Searching Anti-Zika Virus Activity in 1H-1,2,3-Triazole Based Compounds

Abstract

Zika virus (ZIKV) is a mosquito-borne virus belonging to the Flaviviridae family and is responsible for an exanthematous disease and severe neurological manifestations, such as microcephaly and Guillain-Barré syndrome. ZIKV has a single strand positive-sense RNA genome that is translated into structural and non-structural (NS) proteins. Although it has become endemic in most parts of the tropical world, Zika still does not have a specific treatment. Thus, in this work we evaluate the cytotoxicity and antiviral activities of 14 hybrid compounds formed by 1H-1,2,3-triazole, naphthoquinone and phthalimide groups. Most compounds showed low cytotoxicity to epithelial cells, specially the 3b compound. After screening with all compounds, 4b was the most active against ZIKV in the post-infection test, obtaining a 50% inhibition concentration (IC₅₀) of 146.0 µM and SI of 2.3. There were no significant results for the pre-treatment test. According to the molecular docking compound, 4b was suggested with significant binding affinity for the NS5 RdRp protein target, which was further corroborated by molecular dynamic simulation studies.

1. Introduction

Zika virus (ZIKV) is a Flavivirus that belongs to the Flaviviridae family. This epidemic arbovirus is transmitted mainly by the bites of infected mosquitos from the genus Aedes and, possibly, Culex [1]. The virus has icosahedral symmetry and its genome consists of a single strand of positive-sense RNA, which is translated into structural (C, prM, E) and non-structural (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5) proteins [2]. ZIKV was first identified in 1947 when scientists discovered the then unknown virus from a febrile rhesus monkey used in yellow fever research in the Zika forest, Uganda [3]. After 60 years without major outbreaks, ZIKV was responsible for an epidemic in the Yap Islands in 2007 [4] and then in French Polynesia in 2012–2013 [5]. At the beginning of 2015, the first cases in Brazil were reported and the virus caused a serious epidemic resulting in over 200 thousand cases [6]. Several attributes contributed to ZIKV becoming endemic in Brazil, including the abundance of the mosquito vector, the deficiency of basic sanitation, inadequate housing conditions and the deficient public health policies to combat the virus [7].

The most common symptoms developed by patients infected with ZIKV are skin rash, low-grade fever, arthralgia, conjunctivitis, and headache. However, neurological disease of varying degrees of severity have also been reported, including some lethal cases [8,9]. One important neurological disorder associated with ZIKV infection is the Guillain-Barré syndrome, which affects mainly adults and is characterized by the destruction of the myelin sheath of neuronal cells [8]. Importantly, cases of microcephaly and other congenial defects, referred to as congenital Zika syndrome (CZS), have been categorically attributed to ZIKV infection [10]. Despite the severity of associated diseases and its endemicity in tropical areas of the globe, there are neither vaccines nor specific treatment to halt virus replication and alleviate patients’ symptoms. Thus, the development of effective antivirals and vaccines alongside mosquito control strategies are extremely important for combating this important pathogen and preventing new epidemics.

Given their structural diversity, a wide range of biological activities have been explored for the development of therapies. These properties make it possible for these molecules to interact with protein sites and trigger actions at a cellular level. Further structural modification of these compounds may improve the performance of a specific biological activity [11]. Recent studies reported the action of some heterocycles against the Zika virus such as the already known chloroquine [12], sofosbuvir [13], ribavirin, and faviparivir [14], as well as new chemical entities, for example pyrrolo[2,3-d]pyrimidines [15]. A tremendous effort has been and is being invested in the development of effective molecules against ZIKV; however, there no approved anti-ZIKV drugs [16].

Nucleoside analogs are another class of antivirals that have been tested extensively against ZIKV due to their proven role in preventing the replication of RNA viruses. For instance, our research group has identified the thiopurine nucleoside analogue 6-methylmercaptopurine riboside (6MMPr) as a potent inhibitor of ZIKV replication in cell culture [17]. To rationalize our strategy, we consider that phthalimide group is known for its anti-inflammatory activity, an important property in infection events. 1,4-Naphthoquinones have relevant cytotoxic activities. These two parts can be conjugated via click chemistry from organic azides and terminal alkynes generating a variety of 1H-1,2,3-triazole compounds (Figure 1).

