Assessment of the heavy metal bioremediation efficiency of the novel marine lactic acid bacterium, Lactobacillus plantarum MF042018

Heavy metal pollution is one of the most serious environmental and human health risk problem associated with industrial progress. The present study was conducted with the goal of isolation and characterization of metal-resistant lactic acid bacteria (LAB) from the Alexandrian Mediterranean Seacoast, Egypt, with their possible exploitation in metal remediation. Lactobacillus plantarum MF042018 exhibited high degree of resistance, up to 500 and 100 ppm, to both nickel and chromium, respectively, with multiple antibiotic resistance (MAR) index above 0.5. In an attempt to improve chromium removal by L. plantarum MF042018, Plackett-Burman followed by Box-Behnken statistical designs were applied. An initial Cr2+ concentration of 100 ppm and inoculum size of 3% presented the best conditions for the accumulation of chromium by L. plantarum MF042018. The study was also navigated to assess the biosorption capacity of L. plantarum MF042018, the maximum uptake capacity (q) of both Cd2+ and Pb2+ was recorded at pH 2.0 and a temperature of 22 °C after 1 hr. The biosorption process of Cd2+ and Pb2+ was well explained by the Langmuir isotherm model better than the Freundlich isotherm. Furthermore, the results revealed that the use of L. plantarum MF042018 is an effective tool for the treatment of hazardous metal-polluted battery-manufacturing effluent. Therefore, the present study implies that L. plantarum MF042018 can be applied as a promising biosorbent for the removal of heavy metals from industrial wasterwaters.

Bioaccumulation of Ni 2+ and cr 2+ by L. plantarum MF042018. The percentage removal of Ni 2+ and Cr 2+ by L. plantarum MF042018 was calculated after 24 hrs of incubation and the obtained results revealed that L. plantarum MF042018 was able to efficiently remove Ni 2+ and Cr 2+ from the broth medium by 33.8 ± 0.8% and 30.2 ± 0.5%, respectively (Data not shown). Moreover, TEM micrographs of L. plantarum MF042018 revealed obvious changes in the cell surface morphology of both Ni-and Cr-treated cells, compared to the control cells, with the appearance of dense metal deposits uniformly adsorbed to the outer cell surface and accumulated inside the cells (Fig. 1).
According to Wang 19 , there are two main mechanisms involved in bacterial resistance to Cr; biotransformation of Cr(VI) to Cr(III), that is less toxic, and biosorption, where bacteria trap the pollutant in its biomass. In addition, Monachese et al. 1 reported the capability of a number of species belonging to the genus Lactobacillus to bind metals, including Cr, and to detoxify them from different heavy metals contaminated geographical areas. Moreover, concerning EDX analysis, no Ni or Cr signals could be detected in the control samples ( Fig. 2A), whereas clear Ni and Cr peaks were observed at 7.5 and 5.6 keV, respectively, in Ni-and Cr-treated cells (Fig. 2B,C).
Statistical optimization of Cr 2+ removal by L. plantarum MF042018 using RSM. Plackett-Burman design was applied to allow reliable short listing of significant culture conditions affecting maximum metal removal. The results showed that, out of the 7 tested variables, the presence of high levels of both initial Cr 2+ concentration and inoculum size in the culture medium had significant positive effects on metal removal, with confidence levels above 95% (P < 0.05), whereas the other tested variables gave insignificant negative effects toward metal removal (Table 1, Fig. 3). Furthermore, the regression coefficient (R 2 ) value of the model was 0.998 that reflects an excellent fit of the model and is considered reliable as having a very high correlation 24 . Based on these results, the effect of initial Cr 2+ concentration and inoculum size on percent removal of chromium was performed by Box-Behnken design, with all other variables kept at the lowest levels, and a 3-D surface graph was constructed (Fig. 4). The figure shows that the maximum percentage removal of chromium (~50%) can be achieved with the increase in both initial metal concentration (100 ppm) and inoculum size of 3%. This result is in accordance with the results reported by dos Reis et al. 25 , where maximum accumulation of Fe 3+ by Bacillus subtilis UFLA SCF590 was attributed to the effect of lower pH value of 3.5 with high concentration of Fe 3+ (0.75 ppm). Furthermore, the results obtained by El-Naggar et al. 26 showed that the biosorption of Pb 2+ by the marine red alga, Gelidium amansii, reached its maximum with the increase in initial Pb 2+ concentration up to 200 mg/l, beyond which biosorption gradually decreased. On the contrary, Choińska-Pulita et al. 27 reported that the efficiency of Cd 2+ and Zn 2+ biosorption by Pseudomonas azotoformans JAW1 is inversely proportional to initial metal concentration, which could be explained by the insufficient metal binding sites on microbial biomass. Thus, increasing biomass concentration causes an increase in contact surface area, leading to higher heavy metal removal efficiency 28 . www.nature.com/scientificreports www.nature.com/scientificreports/ Biosorption of cadmium and lead by L. plantarum MF042018. Kirillova et al. 29 reported that studies on Pb and Cd are often conducted together, as both elements seem to react with bacterial species in similar ways. The potential binding capacity of L. plantarum MF042018 to remove Cd 2+ and Pb 2+ from aqueous solutions was evaluated under different reaction conditions. A contact time of 1 hr was recommended by Halttunen et al. 14 , wherein rapid binding phenomenon, due to high affinity of free metal ion binding sites, was observed across all studied Lactobacillus strains after 1 hr of incubation whereas prolonged exposure led to reduction in metal removal.

