Targeting bacitracin resistance-associated histidine kinase (BraS) in MRSA with isatin and indazole derivatives

Abstract

The BraSR two-component system plays a crucial role in bacitracin sensing and resistance in Staphylococcus aureus, making the BraS histidine kinase an attractive antibacterial target. In this study, we applied a structure-based drug discovery approach to identify ATP competitive isatin and indazole based hits targeting the BraS kinase domain. A two-stage virtual screening workflow, involving molecular docking, molecular dynamics simulations, MM-PBSA freeenergy estimation, and interaction analyses was used to prioritize compounds with favourable relative binding profiles within each series. Computational analyses suggested that several isatin and indazole derivatives adopt stable binding modes within the ATP-binding pocket of BraS and exhibit favourable predicted binding energetics compared to other screened compounds. Based on these in silico rankings, three isatin and three indazole derivatives were selected for chemical synthesis and experimental evaluation. Interactions between the test compounds and the recombinant BraS kinase domain were evaluated using microscale thermophoresis. Several compounds including IS7, IS12, IN16 and IN21 shows measurable binding affinities in the nanomolar range. The isatin-series compound IS12 exhibited a dissociation constant (Kd) of 2nM while IN21 from the indazole series exhibited a Kd of 251nM against the BraS kinase domain. In parallel, selected compounds were evaluated for their effects on biofilm formation using a crystal violet assay. Although a reduction in biofilm biomass was observed at specific concentrations, these effects are interpreted as indirect modulation of regulatory pathways associated with BraSR signalling rather than direct antibiofilm activity. Overall, this study identifies isatin and indazole derivatives as promising ATP-competitive hits that bind to the BraS kinase domain and provides a framework for further structure-guided optimization. However, functional inhibition, cellular efficacy, and pharmacological properties remain to be established through future biochemical and in vivo studies.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-49362-w.

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) is a superbug responsible for wide range of infections including those affecting the skin, soft tissues, bones and joints, the respiratory system, the urinary tract, as well as hospital-acquired infections and infective endocarditis. MRSA spreading in the bloodstream may trigger septicaemia, which can be lethal if not treated appropriately¹. MRSA infections are difficult to treat due to antibiotic resistance, including resistance to β-lactams, cephalosporins and vancomycin, commonly used as first-line treatments against staphylococcal infections². Bacitracin has been used for many years to treat different types of skin diseases and is also utilized as an alternative treatment option for MRSA infections³. Bacitracin is an antimicrobial polypeptide that interferes with the dephosphorylation of lipid carriers like isoprenyl pyrophosphate and disrupts peptidoglycan synthesis in Gram-positive bacteria⁴. The increasing prevalence of bacitracin-resistant staphylococcal species represents an important global health problem, necessitating the investigation of novel strategies and molecules against these drug-resistant strains as conventional antibiotic choices are becoming less effective^5,6. Targeting the bacitracin-resistance associated TCS and drug efflux could be a viable strategy to address drug resistance in staphylococci.

One of the mechanisms that helps MRSA acquire bacitracin resistance involves the activation of the BraS histidine kinase, which is part of the bacitracin-resistance associated BraS/BraR two-component signal transduction system (TCS)⁷. This system detects bacitracin and expels it from bacterial cells using the ABC transporters BraDE and the efflux pump VraDE. When BraS/BraR TCS senses low antibiotic concentrations, it activates two operons, encoding the ABC transporter braDE and the efflux pump vraDE. BraDE assists BraSR TCS in bacitracin sensing and signaling, while VraDE primarily functions as a detoxifying module, conferring tolerance to bacitracin and nisin⁸(Fig. 1).

Fig. 1.

Bacitracin resistance associated BraSR/BraDE/VraDE module in MRSA. The two-component system BraSR having BraS HK which consists of transmembrane helices, DHp domain and CA domain and BraR response regulator, made up of receiver domain and effector or dna-binding domain. ABC transporter BraDE helps in sensing and signaling of bacitracin while VraDE efflux system removing antibiotics (bacitracin and nisin) from staphylococci and lowering the intracellular levels.

The molecular processes behind signal sensing, signal transduction and activation of the responsible gene transcript needs an in-depth study. It is feasible to disrupt this pathway and reverse bacitracin resistance by targeting and inhibiting BraS histidine kinase. In this context, studies on developing small substrate like kinase inhibitors have gained traction among researchers to propose diverse hits in drug pipeline. Studies on the derivatives of isatin and indazole have exhibited anti-bacterial properties and have been explored in the context of developing novel medications^9–11.

Isatin and indazole derivatives are an important class of heterocyclic compounds that can be exploited as therapeutic precursors^11,12. In various reports, such compounds have been extensively explored as inhibitors of kinases, which regulate signalling pathways and physiological processes such as gene expression, metabolism, and cell division. Targeting these kinases has shown the potential to impair bacterial growth and survival^13–15. Radwan et al. synthesized a new series of isatin derivatives and tested their anti-bacterial activity through in vitro and in silico study and reported that 3-hydroxy-3-(2-(2,5-dimethylthiophen-3-yl)-2-oxoethyl)-in-dolin-2-one was effective against WalK HK in MRSA¹⁶. A study by Czelen et al. focused on the inhibition of the cyclin-dependent kinases by isatin derivatives reported that certain isatin derivatives displayed significant inhibitory activity against these kinases and role as an antitumor agent¹⁷. Watanabe et al. investigated the mechanism by which isothiazolones inhibit P. aeruginosa AlgR2/AlgR1 TCS and discovered that isothiazolone derivatives inhibit the autophosphorylation activity of AlgR2 HK and displayed significant inhibition of alginate production, which serves as a crucial virulence factor for this bacterium’s survival¹⁸. Similarly, several indazole derivatives have also been studied as kinase inhibitors. Wang et al. studied a 3-substituted indazole derivative for targeting kinases and stated it as a promising kinase inhibitor¹⁹. Another work by Chen et al. targeted VEGFR2 tyrosine kinase with a novel class of indazole derivatives, showing promise as a potential therapeutic candidate for arthritis and angiogenic therapy²⁰. These derivatives possess structural similarities to ATP, allowing them to compete for binding to the kinase domain and inhibit their autophosphorylation mechanism, where ATP acts as a phosphate donor to phosphorylate a conserved histidine residue within this domain. By utilizing ATP-competitive inhibition, isatin and indazole derivatives may slow down the activity of the BraS histidine kinase, associated with bacitracin resistance²¹. These findings support the exploration of isatin and indazole derivatives as ATP-competitive inhibitors of BraS in MRSA.

