CHEMSUSCHEM FULL PAPERS DOI: 10.1002/cssc.201402589

Dynamic Potential-Dependent Electron Transport Pathway Shifts in Anode Biofilms of Geobacter sulfurreducens Rachel A. Yoho,*[a, b] Sudeep C. Popat,*[a] and Csar I. Torres*[a, c]

Biofilms of the anode-respiring bacterium Geobacter sulfurreducens (G. sulfurreducens) demonstrate dynamic potential-dependent changes between two electron transport pathways that are used selectively depending on the anode potential. Electrochemical impedance spectroscopy (EIS) measurements suggest that these pathways (both n = 1), with midpoint potentials of 0.155 ( 0.005) and 0.095 ( 0.003) V versus standard hydrogen electrode, are not additive within the biofilm, but are preferentially used depending on the anode potential. Poten-

tial step voltammetry and cyclic voltammetry (CV) confirmed rapid changes between the two pathways in minutes when the anode potential is changed. We confirm that the electrochemical response observed in a slow-scan-rate CV (~ 1 mV s1) is often composed of at least the two pathways characterized. Thus, beyond understanding the electron transport pathways in G. sulfurreducens, this study also has implications on the interpretation of previously collected and future potential-dependent datasets.

Introduction Microbial respiration to solid electron acceptors, such as metals or electrodes, is not completely understood despite its relevance in the environment and biotechnology applications.[1] Electrochemical techniques applied in microbial electrochemical cells (MXCs) provide a great opportunity to study these bacteria (henceforth referred to as anode-respiring bacteria, or ARB), with many possibilities for investigating their bioenergetics and kinetics.[2] MXCs are also touted as a promising bioenergy technology as they can be used to convert the chemical energy in organic compounds directly to electrical energy.[3] The performance of MXCs is evaluated not just in the context of rates of microbial conversion but also overpotential; thus, understanding the response of ARB to the electrode potential is critical for practical applications.[4] Herein, we utilize electrochemical techniques on Geobacter sulfurreducens (G. sulfurreducens) biofilms to elucidate important details of their current–potential (j–V) response. [a] R. A. Yoho,+ Dr. S. C. Popat,+ Dr. C. I. Torres Swette Center for Environmental Biotechnology, Biodesign Institute Arizona State University 1001 S McAllister Avenue, Tempe, AZ 85287 (USA) E-mail: [email protected] [email protected] [email protected] [b] R. A. Yoho+ School of Biological and Health Systems Engineering Arizona State University P.O. Box 879709, Tempe, AZ 85287 (USA) [c] Dr. C. I. Torres School for Engineering of Matter, Transport and Energy Arizona State University 501 E Tyler Mall, Tempe, AZ 85287 (USA) [+] These authors contributed equally to this work. Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cssc.201402589.

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Cyclic voltammetry (CV) is the most widely used technique for this purpose because it measures the current response to a potential sweep, wherein the energy available to the bacteria through respiration is changed.[2, 5] CV has been applied in the characterization of model ARBs in several studies to gain information on their energy metabolism and its relation to the performance of MXCs.[2a–c, 6] G. sulfurreducens is an ARB of interest as it forms biofilms on electrodes wherein long-distance electron transport is achieved through an extracellular matrix that is known to contain pili, cytochromes, and polysaccharides.[7] The exact mechanisms of this long-distance electron transport, however, to this date, remain highly debated.[8] In this work, we explore how a combination of electrochemical techniques can reveal new information on the j–V response of G. sulfurreducens biofilms that could have a potential implication on the understanding of mechanisms of electron transport outside the cells. The CVs of G. sulfurreducens biofilms performed in the presence of the electron donor acetate show a sigmoidal shape for the j–V relationship (see Figure 1 and Refs. [2a,b, 7b]). Considerable efforts have been spent in modeling this sigmoidal response,[2c, 6c, 9] which is consistent with that of an irreversible enzymatic reaction.[10] Thus, most models assume that this sigmoidal response is governed by one limiting redox enzyme along the electron transport pathway of G. sulfurreducens. In 2007, our group developed the Nernst–Monod equation [Eq. (1)], which was confirmed to accurately fit the CVs of G. sulfurreducens.[10] In Equation (1), the current density produced by the biofilm (j) plotted against the anode potential (E or V) is a function of the maximum current density (jmax) and the midpoint potential of the limiting redox process (EKA). Also, n is the number of electrons transferred, which is assumed to be 1 for the rate-limiting redox enzyme. Since then, other ChemSusChem 0000, 00, 1 – 8

&1&

These are not the final page numbers! ÞÞ

CHEMSUSCHEM FULL PAPERS

www.chemsuschem.org

Figure 1. An example of a CV for a fully developed G. sulfurreducens biofilm grown with an anode potential of 0.02 V, showing a sigmoidal j–V relationship (c) that deviates from the Nernst-Monod equation (a).

