Tolerance to and Accumulation of Cadmium , Copper , and Zinc by Cupriavidus necator

Preliminary results of in vitro experiments with multicontaminated soils and solid media indicated that nodulating diazotrophic bacteria of the genus Cupriavidus are promising for the remediation of contaminated environments due to their symbiosis with legumes and metal tolerance. Thus, strains of Cupriavidus spp. (LMG 19424, UFLA 01-659, UFLA 01-663, and UFLA 02-71) were tested for their ability to tolerate and bioaccumulate cadmium (Cd), copper (Cu), and zinc (Zn) in Luria-Bertani broth. Changes in the growth pattern of Cupriavidus strains in the presence or absence of heavy metals were analyzed by scanning electron microscopy and metal allocation by transmission electron microscopy, to clarify the mechanisms of bioremediation. Highest tolerance was detected for strain UFLA 01-659 (minimum inhibitory concentration of 5, 4.95, and 14.66 mmol L of Cd, Cu, and Zn, respectively). Among the removal rates of the metals tested (9.0, 4.6, and 3.2 mg L of Cd, Cu, and Zn, respectively), the bacterial activity was clearly highest for Cd. The efficiency of strain UFLA 01-659 in removing the heavy metals is associated with its high biomass production and/or higher contents of heavy metals adsorbed and absorbed in the biomass. In response to the presence of heavy metals in the liquid culture medium, the bacteria produced exopolysaccharides and small and aggregated cells. However, these responses varied according to the strains and heavy metals. Regarding allocation, all heavy metals were adsorbed on the cell wall and membrane, whereas complexation was observed intracellularly and only for Cu and Zn. These results indicate the possibility of using C. necator UFLA 01-659 for remediation in areas with very high Cd, Cu, and Zn contents.


INTRODUCTION
In environments contaminated with heavy metals, diverse bacteria are found that are tolerant to high heavy metal concentrations in vitro and have a high capacity to remove these metals (Moreira et al., 2008;Chien et al., 2011;Andreazza et al., 2012).This ability of the bacteria is expressed through different mechanisms, such as efflux pumps, thione production, and enzyme activity (Zoropogui et al., 2008;Hynninen et al., 2009), varying according to the strains and metal evaluated.
Efflux pump activity is one of the most important mechanisms of heavy metal tolerance, allowing cells to extract metals taken up through ATP decomposition.Wang et al. (2015) sequenced the genome of Cupriavidus gilardii CR3 (bacteria with multiple metal resistance) and identified diverse operons that codify efflux pumps, such as czc (Cd 2+ , Zn 2+ ) and cus (Cu + , Ag + ) linked to heavy metal resistance.Behavioral variations in strains of the same genus were observed by Bianucci et al. (2011), who found that not all the four strains of Bradyrhizobium were able to accumulate Cd in the biomass, and that in strains with higher Cd tolerance and biomass contents, the glutathione (antioxidant) levels increased in response to the presence of the metal.
Biofilm formation, another mechanism linked to bioaccumulation, was investigated by Chien et al. (2013), who found that strain EJ01 (Pseudomonas sp.) had a greater ability of removing Cd and nickel (Ni) than the mutant strain m-3055, with poor capacity of biofilm formation.
In addition to the direct removal of heavy metals by adsorption and incorporation into plant biomass, some bacteria may contribute indirectly by their relationship with plants grown in contaminated environments.Strains of Pseudomonas inoculated on sunflower and corn plants, for example, reduced Cu toxicity and induced greater phytoextraction of this metal in both crops (Li and Ramakrishna, 2011).The greater heavy metal uptake of plants inoculated with Pseudomonas koreensis was attributed to solubilization of the metals in the rhizosphere (Babu et al., 2015).Heavy metal tolerance of diazotrophic bacteria was also reported in symbiosis with leguminous plants, where the bacteria can contribute to plant establishment in contaminated soils, favoring an increase of N content in these environments (Mahieu et al., 2011;Ferreira et al., 2013).The genus Cupriavidus has shown promising results in various studies on the contribution of nodulating diazotrophic bacteria to phytoextraction of toxic metals.For example, Klonowska et al. (2012) identified Cupriavidus taiwanensis isolates with high tolerance to Ni, Zi, and Cr, and high efficiency in biological N 2 fixation (BNF) in symbiosis with Mimosa pudica.In another study, inoculation with strain TJ208 (Cupriavidus taiwanensis) increased the biosorption efficiency of Mimosa pudica plants with regard to Cd, Cu, and lead (Pb) (Chen et al., 2008).
Given the promising results of the C. necator strains UFLA 01-659, UFLA 01-663, and UFLA 02-71, the objective of this study was to evaluate the heavy metal tolerance and removal capacity of these bacteria in liquid medium, with regard to Cd, Cu, and Zn, and to investigate the allocation of these metals by electron microscopy, deepening the understanding of mechanisms of heavy metal tolerance and removal by the bacteria.

