Exploring the Plant Growth-Promotion Properties of Rhizospheric and Endophytic Bacteria Associated with Robinia pseudoacacia L. in Serpentine Soil

The soil sample was collected from the serpentine outcrop at Donja Paklenica, Bosnia and Herzegovina. From 10 cm depth, we sampled approximately 100 g of A-horizon or topsoil layer of undisturbed plot. Large and small soil particles were mixed and pooled as an individual sample. Samples were collected in triplicate using sterile bags and stored in an opaque portable refrigerator. For the isolation of bacteria, approximately 100 g of bulk soil was collected in the proximity of R. pseudoacacia roots. The sample was taken in triplicate, placed in sterile bags, and stored in a transportable refrigerator.

To remove moisture, soil particles were air-dried on a clean surface in a thin layer. After removing surface debris, heavy metals and macroelements were extracted following ISO 11466:1995 (1995). Briefly, 3 g of soil sample was mixed in 28 ml of aqua regia (21 ml HCl, 7 ml HNO₃) and heated until boiled. The mixture was preserved for 16 h at 20°C and then boiled using a reflux condenser. After cooling, the suspension was filtered through a filter paper and diluted with distilled water up to a final volume of 100 ml. After adjusting the volume, the concentration of the soil was 30 mg/ml, which was utilized as an input in flame atomic absorption spectroscopy (FAAS) (AA-7000; Shimadzu Corporation, Japan) to express elements in mg/kg.

A series of working solutions and standards were prepared using certified standard stocks (Merck KGaA, Germany). According to ISO 11047:1998 (1998), the concentrations of Ca, Mg, K, Cd, Cr, Cu, Co, Fe, Mn, Ni, Pb, and Zn were specified using the atomic absorption spectrophotometer AA-7000 (Shimadzu Corporation, Japan).

Index of geoacummuation (Igeo) and enrichment factor (EF) were used to assess the presence of anthropogenic contamination in soil. EF is a reliable tool to determine the differences between naturally occurring and anthropogenic sources of pollutants (Morillo et al. 2002; Kumar et al. 2012). The EF approach requires normalizing one metal concentration to the reference element to differentiate natural from anthropogenic contaminants. Due to their stability in the soil, Al, Mn, Fe, and Rb are most frequently used as reference elements (Sinex and Wright 1988; Mucha et al. 2005; Lizarraga Mendiola et al. 2008). In our study, we used Fe as a reference element. To calculate the enrichment factor, we used the following equation: EF=CMetalCREsoilCMetalCREcontrol $$EF = {{\left( {{{{C_{Metal}}} \mathord{\left/ {\vphantom {{{C_{Metal}}} {{C_{RE}}}}} \right. \kern-\nulldelimiterspace} {{C_{RE}}}}} \right)soil} \over {\left( {{{{C_{Metal}}} \mathord{\left/ {\vphantom {{{C_{Metal}}} {{C_{RE}}}}} \right. \kern-\nulldelimiterspace} {{C_{RE}}}}} \right)control}}$$

C_Metal denotes a heavy metal concentration, whereas C_RE represents the reference heavy metal element concentration in soil and control samples (Okedeyi et al. 2014). The results were interpreted according to Sutherland’s (Sutherland et al. 2000) scale for evaluating the types of contamination (Table I).

Igeo was employed for the evaluation of the soil contamination using the following equation (Müller 1979): Igeo=log2⋅Cn1.5Bn

C_n represents the heavy metal concentration in the soil, whereas B_n represents the geochemical background concentration of the same element. To limit the influence of any potential variations in the background, we used 1.5 factor (Lizarraga Mendiola et al. 2008). The results of determined Igeo values were interpreted using contamination categories suggested by Huu et al. (2010) (Table II). The method was initially developed to assess the contamination of river bottom sediments, but it can also be used to evaluate soil pollution (Loska et al. 2004).

