Reversible Electrochemistry of Fumarate Reductase Immobilized on an Electrode Surface. Direct...

8/12/2019 Reversible Electrochemistry of Fumarate Reductase Immobilized on an Electrode Surface. Direct Voltammetric Obs

1/11

Biochemistry 1993, 32, 5455-5465 5455

Reversible Electrochemistry of Fu mara te Reductase Immobilized on an ElectrodeSurface. Direct Voltamm etric Observations of Redox Centers and TheirParticipation in Rapid Catalytic Electron T ranspo rt+Artur Suche ta ,*Richard Camm ack,s Joel Weiner, ll and Fraser A . Armstrong*,*

Department of Chemistry , University of California , Iruine, California 9271 7, Division of Biomolecular Sciences,Kings College London, Campden Hill Road, London W8 7AH, United Kingdom, and Department of Biochemistry,University of Alberta, Edmonton, Alberta, Canada T6G 2H7Received December 15, 1992; Revised Manuscript Received February 26, I993

ABSTRACT: Fumarate reductase (Escherichia coli) can be immobilized in an extremely electroactive statea t an electrode, with retention of native catalytic properties. Th e mem brane -extrins ic FrdA B componen tadsorbs to monolayer coverage a t edge-oriented pyrolytic grap hite and catalyzes reduction of fu ma rat e oroxidation of succinate, depending upon the electrode potential. In the absence of substrates, reversibleredox transformations of centers in the enzym e are observed by cyclic voltamm etry. The m ajor componentof the voltamm ogram s is a pair of narrow reduction an d oxidation signals corresponding to a pH -sensitivecouple with formal reduction potential E = -4 8 mV vs SH E a t pH 7 .0 (25 C ) . This is assigned totwo-electron reduction and oxidation of the active-site FAD . A redox couple with Eo = -31 1 m V a t p H7 is assigned to center 2 ( 4Fe-4SI2+/ +) .Voltamm ograms for fumara te reduction a t 25 O C , measured w itha rotating-disk electrode, show high c atalytic activity without t he low-potential switch-off tha t is observedfor the related enzyme succinate dehydrogenase. Th e catalytic electrochemistry is interpreted in terms ofa basic model incorporating mass tra nsp ort of substrate, interfacial electron tran sfer, and intrinsic kineticproperties of the enzyme, each of these becoming a rate-determining factor under certain conditions.Electrochem ical reversibility is appro ached un der conditions of low turnover r ate, for example, as th e supplyof sub strate to th e active site is limited. In this situation, electrocatalytic half-wave potentials, E 1 / 2 , r esimilar for oxidation of bulk su ccinate and reduction of bulk fu ma rat e and coincide closely with the Evalue assigned to the FAD. At 25 C a nd p H 7 , t he a ppa re n t K M or fum arate reduction is 0.16 m M, an dk,,, is 840 s-I. Accordingly the second-order rate constant, kcat/&, is 5.3 X lo6M- s-I. Under the sameconditions, oxidation of succinate is much slower. As th e supply of fum ara te to the enzyme is raised toincrease turnover, the electrochemical reaction eventually becom es limited by the r ate of electron transferfrom the electrode. Und er these conditions a second catalytic wave becomes evident, the value of whichcorresponds to the reduction potential of the redox couple suggested to be center 2. Thi s small boost tothe catalyt ic curren t indicates that the low-potentia l [4Fe-4S] cluster can function as a second center forrelaying electrons to the F AD .

Biological in terconv ersion of fum arate an d su ccina te iscatalyzed by oxidoreductases known as fuma rate red uctase(FRD) andsuccinatedehydrogenase (SD H) (Coleet al., 1985;Ohnishi, 1987; Ackrell et al., 1992 ). These closely relatedenzymes exist as memb rane-bound complexes, respectivelymenaquino1:fumarate oxidoreductase and succinate:ubiquino -ne oxidoreductase (also known as complex 11), that linkoxidation of succinate or reduction of fu ma rate to reductionor oxidation of the quinone pool (Q/QH z) as shown by eq 1.(1)umarate + Q H 2+=uccinate + Q

In Escherichia coli and other bacteria, fum arate can serve asa terminal electron acceptor in anaerobic respiration (HaddockJones, 1977; Ingledew Poole, 1984). The FR D complexis located in the cytoplasmic memb rane and com prises fourThis work was supportedby grants from the Donors of the PetroleumResearch Fund administeredby the American Chemical Society (F.A .A. ),by the National Science Foundation (MCB-9118772) (F.A.A.),and bythe University of California.To whom correspondence should be addressed at the U niversity ofCalifornia.

f Kings College London.University of Alberta.University of California.

0006-2960 I93 10432-5455%04.00/0

nonidentical subunits organized into two domains (A ckrell etal., 1992 ). The four genes encoding these subunits have beencloned and sequenced, and FRD now provides an importantmodel system for studying active-site structure and functionincomplexrespiratoryenzymes (Blau t et al., 1989; Westenberget al., 1990; Schriider et al., 1991; We rth et al., 1990, 1992).The soluble membrane-extrinsiccatalyticdomain of FRDconsists of subunit Frd A (66 kDa), which contains a covalentlybound FAD (Weiner Dickie, 1979; Blaut et al., 1989) andthe site of substrate binding, and subunit FrdB (27 kDa),which contains the iron-sulfur clusters, centers 1 ([2Fe-2SI2+/+), ([4Fe-4SI2+/+),nd 3 ( 3Fe-4SI1+/O)Manodoriet al., 1992). The reduction poten tials of centers 1 and 3 liein the rang e -20 to-70 mV ap propriate for mediating electrontransfer from m enaquinol to fumarate (C amm ack et al., 1984;Simpkin Ingledew, 1985; Cam mac k et al., 1986, 1987;Werth et al., 1990). Conversely, the much more negativereduction potential reported for center 2 (rang e -285 to -320mV ) poses a problem with regard to its role in the functioningof the enzyme (Simpkin Ingledew, 1985; Cam mac k et al.,1986, 1987). The catalytic domain is associated with themem bran e via a mem brane-intrins ic anchor domain con-sisting of subunits FrdC (15 kDa) and FrdD (13 kDa). Theseare essential for reaction with m enaquinone (W einer et al.,

1993 American Chem ical Societv


2/11

5456 Biochemistry, Vol. 32, No. 0, 19931986; Westenberg et al., 1990). However, a soluble enzymeconsisting only of Frd A and FrdB subu nits (FrdA B) can beprepared which is active in fumarate reduction by benzylviologen (Robinson Weiner, 1982;Lemire Weiner, 1986).Recently it was shown that the analogous, soluble two-subunit form of SDH (bovine heart mitochondria) exhibitsdirect unmediated c atalytic electrochemistry when adsorbedat an edge-oriented pyrolytic graph ite electrode (Sucheta e tal., 1992). Such voltammetric methods are proving to haveincreasing applicability for stud ying redox proteins tha t haveunstable, reactive centers (Arm strong, 1990,1992; Armstronget al., 1993a; Butt et al., 199 1, 1993). For an enzyme,moreover, they permit c atalytic performance to be m easuredunder conditions of continuously variable applied potential(Varfolom eev Berezin, 1978; Hill, 1993 ). Detailed infor-mation may thus be obtained on how activity is linked to theenergetics of the catalyzed reaction and to th e redox statusof active sites. From studies with SDH , it was concluded thatthe adsorbed enzyme remains fully functional with regard tosuccinate oxidation. Furthermore, it was found that catalysisin the direction of fum arate reduction (physiologically thereverse process) is controlled by a n interesting potential gatingeffect. Specifically, reductio n of fum ara te is catalyzed onlywithin a narrow region of applied potential, with activity beingswitched off as the electrochem ical driving force is increased.This fully reversible phenomenon has now been observed innonelectrochemical kinetic experiments with SDH fromvarious sources and intact complex I1 (Ackrell et al., 1993).These studies have revealed th at th e rat e of oxidation of benzylviologen radical by fuma rate as catalyzed by SD H increaseswith the form ation (or addition) of th e oxidized product, i.e.,opposite to the profile observed with F RD . Th e physiologicalrelevance of this property is under consideration.In th is paper we describe direct voltamm etric experimentson the FrdA B component of fum arat e reductase. Th e resultsprovide (i) a n unambiguous d emonstration of the capabilityof an enzym e to exhibit reversible electrocatalysis when boundin a near-monolayer film at an electrode surface, (ii) avoltamm etric snapshot of active sites and their participatio nin turnover, and (iii) measure ments of kinetic param eters th atreveal large rate con stants for electron transfer between theelectrode and the enzym e and within the enzyme itself. Resultsare discussed in term s of a generic model capable of wideapplication in fu ture developments.EXPERIMENTAL PROCEDURES

