Computer-Aided Design, Synthesis, and Antiviral Evaluation of Novel Acrylamides as Potential Inhibitors of E3-E2-E1 Glycoproteins Complex from Chikungunya Virus

Chikungunya virus (CHIKV) causes an infectious disease characterized by inflammation and pain of the musculoskeletal tissues accompanied by swelling in the joints and cartilage damage. Currently, there are no licensed vaccines or chemotherapeutic agents to prevent or treat CHIKV infections. In this context, our research aimed to explore the potential in vitro anti-CHIKV activity of acrylamide derivatives. In silico methods were applied to 132 Michael’s acceptors toward the six most important biological targets from CHIKV. Subsequently, the ten most promising acrylamides were selected and synthesized. From the cytotoxicity MTT assay, we verified that LQM330, 334, and 336 demonstrate high cell viability at 40 µM. Moreover, these derivatives exhibited anti-CHIKV activities, highlighting the compound LQM334 which exhibited an inhibition value of 81%. Thus, docking simulations were performed to suggest a potential CHIKV-target for LQM334. It was observed that the LQM334 has a high affinity towards the E3-E2-E1 glycoproteins complex. Moreover, LQM334 reduced the percentage of CHIKV-positive cells from 74.07 to 0.88%, 48h post-treatment on intracellular flow cytometry staining. In conclusion, all virtual simulations corroborated with experimental results, and LQM334 could be used as a promising anti-CHIKV scaffold for designing new drugs in the future.


Introduction
The Chikungunya virus (CHIKV) is an arbovirus from the Alphavirus genus, which belongs to the Togaviridae family [1,2]. It is mainly transmitted by the bite of infected mosquitoes from Aedes aegypti and Ae. albopictus species [3][4][5][6][7][8]; although, Ae. furcifer and Culex spp. mosquitoes have been also reported as vectors [3,[9][10][11]. Since 1990, CHIKV infections have been reported in many countries from South and Central Americas, estimating 11,675 million cases [2,[12][13][14]. Recently, it was verified that In general, nsP2 inhibitors have been broadly explored by different research groups [51,[56][57][58][59]. It is a cysteine protease that corresponds to approximately 20-30% of the viral particle, and also has four binding sites in its surface [1,52]. However, the binding site number 4 is considered the most important, since it contains the catalytic dyad, composed of Cys 1013 and His 1083 amino acids [49]. This cysteine protease has the potential to act as a nucleophile protein since the Cys 1013 could be deprotonated by His 1083 , via an acid-base mechanism, at physiological pH [49,66]. Therefore, Michael's acceptors represent an interesting alternative for developing anti-CHIKV compounds [56,58]. Additionally, the Trp 1084 residue seems to perform an auxiliary role during the proteolytic activity of this protease [66]. As the global impacts of CHIKV and that the development of new antivirals is an unmet need, we aimed to develop new antiviral agents. The molecular docking of small flexible ligands toward macromolecules remains as the most broadly used in silico technique [67][68][69][70][71]. It is a stochastic method that uses scoring functions to find a minimum energy value, based on binding modes of ligands at the active site of macromolecules [72,73]. Essentially, it considers that hit compounds display high-affinity values forward their targets since they have favorable chemical groups for efficient interactions [69]. In this context, molecular docking was applied in this computerguided study. Initially, virtual screening was performed in a dataset containing 132 compounds (including acrylamides and acylhydrazones) toward nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins complex, and C proteins. Meaningful FitScore values for acrylamide derivatives toward mature E3-E2-E1 glycoproteins complex were obtained by using molecular docking, which were analyzed by heat maps. Finally, an acrylamide analog exhibited a remarkable antiviral activity inhibiting the CHIKV infection in vitro, corroborating with our in silico results, suggesting that this compound acts by interaction with mature E3-E2-E1 glycoproteins complex and blocks the CHIKV attachment. Figure 2 presents the workflow including all the steps followed to perform this rational study.  (E b ) value obtained by using Auto Dock Vina software; *: Fluorinated adenosine analog was reported as an RNA replication inhibitor, in which its antiviral activity was associated with an indirect effect on viral methyltransferase (MTase) activity through the inhibition of host S-adenosyl-L -homocysteine (SAH) hydrolase.
In general, nsP2 inhibitors have been broadly explored by different research groups [51,[56][57][58][59]. It is a cysteine protease that corresponds to approximately 20-30% of the viral particle, and also has four binding sites in its surface [1,52]. However, the binding site number 4 is considered the most important, since it contains the catalytic dyad, composed of Cys 1013 and His 1083 amino acids [49]. This cysteine protease has the potential to act as a nucleophile protein since the Cys 1013 could be deprotonated by His 1083 , via an acid-base mechanism, at physiological pH [49,66]. Therefore, Michael's acceptors represent an interesting alternative for developing anti-CHIKV compounds [56,58]. Additionally, the Trp 1084 residue seems to perform an auxiliary role during the proteolytic activity of this protease [66]. As the global impacts of CHIKV and that the development of new antivirals is an unmet need, we aimed to develop new antiviral agents. The molecular docking of small flexible ligands toward macromolecules remains as the most broadly used in silico technique [67][68][69][70][71]. It is a stochastic method that uses scoring functions to find a minimum energy value, based on binding modes of ligands at the active site of macromolecules [72,73]. Essentially, it considers that hit compounds display high-affinity values forward their targets since they have favorable chemical groups for efficient interactions [69]. In this context, molecular docking was applied in this computer-guided study. Initially, virtual screening was performed in a dataset containing 132 compounds (including acrylamides and acylhydrazones) toward nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins complex, and C proteins. Meaningful FitScore values for acrylamide derivatives toward mature E3-E2-E1 glycoproteins complex were obtained by using molecular docking, which were analyzed by heat maps. Finally, an acrylamide analog exhibited a remarkable antiviral activity inhibiting the CHIKV infection in vitro, corroborating with our in silico results, suggesting that this compound acts by interaction with mature E3-E2-E1 glycoproteins complex and blocks the CHIKV attachment. Figure 2 presents the workflow including all the steps followed to perform this rational study.

