Electrochemical Study of Reversible Hydrogenase Reaction of

S-shaped voltammogram with both cathodic- and anodic- catalytic-limiting currents in a methyl viologen-containing buffer saturated with H2. Methyl vio...
0 downloads 0 Views 142KB Size
Download PDF
Anal. Chem. 1999, 71, 1753-1759

Electrochemical Study of Reversible Hydrogenase Reaction of Desulfovibrio vulgaris Cells with Methyl Viologen as an Electron Carrier Hirosuke Tatsumi, Kazuyoshi Takagi, Megumi Fujita, Kenji Kano,* and Tokuji Ikeda*

Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan

An electrode modified with immobilized whole cells of Desulfovibrio vulgaris (Hildenborough) produces an S-shaped voltammogram with both cathodic- and anodiccatalytic-limiting currents in a methyl viologen-containing buffer saturated with H2. Methyl viologen penetrates into the bacterial cells to serve as an electron carrier in the reversible reaction of hydrogenase in the cells and functions as an electron-transfer mediator between the bacterial cells and the electrode, thus producing the catalytic currents for the evolution and consumption of H2. An equation for the catalytic current that takes into account the reversible hydrogenase reaction explains well the shape of the voltammogram. The potential at null current on the voltammogram agrees with the potential determined by potentiometry with the same electrode, which is equal to the redox potential of the H+/H2 couple in the solutionsthe standard potential of a hydrogen electrode at the pH of the solution. When D. vulgaris cells are suspended in an argon-saturated buffer containing methyl viologen, the suspension produces a catalytic current at a bare glassy carbon electrode for the evolution of H2. Analysis of the current by a theory for a catalytic current for a unidirectional nonlinear enzyme catalysis allows us to determine the kinetic parameters of the reaction between methyl viologen and hydrogenase in intact D. vulgaris cells. Thus we obtain the apparent Michaelis constant for methyl viologen cation radical, K′MV•+ ) 0.16 mM, and the apparent catalytic constant (that is, the turnover number per D. vulgaris cell), zkcat,H+ ) 1.2 × 107 s-1, for the H2 evolution reaction at pH 5.5 and at 25 °C, z being the number of hydrogenases contained in a D. vulgaris cell. The bimolecular reaction rate constant, kcat,H+/K′MV•+, of the reaction between methyl viologen cation radical and oxidized hydrogenase in intact D. vulgaris cells is estimated as 4.2 × 107 M-1 s-1. Similarly, the bimolecular reaction rate constant, kcat,H2/ K′MV2+, of the reaction between methyl viologen and reduced hydrogenase is estimated to be 1.2 × 107 M-1 s-1 at pH 9.5 and 25 °C. Both rate constants are large enough for the reactions to be diffusion-limited processes. Hydrogenase catalyzes the reversible reaction:

H2 S 2H+ + 2e-

(1)

in the presence of suitable electron carriers. Ferredoxin, flavodoxin, cytochrome c3, and NAD are known to be the natural * Corresponding authors: (fax) +81-75-753-6128, (e-mail) (K. K.) kkano@ kais.kyoto-u.ac.jp and (T. I.) [email protected]. 10.1021/ac981003l CCC: $18.00 Published on Web 03/27/1999

© 1999 American Chemical Society

electron carriers, while viologen dyes function as artificial electron carriers and are employed in the routine enzyme assay of measuring either the H2 evolution or consumption reactions.1 Electrochemical methods based on enzymatic electrocatalysis have proven suitable for not only assaying the enzyme activity2 but also kinetic measurements of the hydrogenase reaction.3-10 Boivin and Bourdillon have measured a catalytic current for the methyl viologen mediated H2 evolution reaction at the glassy carbon electrode modified with immobilized hydrogenase.3 Hoogvliet et al. have reported catalytic currents at a bare carbon electrode for the methyl viologen mediated H2 evolution in the presence of hydrogenase in solution.6,7 Similar catalytic currents have been observed with the natural electron carriers, cytochrome c3 and cytochrome c553, for the H2 consumption reactions4,5,8,10 and with cytochrome c3 for the H2 evolution reaction.9 Hydrogenase exists in a wide range of both bacterial and algal species, and functions in vivo to either evolve or consume H2.1 We present here the first report of the kinetic measurement of bacterial in vivo hydrogenase reaction with an artificial electron carrier, using an electrochemical method based on bacterial electrocatalysis,11-15 a phenomenon similar to enzymatic electrocatalysis. We use Desulfovibrio vulgaris (Hildenborough), a gramnegative sulfate-reducing bacterium, which has at least three (1) Adams, M. W. W.; Mortenson, L. E.; Chen, J.-S. Biochim. Biophys. Acta 1981, 594, 105-176. (2) Kimura, K.; Inokuchi, H.; Yagi, T. Chem Lett. 1972, 693-697. (3) Boivin, P.; Bourdillon, C. Biochem. Biophys. Res. Commun. 1986, 135, 928933. (4) Haladjian, J.; Bianco, P.; Guerlesquin, F.; Bruschi, M. Biochem. Biophys. Res. Commun. 1987, 147, 1289-1294. (5) Nivie`re, V.; Hatchikian, E. C.; Bianco, P.; Haladjian, J. Biochim. Biphys. Acta 1988, 935, 34-40. (6) Hoogvliet, J. C.; Lievense, L. C.; van Dijk, C.; Veeger, C. Eur. J. Biochem. 1988, 174, 273-280. (7) Hoogvliet, J. C.; Lievense, L. C.; van Dijk, C.; Veeger, C. Eur. J. Biochem. 1988, 174, 281-285. (8) Bianco, P.; Haladjian, J.; Brushi, M.; Guerlesquin, F. Biochem. Biophys. Res. Commun. 1992, 189, 633-639. (9) Moreno, C.; Franco, R.; Moura, I.; Le Gall, J.; Moura, J. J. G. Eur. J. Biochem. 1993, 217, 981-989. (10) Verhagen, M. F. J. M.; Wolbert, R. B. G.; Hagen, W. R. Eur. J. Biochem. 1994, 221, 821-829. (11) Takayama, K.; Kurosaki, T.; Ikeda, T. J. Electroanal. Chem. 1993, 356, 295301. (12) Takayama, K.; Kurosaki, T.; Ikeda, T.; Nagasawa, T. J. Electroanal. Chem. 1995, 381, 47-53. (13) Ikeda, T.; Kato, K.; Maeda, M.; Tatsumi, H.; Kano, K.; Matsushita, K. J. Electroanal. Chem. 1997, 430, 197-204. (14) Ikeda, T.; Kato, K.; Tatsumi, H.; Kano, K. J. Electroanal. Chem. 1997, 440, 265-269. (15) Ikeda, T.; Matsubara, H.; Kato, K.; Iswantini, D.; Kano, K.; Yamada, M. J. Electroanal. Chem. 1998, 449, 219-224.

