Anal. Chem. 2004, 76, 6046-6052
Direct Electron Transfer of Cytochrome P450 2B4 at Electrodes Modified with Nonionic Detergent and Colloidal Clay Nanoparticles Victoria V. Shumyantseva,† Yuri D. Ivanov,† Nikitas Bistolas, Frieder W. Scheller, Alexander I. Archakov,† and Ulla Wollenberger*
Department of Analytical Biochemistry, University of Potsdam, Karl-Liebknecht-Strasse 24-25, 14476 Golm, Germany
A method for construction of biosensors with membranous cytochrome P450 isoenzymes was developed based on clay/detergent/protein mixed films. Thin films of sodium montmorillonite colloid with incorporated cytochrome P450 2B4 (CYP2B4) with nonionic detergent were prepared on glassy carbon electrodes. The modified electrodes were electrochemically characterized, and bioelectrocatalytic reactions were followed. CYP2B4 can be reduced fast on clay-modified glassy carbon electrodes in the presence of the nonionic detergent Tween 80. In anaerobic solutions, reversible oxidation and reduction is obtained with a formal potential between -0.292 and -0.305 V vs Ag/AgCl 1 M KCl depending on the preparation of the biosensor. In air-saturated solution, bioelectrocatalytic reduction currents can be obtained with the CYP2B4-modified electrode on addition of typical substrates such as aminopyrine and benzphetamine. This reaction was suppressed when methyrapone, an inhibitor of P450 reactions, was present. Measurement of product formation also indicates the bioelectrocatalysis by CYP2B4. Cytochrome P450 enzymes are members of a large superfamily of hemoproteins that catalyze the NAD(P)H-dependent metabolism of a wide range of organic substrates estimated to be in the region of 200 000 or more chemicals and involving ∼60 distinct classes of biotransformation reactions.1,2 For the development of individual chemotherapy, in the future it will be important to know the activity of drug-metabolising isoforms of human P450s and their polymorphism. Cytochrome 2B4 (CYP2B4) is the main hepatic P450 isoform inducible by phenobarbital and other barbiturates. Among the preferred substrates of this isoform are phenobarbital and related barbital drugs and also benzphetamine and other aliphatic amines.3 Barbital drugs are still used as sedatives in medical practice. On the other hand, amphetamines are among often abused stimulant drugs which may produce a * To whom correspondence should be addressed. E-mail: uwollen@ rz.uni-potsdam.de. Fax: ++493319775051. † Permanent address: Institute of Biomedical Chemistry, Pogodinskaya Str. 10, 119121 Moscow, Russia. (1) Archakov, A. I.; Bachmanova, G. I. Cytochrome P450 and Active Oxygen; Taylor and FrancisL London, 1990. (2) Ortiz de Montellano, P. R., Ed. Cytochrome P450: Structure, Mechanism, and Biochemistry, 2nd ed.; Plenum Press: New York, 1995. (3) Lewis, D. F. V. Cytochrome P450. Structure, function and mechanism; Taylor and Francis: London 1996.
6046 Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
psychosis that resembles schizophrenia. For these compounds, methods such as GC/MS and HPLC4 that are now available are time-consuming and expensive.5 With colorimetric immunoassays, rapid qualitative in vitro diagnosis can be made. Quantitative measurements with immunoassays require longer time, i.e., up to 3 days. A cytochrome P450 biosensor may be a potential alternative that would allow a quick measurement for those compounds with comparable simple equipment. On the other hand, CYP2B4 can serve as model for other membrane-bound drug-metabolizing P450 isoenzymes, which could lead to an array of drug biosensors. The well-accepted general mechanism3 of the cytochrome P450 reaction entails several reaction steps. After substrate binding, one electron is transferred from a redox partner to the ferric heme iron, which is reduced to the ferrous enzyme. The reduced heme moiety binds molecular oxygen. The next steps include transfer of a second electron and proton to the ferrous-dioxygen species gaining an iron-hydroperoxo intermediate. The peroxo bond is cleaved to release a water molecule and an highly active ironoxo ferryl intermediate, which finally leads to a single oxygen atom insertion into the bound organic substrate, release of products, and regeneration of the ferric form. The two reducing