Rhizospheric bacteria were isolated from soil samples using a sterile 0.85% NaCl solution. The mixture of 1 g of soil and 100 ml of NaCl solution was vigorously shaken and left to stand still for large particles to settle. The supernatant was used to prepare a series of dilutions in 1/10, 1/100, and 1/1000 ratios. To isolate endophytic bacteria, roots were cleaned of soil particles and shortly surface sterilized for 10 min with 5% sodium hypochlorite (NaClO) and washed three times with sterile distilled H₂O afterward. Roots were chopped into smaller pieces (2–3 cm long) with a sterile scalpel and transferred to a Petri dish with sterile 0.85% NaCl. With occasional stirring, roots were left in the solution for 2–3 h, and then the solution was smeared on a solid nutrient medium. On a yeast mannitol agar (YMA) medium, we added between 100–200 μl of previously prepared solutions and incubated at 22–25°C for 24–48 h. After incubation, only distinct, morphologically differentiated and conspicuous colonies were selected and transferred to a tryptone yeast (TY) solid medium to obtain bacterial monocultures. Isolated rhizospheric bacteria are marked as Rb, while endophytic bacteria are marked as Eb. Overall, 25 isolates were selected for the PGP properties analysis, 14 Rb (Rb1–Rb14) and 11 Eb (Eb1–Eb11).

All identified PGPB were grown in the presence of Cu, Ni, and Co at 22°C. CuSO₄ · 5H₂O (100 mg/l and 200 mg/l), NiSO₄ · 6H₂O (100 mg/l and 200 mg/l), and CoCl₂ · 6H₂O (100 mg/l) were each added to the TY solid medium at the specified concentrations. Firstly, bacteria were incubated in TY broth medium with agitation of 130 rpm at 25°C and transferred onto solid TY medium. After 14 days of incubation at 22°C, observable growth was marked with the special character +, whereas the absence of growth was marked with −.

CAS assay was used to analyze siderophores qualitatively (Kumar et al. 2017). Briefly, in prepared King’s B agar (20 g of peptone; 1.5 g K₂HPO₄; 1.5 g MgSO₄ · 7H₂O; 10 ml glycerol and 15 g of agar; pH 7.2 ± 0.2) (King et al. 1954) we added 100 ml of CAS reagent (60.5 mg dissolved in 50 ml H₂O; 10 ml FeCl₃ · H₂O and 72.9 mg HDTMA dissolved in 40 ml H₂O) (Schwyn and Neilan 1987). Selected isolates were incubated in TY broth solution at 25°C for 24 h and 130 rpm. The absorbance of culture was measured at 600 nm, and OD was set at 0.5. To assess siderophore production, we performed the well-diffusion test. In the perforated CAS agar wells (5–6 mm), 20 μl of each bacterial culture was added. The growth was monitored for 7 days, and the radius of orange zones around the bacterial colonies was expressed in mm afterwards.

Isolates were cultivated in the succinate broth medium without Fe (4 g sodium succinate; 1 g (NH₄)₂SO₄; 3 g K₂HPO₄ and 0.2 g MgSO₄ · 7H₂O; pH 7 ± 0.2). After incubation for 48 h at 28°C and 120 rpm, bacterial cultures were centrifuged for 10 min at 5,000 rpm. The supernatant was mixed with CAS in a 1:1 ratio and measured using a microplate reader (Multiskan^™ FC Microplate Photometer; Thermo Scientific^™, Thermo Fisher Scientific, Inc., USA) at 630 nm. The succinate medium without Fe was used as a control (Senthilkumar et al. 2021). The percentage of siderophore units was calculated according to the following equation: Ar−AsAr×100 where Ar is the absorbance of the control sample and As represents the absorbance of the bacterial sample.

