Isolation, characterization, and evaluation of a high-siderophore-yielding bacterium from heavy metal–contaminated soil

Heavy metal–resistant siderophore-producing bacteria (SPB) with plant growth–promoting traits can assist in phytoremediation of heavy metal–contaminated soil. We isolated siderophore-producing bacteria from Pb and Zn mine soil in Shangyu, Zhejiang, China. The isolate with the highest siderophore production, strain SX9, was identified as Burkholderia sp. Burkholderia sp. SX9 produced catecholate-type siderophore, with the highest production at a pH range of 6.0 to 8.0, a temperature range of 20 to 30 °C and NaCl concentration below 2%. Siderophore production was highest without Fe3+ and became gradually lower with increasing Fe3+ concentration. Minimal inhibitory concentrations (MIC) of Pb2+, Zn2+, Cu2+, and Cd2+ were 4000, 22000, 5000, and 2000 μmol L−1, respectively. The strain had a strong metal solubilization ability: the contents of Cu2+, Zn2+, and Cd2+ in the supernatant were 47.4%, 133.0%, and 35.4% higher, respectively, in strain SX9–inoculated cultures than in the not inoculated controls. The siderophore produced by strain SX9 could combine with Fe3+, Zn2+, and Cd2+ with good effectiveness. The plant growth–promoting traits of the strain included indole acetic acid (IAA) production, 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity, and phosphate solubilization capability. Compared to the uninoculated growth medium and SX9 culture supernatant, the germination rate of Lolium perenne seeds was higher when inoculated with strain SX9 culture. In the experiment of seed germination, adding bacterial culture or supernatant could alleviate the toxicity of heavy metals to L. perenne seed germination. Under Cu2+ and Zn2+ stress, strain SX9 promoted the germination rate. Taken together, Burkholderia sp. SX9 had properties beneficial in the microbial enhancement of phytoremediation of soil contaminated with heavy metals.


Introduction
The contamination of soil with heavy metals poses major environmental and human health challenges, calling for the remediation of the contaminated soils. Heavy metals are removed from the soil using physical, chemical, and biological technologies. Each remediation technique has its own advantages and disadvantages in terms of pollutant removal and cost efficiency (Khalid et al. 2016). Phytoremediation technology has attracted attention due to its economic, environmentally friendly, and non-secondary pollution characteristics (Ashraf et al. 2019). The discovery of hyperaccumulator plants has enhanced the prospects for phytoremediation. Unfortunately, most of the hyperaccumulators grow and produce biomass slowly, requiring long remediation periods that diminish their applicability in phytoextraction (Ashraf et al. 2019;Kuppusamy et al. 2016).
Therefore, as an alternative strategy, the phytoremediation capacity of non-hyperaccumulating plants that produce biomass faster could be improved. Tolerance to stressful remediation conditions might be achieved by exploiting plant growth-promoting rhizobacteria (PGPR) (Lucian et al. 2014). A large number of bacterial genera include PGPR, e.g., Pseudomonas, Burkholderia, Bacillus, Azotobacter, Azospirillum, and Klebsiella (Pliego et al. 2011;Ahmad et al. 2008). PGPR promote plant growth through direct and indirect mechanisms. Direct promotion includes the synthesis of substances that have a direct effect on plant growth, e.g., indoleacetic acid (IAA) and gibberellin, and changes in the availability of nutrients, e.g., nitrogen fixation and phosphate solubilization. Indirect promotion includes the synthesis of substances that inhibit or reduce the effects of adverse factors on plant growth and yield, e.g., siderophores and 1aminocyclopropane-1-carboxylate (ACC) deaminase. The siderophore-producing bacteria (SPB) can provide iron to plants and alleviate the toxicity of heavy metals (Sinha and Mukherjee 2008).
Under iron-limiting conditions, microorganisms secrete siderophores that have a high affinity for Fe 3+ (Saha et al. 2016). Most aerobic and facultative anaerobic microorganisms are capable of synthesizing siderophores, and hundreds of different kinds of siderophores have been characterized (Saha et al. 2016). According to the characteristic functional groups, siderophores are classified into four types: hydroxamate, catecholate, hydroxycarboxylate, and mixed types, out of which hydroxamate type is the most common (Khan et al. 2017). In addition to binding Fe 3+ , siderophores can form soluble metal-siderophore chelates with Al 3+ , Cu 2+ , Cd 2+ , Pb 2+ , Zn 2+ , and other metals (Dimkpa et al. 2009;Rajkumar et al. 2010). Siderophore production has positive effects on the plant, and the heavy metal chelating ability of siderophores already has applications in practical remediation; inoculation with SPB and adding the supernatant of a SPB culture promoted the growth of plants under heavy metal stress and increased heavy metal uptake by plants (Jiang et al. 2008;Wang et al. 2011).
Lolium perenne is a kind of perennial herbaceous plant with high tolerance to Cd 2+ , Cu 2+ , and Pb 2+ . L. perenne has large aboveground biomass, fast growth rate, strong environmental adaptability, and the potential of remediation of heavy metal-contaminated soil. Zhao et al. once inoculated B. megacarum LY02 in Cd 2+ -and Cu 2+ -contaminated soils, and combined contaminated soil, which significantly increased the aboveground biomass of L. perenne by 65.0-108.3%, and promoted the absorption of the test metals (Zhao et al. 2017).
The purpose of this study was to isolate high-efficient SPB and exploring its characteristics to provide a suitable strain for remediating heavy metal-contaminated soils. We isolated bacteria from Pb-and Zn-contaminated soil and assessed their siderophore production. The isolate with highest siderophore production capacity, strain SX9, was tested further: we determined its capacity to produce siderophore under different conditions and tolerance to heavy metals. Furthermore, we investigated the complexation capacity of siderophore produced by strain SX9 for metal ions and the plant growth-promoting characteristics of strain SX9, including its ability to promote the growth of L. perenne under heavy metal stress.