equivalents are naturally supplied by NAD(P)H via flavoproteins or ferredoxin-like proteins.1,2 The controlled potential electrolysis method for the delivery of electrons via the NADPH-P450-reductase and cytochrome P450 was applied to a number of P450-catalyzed reactions6,7 and also the transfer of electrons via the iron-sulfur protein from the electrode to P450.8 In this way, it is not necessary to apply NAD(P)H. For a potential application of cytochromes P450 in biosensors and bioreactors, it might be suitable to substitute the biological electron delivery and transport system completely by an electrode. The electron transfer could be realized by means of redox mediators covalently bound to the enzyme.9 The direct (4) Masubuchi, Y.; Fujita, S.; Chiba, M.; Kagimoto N.; Umeda, S.; Suzuki, T. Biochem. Pharmacol. 1991, 41, 861-865. (5) Fujita, K.; Kamataki T. Drug Metabol. Pharmacokinet. 2002, 17, 1-22. (6) Faulker, K. M.; Shet, M. S.; Fisher, C. W.; Estabrook, R. W. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 7705-7709. (7) Estabrook, R. W.; Faulker, K. M.; Shet, M. S.; Fisher, C. W. Methods Enzymol. 1996, 272, 44-51. (8) Reipa, V.; Mayhew, M. P.; Vilker, V. L. Proc. Natl. Acad. Sci. U.S.A. 1997, 94 13554-13558. (9) Shumyantseva, V. V.; Bulko, T. V.; Bachmann, T. T.; Bilitewski, U.; Schmid, R. D.; Archakov, A. I. Arch. Biochem. Biophys. 2000, 376, 43-48. 10.1021/ac049927y CCC: $27.50
© 2004 American Chemical Society Published on Web 09/17/2004
electrochemistry of the soluble bacterial cytochrome P450cam (CYP101) from Pseudomanas putida has been reported several times10-14 but without the proof of electrochemically driven substrate conversion. Human P450 3A4 (quinine 3-monooxygenase) is electrocatalytically active at a polycationic filmloaded gold electrode.15 Addition of verapamil or midazolam resulting in millimolar concentrations to oxygenated solutions increased the reduction current. In the case of the human P450 3A4, addition of peroxide had only a minor effect. However, this immobilization causes a drastic anodic potential shift to ∼+98 mV (vs NHE), indicating conformational changes of the enzyme. Recently, an electrochemical study of human P450 2E1 was published, where the electron-transfer rate for the first electron was between 1 and 10 s-1 in anaerobic solutions and in oxygenated solutions p-nitrophenol conversion to p-nitrocatechol was demonstrated.16 Clay material can form films on conductive material simply by dropping and drying of a colloidal clay suspension. Due to its thixotrophy, montmorillonite is the most frequently used material.17 Clay colloid provides a favorable microenvironment for electron transfer and catalytic reactions on electrodes.13-20 Glassy carbon electrodes modified with colloidal sodium montmorillonite nanoparticles have been reported to facilitate heterogeneous electron-transfer processes to proteins and enzymes.14,21-24 Ionic surfactants are known to generate biomembrane-like structures and were therefore used to create layered protein-polyion films.12 Direct reduction and oxidation of CYP101 in thin films of anionic and cationic surfactants on pyrographite electrodes could be demonstrated using cyclic and square wave voltammetry.10 In recent studies, the electrochemical behavior of myoglobin at surfaces modified with polyelectrolyte-ionic surfactant complexes has been described.25 For membraneous proteins, however, the procedure is not efficient. In the present work, a combination of clay and nonionic detergent is introduced to achieve a efficient direct electron exchange between the membrane enzyme CYP2B4 and electrode, (10) Zhang, Z.; Nassar, A. E. F.; Lu, Z.; Schenkman, J.; Rusling, J. J. Chem. Soc., Faraday Trans. 1997, 93, 1769-1774 (11) Kazlauskaite, J.; Westlake, A. C. G.; Wong, L. L.; Hill, H. A. O. Chem. Commun. 1996, 2189-2190. (12) Lvov, Y. M.; Lu, Z.; Schenkman, J. B.; Zu, X.; Rusling, J. F. J. Am. Chem. Soc. 1998, 120, 4073-4080. (13) Lo, K. W. K.; Wong, L. L.; Hill, H. A. O. FEBS Lett. 1999, 451, 342-346. (14) Lei, C.; Wollenberger, U.; Jung, C.; Scheller, F. W. Biochem. Biophys. Res. Commun. 2000 68, 740-744. (15) Joseph, S.; Rusling, J. F.; Lvov, Y. M.; Friedberg, T.; Fuhr, U. Biochem. Pharmacol. 