Computer-Aided Drug Design
After the analysis of a dataset composed of 132 derivatives (including acrylamides and acylhydrazones) by molecular docking towards nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins complex, E, and C proteins from the CHIKV, it was identified that acrylamide derivatives could be more actives than acylhydrazones. In general, the 10 most favorable binding modes for each compound were generated by molecular docking. Moreover, it was verified that all molecules analyzed in this step demonstrated high affinity (FitScore) values forward the E3-E2-E1 glycoproteins complex (PDB ID: 3N41). Then, this observation was confirmed by using heat maps, in which a "hot-zone" was verified, concentrated on this molecular target. Thereafter, it was revealed that all acrylamides present FitScore values about 5-10 points higher than acylhydrazones. Additionally to this fact, some works have suggested that acylhydrazones and hydrazones could be associated with pan-assay interference scaffolds (PAINS), providing unreliable results in biological tests [74,75]. Combining this information and our virtual screening results, the top 10 most promising (FitScore ≥ 50.0) acrylamide derivatives ( Figure 3) were selected for synthesis and biological evaluation in this study.

Computer-Aided Drug Design
After the analysis of a dataset composed of 132 derivatives (including acrylamides and acylhydrazones) by molecular docking towards nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins complex, E, and C proteins from the CHIKV, it was identified that acrylamide derivatives could be more actives than acylhydrazones. In general, the 10 most favorable binding modes for each compound were generated by molecular docking. Moreover, it was verified that all molecules analyzed in this step demonstrated high affinity (FitScore) values forward the E3-E2-E1 glycoproteins complex (PDB ID: 3N41). Then, this observation was confirmed by using heat maps, in which a "hot-zone" was verified, concentrated on this molecular target. Thereafter, it was revealed that all acrylamides present FitScore values about 5-10 points higher than acylhydrazones. Additionally to this fact, some works have suggested that acylhydrazones and hydrazones could be associated with pan-assay interference scaffolds (PAINS), providing unreliable results in biological tests [74,75]. Combining this information and our virtual screening results, the top 10 most promising (FitScore ≥ 50.0) acrylamide derivatives ( Figure 3) were selected for synthesis and biological evaluation in this study.
Pharmaceuticals 2020, 13, 141 5 of 23 ranging from 61.95 to 83.52% for carbon, from 3.96 to 5.68% for hydrogen, and from 4.67 to 4.94% for nitrogen. For all acrylamides previously synthesized, the corresponding references were provided in order to compare the results obtained in this study (see Materials and Methods section). Finally, all these physicochemical and spectroscopic techniques were sufficient to unequivocally characterize the compounds synthesized in this work. All chromatograms, FT-IR, 1 H and 13 C NMR spectra are available in the supplementary material related to this manuscript.

