Engineering of cytochrome P450 2B1 specificity. Conversion of an androgen 16 beta-hydroxylase to a 15 alpha-hydroxylase

Cytochrome P450 research and The Journal of Biological Chemistry

In honor of the 100th birthday of Dr. Herbert Tabor, JBC's Editor-in-Chief for 40 years, I will review here JBC's extensive coverage of the field of cytochrome P450 (P450) research. Research on the reactions catalyzed by these enzymes was published in JBC before it was even realized that they were P450s, i.e. they have a "pigment" with an absorption maximum at 450 nm. After the P450 pigment discovery, reported in JBC in 1962, the journal proceeded to publish the methods for measuring P450 activities and many seminal findings. Since then, the P450 field has grown extensively, with significant progress in characterizing these enzymes, including structural features, catalytic mechanisms, regulation, and many other aspects of P450 biochemistry. JBC has been the most influential journal in the P450 field. As with many other research areas, Dr. Tabor deserves a great deal of the credit for significantly advancing this burgeoning and important topic of research.

X-ray crystal structure of the cytochrome P450 2B4 active site mutant F297A in complex with clopidogrel: insights into compensatory rearrangements of the binding pocket

Prior X-ray crystal structures of cytochrome P450 2B4 revealed the pivotal role of rearrangement of the side chains of residues F206 and F297 in the active site in accommodating various inhibitors or substrates. To explore the role of these residues, 2B4 F206A and F297A were created by site-directed mutagenesis and characterized functionally. The structure of F297A with clopidogrel demonstrated the reorientation of the ligand such that the methyl ester group is oriented toward the heme, whereas the thiophene moiety now extends to the additional void in the F297A mutant. Most interestingly, movement of the I helix and several amino acid side chains within the active site was observed in apparent response to the altered binding orientation. Results of flexible docking using the 2B4 wild type or the F297A-virtual mutant positioned either the thiophene or chlorophenyl group closer to heme. However, docking of clopidogrel using the real F297A mutant or a virtual mutant with the I-helix re-positioned oriented clopidogrel preferentially with either the methyl ester or the chlorophenyl group closest to heme. The study provides insight into how the altered active site adapts to accommodate and interact with the substrate in a distinct orientation while maintaining the overall closed protein conformation.

Investigation by site-directed mutagenesis of the role of cytochrome P450 2B4 non-active-site residues in protein-ligand interactions based on crystal structures of the ligand-bound enzyme

Residues located outside the active site of cytochromes P450 2B have exhibited importance in ligand binding, structural stability and drug metabolism. However, contributions of non-active-site residues to the plasticity of these enzymes are not known. Thus, a systematic investigation was undertaken of unique residue-residue interactions found in crystal structures of P450 2B4 in complex with 4-(4-chlorophenyl)imidazole (4-CPI), a closed conformation, or in complex with bifonazole, an expanded conformation. Nineteen mutants distributed over 11 sites were constructed, expressed in Escherichia coli and purified. Most mutants showed significantly decreased expression, especially in the case of interactions found in the 4-CPI structure. Six mutants (H172A, H172F, H172Q, L437A, E474D and E474Q) were chosen for detailed functional analysis. Among these, the K(s) of H172F for bifonazole was ∼ 20 times higher than for wild-type 2B4, and the K(s) of L437A for 4-CPI was ∼ 50 times higher than for wild-type, leading to significantly altered inhibitor selectivity. Enzyme function was tested with the substrates 7-ethoxy-4-(trifluoromethyl)coumarin, 7-methoxy-4-(trifluoromethyl)coumarin and 7-benzyloxyresorufin (7-BR). H172F was inactive with all three substrates, and L437A did not turn over 7-BR. Furthermore, H172A, H172Q, E474D and E474Q showed large changes in k(cat)/K(M) for each of the three substrates, in some cases up to 50-fold. Concurrent molecular dynamics simulations yielded distances between some of the residues in these putative interaction pairs that are not consistent with contact. The results indicate that small changes in the protein scaffold lead to large differences in solution behavior and enzyme function.

