the solvent peak
or TMS. Coupling constants (J) are reported in hertz (Hz). Splitting patterns are
described by using the following abbreviations: s, singlet; d, doublet; t, triplet; q,
quartet; br, broad; m, multiplet. LC/MS data were measured on an Agilent
MSD-1100 ESI-MS/MS System. All tested compounds were detected at UV 254nm
unless otherwise stated. Column chromatography was performed with silica gel
(Merck Kieselgel 60, 230-400 mesh). Reactions were monitored by TLC using
Merck 60 F254 silica gel glass backed plates and visualized under ultraviolet
irradiation (254 nm and 360 nm) or by spraying with phosphomolybdic acid reagent
o
(
Aldrich) followed by heating at 80 C. Melting points were determined on an
Electrothermal IA9000 Series Digital Melting Point Apparatus. Purity of the final
compounds was determined on a Hitachi 2000 series HPLC system with a reverse
phase C column (Agilent ZORBAX Eclipse XDB-C18 5 µm, 4.6 mm x 150 mm),
1
8
o
operating at 25 C. Mobile phase A was acetonitrile. Mobile phase B was 10 mM
2
2

ACCEPTED MANUSCRIPT
NH OAc aqueous solution containing 0.1% formic acid. The gradient system started
4
from A/B (10%/90%) at 0 min to A/B (90%/10%) at 45 min. The flow rate of the
mobile phase was 0.5 mL/min, and the injection volume of the sample was 5 µL.
Peaks were detected at 254 nm. The purity of all tested compounds is >95% purity.
4
.1.1. N-[2-(4-Bromophenyl)-2-oxoethyl]-1-(tert-butoxycarbonyl)-L-prolinamide (2)
A solution of N-Boc-L-proline (5.16 g, 24 mmol) and hydroxybenzotriazole
monohydrate (HOBt·H O, 3.67 g, 24 mmol) in 200 mL CH Cl was stirred at room
2
2
2
temperature for 10 minutes, then treated with N-ethyl-N’-(3-dimethylaminopropyl)
carbodiimide hydrochloride (EDC·HCl, 4.6 g, 24 mmol). The resulting mixture was
stirred at room temperature for 30 minutes and then treated with a yellow solution
formed by stirring 2-amino-4-bromoacetophenone hydrochloride (5.0 g, 20 mmol)
and N,N’-diisopropylethylamine (DIPEA, 2.58 g, 20 mmol) in 150 mL CH Cl at
2
2
room temperature for 10 minutes. The resulting mixture was stirred at room
temperature overnight and then filtered through Celite to remove the precipitate.
The filtrate was partitioned between CH Cl and water. The organic layer was
2
2
washed with brine, dried over MgSO , filtered, and concentrated via vacuo. The
4
residue was purified by column chromatography (ethyl acetate : hexane = 2 : 5) to
o
1
yield pure product 2 as a yellow solid (7.39 g, 90% yield): mp 130-131 C; H NMR
2
3

ACCEPTED MANUSCRIPT
(
300MHz, CDCl ) (1 : 2 mixture of rotomers) δ 1.45 (s, 9H), 1.83-2.00 (m, 2H), 2.14
3
(br s, 4/3H), 2.22 (br s, 2/3H), 3.38 (br s, 2/3H), 3.48 (br s, 4/3H), 4.28 (br s, 1/3H),
4
.35 (br s, 2/3H), 4.69 (d, J = 4.5 Hz, 2H), 7.02 (br s, 2/3H), 7.51 (br s, 1/3H), 7.60 (d,
+
J = 8.3 Hz, 2H), 7.80 (d, J = 8.3 Hz, 2H); LC/MS (ESI) m/z: 411.3 [M + H] , 433.0
+
[
M + Na] .
4
.1.2.
tert-Butyl
(2S)-2-[5-(4-bromophenyl)-1,3-thiazol-2-yl]pyrrolidine-1-carboxylate (3)
To a solution of the kitoamide substrate 2 (25.26 g, 61.42 mmol) in THF (300 mL),
Lawesson’s reagent (37.21 g, 92.11 mmol) was added. The resulting mixture was
refluxed for 6 hours, cooled to room temperature, and then concentrated in vacuo.
The residue was purified by column chromatography (ethyl acetate : hexane = 1 : 2)
o
1
to get the product 3 as a yellow solid (19.6 g, 78% yield): mp 84-85 C; H NMR
(
400MHz, CDCl ) (4 : 5 mixture of rotomers) δ 1.33 (s, 5H), 1.47 (s, 4H), 1.94 (br s,
3
2
4
H), 2.22 (br s, 8/9H), 2.27 (br s, 10/9H), 3.43-3.60 (m, 2H), 5.08 (s, 5/9H), 5.19 (s,
/9H), 7.36 (d, J = 6.4 Hz, 2H), 7.47 (d, J = 6.4 Hz, 2H), 7.80 (s, 1H); LC/MS (ESI)
+
+
m/z: 409.0 [M + H] , 431.0 [M + Na] .
4
.1.3.
tert-Butyl
2
4