Source of soil sample and determination of soil basic properties
Soil sample was collected from a Pb and Zn mine in Shangyu, Zhejiang, China (29°59′ N, 120°46′ E) that has East-Asian monsoon climate with an annual average precipitation of 1400 mm and temperature of 16.4°C. The soil sample was packed in a sterile plastic bag and stored at 4°C before use. The physicochemical properties of soil were determined as described previously (Bao 2000). Heavy metals were extracted with aqua regia-HClO 4 and analyzed with an inductively coupled plasma optical emission spectrometer (ICP-OES, Optima 8000, Perkin Elmer Co, USA).

Growth media and buffers
Liquid media and buffers were prepared by dissolving the below-mentioned components into deionized water to make up a total volume of 1000 mL. For solid media, 1.8% (w/v) agar was added to the liquid media. The media were sterilized at 121°C for 30 min.

Isolation of bacteria
Ten grams of soil sample was suspended into 80 mL of sterile water, and the suspension was shaken at 150 rpm for 30 min at 28°C. Tenfold serial dilutions of the suspension were inoculated onto Luria-Bertani (LB) agar plates as described previously (Liu et al. 2019). Plates were incubated at 28°C for 48 h. Well-separated colonies were picked and inoculated onto LB agar plates until a pure culture was obtained.

Qualitative analysis of SPB
The isolates were inoculated onto iron-free modified King B (MKB) medium agar plates and cultured at 28°C for 48 h. Chrome azurol sulphonate (CAS) semisolid medium was cooled to approximately 60°C and poured onto the MKB agar. After 2-4 h, a color change from blue to orange around a colony indicated siderophore production. Forty-four isolates with distinct orange circles around the colonies were chosen for further analyses. The analyses here and further on were done in triplicate.

Quantitative analysis of SPB
The isolates were cultured in an iron-free MKB medium at 28°C for 48 h at 150 rpm. After centrifugation at 10,000 rpm for 10 min, the supernatant was collected and 1 mL supernatant was mixed with 1 mL of CAS reagent (Arora and Verma 2017;Schywn and Nielands 1987). A negative control was made by mixing 1 mL sterile water and 1 mL of CAS reagent. After incubation for 30 min at room temperature, the absorbance value of the mixture (As) was measured at the wavelength of 630 nm. The absorbance of uninoculated iron-free MKB medium mixed with CAS reagent was determined as the reference ratio (Ar). The experiment was carried out in triplicate. The percent siderophore unit (SU) was calculated using the following formula: The larger the SU value, the stronger the ability of an isolate to produce siderophore. The isolate with highest % SU value, referred to as strain SX9 from hereon was chosen for further analyses.

