Microbiological culture media and chemical reagents were procured from HiMedia, Mumbai. Antibiotics were obtained from SRL Fine Chemicals. The LIVE/DEAD BacLight Bacterial Viability Kit (L-7012) was sourced from Thermo Fisher Scientific in the United States.

Bacterial strains SA1199 (wild type) and SA1199B (overexpressing norA) were obtained from BEI Resources, NIAID, NIH, USA. The strains were retrieved from glycerol stock (−80°C) and regularly sub-cultured, maintaining them on Tryptic Soy Agar (TSA). Confirmation of SA1199B as a fluoroquinolone-resistant strain due to NorA overexpression was carried out through antimicrobial susceptibility testing using the broth microdilution method as per Clinical and Laboratory Standards Institute guidelines (Limbago, 2001).

The ligands (QSL_P^A and QSL_P^B) and their palladium(II) metal complexes (QSL_Pd^1A to QSL_Pd^6A and QSL_Pd^1B to QSL_Pd^6B) were developed through collaboration with Dr. D. Suresh, Assistant Professor of Research at the School of Chemical and Biotechnology, SASTRA Deemed University.

The minimum inhibitory concentration (MIC) for the synthesized ligands and their palladium(II) metal complexes (“A” and “B” series) was determined by following CLSI guidelines with slight modifications (Limbago, 2001). A stock solution containing all the synthesized compounds was meticulously prepared in dimethyl sulphoxide (DMSO), achieving a concentration of 22 mg/mL. The strains SA1199 and SA1199B were cultured overnight in tryptic soy broth (TSB). The culture was diluted 1:100 in physiological saline, and the bacterial suspensions were adjusted to 0.5 Mc Farland standard. Employing a 96-well microtiter plate, the adjusted inoculum was added to the wells containing two-fold serial dilutions of the compound covering a concentration range from 600 µg/mL to 1.71 µg/mL in Cation-adjusted Mueller Hinton broth (CAMHB). Following a 24-h incubation at 37°C, cell viability was measured at OD₅₉₅, and the extent of inhibition was quantified using the following formula:

For qualitative assessment of EP inhibition induced by the synthesized ligands and palladium(II) complexes, a cartwheel assay was conducted following the protocol outlined by Martins et al., with minor adaptations (Martins et al., 2013). EPs use ethidium bromide (EtBr) as the substrate to extrude, hence used to detect the efflux activity by EtBr cartwheel assay. Only compounds from the “A” series were selected on the basis of MIC outcomes. A consistent concentration of EtBr at 1 µg/mL and compounds at 0.5× MIC were incorporated into molten TSA at 50°C. The resulting agar mixture was poured into petri plates and allowed to solidify. Cultures of bacterial strains SA1199 and SA1199B were adjusted to a density of 0.5 McFarland units (1.5 × 10⁸ Colony forming units (CFU)/mL) and were systematically swabbed onto the plates in a cartwheel pattern, with triplicate samples. Following incubation for 24 h at 37°C, fluorescence was visualized using a UV-transilluminator.

For a quantitative assessment of EP inhibition induced by the synthesized palladium(II) complexes and ligands, a real-time efflux assay was conducted following the protocol outlined by Costa et al., with minor adaptations (Costa et al., 2013b). The overnight culture of SA1199B was subjected to centrifugation, washed twice with PBS supplemented with 0.4% glucose, and subsequently re-suspended in an equivalent volume of media and the turbidity was adjusted in the uptake buffer, followed by the addition of EtBr (10 µg/mL). The cells were incubated at 37°C for 1 h to facilitate EtBr uptake. After centrifugation at 12,000 rpm for 5 min, the cells underwent two washes with phosphate buffered saline (PBS) (0.4% glucose). Four palladium(II) complexes demonstrating elevated fluorescence in the EtBr cartwheel assay, one ligand(QSL_P^A) and verapamil as control at 0.5× MIC were each introduced into a 96-well black tissue culture plate. Subsequent readings were taken at 5-min intervals over a span of 30 min using the Biotek Synergy H1 multimode reader, with excitation at 530 nm and emission at 600 nm. The compound displaying stronger EP inhibition in the real-time efflux assay was selected for subsequent combinatorial investigations.