Intercept
The results revealed that both Cd 2+ and Pb 2+ removal efficiencies by L. plantarum MF042018 were concentration-dependent where maximum removal efficiencies (MRE) of Cd 2+ (0.18 ± 0.054 mg Cd h −1 mg −1 ) and Pb 2+ (0.07 ± 0.03 mg Pb h −1 mg −1 ) were recorded with initial metal concentrations of 50 ppm and 10 ppm, respectively (Fig. 5A). Whereas nearly negligible fractions of Cd and Pb were removed upon increasing metal concentration to 80 ppm, as similarly reported by Oves et al. 30 , where high metal concentrations lead to substantial decline in metal biosorption capacity due to the saturation of adsorption sites on the peptidoglycan layer and the lack of sufficient free binding sites. However, Samuel et al. 2 showed high adsorptive potential (92%) of Pb 2+ , from an initial concentration of 50 ppm, using magnetic chitosan/graphene oxide (MCGO) composite material within 7 hrs. Moreover, Vilela et al. 31 demonstrated the efficiency of graphene oxide-based microbots (GOx-microbots) for the removal of Pb 2+ from contaminated water by adsorption process and the subsequent reusability of GOx-microbots and recovery of lead. On the other hand, graphene oxide/fungal hyphae (GO-FH) bio-nanocomposite material exhibited excellent performance of adsorption and regeneration of chromium(VI) up to 212.76 mg/g 23 .
Furthermore, monitoring environmental physicochemical parameters; pH and temperature, can significantly influence microbial binding of heavy metals 32 . The results showed that the highest uptake capacities (q) of Cd 2+  www.nature.com/scientificreports www.nature.com/scientificreports/ (1.46 ± 0.022 mg Cd mg −1 ) and Pb 2+ (0.91 ± 0.020 mg Pb mg −1 ) by L. plantarum MF042018 were achieved at 22 °C and pH 2 ( Fig. 5B,C). Whereas, the removal of both Cd 2+ and Pb 2+ was almost negligible with increasing pH and/or temperature beyond pH 5 and/or 30 °C. The obtained results were in agreement with Sofu et al. 33 , where 20 °C was the optimum temperature for maximum removal of Fe(II) and Zn(II) by Lactobacillus delbrueckii ssp bulgaricus (Lb-12) and Streptococcus thermophillus (STM-7). In contrast, Halttunen et al. 34 reported significant increase in Cd removal by Lactobacillus rhamnosus GG when the incubation temperature was raised to 37 °C. Furthermore, Yi et al. 35 and Halttunen et al. 14 indicated that, for in vivo practical use, LAB perform high Pb removal capacity at low pH that could be due to a higher number of phosphate groups available on the www.nature.com/scientificreports www.nature.com/scientificreports/ bacterial surface. Similarly, Hansen et al. 36 revealed that acidic pH was optimum for maximum removal of As(V) by Lessonia nigrescens. However, Topcu and Bulat 37 reported that the lowest level of removal for both cadmium and lead by Enterococcus faecium is typically observed at pH below 3 wherein a linear increase in metal removal took place with an increase in pH.
Biosorption of Cd 2+ and Pb 2+ by L. plantarum MF042018 was verified by TEM and SEM examinations and EDX analysis as presented in Figs. 6 and 7. TEM micrographs clearly confirmed that both Cd 2+ and Pb 2+ formed visible deposits covering the cell surface reflecting high binding capacity (Fig. 6A,B). In addition, SEM micrographs revealed that exposure to Cd 2+ and Pb 2+ led to enormous aggregation of L. plantarum MF042018 cells (Fig. 6C,D) compared to the untreated cells, with no morphological changes. Furthermore, EDX analysis also  www.nature.com/scientificreports www.nature.com/scientificreports/ detected additional Cd and Pb peaks in both Cd-and Pb-treated cells (Fig. 7A,B) due to biosorption, whereas no peaks were observed in the control sample ( Fig. 2A).
Morphological alterations of L. plantarum MF042018 after Cd 2+ and Pb 2+ exposure, observed by TEM and SEM examinations, were in concordance with studies conducted by Halttunen et al. 38 who found large deposits of lead on the surface of lyophilized Bifidobacterium longum 46 and Lactobacillus fermentum ME3 after lead binding. Zhai et al. 15 mentioned that such phenomenon may be a form of self-protection by Lactobacillus plantarum CCFM8610 cells caused by the change in surface charge and the degeneration of surface proteins enhanced by Cd exposure and led to the aggregation of cells.
Cd and Pb biosorption isotherms and kinetics. The Langmuir and Freundlich adsorption isotherm parameters for Cd 2+ and Pb 2+ biosorption by L. plantarum MF042018 biomass were calculated and presented in Table 2. It appears that the biosorption process of both metal ions was more consistent with the Langmuir isotherm model than the Freundlich isotherm model. The regression coefficient (R 2 ) values of the Langmuir equation for both Cd 2+ (R 2 = 0.991) and Pb 2+ (R 2 = 0.957) conform better than those of the Freundlich equation. In addition, R L values for Cd 2+ (0.526) and Pb 2+ (0.613) are greater than "0" indicating that the sorption equilibrium is favorable with the Langmuir model as reported by Khan et al. 39 . These findings are in accordance with the results of Chakravarty and Banerjee 40 and Zhai et al. 15 that suggest the contribution of both electrostatic reaction and complex formation together with the high affinity of cadmium binding to the bacterial biomass.

