15-Redox properties of cytochrome P450BM3 measured by direct methods

Redox properties of cytochrome P450BM3measured by direct methodsBarry D.Fleming1,Yanni Tian1,Stephen G.Bell1,Luet-Lok Wong1,Vlada Urlacher2and H.Allen O.Hill1 1Department of Chemistry,Inorganic Chemistry Laboratory,University of Oxford,Oxford,UK;2Institute of Technical Biochemistry, University of Stuttgart,Stuttgart,GermanyCytochrome P450BM3is a self-sufficient fatty acid mono-oxygenase consisting of a diflavin(FAD/FMN)reductase domain and a heme domain fused together in a single polypeptide chain.The multidomain structure makes it an ideal model system for studying the mechanism of electron transfer and for understanding P450systems in general. Here we report the redox properties of the cyto-chrome P450BM3wild-type holoenzyme,and its isolated FAD reductase and P450heme domains,when immobilized in a didodecyldimethylammonium bromidefilm cast on an edge-plane graphite electrode.The holoenzyme showed cyclic voltammetric peaks originating from both theflavin reductase domain and the Fe III/Fe II redox couple contained in the heme domain,with formal potentials of)0.388and )0.250V with respect to a saturated calomel electrode, respectively.When measured in buffer solutions containing the holoenzyme or FAD-reductase domain,the reductase response could be maintained for several hours as a result of protein reorganization and refreshing at the didodecyldi-methylammonium modified surface.When measured in buffer solution alone,the cyclic voltammetric peaks from the reductase domain rapidly diminished in favour of the heme response.Electron transfer from the electrode to the heme was measured directly and at a similarly fast rate (k s¢¼221s)1)to natural biological rates.The redox potential of the Fe III/Fe II couple increased when carbon monoxide was bound to the reduced heme,but when in the presence of substrate(s)no shift in potential was observed. The reduced heme rapidly catalysed the reduction of oxygen to hydrogen peroxide.Keywords:cytochrome P450BM3;redox properties;electro-chemistry.The cytochrome P450group of enzymes comprises a variety of heme-containing monooxygenases that are present in the majority of prokaryotic and eukaryotic organisms[1].The primary reaction catalysed by these enzymes is the hydroxylation of carbonÆhydrogen bonds. Substrate hydroxylation requires the activation of dioxy-gen.The two-electron reducing agent in biological systems is almost exclusively NAD(P)H,and this reductive activa-tion of oxygen by P450occurs in two separate one-electron additions.Thefirst electron reduces the substrate-bound ferric heme,facilitating the rapid binding of dioxygen and the formation of the ferrous–dioxygen intermediate.The second electron addition,followed by protonation,forms the ferric hydroperoxy complex.The O–O bond is cleaved, with one O inserted into the substrate and the other reduced to form water.Cytochrome P450BM3is a self-sufficient fatty acid mono-oxygenase found in Bacillus megaterium[2,3].The native function of wild-type P450BM3is to oxidize long-chain fatty acids,but it has also been shown to oxidize many other substrates[4,5].The119-kDa molecular mass holoenzyme has its diflavin(FAD/FMN)reductase domain and heme domain fused together in a single polypeptide chain,making the transfer of electrons highly efficient[6].The multi-domain structure has made it an ideal model system for studying the mechanism of electron transfer and for understanding P450systems in general[7–10].Independent expression of the two domains has permitted their study in isolation[11].Redox potentiometric studies have shown the electron flow for P450BM3to follow the path NADPHfiFADfiFMNfiheme[12].The electronflow to the heme centre in P450BM3was presumed to be regulated by a substrate-dependent increase(>100mV)in the redox potential of the heme,with theÔsuitabilityÕof the substrate for catalytic transformation being reflected in the magnitude of the increase in potential.Thermodynamic arguments recently presented by Honeychurch et al.