2003, 65, 1817-1826. (16) Fantuzzi, A.; Fairhead, M.; Gilardi, G. J. Am. Chem. Soc. 2004, 126, 50405041. (17) Navra´tilova´, Z.; Kula, P. Electroanalysis 2003, 15, 837-846. (18) Ege, D.; Ghosh, P. K.; White, J. R.; Equey, J. F.; Bard, A. J. J. Am. Chem. Soc. 1985, 107, 5644-5652. (19) Pinnavaia, T. J. Science 1983, 220, 363-371. (20) Laszlo, P. Science 1987, 235, 1473-1477. (21) Sallez, Y.; Bianco, P.; Lojou, E. J. Electroanal. Chem. 2000, 493, 37-49 (22) Lei, C.; Wollenberger, U.; Bistolas, N.; Guiseppi-Elie, A.; Scheller, F. W. Anal. Bioanal. Chem. 2002, 372, 235-239. (23) Rusling, J. F. In Protein Architecture: Interfacing Molecular Assemblies and Immobilization Biotechnology; Lvov, Y., Mo ¨hwald, H., Eds.; Marcel Dekker: New York, 2000; pp 337-354. (24) Rusling, J. F.; Zhang, Z. In Handbook of Surfaces and Interfaces of materials, Vol. 5, Biomolecules, Biointerfaces, and Applications; Nalwa, R. W., Ed.; Academic Press: San Diego, 2001; pp 33-71. (25) Wang, L.; Hu, N. J. Colloid Interface Sci. 2001, 236, 166-172.
resulting in electrocatalytic substrate conversion. The biosensor function is shown for two selected substrates of CYP2B4. EXPERIMENTAL SECTION The microsomal fraction was isolated from the liver of male New Zealand rabbits, which were treated with 0.1% sodium phenobarbital in drinking water for one week.26 CYP2B4 was purified from the liver microsomes of phenobarbital-treated rabbits according to the literature.26,27 To remove the detergent Emulgen 913 from purified P450, CM-Sephadex C-50 column chromatography was used.26 The peak fractions having the absorbance ratio at 417 nm to that of 276 nm greater than 1.5 were collected. The CYP2B4 concentration was determined by the method of Omura and Sato,28 with the extinction coefficient of the reduced carbon monoxide complex of 450-490 ) 91 mM-1 cm-1. The protein stock solution used for the electrochemical studies was 110 µM CYP2B4 in 100 mM phosphate buffer, pH 7.4, 20% glycerol (v/v), 1 mM EDTA, and 0.1 mM dithiothreitol. For monomerization of CYP2B4, the protein stock solution was 10:1 diluted with 20% Tween 80 and incubated for 10 min at room temperature.29 The clay colloid was prepared from 24 µL of sodium montmorillonite colloid,30 1 µL of colloidal Pt,31 and 85 µL of deionized water. Electrodes were prepared with 5-µL aliquots of 1:1 (v/v) mixtures of the clay colloid with either water or 2% Tween 80 followed by incubation in CYP2B4 solution with or without 2% Tween 80 overnight for protein adsorption. Alternatively, the clay colloid was mixed 1:1 (v/v) with protein stock solution or monomerized protein solution, and 5-µL aliquots of the final mixtures were then spread on a freshly polished glassy carbon electrode (BAS, diameter 3 mm). For film formation, the electrodes were then put in a refrigerator overnight followed by a 30-min final drying under low-pressure atmosphere. Experiments were made with clay mixed with monomerized CYP2B4 if not otherwise stated. All electrochemical experiments were carried out with a computer-controlled Autolab PSTAT10 (Eco Chemie). The electrochemical cell was equipped with an Ag/AgCl 1.0 M KCl reference electrode and a platinum wire auxiliary electrode. Prior to measurement, 2 mL of the working buffer was carefully purged with nitrogen or argon for at least 20 min to remove oxygen if not stated otherwise. The working buffer was 100 mM potassium phosphate, 50 mM KCL, pH 7.4. Amperometric measurements were made in a stirred airsaturated solution at an applied potential of -500 mV (vs Ag/ AgCl). Aliquots of 80 mM aminopyrine or benzphetamine stock solution were added. All electrochemical experiments were carried out at room temperature. The potentials are all referred to a Ag/AgCl 1 M KCl reference electrode. (26) Imai, Y.; Hashimoto, C.; Satake, H.; Girardin, A.; Sato, R. J. Biochem. 1980, 88, 489-503. (27) Karuzina, I. I.; Zgoda, V. G.; Kuznetsova, G. P.; Samenkova, N. F.; Archakov, A. I. Free Radical Biol. Med. 1999, 26, 620-632. (28) Omura, T. ; Sato, R. J. Biol. Chem. 1964, 239, 2379-2385. (29) Kiselyova, O. I.; Yaminsky, I. V.; Ivanov, Y. D.; Kanaeva, I. P.; Kuznetsov, V. Y.; Archakov, A. I. Arch. Biochem. Biophys. 1999, 371, 1-7. (30) Lei, C.; Deng, J. Anal. Chem. 1996, 68, 3344-3349. (31) Gosh, P. K.; Bard, A. J. J. Am. Chem. Soc. 1983, 105, 5691-5693.
