Introduction

Hepatitis C virus (HCV) infection is considered as an immense burden on global health because it is the leading cause of liver cancer. Globally, an estimated 71 million people are infected with chronic HCV and susceptible to develop liver cirrhosis and hepatocellular carcinoma [1]. The high cost of treatment is the major challenge for HCV-infected patients in developing countries. In Egypt, where HCV genotype 4a (GT4a) infection hits 30% of the population and represents 15% of the global HCV infections, the treatment of HCV has become a national health concern [2]. Boceprevir, Telaprevir, Simeprevir, Ombitasvir, Sofosbuvir, and Dasabuvir are directly acting antivirals (DAA), which have been used in the form of combination therapies for HCV treatment [3,4,5]. HCV therapy has been progressed rapidly with particular focus on GT1a and GT1b, while the battle against GT4a requires more efforts because of the high resistance, progressive mutations, and pharmacoeconomic aspects [6, 7].

HCV NS5B is the responsible enzyme for genome replication of the HCV [8]. DAA drugs can manipulate the polymerase function of viruses either through the active (orthosteric) site by providing false nucleotide substrates leading to RNA chain termination or through the allosteric sites to inhibit RNA synthesis [8]. The NS5B structure features a classical “right hand” architecture of the polymerase family, which is characterized by the GDD functional motif (Gly317-Asp318-Asp319), which contains the orthosteric site. The fingers, palm, and thumb domains hold the allosteric sites (Fig. 1) [8, 9].

Fig. 1
figure 1

HCV NS5B crystal structure (PDB: 3FRZ) showing palm (red), thumb (blue), and finger (green) domain. The image generated by PyMOL4.0.4 [43]

Full size image

Non-nucleoside inhibitors are a class of small molecules that target the NS5B allosteric sites that are thumb I, thumb II, palm I, palm II, and palm III [8]. The binding of inhibitors to these sites results in a conformational change of the active site. Thumb II site is shallow and hydrophobic pocket at the base of the thumb domain. It is 35 Å away from the orthosteric site [10].

Thumb II inhibitors (Fig. 2) include dihydropyrone [11], thiophene carboxylic acid [12], and 2-arylimino-4-thiazolidinone [13] scaffolds (pink color). These scaffolds engage the thumb II site by hydrogen bonding with the backbone NH of Ser476 and Tyr477 and the side chain of Arg501. While right side substituents (blue color) provide hydrogen bond donor and acceptor atoms, the left side substituents (green color) offer hydrophobic contacts with the lipophilic amino acids of the pocket [14, 15].

Fig. 2
figure 2

HCV NS5B thumb II inhibitors

Full size image

Filibuvir, with a dihydropyrone scaffold, is a potent and selective inhibitor with IC50 of 0.019 μM against GT1b [16]. The triazolopyrimidine ring forms a π–π stacking interaction with His475; this ring is linked to the dihydropyrone core by a methylene group. Hydrophobic interaction between the thumb II site and one of the ethyl groups of the pyridine has provided good enzymatic potency [14]. Phase II clinical candidate Radalbuvir (GT1b EC50 = 0.006 µM) discovery led by the introduction of heterocyclic ethers on the N-alkyl substituent hydroxyl groups. The heterocyclic ether modulates protein binding and enhances the pharmacokinetic profile [15].

Both compounds 1 (GT1b EC50 = 0.6 µM) [17] and 2 (GT1b IC50 = 11 µM) [17] have similar hydrophobic substituents (green color), while they differ in the right side substituents. Compound 1 has extended pyridine-2-yl and compound 2 contains 1,2,3-thiadiazole to provide hydrogen bond acceptor atoms. Thiazolidinones were reported to have versatile biological activities including antiproliferative, antibacterial, antifungal [18,19,20]. Cyclooxygenase enzymes can be targeted by these compounds, which make thiazolidinones good scaffold to develop anti-tumor and anti-inflammatory agents [21]. Furthermore, thiazolidinone compounds have been shown to inhibit the HIV-1 reverse transcriptase [22].

Recently, it has been reported that coupling of the 1,3,4-thiadiazole ring with 4-thiazolidinone resulted in leads that strongly inhibit HCV GT1b NS5B [23, 24]. These studies have indicated that bioactivity was improved by increasing the lipophilicity of the scaffold. Also studies indicated that condensation of arylidene moiety on 1,3-thiazolidine-4-one ring could be essential for anti-HCV NS5B activity [23, 24]. However, this scaffold has not been tested yet for the resistant GT4a. In this study, the HCV GT4a NS5B model showed a polar region composed of Arg501, Arg505, Arg498, Lys531, and Lys533 that can be targeted by compounds with hydrogen bond acceptor functional groups. Compounds 7b, c, f, g, j, l, m, n, q, and r were synthesized and biologically evaluated to test this hypothesis. Furthermore, compound 7e with an extended alkyl side chain was synthesized and explored for better fitting in the lipophilic pocket. Compound 7a (Fig. 3) was reported to have good activity on HCV NS5B GT1b (IC50 = 25.3 µM) [23] and it is used in this study as a lead for the design of the new compounds.

Fig. 3
figure 3

The designed set of compounds with diverse substitutions on rings A and B

Full size image

Accordingly, a set of compounds with diverse substitutions on ring A and B (Fig. 3) has been designed, taking into consideration the features of the thumb II GT4a binding site and previous reports. Ring A (green color) should have halogens, while ring B (blue color) could be derivatized with halogens, hydrogen bond donor/acceptor functional groups, or extended alkyl side chains. A homology model of HCV NS5B GT4a was constructed and the compounds were docked into thumb II binding site in order to validate the binding affinity of the designed set of compounds. Based on the docking results, 18 compounds were selected for synthesis and were further evaluated for their activity.

