Metabolites produced by inoculated Vigna radiata during bacterial assisted phytoremediation of Pb, Ni and Cr polluted soil

Phytoremediation assisted with plant growth promoting bacteria (PGPB) is a green technology to remediate metal contaminated soils. Plants usually produce secondary metabolites to tolerate metal toxicity. Present study was designed to explore the phytoremediation potential of Vigna radiata var. NM-II in the presence of metal resistant PGPB and comparison of metabolites produced under heavy metal stresses (Pb, Ni, Cr). Three PGPB selected for present study include Bacillus pumilus MB246, Serratia nematodiphila MB307 and Delftia Lacustris MB322. Pot experiments were conducted with inoculated V. radiata NM-II seeds grown in soil artificially contaminated with lead (Pb), Nickle (Ni) and chromium (Cr) at a concentration of 300, 200 and 100 mg/kg respectively. After harvesting various growth parameters were studied (root length, shoot length, fresh weight and dry weight). Bacterial colonization on root surfaces of harvested plants was observed through Scanning electron microscopy (SEM) and Elemental composition was recorded through Energy dispersive X-ray spectroscopy (EDX) attached with SEM. Metabolic response of harvested plants was studied through Gas chromatography Mass spectrophotometry (GC-MS) analysis. Metal accumulation in roots, shoots and soil was analysed by acid digestion method from which Bioaccumulation factor (BF) and Translocation factor (TF) of metal from soil to plant was calculated. Results revealed stimulatory effect of PGPB on growth and phytoextraction ability of V. radiata. Soil metal removal efficiency was in the order Pb>Ni>Cr, whereas metal distribution in each part of plant was root>stem>leaf. The BF and TF values suggested V. radiata as Pb and Ni excluder while moderate accumulator for Cr. Elemental analysis through Energy Dispersive X- ray spectroscopy (EDX) found potassium (K+)and calcium (Ca+)as highly abundant nutrients with least accumulation of sulphur (S). Metabolites study through GC-MS revealed variety of compounds (carbohydrates, amino acids, fatty acids, steroids etc) detected differentially under each metal treatment and their concentration was influenced by different bacterial inoculations. Overall 9-Octadecenamide was found as commonly present lipid compound in most of the treatments which is required for detoxification in plants. The study concluded beneficial role of PGPB for successful phytoremediation of heavy metals and differential response of metabolites towards each metal stress that is related to metal tolerance ability of V. radiata.

The nutrient elements phosphorus (P), sulphur (S) and potassium (K) are known to play an important role in growth and metabolism of plants. They contribute towards survival of plants under various biotic and abiotic stresses [27,28]. HMs severely inhibits the growth of plants and even resulted in plant death by disturbing the nutrients uptake [29]. Several studies have reported reduction in uptake of most mineral elements e.g Ca, Mg, K and Fe under HM stress [30][31][32]. Rhizospheric plant growth promoting bacteria are useful in solubilizing, mobilizing, and transforming nutrients. They improve the availability of minerals through various direct and indirect mechanisms, including atmospheric nitrogen fixation, phosphate and potassium solubilization and by releasing certain volatile organic compounds [33]. The selected PGPB in present study have the capacity to promote growth of V. radiata under Pb, Ni and Cr stress and enhance the phytoextraction of these metals from soil. They can also promote the production of secondary metabolites to mitigate the effect of metal stress. The objectives of current study were to assess the impact of PGPB on phytoremediation ability of V. radiata to remediate Pb, Ni and Cr polluted soil and to identify the predominant metabolites produced in response to metal stress using GC-MS analysis. Current study has novelty and practical applications which are of relevance to food, crops, biodiversity and climate change and will have true value to the public. The study is supportive for agriculturist and environmentalist to clean the agricultural land from metal pollution without any extra cost and make it productive for growth of crops.

