Cell Wall Reactivity of Acidophilic and Alkaliphilic Bacteria

In this study, we used potentiometric titrations and Cd adsorption experiments to determine the binding capacities of two acidophilic (A. cryptum and ...
0 downloads 0 Views 2MB Size
ARTICLE pubs.acs.org/est

Cell Wall Reactivity of Acidophilic and Alkaliphilic Bacteria Determined by Potentiometric Titrations and Cd Adsorption Experiments Janice P. L. Kenney*,†,‡ and Jeremy B. Fein† †

Department of Civil Engineering and Geological Sciences, University of Notre Dame, Notre Dame, Indiana 46556, United States

bS Supporting Information ABSTRACT: In this study, we used potentiometric titrations and Cd adsorption experiments to determine the binding capacities of two acidophilic (A. cryptum and A. acidophilum) and two alkaliphilic (B. pseudofirmus and B. circulans) bacterial species in order to determine if any consistent trends could be observed relating bacterial growth environment to proton and Cd binding properties and to compare those binding behaviors to those of neutrophilic bacteria. All of the bacterial species studied exhibited significant proton buffering over the pH range in this study, with the alkaliphiles exhibiting significantly higher acidity constants than the acidophiles as well as the neutrophilic bacterial consortia. The calculated average site concentrations for each of the bacteria in this study are within 2σ experimental error of each other, with the exception of A. cryptum, which has a significantly higher Site 2 concentration than the other species. Despite differing acidity constants between the acidophiles and alkaliphiles, all bacteria except A. cryptum exhibited remarkably similar Cd adsorption behavior to each other, and the observed extent of adsorption was also similar to that predicted from a generalized model derived using neutrophilic bacterial consortia. This study demonstrates that bacteria that grow under extreme conditions exhibit similar proton and metal adsorption behavior to that of previously studied neutrophilic species and that a single set of proton and metal binding constants can be used to model the behavior of bacterial adsorption under a wide range of environmental conditions.

’ INTRODUCTION The speciation and mobility of metals in the environment can be influenced by many factors, including adsorption onto bacterial cell wall functional groups (e.g., refs 17). A number of studies have documented similarities between bacterial species in terms of their deprotonation and metal uptake behaviors (e.g., refs 5 and 812). Yee and Fein9 investigated proton and metal adsorption onto both Gram-positive and Gram-negative neutrophilic bacteria and found that the differences in cell wall structure do not affect the adsorption behavior. Yee and Fein9 proposed that a wide range of bacteria exhibit similar proton and metal binding behaviors and that this behavior can be modeled with a single set of stability constants for the important proton- and metal-bacterial site complexes. Borrok et al.11 tested the universal adsorption behavior proposed by Yee and Fein9 by examining proton and Cd binding to 8 different consortia of bacteria, observing similar adsorption behavior for all of the consortia studied, and using a single set of stability constants to model the behavior. Johnson et al.12 expanded on the work of Borrok et al.11 by examining a much larger range of bacterial consortia and measuring the extent of adsorption of Ca, Cu, Pb, Sr, and Zn in addition to that of Cd. Johnson et al.,12 like Borrok et al.,11 found r 2011 American Chemical Society

that the calculated stability constants for proton- and metalbacterial complexes do not vary significantly between consortia and that a wide range of bacteria display similar extents of adsorption as a function of pH for experiments conducted at the same bacterial concentrations and ionic strength conditions. Most consortia and monoculture experiments that have measured proton and metal adsorption onto bacteria have involved bacterial species that grow under similar environmental conditions, which do not necessarily reflect the diversity of natural ecosystems. There are some indications that not all microorganisms exhibit similar adsorption behavior. Borrok et al. 13 examined proton and Cd adsorption behavior of bacterial consortia grown from contaminated environments and found that some of these consortia exhibit significantly enhanced adsorption capacities compared to those grown from uncontaminated sites. Ginn and Fein14 studied the binding properties of bacteria with a wide range of genetic diversity and environmental growth conditions. Received: January 17, 2011 Accepted: April 8, 2011 Revised: March 28, 2011 Published: April 18, 2011 4446