Characterization of strain SX9
The morphological and chemotaxonomic characterization of strain SX9 was done as described in Bergey's Manual of Determinative Bacteriology (Holt et al. 1994). Genomic DNA of strain SX9 was extracted using Rapid Bacterial Genomic DNA Isolation Kit (Sangon Biotech Co., Ltd., Shanghai, China). The 16S rRNA gene of the strain was amplified with polymerase chain reaction (PCR) using the universal primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-ACGGCTACCTTGTTACGACTT-3′) (Sayyed et al. 2019). The PCR reaction conditions were as follows: initial denaturation at 94°C for 5 min, 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s and extension at 72°C for 90 s, and a final extension at 72°C for 10 min. The PCR products were checked in 1.0% agarose gel and purified using DiaSpin DNA Gel Extraction Kit (Sangon Biotech Co., Ltd, Shanghai, China). The amplified product was sequenced at Sangon Biotech Co., Ltd (Shanghai, China). The 16S rRNA gene sequence of strain SX9 was compared to closely related sequences in the National Centre of Biotechnology Information (NCBI) GenBank database using BLAST online (https://blast.ncbi.nlm.nih.gov/Blast. cgi). A phylogenetic tree was constructed using neighborjoining method in MEGA 6.0 with a bootstrap value of 1,000 (Koichiro et al. 2013). The sequence was submitted to GenBank with the accession number MN658668.

Siderophore type determination
Arnow's test (Arnow 1937), Csaky's test (Csáky et al. 1948), and Shenker's test (Shenker et al. 1992) were used to determ i n e c a t e c h o l a t e -t y p e , h y d r o x a m a t e -t y p e , a n d hydroxycarboxylate-type siderophores, respectively.
Arnow's test One milliliter of culture supernatant was mixed with 0.1 mL of 5 mol L −1 HCl. Subsequently, 1 mL nitritemolybdate reagent (10.00 g NaNO 2 and 10.00 g Na 2 MoO 4 dissolved in 100 mL of deionized water) was added. In the presence of catechol, the nitrite in the solution is decomposed to form a yellow NO ligand. Next, 0.5 mL of 2 mol L −1 NaOH was added. In the presence of catecholate siderophore, the solution becomes red and remains unchanged for at least 1 h.
Csaky's test One milliliter of culture supernatant was mixed with 1.0 mL of 1 N H 2 SO 4 , boiled for 6 h, 3 mL of 35% sodium acetate solution was added and mixed well, and 1 mL of the solution was diluted with deionized water in a 1:5 ratio. Next, 0.5 mL of sulfanilic acid solution (1.00 g sulfanilic acid dissolved in 100 mL of 30% acetic acid) and 0.2 mL of iodine solution (1.30 g iodine dissolved in 100 mL acetic acid) were added and the solution was incubated at room temperature for 5 min. After this, 0.2 mL of sodium thiosulphate (2.00 g sodium thiosulphate dissolved in 100 mL deionized water) and 0.1 mL of α-naphthylamine solution (3.00 g α-naphthylamine in 1,000 mL of 30% acetic acid) were added and the solution was incubated at room temperature for 20-30 min. In the presence of hydroxamate siderophore, the solution becomes pink.
Shenker's test One milliliter of supernatant was mixed with 1 mL 250 μmol L −1 CuSO 4 and 2 mL acetic acid buffer (pH 4) and incubated at room temperature for 30 min. The presence of hydroxycarboxylate siderophore-copper compound was determined using full wavelength scanning. A maximum absorption peak at 190-280 nm indicated that the solution contains hydroxycarboxylate siderophore.