The membrane-extrinsic catalytic dimer (FrdAB) of fu-mar ate reductase was isolated from E. coli HBlOl /pFRD84according to a previously repor ted method (Lemire Weiner,1986). The enzyme was assayed by monitoring oxidation ofreduced benzyl viologen by f um ara te as described by Dickieand Weiner (1979 ). Th e activity of the sample used in thesestudies was 5 1 1 units mg-*. Because of th e air sensitivity ofsoluble FRD , all experiments and preliminary handling of t heenzyme were undertaken in a glovebox under an atmosph ereof argon or nitrogen (02 < lppm). Concentrated enzymesolutions were prepared by diafiltration with an Amicon 8 M C(YM10 membrane), and the concentration was determinedby spectrophotometry using e = 10 mM-I cm-1 a t 460 nm(We iner Dickie, 1979 ). Purified water of resistivity - 8M Q cm w as used throughout.For all electrochemical studies, a supporting electrolyteconsisting of 0.1 M N aC l (Aldrich) 50 m M in an appropriatebuffer was used. These buffers were TA PS (3-[[tris(hy-droxymethy1)methyll amino] propanesulfonic acid), H EP ES

Sucheta et al.N- 2-hydroxyethyl)piperazine-N-2-ethanesulfonic cid), MES(4-morpholineethanesulfonic acid), and PIP ES (1,4-pipera-zinediethanesulfonic acid]), each purchased from S igma. Allbuffers were prepared, and their pH was adjusted by additionof concentrated N aO H or H Cl as needed, at 25 OC (or 0 OCfor low-temperature studies). Polymyxin B sulfate wasobtained from S igma and prepared as a stocksolution of about30 mg/m L with its pH adjusted to 7.0.

For cyclic voltammetry (CV) at stationary and rotatingdisk electrodes, an Ursar Scientific Instruments (Oxford)potentiostat was used. An EG & G M636 electrode rotatorand control unit were used for rotating disk experiments. Theall-glass voltammetric cell was similar to that describedpreviously (Arm strong et al., 198 9a) except that th e samplecompartment was equipped with a circulating jacket forthermostatting. Th e rotating-disk electrode (RD E) was asmall cylinder of pyrolytic graphite (L e Carbone, France) cutso that the ends exposed the edge plane (hence the termPGE for pyrolytic graph ite edge) which was mounted in aTeflon sheath enclosing a brass contact rod. The sheath wasmachined fo r compatibility with th e cell and featur ed a ferru leequipped with a grub screw for rapid attachment to theelectrode rotator. Prior to each experiment, the electrodewas polished with an aqueous slurry of 0.3-pm alumina(Buehler micropolish) and sonicated thoroughly. A Fishersatur ated calomel electrode (SC E) w as used as the reference.All potentials given ar e with reference to the standard hydrogenelectrode (SHE ); our values are based on Eo(S CE) = 0.241V (Bar d Faulkner, 1980). Th e auxilliary electrode was apiece of platinum gauze.All aspects of the rotating-disk system (cell, electrode,potentiostat) were checked by conducting experiments a t 25OC with solutions of Fe(CN )63- containing 0.1 M NaCl and50 m M P IPE S, pH 7. Da ta yielded Levich plots (Levich,1944; Riddiford, 196 6) that were linear up to a rotation r ateof at least 1800 rpm. Tests were performed under conditionsin which the limiting curre nt, ili,,,,exceeded values observedin experiments with fum arat e reductase. It was confirmedthat the current-potential ( i -E) curves for reduction ofFe(CN)63- were symm etrical about the half-wave potentialand gave linear plots (slope 59 mV) of E against lOg((i1im -i ) / i } (Bond, 1980; Bard Faulkner, 1980; SouthamptonElectrochemistry Group, 1985). Levich plots ( ie , vs squareroot of rotation rate) had slopes directly proportional to t hebulk concentration of ferricyanide, with the proportionalitycoefficient given by 0.620 FAD 2/3rr/6 see eq 4 and definitionsgiven below); this relationship was used to determine theelectroactive surface are a of th e electrodes. Values of andD were taken to be 0.01 cm2 sec- and 0.763 X lo- cm2 ec-I,respectively (Bard Faulkner, 1980; Adams, 1969).The electrochemistry of FR D was enhanced and stabilizedin the presence of polymyxin and (t o a lesser extent) neom ycin,which we have previously shown to induce strong adsorptionof ferredoxins at a P G E electrode (Armstrong et al., 1989,1993a; Butt e t al., 1991, 1993). Unless otherwise stated, allexperiments were performed with electrolyte solutions con-taining 0.1-1.0 g L- (0.08-0.75 mM) polymyxin B sulfate(Ha yash i Suzuki, 1965) which was added as an aliquot ofthe concentrated stock solution.

RESULTS AND DISCUSSIONDetection, Optim ization, and a Preliminary Description ofthe Catal ytic Electrochemistry as Measured at a Station aryElectrode. Following introduction of a freshly polishedelectrode into a solution containing fu ma rate and 0.3 gM


3/11

Electrochemistry of F uma rate R eductase Biochemistry, Vol. 32, No. 20, I993 5457enzyme concentration of 0.8 p M and a polymyxin concen-tration of 0.1 g L lo The influence of the applied potentialupon the adsorption process was assessed by poising theimmersed electrode at various potentials for different timesbefore initiating scanning. In this way it was ascertained thata value of -160 mV gave reproducibly good results. PanelsC and D of Figure 1 show voltammograms obtained afterholding the potential at -160 mV for 2 min prior to scanning.For Figure lC , polymyxin is absent, whereas it is present ata concentration of 0.1 g L-' for th e experiment shown in Figure1D. Solution conditions are the same as for Figure 1A ,Bexcept that the enzyme concentration is 0.8 pM. Closeinspection shows that the peak potential is more positive inFigure 1D (-40mV) tha n in Figure 1C (-59 mV ); moreover,the voltammogram was much more stable over successivescans. Th e observations that the response is both sharpenedand stabilized in the presence of polymyxin indicate that itslikely role is either to promote adsorption of active enzymemolecules in such a manner th at denatu ration is retarded (e.g.,by participating in the enzyme-electrode assembly) or toprevent adsorption of inactive mo lecules tha t would otherwiseblock the electrode surface.

Similar results were obtained over the pH range 7-9, butat p H 6 a m uch h igher level of polymyxin, e.g., 1 g L-I, wasrequired to give a stab le response. Neomy cin also was foundto promote enzyme adsorption. In studies with ferredoxins,we have found th at reagents such as polymyxin or neomycininduce formation of films close to monolayer coverage thatare stable for extended periods in the absence of a bulkconcentration of the protein (A rmstrong, 1992; Armstrong etal., 1989b, 1993a; Butt et al., 1991, 1993).

By contrast with the characteristic voltammetry observedfor the case of succinate dehydrogenase (S ucheta et al., 19 92),there is no evidence for potential-controlled gating of thereduction of fuma rate. Th e reduction peak th at is observedupon scanning in the direction of decreasing potential is asexpected for a reaction con trolled by diffusion of a reduciblesubs trate to a planar electrode surface (Nicholson Shain ,1964; Bard Faulkner 1980). That this peak arises fromdepletion of t he diffusion layer by extremely efficient catalysisis demonstrated by the fact that it transforms to a largesigmoidal wave upon rotatin g the electrode. For satisfactoryda ta analysis and interpretation it was thus essential to achievestrict control over mass transport of substrate, and furtherstudies of th e electrocatalytic reaction were carried ou t witha rotating disk electrode.

