Stock solutions of As(V) (Na₃AsO₄), As(III) (Na₂AsO₂), Cr(VI) (K₂CrO₄), Cu (Cu₂SO₄), Co (CoCl₂.6H₂O), Cd (CdCl₂), Hg (HgCl₂), Se (Na₂SeO₃), Pb (PbCl₂), Ni (NiCl₂), and Zn (ZnSO₄) were prepared and filter sterilized. Contaminated soil samples were collected from the wastewater treatment plant at Kasur, Pakistan (31°05′40.0″N, 74°28′04.7″E), in sterile containers. The wastewater treatment plant receives wastewater from local industries, which are mostly tanneries. The soil samples were transported to the laboratory; pH and temperature were recorded. Multi-metal-resistant bacteria were isolated using an enrichment culture technique by adding 1.0 g of soil in a tryptic soya broth (TSB) containing 1.0 mM each of As(III), As(V), Cr, and Se. After 24 h of incubation at 37 °C and 150 rpm, 1.0 mL of the culture broth was transferred into a fresh TSB flask containing the above-mentioned metals as well as 1.0 mM of Zn, and was incubated at 37 °C under 150-rpm agitation. The above process was repeated until the culture was supplemented with As(III), As(V), Cd, Co, Cu, Pb, Se, Ni, Zn (1.0 mM each) and Cr and Hg (0.5 mM each) [11]. Cultures from every enrichment flask were serially diluted and plated on Luria–Bertani (LB)-agar plates. CFU·g⁻¹ was counted and diverse colonies from the last enrichment culture (containing all the metals) were purified by repeated quadrant streaking on LB-agar.

For the determination of the metal resistance of isolates, mineral agar media plates [12] were prepared containing As(III), As(V), Co, Cd, Cu, Se, Ni, Pb or Zn ranging from 1.0 to 30 mM. For Cr and Hg, concentrations ranging from 0.5 to 30 mM were used [11]. Bacterial isolates were cultured on these plates. After 72 h of incubation at 37 °C, the presence or the absence of growth was recorded.

Selected isolates were characterized morphologically and biochemically [13] (see supplementary material). The optimal growth conditions (pH and temperature) of the isolates were also determined. Overnight bacterial cultures were standardized at 0.2 OD_600nm and inoculated in the LB broth, followed by incubation at 15 °C, 25 °C, 37 °C and 45 °C (pH 7.0). For optimal pH determination, LB broth in four different batches was prepared and set at pH 3, 5, 7 and 9. The medium was inoculated with the same standardized cultures and incubated at 37 °C. After 24 h of incubation, the optical densities were recorded at 600 nm.

The bacterial isolates were cultured on LB-agar media plates supplemented with antibiotics such as ampicillin (1.0 μg·mL⁻¹), chloramphenicol (3.0 μg·mL⁻¹), erythromycin (1.5 μg·mL⁻¹), kanamycin (3.0 μg·mL⁻¹) and tetracycline (3.0 μg·mL⁻¹) and incubated at 37 °C. After incubation, the plates were observed for bacterial growth.

The extent of As-oxidation/reduction was assessed by the assay reported by Cummings et al. [14], with minor modifications. Overnight bacterial cultures were centrifuged, and pellets were washed with and resuspended in 0.9% NaCl (pH 7.0). The suspension was divided into two sets, one was supplemented with 1000 μM of As(V), while the other one was supplemented with 500 μM As(III) and incubated at room temperature for 5 h. Supernatants were collected by centrifugation before the start of incubation as well as after incubation.

To quantify As(V) reduction into As(III) or As(III) oxidation into As(V), 100 μL of each supernatant was acidified with 100 μL of 24 mM HCl. One hundred microliters of the acidified samples were added to 900 μL of the colouring reagent, which contained 9.9 g of ascorbic acid, 0.132 g of potassium antimony tartrate, 6 g of ammonium molybdate, and 500 mL of 5 N sulphuric acid per litre. The reaction mixture was heated at 78 °C for 20 min on water and incubated on ice for 5 min. The absorbance was determined at 865 nm using a spectrophotometer (Cecil CE 7200).