Results and discussion

Homology modeling

The sequence of HCV NS5B GT4a was obtained from the UniProt database [25]. SWISS-MODEL server has been used to build the homology model (Supplementary Fig. S46, Supplementary Material).

A BLAST search of HCV NS5B GT4a against PDB sequence entries shows a match with 77.54 sequence identity. The SWISS-MODEL server was used to generate homology model for HCV NS5B GT4a. Global Model Quality Estimation (GMQE) and Qualitative Model Energy Analysis (QMEAN) scoring functions were used as an initial criterion to discriminate good from bad models. The model shows –0.09 QMEAN score (Supplementary Fig. S45a, Supplementary Material), which indicates good agreement between the homology model and experimental structures of the same size. The local quality estimate plot (Supplementary Fig. S45b, Supplementary Material) shows that there are no residues that gave a score below 0.6, which indicates high model quality. In the comparison plot (Supplementary Fig. S45c, Supplementary Material), the homology model is shown as a red star, and it is aligned with QMEAN score within one standard deviation of the mean |Z-score| as the experimental structures represented in black dots.

After building the homology model, it was subjected to protein refinement using the ModRefiner server [26]. Validation was carried out to explore the stereochemical quality of dihedral angles Ф against Ѱ of amino acid residues in the homology model and identify sterically allowed regions for these angles. Validation was carried out using SWISS-MODEL structure assessment online server [27] (Fig. 4). The Ramachandran plot of modeled HCV NS5B protein (Fig. 4A) shows 0.75 MolProbity score, 98.04% Ramachandran favored, and 0.18% Ramachandran outliers. The glycine and proline Ramachandran plot (Fig. 4B, C) show no outliers in disallowed regions.

Fig. 4
figure 4

The Ramachandran plot of modeled HCV NS5B protein using SWISS-MODEL structure assessment online server, A General Ramachandran plot, B Glycine Ramachandran plot, C Proline Ramachandran plot. Favored regions are shown in green, additional allowed region shown in light green, generously allowed regions are shown in pale green, and disallowed regions shown in white

Full size image

Molecular docking

The docking procedure was validated by using crystal structure of HCV NS5B GT1b in complex with Filibuvir (PDB: 3FRZ) [14]. The redocked binding pose of Filibuvir in GT1b observed overlays the original crystal structure binding pose with the same binding interactions as reported (Fig. 5A). The same established grid box coordinates in HCV NS5B GT1b was taken as a model to expect the binding site of HCV NS5B GT4a.

Fig. 5
figure 5

A Docking pose of Filibuvir to HCV NS5B GT1b thumb II site overlaying on the original crystal structure binding pose. B Docking pose of Filibuvir to HCV NS5B GT4a thumb II site and showing the new features of the allosteric pocket in GT4a

Full size image

The docking score of Filibuvir was decreased from –12.78 against HCV NS5B GT1b to –8.84 against HCV NS5B GT4a. This decline indicates that the features of the binding site in the HCV NS5B GT4a model changed (Fig. 5B). Two lipophilic surfaces were detected in thumb II GT4a, one of them was similar to GT1b. Hydrophilic surface, also, was detected in thumb II GT4a that can be targeted for hydrogen bonding and electrostatic interactions by the studied compounds (Fig. 5B).

The alignment of the 3D structure of GT1b crystal structure and the GT4a homology model was carried out to analyze the differences in the amino acids and the spatial arrangement of the critical residues of interaction (Fig. 6).

Fig. 6
figure 6

Alignment of the 3D structure of HCV NS5B GT1b crystal structure (green color) (PDB: 3FRZ) and HCV NS5B GT4a homology model (cyan color). Ser476, Leu419, Ile482, and Arg501 (red color) are of HCV NS5B GT1b, Gly476, Ile419, Leu482, and Arg501amino acids (blue color) are of HCV NS5B GT4a. The image was generated by PyMOL4.0.4 [43]

Full size image

HCV NS5B GT1b thumb II site consists of Leu497, Leu419, Ile482, Val376, His475, Trp528, Ser476, and Arg501. These amino acids are the key amino acids for interaction with the reported active compounds of dihydropyrone, thiophene carboxylic acid, and 2-arylimino-4-thiazolidinone scaffolds [28]. However, in HCV NS5B GT4a thumb II there are three amino acid mutations reported, S476G, I482L, and L419I that render the binding site more lipophilic and confer a high level of resistance [6, 7]. These findings indicate that the binding site of HCV NS5B GT4a lost the features of interactions that afforded in HCV NS5B GT1b. Thus, the designed compounds should have functional groups that afford more hydrogen bond interactions with optimal lipophilicity. Also, compounds with extended side chains will be investigated for a suitable fitting of the molecule to the binding site.

Docking results showed that the thiadiazole ring with substituted halogenated phenyl derivatives interacts with the hydrophobic region of thumb II site, which indicates that these compounds can be active against HCV NS5B GT4a. The top 18 compounds were selected for synthesis and biological evaluation (Table 1).

Table 1 The docking scores and IC50 results of the synthesized compounds (7a–r)
Full size table

Chemistry

Compounds 7a–r were synthesized according to Scheme 1. Thiosemicarbazide 2 and different aldehydes 1a–e reacted with each other to give the thiosemicarbazone derivatives 3a–e.

