Novel thioglycosyl analogs of glycosyltransferase substrates as antiviral compounds against classical swine fever virus and hepatitis C virus

Article abstract of DOI:10.1016/j.ejmech.2017.05.051

Hepatitis C virus (HCV) and classical swine fever virus (CSFV) are important pathogens for which new therapeutic approaches are in high demand. Herein, we report the synthesis of newly designed thioglycosyl analogs of glycosyltransferase substrates which were evaluated using cell-based assays for cytotoxicity and antiviral activity against both viruses. The antiviral activity of synthesized compounds against CSFV and HCV was confirmed using pseudo-plaque reduction assays where a significant arrest of viral growth was observed in the presence of selected compounds. We showed that compounds 13 and 14 exerted the most significant inhibitory effect on in vitro CSFV and HCV infections in the series. Glycoconjugates 13 and 14 not only inhibited both viral propagation with IC₅₀ values in low micromolar range, but efficiently suppressed the production of viral proteins in a dose-dependent manner. In addition, studies using in vitro HCV infection and replication models have shown that both compounds are able to significantly reduce viral genomic replication. We demonstrated that compounds 13 and 14 showed a strong inhibition, up to 90% of replication which inscribe them in the promising alternative approach for the development of new anti-CSFV and anti-HCV drugs.

Full text of DOI:10.1016/j.ejmech.2017.05.051

Accepted Manuscript
Novel thioglycosyl analogs of glycosyltransferase substrates as antiviral compounds
against classical swine fever virus and hepatitis C virus
Gabriela Pastuch-Gawolek, Binay Chaubey, Boguslaw Szewczyk, Ewelina Krol
PII:
S0223-5234(17)30420-8
10.1016/j.ejmech.2017.05.051
EJMECH 9481
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 25 February 2017
Revised Date: 29 April 2017
Accepted Date: 24 May 2017
Please cite this article as: G. Pastuch-Gawolek, B. Chaubey, B. Szewczyk, E. Krol, Novel thioglycosyl
analogs of glycosyltransferase substrates as antiviral compounds against classical swine fever virus and
hepatitis C virus, European Journal of Medicinal Chemistry (2017), doi: 10.1016/j.ejmech.2017.05.051.
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
Novel thioglycosyl analogs of glycosyltransferase
substrates as antiviral compounds against Classical
Swine Fever Virus and Hepatitis C Virus
1
,2
3,4
3
3*
Gabriela Pastuch-Gawolek , Binay Chaubey , Boguslaw Szewczyk , Ewelina Krol
1
Silesian University of Technology, Faculty of Chemistry, Chair of Organic Chemistry,
Bioorganic Chemistry and Biotechnology, Krzywoustego 4, 44-100 Gliwice, Poland,
2
Biotechnology Center, Silesian University of Technology, Krzywoustego 8, 44-100 Gliwice,
Poland
3
Department of Recombinant Vaccines, Intercollegiate Faculty of Biotechnology, University
of Gdansk and Medical University of Gdansk, Abrahama 58, 80-307 Gdansk, Poland
4
Functional Genomics Lab., Centre for Advanced Study, Department of Botany, University of
Calcutta, 35, Ballygunge Circular Road, 700019 Kolkata, India
*
To whom correspondence should be addressed: Ewelina Krol, Abrahama 58, 80-307 Gdansk
Phone: +48 58 523 63 83, fax: +48 58 305 73 12, e-mail: ewelina@biotech.ug.gda.pl
1

ACCEPTED MANUSCRIPT
ABSTRACT:
Hepatitis C virus (HCV) and classical swine fever virus (CSFV) are important
pathogens for which new therapeutic approaches are in high demand. Herein, we report the
synthesis of newly designed thioglycosyl analogs of glycosyltransferase substrates which
were evaluated using cell-based assays for cytotoxicity and antiviral activity against both
viruses. The antiviral activity of synthesized compounds against CSFV and HCV was
confirmed using pseudo-plaque reduction assays where a significant arrest of viral growth
was observed in the presence of selected compounds. We showed that compounds 13 and 14
exerted the most significant inhibitory effect on in vitro CSFV and HCV infections in the
series. Glycoconjugates 13 and 14 not only inhibited both viral propagation with IC values
5
0
in low micromolar range, but efficiently suppressed the production of viral proteins in a dose-
dependent manner. In addition, studies using in vitro HCV infection and replication models
have shown that both compounds are able to significantly reduce viral genomic replication.
We demonstrated that compounds 13 and 14 showed a strong inhibition, up to 90% of
replication which inscribe them in the promising alternative approach for the development of
new anti-CSFV and anti-HCV drugs.
KEYWORDS:
hepatitis C virus, classical swine fever virus, antivirals, glycoconjugates, analogs of
glycosyltransferase substrates
2