models have been developed, as it was found that the Nernst– Monod equation does not always predict or fit the data.[9a] This is highlighted in Figure 1 where deviations in the top and bottom of the sigmoidal response between the data and the model are evident. j ¼ jmax

1

Results and Discussion

!

 nF  1 þ exp  RT ðE  EKA Þ

ð1Þ

Other models, such as the Butler–Volmer–Monod model and the Nernst–Ping–Pong model, have included new concepts while still adhering to the assumption of a single limiting and/ or governing enzymatic or redox reaction along the electron transport pathway.[9] The only model that goes beyond this assumption is that put forth by Snider et al. in which the authors introduce an empirical parameter “g” to describe the presence of multiple redox co-factors governing electron transport outside the cells.[7c] Even so, a greater fundamental understanding of the respiratory metabolism of G. sulfurreducens is needed to interpret the electrochemical responses that lead to deviation from most current models. Several researchers have also obtained CVs in the absence of acetate that show the involvement of multiple redox processes;[5a, 7b, 11] we show an example in Figure S1 in the Supporting Information. Here, processes could refer to a single redox enzyme or an enzymatic pathway with a characteristic midpoint potential; so far, this distinction has been difficult to make with the techniques used. The previous results summarized here show an incomplete understanding of the electrochemical response of G. sulfurreducens. The discrepancy between experimental data and theoretical calculations suggests a complex electrochemical behavior involving multiple redox processes that cannot be explained with our current mechanistic models. Our inability to represent experimental data with a mathematical model connotes our knowledge gap related to electron transport pathways in anode respiration and impedes our ability to predict ARB behavior in MXCs. A better understanding of electron transport kinetics of ARBs and the relationship with the anode potential presents possibilities of optimizing operating conditions of MXCs in various biotechnology applications and predicting the response of ARB through advanced modeling techniques.  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

In this study, we aimed to evaluate the contribution of multiple governing redox processes to the j–V response of G. sulfurreducens biofilms. Because CV represents a transient electrochemical method, we used it in conjunction with other electrochemical methods that provide more detailed, pseudo-steadystate information to highlight the involvement of multiple governing redox processes shaping the overall j–V relationship. We pay special attention to the use of chronoamperometry where we can observe time-dependent changes as a function of the anode potential and electrochemical impedance spectroscopy (EIS) performed at different anode potentials to evaluate charge transfer resistances. Our results suggest a complex electrochemical response of G. sulfurreducens that has not been characterized in detail before in which at least two major electron transport pathways contribute to current production. The potential-dependent response of G. sulfurreducens biofilms we show herein may fundamentally change our understanding of their metabolism and the interpretation of all previously collected and future datasets.

CV measurements We investigated first if the deviations in CVs of G. sulfurreducens biofilms from the Nernst–Monod equation, as shown in Figure 1, manifest with the maturity of the biofilms. In previous analyses, we have discussed that a deviation in the CV at high current densities could be the consequence of a potential gradient along the biofilm.[2c] This deviation results in a slope less steep at anode potentials above EKA. We performed CV on G. sulfurreducens biofilms grown at an anode potential of 0.02 V as a function of biofilm growth. We show in Figure 2 a the CVs for one biofilm from measurements made when the current densities reached 0.5, 1, 3, 5, and 6 A m2. We repeated these measurements with four different biofilms and observed the same trend (data not shown). It is difficult to observe any major changes in the sigmoidal shape of the CVs through this representation. However, it is evident that the onset potential, where positive current can first be measured, decreases with increasing biofilm growth. We show in Figure S2 Nernst– Monod equation fits for each of the CVs shown in Figure 2 a. Figure S2 suggests that the deviation from the Nernst–Monod equation starts to occur when the current density is > 3 A m2. The changes of patterns in the CVs shown in Figure 2 a with time are not directly apparent. To visualize these changes, we normalized each CV to the maximum current density for each curve, jmax (Figure 2 b). This figure suggests that with increasing growth of the biofilm there is a change in the sigmoidal shape with 1) the midpoint decreasing to more negative values, and 2) a decrease in the overall slope. This change is more pronounced in the region of 0.2 j/jmax at which a 25 mV shift in the anode potential from 0.152 to 0.177 V occurs. Interestingly, the normalized CVs mostly overlap at higher potentials except for a small peak visible in the early growth stages (< 1 A m2) at 0.075 V. This peak is observed in other studies as well,[7b] and its origin has remained unknown. Nonetheless, this region ChemSusChem 0000, 00, 1 – 8

&2&

These are not the final page numbers! ÞÞ

CHEMSUSCHEM FULL PAPERS

Figure 2. (a) CVs at various stages of growth for a G. sulfurreducens biofilm grown with an anode potential at 0.02 V. (b) CVs from (a) normalized to the maximum current density at each stage of biofilm growth, showing shifts in the slope of the sigmoidal j–V relationship. c: 6 A m2 ; c: 5 A m2 ; c: 3 A m2 ; c: 1 A m2 ; c: 0.5 A m2.