Cell Viability and Antiviral Assays
The cytotoxicity was performed in vitro for the ten synthesized acrylamides (LQM328 to LQM337) towards Vero E6 cells at 20 µM concentration by MTT assay [80]. As shown in Figure 4, only the LQM329 was highly cytotoxic, reducing the cell viability to less than 50% (41.5% ± 3.3) after 48h of culture, thus being removed for further analysis. Therefore, the screening of antiviral activity against CHIKV was performed for all other nine acrylamides.

Chemistry
All the chemical intermediates (3a-j) were obtained by the reaction between corresponding aldehydes (1) and malonic acid (2), via Doebner-Knoevenagel condensation ( Figure 3), with yields ranging from 45 to 94% [76,77]. Additionally, 1 H Nuclear Magnetic Resonance (NMR) spectra revealed that the intermediates (3a-j) were obtained in an (E)-configuration, confirmed by large vinylic coupling constant (J) values, ranging from 15.8 to 16.1 Hz [77,78]. Moreover, 1 H NMR analysis showed that the chemical shifts (δ) for the hydroxyl (OH) from the carboxylic acid can appear between 12.38 and 12.71 ppm. Subsequently, the acrylamide derivatives (LQM328-337) were obtained by the TBTU-coupling reaction (Figure 3), using diisopropylethylamine (DIPEA) as a catalyst base [79], with yields from 52 to 92%. For all these final compounds, the purity degree was determined by the HPLC technique, which resulted in purities ranging from 95.3 to 99.9%, with retention time (R T ) between 3.07 and 3.88 min. The analysis of Fourier-Transform Infrared (FT-IR) spectra revealed three characteristic stretches (v) from these chemical molecules, ranging from 3240 to 3356 cm −1 for v(N-H) bond; from 1658 to 1651 cm −1 for v(C=O) bond; also from 1620 to 1612 cm −1 and from 979 to 964 cm −1 v(C=C) ene bond. Additionally, melting points (Mp) are uncorrected and they range from 113 to 227 • C. In the 1 H NMR spectra of the acrylamides, it was observed that the amide (N-H) signal ranges from 10.13 to 10.35 ppm. Equally to their intermediates, acrylamides were obtained in (E)-configuration, also confirmed by the vinylic coupling constant values, J. From 13 C NMR spectra, the carbonyl (C=O) from the amide group has chemical shifts (δ) varying from 163.17 to 164.42 ppm. Additionally, elemental analyses (CHN) were only performed for the new acrylamides synthesized (LQM328, LQM331, and LQM337), in which their chemical compositions ranging from 61.95 to 83.52% for carbon, from 3.96 to 5.68% for hydrogen, and from 4.67 to 4.94% for nitrogen. For all acrylamides previously synthesized, the corresponding references were provided in order to compare the results obtained in this study (see Materials and Methods section). Finally, all these physicochemical and spectroscopic techniques were sufficient to unequivocally characterize the compounds synthesized in this work. All chromatograms, FT-IR, 1 H and 13 C NMR spectra are available in the Supplementary Material related to this manuscript.

Cell Viability and Antiviral Assays
The cytotoxicity was performed in vitro for the ten synthesized acrylamides (LQM328 to LQM337) towards Vero E6 cells at 20 µM concentration by MTT assay [80]. As shown in Figure 4, only the LQM329 was highly cytotoxic, reducing the cell viability to less than 50% (41.5% ± 3.3) after 48h of culture, thus being removed for further analysis. Therefore, the screening of antiviral activity against CHIKV was performed for all other nine acrylamides.