Pot experiment
Pot experiments were conducted in a greenhouse located at Fatimma Jinnah Women University, Rawalpindi, Pakistan under natural conditions in the month of April. The geographical description of study area is Latitude: 33.6007, Longitude: 73.0679, 33˚36' 33" North, 73˚4' 44" East. Altitude is 497 m with humid subtropical climate. The average temperature was 26˚C and the photoperiod was 12 hrs.
Garden soil used for this study (0-20 cm depth) was first air dried and sieved (2 mm). It was then analysed for various physiochemical characteristics. Soil pH and electrical conductivity (EC) were determined in 1:1 soil to water suspension [34]. For the measurement of organic matter, loss on ignition method was used [35]. Moisture content was determined by drying method. For phosphate and sulphate analysis methods of Mussa et al. [36] were used.
Soil was sterilized and artificially contaminated with Pb, Ni & Cr. Metals were added in soluble form to ensure plant availability. Aqueous solutions of PbNO 3 , NiCl 2 and K 2 CrO 4 were prepared in autoclaved distilled water to provide a contamination level of 300, 200 and 100 mg of Pb, Ni and Cr kg -1 of soil, respectively. These solutions were poured on soil and mixed thoroughly for even distribution of each metal. Soil was then air dried in shade and allowed to settle for 2-3 days.
Small sized clay pots (5x6 inches) were used for plant growth. Each pot was filled with half kg of soil. Three strains were selected for seeds' inoculation based on their performance in previous hydroponic experiments [37]. These include Bacillus pumilus MB246, Serratia nematodiphila MB307 and Delftia Lacustris MB322. Seeds of V. radiata (var. NM-II) were obtained from National Agricultural Research Centre, Islamabad, Pakistan. Healthy and surface sterilized seeds (with 0.1% HgCl 2 ) were inoculated with their respective single bacterial suspension. For bacterial suspension twenty four hour freshly prepared bacterial cultures grown on Nutrient agar medium supplemented with 200 μg/ml of respective metal (i.e. Pb, Ni and Cr) was used. Inoculum was prepared by mixing bacterial culture in 10 mL autoclaved distilled H 2 O. The optical density of inoculums was adjusted to 10 6 cells per mL at 600 nm in spectrophotometer in order to give uniform inoculum of each strain.
Control seeds were soaked in autoclaved distilled H 2 O. Initially eight seeds were sown in pot soil and irrigated with water. After seedlings germination, each pot was thinned to five plants per pot. Plants were supplied with water and Nutrient solution [38] as required and allowed to grow under natural conditions. After three weeks of growth plantlets were removed carefully from the pots and washed. Growth was measured in terms of root length and shoot length. Plantlets were separated into roots, stems and leaves and their fresh weights were recorded individually. Some plant material was frozen for metabolic studies while remaining was dried at 70˚C for 24 hrs.

Heavy metal analysis
Heavy metal analysis was done by acid digestion method. Dried plant material (0.5g) was ground and digested with a mixture of HCl/HNO 3 (3:1, v/v). Concentrations of metals in the digested samples were determined using atomic adsorption spectroscopy (AAS). Bioaccumulation factor (BAF) was calculated by dividing the total concentration of metal in plant tissue by the total concentration of metal in targeted soil.

Bioaccumulation factor BF
ð Þ ¼ Average conc: of metal in the whole plant mg=kg ð Þ Metal added in soil mg=kg ð Þ Translocation Factor (TF) i.e. ratio of HMs concentration in plant shoots to its roots was calculated according to Yadav et al. [39].

Efficiency of metal removal from soil
Percentage removal of metal from soil was calculated as follows; Co = initial metal concentration in soil Ce = final metal concentration in soil

Bacterial colonization on root by Scanning Electron Microscopy (SEM)
To study the bacterial colonization on the surfaces of plant roots, the sterilized V. radiata seeds were first germinated and grown in sterilized petri plates. After the growth of seven days in petri plates these seedlings were then allowed to grow in 50 mL tubes each supplied with 25 mL of Hoagland's nutrient solution and 100 μg/mL of respective metal solution (Pb, Ni and Cr). The suspended seedlings in each tube were inoculated with one mL of their respective bacterial suspension. Seven days old seedlings were harvested. Some samples were frozen for GC-MS analysis. For scanning electron microscopy, harvested plant roots were dried and then cut into small pieces of 5-10 mm size. These pieces were mounted onto metal stub with double sided carbon tape and sputter coated with gold palladium. Each sample was observed under SEM (Tescon model) at varying magnifications (500, 1000 and 5000) with 20KV light. Energy dispersive X-ray spectroscopy (EDX) attached with SEM was used for the identification of elemental composition of samples. The atomic percentage of elements in each plant sample was obtained by X-ray emission based spectral peaks.

