Electron Spin Resonance Study of Chromium (V) Formation and

Andronikashvili Institute of Physics, Georgian Academy of Sciences, Tbilisi 0177, Georgia, and Center for Environmental Biotechnology, Lawrence Berkel...
0 downloads 0 Views 212KB Size
Download PDF
Environ. Sci. Technol. 2003, 37, 4678-4684

Electron Spin Resonance Study of Chromium(V) Formation and Decomposition by Basalt-Inhabiting Bacteria TAMAZ L. KALABEGISHVILI,† N E L L Y Y . T S I B A K H A S H V I L I , * ,† A N D HOI-YING N. HOLMAN‡ Andronikashvili Institute of Physics, Georgian Academy of Sciences, Tbilisi 0177, Georgia, and Center for Environmental Biotechnology, Lawrence Berkeley National Laboratory, Berkeley, California 94720

Bacterial reduction of Cr(VI) to Cr(III) compounds may produce reactive intermediates Cr(V) and Cr(IV), which can affect the mobility and toxicity of chromium in environments. To address this important subject, we conducted an electron spin resonance (ESR) study to understand the kinetics of the formation and decomposition of Cr(V) during Cr(VI) reduction by different Gram-positive Cr(VI)-tolerant bacteria, which were isolated from polluted basalts from the United States of America and the Republic of Georgia. Results from our batch experiments show that during Cr(VI) reduction, the macromolecules at the cell wall of these bacteria could act as an electron donor to Cr(VI) to form a stable square-pyramidal Cr(V) complexes, which were reduced further probably via a one-electron transfer pathway to form Cr(IV) and Cr(III) compounds. The Cr(V) peak at the ESR spectrum possessed superhyperfine splitting characteristic of the Cr(V) complexes with diol-containing molecules. It appears that the kinetics of Cr(V) formation and decomposition depended on the bacterial growth phase and on the species. Both formation and decomposition of Cr(V) occurred more quickly when Cr(VI) was added at the exponential phase. In comparison with other Grampositive bacteria from the republic of Georgia, the formation and decomposition of Cr(V) in Arthrobacter species from the Unites States was significantly slower.

Introduction There is a continual influx of heavy metal contaminants into the environment from anthropogenic sources. One of the most common polluting metals is chromium. It has been discharged into the environment primarily from industries such as metal plating and alloying, leather tanning, and wood preservation. The two oxidation states of chromium commonly found in the environment are trivalent [Cr(III)] and hexavalent [Cr(VI)] chromium, which have widely contrasting toxicity and transport characteristics (1). Cr(VI) compounds are highly water soluble and toxic species, while most Cr(III) compounds are less water soluble and less harmful. The genotoxic and carcinogenic effects of Cr(VI) compounds are * Corresponding author phone: +995-32-396716; fax: +995-32536937; e-mail: [email protected]. † Georgian Academy of Sciences. ‡ Lawrence Berkeley National Laboratory. 4678

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 20, 2003

associated with their ability to enter cells rapidly and to be activated through intercellular reduction (2). Reactive intermediates (such as Cr(V), Cr(IV), radicals) generated upon activation of Cr(VI), can attack macromolecules and lead to DNA damage (2, 3). Since metal ions cannot be destroyed in environments, factors which influence the reduction of Cr(VI) can dictate the Cr toxicity to ecological receptors. Indigenous bacteria are found to play important roles in the environmental fate of toxic chromium with a multiplicity of mechanisms effecting transformations between the soluble and insoluble forms (4-6). These mechanisms are of great potential for both in situ and ex situ bioremedial treatment processes for solid and liquid wastes. To date, many of the studies focus on the roles of Gram-negative bacteria in remediating Cr(VI) polluted soils, sediments, and groundwater. Relatively few Gram-positive organisms have been examined for metal-reducing capabilities as possible contributors to bioremediation strategies. Most importantly, the generation of the reactive chromium intermediates during microbial reduction of Cr(VI) has not been investigated systematically. Cr(V) is a known intermediate product during the reduction of Cr(VI) by various biological reductants from mammalian cells (7, 8). Using low frequency ESR spectrometry the appearance of the Cr(V) species in algae and terrestrial plants has been also investigated (9, 10). However, there have been only a few studies that examined the formation of Cr(V) species in bacteria (11-13). Recently, Myers et al. (11) observed Cr(V) intermediate in Shewanella oneidensis MR-1 (Gram-negative bacterium), suggesting a one-electron transfer as the first step during the microbial reduction of Cr(VI) under anaerobic conditions. In this study we investigated systematically the formation and decomposition of Cr(V) compounds during the reduction of Cr(VI) by the Gram-positive endolithic (rock/mineral inhabiting) bacteria under aerobic conditions. Initially we focused on Arthrobacter oxydansswhich was isolated and cultivated from the Columbia basalt samples (in the United States of America). A. oxydans has been demonstrated to be a Cr(VI)-tolerant bacterium that can reduce Cr(VI) to Cr(III) with Cr(V) as its stable intermediate product in the synchrotron radiation-based (SR) Fourier transform infrared (FTIR) spectromicroscopy experiments (12, 13). Mechanisms associated with the Cr(VI) reduction by A. oxydans have been elucidated partly in our recent studies (14-16). Our ESR measurements of A. oxydans cells in the presence of Cr(VI) showed that Cr(V) compounds were intermediate products (15). The purpose of this study was to improve the current understanding of the fate of Cr(V) during bacterial reduction of Cr(VI) to Cr(III) compounds. Specifically, this study was designed to (i) establish the time-course of Cr(V) formation from Cr(VI) by A. oxydans at different phases of growth, (ii) estimate the kinetic parameters of Cr(V) formation and decomposition by A. oxydans, and (iii) compare the values with other Cr(VI)-reducing bacteria isolated from Georgia basalts.