dx.doi.org/10.1021/es200191u | Environ. Sci. Technol. 2011, 45, 4446–4452

Environmental Science & Technology Among the bacteria examined were a thermophile, a psychrophile, and an acidophile. All of the bacteria studied by Ginn and Fein14 displayed similar extents of proton and Cd adsorption, with the exception of the acidophile, Acidophilium angustum, which exhibited significantly lower proton buffering and metal adsorption capacities than the other species. Daughney et al.15 also examined an archaea, Thermococcus zilligii, that exhibited an order of magnitude lower site concentrations and significantly less Cd adsorption capabilities than most bacteria.13,9 In contrast, the acidophile Thiobacillus ferrooxidans exhibits buffering capacity across a wide pH range, more similar to other bacterial species than to A. angustum.16 The buffering capacities of other acidophilic bacterial species need to be studied before a general pattern, if any exists, can be discerned. While it is clear that not all bacteria exhibit similar adsorption behaviors, it is unclear whether bacterial growth environment affects adsorption behavior. The objective of this study is to test whether bacteria that grow under some of the most extreme conditions, for example extremely low pH and extremely high pH, exhibit similar or different proton and Cd adsorption behavior compared to neutrophilic bacteria. We compare our results to those from other studies that have been conducted with the same concentrations of bacteria at the same buffered ionic strength. We interpret the data using a surface complexation modeling approach so that the calculated acidity constants, site concentrations, and stability constants of the important Cd-bacterial surface complexes can be directly compared to those determined using the same approach for a range of bacterial species and consortia already studied.

’ MATERIALS AND METHODS Potentiometric Titrations. Experiments were conducted using washed inactive cells of the Gram-positive bacterial species Bacillus pseudofirmus and Bacillus circulans and the Gram-negative species Acidiphilium cryptum and Acidiphilium acidophilum. All the bacteria in this study were harvested at stationary growth stage. The growth and washing conditions are described in detail in the Supporting Information. Potentiometric titrations were conducted using 3880 g/L (wet mass) suspensions of cells. The ionic strength of the suspensions was buffered using 0.1 M NaClO4 and conducted under a N2 atmosphere. The electrolyte was bubbled with N2 for 1 h prior to suspension, in order to eliminate atmospheric CO2. Titrations were carried out a minimum of three times, each using separate batches of bacterial cultures, using an automated buret assembly. Blank titrations were performed for machine calibration and comparison, using bacteria-free 0.1 M NaClO4. The initial pH of each suspension of acidophilic bacteria was approximately 3. These suspensions were further acidified to pH 2.5 using 1.048 N HCl. Aliquots of 1.030 N NaOH were used to increase the pH of the suspensions to approximately pH 10, and the volume of base added and corresponding pH change were recorded at each titration step. Reverse ‘down-pH’ titrations were also conducted, decreasing the suspension pH to 3 using 1.048 N HCl in order to test the reversibility of proton binding on the cells. The initial pH of each suspension of alkaliphilc bacteria was approximately 10. Prior to the start of the titrations, the pH of these suspensions was increased to pH 11 using 1.030 N NaOH. The titrations of the suspensions of alkaliphilic bacteria involved decreasing the pH to 3.5 using 1.048 N HCl, again recording the volume of acid added and corresponding pH change at each titration step. Reversibility