[13],suggested that the binding of dioxygen to P450cam,another essential step in the P450catalytic process,would be sufficient to enable electron cycling,regardless of whether camphor is present.Research on the electrochemistry of enzymes is driven partly by the desire to understand the details of electron transport in proteins and partly by the great potential uses of enzymes in electrochemically based biosensors and bioreactors.Electron transfer between an electrode and protein was initially accomplished in the pioneering work of Eddowes&Hill[14]and Yeh&Kuwana[15].They showed that the problem of slow electron transfer betweenCorrespondence to B.D.Fleming,Department of Chemistry,Inorganic Chemistry Laboratory,University of Oxford,South Parks Road,Oxford OX13QR,UK.Fax:+441865272690,Tel.:+441865275902,E-mail:barry.fleming@Abbreviations:DDAB,didodecyldimethylammonium bromide;EPG,edge-plane pyrolytic graphite;SCE,saturated calomel electrode.Enzymes:cytochrome P450(CYP102,P450BM3)(EC1.14.14.1).(Received14July2003,revised19August2003,accepted21August2003)Eur.J.Biochem.270,4082–4088(2003)ÓFEBS2003doi:10.1046/j.1432-1033.2003.03799.xan electrode and a metalloprotein could be overcome by use of an electron shuttle or mediator.Since then,much effort has been directed into developing suitably mediated or modified electrode systems that facilitate biological electrochemistry.One such approach is the recently documented technique of using synthetic surfactant-based biomimetic membranes[16].These castfilms have been shown to be useful in attaining direct electrochemistry of heme-containing proteins[17–22].In the present work we employed this method to obtain the direct, quasi-reversible electrochemistry of the wild-type cyto-chrome P450BM3holoenzyme and its isolated FAD-reduc-tase and heme domains.Materials and methodsEnzymes and chemicalsWild-type cytochrome P450BM3was overexpressed in the Escherichia coli strain,DH5a(that contained the gene of the wild-type cytochrome P450BM3)and the bacterial growth and protein purification were carried out using published procedures[23].The FAD-reductase domain was cloned, expressed and purified according to published procedures [24].The P450BM3heme domain was cloned and expressed, using standard procedures,to encompass amino acid residues1–481of P450BM3.The P450BM3heme domain was purified by DEAE Sepharose and Source-Q anion-exchange chromatography,as described previously[23]. After purification,fractions with an A417/A280of>1were collected and stored.All samples were stored at)20°C in 40m M phosphate buffer,pH7.4,containing50%v/v glycerol.Glycerol was removed immediately prior to experiments by gelfiltration on a Amersham-Pharmacia PD-10column equilibrated with40m M phosphate buffer, pH7.4.Chemicals and solvents were of reagent grade and used without further purification unless stated otherwise.Dido-decyldimethylammonium bromide(DDAB)from Aldrich was prepared and used as a0.1-M stock solution in chloroform.Apparatus and proceduresDC cyclic voltammetry experiments were performed at room temperature in a standard two-compartment glass cell with a working volume of0.5mL.The working compart-ment housed the platinum gauze counter electrode in addition to the edge-plane pyrolytic graphite(EPG)work-ing electrode.A saturated calomel electrode(SCE)was used as a reference in a sidearm that connected to the working compartment via a Luggin capillary.All potentials were referred to the SCE.An Autolab potentiostat(Eco Chemie, Utrecht,the Netherlands)was used to record and control the potential of the working electrode.All measurements were made in40m M potassium phosphate buffer,pH7.4. Voltammograms were taken in solutions that had been deoxygenated by purified argon.For buffered protein solutions this was accomplished by blowing argon over the solution for several hours.An argon atmosphere was maintained over the solution during the experiment,unless stated otherwise.Preparation of P450/DDAB/EPG electrodesDDABfilms were made by spreading5l L of the stock solution onto a freshly polished EPG electrode.The chloroform was allowed to evaporate in air at room temperature for%1h.To incorporate the enzyme into the DDAB-modified electrode,the electrode was either placed into a solution of protein(%10l M)for1h at4°C, or directly into the electrochemical cell containing buffered protein solution at room temperature.ResultsA typical cyclic voltammogram,recorded for a P450BM3–DDAB-modified EPG electrode in oxygen-free phosphate buffer,pH7.4,is shown in Fig.1.Three cyclic voltammo-grams are shown:the initial scan soon after electrode immersion,and then scans after30and60min.This series of scans highlights the evolution of the response over a period of1h.Initially,the