Fluorescent Pseudomonas strain Psd, isolated from the roots of Vigna mungo, has been characterized as mentioned before (Upadhyay and Srivastava, 2008). The strain was maintained on Gluconate Minimal medium (GMM; Gilotra and Srivastava, 1997) with and without ZnSO₄.7H₂O. As per the experimental requirement, the media was supplemented with the appropriate concentration of autoclaved metal salt solution (ZnSO₄.7H₂O). The shake cultures were raised on Controlled Environment Shaker Incubator (Kühner, Switzerland) at 200 rpm at 30°C for the required period of time. Growth was determined turbidometrically at 600 nm, as per the protocol described earlier (Upadhyay and Srivastava, 2014). All chemicals used in the study were of analytical grade and purchased from Sigma Aldrich (St. Louis, MO, USA). All the other strains and plasmids used in this study along with their relevant characteristics are listed in Table S1.

EPS extraction was carried from 5 day-spent culture filtrate of strain Psd grown at different Zn²⁺ concentrations in GMM by the method described earlier (Bitton and Freihofer, 1978). Briefly, 2 volumes of 95% ethyl alcohol was added to the culture supernatant and was kept at 4°C overnight. The resulting precipitate was recovered by centrifugation at 8,000 × g for 10 min. The precipitate was dissolved in sterile distilled H₂O and dialyzed at 4°C against distilled H₂O for desalting. The purified EPS was concentrated under vacuum and stored at 4°C until use.

Cells grown at different Zn²⁺ concentrations were analyzed for the ultra-structural changes by Transmission Electron Microscopy (TEM). Pellets of sedimented cells, washed in saline, were fixed with 2.5% gluteraldehyde in 0.1 M phosphate buffer for 6 h at 4°C, and washed thrice with the same buffer. Dehydration was carried out in a graded ascending series of acetone followed by toluene. Specimens were infiltrated with 3:1 (v/v) mixture of toluene and Araldite [50% Epoxyresin + 50% Dodecenyl succinic anhydride (DDSA)] for 2 h, followed by pure Araldite, and were finally embedded in beam capsule with Araldite + Accelerator [(Trimethyl aminomethyl phenol (DMP-30)] and polymerized at 50°C for 24 h and 60°C for 48 h. Thin sections were cut with an Ultracut Microtome E (Ultracut E, Riechert Jung, Germany) and stained with uranyl acetate and lead citrate. The sections were analyzed in Transmission electron Microscope (FEI Electron Optics, USA).

Fourier-Transformed Infrared (FT-IR) spectrum of purified EPS was recorded to elucidate the chemical binding environment of Zn²⁺. Analyses were performed on an FT-IR spectrophotometer (Tensor 37, Bruker Optics, USA), equipped with total attenuated reflectance (ATR) objective. The purified EPS from strain Psd grown at varying Zn²⁺ concentrations was scanned over a wave number range of 4,000–1,000 cm⁻¹ with a resolution of 4 cm⁻¹. Samples were placed over a Zn-Se crystal and the peak emanating from the crystal was obtained between wave number 2400–2200 cm⁻¹. Background spectrum of water was collected and normalized prior to measurement of samples. For each sample, 16 scans were collected in order to evaluate the heterogeneity in the sample. All the spectra obtained were smoothened and their baseline was corrected.

Total RNA was isolated from bacterial cultures with the RNeasy Protect Bacteria Mini Kit for total RNA purification (Qiagen, Netherlands), as per manufacturer's instructions. Traces of DNA were removed from RNA preparations using DNase I, (Amplification Grade, Sigma Aldrich). DNase-free RNA (1 μg) was used in a one-step RT-PCR reaction (RevertAid First Strand cDNA Synthesis Kit, Thermo Scientific, USA) performed with the universal hexamer primers provided with the kit. The cDNA was used as a template for semi-quantitative RT-PCR. The primers used are listed in Table 1.

