Inhibition and disintegration of Bacillus subtilis biofilm with small molecule inhibitors identified through virtual screening for targeting TasA(28-261), the major protein component of ECM

Abstract Microbial biofilms have been recognized for a vital role in antibiotic resistance and chronic microbial infections for 2-3 decades; still, there are no 'anti-biofilm drugs' available for human applications. There is an urgent need to develop novel 'anti-biofilms' therapeutics to manage biofilm-associated infectious diseases. Several reports have suggested that targeting molecules involved in quorum sensing or biofilm-specific transcription may inhibit biofilm formation. However, the possibility of targeting other vital components of microbial biofilms, especially the extracellular matrix (ECM) components, has remained largely unexplored. Here we report targeting TasA(28-261), the major proteinaceous component of Bacillus subtilis ECM with two small molecule inhibitors (lovastatin and simvastatin) identified through virtual screening and drug repurposing, resulted in complete inhibition of biofilm. In molecular docking and dynamics simulation studies, lovastatin was observed to make stable interactions with TasA(28-261), whereas the simvastatin – TasA(28-261) interactions were relatively less stable. However, in subsequent in vitro studies, both lovastatin and simvastatin successfully inhibited B. subtilis biofilm formation at MIC values of < 10 µg/ml. Besides, these potential inhibitors also caused the disintegration of pre-formed biofilms. Results presented here provide ‘proof of concept’ for the hypothesis that targeting the extracellular matrix's vital component(s) could be one of the most efficient approaches for inhibiting microbial biofilms and disintegrating the pre-formed biofilms. We propose that a similar approach targeting ECM-associated proteins with FDA-approved drugs could be implemented to develop novel anti-biofilm therapeutic strategies against biofilm-forming chronic microbial pathogens. Communicated by Ramaswamy H. Sarma


Introduction
Microbial biofilms are communities of aggregated microbial cells communities of microorganisms attached to surfaces and embedded within an extracellular matrix (ECM) comprising polymeric substances including DNA, Proteins, and Polysaccharides (Davey & O'toole, 2000;Yin et al., 2019). Biofilms have been reported to enhance microbial persistence in adverse environments, e.g., extreme temperature, extreme pH, high salinity, high pressure, poor nutrients, etc. (Davey & O'toole, 2000;Yin et al., 2019). In the recent past, biofilms have also got acknowledged as one of the vital mechanisms involved in microbial pathogenesis and resistance towards antimicrobial therapeutics (Costerton et al., 1999;Donlan, 2000;Stewart & Costerton, 2001). According to a recent report presented by the National Institutes of Health (NIH) -USA, $ 80% of all chronic microbial infections involve biofilm-associated forms of infectious microorganisms (Costerton et al., 1999). Although the precise mechanism underlying the enhanced pathogenesis and resistance towards antibiotics of biofilm-associated forms still remains to be completely elucidated, still it is generally agreed that the enhanced resistance in biofilm-associated forms is caused due to poor antibiotic penetration, nutrient limitation, slow growth, adaptive stress responses, and formation of persister cells within microbial biofilms (Jamal et al., 2018;Stewart & Costerton, 2001).
Given the relevance of biofilms in antibiotic resistance and chronic microbial infections, it is critical to characterize the underlying biochemical and molecular mechanisms involved in various stages of biofilm formation, maturation, and dispersion. Similarly, it is critical to determine how microbial biofilms contribute to chronic microbial infections and identify potential drug targets that could be used to discover and develop innovative anti-biofilm therapeutics. Lastly, but probably, most importantly, it is critical to develop new therapeutic strategies to manage and eradicate biofilmassociated microbial infections effectively. Several studies have been carried out in the recent past, wherein autoinducers involved in quorum sensing or cyclic dinucleotide secondary messengers (e.g., c-di-GMP) have been targeted to inhibit microbial biofilms (Brackman & Coenye, 2015;R€ omling & Balsalobre, 2012). Alternatively, a few reports have shown that targeting discreet molecules such as molecular chaperones (e.g., DnaK in E. coli) could prevent biofilm development (Arita-Morioka et al., 2015). Despite success attained with such studies, there is still a lack of FDA-approved antibiofilm drugs for human application. This situation suggests the necessity of pursuing alternative targets for achieving specific inhibition of microbial biofilms. One of the potential alternative targets could be the biomolecular components of the extracellular matrix. There are multiple reasons to argue that inhibitors targeting ECM components might be more effective than those targeting molecules embedded within the biofilms or present inside the biofilm. This argument is supported by studies that have reported successful inhibition of microbial biofilm and the disintegration of pre-formed biofilms using DNases (which potentially target and degrade the DNA present in biofilm ECM). In the same vein, it could be hypothesized that targeted inhibition of protein components with the biofilm ECM may successfully inhibit microbial biofilms and cause the disintegration of pre-formed biofilms.
To test the above hypothesis, we carried out studies using Bacillus subtilis, one of the most commonly used non-pathogenic bacterial model organisms, for studies on microbial biofilms. It forms biofilms with an extracellular matrix predominantly consisting of the processed protein TasA (Branda et al., 2006). TasA has been reported to be vital for biofilm formation in B. subtilis as deletion mutants lacking TasA cannot produce observable biofilm pellicles (Branda et al., 2006). TasA exists as a stable soluble monomeric protein until it is processed by a signal peptidase viz., SipW (St€ over & Driks, 1999). SipW catalyzes the removal of 27 amino acid residues from its N-terminus, leaving the processed form, i.e., TasA  , which undergoes a rapid structural change, get secreted into the extracellular space, acquires a protease-resistant, b-sheet-rich fibrillary amyloid-like structure in the presence of the accessory protein viz., TapA and exopolysaccharides (Romero et al., 2011). Other studies have reported that interaction of TasA  with the processed form of accessory protein, i.e., TapA , leading to the formation of amyloidlike aggregates of TasA (28-261) -TapA  is critical for formation and maturation of biofilm by B. subtilis (Romero et al., 2014). In light of the essential role of TasA  in B. subtilis biofilm and its occurrence as the major proteinaceous component of the biofilm ECM, we conjectured that inhibition of B. subtilis biofilm could be achieved by targeted inhibition of TasA  with small molecule inhibitor(s). With this rationale, small molecule inhibitors/drugs targeting TasA  were identified through virtual screening and drug repurposing. The study involved homology modeling, virtual screening, molecular dynamics simulation, and validation of identified inhibitor(s) with in vitro experiments.
The experimentally determined 3-D structures of either TasA or the processed form, i.e., TasA (28-261), were unavailable when the present study was initiated. Therefore, at this point of the study, several models of TasA  were generated; the best model was validated with Ramachandran Plot analyses, and its quality, stability was assessed in extended molecular dynamics simulation over 100 ns. The best model was selected to predict the binding sites and screen the potential small-molecule inhibitor(s) of TasA  through virtual screening and drug repurposing. For this purpose, the FDA-approved drug database was used. In addition, the validated model of TasA  was also used in another study by our group wherein computational analysis of TasA (28-261) -TapA  interaction was carried out (Verma et al., 2020). While our studies were going on, the X-Ray defined structure of a truncated version of this protein, i.e., TasA  , was determined (Diehl et al., 2018). It became available in Protein DataBank with PDB ID 5OF1. Noticeably, it lacks 29 amino acid residues at the N-terminus and 21 amino acid residues at the Cterminus (Diehl et al., 2018). Upon availability of validated modeled structure of TasA  and the crystal structure of TasA (PDF ID5OF1), both structures were used to screen and identify potential inhibitors during the present study. Results from virtual screening identified simvastatin and lovastatin as potential inhibitors of TasA  . However, in subsequent analyses with extended molecular dynamics (MD) simulation, only lovastatin was observed to form a stable complex with TasA (28-261), whereas simvastatin didn't show a stable trajectory in extended MD simulation; therefore, simvastatin may not be completely claimed as an inhibitor of TasA  . Remarkably yet, in in vitro studies, both lovastatin and simvastatin inhibited biofilm formation by B. subtilis with minimum biofilm formation inhibitory concentrations of < 10 mM. Furthermore, both lovastatin and simvastatin also caused the disintegration of previously formed biofilms of B. subtilis.
These results are rather interesting and present one of the first 'proof of concept' reports for successfully inhibiting and disintegrating pre-formed bacterial biofilms by targeting the protein component of the extracellular matrix with inhibitors selected through a structure-guided virtual screening approach. We advocate that a similar approach could be implemented to discover and develop anti-biofilm therapeutics that inhibit biofilms formation by priority pathogens for efficiently managing the biofilm-associated chronic microbial infections.