A single amino acid residue, Ala 105, confers 16alpha-hydroxylase activity to human cytochrome P450 17alpha-hydroxylase/17,20 lyase

In adrenal steroidogenesis, CYP17 catalyses the 17alpha-hydroxylation of pregnenolone and progesterone and the subsequent 17,20-lyase reaction, yielding adrenal androgens. The enzyme exhibits distinctly different selectivities towards these substrates in various species. CYP17 has also been shown to exhibit 16alpha-hydroxylase activity towards progesterone in some species, with only human and chimp CYP17 catalysing the biosynthesis of substantial amounts of 16-OHprogesterone. The 16alpha-hydroxylase activity was investigated by introducing an Ala105Leu substitution into human CYP17. The converse mutation, Leu105Ala was introduced into the baboon, goat and pig enzymes. Wt human CYP17 converted approximately 30% progesterone to 16-OHprogesterone while the Ala105Leu mutant converted negligible amounts to 16-OHprogesterone ( approximately 9%), comparable to wt CYP17 of the other three species when expressed in COS-1 cells. The ratio of 17-hydroxylated products to 16-OHprogesterone of human CYP17 was 2.7 and that of the mutant human construct 10.5. Similar ratios were observed for human and goat CYP17 with the corresponding Ala or Leu residues. Although the Leu105Ala mutation of both baboon and pig CYP17 exhibited the same trend regarding the ratios, the rate of progesterone conversion was reduced. Coexpression with cytochrome b(5) significantly decreased the ratio of 17-hydroxylated products to 16-OHprogesterone in the Leu105 constructs, while effects were negligible with Ala at this position. Homology models show that Ala105 faces towards the active pocket in the predicted B'-C domain of CYP17. The smaller residue allows more flexibility of movement in the active pocket than Leu, presenting both the C16 and C17 of progesterone to the iron-oxy complex.

Evidence for the functional role of residues in the B'-C loop of baboon cytochrome P450 side-chain cleavage (CYP11A1) obtained by site-directed mutagenesis, kinetic analysis and homology modelling

To gain further insight into the structure/function relationship of cytochrome P450 side-chain cleavage (CYP11A1), this enzyme was investigated in the Cape baboon (Papio ursinus). Four constructs were cloned and characterised in non-steroidogenic mammalian COS-1 cells. Wild type recombinant baboon CYP11A1 cDNA yielded a K(m) value of 1.6 microM for 25-hydroxycholesterol. The single amino acid substitutions, I98Q and I98K resulted in a 1.7- and 2.8-fold increases in K(m) values, respectively. Conversely, the introduction of the mutation, K103A, resulted in a 1.8-fold decrease in K(m). A homology model of CYP11A1, based on the crystal structures of CYP102 and CYP2C5, revealed that residues 98 and 103 lie within the B'-C loop and contribute to the spatial orientation and structural integrity of this domain. Based on these results we propose a topological model of the CYP11A1 active pocket, which is supported by substrate docking analysis and kinetic studies.

Predictive value of comparative molecular field analysis modelling of naphthalene inhibition of human CYP2A6 and mouse CYP2A5 enzymes

The objects of this study were first to compare how well the recently constructed structure-inhibition activity relationship models of mouse CYP2A5 and human CYP2A6 predict the interaction of naphthalene in liver microsomes and secondly to study if these CYP enzymes actually oxidize naphthalene. The CoMFA model of CYP2A5 predicted the IC(50) value of naphthalene to be 42 microM (18-115 microM 95% CL) whereas in the in vitro experiment the result was 74 microM (65-83 microM) with the corresponding values for CYP2A6 being 41 microM (18-112 microM) and 25 microM (21-30 microM), respectively. Naphthalene appeared to be a competitive inhibitor both for mouse and human liver microsomal coumarin 7-hydroxylase, which is the specific probe activity for CYP2A5 and CYP2A6. The K(i)-value for the mouse enzyme was between 12-26 microM and for the human enzyme 1.2-5.6 microM. A 1-h in vitro incubation of naphthalene with human and pyrazole treated mouse liver microsomes produced more 1-naphthol than 2-naphthol. Antibody against the purified CYP2A5 inhibited 50-60% of the formation of 1-naphthol and 30-40% of the formation of 2-naphthol. These results indicate that in silico CoMFA models predict relatively well the interaction of naphthalene with CYP2A5 and CYP2A6 and that these CYPs actually oxidize naphthalene in vitro. CoMFA CYP2A5 and CYP2A6 models are thus useful as a technique for elucidating the interaction and potency of untested chemicals with these CYPs.

Molecular modelling of CYP2B6 based on homology with the CYP2C5 crystal structure: analysis of enzyme-substrate interactions

The results of homology modelling of CYP2B6 based on the CYP2C5 crystal structure is described in terms of substrates and inhibitors binding within the putative active site. In general these results are in agreement with currently available evidence from substrate metabolism, mode of inhibitor action and site-directed mutagenesis experiments within the CYP2B subfamily of enzymes. Consequently, the model based on the CYP2C5 template represents an advance on those models produced from bacterial P450s, such as CYP101 and CYP102. Quantitative Structure-Activity Relationships (QSARs) for substrates binding to CYP2B6 indicate a key role for hydrogen bonding, and lipophilic character, as determined by the log P parameter (where P is the octanol/water partition coefficient), is also of importance for explaining the variation in experimental binding affinity for CYP2B6 substrates. It is possible to estimate the binding energies for typical CYP2B6 substrates based on their properties and interactions with the enzyme, which show good concordance with experimental data in the form of apparent Km values.