Analytical Chemistry, Vol. 71, No. 9, May 1, 1999 1753

distinct hydrogenases, one in the periplasmic space1,16-18 and the other two in the cytoplasmic membrane;19-21 the bacterium possesses the ability to utilize H2 as an initial electron donor to a complex electron-transfer chain which takes sulfate as the terminal electron acceptor.1 We have made a preliminary observation22 that methyl viologen shuttles between hydrogenase in the bacterial cells and an electrode to act as an electron-transfer mediator in producing a catalytic current for the H2 evolution reaction. In this paper, we measure cyclic voltammograms attributable to the methyl viologen mediated hydrogenase reaction, both for the evolution and the consumption of H2, and interpret the voltammograms on the basis of the theory of catalytic current.23-25 The reversible nature of the bacterial hydrogenase reaction with methyl viologen is confirmed by potentiometry, and kinetic parameters for the H2 evolution and consumption reactions are determined from the limiting catalytic currents by an equation applicable to a nonlinear enzymatic electrocatalysis.26 EXPERIMENTAL SECTION Bacterial Cells and Chemicals. Desulfovibrio vulgaris (Hildenborough) (D. vulgaris (H)) was kindly provided by Dr. T. Kakiuchi. The bacterium was grown in 50 mL anaerobic vials at 37 °C on Postgate’s medium C27 with a slight modification: the medium contains 0.1 g L-1 of ascorbic acid and 0.01 g L-1 of sodium dithionite. The cells were harvested in the late logarithmic phase by centrifugation at 2000 × g for 20 min and washed with a saline solution (0.85% NaCl); then the cells were suspended in a saline solution, and the suspension was kept at -30 °C when not in use. Cell population in the suspension was evaluated as an absorbance at 610 nm (A610) of the suspension; unit absorbance corresponded to 2.9 × 108 cells/mL, which was separately determined by the use of a hemacytometer. The cell number can also be expressed as 4.8 × 10-13 mol L-1 (M). Methyl viologen dichloride trihydrate was purchased from Nacalai Tesque Co., and other chemicals were from Wako Pure Chemical Industries. All chemicals were of reagent grade and were used as received. Preparation of An Electrode Modified with D. vulgaris (H). A method of immobilization of bacterial cells on electrodes (16) Huynh, B. H.; Czechowski, M. H.; Kru ¨ ger, H.-J.; Der Vartanian, D. V.; Peck, H. D., Jr.; Le Gall, J. Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 3728-3732. (17) Patil, D. P.; G. Moura, J. J.; He, S. H.; Teixeira, M.; Prickril, B. C.; Der Vartanian, D. V.; Peck, H. D., Jr.; Le Gall, J.; Huynh, B. H. J. Biol. Chem. 1988, 263, 18732-18738. (18) Pierik, A. J.; Hagen, W. R.; Redeker, J. S.; Wolbert, R. B. G.; Boersma, M.; Verhagen, M. F. J. M.; Grande, H. J.; Veeger, C.; Mutsaers, P. H. A.; Sands, R. H.; Dunham, W. R. Eur. J. Biochem. 1992, 209, 63-72. (19) Lissolo, T.; Choi, E. S.; Le Gall, J.; Peck, H. D., Jr. Biochem. Biophys. Res. Commun. 1986, 139, 701-708. (20) Gow, L. A.; Pankhania, I. P.; Ballantine, S. P.; Boxer, D. V.; Hamilton, W. A. Biochim. Biophys. Acta 1986, 851, 57-64. (21) Romao, C. V.; Pereira, I. A. C.; Xavier, A. V.; Le Gall, J.; Teixeira, M. Biochem. Biophys. Res. Commun. 1997, 240, 75-79. (22) Ikeda, T.; Takagi, K.; Tatsumi, H.; Kano, K. Chem. Lett. 1997, 5-6. (23) Bard, A. J.; Faulkner, L. R. Electrochemical Methods, Fundamentals and Applications; Wiley: New York, 1980; p 457. (24) Coury, L. A., Jr.; Oliver, B. N.; Egekeze, J. O.; Sosnoff, C. S.; Brumfield, J. C.; Buck, R. P.; Murray, R. W. Anal. Chem. 1990, 62, 452-458. (25) Ogino, Y.; Takagi, K.; Kano, K.; Ikeda, T. J. Electroanal. Chem. 1995, 396, 517-524. (26) Kano, K.; Ohgaru, T.; Nakase, H.; Ikeda, T. Chem. Lett. 1996, 439-440. (27) Postgate, J. R. The Sulphate-Reducing Bacteria (2nd ed.); Cambridge University Press: Cambridge, 1984.