The objective of the present study is the rational evaluation of isatin and indazole derivatives with ATP-mimetic cores, based on prior kinase inhibition studies, to assess their suitability for targeting the BraS sensor histidine kinase involved in bacitracin resistance. In this study, an integrated strategy was employed to guide the synthesis of potential isatin and indazole derivatives. These compounds were subsequently evaluated through in vitro assays to assess binding affinities and anti-biofilm activity assays. The proposed approach of targeting BraS with isatin and indazole derivatives not only addresses a specific resistance mechanism in MRSA but also provides a framework for the identification and prioritization of ATP-competitive hit-like molecules associated with TCS-mediated responses. This study also highlights the potential for repurposing existing compounds for MRSA infections. The work is focused on structure-guided hit identification and experimental validation using a screening workflow involving iterative in silico evaluation and experimental validation, as summarized in Fig. S1.

Methodology

Ligand and protein preparation, docking, and interaction analysis

All isatin and indazole derivatives were drawn using ChemDraw²² and the Open Babel²³ tool was used for energy minimization using MMFF94 force fields. The Swiss-ADME tool was used to assess small molecules’ pharmacokinetic properties and drug-likeness²⁴. In the absence of a crystallographic structure for S. aureus BraS (Uniprot entry: A0A0H3JNA3), homologous protein structures were identified using BLAST²⁵ and models of BraS_CA were created using SWISS-MODEL²⁶. The modelled structures were energy minimized using 1000 steepest descent minimization steps to avoid distorted geometric and steric clashes, followed by MD simulation of 100 ns.

To gain insights into substrate-binding spots, structural interaction comparisons were carried out using similar homologous HK structures of Bacillus subtilis BceS (PDB ID: 8G3A), Thermotoga maritima HK853 (PDB ID: 3DGE), S. aureus SrrB (PDB ID: 6PAJ), B. subtilis WalK (PDB ID: 3SL2), and Staphylococcus mutans (PDB ID: 4I5S) using of Pymol graphics tool interface²⁷ superimposition and LigPlot⁺²⁸. The conserved domain databases (CDD) were used to identify the substrate binding site and were utilized for generating the grid for docking. Molecular docking of selected ligands with BraS_CA were conducted with AutoDock Tools 1.5.6²⁹. After incorporating the electron density obtained from ligand torsions and non-polar hydrogen bonds, Gasteiger partial charges were allocated to the ligand atoms. In order to compensate for the effect of polar hydrogen atoms, Kollman charges were introduced to the protein. Solvation and charges have been calculated using the AutoDock Tools module. The Auto Dock 4.2 Lamarckian Genetic Algorithm (LGA) was used to investigate the binding sites of the protein. Around the protein binding site, a grid of 24 × 20 × 20 points was made using a 1 Å grid spacing. The dimensions of the grid were − 14.6, − 28.9, and − 9.5 Å along x, y, and z, respectively.

The prepared receptor and ligand files, along with the grid box parameters, were provided to AutoDock Vina³⁰, and docking was executed using a configuration file specifying the receptor, ligand, grid box parameters, and output options. Docking results were analyzed by visualizing binding poses with Pymol to assess key interactions like hydrophobic interactions, π–π stacking, halogen bond and hydrogen bonds and orientations. Docking results were correlated with structural features to provide insights into ligand-receptor interactions. Predicted binding poses were evaluated based on docking energy scores from the docking log files, allowing comparison among derivatives to identify the most promising derivatives.

Molecular dynamics (MD) simulations

GROMACS software (version 2018.1) was used for MD simulations to analyse conformational dynamics of protein-ligand complexes³¹. The protein parameters were generated with the CHARMM36 force field, whereas ligand parameters were obtained using the SwissParam web server using CHARMM General Force Field (CGenFF)^32,33. GROMACS tools were used to construct the coordinate files and topology, and the ligand structure was optimized before parameterization. To solvate the protein-ligand complexes, a cubic box containing four-point TIP3P water molecules was employed, with the solute and box boundaries at least 1.0 nm apart. Using the Monte Carlo approach, the systems were neutralized with Na⁺ and Cl⁺ ions to attain a 0.15 M salt concentration. For structural relaxation, a steep descent strategy was employed to minimize the solvated systems over 5000 steps, and equilibration was performed using 125ps NVT ensembles at 310 K. An NPT equilibration run of 125 ps was performed using Nose–Hoover thermostat using relaxation time of 1.0 ps and isotropic Parrinello–Rahman barostat of compressibility with 4.5 × 10⁻⁵ bar⁻¹ and a coupling constant of 5.0 ps, for obtaining system pressure at 1 bar³⁴. Production MD simulations were run for 100 ns with a 2.0 fs timestep under periodic boundary limitations. Long-range electrostatics has been treated with PME of 12 Å cutoff for short-range interactions, and hydrogen-containing bonds were constrained via LINCS³⁵. MD trajectories were analysed using GROMACS tools, with RMSD and RMSF used to evaluate protein–ligand stability and flexibility. VMD 1.9.3 and Xmgrace 5.1.25 were used for plotting and analysing trajectory data^36,37.

Relative binding free energies of complexes

The relative binding free energy estimations were conducted using MM-PBSA methodology, as implemented within the GROMACS compatible plugin³⁸ integrating MM-PBSA calculations with GROMACS trajectory files, providing an efficient workflow. For MM-PBSA computations, 500 uniformly spaced snapshots from the last 10 ns of the production trajectory have been chosen for each complex. The contributions of molecular mechanical energy (ΔE_MM), polar solvation energy (ΔG_polar), and non-polar solvation energy (ΔG_nonpolar) were separated from total binding free energy (ΔG_bind). The integrated plugin was used to calculate the solvent-accessible surface area (SASA), which has been used to calculate non-polar solvation component as a linear function of SASA. The MM-PBSA outputs provide thorough explanations of the energetics of the complex interactions.

Synthesis of isatin and indazole derivatives

The synthesis of isatin analogs IS7, IS9 and IS12 are shown in Scheme 1.

Scheme 1.

Synthesis of isatin derivatives IS7, IS9, and IS12.

Methyl 4-((5-methoxy-2,3-dioxoindolin-1-yl) methyl) benzoate (IS9)

To a stirred solution of 5-methoxy isatin (i) (300 mg, 1.69 mmol) in anhydrous N, N-dimethylformamide (DMF) (10 mL) at room temperature, was added anhydrous K₂CO₃ (1.2 eq., 281 mg, 2.02 mmol) in one portion, and the dark coloured suspension was stirred for 1 h. Methyl 4-(bromomethyl)benzoate (ii) (1.2 Eq. 462 mg, 2.02 mmol) was added slowly with constant stirring, and the reaction mixture was stirred at room temperature for 4–5 h until the 5-methoxy isatin had been consumed (TLC). The reaction mixture was poured into water and extracted with ethyl acetate. The ethyl acetate layer was washed with water, brine and dried over MgSO₄. The solvent was removed, and the crude product was purified by silica gel column chromatography using 20% ethyl acetate in hexane as eluent to obtain compound IS9 (334 mg, 60%) as orange-red crystals, mp 182–184 °C.