Scheme 1
scheme 1

Synthetic route for compounds 7a–r. (i) Acetic acid, reflux, 12 h; (ii) Ferric chloride, ethanol, reflux,20 h; (iii) Chloroacetyl chloride, triethylamine, dichloromethane, 4 h; (iv) Ammonium thiocyanate, reflux, 20 h; (v) Different aldehydes, CH3ONa, methanol, reflux, 12 h

Full size image

Compounds 3a–e undergo oxidative cyclization into 2-Amino-1,3,4-thiadiazole derivatives 4a–e [23]. A suggested mechanism for the oxidative cyclization is shown in Scheme 2 [23, 29]. According to this mechanism, reaction is initiated by deprotonation, forming a radical at thiosemicarbazone N2. The resonance hybrid thiol radical attacks azomethine carbon to form the ring. The expulsion of another hydrogen radical finally yields 5-aryl-1,3,4-thiadiazole-2-amines [23, 29].

Scheme 2
scheme 2

The mechanism of 1,3,4-thiadiazole ring formation

Full size image

Compounds 4a–e were converted to compounds 5a–e by reaction with chloroacetyl chloride in the presence of trimethylamine. The synthesis of compounds 6a–e was carried out by the reaction of compounds 5a–e with ammonium thiocyanate [23]. Compounds 5a–e react with ammonium thiocyanate to form intermediates that undergo intramolecular cyclization to generate compounds 6a–e as indicated in Scheme 3 [30].

Scheme 3
scheme 3

The mechanism of thiazolidin-4-one ring formation

Full size image

The target compounds 7a–r were obtained by the reaction of compounds 6a–e with different aldehydes in the presence of sodium methoxide [23].

Compounds 3a–e, 4a–e, 5a–e, and 6a–e were previously reported [23, 24], while compounds 7a–r are new and confirmed through elemental and spectroscopic analysis. Elemental analysis (within the accepted limits (±0.4 %)) confirms the purity of the final compounds and the spectroscopic data of the synthesized final compounds 7a–r agreed with the expected chemical structures as given in the experimental section.

IR of compounds 7a–r shows one broad peak in the range (3450–3300 cm−1) of the lactam NH and peak at 1720 cm−1 of the carbonyl functional group in the thiazolidinone ring. 1H NMR shows the characteristic singlet at (12.5–13.5 ppm) that is assigned to NH of 1,3-thiazolidin-4-one ring and the peaks of the aromatic protons and exocyclic =CH- proton were found in the range of (7–8 ppm). The 1H NMR spectra of the compounds 6a–e were characterized by the (singlet, 2H) peak of the active methylene in 1,3-thiazolidin-4-one ring at 4.1 ppm [23]. The absence of this peak in 1H NMR of compounds 7a–r confirms the condensation of the different aldehydes and formation of the target final compounds (7a–r). The exocyclic =CH- proton in compounds 7a–r showed singlet peak with resonances at 7.70–8.10 ppm within the aromatic peaks range of 7–8 ppm. 5-Arylidene-1,3-thiazolidinones exist as Z configurational isomer as evidenced by previous literature reports of the 1H NMR [31] and X-ray crystallography [32].

The 13C NMR spectra of the reported compounds 6a–e characterized by the thiazolidinone C5 signal at the range of (30–38 ppm) [23], while the corresponding carbon signals in the final compounds 7b–f, 7n, and 7q were shifted to the range of (135–155 ppm).

1H-1H correlation spectroscopy (COSY) and heteronuclear single quantum correlation (HSQC) experiments were carried out for compound 7b as a representative example for further structural confirmation. The atom numbering of compound 7b was generated by ChemBioDraw software [33] specifically for NMR analysis. The 1H-1H COSY (Supplementary Fig. S5, Supplementary Material) confirms the 1H NMR assignment as follow; signal at 7.2 ppm (H21) is correlated with signal at 7.3 ppm (H22), signal at 7.7 ppm (H18) is correlated with signals at 7.5 ppm (H17) and 7.6 ppm (H19). HSQC (Supplementary Fig. S6, Supplementary Material) spectrum indicates that the exocyclic H13 singlet signal at 7.8 ppm in 1H NMR is correlated with 134.83 ppm (C13) in 13C NMR that is more de-shielded than C21, C22, and C25, while the signal at 126.18 ppm (C25) is correlated with 7.3 ppm (H25) that confirms the assignment of the signal at 7.3 ppm to H22. Signal at 115.07 ppm (C17) is correlated with 7.5 ppm (H17), signal at 126.21 ppm (C19) is correlated with 7.6 ppm (H19), and signal at 135.86 ppm (C18) is correlated with 7.7 ppm (H18).

Biological evaluation and molecular modeling studies

In vitro HCV NS5B GT4a inhibition assay

The HCV NS5B GT4a inhibition assay is carried out on 18 synthesized compounds 7a–r (Table 1). The results show that compound 7a (IC50 = 1.029 µM) is 25 folds more potent on HCV NS5B GT4a than reported on HCV NS5B GT1b (IC50 = 25.3 µM) [23], from which we conclude that 2-[(5-substituted phenyl-1,3,4-thiadiazol-2-yl)imino]-5(substituted phenylmethylidene)-1,3-thiazolidin-4-one scaffold is more potent on HCV NS5B GT4a.

The NS5B inhibitory assay results illustrate that compounds 7a–d and 7n–r having 2-Chloro-6-fluorophenyl and 2-Chlorophenyl ring A substitution on ring A respectively exhibit better HCV NS5B inhibition activity than 3-Fluoro, 4-Fluoro, and 4-chloro substitutions. Among the series 7a–d and 7n–r, compounds 7b and 7n (IC50 = 0.338 and 0.342 µM, respectively) are the most potent compounds. Compound 7b is considered to be three folds more potent than the lead compound 7a (IC50 = 1.029 µM on HCV NS5B GT4a). Also, compound 7b is considered to be 74 folds more potent than the lead compound 7a (IC50 = 25.3 µM on HCV NS5B GT1b).