1754 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999

has been described.11 Thus, a 10-µL aliquot of the D. vulgaris (H) cell suspension (A610 ) 54) was dropped onto the surface of a glassy carbon electrode (Bioanalytical System Inc., No.11-2012). After the solvent had evaporated, the electrode surface was covered with a polycarbonate membrane (TOYO Filter Co., No. K040A025A), and the whole electrode was covered with a nylon net to fix the bacterial cells on the electrode. Procedures. Electrochemical measurements were carried out with a BAS CV-50W voltammetric analyzer. The D. vulgaris (H) modified electrode or a bare glassy carbon electrode was used as the working electrode. A Pt-disk electrode and an Ag/AgCl/ saturated KCl electrode were used as the counter and reference electrodes, respectively. All potentials are referred to the Ag/ AgCl/saturated KCl electrode (+0.197 V vs NHE). Cyclic voltammograms of methyl viologen at the D. vulgaris (H) modified electrode were recorded in buffer solutions of various pHs; the solutions were saturated with H2 gas, and the measurements were done under the flow of H2 gas over the solutions. Cyclic voltammograms of methyl viologen in the suspensions of D. vulgaris (H) cells were recorded with a bare glassy carbon electrode. Sedimentation of the bacterial cells in the suspensions was not observed during the course of the electrochemical measurements. Potentiometry was carried out with the D. vulgaris (H) modified electrode in solutions containing methyl viologen at various pHs under H2 gas saturated conditions; the measurements were made with an Iwatsu digital multimeter SC-7401 (Iwasaki Electric Co.). All the measurements were carried out at 25 °C. Digital Simulations. The cyclic voltammogram of methyl viologen at a bare glassy carbon electrode in the D. vulgaris (H) cell suspension was analyzed theoretically by normal explicit finite difference digital simulation coupled with a nonlinear least-squares analysis method using a program described in the literature28 with a modification to incorporate the nonlinear character of the catalytic H2 evolution reaction. RESULTS AND DISCUSSION Cyclic voltammograms of methyl viologen at the D. vulgaris (H) modified electrode. In a previous paper,22 we have demonstrated that the D. vulgaris (H) modified electrode produces a methyl viologen mediated catalytic cathodic current for the evolution of H2 in deaerated buffer solutions. Here, we have measured cyclic voltammograms in solutions saturated with H2 gas. Figure 1A shows cyclic voltammograms recorded with the D. vulgaris (H) modified electrode in a H2 saturated buffer (pH 7.5) in the absence (a) and presence (b) of 0.10 mM methyl viologen at the potential scan rate v ) 5 mV s-1. In the absence of methyl viologen, no appreciable waves are observed on the voltammogram, indicating that D. vulgaris (H) cells themselves do not produce any redox waves. When methyl viologen is added to the solution, an S-shaped wave appears with both cathodic and anodic steady-state currents (curve b). This is contrasted with the wave of methyl viologen in Figure 1B, which was taken at a bare glassy carbon electrode in the same buffer saturated with H2. The wave has a shape typical of a diffusion-controlled redox reaction (28) Kano, K.; Sugimoto, T.; Misaki, Y.; Enoki, T.; Hatayama, H.; Oka, H.; Hosotani, Y.; Yoshida, Z. J. Phys. Chem. 1995, 98, 251-258.

