All chemicals used in this study were of 100% purity. Potassium dichromate, lead acetate, and zinc chloride, the heavy metal salts, were sourced from Himedia. The chemicals for media preparation were obtained from Himedia and SRL, India. Reagents for the plant growth-promoting rhizobacteria (PGPR) study were procured from Sigma and SRL, India. Stock solutions for each of the three metals were prepared at a concentration of 50,000mg/l.

Wetland paddy rhizosphere soil samples were collected from a depth of 20cm from the agricultural lands of Ranipet Industrial Estate (12.550°N–79.170°E). The SIPCOT (State Industries Promotion Corporation of Tamil Nadu) industrial corridor, located near Vellore, Tamil Nadu, India, is primarily dominated by leather, chemical, and heavy metal fabrication industries. The study was conducted using a random sampling method. A total of 12 soil samples were collected in sterile polyethylene bags and were transported to the laboratory within 2h for the isolation of bacteria capable of tolerating Cr, Pb and Zn. The physicochemical properties of the soil such as pH, conductivity and other micronutrients were analyzed and are mentioned in Supplementary Table S1 .

The collected soil samples were pooled and serially diluted, plating was performed from 10⁻⁴ and 10⁻⁵ dilutions into modified LGI agar plates supplemented with 100ppm of Cr and 250ppm of Pb and Zn in the form of potassium dichromate, lead acetate and zinc chloride respectively. The concentrations selected was 10 folds higher than the permissible limits suggested by ATSDR, EPA and APHA due to the variation in concentration that is expected to be released from industries. For enrichment technique, the soil samples were transferred into LGI liquid medium supplemented with same metals used for LGI solid medium and the tubes were incubated for 7 days at 120 rpm. Upon enrichment, serial dilution was performed to obtain morphologically distinct colonies which were further purified and maintained at 50% glycerol stock (36).

The catalase activity was estimated using hydrogen peroxide (H₂O₂) as described by Das and Osborne (40). The reaction mixture consisting of 0.1ml of cell free broth, 0.9ml of phosphate buffer and 20mM H₂O₂ solution in sodium potassium phosphate buffer at pH7.0 and the reaction was initiated and incubated for 1min at room temperature. To stop the reaction 1ml of 1mM sulfuric acid (H₂So₄) was added. Initial and final absorbance after 1 min of incubated was recorded spectrophotometrically at 240nm. The reaction mixture without cell free broth served as control. For determination of chromate reductase activity, diphenyl carbazide (DPC) was used. To measure the reduction of hexavalent chromium (Cr VI) to trivalent chromium (Cr III). The 200µl reaction mixture consists of 80 µl of cell-free broth, 0.1 mM NADH, and 12.5mg/l of K2Cr2O7 (Cr VI) in a potassium phosphate buffer at pH 7.0. A reaction mixture without the cell-free broth serves as the control. The mixture is incubated at 30°C for 30 minutes, after which the residual Cr VI is quantified using the DPC reagent (41). Superoxide dismutase (SOD) activity was performed by adding 0.025mM of Nitro Blue Tetrazolium (NBT) solution (0.1ml), 0.1mM of NADH solution (0.1ml), 0.8ml of sodium phosphate buffer at pH 7.4 and 0.1ml of cell free lysate were mixed thoroughly and the reaction mixture was incubated at room temperature for 5 min and the absorbance was measured at 560nm using spectrophotometer (42).

The effective strain VITVJ8 was characterized morphologically and biochemically by the conventional tests such as gram staining, IMViC, catalase and oxidase (43). Genomic DNA is extracted from VITVJ8 using fluorescent dye terminator technique, ensuring the purity and integrity of the DNA for downstream applications. The 16S rRNA gene, which is highly conserved among bacteria, is amplified using polymerase chain reaction (PCR) with universal primers 27F and 1492R.

The amplified 16S rRNA gene PCR product is purified to remove any unincorporated nucleotides, primers, and other contaminants. The purified PCR product is sequenced using automated DNA sequencers to determine the exact nucleotide sequence of the 16S rRNA gene. The obtained sequences are analyzed and compared with NCBI’s GenBank. BLAST (Basic Local Alignment Search Tool) is used to find the closest matches and identify the bacterial isolate. A phylogenetic tree was constructed using the neighbor joining method using mega 2.0 software. The tree was bootstrapped to indicate the evolutionary distance of the effective strain to its neighbor (44, 45).

The impact of heavy metals such as Cr, Pb and Zn on the strain VITVJ8 was studied by comparing the growth conditions in the presence and absence of these metals. Throughout the experiment concentration of Cr was maintained at 100ppm, Pb and Zn concentrations were maintained at 250ppm with 2% bacterial seed culture (0.27 OD). The flask was incubated at 28 ± 2°C and the growth was monitored on 12h intervals until the cells reached the decline phase.