Strains and inoculum preparation for the experiments in liquid medium
In this study, the strains UFLA 01-659, UFLA 01-663, and UFLA 02-71 of the species Cupriavidus necator (Silva et al., 2012) were used, as well as strain LMG 19424 T of the species Cupriavidus taiwanensis (Vandamme and Coenye, 2004), as a representative strain of the genus.All strains studied were preserved by lyophilization in the culture collection of the Soil Microbiology Laboratory of the Soil Science Department of UFLA.
To obtain the inoculum, the strains were cultured in Luria-Bertani (LB) agar containing: 10 g L -1 tryptone, 5 g L -1 yeast extract, and 5 g L -1 NaCl (Sambrook et al., 1989), for two days at 28 °C.Thereafter, the cells were suspended in sterile saline solution (0.85 % NaCl).The cell concentration was standardized to an optical density (OD) of 45 % transmittance, corresponding to a density of colony-forming units (CFU) of 6 × 10 8 CFU mL -1 .For the experiments of minimum inhibitory concentration and metal removal, a proportion of 1 mL of inoculum to 100 mL of liquid culture medium was adopted.
To avoid precipitation of the metals, 6.0 g L -1 TRIS was added to the culture medium (Mergeay et al., 1985).At the end of the incubation period, the colony-forming units of the liquid medium (CFU per mL) were quantified by the microdrop method (Herigstad et al., 2001).

Heavy metal removal
In this experiment, the strains were cultured in 80 mL LB broth under the same conditions of pH, inoculation, shaking, and incubation time as for MIC determination.The culture medium was supplemented with Cd at 1 mmol L -1 and Cu and Zn at 2 mmol L -1 , separately, corresponding to 112.41, 127.09, and 130.76 mg L -1 , respectively.These concentrations were established based on the results of the MIC experiments, adopting the highest possible metal concentrations at which all strains could be cultured.As a negative control, the strains were also cultured in a metal-free culture medium.All treatments were performed in triplicate.
After the period of incubation, the bacterial broth was centrifuged for 10 min at 10,000 rpm.The supernatant was discarded and the pellet washed twice with sodium phosphate buffer (8 mmol L -1 Na 2 HPO 4 .12H 2 O, 1.9 mmol L -1 NaH 2 PO 4 .2H 2 O, and 8 g NaCl with pH 7.3), as described by Moreira et al. (1993) and Pot et al. (1994).The resulting bacterial biomass was weighed and then subjected to nitric-perchloric acid (2:1) wet digestion.After that, the concentration of each based on this result, the rate of metal removal from the culture medium (mg L -1 ) and the metal concentrations in the biomass (mg g -1 ) were calculated.

Scanning electron microscopy
Samples of bacterial suspensions from the experiment of heavy metal removal were analyzed by scanning electron microscopy (SEM).Prior to sample preparation, 10 µL poly-L-lysine (0.1 %) was deposited and spread on glass cover slips to form a basis for bacteria adherence on their surface.After the poly-L-lysine had dried, 20 µL of bacterial suspension was placed on the cover slips and left to dry again.
The samples were prepared according to the protocol: fixation in Karnosvsky solution for 36 h, washing with cacodylate buffer (0.02 mol L -1 ), post-fixation with osmium tetroxide for 1 h, and dehydration in acetone gradient, followed by drying in a critical point apparatus.Gold metallization was performed to increase the electrical conductivity of the samples.Micrographs were taken with a scanning electron microscope LEO EVO 40 XPV.