Metabolite profiling by Gas chromatography Mass spectrometry (GC-MS)
GC-MS is used for quantitative and qualitative detection of compounds present at molecular levels with very high accuracy. In current study the metabolic analysis was carried out by Schimadzu Gas Chromotograph and Mass-spectrometer following the method of Sun et al. [40]. For metabolite extraction, 100 mg of harvested frozen and ground plant material was extracted in 5 mL of extraction mixture (methanol/chloform/water in 2.5:1:1v/v/v) for 2 hrs by continuous shaking at room temperature. The extraction solution was centrifuged at 15000 rpm for 10 minutes. The polar supernatant (methanol/water) phase and non-polar (chloroform/methanol) phase were collected in separate tubes. The polar phase was reduced to dryness and dried polar residue was methoximated with 50 μL of 20 mg/mL of methoxyamine hydrochloride in pyridine for 90 min at 37˚C followed by trimethylsilylation with 80 μL N-methyl-N-trimethyl silyl triflouroacetamide (MSTFA) for 60 min at 37˚C. For metabolite analysis, 1 μl of derivatized sample was injected in GC-MS. Non-polar layer was used for lipid analysis. For this samples were resuspended in 0.4 mL chloroform followed by addition of 0.5 mL of 0.25M HCl in methanol and incubated at 50˚C for 4 hrs. After evaporation of solvents and HCl, the samples were resuspended in 35 μL of pyridine and again incubated at 50˚C until all the residue dissolves. Samples were then treated with 30 μL of MSTFA and incubated at 50˚C for 1 hr. 1 μL of sample was injected in GC-MS for lipid analysis. All compounds were identified by using mass spectrum data of GC-MS coupled with National Institute Standard and Technology (NIST) database attached with GC-MS and from NIST chemistry web book.

Statistical analysis
All measurements were taken as mean of three replicates ± standard error of means. Metabolites Data was analysed by Principal component analysis using PAST software.

Results
The results of various physiochemical characteristics of soil are displayed in S1 Table in S1 File. It was a light brownish soil was sandy loam texture. It has a 9 pH and electrical conductivity 144 (μs/cm), Moisture content of the soil was 1.62%, while it contained 2.11% organic matter.

Growth response of Vigna radiata in Pb, Ni and Cr contaminated soil
The growth response of V. radiata with respect to length (root and shoot) and weight (fresh and dry) were studied and displayed in Table 1. It was observed that all the growth indicators in inoculated V. radiata were higher than uninoculated control. The growth response also varied with different bacterial inoculations. Highest increase in root length was observed with MB246 (15.97%) inoculation followed by MB307 (12.2%). However, the shoot length exhibited maximum increase with MB307 inoculation (11%) followed by MB322 (10.7%). In case of fresh weights and dry weights, maximum increase was observed with MB322 inoculation i.e. 29.1% and 23.3% respectively.

Metal remediation performance of Vigna radiata towards Pb, Ni and Cr
The performance for phytoremediation of Pb, Ni and Cr by V. radiata was evaluated through its Bioaccumulation Factor (BF) and Translocation factor (TF) values as shown in Table 2. It was observed that metal uptake by plant proportionate to its concentration in the soil. It increased with the increase in metal concentration. The BF and TF values varied with each metal. Maximum BF values were observed for Cr followed by Ni and Pb. However, Cr exhibited minimum TF values. Bacterial inoculations also variably affected metal uptake. Maximum BF for Ni and Cr was with MB246 inoculation i.e. 0.6 and 0.71 respectively. Whereas, for Pb maximum BF was observed with MB307 inoculation i.e. 0.52. Results of TF for all metals revealed maximum values with MB322 inoculation i.e. 0.99 for Pb followed by 0.93 for Ni and 0.36 for Cr. The metal remediated from soil was determined by analysis of metal left in soil after harvesting ( Table 3). The efficiency of metal removal from soil (Fig 1) showed Pb as highly remediated metal followed by Cr and Ni.

