Metal contamination of soils has become one of the most significant environmental problems today. Metal uptake by crop plants from the contaminated agricultural lands can have strong adverse impacts on human health through the food chain (Karavoltsos et al., 2002). Moreover, excessive metals in the soil can result in decreased crop yield due to the inhibition of plant metabolic processes (Singh and Aggarwal, 2006). Apart from the metals with unknown biological functions (Cd, Cr, Pb, Co, Ag, Se, and Hg), essential elements (Fe, Mn, Zn, Cu, Mg, Mo, and Ni) also keep accumulating in agricultural soils by means of wastewater irrigation, animal manures and sewage sludge application, use of fertilizer and agrochemicals (Thomas et al., 2012). In the toxicological point of view, the essential elements are also important, because, at higher concentrations they too can be toxic to plants as well as to dietary intake levels (Karavoltsos et al., 2002).

With the continuous addition of undesirable metals into the environment, remediation of contaminated soils receives increasing attention (Cao et al., 2007). However, due to the fact that metals are not easily degraded, remediation of the contaminated soils is always considered being a demanding exercise (Rajkumar et al., 2008). Depending on the resource availability, severity of the problem, nature of the metals and contaminated soil, different methods have been employed in restoring the contaminated lands (Arunakumara et al, 2013). In this context, systematic technologies such as bioremediation, physical/chemical remediation and integrated remediation are among the widely used techniques (Luo, 2009). However, the physical and chemical methods such as physical separation, acid leaching or electrochemical processes, are considered to be ineffective because of high cost, low efficiency, and destruction of soil structure and fertility (Jing et al., 2007). In contrast, phytoremediation, a method which uses plants to extract, sequester and detoxify pollutants has received considerable attention (Arunakumara, 2011). However, the wider application of the technology has been restricted due to the limitations such as low soil thickness that can be treated, low translocation rate of metals from roots to shoots, and the slowness of the treatment (Lebeau et al., 2008).

The amount of heavy metals uptake in plants varies with the mobility and the concentration of metals in soil (Chen et al., 2010) and the interface between soil microbes and plant roots (rhizosphere) is displayed to have a great influence on the uptake of nutrients as well as on the decrease of metal toxicity (McNear, 2013). Since soil microbes could alter the metal status of the soil (Fazal and Bano, 2010), exploitation of such microbes to reduce the metal toxicity to plants is worth investigating (Rajkumar and Freitas, 2008). In this context, some metal resistant bacterial strains were proved exceptional at enhancing the growth of the host plant through different mechanisms such as the production of plant growth promoting substances, nitrogen fixation and phosphate solubilization (Hemambika et al., 2013). As reported by Rajkumar et al. (2008), heavy metal tolerance of the microbes may be attributed to one or several mechanisms including exclusion, active removal, biosorption, and precipitation or bioaccumulation of metals both in external and intracellular spaces. Therefore, microbes having remarkable metal tolerance and plant growth-promoting abilities could play a significant role in remediation of metal-contaminated soils, because bioaugmentation with such microbes could promote phytoextraction of metals (Prapagdee et al., 2013). In the present study, we isolated phosphate solubilizing bacterial strains from metal-contaminated soils and the strain with the highest degree of metal resistance was employed in (i) assessing the potential of mobilization of Co, Pb and Zn, and (ii) evaluating the effects of inoculation with the selected strain on plant growth and uptake of Co, Pb and Zn by Helianthus annuus (sunflower).

Heavy metal contaminated soils collected from abandoned mines of Boryeong area in South Korea were used in isolating phosphate solubilizing bacteria. Aliquots of serially diluted soil samples were inoculated on NBRIP (National Botanical Research Institute Phosphate) medium containing 10 g glucose, 5 g Ca₃(PO₄)₂, 5 g MgCl₂.6H₂O, 0.25 g MgSO₄.7H₂O, 0.2 g KCl, 0.1 g (NH₄)₂SO₄ in 1 L distilled water (Nautiyal, 1999). The pH of the media was adjusted to 7 ± 0.1. The petri plates were incubated at 30⁰C for 7 days. Morphologically distinct colonies with clear halos were purified by repeated sub culturing. A total of 20 isolates were selected based on the greater halo size (>3 mm) and maintained on solid NBRIP agar medium until use.