ARTICLE

was then tested by conducting an ‘up-pH’ titration, increasing the pH back to 11 using 1.030 N NaOH. Due to titrant additions, the ionic strength of our suspensions varied from 0.100 to 0.097 M NaClO4. Each individual suspension was stirred throughout the titration with a magnetic stir bar. Cadmium Adsorption Experiments. Batch Cd adsorption experiments were conducted by measuring the change in aqueous Cd concentration that occurred upon exposure of an aqueous Cd solution to the washed acidophilic or alkaliphilic cells. For each set of experiments with a bacterial species, a parent solution containing approximately 10 ppm Cd(II) dissolved in 0.1 M NaClO4 was prepared from a 1000 ppm Cd standard reference solution. The pH of the Cd-bearing parent solution was adjusted to pH 3 for the experiments with acidophiles and to pH 10 for the experiments with alkaliphiles using aliquots of 0.1 to 1 M HCl or NaOH. A pellet of washed cells of one of the four bacterial species studied, whose wet mass was determined, was suspended in the 10 ppm Cd-bearing solution to achieve a bacterial concentration of 10 g (wet mass)/L in each experiment. The batch adsorption experiments were conducted twice for each bacterial species studied, using two separate growths of cells. The parent suspensions were divided into 10 mL volumes in 15 mL polypropylene test tubes, and the solution pH of each suspension was adjusted to a desired starting pH, ranging from pH values of 2 to 10, using small aliquots of 0.1 to 1 M HCl or NaOH. The systems were allowed to equilibrate via end-overend rotation at 24 rpm for 2 h to allow time for equilibration between the Cd and the cells, after which the final pH was measured, the tubes were centrifuged at 8200 g, and the supernatant was filtered through a 0.45 μm disposable nylon filter to remove cells with adsorbed Cd from solution. The final concentration of dissolved Cd was determined using inductively coupled plasma-optical emission spectroscopy (ICP-OES) with matrix-matched standards, and the concentration of Cd that was adsorbed to the cells was calculated by difference between the initial and final Cd concentrations. Control experiments were conducted without bacteria to determine if Cd was lost to the experimental apparatus, and loss was determined to be negligible. The analytical uncertainties for the ICP-OES results were determined to be approximately (3% by repeat analysis of standards.

’ RESULTS AND DISCUSSION Potentiometric Titrations. The titration data, shown in Figure 1, are plotted as a buffering capacity, calculated as (Ca  Cb  [Hþ] þ [OH])/mb, where Ca and Cb are the total concentrations of acid and base added at each step of the titration, respectively (including initial amounts of acid or base added to the suspension prior to the titration); the brackets represent molar species concentrations, and mb is the wet mass concentration (g/L) of the bacterial suspension. All of the species studied exhibit significant and continuous proton buffering capacity over the entire pH range considered. The shapes of the titration curves for all of the species studied here are similar, although the pH range of the titrations for the acidophiles was lower than that for the alkaliphiles, and the low pH end of the acidiphile titrations tended to exhibit steeper buffering capacity (Figure 1). The differences in the chemical composition of the media used to grow the bacteria prior to these experiments may be responsible for some of the differences that we observe between acidophiles and alkaliphiles. Because of the radically different conditions required for growth of acidophiles and alkaliphiles, 4447

dx.doi.org/10.1021/es200191u |Environ. Sci. Technol. 2011, 45, 4446–4452

Environmental Science & Technology

ARTICLE

Figure 1. Potentiometric titration data for a) A. acidophilium, b) A. crytpum, c) B. curculans, and d) B. pseudofirmus (3880 g/L wet mass) biomass in 0.1 M NaClO4.

the effects of growth medium on these properties cannot be controlled. The average buffering capacities for the species studied here, which are compared in Figure 1, are similar over the pH range of 3.5 to 9.0; A. acidophilum, A. cryptum, B. circulans, and B. pseudofirmus yield a difference in the average value of (Ca  Cb  [Hþ] þ [OH])/mb (which we define as buffering capacity) over this pH range of 1.27 ( 0.20  104, 2.12 ( 0.16  104, 1.49 ( 0.29  104, and 2.23 ( 0.57  104 mol/g, respectively (reported errors represent 1σ uncertainties). These values are within a factor of 2 of the buffering capacities over the same pH range of neutrophilic bacterial consortia (2.7  104 mol/g; 12) and of B. subtilis (3.0  104 mol/g; 17). The buffering capacities of the acidophiles in this study are an order of magnitude greater than that observed for A. angustum over this same pH range (2.5  105 mol/g) by Ginn and Fein14 and are similar to the buffering behavior observed for T. ferrooxidans by Naja et al.16 Figure S1 depicts representative examples of titration reversibility for one of the acidophiles studied and for one of the alkaliphiles. Although some differences exist between the up-pH and down-pH titration curves, the differences in general are smaller than the differences between titration curves that we observed from one replicate to another, and we conclude that within the time scale of these experiments the protonation reactions are fully reversible and that proton adsorption equilibrium was attained. Titration data sets that measure bacterial cell wall protonation behavior have been interpreted using a wide range of modeling approaches,1822 and a range of models can yield equally good fits to bacterial titration data.17 The potentiometric titration data in our study were modeled using a nonelectrostatic, discrete site