Distribution of Pb, Ni and Cr in different parts of Vigna radiata
Concentration of metal that accumulated in each part of Vigna radiata was studied through atomic absorption spectroscopy. Metal uptake by root, stem and leaf varied towards each metal and with each bacterial strain (Fig 2). Overall maximum accumulation of metals was observed in roots as compared to shoots. Metal content accumulated in each part of V. radiata decreased in the order; root>stem>leaf. However, the trend slightly changed with Pb having more accumulation in leaf as compared to stem. The highest metal concentration in root, stem and leaves of V. radiata were found in MB307 inoculated plantlets for Pb i.e. 3.38mg/g, 1.73mg/g for Ni in stem and 0.63mg/g for Cr in leaf.

SEM-EDX studies
Scanning electron microscopy of inoculated V. radiata seedlings showed presence of bacterial strains on root surfaces as seen in Fig 3. Energy Dispersive X-ray (EDX) attached with SEM was used for elemental analysis. Elemental profile (weight %) in different parts of plant is displayed in Table 4. Potassium (K) was recorded as highly abundant element in most of the samples whereas Sulphur (S) was found in minimum percentages. Highest concentration of Ca, K and Mg was found in MB307 inoculated stem. Whereas, Al, Fe and Si were present most abundantly in MB246 inoculated roots.

Metabolic response of V. radiata to metal stress
Metabolites present in plant extracts of uninoculated and inoculated V. radiata seedlings exposed to 100 μg/mL of Pb, Ni and Cr was studied through GC-MS. Both polar and nonpolar phases of plant extracts were analysed to study the variety of lipids and other secondary metabolites present in these layers. Name of compounds along with their molecular formula, concentration (peak area %) and retention time (RT) are displayed in Tables 5 and 6 respectively. Identified compounds were analysed through principal component analysis (PCA). a. Lipophilic (non-polar layer). GC analysis of V. radiata revealed up to 24 metabolites in lipophilic layer (Fig 4). 6-octadecenoic acid, methyl ester (12.87%) was major compound detected in lipophilic layer of MB322 inoculated seedlings under Pb stress. In the presence of Ni, MB307 inoculation resulted in maximum production of Glycine, N-(phenylacetyl)-, methyl ester (54.63%), whereas under Cr stress L-Glutamic acid was identified with maximum peak area (57.8%) in MB246 inoculated seedlings. It was observed that only Picolinamide; Nonadecanoic acid; L-Cysteine and N,S-bis (2,6-difluorobenzoyl)-, methyl ester were produced when plant was under Pb stress whereas, Docosanoic acid; Cyclopropanecarboxylic acid,-2-(2-propynyl) methyl ester; Hexadecanamide; Picrolonic acid; 8-Methyl-6-nonenamide; 4-Dimethylamino-3,5-dinitrobenzoic acid and Glycolic acid, 2TMS derivative were particularly produced under Ni stress. In the presence of Cr, Nickel, (eta-2-2-diallyl ether); Tetradecanamide and cis-5,8,11-Eicosatrienoic acid, methyl ester was produced unlike other treatments. Hexadecanoic acid, methyl ester or pentadecanoic acid methyl ester were commonly produced in control and under all metals i.e. Pb, Ni and Cr (S1 Table in S1 File).