Removal of heavy metals from battery-manufacturing effluent by L. plantarum MF042018.
The potency of L. plantarum MF042018 to facilitate the removal of toxic heavy metals from industrial effluents was examined. After 1 hr exposure of the effluent waste to L. plantarum MF042018 enclosed in a dialysis tube, the percentage removal of metal ions was calculated. The used cells achieved a 100% removal of Ni 2+ , Cr 2+ , Cd 2+ and Pb 2+ ions, indicating the successful ability of L. plantarum MF042018 to remove a mixture of heavy metals from the industrial effluent within 1 hr only. Furthermore, applying the microbial mass hanged in a dialysis tube is considered a significant criterion, to facilitate the collection and separation of cells from the treated industrial effluent for further use 41 .

Materials and Methods
Sample collection and isolation of marine LAB. A total of 20 sediment and seawater samples were aseptically collected from different stations along the Alexandrian Mediterranean Seacoast, Egypt, and kept at 4°C for further studies. Collected samples were used for isolation of LAB by the spread plate technique using de Man, Rogosa and Sharpe (MRS) agar media (Difco, USA) incubated anaerobically at 30 °C for 48 hrs 42 . Selected Gram positive isolates with catalase negative activity were preserved on MRS agar slants at 4°C for routine use or maintained in MRS broth supplemented with 30% glycerol at −80°C for long-term storage.

Screening for potential metal-resistant LAB isolates. Metal solutions, Ni(NO 3 ) 2 , Pb(NO 3 ) 2 ,
Cd(NO 3 ) 2 and Cr(NO 3 ) 2 (Sigma-Aldrich, USA), were sterilized by filtration through 0.45μm Millipore bacterial filters (Advantec, Tokyo, Japan). Metal-resistant profiles of LAB isolates were conducted on MRS agar plates amended with metals concentrations (up to 600 ppm) and incubated under anaerobic conditions at 30 °C for 1-3 days. The minimum inhibitory concentrations (MICs) were assessed by determining the lowest metal concentrations that completely inhibited LAB growth 39 . The higher-growth-exhibiting isolates were selected as potential metal-resistant LAB isolates for further studies.