surface complexation model (e.g., refs 17 and 19) to determine the proton binding constants and site densities for each bacterial species studied. This model was chosen to enable comparison with previously published site densities and proton binding constants from studies that also used the nonelectrostatic modeling approach, and because we did not collect data as a function of ionic strength, which is required to calibrate surface electric field models. In this approach, deprotonation of each of the functional group types on the bacterial cell wall is represented by the following reaction R  SiteðXÞ  H o S R  SiteðXÞ þ H þ

ð1Þ

where R represents the bacterial cell wall macromolecule to which each functional group type, Site(X), is attached, and in this approach each site type has its own distinct acidity constant and site concentration. The equilibrium constant for Reaction 1, or Ka, quantifies the relative activities of the protonated and deprotonated functional groups of a type of site, and can be expressed as follows Ka ¼

½R  SiteðXÞ  3 aH þ ½R  SiteðXÞ  H o 

ð2Þ

where [RSite(X)] and [RSite(X)-H°] represent the molar concentrations of the deprotonated and protonated sites, respectively, and aHþ is the activity of protons in solution. To determine the minimum number of discrete site types that are necessary to account for the buffering observed in this study, and the Ka values and site concentrations for each type of site, we used FITEQL 2.023 to sequentially test models with 1 to 5 proton 4448

dx.doi.org/10.1021/es200191u |Environ. Sci. Technol. 2011, 45, 4446–4452

Environmental Science & Technology

ARTICLE

Figure 2. Histogram of A) acidity constants and B) site concentrations for all the extremophiles. Error bars represent 1σ uncertainties associated with each value.

active sites. The goodness of fit for each model was quantified using the residual function in FITEQL, V(Y), with a V(Y) value of 1 representing a perfect fit to the data.23 The calculated V(Y) values improve significantly with each additional site considered in the model to a minimum for the 4-Site model; models with 5 sites fail to converge in all cases, indicating insufficient constraints. The average V(Y) values for the titrations for the 1-, 2-, 3-, and 4-Site models are 706.8, 216.7, 29.6, and 6.2, respectively. The 4-Site models consistently yield the V(Y) values closest to 1 as well as the best visual fits to the data (e.g., Figure S2), and resulting calculated site concentrations and acidity constants are compiled in Table S1 and shown graphically in Figure 2. The cell wall sites for each species will subsequently be referred to as Site1, -2, -3, and -4, with Site-1 having the smallest Ka value and Site-4 the largest. We cannot assign chemical identities to these cell wall sites based on our data alone. One type of functional group can exhibit different protonation behavior depending on its structural conformation; therefore, spectroscopic data are required to identify the composition of each site type.

A comparison of the calculated acidity constants for the extremophiles indicates that there are significant differences between the Ka values for the acidophiles and the alkaliphiles studied here (Figure 2). Although the uncertainties associated with the Ka values for A. acidophilum are larger than those obtained for the other species, the two alkaliphile species still exhibit significantly higher Ka values for all sites except Site-1, where the Ka for B. pseudofirmus lies within two standard deviation units of the acidiphile values. The Ka values for A. acidophilum and A. cryptum are within one standard deviation unit (1σ) of each other and of the average values for the neutrophilic bacterial consortia studied by Johnson et al.12 The enhanced buffering capacity of the acidophiles studied here relative to that of A. angustum studied by Ginn and Fein14 is reflected in the calculated Ka values as well. Whereas Ginn and Fein14 determined a single Ka value of 2.7 for A. angustum, both of the acidophiles studied here exhibit buffering capacity over a much wider pH range, requiring 4 protonactive sites to account for the observed buffering behavior. The buffering behavior that we observed for the acidophiles is similar 4449