ACCEPTED MANUSCRIPT
(2S)-2-{5-[4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl]-1,3-thiazol-2-yl}p
yrrolidine-1-carboxylate (4)
A flask charged with Pd(PPh ) (0.49 g, 0.43 mmol), potassium acetate (2.09 g, 21.37
3
4
mmol), bis(pinacolato)diboron (5.16 g, 17.1 mmol), compound 3 (3.50 g, 8.55 mmol)
and 1,4-dioxane (100 mL) was flushed with nitrogen. The reaction mixture was
o
stirred at 80 C for 6 hours. After cooling to ambient air temperature, the reaction
mixture was filtered and concentrated under reduced pressure. The residue was
purified by flash chromatography (ethyl acetate : hexane = 1 : 2) to give the product 4
o
1
as a yellow solid (3.88 g, 99% yield): mp 138-140 C; H NMR (400MHz, CDCl ) (4
3
:
5 mixture of rotomers) δ 1.27 (s, 17H), 1.41 (s, 4H), 1.86-1.93 (m, 2H), 1.60-2.22
(m, 2H), 3.37-3.57 (m, 2H), 5.04 (s, 5/9H), 5.16 (s, 4/9H), 7.21 (s, 4/9H), 7.28 (s,
+
5
/9H), 7.46 (s, 2H), 7.75 (s,2H); LC/MS (ESI) m/z: 457.2 [M + H] .
4
.1.4.
di-tert-Butyl
(
2S,2’S)-2,2’-(biphenyl-4,4’-diyldi-1,3-thiazole-5,2-diyl)dipyrrolidine-1-carboxylate
5)
(
A flask charged with PdCl (dppf) (0.48 g, 0.59 mmol), potassium carbonate (5.87 g,
2
4
2.5 mmol), compound 3 (3.75 g, 9.16 mmol), compound 4 (3.88 g, 8.5 mmol) and
1
,2-dimethoxyethane (100 mL) was flushed with nitrogen. The reaction mixture was
2
5

ACCEPTED MANUSCRIPT
o
stirred at 80 C over 24 hours. After cooling to ambient air temperature, the reaction
mixture was filtered and concentrated under reduced pressure. The residue was
purified by flash chromatography (ethyl acetate : hexane = 1 : 2) to give the product 5
o
1
as a white solid (2.66 g, 47% yield): mp 188-190 C; H NMR (400MHz, CDCl ) (4 :
3
5
4
mixture of rotomers) δ 1.37 (s, 10H), 1.51 (s, 8H), 1.95-2.04 (m, 4H), 2.27-2.35 (m,
H), 3.46-3.65 (m, 4H), 5.13 (s, 10/9H), 5.24 (s, 8/9H), 7.63 (s, 8H), 7.90 (s,2H);
+
LC/MS (ESI) m/z: 658.8 [M + H] .
4
.1.5. 5,5’-Biphenyl-4,4’-diylbis{2-[(2S)-pyrrolidin-2-yl]-1,3-thiazole} (6)
Trifluoroacetic acid (5 equiv) was introduced to the solution of compound 5 (2.66 g,
.04 mmol) in CH Cl (40 mL) at ice-bath. The resulting mixture was stirred at
4
2
2
room temperature for 2 hours, and then cooled with ice-bath. The saturated NaHCO3
aq) was introduced to neutralized the reaction mixture until the pH reached to 8.
The resulting mixture was partitioned between CH Cl and water. The combined
(
2
2
organic layers were dried over MgSO , filtered, and concentrated under reduced
4
pressure. The crude residue was purified by flash chromatography (100% ethyl
acetate, then methanol : CH Cl = 1 : 20) to give pure product 6 as a yellow solid
2
2
o
1
(
1.83 g, 99% yield): mp 247-249 C; H NMR (300MHz, DMSO-d ) δ 2.04-2.23 (m,
6
8
H), 3.33-3.41 (m, 4H), 5.12 (t, J = 5.9 Hz, 2H), 7.85 (d, J = 6.3 Hz, 4H), 7.82 (d, J =
2
6