Experimental Section Chemicals. All experimental chemicals were ACS-reagent grade and purchased from Sigma (St. Louis, MO). Bacteria and Treatment. The Gram-positive A. oxydans was isolated from polluted Columbia basalt rocks collected from 75 m below the ground surface of the Eastern Snake River Plain in the United States (12). Three additional Grampositive bacterial strains were isolated from basalt samples that were collected from the most polluted regions in the 10.1021/es0343510 CCC: $25.00

 2003 American Chemical Society Published on Web 09/10/2003

republic of Georgia (17). These bacterial strains were identified as Cr(VI) reducers by ESR method and spectrophotometrically by diphenilcarbazide. Bacterial cells were maintained as a batch culture in a nutrient medium described in ref 14: 2 g of K2HPO4, 0.01 g of FeSO4, 0.2 g of MgSO4‚7H2O, 1 g of C6H11O7N, 1 g of glucose, 1 g of yeast extract, and 1.0 L of distilled water. A. oxydans cells were grown aerobically in 250 mL Erlenmeyer flasks at 21 °C with constant shaking. Cr(VI) [as K2CrO4] was added to the nutrient medium in various growth-phases to provide the chromium concentration of 35 mg/L. The bacterial cells were harvested at various time points by centrifugation (10 000 rpm, 15 min, 4 °C) prior to analysis. Electron Spin Resonance (ESR) Experiments. ESR investigations were carried out on the RE 1306 radio spectrometer (Russia). Water in bacterial samples absorbs microwave radiation in resonator, thus decreases the sensitivity of ESR spectrometer. To overcome this obstacle, we froze the samples, the ice formed did not absorb the microwave radiation, and the sensitivity of ESR spectrometer remained high. Therefore, most of our ESR measurements of Cr(V) were carried out at liquid nitrogen temperature (77 K). We placed each 0.2-0.5 mL bacterial sample in a 4-5 mm diameter quartz tube and placed the tube in a Dewar finger with liquid nitrogen. The ESR spectra of the sample (in the quartz tube in liquid nitrogen) were recorded using a radio spectrometer, which included a Gauss meter for magnetic field calibration and a frequency counter. Some of our samples were stored at liquid nitrogen temperature for 2 years. The ESR line of these Cr(V) samples remained practically unchanged even after such a long period of time, indicating that Cr(V) is stable at 77 K for a long time. The typical settings for the registration of Cr(V) for the spectrometer were as follows: microwave frequency ) 9.3 GHz, modulation frequency ) 100 kHz, microwave power ) 12 mW, modulation amplitude ) 4 G, time constant ) 0.3 s, field set ) 3300 G, sweep width ) 100 G, scan time ) 3 min. To reveal the superhyperfine structure of Cr(V), we also analyzed a portion of the samples at the ambient temperature (300 K) by placing it in a 1 mm diameter quartz tube and setting the modulation amplitude to 0.4 G. The registration of the broad line for Cr(III) was complicated at low temperatures owing to the existence of oxygen impurity in liquid nitrogen which resulted in the displacement of zero line. To avoid this problem, we measured Cr(III) mainly at room temperature after drying the samples at 100 °C. The spectrometric settings for the registration of Cr(III)) spectrometer settings were as follows: microwave power ) 50 mW, modulation amplitude ) 8 G, field set ) 3000 G, sweep width ) 2000 G, scan time ) 20 min. To measure the value of the g-factor accurately, we took into account the difference in values of magnetic field caused by different positions of the samples and the sensor used for magnetic field measurement. We measured the value of the g-factor of the standard, obtained the small difference, and included them in the measurements of g-factor of the samples. The error in the value of g-factor was (0.0003 and (0.005 for Cr(V) and Cr(III), respectively. The ESR signal intensity was measured by the peak-to-peak height of the signal (the width of ESR line does not change and therefore the signal intensity is proportional to the concentration of paramagnetic centers).

