Soil, sewage, and/or water samples were collected in sterilized plastic bottles from the discharge waters of the tannery industry areas of Sialkot, Kasur, and Islamabad in Pakistan. The samples were subsequently brought to the laboratory and stored at 4°C until further use. The effluent samples were analyzed for heavy metals (Pb, Cd, Cu, Cr, and As) using an atomic absorption spectrophotometer (Perkin–Elmer, USA). Standard stock solutions (1,000 mg/L) of the metals were procured from Sigma–Aldrich, USA.

For the isolation of bacterial strains, the samples were diluted in phosphate-buffered saline (PBS) solution supplemented with increasing concentrations of heavy metals (100 mg/L/day). The supernatant was streaked on agar plates containing nutrient agar (NA) from different media, tryptic soya agar (TSA), or marine agar (Difco™, USA) supplemented with 600–1,200 mg/L of heavy metals (Pb, Cd, Cu, Cr, and As). The plates were then incubated at 28°C. The heavy metals were added using the salts lead nitrate (Pb(NO₃)₂). Cadmium nitrate (Cd(NO₃)₂), copper sulphate (CuSO₄.4H₂O), potassium dichromate (K₂Cr₂O₇), and sodium dihydrogen arsenate (NaH₂AsO₄). Growth was observed after 24–72 h or until the appearance of bacterial colonies. The isolated colonies exhibiting distinct morphologies (in terms of texture, shape, margin, color, and elevation) were further purified via the subculturing method. The purified cultures of the bacterial strains were maintained on agar plates and stored at-80°C in 35% glycerol stock solution.

The purified bacterial colonies were morphologically characterized based on colony color, form, elevation, and margin. The cells of the isolates were also analyzed for Gram staining, morphology, and motility using a microscope (Olympus, CX31 equipped with Digital Camera 5A). Growth characteristics of the bacterial strains were determined across a pH range of 4 to 9, a temperature range of 3 to 50°C, and their tolerance to NaCl concentrations of 0–30%. The cells were grown in tryptic soya broth at a range of pH values, and growth was observed after 24 h using a spectrophotometer (IMPLEN, Germany) at a wavelength of 600 nm. The temperature range was determined by growing the cells on TSA and incubating them at various temperatures (4 to 50°C). Salt tolerance was determined by growing bacterial strains on TSA plates supplemented with NaCl at concentrations ranging from 0 to 30%, with 1% increment.

The maximum tolerance limit (MTL) for each heavy metal by the isolated bacterial strains was determined according to the methods described by Malik and Jaiswal (2000), in which the MTL was defined as the highest concentration at which visible bacterial growth was observed, while complete inhibition of growth at the next higher concentration was considered the threshold. For MTL evaluation, the media were supplemented with various heavy metals (Pb, Cd, Cu, Cr, and As) using their salts as mentioned above, initially at a concentration of 300 mg/L, with a gradual increase of 300 mg/L up to the MTL for the tested isolate. The heavy metal-containing plates were subdivided into four equal sectors, and the isolates were streaked separately in each quarter. The same procedure was carried out with control plates (those plates without metal). Each sample was analyzed in triplicate. Finally, the plates were kept at 28°C for 4 to 6 days to observe the growth of bacteria. The MTL concentration at which the tested isolate failed to grow was subsequently determined.

To determine the biosorption of heavy metals, 22 strains (which exhibited maximum tolerance against different heavy metals) were tested in this study, with 3 independent replicates. Escherichia coli was included as a biological reference strain. Each isolate was cultivated in 5 mL of tryptic soy broth (TSB) supplemented with 50 mg/L of a single heavy metal (Pb, Cd, Cu, Cr, or As) and incubated at 30°C with shaking for 48 h. Bacterial growth was monitored by measuring the optical density at 600 nm (OD₆₀₀) to ensure consistent culture density across all assays. After incubation, 2 mL culture samples were centrifuged at 7,500 × g for 10 min. The supernatants were collected to determine the residual metal concentration, while the pellets were dried at 60°C overnight and digested in a 5:3 mixture of nitric acid and perchloric acid. The metal content in both the supernatant and biomass was quantified using inductively coupled plasma–optical emission spectrometry (ICP–OES, Optima 8300, Perkin Elmer, USA). Biosorption was calculated based on the difference between the initial and final metal concentrations. In addition, abiotic controls (metal-containing medium without bacterial inoculation) were included to account for non-biological metal precipitation or adsorption (Chen et al., 2006). Statistical analysis was performed following the GLM procedure in SAS (version 9.4) (SAS Institute, Inc., Cary, USA).