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
6047
Spectrophotometric measurements were performed with a Pharmacia LKB-Biochrom 4060 spectrophotometer and HewlettPackard 8451A diode array spectrophotometer. Product analysis (formaldehyde formation) after electrolysis of aminopyrine was performed with Nash reagent32 using 412 ) 4 mM-1 cm-1. Atomic force microscopy (AFM) experiments were carried out in a tapping mode on a multimode Solver P47H atomic force microscope (NT-MDT, Moscow, Russia). Cantilevers CSC12 produced by MicroMash were used. The resonant frequency of the cantilevers was 75-150 kHz, and the force constant was about 0.6-1.8 N/m. The calibration of the microscope by height was carried out on a calibration TGZ01 grating (MicroMash) with the 22 ( 0.5 HM step height. The total number of measured particles in each sample was no less than 600, and the number of measurements for each sample was no less than 16; i.e., there were four measurements in each of the four series. The density of protein distribution with height, F(h), was calculated as F(h) ) (Nh/N) × 100%, where Nh is the number of imaged proteins with height h and N is the total number of imaged proteins. The calculation was carried out using a step of 0.1 nm and was smoothed off at three points by use of the moving average. RESULTS AND DISCUSSION We recently reported reversible oxidation and reduction for CYP101 adsorbed at clay films on glassy carbon electrodes.14 Using the same approach for the study of the microsomal monooxygenase CYP2B4 under strictly anaerobic conditions, only a small reduction wave was observed that tends to disappear at scan rates higher than 0.1 V/s (not shown). The incorporation of CYP2B4 into clay by mixing clay and protein solution only slightly improves the electrochemical behavior. Again, only at low scan rates is a reduction current measured with a reduction peak at -430 mV (not shown). The reduction current is increased in the presence of oxygen (by stirring the background solution in an open cell) when the scan rates are not higher than 0.05 V/s. Ferrous iron binds dioxygen and forms the unstable FeII-dioxygen complex, which can accept a further electron or release superoxide and ferric iron, which is again reduced. The pronounced reduction currents are typical for an electrocatalytic oxygen reduction and have also been observed with other heme proteins including P450 enzymes.15,22-24 At higher scan rates the peaks disappear, which may be due to a slow redox process. The reason for this inefficient electron transfer to heme iron may be an oligomeric state of CYP2B43,33 since it has been observed that CYP2B4 becomes oligomeric when solubilized from the microsomal membrane.33 Nonionic detergents can easily monomerize membraneous proteins from their oligomeric state,29,32 which is important for proteins bearing a hydrophobic tail as membrane anchoring part such as CYP2B4.1,33,34 AFM data show that CYP2B4 is in its monomeric state when more than 0.33% Tween 80 is present (see Supporting Information). (32) Kanaeva, P.; Dedinskii, I. R.; Skotselyas, E. D.; Krainev, A. S.; Guleva, I. V.; Sevrukova, I. F.; Koen, Y. M.; Sevrukova, G. P.; Bachmonova, G. I.; Archakov, A. I. Arch. Biochem. Biophys. 1992, 298, 395-402. (33) Wagner, S. L.; Dean, W. L.; Gray, R. D. J. Biol. Chem. 1984, 259, 23902395. (34) Werck-Reichhart, D.; Feyereisen, R. Genome Biol. 2000, 1, 3003.1-3003.9.