Demonstration of Electrochemically Near-Reversible Ca-talysis by FR D Using a Rotating Disk Electrode. Figure 2shows RDE voltammograms (i-E curves) o btained a t 25 OCfor solutions of fum arate reductase (0.1 M NaCl and 20 mMP IP ES a t pH 7.0) in th e presence of low levels of su bstrate,respectively 2.6 p M fum arate or 13 pM succinate. In eachcase, scanning was performed a fter inducing optim al activityby poising the potential at -160 mV for 2 min. As the appliedpotential, E , is varied in the direction required to causereduction of succinate or oxidation of fumarate (in this caseat a scan rate u ) of 9.7 mV s-l), the faradaic current, i ,increases, finally reaching a constant value, ili,,,,at which itis clear th at th e electrochemical po tential ceases to be a factorcontrolling the current. Upon changin g the direction ofscanning, the sam e curves are traced w ith the c urren t offsetby the double-layer capacitance (the magnitude of whichdepends directly upon the scan rate). Comp ared to backgroundvoltammograms measured in the absence of substrate, fu-mar ate caused a negative displacement of the cu rrent, while

A C A

400 -200 0 400 -200 0E/mV vs S.H.E.--I

FIGURE: Cyclic voltammograms measured with a stationary PGEelectrode on successive cycles following introduction of the electrodeinto a solution containing 25 pM fumarate and 0.10 M NaCl at 25O C and pH 7.0 (PIPE S). Electrode surface area is 0.09 cm2;scanrate, 9.7mV s-l. (A) Successive cycles in solution containing 0.33pM fumarate reductase (FrdAB) and no polymyxin. (B) Successivecycles in solution containing 0.33pM FrdAB and 0.1 g / L polymyxin.(C) Single cycle following a potential hold for 2 min at -160 mV;0.82 pM FrdAB, nopolymyxin. (D) Singlecycle following a potentialhold for 2 min at -160 mV; 0.82 pM FrdAB, 0.1 g/L polymyxin.FrdA B, a sigmo idal reduction wave develops upon repetitivecycling over the potential range 150 to -550 mV. This isshown in Figure 1A. Th e size of the wave increases as thefum arat e concentration is raised ( in this case the concentrationis 25 pM ut decreases very markedly as the tem peratu re islowered from 25 to 0 OC. N o farad aic response in this regionis observed for a solution of fumarate without the enzyme.These observations ar e consistent with enzymatic catalysis ofthe electrochemically driven reduction of fuma rate . After aperiod of time th e responsedeteriorates, even after brief stirringof the solution to replenish the substr ate close to th e electrode.As shown in Figure lB , if th e same experiment is carried outin the presence of polymyxin, a more rapid and markeddevelopment is observed, with the response changing from asigmoidal wave to a peak. Th e stability of the optimal responseis also considerably grea ter. Increases in the enzyme con-centration (up to 0.8 rM produce a much sharper, peak-likeresponse in a shorter time.If the electrode, at the point of at:aining th e maximalresponse, is rinsed with buffer and then trans ferred to a solutioncontaining all the components except fumarate reductase,electrocatalyticactivity remains detectable for at east 15min.During this time the response reverts to the sigmoidalwaveform; however, the attenua tion occurs more rapidly thanthe deterioration observed when enzyme is present in thesolution. These observations show tha t Fr dA B is reversiblyadsorbed a t the electrode surface. The transition fromsigmoidal to peak-type catal ytic waveform therefore reflectsthe progress of adsorption, with the optimal peak-type responsebeing associated with the highest coverage of active enzymemolecules. Suc h a time dependence resembles the observationmade for electrocatalysis of H202 reduction by cytochromec peroxidase (Armstrong Lannon, 1987; Armstrong et al.,1993b).Several factors were investigated in order to establishconditions for reproducibly obtaining optimal results. W eadopted a working hypothesis to gauge optimization: for agiven subs trate concentration, m aximum coverage of activeenzyme is associated with the most positive peak potentialand highest peak current (Armstrong et al., 1993b). For pH7, it was found that this condition could be achieved with an


4/11

5458 Biochemistry, Vol. 32, No. 20, 1993 Sucheta et al.

-IO - 4 0 -30 -20 - 1 0 0

E /mV vs S . H . E

-4w -2w 0E I m V vs S.H.E.

FIGURE: Cyclic voltammograms demonstrating the reversibility ofelectrochemical turnover of fumarate or succinate by fumaratereductase adsorbed at a rotating PGE disk electrode. Scans shownwere recorded after development of maximum activity at 25 OC. Therespective origins of the current-potential coordinate systems aremarked by a cross. Conditions: 1.2pM FrdAB; 0.46 g/L polymyxin;20 mM PIPE S at pH 7; 0.1 M NaC1; temperature, 25 OC; can rate,9.7 mV S K I For fumarate (2 .6 pM) the electrode rotation rate is 300rpm. For succinate (13 pM) the rotation r ate is 900 rpm. The insertshows plots of Iog(il,m- i)/i V, uccinate oxidation; A umaratereduction) against applied potential, E , for scans in direction ofincreasing potential.succinate caused a positive displacement, th us confirming th eelectrode reactions as reduction and oxidation, respectively.Th e forms of the farad aic component of th e correspondingi-E curves obtained a t a slower scan rate (4.9 mV s-l) wereidentical within experimental error.For reversible electrochemistry with a solution comprisingoxidant 0 under steady-state conditions (as approachedthrough a technique such as RDE voltammetry or polarog-raphy) the relationship between current and potential isgivenby the Heyrovskf-Ilkovie equation (Heyrovskf Kd ta, 1966;Bond, 1980; Bard Faulkner, 1980; Southampton Electro-chemistry Grou p, 1985). In th e specific case of a simple,noncatalytic, diffusion-controlled reaction, this equation canbe written as

(2)Provided reactant 0 and product R have equal diffusioncoefficients, the half-wave potential, E1p coincides withthe form al reduction po tential, EO, of the O / R couple.Equation 2 predicts that a plot of log{(ilim- )/i) against Eshould be a straight line of reciprocal slope 2.3RTlnFintercepting the potential axis at Ell2. The term n in this casereflects the stoichiometry of th e reaction. Th e same value ofE112 s expected if th e experiment is carried out in th e reversedirection, i.e., oxidizing the bulk red uctan t R.The inset to Figure 2 shows the corresponding plots oflog((ili, - )/i) (corrected for background slope) against E.These ar e approximately linear, with slopes of 30- 45 mV perdecade , and yield E1p values that a re in close agreement w itheachother . At pH 7.0 and 25 OC,Elpis -32f Sforfumara tereduction and -26 f 5 mV for succinate oxidation. Forexperiments with fum arate, elevated subs trate concentrationsor higher electrode rotation rates gave increasing distortionof the voltammogram toward the low-potential limit. This

E = E,l2+ (2.3RT/nF) log{(& - )/ i )

A

B

C

currentunit

I , . I . . I I-800 400 0E /mV vs S.H.E.

FIGURE: Voltammograms of a solutioncontaining 0.82 pM FrdAB,0.1 g/ L polymyxin, and 0.10 M NaCl a t pH 7.0 (PIPES) and 25 OC.Electrode surface area is 0.09 cm2. (A) Successive cycles at astationary electrode (scan rate, 50 mV s-I, current unit, 10 PA)following a potential hold for 2 min at -160 mV. The positionsofsignals I, 11, and 111 are indicated. The initial cycle exhibits anadditional broad cu rrent component in the direction of decreasingpotential due to irreversible reduction of electrode surface groups.(B) Voltammogram recorded in the presence of 0.4 pM fumarate astationary electrode following a potential hold a t -160 mV nd briefrotation of the electrode (scan rate, 10mV SKI;current unit, 0.2 PA .(C) Voltammogram recorded in the presence of 0.4 pM fumaratewith an electrode rotation rate of 1200rpm following a potential holdfor 2 min at -160 mV (scan rate, 10 mV s-I; current unit, 0.2 pA ) .observation (which we propose arises as turnover becomeslimited by th e rate of electron transfer to the enzyme fromthe electrode) is discussed later.

We thus conclude that, under low rates of turnover, thevoltammetric response conforms to some of t he basic expec-tations for a reversible electrode reaction. How ever, thr eeobservations reflect the controlling influence of the enzyme.First, the curr ent obtained per concentration unit of succinateis much lower than for fum ara te (n.b., thediffusioncoefficientsof succinate and fum arate should be similar). Second, theE1/2value observed is more negative (by approximately 6 0mV) than the reported reduction potential E O F ~30mV) forthe fumarate-succinate redox couple under similar conditions(Clar k, 1960). Thir d, the n values are below th e value of 2.0expected for equilibrium between fumarate and succinate.