The optimization of heavy metal removal parameters included the factors such as carbon source (glucose, sucrose and glycerol), nitrogen source (yeast extract, ammonium sulphate and urea) and pH (ranging from 4.5–8.5). These factors were designed using RSM, a design expert statistical software (V11, stat-Ease Inc. Minneapolis, Mn, USA). Preliminary experiments were carried out to determine the optimal range for each parameter.

Data analysis was performed using a quadratic model approach where one factor was analyzed in reference to the other two factors at three different level +1, 0 and −1. The percentage of heavy metal removal was designated as the response variable (R1) upon 20 experimental runs. Validity of the model and the output response were analyzed based on the results obtained from ANOVA (Analysis of Variance) and 3D contour plots (46).

Using the optimized parameters, the effective bacterial isolate was inoculated into LGI medium at the specified concentrations. The flasks were incubated at 28 ± 2°C and after incubation the cells were harvested by centrifugation at 11,000 rpm for 20 min at 4°C. Concentration of heavy metals in both supernatant and pellet was estimated using Atomic Absorption Spectrometer (AAS) model variant Spectraa 240 (47, 48).

All the isolates were assessed for PGPR traits such as the production of IAA (indole acetic acid), siderophore, ammonia and the ability of the isolate in solubilizing insoluble phosphate. For the indole acetic acid (IAA) production, to the 48h culture free supernatant, Salkowski’s reagent and orthophosphoric acid were added and change in color from yellow to pink indicated the production of IAA which was further quantified spectrophotometrically at 530nm (50). For the siderophore production, supernatant of 48h cell free supernatant of Fiss minimal medium was added in the wells of Chrome Azurol S (CAS) agar and formation of halozone indicated siderophore production (51). For ammonia production, Nessler’s reagent was added to the 48h culture supernatant, change in color from yellow to brown indicated ammonia production (52). The ability of the isolate in converting insoluble tricalcium phosphate to soluble form was studied in Pikovskaya agar plates where the formation of halozone indicated efficient solubilization (53). The solubilization efficiency (SE) was calculated by the formula (49),

Pot culture study was performed to observe the efficiency of the effective isolate in combination with the plants, in enhancing plant growth and thereby increasing the uptake of heavy metals. C. zizanioides (vetiver) was selected for the study due to their fast-growing potential, fibrous root system and their robust nature to survive in extreme conditions like drought and water logging. Plantlets of vetiver were collected from the VIT nursery and were thoroughly washed to remove soil particles. Before planting, the plantlets were pruned to 5cm of root and shoot. The pruned plantlets were then planted in pots containing 2kg of soil. Soil used for the pot culture study was sieved (2mm) to remove the denser particles and was packed into a 2kg LDPE bags.

All the study was carried out in triplicates and the statistical analysis was conducted for the bioremoval of Cr, Pb and Zn by Bacillus infantis in soil and plants based on the time intervals using Origin software version 8.0. The optimization of data was evaluated using Design Expert software version 12.0. Response Surface Methodology (RSM) was employed with a quadratic model approach. Plant growth parameters was analyzed using Analysis of Variance (ANOVA) for the statistical significance.

A total of eight morphologically distinct colonies VITVJ1 to VITVJ8 were isolated on LGI medium supplemented with 100ppm of Cr, 250ppm of Pb and Zn. Lalhriatpuii et al. (56) revealed the potential of rhizobacteria in the removal of heavy metals such as Cr, Pb and Zn. There are several studies showing the application of MSM for isolation of bacteria capable of surviving high concentrations of heavy metals. Concentration used for the study was based on the previous reports by Das and Osborne (40). For the isolation of heavy metal resistant bacteria wetland paddy rhizosphere soil was chosen due to the presence of increased abundance of bacterial diversity in the rhizosphere (57). Tekaya et al. (58) reported the role of rhizobacteria in restoring metal-contaminated soils, and the mechanisms by which PGPRs assisted in heavy metal bioremediation was studied by Riseh et al. (59).

There was an increase in enzyme activity in response to various concentrations of Cr, Pb and Zn. Notably, SOD and catalase activities were significantly enhanced with the increase in Pb concentration of up to 150ppm. Catalase activity also showed a positive correlation with Zn concentration of up to 150ppm ( Supplementary Figure S2 ). Kumar et al. (63) reported similar pattern, where increase in SOD activity was directly proportional to the increase in Zn concentration. Similarly, reports by Shi et al. (64) and Abedi and Shahpiri (65) highlighted the crucial role of SOD in scavenging superoxide radicals, which acts as a defense mechanism against oxidative stress by breaking down H₂O₂ into H₂O and O₂. Therefore, it can be substantiated that bioremoval process was catalyzed by the activities of superoxide dismutase (SOD), chromate reductase and catalase.

In contrast, chromate reductase activity decreased with a Cr concentration of 150ppm, indicating the inverse proportionality of the isolate in resisting chromium. This finding aligns with the report of Baldiris et al. (66), who also stated the bacterial tolerance to Cr VI at 100ppm (40). The activity of chromate reductase was further elucidated, with expression observed in membrane associated chromate reductase when cells were pre-exposed to Cr VI (67). Therefore, the resistance in isolates is not by a single enzyme but a collective resistance obtained by synthesis of all the three enzymes.