These three classes of compounds, namely 1,4-naphthoquinone, phthalimide and 1,2,3-triazole inspired our research. These moieties have biological activities against several infections [18,19,20], including ZIKV [21,22]. Our group has synthesized several heterocycle derivatives and tested their activities against human cancer cell lines [23,24,25]. The compounds 1–7 (see Figure 2), that contain these cited heterocycles, and will be evaluated, were previously synthesized by our research group, except for compound 5a [25,26,27]. Here, we expand our initial efforts and evaluate their cytotoxicity to epithelial cells and their antiviral activity against ZIKV.

2. Results

2.1. Synthesis of the Compounds 1–7

Compounds 1–7 (Figure 2 and Figure 3,

Table 1. Synthesis of 1H-1,2,3-triazole derivatives 1–7.

Entry	Products	Methods	Yields, %	Reference
1	1	A	92	[26]
2	2a,b	B	85–95	[25]
3	3a,b	C	70–95	[26]
4	4a,b	D	70–88	[27]
5	5a	E	90	New compound
	5b	E	86	[26]
6	6a–d	F	62–85	[25]
7	7	D	85	[27]

Method-A: DMF/CuI//Et₃N/ultrasound/30 min.; Method-B: CH₃CN/CuI/0 °C/12 h/absence of light; Method-C: CH₃CN/CuI/r.t./20 h; Method-D: DMF/CuI/Et₃N/r.t./60 min; Method-E: DMSO/CuI/Et₃N/r.t./2 h; F: CH₃CN/CuI/0 °C/12 h.

) were prepared according to the methodologies described previously by our research group [25,26,27]. Only the compound, 5a, is a new compound. All compounds were characterized through infrared (IR) and nuclear magnetic resonance (NMR) analyses and yields between 62–95% were obtained. For a better discussion of the results, the compounds were classified according to triazole-linked groups.

2.2. Cytotoxicity of Triazole Derivatives in Vero Cells

The cytotoxicity of compounds 1–7 to Vero cells, an epithelial cell line derived from the kidney of an African green monkey, was tested using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) method. Statistical analysis found 20% cytotoxicity concentration (CC₂₀) values between 8.87–527.20 µM and 50% cytotoxicity concentration (CC₅₀) values from 38.01 to 1189 µM for the tested compounds (

Table 2. Cytotoxicity assay in Vero cells (CC20 and CC50).

Compound	CC20 (µM) 1	CC50 (µM) 2	logP 3
1	177.10	680.10	1.44
2a	8.87	38.01	0.99
2b	9.89	38.81	1.26
3a	55.28	469.80	2.21
3b	527.20	1189.00	2.22
4a	40.69	68.78	3.23
4b	136.90	330.20	2.94
5a	74.85	293.90	1.05
5b	66.93	298.20	1.38
6a	39.92	68.05	2.07
6b	21.18	48.69	1.86
6c	25.73	58.14	2.67
6d	38.61	69.50	2.34
7	119.10	334.50	1.80
6MMPr 4	60.5	291	0.55

¹ CC₂₀ (20% cytotoxic concentration) refers to compound concentration that causes a 20% reduction in cell viability. ² CC₅₀ (50% cytotoxic concentration) refers to compound concentration that causes a 50% reduction in cell viability. ³ The logP (the logarithm of the octanol-water partition coefficient): was calculated using the Molinspiration molecular property (http://www.molinspiration.com).⁴ See reference [17].

). Based on CC₂₀ values, which was the maximum concentration used for antiviral screening, 3b compound was the least toxic with a value of 527.2 µM. Compounds 2a and 2b were the most cytotoxic with values of 8.87 µM and 9.89 µM, respectively. Toxic effects of triazole derivates showed dependence with structural modification. Treatment of cells with compounds linked to naphthoquinone 5a,b and 6a–d resulted in low cell viability, whose CC₂₀ ranged between 21.18–74.85 µM. The results from compounds 3a, 3b and 4a, 4b ranged from 40.69 to 527.20 µM. Compounds 1 (phthalimide-triazole-alcohol) and 7 (sugar-triazole-amino-naphthoquinone) showed median results of 177.10 µM and 119.10 µM, respectively. These results show that the presence of only naphthoquinone in the molecule reduces the values of cytotoxic concentration. The presence of phthalimide group or sugar increases the values of CC₂₀ and CC₅₀ (Table 2 and Supplementary Material: Figure S1).