Figure 2. Dependence of (+) Ei ) 0 and (n) Eeq on pH. Solid line is the theoretical line for the standard redox potential of the reversiblehydrogen electrode. Figure 1. (A) Cyclic voltammograms recorded with the D. vulgaris (H)-modified electrode in H2 saturated pH 7.5 buffer in the absence (a) and presence (b) of 0.10-mM methyl viologen. Scan rate: 5 mV s-1. (B) Cyclic voltammograms recorded with a bare glassy carbon electrode under the same conditions as (A).

of methyl viologen written as

MV2+ + e- S MV•+

(2)

where MV2+ and MV•+ stand for methyl viologen and its cation radical, respectively. The results confirm that methyl viologen does not directly react with hydrogen and that D. vulgaris (H) cells catalyze both the evolution and the consumption of H2, with methyl viologen as an electron carrier, to produce the cathodic and anodic limiting currents. The fact that both the cathodic and anodic catalytic currents are observed is significant in view of the thermodynamics of mediated bioelectrocatalysis.29 The direction of an overall redox reaction between a mediator and a substrate is generally governed by the difference in the redox potential between the two redox couples. If the oxidation of a substrate to a product by a mediator is thermodynamically favorable, the reduction of the product by the reduced form of the mediator is thermodynamically unfavorable. Consequently, it is difficult for the reverse reaction to produce a catalytic current of significant magnitude.29 Figure 1A clearly shows that this is not the case; the cathodic and anodic catalytic currents are almost the same in magnitude. Such a novel two-way bioelectrocatalysis is allowed only when the redox potential of a mediator is close to that of a substrate and when the redox reaction is very rapid in both directions. The formal redox potential of the MV2+/MV•+ couple, -0.651 V, determined from Figure 1B is, in fact, close to the standard redox potential of the H+/H2 couple at pH 7.5, which is calculated to be -0.641 V. pH-Dependence of the Potential at Null Current, Ei ) 0. From Figure 1A, the potential at null current, Ei ) 0, is determined to be -0.641 ( 0.002 V by taking the average of the potentials obtained from the negative-going and positive-going voltammograms. This agrees with the redox potential, -0.641 V, of the H+/ H2 couple at this pH. We have measured similar cyclic voltammograms at several pHs in the range between 6.5 and 8.5 and (29) Takagi, K.; Kano, K.; Ikeda, T. J. Electroanal. Chem. 1998, 445, 211-219.

obtained the Ei ) 0 value at each pH in a similar manner. The result is plotted against the solution pH in Figure 2 (depicted by +). It is evident that the Ei ) 0 vs pH plot satisfies the relation Ei ) 0/ mV ) -197 - 59.2 × pH, the relationship expected for the reversible-hydrogen electrode. Actually, however, the redox species communicating with the D. vulgaris (H) modified electrode is methyl viologen. Accordingly, we may consider that the redox state of methyl viologen in the vicinity of the electrode surface at null current is in practical equilibrium with the redox state of the H+/H2 couple in the solution. This implies that the D. vulgaris catalyzed hydrogenase reaction is rapid enough to reach the equilibrium state in the immobilized D. vulgaris (H) layer. To verify the above argument, we have performed potentiometry with the same D. vulgaris (H) modified electrode to measure an equilibrium potential, Eeq, at each pH. The potentiometry gave a stable potential reading within 30 min after the addition of methyl viologen to H2 saturated buffer; the time is obviously too short for the two redox couples, MV2+/MV•+ and H+/H2, to reach an equilibrium over the whole solution. This fact strongly supports the idea of rapid local equilibrium between the two redox couples in the vicinity of the electrode surface in the immobilized layer of D. vulgaris (H), as mentioned above. Thus, the stable potential reading gives the Eeq value in the state of the local equilibrium. The Eeq data denoted by E in Figure 2 lie on the line given by Eeq/mV ) -197 - 59.2 × pH, in agreement with the Ei ) 0 data. Interpretation of the Two-Way Bioelectrocatalytic Voltammograms. We consider the shape of the voltammogram for the mediated catalytic reaction. The current-potential curves after being corrected for the base currents are shown in Figure 3 at pH 6.5 (b), 7.5 (2), and 8.5 (9). This figure shows that the steadystate current-potential curve is shifted upward with the increase in pH. To interpret the current-potential curve, we must consider a mechanism that takes into account the two-way catalytic reactions with nonlinear characteristics, which is inherent in enzyme kinetics. For the sake of simplicity, however, we assume here a linear enzyme reaction, which is satisfied when the concentration of substrate, here methyl viologen, is sufficiently low compared with its Michaelis constant and when the concentration of hydrogen ion and the pressure of H2 are kept constant during the measurement. Then, the rate of the reaction, vB, between methyl viologen and hydrogenase in the bacterial cells Analytical Chemistry, Vol. 71, No. 9, May 1, 1999

1755

redox potential of methyl viologen, [MV2+]Bo and [MV•+]Bo are the concentrations of MV2+ and MV•+ at x ) 0, and R, T, and F have their usual meanings. We may assume that the reaction layer is sufficiently thin compared with the D. vulgaris (H) layer because of the very high hydrogenase activity of the bacterial cells mentioned above. Thus, with increasing x in the D. vulgaris (H) layer, the [MV2+]B/[MV•+]B ratio will approach the value in the redox equilibrium with the H+/H2 couple in solution as written by

EoMV + (RT/F) ln([MV2+]B*/[MV•+]B*) ) EoH2 + (RT/F)ln[H+] (6) Figure 3. Base current-corrected voltammograms obtained with the D. vulgaris (H) modified electrode in H2 saturated buffer containing 0.10 mM methyl viologen at pH 6.5 (B), 7.5 (2), and 8.5 (9). Solid lines are the regression curves drawn by eq 7 with the parameter values given in the text.