Cell Viability and Antiviral Assays
The cytotoxicity was performed in vitro for the ten synthesized acrylamides (LQM328 to LQM337) towards Vero E6 cells at 20 µM concentration by MTT assay [80]. As shown in Figure 4, only the LQM329 was highly cytotoxic, reducing the cell viability to less than 50% (41.5% ± 3.3) after 48h of culture, thus being removed for further analysis. Therefore, the screening of antiviral activity against CHIKV was performed for all other nine acrylamides. Initially, the in vitro anti-CHIKV activity for the acrylamides was evaluated at a 20 µM concentration. For this purpose, CHIKV adsorption was performed in the Vero E6 cells followed by the treatment with pre-selected compounds, and the cell viability was then assessed after 48h. As a result, significant viral inhibition was detected for the LQM328, LQM330, LQM334, LQM336, and LQM337 compounds ( Figure 5).  Initially, the in vitro anti-CHIKV activity for the acrylamides was evaluated at a 20 µM concentration. For this purpose, CHIKV adsorption was performed in the Vero E6 cells followed by the treatment with pre-selected compounds, and the cell viability was then assessed after 48h. As a result, significant viral inhibition was detected for the LQM328, LQM330, LQM334, LQM336, and LQM337 compounds ( Figure 5). To investigate the improvement in the antiviral activity of the compounds due to a higher concentration of the compounds, both in vitro cytotoxicity and anti-CHIKV assays were evaluated at 40 µM concentration, after 72h for the five most promising acrylamides (LQM328, LQM330, LQM334, LQM336, and LQM337). Although high cytotoxicity was detected for the LQM328 and LQM337, no toxicity was detected for LQM330, LQM334, and LQM336 at this concentration ( Figure 6A). Regarding anti-CHIKV activity, a significant viral inhibition was detected for LQM330, LQM334, and LQM336, with the highest antiviral activity detected for the LQM334 (viral inhibition = 81.1% ± 6.4 for LQM334 vs. 49.1% ± 11.1 for LQM330, and 32.2% ± 2.4 for LQM336), as shown in Figure 6B. To investigate the improvement in the antiviral activity of the compounds due to a higher concentration of the compounds, both in vitro cytotoxicity and anti-CHIKV assays were evaluated at 40 µM concentration, after 72h for the five most promising acrylamides (LQM328, LQM330, LQM334, LQM336, and LQM337). Although high cytotoxicity was detected for the LQM328 and LQM337, no toxicity was detected for LQM330, LQM334, and LQM336 at this concentration ( Figure 6A). Regarding anti-CHIKV activity, a significant viral inhibition was detected for LQM330, LQM334, and LQM336, with the highest antiviral activity detected for the LQM334 (viral inhibition = 81.1% ± 6.4 for LQM334 vs. 49.1% ± 11.1 for LQM330, and 32.2% ± 2.4 for LQM336), as shown in Figure 6B. 40 µM concentration, after 72h for the five most promising acrylamides (LQM328, LQM330, LQM334, LQM336, and LQM337). Although high cytotoxicity was detected for the LQM328 and LQM337, no toxicity was detected for LQM330, LQM334, and LQM336 at this concentration ( Figure 6A). Regarding anti-CHIKV activity, a significant viral inhibition was detected for LQM330, LQM334, and LQM336, with the highest antiviral activity detected for the LQM334 (viral inhibition = 81.1% ± 6.4 for LQM334 vs. 49.1% ± 11.1 for LQM330, and 32.2% ± 2.4 for LQM336), as shown in Figure 6B.

Structure-Activity Relationship (SAR) Analysis
Posteriorly to the cell viability and antiviral assays, a structure-activity relationship (SAR) analysis was performed for this small series of acrylamide analogs (LQM328-337). As a rule, this will be discussed in the following sequence: electron-withdrawing; electron-donating; and aromatic ring substituents, considering the results at 20 and 40 µM, respectively.
Considering Figures 4 and 5, the analog containing trifluoromethyl substituent at position 4 (LQM330) presents good cell viability (88.3% ± 2.8) and viral inhibition activity value of 22.4% ± 10.6. When this group is replaced with a fluorine atom results in an analog (LQM332) with better cell viability (96.4% ± 3.3), however, it abolishes the antiviral effect. Considering the chloro-containing derivatives, it is verified that the chlorinated 2,3-disubstituted analog (LQM331) demonstrates good cell viability; in contrast, it is completely inactive. When the chlorine atom at position 3 is shifted to position 4, an analog (LQM333) non-cytotoxic and inactive against CHIKV is obtained. Moreover, when the chlorine atom is shifted from position 2 to 3, a derivative (LQM328) which is slightly more cytotoxic is generated, presenting a cell viability value of 75.5% ± 7.2. Additionally, a small improvement in the antiviral activity is observed, with an inhibition value of 33.1% ± 3.9. However, it is important to note that LQM328 was highly cytotoxic at 40 µM. Finally, when the chlorine atom at position 4 is removed, with only one remaining chlorine atom at position 3 (LQM334), the most active acrylamide is obtained, with a viral inhibition value of 36.3% ± 3.3 at 20 µM and 81.1% ± 6.4, at 40 µM concentration. Regarding the acrylamides containing electron-donating substituents, the 3,4-disubstituted methoxyl compound (LQM335) showed no cytotoxicity and it was not active against CHIKV. Furthermore, its analog substituted only at position 2 (LQM336) shows a slight antiviral activity at 20 µM (18.5% ± 6.1). Finally, the acrylamide derivative containing a phenyl ring as a substituent at position 4 (LQM329) has high cytotoxicity, exhibiting a poor cell viability value of 41.5% ± 3.3. When this phenyl group is shifted to position 2 (LQM337), its cytotoxicity at 20 µM is abolished.
Concerning the results at 40 µM concentration ( Figure 6), it is possible to verify that the 3,4-disubstituted chlorine compound (LQM328) becomes highly toxic (5.7% ± 0.6), in comparison with its results at 20 µM. In contrast, when the chlorine atom is removed from position 4, the cell viability is strongly increased (LQM334). Additionally, the analog containing a 4-trifluoromethyl substituent (LQM330) exhibited good cell viability. Similarly, the compound with a strong electron-donating group, such as 2-methoxyl substituent (LQM336), shows good cell viability. In contrast, a phenyl ring at position 2 (LQM337) showed high cytotoxicity. Interestingly, compounds presenting electron-withdrawing groups (LQM330 and 334) demonstrated better results in antiviral assays, with inhibition values of 49.1% ± 11.1 and 81.1% ± 6.4, respectively. Finally, LQM336 can be considered as a weak inhibitor against CHIKV, exhibiting an inhibition value of 32.2% ± 2.4, at 40 µM concentration.
In brief, Figure 7 summarizes the SAR analysis for the acrylamide derivatives at 20 µM, since the number of compounds was higher than those at 40 µM concentration.

