Characterization of human liver cytochromes P450 by combining the biochemical and proteomic approaches.

Highly purified human liver microsomes were processed by a combination of the biochemical and proteomic methods. Microsomes were purified from the morphologically normal liver tissue obtained from the resected and discarded masses of surrounding liver upon surgical treatment for hemangioma (control) or hepatic metastases arising from colon cancer (pathology). Proteins of each sample were separated by two-dimensional (2-DE) and one-dimensional electrophoresis (1-DE); selected gel regions were excised, in-gel digested and analyzed by matrix-assisted laser desorption-ionization time-of-flight (MALDI-TOF) mass spectrometry. Analysis of collected fingerprints has revealed a total of 13 microsomal membrane proteins involved in the biotransformation of xenobiotics. These were disulfide isomerase, flavine monooxygenase, NADPH-cytochrome P450 reductase and 10 cytochrome P450 forms, namely: CYPs 1B1, 2A6, 2E1, 2C8, 2C9, 2C10, 2D6, 3A4, 4A11, 4F2. These same samples were characterized by the enzymatic assays using the marker substrates for CYPs 1A, 2B, 3A4, 2C and 2E1. Correlations between mass spectrometric data and enzymatic activities were investigated to demonstrate the manner in which the functional and structural aspects of proteomics meet each other in the field of cytochromes P450.

Proteomic and biochemical analysis of the mouse liver microsomes.

The efficiency of the proteomic approach for the revelation of proteins, including components of the liver microsomal monooxygenase system (cytochromes b5 and P450) was demonstrated. The liver microsomes and their ghosts (i.e. membranes devoid of "ballast" proteins) were prepared from the control and phenobarbital-treated mice. Microsomes and their ghosts were characterized using the conventional biochemical assay and analysed by one- and two-dimensional electrophoresis (1-DE and 2-DE, respectively) coupled with MALDI-TOF peptide mass fingerprinting procedure. Catalytic activity of cytochromes P450 was measured using specific fluorogenic substrates for CYP1A, CYP2A, CYP2B and CYP2C families. The protein composition of control and phenobarbital-induced ghosts was analysed. The proteomic 2D-based protein separation method enabled us to reveal up to 1005 proteins, the majority of them being soluble. Among the 34 identified proteins, the cytochrome b5-like protein was revealed; however, cytochromes P450 appeared to be undetectable under 2-DE separation conditions. The separation of microsomal ghosts proteins by 1-DE, followed by mass-spectrometric analysis of bands from the 45 to 66 kDa gel range made it possible to identify hydrophobic proteins including cytochromes P450 (CYP2A4 and CYP2A5) and dimethylaniline monooxygenase. The high O-deethylation rate of 7-ethoxycoumarin-a substrate for rodent CYPs 2A and 2B, in particular for CYP2A5-was observed, in agreement with the results of mass-spectrometric identification. Collectively, the data obtained indicate that a combination of enzyme activity assays and various protein separation techniques coupled with mass-spectrometric protein identification allows a more comprehensive insight into the machinery of the cellular detoxifying system.

[Study of the mouse liver microsomes by the methods of proteome analysis]. / Issledovanie mikrosom kletok pecheni myshi metodami poreomnogo analiza.

Proteome maps of microsomes and their ghosts (i.e. membranes purified from "ballast" proteins) were obtained using highly purified mouse liver microsomes. Comparative analysis of protein composition of ghosts without and after the induction with phenobarbital (cytochromes P450 inducer) by using 1D- and 2D-electrophoresis and MALDI-TOF-mass-spectrometry revealed more than 30 new proteins, in the course of induction in the 45-60 kDa range (corresponding to the mol. weights of cytochromes P450). In the 17 kDa range (corresponding to the mol. wt. of cytochrome b5) there were 4 additional protein stains about 20 proteins disappeared over the entire electrophoregram). Separation of microsomal ghosts proteins by 1D electrophoresis followed by mass-spectrometric analysis allowed to identify cytochromes P450. The present investigation demonstrates the efficiency of different proteomic methods combination (1D- and 2D-electrophoresis, mass-spectrometry, bioinformatics and determination of the enzyme activities) for cytochromes P450 identification and elucidation of their functioning in different animal tissues and then extrapolating this approach to humans.

