Characterization of the RNA-dependent RNA polymerase from Chikungunya virus and discovery of a novel ligand as a potential drug candidate

Chikungunya virus (CHIKV) is the causative agent of Chikungunya fever, an acute febrile and arthritogenic illness with no effective treatments available. The development of effective therapeutic strategies could be significantly accelerated with detailed knowledge of the molecular components behind CHIKV replication. However, drug discovery is hindered by our incomplete understanding of their main components. The RNA-dependent RNA-polymerase (nsP4-CHIKV) is considered the key enzyme of the CHIKV replication complex and a suitable target for antiviral therapy. Herein, the nsP4-CHIKV was extensively characterized through experimental and computational biophysical methods. In the search for new molecules against CHIKV, a compound designated LabMol-309 was identified as a strong ligand of the nsp4-CHIKV and mapped to bind to its active site. The antiviral activity of LabMol-309 was evaluated in cellular-based assays using a CHIKV replicon system and a reporter virus. In conclusion, this study highlights the biophysical features of nsP4-CHIKV and identifies a new compound as a promising antiviral agent against CHIKV infection.

The Chikungunya virus (CHIKV) belongs to the Togaviridae family and is the causative agent of Chikungunya fever. The main transmission route occurs through the bite of infected female mosquitoes of the Aedes sp. Genus. After CHIKV infection, the proportion of individuals who develop clinical and debilitating symptoms is considered the highest compared to other arboviruses, with an average of 80% of symptomatic cases 1,2 . The control of the mosquito vector remains the best prophylaxis since there are no licensed vaccines or efficient antivirals available 3 . In this scenario, the infection caused by CHIKV has a high social impact and constitutes a serious public health issue 3 .

nsP4-CHIKV purification and SEC-MALS analysis.
The nsP4-CHIKV was bacterially expressed and purified using chromatography systems, and the purity was confirmed by acrylamide gel electrophoresis ( Supplementary Fig. 1). The nsP4-CHIKV is formed by 492 amino acids and has a theoretical molecular mass (MM) of 54.54 kDa. This construction covers the entire region of the RNA-dependent RNA polymerase (RdRp) domain, responsible for the nsP4-CHIKV function and where the catalytic aspartic acid residues (Asp371 and Asp466) are located 23 . The Asp466 is in the GDD motif, a highly conserved sequence of viral polymerases 23 .
In order to determine the oligomeric state of nsP4-CHIKV in the working buffer solution, Size Exclusion Chromatography coupled with Multi-angle Light Scattering (SEC-MALS) was employed 24 . The SEC-MALS data showed a low polydispersity index and yielded a MM of (60 ± 1)kDa for nsP4-CHIKV (Fig. 1A). The proximity of the experimental value with the theoretical molecular mass of the protein suggests that the monomeric state is the most populated oligomer under the evaluated conditions. Evaluation of nsP4-CHIKV secondary structure profile. Circular dichroism (CD) spectroscopy was used to estimate the secondary structure content of nsP4-CHIKV in solution 25 . The CD spectrum of nsP4-CHIKV is characteristic of an α-helical rich protein, with two negative minima at 208 and 222 nm and one positive maximum around 195 nm (Fig. 1B). This profile corroborates with structural features described for viral polymerases and other polymerase structures solved experimentally 23  www.nature.com/scientificreports/ Dichroweb was used in the quantitative analysis of the CD spectrum 26 . The best-fit spectra were obtained with the CDSSRT method 27 , which displayed an NRMSD of 0.009 for the SP175 database and 0.012 for Sets 4 and 7, considered the most representative of the secondary structure content of nsP4-CHIKV (Supplementary Table 1). Secondary structure content estimation from the CD spectrum confirmed the higher percentage of α-helix in the nsP4-CHIKV (60%). The remaining fractions were: 6.3% of sheets, 9.6% of turns and 24% of disordered structure. nsP4-CHIKV thermal stability profile through CD spectroscopy and DSC. The CD technique was also applied to study the thermal behavior of this protein when subjected to temperature variations 28 . The unfolding transition of nsP4-CHIKV was studied by monitoring the ellipticity at 222 nm as a function of temperature 28 . The results showed that the thermal denaturation process occurred cooperatively, exhibiting the transition from folded to unfolded states in a defined way (Fig. 1C).