dx.doi.org/10.1021/es200191u |Environ. Sci. Technol. 2011, 45, 4446–4452

Environmental Science & Technology

Figure 3. Cd adsorption data for A. acidophilium (black triangles), A. crytpum (black circles), B. circulans (open squares), and B. pseudofirmus (open diamonds) showing the percentage of adsorbed Cd as a function of pH. All experiments contained 10 g/L (wet mass) of biomass and 10 ppm Cd in 0.1 M NaClO4. The dashed curve was calculated from the generalized model of Johnson et al. (2007), derived from adsorption experiments involving neutrophilic consortia grown from a range of natural environments.

to that of T. ferrooxidans studied by Naja et al.16 (with 2 sites with Ka values of 5.3 and 9.8), and our results suggest that A. angustum is anomalous in its limited buffering capacity. The Ka values for the alkaliphiles, B. circulans, and B. pseudofirmus, are within one standard deviation (1σ) of each other. It should be noted that the differences between the Ka values of the alkaliphiles and the acidophiles studied here may arise from the different pH ranges studied for each type of species. This is a common difficulty in quantifying proton-binding by bacterial cells, as there appears to be proton activity that extends outside the pH range of cell wall integrity, making modeling of these Ka values problematic (e.g., ref 17). The calculated site concentrations of the acidophiles and alkaliphiles in this study are within two standard deviation units of each other (Figure 2). The magnitudes of the uncertainties vary significantly between species studied, with the largest being associated with the site concentrations of B. pseudofirmus. One B. pseudofirmus titration exhibited significantly different buffering behavior than the others, and it is not clear whether this difference is due to experimental uncertainties or to real variation in cell wall chemistry during growth. The average calculated site densities for Sites 14 for all of the species in this study are shown in Table S1, and these values are similar to what was observed for natural consortia by Johnson et al. 12 as well as similar and within experimental uncertainty of the average site concentrations for similar models of potentiometric titration data for 36 neutrophilic bacterial species and consortia, analyzed Borrok et al.24 Cd Adsorption Experiments. Cd adsorption experiments were conducted as a function of pH using the two acidophiles and two alkaliphiles examined in this study. The concentration of Cd adsorbed to the cell walls is similar for all the bacteria in the study above pH 5.5, where in general an increasing amount of Cd was removed from solution with increasing pH (Figure 3). At pH values lower than 5.5, all of the bacteria except A. cryptum show a decreasing concentration of Cd adsorbed to the cells with decreasing pH. In contrast, below pH 5.5, A. cryptum exhibits markedly different Cd adsorption behavior than the other species

ARTICLE

studied, with increasing extents of Cd adsorption onto A. cryptum with decreasing pH, from approximately 60% removal at pH 4 to approximately 90% removal at pH 2. The anomalous low pH Cd adsorption behavior of A. cryptum is similar to the low pH Cd adsorption behavior of Deinococcus radiodurans,14 a radioactivityresistant bacterial species, and may be attributable to anomalous sensitivity of the cell wall to Cd2þ or to internalization or precipitation of Cd. Active internalization is unlikely during these experiments in general; the biomass used in the experiments was thoroughly washed and suspended in an electrolyte absent of nutrients and electron donors and hence was not actively metabolizing. Furthermore, the experimental duration was short, and previous experiments conducted under these conditions exhibited complete and rapid reversal of adsorption reactions,25 strongly suggesting that Cd adsorption was onto cell walls and that Cd was not internalized. Except for the low pH adsorption behavior of A. cryptum, the species studied here exhibit similar extents of adsorption over a wide range of pH (Figure 3). In fact, the extent of adsorption observed here is in reasonable agreement with the concentration of Cd that is predicted using the generalized model of Johnson et al.,12 although the extents of Cd adsorption observed in this study are on average lower than those measured by Johnson et al.12 The Johnson et al. 12 model involves Ka values, site concentrations, and Cd-bacterial site stability constant values calculated from adsorption experiments involving neutrophilic bacterial consortia. The agreement between the predicted extent of Cd adsorption and the extent observed in our experiments suggests that the generalized thermodynamic model of Johnson et al. 12 is more widely applicable than just to neutrophilic bacterial species. Stability constants for the important Cd-bacterial cell wall complexes were determined from the Cd adsorption data in this study using a nonelectrostatic surface complexation approach. We included data from the entire pH range studied for each species in these models, except for the model of the A. cryptum data which only included data in the pH range of 5.5 to 9 due to the anomalous low pH adsorption behavior. We modeled the adsorption of Cd as the formation of a bacterial surface complex with a 1:1 Cd:site molal ratio19 R  SiteðXÞ þ Cd2þ S R  SiteðXÞ  Cdþ