ACCEPTED MANUSCRIPT
ABBREVIATIONS:
Ac, acetyl; Bz, benzoyl; BAIB, [bis(acetoxy)-iodo]benzene; CSFV, classical swine fever
virus; DAAs, direct-acting antivirals; DMTMM, 4-(4,6-Dimethoxy-1,3,5-triazin-2-yl)-4-
methylmorpholinium chloride; DMSO, dimethyl sulfoxide; GTs, glycosyltransferases; HCV,
hepatitis C virus; TBDMS, tert-butyldimethylsilyl; TEMPO, 2,2,6,6-tetramethyl-1-
piperidinyloxy
1
. Introduction
The Flaviviridae family comprises many viruses associated with human and animal
diseases of great impact on public health. The most well-known members of this family are
important human pathogens including hepatitis C virus (HCV), dengue fever virus, yellow
fever virus as well as viruses of veterinary importance causing economically important
diseases like bovine viral diarrhea virus and classical swine fever virus (CSFV). In spite of
effective vaccines and therapies for some of these pathogens, viral mutations, drug resistance
or viral re-emergence generate problems for global control and eradication.
HCV, a member of the Hepacivirus genus and CSFV, a Pestivirus member, show high
degree of homology in genomic organization, replication and protein function. They are
enveloped viruses with single-stranded RNA genomes encoding a single polyprotein that is
cleaved into different structural and non-structural proteins by host and viral proteases. Due to
the similarity of CSFV and HCV, the former was frequently used as a surrogate model to
study the role of envelope glycoproteins of HCV [1,2].
CSFV causes an acute, highly infectious and economically damaging disease in swine
and wild boars in many countries [3,4]. Classical swine fever (CSF) is a major threat to
3

ACCEPTED MANUSCRIPT
commercial pig production worldwide [5]. CSF outbreaks among wild boars present a
constant threat for domestic pigs [6]. There is no treatment for classical swine fever, other
than supportive care. Several vaccines against CSFV are currently available, however the
most efficient vaccines are live attenuated ones developed over 50 years ago, which do not
allow for differentiating between infected and vaccinated animals [7]. The virus is still widely
circulating in endemic regions, therefore, the use of antiviral drugs could be a good control
strategy to prevent transmission of the virus in case of an outbreak.
HCV is a major cause of liver diseases in human, like chronic hepatitis, liver cirrhosis
and hepatocellular carcinoma. So far, there is no vaccine available [8]. It is estimated that
around 180 million people are chronically infected worldwide, which makes the virus one of
the most serious burdens to public health. Since the introduction of PEGylated interferon
alpha and Ribavirin as the standard drugs for clinical therapy of HCV in 1990, significant
efforts have been made to discover new and more effective treatments. New potent HCV
inhibitors belonging to direct-acting antivirals (DAAs) such as Boceprevir, Telaprevir,
Sofosbuvir, Simeprevir, Daclatasvir, and Ledipasvir have been developed in recent years [9].
These DAAs directly target enzymatic functions of viral proteins such as NS3/4A proteases,
NS5A or NS5B polymerase. New formulations based on combination of the new agents and
the traditional HCV inhibitors as well as multiple DAAs and interferon-free therapies have
improved the rates of sustained virologic response (SVR) and greatly reduced the side effects
[
10]. However, although the combination therapy based on Boceprevir or Telaprevir
inhibitors of NS3/4A protease) together with Ribavirin and PEGylated interferon alpha
(
improves antiviral response, it is limited to hepatitis C genotype 1. Introduced in 2013
Sofosbuvir (NS5B polymerase inhibitor) is the first drug used in combination with Ribavirin
in interferon-free therapy, however it is effective only for genotype 2 and 3 HCV patients. In
4

ACCEPTED MANUSCRIPT
2
013 another drug Simeprevir (protease inhibitor) has been approved by FDA in interferon-
free regimen. Combination therapy containing two other NS5A inhibitors, Daclatasvir and
Ledipasvir, show high success rates (even over 90%) depending on viral genotype [11].
Although a significant progress in anti-HCV therapy has been undoubtedly attained, many
limitations still exist. Monotherapy with DAAs are associated with rapid emergence of drug-
resistant viral mutants, therefore clinical application of DAAs is mostly limited to
combination regimens which is associated with additional side effects, drug-drug interactions,
high costs and availability. Thus, new therapeutic strategies consisting of new HCV inhibitors
targeting different stages of HCV life cycle, with increased effectiveness and wider
availability are still needed to overcome these limitations.
Proteins protruding from the viral envelope of many viruses are usually highly
glycosylated. Protein glycosylation is carried out by a family of glycosyltransferases (GTs),
which catalyze the transfer of a sugar moiety from an activated nucleotide sugar donor to a
hydroxyl or an amino groups of acceptor substrate [12,13]. GTs are involved in many
fundamental biological processes and modulation of their activities by efficient inhibitors is a
potential way for the control of certain cellular functions. A large number of potent natural as
well as synthesized GT inhibitors based on 3D structures of several GTs have been identified.
Most of synthesized GT inhibitors belong to donor substrate analogues, acceptor substrate
analogues or bisubstrate analogues.
In case of metal-dependent GTs, the pyrophosphate moiety of the donor-type substrate
2
+
2+
interacts with bivalent cations such as Mn or Mg with coordination of two aspartate
residues within a DXD motif in the enzyme active site. Numerous analogues of
nucleosidediphosphate (NDP) sugars with modifications of the diphosphate bridge have been
described. Some of them were obtained by varying the pyrophosphate moiety on phosphonate
5