The thermodynamic parameters of the nsP4-CHIKV unfolding were obtained by adjusting a two-state equilibrium model to the experimental data. Changes in the heat capacity were not considered, and the fitting was performed taking into account the linear changes in pre-and post-transition ellipticity as a function of the temperature 28 . Thus, the melting temperature (T m ), the apparent enthalpy change (ΔH app ), and the apparent entropy change (ΔS app ) for the same protein concentration were obtained at two different heating rates (Table 1). The proximity of the estimated molecular weight of (60 ± 1) kDa with the molecular weight of the protein suggests that nsP4-CHIKV is mostly at a monomeric state under the evaluated conditions. (B) Secondary structure profile of nsP4-CHIKV by CD spectroscopy. The nsP4-CHIKV spectrum suggests the predominance of helical secondary structures. (C) Thermal stability of nsP4-CHIKV probed by CD spectroscopy. The ellipticity at 222 nm as a function of temperature for nsP4-CHIKV was recorded at 0.5 °C/min (left) and 1.0 °C/min (right) and transformed to the protein unfolded fraction. Solid lines are best fits to the CD data using a two-state equilibrium model. The thermodynamic parameters of the protein unfolding transition are summarized in Table 1. (D) Scan-rate normalized (13 °C/h) DSC data of the irreversible thermal denaturation of 11.9 μM of nsP4-CHIKV and the corresponding instrumental buffer (50 mM Tris-HCl pH 8,0, 200 mM NaCl e 5% glycerol) baseline. (E) Excess heat capacity of nsP4-CHIKV at the indicated scan rates obtained after normalization by protein concentration and subtraction of the buffer baseline. (F) Arrhenius-type plot showing the scan rate dependence of the nsP4-CHIKV unfolding temperature, T m . The slope yields the activation energy for the irreversible denaturation of nsP4-CHIKV. Table 1. Thermodynamic parameters of nsP4-CHIKV unfolding by CD spectroscopy. The T m and the ΔH app were determined by fitting the CD data to a two-state equilibrium model 28  www.nature.com/scientificreports/ The thermodynamic parameters of the nsP4-CHIKV unfolding transition exhibited a dependence on the heating rate. This result suggests that the irreversible nsP4-CHIKV unfolding transition is kinetically dependent 29,30 . The values of the enthalpy and entropy changes obtained from the CD data agree well with those observed for globular proteins 31 . The thermal behavior of nsP4-CHIKV was also evaluated using Differential Scanning Calorimetry (DSC) 32 . Figure 1D shows the temperature-dependence of the heat capacity profile (C p -sample minus reference) of nsP4-CHIKV and the instrumental buffer baseline, acquired with a scan rate of 13 °C/h. The protein C p was subtracted from the buffer baseline C p and normalized to the protein concentration. DSC experiments were performed at different scan rates to investigate its effect on the DSC profile and the reversibility of the transitions. The temperature-and scan rate dependence of the excess heat capacity profile of nsP4-CHIKV are illustrated in Fig. 1E. The nsP4-CHIKV undergoes an irreversible thermal denaturation, and the values of the calorimetric enthalpy change (ΔH cal ) obtained from the analysis of the thermograms are within the range of values observed for other globular proteins [33][34][35][36] . Moreover, the transition peak shows a clear scan rate dependence, confirming that all thermodynamic parameters associated with nsP4-CHIKV thermal denaturation depend upon the heating rate (Table 2), as observed in our previous CD analysis. Except for the van't Hoff enthalpy change (ΔH vH ), the dependences of the thermodynamic parameters on the heating rate are markedly non-linear (Table 2 and Supplementary Fig. 2). This feature illustrates the non-equilibrium character of the protein denaturation process.