The cytotoxicity of phthalimide-amino-naphthoquinone-triazoles 3a, 3b and 4a, 4b showed dependence with the size of the alkyl chain and presence/absence of the bromine at the naphthoquinone moiety.

For instance, the compounds 3a (short alkyl chain, CC₂₀ = 55.28 µM) and 4a (long alkyl chain and bromine, CC₂₀ = 40.69 µM) were most cytotoxic than compound 3b (long alkyl chain, CC₂₀ = 527.20 µM) and 4b (short alkyl chain and bromine, CC₂₀ = 136.90 µM). With these data, we observed that the size of the chain and the presence of the bromine atom are decisive for the CC₂₀ results. Perhaps the access of the compounds to the cell change accord to these structural differences and chemical properties, as with hydrogen-bond interaction.

In compound 7, the phthalimide group was exchanged for glucose from compound 4b and no major changes were found in CC₂₀ (119.00 µM).

In general, the presence of phthalimide and/or amino-naphthoquinone groups made the molecules less cytotoxic, to mention some comparisons: (i) 1 (177.10 µM) and 2a, 2b (8.87–9.89 µM): 2-triazole-naphthoquinone is more toxic than phthalimide; (ii) 2a, 2b and 6a–d (21.18–39.92 µM), and 5a, 5b (74.85–66.93 µM): 2-amino-alkyl-naphthoquinone is less toxic than 2-triazole-naphthoquinone; (iii) 5a, 5b with 3b,4b (527.20–136.90 µM): the presence of phthalimide group decrease the cytotoxicity.

2.3. Screening of Triazole Derivates

The antiviral activities of the compounds were evaluated by infecting Vero cells with ZIKV and then treating them with the testing molecules. Virus titer in the supernatants of treated wells and its comparison with the negative (cell + virus) and the positive 6-methylmercaptopurine riboside (6MMPr) controls allowed for the determination of the antiviral effect. Only compound 4b was found to result in a statistically significant reduction in viral titer (Figure 4).

According to the post-infection test performed, compound 4b proved to be effective in inhibiting the virus at the lowest concentration used. The % viral inhibition showed in

Table 3. % Viral Inhibition.

Compound	4b			6MMPr
Concentration (µM)	3 d.p.i. 1 %	5 d.p.i %	Concentration (µM)	3 d.p.i. %	5 d.p.i. %
CC20 = 136.90	94.9	97.5	60.50	99.8	99.7
CC20/2 = 68.45	85.6	93.7	30.25	54.4	96.8
CC20/4 = 34.23	85.8	90.8	15.13	42.7	95.1
CC20/8 = 17.11	88.8	96.8	7.6	−93.4 *	81.1

¹ d.p.i.: days post-infection. * The negative sign refers to increased virus titer compared to control.

, demonstrates that at 17.11 µM (96.8%) 4b was most effective than 6MMPr at 7.56 µM (81.1%). Then, using less compound 4b yielded significant results in inhibiting ZIKV.

2.4. Pre-Treatment with Compound 4b

Since compound 4b was the only one in the series with activity against ZIKV, further experiments were conducted to elucidate its mechanism of action in cells that were pre-treated. To this, cells were pre-treated with varying concentrations of compound 4b, washed, and then infected with ZIKV. The concentrations used were chosen with the aim to preserve cellular viability and, thus, the maximum concentration was the previously determined CC₂₀ value (136.90, 68.45, 34.22 and 17.11 µM). There was no significant change in virus titers when compared with negative control (cells + virus, untreated) (Figure 5), suggesting that compound 4b cannot prevent the entry of ZIKV into Vero cells.