¹H NMR (500 MHz, CDCl₃): δ = 8.053 (d, J = 8.5 Hz, 2 H), 7.42 (d, J = 8.5 Hz, 2 H), 7.20 (d, J = 2.5 Hz, 1H), 7.05 (dd, J = 8.5 Hz, J = 2.6 Hz, 1H), 6.64 (d, J = 8.5 Hz, 1H), 4.98 (s, 2 H), 3.93 (s, 3 H), 3.80 (s, 3 H). ¹³C NMR (125 MHz, CDCl₃): δ = 183.27, 166.50, 158.37, 156.70, 144.20, 130.34, 130.09, 129.28, 124.70, 118.14, 111.85, 109.74, 55.97, 52.24, 43.79. MS (ESI) m/z 326 [M + H]+; HRMS (ESI) m/z for C₁₈H₁₅NO₅ calculated 326.0950, found m/z 326.1025.

4-((5-Methoxy-2,3-dioxoindolin-1-yl) methyl) benzaldehyde (IS7)

The procedure used for IS9 was repeated with compound 4-(bromomethyl) benzaldehyde (403 mg, 0.41 mmol) and K₂CO₃ (1.2 eq., 403 mg, 2.02 mmol) in anhydrous DMF (10 ml). The compound IS7 was obtained as an orange solid; yield, 245 mg (49%); mp: 146–148 °C; ¹H NMR (500 MHz, CDCl₃): δ = 10.02 (s, 1H), 7.91 (d, J = 8.27 Hz, 2 H), 7.52 (d, J = 8.27 Hz, 2 H), 7.21 (d, J = 2.64 Hz, 1H), 7.07 (dd, J = 8.71 Hz, J = 2.64 Hz, 1H), 6.65 (d, J = 8.5 Hz, 1H), 5.0 (s, 2 H), 3.81 (s, 3 H). ¹³C NMR (125 MHz, CDCl₃): δ = 191.49, 183.14, 158.38, 156.77, 144.07, 141.42, 136.21, 130.45, 127.89, 124.75, 118.17, 111.74, 109.83, 55.98, 43.82. MS (ESI) m/z 296 [M + H]+; HRMS (ESI) m/z for C₁₇H₁₃NO₄ calculated 296.0845, found m/z 296.0918.

1-(4-Acetylbenzyl)-5-methoxyindoline-2,3-dione (IS12)

The procedure used for IS9 was repeated with 1-(4-(bromomethyl) phenyl) ethan-1-one (431 mg, 2.03 mmol) and K₂CO₃ (1.2 eq., 403 mg, 2.02 mmol) in anhydrous DMF (10 ml). The product IS12 was obtained as an orange solid; Yield, 231 mg (74%); mp: 126–128 °C; ¹H NMR (500 MHz, CDCl₃): δ = 7.97 (d, J = 8.2 Hz, 2 H), 7.45 (d, J = 8 Hz, 2 H), 7.20 (d, J = 2.5 Hz, 1H), 7.06 (dd, J = 8.85 Hz, J = 1.99 Hz, 1H), 6.65 (d, J = 8.5 Hz, 1H), 4.98 (s, 2 H), 3.80 (s, 3 H), 2.60 (s, 3 H). ¹³C NMR (125 MHz, CDCl₃): δ = 197.70, 183.25, 158.39, 156.71, 144.17, 139.88, 136.93, 129.09, 127.50, 124.71, 118.14, 111.83, 109.77, 55.97, 43.74, 26.65. MS (ESI) m/z 310 [M + H]+; HRMS (ESI) m/z for C₁₈H₁₅NO₄ calculated 310.1001, found m/z 310.1073.

Methyl 4-((3-bromo-5-(trifluoromethyl)-1 H-indazol-1-yl) methyl) benzoate (IN16)

The synthesis of indazole derivative IN16 is shown in Scheme 2.

Scheme 2.

Synthesis of indazole derivative IN16.

Sodium hydride (60% in oil, 57 mg, 1.41 mmol) was added slowly to a solution of compound (iii) (330 mg, 0.87 mmol) in anhydrous DMF (6 mL) at room temperature. The reaction mixture was stirred for 30 min at 60 °C. It was stirred for another 4 h at 60 °C after adding the methyl 4-(bromomethyl) benzoate (iv) (224 mg, 1.13 mmol). The reaction mixture was poured into ice-cold water. The product was extracted with ethyl acetate, washed with water, and purified using silica gel column chromatography by eluting with 10% ethyl acetate in hexanes to obtain IN16 as a white solid; Yield, 287 mg (55%); mp: 123–125 °C; ¹H NMR (500 MHz, CDCl₃): δ = 8.02 (d, J = 8.2 Hz, 2 H), 7.99 (s, 1H), 7.63 (d, J = 8.7 Hz, 1H), 7.41 (d, J = 8.7 Hz, 1H), 7.29 (d, J = 8.2 Hz, 2 H), 5.66 (s, 2 H), 3.92 (s, 3 H). ¹³C NMR (125 MHz, CDCl₃): δ = 166.45, 141.62, 140.48, 130.25, 130.16, 127.14, 124.47, 123.58, 122.44, 119.03, 110.25, 53.32, 52.21. MS (ESI) m/z 414 [M + H]+; HRMS (ESI) m/z for C₁₇H₁₂BrF₃N₂O₂ calculated 414.0034, found m/z 414.0140.

Methyl 4-((3-amino-5-fluoro-1 H-indazol-1-yl) methyl) benzoate (IN20) & 4-((3-amino-5-fluoro-1 H-indazol-1-yl) methyl) benzoic acid (IN21)

The synthesis of indazole derivatives IN20 and IN21 are shown in Scheme 3.

Scheme 3.

Synthesis of indazole derivatives IN20 and IN21.

Sodium hydride (60% in oil, 153 mg, 6.36 mmol) was added slowly to a solution of compound (v) (800 mg, 5.3 mmol) in anhydrous DMF (15 mL) at room temperature. The reaction mixture was stirred for 30 min at 60 °C. It was stirred for another 4 h at 60 °C after adding methyl 4-(bromomethyl) benzoate (iv) (1.46 g, 6.36 mmol). The reaction mixture was poured into ice-cold water. The product was extracted by ethyl acetate, washed with water, and purified by silica gel column chromatography by eluting with 30% ethyl acetate in hexanes to obtain IN20 as a white solid; Yield, 760 mg (48%); mp: 124–126 °C; ¹H NMR (500 MHz, CDCl₃): δ = 7.98 (d, J = 8 Hz, 2 H), 7.23 (m, 3 H), 7.12 (d, J = 7 Hz, 2 H), 5.41 (s, 2 H), 3.91 (s, 3 H). ¹³C NMR (125 MHz, CDCl₃): δ = 166.93, 157.95, 156.07, 147.66, 142.51, 138.60, 130.24, 129.76, 127.09, 117.07, 116.85, 114.87, 110.11, 110.03, 104.36, 52.23; MS (ESI) m/z 414 [M + H]+; HRMS (ESI) m/z for C₁₆H₁₄FN₃O₂ calculated 300.1070, found m/z 300.1149.