The bacterial strains were identified based on the sequence analysis of the 16S rRNA gene, as described previously (Ahmed et al., 2007). For this purpose, the DNA template was extracted from fresh cells of the strain by colony PCR at 94°C for 10 min. The 16S rRNA gene was amplified in a thermal cycler (Applied Biosystems, Veriti, USA) by using a Premix Ex-Taq Kit (Takara Cat # RR003A, Japan) with forward primer 9F (5′-GAG TTT GAT CCT GGC TCA G-3′) and reverse primer 1510R (5′-GGC TAC CTT GTT ACG A-3′) under the following PCR conditions: pre-denaturation for 2 min at 94°C (1 cycle), denaturation for 1 min at 94°C (30 cycles), primer annealing for 1 min at 50°C, extension for 1.30 min at 72°C, and a final extension for 5 min at 72°C. The amplified 16S rRNA gene was visualized on a 0.8% (w/v) agarose gel and subsequently purified using a purification kit (Invitrogen, USA), according to the manufacturer’s protocol. The purified PCR product of the 16S rRNA gene was sent for sequencing by Macrogen, Korea,¹ using the forward primer 27F (5′-AGA GTT TGA TCM TGG CTC AG-3′) and the reverse primer 1492R (5′-ACC TTG TTA CGA CTT-3′). The software package ‘BioEdit’ was used to edit and construct an assembled consensus sequence, which was subsequently subjected to a BLAST search against the DNA Data Bank of Japan (DDBJ) and the EzTaxon server for identification of the strains. For the phylogenetic analysis of the isolated strains, the sequences of the closely related 16S rRNA gene were retrieved from validly published databases. Sequence alignment was carried out by using ClustalW (version 1.6) (Thompson et al., 1997), and phylogenetic analysis was performed to determine the evolutionary relationship of the strain with other validly published strains. Phylogenetic trees were constructed using three algorithms (data not shown in this study): maximum parsimony (MP), maximum likelihood (ML), and neighbor-joining (NJ), which were generated with the software package MEGA-6 (Tamura et al., 2013). The DNA sequences were submitted to the DNA databank of Japan, whose accession numbers are listed in Supplementary Table S1. The strains were submitted to the National Culture Collection of Pakistan (NCCP), and the most important strains (with respect to biosorption capacity or novel species) were also deposited in the Japan Culture of Microorganisms (JCM).

The screening of isolates for the presence of the nifH and acdS genes was performed using different primer sets to detect the specific amplicon of the respective gene. For this purpose, genomic DNA was extracted from young bacterial cells (14–16 h) using a QIAamp DNA Mini Kit following the manufacturer’s instructions (Qiagen Cat # 51304, Germany).

Amplification of the acdS gene was performed in a 50 μL volume using a Premix Ex-Taq kit (Takara Cat # RR003A, Japan) with four sets of primers (Supplementary Table S2) and 50–100 ng of genomic DNA as a template. Amplification was performed with the following PCR procedure: pre-denaturation for 5 min at 94°C (1 cycle), denaturation for 30 s at 94°C (30 cycles), annealing for 30 s at 50°C, extension for 30 s at 72°C, and a final extension for 7 min at 72°C (Blaha et al., 2006). The DNA band of the expected amplicon size was analyzed with each primer set on a 0.8% agarose gel.

Amplification of the nifH gene was performed using a Premix Ex-Taq kit (Takara Cat # RR003A, Japan) with three sets of primers (PolF/PolR, nifHF/nifHI, and nifHfor/nifnai; Supplementary Table S2) and 50–100 ng of genomic DNA as a template using the aforementioned PCR conditions for each set of primers (Supplementary Table S2) (Laguerre et al., 2001; Poly et al., 2001; Sarita et al., 2008). The DNA band sizes of the expected amplicons with each primer set were analyzed on a 0.8% agarose gel.