Analysis of the results focusing on ring B and considering compounds 7a–d, 7h, 7k, 7l, 7n–r reveals that the substitution on meta or para positions is more favorable than substitution on ortho position, as exhibited by the IC50 values of compounds 7h and 7i showing IC50 2.401 and 6.66 μM, respectively. The results show that the most active four compounds are 7b, 7c, 7k, 7n, which can be explained by the docking studies. Generally, the most prominent interaction type of compounds 7a–r (Figs. 7–10, and S47-S60, Supplementary Material) with the surrounding amino acids is in the form of hydrogen bonding through the carbonyl oxygen and the NH group of the thiazolidinone ring. Analysis of the binding interactions of 7b, 7c, 7k, and 7n with the thumb II site is summarized in Table 2.

Fig. 7
figure 7

The 3D (left) and 2D (right) of the docking interactions of compound 7b with HCV NS5B GT4a thumb II site

Full size image
Fig. 8
figure 8

The 3D (left) and 2D (right) of the docking interactions of compound 7c with HCV NS5B GT4a thumb II site

Full size image
Fig. 9
figure 9

The 3D (left) and 2D (right) of the docking interactions of compound 7k with HCV NS5B GT4a thumb II site

Full size image
Fig. 10
figure 10

The 3D (left) and 2D (right) of the docking interactions of compound 7n with HCV NS5B GT4a thumb II site

Full size image
Table 2 Analysis of binding interactions of 7b, 7c, 7k, and 7n with HCV NS5B GT4a thumb II site
Full size table

The most active compound 7b forms two hydrogen bonds with Lys533 and Thr473, and it shows hydrophobic contacts with many hydrophobic amino acids in the binding site (Fig. 7 and Table 2).

The decrease in the biological activity of compound 7c could be explained by the inability to form two hydrogen bonds compared to compound 7b. Compound 7c forms one hydrogen bond with His475 amino acid. Also, it has fewer nearby hydrophobic interactions (Fig. 8 and Table 2).

Compound 7k forms three hydrogen bonds with His475, Arg422, Thr473 amino acids, and shows less hydrophobic interactions through aromatic rings. Also, hydrogen bond distance with Arg422 is less than optimum (Fig. 9 and Table 2).

Also, compound 7n forms three hydrogen bonds with His475, Arg422, Thr473 amino acids, and this can explain its high activity. 7n shows hydrophobic interactions through aromatic rings with hydrophobic amino acids (Fig. 10 and Table 2).

Antiviral activity on HCV GT4a

Compounds 7a–d, 7k, 7n, 7q, and 7r showed the most potent activities with IC50 < 2 µM. Thus, they have been selected for antiviral activity testing. The results illustrated in Table 3 show that all eight tested compounds exhibited more than 79% inhibition of the HCV GT4a viral activity, which indicates the good activity of these compounds. Compounds 7b (0.338 µM, 92.88%) and 7c (0.634 µM, 94.77%) have high % inhibition compared to Beclabuvir reference compound (92.72%). Compound 7c exhibited the highest % inhibition among the tested compounds, while 7b showed higher IC50 with good percentage of inhibition too.

Table 3 HCV NS5B GT4a antiviral assay of the selected compounds
Full size table

Cytotoxicity study on human hepatoma (Huh7.5) cell line

Compounds 7b and 7c, the most active inhibitors, and the lead compound 7a were subjected to cytotoxicity study on Huh7.5 cell line along with Beclabuvir as a reference compound (Table 4). The assay was carried out using MTT assay kit, in which the detection of cell metabolic activity is based on the purple color that results from biochemical reaction. The developed color is directly proportional to the cell number, which indicates the cell viability.

Table 4 Cytotoxic activity (CC50) of 7a, 7b, and 7c against human hepatoma (Huh7.5) cell line
Full size table

The results show that all tested compounds exhibited low cytotoxicity on Huh7.5 cell lines compared to Beclabuvir reference compound. Moreover, compound 7b exhibited the least cytotoxicity by more than six folds compared to Beclabuvir.

Conclusion

Our efforts are directed to develop potent inhibitors of HCV NS5B GT4a for treating HCV GT4a infection. A set of compounds with 4-thiazolidinone derivatives has been designed. To validate the binding affinity of this set of compounds, homology model of HCV NS5B GT4a was constructed and the compounds were docked. Based on the docking results, 18 compounds (7a–r) were selected for synthesis and were further evaluated for their activity. Compounds (7a–r) were subjected to in vitro HCV NS5B GT4a inhibition assay. Compounds 7b and 7n showed the highest inhibitory activity (IC50 = 0.338 and 0.342 µM, respectively). Compounds 7a, 7b, 7c, 7d, 7k, 7n, 7q, and 7r with IC50 < 2 µM were selected for the assessment of the anti-HCV GT4a activity using human hepatoma cell line (Huh 7.5), the percentage of inhibition of the tested compounds was between 79.67 and 94.77%. Compounds 7a, 7b, and 7c were tested for their cytotoxicity on the Huh7.5 cell line and exhibited low cytotoxicity compared to Beclabuvir reference compound. Compound 7b was identified through this study as a potential new lead for future HCV inhibition studies.