ACCEPTED MANUSCRIPT
[
14,15], methylenediphosphonate [16] or ethyl phosphonophosphate [17]. However, the main
drawback of such compounds is their anionic character, which precludes their entry into cells
through the phospholipid bilayer. Due to this fact, the promising strategy for in vivo biological
applications is the preparation of GT inhibitors containing a neutral diphosphate surrogate
[
18,19].
Previously, a new kind of sugar nucleotide analogues, which were designed to act as GT
inhibitors were reported [20]. When compared to the natural GT substrates, the anomeric
oxygen atom was replaced in these structures by the sulfur in order to increase their resistance
to enzymatic hydrolysis. Another change in the structure was the replacement of
pyrophosphate bridge with a succinic linkage. The choice of such linker was based on the
ability to coordinate divalent metal ions (Scheme 1) [21]. Based on the foregoing
assumptions, glycoconjugates 8−10 where aminopyridyl 1-thioglycosides (derivatives of D-
glucose 5 or D-galactose 6) were connected to selectively protected uridine with succinic
spacer 1 or 2 through the amide bond were prepared [20].
O
O
O
NR3
O
NH
O
O
P
O
O
P
O
S
NH
O
N
HO
N
O
N
O
O
O
O
O
1
O
O
R O
R O OR2
sugar nucleotide analogues
2
HO OH
natural GT substrate
Scheme 1. Natural GT substrates and sugar nucleotide analogues.
In this paper, we described the synthesis and evaluation of antiviral activity of novel
thioglycosyl analogs of glycosyltransferase substrates against two viruses belonging to the
Flaviviridae family – CSFV and HCV in cell culture system. To investigate the effect of the
presence of the succinic linker on glycoconjugates biological activity, a series of new
glycoconjugates (11-17) were synthesized based on the same structural fragments as in case
6

ACCEPTED MANUSCRIPT
of glycoconjugates 8-10 but omitting the succinic linker. For the propose of this study,
glycoconjugates 8-10 were additionally prepared using improved method using microwave
irradiation. Two promising compounds with novel properties were chosen for further
development as lead hits. The broad spectrum activity exhibited by selected compounds
against these two viruses suggests their suitability as potential inhibitors against infections
caused by other viruses from the Flaviviridae family.
2
. Results and discussion
2
.1. Chemistry
Our preliminary studies on biological activity of compounds 8−10 indicated that some of
them were able to inhibit cell proliferation of the classical swine fever virus. These results
encouraged us to synthesize other glycoconjugates in which succinic fragment has been
omitted and N-(6-mercaptopyridyn-3-yl)formamide part was the only substitute of
diphosphate bridge (Scheme 2).
Scheme 2. General structure of tested glycoconjugates used in these studies.
7

ACCEPTED MANUSCRIPT
The effect of the presence and the type of protecting groups in both parts of
glycoconjugate: in the sugar ring and in the uridine moiety was also examined. As protecting
groups in the sugar ring acetyl, benzoyl and TBDMS groups were selected. Ribose in uridine
moiety was protected using the isopropylidene group or more hydrophobic TBDMS groups.
Ester type protecting groups were selected due to the possibility of their hydrolysis by
enzymes present within the cells. These groups only increased the hydrophobicity of
glycoconjugates and allowed them to penetrate into the cell. In turn, protecting groups in the
uridine part were chosen not only to allow the regioselective synthesis of glycoconjugates but
also to improve the hydrophobicity of the products and stability within the cell.
A new group of glycoconjugates, potential glycosyltransferases inhibitors, were
synthesized using improved, compared to the previously described, method of condensation.
The substrates used in condensation reactions were uridine derivatives containing carboxyl
group 1−4 and (5-amino-2-pyridyl) β-D-1-thioglycosides derivatives of D-glucose and D-
galactose 5−7 (Fig. 1).
Figure 1. Structure of substrates 1-7.
8

ACCEPTED MANUSCRIPT
The synthesis of uridine derivatives 1−2 [20] and 2’,3’-O-isopropylideneuridine-5’-
carboxylic acid 3 [22] was described earlier. 2’,3’-Di-O-tert-butyldimethylsilyluridine-5’-
carboxylic acid 4 was synthesized by oxidation of the corresponding 2’,3’-di-O-tert-
butyldimethylsilyluridine 4a with TEMPO/BAIB system in CH CN-water 1:1 solvents
3
1
13
system. H NMR and C NMR spectra confirmed the structure of obtained compound.
A second group of structural components of glycoconjugates was (5-amino-2-pyridyl) β-
D-1-thioglycosides 5−7. A simple and efficient synthesis of per-O-acetylated (5-amino-2-
pyridyl) β-D-1-thioglycosides 5 and 6 was described a few years ago [23]. (5-Amino-2-
pyridyl) 2,3,4-tri-O-benzoyl-6-O-tert-butyldimethylsilyl-1-thio-β-D-glucopyranoside 7 was
synthesized in a four-step procedure (Scheme 3) consisting of de-O-acetylation of (5-nitro-2-
pyridyl) 2,3,4,6-tetra-O-acetyl-1-thio-β-D-glucopyranoside 7a, selective silylation of 6-OH
group in compound 7b, benzoylation of the other three OH groups in the compound 7c and
reduction of the nitro group in aglycon of compound 7d. The nitro group reduction procedure
with zinc powder/acetic acid system in CH Cl was previously successfully used for the
2
2
preparation of derivatives 5 and 6. The reaction proceeded at room temperature. It is worth
noting that desilylation reaction at C-6 position was not observed under the reaction
conditions despite acidic reaction medium. This method allows to obtain the compound 7 in
good overall yield for the four step (50%).
9