Materials and methods
Homology modelling of TasA

and protein preparation
The crystal structure of either TasA or the processed TasA (i.e., TasA  ; which consists of amino acid residues 28-261, was not available in the protein data bank (PDB) (http:// www.rcsb.org; therefore, multiple models of TasA  were generated with a homology modeling approach, validated with Ramachandran Plot Analyses and Molecular Dynamics Simulation for 100 ns. The selected and validated model was used in a previous study performed by our group (Verma et al., 2020). The same validated model was used during the present study. In addition, the crystal structure(s) of the folded monomer of TasA (PDB IDs: 5OF1 and 5OF2; retrieved from PDB) were also used during the present study. The modeled structure of TasA  , as well as the crystal structures 5OF1 and 5OF2, were subjected to energy minimization, refinement, and protein preparation using protein preparation wizard of Glide software version 6.9 (https:// www.schrodinger.com/) (Schrodinger LLC, New York, NY, USA) by applying an OPLS-2005 force field.

Binding site prediction
The binding sites of the modeled structures and crystal structures were predicted by using Sitemap (Sitemap, version 3.7, Schr€ odinger, LLC, New York, NY, 2015). The PDB files of query proteins were uploaded onto the site map, and the OPLS_2005 force field was used to crop site maps at 4 Å from the nearest site point. For each site calculation, physical descriptors, e.g., size of the site, the degree of the enclosure by the protein, the degree of exposure to solvent, the tightness with which the site points interact with the protein, the hydrophobic and hydrophilic character of the site as well as the balance between these terms, and the degree to which a ligand has the possibility of donating or accepting hydrogen bonds were used for prediction of the binding sites as recommended previously (Halgren, 2009).
Virtual screening for potential inhibitors of TasA  and their binding energy calculation To screen the potential inhibitors of TasA  , it was reasoned that screening FDA-approved drugs/small molecules would be more fruitful. During the past few decades, de novo drug development has become extremely costly and time-consuming, whereas virtual screening of potential inhibitors from the FDA-approved drugs (also referred to as drug repurposing) has emerged as an approach projected to accelerate scientific findings into clinical conversions (Park, 2019;Pushpakom et al., 2019). With this rationale, the 2 D-FDA-approved ligands data file was obtained from ZINC databases (https://zinc.docking.org/). The ligand library of compounds was prepared by LigPrep, and screened compounds were filtered at three different Glide Docking Levels as mentioned above. The potential inhibitors were ranked by assessing the binding energies of ligand upon their binding onto the predicted binding sites of the TasA  . The virtual screening workflow of GLIDE 6.9 (Schr€ odinger, LLC, New York, NY, 2015) was used. According to the rankings, the 50% of Glide docked ligands that precede at the three different levels (HTVS, SP, and XP) were used for further analyses. The final selection of ligands was made based on the Glide extra-precision (XP) mode, which is designed to determine all reasonable conformations and produce the most accurate binding poses. The best screen compounds by virtual screening were further subjected for free energy calculation for respective docking poses using Prime MM/GBSA (Molecular mechanics/generalized born model and solvent accessibility) with OPLS2005 from Maestro version 10.4 (Schr€ odinger, LLC, New York, NY, 2015). The binding energy computes the force field energies in bounded and unbounded molecules involved in docking processes, and the OLPS2005 includes Van der Waals and electrostatic interactions.