Results and Discussion Cr(V) Formation and Decomposition by A. oxydans. ESR was the key technique employed in our study to understand the structures of Cr(V) complexes and their behavior in bacteria. Control experiments showed that no ESR signal typical of Cr(V) species was detected: neither in the pure A. oxydans cell culture, nor in the pure K2CrO4 solution and the

FIGURE 1. ESR spectrum of Cr(V) from A. oxydans recorded 20 min after adding 0.7 mM chromate solution to A. oxydans cells. ESR measurements are performed in a 4 mm diameter quartz tube at 77 K. Spectrometer settings: microwave frequency ) 9.3 GHz, modulation frequency ) 100 kHz, microwave power ) 12 mW, modulation amplitude ) 4 G, time constant ) 0.3 s, field set ) 3300 G, sweep width ) 100 G, scan time ) 3 min. [In some cases we registered the formation of Cr(V) after only 2 min of reaction time. It was technically difficult for us to register Cr(V) signal for less than 2 min.] pure culture growth medium. The ESR signal of Cr(V) was also not observed after the adding of pure K2CrO4 solution to the pure growth medium. The singlet ESR line with a g-factor of 1.980 and a width of 12 G appeared only after we mixed the chromate solution with A. oxydans cells (Figure 1). The Cr(V) ESR line registered in this study was similar with the line detected in S. oneidensis MR-1 (with a g-factor ) 1.98) (11). After mixing the A. oxydans cells with the chromate solution of different concentrations, the Cr(V) ESR line appeared immediately. To further the understanding of the production of Cr(V) in our experimental system, we performed an additional set of experiments. A. oxydans cells grown in the nutrient medium containing chromate solution were harvested at various time points by centrifugation. Then a portion of the harvested cells was washed three times with 0.15 M NaCl after centrifugation (washed cells). The presence of Cr(V) in the washed cells and the unwashed cells were analyzed by ESR spectroscopy. We found that, except for the samples of washed cells, similar ESR signals (Figure 1) corresponding to Cr(V) were detected within minutes in all other samples of unwashed cells. Most importantly, the g-factor detected for unwashed cells was 1.980 and not 1.9787 (a value of 1.9787 is a characteristic of Cr(V)-glutathione complex (18)). This implies that Cr(V) compounds were formed predominantly on the bacterial surface and not inside the cells (at least at the initial state). This conclusion is in agreement with our earlier data obtained by the method of microcalorimetry (15) and the method fluorescence microscopy (see below). The microcalorimetric measurements showed that the melting temperature of the DNA-protein (DNP) complex of A. oxydans was relatively constant under chromium action over a 10 days exposure time at a concentration of 35 mg/L chromate. This implied that chromate was not readily crossing the bacterial cell wall and thus produced no detectable DNA damage, such as strand breaks, DNA-protein cross-links, DNA-DNA cross links, CrDNA adducts, and base modifications in cells (2, 3). The identification of Cr(V) complexes by means of ESR spectral parameters (in solution) is complicated in heterogeneous systems because of the different tumbling rates compared to homogeneous species (2, 19). However, a recent analysis has shown that the measured ESR spectral parameters of Cr(V) complexes in solution can be used to elucidate structures associated with Cr(V) signals (20). Testa et al. (19) used these parameters to shed light on the mechanism of VOL. 37, NO. 20, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4679

Cr(VI) heterogeneous photocatalytic reduction and to obtain a clear evidence of an initial one-electron transfer to Cr(V). The g-factor of the Cr(V) intermediates formed during their photocatalytic reduction experiment was found to be very close to that for a freely tumbling in solution. A similar approach was applied to our analysis. A. oxydans is a Gram-positive bacterium (13). The cell wall of Gram-positive bacteria is a peptidoglycan macromolecule with attached accessory molecules such as teichoic acids, teichuronic acids, polyphosphates, carbohydrates, and proteins (21). Hydrophilic components of the accessory molecules directed outward can act as an electron donor to Cr(V) in a number of ways to form different Cr(V) complexes. According to the empirical relationship established by Barr-David et al., if Cr(V) complexes with ed [ed ) 1,2ethanedioloato(2-)], two ed ligands are bound to Cr(V). This yields a 1.980 g-factor, which corresponds to a [Cr(O)(ed)2]structure (20). The same g-factor has been assigned to Cr(V) complexes with the following donors: pin [pin ) pinacolato(2-)], to form [Cr(O)(pin)2]-; pd [pd ) 1,2-propanaldiolato(2-)], to form [Cr(O)(pd)2]-; and glyc ) glycerol [glycerol ) 1,2,3-propantriolato(2-)], to form [Cr(O)(glyc)2]-. Thus, our measurements imply that it is possible that biomolecules at the surface of A. oxydans which contain 1,2-diol and 1,2,3triol moieties are likely to act as an electron donor to Cr(V) to form a stable square-pyramidal complexes. Our measurements were carried out mainly at low temperature (77 K). In this case the ESR line of Cr(V)-diol complexes is broadened and the superhyperfine splitting characteristic of nearby hydrogen in Cr(V) complexes of diolcontaining molecules is not observed (22). Therefore, we decided to make measurements at room temperature in order to ascertain whether the ESR signal in our samples was due to the Cr(V)-diol complexes. This technique was utilized by Shi and Dalal (23) to characterize Cr(V) that complexes with biologically relevant diols. They reported that the main peak (with g ) 1.9792) in the Cr(V)-NADPH complex showed a multiplet having at least five principal components with a 0.84 spacing. Similar splittings have also been observed for other Cr(V)-diol complexes. The Cr(V) peak generated in living plants exhibited hyperfine splittings of about 0.79 G (9). For our ESR measurements at room temperature, we added 7 mM chromate solution to the A. oxydans cells free from growth medium. Such high concentrations of chromium could allow us to obtain the ESR signal of Cr(V) complexes (Figure 2). As it can be seen in Figure 2, the detected ESR line was strong enough that the super hyperfine splitting could be examined. On scale expansion, at the modulation amplitude of 0.4 G, a superhyperfine splitting of about 1 G could be resolved (Figure 2b). [The further decrease of modulation strongly increased the signal/noise ratio, and therefore we did not need to use it here.] Accordingly, the reduction of Cr(VI) in A. oxydans began with the formation of Cr(V)-diol complexes at the surface of bacteria. To evaluate this observation and to further the understanding of the real-time behavior of Cr(V) complexes, we added 35 mg/L of Cr(VI) to the growth medium at the mid-exponential, early stationary, and middle stationary phases of growth (the growth curve of A. oxydans is given in ref 15). The concentration of 35 mg/L was chosen for the following reasons. First, our earlier results showed that the protein composition of A. oxydans began to change at 35 mg/L (15). Second, by using atomic absorption spectrometry and epithermal neutron activation analysis methods, we have established that chromate accumulation is dose-dependent, and it was most intensive in the interval of Cr(VI) concentrations (10-50 mg/L) (14, 16). Finally, according to our earlier ESR measurements, at lower concentrations of Cr(VI) (up to 50 mg/L) the rate of Cr(V) formation was also maximal (16). 4680