The effectiveness of QSL_Pd^5A in inhibiting biofilm formation was assessed against SA1199B using the crystal violet assay. The overnight culture of SA1199B was suitably diluted, following the procedure detailed in the MIC determination, and introduced into a 96-well plate containing sub-inhibitory concentrations of QSL_Pd^5A (ranging from 9.38 µg/mL to 0.29 µg/mL in two-fold serial dilutions). The culture medium employed was TSB supplemented with 1% NaCl and 1% glucose. After 24 h of incubation at 37°C, the planktonic cells were measured for their optical density at OD₅₉₅. The next steps involved gently tapping to remove the planktonic cells, followed by two PBS rinses, and air-drying for 5–10 min. Subsequently, a 0.2% crystal violet (120 µL) was added and incubated for 15 min to achieve cellular staining. After removing the excess stain by washing it with water, the plate was air-dried and eluted with 33% acetic acid. The absorbance was quantified at 595 nm (Vasudevan et al., 2020). This meticulous process was conducted in triplicates to ensure robust results.

Fluorescence microscopy imaging was employed to visually investigate the biofilm inhibition potential of QSL_Pd^5A. In a concise description of the method, the inoculum was cultured in a six-well plate with 2 mL of TSB supplemented with 1% glucose and 1% NaCl. Sterile coverslips were positioned, both in the presence and absence of the compound. Different concentrations, including 0.5×, 0.25×, and 0.125× the MIC of QSL_Pd^5A, as well as 0.5× MIC of QSL_P^A and 0.5× MIC of verapamil (used as a positive control), were introduced into appropriate wells. Untreated wells were retained as controls. After 24 h of incubation at 37°C, non-adherent planktonic cells were carefully removed, and the coverslips were gently washed twice with PBS. Utilizing the Bac Light Bacterial Viability Kit (L7012), coverslips were stained as per the manufacturer’s instructions using propidium iodide and Syto9 dyes, followed by a 10-min dark incubation. After eliminating excess stains and conducting two PBS washes, the coverslips were meticulously mounted onto glass slides. Subsequently, fluorescence excitation images were adeptly captured at a ×20 magnification using the Nikon Eclipse Ts2 microscope.

The checkerboard assay was performed to assess the potentiating activity of QSL_Pd^5A with ciprofloxacin, following the guidelines outlined by the CLSI (Limbago, 2001). Briefly, QSL_Pd^5A starting from its sub-inhibitory concentration of 9.38 µg/mL was two-fold serially diluted in the wells containing 100 µL of CAMHB. This was combined with an addition of ciprofloxacin concentrations spanning from 4 µg/mL to 0.125 µg/mL. The individual concentrations of both compounds were also plated to serve as controls and references for the synergistic effects observed. Following the approach described in the MIC experiment, the overnight culture of SA1199B was suitably diluted and inoculated in 96-well plates (Thamilselvan et al., 2021). The plates were incubated for 24 h at 37°C, the optical density was measured at 595 nm and, subsequently, the percentage of inhibition was calculated. The obtained data were analyzed using SynergyFinder Plus, utilizing the bliss independence model. This model calculates the expected effect based on the product of individual effects, represented by the equation Ei = EA * EB.

The time-kill efficacy of the combination of ciprofloxacin (CIP) with QSL_Pd^5A was investigated to gain insights into CIP’s impact on SA1199B in the presence of QSL_Pd^5A over 24 h. The experimental protocol closely mirrored that of the checkerboard assay. Succinctly, CIP at a concentration of 0.125 µg/mL was combined with QSL_Pd^5A at a concentration of 4.69 µg/mL, as determined from the outcomes of the checkerboard assay. These concentrations were added in volumes of 100 µL to each well of a 96-well plate, along with 10 µL of the bacterial culture. The untreated bacterial culture was treated as a negative control. At intervals of every 2 h, readings were taken at OD₅₉₅ from the 0th hour through the 10th hour, culminating in a final reading at the 24th hour.