Discussion
Phytoremediation is a green technology for the removal of heavy metal contaminants. Mung bean has been reported as an efficient plant for phytoremediation of various contaminants [41][42][43]. Phytoremediation efficiency of hyper-accumulating plants can further be improved by using metal resistant PGPB that alleviated metal stress and improved plant growth [44,45]. Gupta [46] reported Pseudomonas sp. for successful phytoremediation.
In our study, metal resistant PGP strains Bacillus pumilus (MB246), Serratia nematodiphila (MB307) and Delftia Lacustris (MB322) were tested for bacterial assisted phytoremediation of Pb, Ni and Cr polluted soil by Vigna radiata. Growth of inoculated plants appeared healthy in metal stress soil as compared to uninoculated control. Improved growth under metal stress  condition might be attributed towards PGP activities of these strains. Our results are in accordance with some other studies which reported enhanced growth of inoculated Vigna radiata in the presence of metals [47,48]. Plants respond to abiotic stressors through activation of stress response genes. Expression of these genes increase by PGPB [49]. The Translocation factor and Bioaccumulation factors are two key indicators that determine the phytoremediation efficiency of plant. Major accumulation of metals in plant roots is one of the strategies used by plants to tolerate heavy metal stress. Metals are translocated from roots to shoot by membrane metal transporters and metal-binding proteins. Through certain mechanisms a balance is maintained between having enough essential metals available for metabolic functions and at the same time avoiding toxicity and to keep nonessential metals below their toxicity thresholds. Therefore excess amount of metals are stored in where they are less toxic for cellular metabolism e.g. in roots. Roots are expected to take up a greater amount of contaminants, and they represent a significant metal sink [50]. In current experiment, higher BF and low TF values for Cr showed that Cr retained more in roots as compared to shoots which indicated presence of plant mechanisms that prevent Cr translocation to shoots. Prapagdee et al. [51] also reported high BF and Low TF values for Cd by Vigna radiata. A study by Anjum et al. [52] presented Vigna radiata as good Cd stabilizer and Cd extractor. In our study, lowest BF values with highest TF values were obtained for Pb and Ni that indicated more translocation of Pb and Ni in shoot. This also relates to maximum metal removal of Pb from soil followed by Ni and Cr. According to Kumar et al. [53], Pb translocation from root to shoot is generally low in plants. However, in hyper-accumulating plants a higher TF is important. Results depicted stimulatory effect of PGPB on metal extraction and translocation ability of Vigna radiata as observed with greater BF and TF values in most of the inoculated plants. This might be attributed towards increased bioavailability of metals through the synthesis of chelators and siderophores to solubilize and sequester metals from soil [54]. Our observations for improved phytoremediation with selected PGPB related to findings in our previous study under in vitro conditions. Moreover, use of Serratia sp. Pseudomonas sp. and Bacillus pumilus for improved phytoremediation of HM contaminated soil has been recognized in several studies [55][56][57].
Distribution of metals in different parts of plant exhibited maximum metal content in roots followed by stems and leaves. Maximum metal accumulation in roots is a consequence of metal tolerance by plant to reduce the effect of metal toxicity [58]. In current study, presence of higher Pb content in leaf relative to stem attributed towards high availability of Pb in the soil. Aransiola et al. [59] reported more accumulation of Pb in above ground biomass of G. max L. in the order of seeds> leaves> root>stem.
Uptake of mineral nutrients like, Ca, P, K, Mg, Mn, Na Fe, B and Zn is severely retarded in metal stress plants [32,60]. Results of elemental analysis revealed Ca and K as highly abundant minerals, whereas S was present in minimum quantity. The elemental composition of metal stress V. radiata is not effected in the presence of PGPB. Bacterial inoculations enhanced the elemental composition in most of the cases. Desai et al. [61] also reported enhanced uptake of nutrients in V. radiata inoculated with PGPB. Plant growth promoting bacteria help plant hosts in the absorption of mineral nutrients in metal polluted soils [62]. Serratia sp. inoculation to linseed under Cd stress maintains an optimum level of nutrients [55]. Inoculation of Pseudomonas sp. enhanced nutrient uptake in sunflower and tomato [45]. Elemental concentration also varied in different parts of plant as depicted in results of current experiment. Ashraf et al. [63] reported higher concentrations of all the metals in the roots except for K, which was greater in the shoots of mung bean.