Molecular dynamics simulation on protein-inhibitor complexes
To calculate the physicochemical properties and stability of proteinpotential inhibitor complexes, the Molecular Dynamics (MD) simulation was performed by using GROMACS-2019 (http://www.gromacs.org/) (Pronk et al., 2013). The top ten hits of protein-potential inhibitor docked complexes (sorted based on binding energy) were subjected to MD simulations for the periods of 10 nanoseconds with GROMOS96 43A1 force field. Further, the MD simulations for best hits (i.e., top two inhibitors) were re-performed over the simulation periods of 50 ns each. For MD simulations, the ligand topologies were prepared using the PRODRG2 server (http://davapc1.bioch.dundee.ac.uk/cgi-bin/ prodrg) (Sch€ uttelkopf & Van Aalten, 2004). The simulations were conducted in a solvent environment using the bt-dodecahedron box and TIP3P water model. Simulations were neutralized by adding positive ion (Na þ ) and negative ion (Cl -) ions. The energy minimizations during MD simulations were performed for 5000 cycles, and these cycles were carried out through the steepest descent method. Subsequently, the equilibration for dynamics was performed by heating the complexes to the temperature of 300 K for 100 picoseconds using the NVT Berendsen thermostat algorithm. The equilibration followed heating for 100 picoseconds steps using NPT Ensemble of Nose-Hoover thermostat. A linear constraint solver for molecular simulations (LINCS) algorithm was used to constrain all bonds involving hydrogen atoms, whereas Particle-Mesh Ewald (PME) method was used for electrostatic calculations. The stability of the ligand-bound complex with the druggable site of TasA (28-261) obtained from 50 ns Molecular dynamics calculation was further checked by extending the molecular simulation analysis at 100 ns. The MD simulation studies at 100 nanoseconds were performed using Desmond with the same set of dynamics simulations performed by Gromacs. All Molecular dynamics simulation trajectories were visualized and analyzed using the Visual Molecular Dynamics (VMD) (https://www.ks. uiuc.edu/Research/vmd/) as developed and described earlier (Humphrey et al., 1996). All graphs of different simulations such as Root Mean Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF), solvent accessible surface area (SASA), and radius of gyration (Rg) were plotted using Graph-pad Prism (https://www.graphpad.com/scientific-software/prism/) and Grace (http://plasma-gate.weizmann.ac.il/Grace/).

Principal component analysis (PCA) and free energy landscape (FEL)
The principal component analysis (PCA) of best-docked complexes selected based on molecular docking scores and molecular dynamics simulation was performed on C-alpha atoms of TasA (28-261) using GROMACS tools. PCA analysis was carried out to obtain a linear transformation with all the basic motions governing conformational transitions via covariance matrix throughout the overall molecular simulations according to the method described by David and Jacobs (David & Jacobs, 2014). The g_anaeig and g_covar tools (built in within the Gromacs) were used for the generation of the co-variance matrices and their diagonalization. The Free Energy Landscapes (FELs) were determined to understand the conformational changes in protein structures by estimating joint probability distribution. As described earlier, the proteins associated in different energy states or distributed onto the 3 D space were analyzed by principal component analysis, i.e., PC1 and PC2 (Frauenfelder et al., 1991). The g_sham module in Gromacs was used for FEL analysis, and the plots were viewed by Ghostscript version 9.26 (https:// www.ghostscript.com/Ghostscript_9.26.html) and XmGrace (http://plasma-gate.weizmann.ac.il/Grace/).

Chemicals
The small molecule inhibitors (lovastatin and simvastatin) selected through in silico analyses described above were purchased from Tokyo Chemical Industry (TCI, Chennai, TN, India). Bacterial growth media or media components, e.g., Nutrient Broth, Luria Broth, Agar, Ammonium Sulfate, Magnesium Sulfate, Potassium Phosphate, Sodium Citrate, and Glucose were procured from HiMedia Laboratories Pvt. Ltd. (HiMedia, Mumbai, MH, India). All other chemicals (e.g., Methanol, Acetic Acid, DMSO, Safranin, etc.) used during the present study were of the purest grade available.

Bacterial strain and growth conditions
The model organism used during the present study, i.e., Bacillus subtilis strain MCC2049 (T), was procured from National Centre for Microbiol Resources (NCMR) -NCCS (Pune, MH, India). Cultures were grown and maintained on Nutrient Broth (NB), or Luria Broth (LB) prepared and sterilized as per the manufacturer's recommendation. Cultures were grown by incubation overnight at 37 C with aeration at 200 rpm.

Assessing the antimicrobial effects of selected potential inhibitors
The antimicrobial assay was carried out as discussed by Balouiri et al. (2016) to assess the possible antimicrobial effects of lovastatin and simvastatin. Briefly, lovastatin and simvastatin stock solutions were prepared by dissolving 20 and 25, mg respectively, in 1 ml of dimethyl sulfoxide (DMSO). Stock solutions were filter sterilized and stored at 30 C until further usage. Stocks of both the potential inhibitors were used within a week and then replaced with fresh stocks. The antimicrobial effects of potential inhibitors on B. subtilis strain MCC2049 (T) were determined by monitoring bacterial growth (measured with an optical density at 600 nm) at regular time intervals. Overnight grown primary culture (with OD 600 of 1.5) was added to 20 ml of LB medium at 1% or 4% (v/v) inocula, and cultures were incubated at 37 C with aeration at 200 rpm. For minimum inhibitory concentrations (MICs) calculations, inhibitor dose responses were tested over 6.25 À 100 mg/ml with intermediate concentrations used with a twofold increase in the concentration. The lowest concentrations of potential inhibitors that inhibited bacterial growth were recorded as MIC. The non -inoculated media were taken as abiotic control, whereas media inoculated with B. subtilis in the absence of inhibitor were taken as the positive controls for bacterial growth. Samples were collected at regular intervals of every 2 h for the total duration of incubation of 10 h.