A greenhouse experiment was performed to test the PGPR activity of selected heavy metal-tolerant strains (NCCP-602, NCCP-614, NCCP-644^T, and NCCP-650^T) and a reference nitrogen-fixing strain, Bradyrhizobium diazoefficiens JCM 10833, in soil-packed plastic pouches. Brassica napus seeds were grown in plastic pouches filled with soil and were watered with 50 mg/kg of each of five heavy metals, namely Pb, Cu, Cr, Cd, and As, separately (using salts Pb(NO₃)₂, CuSO₄.4H₂O, K₂Cr₂O₇, Cd(NO₃)₂, and NaH₂AsO₄, respectively) during the whole growth period. The experiment was performed according to a complete random design (CRD) with three independent replications, keeping the strains as a significant factor than those in the heavy-metal treatment. The plants were harvested after 68 days of growth, and plant growth parameters (shoot length and shoot dry weight) were recorded. The statistical analysis was performed following the GLM procedure in SAS version 9.4 (SAS Institute, Inc., Cary, USA). Least squares means were estimated for the main effects of the metal and strain as well as their interaction effect. The standard error of the difference between means was calculated using the estimate statement in the model. In Figure 1, statistical differences among treatment groups are indicated by alphabetic letters, based on post-hoc comparison of least squares means.

Effluent samples were analyzed for various physicochemical parameters, including metal ion discharge, pH of the effluent samples, and physical appearance. The pH of the samples was mostly alkaline (pH 7.5 to 8.5), while the majority of these samples were highly colored and had a foul smell. The Cr and Pb concentrations in the majority of the samples were above the permissible limits of 0.1 mg/L.

A total of 68 strains were isolated from the discharge of industrial areas (Islamabad, Sialkot, and Kasur) in Pakistan, and the purified strains were subsequently designated as NCCP-601 and onward. The isolated strains were enriched with different heavy metals and differentiated based on colony morphology (Supplementary Table S1). The majority of the strains were round, lobed, and filamentous in shape, with entire margins; however, some strains had irregular margins. The colony colors of the majority of strains were white and pale yellow, while some strains were peach and off-white. The morphologically different strains were further subjected to other experiments and stored in a freezer at −80°C.

All the enriched bacterial strains were found to be highly tolerant to different heavy metals, including Cr, Cu, Cd, Pb, and As (Figure 2). The results demonstrated that the majority of the isolated strains tolerated 3,600 mg/L Cr, 3,300 mg/L Cu, 3,000 mg/L Cd, 1,500 mg/L Pb, and 1,200 mg/L As. Among these isolates, NCCP-601, 602, 603, 621, 627, 647, 653, 657, 660, and 661 had maximum tolerance limits (MTLs) for Cr (3,600 mg/L), Cd (2,400–3,000 mg/L), Cu (2,100–3,300 mg/L), Pb (1,200–1,500 mg/L) and As (900–1,200 mg/L). The isolates reported in these studies showed the highest tolerance to Cd and As compared to previous reports. The majority of the bacterial isolates were also found to grow over a wide range of sodium chloride (NaCl) concentrations, ranging from 0 to 20%, as shown in Supplementary Table S1.

Based on the MTL, the isolated strains were further analyzed for possible use in heavy metal biosorption from 50 mL of TSB containing 50 mg/L of the corresponding heavy metals. All the isolated bacterial strains showed significant reductions in heavy metal concentrations in the TSB medium. The maximum biosorption occurred for Pb, followed by Cd and Cu. However, no significant reduction in Cr or As concentration was observed for any of the isolated strains (Figures 3d,e). The isolates significantly differed in their biosorption of Pb, Cd, and Cu (Figures 2a–c), but there was no significant difference in the biosorption of As or Cr (Figures 3d,e). The same alphabetic letter is used for each metal treatment, and the bars in Figure 3 indicate statistically non-significant differences (p = 0.05). Among these, three isolates, namely NCCP-614 (99%), NCCP-605 (96%), and NCCP-655 (91%), exhibited maximum biosorption of Pb (Figure 3a). Similarly, strains NCCP-614 and NCCP-655 also exhibited maximum biosorption capacities of 89 and 59% for Cd, respectively (Figure 3b); however, the maximum biosorption of Cu was achieved by NCCP-625 (42%), followed by NCCP-619 (38%) and NCCP-647 (36%) (Figure 3c). Our results indicated that these isolates can be used for the bioremediation of soil and water contaminated with heavy metals such as Pb, Cd, and Cu.