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 20, 2003

FIGURE 2. ESR spectrum of Cr(V) from A. oxydans recorded 20 min after adding 7 mM chromate solution to A. oxydans cells. ESR measurements are performed in a 1 mm diameter quartz tube at 300 K. Spectrometer settings: microwave frequency ) 9.3 GHz, modulation frequency ) 100 kHz, microwave power ) 12 mW, time constant ) 0.3 s, field set ) 3300 G, sweep width ) 100 G, scan time ) 3 min. (a) modulation amplitude ) 0.6 G, (b) modulation amplitude ) 0.4 G.

FIGURE 3. Time-course of Cr(V) formation and decomposition in A. oxydans. A. oxydans cells were grown in the nutrient medium in which 35 mg/L of Cr(VI) was added in the middle exponential (a), early stationary (b), and middle stationary phases (c), respectively. After certain periods of time cells were harvested by centrifugation. ESR measurements were performed at 77 K in a 4 mm quartz tubes. The signal intensity was measured by the peak-to-peak height of the signal. Spectrometer settings are the same as in Figure 1. The solid lines represent the simulated results based on the kinetic model. Figure 3 is a time-course of the relative intensity of Cr(V) signal, which is in direct proportion to of the concentration of Cr(V) complexes. It appears that the behavior of Cr(V) complexes was affected by when Cr(VI) was introduced into the A. oxydans culture. Both the formation and decomposition of Cr(V) seem to occur more rapidly when Cr(VI) was added to the culture during its mid-exponential phase of growth. In this case the ESR signal of Cr(V) increased very rapidly and then disappeared within 50 h. Figure 3 also depicts that the formation of Cr(V) occurred faster than the decomposition of Cr(V), independent of the growth phase at which chromium was added. To quantify this behavior, we performed the kinetic analysis of this formation/decomposition

TABLE 1. Estimated Kinetic Parameters of Cr(V) Formation and Decomposition by Cr(VI)-Reducing Basalt-Inhabiting Bacteriaa Cr(VI) was added in half-time of Cr(V) formation, (tf )0.5 (min) bacteria

A. oxydans #1 #2 #3

half-time of Cr(V) decomposition, (td)0.5 (min)

exponent phase

early stationary phase

middle stationary phase

exponent phase

4.6 ( 0.8

11 ( 1.6 9.1 ( 0.6

15 ( 5.9 12 ( 2.2 5.4 ( 1.9

960 ( 180 6.2 ( 1.2 13.5 ( 1.2

11.6 ( 2.5

early stationary phase

middle stationary phase

44 ( 8.7 41 ( 8.3 1050 ( 320

490 ( 32 108 ( 35 -

a

(tf )0.5 was not calculated when the formation of Cr(V) proceeded very quickly; (td )0.5 was not calculated when the reduction of Cr(V) proceeded very slowly. A. oxydans was isolated from Columbia basalts (U.S.A.) and identified by T. Torok. Isolates #1 and #2 were isolated from Georgian basalts by M. Abuladze. Isolate #3 was isolated from Georgian basalts by D. Pataraya.

process. For this purpose we used an equation given in refs 9 and 10 which was applied to describe the Cr(V) formation and reduction in the circulating blood of rats and in the fronds of the duckweed Spirodela polyrhiza

I ) A[exp(- kdt) - exp(- kft)]