proceeding in the D. vulgaris (H) layer on the electrode surface may be written by

vB ) k′MV•+zox[B][MV•+]B - k′MV2+ zred[B][MV2+]B (3) where [B] is the concentration of the bacterial cells zox and zred, respectively, are the number of hydrogenases in the oxidized and reduced states in a D. vulgaris (H) cell, the sum z ) zox+ zred being the total number of active hydrogenases per cell, [MV•+]B and [MV2+]B are the concentrations of MV•+ and MV2+, respectively, in the D. vulgaris (H) layer, k′MV•+ is the apparent rate constant of the reaction between MV•+ and oxidized hydrogenase for the H2 evolution reaction, and k′MV2+ is the apparent rate constant of the reaction between MV2+ and reduced hydrogenase for the H2 consumption reaction. Relative to the latter, “apparent” means that the constants may be functions of solution pH30 and reflect the effect of permeability of the bacterial outer membrane and/or the equilibrium distribution.31,32 Methyl viologen undergoes an electron-transfer reaction with the D. vulgaris (H) modified electrode as written by eq 2. In the steady state, the diffusion of methyl viologen coupled with the chemical reaction in the D. vulgaris (H) layer can be described by

d[MV•+]B/dt ) D(d2[MV•+]B/dx2) - vB ) 0

(4)

where D is the diffusion coefficient of MV•+. At the electrode surface (at x ) 0), the concentration ratio [MV2+]B/[MV•+]B is determined by the electrode potential as given by

E ) EoMV + (RT/F) ln([MV2+]Bo/[MV•+]Bo)

(5)

in case of a reversible electrode reaction, where EoMV is the formal (30) Segel, I. H. Enzyme Kinetics; Wiley: New York, 1993; p 608. (31) Ikeda, T.; Kurosaki, T.; Takayama, K.; Kano, K.; Miki, K. Anal. Chem. 1996, 68, 192-198. (32) Ikeda, T.; Kano, K.; Kato, K. Electrochem. Soc., Proc. 1996, 96, 274-283.

1756 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999

where [MV2+]B* and [MV•+]B* are the concentrations of MV2+ and MV•+ in equilibrium with those of the redox couple H+/H2 in solution and EoH2 is the standard potential of hydrogen. Equation 4 can be solved with the boundary conditions, eq 5 and eq 6, under the condition [MV2+]Bo + [MV•+]Bo ) [MV2+]B + [MV•+]B ) [MV2+]B* + [MV•+]B* ) cMV, in which cMV is the total concentration of methyl viologen added to the solution. Finally, the catalytic current, I, which is given by I/nFA ) D[d[MV•+]B/dx]x ) 0, can be expressed as

I ) nFA{D(zoxk′MV•+ + zredk′MV2+)[B]}1/2cMV {1/(1 + K) 1/(1 + η)} (7) where n and A are the number of electrons and the electrode surface area, respectively, η and K are defined by η ) [MV2+]Bo/ [MV•+]Bo ) exp[F(E - EoMV)/RT] and K ) [MV2+]B*/[MV•+]B* ) exp[F(EoH2 - 59.2 × pH - EoMV)/RT], respectively. The expressions for the cathodic limiting current, Ic,l, and anodic limiting current, Ia,l, are easily derived from eq 7:

Ic,l ) -nFA{D(zoxk′MV•+ + zredk′MV2+)[B]}1/2cMV K/(1+ K) (8) 1/2

Ia,l ) nFA{D(zoxk′MV•+ + zredk′MV2+)[B]}

cMV/(1+ K) (9)

The solid lines in Figure 3 are the regression curves drawn by eq 7 with n ) 1, FA(D)1/2 ) 22.4 µC s-1/2 mM-1 (from chronoamperometry of MV2+; D value of MV•+ is assumed to be equal to that of MV2+), EoMV ) -0.651 V (from cyclic voltammetry), cMV ) 0.10 mM, [H+] ) 3.16 × 10-7 M (curve a), 3.16 × 10-8 M (curve b), and 3.16 × 10-9 M (curve c). The unknown quantity, (zoxk′MV•+ + zredk′MV2+)[B], is used as an adjustable parameter to get best fit with the experimental data points (zoxk′MV•+ + zredk′MV2+)[B] ) 0.36, 0.26, and 0.26 s-1 at pH 6.5, 7.5, and 8.5, respectively. The regression curves reproduce the experimental data well despite the assumptions (eq 3) made in deriving eq 7. The pH dependence of the parameter values may be explained from the pH dependence of the ratio zox/zred. We will discuss this problem later. The quantity [D/{(zoxk′MV•+ + zredk′MV2+)[B]}]1/2 corresponds to the thickness of the reaction layer and is calculated to be 55-65 µm, which is sufficiently thinner than the D. vulgaris (H) layer that can be estimated to be (1.4-1.7) × 102 µm on the basis of the volume of immobilized D.