electrode response was domin-ated by a well-defined reversible couple centred at)0.388V, with a broader,less intense couple observed at)0.250V. With time the response at the more negative potential diminished,whilst the intensity of the other increased,until after1h the current response of the couples,relative to each other,remained stable(over longer periods of time both couples decreased at approximately the same rate).A significant response at)0.250V could still be observed for up to1week for electrodes maintained in buffer solution at 4°C.A different result was observed for voltammograms measured using a DDAB-modified EPG electrode in a buffered solution of the holoenzyme(Fig.2).The intensity of the couple at)0.388V could be maintained for longer time-periods(up to3h in this case),with a slight increase in the signal centred on)0.250V.When the electrode was removed from the enzyme solution and placed into enzyme-free buffer solution,the electrode response observedwas Fig.1.Cyclic voltammetry of wild-type P450BM3holoenzyme immobi-lized at a didodecyldimethylammonium bromide(DDAB)-modified edge-plane pyrolytic graphite(EPG)electrode.Cyclic voltammograms were recorded in deoxygenated phosphate-buffer solution(pH7.4),at a scan rate of0.1VÆs)1,for wild-type P450BM3holoenzyme immobilized at a DDAB-modified EPG electrode after being initially immersed (darkest line)and then30and60min(lightest line)later.The arrows indicate the direction of current change for the respective peaks over the duration of the experiment.ÓFEBS2003Redox properties of cytochrome P450BM3(Eur.J.Biochem.270)4083similar to that shown in Fig.1.If,after removing the electrode from the enzyme solution it was stored in buffer for any length of time,then the response was essentially dominated by the signal at )0.250V.To help identify the origins of the cyclic voltammetric peaks observed with the holoenzyme,the individual reduc-tase and heme domains were tested separately.Figure 3shows that when the FAD-reductase domain was immobi-lized at a DDAB-coated EPG electrode,quasi-reversible electrochemistry in buffer was possible.The mid-point potential of the reductase domain was )0.405V.As with the holoenzyme,the reductase domain response diminished quite rapidly over a period of 1h when in buffer solution,but could be maintained for a longer duration when in protein solution.A typical cyclic voltammogram recorded for the heme-domain DDAB-modified EPG electrode in buffer is shown in Fig.4.The response consisted of one Faradaic couple centred on )0.244V,probably from the Fe III /Fe II redoxcouple.This strong response could be maintained for many hours and for up to 1week when the electrode was maintained in buffer solution at 4°C.The cyclic voltam-metric peaks were approximately symmetrical,having equal areas under both the reductive and oxidative cycle,and showed a linear current dependence with scan rate from 0.01to 1V Æs )1,as expected for thin films of electroactive ing these peak areas,the concen-tration of the electroactive enzyme at the electrode surface was calculated to be of the order of nmoles Æcm )2.The peak separation was measured as a function of scan rate for the Fe III /Fe II redox couple in both the holoenzyme and the heme domain.The Ôtrumpet Õplots,so-formed in this case,are shown in Fig.5.Significant peak separation was only observed when the scan rate exceeded %10V Æs )1.From these data,a value for the apparent average electron-transfer rate constant (k s ¢)could be calculated.Generally,k s ¢values are obtained from the peak (E pa )to peak (E pc )potential separation values in cyclic voltammograms based on Laviron’s approach for diffusionless thin-layer voltam-metry [25].The k s ¢value for the Fe III /Fe II redox coupleinFig.2.Cyclic voltammetry of wild-type P450BM3holoenzyme in solu-tionusin g a DDAB-modified EPG elec trode.Cyclic voltammograms were recorded in a pH 7.4-buffered solution of wild-type P450BM3holoenzyme at a scan rate of 0.1V Æs )1for a DDAB-modified EPG electrode after being immersed for 1(darkest line),2or 3h (lightest line),respectively.The arrows indicate the direction of current change for the respective peaks over the duration of theexperiment.Fig.3.Cyclic voltammetry of the P450BM3FAD-reductase domain immobilized at a DDAB-modified EPG electrode.The cyclic vol-tammogram was recorded in deoxygenated phosphate-buffer solution,pH 7.4,at a scan rate of 0.1V Æs )1for the P450BM3reductase domain immobilized at a DDAB-modified EPGelectrode.Fig.4.Cyclic voltammetry of the P450BM3heme domainimmobilized at a DDAB-modified EPG electrode.The cyclic voltammogram was recorded in deoxygenated phosphate-buffer solution,pH 7.4,at a scan rate of 0.1V Æs )1for the P450BM3heme domain immobilized at a DDAB-modified EPGelectrode.Fig.5.