Revelation of ternary complexes between redox partners in cytochrome P450-containing monooxygenase systems by the optical biosensor method.

Formation of binary and ternary complexes in the water-soluble cytochrome P450cam (P450cam)-containing as well as in the membrane P4502B4(2B4)- and the mixed P450scc-containing monooxygenase systems was investigated in real time by the 'resonant mirror' optical biosensor method. It was shown that the inter-protein electron transfer occurs not only during complex formation but also upon random collision--as was the case with the d-Fp/d-b5 pair (2B4 system). Binary complexes may be either facilitative to electron transfer (electron-transfer complexes) or prohibitive to it (non-productive complexes). Although the binary PdR/Pd and P450cam/Pd complex formation (within the P450cam-system) as well as the binary AdR/Ad and P450scc/Ad complex formation (within the P450scc-system) does occur, the lifetimes of these complexes formed are several orders of magnitude higher than the time required for realization of a complete hydroxylation cycle. At the same time, the lifetimes of the ternary PdR/Pd/P450cam and AdR/Ad/P450scc complexes are sufficient to permit the realization of a complete hydroxylation cycle in either of these systems. For the membrane P450 2B4 system, the formation of both the binary (Fp/2B4 and 2B4/b5) and ternary (Fp/2B4/b5) complexes was registered. The lifetimes of the binary Fp/2B4 and the ternary Fp/2B4/b5 complexes are sufficient for realization of a complete hydroxylation cycle in each of them.

Molecular recognition in the p450cam monooxygenase system: direct monitoring of protein-protein interactions by using optical biosensor.

A real-time optical biosensor study on the interactions between putidaredoxin reductase (PdR), putidaredoxin (Pd), and cytochrome P450cam (P450cam) within the P450cam system was conducted. The binary Pd/P450cam and Pd/PdR complexes were revealed and kinetically characterized. The dominant role of electrostatic interactions in formation of productive electron transfer complexes was demonstrated. It was found that Pd/P450cam complex formation and decay obeys biphasic kinetics in contrast to the monophasic one for complexes formed by other redox partners within the system. Evidence for PdR/P450cam complex formation was obtained. It was found that, in contrast to Pd, which binds only to its redox partners, PdR and P450cam were able to form PdR/PdR and P450cam/P450cam complexes. A ternary PdR/Pd/P450cam complex was also registered. Its lifetime was sufficient to permit up to 60 turnovers to occur. The binding of Pd to P450cam and to PdR within the ternary complex occurred at distinct sites, with Pd serving as a bridge between the two proteins.

Optical biosensor investigation of interactions of biomembrane and water-soluble cytochromes P450 and their redox partners with covalently immobilized phosphatidylethanolamine layers.

A phospholipid-containing biochip was created by covalently immobilizing phospholipids on the optical biosensor's aminosilane cuvette and employed to monitor the interactions of the membrane and water-soluble proteins in cytochrome P450-containing monooxygenase systems with planary layers of dilauroylphosphatidylethanolamine (DLPE) and distearoylphosphatidylethanolamine (DSPE), differing in acyl chain length. It was shown that the full-length membrane proteins-cytochrome P4502B4 (d-2B4), cytochrome b5 (d-b5) and NADPH-cytochrome P450 reductase (d-Fp)-readily incorporated into the phospholipids. The incorporation was largely due to hydrophobic interactions of membranous protein fragments with the phospholipid layer. However, electrostatic forces were also but not always involved in the incorporation process. They promoted d-Fp incorporation but had no effect on d-b5 incorporation. In low ionic strength buffer, no incorporation of these two proteins into the DSPE lipid layer was observable. Incorporation of d-b5 into the DLPE layer was abruptly increased at temperatures exceeding phospholipid phase transition point. Incorporation of d-2B4 was dependent on its aggregation state and decreased with increasing protein aggregability. Water-soluble proteins either would not interact with the phospholipid layer (adrenodoxin) or would bind to the layer at the cost of only electrostatic (albumin) or both electrostatic and hydrophobic (P450cam) interactions.