Compound IN20 (260 mg) was stirred with 1 N sodium hydroxide (5 ml) in methanol for 4 h and neutralized using 1 N HCl. The product was extracted with ethyl acetate, washed with water, dried, and purified by silica gel column using 5% methanol in methylene chloride to obtain IN21 as white solid; Yield, 188 mg (76%); mp: 245–247 °C; ¹H NMR (500 MHz, CDCl₃): δ = 7.86 (d, J = 8 Hz, 2 H), 7.50 (d, 2 H, J = 9 Hz), 7.26 (d, J = 9 Hz, 2 H), 7.17 ( m, 1H), 5.40 (s, 2 H). ¹³C NMR (125 MHz, CDCl₃): δ = 167.11, 156.46, 154.61, 149.06, 143.24, 138.17, 129.43, 127.46, 115.76, 114.22, 110.27, 104.88, 50.84, MS (ESI) m/z 286 [M + H]+; HRMS (ESI) m/z for C₁₅H₁₂FN₃O₂ calculated 286.0914, found m/z 286.0992.

Cloning, expression and purification of BraS kinase domain (BraS_CA)

To investigate ligand affinity with the BraS_CA, the gene fragment encoding the BraS_CA (residues 142–295) from S. aureus strain N315 was amplified by PCR using gene-specific primers: forward 5′-CAG CCA TAT GAA TGA GAC TTC TGA TAT TTC T-3′ and reverse 5′-GAT TCT CGA GTT ATT CAT CTG GAA ATT G AA TC-3′. An N-terminal 6×His-tagged BraS_CA construct was created by ligating the PCR product into the pET-28a (+) expression vector after it had been digested with NdeI and XhoI. Restriction digestion followed by DNA sequencing were used to validate the recombinant plasmid. The recombinant plasmid was then transformed into Escherichia coli BL21(DE3) cells for expression of protein on Luria–Bertani (LB) agar plates with kanamycin. One transformed bacterial colony was then inoculated overnight in a small culture medium. This overnight grown bacterial culture has been diluted 1:100 into new LB medium with kanamycin and then incubated at 37 °C until it reached the mid-logarithmic growth phase having an optical density (OD₆₀₀) of 0.6–0.8. At this point, 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) was added to the culture to induce protein expression, and the culture was placed at 16 °C and incubated for 16 h with continuous shaking. The reduced temperature during post-induction was employed to minimize protein misfolding and inclusion body formation.

Protein purification was carried out using a His-tag–based affinity chromatography followed by size-exclusion chromatography, as described previously with minor modifications^39–41. The cells were harvested from culture media using a centrifuge machine at centrifugal force of 5000 g at 4 °C for 10 min, and then the harvested cell pellet were suspended in a buffer consisting of 20 mM Tris-HCl at pH 8.0, 200 mM NaCl and 5% (v/v) glycerol. The resuspended cells were sonicated on ice to rupture the membranes and release intracellular contents. The sonicated lysate was then centrifuged at a centrifugal force of 15,000 g at 4 °C for 30 min to separate soluble proteins from cell debris and insoluble substances. The cleared supernatant was applied into a Ni-NTA agarose column equilibrated with the lysis buffer of the same composition. A wash buffer containing lysis buffer with 20 mM imidazole was used to wash out proteins that has not bound to the resins, and bound fractions were eluted using 250 mM imidazole. Pure fractions were concentrated using centricon filters (10 kDa cut off, Amicon). The concentrated protein was further purified using size-exclusion chromatography with a Superdex 200 Increase 10/300 GL column on ÄKTA purifier platform. The system is washed and well equilibrated with 20 mM Tris buffer (pH 8.0) at a flow rate of 0.5 mL/min before loading the sample. A 100 µL sample of protein (5 mg/mL) was loaded and elution was monitored at 280 nm. The column was calibrated using standard markers set to estimate the apparent molecular weight of the purified protein based on retention volume. An SDS-PAGE gel was run to check protein purity and confirm the molecular weight of the purified protein. Purified fractions were pooled and concentrated for downstream applications.

Microscale thermophoresis analysis of ligand binding to the BraS_CA

A Monolith NT.115 device (NanoTemper Technologies) was used to conduct microscale thermophoresis. RED-Tris-NTA (2nd Generation) labelling kit manufactured by Monolith for fluorescently labelling the purified his-tagged BraS_CA protein. Multiple ligand compounds were screened for interaction with BraS_CA. Dimethyl sulfoxide was used to dissolve the compounds, and they were then diluted with Tris buffer (20 mM Tris-HCl, pH 8.0, 200 mM NaCl, and 0.05% Tween-20) such that the DMSO content in the MST reaction did not exceed 5%. Each ligand was serially diluted, then equal amounts of the ligand and tagged protein dilutions were mixed and left at normal temperature for 5 min. Thermophoretic fluorescence were recorded using 40% LED and 40% MST power after samples were loaded in standard Monolith capillaries. Laser exposure and rest durations were 30s and 5s, respectively. Affinity Analysis software (NanoTemper Technologies) was used to analyze the data, and dissociation constants (K_D) were computed using the normalized fluorescence (F_norm) values. The normalized fluorescence was calculated using the formula:

1

Normalized fluorescence values were plotted against ligand concentrations using a semi-logarithmic scale. The following formula was used to fit the data to a binding isotherm model:

2

In this equation and represent the fluorescence at minimum and saturation binding, respectively. The ligand concentration is denoted by [L]. The dissociation constant is denoted by K_D obtained using non-linear regression. Negative control samples containing no ligand were included to account for baseline fluorescence^42–44.

Biofilm inhibition assay

Inoculum was prepared by incubating the S. aureus N315 strain overnight in tryptic soy broth medium containing 1% glucose, followed by dilution to 0.2 OD at 600 nm in sterile tryptic soy broth containing 1% glucose. Serial dilutions of the compounds were prepared in 96-well flat-bottom plate, and the diluted culture was inoculated. Plates were incubated for 24 h at 37 °C. Briefly, planktonic cells were removed, and wells were washed 3 times with sterile distilled water. Wells were fixed with 99% ethanol, followed by that 0.1% crystal violet was added and the plates were incubated at room temperature for 20 min. Crystal violet stain was removed, and wells were gently washed for 3 times with sterile distilled water. Dried and the stain was solubilised in 30% acetic acid. Samples were transferred to another clean 96-well plate, and the absorbance was measured at 570 nm. Aseptic techniques were maintained, and parameters adjusted as needed⁴⁵. The experiment was performed in triplicate and data analysed using one-way ANOVA test with Dunnett’s multiple comparison using GraphPad Prism 8 software. The percentage of biofilm formation was calculated after the subtraction of the media blank using following formula.

The DMSO control was considered as 100% biofilm formation. All experiments were performed in triplicate, and data are presented as mean ± standard deviation.