2.5. Post-Infection Studies of Compound 4b

We then investigated the activity of compound 4b after ZIKV infection in cells. As already noted in the screening test, 4b reduced ZIKV growth to levels similar to 6MMPr, a positive control previously described by our group [17]. As in the pre-treatment test, four different concentrations were used, with the maximum being CC₂₀ (136.90, 68.45, 34.22 and 17.11 µM). According to the statistical analysis (Figure 5), 4b inhibits the growth of the virus at all concentrations used, especially in the highest one. It is worth highlighting the constant decrease in the viral titer, even in lower concentrations, while the positive control 6MMPr reduces it only in the highest concentrations. The 4b half maximal inhibitory concentration (IC₅₀) and selectivity index (SI) was 146.0 µM and 2.3, respectively (

Table 4. Results for post-infection test with 6MMPr and 4b.

	IC50 1 (µM)	SI 2	% Viral Inhibition
4b	146.0	2.3	91.1
6MMPr	24.5	11.9	97.8

¹ IC50 (50% inhibition concentration) refers to compound concentration that caused a 50% inhibition in virus replication. ² SI (selectivity index) refers to the ratio between the CC₅₀ and IC₅₀ values.

).

The treatment time of the compounds for three days post-infection (d.p.i.) and five days post-infection (Figure 6) was also assessed; 6MMPr prevents the growth of the virus in both cases, but with a noticeable decrease if the treatment is longer. Compound 4b continues to behave similarly in both situations, decreasing the viral titer at all concentrations and at both times of treatment, suggesting that its antiviral activity lasts longer than the reference drug used in our tests (see Table S1). Statistical analysis compared the concentrations of each compound at each collection time (3 d.p.i. and 5 d.p.i.) with a t-test.

2.6. Molecular Docking

After discarding the hypothesis that 4b acts via a viral entry process and confirming instead that the viral inhibition of compound 4b occurs in some stage of Zika replication, an investigation of the possible binding site was conducted. The compound, 6MMPr, used as a reference drug, is a riboside analogue, so it was speculated its action would interfere with the RNA processing machinery of the virus. The two viral proteins that bind to RNA are NSP3 helicase and the NS5 RdRp proteins; PDB ID: 6MH3 and PDB ID: 5U04, respectively, were selected to predict the potential target of hit compound 4b through extensive in-silico studies.

The RdRp protein has several missing residues on many locations as depicted in Table S1 (Supporting Information). These missing residues were filled in the form on loops by Modeller. Among the 9 models, a top model (model 9) was selected based on the ERRAT score, PROSA, Verify 3D, and Procheck (Table S2, entry 9). The ERRAT score, PROSA, and Verify 3D score of selected models were measured as −70.5882, −7.69, and 84.95, respectively. The model 4 with an ERRAT score of 70.7612 was excluded because of poor <80% Verify 3D score (Table S2, entry 4). The selected model has only 1.1% residues in a disallowed region (Table S2, entry 9), which included only 6 residues, Asp346, Ser406, Glu415, Ala537, Arg601, and Lys688, as depicted in Figure S2. Compound 4b in complex with the helicase (6MH3) protein showed docking score, XP Gscore, and binding free energy values of −3.481 kcal/mol, −3.481 kcal/mol and −47.11 kcal/mol, respectively. On the other hand, compound 4b complexed with the RdRp (5U04) protein displayed a docking score, XP Gscore, and binding free energy of −5.348 kcal/mol, −5.348 kcal/mol and −52.91 kcal/mol, respectively. In 4b-helicase complex, compound 4b interacted to the binding site residues through H-bond (Asn463, Arg462), halogen bond (Trp403, Lys419), salt bridge interaction (Lys358), and pi-cation interaction (Lys419), as represented in Figure 7A. Likewise, in 4b-RdRp complex, compound 4b revealed interactions through H-bond (Asp665, Ile799, Arg731), salt bridge interaction (Arg739), halogen bond (Arg731), and pi-pi interaction (Tyr609, Trp797), which are presented in Figure 7B. Molecular docking results suggested that compound 4b possesses strong binding affinity towards RdRp protein over the helicase.