A rational approach to Re-engineer cytochrome P450 2B1 regioselectivity based on the crystal structure of cytochrome P450 2C5

The regioselectivity for progesterone hydroxylation by cytochrome P450 2B1 was re-engineered based on the x-ray crystal structure of cytochrome P450 2C5. 2B1 is a high K(m) progesterone 16alpha-hydroxylase, whereas 2C5 is a low K(m) progesterone 21-hydroxylase. Initially, nine individual 2B1 active-site residues were changed to the corresponding 2C5 residues, and the mutants were purified from an Escherichia coli expression system and assayed for progesterone hydroxylation. At 150 microm progesterone, I114A, F297G, and V363L showed 5-15% of the 21-hydroxylase activity of 2C5, whereas F206V showed high activity for an unknown product and a 13-fold decrease in K(m). Therefore, a quadruple mutant, I114A/F206V/F297G/V363L (Q), was constructed that showed 60% of 2C5 progesterone 21-hydroxylase activity and 57% regioselectivity. Based on their 2C5-like testosterone hydroxylation profiles, S294D and I477F alone and in combination were added to the quadruple mutant. All three mutants showed enhanced regioselectivity (70%) for progesterone 21-hydroxylation, whereas only Q/I477F had a higher k(cat). Finally, the remaining three single mutants, V103I, V367L, and G478V, were added to Q/I477F and Q/S294D/I477F, yielding seven additional multiple mutants. Among these, Q/V103I/S294D/I477F showed the highest k(cat) (3-fold higher than that of 2C5) and 80% regioselectivity for progesterone 21-hydroxylation. Docking of progesterone into a three-dimensional model of this mutant indicated that 21-hydroxylation is favored. In conclusion, a systematic approach to convert P450 regioselectivity across subfamilies suggests that active-site residues are mainly responsible for regioselectivity differences between 2B1 and 2C5 and validates the reliability of 2B1 models based on the crystal structure of 2C5.

Essential requirements for substrate binding affinity and selectivity toward human CYP2 family enzymes

A detailed analysis of substrate selectivity within the cytochrome P450 2 (CYP2) family is reported. From a consideration of specific interactions between drug substrates for human CYP2 family enzymes and the putative active sites of CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP2E1, it is likely that the number and disposition of hydrogen bond donor/acceptors and aromatic rings within the various P450 substrate molecules determines their enzyme selectivity and binding affinity, together with directing their preferred routes of metabolism by the CYP2 enzymes concerned. Although many aliphatic residues are present in most P450 active sites, it would appear that their main contribution centers around hydrophobic interactions and desolvation processes accompanying substrate binding. Molecular modeling studies based on the recent CYP2C5 crystal structure appear to show close agreement with site-directed mutagenesis experiments and with information on substrate metabolism and selectivity within the CYP2 family.

Quantitative structure-activity relationship (QSAR) and three-dimensional QSAR analysis of a series of xanthates as inhibitors and inactivators of cytochrome P450 2B1

1. Various xanthates (R-OCS2) were found to be mechanism-based inactivators of cytochrome P450 2B1 (CYP2B1) and CYP2B6 via formation of reactive metabolites. 2. In the present study, quantitative structure-activity relationships (QSARs) were derived with inhibitory and inactivation potencies of 15 xanthates (R = two to 20 methylene groups, allyl, cyclohexyl or O-tricyclo[5.2.1.0(2,6)]dec-9-yl (D609)) against purified, reconstituted rat liver CYP2B1. Factor, regression and comparative molecular field analyses (CoMFA) were used. 3. The compounds formed two groups whose activities depended on different structural features: the first group consisted of compounds with ethyl, propyl, allyl, cyclohexyl and D609 substituents; the second involved compounds with eight to 20 methylene groups. 4. High correlation between the molecular volume and inhibitory potency of the xanthates of the second group was found. The inactivation potency in the first group correlated with the charge of the first carbon atom of R, identifying this atom as a potential target for metabolic attack. A decrease in the inactivation potency with an increase in the size of R was observed in the second group. This finding could be explained by a decreased rate of metabolism of the long alkyl chain compounds and/or by difficulty in binding of the resulting metabolite(s) to the enzyme molecule.