Results and discussion

Homology modelling and binding site analysis of S. aureus BraS_CA

Multiple sequence alignment of S. aureus BraS_CA with homologous histidine kinases showed conserved regions at different homology boxes like N-box, G1-box, G2-box, and F-box with variation in ATP-lid region (Fig. S2a). Based on selected template HK853_CA (38.74% sequence identity over 71% query coverage, with an E-value of 7 × 10⁻²⁰), a homology model of BraS_CA was constructed. It displayed stable structural behaviour through 100 ns MD simulation in comparison to homologous T. maritima sensor kinase HK853, supporting its reliability for further in-silico screening (Fig. S2b). Structural superimposition of the BraS_CA model with homologous crystal structure HK853 (3DGE) and YycG (3SL2) revealed that BraS_CA shares a conserved canonical histidine kinase domain (Fig. S2c–e). Comparative interaction analysis of ATP with homologous crystal structure YycG confirms BraS have similar substrate binding region and ligand-interaction geometry indicating conservation of ATP recognition region across this system (Fig. 2). Several conserved N-box residues Gln199, Asn203, Tyr207 of BraS, Asn499, Asn503, Tyr507 of YycG, were found to be interacting with ATP. Conserved Domain Database (CDD) predictions further confirmed a conserved substrate-binding pocket across Gln199 to Leu 262, overlapping 95% of superimposed nucleotide binding sites (Fig. S3).

Fig. 2.

Structural interaction comparison and identification of substrate-binding sites of BraS_CA: (a) Superimposition of homology model of BraS_CA (light pink) with the crystal structure of YycG (3SL2) bound to ATP (cyan). (b) Interaction of ATP with YycG. (c) Interaction of ATP with BraS_CA.

Preliminary computational screening and prioritisation of hits

Overview of compound selection

An initial set of six isatin and fourteen indazole derivatives were selected from the literature for virtual screening against BraS_CA based on previous studies highlighting the kinase-inhibitory potential of isatin and indazole scaffolds. These compounds have varying degrees of biological activities against two-component systems such as WalK, as well as cancer associated targets including ADH, CDK2, EGFR tyrosine kinase, PI3 kinase, and RAF kinase^46–51. The rationale for selection was to test the inhibitory potential of as they possess ATP-mimetic heterocyclic core and may compete for the ATP-binding site in BraS_CA. This competitive binding may inhibit the autophosphorylation activity of BraS by occupying the ATP-binding pocket and thereby suppressing its catalytic activity. The objective of this initial screening was to evaluate whether previously reported isatin and indazole derivatives could adopt favorable binding poses within the ATP-binding pocket of BraS_CA and serve as effective hits against BraS. All compound structures included in this screening set are presented in Fig. S4.

Molecular docking, interaction analysis and identification of initial hits

Molecular docking analyses were performed to evaluate the binding poses and interaction patterns of the initial set of isatin and indazole derivatives with BraS_CA. The docking was validated through re-docking of ATP with the crystal structure of YycG bound to ATP (3SL2), which reproduced the experimentally observed binding pose (Fig. S5). The docking energies and key interacting residues for isatin derivatives are summarized in Table S1.

Among the isatin derivatives, IS2 and IS3 showed the most favorable docking scores, comparable to ATP. The docked IS2 and IS3 formed multiple stabilizing interactions with conserved catalytic residues in the BraS_CA ATP pocket. Both compounds form stabilizing interactions with conserved catalytic residues and occupied the canonical ATP-binding pocket of BraS suggesting favourable occupancy (Fig. 3).

Fig. 3.

Docking interaction of initial hits of isatin derivatives with BraS_CA: Isatin-2 (top panel) and Isatin-3 (bottom panel) are shown bound within the active site pocket of BraS_CA on the left, with detailed 3D interaction views on the right. Key interactions include hydrogen bonds (blue lines), hydrophobic interactions (grey dashed lines), and a salt bridge (orange dashed line) with distance in Å.

For indazole derivatives, the docking analysis identified IN12 and IN13 as top-ranked compounds based on their predicted binding poses and interaction profiles within the BraS_CA ATP pocket (Table S2). Both compounds adopted orientations similar to ATP and engaged conserved residues through hydrogen bonding and hydrophobic interactions. The presence of polar and halogen-substituted functional groups contributed to stabilization of interaction within the pocket (Fig. 4).

Fig. 4.

Docking interaction of initial hits of indazole derivatives with BraS_CA: Indazole-12 (top panel) and Indazole-13 (bottom panel) are shown bound within the active site pocket of BraS_CA on the left, with detailed 3D interaction views on the right. Key interactions include hydrogen bonds (blue lines), hydrophobic interactions (grey dashed lines), halogen bonds (cyan lines), and a salt bridge (orange dashed line) with distance in Å.

IS2 andIS3 from isatin derivatives and IN12 and IN13 from indazole derivatives were selected based on docking performance, pose consistency and interactions as initial hits for further evaluation through MD simulations to assess the stability of their binding within the pocket.

Molecular dynamics simulations of selected hits

MD simulations of the BraS_CA were conducted in complex with ATP and initial isatin and indazole hits IS2, IS3, IN12 and IN13 over a 100 ns simulation period to evaluate the dynamic behaviour and conformational stability of the complexes. The C_α RMSD plots showed stabilization of all protein-ligand complex systems within 20–25 ns. BraS_CA showed higher backbone fluctuations, indicating increased conformational flexibility in the absence of a bound ligand while the BraS_CA-ATP complex showed a stable C_α RMSDs, consistent with substrate ATP. Among the ligand-bound complexes, BraS_CA-IS2 and BraS_CA-IN12 showed stable C_α RMSDs, while IS3 and IN13 complexes showed moderately higher fluctuations (Fig. S6a, g). The ligand RMSD analysis provided additional support with IS2 and IN12 showing early stabilization and remaining consistently low compared to IS3 and IN13 which required a longer equilibration period. ATP in comparison has a higher ligand RMSD fluctuation relative to screened compounds, reflecting the dynamic binding behaviour of the native substrate (Fig. S6b, h).

The effect of ligand binding and interactions within the substrate binding cleft was studied using RMSF plots. revealed that BraS_CA is more flexible in the absence of ATP, showing greater RMSF values across different loop regions which play important role in ATP binding. The fluctuations of these regions were reduced in complexes of ATP and initial hits. IS2 and IN12 complexes showed a substantial drop in RMSF values in the key loop region, including the ATP-lid while IS3 and IN13 complexes showed comparable or moderately reduced fluctuations relative to ATP complex (Fig. S6c, i).

Conformational snapshots and interaction analysis of all systems from the 100 ns simulation trajectory reveal a similar binding pattern of each ligand within the ATP-binding domain. These analyses suggest that ATP and the screened compounds engaged conserved catalytic residues through a combination of hydrogen bonding and hydrophobic interactions, supporting the retention of ATP-competitive binding modes. Detailed residue-level interactions and distances are provided in Fig. S6.