ACCEPTED MANUSCRIPT
Scheme 3. Preparation of compound 7.
For amine and uridine derivatives, the next step was to carry out the condensation
reaction. The simple approach for the amide bond formation by mixing an amine and
carboxylic acid results in formation of a stable salt. Therefore, activation of carboxylic acid is
necessary [24]. Numerous condensing agents for efficient formation of an amide bond have
been described, of which the most common are carbodiimide derivatives (generally DCC)
[
25] alone or with additives such as DMAP or HOBt [26]. Another method for amide bond
formation utilizes an ethyl chloroformate generation of mixed anhydride [27]. Effective way
to synthesize amides in the presence of 2-chloro-4,6-disubstituted-1,3,5-triazines and tertiary
amines (e.g., N-methylmorpholine) has been published by Kaminski and co-workers [28].
These methods were previously used to form uridine glycoconjugates 8−10 [20]. The best
results were obtained using DMTMM (4-(4,6-Dimethoxy-1,3,5-triazin-2-yl)-4-
methylmorpholinium chloride) as a condensing agent. The only drawback of this method was
a long reaction time (up to two days). In the currently described experiments, in order to avoid
long reaction time microwave irradiation was used. For condensation of the amine group in
sugar derivatives 5−7 with carboxylic group in uridine derivatives 1−4 (Scheme 4), DMTMM
1
0

ACCEPTED MANUSCRIPT
was used in combination with microwave irradiation (Standard Program, 50°C) (Table 1,
Procedure B). Parallel condensation reactions using DMTMM at room temperature on a
magnetic stirrer were conducted (Table 1, Procedure A). As results of these reactions,
glycoconjugates 8−15 were obtained. The yields of the respective products are presented in
Table 1.
1
1

ACCEPTED MANUSCRIPT
Scheme 4. A. Synthesis of glycoconjugates 8-15. B. Structures of compounds 8-15.
Table 1. Yields of glycoconjugates 8-15.
Substrate 1
Substrate 2
Product Procedure Reaction time
[h]
Yield
[%]
53
Entry
Amine
Uridine deriv.
1
2
3
4
5
6
5
5
6
6
5
5
1
1
1
1
2
2
8
8
A
B
A
B
A
B
48
3
56
9
24
4
39
9
41
10
10
48
4
51
58
1
2

ACCEPTED MANUSCRIPT
7
8
9
5
5
6
6
5
5
6
6
7
7
3
3
3
3
4
4
4
4
4
4
11
11
12
12
13
13
14
14
15
15
A
B
A
B
A
B
A
B
A
B
96
2
38
49
29
33
32
53
43
61
52
54
96
2
1
0
1
2
3
4
5
6
1
1
1
1
1
1
96
2
96
2
120
4
Both condensation procedures yielded the desired glycoconjugates. It should be noted,
that the use of microwave irradiation not only allowed for the shortening of the reaction time
but also led to higher yield of products. This is particularly evident in the case of
glycoconjugates 13 and 14 containing uridine part protected with TBDMS groups. Most
likely, bulky TBDMS group in uridine derivative 4 causes a reduction of its reactivity in
condensation reaction. It is worth noting that the introduction of large protective groups to the
amine derivative 7 also decreased its reactivity. In the case of the reaction between amine 7
and uridine derivative 4, the reaction time for the standard procedure B (2 hours) was
insufficient. TLC analysis of the reaction mixture showed the presence of unreacted substrates
and therefore the reaction was continued for additional 2 hours.
1
3

ACCEPTED MANUSCRIPT
Scheme 5. Synthesis of glycoconjugates 16 and 17.
The final step of glycoconjugate synthesis was the hydrolysis of the protecting groups.
Compound 11 was selected as the substrate to provide selectively or fully deprotected
glycoconjugates (Scheme 5). The hydrolysis of the isopropylidene group from the uridine part
is facilitated in acidic conditions. In this case an acidic ion exchange resin Amberlyst-15 in
MeOH/H O system was used for deprotection to give glycoconjugate 16 (73% yield).
2
Complete removal of the protecting groups from glycoconjugate 11 was conducted in two
steps. The first step was the methanolysis of acetyl groups in the sugar part by the action of
sodium methoxide (NaOMe) in methanol. It should be noted that the same reagent was used
in the neutralization step of acetyl removal and in the reaction of isopropylidene group
removal. To explore fully the advantages of this convenient reagent, Amberlyst-15 was
employed in the two-step, one-pot deprotection of derivative 11 to give glycoconjugate 17 at a
moderate yield (38%).
1
4