Structural determinants of aromatase cytochrome p450 inhibition in substrate recognition site-1

The porcine gonadal form of aromatase cytochrome P450 (P450arom) exhibits higher sensitivity to inhibition by the imidazole, etomidate, than the placental isozyme. The residue(s) responsible for this functional difference was mapped using chimeragenesis and point mutation analysis of the placental isozyme, and the kinetic analysis was conducted on native and mutant enzymes after overexpression in insect cells. The etomidate sensitivity of the placental isozyme was markedly increased by substitution of the predicted substrate recognition site-1 (SRS-1) and essentially reproduced that of the gonadal isozyme by substitution of SRS-1 and the predicted B helix. A single isoleucine (I) to methionine (M) substitution at position 133 of the placental isozyme (I(133)M) was proven to be the critical residue within SRS-1. Residue 133 is located in the B'-C loop and has been shown to be equally important in other steroid-metabolizing P450s. Single point mutations (including residues 110, 114, 120, 128, 137, and combinations thereof among others) and mutation of the entire B and C helixes were without marked effect on etomidate inhibitory sensitivity. The same mutation (I(133)M) introduced into human P450arom also markedly increased etomidate sensitivity. Mutation of Ile(133) to either alanine (I(133)A) or tyrosine (I(133)Y) decreased apparent enzyme activity, but the I(133)A mutant was sensitive to etomidate inhibition, suggesting that it is Ile(133) that decreases etomidate binding rather than Met(133) increasing enzyme sensitivity. Androstenedione turnover and affinity were similar for the I(133)M mutant and the native placental isozyme. These data suggest that Ile(133) is a contact residue in SRS-1 of P450arom, emphasize the functional conservation that exists in SRS-1 of a number of steroid-hydroxylating P450 enzymes, and suggest that substrate and inhibitor binding are dependent on different contact points to varying degrees.

Site-directed mutagenesis of cytochrome P450eryF: implications for substrate oxidation, cooperativity, and topology of the active site

The role of five active-site residues (Phe-78, Gly-91, Ser-171, Ile-174, and Leu-175) has been investigated in P450eryF, the only bacterial P450 known to show cooperativity. The residues were selected based on two-ligand-bound P450eryF structures and previous mutagenesis studies of other cytochromes P450. To better understand the role of these residues in substrate catalysis and cooperativity, each mutant was generated in the wild-type and A245T background, a substitution that enables P450eryF to oxidize testosterone and 7-benzyloxyquinoline (7-BQ). Replacement of Phe-78 with tryptophan decreased cooperativity of 9-aminophenanthrene binding, with little effect on testosterone binding or oxidation. Interestingly, substitution of Gly-91 with alanine or phenylalanine abolished the type-I spectral change elicited by testosterone and significantly decreased testosterone hydroxylation. However, G91A/A245T showed a 4-fold higher k(cat) value with 7-BQ compared with A245T. Replacement of Ser-171 with alanine or phenylalanine did not alter cooperativity of testosterone binding but significantly decreased binding affinity and oxidation of testosterone and 7-BQ. The only mutant that exhibited an increased testosterone binding affinity and increased rates of testosterone and 7-BQ oxidation was I174F. Substitution of Ile-175 with phenylalanine decreased testosterone and 7-BQ oxidation. Reaction with phenyldiazene showed that P450eryF may be much more open above pyrrole ring B than other cytochromes P450 and indicated significant changes in active-site topology in some of the mutants. The study suggests a crucial role of residues Ser-171, Ile-174, and Leu-175, which are part of a distal ligand site, in addition to the proximal Gly-91 in determining the oxidative properties of P450eryF.

The role of cytochrome 2B1 substrate recognition site residues 115, 294, 297, 298, and 362 in the oxidation of steroids and 7-alkoxycoumarins

At least two substitutions were made at each of five amino acid residues in rat cytochrome P450 2B1 that align to residues of known importance in other P450s. The mutants were histidine tagged for purification from Escherichia coli, and the proteins were assessed for testosterone and 7-alkoxycoumarin oxidation. Alteration of each of the sites studied, Phe-115, Ser-294, Phe-297, Ala-298, and Leu-362, was found to affect overall enzyme activity or the metabolite profile. In particular, most of the mutants, excluding F297A, A298G, and L362F, exhibited significantly altered ratios of 16alpha-hydroxytestosterone:16beta-hydroxytestosterone, with the most dramatic alteration being displayed by A298V. Four 7-butoxycoumarin metabolites were produced by CYP2B1, of which two, 7-hydroxycoumarin and 7-(3-hydroxybutoxy)coumarin, were formed at nearly equal rates. Several mutants, F115A, F297A, F297I, and A298V, exhibited an increased predominance of one of the metabolites. The results from this study illustrate the conservation of functionally important residues across P450 subfamilies and families.