is where, as discussed above, a potential gradient would change the slope of the CVs the most.[2c] Thus, the shift we observed here cannot be attributed directly to a strong potential gradient along the biofilm. Other studies focusing on electrochemical analyses of G. sulfurreducens biofilms also suggest such changes in the CVs with biofilm growth. For example, Marsili et al. also showed similar changes,[2b] except that they showed derivative CVs instead of normalized CVs as we do here. Strycharz-Glaven and Tender also presented CVs as a function of biofilm growth wherein changes in the CVs only in the low current region (< 0.15 j/jmax) are apparent.[7b] The normalized CVs suggest that there are features in the respiratory metabolism of G. sulfurreducens that are not captured by an assumption of one governing redox process (represented with n = 1). In fact, these normalized CVs suggest an adaptation to lower potentials within the biofilm, with more current produced at lower anode potentials. This too refutes the possibility of a direct potential-gradient-related shift in the shapes of the CVs as one would expect the potential gradient to decrease the normalized current produced at a given anode potential, not increase it. The shift also supports our observation of deviation of the CV from the Nernst–Monod relationship on fully developed biofilms, such as that shown in Figure 1. EIS measurements To further elaborate on this observation, we performed EIS measurements on fully developed G. sulfurreducens biofilms  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org grown at 0.02 V over a range of discrete anode potentials. Because CV is a transient method that may not allow obtaining a pseudo-steady-state j–V response, we used EIS to analyze the relationship of charge-transfer resistance (Rct, W cm2) with the anode potential. Unlike direct-current techniques such as CV, EIS is an alternating-current method. We used potentiostatic EIS, which involves applying a small sinusoidal amplitude over a fixed anode potential at a range of frequencies and measuring the resulting magnitude and phase shift in the current. Because different processes, such as charge transport, charge transfer, and mass transfer, govern the response in current to changes in potential at different frequencies, it is thus theoretically possible to quantify individual contributions of different processes to the overall resistance. The Nernst–Monod equation can be used also to represent the Rct–V response. Rct is related to j through Ohm’s law (Rct = dV/dj), so the sigmoidal shape for a j–V relationship, changes to a U-shape for Rct–V relationship similar to the derivative of a CV as shown in Ref. [2a and b]. Because EIS is performed at a single anode potential at a time, it should provide pseudosteady-state responses for the Rct–V relationship. An example of an experimental dataset of Nyquist plots obtained through EIS measurements is shown in Figure 3 together with the equivalent circuit model (ECM) used (inset; three more examples are shown in Figure S3). Fitting the Nyquist plot to an ECM is a generally accepted method to obtain values for particular resistances within an electrochemical system, provided that the model accurately depicts the physicochemical–biological phenomena occurring. We selected a circuit that contains three Rcts. This selection was made on the basis of Bode plots obtained from several EIS measurements that showed three peaks in the phase shift (data not shown). As we anticipate that these responses come from the electron transport chain, we used an ECM with three resistances in series. When fitting to the Nernst–Monod equation, we used the sum of the values for all the three resistances, R1, R2, and R3, because it models the overall response of the cells, not just an individual resistance. In Figure 4 a, we show the total Rct (i.e., R1 + R2 + R3) versus the anode potential for a G. sulfurreducens biofilm grown at

Figure 3. An example of the Nyquist plot from EIS measurement on a G. sulfurreducens biofilm grown at an anode potential of 0.02 V (data: *; ECM fit: a. Conditions used for this example were: anode potential of 0.12 V, amplitude of 10 mV, and frequencies ranging from 200 kHz to 10 mHz. The ECM is shown as the inset to the figure.

ChemSusChem 0000, 00, 1 – 8

&3&

These are not the final page numbers! ÞÞ

CHEMSUSCHEM FULL PAPERS

www.chemsuschem.org We can estimate jmax for each of the processes (7 and 5.5 A m2) from the model fits performed by minimizing the sum least squares between the data and the Nernst–Monod equation to represent the Rct–V response. Additionally, the two processes appear to have different EKA values for the governing redox co-factors (0.157 V for P1 and 0.097 V for P2) whereas the often reported single EKA is 0.14 V.[2a–c, 6a, 11] Averaging all replicate measurements, even at different current densities, we obtain EKA values of 0.155 ( 0.005) and 0.095 ( 0.003) V for P1 and P2. Interestingly, the two processes do not appear to be additive as the overall jmax (6.9 A m2 during EIS measurements) is significantly lower than the sum of jmax for both processes. This suggests that the measured overall current density is the result of a shift from one pathway to another as a function of anode potential. Thus, from now on, we will refer to these processes as electron transport pathways. Biofilm experiments at low anode potentials

Figure 4. Charge transfer resistance as a function of anode potential for a fully developed G. sulfurreducens biofilm grown with an anode potential of 0.02 V. The experimental data is shown in blue dots, where Rct is the sum of three charge-transfer resistances. The dashed lines show best fits with the Nernst-Monod equation considering (a) a single pathway with the same parameters used to fit the CV in Figure 1, and (b) two pathways, P1 and P2, with jmax = 7 and 5.5 A m2 respectively and EKA = 0.157 and 0.097 V respectively. Parameters were obtained through a best fit by minimizing the sum of least squares using Excel Solver.