Assessing the antibiofilm effect of selected potential inhibitor
The assay for biofilm formation by B. subtilis and its inhibition in the presence of selected potential inhibitors was carried out using the protocol described earlier (Hamon & Lazazzera, 2001). Briefly, the biofilm growth media (BGM) containing LB medium supplemented with 0.15 M ammonium sulfate, 100 mM potassium phosphate, 34 mM sodium citrate, one mM magnesium sulfate, and 0.1% glucose was used for biofilm formation assay. Culture medium was taken in 24 well plates (HiMedia, Mumbai, MH, India) at the total culture volume of 2 ml and inoculated with 1% of overnight grown primary culture. Inoculated multi-well plates were incubated at 30 C under static conditions. Increasing concentrations of potential inhibitors were added to BGM to attain the working concentrations of 5 À 25 mg/ml with regular increments of 5 mg/ml. Alternatively, biofilm formation and inhibition were also assessed in 15 ml glass test tubes. Biofilm growth medium was added to glass test tubes to the total volume of 5 ml; other parameters, e.g., inoculum density, the concentration of the potential inhibitors, were identical to those used in 24 well-plates assays. Cultures were incubated at 30 C for 48 h without agitation. Cultures with adequate volumes of DMSO were taken as the solvent control group.
At the completion of incubation, a quantitative assessment of biofilm formation was done according to the method described earlier (Ommen et al., 2017). Briefly, cells attached to the vessel wall and surface were washed with distilled water, fixed with 100% methanol, and subsequently stained with safranin. The excess amounts of safranin were washed with distilled water, and bounded safranin was eluted with 30% glacial acetic acid followed by estimation of Optical Density at 492 nm by spectrophotometer (Halo DB-20, Dynamica Scientific Ltd. Livingston, UK) or ELISA plate reader (MULTISKAN GO, Thermo Fisher Scientific, MA, USA). The non-inoculated BGM was taken as abiotic control; BGM inoculated with 1% overnight growth cultures of B. subtilis without any inhibitor was taken as a positive control for biofilm formation. Incubations were carried out in triplicates, and data collected were subjected to appropriate statistical analyses.
For qualitative assessment of the biofilm inhibition, B. subtilis biofilms were grown according to the procedure described earlier. Briefly, 1% (v/v) of the overnight grown primary culture of B. subtilis was inoculated on microscopic glass slides (submerged in BGM using 96 mm Petri-plates) in the presence or absence of 25 mg/ml of the potential inhibitors. These plates were incubated for 48 h at 37 C. After incubation, glass slides were treated with safranin staining followed by washing with distilled deionized water and air drying at room temperature for 15 min. Subsequently, the slides were observed under the bright field light microscopy using Axio-Lab A1 Compound Microscope (Carl Zeiss, Germany) at 40X and 100X magnification.
Assessing the effect of the selected potential inhibitor on pre-formed biofilms To assess the effect of selected potential inhibitors on the disintegration of pre-formed biofilms, the mature biofilms were allowed to form according to the method described above in the absence of the inhibitors with incubation at 30 C for 24 À 48 h. Successful formation of biofilm was assessed by visual observation. At this point, 25 mg/ml of the potential inhibitor were added to the pre-formed biofilms and further incubated for additional 24 h, at 30 C. Sample left untreated and treated with appropriate volumes of DMSO were taken as the negative control and solvent control, respectively. After completion of additional incubation of 24 h, glass slides were treated with safranin staining, washed with distilled deionized water, air-dried at room temperature for 15 min, and then observed under the microscope.

Statistical analysis
Statistical analysis, e.g., calculation of average mean values, standard deviations, comparison of the control and test groups using one-way ANOVA and regression analyses, etc., were performed with GraphPad Prism. The p values of < 0.0001, indicating statistical significance of the variation, were represented with Triple Stars ( ÃÃÃ ). For one-way ANOVA, the data comparison analyses were carried out using Tukey's Multiple Comparison Test. The statistical analyses of the dose-response of the various concentrations of the inhibitor for inhibition of biofilm formation for assessment of IC 50 were also carried out with Non-linear regression (curve fit) using GraphPad Prism. Each experiment carried out during this study was performed using nine replicates, and data values were recorded in the form of mean ± standard deviation.
A graphical representation showing a flowchart of the scheme used for homology modeling of TasA  , virtual screening for selection of potential inhibitors, and experimental validation of selected potential inhibitors is presented in supplementary Figure S1.