Molecular analysis of multiple CYP6B genes from polyphagous Papilio species

Papilio glaucus (eastern tiger swallowtail) and Papilio. canadensis (Canadian tiger swallowtail) are two closely related species with broad but overlapping hostplant ranges. P. glaucus encounters toxic furanocoumarins occasionally in its diet in its rutaceous hostplants, whereas P. canadensis rarely if ever encounters these compounds. Analysis of their furanocoumarin-metabolic profiles indicates that these species induce cytochrome P450 monooxygenases (P450s) capable of metabolizing linear and angular furanocoumarins to varying degrees in response to dietary supplementation with xanthotoxin (a linear furanocoumarin). In P. glaucus, metabolism is induced to a significantly higher level than in P. canadensis. Cloning of multiple P450 genes from each species has revealed that both species contain and express two groups of P450s, designated CYP6B4 and CYP6B17, that are related to the P. glaucus CYP6B4v1 enzyme known to metabolize an array of furanocoumarins. Expression patterns of the CYP6B4 and CYP6B17 group transcripts differ in these species in both their basal and furanocoumarin-inducible levels. In P. glaucus, CYP6B4 transcripts, which are not detectable constitutively, are 311-fold induced by xanthotoxin and CYP6B17 transcripts, which are detectable constitutively, are 3-fold induced by xanthotoxin. In P. canadensis, CYP6B4 transcripts are only 8-fold induced and CYP6B17 transcripts are 13-fold induced. These findings are consistent with the postulated evolutionary history of these two species, according to which P. glaucus maintains its association with rutaceous hostplants and P. canadensis has differentiated to utilize hostplants in other families more extensively.

Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity

Cytochrome P450 (P450) enzymes catalyze a variety of reactions and convert chemicals to potentially reactive products as well as make compounds less toxic. Most of the P450 reactions are oxidations. The majority of these can be rationalized in the context of an FeO(3+) intermediate and odd electron abstraction/rebound mechanisms; however, other iron-oxygen complexes are possible and alternate chemistries can be considered. Another issue regarding P450-catalyzed reactions is the delineation of rate-limiting steps in the catalytic cycle and the contribution to reaction selectivity. In addition to the rather classical oxidations, P450s also catalyze less generally discussed reactions including reduction, desaturation, ester cleavage, ring expansion, ring formation, aldehyde scission, dehydration, ipso attack, one-electron oxidation, coupling reactions, rearrangement of fatty acid and prostaglandin hydroperoxides, and phospholipase activity. Most of these reactions are rationalized in the context of high-valent iron-oxygen intermediates and Fe(2+) reductions, but others are not and may involve acid-base catalysis. Some of these transformations are involved in the bioactivation and detoxication of xenobiotic chemicals.

Unusual oxidative chemistry of N(omega)-hydroxyarginine and N-hydroxyguanidine catalyzed at an engineered cavity in a heme peroxidase

Heme enzymes are capable of catalyzing a range of oxidative chemistry with high specificity, depending on the surrounding protein environment. We describe here a reaction catalyzed by a mutant of cytochrome c peroxidase, which is similar but distinct from those catalyzed by nitric-oxide synthase. In the R48A mutant, an expanded water-filled cavity was created above the distal heme face. N-hydroxyguanidine (NHG) but not guanidine was shown to bind in the cavity with K(d) = 8.5 mM, and coordinate to the heme to give a low spin state. Reaction of R48A with peroxide produced a Fe(IV)=O/Trp(.+) center capable of oxidizing either NHG or N(omega)-hydroxyarginine (NHA), but not arginine or guanidine, by a multi-turnover catalytic process. Oxidation of either NHG or NHA by R48A did not result in the accumulation of NO, NO(2)(-), NO(3)(-), urea, or citrulline, but instead afforded a yellow product with absorption maxima of 257 and 400 nm. Mass spectrometry of the derivatized NHA products identified the yellow species as N-nitrosoarginine. We suggest that a nitrosylating agent, possibly derived from HNO, is produced by the oxidation of one molecule of substrate. This then reacts with a second substrate molecule to form the observed N-nitroso products. This complex chemistry illustrates how the active sites of enzymes such as nitric-oxide synthase may serve to prevent alternative reactions from occurring, in addition to enabling those desired.