0.02 V. We repeated these measurements on four different biofilms two times each and obtained the same trends as shown here. We show three additional plots in Figure S4. As mentioned earlier, this plot can be considered as the derivative of the CV (R = dV/dj according to Ohm’s Law). Additionally in Figure 4 a, we overlay a black dotted line, which represents the best fit of the Nernst–Monod equation to show dV/dj. The curve shows a minimum at EKA with an Rct value at this potential that is determined by jmax. It also shows two slopes (one at each side of the EKA), which should be symmetrical and which are associated with the n value (n = 1 in Figure 4 a). In Figure 4 a, we observe that the Nernst–Monod equation does not accurately describe the Rct–V relationship as determined through EIS measurements. However, although the apparent deviations in experimental and modeling results are small for the CV in Figure 1, those in the Rct–V relationship are more evident (Figure 4 a). In particular, we observe that the width of the U-shape is significantly underestimated by the Nernst–Monod equation. In fact, the experimental data suggest that there are two minima in the Rct, which can be modeled by two separate Nernst–Monod equations using different EKA and jmax values, each with n = 1 (black and red dotted lines in Figure 4 b). This suggests that there are two major electron transport processes, henceforth denoted P1 and P2, each with different characteristic midpoint potentials.  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

To investigate further whether a pathway shift is responsible for the observed response of G. sulfurreducens, we grew biofilms at a relatively low anode potential (0.145 V) in an effort to obtain active expression of only P1. As shown in Figure 4 b, the selected potential is close to the EKA value of P1 while giving a very large Rct for P2. We recorded chronoamperometric j–V curves in which we stepped the anode potential through a defined range, allowing sufficient time (at least 10 min) for the biofilm to equilibrate at each potential. This should allow a shift from one pathway to the other even if it implies expression of new proteins. Figure 5 a shows the results of this experiment: the curves resemble a slow-scan-rate simulated CV. The results suggest that there is a decrease in current production as there is a shift from P1 to P2. This further confirms why a simple additive model for the two pathways does not explain the Rct–V data we show in Figure 4. We performed a potential-step experiment on a biofilm grown at 0.145 V to track the time it takes to change from one pathway to the other. In the experiment, the potential was changed from the poised growth potential (0.145 V, shown in blue) to 0.02 V (shown in red) for approximately 30 min and then changed back to the original growth potential (shown in blue). We show the data in Figure 5 b, with a replicate dataset also shown in Figure S5. When the anode potential was stepped up to 0.02 V, the biofilm responded by increasing current production, as expected by the Nernst– Monod equation. However, this increase only lasted approximately 2 min; then, the current started to decrease, finally equilibrating to a lower value, a behavior similar to that shown in Figure 5 a. We must point out that the observed response is that of a biofilm grown at a relatively low potential; a biofilm grown at 0.02 V would be able to produce > 6 A m2 as shown in Figure 2 a. A similar but opposite response is observed when the potential is stepped back down to 0.145 V; a lower current is observed for approximately 20 min and is then followed by an increase, leading to the original steadystate current. These results suggest that the shift in pathways ChemSusChem 0000, 00, 1 – 8

&4&

These are not the final page numbers! ÞÞ

CHEMSUSCHEM FULL PAPERS

Figure 5. (a) j–V curve for a fully developed G. sulfurreducens biofilm grown with an anode potential of 0.145 V. (b) Chronoamperometric measurements on a fully developed G. sulfurreducens biofilm grown with an anode potential of 0.145 V, showing shifts in the current density when the anode potential is shifted from 0.145 V (blue) to 0.02 V (red). Note that (a) and (b) are from two different biofilms grown with an anode potential of 0.145 V at different time points, and thus the absolute current densities are different.

is also associated with a change in the current density, which is an indirect measurement of G. sulfurreducens respiration rate. It also suggests that the pathway shifts occur in a matter of minutes. To confirm further that there is a shift in electron transport pathways of G. sulfurreducens, we measured the CVs shown in Figure 6 a. A replicate is also shown in Figure S6. In this experiment, CVs were performed on a biofilm grown at 0.145 V at a scan rate of 5 mV s1, for a total scan time of 80 s, which is lower than the time observed in Figure 5 b for the pathway change. After a CV was performed at 0.145 V, the potential was changed to 0.02 V for 1 h, allowing for the shift in the electron transport pathway. Then, we performed a second CV at 5 mV s1, which shows a clear shift in the midpoint potential. This was also accompanied by a lower current density (jmax = 5.7 A m2 at 0.145 V vs. jmax = 4.8 A m2 at 0.02 V). These experiments confirm that there is a shift between the two distinct pathways that have governing redox co-factors with different characteristic midpoint potentials. The derivatives of the CVs (Figure 6 b) reveal a minimum at 0.105 and 0.159 V, values that are similar to those observed in EIS measurements (Figure 4). Although the derivatives show primarily only one pathway for each CV scan, the lack of symmetry in both curves suggests that we did not completely eliminate the other pathway in each case, but only suppressed it so that we can observe the signature of the major pathway. Similar responses  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemsuschem.org