Intracellular Flow Cytometry Staining for CHIKV after Treatment with LQM334
In order to confirm the promising anti-CHIKV activity of LQM334, investigating its ability to inhibit the viral infection in Vero E6 cells, the intracellular labeling of CHIKV was performed 48 h post-treatment and the percentage of CHIKV-positive cells was detected by flow cytometry. As shown in Figure 8A, LQM334 reduced the cytopathogenic effect induced by the virus compared to untreated cells (CHIKV). Moreover, LQM334 was able to significantly reduce the percentage of CHIKV-positive cells from 74.07% ± 1.19 to 7.38% ± 1.96 at 20 µM and to 0.88% ± 0.29 at 40 µM ( Figure  8B and 8C).

Molecular Docking Studies for LQM334
After the obtainment of interesting results from the intracellular flow cytometry for the compound LQM334, a deep molecular docking analysis was performed throughout the possible six CHIKV targets, which were nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins

Intracellular Flow Cytometry Staining for CHIKV after Treatment with LQM334
In order to confirm the promising anti-CHIKV activity of LQM334, investigating its ability to inhibit the viral infection in Vero E6 cells, the intracellular labeling of CHIKV was performed 48 h post-treatment and the percentage of CHIKV-positive cells was detected by flow cytometry. As shown in Figure 8A, LQM334 reduced the cytopathogenic effect induced by the virus compared to untreated cells (CHIKV). Moreover, LQM334 was able to significantly reduce the percentage of CHIKV-positive cells from 74.07% ± 1.19 to 7.38% ± 1.96 at 20 µM and to 0.88% ± 0.29 at 40 µM ( Figure 8B,C).

Intracellular Flow Cytometry Staining for CHIKV after Treatment with LQM334
In order to confirm the promising anti-CHIKV activity of LQM334, investigating its ability to inhibit the viral infection in Vero E6 cells, the intracellular labeling of CHIKV was performed 48 h post-treatment and the percentage of CHIKV-positive cells was detected by flow cytometry. As shown in Figure 8A, LQM334 reduced the cytopathogenic effect induced by the virus compared to untreated cells (CHIKV). Moreover, LQM334 was able to significantly reduce the percentage of CHIKV-positive cells from 74.07% ± 1.19 to 7.38% ± 1.96 at 20 µM and to 0.88% ± 0.29 at 40 µM ( Figure  8B and 8C).