Next, MD simulation for 100 ns was performed to analyse the stability of the compound 4b within the binding pocket of both proteins, helicase and RdRp. In 4b-helicase complex, protein Cα-RMSD reached stability within 10 ns and the results are showcased in Figure 7. The RMSD of compound 4b achieved stability nearby 28 ns and sustained up to 70 ns. Afterwards, RMSD of 4b became unstable and escaped from the binding pocket as indicated by the remarkably high RMSD and trajectory analysis (Figure 8). The ligand RMSF showed that its each atom fluctuated more than 6Å, as presented in Figure S3a. The interaction with water may be responsible for the instability of compound 4b, as displayed in Figure S3b.

On other hand, protein Cα-RMSD of 4b-RdRp complex, stabilized within 10 ns and remained stable throughout the simulation period (100 ns) as presented in Figure 8. Compound 4b RMSD suggested the structural and conformational stability, where fluctuation was observed within an acceptable range, 3 Å (Figure 9A). Each atom of the compound 4b fluctuated within 3.5 Å (Figure 9B). Compound 4b showed interactions with residues Tyr609, Trp797, Arg731, and Ile799 and thus supported docking results (Figure 9C). The Ramachandran mapping of compound 4b-RdRp complex after MD simulation showed only 0.6% (Asn454, Glu695, and Trp748) residues in the disallowed region for the protein RdRp, which indicated good stereo-chemical geometry of the residues, as presented in Figure S4.

3. Discussion

The search for effective treatments to combat symptoms and protect against ZIKV remains in progress, since there are no options available in the market. This work explored the potential of synthetic compounds in the search of viable treatment options against this devasting virus.

According to the structural analysis of the compounds, some aspects related to the cytotoxic effect on Vero cells were highlighted. The compounds 2a and 2b (naphthoquinone-triazole-alcohol), whose cell viability results were more cytotoxic based on their CC₂₀ value, presents the 1,4-naphthoquinone directly linked with 1H-1,2,3-triazole nucleus; on the other hand, in general, 2-amino-1,4-naphthoquinone decrease the toxicity level [27]. We believe that some conformational factor hydrophobic effects are present. Michael acceptor group (3a/3b), and “halogen bond”-bromide- (4a/4b) can help with protein-ligand complex interactions. All these factors can promote different results in cytotoxicity. The 4a cytotoxicity is likely due to this compound being more lipophilic (logP = 3.23, Table 2) and may accumulate in the cell membrane with toxic consequences. This effect is shown in CC50 (4a = 68.78 µM). The phthalimide group, represented by 4-phenyl-1-[2-(phthalimido-2-yl)ethyl]-1-H-1,2,3-triazole, showed low cytotoxicity for macrophages (peritoneal and J774A.1) and fibroblast cells [28].

Naphthoquinone-triazoles 6a–d linked to carbohydrates decreased the cytotoxicity, i.e., at 50 µM they maintained cell viability around 80%, except for 6b. The amino-naphthoquinone-triazoles 5a and 5b behaved contrarily compared to 2a, 2b and 6a–d, being less cytotoxic, with cells remaining viable until the concentration of 400 µM.

Naphthoquinone compounds have a variety of activities described in the literature, including cytotoxicity in Vero cells. Gonzaga et al. (2019) used the MTT method to test the cytotoxicity of compounds derived from bis-naphthoquinones [21]. In general, the presence of the group, even in duplicate, does not increase cytotoxicity to Vero cells.

The screening reveals little or no action of the compounds 1–7 against ZIKV, except for 4b, which was the only one with statistically significant results when compared to the non-treated control. Pre-treatment and post-infection tests were performed with 4b to elucidate the possible mechanism of action and determine the IC₅₀ and SI of values, 146.0 µM and 2.3, respectively. The results indicate that the compound 4b probably acts in some stage of the virus replication after it has entered the cell. Overall, compound 4b exhibited significant anti-ZIKV activity without any apparent cytotoxicity.

Lima et al. (2019) tested triazole compounds against ZIKV with Vero cells. The method used was MTT in both analyses, thus verifying the viability of the cells in the tests. Lima’s triazole series showed high cytotoxicity. The antiviral activity were the best ZIKV inhibitors, considering at least 50% of cell viability [22].