As this assay measures As(V) rather than As(III), the samples were oxidized to confirm As(V)-reduction. As(III) produced in the samples [due to As(V) reduction] was oxidized back to As(V) by addition of 100 μL of 5 mM KIO₃ and 48 mM of HCl. The mixture was heated at 78 °C for 5 min and 100 μL of the oxidized samples were added to 900 μL of colouring reagent and processed as described above. As(V) reduction was estimated by subtracting the As(V) concentration of oxidized samples from the As(V) concentration of unoxidized samples. The concentration of reduced As(V) or oxidized As(III) was calculated by generating a standard curve.

Genomic DNA was isolated using the method described by Wilson [15]. The 16S rRNA gene was amplified using the universal primers 8F (AGAGTTTGATCCTTGGCTCAG) and 1492R (GCYTACCTTGTTACGACTT). Amplicons were confirmed using agarose gel electrophoresis and were sequenced from Macrogen (Korea). The quality of the base calling was checked in FinchTV (Geospiza) and contigs were created through BLAST tool of NCBI. Closest matching sequences from the BLAST results were downloaded, aligned, and a neighbour-joining tree was made in MEGA 5 [16].

Plasmid isolation was carried out using the GF-1 plasmid DNA extraction kit (Vivantis, USA) following the manufacturer's instructions, and the yield was confirmed using agarose gel electrophoresis. Plasmids isolated from multi-metal-resistant bacteria were used to transform Escherichia coli DH5α. CaCl₂-based method was used to make competent cells of E. coli DH5α [17]. DH5α was cultured in a LB broth to 0.6 OD_600 nm and cells harvested by centrifugation at 4 °C. The pellets were suspended in 0.1 M MgCl₂ and incubated on ice for 10 min. The suspension was centrifuged at 4 °C and pellets were resuspended in 0.05 M CaCl₂ followed by incubation on ice for 20 min. After incubation, the pellets were obtained by centrifuging the suspension at 4 °C. The pellets were dissolved in 0.5 M CaCl₂ with glycerol solution. Competent cells (200 μL) were mixed with plasmid (10 μL), and the mixture was incubated for 30 min on ice. The reaction mixture was given a heat shock at 42 °C using a water bath for 45 s and incubated again on ice for 2.0 min. The reaction mixture was provided with 1.0 mL of LB broth and incubated at 37 °C for 30 min. After incubation, transformants were screened on LB-agar plates containing the metals mentioned previously.

Indole-3-acetic acid (IAA) production, phosphorous solubilisation and hydrogen cyanide (HCN) production by the bacterial isolates was determined. Bacterial cultures MX-2 and MX-6 were inoculated in an L-broth supplemented with l-tryptophan. After 72 h of incubation at 37 °C, 120 rpm, 100 μL of the supernatant were mixed with 200 μL of Salkowski's reagent [18] in a microtitre plate. After 1.0 h of incubation in the dark, the optical density was recorded at 535 nm using a Biotek Epoch plate reader. The concentration of IAA in the culture was measured using standard curve achieved using IAA standards.

The phosphate solubilisation ability of MX-2 and MX-6 was determined by culturing the bacteria in the phosphate agar media of the National Botanical Research Institute (NBRI) [19]. The plates were incubated for 5 to 7 days at 37 °C. The presence of clear zones around bacterial growth indicated phosphate solubilisation.

To evaluate the ability of isolates to produce hydrogen cyanide, bacteria were inoculated on nutrient agar media plates supplemented with glycine (4 g L⁻¹) [20]. After 4 days of incubation at 37 °C, the development of a yellow to orange colour was observed.

The experiment to study the phytostimulatory effect of the bacterial isolates on plant growth in the presence and absence of metals was analysed in pots [5,11]. Seeds of Vigna radiata (mung bean plant) were washed with autoclaved distilled water and were surface-sterilized with 0.1% HgCl₂ for 2 min followed by washing thrice with distilled water. Pots were filled with sieved soil and divided into four sets; plants, plants + bacteria, plant + metals and plants + metals + bacterial strains.