The cDNA prepared as mentioned above was subjected to quantitative PCR (qPCR) using SYBR GREEN reaction mix in a 7900HT Fast Real-Time PCR System (Applied Biosystems, USA). Thermal cycling conditions were as follows: 10 min at 95°C followed by 40 repeats of 30 s at 95°C, 30 s at 58°C. Following PCR amplification, the reactions were subjected to temperature ramping to create the dissociation curve, measured in terms of changes in fluorescence intensity as a function of temperature, by which non-specific products can be detected. The dissociation program was 95°C for 1 min, 60°C for 10 s, and 95°C for 30 s. The experiment included three biological replicates and each biological replicate was evaluated by three technical replicates. The constitutively expressed gene of Pseudomonas sp., for 16s rDNA, was taken as the calibrator in order to normalize gene expression levels. The MIQE guidelines (Bustin et al., 2009) for qPCR recommend the more generic quantification cycle (C_q). However, as C_t and C_q are used interchangeably and refer to the same value, C_t was used in this study. The C_t-values obtained in the experiment were used for quantification of relative change in gene expression, by the 2^−ΔΔCt method (Livak and Schmittgen, 2001).

Alginate content in EPS was estimated as per the protocol described by Richardson et al. (2004). Briefly, to 1 mL of purified EPS, 1 mL of 0.8 M NaOH was added and neutralized with 120 μL 2.25 M citric acid after 5 min of incubation. To this, 40 μL of DMMB (1,9-dimethyl methylene blue) was added. The solution was vortexed vigorously and incubated at room temperature for 45 min. Thereafter, UV-visible spectrum of the samples was recorded between 500 and 700 nm using microtiter plate reader (Ultramark, Bioplate Imaging system, Biorad, USA). The absorbance intensities at 520 and 650 nm, signifying alginate-bound and unbound DMMB were used to estimate the respective alginate concentrations.

The alg8 gene knockout strategy involved gene replacement by homologous recombination, wherein the native alg8 gene was replaced with an antibiotic resistance (kan^r) cassette. For homologous recombination, a construct consisting of kan^r gene flanked by the regions upstream and downstream to alg8 gene was generated, providing homology, as shown in Figure S2A. To assemble this construct, primers were designed to amplify ~500 bp upstream and downstream regions of alg8 from strain Psd, as XhoI/HindIII and PstI/BamHI fragments, respectively. The primer sequences have been mentioned in Table 1. These fragments were sequentially cloned in pBlueScript (pBKS+) vector, using the aforementioned restriction sites to generate a construct termed as pBKS_alg8_up/down. Subsequently, kan^r, was cloned between the upstream and downstream regions. The recombinant plasmid, pBKSKmΔAp (~3.3 kb) carrying kanamycin resistance gene, kan^r (~800 bp) cloned in SacI restriction site, served as the source of antibiotic resistance gene. The vector construct, pBKS_alg8_up/down was digested with EcoRV to generate a blunt-ended linear DNA fragment. Both of these fragments were then ligated and transformed into E. coli XL1-Blue to generate a construct carrying kan^r gene flanked by regions upstream and downstream to alg8 (pBKS_alg8_up/kan/down). Screening of transformants was done on LB Agar supplemented with 50 μg mL⁻¹ of kanamycin. This was followed by plasmid isolation and PCR amplification of three regions, viz. upstream, downstream and kan^r using their respective primers. Amplification profile confirmed successful cloning of these regions in pBKS+ (Figure S2C). One of the positive constructs was transferred to strain Psd by electroporation and putative transformants were selected on kanamycin-containing agar medium. Due to the narrow host-range of pBKS+, the construct pBKS_alg8_up/kan/down was less likely to be maintained as a plasmid in strain Psd. Hence, growth of clones on kanamycin-supplemented medium is explained by the integration of kan^r cassette in place of alg8 gene through homologous recombination (Figure S2B). In order to confirm stable insertion of kan^r cassette, positive transformants were transferred to non-selective conditions (without kanamycin) and grown for a few generations. Transformants with a stable integration were able to resist kanamycin when transferred back to the selection medium containing kanamycin.