(1)

where I ) Cr(V) ESR signal intensity (arb. units), A ) the limiting concentration (intensity) of Cr(V) (which is obtained only after the formation of Cr(V), if their decomposition does not take place) (arb. units), kd ) first-order decomposition rate constant (min-1), kf ) first-order formation rate constant (min-1), and t ) time (min). Here kd and kf were calculated using the curve fitting (nonlinear Levenberg-Marquardt) method of the eq 1. The half-time of Cr(V) formation (tf)0.5 and decomposition (td)0.5 were calculated from (tf)0.5 ) 0.693/kf and (td)0.5 ) 0.693/kd, respectively. Equation 1 is true only when kf > > kd. The estimated kinetic values are presented in Table 1 (with the corresponding 95% confidence intervals). By comparing these values with the data shown in Figure 3, we suggest that the model simulation is in good agreement with the data points. The values of the kinetic parameters in the table indicate that the rate of Cr(V) formation kf was high for all cases, with a maximum value of kf when Cr(VI) was added to the A. oxydans at the middle exponential phase, which was about two times higher than those at other phases. On the contrary, the reduction of Cr(V) proceeded very slowly. Even when Cr(VI) was added to the exponential phase, the ratio of kf/kd was approximately 200. In the stationary phase the decomposition process proceeded significantly slower. Consequently we were not able to calculate kd using eq 1 during our experimental duration. When Cr(VI) was introduced into the exponential phase culture, the highest value of Cr(V) signal intensity was at least two times smaller than those at the stationary phase (Figure 3). This variation can be explained using fluorescent microphotographs of A. oxydans cells shown in Figure S1 in the Supporting Information. These microphotographs, taken by the method described in ref 24, demonstrate the typical changes in the morphology of cells at different phases of growth. They show that the life cycle of A. oxydans cells was characterized by the changing rod-cocci forms (25). At the mid-exponential phase, approximately one-fifth of the A. oxydans cells were in the form of cocci with an average diameter 2.5 times larger than that in the stationary phase (Figure S1). The remaining four-fifth of the cells was in the rod shape with lengths almost three times larger than the diameters of the cocci in the stationary phase. The ratio of the surface area of the cells in these two cases for the same masses yields Sexp/Sstat = 0.5 (the calculations were carried out with the assumption that densities of each cell differ only slightly in the stationary and exponential phases, so we consider them to be equal). This value is similar to the ratio of the ESR signal of Cr(V) for different growth phases, which

confirms that Cr(V) compounds were produced mainly on the surfaces of A. oxydans. According to these results, the reduction of Cr(VI) begins with rapid formation of Cr(V) in all stages of growth cycle of A. oxydans. This process occurred more quickly when Cr(VI) was added at the exponential phase. This corroborates with results from our earlier study of chromate reduction in A. oxydans, which revealed that Cr(VI) reduction rate was maximal when K2CrO4 solution was added to the culture either at the exponential (middle/late) phase (14). It also revealed that chromate reduction occurred initially at a high rate followed by a decrease in rate until chromate reduction ceased within 8 days. The presence of growth-phase dependent Cr(V) intermediates has also been detected in other microbial systems, for example, at the mid-exponential phase of the fumaratereducing S. oneidensis MR-1 (11). The prominent Cr(V) signal in S. oneidensis MR-1 was detected at the earliest time point, and this signal persisted over a 2 h time course at the end of which Cr(VI) reduction was not complete. It was also observed that the Cr(VI) reduction rate reached a maximum at the late exponential or early stationary phase (26, 27). The Cr(V) signal was absent when either Cr(VI) or formate or cytoplasmic membrane was excluded. This implies that the Cr(V) signal is due to the reduction of Cr(VI) by formatedependent electron transport components in the cytoplasmic membrane of S. oneidensis MR-1. A multicomponent electron transport mechanism could possible include cytochromes, quinones, and flavoproteins with iron-sulfur centers (11). Much previous research on microbial Cr(VI) reduction has focused on the quest for, and the characterization of, a terminal reductase responsible for Cr(VI) reduction. Cytochrome c3 has been proposed as an useful enzyme for bioremediation (28, 29). Lovley has proposed that in Desulfovibrio vulgaris Cr(VI) reduction is catalyzed by cytochrome c3 coupled to hydrogenase, with both proteins being localized in the periplasm of the bacteria (28). In contrast, the activities in Bacillus strain QC1-2 and Pseudomonas putida are associated with the cytosolic and soluble fractions, respectively (30, 31). In our work we have shown, that diolor triol-containing biomolecules on the cell surface of A. oxydans initiated the single-electron reduction to form at the first stable square-pyramidal Cr(V) complexes (g ) 1.980), which were reduced further probably via a one-electrontransfer pathway to form Cr(IV) and Cr(III) compounds. Complete identification of biomolecules taking part in an electron-transfer mechanism in our bacterial system needs further investigations. The following experiments were focused on the detection of the next oxidation states of chromiumsCr(IV) and Cr(III). ESR signal with a g-factor of 2.02 and a line width of 650 G was detected at 77 K during the decomposition of Cr(V) Figure 4a). At liquid nitrogen temperature the zero line displacement took place, but this ESR signal was easily detected at 300 K VOL. 37, NO. 20, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4681