ACCEPTED MANUSCRIPT
All glycoconjugates were purified by column chromatography and their structures were
determined by NMR spectra. The obtained compounds were evaluated for their inhibitory
properties against classical swine fever virus and hepatitis C virus.
2
.2. In vitro antiviral activity.
In primary screening we evaluated the in vitro antiviral activities of all synthesized
glycoconjugates using classical swine fever virus and hepatitis C virus. The measurements
included the cytotoxic concentration CC , which was defined as the compound concentration
5
0
that reduced cell viability by 50% measured using MTS assay for SK6 cells and MTT assay
for Huh-7.5 cells, the 50% inhibitory concentration (IC ), which was defined as the
5
0
compound concentration that reduced virus plaque formation by 50% and the selectivity index
SI (the ratio of CC to IC ). Due to the lack of cytopathic effect caused by both CSFV and
5
0
50
HCV viruses, accurate virus propagation measurement was based on visualization of the foci
caused by the virus (pseudo-plaques) using immunoperoxidase monolayer assay (IPMA)
which detects the areas of maximum concentrations of viral glycoproteins. The results of
these experiments are summarized in Table 2.
Table 2. Inhibitory effects of all synthesized compounds on CSFV and HCV replication in
SK6 and Huh-7.5 cells.
CSFV
HCV
Compound
a
b
c
a
b
c
CC (μM)
IC (μM)
SI
2
CC (μM)
IC (μM)
SI
50
50
50
50
8
9
207
194
45
103 ± 2
49 ± 1.18
13 ± 0.6
>325
328
340
71
231 ± 3.9
213 ± 2.8
45 ± 1.15
244 ± 3.4
1.4
1.6
1.6
1.1
4
1
1
0
1
3.5
>380
>1.2
271
1
5

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
2
3
4
5
6
7
>353
86
>299
3 ± 0.1
6 ± 0.4
97 ± 2.15
320
>1.2
28.7
25.2
1.9
1.1
1
312
135
173
>270
402
420
31
271 ± 2.4
7 ± 0.7
7 ± 0.4
>236
1.2
19.3
24.7
>1.1
1.1
151
186
350
380
ND
373
370
402
1.05
119.23
SOFOSBUVIR
ND
ND
0.26
a
Concentration required to reduce cell viability by 50%.
b
Concentration required to reduce virus plaque formation by 50%.
Expressed as the mean ± S.D. from three independent experiments.
In vitro selectivity index (CC /IC ).
c
5
0
50
ND: not determined
In the in vitro antiviral screening of the compounds, cytotoxicity analysis and antiviral
assay using CSFV were performed according to the procedures previously reported [29,30].
MTS-based cell proliferation assay (CellTiter 96 AQueous nonradioactive cell proliferation
assay) was carried out in order to determine the maximum concentrations of glycoconjugates
that could be used without causing a significant effect on cell viability. The toxicity analysis
of the compounds in SK6 cells demonstrated that tested compounds produced a dose-
dependent toxic effect, however no obvious cytotoxicity was observed with selected working
doses of the compounds in further experiments. Moreover, there was no cytotoxicity observed
in cells treated with the highest concentrations of the solvent (DMSO) used in the screening.
Anti-CSFV activity of all compounds was determined by pseudo-plaque reduction
assay where SK6 cells were infected with low MOI of the virus to visualize single plaques
and incubated for 48 h with varying amounts of tested glycoconjugates. Among ten
1
6

ACCEPTED MANUSCRIPT
synthesized compounds, some compounds were found to inhibit CSFV replication as
rns
evidenced by immunoperoxidase monalayer assay using rabbit polyclonal anti-E serum to
detect the viral antigen. These compounds had IC values < 50 µM showing selectivity
5
0
indexes (SI), defined as the CC /IC ratio, from 3.5 to 28.7. Compound 13 exhibited the
5
0
50
most potent antiviral activity from all synthesized compounds, with an IC value of 3 µM and
5
0
low cytotoxicity (CC50 of 86 µM), resulting in a selectivity index of 28.7 (Table 2).
Compound 14 also showed a very good activity against CSFV virus and had an IC value of
5
0
6
µM with similar to 13 selectivity index of 25.2. The typical IPMA reaction corresponding to
the dose-dependent inhibition of the hit compound 13 on CSFV infectivity is shown in Fig.
A. This compound caused more than 90% inhibition of CSFV infection at 9 µM
2
concentration, which was evaluated by reduction in the average size and number of pseudo-
plaques compared to the control.
Fig. 2. Dose-dependent effect of compound 13 on CSFV (A) and HCV (B) pseudo-plaque
formation in SK6 and Huh-7.5 cells, respectively. SK6 and Huh-7.5 cells were mock
infected or infected with CSFV at an MOI of 0.001 (A) or with HCV at an MOI of 0.1 (B). At
2
-3 h p.i. cells were treated with various concentrations of compound 13 or left untreated
(positive control). Two or three days post infection, cells were fixed and IPMA was
1
7