ð3Þ

The mass action equation for Reaction 3 can be expressed as Kads ðXÞ ¼

½R  SiteðXÞ  Cdþ  ½R  SiteðXÞ  3 aCd2þ

ð4Þ

where the brackets represent the concentration in moles of sites per liter, Kads is the equilibrium constant for the Cd adsorption reaction, and aCd2þ is the activity of Cd2þ in solution. Using the average site concentrations and Ka values for each species that we calculated from the titration modeling, we tested the goodness of fit of a range of Cd adsorption models, simultaneously modeling all data for each individual species. For each species, we determined the simplest model that could account for the observed adsorption. In each case, 1-Site models involving Cd adsorption onto any one of the four sites yielded poor fits to the data as determined by the resulting V(Y) values from FITEQL. We next constructed 2-Site models, involving Cd adsorption onto sites with sequential Ka values (e.g., Sites 1 and 2; Sites 2 and 3, and Sites 3 and 4). All these possibilities were tested for each species. The A. cryptum data were best fit by a model involving Cd 4450

dx.doi.org/10.1021/es200191u |Environ. Sci. Technol. 2011, 45, 4446–4452

Environmental Science & Technology

ARTICLE

other, and the extent of adsorption was also similar to the concentration of Cd predicted to adsorb based on the generalized model derived using neutrophilic bacterial consortia. The range of microbial diversity to which these average models of proton and metal adsorption apply remains to be determined. However, the similarities in Cd adsorption that we observed suggest that a wide range of bacteria, even those grown under extremes of pH conditions, exhibit similar binding sites and binding properties and that a generalized model of proton and metal binding onto these species can provide reasonable estimates of adsorption behavior. Figure 4. Histogram depicting the calculated K values for Reaction 3 for Site 13, and the calculated K values for Reaction 5 (labeled Rxn 5). Error bars represent 1σ uncertainties associated with each value.

adsorption onto Sites 2 and 3 (log Kads values of 3.6 and 4.2, respectively); the A. acidophilum data were best fit by a model involving Cd adsorption onto Sites 1 and 2, with log Kads values of 3.0 and 3.5, respectively. In both of these cases, adding a third site did not significantly improve the fit or resulted in a lack of model convergence due to insufficient constraints on the system. Cd2þ binding onto Sites 1 and 2 and onto Sites 1, 2, and 3 yielded the best fits to the B. circulans and B. pseudofirmus data, respectively; however, in both cases the best fitting models do not account for the observed extent of Cd adsorption at high pH. Under these conditions, CdOHþ becomes the dominant aqueous Cd species in solution and the concentration of Cd2þ decreases dramatically. Therefore, to account for the observed Cd adsorption under high pH conditions in the experiments involving the alkaliphiles, we included the following reaction in the data models R  SiteðXÞ þ Cd2þ þ H2 O  H þ S R  SiteðXÞ  CdOH o