Materials and methods

Chemistry

General

Chemical reagents and materials were supplied by Sigma-Aldrich and Alfa Aesar companies and were used without further purification. Electrothermal melting point device (Stuart SMP30) was used for melting points detection, and the results were reported uncorrected. TLC aluminum sheets Silica 60 F254 (Merck Co.) were used for reactions monitoring and were visualized with UV light at 254 nm, using TLC system:chloroform/methanol/acetic acid (9.3:0.5:0.2 v/v/v). The PerkinElmer 2400 series II CHNS elemental analyzer was used for elemental analysis. IR spectra of all compounds were recorded on a PerkinElmer FT-IR spectrophotometer using KBr discs with a frequency range of (4000–400 cm−1). 1H NMR spectra were processed on Bruker FT-NMR (400 MHz) spectrometer using dimethyl sulfoxide (DMSO) as a solvent, and the chemical shifts were reported as ppm. 13C NMR spectra of the representative examples (7b–f, 7n, 7q) were recorded on Bruker FT-NMR (100 MHz) using DMSO-d6 as a solvent with tetramethylsilane as an internal standard. Pyridine was used to dissolve compounds 7b, 7n, and 7f duo to their poor solubility in DMSO. All chemical shift values and multiplicity were quoted in ppm.

General synthetic procedure of 2-(arylmethylidene)hydrazinecarbothioamides (3a–e) [23]

Solutions of thiosemicarbazide 2 (30 mmol) were added to solutions of aldehydes (1a–e) (30 mmol), and five drops of acetic acid were added then the reaction mixtures were heated under reflux for 12 h. The products (3a–e) were filtered, and the recrystallization was performed using ethanol.

Compounds 3a, 3d, and 3e [23] and compounds 3b and 3c [34] were reported.

General synthetic procedure of 5-(substituted phenyl)-1,3,4-thiadiazol-2-amines (4a–e) [23, 24]

Solutions of compounds 3a–e (2 mmol) in ethanol were added to solutions of ferric chloride (8 mmol). Then the reactions were heated under reflux for 20 h. The reactions were neutralized by ammonia solution, filtered, dried, and the recrystallization was performed using ethanol.

Compounds 4a, 4d, and 4e [23] and compounds 4b and 4c [24] were reported.

General synthetic procedure of 2-chloro-N-[5-(substituted phenyl)-1,3,4-thiadiazol-2-yl]acetamides (5a–e) [23, 24]

Compounds 4a–e (10 mmol) were dissolved in dichloromethane, and triethylamine (12 mmol) was added to the reaction pots. Chloroacetyl chloride (20 mmol) was added to the reaction pots slowly under anhydrous condition, followed by stirring for 4 h under reflux. The reactions were filtered then dried, and the recrystallization was performed using ethanol.

Compounds 5a, 5d, and 5e [23] and compounds 5b and 5c [24] were reported.

General synthetic procedure of 2-{[5-(substituted phenyl)-1,3,4-thiadiazol-2-yl]imino-1,3-thiazolidin-4-ones (6a–e) [23, 24]

Solutions of compounds 5a–e (2.5 mmol) in 1,4-dioxane were added to ethanolic solutions of ammonium thiocyanate (5 mmol). The reactions were stirred under reflux for 20 h. The solvent was evaporated. The reactions were filtered then dried, and the recrystallization was performed using ethanol.

Compounds 6a, 6d, and 6e [23] and compounds 6b and 6c [24] were reported.

General synthetic procedure of 5-(substituted arylidene)-2-{[5-(substituted phenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-ones (7a–r) [23, 24]

Compounds 6a–e (1.5 mmol) were dissolved in methanol, then sodium methoxide (3 mmol) in methanol was added, and the mixture was heated for 10 min. The aldehyde (1.8 mmol) was dissolved in methanol in excess amount and added to the mixture. The reaction mixtures were heated under reflux for 12 h. The reactions were left to cool at 25 °C. Then 40 mL ice-cold water was added. Then the mixtures were neutralized with glacial acetic acid to give compounds 7a–r. The precipitated products were filtered, washed with water, and the recrystallization was performed using Dimethylformamide.

5-(3-Fluorobenzylidene)-2-{[5-(2-chloro-6-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7a)

The results of the characterization of the compound 7a (shown in Page 2, Supplementary Material) were as reported [23].

5-(1,3-Benzodioxol-5-ylmethylidene)-2-{[5-(2-chloro-6-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7b)

Pale yellow solid; yield (0.065 g, 20.2%) with mp 285–289 °C. IR (KBr, cm−1): 3449 (N-H str.), 3145 (Ar-H), 2902 (C-H str. of CH2), 1705 (C=O, lactam), 1593 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 6.17 (s, 2H, CH2), 7.17–7.19 (d, 1H, Ar-H21), 7.2 (s, 1H, Ar-H25), 7.27–7.29 (d, 1H, Ar-H22), 7.49–7.53 (t, 1H, Ar-H17), 7.59–7.61 (d, 1H, Ar-H19), 7.67–7.71 (q, 1H, Ar-H18), 7.78 (s, 1H, =CH13), 13.04 (broad s, 1H, NH of lactam). 13C NMR (100 MHz, pyridine) δ (ppm): 102.28 (C27), 109.16 (C21), 109.71 (C22), 114.85 (C6), 115.07 (C17), 122.78 (C15), 123.81 (C20), 126.18 (C25), 126.21(C19), 132.67 (C23), 132.79 (C24), 134.83 (C13), 135.86 (C18), 148.74 (C16), 149.08 (C10), 150.15 (C2), 159.70 (C8), 162.21 (C4), 198.20 (C11). Anal. Calcd. for C19H10ClFN4O3S2: C, 49.51; H, 02.19; N, 12.16; S, 13.91%; Found: C, 49.80; H, 02.42; N, 12.35; S, 14.15%.