6048
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
Therefore, protein films were prepared in the presence of the nonionic detergent Tween 80. Other authors 10 described “biomembrane-like films” of synthetic lipids such as didodecyldimethylammonium bromide for the electrochemical reduction of the water-soluble nonmembranous protein CYP101. It was also shown that synthetic lipids stabilize and protect inorganic nanoparticles35,36 and enhance electron transfer for hemoglobin intercalated in such a surfactant-clay.37 Thus, such nanostructured biomembrane-like environments should facilitate the electron exchange with the protein and the electrode. We therefore combined colloidal clay nanoparticles and detergent to improve the electron transfer between glassy carbon electrode and CYP2B4. Cyclic voltammogram of a modified electrode prepared by forming a clay-detergent film to which CYP2B4 has been adsorbed shows small oxidation and reduction peaks even at scan rates above 0.5 V/s, which could not be obtained with detergentfree clay (Figure 1a). This indicates a promoting effect of Tween 80. Figure 1b shows cyclic voltammograms of electrodes prepared with CYP2B4 pretreated with nonionic detergent in oxygen-free solutions. Peaks corresponding to the one-electron heme iron P-FeII/P-FeIII conversion are obvious. Highest currents are obtained when monomerized CYP2B4 is mixed with the clay colloid. Although the peak currents are much higher than for the electrode with adsorbed monomerized CYP2B4, the peak potentials are similar. From the peak potentials, an average midpoint potential of Em ) -305 mV was calculated (Table 1). The voltammograms in the scan rate range from 0.5 to 10 V/s are almost symmetrical with equal reduction and oxidation peak heights (Figure 2). The reduction peak currents increase linearly with scan rates up to 2 V/s, which indicates that CYP2B4 performed a surface electrode reaction with a rate constant of 80 s-1 under these conditions.38 At slow scan rates (0.01-0.1 V/s), reduction is more pronounced, most probably due to residual oxygen in the layer. Integration of the oxidation peaks permits calculation of the charge and thus the concentration of electroactive molecules on the surface of the electrode. A value of 2.8 pmol (40.5 pmol/cm2, when assuming a plane surface) was estimated, corresponding to 2.6% of the total amount of the loaded enzyme to be electroactive. The charge transferred upon reduction of CYP2B4 measured at 10 V/s is only 40% of the value obtained at 0.05 V/s, but still 12% higher than the anodic value (at the same scan rate). This behavior may again be caused by residual oxygen. The amount of electroactive enzyme depends on the preparation method used. A comparison of the results for the different electrode preparations shows clearly that a film formed by a mixture of clay and monomerized CYP2B4 gives the highest amount of electroactive enzyme (Table 1). The calculated amounts of electroactive species represents only a small part of the total protein immobilized. Thus, the integration (35) Murray, R. W. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1984; Vol. 13, pp 192-238. (36) Han, X.; Chen, W.; Zhang, Z.; Dong, S.; Wang, E. Biochim. Biophys. Acta 2002, 1556, 273-277. (37) Chen, X.; Hu. N.; Zeng, Y.; Rusling, J. F.; Yang, J. Langmuir 1999, 15, 7022-7030. (38) Laviron, E. J. Electroanal. Chem. 1979, 101, 19-28.
In the presence of oxygen, the reduction peaks increased and oxidation peaks vanished completely (Figure 3). This experiment can serve as evidence of the electrocatalytic activity of cytochrome P450 based on the fast binding of oxygen to reduced ferrous cytochrome P450 (>106 M-1 s-1)3. The resulting ferrous-dioxygen complex is unstable.39 It can easily accept a second electron but may also release superoxide or alternatively peroxide, while it is itself oxidized back to the ferric form. The latter is reduced again electrochemically, which generates the increase in reduction current. The following reaction sequence describes this process (P stands for protein) and thus explains Figures 1-3: The reversible electrode reaction in absence of oxygen is shown in eq 1:
P-FeIII + 1e- a P-FeII
(1)
In the presence of oxygen, the one-electron reduction (eq 1) is followed by fast oxygen binding and introduction of a second electron (eq 3) or abstraction of reactive oxygen species (eq 4) and regeneration of the ferric state. The autoxidation (eq 4) is a very slow process15 and will not contribute substantially.