‘Trumpet’plots for the heme-domainrespon se.The response from the P450BM3holoenzyme (j )and the heme domain (m )are shown.The reductive and oxidative peak potentials are plotted against the scan rate.This type of plot can be used to calculate the electron transfer rate constant.4084B.D.Fleming et al.(Eur.J.Biochem.270)ÓFEBS 2003the holoenzyme and heme domain were determined to be 138and 221s )1,respectively.The influence of pH on the heme redox potential,measured by cyclic voltammetry,is shown in Fig.6.These data are representative for the heme response from both the holoenzyme and isolated heme domain.The mid-point potential became increas-ingly more negative as the pH was increased from 3to 10.Two linear regions of different slopes were observed,one between pH 3and pH 8,and the other between pH 8and pH 10.The slope of these regions was )33and )126mV ÆpH unit )1,respectively.After bubbling the buffer solution with CO for 15min,the mid-point potential of the heme domain was positively shifted by 50mV (Fig.7).When purged with argon,the original formal potential returned.The effect of substrate binding on the redox potential of the heme domain was also investigated.When any of the substrates lauric acid,palmitic acid or octane were added to the buffer solution,there was typically no change in the cyclic oltammetric peaks.The heme redox couple was very sensitive to the presence of molecular oxygen.Figure 8shows the effect,on the cyclic voltammogram for the heme domain DDAB-modified EPG electrode,of adding 1,3or 5mL of air into the buffer solution.A new couple,at a potential slightly positive of the Fe III /Fe II couple,was observed.The reduction of O 2by Ôbare Õor DDAB-coated EPG electrodes occurred at more negative potentials ()0.5to )0.7V).Thus,the presence of the heme significantly lowers the overpotential required for O 2reduction.The magnitude of the reduction peak was related to the amount of O 2added.The oxidation peak height was less intense than the reduction peak,which is characteristic of a mechanism involving the rapid electrocatalytic reduction of O 2to H 2O 2by the reduced heme-containing films as per Eqn (1)and Eqn (2):P450Fe II þO 2!P450Fe II ÀO 2ð1ÞP450Fe II ÀO 2þ2H þþ2e !P450Fe II þH 2O 2ð2ÞDiscussionThe isolated FAD reductase and heme domains of P450BM3have been useful in identifying the voltammetric response observed with the P450BM3holoenzyme.It was clear that the Faradaic couples centred on )0.388and )0.250V had their origin in the reductase and heme components,respectively.The peak identification was aided by the fact that there was no significant shift in potential for the individual electroactive components compared with when they were fused together in the holoenzyme.A close similarity in the redox potential has also been measured by redox potentiometry for FAD and FMN in the isolated domains or the P450BM3holoenzyme [12].These solution measurements were explained in terms of there being no significant change in the domain environments whether isolated or fused together.The same could also be said here of the reductase and heme domains when they are incorporated in the DDAB film at the electrodesurface.Fig.6.Influence of pH on the heme redox potential.Cyclic vol-tammograms at 0.1V Æs )1for wild-type P450BM3holoenzyme were measured at different pH values.The mid-point potentials observed for each voltammogram were plotted against pH.A similar trend was observed for the isolated hemedomain.Fig.7.The effect of CO binding on the P450BM3heme redox potential.Cyclic voltammograms were recorded in deoxygenated phosphate-buffer solution (at pH 7.4),before and after bubbling with CO for 15min,at a scan rate of 0.1V Æs )1for the P450BM3heme domain immobilized at a didodecyldimethylammonium bromide (DDAB)-modified EPG electrode.Peaks shifted to the right in the presence ofCO.Fig.8.The effect of O 2binding on the electrochemistry of the P450BM3heme domain.Cyclic voltammograms were recorded in a phosphate-buffer solution,pH 7.4,after adding 0,1,3or 5mL of air,at a scan rate of 0.1V Æs )1for the P450BM3heme domain immobilized at a DDAB-modified EPG electrode.The height of the reduction peak increased with the amount of air injected into the solution.