For inoculation, fresh cultures of MX-2 and MX-6 were taken and bacterial suspension was prepared in a normal saline solution (0.9% NaCl). Surface-sterilized seeds were soaked in bacterial suspension (0.5 OD_600 nm) for 20 to 25 min, and eight seeds were sown at equal distances in each pot. Thinning was done after germination, leaving five plants in each pot. A metal solution was used for watering where needed. Each pot contained only one of the metals (0.05 mM Hg, Cr, 0.1 mM As(III), As(V), Co, Cd, Cu, Se, Ni, Pb or Zn). The pots were placed at a temperature of 25 ± 1 °C during a 12-h photoperiod. After two weeks, root and shoot lengths, number of leaves and roots, wet and dry weights were recorded. There were at least ten pots of each condition, the mean was taken and standard deviation was applied.

All the experiments were performed in triplicates unless mentioned otherwise. The mean of the results was plotted and the standard deviation was shown as error bars. The plant–microbe interaction data were analysed and compared statistically by using ANOVA (analysis of variance) followed by two-tailed t-test post-hoc analysis (P = 0.05). Bonferroni correction was also applied to adjust the threshold of significance. The analysis was done with Microsoft Excel 2013.

The pH of the sample was 7.0, while the temperature was 40 °C. Bacteria were isolated using an enrichment culture technique and enumerated using a serial dilution technique on LB-agar plates. Six bacterial isolates (MX-1, MX-2, MX-3, MX-4, MX-5 and MX-6) resistant to all the metals tested [1.0 mM As(III), As(V), Cd, Cu, Co, Se, Pb, Ni, Zn and 0.5 mM Cr and Hg] were selected for further analysis.

The CFU count in the presence of As(III), As(V), Cr and Se was found to be 3.2 × 10⁵. The CFU count was found to be 3.7 × 10⁴ when As(III), As(V), Cr, Se and Zn were added to the next enrichment flask. CFU further decreased to 2.8 × 10⁴ when supplemented with As(III), As(V), Cr, Se, Zn and Cd. Further enrichment of bacteria in the presence of As(III), As(V), Cr, Se, Zn, Cd and Ni resulted in a CFU count of 2.5 × 10³. With the addition of Co, the CFU count dropped further to 2.2 × 10³. Further addition of Pb and Cu resulted in the decrease of CFU count to 2.8 × 10². In the presence of As(III), As(V), Cr, Se, Cu, Zn, Co, Cd, Ni, Pb and Hg, the final CFU count was 1.2 × 10².

Selected bacteria were subjected to morphological and biochemical analyses (see supplementary material). The temperature and pH optima for these isolates (for which the best growth was observed) were 37 °C and 7.0, respectively. The lesser growth was observed at pH 3 and at 45 °C.

Based on morphological and biochemical characteristics as well as 16S rRNA gene sequencing, the isolate MX-6 was identified as Pseudomonas sp. The neighbour-joining tree of 16S rRNA sequences of MX-6 and closely matched NCBI BLAST results was made in MEGA 5.0 to confirm the identity of the bacteria (Fig. 1). The sequence was submitted to NCBI GenBank with the accession number KU195676.

A high degree of resistance to all the selective metals was exhibited by MX-2 and Pseudomonas sp. MX-6. MIC values varied from 5.0 mM to 30 mM. Among the selected metals, Cr and Hg proved the most toxic ones. MIC of Hg and Cr for MX-2 and Pseudomonas sp. MX-6 were 5.0 mM and 2.5 mM, respectively. The least toxic metals for MX-2 were Co and Cu with an MIC value of 30 mM, while the least toxic metal for Pseudomonas sp. MX-6 was Zn as it resisted Zn up to 30 mM (Fig. 2).