The analysis with the compound-based 1H-1,2,3-triazoles strengthens the principle that the presence of nitrogen groups increases the biological activity. The structural modification of the tested groups driven by in-silico techniques helped us to understand the negative response of the tested compounds against ZIKV, as well as the improvement of the 4b response in viral inhibition.

Concerning the docking score, XP Gscore and binding free energy, compound 4b was suggested with a significant binding affinity towards RdRp protein when compared with helicase. These results were further corroborated by MD simulation studies. However, validation experiments are necessary before concluding the possible target for the hit compound 4b.

4. Materials and Methods

4.1. Chemistry

Drugs: the compounds 1–7 (Figure 1) were synthesized by our research group. The formation of the triazole derivatives was carried out by copper(I)-catalysed azide-alkyne cycloaddition (CuAAC) through the reaction between phthalimide-azide or naphthoquinone-azide and 2,3-unsaturated glycosides or alkynyl alcohols. The synthesis of 2,3-unsaturated glycosides 6a–d was carried out using the montmorillonite K-10/iron(III) chloride hexahydrate catalysis described by Melo et al. (2015) [25].

The new compound 5a was synthesized using Method E (Table 1) [26]. The ¹H and ¹³C NMR are showed in the Figures S5 and S6. Characterization of 2-[2-(4-(3-Hydroxypropyl)-1H-1,2,3-triazol-1-yl)]-ethylamino-1,4-naphthoquinone (5a): Yield = 90%; red solid; Mp 143–145 °C. ¹H NMR (400 MHz, DMSO-d₆): δ 7.98–7.92 (m, 2H, H_Naph), 7.87 (s, 1H, H_Triaz), 7.84–7.72 (m, 2H, H_Naph), 7.49 (br s, 1H, NH), 5.70 (s, 1H, H-3_Naph), 4.54 (t, 2H, J = 6.0 Hz, CH₂N_Triaz), 4.39 (t, 1H, J = 5.0 Hz, OH), 3.66–3.64 (m, 2H, CH₂NH), 3.40 (m, 2H, CH₂OH), 2.60 (m, 2H, CH₂-C4_Triaz), 1.70 (m, 2H, CH₂). ¹³C NMR (100 MHz, DMSO-d₆): δ 202.2, 182.1, 148.7, 145.0, 135.3, 134.8, 132.7, 130.8, 126.3, 125.8, 122.8, 100.5, 60.5, 47.7, 42.4, 32.7, 22.1. Anal. calcd. for C₁₇H₁₈N₄O₃: C, 62.57; H, 5.56; found: C, 62.71; H, 5.49.

After molecular structure characterization by infrared (IR), and ¹H and ¹³C NMR, the compounds were tested for ZIKV infection. The stock solutions were made at a concentration of 200 mM with DMSO (Sigma-Aldric, San Luis, MO, USA). The compounds dilutions were made from the stock solution with Dulbecco’s Modified Eagle Medium (DMEM) (Thermo Fisher Scientific, Waltham, MA, USA) reaching the maximum concentration of 0.8% DMSO.

A stock solution of the riboside nucleotide analog 6MMPr (Sigma-Aldrich, Louis, MO, USA) of 1 mM was made in Milli-Q water and the compound was used as a positive control in antiviral tests. Dilutions of 60.50, 30.25, 15.13 and 7.56 µM were used according to the cytotoxicity methodology and CC₂₀ values used by de Carvalho and co-workers [17].

4.2. Cells and Viruses

Vero cells (CCL-81) were used in all phases of the in vitro testing. The cultivation was performed with products from the company, Thermo Fisher Scientific. DMEM supplemented with 10% foetal bovine serum (FBS) and 1% penicillin/streptomycin and incubated at 37 °C and 5% CO₂ atmosphere on incubator Heracell VIOS 250i (Thermo Fisher Scientific).

A strain of ZIKV/H.sapiens/Brazil/PE243/2015 (abbreviated to ZIKV PE243; GenBank accession number KX197192.1) was used in antiviral tests. The viral stock was made in Vero cells and grown at 37 °C and 5% CO₂ atmosphere. The endpoint dilution assay was the method used to obtain the TCID₅₀/mL titer of this stock.