The dependence of T m on the scan rate was used to calculate the kinetic activation energy, E a , for the irreversible nsP4-CHIKV thermal denaturation. According to Sanchez-Ruiz et al. 30 , the T m shifts induced by different heating scan rates, ν, for an irreversible two-state process can be modeled by the following equation: where A is the pre-exponential factor in the Arrhenius equation, and R is the gas constant. Thus, by plotting ln ν/T 2 m against 1/T m , the apparent activation energy can be determined from the slope of the curve. The Arrhenius plot showing the scan rate-dependent changes in the T m is illustrated in Fig. 1F, from which E a was determined as (110 ± 4) kcal/mol. Evaluation of nsp4-CHIKV interaction with compounds. In the search for new compounds able to interact and inhibit the nsP4-CHIKV in solution, an initial experimental screening with a series of 12 compounds (Supplementary Table 2) was selected from the OpenZika project 22,37 and performed using differential scanning fluorimetry (DSF or ThermoFluor assay). For nsP4-CHIKV, the T m in the absence of compounds (only with DMSO) was 37.7 ± 0.4 °C. Among the compounds, LabMol-309 ( Fig. 2A) caused the highest thermal shift (~ 4 °C), suggesting the occurrence of interaction with nsP4-CHIKV. Therefore, this compound was chosen to proceed with the other assays.
The interaction between LabMol-309 and nsP4-CHIKV was later analyzed by MicroScale thermophoresis (MST) and solution nuclear magnetic resonance (NMR). The MST data showed an affinity curve with the occurrence of well-defined bound and unbound states (Fig. 2B). From that, the dissociation constant (K D ) for the interaction of nsP4-CHIKV with LabMol-309 was estimated as (6 ± 1) µM.
The interaction of LabMol-309 with nsP4-CHIKV was further evaluated by solution NMR, monitoring the chemical shift perturbation (CSP). Figure 2C shows the spectra obtained for the compound LabMol-309 in the presence (red line) and absence of the protein (blue line). These spectra were superimposed, and the chemical shift differences were identified and mapped according to the respective positions of the proton resonances ( Fig. 2C) previously identified in the LabMol-309 assignment ( Supplementary Fig. 3).
Therefore, detecting these chemical shifts perturbations is additional evidence for the interaction between nsP4-CHIKV and compound LabMol-309, reinforcing the results obtained using DSF and MST.
The nsP4-CHIKV three-dimensional model and structural analysis. A combined analysis of several biophysical techniques suggests that the nsP4-CHIKV is a monomeric α-helical rich protein capable of forming a complex with the compound LabMol-309 within a moderate dissociation constant (10 −6 M). The nsP4-CHIKV 3D structural model was obtained using AlphaFold 38 . The model obtained showed a very high Table 2. Thermodynamic parameters associated with the nsP4-CHIKV thermal denaturation by DSC. T m represents the temperature where C P reaches its maximum value. ΔH cal was calculated as the area under the DSC trace. ΔT 1/2 corresponds to the linewidth at half the height of the transition peak. The entropy change at T m , ΔS, was calculated as ΔS = ΔH cal /T m . ΔH vH was calculated as 4RT m 2 C p,max /ΔH cal . Analyses of the thermograms were performed with MicroCal Origin software. Uncertainties: T m (± 0.2 °C), ΔH cal (± 1-3 kcal/ mol), ΔT 1/2 (± 0.2 °C). www.nature.com/scientificreports/ per-residue confidence score (pLDDT) for more than 90% of the covered sequence, and only for the N-and C-terminal regions pLDDT were low (Fig. 3A). The full-length nsP4-CHIKV structural model is illustrated in Fig. 3.
In the model of nsP4-CHIKV, the electrostatic surface potential of the active site cavity where are located the aspartic acid dyad (Asp371 and Asp466) is remarkably positive, a signature of nucleic acids-interacting motifs ( Fig. 3C) 39 . Additionally, the analysis using ConSurf reveals that the catalytic site region and its surroundings are highly conserved (Fig. 3D). These results serve as corroboration for the robustness of the AlphaFold model of nsP4-CHIKV.