The bacterial strain VITVJ8 was identified as a Gram-positive, rod-shaped endospore forming motile bacteria. It was also found to be oxidase negative and catalase positive. The isolate was further characterized and identified through 16S rRNA gene sequencing. Molecular sequencing and phylogenetic tree analysis revealed that VITVJ8 has a close evolutionary relationship with Bacillus infantis. The sequence has been submitted to GenBank with the accession number OM250034 ( Supplementary Figure S3 ).

Growth kinetics of the isolated bacteria, both in the presence and absence of heavy metals (100ppm Cr, 250ppm Pb and Zn), showed an increased lag phase, followed by the steady exponential phase with increased cell growth ( Supplementary Figure S4 ). In the presence of Cr, Pb and Zn, there was no significant changes observed in growth pattern of VITVJ8. There are several studies indicating the presence of extended lag phase upon exposure to heavy metals. Zn and Cu uptake by Pseudomonas sp. has been reported to have a prolonged lag phase upon treatment (68–72).

The optimization of bacterial growth was primarily influenced by factors such as carbon (glucose, sucrose and glycerol) and nitrogen sources (yeast extract, ammonium sulphate and urea) and pH levels (ranging from 4.5 to 8.5). Among the tested carbon sources, the highest degradation was achieved when the media was supplemented with 1% of glucose as the primary carbon source and the optimal pH for maximum degradation was determined to be 6.5. Further, for nitrogen source, 1% yeast extract yielded the highest degradation. The results obtained was validated using a Response Surface Methodology (RSM) model. The results corresponded with the experimental trials, confirming the reliability of the optimization study carried out through both experimental and RSM models ( Figure 1 ).

The 3D surface diagrams and contour plots of the Box Behnken Design (BBD) model are illustrated in Figures 1A–F . The actual and predicted output responses were validated using ANOVA quadratic model ( Table 2 ). Table 2 also shows the F values for pH (745.87), carbon source (17.52) and nitrogen source (1.02) with corresponding P values being 0.0001, 0.0019, 0.3354 respectively. The model was validated with two among three responses having P values less than 0.05. The pr.ecision value, calculated as the ratio between signal and noise, was 350.410 which exceeds the adequate precision threshold of 4, indicating a strong signal strength. The 3D diagrams ( Figures 1A–C ) and contour plots ( Figures 1D–F ) showed that optimal input variables significantly influenced HM removal, both independently and in combination. Supplementary Table S2 provides the RSM design data, consisting of 20 experimental runs. Among the various factors studied which includes A (pH), B (carbon source) and C (nitrogen source), it was observed that the interactions AB, AC, and BC, as well as the quadratic terms A², B², and C², were significant. The following BBD equation was derived based on the coded factors for heavy metal removal Percentage (%): 97.08 – 1.19A + 0.1817B – 0.0439C – 0.1250AB – 0.2500AC – 0.2250BC – 13.41A2 – 6.90B2 − 0.0086C2.

This equation was used to predict the responses and assess the relative impact of the optimal factors. The results obtained from RSM and the derived statistical values were found to have relative impact within standard parameters, affirming the effective growth of the bacterial isolate VITVJ8 with the optimized parameters. The optimization using RSM for bioremoval studies has been reported in several studies (32, 46, 73, 74).

The strain VITVJ8 showed the ability to produce siderophores, IAA and solubilize the insoluble phosphate, exhibiting its plant growth promoting traits (PGPR) ( Table 1 ). The quantitative assay revealed that 59μg/ml of IAA was produced upon supplementation with 5mg/ml of tryptophan. Ahemad and Kibert (83) reported that IAA can also be produced by bacteria such as Azospirillum, Pseudomonas, Bacillus and Cyanobacteria. The presence of halozone around VITVJ8 in NBRIP medium indicated the solubilization of insoluble phosphate to its soluble form with an efficiency of 61.5%. Similarly, halozone formation of 1.165cm in CAS agar plates indicated siderophore production, which can be used for mobilization of iron. Hou et al. (84) reported an increase in the siderophore production in iron-depleted conditions. Tsotetsi et al. (85) and Gupta et al. (11) reported that Bacillus sp. are prominent PGPR which have the ability to tolerate stress and produce various biomolecules that can contribute to PGPR traits. Similarly, Itusha et al. (54) reported Aeromonas sp. (VITJAN13) as an efficient plant growth-promoting rhizobacterium (PGPR) capable of producing 46.5µg/ml of indole-3-acetic acid (IAA). The bacteria also demonstrated the ability to produce siderophores, which was confirmed by the formation of halo zones. Several strains such as Enterobacter cloacae (VITPASJ1), Bacillus xiamenensis (VITMSJ3), Klebsiella sp. (VITAJ23) have proven to be effective PGPR (43, 49, 86).

Pot culture study was performed using C. zizanioides to understand the uptake of heavy metals in the presence and absence of VITVJ8 in the rhizospheric region.