Isolated bacterial strains were assessed for their resistance to heavy metals using the agar dilution method (Cervantes et al., 1986). Freshly prepared agar plates were amended with 3 different heavy metals; CoCl₂.6H₂O, 2PbCO₃.Pb(OH)₂, and ZnCl₂ at various concentrations ranging from 100-400 μg/mL (Co, Pb and Zn). They were inoculated with isolated strains and heavy metal tolerance was determined by the appearance of the bacterial growth after 2 days of incubation at 30⁰C. The bacterial strain showing the highest degree of metal resistance was selected for further studies.

The partial sequencing of 16S rRNA for the bacterial strain was done with the help of DNA sequencing service, SOLGENT, Daejeon, South Korea using universal primers, 27F (5′-AGAGTTTGATCCTGGCTCAG -3′) and 1492R (5′-GGTTACCTTGTTACGACTT -3′). The online program BLAST was used in identifying the related sequences with known taxonomic information available at the databank of NCBI (http://www.ncbi. nlm.nih.gov/BLAST). A Phylogenetic tree was constructed using CLUSTAL X program (Thompson et al., 1997), which involved sequence alignment by neighbor joining method (Saitou and Nei, 1987) and maximum parsimony using the MEGA4 program (Kumar et al., 2001). Grouping of sequences was based on confidence values obtained by bootstrap analysis of 1,000 replicates. Gaps were edited in the BioEdit program and evolutionary distances were calculated using Kimura two parameter model. Reference sequences were retrieved from GenBank under the accession numbers indicated in the trees.

NBRIP medium supplemented with heavy metals (Co, Pb and Zn) at the concentration of 200 mg/L was inoculated with bacterial suspension (10⁶ colony forming units/ml) and incubated with continuous shaking at 30⁰C. Optical density of culture supernatant was measured at definite time intervals using UV spectrophotometer at 660 nm to estimate the cell growth.

Bacterial culture having 10⁶ colony forming units/mL (2 days old) was inoculated in sterilized liquid NBRIP medium (250 mL) supplemented with different heavy metals (Co, Pb and Zn) at the concentration of 200 mg/L and incubated with continuous shaking at 30⁰C. Sterilized liquid NBRIP medium without supplemented with heavy metals was served as a control. A sample (10 ml) of each cultured and control were taken and centrifuged at 12000 ×g for 15 min. The clear supernatant was used in determining the pH and amount of phosphorous released into the medium.

A pot experiment was conducted under green house conditions at the College of Agriculture, Chungnam National University in October 2012. The soils collected from several locations of a nearby forest and a waste button mushroom bed in Buyeo-gun area, Chungchungnam-do, South Korea, were mixed with the ratio of 1:1, air dried and sieved (2 mm). Sterilized soil (by steaming at 100℃ for three consecutive days) was amended with aqueous solutions of different heavy metals (Co, Pb and Zn) to achieve the final concentrations of 200 mg/kg soil. They were then kept for 2 weeks in a greenhouse for metal stabilization and used in filling the plastic pots (25 cm diameter, 35 cm height). Seeds of H. annuus were surface sterilized by immersing in alcohol (70%) for 40 s, NaClO (1.0%) for 15 min followed by rinsing several times with sterile distilled water. Seeds sown in germination trays containing sterilized non-contaminated soil were provided with 14/10 light/ dark regime and kept at 25℃ for germination. Bacterial cultures grown under standard conditions for 2 days were harvested by centrifugation at 12000 ×g at 15 min. The cells were washed twice with sterile distilled water and resuspended in biological saline (0.85% KCl) to be used in inoculation. Three weeks old seedlings were carefully uprooted from the germination bed and their roots were dipped in the bacterial culture (10⁹ colony forming units/mL) for 2 h. They were transplanted into the plastic pots (five plants/pot) containing 300 g of metal contaminated or non- contaminated soil and allowed to grow at 25℃ and 14/10 light/dark regime. The average pH of soil at the time of planting was recorded as 6.65. Three weeks later, the plants were carefully uprooted and cleaned the root surface thoroughly with distilled water. As growth parameters, fresh and dry biomass was measured and accumulation of metals in plant biomass was quantified as described by Freitas et al. (2004). Each treatment had three replicates.