ACCEPTED MANUSCRIPT
rns
performed using rabbit polyclonal anti-E serum (A) or anti-HCV Core antibody (B) to
detect CSFV and HCV pseudo-plaques, respectively.
The analysis of anti-CSFV activity of all synthesized compounds revealed that the
sugar moiety type (D-glucose or D-galactose) does not significantly affect the biological
activity. In all the examined structures, amidopyridyl aglycon is linked with the sugar unit by
thioglycosyl linkage. In case of compounds 8−10 this heteroaromatic part is connected to
uridine derivative through succinic linker. Compounds 8 and 9, in which positions 2’-OH and
3
’-OH in ribose were protected using isopropylidene group, showed low activity against
CSFV virus with IC50 of 103 and 49 µM, respectively. Replacement of isopropylidene
protecting group with TBDMS groups and introduction of additional benzoyl protection at
N−3 position in uracil ring in compound 10 resulted in slight improvement in anti-CSFV
activity as compared to the parent compound 8. However, compound 10 had a higher toxicity
to the virus host cells resulting in moderate in vitro selectivity index of 3.5. Significant loss of
activity was observed for compounds 11 and 12 in which succinic linker was omitted and
aminopyridyl 1-thioglycoside was directly connected to the uridine part. These structural
modifications reduced the lipophilicity of compounds. Such a situation occurred when
hydroxyl groups in ribose were protected with an isopropylidene group. Both compounds did
not show any inhibition against CSFV in antiviral assay showing very low S.I. about 1.
Changing the isopropylidene protective group in compounds 11 and 12 to TBDMS groups
yielded respectively compounds 13 and 14, which had a slightly higher toxicity to virus host
cells however showed definitely higher activity. The increase of lipophilicity of compounds
had beneficial effect on their antiviral activity. At this point it is worth mentioning that earlier
observed increased toxicity of the compound 10 may be related to the fact that it may well act
as a benzoylation agent. This observation was supported by the fact that compounds 13 and
1
4, which lack the benzoyl group at the N−3 imide, do not show the increase in toxicity.
1
8

ACCEPTED MANUSCRIPT
These two compounds exhibited the most potent antiviral activity in the series with IC50
values of 3 and 6 µM, respectively and significant selectivity indexes of 28.7 and 25.2. In case
of compound 15, which showed no activity in the antiviral screening, acetyl protecting groups
in D-glucose part were replaced with benzoyl groups (positions 2-OH, 3-OH and 4-OH in
sugar part) and with TBDMS group (position 6-OH in sugar part). In this case, all protective
groups make compound 15 less toxic, however they are probably too bulky. These data
suggest that TBDMS group, especially inserted in the uridine moiety, may be important for
antiviral activity. Further experiments with the compound 16 showed, that the removal of
protective groups from the positions 2−OH and 3−OH in the uridine moiety leads to a loss of
antiviral activity. This is another proof of the importance of TBDMS groups in uridine part.
The complete removal of protective groups in compound 17 did not influence substantially
the results of antiviral activity with respect to compound 16. It was also shown that the acetyl
protective groups in sugar part affected the increase of glycoconjugate hydrophobicity but
inside cells most likely they undergo enzymatic hydrolysis.
All synthesized compounds were also evaluated for anti-HCV activity in HCV cell culture
system measured by HCVcc pseudo-plaque reduction assay (HCVcc, Jc1/JFH genotype 2a) in
human hepatoma Huh-7.5 cells, with Sofosbuvir, an inhibitor of the HCV NS5B RNA-
dependent RNA polymerase, as a positive control [31,32]. This system allows for complete
replication of HCV including in vitro production and secretion of HCVcc. Infected cells were
treated with increasing concentrations of the compounds, DMSO or Sofosbuvir and IPMA
assay using the specific anti-HCV Core antibody was performed. All the compounds were
used below their CC values as determined by MTT assay on Huh-7.5 cells after 48 h (Table
50
2
). The analysis showed that two compounds (13 and 14) inhibited the Jc1/JFH virus
propagation in a dose-dependent manner in Huh-7.5 cells and showed IC values of 7 µM
5
0
1
9

ACCEPTED MANUSCRIPT
which led to selectivity index of 19.3 and 24.7, respectively. These results were in agreement
with the experiments using CSFV indicating that synthetic compounds which showed the
inhibitory effect against CSFV were also active against hepatitis C virus. Although,
compound 13 showed poorer activity against HCV infection compared to its activity against
CSFV, compound 14 enhanced its potency. It can be concluded that these two compounds
effectively inhibit both flaviviruses. As shown in Fig. 2B the treatment with compound 13
resulted in a dose-dependent inhibition of viral replication, as monitored by IPMA, where
significant reduction in the size as well as number of infected foci among the drug treated
cells as compared to the untreated infected cells was observed. Compound 13 showed more
than 90% inhibition of HCV infection at a concentration of 80 µM.
2
.3. Inhibitory properties of selected compounds against CSFV and HCV viral
glycoprotein synthesis.
In the light of promising activity of 13 and 14, the action of these two compounds on
proliferation of CSFV and HCV was more thoroughly examined. The compounds were
synthesized as glycosyltransferase inhibitors, the enzymes involved in the glycosylation
process. Therefore, the effect of selected compounds on the glycoprotein synthesis of both
viruses was investigated using Western blot analysis under nonreducing and reducing
conditions. The levels of CSFV E2 protein in SK6 infected cells and HCV E1 protein in Huh-
7
.5 infected cells treated with compounds 13 and 14 were measured by Western blot analysis
using rabbit monospecific polyclonal serum anti-E2 CSFV and MAb anti-E1 HCV antibody,
respectively. These two proteins are highly glycosylated containing 6 and 5 glycosylation
sites, respectively.
2
0