5-(4-Dimethylaminobenzylidene)-2-{[5-(2-chloro-6-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7c)

Orang solid; yield (0.132 g, 38.27 %) with mp 270–273 °C. IR (KBr, cm−1): 3439 (N-H str.), 3142 (Ar-H), 2926 (C-H str. of CH3), 1699 (C=O, lactam), 1583 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.04 (s, 6H, 2-CH3), 6.89–6.91 (d, 2H, Ar-H), 7.48–7.72 (m, 6H, Ar-H, =CH), 12.84 (s, 1H, NH of lactam). 13C NMR (100 MHz, DMSO) δ (ppm): 40.04 (2 CH3), 112.66, 115.68, 115.89, 116.27, 120.29, 126.77, 133.05, 134.37 (Ar-C), 135.05 (thiazolidinone C5), 152.14 (=CH-Ar), 160.51, 161.90 (thiadiazole C5, thiazolidinone C2), 154.40 (thiadiazole C2), 167.75 (thiazolidinone C4). Anal. Calcd. for C20H15ClFN5OS2: C, 52.23; H, 03.29; N, 15.23; S, 13.94%; Found: C, 52.51; H, 03.49; N, 15.45; S, 13.76%.

5-(3-Chlorobenzylidene)-2-{[5-(2-chloro-6-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7d)

Yellow solid; yield (0.443 g, 65.44%) with mp 284–287 °C. IR (KBr, cm−1): 3437 (N-H str.), 3095 (Ar-H), 1714 (C=O, str.), 1598 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.48–7.84 (m, 7H, Ar-H), 7.79 (s, 1H, =CH), 13.17 (s, 1H, NH of lactam). 13C NMR (100 MHz, DMSO) δ (ppm): 115.68, 115.89, 126.23, 126.73, 128.42, 130.66, 131.77, 132.06, 134.02, 134.31, 134.34, 134.46, 135.76, 155.18 (Ar-C, =CH-Ar, thiazolidinone C5), 159.39, 159.69 (thiadiazole C5, thiazolidinone C2), 161.89 (thiadiazole C2), 167.33 (thiazolidinone C4). Anal. Calcd. for C18H9Cl2FN4OS2: C, 47.90; H, 02.01; N, 12.42; S, 14.21%; Found: C, 47.73; H, 01.83; N, 12.71; S, 13.97%.

5-(3-Phenylpropylidene)-2-{[5-(4-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7e)

Yellow solid; yield (0.063 g, 20.46%) with mp 224–226 °C. IR (KBr, cm−1): 3447 (N-H str.), 3084 (Ar-H), 2923 (C-H str. of CH2), 1728 (C=O, str.), 1585 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 2.61–2.62 (m, 2H, CH2CH2-Ph), 2.84–2.89 (m, 2H, CH2CH2-Ph), 6.92–6.94 (t, 1H, =CH), 7.20–7.42 (m, 7H, ArH), 7.97–7.99 (t, 2H, ArH), 12.78 (s, 1H, NH of lactam). 13C NMR (100 MHz, DMSO) δ (ppm): 33.09, 33.59 (2 CH2) 116.89, 117.11, 126.68, 127.06, 127.83, 128.83, 129.93, 138.09, 140.85 (Ar-C), 159.32 (thiazolidinone C5), 162.82 (=CH-Ar), 164.14, 165.30 (thiadiazole C5, thiazolidinone C2), 166.09 (thiadiazole C2), 170.32 (thiazolidinone C4). Anal. Calcd. for C20H15FN4OS2: C, 58.52; H, 03.68; N, 13.65; S, 15.62%; Found: C, 58.78; H, 03.38; N, 13.89; S, 15.37%.

5-(3,4-Dimethoxybenzylidene)-2-{[5-(4-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7f)

Yellow solid; yield (0.170 g, 38.49%) with mp 266–268 °C. IR (KBr, cm−1): 3452 (N-H str.), 1640 (C=O st., lactam) 1587 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.84 (s, 6H, CH3), 7.22–8.02 (m, 7H, Ar-H), 8.02 (s, 1H, =CH), 12.92 (s, 1H, NH of lactam); 13C NMR (100 MHz, Pyridine) δ (ppm): 55.66 (2 OCH3), 112.17, 114.23, 116.26, 116.48, 123.83, 127.60, 129.53, 129.61, 134.84 (Ar-C, =CH-Ar), 135.61 (thiazolidinone C5), 151.52 (thiazolidinone C2), 162.30 (thiazolidinone C4). Anal. Calcd. for C20H15FN4O3S2: C, 54.29; H, 03.42; N, 12.66; S, 14.49%; Found: C, 54.51; H, 03.62; N, 12.99; S, 14.98%.

5-{[4-(Benzyloxy)-3-methoxyphenyl]methylidene}-2-{[5-(4-chlorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7g)

Yellow solid; yield (0.085 g, 23.72%) with mp 259–261 °C. IR (KBr, cm−1): 3451 (N-H str.), 1641 (C=O st., lactam), 1587 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.85 (s, 3H, CH3), 5.19 (s, 2H, CH2), 7.23–7.78 (m, 12H, Ar-H), 7.97 (d, 1H, =CH), 12.96 (s, 1H, NH of lactam). Anal. Calcd. for C26H19ClN4O3S2: C, 58.37; H, 03.58; N, 10.47; S, 11.99%; Found: C, 58.56; H, 03.79; N, 10.62; S, 12.26%.