Insights from initial screening and shift toward novel analogues

To explore inhibitors of BraS_CA, we initially adopted a repurposing approach by screening known isatin and indazole derivatives, which had been previously reported as kinase-targeting agents, primarily in anticancer studies. These scaffolds have favourable binding orientations and compatibility with the ATP-binding cleft of BraS_CA, interacting with several key residues such as Gly251, Asn202, and Ser247. However, in comparison to the ATP, these compounds formed fewer and less persistent interaction networks, particularly with respect to electrostatic and hydrogen-bonding interactions. Trajectory analyses revealed that while ATP maintained a well-stabilized binding pose with minimal drift and preserved conformational integrity of the protein, some derivatives, like IS3 and IN13, showed higher ligand RMSD values and were associated with increased protein backbone fluctuations based on C_α RMSD/RMSF analysis, suggesting less optimal binding stability in the ATP-binding pocket. The binding conformations captured at the 100 ns further supported this observation, showing partial displacement or reduced contact surface relative to ATP.

In parallel, in-silico ADMET profiling indicated that several candidates displayed suboptimal oral bioavailability, poor absorption predictions, and potential toxicological concerns, such as mutagenicity and hepatotoxicity, posing limitations for further prioritization as BraS_CA inhibition (Supplementary Table S3).

To address these limitations, a structure-guided analogue design approach was adopted to generate a new series of isatin derivatives and indazole derivatives (Fig. 5). In this refined set, core scaffolds were retained, but polar and hydrogen-bond-forming groups were modified to achieve improved binding parameters with conserved residues involved in phosphate and nucleotide recognition. These derivatives were subsequently evaluated through in silico docking, simulations and energy calculations to assess predicted binding behaviour, interaction stability, and pocket occupancy. Selected compounds were synthesized and experimentally evaluated using microscale thermophoresis (MST) with recombinant BraS_CA. This transition in compound selection ensures an integrated structure-based design strategy to move toward prioritized hit-like candidates with improved binding efficacies and BraS_CA inhibition.

Fig. 5.

New set of isatin and indazole derivatives selected for screening: The upper panel shows six isatin derivatives (IS7–IS12), with R groups varied as indicated (CHO, CN, COOCH₃, COOH, N(CH₃)₂, C(=O)–CH₃), based on two prior isatin hits (IS2 and IS3) shown on the left. The lower panel depicts nine indazole derivatives (IN15–IN23) divided into two groups, derivatives containing Br/F3C substitutions (IN15–IN18) and those with F/NH2 groups (IN19–IN23), based on two prior hit compounds (IN12 and IN13) shown on the left. Each group includes substitutions at the R position similar to the isatin series. Previously identified hit compounds (IS2, IS3, IN-12 and IN-13) served as templates for designing these new derivatives for screening.

Molecular docking analysis of secondary derivatives

Molecular docking was carried out for new derivatives using the same receptor grid and protocol established for the initial set to ensure consistency in analysis. The docking analysis indicated that these derivatives formed multiple stabilizing interactions with conserved catalytic residues within the ATP-binding site of BraS_CA and adopted stable ATP-competitive binding poses. The refined isatin derivatives (IS7–IS12) demonstrated docking scores ranging from − 7.6 to − 8.0 kcal/mol, which are comparable to ATP (− 8.1 kcal/mol) (Table S4). Among these, IS10 and IS12 emerged as top-ranked candidates based on binding poses and relative docking scores. The docking interaction analysis reveals that isatin derivatives were accommodated within the ATP-binding cleft of BraS_CA and stabilized through polar and hydrophobic interactions. IS9, IS10, and IS12 primarily formed hydrogen bonds with conserved residues in the N-box region such as Gln199 and Asn202 while other derivatives also engaged residues of ATP-lid region (Fig. 6). Detailed docking score and residue-level interactions of these isatin derivatives are provided in the supplementary Table S4.

Fig. 6.

Interaction analysis of novel isatin derivatives (IS7–IS12) with BraS_CA: Isatins (IS7-IS12) are superimposed within the active site pocket (left), with detailed 3D interaction views on the right. Key interactions include hydrogen bonds (blue lines), hydrophobic interactions (grey dashed lines), and a salt bridge (orange dashed line) with distance in Å.

Similarly, docking analysis of the new indazole subset (IN15–IN23) demonstrated scores ranging from − 7.6 to − 8.8 kcal/mol with several derivatives showing comparable docking scores with respect to ATP (− 8.1 kcal/mol) (Table S5). The indazole derivatives adopted ATP-like orientations within the binding pocket and were stabilized by a combination of polar, hydrophobic, halogen-mediated, and π-cation interactions with conserved catalytic regions. Recurrent engagement of residues within the conserved N and G-boxes was observed across the indazole series, underscoring their importance in ligand interaction and stabilization (Fig. 7). Docking score and detailed interaction patterns for individual compounds are presented in the Supplementary Table S5.

Fig. 7.

Interaction analysis of novel indazole derivatives (IN15–IN23) with BraS_CA: Indazoles (IN15–IN23) with detailed 3D interaction views. Key interactions include hydrogen bonds (blue lines), hydrophobic interactions (grey dashed lines), halogen bonds (cyan lines), and a salt bridge (orange dashed line) with distance in Å.

MD simulations and MM-PBSA analysis of secondary derivatives

Taken together, the docking outcomes from initial hits and novel isatin and indazole scaffolds demonstrated consistent targeting of conserved catalytic residues within the BraS_CA catalytic cleft, with no substantial differences observed in their docking scores. Therefore, all secondary derivatives were subjected to 100 ns MD simulation studies to assess binding pose retention, conformational stability, interaction dynamics, and MM-PBSA-based relative binding free energies.

Trajectory analysis of isatin and indazole complexes

C_α RMSD values provide insights into the overall backbone stability of the BraS_CA complexes throughout the 100 ns simulation. Across both scaffolds, most ligand-bound complexes maintained stable RMSD profiles comparable to the ATP bound complex. In particular, isatin derivatives IS7, IS10, IS11, and IS12 exhibited stable RMSD values whereas IS8 and IS9 showed moderately higher fluctuations reflecting increased backbone flexibility (Fig. S7a). Similarly, in case of indazole derivatives most complexes displayed stable RMSD profiles, while IN15, IN21 and IN23 showed higher fluctuations (Fig. S8a). These observations suggest that ligand binding does not substantially disrupt the overall conformation of BraS_CA aside from a few outliers showing increased backbone flexibility.

Positional stability of each ligand within the binding pocket was assessed through ligand RMSD profile which indicated that most of the isatin derivatives including ATP, IS7, IS9, IS10, IS11, and IS12 exhibited low and stable profiles indicating strong retention and minimal displacement from the binding site. In contrast, IS8 demonstrated increased ligand RMSD after equilibration indicating poor positional stability within the binding pocket (Fig. S7b). Among the indazole derivatives, several complexes displayed stable ligand RMSDs while IN19, IN20 and IN23 exhibited increased ligand RMSD suggesting reduced positional stability in binding pocket of BraS_CA (Fig. S8b).