Transmission electron microscope
An aliquot of 1 mL of bacterial broth (only of strain UFLA 01-659) was taken from the treatments of the metal removal experiments, centrifuged for 10 min at 10,000 rpm in a microcentrifuge, and the supernatant discarded.The pellet obtained was fixed in Karnovsky solution, centrifuged again, and polymerized with agarose (2 %) to facilitate preparation.Agarose cubes containing the pellet were removed, followed by washing in cacodylate buffer, and post-fixing in osmium tetroxide for 2 h and uranyl acetate (0.5 %) for 12 h.Dehydration in acetone gradient (25,30,40,50,60,70,75,80,90,95, and 100 %) followed by embedding in Spurr's resin.Ultra-thin sections were cut with an ultramicrotome, followed by contrast staining with uranyl acetate and lead citrate.Sections were examined with a Tecnai G2-12 transmission electron microscope (120 kV).

Statistical analyses
Software R was used for all statistical analyses.The MIC values were predicted by regression equations of the CFU based on increasing rates of each metal (p<0.05).The data regarding cell biomass, metal removal rate, and metal content in the biomass were evaluated in a (4 × 4) factorial arrangement; the first factor consisted of four bacterial strains and the second of the LB broth composition (control without metals, broth supplemented with Cd, Cu, and Zn).The means in the analysis of variance were compared by the Tukey test, at a significance degree of 5 %.

Minimum inhibitory concentration
For strain UFLA 01-659, greater tolerance to all metals was observed (MIC of 5.00 mmol L -1 Cd, 4.95 mmol L -1 Cu, and 14.66 mmol L -1 Zn) (Table 1).These values correspond to 562, 314, and 958 mg L -1 of the respective metals.A comparison of the metal tolerance ability of the strains studied shows a 2.5 times greater MIC of UFLA 01-659 for Cd than the other strains, while the MIC for Cu and Zn were 1.7 and 3 times greater, respectively.
The MIC for Zn (the most tolerated metal) differed considerably among the strains (in decreasing order: UFLA 01-659 > UFLA 02-71 = UFLA 01-663 > LMG 19424 T ).The MIC for Cd and Cu were quite similar among the three strains, with lower tolerance than for Zn.

Removal of heavy metals
With regard to the effect of metals on the strains, a higher biomass production of C. taiwanensis LMG 19424 T was observed in the medium cultured with zinc (p<0.05),exceeding the control treatment without addition of any metal (Figure 1a).A comparison of biomass production between the Cu and control treatments showed that the presence of the metal did not affect the biomass production of strain UFLA 01-659.For all strains, Cd was the metal most detrimental to biomass production.
Rev Bras Cienc Solo 2018;42:e0170080 Comparing the biomass production of the strains (Figure 1b), the strain UFLA 01-659 had a higher biomass production when Cd was added to the culture medium, whereas in the presence of Cu, strains UFLA 01-659 and UFLA 02-71 produced more biomass.When the strains were cultured with Zn, strain LMG 19424 T had the highest biomass production.
The metal removal rates of strain UFLA 01-659 were higher (9, 4.6, and 3.2 mg L -1 of Cd, Cu, and Zn, respectively) than those of the other strains (Figure 2a).In the treatments with Cu and Zn, strain UFLA 01-659 removed up to six times more than the others and for Cd, the removal was around four times greater.
In relation to the metal concentrations in the biomass (Figure 2b), in the Cd treatment, the highest metal concentrations in the biomass were found in the strains UFLA 01-659 and UFLA 01-663s, and in the treatment with Cu, in strains UFLA 01-659 and LMG 19424 T .In the Zn treatment however, high metal concentrations in the biomass were only found in strain UFLA 01-659.