Assessment of phosphate solubilization was carried out as per the protocol of Pikovskaya (1948). Extraction and detection of IAA was carried out using Ultra Performance Liquid Chromatography (UPLC), according to the modified protocol of Malhotra and Srivastava (2008). The secretion of siderophores was observed by spectral analysis at 400 nm (Schwyn and Neilands, 1987). Phenazine was extracted and quantified by the method described by Whistler and Pierson (2003). The antifungal activity of the strain was tested against two known plant pathogenic fungi, namely, Fusarium oxysporum and Fusarium graminearum by the following assays:

Dual culture assay

For dual culture assay, the Potato Dextrose Agar (PDA) plate was divided in two-halves. Inhibition assays were carried out by placing an agar-block from fully grown fungal plate on one half of the plate and streaking the biocontrol bacterium on the other half. The plates were incubated at 28°C for 5-days. Fungal inhibition was obtained as a zone of clearance on the plate in the area where bacteria were growing.

The plates with bacterial growth and fungal block alone were taken as reference.

Biofilm formation was measured using microtiter plate biofilm assay (Merritt et al., 2005). Briefly, 1% of the overnight grown cultures were subcultured in 10 mL GMM supplemented with increasing Zn²⁺ concentrations (0, 1, 2, 5 mM). From this, 100 μL culture was pipetted in a fresh 96-well microtiter plate. The plate was covered and incubated at 30°C for 48 h under static conditions. After incubation, the wells were washed thoroughly with distilled H₂O to remove planktonic bacteria. Following this, 125 μL of crystal violet stain (0.1%) was added to each well. The stain was removed after 10 min of incubation at room temperature and plates were allowed to air-dry. Thereafter, 200 μL of 95% ethyl alcohol was added to each stained well in order to solubilize the dye. The contents of each well were mixed well and 125 μL of the crystal violet/ethyl alcohol solution was transferred to a fresh microtiter plate. The absorbance at 560 nm was measured using a microtiter plate reader (Ultramark, Bioplate Imaging system, Biorad, USA). All values were normalized with cellular OD₆₀₀.

Wheat (Triticum aestivum var. HD2851, IARI, New Delhi) seeds were surface sterilized with 0.1% (w/v) HgCl₂ for 5 min and washed thoroughly with sterile distilled water. For bacterial inoculation, stationary phase bacterial cells, raised in GMM supplemented with different Zn²⁺ concentrations, were washed twice with saline to remove the residual medium and re-suspended in saline at a cell density of ~10⁸ cfu mL⁻¹. To this suspension, 10 surface-sterilized wheat seeds were added and flasks were incubated at 30°C, 70 rpm for 2 h. After this treatment, seeds were thoroughly rinsed with sterile water. Seeds were sown in pots and growth of seedlings was monitored for 6-days after treatment.

Confocal laser scanning microscopy of root samples from 6-day-old plantlets was performed according to the modified protocol by Bianciotto et al. (2001). Briefly, the root samples from 6-day-old plantlets were excised and washed with 100 mM cacodylate buffer (pH 7.40). Thereafter, the samples were stained with LIVE/DEAD® BacLight™ Bacterial Viability Kit (ThermoFisher Scientific, USA) for 25 min at room temperature. The kit contains a mixture of SYTO 9 and propidium iodide stains, with excitation/emission maxima of 480/500 and 490/635 nm, respectively. After incubation at room temperature for 15 min, samples were mounted on clean glass slides using the mounting oil provided in the kit and observed under Leica TCS SP5 confocal microscope (Leica Microsystems, Germany).

Scanning electron microscopy (SEM) of root samples from 6-day-old plantlets was performed as described by Koul et al. (2015). Briefly, root samples (1–1.5 cm) were fixed with 2.5% gluteraldehyde in 0.1 M phosphate buffer for 6 h at 4°C. Following this, post-fixation was done with osmium tetraoxide (1% v/v in 0.1 M phosphate buffer) for 30 min. Thereafter, the specimens were washed thrice with milliQ water and treated with 2% (w/v) uranyl acetate for 40 min. The samples were dehydrated through a graded ethanol series (30–100%) followed by acetone (100%). The treated specimens were mounted on aluminum stubs, coated with gold-palladium and examined under a scanning electron microscope (EVO MA10, Carl Zeiss).