The importance of SRS-1 residues in catalytic specificity of human cytochrome P450 3A4

The structural basis for the regioselective hydroxylation of Delta-4-3-ketosteroids by human CYP3A4 was investigated. Prior studies had suggested that the chemical reactivity of the allylic 6beta-position might have a greater influence than steric constraints by the enzyme. Six highly conserved CYP3A residues from substrate recognition site 1 were examined by site-directed mutagenesis. F102A and A117L showed no spectrally detectable P450. V101G and T103A exhibited a wild-type progesterone metabolite profile. Of five mutants at residue N104, only N104D yielded holoenzyme and exhibited the same steroid metabolite profile as wild-type. Of four mutants at position S119 (A, L, T, V), the three hydrophobic ones produced 2beta-OH rather than 6beta-OH progesterone or testosterone as the major metabolite. Kinetic analysis showed S(50) values similar to wild-type for S119A (progesterone) and S119V (testosterone), whereas the V(max) values for 2beta-hydroxysteroid formation were increased in both cases. All four mutants exhibited an altered product profile for 7-hexoxycoumarin side-chain hydroxylation, whereas the stimulation of steroid hydroxylation by alpha-naphthoflavone was similar to the wild-type. The results indicate that the highly conserved residue S119 is a key determinant of CYP3A4 specificity and reveal an important role of the active site topology in steroid 6beta-hydroxylation.

The use of random chimeragenesis to study structure/function properties of rat and human P450c17

The microsomal 17alpha-hydroxylase/17,20-lyase cytochrome P450 (P450c17) catalyzes the 17alpha-hydroxylase reaction required to produce cortisol, the major glucocorticoid in many species and the 17,20-lyase activity required for the production of androgens in all species. Utilizing the technique of random chimeragenesis we have attempted to map regions of primary sequence that contribute to the species-specific biochemical differences between rat and human P450c17. We have previously reported significant differences between rat and human P450c17 in their activities, stability and substrate-dependent coupling efficiencies even though they share 68% amino acid identity. Identification of the regions of primary sequence that contribute to each of these properties would be helpful in understanding the structure/function relationships in this enzyme. A single plasmid containing the cDNAs encoding both enzymes in a tandem orientation was constructed. This plasmid was linearized at unique restriction sites and used to transform Escherichia coli. A three-step screening protocol identified five chimeras with a uniform distribution of 5' rat and 3' human sequence. All chimeric proteins yield the characteristic reduced-CO difference spectra, indicating proper folding. The chimeras exhibit a range of stability and activities that are not consistent with the degree of parental primary sequence. A chimera containing 301 N-terminal rat P450c17 amino acids and lacking the rat P450c17 phenylalanine 343, had the highest lyase activity. Generation of these functional rat/human chimeras suggests that the tertiary structures of rat and human P450c17 are sufficiently conserved to allow proper folding of chimeric enzymes. However, the properties of these chimeras did not permit identification of a region of primary sequence that contributes to a species-specific property of rat and human P450c17. Stability of these chimeras and insight into the presence of secondary structural elements is discussed.

2,2',3,3',6,6'-hexachlorobiphenyl hydroxylation by active site mutants of cytochrome P450 2B1 and 2B11

The structural basis of species differences in cytochrome P450 2B-mediated hydroxylation of 2,2',3,3',6,6'-hexachlorobiphenyl (236HCB) was evaluated by using 14 site-directed mutants of cytochrome P450 2B1 and three point mutants of 2B11 expressed in Escherichia coli. To facilitate metabolite identification, seven possible products, including three hydroxylated and four dihydroxylated hexachlorobiphenyls, were synthesized by direct functionalization of precursors and Ullmann and crossed Ullmann reactions. HPLC and GC/MS analysis and comparison with authentic standards revealed that 2B1, 2B11, and all their mutants produced 4, 5-dihydroxy-236HCB and 5-hydroxy-236HCB, while 2B11 L363V and 2B1 I114V mutants also catalyzed hydroxylation at the 4-position. The amount of products formed by 2B1 mutants I114V, F206L, L209A, T302S, V363A, V363L, V367A, I477A, I477L, G478S, I480A, and I480L was smaller than that of the wild type. I477V exhibited unaltered 236HCB metabolism, and I480V produced twice as much dihydroxy product as the wild type. For 2B11, substitution of Val-114 or Asp-290 with Ile decreased the product yields. Replacement of Leu-363 with Val dramatically altered the profile of 236HCB metabolites. In addition to an increase in the overall level of hydroxylation, the mutant mainly catalyzed hydroxylation at the 4-position. Incubation of P450 2B1 with 5-hydroxy-236HCB produced 4,5-dihydroxy-236HCB, which indicates that 4,5-dihydroxy-236HCB may be formed by a direct hydroxylation of 5-hydroxy-236HCB. The findings from this study demonstrate the importance of residues 114, 206, 209, 302, 363, 367, 477, 478, and 480 in 2B1 and 114, 290, and 363 in 2B11 for 236HCB metabolism.