Electron microscopy of Cupriavidus cells cultured with heavy metals
Scanning electron microscopy showed changes in cell growth patterns (size and shape) and in exopolysaccharide production in response to metal exposure (Figure 3 and 4).In the control treatment of strain LMG 19424 T , exopolysaccharide production was not observed (Figure 3a).However, when cultured with Cd (Figure 3b) or Cu (Figure 3c), this strain maintained the same growth pattern, however produced exopolysaccharide.When cultured with zinc, it had small and aggregated cells (Figure 3d).
No visible modifications in C. necator strain UFLA 01-659 were detected by SEM at the metal concentrations tested (Figures 3e,3f,3g,and 3h).In all treatments, strain UFLA 01-663 showed characteristic biomass production.In the control treatment (Figure 4a) and in the presence of Cu (Figure 4c), exopolysaccharide production was observed.However, this effect was absent in the presence of Cd (Figure 4b) or of Zn (Figure 4d).In the presence of Cd (Figure 4f) and Cu (Figure 4g), the C. necator UFLA 02-71 samples had similar growth patterns and visibly greater exopolysaccharide production than in the control treatment (Figure 4e).In the presence of Zn, this strain maintained the same cell shape, though with high exopolysaccharide production (Figure 4h).
In the images taken with the transmission electron microscope (TEM), an accumulation of all metals can be seen (Figure 5b, 5c, and 5d), both in the cytoplasmic membrane and the cell wall.The metals Cu (Figure 5c) and Zn (Figure 5d) were also accumulated within the cells of strain UFLA 01-659.

DISCUSSION
The literature reports values for MIC, metal removal rate, and metal concentration in the bacterial biomass for heavy metal-resistant strains.the MIC values registered for Cd were 0.75 -3 mmol L -1 for Pseudomonas, Enterobacter sp., Acinetobacter sp., and Cupriavidus metallidurans CH34; the last mentioned is considered highly tolerant (Chien et al., 2008;Chien et al., 2011;Klonowska et al., 2012).With regard to accumulation in biomass, Bianucci et al. (2011), studied Bradyrhizobium strains and observed Cd concentrations in the biomass of 7.7 mg g -1 .
The C. necator strain UFLA 01-659 in this study had MIC values of approximately 5.00 mmol L -1 Cd in liquid medium, a removal rate of 9.00 mg L -1 Cd in culture medium, and a high concentration of 16.3 mg g -1 of the metal in the biomass.These values are higher than those previously reported in the literature.In relation to Cu, MIC values of 5.00 mmol L -1 for Pseudomonas, Cupriavidus taiwanensis, and endophytic bacteria, and concentrations of 0.838 mg g -1 in the biomass of Pseudomonas sp. are registered (Chen et al., 2008;Li and Ramakrishna, 2011;Luo et al., 2011).In our studies, we registered a MIC value similar to that already reported (5.00 mmol L -1 ); however, the indices of removal rates of 4.6 mg L -1 and Cu content in the biomass of 4.4 mg g -1 in strain UFLA 01-659 were considerable.
Strains of Pseudomonas were isolated from lake sediments by Li and Ramakrishna (2011), with MIC values of 6 mmol L -1 Zn in solid medium and a content of 15.877 mg g -1 Zn in the biomass.For some Pseudomonas isolates, Klonowska et al. (2012) obtained MIC values of 15 mmol L -1 Zn, while in the same study, strain LMG 19424 T exhibited a high MIC index (10 mmol L -1 ), much higher than the values found in this study (3.59 mmol L -1 ).This low tolerance in our experiments may be due to differences in composition of the medium because we used LB broth supplemented with TRIS (buffer), which prevents precipitation of the compound by changing the pH and, consequently, reducing its bioavailability.In the study of Klonowska et al. (2012) however, YM medium was used, without buffer.In our study, the tolerance of strain UFLA 01-659 to Zn was also high (14.66mmol L -1 ), while the indices of removal (3.24 mg L -1 ) and content (4.1 mg g -1 ) of Zn in the biomass were less marked.
The superiority of strain UFLA 01-659 in tolerating the metals studied was also noted in the images taken by in SEM (Figure 3), in which no changes were observed in the development pattern (cell size and exopolysaccharide production) in a comparison of the cultures with metals and the control.Their high efficiency in removing metals from the culture medium was expressed either through higher biomass production (observed in the Cd and Cu cultures), or by higher metal content in the biomass (when cultured with Zn) (Figures 1 and 2).According to Babu et al. (2015), the efficiency in heavy metal removal from the culture medium is due to the ability of the strains in increasing the cell density and to the saturation of metal adsorption sites on the cell surface.
Increased exopolysaccharide production by bacterial strains with the presence and increase in Cd and Ni concentrations in culture medium was reported by Chien et al. (2013).Nevertheless, no difference in exopolysaccharide production was observed between the strains with the highest and lowest metal removal rates.This confirmed that greater removal may be linked to the composition of these exopolysaccharides and their capacity of metal adsorption.
In our study, the presence of metals in the culture medium induced exopolysaccharide production in the strains LMG 19 424 T and UFLA 02-71 (Figure 3 and 4); however, these strains had a lower removal rate.In relation to tolerance mechanisms, Babu et al. (2015) confirmed that multimetal complexes (As, Cd, Cu, Pb, and Zn) were found on the surface of and outside the cells, and attributed an important role to exopolysaccharide   production with regard to extracellular metal complexation.However, in experiments with isotherms for Zn adsorption by Pseudomonas aureofaciens biomass, greater metal adsorption by the strain with low exopolysaccharide production was observed as of the concentration of 0.40-0.45mmol L -1 Zn in equilibrium solution (Drozdova et al., 2014).Considering the cited studies and what we found in the images taken by SEM, it may be said that exopolysaccharide production is a response to the presence of metals, but does not always contribute to heavy metal removal from the environment.
The presence of Cd associated with bacterial cells of Stenotrophomonas maltophilia by SEM coupled with energy-dispersive X-ray spectroscopy (EDX) was detected by Pages et al. (2008); however, an exact determination of metal allocation was not possible by this technique.For these authors, the presence of Cd redirects the bacterial metabolism to cysteine production for later use as a precursor in formation of cadmium sulfate particles (CdS).For Wang et al. (2015), the high resistance of Cupriavidus gilardii CR3 to various metals is mainly due to the efficient system of ion efflux and metal complexation and reduction, and indirectly, to its self-repair ability.
Thus, it may be inferred that metal allocation is determined by the tolerance mechanisms of each bacterial strain.Allocation of Cu and Zn within the cells of UFLA 01-659 may indicate that the strain has low capacity of metal regulation by ion efflux pumps and/or that it complexes metals, forming particles intracellularly.The presence of Cd only in the cytoplasmic membrane and in the cell wall may indicate that there is efficient extraction of this metal and that its high removal rate is mainly due to physical phenomena of adsorption.