5-(3-Bromobenzylidene)-2-{[5-(3-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7h)

Yellow solid; yield (0.075 g, 21.74%) with mp 274–276 °C. IR (KBr, cm−1): 1721 (C=O st., lactam), 1610 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.40–7.80 (m,8H, Ar-H), 7.90 (s,1H, =CH), 13.12 (s, 1H, NH of lactam). Anal. Calcd. for C18H10BrFN4OS2: C, 46.86; H, 02.19; N, 12.15; S, 13.89%; Found: C, 46.99; H, 01.92; N, 11.98; S, 14.21%.

5-(2-Bromobenzylidene)-2-{[5-(3-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7i)

Yellow solid; yield (0.058 g, 16.76%) with mp 267–268 °C. IR (KBr, cm−1): 3434 (N-H str.), 1724 (C=O st., lactam), 1603 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.41–7.85 (m, 8H, Ar-H), 7.90 (s, 1H, =CH), 13.18 (s, 1H, NH of lactam). Anal. Calcd. for C18H20BrNO4: C, 46.86; H, 02.19; N, 12.15; S, 13.89%; Found: C, 46.57; H, 02.48; N, 12.42; S, 13.61%.

5-(4-Dimethylaminobenzylidene)-2-{[5-(3-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7j)

Orange solid; yield (0.093 g, 29.14%) with mp 255–258 °C. IR (KBr, cm−1): 3448 (N-H str.), 3145 (Ar-H), 1699 (C=O st., lactam), 1583 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.06 (s, 6H, 2 CH3), 7.71 (s, 1H, =CH), 6.90–7.82 (8H, Ar-H), 12.81 (s, 1H, NH of lactam). Anal. Calcd. for C20H16FN5OS2: C, 56.45; H, 03.79; N, 16.46; S, 15.07%; Found: C, 56.68; H, 04.10; N, 16.71; S, 15.32%.

5-(3-Chlorobenzylidene)-2-{[5-(3-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7k)

Yellow solid; yield (0.158 g, 50.54%) with mp 275–278 °C. IR (KBr, cm−1): 3448 (N-H str.), 3151 (Ar-H), 1709 (C=O str., lactam), 1593 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.41–7.95 (m,9H, ArH, =CH), 13.13 (s, 1H, NH of lactam). Anal. Calcd. for C18H10ClFN4OS2: C, 51.86; H, 02.42; N, 13.44; S, 15.38%; Found: C, 51.56; H, 02.14; N, 13.14; S, 15.11%.

5-(4-Methoxybenzylidene)-2-{[5-(3-fluorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7l)

Yellow solid; yield (.050 g, 24.25%) with mp 267–270 °C. IR (KBr, cm−1): 3448 (N-H str.), 3100 (Ar-H), 1702 (C=O st., lactam), 1587 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.84 (s, 3H, CH3), 7.16–7.80 (m, 9H, Ar-H, =CH), 12.96 (s, 1H, NH of lactam). Anal. Calcd. for C18H13FN4O2S2: C, 55.33; H, 03.18; N, 13.59; S, 15.55%; Found: C, 55.02; H, 02.95; N, 13.83; S, 15.23%.

5-{[4-(Benzyloxy)-3-methoxyphenyl]methylidene}-2-{[5-(2-chlorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7m)

Yellow solid; yield (0.420 g, 52.33%) with mp 250–252 °C. IR (KBr, cm−1): 3450 (N-H str.), 1688 (C=O st., lactam), 1577 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.85 (s, 3H, CH3), 5.19 (s, 2H, CH2), 7.26–8.20 (m, 12H, Ar-H), 8.2 (d, 1H, =CH), 12.98 (s, 1H, NH of lactam). Anal. Calcd. for C26H19ClN4O3S2: C, 58.37; H, 03.58; N, 10.47; S, 11.99%; Found: C, 58.68; H, 03.77; N, 10.19; S, 12.28%.

5-(4-Dimethylaminobenzylidene)-2-{[5-(2-chlorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7n)

Orange solid; yield (0.076 g, 22.98%) with mp 291–294 °C. IR (KBr, cm−1): 3450 (N-H str.), 1697 (C=O st., lactam), 1577 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.04 (s, 6H, 2 CH3), 8.10 (d, 1H, =CH), 6.89–7.73 (8H, Ar-H), 12.82 (s, 1H, NH of lactam). 13C NMR (100 MHz, Pyridine) δ (ppm): 33.17 (2 CH3), 106.02, 114.73, 116.54, 117.58, 121.37, 123.7, 124.55, 125.32, 125.92(Ar-C), 145.4 (thiazolidinone C5), 126.52 (=CH-Ar), 129.6, 129.35 (thiadiazole C5, thiazolidinone C2), 128.28 (thiadiazole C2), 153.91 (thiazolidinone C4). Anal. Calcd. for C20H16ClN5OS2: C, 54.35; H, 03.65; N, 15.85; S, 14.51%; Found: C, 54.59; H, 03.87; N, 16.10; S, 14.81%.

5-(3-Bromobenzylidene)-2-{[5-(2-chlorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7o)

Yellow solid; yield (0.108 g, 30.14%) with mp 280–283 °C. IR (KBr, cm−1): 3449 (N-H str.), 1704 (C=O st., lactam), 1594 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.57–7.91 (m, 8H, Ar-H), 8.20 (d, 1H, =CH), 13.15 (s, 1H, NH of lactam); Anal. Calcd. for C18H10BrClN4OS2: C, 45.25; H, 2.11; N, 11.73; S, 13.42%; Found: C, 45.51; H, 1.84; N, 11.44; S, 13.17%.