Optical biosensor study of ternary complex formation in a cytochrome P4502B4 system.

The optical biosensor method was used for the revelation of ternary complexes, formed by the full-length NADPH-cytochrome P450 reductase (d-Fp) and cytochromes P4502B4 (d-2B4) and b5 (d-b5) in the course of their interactions within the reconstituted d-2B4-containing system. Based on the lack of competition between d-b5 and d-Fp for the binding sites on immobilized 2B4 (3) as well as on the analysis of data obtained in the three proteins' dissociation reactions, the possibility of formation of ternary complexes through the interactions between membranous hydrophobic fragments of proteins was substantiated. All the complexes obtained were productive.

AFM study of membrane proteins, cytochrome P450 2B4, and NADPH-cytochrome P450 reductase and their complex formation.

The application of the AFM technique for visualization of membrane proteins and for measuring their dimensions was demonstrated. The AFM images of the microsomal monooxygenase system components-cytochrome P450 2B4 and NADPH-cytochrome P450 reductase-were obtained by using two types of supports-hydrophobic, highly oriented pyrolytic graphite (HOPG) and hydrophilic mica. It was shown that hemo- and flavoprotein monomers and oligomers can be adsorbed to and visualized on HOPG. On the negatively charged mica matrix, flavoprotein oligomers dissociated to monomers while hemoprotein oligomers dissociated into less aggregated particles. The images of cytochrome P450 2B4 and NADPH-cytochrome P450 reductase monomers were about 3 and 5 nm high, respectively, while the images of oligomeric forms of these proteins were about 10 and 8 nm high, respectively. We were able to observe the binary complexes composed of monomeric proteins, cytochrome P450 2B4 and its reductase and to measure the heights of these complexes (7 nm). The method is applicable for visualization of not only individual proteins but also their complexes.

The optical biosensor studies on the role of hydrophobic tails of NADPH-cytochrome P450 reductase and cytochromes P450 2B4 and b5 upon productive complex formation within a monomeric reconstituted system.

The optical biosensor study of interaction between microsomal proteins-NADPH-cytochrome P450 reductase, cytochrome P450 2B4, and cytochrome b5-was carried out in the monomeric reconstituted system in the absence of phospholipids. The formation of individual complexes was kinetically characterized and their association and dissociation rate constants were determined. The association rate constants for the complexes formed were found to be close to the diffusiion limit-(0.5-4) x 10(6) M-1 s-1-while their dissociation rate constants did not exceed 0.5 s-1. It was shown that the interprotein electron transfer can occur both through complex formation and due to random collision. The dominant role of hydrophobic membraneous protein fragments in formation of productive electron transfer complexes was demonstrated.

An optical biosensor study of the interaction parameters and role of hydrophobic tails of cytochrome P450 2B4, b5 and NADPH-flavoprotein in complex formation.

The real-time interactions of membrane proteins - cytochrome P450 2B4, NADPH cytochrome P450 reductase and cytochrome b5 - were studied by use of an optical biosensor system. The association and dissociation rate constants for the individual complexes were measured and the affinities of the redox partners for each other were estimated. The association rate constants of these complexes were found to be close to the diffusion limit and their dissociation rate constants were in the order of 1s-1. A dominant role of the interaction of the membraneous hydrophobic fragments in the formation of productive electron transferring complexes between the proteins was demonstrated.

Interactions of cytochrome P450 2B4 with NADPH-cytochrome P450 reductase studied by fluorescent probe.