4.3. Titration of Stock and Samples Virus

96-well plates with 1 × 10⁴ cell/well were used in this assay and prepared 24 h previously. Seven serial dilutions of the virus were prepared in DMEM + 2% FBS culture medium. After removing the culture medium from the plate, 50 µL of each dilution was added to the wells in several repetitions. The plate was incubated at 37 °C and 5% atm CO₂ for 1 h. After this interval, each well of the plate received an additional 100 µL of DMEM + 2% FBS culture medium. The plate was incubated again, in the same parameters, for 5 days. Some columns of the plate were selected for use as a cell control (culture medium only, no virus). The reading was performed in an inverted optical microscope AE2000, Motic (Motic, Kowloon Bay, Hong Kong) counting positives wells by the presence of cytopathic effect and the determination of the sample title was performed with the aid of the Microsoft Office Excel program (Microsoft^® Office, Redmond, WA, USA).

4.4. Cytotoxicity Test with MTT Method

96-well plates with 1 × 10⁴ cell/well were used in this assay and prepared 24 h previously. The compounds were diluted in a 2:1 ratio so that a maximum of 0.8% DMSO was used. The concentrations used were 800, 400, 200, 100, 50, 25 and 12.5 µM. Dilutions were added to the wells in triplicate and incubated at 37 °C and 5% CO₂ atmosphere. After 5 days, the wells were emptied and a 50 µL of 1 mg/mL MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (Sigma-Aldrich) solution was added, and the cells were incubated for 3–4 h for formation of formazan crystals through the mitochondria of viable cells. Then, medium was removed and 100 µL DMSO was added to the wells to solubilize the formazan crystals. With the aid of an incubator Tecnal TE-420 (Tecnal Equipamentos Científicos, Piracicaba, São Paulo, Brazil), the plates were homogenized at 120 RPM at 37 °C and the optical densities were read at 570 nm on a spectrophotometer BioTekTM ELx800TM (BioTek Instruments, Winooski, VT, USA). The statistical analysis of these data revealed the values of CC₂₀ and CC₅₀, both in µM, which was used in the screening test.

4.5. Screening Test

The screening of the compounds was carried out with the CC₂₀ value (µM) to determine the most viable compounds for more detailed antiviral tests. 48-well plates with 2.5 × 10⁴ cells/well were prepared with a previous 24 h. The culture medium from the plate was removed and the wells were inoculated with a viral solution at MOI 0.1, except cell control (which received only culture medium + 2% FBS). During 2 h of incubation at 37 °C and 5% CO₂ atmosphere, the solutions with the compounds are prepared with DMEM + 2% FBS. After incubation, the medium was discarded, and the wells were washed with 1X phosphate buffer solution (PBS). The solutions with the compounds were added to the wells, in triplicate, and then incubated for 5 days under the same conditions. After incubation, the supernatant from each well was collected and stored at −80 °C. The samples were then titrated and analysed statistically.

4.6. Pre-Treatment Activity

To prepare, 48-well plates with 2.5 × 10⁴ cells/well were prepared 24 h previously. Compound 4b was diluted in a 2:1 ratio in culture medium with concentrations of 136.90 µM (CC₂₀), 68.45 µM (CC₂₀/2) 34.22 µM (CC₂₀/4) and 17.11 µM (CC₂₀/8). The dilutions were inoculated into the wells in triplicate and the cells incubated at 37 °C and 5% CO₂ atmosphere. After 1 h, the culture medium with the compounds was removed and the wells were washed with 1X PBS. A ZIKV solution at MOI 0.1 was inoculated into the treated and control wells. The cells were again incubated for 5 days. The supernatants from the wells were collected and titrated.

4.7. Post-Infection Activity

48-well plates with 2.5 × 10⁴ cells/well were prepared 24 h previously. A ZIKV solution at MOI 0.1 was inoculated into the wells and the cells incubated at 37 °C and 5% CO₂ atmosphere. After 2 h, the culture medium was removed, and the wells were washed with 1X PBS. 2:1 dilution of 4b (136.90 µM, 68.45 µM, 34.22 µM and 17.11 µM) and 6MMPr (60.50 µM, 30.25 µM, 15.13 µM and 7.56 µM) were added in triplicate. The cells were again incubated for 72 h and 120 h. The supernatants from the wells were collected and titrated.