Figure 6. (a) A comparison of the CVs obtained for a fully developed G. sulfurreducens biofilm grown at an anode potential of 0.145 V (blue) and the same biofilm 1 h later after it has been adapted to an anode potential of 0.02 V (red). (b) Derivatives of the CVs in (a).

have been observed in mixed-culture biofilms, where pathways from different microbial species or strains are selected based on the anode potential,[12] but this pathway shift has not been observed on the same biofilm of a pure-culture ARB until now. From these results, it is evident that G. sulfurreducens biofilms can change electron transport pathways dynamically as a response to the anode potential. We hypothesize that the most likely reason for this pathway change is due to an energy optimization in their respiration and, thus, is probably similar to how facultative anaerobic bacteria shift between oxygen and lower potential electron acceptors to optimize their growth and yield.[13] If this is true for G. sulfurreducens, the shift in pathways would occur inside the cells and be associated to a proton-pumping protein in the electron transport chain, which increases the proton-motive force (PMF) and adenosine triphosphate (ATP) production. If our hypothesis is correct, G. sulfurreducens should be able to grow at a faster rate at higher potentials as ATP can be produced at a faster rate. In fact, we estimated the doubling times for the two different potentials studied based on the exponential increase in current density after inoculation (see Figures S7 and S8). Based on these, we calculated the doubling time to be approximately 7 h at 0.02 V while it was approximately 14 h at 0.145 V. It is known that the doubling times for different electron acceptors differ for G. sulfurreducens;[14] we show here that this difference could arise even from a fairly small increase in redox potential from 0.145 to 0.02 V. Our hypothesis of energy optimization would explain also the change in current density observed in Figures 5 and 6 as ChemSusChem 0000, 00, 1 – 8

&5&

These are not the final page numbers! ÞÞ

CHEMSUSCHEM FULL PAPERS

www.chemsuschem.org

the potential is changed. As the cells obtain more PMF per electron respired at higher potentials, the cells reaches a maximum ATP production rate with a lower respiration rate. This would manifest as a lower current density at the higher-potential pathway. While our data support the mentioned hypothesis, it is rather difficult to confirm that the pathway change is associated with a shift in proteins within the electron transport chain. To date, we do not have a full understanding of the electron transport chain in G. sulfurreducens, which contains over 100 cytochromes that could be involved in different respiratory pathways,[15] including the ones characterized here. The dynamic pathway shift observed, in the order of minutes, can explain many observations related to the electrochemical responses of G. sulfurreducens. Non-turnover CVs already show the presence of two redox peaks, which we now suggest could be related to two separate pathways instead of different redox proteins within a single pathway, as previously hypothesized.[11] Under turnover conditions, CVs often show a higher current at lower potentials (such as that shown in Figure 5 a); this observation has been attributed to negative Faradaic resistances[16] and can now be explained more mechanistically through the pathway shift. The shifts in CVs as a function of growth, such as the ones shown in Figure 2, suggest an adaptation of the biofilm through a pathway optimization as the biofilm matures. Thus, the dynamic pathway shift characterized in this work affects the interpretation of most electrochemical data collected on G. sulfurreducens so far.

scale of minutes (seen in Figure 5 b) suggest that the changes are inside the cells. Changes outside of the cells would be more energetically “expensive” changes and would, likely, be slower than what we observed. Although changes in the slope of the CVs we observe in Figure 2 a as a function of biofilm growth can now primarily be attributed to the two pathways, and not necessarily to potential gradients, we should acknowledge that the use of a second pathway could be an implication of a potential gradient. Several studies have shown that the anode biofilms become more reduced on the outside,[17] which could trigger the preferred use of one process (P1) over another (P2) in biofilms grown at a relatively high anode potentials. Even so, because we anticipate that the pH would also play a major role in the energetics of a given pathway, pH gradients already shown to occur in these biofilms[17] could also influence changes in the shape of the CVs and, thus, should be a topic of studies focused on gradients. The expression of different pathways depending on the anode potential for G. sulfurreducens has far-reaching implications on the interpretation of all previously collected and future datasets. At present, we cannot determine whether these results are the work of multiple proteins or a single protein with multiple “exit pathways.” These results and conclusions open up more research questions as to the mechanisms and the identity of these pathways for the electron transport chain in G. sulfurreducens.