A new method for monitoring the formation of the cytochrome P450 complexes with NADPH-cytochrome P450 reductase (NCPR) is introduced. The method is based on the quenching of fluorescence of NCPR labelled with 7-ethylamino-3-(4'-maleimidilphenyl)-4-methylcoumarin maleimide (CPM). In a monomerized soluble reconstituted system in the absence of phospholipid, cytochrome P450 2B4 and NCPRcpm were shown to form 1:1 complexes with a Kd of 0.038 microM. Formation of the complex follows the kinetics of reversible second order transition with k(on) = 6.5 10(5) M-1 s-1. Application of high hydrostatic pressure induces dissociation of the complex (delta V degrees = -65 mL/mol). Succinylation of the hemoprotein increases the value of Kd to 0.5 microM primarily by decreasing k(on). In contrast to what was shown for intact 2B4, rising pressure does not take apart succinylated hemoprotein and NCPRcpm molecules, but causes some internal transition in their complex that diminishes the quenching. This transition is characterised by a very large volume change (delta V degrees = -155 mL/mol). The following conclusions were drawn: 1) a molecule of 2B4 contains two distinct contact regions involved in the interactions with NCPR. Only one of these regions is polar and highly hydrated in unbound hemoprotein; 2) interactions of the polar regions of 2B4 and NCPR are necessary to bring CPM-labelled cysteine of NCPR in short distance of the heme of 2B4; and 3) some of the lysine residues located in the proximity of the polar binding regions are apparently involved in the formation of the internal salt bridges in the molecule of 2B4.

Catalytic activity of cytochrome P4501A2 in reconstituted system with Emulgen 913.

Physicochemical properties and catalytic activity of cytochrome P4501A2 in the reconstituted system with Emulgen 913 were studied. The formation of cytochrome P4501A2 monomers was shown by gel filtration at an Emulgen concentration of 8 g/liter. The catalytic activity of the monomeric monooxygenase system was low. The maximum rate of the 7-ethoxyresorufin O-deethylation reaction, 150 pmol resorufin min-1 nmol-1 cytochrome P4501A2, was observed at an Emulgen concentration of 0.1 g/liter when cytochrome P4501A2 pentamers predominated. The effects of Emulgen on cytochrome P4501A2 cumene hydroperoxide-dependent peroxidase activity, on its affinity to substrate and to cytochrome b5, and on the rate constants of dithionite-dependent reduction were insignificant. Study of the NADPH-dependent reduction of cytochrome P4501A2 in the reconstituted system showed that the rate constant and reduction level of cytochrome P4501A2 were always higher when the reaction was initiated by NADPH than when it was initiated by NADPH-cytochrome P450 reductase. This indicated that the reduction reaction initiated by the reductase was limited by a step in the cytochrome P4501A2 and reductase interaction. Correlation between 7-ethoxyresorufin O-deethylase activity and reduction level of cytochrome P4501A2 was found.

Comparative study of monomeric reconstituted and membrane microsomal monooxygenase systems of the rabbit liver. II. Kinetic parameters of reductase and monooxygenase reactions.

The kinetic parameters of NADPH-dependent cytochrome P450 LM2 (2B4) reduction and substrate oxidation in the monomeric reconstituted system, consisting of purified NADPH-cytochrome P450 reductase and cytochrome P450 LM2 monomers, and in phenobarbital-induced rabbit liver microsomes were compared. In the absence of benzphetamine, NADPH-dependent reduction of cytochrome P450 LM2 was monophasic in the monomeric reconstituted system and biphasic in the microsomes. The presence of the substrate in the monomeric reconstituted system caused the appearance of the fast phase. In this system substrate-free cytochrome P450 LM2 was entirely low-spin, and the addition of benzphetamine shifted the spin equilibrium to a high state very weakly. No correlation between high-spin content and the proportion of the fast phase of NADPH-dependent LM2 reduction was found in the system. Vmax values for the oxidation of type I substrates (benzphetamine, dimethylaniline, aminopyrine) in the monomeric reconstituted system were higher or the same as in the microsomes, whereas Km values for the substrates and NADPH were lower in the microsomes. Maximal activity of the monomeric reconstituted system was observed at a 1:1 NADPH-cytochrome P450 reductase/cytochrome P450 LM2 ratio. Measurements of benzphetamine oxidation as a function of NADPH-cytochrome P450 reductase/cytochrome P450 LM2 ratio at a constant total protein concentration allowed the Kd of the NADPH-cytochrome P450 reductase/cytochrome P450 LM2 complex to be estimated as 6.4 +/- 0.5 microM. Complex formation between the NADPH-cytochrome P450 reductase and cytochrome P450 LM2 monomers was not detected by recording the difference binding spectra of the reductase monomers with LM2 monomers or by treatment the mixture of the monomers of the proteins with the crosslinking reagent, water-soluble carbodiimide.