The impact of bacterial inoculation on the mobility of metals in soil was investigated under laboratory conditions with 50 mL scaled polypropylene centrifuge tubes. The bacterial strain transferred into 100 mL flasks containing LB broth was cultured aerobically on a rotating shaker (150 rpm) at 30℃ until reaching the final concentration of 10⁶ colony forming units/mL. The bacterial cells were then harvested by centrifugation at 10000 × g for 15 min and washed in phosphate buffer (pH 7.0) twice. The bacterial pellet was washed in sterile water, re-centrifuged, and finally re-suspended in 5 mL sterile water. Artificially contaminated soil (1 g) in the centrifuge tubes was inoculated with small aliquots (up to 1 mL) of the final washed bacterial culture. After taking the weight the tubes, they were wrapped with brown paper and placed on an orbital shaker at 200 rpm at 25 ℃. At the end of the period of 10 d, the weight of the tubes was recorded and 10 mL of sterile water were added to each tube to extract the soil water soluble heavy metals. The extracts were centrifuged at 10000 × g for 10 min and filtered and the metal contents in the filtrate were determined using an atomic absorption spectrophotometer (Perkinelmer, Aanalyst 800, USA). Artificially contaminated soil without inoculation with the strain served as the control after centrifugation.

A total of 20 bacterial strains with the potential ability to solubilize inorganic phosphates were screened based on the greater halo size (>3 mm). They were then assessed for their resistance to heavy metals and the bacterial strain showing the highest degree of metal resistance was selected for the study. According to 16S rRNA sequence analysis, the selected strain showed close proximity with Enterobacter ludwigii DSM 16688. Phylogenetic tree (Fig. 1) shows the position of the isolated phosphate solubilizing bacterial strain with respect to related species.

[Fig. 1.] Phylogenetic tree based on 16S rRNA gene sequences, showing the position of isolated phosphate solubilizing bacterial strain (PSB-28) with respect to related species. The scale bar indicates 0.002 substitutions per nucleotide position and accession numbers are given in parenthesis.

The growth of the strain measured as the optical density of culture supernatant at definite time intervals is given in Fig. 2. During the incubation period of 36 h, none of the metal was found to be highly toxic to the strain. However, compared to the metal free culture medium, slight reductions in bacterial growth were observed in metal supplemented media.

[Fig. 2.] Growth of Enterobacter ludwigii PSB 28 on sterilized liquid NBRIP medium supplemented with metals (Co, Pb and Zn) at the concentrations of 200 mg/L. Sterilized liquid NBRIP medium without supplemented with heavy metals was served as a control. Values are the means of three replicates. Error bars represent standard deviation.

Phosphate solubilization as measured by the amount of phosphorous released into the medium is depicted in Fig. 3. As indicated by the results, the strain was shown to be capable of utilizing tricalcium phosphate as the sole source of phosphate. However, the presence of metals in NBRIP medium (200 mg/L) caused reduction in phosphate solubilizations. Compared with the control, reductions were 26, 56, and 19%, respectively for Co, Pb and Zn.

[Fig. 3.] Phosphate solubilization by Enterobacter ludwigii PSB 28 on NBRIP medium supplemented with heavy metals (Co, Pb and Zn) at the concentration of 200 mg/L. Sterilized liquid NBRIP medium without supplemented with heavy metals was served as a control. Values are the means of three replicates. Error bars represent standard deviation.