Molecular modelling of CYP2B6, the human CYP2B isoform, by homology with the substrate-bound CYP102 crystal structure: evaluation of CYP2B6 substrate characteristics, the cytochrome b5 binding site and comparisons with CYP2B1 and CYP2B4

1. Molecular modelling studies of CYP2B isoforms from rat (CYP2B1), rabbit (CYP2B4) and man (CYP2B6) are reported, with particular emphasis on substrate interactions with the human CYP2B isoform, CYP2B6. 2. The findings represent an advance on our previous study that focused primarily on the rat CYP2B isoform, CYP2B1, and involved homology modelling with substrate-free CYP102. 3. The current work utilizes the recently published substrate-bound CYP102 crystal structure as a template for construction of the CYP2B subfamily isoforms and shows, in particular, that known CYP2B6 substrate specificity and regioselectivity can be rationalized by putative active site interactions.

Enzymatic determinants of the substrate specificity of CYP2C9: role of B'-C loop residues in providing the pi-stacking anchor site for warfarin binding

Previous modeling efforts have suggested that coumarin ligand binding to CYP2C9 is dictated by electrostatic and pi-stacking interactions with complementary amino acids of the protein. In this study, analysis of a combined CoMFA-homology model for the enzyme identified F110 and F114 as potential hydrophobic, aromatic active-site residues which could pi-stack with the nonmetabolized C-9 phenyl ring of the warfarin enantiomers. To test this hypothesis, we introduced mutations at key residues located in the putative loop region between the B' and C helices of CYP2C9. The F110L, F110Y, V113L, and F114L mutants, but not the F114Y mutant, expressed readily, and the purified proteins were each active in the metabolism of lauric acid. The V113L mutant metabolized neither (R)- nor (S)-warfarin, and the F114L mutant alone displayed altered metabolite profiles for the warfarin enantiomers. Therefore, the effect of the F110L and F114L mutants on the interaction of CYP2C9 with several of its substrates as well as the potent inhibitor sulfaphenazole was chosen for examination in further detail. For each substrate examined, the F110L mutant exhibited modest changes in its kinetic parameters and product profiles. However, the F114L mutant altered the metabolite ratios for the warfarin enantiomers such that significant metabolism occurred for the first time on the putative C-9 phenyl anchor, at the 4'-position of (R)- and (S)-warfarin. In addition, the Vmax for (S)-warfarin 7-hydroxylation decreased 4-fold and the Km was increased 13-fold by the F114L mutation, whereas kinetic parameters for lauric acid metabolism, a substrate which cannot interact with the enzyme by a pi-stacking mechanism, were not markedly affected by this mutation. Finally, the F114L mutant effected a greater than 100-fold increase in the Ki for inhibition of CYP2C9 activity by sulfaphenazole. These data support a role for B'-C helix loop residues F114 and V113 in the hydrophobic binding of warfarin to CYP2C9, and are consistent with pi-stacking to F114 for certain aromatic ligands.

Nonsubstrate recognition site residues are involved in testosterone hydroxylation by cytochrome P450 CYP 2C11

We have previously characterized an allelic variant of cytochrome P450 CYP 2C11 from the Gunn rat that differs at three positions (amino acids 4, 116, and 187) from the predominant allele from Wistar rats and that displays a dramatically reduced testosterone hydroxylation activity. To assess the relative contribution of these mutations to the decrease in the enzymatic activity we constructed single and double mutants and coexpressed them with reductase. Testosterone metabolism was determined with a baculovirus/insect cell expression system. None of the identified positions alone is critical for the activity since the reversion of one of these mutations is unable to restore fully the Wistar-type activity. The activity of CYP 2C11 containing either the Asn116Ser substitution or the Phe187Leu represents congruent with30% of the activity of the CYP 2C11 Wistar-type protein. In contrast, the activity of the Val4Ala mutated protein is only 10% that of the Wistar-type protein, close to that of the Gunn-type protein. This study reevaluates the contribution of amino acid 4 to the catalysis by cytochrome P450 2C11 and points out the role of extra SRS residues.

Differential modulation of CYP2E1 activity by cAMP-dependent protein kinase upon Ser129 replacement