ACCEPTED MANUSCRIPT
Both analysis showed that the level of synthesis of structural proteins for CSFV- and HCV-
infected cells was downregulated upon treatment with these compounds. Compounds 13 and
1
4 reduced CSFV E2 and HCV E1 protein production in a dose-dependent manner relative to
the positive DMSO-treated control, confirming that these two selected compounds exerted an
antiviral action against both CSFV (Fig. 3) and HCV (Fig. 4). The extent of reduction in
expression of CSFV E2 and HCV E1 protein treated with both compounds was consistent
with the results obtained in pseudo-plaques reduction assays.
Compound 13 completely inhibited CSFV E2 (Fig. 3A) and HCV E1 (Fig. 4A) protein
synthesis at concentrations of 10 µM and 75 µM, respectively. Similarly, treatment with 15
µM for CSFV (Fig. 3B) and 75 µM for HCV (Fig. 4B) of compound 14 resulted in a
complete arrest of CSFV E2 and HCV E1 glycoproteins synthesis compared to the untreated
control samples. However, no unglycosylated or underglycosylated forms of CSFV E2 and
HCV E1 proteins were detected after treatment with all, even the highest doses of both
compounds. The CSFV and HCV glycoproteins synthesized in cell cultures without the
addition of inhibitors and deglycosylated with PNGase F and Endo H were detected in
Western blot using the same antibodies, suggesting that the disappearance of glycoproteins in
the presence of inhibitor 13 and 14 was not related to the lack of reactivity with specific
antibodies [29].
The consistent synthesis of internal control (housekeeping gene - β-actin) in all samples
indicated no effect of tested concentrations of the compounds on the level of host cellular
proteins. These results demonstrate that compounds 13 and 14 inhibit protein synthesis
without cytoxicity.
The obtained results suggest that either the less glycosylated viral glycoproteins are not
detectable due to quick degradation process of incorrectly matured proteins or they are not
2
1

ACCEPTED MANUSCRIPT
produced due to other mechanism of action of synthesized compounds which can be
associated with another step of the viral life cycle.
Fig. 3. Effect of compounds 13 and 14 on the E2 CSFV protein accumulation.
CSFV-infected SK6 cells were treated with varying concentrations of compound 13 (A) or 14
(B). At 48 h p.i., cells were lysed and 20 µg protein was analysed by Western blot with
CSFV anti-E2 and anti-actin antibodies after SDS-PAGE under non-reducing conditions.
Fig. 4. Effect of compounds 13 and 14 on the E1 HCV protein accumulation.
HCV-infected Huh-7.5 cells were treated with varying concentrations of compound 13 (A) or
1
4 (B). At 72 h p.i., cells were lysed and 20 µg protein was analysed by Western blot with
HCV E1 and anti-actin antibodies after SDS-PAGE under reducing conditions.
2
2

ACCEPTED MANUSCRIPT
2
.4. Inhibitory properties of 13 and 14 against HCV depending on the timing of
infection and the exposure to compounds.
To further elucidate the mechanism of action and to assess the antiviral effect of the
most active compounds (13 and 14) on HCV infection, they were tested in the three different
models of infection according to Magri et al., 2016 using human hepatoma cell line Huh7-J20
generated to facilitate in vitro studies of HCV infection and replication [33]. This cell line
stably expresses enhanced green fluorescent protein (EGFP) fused in-frame to secretory
alkaline phosphatase (SEAP) via a recognition sequence of the viral NS3/4A serine protease
as a reporter system [34]. The SEAP reporter is released from the fusion protein and secreted
into the extracellular culture medium following cleavage by viral NS3/4A protease, produced
during viral replication. The level of SEAP activity in the culture medium directly correlates
with the level of intracellular viral RNA replication. In Model 1, Huh7-J20 cells were pre-
treated for 1 h and infected with cell culture infectious HCV (HCVcc) genotype 2a strain
JFH-1 in the presence of different concentrations of the compound or DMSO as a control for
3
h. The cells were then washed and fresh medium without the inhibitor was added for 72 h.
This model was used to investigate a possible effect on HCV entry. In Model 2, Huh7-J20
cells were pre-treated for 1 h, infected with JFH-1 HCVcc for 3 h in the presence of various
1
3 and 14 concentrations or DMSO and then fresh medium with compound or DMSO was
added for further 72 h. This model was used to evaluate the effect of inhibitors on full viral
life cycle. In Model 3, Huh7-J20 cells were first infected for 3 h with the virus and then the
fresh medium containing various concentrations of the inhibitor or DMSO was added in order
to determine the effect of drugs on post-viral entry (e.g. RNA replication and/or virus
2
3