FIGURE 4. ESR spectra of Cr(III) from A. oxydans cells recorded 6 days after growing in the nutrient medium containing 35 mg/L of Cr(VI). Chromate solution was added to the nutrient medium at the early stationary phase. Spectrometer settings: microwave frequency ) 9.3 GHz, modulation frequency ) 100 kHz, microwave power ) 50 mW, modulation amplitude ) 8 G, time constant ) 0.3 s, field set ) 3000 G, sweep width ) 2000 G, scan time ) 20 min. The dashed line indicates the zero line displacement at liquid nitrogen temperature. Measurements were performed at 77 K in (a) and at 300 K in (b). after drying of samples at 100 °C (Figure 4b). The observed line was consistent with the reported signal of Cr(III) complexes generated in Chinese hamstar V-79 cells (32). Cr(III) ions are characterized with broad ESR lines, and respectively a large amount of paramagnetic centers are necessary for its registration. For this reason, in bacterial cells, the prominent Cr(III) ESR signal was detected only after 1 day of exposure to Cr(VI) (not shown). We also observed Cr(V) and Cr(III) complexes simultaneously in the course of our experiments at 77 K. In some cases we observed broad signals with g-values changing in the range of 2.02 ÷ 1.99 (in some samples we observed a broad line with a g-factor 2.02, in some other samples it was 1.99 and so on in this range of 2.02 ÷ 1.99). The appearance of different ESR lines after decomposition of Cr(V) suggests the involvement of different mechanisms in the reduction of Cr(V) with the formation of Cr(III) and Cr(IV). The formation of various Cr(III) complexes by different cellular reductants is one of the possible mechanisms. The complexes of various types due to Cr(III) ions are well determined in different medium (32-35). The other mechanism contributing to the reduction of Cr(V) may include the generation of Cr(IV) complexes. Cr(IV) complexes can be formed with a wider variety of ligands (2, 36). The generation of Cr(IV) complexes with ox [ox ) oxalate (ethanedioate(2-)] and pic [pic ) picolinate (2pyridinecarboxylate(-)] ligands, which are relatively stable under biologically relevant conditions, was achieved by Codd et al. (37). Stable oxochromium(IV) tetraphenylporphyrin pyridine complexes were isolated and characterized by Jin et al. (38). The X-band ESR spectra of these complexes showed the line near g ) 2.005, within the range of g-factors measured by us. It seems that one electron transfer, with the formation of Cr(V) and Cr(IV) intermediates, is an additional possible pathway for Cr(VI) reduction to Cr(III). We suppose that these two types of mechanismssgeneration of Cr(III) and Cr(IV) complexessare not mutually exclusive. For an illustration of this hypothesis we computed the sum of Cr(III) and C(IV) signals. The obtained ESR line is presented in Figure 5. Here we used the ESR spectrum of Cr(III) in A. oxydans and the spectrum of the model Cr(IV) complex (g ) 1.98, line width ) 450 G). The model Cr(IV) complex was synthesized according to the method described by Liu et al. (39). It was found that the sum of Cr(III) and C(IV) spectra is similar to the ESR line detected by us experimentally. 4682

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 20, 2003

FIGURE 5. ESR spectra of Cr(III) (g ) 2. 02), Cr(IV) (g ) 1.98), and the sum of Cr(III) + Cr(IV) (g ) 2.004). Measurements of Cr(III) and Cr(IV) were performed at T ) 300 K Spectrometric settings are the same as in Figure 4b.

FIGURE 6. Time-course of Cr(V) formation and decomposition in isolate #1. Bacterial cells were grown in the nutrient medium in which 35 mg/L of Cr(VI) was added in the middle exponential (a), early stationary (b), and middle stationary phases (c), respectively. Spectrometer settings are the same as in Figure 1. The solid lines represent the simulated results based on the kinetic model. The formation of Cr(IV) complexes require further investigations in terms of their possible involvement in Cr(VI) induced processes. These experiments are underway. Cr(V) Formation and Decomposition by Other GramPositive Cr(VI)-Reducing Bacteria. In the second set of experiments the kinetics of Cr(V) formation by other Grampositive Cr(VI)-reducing bacteria isolated from polluted basalts in Georgia Republic was studied. Measurements were then employed to compare the processes of Cr(VI) reduction by A. oxydans with the Cr(VI) reduction by other Cr(VI)reducing bacteria. After exposing cells of bacteria #1 and #2 (in Table 1) to chromate solution we again observed the singlet ESR signal with g-factor ) 1.980 and width ) 12 G. As it was mentioned above, the similar Cr(V) ESR signal (g ) 1.98) was detected in Gram-negative bacterium S. oneidensis MR-1 as well. The real-time behavior of Cr(V) for both bacteria #1 and #2 from the Republic of Georgia revealed again the dependency on the growth stage at which chromium was added (Figures 6 and 7). Similarly, the formation and decomposition of Cr(V) occurred faster when Cr(VI) was added at the exponential phase. In contrast to A. oxydans, the formation of Cr(V) happened so quickly (less than 2 min) at both types of bacteria that one could not estimate the kf for bacteria #2 (Table 1). The rapid reduction of Cr(V) occurred in all growth phases. For example, when Cr(VI) was added in the mid-

FIGURE 7. Time-course of Cr(V) formation and decomposition in isolate #2. Bacterial cells were grown in the nutrient medium in which 35 mg/L of Cr(VI) was added in the middle exponential (a), early stationary (b), and middle stationary phases (c), respectively. Spectrometer settings are the same as in Figure 1. The solid lines represent the simulated results based on the kinetic model.