Siderophore production by strain SX9
To study siderophore production over time, strain SX9 was grown for 24 h, and 2.5 mL of the culture, adjusted to A 600 of 1.0, was inoculated into 250 mL iron-free MKB and incubated at 28°C for 48 h at 150 rpm. Cell density as OD 600 and siderophore production as % SU were recorded at 4-h intervals.
The effects of pH and NaCl and Fe 3+ concentrations on the growth and siderophore production of strain SX9 were studied using MKB medium, inoculation volume of 1% (v/v), and incubation at 28°C and 150 rpm. The effect of temperature was studied similarly, except that the cultures were incubated at temperatures from 15 to 35°C. The effect of pH was examined in medium with the initial pH adjusted to 5.0, 6.0, 6.8, 7.4, 8.0, and 9.0. Salt tolerance was assessed at 0, 0.5, 1, 2, 3, and 5% (w/v) NaCl concentrations. The effect of Fe 3+ concentration was assessed at 0, 0.5, 1, 2, 3, 5, 10, and 20 μmol L −1 Fe 3+ concentrations.

Minimal inhibitory concentrations (MICs)
To determine the MICs of twelve metal salts, strain SX9 was grown in LB medium with predetermined concentrations of metal salts (Table 1) at 28°C for 48 h at 150 rpm (Yu et al. 2013).

Solubilization of heavy metals
Liquid MKB was divided into 50 mL with 50 mg of Cu 2+ as CuO, 50 mg of Zn 2+ as ZnO, or 50 mg of Cd 2+ as CdCO 3 , which was inoculated with 1 mL of logarithmic-phase strain SX9 inoculum and incubated at 28°C and 150 rpm. After 72 h, 5 mL of the culture was centrifuged at 8,000 rpm for 10 min and filtered through a 0.22-μm needle cartridge membrane filter (Jin Teng, China). The Cu 2+ , Zn 2+ , and Cd 2+ concentrations in the filtrate were determined using inductively coupled plasma optical emission spectrometry (ICP-OES).

Complexation capacity assays
To assess the ability of siderophore to chelate metal ions under alkaline conditions, a standard curve was made using demertamine mesylate, and the content of siderophores was determined using the CAS method (Ferreira et al. 2019). FeCl 3 ·6H 2 O, CuSO 4 ·5H 2 O, ZnSO 4 ·7H 2 O, and Cd(NO 3 ) 2 · 4H 2 O were added to a fixed volume of the culture filtrate (Ferreira et al. 2019). The pH of the solution was set to 9.0 ± 0.1, the solution was incubated at room temperature for 30 min, the pH of solution was adjusted to 9.0 again, and the solution was incubated for 3 h. The solution was centrifuged at 3,000 g for 10 min and filtered through a 0.45-μm pore size membrane. The concentrations of metal ions in the solution were determined using ICP-OES. The iron complexation capacity of the siderophore in solution was calculated by plotting the ratio [ Plant growth-promoting characteristics IAA synthesis assay Strain SX9 was inoculated into LB liquid medium containing 200 mg L −1 L-tryptophan and incubated at 150 rpm for 48 h at 30°C. Five milliliters of the culture was centrifuged at 8,000 rpm for 10 min. Two milliliters of the supernatant was mixed with 4 mL Salkowski's reagent, incubated for 30 min at room temperature in the dark, and the absorbance of the solution was measured at 530 nm (Bahadur et al. 2020). A standard curve was made by measuring the absorbance of IAA standard solution diluted to concentrations of 0, 3, 5, 8, 10, and 15 mg L −1 .
ACC deaminase activity assay Strain SX9 was inoculated into 20 mL Dworkin and Foster (DF) medium. After incubation at 150 rpm for 24 h at 30°C, the cells were collected by centrifugation at 8,000 rpm for 10 min at 4°C. The cells were washed twice with DF liquid medium without (NH 4 ) 2 SO 4 , suspended in 10 mL of ADF medium, and incubated at 150 rpm for 24 h at 30°C . Protein content was determined using colorimetry with bovine serum albumin as the standard (Penrose and Glick 2010). The ACC deaminase activity was calculated using the standard curves of α-ketobutyric acid (α-KA) and protein as the amount  (Kong and Hong 2020). After centrifugation at 6,000 rpm for 10 min, phosphorus content of 1 mL supernatant was determined using the Mo-Sb colorimetric method. A not inoculated culture was used as a negative control.