ð5Þ

We tested models that involved CdOHþ adsorption onto different sites. For the B. circulans data, the best fitting model involved Cd adsorption onto three sites: Cd2þ adsorption onto Sites 1 and 2 and CdOHþ adsorption onto Site 4. The B. pseudofirmus data require a 4-Site model involving Cd2þ adsorption onto Sites 1, 2, and 3 and CdOHþ adsorption onto Site 4. The model fits for each bacterial species are shown in Figure S3, and calculated Kads values are compiled in Table S2 and shown in Figure 4. The histogram depicts error bars which represent 1σ uncertainty associated with each value. In each case, the best fitting model yields a good fit to the pH dependence of the adsorption data. Despite the dramatic differences in growth environments of the species studied here, the stability constants that we calculate for the important Cd-bacterial surface complexes are remarkably similar. The stability constants for Reaction 3 for Sites 13 and the calculated K values for Reaction 5 for each of the species studied are the same within experimental uncertainty, with average values for log Kads(13 and 5) of 3.4 ( 0.4, 3.5 ( 0.1, 4.6 ( 0.4, and 4.1 ( 1.0. The uncertainties associated with the calculated K values for Reaction 5 are large because these values are relatively poorly constrained by the data, and more of the high pH adsorption data would be required to determine if this K value varies significantly for different bacterial species. Regardless of differing acidity constants between the acidophiles and alkaliphiles, the bacteria studied here except A. cryptum exhibited remarkably similar Cd adsorption behavior to each

’ ASSOCIATED CONTENT

bS

Supporting Information. (1) Methods for bacterial growth and washing procedures, (2) table showing binding constants and site concentrations, (3) table showing Cd-functional group binding constants, (4) figure of representative uppH and down-pH titrations, (5) figure of best-fitting model of the potentiometric titration data, (6) figure of best-fitting models of the measured extents of Cd adsorption. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Present Addresses ‡

 Department of Chemistry, Umea University, SE - 90183 Umea, Sweden.

’ ACKNOWLEDGMENT The National Science Foundation through an Environmental Molecular Science Institute grant (EAR02-21966) provided research funding. We would also like to acknowledge the Center for Environmental Science and Technology at the University of Notre Dame, where all of the ICP-OES work was completed. Three thorough journal reviews significantly improved the presentation of the research. ’ REFERENCES (1) Beveridge, T. J.; Murray, R. G. E. Uptake and retention of metals by cell walls of Bacillus subtilis. J. Bacteriol. 1976, 127, 1502. (2) Beveridge, T. J.; Murray, R. G. E. Sites of metal deposition in the cell wall of Bacillus subtilis. J. Bacteriol. 1980, 141, 876–887. (3) Ledin, M.; Krantz-R€ulcker, C.; Allard, B. Zn, Cd and Hg accumulations by microorganisms. Organic and inorganic soil components in multicompartment systems. Soil Bio Biochim. 1996, 28, 791–799. (4) Ledin, M.; Krantz-R€ulcker, C.; Allard, R. Microorganisms as metal sorbents: comparison with other sil constituents in multicompartment systems. Soil Bio Biochim. 1999, 31, 1639–1648. (5) Daughney, C.; Fein, J. B.; Yee, N. A comparison of the thermodynamics of metal adsorption onto two common bacteria. Chem. Geol. 1998, 144, 161. (6) Ohnuki, T.; Yoshidaa, T.; Ozakia, T.; Samadfamb, M.; Kozaib, N.; Yubutac, K.; Mitsugashirac, T.; Kasamad, T.; Francis, A. J. Interactions of uranium with bacteria and kaolinite clay. Chem. Geol. 2005, 220, 237–243. (7) Neu, M. P.; Icopini, G. A.; Boukhalfa, H. Plutonium speciation affected by environmental bacteria. Radiochimica Acta 2005, 93, 705–714. 4451