Molecular Docking Studies for LQM334
After the obtainment of interesting results from the intracellular flow cytometry for the compound LQM334, a deep molecular docking analysis was performed throughout the possible six CHIKV targets, which were nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins complex, and C proteins. As a result, it was identified that LQM334 possibly binds more efficiently to the E2 domain A from the mature E3-E2-E1 glycoproteins complex (PDB: 3N41) ( Figure 9A),

Molecular Docking Studies for LQM334
After the obtainment of interesting results from the intracellular flow cytometry for the compound LQM334, a deep molecular docking analysis was performed throughout the possible six CHIKV targets, which were nsP2, nsP2/helicase, nsP3, immature and mature E3-E2-E1 glycoproteins complex, and C proteins. As a result, it was identified that LQM334 possibly binds more efficiently to the E2 domain A from the mature E3-E2-E1 glycoproteins complex (PDB: 3N41) (Figure 9A), exhibiting a FitScore value of 62.3293. Additionally, it is placed into the central cleft from the protein ( Figure 9B). In order to obtain a FitScore parameter for comparison, the co-crystallized ligand from the mature E3-E2-E1 glycoproteins complex (a NAG molecule) was redocked using the same docking parameters, as described in methods' section. To validate this approach, the root-mean-square deviation (RMSD) was used to evaluate how different the obtained docking orientation is from the corresponding co-crystallized pose of the NAG molecule. Concerning docking solutions, RMSD values allow to classify them as: (a) good solution when RMSD ≤ 2.0 Å; (b) acceptable solutions when RMSD is between 2.0 and 3.0 Å, and (c) bad solutions when RMSD ≥ 3.0 Å [81][82][83]. As a result, an RMSD value of 1.132 Å was obtained for the NAG redocking solution, suggesting a reliable docking protocol. Then, it was revealed that NAG had a FitScore value of 46.4102, suggesting that the LQM334 has a high affinity to this CHIKV target. According to Voss et al. (2010), the NAG molecule hydrophobically interacts with Lys 115 , Thr 116 , Phe 118 , Lys 181 , Leu 261 , Ala 262 , and Asn 263 amino acid residues [84]. Regarding these interactions, LQM334 also interacts with Phe 118 and Lys 181 residues. Still, it hydrophobically interacts with Leu 42 , Val 179 , Tyr 180 , Asn 264 , Pro 265 , and Val 266 , while also interacting with Ser 120 and Tyr 122 amino acids, via hydrogen bonding interactions at distances of 2.07 and 2.1 Å, respectively ( Figure 9C). Additionally, in silico studies involving E3-E2-E1 glycoproteins have been developed focusing on virtual screening of phenothiazines, bafilomycin [85], and FAD-approved antimicrobial agents, such as cefmenoxime, ceforanide, cefotetan, cefonicid sodium, and cefpiramide [86]. Recently, Song et al. (2019) [87] reported the crystal structures of the free mouse MXRA8 (mMXR8) receptor and the complex between human MXRA8 (hMXRA8) and the CHIKV E3-E2-E1 glycoproteins complex. From this study, the authors verified that the interaction between hMXRA8 and E3-E2-E1 glycoproteins complex occurs via 25 hydrophobic contacts, including Asn 264 and Pro 265 residues, as observed for LQM334 interactions. This fact reinforces that LQM334 possibly binds to an important binding site in the E3-E2-E1 glycoproteins complex and thus, it could suggest this molecule as a potential virus entry inhibitor.  Figure 9B). In order to obtain a FitScore parameter for comparison, the co-crystallized ligand from the mature E3-E2-E1 glycoproteins complex (a NAG molecule) was redocked using the same docking parameters, as described in methods' section. To validate this approach, the root-mean-square deviation (RMSD) was used to evaluate how different the obtained docking orientation is from the corresponding co-crystallized pose of the NAG molecule.  Figure 9C). Additionally, in silico studies involving E3-E2-E1 glycoproteins have been developed focusing on virtual screening of phenothiazines, bafilomycin [85], and FADapproved antimicrobial agents, such as cefmenoxime, ceforanide, cefotetan, cefonicid sodium, and cefpiramide [86]. Recently, Song et al. (2019) [87] reported the crystal structures of the free mouse MXRA8 (mMXR8) receptor and the complex between human MXRA8 (hMXRA8) and the CHIKV E3-E2-E1 glycoproteins complex. From this study, the authors verified that the interaction between hMXRA8 and E3-E2-E1 glycoproteins complex occurs via 25 hydrophobic contacts, including Asn 264 and Pro 265 residues, as observed for LQM334 interactions. This fact reinforces that LQM334 possibly binds to an important binding site in the E3-E2-E1 glycoproteins complex and thus, it could suggest this molecule as a potential virus entry inhibitor.