Bacterial isolate MX-2 was found to resist chloramphenicol (3.0 μg·mL⁻¹), ampicillin (1.0 μg·mL⁻¹), kanamycin (3.0 μg·mL⁻¹), and erythromycin (1.5 μg·mL⁻¹) while sensitive to tetracycline (3.0 μg·mL⁻¹). On the other hand, Pseudomonas sp. MX-6 was found resistant to chloramphenicol (3.0 μg·mL⁻¹), but sensitive to all other antibiotics tested.

Bacterial isolates were screened for their ability to oxidize/reduce As. The results showed that both isolates had the ability to reduce As(V) into As(III) as well as to oxidize As(III) into As(V). The best reduction potential was exhibited by isolate Pseudomonas sp. MX-6. After 5 h of incubation in a normal saline solution, Pseudomonas sp. MX-6 reduced 506 μM As(V). Isolate MX-2 showed As(V) reduction up to 484 μM (Fig. 3).

The As(III) oxidation potential of the isolates was also determined. After 5 h of incubation, it was found that best oxidation ability was exhibited by the isolate Pseudomonas sp. MX-6 (160 μM). On the other hand, MX-2 oxidized 150 μM As(III) into As(V) (Fig. 3).

Plasmids were found to be present in both the isolates. The isolated plasmids were used to transform DH5α cells. Non-transformed DH5α was unable to resist all metals at a concentration of 5 mM and was sensitive to kanamycin and ampicillin. Thus, these two criteria were used for the screening of the transformants. In MX-2, resistance genes for As(III), As(V), Ni and Zn were found to be present on plasmids, as indicated by the transformation, while in Pseudomonas sp. MX-6 the genes for As(V), Cd, Se and Zn resistance were plasmid borne. We also analysed the resistance to antibiotics of the transformants, and it was observed that kanamycin- and ampicillin-sensitive DH5α was able to resist these two antibiotics after transformation. Pseudomonas sp. MX-6 plasmid conferred resistance against ampicillin as the corresponding DH5α transformants were able to resist ampicillin. Similarly, MX-2 plasmid was found to carry kanamycin resistance genes as kanamycin resistance was observed in the subsequent transformants. Non-transformed competent cells plated on LB-agar plates containing metals showed no growth, while colonies were observed on LB-agar plates not containing any metals.

Both isolates MX-2 and Pseudomonas sp. MX-6 were able to produce IAA and HCN. Maximum concentration of auxin (49.37 μg·mL⁻¹) were produced by Pseudomonas sp. MX 6, while MX-2 produced auxin at a concentration of 34.06 μg·mL⁻¹. MX-2 also showed phosphate solubilisation ability as clear zones were observed around the streaked lines.

Shoot and root lengths, wet and dry weights, and the number of leaves of the plants were decreased in the presence of metals as compared to control plants. The shoot length of the plants was highly affected in the presence of As(V), As(III), Cr, Hg, and Pb. The mean shoot length decreased from 15.9 cm (control plants) to 11.33, 9.0, 11, 10.34 and 10.3 cm in the presence of As(V), As(III), Cr, Hg and Pb, respectively. The presence of MX-2 and Pseudomonas sp. MX-6 increased the shoot length of the plants to 19 and 19.2 cm, respectively. The shoot length was also increased by bacteria in the presence of metals as well, and most prominent results were shown by MX-2 in the presence of Hg (25 cm) and Pseudomonas sp. MX-6 in the presence of Se (23 cm) (Fig. 4).

The root length of the plants was most affected in the presence of As(V), As(III), Co, Cr and Hg. The mean root length of the control plants was 1.57 cm; it increased to 3 and 2 cm in the presence of MX-2 and Pseudomonas sp. MX-6, respectively. The root length decreased to 0.85, 0.65, 0.75, 0.72 and 0.5 cm in the presence of As(V), As(III), Co, Cr, and Hg, respectively. However, the presence of bacteria increased root length, even in the presence of metals, and most prominent results were shown by MX-2 in the presence of Zn (3.3 cm) and Pseudomonas sp. MX-6 in the presence of Se (3.1 cm) (Fig. 5). Moreover, an increase in the number of leaves as well as an increase in wet and dry weights was also observed in the presence of the bacteria (Table 1).