dx.doi.org/10.1021/es200191u |Environ. Sci. Technol. 2011, 45, 4446–4452

Environmental Science & Technology

ARTICLE

(8) Small, T. D.; Warren, L. A.; Roden, E. E.; Ferris, F. G. Sorption of strontium by bacteria, Fe(III) oxide, and bacteria-Fe(II) oxide composites. Environ. Sci. Technol. 1999, 33, 4465–4470. (9) Yee, N.; Fein, J. B. Cd adsorption onto bacterial surfaces: A universal adsorption edge? Geochim. Cosmochim. Acta 2001, 65, 2037. (10) Yee, N.; Fein, J. B. Quantifying metal adsorption onto bacteria mixtures: A test and application of the surface complexation model. Geomicrobiol. J. 2003, 20, 43. (11) Borrok, D.; Kulpa, C. F.; Fein, J. B. Proton and Cd adsorption onto natural bacterial consortia: testing universal adsorption behaviour. Geochim. Cosmochim. Acta 2004, 68, 3231. (12) Johnson, K. J.; Szymanowski, J. E. S.; Borrok, D.; Huynh, T. Q.; Fein, J. B. Proton and metal adsorption onto bacterial consortia: Stability constants for metal-bacterial surface complexes. Chem. Geol. 2007, 239, 13. (13) Borrok, D.; Kulpa, C. F.; Fein, J. B. Cd and proton adsorption onto bacterial consortia grown from industrial wastes and contaminated geologic settings. Environ. Sci. Technol. 2004, 38, 5656. (14) Ginn, B.; Fein, J. B. The effect of species diversity on metal adsorption onto bacteria. Geochim. Cosmochim. Acta 2008, 72, 3939. (15) Daughney, C. J.; Hetzer, A.; Heinrich, H. T. M.; Burnett, P. G. G.; Weerts, M.; Morgan, H.; Bremer, P. J.; McQuillan, A. J. Proton and cadmium adsorption by the archaeon Thermococcus zilligii: Generalising the contrast between thermophiles and mesophiles as sorbents. Chem. Geol. 2010, 273, 82–90. (16) Naja, G.; Mustin, C.; Volesky, B.; Berthelin, J. A high resolution titrator: a new approach to studying binding sites of microbial biosorbents. Water Res. 2005, 39, 579–586. (17) Fein, J. B.; Boily, J. F.; Yee, N.; Gorman-Lewis, D.; Turner, B. F. Potentiometric titrations of Bacillus subtilis cells to low pH and a comparison of modeling approaches. Geochim. Cosmochim. Acta 2005, 69, 1123. (18) Plette, A. C. C.; Benedetti, M. F.; van Riemsdijk, W. H. pH dependent charging behavior of isolated cell walls of a gram-positive soil bacterium. J. Colloid Interface Sci. 1995, 173, 354–363. (19) Fein, J. B.; Daughney, C.; Yee, N.; Davis, T. A chemical equilibrium model for metal adsorption onto bacterial surfaces. Geochim. Cosmochim. Acta 1997, 61, 3319–3328. (20) Cox, J. S.; Smith, D. S.; Warren, L. A.; Ferris, F. G. Characterizing heterogeneous bacterial surface functional groups using discrete affinity spectra for proton binding. Environ. Sci. Technol. 1999, 33, 4514–4521. (21) Martinez, R. E.; Smith, D. S.; Kulczycki, E.; Ferris, F. G. Determination of intrinsic bacterial surface acidity constants using a Donnan shell model and continuous pKa distribution method. J. Colloid Interface Sci. 2002, 253, 130–139. (22) Pagnanelli, F.; Veglio, F.; Toro, L. Modelling of the acid-base properties of natural and synthetic adsorbent materials used for heavy metal removal from aqueous solutions. Chemosphere 2004, 54, 905–915. (23) Westall, J. C. FITEQL, a computer program for determination for chemical equilibrium constants from experimental data. Version 2.0. Oregon St. University, Corvallis, OR, 1982. (24) Borrok, D.; Fein, J. B. The impact of ionic strength on the adsorption of protons, Pb, Cd, and Sr onto the surfaces of Gram negative bacteria: Testing non-electrostatic, diffuse, and triple layer models. J. Colloid Interface Sci. 2005, 286, 110–126. (25) Fowle, D. A.; Fein, J. B. Experimental measurements of the reversibility of metal-bacteria adsorption reactions. Chem. Geol. 2000, 168, 27–36.

4452

dx.doi.org/10.1021/es200191u |Environ. Sci. Technol. 2011, 45, 4446–4452