chromium reduction by bacterial isolates from Georgia Republic are underway. Our ESR studies have shown that A. oxydans from Columbia basalt and the isolates from basalt rocks in Georgian Republic can reduce Cr(VI) to Cr(III) through Cr(V) complexes. The Cr(V) peak from these measurements possess superhyperfine splitting that is thought to be characteristic of the Cr(V) complexes with diol-containing molecules. We also have shown in our batch experiments that the kinetics of Cr(V) formation and decomposition depends on the growth stage at which chromium was added. It appears that a sequential one- and two-electron transfer and/or a sequential one-electron transfer could be the pathway(s) for Cr(V) reduction from Cr(VI) by A. oxydans. These chromium intermediate products Cr(V) and Cr(IV) are known to play an important role in chromium-induced DNA damages; the formation of these compounds and their persistance in environments need to be evaluated carefully should one wish to employ intrinsic microorganisms as a means to remediate chromium polluted subsurface environments.

Acknowledgments This work was supported by International Science and Technology Center (ISTC) Grant G-348. We gratefully acknowledge Dr. M. Abuladze and Prof. D. Pataraya for providing bacterial samples and Dr. E. Namchevadze for fluorescent microscopy measurements. We thank Dr. I. Murusidze for helpful discussions and comments. We also thank Dr. M. Janjalia for her help during ESR experiments.

Supporting Information Available Figure 1S presents fluorescence microphotographs of A. oxydans: (a) cells representative of stationary phase and (b) cells representative of exponential phase. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited FIGURE 8. Time-course of Cr(V) formation and decomposition in isolate #3. Bacterial cells were grown in the nutrient medium in which 35 mg/L of Cr(VI) was added in the early stationary phase. Spectrometer settings are the same as in Figure 1. The solid lines represent the simulated results based on the kinetic model. exponential phase culture Cr(V) almost disappeared in less than 1 h for both isolates. These measurements show that the rate of reduction of Cr(VI) by A. oxydans is less than the rates of reduction by the other Gram-positive bacteria from Georgia Republic. In our laboratory it was established that the bacterial strains #1 and #2 did not belong to Arthrobacter genera (not shown). Because of this we also measured in addition the kinetics of Cr(V) formation and decomposition by isolate #3, which was preliminary assigned to the genus Arthrobacter from its growth properties and morphology (17). During this analysis chromate was added to the early stationary phase culture. ESR measurements showed that the kinetics of Cr(V) formation and decomposition associated with isolate #3 were similar to A. oxydans. Results are presented in Table 1 and Figure 8. We propose the slow reduction rate in A. oxydans may be a result of the special structure of the cell wall of Arthrobacter genera. Previous studies reported that the cell wall of Arthrobacter species contains a second bilayer of waxy lipids in the form of mycolic acids in addition to the thicker peptidoglycan layer, polysaccharides, menaquinones, teichoic acids, and proteins (21). The proposed hypothesis needs to be evaluated by further experiments. Detail investigations of the mechanisms of

(1) Nriagu, J. O. In Chromium in the Natural and Human Environments; Nriagu, J., Nieboer, E., Eds.; Wiley: New York, 1988; Vol. 20, p 81. (2) Codd, R.; Dillon, C. T.; Levina, A.; Lay, P. Coord. Chem. Rev. 2001, 216-217, 537-582. (3) Tsou, T.; Lai, H.; Yang, J. J. Chem. Res. Toxicol. 1999, 12, 10021009. (4) Bruins, M.; Kapil, S.; Oehme, F. Ecotoxicol. Environ. Saf. 2000, 45, 198-207. (5) McLean, J.; Beveridge, T. Appl. Environ. Microbiol. 2001, 67, 1076-1084. (6) Sani, R.; Peyton, B.; Smith, W.; Apel, W.; Peterson, J. Appl. Microbiol. Biotechnol. 2002, 60, 192-199. (7) Ueno, S.; Susa, N.; Furukawa, Y.; Sugiyama, U. Toxicol. Appl. Pharmacol. 1995, 51, 421-427. (8) Sakurai, H.; Takechi, K.; Tsuboi, H.; Yasui, H. J. Inorg. Biochem. 1999, 76, 71-80. (9) Liu, K. J.; Jiang, J.; Shi, X. Biochem. Biophys. Res. Commun. 1995, 206, 829-834. (10) Appenroth, K.; Bishoff, M.; Gabrys, H.; Stoeckel, J.; Swartz, H. M.; Walczack, T.; Winnefeld, K. J. Inorg. Biochem. 2000, 78, 235242 (11) Myers, C.; Carstens, B.; Antholine, W.; Myers, J. J. Appl. Microbiol. 2000, 88, 98-106. (12) Holman, H.-Y. N.; Perry, D.; Martin, M.; Lamble, G.; McKinney, W.; Hunter-Cevera, J. Geomicrobiol. J. 1999, 16, 307-324. (13) Holman, H.-Y. N.; Perry, D. L.; Martin, M. C.; McKiney, W. R. In Application of Synchrotron Radiation Techniques to Materials Science IV; Mini, S. M., Stock, S. R., Perry, D. L., Terminello, L. J., Eds.; Berkeley, 1998; Vol. 524, p 9. (14) Tsibakhashvili, N.; Abuladze, M.; Asatiani, N.; Bakradze, N.; Holman, H.-Y.; Kalabegishvili, T.; Kartvelishvili, T.; Rcheulishvili, A.; Sapojnikova, N.; Mosulishvili, L. In Protection and Restoration of the Environment VI; Kungolos, L., Liakopolus, A. S., Korfiatis, G. P., Eds.; Skiathos, Greece, 2002; Vol. 2, pp 755-762. VOL. 37, NO. 20, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4683