Molecular docking and molecular dynamics of LabMol-309 against nsP4-CHIKV. Docking cal-
culations were used to investigate the binding mode of LabMol-309 at the nsP4-CHIKV. The docking results suggest that LabMol-309 binds to the nsP4-CHIKV active site, interacting with the GDD catalytic triad (Asp466 and Asp467), with a docking score of -7.15 kcal/mol −1 . LabMol-309 makes H-bonds with Glu369, Asp466, Asp467, www.nature.com/scientificreports/ Gly507, Arg573 and cation-π interactions with Lys295 residue (Fig. 4). The nitrogen atom of the indole group and amine of the pyridine group of LabMol-309 make relevant interactions with Asp466 and Asp467, respectively. Additionally, the indole group makes cation-π interaction with Lys295. The structural stability of the structural model of the nsP4-CHIKV/LabMol-309 complex calculated by docking was evaluated using 100 ns molecular dynamics (MD) simulation. Figure 5A presents the values of root mean square deviation (RMSD) for the backbone atoms of the protein and non-hydrogen atoms of the ligand from the initial structure. It is possible to observe that RMSD values are stable after 5 ns of simulation and reach plateaus around 0.3 and 0.5 nm for the protein and ligand, respectively. Figure 5B shows that the number of contacts < 0.6 nm between nsP4-CHIKV and LabMol-309 does not drop down to zero throughout the MD simulation, indicating that the ligand interacts with the protein is persistent. The number of hydrogen bonds is stable throughout the MD simulations and presents an average value of three (Fig. 5C). An evaluation of the hydrogen bonds with significant percentage occupancy (< 5%, Supplementary Table 3) during the MD trajectory reveals that Glu369, Asp371, Asp466, Asp467, Asn468, Lys501, and Arg573 are important for the stabilization of the protein-ligand complex, and further mutagenesis studies may be relevant for confirmation of these interactions. It is worth noting that Glu369 and Asp371 presented percentage occupancies higher than 70%. Considering all MD analyses, it can be concluded that the structural model of the nsP4-CHIKV/LabMol-309 complex is stable throughout the simulation.

Inhibition of CHIKV replication by LabMol-309 through replication-based and viral infection assays.
The inhibitory activity of the compound LabMol-309 was evaluated through the replicon-based screenings in a dose-dependent manner to determine its effective and cytotoxic concentrations (EC 50 and CC 50 , respectively). Replicon cells were incubated with twofold serial dilutions of compound (from 20 to 0.03 µM for EC 50 and from 100 to 0.30 µM for CC 50   www.nature.com/scientificreports/ To confirm the antiviral activity of the LabMol-309, we carry out the effective concentration of 50% (EC 50 ) and cytotoxic concentration of 50% (CC 50 ) using BHK-21 cells infected with CHIKV-nanoluciferase, a recombinant CHIKV that express the Nanoluciferase reporter, at a multiplicity of infection (MOI) of 0.1, with a two-fold serial dilution of LabMol-309 at concentrations ranging from 0.78 to 100 µM. Nanoluciferase activity levels, proportional to viral replication, were assessed 16 h post-infection (adapted from 40,41 ). The cytotoxic concentration of 50% (CC 50 ) was determined in parallel experiments (Fig. 6C). As a result, these assays demonstrated that the LabMol-309 has a EC 50 of 5.2 µM on BHK-21 cells infected with CHIKV-nanoluc and CC 50 of 52 µM on naive BHK-21 cells, over a period of incubation of 48 h, resulting in a Selectivity Index (SI) of 10 (Table 3; Fig. 6C;  Supplementary Table 6).

Discussion
The nsP4-CHIKV polymerase plays a crucial role in viral replication and has been considered a promising target for the search and development of new drugs. Thus, understanding its dynamic and structural features is an important step for studies with this target.
Our biophysical data agree with the structure prediction using Alphafold and point out that nsp4-CHIKV is a monomeric protein enriched with alpha-helix content. These observations also agree with experimental data    20,23 . The thermodynamic data showed that when in the in vivo host, this protein can be stable in the vicinity of its thermal denaturation (T m ~ 40 °C and ΔT 1/2 ~ 4-5 °C), and its unfolding is both scan-rate dependent and irreversible under all conditions tested. It was shown before that the T m might be scan-rate dependent if the scan rate exceeds the unfolding rate 42 . The calorimetric thermograms for the irreversible denaturation of proteins are highly scan-rate dependent, and their shapes are normally asymmetric 43 , exactly what is observed for nsp4-CHIKV. Therefore, the kinetics of the thermal denaturation could be treated as a single first-order irreversible step N → D, whose rate of temperature dependence obeys the Arrhenius equation.  www.nature.com/scientificreports/ The effective activation energy derived from this equation was (110 ± 4) kcal/mol, which fits well within the wide range of the values reported for the thermal denaturation of mammalian tissues and strengthened the thermodynamic data collected for this protein 44 .