4.8. Molecular Docking

The coordinate structure of NSP3 (PDB: 6MH3) and NSP5 (PDB: 5U04) proteins were collected from RCSB website (https://www.rcsb.org, accessed on 10 February 2021). The NSP3 and NSP5 also known as helicase and RNA dependent RNA polymerase (RdRp) protein, respectively. The RdRp protein had several missing residues (Table S1), so missing loops were filled with the help of modeller software [29]. The best model was selected among 15 models based on ERRAT [30], PROSA [31], Verify3D [32], and PROCHECK [33] (Table S2). The computational work performed using Schrodinger software (Schrodinger release 2020-1 license dated 20 November 2020). Both protein structures were prepared to minimize the structural defects by using protein preparation wizard before docking studies [34,35,36]. Before docking compound 4b structure was also prepared by the Ligprep [35,37]. The Epik module of Schrodinger was used to predict the ionization states of compound 4b at pH 7 ± 2 as well as tautomers generated [38]. In-silico study was carried out under OPLS2005 forcefield.

4.8.1. Molecular Docking of Designed Chemical Library

The binding site of both the proteins was predicted by sitemap [39]. The best site was selected based on site score (Tables S3 and S4). The co-ordinate of binding site for RdRp protein includes 26.17, 66.84, and 103.42, while the helicase protein includes −5.54, 2.84, and −6.55. Compound 4b was docked at XP precision to both the proteins in a site-specific manner using Glide module of Schrödinger suite [40]. The Van der Waals radii scaling factor and partial charge cut-off was 0.8 and 0.15, employed for docking study, respectively. The binding free energy for both the complexes were also calculated by prime MMGBSA [41].

4.8.2. Molecular Dynamics (MD) Simulation

To validate the docking results, both the protein complexes of compound 4b with helicase and RdRp protein were selected for 50 ns MD simulation [42]. These complexes were solvated in TIP3P [43] water model and 0.15 M NaCl to mimic a physiological ionic concentration. The stereo-chemical geometry of 5HGL protein residues was measured by Ramachandran map by procheck [33].

4.9. Statistical Analysis

After reading the optical densities of the cytotoxicity test revealed by the MTT method, the calculation of the cell viability percentage was performed in the Microsoft Office Excel 15 (Microsoft^® Office) program using the following formula:

% Viability = OD _sample × 100/OD_{cellular control}

(1)

The determination of the CC₂₀ and CC₅₀ values was obtained through the linear regression analysis of the XY graph generated by the values of the concentrations used in the test and the triplicate values of the viability. In the post-infection and pre-treatment tests, ANOVA and Dunnett tests were used, considering p > 0.05 as the minimum significance value. All tests were performed in GraphPad Prism v.6.0 program (GraphPad Software, Inc., San Diego, CA, USA).

5. Conclusions

Heterocyclic compounds have several activities recorded in the literature. The triazole and naphthoquinone groups have already been reported with antiviral activity against the Zika virus, while the phthalimide group has no previously reported activity against the virus. Fourteen compounds were tested in this study. In the cytotoxicity test, the CC₂₀ (8.87–527.20 µM) and CC₅₀ (38.01–1189.00 µM) values were quite diverse, showing high to low values, but all of them were feasible for the antiviral screening. Compound 3b stood out with the highest values of CC₂₀ and CC₅₀. The screening revealed compound 4b as the most active against ZIKV, and the results confirm a post-infectious antiviral activity. Docking results coupled with score, XP Gscore and binding free energy suggested compound 4b with strong binding affinity for NS5 RdRp protein target of ZIKV, which was further supported by MD simulations. However, validation experiments are necessary before concluding the possible target for the hit compound 4b, which could be a part of our future investigations. Lead optimization, in conjunction with in-silico analysis, could deliver new versions of the molecule with significantly increased activity against ZIKV. In addition, in vivo tests will be important to confirm the efficacy of these promising molecules in adequate ZIKV animal models.