Results
Molecular modeling and molecular dynamics simulation of TasA  The target protein (TasA) 's primary amino acid sequences were retrieved from NCBI and subjected to PDB BLAST analyses. The PDB BLAST analyses showed that TasA has the highest sequence similarity with a conserved exported protein from Bacteroides fragilis (PDB ID: 3CLW) (Bonanno et al., XXXX) and D-isomer specific 2-hydroxy acid dehydrogenase from Lactobacillus delbrueckii ssp. Bulgaricus (PDB IDs: 2YQ4) (Holton et al., 2013). A total of twenty 3-D models for TasA  were generated with homology modeling. Each model had unique Discrete Optimized Protein Energy (DOPE) scores. The models were compared for their DOPE scores, and the best model was selected based on the lowest DOPE scores. The model with DOPE scores of À 56877.059 was found to be the best, and it was selected for quality assessment and model refinements. The refined model is shown in Figure S2A. It was subjected to validation with Ramachandran Plot Analysis. Results from this analysis showed that most of the amino acid residues lie in the favored, allowed, and generously allowed regions and only 3.5% residues were found to lie in the disallowed region. This observation suggested that the selected model is valid and it could be used for subsequent studies.
As an additional step for validating the selected model and assessing its stability, it was subjected to extended molecular dynamics simulations performed over a time scale of 100 ns. During MD simulation, nearly all the residues remained in the most favored or additionally allowed or generously allowed regions, and only 0.9% of residues were found to be in the disallowed region. The structural integrity of most of the regions was sustained during molecular dynamics simulation except for the residues Glu-50, Ser-77, Val-116, Leu-140, Tyr-181, His-242, and Asn-252 ( Figure S2B). Results obtained from MD simulations were subjected to position restraining for 100 ns, and RMSD of the backbone atoms was calculated over the entire duration of the MD simulation. The RMSD values for the selected model of TasA  range between 0.3 and 1 nm ( Figure S2C). These values indicated stable trajectories of the selected model of TasA (28 À261) during extended MD simulation. The RMSF values also corroborated this observation as the RMSF values ranged between 0.3 and 1.5 nm ( Figure S2D). One of the noteworthy observations with RMSF calculation during MD simulations was that the residues present at the N terminal and C-terminal of TasA  showed greater values for RMSF, suggesting that these residues are more flexible compared to other regions of the modeled protein.
Virtual screening for selection of potential inhibitor(s) of TasA  To explore the possibility of virtual screening of small-molecule inhibitors of TasA, the putative binding sites of TasA  were predicted with the help of the SiteMap tool available in the Schrodinger suit. The binding pockets were defined only when the threshold distance was 4 Å from the nearest site point. The identified sites with a site score ! 0.5 and D-score ! 0.5 were considered as potentially targetable pockets. Binding pocket identification was also performed with the PDB ID -5OF2 (corresponding to the globular, non-processed form of TasA). These analyses led to identifying 4 and 1 binding pockets no the TasA (28 À 261) model and PDB ID -5OF2, respectively. The druggable probability of each of these binding pockets was found to be greater than 0.5. Other characteristics of the binding pockets were also suitable for small molecule inhibitor mediated inhibition. The list of the detailed characteristics of the identified binding pockets is presented in Table S1.
After predicting the binding pockets/druggable sites in the modeled structure of TasA  and PDB-ID structure, the structure-guided virtual screening was performed with 3180 FDA-approved drugs. Results from preliminary screening showed that 280 compounds could dock within predicted binding sites when selected with 3 rounds of successive selections, using the cut-offs of selecting only the top 50% compounds during each selection. The dock-Score and glide-Score for the 280 screened compounds complexed with modeled TasA  were À10.60 to À6.40 and À10.87 to À6.40, respectively. These scores are within the acceptable range for protein-ligand or protein-inhibitor interaction (Meng et al., 2011;Richard et al., 2004). In contrast, the dockscore and glide-score for screened compounds complexed with TasA crystal structure, i.e., PDB ID 5OF2, were found to be À 8.45 to À4.33 8.45 to À 4.83, respectively. These values are slightly higher than the acceptable limits for a stable protein-inhibitor complex. The values for binding characteristics were reassessed after refinement of docked complexes of potential inhibitors with modeled TasA  and TasA crystal structure (PDB ID 5OF2). The refinements were done to calculate Molecular Mechanics Generalized Born Surface Area (mmGBSA) binding energies. It was found to be in the range of À103.63 to À24.42 Kcal/mol for TasA  and À81.31 to À10.55 Kcal/mol for PDB ID 5OF2, respectively. Again, these values are considered suitable for protein-ligand or protein-inhibitor interactions (Genheden & Ryde, 2015).
At this point of the study, the top 10 potential inhibitors were selected for ranking and further analyses based on the binding energy properties (i.e., mmGBSA binding energies, dock-scores, glide-scores, and XP-Scores. The list of binding energy properties obtained for the top 10 potential inhibitors complexed with modeled TasA (28 À261) is presented in Table  1. Similarly, a list of binding energy properties obtained for the top 10 potential inhibitors complexed with TasA crystal structure (PDB ID 5OF2) is presented in Table 2. Amongst the top 10 potential inhibitors, simvastatin, epoprost, and lovastatin exhibited the lowest binding energies. Therefore, these inhibitors were selected for further studies.
The commercial supply of the selected top 3 potential inhibitors was surveyed from prospective vendors and API manufacturers. While simvastatin and lovastatin could be procured from a reputed fine chemical OEM, i.e., Tokyo Chemical Industry (TCI), Chennai, India; however, epoprost could not be obtained. Subsequently, the binding characteristics of available potential inhibitors simvastatin and lovastatin were determined for possible binding at the alternative druggable sites predicted on the modeled TasA (28-261) and the single docking site on TasA crystal structure (PBD ID 5FO2). Results obtained from this analysis indicated that the docking scores and mmGBSA for complexes formed at alternative druggable sites (i.e., sites 2, 3, 4 on TasA (28-261) and  TasA crystal structure (PBD ID 5FO2) were higher than the scores obtained for the complex formed at druggable site no. 1 of TasA (28-261) ( Table 3). This observation suggested that the complexes formed by docking of simvastatin and lovastatin at alternative docking sites would be less stable. Apart from determining the binding characteristics, results from docking analyses of potential inhibition -TasA  and TasA crystal structure (PBD ID 5FO2) were also used for defining the amino acid residues that might be involved in protein-inhibitor interaction. It was observed that Glu232 and Lys245 present at the docking site-1, and Gly117, Lys118, and Asn220 present on site-2 of TasA  are involved in hydrogen bond formation with a hydroxyl group and oxygen atom of simvastatin. In contrast, site-3 on TasA (28-261) was found to have no hydrogen bond interaction with simvastatin, and site-4 did not form a stable complex with simvastatin. A graphical representation showing interactions in simvastatin -TasA (28-261) docked complexes for sites 1-3 is presented in Figure 1A-C, respectively.
The interactions of lovastatin with TasA (28-261) was found to be comparatively more stable, and it was mediated via hydrogen bond formation by Ser142, Lys176, Thr243, and Lys245 of site-1; Asp91 and Asn220 of site-2, Glu À103 on site-3, and Asn122 of site-4 ( Figure 1D-G). In the case of TasA crystal structure (PDB ID 5OF2), Thr189 was observed to be involved in hydrogen bond formation with simvastatin, whereas Ala 178 and Asp182 were responsible for hydrogen bond formation with Lovastatin ( Figure 1H -I).