In the search for new nsp4-CHIKV ligands as potential inhibitors, the compound LabMol-309 was identified as a promising candidate through DSF screening. The validation of this interaction through biophysical methods demonstrated that the compound interacts in the low micromolar range. This compound had already been evaluated in virtual screenings for other viral proteins such as ZIKV and other Flaviviruses 21,22 , but it was the first time it was reported targeting the nsP4-CHIKV protein.
A three-dimensional model was generated using AlphaFold to analyze the LabMol-309-nsP4-CHIKV complex and their mode of interaction. In this sense, the nsP4-CHIKV model presented the regions corresponding to the fingers, palm and thumb domains, which are characteristics of viral polymerases, and the active site region was remarkably positive and conserved. These structural features were equivalent to the regions presented in the RdRp domain of nsP4 from both RRV and SINV, which were experimentally solved recently (PDB ID: 7F0S, 7VB4, 7VW5) 20 . Therefore, the suggested binding mode for LabMol-309 is through the interaction with residues of the nsP4-CHIKV active site.
Although the data involving the complex formation between nsP4-CHIKV and LabMol-309 are solid, it is still not possible to conclude that this compound has inhibitory activity against this enzyme. The development of an enzymatic assay for the nsP4 polymerase from Alphaviruses in general, has been challenging because this protein cannot effectively perform its function on its own, as previously shown by others 45,46 . Different regions of nsP4 recognize the promoters for minus and plus strands. However, the binding requires the presence of the other non-structural proteins to form the replication complex and enable the de novo RNA synthesis 11,[45][46][47] . Moreover, the interactions between these proteins with host components during replication have been studied but remain limited and not completely understood 11,46 .
Tomar et al. 48 reported that template recognition and the nsP4 activation through protein-protein interactions requires the presence of viral polyprotein P123 48 . In another work, the SINV nsP4 was expressed in E. coli, and the polymerase activity was observed only when supplied with the viral polyprotein P123 47 . Recently, Lello et al. 46 demonstrated that nsP4 of SINV, CHIKV, ONNV, BFV, RRV, SFV, MAYV, VEEV, and EILV on their own have minimal RNA polymerase activity 46 . Using a trans replicase system consisting of two relatively independent functional modules (nsP4 and P123), they have shown that the nsP4 of all these Alphaviruses was active only when combined with the corresponding P123 polyproteins 46 . Furthermore, Tan et al. 20 evaluated the polymerase activity of SINV and RRV nsP4 and as a result, the isolated proteins showed less efficient polymerase activity than the dengue virus RdRp used as the positive control 20 . Altogether these findings corroborate that bacterially produced nsP4 could not efficiently synthesise RNA unless combined with the viral polyprotein P123 obtained from animal cell extracts 47,49,50 .
Given these limitations in establishing an efficient method for evaluating the enzymatic activity of purified recombinant nsP4-CHIKV, in this work the inhibitory effect of the compound LabMol-309 was evaluated using both replicon-based assays and cells infected with the CHIKV expressing the nanoluciferase reporter (CHIKVnanoluc). Replicon-based systems have been widely used as tools for drug discovery of antiviral agents, and promising replication inhibitors were identified by this method 51 . Specifically to CHIKV, BHK-21 cells harboring other CHIKV replicon constructs were reported for the high-throughput screening of viral replication inhibitors 52,53 . The same system was also used to evaluate the anti-CHIKV activity of other compounds and different flavonoids [54][55][56] .
The evaluation of LabMol-309 using a replicon-based system was performed in a dose-dependent manner, and its inhibition clearly occurred. Comparing with studies that also used CHIKV replicon, the EC 50 obtained for LabMol-309 was lower than the values already reported for other compounds 52,55 , reinforcing the antiviral potential of this compound. LabMol-309 showed toxicity to the cells, and the resulting low selectivity index of 1.7 may be correlated to a possible negative impact in the cellular factors associated with the viral genome replication. These data suggest that, even with inhibitory activity, chemical modifications would be required to optimize this compound's efficiency and reduce its toxicity. Furthermore, antiviral assays performed with cells infected with a recombinant CHIKV demonstrated that the LabMol-309 decreased CHIKV replication with an EC 50 of 5.2 µM and an CC 50 of 52 µM, with a SI of 10 in BHK-21 cells.