Computational Details and Computer-Aided Drug Design
All in silico experiments involving molecular docking were performed in a Dell ® notebook, (Texas, USA), model 5500U, with an Intel ® Core TM 4th generation i-7 processer, CPU 2.4 GHz, 16 GB RAM, and running at Windows ® 8.1 platform (Redmond, USA).

Reagents and Solvents
All starting reagents and solvents were purchased from Merck/Sigma-Aldrich ® Company (St. Louis, MO, USA), and they were commercial products of high purity (>98%). Additionally, the solvents used in reactions and column chromatography were subjected to rotary evaporation before use to remove impurities. Finally, in high-performance liquid chromatography (HPLC) experiments, methanol HPLC degree from Tedia ® High Purity Solvents Company (Fairfield, OH, USA) was used as eluent.

Chemical Characterization and Apparatus
For the intermediate products (3a-j), information about yields and physical aspects was provided. Additionally, these were only characterized by hydrogen Nuclear Magnetic Resonance ( 1 H NMR) since they are not new in the literature. For the final products (LQM328-LQM337), information about yields and physical aspects, purity, retention time (R T ), melting point (Mp) or degradation point (Dp), Fourier-Transform Infrared (FT-IR) spectra, 1 H and 13 C NMR spectra, and elemental analyses (CHN) were provided, since most of the final products are completely new in the literature. Finally, for the acrylamides previously synthesized, the corresponding references were also provided in this section.

Melting Point Determination
All melting point (Mp) for final compounds were determined by using an MSTecnopon ® (Piracicaba, Brazil), model PFMII Digital, with maximum temperature at 330 • C, utilizing glass capillaries containing the samples. Initially, 40 • C was admitted as a starting temperature and then a temperature increase by 1 • C/min was allowed. In some cases, Mp due to the degradation of the sample was not observed. Thus, the degradation point (Dp) was computed. Finally, all Mp or Dp are uncorrected and were assumed in a range of 1 • C between the values [102].

Elemental Analysis (CHN)
Samples containing 1-2 mg of LQM328, LQM331, and LQM337 were placed into tin capsules for solids, specific for elemental analyses. All determinations were carried out in a Perkin Elmer ® equipment, model CHNS/O Analyzer 2400 series II. For combustion and reduction columns, the temperatures of 950 and 640 • C were used, respectively. Gas pressures for O 2 e He were admitted as 140 and 105 KPa, respectively. Additionally, a combustion column filling-time of 30 s was assumed. Finally, a total run-time of 5 min for each sample was allowed. These procedures are adaptations from works previously reported [107][108][109].
3.9. Synthesis of Cinnamic Acid and Acrylamide Derivatives 3.9.1. General Procedures for the Obtainment of Cinnamic Acids (3a-j) In general, an adaptation of methods described by Luo and collaborators (2015) was used [110], through a Knoevenagel condensation Doebner modification reaction [77]. In a bottom flask (50 mL) containing 6 mL pyridine, the corresponding aldehydes (1.0 eq.) were added. Subsequently, malonic acid (1.1 eq.) was also added into the solution. After 15 min under reflux and stirring, N-methyl piperazine (10 mol%) was added as a catalyst base. Then, the reactional mixture was kept at these conditions overnight. After the reaction completion (verified by TLC), 10 mL distilled water was added to the crude mixture, providing a white powder precipitated. The heterogeneous mixture was refrigerated (2 • C) by 30 min. Posteriorly, the mixture was stirred for 10 min and, then, 15 mL concentrated HCl (37%) was added into the flask until the pH of 1. Finally, the resulting precipitated was filtered and washed by distilled water (2 × 50 mL), affording the corresponding cinnamic acids. In general, an adaptation of methods described by Tanja and collaborators (2019) was used [79]. Initially, aniline (1.0 eq.) was added into a bottom flask (50 mL) containing 5 mL dimethylformamide (DMF) as the solvent. The corresponding cinnamic acid derivatives (1.1 eq.) were posteriorly added. Then, 2-(1H-Benzotriazole-1-yl)-1,1',3,3'-tetramethyluronium tetrafluoroborate-TBTU (1.0 eq.) was of 0.5 mg/mL followed by incubation for 3h. The culture medium was removed and 150 µL of dimethyl sulfoxide (DMSO) was added to each well leading to formazan solubilization. The value of blank control absorbance (only the used culture medium in the absence of cells) was subtracted from all samples. The absorbance of each well was measured at a 492 nm wavelength and the percentage of cell viability was calculated as follows: Cell viability (%) = [sample absorbance/average of cell control absorbance] × 100. (1)