Detection of Reversible, Nonc atalytic Electron E xchangebetween the Electrode and the Redo x Centers of AdsorbedEnzyme Molecules. A cyclic voltammogram of an enzymesolution at 25 OC recorded in th e absence of fum ara te is shownin Figure 3A . Several isolated pairs of oxidation and reductionwaves are ob served following immersion of a freshly p olishedelectrode and poising of the potential initially at -160 mV for2 min prior to cycling. The peaks ar e easily visible above th e


5/11

Electrochemistry of Fumarate ReductaseE /mV vs S.H.E.

-600 -400 -200 0

Biochemistry, Vol. 32, No. 20, 1993 5459A

600 -400 -200 0E /mV vs S.H.E

FIGURE 4: Stationary-electrode cyclic voltammograms of solutionscontaining 0.82 pM FrdAB and 0.10 M NaCl at 0 C. No fumaratewas present in solutions. Electrode surface area is 0.09 cm2; scanrate, 50 mV s - I Voltammograms were recorded after a potentialhold for 2 min at -160 mV. Conditions: pH 6 (6.0), pH 7 7.0), pH8 8.0), and pH 9 (9.1).large but relatively uniform background from double-layercapacitance. T he amplitudes of three pairs of such peaks,which we shall refer to as signals I, 11, and I11 in order ofdecreasing potential, diminish slowly with time coin cidentallywith the appearan ce of a weak, broad featu re in the regionof -1 50 to -200 mV, and on the sam e time scale as the decreasein catalytic activity mentioned above. Signal I is the mostprominent feature, w ith oxidation and reduction peaks beingof equa l size and coinciding in po tential position to within 25mV even at a scan rate of 470 mV s-l. Signal I1 is lesspronounced and noticeably broader a lthough oxidation andreduction peaks have similar shapes and again appear of equalsize. On the other hand, signal I11 clearly arises from acomp licated redox process since the redu ction wave is broaderthan the oxidation wave, and this asymm etry is particularlymarked a t high scan rate. Impo rtantly, the entire voltam-mogram is unchanged upon rotation of the electrode. W ethus conclude that the signals arise from sites in moleculesthat are confined to the electrode surface and for whichdiffusional exchange with bulk solution is negligible duringthe time course of the scan. Th e formal reduction potentialsof signals I and I1are -48 and -3 11 10 mV , respectively,as determined from the average of oxidation and reductionpeak potentials. W e did not attemp t to determine a value forsignal I11because of the ap pare nt complexity of its dependenceon experimental parameters.Addition of fum arate to a final concentration even as lowas 0.4 pM produces marked changes in the voltammetry.Examined with a stationary electrode (as shown in Figure3 B ) , he reduction peak of signal I is enlarged. Little changeoccurs elsewhere. However, upon rotatin g the electrode(Figure 3C ) ,both the reductio n and oxidation peaks of signalI vanish entirely to be replaced by a sigmoidal wave, the formand position of which (E?p= -31 mV, n = 1.7-1.8) agreewith the result observed with the severalfold higher fum ara teconcen tration described above. Each of the other signals (aswell as the redox couple in the region of -200 mV thateventually dominates at longer times) remain peak-like,although signal I1 appears attenuated compared to the zero-

50, I50 4

4506 7 6 9

PFIGURE: Graph showing the variation of E 'I 0)and E O ' I I 0 )with pH for fumarate reductase adsorbed at PGE electrodes at 0 OC(A) and 25 C (B). Other conditions are as for Figure 4. Graphsalso show the positions of E 1 p A) bserved for reduction of lowconcentrations of fumarate. Low-temperature E 1 / 2 alues are for 3C. Errors are f5 mV for E ', and E l l 2 , nd A10 mV for E ' ~ I .substrate/zer o-rotation experiment. These results specificallyidentify the entir e population of centers giving rise to signalI as being directly involved in cataly tic electron transpo rt andindicate a high catalytic rate co nstant for the adsorbed enzyme.Figure 4 shows cyclic voltammograms recorded for enzymesolutions over a range of p H a t 0 OC, in the absence of fumar ateor succinate. As before, these were recorded after the potentialwas in itially poised a t -1 60mV for 2 min to develop maximumsignal size. For pH 6, a greater concentration of polymyxinwas required to provide stability, and the curre nt peaks wereconsistently of somewh at lower amp litude. Gener ally, how-ever, the peaks are sharper and permit more detailedexamination compared to the results obtained at 25 OC.Voltammogram s were stable for a period of at least 20 min.Peak currents were proportional to scan rate, although weagain observed tha t the waves of signal I11 are broadened athigh scan rate, with the effect being particularly marked forthe reduction component.From the following considerations, we conclude that signalI arises from a redox couple involving concerted transfer ofmore than oneelectron. Theory predicts that thevoltammetricresponse arising from reversible electrochemistry of an ideal ,non-interacting, homogeneous population of d iffusionless edoxcouples should consist of reduction and oxidation peaks thatare symmetrical about the formal reduction potential, E O ' ,and have a half-height width, 6, of 3.52RTlnF volts ( 8 3 / nmV at 0 C) Laviron, 1979 a,b). Even though the baselineof these peaks cannot be determined with certainty (and webelieve, as m entioned below, that they m ust overlay two m uchbroader signals from centers 1 and 3 ) , it is clear tha t 6 mustbe considerably less than 8 3 mV. Adopting a reasonablebaseline by inspection, we estimate that 6 is in the order of50mV or less. Since environmental heterogeneity must serveonly to increase 6, and considering that interactions betweenlike centers in adjacent macromolecules are extremely unlikely(by contr ast with th e case for small adsorbed molecules), weconclude tha t n mus t lie close to 2 (Laviron, 1979a). The pHdependence of th e form al reduction potential, E O ' , , a t 0 OCis shown in Figu re 5A, and d ata a re displayed in Table I. Theplot is linear with a gradient of -29 mV/pH unit, thul


6/11

5460 Biochemistry, Vol. 32, No. 20, 1993 Sucheta et al.titrations (Simpkin Ingledew, 1985;Cam mac ket al., 1986;We rth et al., 1990). Th e second observation is that theamplitud e of signal I1 always appears related to th e size ofsignal I and to the corresponding catalytic activity. The thirdobservation (to be described later) is that, under conditionsof high turnover rate, the catalytic current for fumaratereduction shows a sm all but reproducible sigmoidal increasewith an E112value tha t coincides with E O ' , , . Direct inves-tigations of center 2 in FRD and SDH by spectroscopicmethods have been problematic. The EP R spectrum of [4Fe-4S l1+ is very broad, ind icative of interaction with theparama gneticcenters 1 and 3 (Cam mac ket al., 1986; Mag uireet al., 1985). Center 2 in succinate dehydrogenase has beencharacterized also by MCD spectroscopy (Johnson et al.,1985).

Signal 111, which is at unusually low potential and isparticularly sensitive o pH and scan rate, does not correspondto any previously measured redox process of th e enzyme. A tpH 8 and below, it appears as an asym metric pair of waveswhich sharpen up as the scan rate or pH is lowered. W esuggest that it corresponds o furth er proton-linked reductionof the [3F e-4SIo cluster (cen ter 3). Ou r rationale for thisassignment stems from o bservations tha t we h ave made involtammetric experiments on a number of smaller proteinscontaining [3Fe-4S] clusters. In every case thus far studiedthese voltammograms reveal an additional redox couple atlow potential, whose wave form and effective reductionpotential are also pH dependent (Armstrong et al., 1989b;Butt et al., 1991a; Iismaa et al., 19 91). Furtherm ore, forferredoxin I11from Desulfiovibrio africanus it was noted thatthe signal vanishes as the [3Fe-4S] cluster is converted into[4Fe-4S] or a heterometal analog (Butt et al., 1991a).Integrations of the low-potential signal and comparison withthe charg e passed for the n ormal [3Fe-4SI1+/O rocess led usto conclude that fu rther reduction of [3Fe-4S Iooccurs to givea novel state formally at the 2- oxidation level (Arm stronget al., 1989 b). Suc h redox activity has not been observed forproteins containing only [2Fe-2S ] or [4Fe-4S] clusters.