Figure 4. Base current-corrected cyclic voltammogram (broken line) obtained with a bare glassy carbon electrode in the suspension of 2.4 × 10-13 M D. vulgaris containing 0.25 mM methyl viologen at pH 5.5 under argon-saturated conditions. Scan rate: v ) 5 mV s-1. Solid line is obtained by the digital simulation of the voltammogram with the parameter values given in the text.

vulgaris (H) cells. This calculation ascertains that the assumption made in deriving eq 7, local redox equilibrium exists in the D. vulgaris (H) layer, is reasonable. Catalytic Currents for the Evolution of H2 Produced by the D. vulgaris (H) Cells Suspended in Solution. Figure 4 illustrates the base current-corrected cyclic voltammogram (given by the broken line) obtained with a bare glassy carbon electrode in the suspension of D. vulgaris (H) containing methyl viologen at 0.25 mM and at pH 5.5 under argon-saturated conditions. The S-shaped voltammogram with a cathodic limiting current indicates that the electrocatalytic reaction for H2 evolution occurs with the D. vulgaris (H) cells suspended in solution. This is significant in view of the fact that the bacterial concentration in the vicinity of the electrode surface must be much lower than that in the immobilized layer at the D. vulgaris (H) modified electrode. The appearance of only a cathodic limiting current indicates that the D. vulgaris (H) cells catalyze the reaction predominately in the direction of H2 evolution. This is understandable if there is a relatively large difference between the redox potentials of MV2+/ MV•+ and H+/H2. The formal redox potential of MV2+/MV•+ couple is -0.651 V, and that of the H+/H2 couple at this pH and under atmospheric pressure of H2 is -0.523 V; the potential difference, -128 mV, strongly favors the direction of H2 evolution. Since the measurement has been done under argon-saturated conditions, the actual potential difference should be much larger; thus, the rate of the backward uphill reaction is reasonably ignored.29 Voltammograms obtained under such conditions are allowed to be analyzed on the basis of either the theory of a catalytic current23-25 under the conditions of linear enzyme kinetics with respect to a mediator or of a modified theory26 under nonlinear conditions. Figure 5 shows linear-sweep voltammograms for the methyl viologen mediated catalytic H2 evolution reaction at v ) 5 mV s-1 and at cMV ) 5.0 µM to 2.1 mM in the suspension of D. vulgaris (H) cells. At lower cMV, lower than 0.25 mM, the limiting currents are in steady states, indicating that the currents are governed by the catalytic process. At higher cMV, the diffusional effect on the limiting currents becomes significant, but the currents approach steady-state when measured at slower scan rates, that is, at increased times of measurement as shown in the inset of Figure

Figure 5. Linear-sweep voltammograms of methyl viologen at a bare glassy carbon electrode in the suspension of 8.2 × 10-13 M D. vulgaris containing methyl viologen at the concentrations (A) 30, (B) 60, (C) 130, (D) 250, (E) 370, (F) 500, (G) 740, (H) 990, (I) 1500, and (J) 2100 µM. pH ) 5.5. Scan rate: v ) 5 mV s-1. Inset shows the scan-rate dependence of the voltammogram at 0.37 mM methyl viologen. v ) (a) 1, (b) 2, (c) 5, and (d) 10 mV s-1.

Figure 6. Dependence of the steady-state limiting current on the concentration of methyl viologen, cMV. Solid lines are the regression curves with the kcat,H+ and K′MV•+ values given in the text.

5. To obtain steady-state limiting currents over the high cMV, we performed chronoamperometry at -0.80 V. Figure 6 plots the steady-state current against cMV and clearly shows that the plot bends downward with increasing cMV; a linear relationship is satisfied only at low cMV as illustrated in the inset. This nonlinearity is inherent in the enzyme kinetics when measured in a wide range of the substrate concentration. We have taken this into account in deriving an equation of the current for a mediated enzyme electrocatalysis,26 which can be expressed, in the present case, by

I ) FA

x

{

2Dzkcat,H+[B] c / 1+ K′MV•+ + cMV/2 MV cMV/K′MV•+

0.848(cMV/K′MV•+)2 + 11.4(cMV/K′MV•+) + 17.9

}

(10)

where kcat,H+ is the catalytic constant, that is, the turnover number Analytical Chemistry, Vol. 71, No. 9, May 1, 1999