We employed the phosphate solubilization test by Ambrosini and Passaglia (2017) for the phosphate solubilization properties. After incubation for 24 h at 25°C in TY broth medium, the absorbance of the bacterial cultures was measured at 600 nm, and OD was set to 0.5. In glucose yeast (GY)/tricalcium phosphate agar solid medium (10 g glucose; 2 g yeast extract; 15 g of agar; pH 7.2 ± 0.2), we added 100 ml of bromophenol blue solution (0.025 g of bromophenol blue in 100 ml of distilled H₂O). After medium autoclavation, sterile K₂HPO₄ (50 ml) and 10% CaCl₂ (100 ml) were added. A well-diffusion test was employed to determine the phosphate solubilization properties. In the perforated GY/tricalcium phosphate agar wells (5–6 mm), 20 μl of each bacterial culture was added. The growth was monitored for 7 days, and the radius of clear zones around the bacterial colonies was expressed in mm afterwards.

Isolates were inoculated into semi-liquid medium without N₂ (5 g of malic acid; 4 g KOH; 0.5 g K₂HPO₄; 0.05 g FeSO₄ · 7H₂O; 0.1 g MnSO₄ · 7H₂0; 0.02 g NaCl; 0.01 g CaCl₂; 0.0002 g Na₂MoO₄; 2 ml of ethanolic solution bromothymol blue as an indicator; 1.75 g agar; pH 6.6–7). After 7 days of incubation at 33 ± 2°C, isolates with nitrogen-fixation ability are characterized by white pellicles under the surface of the medium and the change of color from green to blue due to an increase in pH (Latt et al. 2018). Observable change of color was marked with the special character +, whereas the absence of color change was marked with -.

Bacterial isolates were grown in TY broth medium amended with 1 mg/ml of tryptophane. After an incubation period of 96 h at 32 ± 2°C and 130 rpm, bacterial cultures were centrifuged at 5,000 rpm for 10 min. In a 1:1 ratio, Salkowski reagent (150 ml H₂SO₄; 250 ml ddH₂O; 7.5 ml 0.5 M FeCl₃) was mixed with grown cultures and measured at 530 nm on a spectrophotometer UVmini-1240 (Shimadzu Corporation, Japan). The concentration of produced IAA was determined according to the standard curve, which was produced by the series of IAA dilutions (10–200 μg/ml). The obtained values are expressed in μg IAA/ml (Goswami et al. 2015).

Before the initial sterilization and experiment, all laboratory materials were washed thrice in 6M HCl and distilled water afterward. The activity of 1-aminocyclopropane-1-carboxylate (ACC) deaminase was determined by the ability of bacteria to utilize ACC as a source of N₂. Isolates were grown in King’s B broth medium for 48h at 28–30°C and 160–180 rpm. After the cultivation period, cultures were centrifuged at 5,000 rpm for 10 min. The supernatant was removed, and the bacterial pellet was resuspended in sterile 0.85% NaCl. This step was repeated three more times to remove leftovers of nutrition from King’s B medium. DF agar solid plates (2 g of glucose; 2 g of sodium gluconate; 4 g K₂HPO₄; 7.5 g Na₂HPO₄ · 2H₂O; 0,2 g MgSO₄ · 7H₂O; 2 g of citric acid; 100 μl DF solution 1; 100 μl DF solution 2 and 20 g of agar; pH 7.2 ± 2°C) with and without added ACC were used for the inoculation of 2 μl bacterial suspension. After an incubation period of 7 days, differences in growth intensity were observed what indicates ACC deaminase activity (Ambrosini and Passaglia 2017). Observable growth was marked with the unique character +, whereas the absence of growth was marked with −.