ÓFEBS 2003Redox properties of cytochrome P450BM3(Eur.J.Biochem.270)4085The change in the peak intensities for the holoenzyme, when measured in enzyme-free buffer solution,is of particular interest and several explanations are considered. First,some reorganization at the electrode/solution inter-face,presumably within the DDABfilm,had taken place. The initially large response from the reductase domain indicates that it is preferentially bound/oriented closest to the electrode surface.The further development of the electrode response,lowering of the reductase peak vs. increase of the heme peak,indicates that this orientation may be reversed,with the enzyme seemingly rotating to allow the heme domain to take up a more favourable electron transfer position near the electrode surface.It is also possible that the short-lived response in buffer is caused by denaturation at the electrode surface.This is suggested by the results observed with the isolated reductase domain. Its response,both in the presence of protein and in buffer only,was similar to that for the holoenzyme(except for the heme component).When the reductase domain is present in solution,it is plausible that the surface can be refreshed, effectively replacing the denatured protein.This could account for the longer duration of the large Faradaic currents observed.When the protein is not present in solution,denaturation takes place with noÔrefreshmentÕof the electrode surface and hence the relatively rapid decrease in current.With bare EPG electrodes,the direct electron transfer to P450BM3is slow and often not observed.The heme group is deep within the protein structure and favourable orientation at the electrode surface must occur to ensure electron transfer.When incorporated into the surfactant layer,the direct,rapid and quasi-reversible electron transfer between the P450heme and electrode was observed.The P450heme redox potential measured here(in both the holoenzyme and the heme domain only)is of a similar value to previous measurements for other P450enzymes and heme-containing proteins incorporated in DDABfilms[16,17,21,22].The broadness of the heme peak is considered to be caused by dispersion of E°values resulting from slight variations in protein orientation at the electrode surface[17].The differences in broadness between the reductive and oxidative waves,evident in Figs3and4,is typically the result of nonideality observed with thin-film systems[26].The potential of the heme Fe III/Fe II redox couple()0.252V) was much more positive than that measured in solution ()0.609V)[12].This type of behaviour has been reported for all cases where DDAB has been present as the biomimetic membrane.This is the result of interactions between the protein and surfactant and/or surfactant-related electrical double-layer effects on electrode potential. This effect was also evident,albeit to a lesser extent,with the reductase response.Our results show that the process of immobilization provides a very favourable environment for electron transfer to the heme to occur.The k s¢values calculated for the P450 heme domain(%200s)1)were similar in magnitude to the k s¢value measured for the natural electron transfer process–that between the FMN and heme for the P450BM3 holoenzyme in solution(223s)1with myristate)[7].These electron transfer results were dependent on the nature of the substrate–with the less favoured substrate lauric acid showing a lower k s¢value of130s)1.It has also been shown that it takes more energy to transfer an electron to the P450 heme when no substrate is bound[12].Our results are all the more interesting given that no substrate was present.Other k s¢values for heme-containing proteins,measured using electrode systems,are well below those reported here for the P450BM3heme domain.For example,P450cam in a DDAB film on a PG electrode was25s)1[17],whilst a variety of modified PG electrodes containing myoglobin showed k s¢values ranging from27to86s)1[27].The pH-dependent potential change observed in Fig.6 has been shown previously for other heme-containing proteins immobilized at electrode surfaces[18,19,22,27–30].However,in contrast to these previously published data, the