ACCEPTED MANUSCRIPT
assembly). All of the doses of both compounds used in the experiments exerted no or little
effect on Huh7-J20 cell viability as determined by MTT assay (Fig. 5A).
The results showed that none of the tested compounds showed inhibitory effect on
virus entry (Model 1). Both compounds showed a strong anti-HCV activity on both the full
life cycle (Model 2) or post-entry model (Model 3) observed by a significant reduction in the
SEAP level, which suggest that they can affect virus genome replication. The results are
summarized in Fig. 5. The same modes of action were obtained for Sobosbuvir used as a
positive control, which is a licensed drug targeting viral replication (Supplementary
Information).
A.
1
1
20
00
80
60
40
20
0
compound 13
compound 14
7
5
50
25
12.5 6.25 3.125 1.56 DMSO
µM
B.
2
4

ACCEPTED MANUSCRIPT
1
1
20
00
80
60
40
20
0
Model 1 - entry
Model 2 - life cycle
Model 3 - post entry
7
5
50
25 12.5 6.25 3.125 1.56 DMSO
compound 13 (µM)
C.
1
1
20
00
80
60
40
20
0
Model 1 - entry
Model 2 - life cycle
Model 3 - post entry
7
5
50
25
12.5 6.25 3.125 1.56 DMSO
compound 14 (µM)
Fig. 5. Huh7-J20 cell viability (A) and antiviral activity of compounds 13 (B) and 14 (C)
on HCV infection. (A) Cell viability analysis after treatment with different doses of
compound 13 and 14 for 72 h without viral infection. The values of compound-treated cells
are expressed as percentage relative to DMSO-treated cells expressed as 100% (control). (B,
C) Huh7-J20 cells were pre-treated for 1 h and infected with cell culture infectious HCV in
the presence of different concentrations of 13 (B), 14 (C) or DMSO as a control for 3 h. Then,
the inoculum was removed and fresh medium without compound was added for 72 h (Model
1
, white bars). Huh7-J20 cells were pre-treated for 1 h, infected with JFH-1 for 3 h in the
2
5

ACCEPTED MANUSCRIPT
presence of various concentrations of 13 (B), 14 (C) or DMSO and then incubated for 72 h
with fresh medium with inhibitor or DMSO (Model 2, grey bars). Huh7-J20 cells were
infected for 3 h with JFH-1 and then treated with various concentrations of 13 (B), 14 (C) or
DMSO for 72 h (Model 3, black bars). All inhibitory effects were determined by measuring
SEAP assay performed on infected cell medium. Errors bars represent the SD of the means
for 3 experiments.
2
.5. The inhibitory effect of selected compounds on HCV replication.
Due to the fact that previous experiments indicated that both compounds could
interfere with viral replication, to confirm this hypothesis, both compounds were tested for the
ability to inhibit HCV replication using Huh7-J17 stable cell line, which constitutively
harbors the subgenomic HCV RNA. This is the puromycin-resistant cell line expressing
monocistronic replicon encoding non-structural proteins, structural core protein and firefly
luciferase as a reporter gene, the levels of which directly correlate with virus RNA replication
[
35].
These cells were plated on 96-well plates together with the various concentrations of both
drugs or Sofosbuvir as a positive control, incubated for 72 h and inhibition of HCV RNA
replication was estimated by measuring luciferase activity according to the protocol described
in Materials and Methods. Huh7-J17 cell viability was measured by MTT assay (Fig. 6A).
The obtained results confirmed the previous observation that both compounds highly
inhibited viral RNA replication (Fig. 6B). The highest doses of both compounds were able to
block the replication nearly in 90%. Half maximal inhibitory concentration (IC ) values for
5
0
2
6

ACCEPTED MANUSCRIPT
compound 13 and 14 were 4.005 µM and 3.741 µM, respectively. The half maximal cytotoxic
concentration (CC ) values were 185 µM for 13 and 245 µM for 14. Sofosbuvir displayed an
5
0
IC of 0.028 µM and CC of 23.5 µM in our assay. All the data obtained using Huh7-J17
5
0
50
cells confirmed the hypothesis previously observed in HCVcc system that compounds 13 and
4 are able to inhibit HCV replication.
1
A.
1
1
20
00
80
60
40
20
0
Compound 13
Compound 14
7
5
50
25
12.5 6.25 3.125 1.56 DMSO
µM
B.
2
7

ACCEPTED MANUSCRIPT
1
1
20
00
80
60
40
20
0
Compound 13
Compound 14
7
5
50
25
12.5 6.25 3.125 1.56 DMSO
µM
Fig. 6. Inhibitory effect of compounds 13 and 14 on viral replication. Huh7-J17 cells were
seeded in the presence of various concentrations of 13 and 14. After 72 h, cell viability was
determined using MTT assay (A) or cells were lysed and luciferase levels were measured (B).
The values of drug-treated cells are expressed as percentage relative to DMSO-treated cells
expressed as 100% (control). Errors bars represent the SD of the means for 3 experiments.
Additionally, the effect of compounds 13 and 14 against viral RNA production was
also determined using RT-PCR for NS5B gene performed on viral RNA isolated from Huh-
7
.5 HCV-infected cells treated with various concentrations of both compounds. As shown in
Fig. 7 treatment with these compounds resulted in a significant dose-dependent reduction in
the amount of viral RNA level compared to untreated infected cells.
2
8