It is accepted that culturable microorganisms from any given sample taken from an environment represent only a small portion of the total population that is present. In this study, 68 bacterial strains were identified taxonomically based on the sequence of the 16S rRNA gene (Supplementary Table S1). These 68 strains were isolated from effluent samples and found to be highly tolerant to heavy metals. Based on comparative 16S rRNA gene sequence data, a diverse bacterial community was observed (Figures 4a,b). The isolated population (Figure 4a) belonged to three phyla, Actinobacteria (6%), Firmicutes (38%), and Proteobacteria (56%). These heavy metal-tolerant strains (Figures 4a,b) were related to 19 different genera. The dominant strains were Bacillus (21%), Pseudomonas (12%), and Staphylococcus (10%).

The percent 16S rRNA gene sequence similarity with those of closely related species in their respective genera was 97.9 to 100% for Bacillus, Alcaligenes, Acinetobacter, Pseudomonas, Citrobacter, and Bravibacterium (Figure 4b; Supplementary Table S1), which indicates that some of these strains could be characterized taxonomically to delineate them as novel species. Taking these results into account, along with phylogenetic analyses, DNA–DNA homology, and phenotypic and chemotaxonomic data, three isolated strains have been characterized as novel species: Acinetobacter pakistanensis sp. nov. (Abbas et al., 2014), Alcaligenes pakistanensis sp. nov. (Abbas et al., 2015a), and Bacillus malikii sp. nov. (Abbas et al., 2015b). However, the 16S rRNA gene sequences of the other strains had high similarity (>99%) with those of the closely related taxa in their respective clusters; therefore, these strains were not included in the taxonomic characterization studies.

To determine the potential use of isolated heavy metal-tolerant isolates in agriculture, phylogenetically different isolates were also analyzed for the screening of the nitrogen fixation gene nifH using different primer sets, such as PolF/PolR, nifHF/nifHI, and nifHfor/nifHrev (Supplementary Table S2). Rhizobium etli JCM 21823^T and Bradyrhizobium japonicum JCM 10833^T were used as positive controls. The nifH gene was amplified by these primers, and amplicons of approximately 360, 420, and 780 bp, previously reported to be associated with the presence of the nifH gene, were identified (Supplementary Table S3). The nifH gene was amplified in at least 15 isolates with one or two primer sets (Figure 5). Similarly, the presence of the acdS gene in the genome was analyzed using four sets of primers, F1936f/F1938r, F1936f/F1939r, and F1937f/F1939r (Supplementary Table S2). The ACC deaminase gene acdS was amplified in at least 8 strains by these primers, with approximate sizes of 792, 558, and 516 bp. A careful analysis of the results showed that at least 7 strains (NCCP-650^T, NCCP-611, NCCP-660, NCCP-635, NCCP-622, NCCP-614, and NCCP-605) were found to have both nifH and acdS genes (Figure 5). Among them, strain NCCP-650ᵀ, previously described as a novel species (Alcaligenes pakistanensis), exhibited the most pronounced plant growth-promoting effect and carried the nirK gene (Abbas et al., 2015a), potentially contributing to nitrogen cycling and environmental adaptation. The cooccurrence of functional marker genes with observed biosorption and PGPR traits suggested a possible role for these genes in the adaptive performance of the strains. Overall, the identified strains, particularly those with combined biosorption capacity, PGPR traits, and genetic potential, represent strong candidates for microbial-assisted bioremediation and development as bioinoculants in metal-contaminated agroecosystems.

Based on the presence and absence of the nifH and acdS genes, as well as MTL, four isolates (NCCP-650^T, NCCP-614, NCCP-644, and NCCP-602) and the reference nitrogen-fixing strain JCM 10833 were evaluated for their impact on Brassica napus growth under axenic conditions when treated with a 50 mg/L solution for each studied metal separately. The statistical analysis revealed that the effects of strain, heavy metals, and their interaction were significant (p < 0.05) on shoot fresh weight and length. Additionally, all the strains used in this study played a significant role in increasing the fresh shoot weight and length of Brassica plants compared to those of the control plants (with no addition of strain or metal) when irrigated with water containing 50 mg/L Pb or Cd (Figures 1a,b). However, the application of water containing Cr or As salts significantly reduced plant growth, and no positive response was observed in plants irrigated with water containing 50 mg/L As because shoot length and shoot fresh weight were negligible in comparison to those in the other heavy metal treatments. Among the strains tested, NCCP-650^T demonstrated the greatest improvement in Brassica growth. Additionally, many strains of this genus have been reported to have PGPR activity (Duca et al., 2014).