Molecular dynamics simulations of TasA (28-261) docked with potential inhibitors
To confirm the thermodynamic stability of docked complexes of simvastatin and lovastatin with TasA  and 5OF2, all of the top 10 potential inhibitors were initially subjected to molecular dynamics (MD) simulation studies using the time scale of 10 nanoseconds. This analysis suggested that only the top 3 potential inhibitors may form stable complexes with TasA (28-261) during MD simulations. The RMSD values of these complexes showed a maximum deviation of up to 0.64 nanometers during MD simulations carried out for 10 nanoseconds ( Figure S3A). The RMSF values for these complexes also showed only minor fluctuations, as indicated by values ranging between 0.08 and 0.6 nm ( Figure S3B). Results obtained from MD simulations also showed that, although Epoprost makes a stable complex TasA (28-261) , its stability is comparatively lesser than the complexes formed by either simvastatin or lovastatin. It was also observed that complexes formed by any 3 potential inhibitors with TasA crystal structure (PDB ID 5OF2) were not stable in MD simulations. These complexes were also subjected to calculation of solvent accessible surface area (SASA) values and potential energy of docked complexes. Results from these calculations indicated that the complexes formed by the top 3 potential inhibitor had very similar SASA values ( Figure S3C) and potential energies ( Figure S3D).
To further validate these results, MD simulations were repeated with complexes formed by simvastatin and lovastatin with all of the four predicted binding pockets of the TasA (28-261) over the time duration of 50 ns. Importantly, results obtained with MD simulations carried out for 50 ns corroborated with those results obtained with MD simulations carried out for 10 ns. To highlight, the complex formed by simvastatinsite 1 of TasA (28-261) exhibited the most stable trajectories and showed the lowest RMSD values ranging between 0.0004 and 1.1 nm (Figure 2A). On the other hand, complexes formed by simvastatin -site 2 of TasA  showed the highest RMSD values of 9.8 nm (Figure 2A). This complex was rather unstable towards the later time points (38 À 45 ns) of MD simulation (Figure 2A). The complex formed by simvastatinsite 3 of TasA (28-261) was quite stable up to 45 ns when it suddenly broke ( Figure 2A). The calculations of RMSF values were found to agree with the assessment of RMSD values; the overall fluctuations of simvastatin site 1 of TasA  were observed to have minor fluctuation (ranging from 0.09 to 0.5 nm) in comparison to other complexes formed by binding of simvastatin to other docking sites of TasA (28-261) ( Figure 2B). Noticeably, the RMSF values for simvastatinsite 3 of TasA  were highest (ranging from 0.08 to 0.9 nm).
To further validate the observed MD simulations trajectories, analyses, the docked complexes of simvastatin -TasA  were subjected to the calculation of the average radius of gyration (Rg) values and SASA values. The Rg values indicate the compactness of the docked inhibitor-protein complex, whereas the SASA value indicates the energy of solvation at the site of complex formation. These values reflect upon the stability of the complex; the lower the values for Rg and SASA, the more stable to complex(s) would be. The average Rg value for the simvastatinsite 1 of TasA  was found to be 1.9 nm. This indicates low to minor conformational changes in protein during its interaction with the docked inhibitors. The average Rg value for complexes formed by simvastatin with site À2 and site 3 was even lesser (i.e., 1.8 nm) ( Figure 2C). The values for SASA were found to be comparable amongst all the complexes formed by simvastatin with site 1, site 2, and site 3 of TasA (28-261), respectively. The average values were 168 nm 2 , 156 nm 2 , and 158 nm 2 for complexes formed ( Figure 2D). The extended MD simulations were also performed for complexes formed by docking of lovastatin on different binding sites of TasA  . The RMSD values obtained for different complexes formed by lovastatin -TasA  suggested that sites 1, 3, and 4 were quite stable, while the complex formed at site 2 was unstable ( Figure 2E). The RMSF of the Ca-atom of each docked complex was also calculated. Interestingly, the RMSF values for complexes formed at different sites were found to be rather similar and showed maximum average fluctuations of 0.85, 0.75, 0.68, and 0.81 for the four sites, respectively ( Figure 2F). It was also observed that the major fluctuations occur at amino acid residues close to the protein's Nterminus and Cterminus. The amino acid residues present close to the central core of the protein (amino acid residues 80 À 200) were observed to be quite stable and had minor fluctuations ( Figure   2F). The average Rg values for lovastatin-TasA (28-261) complexes were 1.89, 1.73, 1.89, and 1.82 nm, respectively, for the four binding sites of TasA   (Figure 2G). These Rg values are slightly higher than those observed for simvastatin -TasA  complexes. The values for SASA of lovastatin -TasA (28 À261) complexes were observed to be smaller (i.e., 138, 136, 134, and 138 nm for complexes formed at site-1, site-2, site-3, and site-4, respectively) ( Figure 2H). Such lower values for RMSD, RMSF, Rg, and SASA for of selected potential inhibitors on docking sites of TasA  strongly suggested that the complexes formed by identified potential inhibitors -TasA (28-261) would be thermodynamically stable; therefore, these potential inhibitors may. The MD simulations were also performed with the APO form of modeled TasA  . The comparison of RMSD trajectories obtained for APO form were significantly more stable and had deviation only in the range of 0 À 0.08 nm during the 50 nm (50,000 time steps of 1 ps each). In comparison the RMSD trajectories for Holo forms i.e., TasA  bound to either lovastatin or simvastatin indicated for much higher deviations ( Figure S4). Table 3. Docking properties of complexes formed by docking of simvastatin and lovastatin at all of the predicted binding sites of TasA    Results from MD simulations analyses at 50 ns showed that both the selected potential inhibitors (i.e., simvastatin and lovastatin) might form the most stable complexes with druggable site no. 1 of TasA  . The complexes formed by simvastatin and lovastatindruggable site no. 1 of TasA  were scrutinized with extended MD simulations carried out for 100 ns. The RMSD values in the case of simvastatin bound site 1 of TasA  were in the range of 0 to 16 Å ( Figure 3A). The RMSF values for this complex were found to corroborate with the RMSD values and showed maximum average fluctuations of 15 À 16 Å ( Figure 3B). In contrast, the RMSD values for complex formed by lovastatinsite 1 of TasA  in extended MD simulations were observed to be in the range of 0-8.4 Å ( Figure 3C). The RMSF values for these complexes were observed to have a maximum fluctuation of 7.8 Å ( Figure 3D). Results obtained from extended MD simulations strongly indicated that the simvastatin -TasA (28-261) complex does not have stable trajectories during extended MD simulations; therefore, it could be suggested as a potential candidate/inhibitor of TasA  and Bacillus subtilis biofilms only with further studies. These results also indicated that at the computational analyses level, only lovastatin might form a stable complex with target protein TasA (28-261), and it could be claimed as a potential inhibitor for further validation with in vitro studies/wet-lab experiments.
The stable 2-D docking poses were obtained for simvastatin and lovastatin when complexed with druggable sites 1 of TasA  and assessed over different time scales (10, 50, and 100 ns) of MD simulations. The interaction of simvastatin with TasA (28-261) was mediated through hydrogen bond formation by Lys144, Lys176, Gln232, and Lys245 at 10, 50, and 100 ns of molecular simulation studies ( Figure 4A-C). Similarly, lovastatin bound TasA (28-261) through amino acid residues Ser142, Lys144, Lys176, Gln232, and Lys245 interact by hydrogen bond formation at the different time scales of molecular simulations (Figure 4-F).