From repo rted potentiometric data for centers 1 and 3, weexpect signals due to the couples [2Fe-4SI2+/l+ nd [3Fe-4S]'+/O to lie in the region of signal I. How ever, these areone-electroncouples, and thus the ir half-height widths cannotbe less than 8 3 mV; since the peak curren t for a diffusionlessredox coupleis proportional o n2 (Laviron, 1979b), hey shouldappear as broad, shallow peaks by comparison with thecontribution from FAD. Considering the reported values ofEo' for center 1 (-20 mV) an d center 3 (-70 mV), they mighteither be extremely difficult to observe under the peaks ofsignal I or ap pear a s Yw ingsn o one or both sides of signal I,depending upon pH (Simpkin et al., 1985; Camm ack et al.,1986; We rthet al., 1990). Examinationofthevoltammogram sobtained at pH 6 and 9 does indeed provide some indicationfor th e existence of the expected waves, respectively on th enegativ e and positive sides of sign al I.The am ount of active enzyme on the electrode surface wasevaluated by three independent methods, each of whichinvolved integration of th e areas u nder waves or fractions ofwaves. Integrations were performed manually by cutting outthe required sections of the voltammograms (followingphotocopy enlargement) and com paring their weights with astand ard 'square' of known charge. For method 1, it wasassum ed that the clearly visible portion of signal I is dominatedby a component with n = 2 (assigned as the FAD). Sincepeak curre nt is proportional to n2 Laviron, 1979b), it followsthat the fraction of the signal above the half-height width

Table I: Formal Reduction Potentials (E ' = {Epa- E,)/2)for Redox Couples Observed in the Absence of Fumarate andHalf-Wave Potentials ( E l p ) Observed for Catalysis underConditions of Low Fumarate Concentrationatemp ( C) PH Eo'[ E 'II E112

3 6.0 -5 -273 -56.4 -146.7 -257.0 -30 -299 -347.2 -387.5 408.0 -56 -307 -6 48.8 -799.0 -86 -319 -84

25 6.0 -24 -284 -1 17.0 -4 8 -31 1 -327.5 -588.0 -8 1 -326 -118.6 -1079.0 -133 -354 -1 34Error margins are 5 mV for Eo'l and E , / *and A10 mV for EO'll.

suggesting a e-/H+ transfer ratio close to 2. The small peak-to-peak separation Up E,, - E F ) , which increases from5 to 25 mV over the scan rate range 10-470 mV s-l, showsthat the rate constant for electron exchange between theelectrode and the group responsible must be quite large(Laviron, 1979b). The results of several experiments carriedout a t 25 OC indicate that a m ore complex situation developsas the tempera ture is raised. As evident from Figure 5B, thecorresp ondin g plot of Eo'l against pH shows some curvature.Furthermore, a t pH 9, the signal I peaks appear broadened.Data are included in Table I.The pH dependence and narrow peak shape (apart fromthe singular result a t 25 OC and p H 9) sug gest strongly thatsignal I corresponds to the covalently bound FA D. In so faras this group is assumed to be the im mediate redox catalystfor succ inat efu ma rate interconversion, this proposal issupported by th e observation of its complete collapse duringturnover as the electrode is rotated. Previously, the F AD infumar ate reductase has been determined to have a two-electronreductio n potential of -55 mV at 25 OC and pH 7 and to showa p H dependence consistent with a 2e-/ 1H + process (Ackrellet al., 198 9). In the case of ma mm alian succinate dehydro-genase, potentiometric EP R measurem ents have shown thatreduction of the FAD occurs in two consecutive one-electronsteps with E O r values of -127 and -3 1 mV, respectively, at pH7 (Ohnishi et al., 19 78, 1981). The semiquinon e form istherefore unstable, and Eo' = -79 mV for the effective n =2 reaction. The high potential of the FAD in F RD and S DHcompared to tha t of free flavin and other flavoenzym es s tobe noted. For comparison, the reduction potential of freeFA D is -210 mV at pH 7 (Clarke, 19 60), while that of ahistidyl-FAD peptide fragm ent from sarcosinedehydrogenaseis -150 mV (Pate k Frisell, 1972). It is plausible that th ebroad redox signal in the region of -200 mV ap pearing atlonger times and as electroactivity declines may arise fromdenatured forms of the enzym e in which the environment ofthe FAD is perturbed.Signal I1is of lower peak current am plitude and markedlybroader than signalI,and th e form al reduction potential showsa small dependence upon pH (see Figure 5 and Table I fordata obtained at 0 and 25 C). Several items of evidenceprovide support for its assignment as the [4Fe-4S I2+/l+ luster(center 2) in catalytically active FrdA B. Th e first observ ationis that the reduction potential, E O ' I ~ , s in the rang e (-285 to-320 mV) reported for center 2 on the basis of potentiometric


7/11

Electrochemistry of Fumarate Reductase Biochemistry, Vol. 32, No. 20, 1993 5461.............. ATable 11: Amount of Fumarate Reductase (FrdAB) Adsorbed onPGE Electrodes (pmol cm-2 ) Evaluated by Three IndependentMethodsa

amt of FrdAB at pHb (temp, C)1 3.90 3.51 3.90 3.122 2.56 3.98 3.95 3.093 3.17 3.67 3.45 2.95

1 See text. Determinations were not carried out for pH 6 ecause thesignals were less stable and indicated significantly smaller coverage.effectively derives entirely from the contribution from FAD .This fraction corresponds to 25%of the to tal charge passedin an n = 2 wave envelope. Taking 6 = 41 and 45 mV,respectively, at 0 and 25 OC, we could thus ob tain values forthe numb er of electrochemically active F AD centers presentand hence determine th e number of enzyme molecules makingelectronic contact with the electrode. In method 2, it wasassumed that the en tire signal I envelope (including wings atthe side) comprises couples undergoing total transfer of fourelectrons. In method 3 it was assumed that signal 11,assignedto center 2, represents transfer of one electron per enzymemolecule. Values obtained by using methods 1 ,2 , and 3 arecompared in Tab le 11. Considering th e different assumptionsmade, the results are remarkably sim ilar, and we estim ate theerror to be within a factor of 25%. Taking the average ofvalues determined for pH 7 a t 25 OC (0.29 pmol) and usingthe electrode surface area 0.090 cm2,we calculate the coverager to be 3.2 f 0.8 pmol cm-2. Consequently, each FrdABmolecule effectively occupies an area of 5200 f 1300 A2,equivalent to a square with sides of 72 f A. Consideringthese dimensions and th e general agreement between resultsobtained over a range of pH and temperature (notingparticularly tha t r a t 25 OC is not significantly lower than rdetermined at 0 C), we conclude that the coverage ofelectroactive enzyme molecules achieved in ou r experimentsis virtually satura tive and probably close to one monolayer.

Formulation of a Basic M odel fo r Electrocatalysis andDetermination ofKinetic Parametersfrom RDE Experiments.The catalytic activity of adsorbed enzyme molecules forfumarate reduction was studied in some detail by RDEvoltammetry, w ith attention focused upon th e results obtaineda t pH 7 and 25 OC. So me experiments were also conductedfor fumar ate reduction at 3 O C and for succinate oxidationa t 25 OC, in each case (as measured a t pH 7) revealing muchlower catalytic activities, while preliminary studies were carriedout to assess catalysis of fum arate reduction a t pH 6, 8, and9. Figure 6 shows a typical set of RDE voltammogramsobtained for various fumarate concentrations at 25 OC andpH 7. As described earlier, it can be seen that, for lowconcentrations of fum arate, th e transition from baseline to aconstant limiting reduction cu rrent occurs within a narrowpotential range as expected for a reaction that approaches th eelectrochemically reversible case with n = 2. However, ashigher catalytic curren ts are induced, by increasing either thefum arate concentration or the rotation rate , the potential atwhich the curr ent reaches i l im becomes more negative. Thiseffect appears to evolve in two stages. Th ere is first of all agradual negative shift in E112 from the reversible valuedetermined (see above) for low concentrations of fum arat e orsuccinate, while the n value decreases. More dram atic,however, is the gross distortion th at o ccurs as mu ch highercataly tic currents are drawn. Under th e conditions of highestenforced curre nt, where the i-E curve appears drawn out

. .