Inoculation with the strain resulted in increased fresh and dry biomass of H. annuus plants, compared to non-inoculated plants (Table 1). When non-inoculated plants were exposed to heavy metal stress, the growth was inhibited in a significant level with p < 0.05. For instance, Pb toxicity caused 23 and 20% reductions in fresh and dry weight of the plant, respectively. Inoculation however led to increase in plant fresh and dry weight in the presence of heavy metals. The fresh and dry weight of the plants grown in Pb contaminated soils were respectively 25 and 26% higher than those of non-inoculated plants. Similarly, in Zn contaminated soil, the percent increments were recorded as 31 and 46% respectively, and in Co contaminated soil, the corresponding figures were 33 and 49%.

[Table 1.] Effect of inoculation with Enterobacter ludwigii PSB 28 on shoot and root weight of Helianthus annuus

The amounts of Co, Pb and Zn accumulated in the roots and shoots of H. annuus grown under inoculated and non-inoculated conditions are given in Table 2. Inoculation with Enterobacter ludwigii PSB-28 resulted in increased accumulation of metals both in the shoots and roots. The accumulations of Co, Pb and Zn in shoots were respectively 45, 79 and 27% higher than those of non-inoculated plants. The corresponding accumulations for Co, Pb and Zn in roots were 21, 18 and 17% higher than those of non-inoculated plants. Regardless of inoculation or non-inoculation, the accumulation of metals in root system was found to be considerably higher than that of in shoots, which has been further confirmed by the low translocation factor (TF) for all the metals. However, TF of Zn was significantly higher than that of the other two metals. Similarly low bioconcentration factor (BCF) was also recorded from Co and Pb. However, the results showed a good agreement and demonstrated that inoculation of the bacterial strain led to increase both TF and BCF of the three metals distinctly.

[Table 2.] Effect of inoculation with Enterobacter ludwigii PSB 28 on accumulation and translocation of Co, Pb and Zn in Helianthus annuus

Metal mobilization potential of the strain was assessed in a batch experiment and the results showed that inoculation of the strain could increase the contents of water soluble metals in the soil. The mobilization of Co, Pb and Zn was respectively 41, 27 and 48% higher than those of the control soil (Fig. 4).

[Fig. 4.] Effect of inoculation with Enterobacter ludwigii PSB 28 on the mobilization of Co, Pb and Zn in soil. Soil without inoculation the strain served as the control. Values are the means of three replicates. Error bars represent standard deviation.

Growth response of the present strain under metal contamination conditions is in line with Rajkumar et al. (2008) and Prapagdee et al. (2013), who observed Zn, Cu and Ni resistance in Bacillus weihenstephanensis and Cd resistance in Klebsiella sp. BAM1. Generally microorganisms isolated from heavy metals contaminated soils possess the ability to withstand against multiple pollutants as they have adapted to such environments (Pal et al., 2005; Abou-Shanab et al., 2007).