To determine the cytotoxic effect of QSL_Pd^5A, an 3-(4,5-Dimethylthiazol-2 yl)-2,5-Diphenyltetrazolium Bromide (MTT) assay was performed (Balamurugan et al., 2017). Eagle’s Minimum Essential Medium supplemented with non-essential amino acids, 10% fetal bovine serum, and penicillin-streptomycin (100 U/mL to 100 µg/mL) was used as culture medium. In brief, 200 µL of viable human kidney epithelial cells (NRK-62E) in a concentration of 1.5 × 10⁴ cells/mL were seeded in 96-well tissue culture plates. At 80% confluence, the varying concentrations of QSL_Pd^5A (1×, 10×, and 100× MIC) were added to each well. The plate was incubated at 37°C for 24 h with 5% CO₂. Furthermore, 20 µL of MTT solution (5 mg/mL) was added to each well and incubated for 3 h. Following the incubation period, 200 µL of DMSO was added to the wells, and optical density (OD) was taken at 570 nm. The percentage cell viability was calculated with reference to the untreated control cells using the below-given formula:

Docking simulations were conducted between the NorA EP (PDB ID: 7LO7) and QSL_Pd^5A using the Autodock 4.2 suite. The NorA protein structure model, acquired from a previous study conducted by our lab using the I-TASSER (Iterative Threading ASSEmbly Refinement) server (https://zhanglab.dcmb.med.umich.edu/I-TASSER/) (Thamilselvan et al., 2021), was employed. The protein preparation phase involved reading the NorA protein protein data bank (PDB) file, addressing missing atoms and adding polar hydrogen atoms. Kollman charges were subsequently incorporated into the protein, and the resulting file was saved as a.pdbqt file. During ligand preparation, the ligand molecule’s 2D structure was drawn using MarvinSketch (https://chemaxon.com/products/marvin) and saved as a 3D coordinate.pdb file. The ligand, QSL_Pd^5A, underwent optimization and energy minimization via Avogadro 1.2, employing the UFF force field and a conjugate gradient approach with 2,000 steps. The optimized ligand configuration was integrated into Autodock 4.2, followed by the addition of missing atoms and polar hydrogen. Gasteiger charges were applied to the ligand, and the result was saved as a.pdbqt file. Subsequent steps encompassed grid box generation, centered on the ligand binding site featuring residues Ile23, Pro24, Leu26, Pro277, Phe47, Arg98, Val144, Tyr225, and Gly348. The parameters for palladium(II) metal were sourced from the Autodock website. Docking procedures utilized the genetic algorithm search method with 50 conformations and 300 steps, yielding an output file derived from the Lamarkian genetic algorithm. The output file was scrutinized to identify the most optimal docking complex, as indicated by lower binding energy. The resulting complex was visualized using PyMOL, with interaction visualization facilitated by LigPlus Version V.2.2.8. Similar methodologies were applied to perform docking simulations for the positive control, verapamil, and the negative control ligand, QSL_P^A, aiming to comparatively assess the binding affinity of QSL_Pd^5A.

GraphPad Prism software version 8.0.2 (GraphPad Software Inc., San Diego, CA, United States) was employed for conducting the statistical analysis. To assess significance, a Student’s t-test was executed, with the minimum level of significance established at p ≤ 0.05. All assays were performed in triplicate, and the outcomes were presented as mean ± standard deviation (SD).