RMSF analysis shows that both isatin and indazole complexes shows low RMSF values across the kinase core, with higher flexibility localized to loop regions associated with the ATP-binding cleft. Isatin complexes such as IS7, IS9, and IS12 exhibited lower fluctuations across most region whereas IS8 showed comparatively higher flexibility in the ATP-lid region (Fig. S7c). A similar trend was observed for indazole complexes, where most derivatives maintained lower residue fluctuation after ligand interaction while increases in residue flexibility observed for IN19 and IN22 (Fig. S8c).

The radius of gyration (R_g) analysis shows that most isatin and indazole complexes maintained compact conformations comparable to BraS_CA and its ATP complex. IS7, IS9 and IS12 show stable R_g value throughout the simulation while IS8 complex has increased R_g values indicating a decrease in compactness (Fig. S7d). Similarly, in the indazole group IN16, IN20 and IN21 show stable compactness while complex IN17, IN22 and IN23 showed slightly increased R_g value during simulation period (Fig. S8d).

Hydrogen bond analysis revealed that stable isatin and indazole complexes maintained persistent hydrogen-bonding interactions throughout the simulation. Among the isatin derivatives IS7, IS9 and IS12 complexes showed higher hydrogen-bond occupancy compared to IS8 and ATP throughout the simulation trajectory indicating relatively stable binding interactions (Fig. S7e). In indazole derivatives, most complex demonstrated similar H-bonds pattern while IN18 and IN 23 shows less and variable hydrogen bond numbers (Fig. S8e).

The SASA of derivatives IS7, IS12, IN17 and IN23 clustered at lower values demonstrating compact folding and stable ligand binding, while IS8, IN15 and IN22 display higher SASA indicating increased solvent exposure (Figs. S7f, S8f).

Relative binding free energy calculations by MM-PBSA

MM-PBSA analyzes each ligand-protein complex’s relative binding free energy, decomposing contributions of polar solvation, non-polar solvation, van der Waals, and electrostatic energy. These results provide insights into the binding energetics of isatin and indazole complexes compared to ATP. ATP showed the highest relative binding energy driven by van der Waals and electrostatic contributions.

Among the isatin derivatives, IS7 showed the most favourable relative binding free energy primarily due to van der Waals interactions and electrostatic forces followed by IS9 and IS12. However, the relative binding energies of the isatin series were weaker than that of ATP consistent with higher polar solvation penalties and reduced electrostatic complementarity within the ATP-binding cleft. In comparison to the isatin series, several indazole derivatives showed more favourable relative free binding energy. Among these IN16 showed favourable relative binding energy contributed by electrostatic and van der walls energies which were comparable to that of ATP ligand. Additionally, IN20 and IN21 also showed favourable relative binding energies despite van der Waals contributions which was offset by reduced electrostatic energy contributions.

Overall, simulations and binding energetics data suggest ATP as the naturally most potent binder, but among novel derivatives, several indazole compounds outperformed the isatin series in total binding energy, suggesting better shape complementarity and hydrophobic packing within the ATP-binding cleft. IS7 is the only isatin derivative approaching reasonable binding strength, largely due to its strong electrostatic stabilization, but still lagged the best indazole analogues. The top three isatin (IS7, IS9 and IS12) and top three indazole (IN16, IN20 and IN21) derivatives demonstrated more favorable binding energies than other compounds in their respective sets and selected for further synthesis and in vitro binding assay These trends reflect relative energetic ranking within the compound sets and do not represent absolute binding affinities.

Selection, synthesis and in vitro binding assay

A subset of isatin (IS7, IS9 and IS12) and indazole (IN16, IN20 and IN21) derivatives prioritized based on in silico evaluation were synthesized. The synthesis of compounds IS9, IS7 and IS12, is depicted in Scheme 1. These compounds were prepared through a one-step reaction in which 5-methoxy isatin was reacted individually with methyl 4-(bromomethyl) benzoate, 4-(bromomethyl) benzaldehyde, and [1-(4-(bromomethyl) phenyl] ethan-1-one, using potassium carbonate in DMF, which yielded the corresponding N-substituted isatin IS9, IS7 and IS12. The indazole derivative IN16, as shown in Scheme 2, was synthesized by treating 3-bromo-5-(trifluoromethyl)-1 H-indazole with methyl 4-(bromomethyl) benzoate in the presence of sodium hydride in DMF. Under similar conditions, 5-fluoro-1 H-indazol-3-amine was reacted with 4-(bromomethyl) benzoate to synthesize compound IN20 as shown in Scheme 3, and subsequent alkali hydrolysis produced compound IN21. The structures of all isatin and indazole derivatives were confirmed through 1 H NMR, 13 C NMR, and HRMS analysis. Furthermore, the purity of the compounds, which exceeded 98%, was validated using analytical high-performance liquid chromatography (HPLC) before their application in downstream binding analysis.

To check the binding affinity of these derivatives with BraS_CA, the C-terminal domain of BraS_CA (residues 142–295) was successfully cloned (Fig. S9), expressed, and purified as an N-terminal 6×His-tagged recombinant protein. The size exclusion chromatogram showed a distinct major elution peak at 17.09 ml (Fig. 8a) using a calibrated Superdex 200 Increase 10/300 GL column. Compared to the standard calibration curve, the BraS_CA was estimated to have an apparent molecular weight of approximately 19.2 kDa. This elution volume is consistent with the theoretical size expected for the BraS_CA fragment, suggesting the protein behaving as monomer under the experimental conditions used. SDS-PAGE result of the purified fractions confirmed a single protein band migrating near 19 kDa, which correlates well with the molecular weight estimated from SEC. The absence of additional bands or smearing indicates high purity of the recombinant BraS_CA. The SEC results, consistent with the SDS-PAGE profile, indicate the protein is likely to retain its native fold and behave as a monomer in solution, suitable for further ligand-binding biophysical studies.

Fig. 8.

Size exclusion chromatography and microscale thermophoresis (MST) analysis of BraS_CA with novel isatin and indazole derivatives: (a) size exclusion chromatogram showing the elution of the purified BraS_CA and corresponding SDS-PAGE having Lane “M” represents molecular weight markers and lane “A” for the AKTA-purified fraction, confirming the protein’s purity and molecular weight. (b) MST binding curve of ATP as the reference ligand, demonstrating its affinity with the BraS_CA (c–f) MST binding curves for compounds IS7, IS12, IN16, and IN21, showing concentration-dependent binding to BraS_CA. The structure of all compounds is also shown along with a curve in the top left. The y- axis depicts the normalized fluorescence percentage, and the x- axis represents compound concentration in mole. *MST binding curves represent a single experimental measurement; therefore, error bars and statistical analysis are not shown.