Zn²⁺ content was estimated using atomic absorption spectrophotometer (Perkin Elemer model AAnalyst400) at 219.86 nm, as described earlier (Upadhyay and Srivastava, 2014). Flow diagram of the detailed protocol is shown in Figure S3.

All bacterial culture experiments were carried out in three independent sets, each consisting of three replicates. Values shown here represent mean ± standard deviation (SD). All bacterial inoculation experiments were carried out in three independent sets containing 10 seeds each and values shown are mean ± SD. Data were tested at a significance level of P < 0.05 using one-way ANOVA followed by Dunnett's t-test and expressed as mean ± SD.

Fluorescent Pseudomonas strain Psd can sustain an external Zn²⁺ concentration of 5 mM (Upadhyay and Srivastava, 2014). We have used transmission electron microscopic (TEM) analysis to monitor the changes induced by Zn²⁺ accumulation by strain Psd. In comparison to cells growing without extra Zn²⁺ supplementation in the medium (Figure 1A), cells exposed to increasing Zn²⁺ concentration displayed thickening of the outer membrane, which pointed toward extracellular biosorption of Zn²⁺ on to the cell surface (Figures 1B,C). In addition to the extracellular accumulation, some electron dense aggregates (EDA) were also found in the cytoplasm of the cells. Although most of the Zn²⁺ is located on the outer surface, internally cytoplasmic granules may provide additional sites for deposition of the heavy metal.

For the qualitative analysis of functional groups facilitating Zn²⁺ biosorption, EPS from the cells of strain Psd grown at increasing Zn²⁺ concentrations was subjected to FT-IR spectroscopy. The presence of functional groups such as C-O and C-O-C, signified by stretching between 1,200 and 1,000 cm⁻¹ and O-H, signified by elongation between 3,700 and 3,200 cm⁻¹, was indicative of the presence of carbohydrates, which is the major component of the EPS biopolymer. Sharp peaks in the range of 1,125 to 1,000 cm⁻¹ indicated toward the presence of uronic acids in the EPS. The spectrum also revealed the presence of mannose in the EPS, signified by a peak at ~2,900 cm⁻¹. The mannose peak, however, was weaker in comparison to the uronic acids. Besides, the peaks in the range of 1,210 to 1,140 cm⁻¹ revealed the presence of phosphoryl groups in the EPS. The presence of protein in the EPS was also implicated. Generally, proteins are detected by C = O (Amide I) stretching between 1,680 and 1,630 cm⁻¹, N-H bending vibration (Amide II) between 1,650 and 1,550 cm⁻¹ and N–H stretching (Amide A) vibration between 3,290 and 3,300. Peaks corresponding to Amide I and Amide II bonds were observed in untreated control and cells exposed to 2 mM Zn²⁺ (Figures 2A,B). In the cells exposed to 5 mM Zn²⁺, however, only peak corresponding to Amide I bond was observed (Figure 2C).

Based on the identification of uronic acids and mannose by FT-IR spectroscopy, the relative expression of two genes, alg8 and pslA, involved in biosynthesis of two principle exopolysaccharides, alginate and Psl, were studied in strain Psd exposed to varying Zn²⁺ concentrations, by real-time PCR using the gene encoding 16s rRNA as the reference gene or calibrator. The alg8 gene codes for alginate polymerase and pslA, encoding a sugar transferase, is responsible for the production of the mannose and galactose-rich Psl polysaccharide. When compared to the untreated control, a significant up-regulation was obtained in alg8 expression in the presence of Zn²⁺ (Figures 3A,B). Cells exposed to 5 mM Zn²⁺ displayed 8-fold increase in alg8 expression when compared to control or untreated cells (Figure 3B). On the other hand, no significant changes in the expression of pslA were obtained between control and Zn²⁺—grown cells (Figures 3A,C).