Conformational transitions assessed with PCA analyses, co-variance analyses and FEL calculation
To assess the conformational transitions of simvastatin and lovastatin bound TasA  , the Principal Component Analyses (PCA) was performed on results obtained with Molecular Dynamics simulation of the complexes. The PCA of equilibrated trajectories demonstrated that the complexes formed with simvastatin and lovastatin at different sites on TasA  had significantly high coverage of the eigenvectors. More than 98% of total fluctuations were found to be governed by the initial 10 PCs ( Figure 5A,B). These results indicate that the overall dynamics of simvastatin or lovastatin bound to TasA  are regulated by high frequency & low magnitude motions that probably cause the easy conformational transition. The conformational transitions observed for the unbound form of TasA (28-261) were found to have relatively lower levels of coverage with the eigenvectors as only 83.4% of total fluctuations were found to be governed by the initial 10 PCs.
Further, the correlation matrix plot obtained from co-variance analyses showed how the C-a atom of the backbone of the TasA (28-261) moved relative to each other upon binding with simvastatin or lovastatin. The motions were either positively correlated (moved in the same direction) or negatively correlated (moved in opposite direction). Results obtained from co-variance analyses for simvastatin bound to three sites of TasA  along with the positive and negative limits of co-variance are presented in Figure S5A-C. Results of the co-variance analyses for lovastatin bound to four sites are presented in Figure S5D-G. It was observed that the anti -correlation motions were dominant in all interaction with maximum anticorrelation observed in case of simvastatin bound to the site 2 ( Figure S4B).
Results obtained for deviations in protein motions observed with PCA and co-variance analyses were assessed by Gibbs Energy Landscape (GEL) analysis. The sub-conformational changes of simvastatin or lovastatin bound to druggable sites were plotted against two principal components PC1 and PC2 to generate the 2 D plot of the GEL diagrams. Results obtained with this analysis are presented in Figure  6A-C for complexes formed upon binding of simvastatin with sites 1, 2, 3 and Figure 6D-G for complexes formed and lovastatin with sites 1, 2, 3, 4 of TasA  , respectively. These GEL plots confirmed that with binding of either simvastatin or lovastatin with TasA (28-261) , there is adaptation of one main low-energy region by the backbone atoms. From this observation, it could be interpreted that during the interaction with potential inhibitors, there were only limited structural variations in the protein and these variations were restricted to a certain local structural element only. On a comparative account, it could be suggested that complexes formed by simvastatin and lovastatin -TasA (28-261) interaction quite similar. The GEL values for simvastatin -TasA (28-261) ranged from the lowest of 0 KJ/mol to highest of 13.6 KJ/ mol; whereas the values for lovastatin ranged from the lowest of 0 KJ/mol to highest of 15.8 KJ/mol. These values effectively corroborate with and validated the results obtained with PCA and co-variance analyses.
As this point of the study, it was reasoned that results obtained with various in silico analyses sufficiently advise that the identified potential inhibitors i.e., simvastatin & lovastatin may cause successful inhibition of formation and maturation of B. subtilis biofilm. Therefore, in vitro experiments were carried for determination of 'anti-biofilm' effects of simvastatin and lovastatin on biofilm formation by B. subtilis.

Growth inhibition of planktonic cells of B. subtilis by simvastatin & lovastatin
A few of the statins have been previously reported for potential antimicrobial activities; therefore, it was necessary to test the selected statins for antimicrobial activities on B. subtilis. Otherwise, the 'antimicrobial' activity of selected statins might be wrongly interpreted as 'anti -biofilm' in biofilm inhibition assay. The antimicrobial activities of selected inhibitors against B. subtilis were evaluated through 'bacterial   growth kinetics assay', wherein increasing concertation of the simvastatin and lovastatin were tested over concentration range of (6.25 À 100 mg/ml). The antimicrobial effects were determined by assessing inhibition of bacterial growth measured in terms of OD 600 . Results obtained with this analysis suggested that both simvastatin and lovastatin have antimicrobial on planktonic forms of B. subtilis at concentrations > 50 mg/ml. They exhibit IC 50 values of $ 35.94 and 40.72 mg/ml, respectively ( Figure 7A,B). This observation that simvastatin and lovastatin exhibit antimicrobial activities on B. subtilis at concentration > 50 mg/ml is comparable with a previous report wherein the MIC of simvastatin was found to be in the range of 26.04 À 166.67 mg/ml against tested standard bacterial strains i.e., E. coli, P. aeruginosa, S. pneumoniae, A. baumannii, P. mirabilis, K. pneumoniae, S. pyogenes, and MSSA, MRSA, VSE, VRE, etc (Masadeh et al., 2012). The molar values for IC 50 values of simvastatin and lovastatin against B. subtilis were calculated to be 86.28 and 100.8 mM ( Figure 7C,D).

Inhibition of B. subtilis biofilm formation by simvastatin & lovastatin
The 'anti-biofilm activity' assays were carried out at sub-MIC values using 5, 10, 15, 20 and 25 mg/ml of working concentrations of simvastatin and lovastatin using standard safranin staining. Results obtained from this assay showed that both simvastatin and lovastatin exhibit antibiofilm activity against B. subtilis; wherein > 95% percent and > 90% biofilm reduction was observed at 25 mg/ml of simvastatin and lovastatin, respectively ( Figure 8A,B). Noticeably, at this concentration both Simvastatin and Lovastatin did not adversely affect the bacterial growth. Therefore, it could be argued that the inhibition of biofilm formation by B. subtilis by this concentration of simvastatin and lovastatin occurs through biofilm specific mechanism rather than through non-specific inhibition of Bacillus cell growth. The efficiency of biofilm inhibition was observed to be concentration dependent as statistically significant biofilm inhibition was also observed at concentrations of 15 and 20 mg/ml. At lower concentrations (5 and 10 mg/ml) of the inhibitors, the biofilm inhibition was observed but it was statistically non-significant ( Figure 8A,B). Determination of IC 50 for B. subtilis biofilm inhibition activities of simvastatin and lovastatin showed the values of 12.06 and 9.73 mg/ml, respectively. These values correspond to 28.86 and 24.10 mM ( Figure 8C,D). The qualitative analyses of biofilm inhibition with Simvastatin and Lovastatin corroborated well above observation as significant biofilm reduction was seen in treated group as compared to the control groups.

Disintegration of pre-formed B. subtilis biofilm by simvastatin & lovastatin
Disintegration of pre-formed biofilms has been regarded as one of the desired characteristics of any anti-biofilm therapeutics for its application to mitigate the resistant forms of Figure 8. Effect of sub-MIC doses of (a) compound 1 and (b) compound 2 upon percentage of Biofilm formation in Bacillus subtilis. Statistical analysis was done one-way ANOVA and ÃÃÃ express p value ( 0.0001 (n ¼ 9) (C,D) represents the effect of simvastatin and lovastatin concentrations onto the B.subtilis biofilm formation. The concentration response curve was generated in n ¼ 9, and each point represents the mean value and standard deviation.
the biofilm associated microorganisms. Therefore, during the present study, we also assessed the effect of simvastatin and lovastatin on preformed-mature biofilms. The results from microscopic observation clearly indicated that treatment with the inhibitors resulted in disintegration of biofilm matrix, whereas biofilm disintegration in untreated or solvent treated control groups were non-significant ( Figure 9A-D). Between simvastatin and lovastatin, the latter showed a stronger biofilm disintegration effect ( Figure 9C,D). Antimicrobial and antibiofilm activity of different statins have been reported previously (Masadeh et al., 2012;Jerwood & Cohen, 2008;Graziano et al., 2015); however, to the best of our knowledge, their ability to disintegrate previously formed biofilms has not been reported. .