Seed germination tests
To assess the effect of strain SX9 inoculation on L. perenne seed germination in the presence of heavy metal contaminants, L. perenne seeds were soaked in 75% alcohol for 5 min; washed 4 times with sterile deionized water, wherein 50 seeds per plate were spread on filter paper in glass petri dishes; and inoculated with 5 mL of bacterial culture, bacterial culture supernatant, or uninoculated growth medium. Metal salt solutions with increasing concentrations of Cu 2+ (0, 100, 200, and 400 mg L −1 ), Zn 2+ (0, 200, 500, and 1,000 mg L −1 ), and Cd 2+ (0, 5, 10, and 20 mg L −1 ) were added to the plates. The plates were kept at a constant temperature of 28°C. Seeds were considered germinated when the root protruded from the seed coat by at least 2 mm. The final germination rate was calculated after 7 days. Five seedlings per plate were randomly chosen to determine the shoot and root lengths (Aka and Babalola 2016).

Statistical analysis
Statistical analyses and visualization were done using Microsoft Excel (version 2016) and Origin Pro 2018C (OriginLab Corporation, Northampton, USA), respectively. Statistical significance of the differences was tested using one-way analysis of variance (ANOVA) and Fisher's least significant difference (LSD) test. Differences were considered statistically significant at P < 0.05.

Results and discussion
Characteristics of soil sample

Isolation of SPB and characterization of strain SX9
A total of 44 bacterial strains were isolated from the soil sample. Six of them did not detect siderophore production. Numbers for SU values between 10 and 50% and 50 and 80% are twenty-eight and nine, respectively. The isolate, strain SX9, with the highest SU value (93.7%) was selected for further investigation. The colonies of strain SX9 were round, yellow, and nontransparent with a smooth surface on LB plates (Fig. 1a). Transmission electron microscopy (TEM) showed that SX9 cells were rod-shaped with multiple polar flagella (Fig. 1b). Strain SX9 was Gram-negative, methylene red, fluorochrome, oxidase-and catalase-positive, and utilized fructose, mannose, xylose, maltose, lactose, and galactose. In the 16S rRNA gene sequence analysis, the strain was assigned to Burkholderia sp. (Fig. 2). Based on a positive Arnow's test and negative Casky's and Shenker's tests, strain SX9 produced a catecholate-type siderophore. In an iron-free MKB medium, the growth of strain SX9 reached a stationary phase after 20 h of incubation. Similar to previous studies (Patel et al. 2018;Sayyed et al. 2019), siderophore production reached a maximum in the stationary phase (Fig. 3).
Effects of pH, temperature, and NaCl and Fe 3+ concentrations on siderophore production by strain SX9 The solubility of substances, the permeability of cell membrane, and the rates of enzymatic reactions are affected by pH, temperature, and NaCl. Both the OD 600 and SU values of strain SX9 were highest at pH 7.4 (P < 0.05) (Fig. 4a). Similarly, the production of siderophores by Achromobacter sp. RZS2 was highest at pH 7.5 (Sayyed et al. 2019). The OD 600 and SU values were lowest at pH 5.0 and second lowest at pH 9.0 (P < 0.05), yet relatively stable between pH 6.0 and 8.0. The optimum temperature range for cell growth and siderophore production was 25 to 30°C (P < 0.05) (Fig. 4b). At 20°C and at 35°C, growth was slightly lower than at the optimum range (P < 0.05). Siderophore production was slightly lower at 20°C than at the optimum range (P < 0.05) and second lowest at 35°C (P < 0.05). Both the OD 600 and SU values of strain SX9 were lowest at 15°C (P < 0.05). The growth and siderophore production of strain SX9 was gradually inhibited at NaCl concentrations higher than 2% and 1%, and no growth was observed at 5% NaCl (P < 0.05) (Fig. 5a). Thus, when applying strain SX9 to bioremediation in field, maintaining the pH from 6.0 to 8.0 would be advisable. Based on the effects of environmental factors on siderophore production, the bioremediation efficiency of strain SX9 is at its highest at temperatures from 20 to 30°C and NaCl concentration below 2%.
Iron is an essential element of living organisms and plays an important role in metabolism, and free Fe 3+ concentration in the environment is the primary factor affecting the synthesis of siderophore. The OD 600 value of strain SX9 was highest at 3 μmol L −1 Fe 3+ (P < 0.05), yet the tested Fe 3+ concentrations had only a minor effect on the growth of strain SX9 (Fig. 5b). The siderophore production by strain SX9 was highest without Fe 3+ , and the SU values became gradually lower with increasing Fe 3+ concentration (P < 0.05) (Fig.  5b). The siderophore production by P. fluorescens and P. putida was at the maximum at 1 μmol L −1 Fe 3+ (Sayyed et al. 2005). Increasing Fe 3+ concentrations repressed siderophore synthesis by fluorescent pseudomonads, and at Fe 3+ concentrations from 20 to 27 μmol L −1 siderophore production was stopped (Dave and Dube 2000;Sayyed et al. 2005). Similarly, the SU value of strain SX9 was only 2.7% at 20 μmol L −1 Fe 3+ .