P-FeII + O2 f P-FeII-O2 Figure 1. Effect of detergent and immobilization on the cyclic voltammograms of CYP2B4 electrodes in nitrogen atmosphere. (a) CYP2B4 was adsorbed to clay-modified glassy carbon electrode, which was prepared with and without Tween 80. (b) CYP2B4 was preincubated with Tween 80 and then adsorbed to clay-modified glassy carbon or mixed with clay and then dropped onto glassy carbon. For comparison, CYP2B4 adsorbed to clay-modified glassy carbon in the absence of detergent (clay, 2B4 ads) is shown in both pictures. Measurements were performed with 100 mM phosphate buffer, 50 mM KCl, pH 7.4 at 1 V/s scan rate.
e- + H
+
(2)
H+
P-FeII-O2 98 P-FeIII-OOH 98 P-FeIII + H2O2 (3) P-FeII-O2 f P-FeIII-O2•- f P-FeIII + O2•-
(4)
Electrocatalytic reduction of the generated peroxide will also lead to a catalytic current (eq 5). 2e- + 2H +
P-FeIII + H2O2 f PsFeIVdO 98 P-FeIII + H2O (5)
Table 1. Comparison of Electrochemical Parameters of CYP2B4-Modified Electrodes Prepared with and without Tween 80 by Adsorption to a Clay Film on Glassy Carbon or Mixing the Components Prior to Film Formation
electrode modification
formal potential E°′, mV vs Ag/AgCl
heterogen electrontransfer rate const, ks, s-1
surface loading,a Γ, × 1012 mol cm-2
clay, 2B4 adsorbed clay/2B4 mixed clay/Tween, 2B4 adsorbed clay, 2B4/Tween adsorbed clay/2B4/Tween mixed
ndb nd -292.5 ( 12.7 -305.5 ( 2.1 -295.2 ( 9.6
>144c >114c 80
4.3 6.3 40.5
a Amount of electroactive CYP2B4; average of six peaks. b nd, not detectable. c No maximum value reached in the scan range 0.05-10 V/s; value calculated for 10 V/s.
of CYP2B4 into clay results either in an inactivation or only few molecules are in the right orientation for electron transfer. Rusling and Zhang24 have described that there is a distribution of electrontransfer rates resulting from the distribution of orientations of the protein in a film.
The ratio of the (catalytic) oxygen reduction current (in the presence of oxygen, I(O2) to the reduction current in argon-purged buffer (Ired) is a measure of the catalytic efficiency of P450 because only the reduced (ferrous) hemoprotein can bind oxygen for the second electron transfer. The catalytic efficiency I(O2)/Ired of CYP2B4 decreases exponentially with increasing scan rate (Figure 4) since the catalytic oxygen reduction is rate limiting. It is known that montmorillonite nanoparticles are planar polyanions40 and that the clay matrix is ionically conductive. The interaction of charged species with clay is based on the ionexchange properties of the clay material. It has been discussed that the charges are most likely transferred by ions but the lattice metal ions may also participate in redox reactions. Earlier we showed that membraneous CYP2B4 can be reduced electrochemically in the presence of riboflavin, which serves as an electron-tunneling relay. Such riboflavin-modified flavocytochrome CYP2B4 was reduced at the potential -500 mV (vs (39) Guengerich, F. P. Chem. Res. Toxicol. 2001, 14, 611-650. (40) Bard, A. J.; Mallouk, T. In Molecular design of electrode surfaces; Murray, R. W., Ed.; Techniques of Chemistry Series 22; John Wiley: New York, 1992; pp 271-312.
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
6049
Figure 2. Cyclic voltammograms of CYP2B4-electrode (monomerized CYP2B4 mixed with clay) at different scan rates. Nitrogen-saturated background solution: 100 mM phosphate buffer, 50 mM KCl, pH 7.4, and the resulting scan rate dependence of the peak currents.
Figure 3. Effect of oxygen and aminopyrine on the CYP2B4 electrode with mixed film of CYP2B4, Tween 80, and clay. Cyclic voltammograms in buffer that was argon saturated, air saturated, and air saturated after introduction of 8 mM aminopyrine. Measurements in 100 mM phosphate buffer, 50 mM KCl, pH 7.4; scan rate 10 V/s.
Figure 4. Influence of scan rates on catalytic efficiency I(O2)/Ired of CYP2B4 in 100 mM phosphate buffer, containing 50 mM KCl (9), and 20 mM phosphate buffer, 10 mM KCl (O), pH 7.4.