Discussion
Microbial biofilms have been recognized as a major cause of concern, specifically with regards to chronic microbial infections. Additionally, biofilm associated infections are often difficult to diagnose; therefore, biofilms also limit the selection of the appropriate choice of treatment6 (R€ omling & Balsalobre, 2012). Consequently, several studies have been carried out in recent past for discovery and development of therapeutic strategies for biofilm inhibition and disintegration (Rabin et al., 2015). Noticeably, most of these studies have reported successful inhibition of microbial biofilm by using natural extracts, bioactive secondary metabolites, metallic and polymeric nanoparticles selected through large scale high throughput screening bioassays (Lu et al., 2019;Qvortrup et al., 2019;Kayumov et al., 2014). However, despite success reported with such studies, till date there is no FDA approved 'anti-biofilm therapeutics' for human application. It could be argued that the lack of comprehension about the (i) mode of action; (ii) potential drug target(s); and (iii) the possible off target activities of the inhibitors selected through screening bioassays are the major reasons for limitation (R€ omling & Balsalobre, 2012). On the other hand, relatively fewer studies have been carried out wherein anti-biofilm therapeutics have been selected on the basis of mechanistic understanding (e.g., identification of different druggable target) of biofilm formation (Lu et al., 2019). Furthermore, targeted inhibition of biofilm ECM associated protein(s) has remained elusive as an approach for screening, selection and characterization of potential 'antibiofilm strategies. The non-availability or relatively meager understanding of the 'structurefunction' characteristics of biofilm associated proteins has hampered the progress in this direction. The complex molecular nature of biofilm ECM associated proteins has delayed the progress in development of ECM protein target specific 'anti-biofilm therapeutics'.
During the present study, we ventured into structure/ model driven screening of specific inhibitor(s) of biofilm formation associated protein(s) of B. subtilis. For the purpose, we selected TasA 28-261, the major proteinaceous component of B. subtilis biofilm and previously reported to be pivotal in biofilm formation by B. subtilis) as the potential druggable target protein. Through initial in silico analyses carried out with virtual screening, we found that simvastatin and lovastatin could be potential inhibitors of TasA 28-261 . Noticeably, results from subsequent extended molecular dynamics simulation analyses showed that the RMSD and RMSF values for simvastatin -TasA (28-261) complex were significantly higher than those obtained for lovastatin -TasA  complex. This observation indicated that the complex formed by simvastatin -TasA (28-261) interaction may not be as stable as the complex formed by lovastatin -TasA (28-261) interaction. Therefore, it could be suggested that on the basis of results obtained from computational analyses carried out during present study, simvastatin could only be suggested as a potential candidate for further studies, whereas lovastatin could be claimed as the potential candidate for further wet lab validation studies for inhibition of TasA  and B. subtilis biofilm formation. In spite of the observation that simvastatin may not form as stable complex with TasA  , both lovastatin simvastatin should be assessed for anti-biofilm activity in whole cell based biofilm inhibition studies.
Results from in vitro biofilm inhibition assay studies showed that both simvastatin and lovastatin have inhibitory effect on growth of planktonic form as well as biofilm formation by the adherent form of B. subtilis. Amongst simvastatin and lovastatin, the later was found to be more efficient for inhibition of biofilm formation as shown by lower IC50 of biofilm inhibition. In comparison, the biofilm inhibition IC50 for simvastatin was slightly higher. This observation is consistent with results obtained from in silico analyses, specifically the results of extended MD simulations carried out for 100 ns. The observation that in spite of forming a less stable complex with TasA (28-261), simvastatin could inhibit B. subtilis biofilm formation indicates for the possible involvement of additional mechanism/target through which simvastatin mediates biofilm inhibition. Determination of this mechanism will be quite pertinent and it will provide vital information for future studies. We propose to carry out future studies to determine the potential offtargets of simvastatin. Results from this study may provide essential insight into the alternative mechanism for anti-biofilm of simvastatin.
Meanwhile, the observations that both simvastatin and lovastatin inhibited biofilm formation by B. subtilis are consistent with earlier reports wherein microbial growth and microbial biofilm formation were shown to be inhibited with 'cholesterol lowering drugs' including statin and its derivatives (Jerwood & Cohen, 2008;Masadeh et al., 2012). Nevertheless, such reports are rather few in number and they are quite discreet. Most of these reports have emerged from large scale screening-based bioassays. Therefore, such studies are similar to other studies that have reported inhibition of B. subtilis biofilm formation with screened inhibitors e.g., furanone derivatives, parthenolide and D-tyrosine etc. selected through large-scale screening bioassay (Kayumov et al., 2015;Yu et al., 2016). Results from screening bioassay are quite important, yet they may be significantly difficult to pursue further development and implementation of 'anti-biofilm strategies. On the contrary, results obtained during the present study show that inhibitors selected through in silico studies and validated through in vitro studies could achieve inhibition B. subtilis biofilm formation and caused disintegration of pre-formed of biofilm, through potential inhibition of TasA  . Furthermore, these inhibitors also induced disintegration of the pre-formed biofilm of B. subtilis.
Although, the precise mode of action for anti-biofilm activity of simvastatin and lovastatin remains to be elucidated, yet on the basis of results presented in this study, it could be argued that these inhibitors bind to one of the possible druggable sites of TasA (28-261) causing significant change in its physiological function. The most important physiological activity carried out TasA 28-261 with respect to biofilm formation is its proteinprotein interaction with accessory protein TapA 33-253 which results in formation of amyloid like structures. It could be proposed that simvastatin and lovastatin potentially interfere with the proteinprotein interaction through allosteric mechanism and prevent amyloid formation. Further studies would be needed to evaluate this hypothesis. Importantly, apart from the basic studies on determination of the molecular mechanism involved in inhibition of biofilm formation by simvastatin and lovastatin, these drugs may also be tested as 'anti-biofilm' treatment for taxonomically related, antibiotic resistant pathogenic bacteria.

Conclusions
In conclusion, it is proposed that the results obtained during the present study clearly demonstrate that the FDA approved drugs/small molecular inhibitors (simvastatin and lovastatin) selected through in silico analyses (molecular modeling, virtual screening, molecular docking and dynamics etc.) for targeting biofilm ECM associated major protein (viz, ) successfully inhibited the biofilm formation. We propose that similar strategies could be used as a simple yet reliable tool for identification of novel anti-biofilm therapeutics against biofilm forming pathogenic microorganisms.

Disclosure statement
No potential conflict of interest was reported by the authors.

Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.