Heavy metal tolerance and solubilization
Knowledge on the ability of bacterial strains to grow in the presence of contaminants is essential to estimate the applicability of the strains in bioremediating the contaminated soils. Strain SX9 was resistant to multiple metals, yet the strain was sensitive to Hg 2+ ( Table 2). The MIC values of Pb 2+ , Cu 2+ , and Cd 2+ were 4000, 5000, and 2000 μM, respectively. The high tolerance to Zn 2+ , with a MIC of 22000 μM, may have been due to adaptation to the high level of Zn pollution in the sampled soil. Through comparison with other reported strains, strain SX9 had a superior tolerance to Zn 2+ and Cu 2+ , and a moderate tolerance to Pb 2+ and Cd 2+ (Guo et al. 2011;Jiang et al. 2008;Jiang et al. 2017).
In bioremediating heavy metal-contaminated soil, the availability of the metals directly affects the efficiency of remediation. The SPB may increase the solubility of heavy metals (Guo et al. 2011;Jiang et al. 2008 Fig. 4 Effects of pH (a) and temperature (b) on the growth and siderophore production of strain SX9. SU, percent siderophore unit. Different lowercase and capital letters above points indicate statistically significant differences between growth and SU values, respectively Fig. 5 Effects of NaCl (a) and Fe 3+ (b) concentration on the growth and siderophore production of strain SX9. SU, percent siderophore unit. Different lowercase and capital letters above points indicate statistically significant differences between growth and SU values, respectively   The values are mean ± standard deviation (n = 3). Different capital and lowercase letters indicate statistically significant differences between Cu 2+ concentrations and inoculants, respectively (P < 0.05) that maximum values of water-soluble Cd 2+ , Zn 2+ , and Pb 2+ in the solution inoculated with strain D54 were 28.3, 36.9, and 8.2 mg L −1 , and it was concluded that inoculation with strain D54 could help Sedum alfredii remove more metals from heavy metal-contaminated soils (Guo et al. 2011). In agreement, strain SX9 had a strong metal solubilization ability. In the SX9-inoculated cultures containing CuO, ZnO, and CdCO 3 , the concentrations of Cu 2+ , Zn 2+ , and Cd 2+ in the supernatant were 25.94 ± 0.54 mg L −1 , 2.21 ± 0.31 mg L −1 , and 17.35 ± 0.37 mg L −1 , respectively. The contents of Cu 2+ , Zn 2+ , and Cd 2+ in the supernatant were 47.4%, 133.0%, and 35.4% higher, respectively, in the strain inoculated cultures than in the not inoculated controls (P < 0.05).