The effectiveness of the strain as a plant growth-promoter was assessed with Helianthus annuus, a species known to have an ability to accumulate biomass rapidly and take up substantial amounts of metals (Turgut et al., 2004; Prapagdee et al., 2013). As reported by Ouzounidou et al. (2005) and El-Tayeb et al. (2006), accumulation of plant biomass could be affected by excessive concentrations of heavy metals, which exert adverse impacts on growth and function of root system resulting in poor uptake of water and nutrients. As reported by Jiang et al. (2008), inoculation with Burkholderia sp. J62 led to increase shoot and root dry weights of corn and tomato plants. Inoculation with Pseudomonas fluorescens PsIA12 resulted in enhanced growth of Zea mays and its uptake of N, P and K (Egamberdiyeva et al., 2002). The content of P, K, S and Ca was reported to be increased by the inoculation of rhizobacteria in barley plant grown in metal contaminated soil (Belimov et al., 2004). According to them, inoculation with rhizobacteria resulted in 42% increase in growth of the barley plant compared to the control. Based on the results, they further stated that nutrients play an important role in the detoxification of heavy metals. Their findings were in line with Lebeau et al. (2008), who reported that rhizobacteria could have strong impacts on the nutritional status and the plant resistance to heavy metals. Most recently, Prapagdee et al. (2013) reported that growth of H. annuus could be enhanced by the inoculation of Micrococcus sp. MU1 and Klebsiella sp. BAM1 under Cd contaminated conditions. Belimov et al. (2001) also observed bacterial-assisted growth enhancement in Brassica napus grown in a soil contaminated with Cd. The plant growth-promoting potential of the present strain could be attributed at least partly to the phosphate solubilization ability of the strain under metal stress conditions. In this regards, Rajkumar et al. (2005) also reported that phosphate solubilization ability of Pseudomonas sp. could be contributed to the growth enhancement of the inoculated plants. Inoculation of phosphate solubilizing Bacillus subtilis SJ-101 resulted in higher shoot and root length and biomass with or without Ni (Zaidi et al., 2006). Bacteria is reported to promote the growth of plants (i) indirectly through producing antibiotics to inhibit soil pathogens, and (ii) directly through increasing nutrient and water uptake and thereby the plant biomass (Belimov et al., 2004). Through the production of siderophores, specific enzymes, and organic acids involved in phosphorus solubilization, and fixation of atmospheric N₂, bacteria could assist plants to withstand against metal toxicity (Kloepper, 2003). In this regards, Borgmann (2000) reported that Kluyvera ascorbata SUD165 protected Brassica juncea and Brassica campestris against Ni, Pb and Zn toxicity through the production of enzyme ACC deaminase. Plant growth promoting rhizobacteria was reported to enhance root elongation of Brassica napus by stimulating IAA synthesize (Sheng and Xia, 2006). In Brassica juncea, root elongation was reported to be enhanced by non-identified rhizobacteria (Belimov et al., 2005), Variovorax paradoxus 5C-2 (Belimov et al., 2005) and root dry weight was increased by rhizobacteria (Sheng and Xia, 2006).

Regardless of inoculation or non-inoculation, the accumulation of metals in root systems was found to be considerably higher than that of in shoots. This could primarily be attributed to the poor translocation of heavy metals from roots to shoots (Rajkumar et al., 2006). However, as shown in Table 2, translocation factor of the each metal was increased with the inoculation of the strain, which was of enormous practical significance. Furthermore, metal accumulations in both shoots and roots were found to be higher in inoculated plants than those of non-inoculated plants. Similar observations were made by Rajkumar et al. (2008) for Zn accumulation in H. annuus inoculated with Bacillus weihenstephanensis. However, according to Wani et al. (2007), inoculation of Bradyrhizobium sp. on surface sterilized seeds of Vigna radiate reduced the concentration of Ni in roots, shoots and grains by 15, 19 and 22%, respectively, compared with non-inoculated plants.