slopes measured in this study for the linear regions were both quite different to the)59mVÆpH unit)1expected for a reversible one-electron transfer coupled to a single proton transfer.A similar low-slope region has also been reported for myoglobin and hemoglobin in polyacrylamidefilms,but this was for pH values of<5,a region where protein integrity might be questioned[27,28].Previous redox potentiometry experiments on P450BM3 showed that the presence of a suitable substrate results in an anodic(or positive)shift in redox potential in excess of 100mV[12].A similar substrate-dependent anodic shift was reported for P450cam from electrochemical data[31]. Our results indicate that no shift occurs in the formal potential of heme when in the presence of the substrates lauric acid,palmitic acid or octane.Similar results,based on cyclic voltammetric data,were reported recently for P450cam and P450cin in the presence of their natural substrates,camphor and cineole,respectively[22,32].The work with P450cin employed a similar procedure for enzyme immobilization as reported in this work[22]. Interestingly,redox potentiometric data for P450cin also showed no indication of a substrate-dependent anodic shift.There are reasonable thermodynamic arguments to suggest that substrate binding is not the only or main consideration in determining whether electron transfer to P450will occur[13].CO is known to rapidly bind specifically as a sixth ligand to the reduced heme iron of P450BM3[33].The fact that addition of CO to the buffer solution in our electrochemical experiments resulted in a peak shift of50mV confirmed that the observed response is from the heme domain. Similar results were obtained for P450cam immobilized at a DDAB-modified PG electrode[17]and a glassy carbon electrode modified with sodium montmorillonite[32].For catalytic reactions involving P450enzymes,the reduction of molecular oxygen to reactive oxygen species, such as H2O2,is typically an unwanted occurrence which dramatically reduces the efficiency of the desired catalytic process.As shown in Fig.8,once generated,the ferrous heme rapidly binds dioxygen,but unfortunately catalytic reduction to H2O2usually quickly follows.The real challenge then,in any development of electrode-based bioreactors designed to utilize the monooxygenase capabi-lities of P450,is getting the second electron to be used in peroxoiron complex formation and not in H2O2dissoci-ation.Several groups have recognized the difficulties associated with overcoming this problem and have attemp-ted to utilize mediator-promoted and H2O2-driven pathways to achieve their desired oxidation reactions4086B.D.Fleming et al.(Eur.J.Biochem.270)ÓFEBS2003[30,34,35].Given the fast electron transfer rates and low potentials necessary for thefirst electron reduction of the P450BM3heme domain in this surfactant-electrode ensem-ble,it follows that it should be the subject of ongoing study, as in our laboratory.However,it remains to be seen whether an effective electrochemically driven bioreactor,fully utili-zing Nature’s enzyme technology,can be achieved. ConclusionsWe determined the redox properties of cyto-chrome P450BM3by direct electrochemistry.The holo-enzyme response at a DDAB-modified EPG electrode was characterized by redox couples at)0.388V and)0.250V. These were identified as being direct electron transfer to the individualflavin reductase domain and P450heme domain, respectively.We have also shown that,although electron transfer in the biological system is from FAD/FMN to the heme,electron transfer can occur directly from the electrode to the heme under electrochemical conditions.The rate of this electron transfer is very rapid,of the order of the rates observed in the natural donor system.The redox potential of the heme did not appear to be affected by substrate binding,but it was possible that substrate inclusion in the surfactant layer may alter the surrounding charge environ-ment experienced by the protein.The reduction of mole-cular oxygen was readily catalysed by the P450BM3 heme domain immobilized at the DDAB-modified elec-trode.AcknowledgementsWe would like to thank the European Union for the collaborative grant.We also thank ECEnzymes,BASF and Dr T.Habicher. References1.OrtiZ de Montellano,P.R.(1995)Cytochrome P450:Structure,Mechanism and Biochemistry,2nd edn.Plenum Press,New York.2.Narhi,L.O.&Fulco,A.J.