Comparative study of monomeric reconstituted and membrane microsomal monooxygenase systems of the rabbit liver. I. Properties of NADPH-cytochrome P450 reductase and cytochrome P450 LM2 (2B4) monomers.

Oligomers and monomers of NADPH-cytochrome P450 reductase and cytochrome P450 LM2 (2B4) isolated from the liver microsomes of phenobarbital-treated rabbits were examined for physicochemical properties and catalytic activities. As measured using laser correlation spectroscopy the particle sizes of NADPH-cytochrome P450 reductase and cytochrome P450 LM2 oligomers were 14.8 +/- 1.7 and 19.2 +/- 1.4 nm, respectively. Twenty-four-hour incubation with Emulgen 913 at 4 degrees C at a molar ratio of 1:100 led to the monomerization of NADPH-cytochrome P450 reductase and cytochrome P450 LM2 oligomers, the particle sizes diminishing to 6.1 +/- 1.3 and 5.2 +/- 0.4 nm, respectively. The thermal stability of NADPH-cytochrome P450 reductase monomers was the same as that of oligomers, whereas cytochrome P450 LM2 monomers were less thermostable than oligomers and cytochrome P450 in microsomes. Similar to cytochrome P450 LM2 oligomers and the microsomal hemoprotein, cytochrome P450 LM2 monomers formed complexes with type I and II substrates, but with Kd values higher than those of microsomes and cytochrome P450 LM2 oligomers. Kinetic parameters (Vmax and Km) of H2O2- and cumene hydroperoxide-dependent oxidation of benzphetamine and aniline in the presence of cytochrome P450 LM2 oligomers, monomers, and microsomes were determined. Peroxidase activities of the oligomers and monomers were the same, but were lower than those of microsomes. Thus the substitution of protein-protein interactions in cytochrome P450 LM2 oligomers with protein-detergent interactions in the monomers did not influence the catalytic properties of the hemoprotein.

Cytochrome C (Fe2+) as a competitive inhibitor of NADPH-dependent reduction of cytochrome P450 LM2: locating protein-protein interaction sites in microsomal electron carriers.

The kinetics of NADPH-dependent reduction of cytochrome P450 LM2 in the soluble monomeric reconstituted system in the absence of any substrate is shown to be monophasic. We show that ferrous cytochrome c acts as a competitive inhibitor of the reduction. In the presence of 1 mM benzphetamine an additional extremely fast phase was observed. Under these conditions ferrous cytochrome c was found to be a competitive inhibitor of the slow phase of the reduction process, which accounted for 80% of the total reduction amplitude. Inhibition experiments yield a dissociation constant for the LM2-reductase complex of 3.0 +/- 1.5 microM. This constant was the same both in the presence and in the absence of benzphetamine. Based on these data we conclude that cytochromes P450 and c bind to the same center on the NADPH-cytochrome P450 reductase molecule. Comparative analysis of the amino acid sequences reveals a detectable similarity between cytochrome c and cytochrome P450 LM2 at positions 68-87 and 121-145, respectively. In addition, a substantial similarity was shown for sequence fragments 204-224 of NADPH-cytochrome P450 reductase and 40-60 of cytochrome b5. Based on these findings a hypothesis for the location of the centers of intermolecular interactions on the molecules of cytochrome P450 LM2 and NADPH-cytochrome P450 reductase is proposed.