Based on the above results, it was concluded that alginate is the principle component of EPS produced by strain Psd as a direct correlation with its exposure to varying concentrations of Zn²⁺, and, thus may be involved in Zn²⁺ biosorption. To confirm this, spectrophotometric quantification of alginates, based on 1,9-dimethyl methylene blue (DMMB) complexation was used to estimate the amount of alginates in the EPS produced by the bacterial cells growing in different Zn²⁺ concentrations. As shown in Figure S4, strain Psd was able to produce 0.6 mg/mL of alginates after 5-days of incubation. In the presence of Zn²⁺, up to 3-fold increase was recorded in the amount of alginate produced. This result complied with the earlier hypothesis of the involvement of alginate in Zn²⁺ biosorption by strain Psd. Further evidence for the same was provided by generation of an alg8 knockout of strain Psd.

In order to ascertain the involvement of alg8 during Zn²⁺ biosorption by strain Psd, it was important to generate a mutant strain devoid of this function. The alg8 gene knockout strategy involved gene replacement by homologous recombination, wherein the native alg8 gene was replaced with an antibiotic resistance (kan^r) cassette (Figure S2).

The mutant strain Psd Δalg8::kan had a similar growth profile as the wild-type strain Psd in the absence of Zn²⁺. This showed that growth of the mutant strain was not affected by insertion of the kanamycin cassette. Supplementation of medium with 2 mM Zn²⁺, however, led to up to 20% drop in cell viability of the mutant. The survival further declined with the increase in Zn²⁺ concentration to 5 mM, wherein, 44% decrease in cell survival was obtained (Figure 4A). Besides, alg8 deletion resulted in 80% decrease in the levels of EPS synthesized at external Zn²⁺ concentrations of 2 mM, which increased to 97% at 5 mM Zn²⁺ (Figure 4B). Further, since the EPS production in the alg8-negative mutant was compromised, Zn²⁺ accumulation by the mutant was compared with that of the wild-type strain. For this purpose, the mutant strain (Psd Δalg8::kan) was grown in GMM supplemented with different Zn²⁺ concentrations (1, 2, 5 mM) for 24 h, along with a set containing no Zn²⁺. Wild-type strain Psd was used as control. It was observed that besides a reduced EPS secretion, the deletion of alg8 also led to a drastic decrease in Zn²⁺ accumulation potential of the strain (Figure 4C).

Morphological examination of the mutant cells through TEM revealed a comparatively thin outer membrane along with alterations in membrane architecture, supporting the above observations. There was an evident reduction in the cell wall thickening in the untreated control cells as compared to the wild-type strain (Figure 5A, also refer Figure 1). Further, exposure of cells to higher Zn²⁺ concentrations did not show traces of extracellular accumulation of Zn²⁺ (Figure 5B). However, cytoplasmic granules were observed in the presence of extracellular Zn²⁺, indicating that intracellular Zn²⁺ entry was not affected (Figure 5B). This was suggestive of a regulated intracellular entry of Zn²⁺ in the strain. When EPS is present, the cell is able to keep the high concentration of Zn²⁺ outside. Taken together, all these observations supported the fact that alg8 plays an important role in EPS production by strain Psd and is, also, the primary component involved in Zn²⁺ biosorption by the strain.

Formation of static biofilms was analyzed by crystal violet binding assay. As shown in Figure S5, increased Zn²⁺ concentration aided in biofilm formation by the strain. This was in accordance with the earlier observation that Zn²⁺ biosorption led to increased EPS biosynthesis by the strain. On the other hand, PsdΔalg8::kan produced significantly low amount of biofilm as compared to wild-type. This could be attributed to compromised EPS production by the mutant.

In the context of PGP bacteria, biofilm formation aids in effective root colonization, thus playing a crucial role in plant-microbe interactions. The association of strain Psd and its alg8 deletion variant with 6-day-old plantlets of T. aestivum was visualized by means of confocal microscopy, followed by SEM. As shown in Figures 6A, 7A, no bacteria were observed on the control roots. On the other hand, the wild-type strain Psd was able to associate with the roots (Figure 6B), and form thick biofilms associated with EPS matrix (Figure 7B). Additionally, Zn²⁺ accumulation also aided formation of bacterial aggregates and biofilm (Figures 6C,D, 7C). In case of the mutant strain, however, biofilm formation was drastically reduced, and mostly planktonic cells were observed on the roots (Figures 6E,F, 7D,E). In certain cases, however, presence of bacterial clusters was detected (Figure 7E).