As observed by Walpola and Yoon (2013), acidification of the medium could facilitate the inorganic phosphate solubilization by the PSBs. According to them, the acidification occurs mainly through the production of low molecular weight organic acids such as oxalic acid and gluconic acid. Analogous to their findings, an inverse relationship between pH and soluble phosphorus concentration from Ca₃(PO₄)₂ by Enterobacter ludwigii PSB-28 was also observed in the present study. The inter-relationships among soil pH, solubility and speciation of metals have been intensively investigated (Gadd, 2004). Bacteria such as Azotobacter chroococcum (N-fixing bacteria), Bacillus megaterium (P-solubilizer) and Bacillus mucilaginosus (K-solubilizer) (Wu et al., 2006) and Bacillus sp. RJ16 (Sheng and Xia, 2006) were reported to decrease the pH, enhancing the bioavailability of Cd, Pb and Zn (Chen et al., 2005). As stated by Zaidi et al. (2006), reduction in pH from 7.5 to 4.8 with the inoculation of phosphate solubilizing Bacillus subtilis SJ-101 possibly created favourable conditions for the solubilization of metals and their subsequent uptake by the plants. The increased accumulation of metals in the presence of bacterial strain might be due to the increased uptake of metals under acidic soil conditions created by the phosphate solubilization (Rajkumar et al., 2008). Inoculation of Cd-resistant bacterial strains to Brassica napus to a metal contaminated soil significantly increased the plant uptake of Cd when compared with the non-inoculated controls, as a result of pH reduction (Sheng and Xia, 2006). However, on the contrary, Glomus caledonium (Chen et al., 2004) and Glomus mosseae (Citterio et al., 2005) were reported to have no effect on the speciation of Cd and Zn, and Cr and Ni, thus no effect of bioaugmentation by these arbuscular mycorrhizal (AM) fungi on the rate of phytoextraction has been observed, which could be attributed to whether strong symbiotic relationships between AM fungi and host plants.

The present findings of metal mobilization are in agreement with Wu et al. (2006) and Prapagdee et al. (2012) who also reported bacteria-assisted increase in heavy metal mobilization. Generally, the low amount of metals extracted by plants from a soil is attributed mainly to the low availability of metals. As reported by several authors, the available metal content in a soil is less than 1% of the total metal content (Whiting et al., 2001; Braud et al., 2006). Metal availability is influenced by the nature of the metal and soil characteristics such as pH, CEC and organic matter (Kayser et al., 2001; Lebeau et al., 2008). Bioaugmentation could enhance metal bioavailability by increasing the concentration of the available fractions. As revealed by the present results, the release of heavy metals from the non-soluble phases to soluble phases could be facilitated by the bacterial strain. Therefore, increased accumulation of metals, in particular Zn in both the shoots and roots of H. annuus could be attributed to the higher water soluble metal contents in soil inoculated with bacterial strain. As reported by the results of previous studies, H. annuus is capable of accumulating high amounts of Pb, Cd, Cu, Zn and Co, in both the shoots and the roots (Boonyapookana et al., 2005; Marchiol et al., 2007). According to Braud et al. (2006), inoculation of Pseudomonas aeruginosa and Pseudomonas fluorescens has resulted in 113% increment of Pb content in the exchangeable fraction of the soil. However, the Pb concentration bound to free Mn oxides, organic matter and in the residual fraction remained stable. Abou-Shanab et al. (2006) observed an increase of extractable Ni with Microbacterium arabinogalactanolyticum by a factor up to 15. As reported by Baum et al. (2006), the concentrations in NH₄NO₃-extractable Cd, Cu, Pb and Zn in a soil bioaugmented with ectomycorrhizal fungus Paxillus involutus, were 1.22-, 1.11-, 1.33- and 1.33-fold higher than those of non-bioaugmented soil, depending on the soil composition. However, comparing and contrasting of the results of bioaugmentation studies are hard to perform, because the estimation of bioavailable fraction of metals has been done under different conditions with different extractants such as water (Di Gregorio et al., 2006; Wu et al., 2006), MgCl₂ (Braud et al., 2006), NH₄NO₃ (Baum et al., 2006), NH₄O-Ac (Wu et al., 2006), DTPA (Di Gregorio et al., 2006; Wu et al., 2006), KNO₃ (Di Gregorio et al., 2006) and HCl (Wang et al., 2007).

Inoculation with Enterobacter ludwigii PSB 28 resulted in increased shoot and root biomass and enhanced accumulation of Co, Pb and Zn in Helianthus annuus plants. The strain was found to be capable of promoting metal translocation from the roots to the shoots of H. annuus. The beneficial effects of E. ludwigii PSB 28 further demonstrated by metal mobilization potential of the strain. Therefore, Enterobacter ludwigii PSB 28 could be identified as an effective promoter of phytoextraction of Co, Pb and Zn from metal-contaminated soils.