The differences in the obtained values using naive BHK-21 or BHK-CHIKV cells (Table 3) are understandable since different factors are involved in these assays. For example, in the infection system the virus is effectively infecting the cells and performing all the stages of the virus replicative cycle. It means that the treatment with Lab-Mol-309 may be acting even before the formation of the replication complexes. Alternatively, in the BHK-CHIKV replicon system, the replication complexes are already formed when the treatment starts, which can impact on the effectiveness of the antiviral activity in a short period of treatment. Additionally, the presence of the replicon might change the cell response to the compound, and explain the higher cytotoxicity shown in the results. This isolated effect predominantly observed in the replicon cells can be explained by the differences in incubation periods used in the antiviral activity experiments (48 h for replicon-based screenings compared to 16 h for the viral infection assays). The prolonged exposure of cells to the compound can result in higher cytotoxicity 57 , reinforcing the importance of further studies of the ADME-Tox profile in animal models. Additionally, to the best of our knowledge, this is the first description of LabMol-309 as inhibitor of CHIKV replication, and its low EC 50 value is in similar level with other inhibitors reported to block CHIKV replication, emphasizing the antiviral potential of this compound 58,59 . In this context, the results obtained from the antiviral assays suggest that LabMol-309 is a potential molecule to be further optimized to reduce its cytotoxicity and increase the selectivity index in cellbased antiviral assays. In summary, this study highlights biophysical features of nsP4-CHIKV, contributing to basic research on alphaviruses polymerase, and identified a new compound as a promising antiviral agent against nsP4-CHIKV purification. nsP4-CHIKV was purified using an AKTA Purifier System (GE Healthcare).
The first step was affinity chromatography, using a HisTrap HP 5.0 mL column (GE Healthcare) pre-equilibrated with buffer A (50 mM Tris pH 8.0, 500 mM NaCl, 10% glycerol). The elution was performed using 50 mM Tris pH 8.0, 500 mM NaCl, 250 mM imidazole, 10% glycerol. The buffer was exchanged through dialysis to eliminate the imidazole excess. The 6xHis-tag-SUMO was cleaved by TEV protease during overnight incubation at 4 °C. A second affinity chromatography step was performed using the same system to collect the HisTag-less protein obtained after TEV treatment. A final purification step was done using size-exclusion chromatography on an XK 26/1000 Superdex 75 column (GE Healthcare) pre-equilibrated in gel filtration buffer (50 mM Tris pH 8.0, 200 mM NaCl and 5% glycerol). The eluted fractions were collected and analyzed by SDS-PAGE to confirm their purity and mass spectrometry was performed to confirm the presence of nsP4-CHIKV. The final protein sample was concentrated using 30 kDa MWCO centrifugal concentrators (Vivaspin, Sartorius). Protein concentrations were determined spectrophotometrically in a Nanodrop 1000 spectrophotometer, using the measured absorbances at 280 nm and the theoretical extinction coefficient of 36,495 M −1 cm −1 .

Size exclusion chromatography coupled with multi-angle light scattering (SEC-MALS).
The oligomeric state of the purified nsP4-CHIKV was evaluated by size exclusion chromatography coupled with multi-angle light scattering (SEC-MALS) in running buffer composed of 50 mM Tris-HCl pH 8.0 and 200 mM NaCl. For that, 50 µL of purified nsP4-CHIKV at a concentration of 1.5 mg/mL was injected in a Waters 600 HPLC system (Waters) coupled in-line with a UV detector, a mini DAWN TREOS multi-angle light scattering apparatus (Wyatt Technology), a column Superdex 75 Increase 10/300 GL (GE Healthcare), and a refractive index detector Optilab T-rEX (Wyatt Technology). The light scattering detectors were normalized with bovine serum albumin (Sigma-Aldrich) and the flow rate used was 0.5 mL/min. The data were processed using ASTRA7 software (Wyatt Technology) with the following parameters: refractive index of 1.331, 0.890 cP for the viscosity of the solvent, and a refractive index increment of 0.1850 mL/g. Protein solutions were centrifuged for 10 min at 10,000 × g at a controlled temperature of 4 °C immediately before use.