Metal complexation capacity
Even though siderophores help the SPB to acquire iron, the complexation of siderophores with heavy metals decreases their free concentration and thus their toxicity (Schalk et al. 2011). In general, 1 μmol L −1 of siderophore was expected to complex 1 μmol L −1 of metal ion; the catechol-type siderophore (bacillibactin) produced by Bacillus subtilis can bind Fe 3+ at 1∶1 ratio (Dertz et al. 2006). However, the complexation capacity of strain SX9 siderophore was relatively low for Fe 3+ ; the increase in complexed Fe 3+ was only a fraction of the added (Fig. 6a). The complexation capacity of different siderophore types varies, and variation exists even within the same type (Ferreira et al. 2019). Possibly the complex formed by strain SX9 catecholate-type siderophore was not stable under alkaline conditions, or the added Fe 3+ had precipitated.
According to the results of complexation of Cu 2+ (data not shown), it is impossible to judge the Cu 2+ complexation capacity of strain SX9. For both Zn 2+ and Cd 2+ , the ratio of ions in complex versus ions added increased linearly until [ion]added /[S] ratio of approximately 2, after which the adding of ions did not result in increases in complexed ions (Fig. 6b, c), indicating that the ions were complexed with fairly good effectiveness and were stable up to that amount of ions added.

Plant growth-promoting characteristics of strain SX9
The plant growth-promoting characteristics of bacteria include the production of plant growth hormone IAA, the control of ethylene levels through ACC deaminase activity, and the secretion of organic acids to dissolve nutrients, especially phosphorus. PGPR protect plants from biological and abiotic stresses (Tiwari et al. 2016). In this study, strain SX9 produced 9.50 ± 0.40 mg L −1 IAA in L-tryptophan containing LB medium, and the ACC deaminase activity of strain SX9 was 19.13 ± 1.50 μmol α-KA (mg h) −1 . Thus, the IAA synthesis and ACC deaminase capabilities of strain SX9 were superior to those of Burkholderia sp. J62 and D54 that promoted the growth and heavy metal accumulation capacity of Zea mays L., Lycopersicon esculentum Mill., and Sedum alfredii Hance (Guo et al. 2011;Jiang et al. 2008). The phosphate solubilization capacity of strain SX9 was 106.60 ± 6.65 mg L −1 . The phosphate solubilization activity of the 8 strains isolated by Liu et al. ranged from 12.80 to 140.48 mg L −1 (Liu et al. 2019); this indicated that the phosphate solubilization capacity of strain SX9 reached a medium level. The plant growth-promoting characteristics of strain SX9 suggest that the strain might be applicable in the plant-assisted bioremediation of contaminated sites.
L. perenne can be applied in bioremediation due to its tolerance to heavy metals, fast growth rate, and heavy metal accumulation capacity (Zhao et al. 2017). SPB may contribute to the tolerance, for example, the siderophores of B. tequilensis CD36 The values are mean ± standard deviation (n = 3). Different capital and lowercase letters indicate statistically significant differences between Zn 2+ concentrations and inoculants, respectively (P < 0.05) combined with Ni 2+ and Pb 2+ and alleviated the metal stress of Miscanthus seeds (Li 2018). Therefore, we determined the effect of strain SX9 on the germination and growth of L. perenne under Cu 2+ , Zn 2+ , and Cd 2+ stress. Without the heavy metals, the germination rate of seeds was the highest and roots were shortest when inoculated with Burkholderia sp. SX9 culture (P < 0.05), and inoculation did not result in differences in shoot length (Table 3). Similar to a previous study (Wahyudi et al. 2011), the ability to promote seed germination may have been due to the IAA produced by strain SX9. Some metal ions mainly inhibit the abscisic acid (ABA) catabolism of plants and cause oxidative stress (Ye et al. 2014). In the Cu 2+ stress negative control, inoculated with the growth medium, the seed germination rate was lower the higher the Cu 2+ concentration (P < 0.05), and at the highest concentration of 400 mg L −1 , the seeds did not germinate (Table 3). Similarly, for the seeds inoculated with strain SX9 culture and culture supernatant, the germination rate decreased with increasing Cu 2+ concentration (P < 0.05), yet the seeds germinated even at the highest Cu 2+ concentration. SPB can provide sufficient bioavailable Fe 3+ to plants, which can promote plant growth and reduce the toxicity of heavy metals to plants (Marcus and Mohamed 2007). Interestingly, even though without the heavy metals the germination rate was promoted more by strain SX9 culture than by the supernatant, the supernatant promoted the germination rate more than the culture at all Cu 2+ concentrations tested (P < 0.05). At the concentrations of 200 mg L −1 and 400 mg L −1 Cu 2+ , the roots of L. perenne inoculated with strain SX9 culture were shorter than the roots of L. perenne inoculated with the supernatant (P < 0.05). Thus, the growth promotion under Cu 2+ stress was likely not due to the siderophore that was expected to remain in the supernatant. Characterizing the plant growth-promoting property effective under Cu 2+ stress requires further analyses.
In the Zn 2+ stress negative control, inoculated with the growth medium, the seed germination rate was lower above 200 mg L −1 Zn 2+ concentration (P < 0.05) ( Table 4). At the concentrations of 500 mg L −1 and 1000 mg L −1 Zn 2+ , the germination rates of seeds inoculated with strain SX9 culture and supernatant were higher than that in the negative control (P < 0.05). Possibly the siderophore effectively bound Zn 2+ , making it less toxic. At the highest Zn 2+ concentration of 1000 mg L −1 , inoculation did not result in differences in shoot and root lengths.
Low concentrations of Cd 2+ have been found to promote the germination of seeds (Chen et al. 2013). Similarly, in the Cd 2+ stress experiment, the seed germination rate was higher with than without Cd 2+ (P < 0.05) ( Table 5). At the highest Cd 2+ concentration of 20 mg L −1 , inoculation did not result in differences in germination rates, yet shoots were shortest in the negative control and roots were longest when inoculated with the culture supernatant (P < 0.05). In addition to the siderophore, the growth promotion may have been due to the production of IAA, ACC deaminase, and other substances (Li 2018). Bacterial IAA can help break seed dormancy by stimulating the division of plant cells to promote the development of root and shoot (Aka and Babalola 2016). Similar enhancement of seed germination by inoculation with Bacillus sp. RM-2 and bacilli formulations has been observed in food crops such as cowpea, maize, and pearl millet (Minaxi et al. 2012). Taken together, the results indicated that strain SX9 may be applicable to support the growth of L. perenne in remediating heavy metal-contaminated soil.