The MST data suggests concentration-dependent interactions between the BraS_CA and isatin and indazole derivatives, with dissociation constants (K_d) in the low nanomolar range, indicating relative differences in binding affinities of selected compounds. ATP, as a reference ligand, displayed a typical sigmoidal binding curve (Fig. 8b) with a K_d of 56 nM. This curve exhibited a clear response with saturation at higher ATP concentrations, indicating specific and tight binding to BraS_CA, and establishing a baseline for comparative analysis for other compounds. Among the isatin derivatives, IS7 (Fig. 8c) showed a concentration-dependent binding curve similar to ATP, with a calculated K_d of 58 nM, showing binding affinity in the same order of magnitude as that of ATP while IS12 (Fig. 8d) with a sharply shifted MST curve and an estimated K_d of 2 nM, indicated strong binding affinity under the tested conditions for BraS_CA. IS9 binding affinity cannot be determined under these assay conditions. The indazole derivatives demonstrated more variable affinities, IN16 (Fig. 8e) produced a classic MST binding isotherm, with a K_d of 251 nM, reflecting moderate affinity for BraS_CA compared to ATP and isatin analogues, whereas IN21 (Fig. 8f) affinity is similar to that of ATP, with a K_d of 56 nM, indicating a strong and specific BraS_CA interaction. IN20 showed no detectable binding, and the K_d could not be determined (Table 1). For some compounds, binding affinity could not be determined under the tested concentration range, related to compound solubility or signal response.

Table 1.

Binding affinities of isatin and indazole derivatives with BraS_CA.

Molecule	Name	Kd values
ATP	Adenosine triphosphate	56nM
IS-7	4-((5-Methoxy-2,3-dioxoindolin-1-yl) methyl) benzaldehyde	58nM
IS-9	Methyl 4-((5-methoxy-2,3-dioxoindolin-1-yl) methyl) benzoate	NA
IS-12	1-(4-Acetylbenzyl)-5-methoxyindoline-2,3-dione	2nM
IN-16	Methyl 4-((3-bromo-5-(trifluoromethyl)-1 H-indazol-1-yl) methyl) benzoate	251nM
IN-20	Methyl 4-((3-amino-5-fluoro-1 H-indazol-1-yl) methyl) benzoate	NA
IN-21	4-((3-amino-5-fluoro-1 H-indazol-1-yl) methyl) benzoic acid	56nM

*NA- Not Active.

These MST results provide experimental validation of the computational pipeline and confirm that tested derivatives engage the BraS_CA in-vitro and establish these compounds as promising starting points for further functional or inhibitory assays.

Biofilm inhibition

Biofilm formation by S. aureus N315 was assessed in response to selected isatin and indazole derivatives (Fig. 9). As no antibacterial activity was detected for the tested compounds (Fig. S11), the observed effects reflect non-bactericidal modulation of biofilm formation. Biofilm formation analysis suggest that all tested compounds have effect on biofilm production. Within the isatin series, IS7 and IS12 showed significant decrease in biofilm formation in the range of approximately 85% and 65% respectively in the concentration range of 3.1–50mM as compared to control, while IS9 exhibited approximately 70% reduction in the biofilm formation at some concentrations (**p < 0.01), (Fig. 9a). Among the indazole derivatives, IN16 caused a significant decrease (70%) in biofilm formation at all tested concentrations. IN20 displayed variable effects with 70% decrease in the concentration range of 6–25 mM. IN21 showed 20% reduced biofilm formation at the lowest tested concentration of 3.125 mM, whereas higher concentrations resulted in biofilm enhancement as comparable to the control, indicating a loss of antibiofilm effect at elevated concentrations (Fig. 9b). Overall, isatins have more profound effect on biofilm inhibition compared to indazoles which showed inconsistent dose-dependent behaviour. Cytoscape-based interaction analysis of pathways associated with BraSR signalling and other two-component systems linked to biofilm regulation indicated that the observed biofilm modulation is likely mediated through indirect effects on regulatory networks, including BraSR/NsaRS-associated signaling pathways previously implicated in biofilm control^8,52(Fig. S12).

Fig. 9.

Modulation of biofilm formation by selected isatin and indazole derivatives in S. aureus: (a) Effects of IS7, IS9, and IS12 on biofilm formation; (b) effects of IN16, IN20, and IN21 on biofilm formation. Biofilm biomass was quantified by crystal violet staining and absorbance measured at 570 nm. Data are expressed as percentage biofilm formation relative to the DMSO control (100%) after media blank subtraction. Bars represent mean ± SD of three independent experimental replicates. Statistical significance was assessed by one-way ANOVA followed by Dunnett’s test versus DMSO (ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

Conclusion

In this study, we applied a structure-based drug discovery approach involving molecular docking, molecular dynamics simulations, MM-PBSA free-energy estimation to explore ATP-competitive isatin and indazole based derivatives as inhibitors of the BraS histidine kinase, a key signaling component of the BraSR two-component regulatory system in MRSA to prioritize compounds exhibiting favorable relative binding energetics. ADMET predictions were used further as in-silico filters to deprioritize compounds with apparent liabilities to take forward probable scaffolds as preliminary chemical hits. Overall, the in-silico analyses revealed that several isatin and indazole derivatives adopt stable binding conformations within the binding pocket of BraS kinase domain and showed favorable predicted binding energy profile. These computations resulted in prioritizing hits within each chemical series of compounds. Based on structure guided screening, three isatin and three indazole derivatives were selected and synthesized for experimental evaluations. In vitro binding experiments with the recombinant BraS kinase domain confirmed strong binding in the range of 2nM for IS12 and moderate affinities in the range of 50nM for IS7, IN21 and low affinity for IN16 at 250nM. Biofilm inhibition analysis indicated a reduction in biofilm biomass at specific concentrations for certain compounds; however, this response could be due to indirect effect of regulatory pathways, potentially involving BraSR/NsaRS-associated signaling pathways reflecting phenotypic correlations.

Given the conserved nature of ATP-binding domains across bacterial histidine kinases, the isatin and indazole scaffolds esp. IS12 and IN21 identified here could provide promising foundation for further structure-guided optimization with more in-depth cellular level validations and toxicity profiling to improve potency and to establish functional relevance against two-component signaling pathways in pathogenic bacteria.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Abbreviations

TCS

Two-component system

HK

Histidine kinase

RR

Response regulator

CA domain

Catalytic and ATP-binding domain(kinase) domain

BraS_CA

BraS kinase domain

BraSR-TCS

Bacitracin-resistance-associated two component system

MST

Microscale thermophoresis

MD

Molecular dynamics

Author contributions

“formal analysis, investigation- GJ, KG, ST, SKS; writing-review and editing- GJ, ST, SA, AKS, TS, KG, SKS; conceptualization, supervision, funding acquisition- SKS. All authors read and agreed to the content of the manuscript.”

Funding

Open access funding provided by Manipal University Jaipur. The work was financially supported by Science and Engineering Research Board (SERB), India grant CRG/2020/002008 and Department of Biotechnology, Govt. of India grant BT/PR40197/BTIS/137/68/2023 to SKS. Department of Biosciences, Manipal University Jaipur is supported by DST-FIST project (DST/2022/1012) from Govt. of India.

Data availability

All data generated or analysed during this study are included in this published article [and its supplementary information files]. Any files analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.