The exposure to heavy metal toxicity generates wide-ranging metabolic and physiological changes in plants [16]. Plants can cope with these heavy metal stresses by secondary metabolites production. Several plant secondary metabolites such as ascorbic acid, lipoic acid, αtocopherol, and melatonin can be applied to improve tolerance to abiotic stress [64]. Plant growth promoting bacteria can contribute to the productions of plant secondary metabolites. Plants can detect certain molecules released by microbiomes through a chemical recognition system, which activate the plants to generate signal transduction networks and make corresponding changes in related gene activities, leading to the production of certain plant secondary metabolites. Horizontal gene transfer in plants-endophytes may also lead to changes in plant secondary metabolites [65].
In current study, various metabolites were identified through GC-MS in metal stress Vigna radiata. These metabolites belonged to various groups (organic acids, fatty acids, amino acids, amines, phenol, steroids etc) that might play a vital role in metal resistance and accumulation. Results of metabolic analysis revealed varied metabolic response of plants towards each metal as different heavy metals have different sites of action within the plant. Manipulation of genes responsible for production and action of secondary metabolites can improve the tolerance level and adaptability of plants under stress conditions [3]. The differences and separation of metabolites under each metal treatment was studied through principal component analysis (PCA).
Majority of metabolites in polar layer were absent in control sample. However, most of them appeared in samples treated with different metals. Principal component analysis (PCA) of metabolites (Fig 6) showed that control and sample treated with Cr are closely related and produced few similar metabolites. Similarly, PCA biplot also revealed that Ni and Pb treated samples are somehow related to each other and produced few similar metabolites e.g. Octadecane nitrile. 1-Heptadecanol, TMS was found only in Ni and Cr stress seedlings.
In lipophilic layer, the metabolites detected in control were different from other samples. However, most of them appeared in samples treated with Ni followed by Pb and Cr. PCA biplots (Fig 7) showed that control and sample treated with Cr are closely related and produced few similar metabolites. Lead and Cr also produced few similar metabolites e.g. Eicosatrienoic acid, methyl ester. Metabolites produced under Ni stress are different from those produced under lead and chromium stress (Docosanoic acid, hexadeccaanamide etc). Octadecenoic acid methyl ester was identified in all metal stress (Pb, Ni and Cr). While 11-Octadecenoic acid, methyl ester was detected only in Cr stress seedlings. 9-octadecenoic acid has antimicrobial activity [66]. Azar et al. [67] reported fatty acid compounds; n-hexadecanoic acid, 9-octadecenoic acid (Z)-and octadecanoic acid found abundantly in Leucobryum javense. In a study by [20] organic acid like hexadecanoic acid, linoleic acid, α -linolenic acid and octadecanoic acid were significantly increased in response to Pb stress. Amino acids and derivatives can chelate metals conferring in plants resistant to toxic levels of HMs [68]. Amino acids like proline produced under heavy metals exposure effects the synthesis and activity of the antioxidant enzymes catalase, peroxidase. Some studies have reported for Ni complexation with phosphate, histidine and phytate [25]. In Vinca rosea the alleviated Ni stress levels attributed towards the enhanced synthesis of flavonoids, protein, proline and phenols [69].

Conclusion
The current findings confirm that the application of PGPB is effective for phytoremediation of Pb, Ni and Cr polluted soil by V. radiata. The uptake of metals and its transport in various parts of plant with least accumulation in shoots make them safe for eating purpose keeping in view that the accumulated metal is not able to transport in the extreme upper part of plant shoot where grains are present. GC-MS analysis demonstrated a range of compounds (carbohydrates, fatty acids, amino acids, etc) produced distinctly under each metal stress which might be associated with metal tolerance ability of V. radiata. In future, further insight into direct relation of these metal responsive metabolites with metal tolerance is required at molecular and cellular level to investigate the mechanism of plant adaptation and genes involved in it which will give further advancement in gene manipulation of plant and bacteria for efficient phytoremediation.