(1986)Characterization of a catalyti-cally self-sufficient119,000-dalton cytochrome P-450mono-oxygenase induced by barbiturates in Bacillus megaterium.J.Biol.Chem.261,7160–7169.3.Ravichandran,K.G.,Boddupalli,S.S.,Hasemann,C.A.,Peter-son,J.A.&Deisenhofer,J.(1993)Crystal structure of hemopro-tein domain of P450BM-3,a prototype for microsomal P450s.Science261,731–736.4.Guengerich,F.P.(2001)Common and uncommon cytochromeP450reactions related to metabolism and chemical toxicology.Chem.Res.Toxicol.14,611–650.5.Guengerich, F.P.(2002)Cytochrome P450enzymes in thegeneration of commercial products.Nat.Rev.Drug Discov.1, 359–366.6.Narhi,L.O.&Fulco,A.J.(1987)Identification and character-ization of two functional domains in cytochrome P-450BM-3,a catalytically self-sufficient monooxygenase induced by barbitu-rates in Bacillus megaterium.J.Biol.Chem.262,6683–6690.7.Munro, A.W.,Daff,S.,Coggins,J.R.,Lindsay,J.G.&Chapman,S.K.(1996)Probing electron transfer inflavocyto-chrome P-450BM3and its component domains.Eur.J.Biochem.239,403–409.8.Sevrioukova,I.F.,Hazzard,J.T.,Tollin,G.&Poulos,T.L.(1999)The FMN to heme electron transfer in cytochrome P450BM-3.J.Biol.Chem.274,36097–36106.9.Sevrioukova,I.F.,Li,H.,Zhang,H.,Peterson,J.A.&Poulos,T.L.(1999)Structure of a cytochrome P450-redox partner electron-transfer complex.Proc.Natl A96,1863–1868.10.Munro,A.W.,Leys,D.G.,McLean,K.J.,Marshall,K.R.,Ost,T.W.B.,Daff,S.,Miles, C.S.,Chapman,S.K.,Lysek, D.A., Moser,C.C.,Page,C.C.&Leslie Dutton,P.(2002)P450BM3:the very model of a modernflavocytochrome.Trends Biochem.Sci.27,250–257.es,J.S.,Munro,A.W.,Rospendowski,B.N.,Smith,W.E.,McKnights,J.&Thomson,A.J.(1992)Domains of the catalyti-cally self-sufficient cytochrome P-450BM-3.Genetic construction, overexpression,purification and spectroscopic characterisation.Biochem.J.288,503–509.12.Daff,S.N.,Chapman,S.K.,Turner,K.L.,Holt,R.A.,Govinda-raj,S.,Poulos,T.L.&Munro,A.W.(1997)Redox control of the catalytic cycle offlavocytochrome P-450BM3.Biochemistry36, 13816–13823.13.Honeychurch,M.J.,Hill,H.A.O.&Wong,L.-L.(1999)Thethermodynamics and kinetics of electron transfer in the cyto-chrome P450cam enzyme system.FEBS Lett.451,351–353. 14.Eddowes,M.J.&Hill,H.A.O.(1977)Novel method for theinvestigation of the electrochemistry of metalloproteins:cyto-chrome c.J.Chem.Soc.,mun.21,771–772.15.Yeh,P.&Kuwana,T.(1977)Reversible electrode reaction ofcytochrome c.Chem.Lett.10,1145–1148.16.Rusling,J.F.&Nassar,A.-E.F.(1993)Enhanced electron transferfor myoglobin in surfactantfilms on electrodes.J.Am.Chem.Soc.115,11891–11897.17.Zhang,Z.,Nassar,A.-E.F.,Lu,Z.,Schenkman,J.B.&Rusling,J.F.(1997)Direct electron injection from electrodes to cytochrome P450cam biomembrane-likefilms.J.Chem.Soc.,Faraday Trans.93,1769–1774.18.Rusling,J.F.(1998)Enzyme bioelectrochemistry in cast bio-membrame-likefilms.Acc.Chem.Res.31,363–369.19.Chen,X.,Hu,N.,Zeng,Y.,Rusling,J.F.&Yang,J.(1999)Ordered electrochemically activefilms of hemoglobin, didodecylmethylammonium ions,and ngmuir15,7022–7030.20.Boussaad,S.&Tao,N.J.(1999)Electron transfer and adsorptionof myoglobin on self-assembled surfactantfilms:an electro-chemical tapping-mode AFM study.J.Am.Chem.Soc.121, 4510–4515.21.Koo,L.S.,Immos,C.E.,Cohen,M.S.,Farmer,P.J.&OrtiZ deMontellano,P.R.(2002)Enhanced electron transfer and lauric acid hydroxylation by site-directed mutagenesis of CYP119.J.Am.Chem.Soc.124,5684–5691.22.Aguey-Zinsou,K.-F.,Bernhardt,P.V.,De Voss,J.J.&Slessor,K.E.(2003)Electrochemistry of P450cin:new insights into P450 electron mun.3,418–419.23.Carmichael,A.B.&Wong,L.-L.(2001)Protein engineering ofBacillus megaterium CYP102.The oxidation of polycyclic aromatic hydrocarbons.Eur.J.Biochem.268,3117–3125.24.Sevrioukova,I.,Truan,G.&Peterson,J.A.(1996)Theflavo-protein domain of P450BM-3:expression,purification,and properties of theflavin adenine dinucleotide-andflavin mono-nucleotide-binding subdomains.Biochemistry35,7528–7535. viron,E.(1979)General expression of the linear potential sweepvoltammogram in the case of diffusionless electrochemical sys-tems.J.Electroanal.Chem.101,19–28.26.Armstrong,F.,Heering,H.A.&Hirst,J.(1997)Reactions ofcomplex metalloproteins studied by protein-film voltammetry.Chem.Soc.Rev.26,169–179.27.Shen,L.,Huang,R.&Hu,N.(2002)Myoglobin in poly-acrylamide hydrogelfilms:direct electrochemistry and electro-chemical catalysis.Talanta56,1131–1139.ÓFEBS2003Redox properties of cytochrome P450BM3(Eur.J.Biochem.270)4087。