Biochemical and molecular characterization of metal-resistant LAB isolates.
Potential metal-resistant LAB isolates were biochemically characterized by API 50 CHL test kits (BioMérieux, Marcy I´Etoile, France) according to the manufacturer's instructions and interpretation of results was performed using the computer-aided database API-WEB ™ V.5.0 software as mentioned by Khalil et al. 43 . Furthermore, selected isolates were identified by 16S rRNA gene sequencing, following Hamdan et al. 44 , analyzed in the Genbank DNA database using the online tool (BLAST) at NCBI (http://www.ncbi.nlm.nih.gov/BLAST/), and deposited in the DNA Data Base of Japan (DDBJ) for serial accession numbers.

Bioremediation of heavy metals by LAB isolates. Metal-resistant LAB isolates were cultured in MRS
broth and incubated at 30 °C for 24 hrs on a rotary shaker at 120 rpm. Triplicate sets of MRS broth with initial concentration of 500 ppm for Ni or 100 ppm for Cr were inoculated with 1% freshly prepared LAB isolates and incubated at 30 °C. After 24 hrs, cells were harvested by centrifugation at 6,000 rpm for 10 min to remove metal-bound bacterial cells and supernatants were analyzed by atomic absorption spectrophotometer (Shimadzu

Optimization of Cr 2+ removal by Response Surface Methodology (RSM). Plackett-Burman design
(PBD). Plackett-Burman experiment 45 was performed to select the most significant variable(s) that influence the process of Cr 2+ removal under metal stressed culture conditions. Seven independent variables were examined (evaluated) in two levels, low (−1) and high (+1) levels, including: initial pH (X 1 ), incubation temperature (X 2 ), rate of shaking (X 3 ), inoculum size (X 4 ), NaCl concentration (X 5 ), incubation time (X 6 ) and initial metal concentration (X 7 ) ( Table 3). The selected variables were organized in 8 different trials, performed in triplicate sets, and the effect of each variable on the percentage of metal removal was determined based on the following equation 27 : Where; "Y" is the response (Cr 2+ removal, %), "β 0 " is the model intercept and "β 1 " is the linear coefficient, and "X 1 " is the level of the independent variable.
Box-Behnken design. Based on the results of PBD, two significant variables; Cr 2+ concentration (ppm) and inoculum size (%), with high influence on Cr 2+ removal were selected for further optimization using Box-Behnken design 46 . In triplicate sets, each variable was tested in 15 experimental trials at three different levels; low (−1), middle (0) and high (+1), as shown in Table 4. The interaction effect between the percentages of Cr 2+ removal and the significant independent variables were estimated using the following second order polynomial equation 17 : Where; "Y" is the predicted response (Cr 2+ removal, %), "β 0 " is the regression coefficient, "β 1 " and "β 2 " are the linear coefficients, "β 12 " is the interaction coefficient, "β 11 " and "β 22 " are the quadratic coefficients, and "X 1 " and "X 2 " are the independent variables.
Metal biosorption assay. LAB was anaerobically incubated at 30 °C for 48 hrs in MRS broth and cells were harvested by centrifugation at 6,000 rpm for 10 min and washed three times with sterile ultra-pure Milli-Q (MQ) water. Biosorption experiments were performed following Halttunen et al. 14 with slight modifications. Briefly, cell pellets [30 mg ml −1 (wet weight)] were re-suspended in 5 ml sterile MQ water supplemented with various concentrations (10, 50 and 80 ppm) of Cd 2+ or Pb 2+ and incubated at 30 °C for 1 hr. Furthermore, temperature and pH effects on metal binding capacity of LAB isolates were assessed by adjusting the pH of the metal-amended solutions to pH 2.0, 5.0 and 7.0, using 1 N NaOH or HCl, and incubation temperatures were carried out at 22, 30 and 37 °C. Samples were collected, centrifuged and concentrations of residual, non-adsorbed, metals were measured in supernatants using atomic absorption spectrophotometer. LAB pellets suspended in metal-free MQ water were considered as negative controls, whereas positive controls of only Cd or Pb solutions were included in each experiment. All experiments were performed in triplicates and average values were used to calculate the metal uptake capacity (q, mg mg −1 ) 47 and the metal-removal efficiency (MRE, mg h −1 mg −1 ) 48 by LAB isolates as follows:

Independent Variables Codes
Experimental levels
Where; "t i " and "t f " are the initial and final contact time, respectively, and "M" is the cell biomass (mg).
Antibiotic sensitivity and resistance pattern of LAB isolates. Antibiotic susceptibility assay of metal-resistant LAB isolates was performed on MRS agar plates seeded with 1% overnight cultured LAB against 20 different commercial antibiotics; Neomycin (30 μg), Erythromycin (15 μg Where; "a" is the number of resistant antibiotics, and "b" is the total number of antibiotics used. Biosorption isotherm and kinetics of LAB isolates. The metal biosorption capacity at equilibrium (q e ) was calculated using the following equation 15 Where; "C e " is the metal concentration at equilibrium. Both the Langmuir and the Freundlich isotherm models were applied to the metal uptake data of the LAB isolates as described by Khan et al. 39 and Yi et al. 35 . The Langmuir isotherm model is represented using the following equation; Where; "q max " is the maximum adsorption capacity (mg/g), and "K L " is the Langmuir adsorption constant (l/mg).
In order to express the feasibility of the Langmuir isotherm model, a dimensionless separation factor (R L ), or equilibrium parameter, was calculated using the following equation; The Freundlich isotherm model is represented using the following equation; = q K C e F e n 1/ Where; "K F " is the Freundlich adsorption constant (l/mg), and "n" is the adsorption intensity.
Electron Microscopy and Energy Dispersive X-ray (EDX) analysis. In order to verify the adverse effects of metal stress on cell structure and morphology and investigate the possible localization of accumulated metals within LAB isolates, collected cell pellets of both metal-free (control) and metal-treated isolates were fixed with 2.5% (w/v) glutaraldehyde buffered in 0.1 mol l −1 phosphate buffer (pH 7.2). For Scanning Electron Microscopy (SEM; JEOL JSM 5400 LV, japan) and Transmission Electron Microscopy (TEM; JEOL JSM 5300, Japan), fixed samples were prepared as described by Khan et al. 39 and Zhai et al. 15 , respectively. Furthermore, detection of metal elements present in LAB cells was carried out using Energy Dispersive X-ray (EDX) (JEOL JSM 6360 LA, Japan).
Biosorption of heavy metals from battery-manufacturing effluent. Based on testing the capability of the selected LAB to remove toxic heavy metal ions from the prepared metal solutions, biosorption experiment was performed using an industrial waste effluent to confirm the capability to remove a mixture of heavy metals from such waste. Effluent samples were collected from the outlet pipes of a battery-manufacturing factory in Alexandria city, Egypt, and stored at 4 °C. The effluent was filter sterilized using 0.45μm Millipore bacterial filters and was initially adjusted to pH 5.0 using 1 N HCl. Cells of LAB were harvested by centrifugation, washed with sterile MQ water, confined in dialysis tubing and suspended into effluent samples at an inoculum level of 2% (w/v). Triplicate tubes were incubated at 22 °C for 1 hr and effluent tubes devoid of bacterial inoculum were used as controls. Residual metal ions concentrations were measured by atomic absorption spectrophotometer.
Statistical analysis of data. All experiments were carried out in triplicates and the results were expressed as the mean ± standard deviation (SD). The obtained data were subjected to one-way analysis of variance (ANOVA) followed by Student's t-test to estimate t-value, P-value and confidence levels and results were considered statistically significant when P < 0.05. All statistics were performed using Statistical Package for the Social Sciences (SPSS) program (Version 12.0, SPSS Inc., Chicago, IL). The three-dimensional (3-D) surface plots were generated using STATISTICA software (Version 5.0, StatSoft Inc., Tulsa, USA).

Conclusion
The preliminary obtained results revealed that Lactobacillus plantarum MF042018 has high tolerance against Ni 2+ and Cr 2+ with potential bioremediation capacity and antibiotic-resistant criteria. Moreover, efficient binding ability of Cd 2+ and Pb 2+ , from aqueous solutions was achieved at pH 2 and low incubation temperature (22 °C), and closest fit the Langmuir isotherm model. Based on these findings, further investigations should focus on the feasibility of Lactobacillus plantarum MF042018 biomass as an appropriate, inexpensive tool for detoxification of heavy metal-contaminated environments and/or foodstuffs that would be a reliable area for industrial-scale applications.