In continuation with the observation that alg8 deletion severely affects the biofilm formation by the strain Psd, we were interested in observing its subsequent effect on the PGP potential. Increased accumulation of Zn²⁺ by strain Psd interestingly led to an increase in the phosphate-solubilization and siderophore production by the strain (Figures 8A,B). On the other hand, in strain Psd Δalg8::kan, both these activities were severely compromised. A different response was observed during the assessment of IAA production. In the absence of added Zn²⁺ in the medium, a 24 h-grown culture of strain Psd produced 164 μg IAA/OD₆₀₀. With increase in medium Zn²⁺ concentration, however, there was a significant drop in the amount of IAA produced. Strain Psd grown in the presence of 2 and 5 mM Zn²⁺ concentrations produced ~70% less IAA than the control. The mutant PsdΔalg8::kan produced ~67% less IAA than the wild-type strain Psd in the absence of Zn²⁺ (Figure 8C). Zn²⁺ supplementation, however, did not have a significant effect on IAA produced by the mutant. Increased Zn²⁺ accumulation also had a stimulatory effect on phenazine production by strain Psd (Figure 8D). This indicated that Zn²⁺ accumulation by strain Psd will enhance its biocontrol activity. On the other hand, in PsdΔalg8::kan, no phenazine production was detected.

The secondary metabolites produced by Pseudomonas sp. contribute to their biocontrol properties by inhibition of phytopathogens. Strain Psd showed an increase in the production of siderophores and phenazine in the presence of added Zn²⁺ in the medium, suggesting that Zn²⁺ modulates the production of these crucial secondary metabolites. On the contrary, deletion of alg8 resulted in decreased production of Fe³⁺-chelating siderophores by the strain. Additionally, phenazine production was also hampered in the alg8 deletion mutant. The major implication of these differences was reflected in the antifungal properties of strain Psd and its alg8-negative variant. Antifungal assays to test the ability of the strains to inhibit two phytopathogenic fungi, F. oxysporum, F. graminearum, comprised dual-culture tests followed by biomass inhibition experiments. Strain Psd was able to inhibit the growth of both the fungal strains growing on PDA (Figure 9A). In contrast, PsdΔalg8::kan showed inhibition of only F. oxyporum to some extent. No inhibition, however, was obtained in the case of F. graminearum. This clearly indicated that both strains differed in their antifungal spectrum, and can be explained by the fact that these Pseudomonads are known to produce a variety of antifungal metabolites. To corroborate the above observation, inhibition of fungal biomass was studied. The culture filtrate from wild-type strain Psd led to 70 and 90% reduction in biomass of F. oxysporum and F. graminearum, respectively. The potential of strain Psd as an antifungal agent was supported by the observation that the culture filtrate was able to cause 70 and 90% reduction in biomass of F. oxysporum and F. graminearum, respectively (Figures 9B,C). Inhibition by strain Psd grown in the presence of added Zn²⁺ in the medium increased up to 80% in case of F. oxysporum (Figure 9B). No significant difference, however, was obtained in case of F. graminearum (Figure 9C). In contrast, inhibition by PsdΔalg8::kan reduced to 60% in F. oxysporum (Figure 9B). The inhibition efficiency declined further with increase in Zn²⁺ concentration of the medium, with 40% inhibition at 5 mM Zn²⁺. The mutant strain was not able to inhibit F. graminearum at all (Figure 9C).

The above results highlighted the Zn²⁺ accumulation potential of strain Psd and, subsequently, its beneficial effect on PGP and biocontrol potential of the strain. Furthermore, it was observed that this effect translated into improved plant growth. The treatment of wheat seeds with strain Psd improved root growth and proliferation, as is evident from Figure S6A. Zn²⁺-laden biomass of strain Psd led to improved seedlings growth as higher number of root hairs was observed in these cases (Figure S7). On the other hand, PsdΔalg8::kan-treated wheat seeds did not show any differences with respect to the untreated control (Figure S6B), indicating that lack of alginate has affected the PGP potential of the strain.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.