Substrate access channel geometry of soluble and membrane-bound cytochromes P450 as studied by interactions with type II substrate analogues.

The problems of alterations in the tertiary structure at the cytochrome P450 active site after isolation from the microsomal membrane and comparative analysis of the structures of the active sites of membrane-bound P450 and soluble P450cam have been studied in terms of using bifunctional compounds (I-IV). These amphiphilic compounds contain a pyridine radical, an aliphatic chain of variable length (n), and diphosphonic acid at the end of the molecule. There exists an optimal length (n) at which the interaction between I-IV and P450 is rather efficient. Comparison of the data on such interactions with microsomal P450, as well as P450 isolated from the membrane in oligomeric and monomeric states, and P450cam allows the estimation of the distance between the Fe3+ ion in the active site and the charged residues (Lys/Arg) on the enzyme surface (approximately 17 A for all P450s).

Protein-protein interactions in microsomal cytochrome P-450 isozyme LM2 and their effect on substrate binding.

The effects of protein-protein interactions and substrate binding on the structure of the active site of rabbit liver microsomal cytochrome P-450 LM2 have been analyzed by resonance Raman spectroscopy of the monomeric and oligomeric protein in solution. Also H2O2-dependent catalytic activities of the two states have been compared. The two vinyl substituents of the heme exhibit different orientations, as indicated by the frequencies and intensities of their stretching vibrations. One group lies in the plane of the heme and remains unchanged in the two states of cytochrome P-450 LM2, the other is tilted out of the plane. The tilting angle in oligomers was smaller than in monomers. These vinyl stretching modes together with some porphyrin modes, were found to be sensitive indicators of the quaternary structure and of substrate binding. In both the oligomer and the monomer, substrate binding causes changes of the relative intensities of some porphyrin modes and the vinyl stretching vibrations which may reflect modifications of the electronic transitions due to hydrophobic interactions between the bound substrate and the heme. In contrast to the monomeric cytochrome P-450 LM2, benzphetamine binding to the oligomers of this isozyme additionally produces a shift of the spin-state equilibrium. This indicates that in the oligomer the substrate-binding pocket is converted by protein-protein interaction to a structure that forces substrates to interfere with the sixth ligands, inducing an increase of the five-coordinated high-spin configuration. In the monomer the substrate-binding pocket can accommodate benzphetamine without affecting the spin state. Binding of imidazole to the monomeric and oligomeric cytochrome P-450 LM2 produces essentially the same resonance Raman spectra. Apparently the replacement of the native sixth ligand by imidazole disturbs the structure of the active site in such a way that it becomes insensitive to protein-protein interactions. H2O2-dependent N-demethylation of benzphetamine and aniline p-hydroxylation by cytochrome P-450 LM2 did not depend on its state of aggregation.

Reduction and catalytic properties of cytochrome P-450 LM2 in reconstituted system containing monomeric carriers.

The membrane microsomal monooxygenase system can be reconstituted in solution from NADPH-specific flavoprotein and cytochrome P-450 which exist in the monomeric state in the presence of Emulgen 913 at molar ratio of the proteins and detergent of 1:1:300. Oxidized and dithionite-reduced monomers of cytochrome P-450 were much less thermostable than its initial aggregates, while thermal stability of NADPH-specific flavoprotein did not depend on its aggregation state. Binding spectra of cytochrome P-450 monomers with benzphetamine were atypical and had an absorbance minimum at 422 nm only. The addition of benzphetamine and/or flavoprotein to cytochrome P-450 monomers did not cause the spin equilibrium shift and the low-spin form content was higher than 85% in all cases. Investigation of the dependence of the initial rates of NADPH-dependent cytochrome P-450 reduction and benzphetamine oxidation on the stoichiometry of the flavoprotein and cytochrome P-450 at their constant total concentration showed that the molar ratio of 1:1 was required for maximal activity. Thus this system works in full accordance with the mass action law.