Conclusions

Experimental Section

From the experimental data obtained using several electrochemical techniques, we observe that the biofilms of G. sulfurreducens exhibit at least two major pathways for electron transport during anode respiration. We have observed also that there can be a shift between these pathways on a time scale of minutes. The quick response to a potential change suggests that there is an efficient capability in G. sulfurreducens to sense the redox potential. This sensing mechanism has not yet been discovered for G. sulfurreducens and is most likely used for respiring to solid metal oxides that have different redox potentials in nature. The dynamic shift in pathways is associated also with a change in respiratory rates, as observed by a change in current density. The Nernst–Monod equation was useful in determining the midpoint of the governing redox co-factor in each pathway, but as of yet does not model the overall j–V response, especially as changes in pathway are on the order of minutes, which is similar to that of a slowscan-rate cyclic voltammograms (CVs). We have confirmed that the deviations from theoretical Nernst–Monod calculations observed in the CVs of G. sulfurreducens (Figure 1) are the result of transient changes in pathways as the CV is recorded. We propose that the manifested changes are inside of the cell as part of their electron transport chain, especially because of the differences in doubling times observed when grown under conditions that result in preferable expression of only one pathway. Additionally, the rapidity of the expression changes and equilibration to different potentials on a time

Bacterial strain and culture media: We subcultured G. sulfurreducens from a stock of microorganisms isolated in our laboratory and determined that they were genetically similar to previously isolated G. sulfurreducens.[18] The culture was grown in an anaerobic medium containing per 1 L: NaHCO3 (2.50 g, 30 mm), NH4Cl (1.50 g, 20 mm), NaH2PO4 (0.60 g, 4 mm), KCl (0.10 g, 1 mm), sodium acetate trihydrate (1.36 g, 10 mm), vitamin mix (10 mL), and trace minerals (10 mL). The trace-mineral solution contained per 1 L: nitrilotriacetic acid, trisodium salt (1.5 g, 5.5 mm), MgSO4·7 H2O (3.0 g, 12 mm), MnSO4·H2O (0.5 g, 2.9 mm), NaCl (1.0 g, 17 mm), FeSO4·7 H2O (0.1 g, 0.36 mm), CaCl2·2 H2O (0.1 g, 0.68 mm), CoCl2·6 H2O (0.1 g, 0.42 mm), ZnCl2 (0.13 g, 0.95 mm), CuSO4·5 H2O (10 mg, 0.4 mm), AlK(SO4)2·12 H2O (10 mg, 0.2 mm), H3BO4 (10 mg, 0.16 mm), Na2MoO4·H2O (2.5 mg, 0.01 mm), NiCl2·6 H2O (2.4 mg, 0.01 mm), Na2WO4·2 H2O (0.25 mg, 8.5 mm). We provided acetate as the electron donor at 30 mm. We bubbled all media with N2/CO2 (80:20 v/v) to remove oxygen before autoclaving. Microbial electrochemical cells: We constructed three-electrode MXCs (schematic shown in Figure S9 and photograph in Figure S10) from VWR bottles (100 mL) that were modified to have an opening on one side for the working electrode. We constructed rubber stoppers for the gas inflow and outflow, reference electrode, and nickel wire counter electrodes. We used gold electrodes of a 2 mm diameter (3.14 mm2 area, BASi, West Lafayette, IN), and Ag/AgCl reference electrodes (+ 0.28 V vs. SHE in the medium used as determined through the procedure described earlier,[19] BASi, West Lafayette, IN), as the working and reference electrodes, respectively. All potentials reported are versus SHE. The MXCs were magnetically stirred at 250 rpm and continuously flushed with humidified N2/CO2 gas (80:20 v/v). We sterilized the MXCs by auto-

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemSusChem 0000, 00, 1 – 8

&6&

These are not the final page numbers! ÞÞ

CHEMSUSCHEM FULL PAPERS claving prior to filling media through sterile Norprene tubing by using filtered N2 gas (0.22 mm pore size) in a biosafety cabinet. We inoculated the MXCs with approximately 2 mL volume inoculum from suspension of larger H-type MXCs, which were set up as described before.[20] Electrochemical measurements: We performed all electrochemical measurements using a VMP3 digital potentiostat (Bio-Logic USA, Knoxville, TN). We used chronoamperometry to poise the anode potential at different fixed values (for details see the Results and Discussion section). We performed CV by scanning the anode potential forward and reverse twice starting from the open circuit potential at a scan rate of 1 or 5 mV s1 over the potential window usually of 0.25 to + 0.05 V. We used the second scans for all data analysis. We corrected the CVs for IR loss as described earlier.[21] We performed EIS measurements at different anode potentials (depending on experiment, see the Results and Discussion section) following an equilibration time at the set potential (minimum 10 min) to ensure steady-state conditions for the experiment. We used an amplitude of 10 mV and a frequency range of 200 kHz to 10 mHz with ten measurements per decade. We fit the EIS measurement data in the form of the Nyquist plots to an ECM using the EC-Lab software. The circuit for the ECM we used is presented as an inset in Figure 3. We performed Kramers–Kronig (KK) tests on EIS data to ensure data validity by using the KK test software.[22] Within the range of the anode potentials tested, all data conformed to the KK tests with residuals of < 3 %.

www.chemsuschem.org

[6]