1757

of hydrogenase33 and K′MV•+ is the apparent Michaelis constant for MV•+, in which “apparent” means that the constant may reflect the effect of permeability of the bacterial outer membrane and/ or the equilibrium distribution.30,31 It is easily seen that when cMV/ K′MV•+ , 1, eq 10 is reduced to an equation of the steady-state catalytic current in linear enzyme kinetics.23-25 The two parameters, zkcat,H+ and K′MV•+, were determined by adjusting these parameter values to give the best fit to the plot in Figure 6 by the nonlinear least-squares method with the experimental values FA(D)1/2 ) 22.4 µC s-1/2 mM-1 (from chronoamperometry) and [B] ) 8.2 × 10-13 M. The evaluated zkcat,H+ and K′MV•+ values are 1.2 × 107 s-1 and 0.16 mM, respectively, and the solid lines in Figure 6 are the regression lines with these zkcat,H+ and K′MV•+ values. We also tried to simulate the cyclic voltammogram in Figure 4, using the zkcat,H+ and K′MV•+ values together with the experimental values of A ) 7.9 × 10-2 cm2, D ) 8.6 × 10-6 cm2 s-1,34 cMV ) 0.25 mM, [B] ) 2.4 × 10-13 M, EoMV ) -0.651 V, v ) 5 mV s-1 and the standard rate constant, ko, of the electrode reaction of methyl viologen as an adjustable parameter with an assumption that the transfer coefficient R ) 0.5. The ko value of 3.0 × 10-3 cm s-1 gave the best fit to the experimental data, and is very close to that ((3-6) × 10-3 cm s-1) obtained separately from the voltammogram of methyl viologen in the absence of D. vulgaris (H) cells.35 The regression curve is drawn by the solid line in Figure 4, reproducing the experimental voltammogram well. In a similar manner, we have analyzed anodic steady-state limiting currents obtained at pH 9.5 under H2 saturated conditions. This alkaline pH has been chosen to minimize the effect of the H2 evolution reaction; at this pH, the difference of the standard redox potentials between the MV2+/MV•+ and H+/H2 couples is 108 mV, strongly favoring the direction of H2 consumption. In fact, we observed only an anodic limiting current on the voltammogram of methyl viologen at pH 9.5 under H2 saturated conditions. One problem in the kinetic measurements of the H2 consumption reaction is that MV•+ prevails in the bulk of the test solution during the electrochemical measurements and it dimerizes at higher concentrations, the dimerization constant being determined to be about 2 mM.36,37 To avoid ambiguity in the actual concentration of MV•+, due to the dimerization reaction, we have made the measurements at low cMV, lower than 0.05 mM. Accordingly, we analyzed only a linear part of the dependence of the catalytic current on cMV, which resulted in the apparent bimolecular reaction rate constant zkcat,H2/K′MV2+ ) 2.2 × 1010 M-1 s-1 for the H2 consumption reaction, where kcat,H2 and K′MV2+ are the catalytic constant and the apparent Michaelis constant for MV2+, respectively, for the reaction between methyl viologen and hydrogenase in the bacterial cells. The apparent bimolecular reaction rate constant for the H2 evolution reaction, zkcat,H+/K′MV•+, (33) The quantity zkcat,H+ may be regarded as the turnover number of a single D. vulgaris (H) cell and the quantity zkcat,H+/K′MV•+ as the apparent bimolecular rate constant of the reaction between the cell and MV•+. These are useful quantities in estimating catalytic power of a suspension containing a given amount of bacterial cells. (34) Steckhan, E.; Kuwana, T. Ber. Bunsen-Ges. Phys. Chem. 1974, 78, 253259. (35) We also attempted to adjust the values of zkcat,H+ and K′MV•+ by a nonlinear least-squares method. However, stable convergence was not attained most probably as a result of the statistical correlation between zkcat,H+ and K′MV•+. (36) Kosower, E. M.; Cotter, J. L. J. Am. Chem. Soc. 1964, 86, 5524. (37) Thorneley, R. N. F. Biochim. Biophys. Acta 1974, 333, 487-496.