........____.._..... c

E/mV vs S.H.E.FIGURE: Variation of cy clic voltammetry responses at a rotatingdisk PGE electrode as a function of fumarate concentration.Fumarate reductase (FrdAB ) 0.83 pM, olymyxin 0.2 g/L, pH 7.0(PIPES), emperature 25 OC. R otation rate 1200 rpm, scan rate200 mV s-l. Concentrations of fumarate are 0 pM (A), 49pM (B) ,196 pM (C), and 790 pM (D). Solid arrows indicate the positionof the second wave. The broken vertical arrow indicates the decreasein activity observed for high fumarate concentration. Horizontalbroken lines show the position of 0 absolute current.toward neg ative potential, a second wave appears as a distinctripple on the approach to ilim. The E112value of this waveis -300 mV. For any particular rotation rate, ilim increaseswith fum arate concentration to approach a maximum value.For studies with the highest levels of fumarate, e.g., 5 mM,ilimbecomes independent of rotation rate, including the caseof a stationary electrode. At low fuma rate concentration thevoltammetry was found to be independent of scan rate (andthus a t steady-state) over the range 2-50 mV s-l. At veryhigh fum arat e concentration, this range could be extended,for example, to a t least 470mV s-l for a fumarateconcentrationof 4 mM. This was desirable because it was found that , withsuch high fum arate co ncentrations, activity w as lost rapidlybelow a potential of -400 mV. It is this effect which givesrise to th e decrease in cata lytic curren t on successive cyclesobserved for the hig h-fu mar ate voltammogram shown inFigure 6D; at this fum arate concentration, voltammogramsscanned down to a less negative potential were much morestable but were not analytically useful because the limitingcurre nt was not attained . For the high-fumarate studies, ilimwas thus measured on th e first cycle (at -550 mV) for whichthe effective decrease in activity is negligible at fast scan r ate(hence we employed 200 mV s-I for these measurements).We have no clear explanation for th e loss of activity, exceptfor th e possibility of fum arate destabilizing th e enzyme layerin this potential region.Th e voltammetry clearly contains a wealth of informationon the electron-transport processes. In order to establish abasis for interpreting th e results, it is first of all necessary toformulate a skeleton model which can be elaborated toaccoun t for all aspects of th e observed behavior. In this paper,we shall focus upon this task by considering the system to beresolvable into three separate, independent processes asillustrated in Figure 7 . Under steady-state conditions sucha reaction sch eme gives rise to eq 3, which is analogous to anexpression giving the resultant conductance (reciprocal re-


8/11

5462 Biochemistry, Vol. 32, No. 20, 1993 Sucheta et al.ELECTRODE ENZYME SOLUTION

electron transferiE

Electronentrvf exitsite )

/-=I -elaySubstratetransformationsite

Mass transportCatalysis of substratek a t kev

FIGURE: Scheme showing the sequenceof steps in electron transportbetween electrode and substrate molecules during catalysis by anenzyme molecule adsorbed at the electrode surface.( 3 )

sistance) of a gro up of resistors connected in series. Th e firstterm deals with transport of substrate molecules between theenzyme and bulk solution. This process is addressed byexamining the the rotation rate dependence of th e current,which for a diffusion-con trolled reaction is desc ribed by theLevich equation:iLev= 0.62nFAD2/3Cv- /6w1/2 4)

Here, A is the electrode surface area, C is the bulk concen-tration of substrate (mol per unit volume), D is the diffusioncoefficient of substrate, v is the kinematic viscosity of thesolution (we used = 0.01 cm2 s-l a t 25 OC, a typical valuefor aqueous solutions), and w is the electrode rotation rate(specifically, the ang ular frequen cy in rad s-l). In the limitof a n electrode reaction occurring under such conditions thatit is entirely diffusion-controlled, only this term contributesand the limiting current is directly proportional to the sq uareroot of w (Levich, 1944; Riddiford, 1965).The second term is the exchange current contributiondue to interfacial electron transfer between the electrode andthe primary electron entry/exit site on the enzyme. In themembrane-bound complexes, the primary electron donor/acceptor site for reaction of S D H (and by analogy FRD ) withQ/Q H2 is believed to be center 3 (Ohnishi et al., 1976;Beinertet al., 1977). Electrons are thus relayed between center 3 andthe FA D, with center 1 probably serving as an intermediaryor storagesite. The questionof whether this mech anism appliesin the electroch emical experim ent is of inter est bu t is notaddressed directly at this stage. The p otential dependence ofiE is given by the current-potential characteristic shown byeq 5 . This expression simplifies o the Butler-Volmer equatio ni , = nFAk,[F0 exp(-anF(E -E o ) / R T )-

r R exp((1 - ) n F ( E- E o ) / R T ) ] 5 )for diffusing redox couples under con ditions n w hich the bulksubstrate concentration equals that at the electrode surface(Laviron 1979; Bard Faulkner, 1980; Bond, 1980;Southampton E lectrochemistry Group, 1985). Here, Poandr R are the respective surface concentrations of the oxidizedand the reduced form of the enzyme (expressed as mol perunit surface area) ,a s the transfer coefficient,n is the number

of electrons transferred in the transition state, E is the appliedelectrode potential, and Eo is the appropriate apparentstandard reduction potential of the system. The standardfirst-order rate constant, k, (the exchange rate constant),reflects the ease of moving electrons between the electrodeand the enzyme. At applied potentials much more positiveor negative than E O only oxidation or reduction components(respectively)become significant. Equation 6 shows how theliEl = n F A r k , ( 6 )

potential-dependent terms iE and kE (the first-order electro-chemical rate constant) are interrelated. Th e quantity I isthe total number of enzyme molecules adsorbed per unit a reaof electrode and thus equals r o + r R .The third term of eq 3 describes the intrinsic catalyticproperties of the enzyme in its reaction with su bstrate and isassumed to be independent of driving force and electroderotation rate. It can be expressed as an electrochemical formof the Michaelis-Menten expression as given in eq 7 . Theicat= nFAIkcatC/(C+ KM (7)

quantities KM nd kcat are the app arent M ichaelis-Mentenparameters, which a re assumed to be independent of appliedpotential. In the limit of C >> K M , catbecomes independentof cEquations 3-7 predict an interrelationshipbetween curre nt,applied potential, substrate concentration, and electroderotation rate. For a wide range of experim ental conditions,furt her an alysis can be facilitated by removing from consid-eration th e influenc e of electrochem ical driving force (andthus k ~ ) . e note from eq 3 that the term l / i E vanishesunder the condition i E >> icat and iLev, which is favored by t heintrinsic quantities k , (large), kcat (small), and K M large)and by the exp erimental variables decreasing C (eqs 4 and 7 ) ,decreasing w (eq 4), and s ufficiently large values of th eelectrochemical driving force,IE-EOI eq. ) . In the resultingi-E profile, the cur rent reaches a constant value, ilim once theelectrochemicaldriving force ceases to be a controlling factor.We measured ilim as a function of w1/2for a range of fum arateconcentrations at 25 OC and an alyzed da ta acco rding to eq4. Results of four experiments are shown in Figure 8A. Th enonlinear appearance of the Levich plots show that as w isincreased, the current is determined increasingly by the rateof catalysis as opposed to mass transport. Comb ining eqs 3 ,4, and 7 yields a form of the Koutecky-Levich equation (eq8), which describes hevariatio n of 1 /iIim withd or differentl / i l i m= (C + K M ) / ( n F A I kc a tC )

/ (0.62nFAD2/3Cv-/6w/2)8 )value s of C. Th e correspo nding plot of 1 / i l i m against w-1/2 isshown in Figure 8B. The intercept on the l/iIim axis (thereciprocal current extrapolated to infinite rotation rate) isnow identified with l / i c a t eq 7 ) . Analysis of the resultantslopes from the K outecky-Levich plots gave a value of D forfumar ate of 1 . 1 XIn Figure 8C we have plotted th e resulting values of i c a t /A(catalytic current density) against C. As the substrateconcen tration is increase d, i c a t /A pproa ches a limiting value,imax/A. At low Cv alue s the limiting slope is 2.56 f .01 Acm-2 M-I (M-l = per mol fumarate per liter). Insertingnumerical values, including = 3.2 X mol cm-2determ ined as described earlier, we obtain the value 5 . 3 X l o 6M-1 s-l for the second-order rate constant (kcat/&) forfumar ate reduction by the enzyme. Expressed in the familar

cm2 s-l, using n = 2.