5-(3-Chlorobenzylidene)-2-{[5-(2-chlorophenyl)-1,3,4-thiadiazol-2-yl]-imino}-1,3-thiazolidin-4-one (7p)

Yellow solid; yield (0.107 g, 32.92%) with mp 270–271 °C. IR (KBr, cm−1): 3449 (N-H str.), 1709 (C=O st., lactam), 1592 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.57–7.81 (m, 8H, Ar-H), 8.20 (d, 1H, =CH), 13.14 (s, 1H, NH of lactam). Anal. Calcd. for C18H10Cl2N4OS2: C, 49.89; H, 02.33; N, 12.93; S, 14.79%; Found: C, 50.21; H, 2.11; N, 12.75%; S, 15.10%.

5-(3-Nitrobenzylidene)-2-{[5-(2-chlorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7q)

Yellow solid; yield (0.165 g, 49.56%) with mp 272–274 °C. IR (KBr, cm−1): 3419 (N-H str.), 3121 (Ar-H), 1726 (C=O st., lactam), 1596 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 7.54–8.32 (m, 8H, Ar-H), 8.52 (s,1H, =CH), 13.19 (s, 1H, NH of lactam). 13C NMR (100 MHz, DMSO) δ (ppm): 125.04, 125.15, 127.44, 128.47, 128.98, 131.05, 131.19, 131.46, 131.71, 132.79, 135.26, 136.00, (Ar-C), 148.73 (thiazolidinone C5), 160.72 (=CH-Ar), 158.87 (thiadiazole C5), 158.80 (thiazolidinone C2), 167.18 (thiadiazole C2), 171.36 (thiazolidinone C4). Anal. Calcd. for C18H10ClN5O3S2: C, 48.70; H, 02.28; N, 15.78; S, 14.45%; Found: C, 49.01; H, 02.56; N, 15.49; S, 14.72%.

5-(4-Methoxybenzylidene)-2-{[5-(2-chlorophenyl)-1,3,4-thiadiazol-2-yl]imino}-1,3-thiazolidin-4-one (7r)

Yellow solid; yield (0.133 g, 41.36%) with mp 279–281 °C. IR (KBr, cm−1): 3422 (N-H str.), 3097 (Ar-H), 1701 (C=O st., lactam), 1587 (C=N str.). 1H NMR (400 MHz, DMSO-d6) δ (ppm): 3.84 (s, 3H, CH3), 7.17–7.78 (m, 8H, Ar-H), 8.10 (d, 1H, =CH), 12.98 (s, 1H, NH of lactam). Anal. Calcd. for C19H13ClN4O2S2: C, 53.21; H, 03.06; N, 13.07; S, 14.95%; Found: C, 53.52; H, 02.85; N, 13.32; S, 15.23%.

Biology

In vitro HCV NS5B inhibition assay

Scintillation proximity assay (SPA) was used to evaluate the HCV NS5B inhibition using SPA beads. The beads emit an amount of light that is detected using a top count plate reader (Packard Top Count NXT scintillation counter, PerkinElmer). The detailed assay protocol was carried out as described (Page 31 and Supplementary Fig. S61–S78, Supplementary Material), and as reported [35, 36].

Antiviral activity

Huh-7.5 cells (ATCC code PTA-8561) were cultured as reported [36], the detailed procedure was carried out as described (Page 40, Supplementary Material). The infected Huh-7.5 cells were cultured and then serial concentrations of the tested compounds were added after incubation. The level of HCV RNA in each sample was assessed by quantitative real-time PCR on Rotor-Gene Cycler (Qiagen). The test was carried out using a commercially available kit supplied by “DNA-Technology, Research & Production” LLC (Moscow, Russia) according to the manufacturer’s instructions. The primers used for HCV and GAPDH internal control were previously reported [37]. The percent inhibition was calculated for each sample (Supplementary Fig. S79, Supplementary Material) as previously described [36].

Cytotoxicity study

Huh-7.5 cells were cultured, and cytotoxicity of the tested compounds was measured using MTT kit (Sigma-Aldrich) as described by the manufacturer. The detailed procedure was carried out as described (Page 42, Supplementary Fig. S80–S83, Supplementary Material) [35].

Computational studies

Homology modeling

The sequence for HCV NS5B domain, which consists of 591 amino acids, was retrieved from the UniProtKB database [25] with feature key: (ChainiPRO_0000045543) and entry number (O39929) for HCV GT4a. The sequence was used to build a model using SWISS-MODEL [27] online server. The FASTA sequence was used to identify the templates with high identity using BLAST server [38], then selection of the highest identity template with reliable coverage and GMQE to be used in the second step, which is model building. After the model was built, the model is subjected to protein refinement using the ModRefiner server [26].

Molecular docking

A set of compounds was sketched using MarvinSketch 5.11.5 [39]. OMEGA-3.1.0.3 [40] was used to generate 3D conformations of compounds using torsion driving and distance geometry algorithms. The output library file of OMEGA-3.1.0.3 [40] was ready for docking using OEDocking 3.3.0.3 [41] software against the HCV NS5B GT4a homology model with box dimensions 33.67 Å × 35.00 Å × 25.33 Å, the software set to generate five poses for each compound.

Discovery Studio 2019 Client software [42] was used for 2D and 3D visualization of the docking poses and the interactions between the target and the docked compounds. HCV NS5B protein binding site is shown in wireframe presentation. The atoms color scheme is that gray represents carbon, white represents hydrogen, blue represents nitrogen, red represents oxygen, yellow represents sulfur, and H bonds are represented as dotted green lines.