1758 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999

is calculated as 7.5 × 1010 M-1 s-1 from the zkcat,H+ and K′MV•+ values obtained above. Hydrogenase Activity of D. vulgaris (H) Cells. We have observed the voltammogram due to the methyl viologen mediated D. vulgaris (H) catalyzed evolution and consumption of H2. This is the first clear demonstration of the two-way bioelectrocatalysis with a single mediator compound in the studies of bioelectrocatalysis, though we have reported a similar two-way bioelectrocatalysis using a combination of two kinds of enzymes.29 Similar catalytic currents have been reported in the direct electrochemical measurements of hydrogenases isolated from Thiocapsa roseopersicina38 and Megasphaera elsdenii.39 The potential at null current on the voltammogram agrees with the equilibrium potential determined by potentiometry with the same D. vulgaris (H) electrode, which is equal to the potential of the H+/H2 couple in solution. The novel character of the electrode is ascribed to the very high hydrogenase activity of D. vulgaris (H) cells. D. vulgaris (H) contains at least three types of hydrogenase: periplasmic Fe hydrogenase16-18 and two types of membranebound Ni-containing hydrogenases.19-21 Among them, the periplasmic Fe hydrogenase has a very high H2 evolution activity, 6900 U mg-1,18 and may be assigned as the hydrogenase functioning in the present system. We have demonstrated a number of examples of rapid-electron-transfer communication between periplasmic enzymes in bacterial cells and an electrode.11-15,31,32 The Michaelis constant for methyl viologen of the isolated Fe hydrogenase has been reported to be 0.2 mM at pH 8 and at 30 °C.40 Our K′MV•+ value of 0.16 mM at pH 5.5 and 25 °C is close to the Michaelis constant of Fe hydrogenase. The turnover number of Fe hydrogenase for H2 evolution is calculated to be 6.8 × 103 s-1 from the reported activity of 6900 U mg-1 at pH 8.0 and 30 °C18 and the molecular weight of the enzyme, 59 500.41 Our zkcat,H+ value of 1.2 × 107 s-1 is 1.8 × 103 times larger than the calculated turnover number. If we assume that the intact hydrogenase in the D. vulgaris (H) cell has the same activity as the isolated Fe hydrogenase, then z is calculated to be 1.8 × 103. This amount of the enzyme is calculated to be only about 0.09% of the dry weight of a single bacterial cell (∼ 0.2 pg42), which is a conceivable amount. The ratio kcat,H+/K′MV•+ gives the bimolecular reaction rate constant for the reaction between MV•+ and hydrogenase and can be calculated as 4.2 × 107 M-1 s-1 with z ) 1.8 × 103. The bimolecular reaction rate constant for the reaction between MV•+ and Fe hydrogenase isolated from D. vulgaris (H) has been reported to be (1.7-5.0) × 106 M-1 s-1 at pH 7.5 and 20 °C.6,7 The authors obtained these values from the analysis of the catalytic currents for the methyl viologen mediated hydrogenasecatalyzed H2 evolution reaction, using the theory of a catalytic current applicable to the case of linear kinetics.23 However, the concentration of methyl viologen employed, 0.25 mM, is higher (38) Varfolomeyev, S. D.; Yaropolov, A. I.; Karyakin, A. A. J. Biotechnol. 1993, 27, 331-339. (39) Butt, J. N.; Filipiak, M.; Hagen, W. R. Eur. J. Biochem. 1997, 245, 116122. (40) Van der Westen, H. M.; Mayhew, S. G.; Veeger, C. FEBS Lett. 1978, 86, 122-126. (41) Hagen, W. R.; van Berkel-Arts, A.; Kruse-Wolters, K. M.; Voordouw, G.; Veeger, C. FEBS Lett. 1986, 203, 59-63. (42) Stainier, R. Y.; Ingraham, J. L.; Wheelis, M. L.; Painter, P. R. The Microbial World; Prentice-Hall: New Jersey, 1986; p 186.

than the value of the Michaelis constant; consequently, the analysis would lead to underestimation of the rate constant. If we tentatively multiply the reported value by a factor of 10, the rate constant of the reaction between MV•+ and Fe hydrogenase has the value of (1.7-5.0) × 107 M-1 s-1, which is consistent with our kcat,H+/K′MV•+ value. In the same way, we also obtain the bimolecular reaction rate constant, kcat,H2/K′MV2+, 1.2 × 107 M-1 s-1, for the H2 consumption reaction. As far as we are aware, there is no data on the rate constant for the H2 consumption reaction of Fe hydrogenase. The kcat,H+/K′MV•+ and kcat,H2/K′MV2+ values are close to each other and are within an order of magnitude of the diffusion-limited rate constant in enzyme-catalyzed reactions,43,44 which is consistent with the observed reversible electrocatalytic behavior. Finally, we will discuss the redox potential of hydrogenase, E°enz, on the basis of the (zoxk′MV•+ + zredk′MV2+)[B] values obtained above: 0.36, 0.26, and 0.26 s-1 at pH 6.5, 7.5, and 8.5, respectively. We assume that the redox ratio of the enzyme (given by Kenz ) zox/zred) is determined by the solution pH as written by Kenz ) zox/zred ) exp[F(E°H2 - 59.2 × pH - E°enz)/RT] and that the rate constants k′MV•+ and k′MV2+, which are equivalent to kcat,H+/K′MV•+

and kcat,H2/K′MV2+, respectively, are independent of solution pH. Then we can calculate E°enz values from the above data by knowing the value of [B]. Using [B] ) 1.1 × 10-11 M, estimated from the immobilized amount of the bacterial cells and the volume of the immobilized layer, we obtain E°enz ) -0.55, -0.55, and -0.61 V from the values at pH 6.5, 7.5, and 8.5, respectively. These E°enz values are unexpectedly positive compared with the redox potential of methyl viologen, EoMV ) -0.651 V, but are close to the reported value of the midpoint potential of the two F clusters of Fe hydrogenase: -0.330 and -0.340 V vs NHE at pH 7.5 (that is, -0.527 and -0.537 V vs Ag/AgCl).18

(43) Hiromi, K. Kinetics of Fast Enzyme Reactions; John Wiley & Sons: New York, 1979; p 259. (44) Ogino, Y.; Takagi, K.; Kano, K.; Ikeda, T. J. Electroanal. Chem. 1995, 396, 517-524.

Received for review February 10, 1999.

ACKNOWLEDGMENT We thank the reviewer for his valuable suggestions in describing eq 3. We also thank Dr. Takashi Kakiuchi for his kind donation of the bacterial strain of D. vulgaris (Hildenborough) and Dr. Byung-Hong Kim for his helpful advice in anaerobic cultivation of the bacterium during his stay in this laboratory. This research has been supported in part by Grant-in-Aid from the Ministry of Education, Science, Sports, and Culture, Japan. September

9,

1998.

Accepted

AC981003L

Analytical Chemistry, Vol. 71, No. 9, May 1, 1999

1759