The ligands, QSL_P^A (“A” series) and QSL_P^B (“B” series), and their palladium(II) complexes, QSL_Pd^1A to QSL_Pd^6A and QSL_Pd^1B to QSL_Pd^6B, were synthesized through coupling reactions of commercially available α–picolinic acid with 2-(methylthio)aniline (for “A” series) or 2-(phenylthio)aniline (for “B” series) ( Figure 1 ). The newly synthesized ligands and their complexes were characterized through ¹H and ¹³C{1H} NMR spectroscopic techniques and few of them were confirmed through single-crystal X-ray diffraction studies. The details of the synthesis and characterization will be disclosed elsewhere.

The ligand of both “A” series (QSL_P^A) and “B” series (QSL_P^B) of palladium(II) complexes had no inhibition activity against SA1199 and SA1199B. The palladium(II) ion-centered complexes, “A” series (QSL_Pd^1A - QSL_Pd^6A) and “B” series (QSL_Pd^1B - QSL_Pd^6B) exhibited minimum inhibitory concentration (MIC)₅₀ across a varying concentration of 600 µg/mL to 1.71 µg/mL in both the strains and dose–response graph was plotted to mark the MIC₅₀ ( Figures S1, S2 ). A heat map for the MIC was plotted for both strains to correlate the concentrations ( Figure S3 ). The palladium(II) complexes belonging to the A series were identified to display a lower MIC₅₀ than the “B” series ( Tables 1 , 2 ). Thus, the “A” series of palladium(II) metal complexes were considered for further assays.

The MDR EP in S. aureus is a major contributor to the antibiotic resistance mechanism. NorA-overexpressing strains are known to be less susceptible to antibiotics, as they expel various classes of antibiotics, notably fluoroquinolones. The Pd “A” series was qualitatively evaluated for their EP inhibitory activity, and the basal level fluorescence emitted by the control was measured and compared with the Pd “A”-series complexes (QSL_Pd^1A to QSL_Pd^6A) and their ligand (QSL_P^A). The ligand and the untreated plates were taken as a negative control. High fluorescence emission indicates that the compound inhibits the EP and accumulates EtBr inside the cell and vice versa. From Figure 2 , the results were categorized into three groups on the basis of the fluorescence intensity. The first group consisted of only one compound QSL_Pd^2A that displayed high fluorescence emission and, hence, had a negative efflux. The second group and third group consisted of QSL_Pd^1A, QSL_Pd^3A, and QSL_Pd^5A and of QSL_Pd^4A and QSL_Pd^6A, respectively ( Table S1 ).

The quantitative EP inhibition activity of palladium(II) complexes that exhibited high and intermediate fluorescence (QSL_Pd^2A, QSL_Pd^1A, QSL_Pd^3A, and QSL_Pd^5A) in EtBr cartwheel assay against SA1199B was tested. A time-dependent decrease in the fluorescence was noted in control (untreated) and cells treated with ligand (QSL_P^A) due to the action NorA-mediated efflux of EtBr. On the contrary, the cells treated with the complex exhibited fluorescence retention. The palladium(II) complex QSL_Pd^5A and QSL_Pd^1A had higher relative fluorescence than the verapamil, a known EPI, which indicates that the inhibition of EP activity is more by QSL_Pd^5A and QSL_Pd^1A ( Figure 3 ). The cells treated with QSL_Pd^5A and QSL_Pd^1A exhibited 1.05-fold and 1.01-fold increase in fluorescence compared to verapamil. Thus, QSL_Pd^5A was identified as a potent NorA EPI and was taken for further studies.

A significant reason for chronic and hospital-acquired infections by S. aureus is attributed to biofilm formation. Hence, we examined the effect of QSL_Pd^5A in their sub-MIC (9.38 µg/mL to 0.29 µg/mL) on biofilm using crystal violet assay. A dose-dependent inhibition was obtained with 9.38 µg/mL and 4.69 µg/mL exhibiting highly significant biofilm reduction (∼99%) compared to the control ( Figure 4A ). The Live/Dead staining and fluorescence microscopy images visually substantiated the biofilm assay results. There was a significant biofilm reduction (>90%) with the treatment of QSL_Pd^5A at 0.5× and 0.25× MIC concentrations. A strong biofilm network was not visually observed in 0.125× MIC, although the percentage of biofilm inhibited was less. The ligand QSL_P^A had no effect on the inhibition of biofilm ( Figure 4B ).