Ag/AgCl) on rhodium-graphite screen-printed electrodes.41 In the presence of substrates such as aminopyrine, aniline, and 7-pentoxyresorufin, N-demethylation, para hydroxylation, and O-dealkylation reactions proceeded as was confirmed by product analysis. With the present approach, we established mediator-free electrochemically driven catalysis of CYP2B4 in the presence of aminopyrine in air-saturated buffer. Figure 3 shows cyclic voltammetry of CYP2B4 in the presence of argon, oxygen, and amino(41) Shumyantseva, V. V.; Bulko, T. V.; Usanov, S.; Schmid, R. D.; Nicolini, C.; Archakov, A. I. J. Inorg. Biochem. 2001, 87, 185-190.
6050 Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
Figure 5. Dependence of the amperometric response of aminopyrine and benzphetamine when a constant potential of E ) -500 mV was applied to the CYP2B4 electrode.
pyrine. When aminopyrine was added to the air-saturated buffer solution, there was an increase of the reduction current. Without enzyme, aminopyrine does not cause any voltammetric or amperometric response. Experiments with aminopyrine in 100 mM phosphate buffer gave a ratio of the catalytically enhanced current in the presence of aminopyrine and oxygen to the uncatalyzed current Icat(aminopyrine)/Ired ) 2.62 and I(O2)/Ired ) 2.21. The efficiency of this biosensor is Icat(aminopyrine)/Ired:I(O2)/Ired ) Icat(aminopyrine)/I(O2) ) 1.18. This value is comparable with the efficiency of the CYP3A4 biosensor where the ratio of Icat (verapamil)/I(O2) was 1.38.15 The chronoamperometric approach can be an elegant way for the construction of drug biosensors. This is shown for two typical substrates of CYP2B4 (Figure 5). Reduction currents are measured as a function of aminopyrine and benzphetamine concentration. The response time to reach 95% of the steady-state value was ∼30 s. The detection limit of the CYP2B4-electrode is 0.4 mM aminopyrine and 0.2 mM benzphetamine. The biosensor response approaches saturation at ∼8 mM aminopyrine, and maximum response for benzphetamine is reached at 1.2 mM. The sensitivity is 0.2 µA/mM aminopyrine and 0.02 µA/mM benzphetamine. The Kmapp are 1.13 and 0.37 mM for aminopyrine and benzphetamine, respectively. For the native microsomal system the Km values are 1.8 mM aminopyrine and 0.05 mM benzphetamine. The sensor can be used for ∼15 measurements and should therefore be considered disposable. To verify that the response is due to an enzymatic conversion of aminopyrine, a product analysis was done after controlled
Figure 6. Inhibition of aminopyrine reduction current by methyrapone. Cyclic voltammogram in (a) aerated 100 mM phosphate buffer, 50 mM KCl, pH 7.4, with (b) 0.5 mM metyrapon and (c) 4 mM aminopyrine and 0.5 mM metyrapon at scan rate of 1 V/s.
potential electrolysis at the potential E ) -500 mV in the presence of 8 mM aminopyrine. The product of aminopyrine N-demethylation is formaldehyde, which was determined with Nash reagent.32 After 1 h of controlled potential electrolysis, formaldehyde could be measured and the electro(chemical) catalytic constant was calculated to be kcat ) 0.04 min-1. Taking into consideration that only 2.6% of the total amount of enzyme is electroactive, the catalytic constant is kcat )1.54 min-1, which is almost comparable with the natural microsomal system (kcat )3.5 min-1).41 This suggests that CYP2B4 possesses catalytic activity in the presence of substrate. In a further experiment, an inhibitor to the enzyme was added to the aerated solution. A blocking of the catalytic substrate conversion would support that the aminopyrine response is due to the action of CYP2B4. Metyrapon is an inhibitor of CYP2B4, and addition of aminopyrine does not increase the current significantly when the enzyme is inhibited (Figure 6) as it does in the noninhibited case depicted in Figure 3. The catalytic oxygen reduction is also not reduced as was observed earlier. Furthermore, if a substance that is not a substrate of CYP2B4, such as camphor, is added (no inhibitor present), no effect on the reduction current is observed. Therefore, it may be assumed that the response upon substrate addition to the inhibitor-free aerated solution (Figure 3 and Figure 5) is due to bioelectrocatalytic substrate turnover. Thus, a (simplified) reaction scheme can be drawn that illustrates possible reaction ways and also explains both the bioelectrocatalytic oxygen reduction current and the bioelectrocatalytic substrate conversion (Figure 7). After substrate binding (reaction 1), the first electron is introduced (reaction 2) followed by oxygen binding (reaction 3). The ferrous dioxygen complex accepts a second electron (reaction 4), and the highly reactive peroxy intermediate is formed. The input of protons to this intermediate can result in the cleavage of an O-O bond, producing a high-valence iron-oxygen complex, which is now reactive enough to insert an oxygen atom into the substrate. Aminopyrine itself is unable to reduce the high-valence PsFeVdO species directly.1 Dissociation of ROH (product) then restores the P450 to the starting ferric state (reaction 5). In the NADPH-dependent P450 reaction cycle, in a homogeneous system a low efficiency of substrate conversion is associated with the release of reactive oxygen species.39 The uncoupled formation of peroxide (reaction 7) has also been observed with other isoenzymes.3,15 The uncoupling occurs on the level of the
Figure 7. General scheme of the electrocatalytic reaction of CYP2B4 in the presence of oxygen. The presence of bound substrate is marked by (RH). P is for the protein. Several reaction steps of the catalytic P450 cycle are reduced to a single reaction for better illustration of the possible processes.