[7]

[8]

[9]

[10] [11] [12]

Acknowledgements We acknowledge support from a number of sources, including an NSF Graduate Research Fellowship for R.A.Y. (DGE-1311 230), support for C.I.T. and S.C.P. from the Office of Naval Research (grant N000141210344). We also acknowledge and appreciate the assistance of Kassandra Nikkhah in the laboratory. Additionally, we acknowledge the work of Oluyomi Ajulo and her helpful discussions on the behavior of various anode-respiring bacteria. Additionally, we acknowledge the Swette Center for Environmental Biotechnology and the generous endowment from Mr. Brian Swette to the Center, which supported part of this research. Keywords: anodes · biofilms · cyclic voltammetry · electrochemical impedance spectroscopy · electron transport [1] a) D. R. Lovley, Microbiol. Rev. 1991, 55, 259 – 287; b) D. R. Lovley, Curr. Opin. Biotechnol. 2006, 17, 327 – 332; c) L. J. Bird, V. Bonnefoy, D. K. Newman, Trends Microbiol. 2011, 19, 330 – 340. [2] a) E. Marsili, J. B. Rollefson, D. B. Baron, R. M. Holzalski, D. R. Bond, Appl. Environ. Microbiol. 2008, 74, 7329 – 7337; b) E. Marsili, J. Sun, D. R. Bond, Electroanalysis 2010, 22, 865 – 874; c) C. I. Torres, A. K. Marcus, P. Parameswaran, B. E. Rittmann, Environ. Sci. Technol. 2008, 42, 6593 – 6597. [3] a) K. Rabaey, W. Verstrate, Trends Biotechnol. 2005, 23, 291 – 298; b) B. E. Logan, B. Hamelers, R. Rozendal, U, Schrçder, J. Jeller, S. Freguia, P. Aelterman, W. Verstraete, K. Rabaey, Environ. Sci. Technol. 2006, 40, 5181 – 5192; c) B. E. Logan, D. Call, S. Cheng, H. V. M. Hamelers, T. H. J. A. Sleutels, A. Q. Jeremiasse, R. A. Rozendal, Environ. Sci. Technol. 2008, 42, 8630 – 8640; d) B. E. Logan, K. Rabaey, Science 2012, 337, 686 – 690. [4] a) R. A. Rozendal, H. V. M. Hamelers, J. Rabaey, J. Keller, C. J. N. Buisman, Trends Biotechnol. 2008, 26, 450 – 459; b) C. I. Torres, Curr. Opin. Biotechnol. 2014, 27, 107 – 114. [5] a) E. LaBelle, D. R. Bond, Bio-Electrochemical Systems: From Extracellular Electron Transfer to Biotechnological Application. Integrated Environmen-

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

[13] [14] [15]

[16] [17]

[18] [19] [20] [21] [22]