A checkerboard assay was performed to evaluate the effect of CIP in the presence of QSL_Pd^5A, and the bliss independence model was used to assess the combinatorial effect. According to the model, the bliss score > 0 is synergistic, bliss score < 0 is an antagonistic effect, and bliss score = 0 is called an additive effect. The combinatorial action with varying concentrations of QSL_Pd^5A (9.38 µg/mL to 0.29 µg/mL) and CIP (4 µg/mL to 0.125 µg/mL) was analyzed by fitting into the bliss model ( Figure 5A ). A synergistic effect was observed across lower concentrations of QSL_Pd^5A and CIP and antagonistic effect was observed in higher concentrations of QSL_Pd5A (9.48 µg/mL) and CIP (4 µg/mL) with a bliss score of −104. Despite synergy being evident at 0.125 µg/mL with different QSL_Pd^5A concentrations, the most significant synergy score of 41.1 occurred when CIP was at 0.125 µg/mL and QSL_Pd^5A was at 4.69 µg/mL ( Figures 5B ). This resulted in a remarkable 64-fold reduction in the MIC of CIP against SA1199B. Thus, the bactericidal activity of CIP is significantly enhanced by QSL_Pd^5A compared to the individual effects of each drug on SA1199B.

The time-kill kinetics assay was performed to understand the time-based effect of CIP along with QSL_Pd^5A. SA1199B was exposed to CIP (at 0.125 µg/mL) along with QSL_Pd^5A (at 4.69 µg/mL) based on the results obtained from the combinatorial assay. A time-dependent decrease in the growth of the organism was observed. At the 8th hour, a significant inhibition in the growth (∼80%) was seen, and the effect tend to be retained for a short period ( Figure 6 ).

The MTT assay revealed dose-dependent toxicity in NRK-62E cells when exposed to the compound at its MIC concentration. However, the rate of cell death was insignificant. The percentage of cell viability remains similar to the control (untreated cells) even when cells were exposed to compound concentrations at 10× and 100× the MIC ( Figure 7 ).

Docking was performed between NorA protein and QSL_Pd^5A and docking with ligand QSL_P^A and Verapamil (positive control) ( Figures 8A, B ). The binding energy score is given in Table 3 . The binding energy of QSL_Pd^5A was greater than the positive control verapamil. The binding energy of QSL_Pd^5A was −8.41 kcal/mol, indicating higher stability and affinity of the palladium(II) complex toward the NorA protein. Thus, it proves that QSL_Pd^5A inhibits NorA EP. The interaction sites and amino acid residues at the binding sites are visualized using LigPlus ( Figure 8C ). There are more hydrophobic interactions between QSL_Pd^5A and NorA protein’s residues in the hydrophobic binding pocket like 16Phe, 19Ile, 23Ile, 47Phe, 140Phe, 222Glu, 306Phe, and 336Thr along with few pi-pi stacks, which are also observed ( Table S2 ). Table S3 depicts the amino acids and their position with their interaction type.

The preliminary studies on QSL_Pd^5A demonstrated a substantial decrease in efflux activity and the effective prevention of biofilm formation. QSL_Pd5A’s inhibition of biofilm formation exposes planktonic cells to antibiotics at a remarkably low concentration. In addition, the compound impedes the extrusion of antibiotics, affirming a prolonged and sustained effect at a lower concentration of the conventional antibiotic. Importantly, the complex demonstrates no cytotoxicity, positioning it as a promising therapeutic molecule for topical applications. This opens avenues for further exploration of the compound through pharmacokinetic and pharmacodynamic studies. An in-depth in vivo investigation is required to assess the clinical efficacy of the compound as a pharmaceutical agent.