CONCLUSIONS
The greatest capacity of tolerating and removing Cd was detected in C. necator strain UFLA 01-659, due to the high biomass production and adsorption to the cell membrane and wall.
Efficiency in Cu and Zn removal by C. necator UFLA 01-659 is expressed by its ability of intracellular complexation and adsorption to the cell membrane and wall.
The presence of metals in the culture medium induces exopolysaccharide production of the strains LMG 19424 T and UFLA 02-71, which contribute little to metal removal.

Figure 1 .
Figure 1.Biomass of bacterial strains (LMG 19424 T , UFLA 01-659, UFLA 01-663, and UFLA 02-71) cultured in the absence of metals (control) and in the presence of the metals Cd (1 mmol L 1 ) and Cu and Zn (2 mmol L -1 ) in liquid LB culture medium.Mean values followed by the same letter do not differ from each other by the Tukey test (p<0.05).

Figure 3 .
Figure 3. Scanning electron micrograph of the microbial biomass of the strains LMG 19424 T (a, b, c, and d) and UFLA 01-659 (e, f, g, and h) cultured in liquid LB medium in the following treatments: control -absence of metals (a and e), Cd at 1 mmol L -1 (b and f), Cu at 2 mmol L -1 (c and g), and Zn at 2 mmol L -1 (d and h).

Figure 4 .
Figure 4. Scanning electron micrograph of the biomass of the strains UFLA 01-663 (a, b, c,and d) and UFLA 02-71 (e, f, g, and h) cultured in liquid LB medium in the following treatments: control -absence of metals (a and e), Cd at 1 mmol L -1 (b and f), Cu at 2 mmol L -1 (c and g), and Zn at 2 mmol L -1 (d and h).

Figure 5 .
Figure 5. Transmission electron micrograph of the C. necator strain UFLA 01-659 cultured in liquid LB medium in the following treatments: control -absence of metals (a), Cd at 1 mmol L -1 (b), Cu at 2 mmol L -1 (c), and Zn at 2 mmol L -1 (d).Arrows indicate the locations of clearest contrasts of metal accumulation.

Table 1 .
Minimum inhibitory concentration (MIC) of Cd, Cu, and Zn for the growth of different bacterial strains of the genus Cupriavidus spp. in liquid LB culture medium, calculated by quadratic equations * and **: significant at the level of 5 and 1 %, respectively.