dioxygen (reaction 6) or peroxy complex (reaction 7). This catalytic oxygen reduction regenerates the ferric enzyme, which may again be reduced (reaction 2). As a result, the electrochemical reduction will be enhanced. Furthermore, peroxide may substitute for oxygen and reducing equivalents in what is termed the shunt pathway.42 This reaction, however, runs with no consumption of electrons and will therefore not lead to a catalytic reduction current. From the net current alone it is not possible to elucidate the reaction way. The two reactions, catalytic oxygen reduction and catalytic substrate conversion, consume the same number of electrons per mole. Therefore, the substrate conversion will not result in a higher electron uptake per cycle and increase of reduction current should be a result of a increase in rate. Substrate binding (reaction 1) stimulates the first reduction step (reaction 2). The rates are about 50 and 18 s-1 for reactions 1 and 2 in the NADPH-dependent reaction of the microsomal system respectively.3 At our modified electrode, comparable values were determined (Table 1). The oxygen binding is almost diffusion limited3,39 (>106 M-1 s-1)3 and thus not rate-limiting at all. The introduction of the second electron (reaction 4) is ∼10 times slower (2-7s-1)3 than the first reduction (reaction 2). In a recent model of rate-limiting steps in cytochrome P450 catalysis43 of CYP2D6-catalyzed oxidation of 4-methoxyphenethylamine, it was proposed that the “uncoupled” oxygen reduction rates (50-65 min-1) are ∼1 order of magnitude smaller than the rate of product formation (170-800 min-1). In addition, it was observed that the NADPH-dependent CYP2B4 reaction in homogeneous solution shows 40 and 27% coupling for benzphetamine and aminopyrine conversion, respectively,44 when the heme was stimulated to be in high-spin state (for example, by cytochrome b5 or lipids). In anology to this work, the small difference between catalytic oxygen reduction current in the absence and in the presence of aminopyrine indicates that “futile” oxygen reduction is slower than the substrate conversion, which also consumes electrons. Further studies will be required to understand the details of the electrocatalytic P450-reaction. (42) Porter, T. D.; Coon, M. J. J. Biol. Chem. 1991, 266, 13469-13472. (43) Guengerich, F. P. Biol. Chem. 2002, 383, 1553-1564. (44) Reed, J. R.; Hollenberg, P. F. J. Inorg. Biochem. 2003, 93, 152-160.
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
6051
In summary, with the presented bioelectrochemical technique it is possible to investigate potential substrates or inhibitors of appropriate isoforms of P450 enzymes. These results indicate that for a membrane protein the hybrid clay-detergent-film electrode is most effective when the protein is incubated first in the detergent and then a mixture with clay is subjected onto an electrode. Further investigations will be carried out also on structural aspects of film formation, effects of detergent charge and cations, and the extension to other membrane proteins. CONCLUSION The clay-modified film electrode resulting from the colloidal clay nanoparticles in the presence of detergent is an effective approach for the investigation of the direct electron transfer between cytochromes P450 and electrode. Potential applications of P450 electrochemical activity are catalytic reactions and the
6052
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
construction of biosensors for screening potential substrates and inhibitors of these hemoproteins. ACKNOWLEDGMENT The authors are grateful to Fonds der Chemischen Industrie (0400137) and the European Commission for financial support (QLK3-CT-2000-01481). U.W. thanks Chenghong Lei for providing montmorillonite. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review January 12, 2004. Accepted August 5, 2004. AC049927Y