tal Technology Series, 2008, pp. 1 – 16; b) F. Harnisch, S. Freguia, Chem. Asian J. 2012, 7, 466 – 475. a) E. Marsili, D. B. Baron, I. D. Shikhare, D. Coursolle, J. A. Gralnick, D. R. Bond, Proc. Natl. Acad. Sci. USA 2008, 105, 3968 – 3973; b) H. Richter, K. P. Nevin, H. Jia, D. A. Loqy, D. R. Lovley, L. M. Tender, Energy Environ. Sci. 2009, 2, 506 – 516; c) S. M. Strycharz, A. P. Malanoski, R. M. Snider, H. Yi, D. R. Lovley, L. M. Tender, Energy Environ. Sci. 2011, 4, 896 – 913; d) A. A. Carmona-Martinez, F. Harnisch, L. A. Fitzgerald, J. C. Biffinger, B. R. Ringeisen, U. Schrçder, Bioelectrochemistry 2011, 81, 74 – 80; e) A. A. Carmona-Martnez, F. Harnisch, U. Kuhlicke, T. R. Neu, U. Schrçder, Bioelectrochemistry 2013, 93, 23 – 29. a) N. S. Malvankar, M. Vargas, K. P. Nevin, A. E. Franks, C. Leang, B.-C. Kim, K. Inoue, T. Mester, S. F. Covalla, J. P. Johnson, V. M. Rotello, M. T. Tuominen, D. R. Lovely, Nat. Nanotechnol. 2011, 6, 573 – 579; b) S. M. Strycharz-Glaven, L. M. Tender, ChemSusChem 2012, 5, 1106 – 111; c) R. M. Snider, S. M. Strycharz-Glaven, S. D. Tsoi, J. S. Erickson, L. M. Tender, Proc. Natl. Acad. Sci. USA 2012, 109, 15467 – 15472; d) D. R. Bond, S. M. Strycharz-Glaven, L. M. Tender, C. I. Torres, ChemSusChem 2012, 5, 1099 – 1105. a) N. S. Malvankar, M. T. Tuominen, D. R. Lovley, Energy Environ. Sci. 2012, 5, 6247 – 6249; b) S. M. Strycharz-Glaven, L. M. Tender, Energy Environ. Sci. 2012, 5, 6250 – 6255. a) H. V. Hamelers, A. Ter Heijne, N. Stein, R. A. Rozendal, C. J. Buisman, Bioresour. Technol. 2011, 102, 381 – 387; b) S. Peng, D. W. Liang, P. Diao, Y. Liu, F. Lan, Y. Yang, S. Lu, Y. Xiang, Bioresour. Technol. 2013, 136, 610 – 616. A. K. Marcus, C. I. Torres, B. E. Rittmann, Biotechnol. Bioeng. 2007, 98, 1171 – 1182. K. Fricke, F. Harnisch, U. Schrçder, Energy Environ. Sci. 2008, 1, 144 – 147. a) A. S. Commault, G. Lear, M. A. Packer, R. J. Weld, Bioresour. Technol. 2013, 139, 226 – 234; b) X. Zhu, M. D. Yates, M. C. Hatzell, H. A. Rao, P. E. Saikaly, B. E. Logan, Environ. Sci. Technol. 2014, 48, 1352 – 1360; c) C. I. Torres, R. Krajmalnik-Brown, P. Parameswaran, A. K. Marcus, G. Wagner, Y. A. Gorby, B. E. Rittmann, Environ. Sci. Technol. 2009, 43, 9519 – 9524. B. E. Rittmann, P. L. McCarty, Environmental Biotechnology: Principles and Applications, McGraw-Hill Book Co., New York, 2003. A. Esteve-Nunez, M. Rothermich, M. Sharma, D. Lovley, Environ. Microbiol. 2005, 7, 641 – 648. B. A. Meth, K. E. Nelson, J. A. Eisen, I. T. Paulsen, W. Nelson, J. F. Heidelberg, D. Wu, M. Wu, N. Ward, M. J. Beanan, R. J. Dodson, R. Madupu, L. M. Brinkac, S. C. Daughtery, R. T. DeBoy, A. S. Durkin, M. Gwinn, J. F. Kolonay, S. A. Sullivan, D. H. Haft, J. Selengut, T. M. Davidsen, N. Zafar, O. White, B. Tran, C. Romero, H. A. Froberger, J. Weidman, H. Khouri, T. V. Feldblyum, T. R. Utterback, S. E. Van Aken, D. R. Lovely, C. M. Fraser, Science 2003, 302, 1967 – 1969. S. Matsuda, H. Liu, S. Kato, K. Hashimoto, S. Nakanishi, Environ. Sci. Technol. 2011, 45, 10163 – 10169. a) L. Robuschi, J. P. Tomba, G. D. Schrott, P. S. Bonnani, P. M. Desimone, J. P. Busalmen, Angew. Chem. 2013, 125, 959 – 962; Angew. Chem. Int. Ed. 2013, 52, 925 – 928; Angew. Chem. Int. Ed. 2013, 52, 925 – 928; b) N. Lebedev, S. M. Strycharz-Glaven, L. M. Tender, ChemPhysChem 2014, 15, 320 – 327; c) B. Virdis, D. Milo, B. C. Donose, D. J. Batstone, PLOS One 2014, 9, e89918. J. P. Badalamenti, R. Krajmalnik-Brown, C. I. Torres, mBio 2013, 4, e00144-13. P. Parameswaran, T. Bry, S. C. Popat, B. G. Lusk, B. E. Rittmann, C. I. Torres, Environ. Sci. Technol. 2013, 47, 4934 – 4940. C. I. Torres, A. K. Marcus, B. E. Rittmann, Biotechnol. Bioeng. 2008, 100, 872 – 881. S. C. Popat, D. Ki, B. E. Rittmann, C. I. Torres, ChemSusChem 2012, 5, 1071 – 1079. B. A. Boukamp, J. Electrochem. Soc. 1995, 142, 1885 – 1894.

Received: May 25, 2014 Revised: August 31, 2014 Published online on && &&, 0000

ChemSusChem 0000, 00, 1 – 8

&7&

These are not the final page numbers! ÞÞ

FULL PAPERS R. A. Yoho,* S. C. Popat,* C. I. Torres* && – && Dynamic Potential-Dependent Electron Transport Pathway Shifts in Anode Biofilms of Geobacter sulfurreducens

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

A small change in pathway: Biofilms of Geobacter sulfurreducens exhibit two pathways of electron transport to the anode that are used preferentially depending on the anode potential; this observation has implications not only on the understanding of electron transport in this model anode-respiring bacterium, but also has implications on interpretation of past and future electrochemical data.

ChemSusChem 0000, 00, 1 – 8

&8&

These are not the final page numbers! ÞÞ

Dynamic potential-dependent electron transport pathway shifts in anode biofilms of Geobacter sulfurreducens.

Biofilms of the anode-respiring bacterium Geobacter sulfurreducens (G. sulfurreducens) demonstrate dynamic potential-dependent changes between two ele...
1MB Sizes 0 Downloads 5 Views