Circular dichroism (CD).
Far UV-CD spectra (195-280 nm) were measured in a Jasco J-810 spectrometer (Jasco Corporation, Japan) equipped with a Peltier control system and using a quartz cell with a 1 mm pathlength. The spectra were recorded from 280 to 195 nm, with a scanning speed of 100 nm/min, a spectral bandwidth of 1 nm and a response time of 0.5 s. All the protein samples were in a final concentration of 2.5 µM diluted in water. Spectral deconvolution was applied to estimate the secondary structure content using the DICHROWEB web server 26 . Three different methods were combined with three different databases to improve the reliability of the results. The detailed analysis of the results generated by these combinations is provided in Supplementary material (Supplementary Table 1). The estimated values of secondary structure fractions were averaged from each database used. The best fit was determined from the analysis of the NRMSD parameter, which was considered satisfactory when closer to 0 61  The nsP4-CHIKV tridimensional model and structural analysis. The nsP4-CHIKV sequence (residues 1 to 611) was used to generate the 3D model by AlphaFold2, developed by DeepMind (https:// alpha fold. ebi. ac. uk/) 38 . The nsP4-CHIKV model was structurally refined for docking calculations at GalaxyRefine server 63 . Surface charge was calculated using APBS 64 and residues conservation was analyzed with ConSurf, following the default parameters 65 . Pymol 66 was used to render the 3D images.
Molecular docking of nsP4-CHIKV and LabMol-309. The docking calculations were performed using the DockThor VS web 67,68 , focusing on the active binding site (Asp371 and Asp466 residues). The nsP4-CHIKV and LabMol-309 structures were prepared using the Protein Preparation Wizard 69 and LigPrep tool 70 . The docking grid was centered at the active binding site; grid size 20 Å; and grid coordinates x, y and z of − 27.84 Å, 12.89 Å and 28.25 Å, respectively. The search algorithm precision mode was set up in the standard configuration of genetic algorithm parameters, with the soft docking mode activated. The PLIP server 71 was used to analyze the protein-ligand patterns (hydrogen bonds, hydrophobic interaction, cation-π, π-stacking, water and salt bridge interactions). Poseview server 72,73 was used to generate 2D interaction diagram and VMD program was used to render the 3D images 74 .

Molecular dynamics simulations.
The initial positions of the nsP4-CHIKV-bound LabMol-309 for the molecular dynamics (MD) simulations were obtained by the molecular docking results, and its topology parameterizations (Molid 814093) were obtained from the ATB server 75 . The MD simulations were performed using GROMACS package version 5.0.7 76 . The molecular system of the protein-ligand complex was modeled with the GROMOS54A7 force field 77 and SPC water model 78 , using a cubic box solvated with 200 mM NaCl. The simulation was realized in ensemble NPT at 25 °C and 1.0 bar using a modified Berendsen thermostat with τ T = 0.1 ps and Parrinello-Rahman barostat with τ P = 2.0 ps and compressibility = 4.5 × 10 -5 ·bar -1 . A cutoff value of 12 Å was used for both Lennard-Jones, and Coulomb potentials and long-range electrostatic interactions were calculated using the Particle Mesh Ewald algorithm (PME) 79 . Energy minimization was performed with the steepest descent integrator and conjugate gradient algorithm, using 1000 kJ·mol −1 ·nm −1 as the maximum force criterion. One hundred thousand molecular dynamics steps were performed for each NVT and NPT equilibration, applying force constants of 1000 kJ·mol −1 ·nm −2 to all heavy atoms of the protein-ligand complex. At the end of preparation, a 100 ns MD simulation of the structural model of the protein-ligand complex was carried out for data acquisition. Next, the trajectory was aligned and analyzed according: RMSD of backbone atoms for protein and nonhydrogen atoms for the ligand, number of hydrogen bounds (cutoff distance of 3.5 Å and maximum angle of 30°) between protein and ligand, and number of contacts < 0.6 nm between all atoms of the protein and of the ligand.