For the identification of bacterial isolates with PGP properties, we used the 16S rRNA gene sequencing. Considering the ubiquity among bacterial species and the presence of highly conserved and variable regions, this gene was successfully employed in detection, identification, and classification. The PCR reaction was prepared according to the universal protocol established by Woese (1987). The following universal primers were employed for 16S rDNA: 16S-FD1 (5’-AGAGTTTGATCCTGGCTCAG-3’) and 16S-RP2 (5’-ACGGCTACCTTGTTACGACTT-3’). The overall volume of the PCR reaction was 50 μl and it contained 0.4 μl forward and reverse primer, 1 × RedTaq, 0.5 mM MgCl₂ and 1 μl of the sample (bacterial colonies lysate in distilled H₂O at 95°C for 15 min). The PCR reactions for 35 cycles were used according to the following thermal protocol: initial denaturation at 94°C for 10 min, denaturation at 94°C for 1 min, annealing at 45°C for 1.5 min, elongation at 94°C for 1.5 min, and final elongation at 72°C for 7 min., Gel electrophoresis was performed on 1.5% agarose gel for the validation of PCR products. In the further analysis, we included only PCR products with a specific band size of ≈ 1,500 bp. GenElute^™ PCR Clean-Up (Sigma-Aldrich^®, Merck KGaA, Germany) was used to purify PCR products. After measuring the concentrations of PCR products on Qubit^® 2.0 Fluorometer (Invitrogen^™, Thermo Fisher Scientific, Inc., USA) using HS (high sensitivity) Qubit assay, we prepared Sanger sequencing reactions (10–40 μl) using BigDye^™ Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems^™, Thermo Fisher Scientific, Inc., USA). The overall volume of the PCR reaction was 20 μl containing 4 μl BigDye^™ Terminator v3.1 Ready Reaction Mix, 2 μl BigDye^™ Terminator v3.1 Sequencing buffer, 2 μl 3.2 μM forward primer, 3 μl H₂O, and 9 μl of product. The concentration of PCR products in reaction varied between 3–10 ng. Sequencing PCR reactions for 25 cycles were used according to the following BigDye^™ Terminator v3.1 thermal protocol: initial denaturation at 96°C for 1 min, denaturation at 94°C for 10 sec, annealing at 50°C for 5 sec, elongation at 60°C for 4 min, and final store at 4°C. After the sequencing PCR reaction, samples were purified using SigmaSpin^™ Sequencing Reaction Clean-Up (Sigma-Aldrich^®, Merck KGaA, Germany). After drying the elution, we added 10 μl of formamide for sequencing (3500 Genetic Analyzer; Applied Biosystems^™, Thermo Fisher Scientific, Inc., USA). Due to the significance of variable region 3 16S rDNA (V3), the amplicons were sequenced using a forward primer (5’-AGAGTTT-GATCCTGGCTCAG-3’). DNA sequences were analyzed in Data Collection Software 3 on a 3500 Genetic Analyzer (Applied Biosystems^™, Thermo Fisher Scientific, Inc., USA). BioEdit 7.7 software was employed to edit raw sequences. The obtained sequences were identified by comparing them with sequences deposited in NCBI GenBank (Sayers et al. 2020). For sequence similarity, we used BLAST (Basic Local Alignment Search Tool) (Altschul et al. 1990), paying attention to two parameters: query size and sequence similarity.

All obtained concentrations for quantitative analysis are performed in triplicate and given as an average value in the results. We employed Student’s t-test to compare Rb and Eb using PAST 4.10 (Hammer et al. 2001). Isolates that did not exhibit certain PGP characteristics were excluded from the particular statistical analysis. Even though we first tested the PGP traits of isolates and then performed 16S sequencing, for simplified interpretation and discussion of the results, we included the identified sequenced bacteria starting from the first interpretation of the presented PGP characteristic.

The concentrations of heavy metals in soil are presented in descending order: Fe > Ni > Mn > Cr > Co > Zn > Cu > Cd. The concentration of Ni (1,912.45 mg/kg) in soil exceeded the target value (TV) and intervention value (IV) by more than ninefold. As predicted from serpentine soil, a high level of Fe (16,983 mg/kg) concentration was observed as well. Although not above IV values, Co (87.7 mg/kg) and Cr (203.42 mg/kg) also exceeded TV. The EF ladder revealed extremely high enrichment for Ni (56.305), while considerable enrichment was detected for Co (8.607) and Cr (5.208). Moderate enrichment was measured for Cu (2.904). Other heavy metals (HMs) except Fe and Cd (not calculated) showed deficient to low enrichment (Mn, Pb and Zn). Igeo index uncovered extreme contamination for Ni (5.202) and moderate to strong contamination for Co (2.634). Other HMs, except for Fe and Cd (not calculated), showed deficiency in low enrichment (Mn, Pb, and Zn). Igeo index uncovered extreme contamination for Ni (5.202) and moderate to strong contamination for Co (2.634). Other HMs, according to Igeo, exhibited moderate contamination (Cr and Cu), or were unpolluted (Fe, Zn, Pb) to moderately contaminated (Mn). The concentrations of heavy metals in soil, EF, and Igeo index are presented in Table III. The content of macroelements indicated low levels of Ca and K in the soil (Table IV). In contrast, high Mg and Fe content were observed. The Ca/Mg ratio was extremely low indicating severe imbalance in which Mg suppresses Ca uptake.