(15) Abuladze, M.; Asatiani, N.; Bakradze, N.; Holman, H.-Y.; Kalabegishvili, T.; Kartvelishvili, T.; Mosulishvili, L.; Rcheulishvili, A.; Sapojnikova, N.; Tsibakhashvili, N. Fresenius Environ. Bull. 2002, 11, 562-567. (16) Tsibakhashvili, N.; Mosulishvili, L.; Kalabegishvili, T. L.; Kirkesali, E.; Frontasyevs, M.; Pomyakushina, E.; Pavlov, S.; Holman, H.Y. J. Radioanal. Nucl. Chem. 2004, 259(1/2). (17) Tsibakhashvili, N.; Mosulishvili, L.; Kalabegishvili, T.; Pataraya, D.; Gurielidze, M.; Nadareishvili, G. Fresenius Environ. Bull. 2002, 11, 352-361. (18) Shi, X.; Dalal, N. Biochem. Biophys. Res. Commun. 1988, 156, 137-142. (19) Testa, J.; Grela, M.; Litter, M. Langmuir 2001, 17, 3515-3517. (20) Barr-David, G.; Charara, M.; Codd, R.; Farrell, R. P.; Irwin, J. A.; Lay, P. A.; Bramley, R.; Brumby, S.; Ji, J.-Y.; Hanson, G. R. J. Chem. Soc., Faraday Trans. 1995, 91, 1207-1216. (21) Salzman, G.; Holz, O. The Bacterial Cell Wall, 4th ed.; SpringerVerlag: Berlin, Heidelberg, New York, 2002; pp 211-217. (22) Malinovski, A.; Bontchev, P.; Kabassanov, K. Inorg. Chim. Acta 1974, 8, 17-23. (23) Shi, X.; Dalal, N. S. Arch. Biochem. Biophys. 1990, 277, 342-350. (24) Bakradze, N.; Abuladze, M.; Sokhadze, V.; Asatiani, N.; Sapojnikova, N.; Kartvelishvili, T.; Namchevadze, E.; Tsibakhashvili, N.; Tabatadze, L.; Lejava, L.; Holman, H.-Y. Sci. World J. 2003, 3, 432-442. (25) Tsibakhashvili, N.; Asatiani, N.; Abuladze, M.; Birkaya, B.; Sapojnikova, N.; Mosulishvili, L.; Holman, H.-Y. Biomed. Chromatogr. 2002, 16, 327-331. (26) Viamajala, S.; Peyton, B.; Apel, W.; Petersen, J. Biotechnol. Bioeng. 2002, 78, 770-778.

4684

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 20, 2003

(27) Viamajala, S.; Peyton, B.; Apel, W.; Petersen, J. Biotechnol. Progress 2002, 18, 290-295. (28) Lovley, D. Phillips, E. Appl. Environ. Microbiol. 1994, 60, 726728. (29) Michel, C.; Brugna, M.; Aubert, C.; Bernadac, A.; Brushi, M. Appl. Microbiol. Biotechnol. 2001, 55, 95-100. (30) Campos, J.; Martinez-Pacheco, M.; Cervantes, C. Antonie van Leewenhoek 1995, 68, 203-208. (31) Ishibashi, Y.; Cervantes, C.; Silver, S. Appl. Environ. Microbiol. 1990, 56, 2268-2270. (32) Sugiyama, M.; Tsuzuki, K. FEBS Lett. 1994, 341, 273-276. (33) Cozar, O.; Micu-Semeniuc, R.; David, L.; Haiduc, I. Modern Phys. Lett. B 2002, 16, 401-407. (34) Summerville, D. A.; Jones, R. D.; Hoffman, B. M.; Basolo, F. J. Am. Chem. Soc. 1977, 99, 8195-8201. (35) Glerup, J.; Weihe, H. Inorg. Chem. 1997, 36, 2816-2819. (36) Sugden, K. D.; Wetterhahn, K. E. J. Am. Chem. Soc. 1996, 118, 10811-10818. (37) Codd, R.; Lay, P. A.; Levina, A. Inorg. Chem. 1997, 36, 54405448. (38) Jin, R.; Cho, C. S.; Jiang, L. H.; Shim, S. C. Bull. Korean Chem. Soc. 1996, 17, 657-659. (39) Liu, K. J.; Shi, X.; Dalal, N. S. Biochem. Biophys. Res. Commun. 1997, 235, 54-58.

Received for review April 15, 2003. Revised manuscript received July 23, 2003. Accepted August 4, 2003. ES0343510