Cells and virus. BHK-21 cells were purchased from The Global Bioresource Center (ATCC) and maintained
in Dulbecco's modified Eagle's medium (DMEM, Sigma-Aldrich) supplemented with 100U/mL of penicillin (Hyclone Laboratories), 100 mg/mL of streptomycin (Hyclone Laboratories), 1% dilution of stock of non-essential amino acids (Hyclone Laboratories) and 10% of fetal bovine serum (FBS, Hyclone Laboratories) in a humidified 5% CO2 incubator at 37 °C. BHK-21-Gluc-nSP-CHIKV-99659 cell line, harboring a replicative CHIKV replicon expressing Gaussia luciferase (Gluc) as a reporter gene, was maintained in DMEM 10% FBS with 500 µg/ ml G418 (Sigma-Aldrich). The CHIKV replicon construct includes a T7 bacteriophage promotor followed by the viral 5' UTR region, the nsp1-4 coding sequence, the CHIKV subgenomic promoter (Sg) followed by the www.nature.com/scientificreports/ GLuc sequence and the expression cassette containing a ubiquitination sequence (Ubi) and the neomycin phosphotransferase gene (Neo-resistance gene), and the viral 3' UTR region. This construction and the development of this replicon cell line will be described elsewhere. The CHIKV expressing nanoluciferase reporter (CHIKVnanoluc) used for the antiviral assays is based on the CHIKV isolate LR2006OPY1 (East/Central/South African genotype) and was produced, rescued, and titrated as previously described 40,41 .
CHIKV replicon-based screenings. LabMol-309 at 200 mM in 100% DMSO was diluted with assay media to a final concentration of 1% (v/v) DMSO and was evaluated in a dose-dependent manner to determine its effectiveness (EC 50 ) and cytotoxic (CC 50 ) concentrations, as described in 80 . Approximately 2 × 10 4 replicon cells/well in DMEM 10% FBS were seeded in a 96-well plate. After 16 h of incubation at 37 °C with 5% CO 2 , the medium was replaced with fresh DMEM supplemented with 2% FBS and compound was added to the cells at twofold serial dilutions. After a 48 h-incubation, 40 µL of the cells' supernatant containing secreted Gluc were mixed with 50 μl of Renilla luciferase Assay Reagent (Promega). The Gluc activity was measured using Spec-traMax i3 Multi-mode Detection Platform (Molecular Devices). Replicon cells in 1% DMSO were used as negative control (0% inhibition). The compound concentration required to inhibit 50% of the Gluc activity (EC 50 ) was estimated using the OriginPro 9.0 software. The cytotoxicity was evaluated through a cell proliferationbased MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, as described in 81 . The compound concentration required to cause 50% cytotoxicity (CC 50 ) was estimated using the OriginPro 9.0 software. The dose-response curves were performed twice in duplicates. The EC 50 and CC 50 values were used to determine the compound's selectivity index (SI = CC 50 /EC 50 ).
Infection assays using CHIKV-nanoluc. To  Cell viability assays in BHK-21 cells. As previously described 40,41 , cell viability was measured by MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] assay (Sigma-Aldrich®). After this, the medium was replaced with the MTT solution at 1 mg/mL, cells were incubated for 30 min, after which the MTT solution was removed and replaced with 300 μL of DMSO (dimethyl sulfoxide) to solubilize the formazan crystals. The absorbance was measured at 490 nm on the Glomax microplate reader (Promega®). Cell viability was calculated according to the equation (T/C) × 100%, where T and C represent the mean optical density of the treated and untreated control groups, respectively. The values of CC 50 and EC 50 were used to calculate the selectivity index (SI = CC 50/ EC 50 ). The cytotoxic concentration of 50% (CC 50 ) and the effective concentration of 50% inhibition (EC 50 ) were calculated using GraphPad Prism 8.0.0 for Windows (GraphPad Software, San Diego, California USA, www. graph pad. com).

Data availability
The datasets generated and/or analysed during the current study are included in the are included in this published article [and its supplementary information files]. The raw data of all cellular assays presented in the manuscript were available in the Supplementary information. Additionally, the three-dimensional model of the protein generated using Alphafold is available upon request to the corresponding author. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.