All Rb isolates exhibited growth on Ni 100 mg/l, Ni 200 mg/l, Cu 100 mg/l, and Co 100 mg/l except Rb9-Bacillus sp., which only grew on TY medium supplemented with Ni 100 mg/l. Eb isolates were generally more susceptible to the mentioned HM concentrations, with only four isolates on Cu 100 mg/l and Co 100 mg/l and two isolates on Ni 100 mg/l. HM resistance results are presented in Table V.

All selected isolates except Eb5 and Eb8 showed nitrogen fixation properties, which were observed in color change from green to blue and pellicle formation (Table V). Although Rb, compared to Eb isolates, reached the final change after 4 days, this qualitative method did not provide us with a distinction between isolates within a group.

Nineteen isolates, 13 Rb and 6 Eb (Table V), produced ACC deaminase, which was detected throughout their ability to grow on DF + ACC agar plates.

Statistical analysis of diameter size showed significant differences (p < 0.0001) between Rb and Eb isolates. After 7 days, the average values of diameter size for Rb and Eb were 13.29 mm and 5.5 mm, respectively. The largest diameter was observed by Rb9-Bacillus sp. (20 mm), while only Eb9-Lysinibacillus sp. did not show any activity regarding siderophore production using the qualitative method (Fig. 1). A significant difference (p < 0.0004) between Rb and Eb was observed in the quantitative analysis of the siderophore production rate as well (Fig. 2). The average percentage of siderophore units for Rb and Eb was 72.48% and 38.64%, respectively. The highest production showed Rb1-Pseudomonas sp. (96.70%), while the lowest was detected for Eb11-Bacillus sp. (11.46%).

Fig 1.

Qualitative analysis of siderophore production of Rb and Eb isolates. Different characters indicate statistical differences between groups (p < 0.05).

Fig 2.

Quantitative analysis of siderophore production of Rb and Eb isolates. Different characters indicate statistical differences between groups (p < 0.05).

Out of 25 selected isolates, 17 isolates showed phosphate solubilization activity. Between Rb and Eb isolates, there was no significant difference (Fig. 3). The average diameter size for Rb and Eb was 5.16 mm and 5.75 mm, respectively. The largest diameter of 8 mm was observed at Rb14-Pseudomonas sp. and 7 mm at Eb4-Bacillus sp.

Fig 3.

Qualitative analysis of phosphate solubilization of Rb and Eb isolates.

No statistical difference (p < 0.73) was observed between Rb and Eb isolates (Fig. 4). The highest concentration of IAA production in Rb with 614.66 μg/ml was detected in Rb13-Pseudomonas sp. Similarly, in the Eb group, a concentration of 699.89 μg/ml was observed in Eb2-Staphylococcus sp.

Fig 4.

Quantitative analysis of indol-3-acid (IAA) production of Rb and Eb isolates.

After editing with BioEdit software, the sequences were approximately 550 bp in length. The 16S rRNA gene sequencing of isolates clearly showed differences in genus diversity between Rb and Eb. All isolates from the rhizosphere except Rb9 (Bacillus sp.) belong to the Pseudomonas genus. On the other hand, endophytic isolates are more diverse and related to five different genera: Bacillus, Pseudomonas, Staphylococcus, Lysinibacillus, and Brevibacterium/Peribacillus. The results of BLAST results after sequence editing are presented in Table VI.