ACCEPTED MANUSCRIPT
+
6
.3 Hz, 4H), 8.40 (s, 2H), 9.60 (br s, 2H); LC/MS (ESI) m/z: 459.0 [M + H] .
4
.1.6.
di-tert-Butyl
(biphenyl-4,4’-diylbis{1,3-thiazole-5,2-diyl(2S)pyrrolidine-2,1-diyl[(1R)-2-oxo-1-phe
nylethane-2,1-diyl]})biscarbamate (7)
A suspended mixture of N-Boc-D-phenylglycine (2.21 g, 8.8 mmol) and
hydroxybenzotriazole monohydrate (HOBt·H O, 1.35 g, 8.8 mmol) in DMF (30 mL)
2
was stirred at room temperature for 10 minutes. Then, EDC (1.68 g, 8.8 mmol) was
added and the clear solution was continuously stirred for 30 minutes. To the former
solution, compound 6 (1.83 g, 4.0 mmol) was added. The resulting mixture was
stirred at room temperature overnight and then partitioned between ethyl acetate and
water. The organic layer was dried over MgSO , filtered, and concentrated. The
4
residue was purified by column chromatography (methanol : CH Cl = 1 : 20) to yield
2
2
o
1
the pure product 7 as a white solid (2.77 g, 75% yield): mp 113-115 C; H NMR
(
(
(
300MHz, CDCl ) δ 2.06 (s, 18H), 3.33-3.43 (m, 4H), 3.60-3.68 (m, 4H), 3.89-3.99
3
m, 4H), 5.39-5.49 (m, 2H), 7.28-7.44 (m, 14H), 7.53 (d, J = 8.4 Hz, 2H), 7.59-7.68
+
m, 4H), 8.00 (s, 2H), 9.99 (br s, 2H); LC/MS (ESI) m/z: 925.3 [M + H] .
4
.2. Chemistry
2
7

ACCEPTED MANUSCRIPT
Synthetic Method A: General procedure for the synthesis of compounds 8-17 and
5
0-57
suspended mixture of N-substituted amino acid (2.2 equiv) and
A
hydroxybenzotriazole monohydrate (HOBt·H O, 2.2 equiv) in DMF was stirred at
2
room temperature for 10 minutes. Then, EDC (2.2 equiv) was added and the clear
solution was continuously stirred for 30 minutes. To the former solution, compound
6
(1 equiv) was added. The resulting mixture was stirred at room temperature
overnight and then partitioned between ethyl acetate and water. The organic layer
was dried over MgSO , filtered, and concentrated. The residue was purified by
4
column chromatography (methanol : CH Cl = 1 : 20) to afford the desired product.
2
2
Synthetic Method B: General procedure for the synthesis of compounds 18a-49a and
8b-49b
1
Trifluoroacetic acid (5 equiv) was introduced to the solution of compound 7 (1 equiv)
in CH Cl at ice-bath. The resulting mixture was stirred at room temperature for 2
2
2
hours, and then cooled with ice-bath. The saturated NaHCO (aq) was introduced to
3
neutralized the reaction mixture until the pH reached to 8. The resulting mixture was
partitioned between CH Cl and water. The combined organic layers were dried
2
2
2
8