In Vitro Antiviral Assay
Initially, a serial dilution of CHIKV stock was performed and the viral dilution that has been reduced cell viability at least 80% was used in antiviral assays (data not shown). After that, Vero E6 cells were plated at 2 × 10 4 cells/well in a 96-well microplate and maintained at 37 • C and 5% CO 2 atmosphere until reaching the confluence of~80-90%. The virus adsorption was then performed by incubating the cells with CHIKV diluted 1:200 in DMEM-low glucose medium/2% bovine fetal serum for 2 h with homogenization every 15 min. Thereafter, the medium was removed and the cellular monolayers were washed with phosphate-buffered saline, and several synthesized acrylamides were added at 20 or 40 µM. The cell viability was assessed after 48 or 72 h by MTT cell viability assay as previously described (topic 3.9). The percentage of viral inhibition was calculated as follows: Inhibition (%) = [sample absorbance − average of viral control absorbance/average of (2) cellular control absorbance − average of viral control absorbance] × 100.

Intracellular Flow Cytometry Staining for CHIKV
The antiviral activity of the LQM334 compound was confirmed by intracellular flow cytometry staining [119]. Briefly, after CHIKV adsorption in Vero E6 cells for 2 h, the medium was removed, the cell monolayer was washed with PBS. The LQM334 compound was added at 20 or 40 µM concentrations and the cells were maintained at 37 • C/5% CO 2 atmosphere for 48h. The cells were detached by using a trypsin/EDTA solution and submitted to fixation and permeabilization using the BD Cytofix/ Cytoperm TM Fixation/Permeabilization solution kit (BD Biosciences ® , San Jose, CA, USA) according to the manufacturer's recommendations. Subsequently, the cells were incubated with an anti-CHIKV monoclonal antibody (1:50; A54Q clone; Invitrogen, Carlsbad, CA, USA) for 1h at 4 • C. The cells were washed with the BD Perm Wash solution (BD Biosciences ® , San Jose, CA, USA) and then incubated with the goat anti-mouse IgG (H + L) cross-adsorbed secondary antibody conjugated with Alexa Fluor 488 (1:200; Invitrogen) for 1h at 4 • C. A total of 20,000 events were acquired in the BD FACS Canto TM II flow cytometer (BD Biosciences ® , San Jose, CA, USA) and the results were analyzed by using FlowJo TM v. 10 software.

Statistical Analysis
The statistical analyses were performed in the GraphPad Prism ® v.6.0 software (San Diego, CA, USA) using One-Way ANOVA followed by Dunnett multiple comparison tests, and the p ≤ 0.05 was considered statistically significant.

Conclusions
A dataset of 132 compounds (including acrylamides and acylhydrazones) was built for a virtual protocol. Then, the computer-aided drug design protocol applied towards six CHIKV targets was able to identify 10 promising acrylamides to be synthesized and biologically evaluated. In the cytotoxicity assay, nine of 10 synthesized compounds presented a cell viability higher than 75%, at 20 µM concentration, with promising results in regards to the challenge in infected Vero E6 cells with CHIKV. As a result, the compound LQM334 was found to be the most active analog, even though in a higher concentration, it kept the cell viability and viral inhibition. Additionally, the intracellular flow cytometry staining demonstrated that LQM334 inhibited cell infection by the CHIKV. Regarding the antiviral activity of LQM334, the deep molecular docking analysis pointed to a possible virus target for LQM334. In this regard, it was identified that LQM334 preferably interacts with the E2 domain A from the mature E3-E2-E1 glycoproteins complex from CHIKV. Additionally, it displays hydrogen bonding interactions with Ser 120 and Tyr 122 amino acid residues. Therefore, our virtual pipeline pointed out a potential inhibitor of E3-E2-E1 glycoproteins complex from CHIKV with antiviral activity. It suggests that the medicinal chemistry of CHIKV should be more explored in order to provide more information about the most relevant chemical classes of compounds for designing new antiviral agents. Finally, this work represents the emergence of a new potential E3-E2-E1 glycoprotein complex inhibitor, which is in continuous development and several structural optimizations are being performed currently in order to identify new potent candidates with low toxicity. Concerning all of the results for LQM334, it is possible to suggest that this acrylamide analog could be used as a promising anti-CHIKV scaffold for the design of new antiviral agents in the future.