9/11

Electrochemistry of Fumarate Reductase Biochemistry, Vola32, No. 20, 1993 5463of high current (high C, high w ) . Much of this may beattributed to the increasing influence of the potential-dependent term iE in eq 3. W e determined (see above) thatthe catalytic activity of adsorbed FrdA B is extremely high.Th e apparent retardation thus arises from the limitation thatoriginates in the p otential-dependent current term, iE (eq 5),and, through eq 6, in the electrochemical rate constant, kE,for interfacial electron transfer between electrodeand electronentry /exit site in the enzyme. Und er conditions wherebyturnover is limited by the rateo f arrivalof su bstrate (as favoredby low substrate concentration and low rotation rate), anadeq uate rate of electron transfer to t he enzyme is achievedwithout the need for an elevated driving force. In this casethe electrochemical reaction approaches reversibility. How -ever, an increase in the su bstrate concentration or the rotationrate stimulates a higher demand for electrons. The systemnow becomes sensitive to factors that will increase iE andtherefore responds to the electrochemical driving force. Wethus exp ect distortion of the i-E profile, as is indeed observed.

Two comm ents can be ma de about rates of electron transferin this system. First, the observation tha t the distortion isonly severe at high curre nts illustrates how conductive a con tactthe enzyme makes with the electrode. From the small valueof the peak separation, AE,, observed for signal I at scan ratesup to 470 mV s-I we can estimate a lower limit for k , usingthe treatment described by Laviron (1979b ). This is of theorder of 100 s-l, which is sufficiently large to ensure anadeq uate supply of electrons at low subs trate levels (thus theelectrochemistryappear s to approach reversibility) but certainto be restrictive once deman d is increased. Second, the largekcat value requires that any intramolecular electron-relayprocess must be very fast, i.e., with a rate constant a t leastof the order of 1600 s-l for a single pathway.

With this reasoning, we can now propose an explanationfor the second sigmoidal wave that appears in rotating-diskvoltammogram s at high sub strate concentration. The positionof this wave coincides with the formal redu ction potential,E O ] ) , of signal I1which we have tentatively assigned to center2. If the [4Fe-4S ] cluster is also capable of functioning asan electron entry site on the enzyme, then a second pathwayshould become available to relay electrons from th e electrodeto the F AD . This could be utilized in parallel with the primarypathway, boosting the supply and thus increasing he catalyticcurrent. Participation of this auxilliary electron entry sitewould be detectable under cond itionswhere substrate urnoveris limited by the rate of sup ply of electr ons and should becomeeffective as the applied potential approaches its formalreductio n potential,EOll. These features are indeed observed;the sm all boost results in th e current reaching ili, (in thiscase almost at the level of imaxt a potential more positivethan would otherwise be the case.

Experiments conducted for fumarate reduction (pH 7) a t3 OC showed sim ilar evidence for fast inte rfaci al electrontransfer, b ut values of ilim were much smaller, and Levichplots displayed appreciable curvature even for low concen-trations of fum arate. An analysis of data according to theabove procedure gave k,,, = 14 s-l and K M= 33 pM . Theseresults show a large temperature coefficient for the kineticsand suggest that the enzymes intrinsic ca talytic propertiesplay a greater part in determining the rate as the temperatu reis lowered.A limited amount of data were obtained for succinatedehydrogenation over the concentration range M O O pM a tpH 7 and 25 C. atalytic currents were much smaller thanobserved for fum arate reduction under th e same conditions;

I V Y

a00

i o 20I / ? rad/sec)

O 8 I

0 6

4

2

000010 2 030 040 5

r adked - 1

0 I O 0 0 Zoo0 3 1oo0 5000 6000concentration of fumarate WMF I G U R E: (A) Levich plot for reduction of fumarate by fumaratereductase adsorbed a t a PGE electrode (see text for explanation):temperature,25 O C 0.7 pM FrdAB;0.1-0.2 g/L polymyxin; pH 7.0(PIPE S). Fumarate concentrations are40 pM (+),200 pM (A),1.0mM O),and 5.0 mM (+). (B) Koutecky-Levich plot of data usedin (A). (C) Dependence of cataly tic current density i c a t / A ) ponfumarate concentration: temperature,25.0 O C ; pH 7.0 (PIPES); .1or0.2g/Lpolymyxin;0.8pMFrdAB. Thelineshownis henonlinearregression fit giving K M= 161 7 pM, and i m a x / A 520 pA cm-*(thus k,,, = 840 s- I ) .

formalism of enzym e kinetic theory, the term s imax/nF ndA rk , , , each represent the electrochemical analog of Vm,,.Proceeding further, by independently fitting the data shownin Figure 8C to the Michaelis-Menten model (eq 7), we obtainthe values k,,, = 840 f 13 s-l and K M = 161 f pM. Theerrors given arise in the d ata analysis, and the ra te constantsdo not reflect the estimated 25% uncertainty in r.Comparisons can be made with results from nonelectro-chemical assays using benzyl viologen as electron donor.Apparent K Mvalues of 410-470 p M have been obtained forTriton-solubilized and membrane-bound preparations offumarate reductase at room temperature and p H 7 (DickieWeiner, 1 979). From th e specific activityassay with benzylviologen (511 units mg-I) we calculate the equivalent kc,,value to be 396 s-l. At 38 OC, fum arate reduc tase gives k,,,valuesof ca. ~ O O O S ~Ackr ellet al., 1992). T hus, whenviewedelectrochemically, FRD is at least as active as it app ears fromtraditional methods.We now consider the likely cause of the asymmetricdistortion of the voltammogram s obtained under cond itions


10/11

5464 Biochemistry, Vol 32, No. 20, 1993indeed we noted th at even trace amounts of fum arate presentin the succinate solution produced a reduction cu rrent w hichinterfered with measurements. Sigmoidal i-E profiles wereobtained w ith values of ilim ha t were independent of electroderotation rate over the range 0-1800 rpm. Evidently, forcatalysis of succinate oxidation, the catalytic term (eq 7) sthe sole rate-determining factor u nder these conditions.

Additionally, preliminary studies were carried out to assessthe pH dependence of fum arate reduction at 3 and 25 O C . W erestricted our experiments to low fumarate concentrations,noting the position of theca talytic E1/ 2va lueand the magnitudeof catalytic currents. Two general effects were noted as thepH was increased from 6 to 9: catalytic currents decreased,and E l pvalues became more negative. In no case could wedetec t any evidence for a potential-dependent activity switch-off as observed for SDH (Sucheta et al., 1992). Thus animpo rtant difference exists between the tw o enzymes, perhapsthe result of evolutionary changes tha t have altered the orderor relative rates of intern al catalytic events.

Data for E112are listed in Tab le I and displayed in Figur e5 . For the low-temperature studies, the correspondencebetween E O , and E l p is extremely close and implicates thetwo-electron reduced form of FA D as a key intermed iate inthe catalytic reaction. However, at higher tempera ture, adifferent pattern of behavior is found, in that E lpv ari es inearlywith pH (slope -49 mV) and precedes E O I for p H values of7 and below. Sub tle aspects of a com plicated electron-transport reaction are thus immediately in evidence. W ebelieve tha t these results reveal a shift in the rate-de termin ingstep; for exam ple, a bond -breaking process (large AH* hichoccurs at the active site (and which controls the rate a t lowtempera ture) could give way to an intramo lecular electron-transfer step (small AH* s the temperature is increased.Although we have not made a detailed study of the effect oftemperatu re on rates, the large decrease in k,,,observed uponlowering the temperature from 25 o 3 O C provides independentevidence for the presence of a high-AH* step in the catalyticreaction. Furth er studies are required in order to form ulatean interpr etation of the variation of E1p and n values observedunder different conditions.

Th e results that we have presented now d emon strate thefeasibility of conducting an in-depth analysis of a complexelectron-transport enzyme using voltamm etric methods. Dy-namic electrochemistry exploiting rapid, direct electronexchange with redox centers in an enzym e that is attached tothe electrode offers a unique way to address the variouscomponents of catalysis. Th e approach provides exquisitecapabilities for precise control of impor tant reaction variable s(particularly potential control, driving force, and masstransport) and for viewing active-site redox transform ationsand revealing their participation in turnover. By such meanswe should be able to examine the intricate relationshipsbetween overall driving force and internal energetics. Fu-mar ate reductase and succinate dehydrogenase can now berecognized a s excellent sub jects for electroc hem ical studies ofenzym atic electron-transport systems.ACKNOWLEDGMENT

W e are grateful to Julea Bu tt, who conducted preliminaryelectrochemical experiments with fu mar ate reductase.REFERENCESAckrell, B. A. C., Cochran , B., Cecchini, G. (1989)Arch.Biochem. Biophys. 268, 26-34.