Conclusion
Forty-four bacterial strains were isolated from a heavy metalcontaminated soil sample. A high-yielding SPB, designated as The values are mean ± standard deviation (n = 3). Different capital and lowercase letters indicate statistically significant differences between Cd 2+ concentrations and inoculants, respectively (P < 0.05) strain SX9, was identified as Burkholderia sp. Its SU value was 93.7%, which is at a higher level as compared with other Burkholderia spp. The bacterium showed good growth and siderophore production capacities under different temperatures, pH, Fe 3+ , and NaCl concentrations. Burkholderia sp. SX9 owns prominent tolerances to test heavy metals. Results also show that Burkholderia sp. SX9 could produce IAA and ACC deaminases, dissolve phosphate, and promote germination of L. perenne seed under heavy metal stresses. In general, Burkholderia sp. SX9 has the potential as a candidate for microbe-assisted phytoremediation of heavy metalcontaminated soils. However, further studies are still needed, such as siderophore structural identification, analysis of factors affecting the process of combined remediation, cloning, and analysis of relevant functional genes, especially the evaluation on inoculum survival and colonization for practical application. Data availability The 16S rRNA gene sequence of strain SX9 was submitted to the National Center for Biotechnology Information (NCBI) GenBank with the accession number MN658668. The dataset generated during and/or analyzed within the current study is available from the corresponding author on reasonable request.

Declarations
Ethics approval Not applicable.
Consent to participate Not applicable.
Consent to publish All authors read and approved the final manuscript. The work described has not been published. It is not under consideration for publication elsewhere. Its publication has been approved by all coauthors. If the manuscript is accepted for publication, the authors agree to automatic transfer of the copyright to the publisher.

Competing interests
The authors declare no competing interests.