Sucheta et al.Ackrell, B. A. C., Johnson, M. K., Gunsalus, R. P., Cecchini,G. (1992) n Chemistr y and Biochemistry of Flavoenzymes(Mueller, F., Ed.)Vol. 3,pp 229-297,CR C Press, Boca R aton,FL.Ackrell, B. A. C., Arm strong, F.A,, Sucheta, A., Yu, T. (1993)(submitted for publication).Adams, R. N (1969)Electrochemistry at Solid Electrodes,Marcel Dekker, New York.Armstrong, F. A. (1990)Struct . Bonding 72, 137-221.Armstrong, F.A. 1 992) n Advances in Inorganic Che mist ry 38,

(Sykes, A. G., Camm ack, R., Eds.) pp 117-163,AcademicPress, New York.Armstrong, F. A., Lannon, A. M. (1987) .Am . Chem. SOC.Armstrong, F. A., George, S. J., Camm ack, R., Hatchikian, E.C., Thomson, A. J. (1989a)Biochem. J. 264, 265-273.Armstrong, F. A ,, Butt, J. N., George, S. ., Hatchikian, E. C.,Thomson, A. J. (1989b)FEBS Lett. 259, 15-18.Armstrong, F.A,, Butt, J. N., Sucheta, A. (1993a)MethodsEnzymol. (in press).Armstrong, F. A., Bond, A. M., Biichi, F.N., Hamn ett, A,, Hill,H. A. O., Lannon, A. M., Lettington, 0.C., Zoski,C. G.(1993b)Analyst (in press).Bard, A. J., Faulkner, L. R. (1980) lectrochemicalMethods.Fundamentals and Applica tions, Wiley, New York.Beinert, H., A ckrell, B. A. C., V inogradov, A. D., Kearney, E.B., Singer, T. P. (1977)Arch. Biochem. Biophys. 182,95-106.Blaut, M., Whittaker, K., Valdovinos, A., Ackrell, B. A. C.,Gunsa lus, R. P., Cecchini, G. (1989)J. Biol. Chem. 264,Bond, A. M . (1980)ModernPolarographicMethods in AnalyticalChemistry, Marcel Dekker, New York.Butt, J. N., Armstrong, F. A., Breton, J., George ,S. J.,Thomson,A. J., Hatchikian, E. C. (1991a)J. Am. Chem. SOC. 13,6663-6670.Butt, J. N., Sucheta, A., Armstrong, F. A., Breton, J., Thomson,A. J., Hatchikian, E. C. (1991b)J. Am. Chem.SOC. 13,Butt, J. N., Sucheta, A., Armstrong, F. A., Breton, J., Thomson,

A. J., Hatchikian, E. C. (1993) . Am. Chem. SOC.15,Cammack, R., Patil, D. S., Condon, C., Owen, P., Cole, S. T.,Weiner, J. H. (1984) n Flavins and Flavoproteins (Bray,R. C., Engel, P. C., Mayhew, S. G., Eds.) pp 551-554,Walter de Gruyter, New York.Cammack, R., Patil, D. S., Weiner, J. H. (1986)Biochim.Biophys. Acta 870, 545-551.Cammack, R., Maguire, J. J., Ackrell, B. A. C. (1987) nCytochrome Sy stems ; Molecular Biology and Bioenergetics(Papa, S. , Chance, B., Erns ter, L., Eds.) p 485,Plenum,New York.Clark, W. M. 1 96 ) Oxidation-reduction Potentials of OrganicSystems, Williams and Wilkins, Baltimore.Cole, S. T., Condon, C., Lemire, B. D., Weiner, J. H. (1985)

Biochim. Biophys. A cta 811, 81-403.Dickie, P., Weiner, J. H. (1979)Can. J. Biochem. 57 , 813-821.Had dock , B. A., Jones, C. W. (1977)Bacteriol. Rev. 41,47-99.Hayashi, K., Suzuki, T. (1965)Bulletin Znst. Chem. Res.,Kyoto Uniu. 43, 259-277.Heyrovskq, J., Kfita, J. (1966)Principles of Polarography,Academic Press, New York.Hill, H. A. 0. (1993)Methods Enzymol. (in press).Iismaa, S. E., Vazquez, A. E., Jensen, G. M., Stephens, P. J.,Butt, J. N., Armstrong, F. A., Burgess, B. K., (1991)J.Biol. Chem. 266, 21563-21571.Ingledew, W. J., Poole, R. K. (1984)Microbiol. Rev. 48,222-271.

109, 7211-7212.

13599-13604.

8948-8950.

141 5-142 .


11/11

Electrochemistry of Fumarate ReductaseJohnson, M. K., Morningstar, J. E., Bennett, D. E., Ackrell, B.A. C., Kearney , E. B. (1985)J . Biol. Chem. 260, 7368-7378.Laviron, E. 1 979a) . Electroanal. Chem. Interfacial Electro -chem. 100, 63-270.Laviron,E. 1979b) . Electroanal, Chem. Interfacial Electro -chem. 101,19-28.Lemire, B. D., Weiner, J. H. (1986)Methods Enzymol. 126,Levich, V. G. (1944)Acta Physicochim. URSS 19,117-132.Maguire, J. J., Johnson, M. K., Morningstar, J. E., Ackrell, B.A. C., Kearney, E. B., (1985) . Biol. Chem. 260, 10909-10912.Manodori, A., Cecchini, G., SchrM er, I., Gu nsalus, R. P., Werth,M. T., Johnson, M. K. (1992) iochemistry 31,2703-2712.Nicholson, R. S. Shain, I. (1964)Anal. Chem. 36, 706-723.Ohnishi, T. (1987)Curr. T op. Bioenerg. 15, 37-65.Ohnishi, T., Salerno, J. C., Maida, T., Yu, C. A., Nagaoka, S.,King, T. E. (1978) n Frontiers of Biological Energetics(Dutton, P. L., Leigh, J. S., Scarpa, A, , Eds.) pp 165-173,Academic Press, New York.Ohnishi, T., King, T. E., Salerno, J. C., Blum , H., Bowyer, J. R.,Maida, T. (1981)J. Biol. Chem. 256, 5577-5582.Patek , D. R., Frisell, W. R. (1972)Arch . Biochem. Biophys.150, 347-354.Riddiford, A. C. (1965)Adv. Electrochem. Electrochem. Eng.

4 , 47-97.

377-386.

Biochemistry, Vol. 32,No. 20,1993 5465Robinson, J. J., Weiner, J. H. (1982)Can. J . Biochem. 60,Schrader, I., Gunsalus, R. P., Ackrell, B. A. C., Cochran, B.,Cecchini, G. (1991) . Biol. Chem. 266, 13572-13579.Simpk in, D., Ingledew, W. J. (1985)Biochem. SO C. rans.13,603-607.Southampton Electrochemistry Group (1985) InstrumentalMethods in Electrochemistry, Ellis Horwood series n PhysicalChemistry,) Ellis HorwOOd, Chichester.Such eta, A., Ackrell, B. A. C., Cochran, B., Armstrong, F. A.(1992)Nature 356, 362-363.Varfolomeev, S.D., Berezin, I. V. (1978) . Mol. Catal. 4 ,

Weiner, J. H., Dickie, P. (1979) . Biol. Chem. 254, 8590-8593.Weiner, J. H., Cammack, R., Cole, S. ., Condon, C., Honore,N., Lemire, B. D., Shaw, G. (1986) roc. NatL Acad. Sci.U.S.A.83, 2056-2060.Werth, M. T.,Cecchini, G., Manodori, A., Ackrell, B. A. C.,SchrMer, I., Gunsalus, R. P., Johnson, M. K. (1990) roc.Natl. Acad. Sci. U.S.A.87, 8965-8969.Werth, M. T., ices, H., Cecchini, G., S ch ra er , I., Lasage, S.,Gunsalus, R. P., Johnson, M. K. 1992)FEBS Lett. 299,1-4.Westenberg, D. J., Gunsalus, R. P., Ackrell, B. A. C., Cecchini,G. (1990) . Biol. Chem. 265, 19560-19567.

811-816.

387-399.

Reversible Electrochemistry of Fumarate Reductase Immobilized on an Electrode Surface. Direct...

Documents

Transcript of Reversible Electrochemistry of Fumarate Reductase Immobilized on an Electrode Surface. Direct...