Many toxic compounds are activated by cytochrome P450 (CYP) 2E1 to reactive metabolites, which represents a potential hazard for cellular homeostasis. Therefore knowledge about CYP2E1 regulation could be of great biological importance. It has been shown that CYP2E1 is controlled transcriptionally and post-translationally by phosphorylation. In the present study we investigated the role of serine-129 (Ser129) in the protein kinase A (PKA) recognition sequence motif Arg-Arg-Phe-Ser129. To gain further insights into the possible relevance of Ser129 for CYP2E1 function, Ser129 was replaced by alanine (Ala) or glycine (Gly) by site-directed mutations of the cDNA coding for CYP2E1. The mutant cDNAs were transfected into Chinese hamster lung fibroblast V79 cells. Despite the mutation in the PKA phosphorylation motif, all strains produced catalytically active CYP2E1. However, there was a marked change in the substrate preference: The Gly129-containing strains hydroxylated p-nitrophenol (PNP) to a markedly higher extent than the wild-type cDNA-containing cells, while they demethylated N-nitrosodimethylamine (NDMA) to a markedly lower extent than the wild-type cells. All the strains activated NDMA to mutagenic products. Treatment with the membrane-permeating cAMP derivative db-cAMP reduced markedly both the PNP hydroxylase and the NDMA demethylase activities as well as the mutation frequency induced by NDMA in the Ser129-containing strain. This decrease in activity was not accompanied by a decrease in CYP2E1 content. In addition, the catalytic activities of CYP2E1 were decreased in microsomes from rat hepatocytes treated with db-cAMP. Also in this case, the decrease in activities was not accompanied by a decrease in enzyme protein. These findings argue that involvement of Ser129 and its phosphorylation is not in determining CYP2E1 protein level, but rather in controlling its catalytic activity. In contrast, in the strains containing Ala129 or Gly129, treatment with db-cAMP caused a marked increase in both PNP hydroxylase and NDMA demethylase. In these strains a similar db-cAMP-mediated increase was also observed in the mutation frequency, resulting from the treatment with the promutagen NDMA, which is activated by CYP2E1. Our results show that CYP2E1 in V79 cells responds in two separate ways to db-cAMP exposure depending on the amino acid residue present in the PKA recognition sequence. The enzyme is committed to a negative regulation by db-cAMP if Ser129 is the target amino acid for PKA, leading to a decrease in the metabolic activation to mutagenic and carcinogenic species. On the other hand, Ala129 or Gly129 substitution directed CYP2E1 toward a positive regulation by increasing its catalytic activities and metabolic activation to mutagenic intermediates in the presence of db-cAMP. We also obtained evidence that cAMP-mediated downregulation of wild-type (Ser129) CYP2E1 was not accompanied by its destruction but instead by its stabilization, which shows that Ser129 is not involved in CYP2E1 degradation but dictates requirements for its specific activities.

Probing the active site of cytochrome P450 2B1: metabolism of 7-alkoxycoumarins by the wild type and five site-directed mutants

A series of 7-alkoxycoumarins (chain length of 1-7 carbon atoms) was utilized as active site probes of purified Escherichia coli-expressed cytochrome P450 2B1 wild type and five site-directed mutants (I114V, F206L, V363A, V363L, and G478S). The production of 7-hydroxycoumarin, the O-dealkylation product, by the wild-type enzyme exhibited a rank order of C2 > C4 > C3 > C1 > C5 > C6 = C7. The pattern observed for the P450 I114V mutant was similar to that of the wild-type enzyme, whereas with F206L and G478S mutants, the rate of O-dealkylation was low with all the compounds. In contrast, with V363A, the highest rate of product formation was observed with 7-butoxycoumarin. The V363L mutant preferentially catalyzed the O-dealkylation of 7-methoxy- and 7-ethoxycoumarin, and a further increase in the length of the alkyl chain led to a marked decrease in product formation. The stoichiometry of 7-butoxycoumarin oxidation by V363L suggested that products other than 7-hydroxycoumarin were also formed. HPLC and GC-EIMS analyses revealed that P450 2B1 V363L produced 7-(3-hydroxybutoxy)coumarin and 7-(4-hydroxybutoxy)coumarin as major oxidation products, while the V363A mutant mainly catalyzed the O-dealkylation of 7-butoxycoumarin. Docking of alkoxycoumarins into the active site of a P450 2B1 homology model confirmed the importance of the studied residues in substrate dealkylation and explained the formation of novel 7-butoxycoumarin products by the V363L mutant.

Analysis of four residues within substrate recognition site 4 of human cytochrome P450 3A4: role in steroid hydroxylase activity and alpha-naphthoflavone stimulation

Sequence alignment of human cytochrome P450 3A4 with bacterial enzymes of known structure has provided a basis from which to predict residues involved in substrate oxidation. Substitutions were made at four residues (I301, F304, A305, and T309) predicted to be located within the highly conserved substrate recognition site 4. Site-directed mutants engineered to contain carboxy-terminal histidine tags were expressed in Escherichia coli and purified on a metal affinity column. The integrity of each protein was assessed by SDS-polyacrylamide gel electrophoresis and immunoblotting. Functional analysis was performed using progesterone and testosterone as substrates and alpha-naphthoflavone as an activator. In testosterone hydroxylase assays, all of the mutants displayed rates of total product formation similar to wild-type 3A4, with several mutants showing small differences in specific products formed. However, with progesterone as the substrate, mutants F304A, A305V, and T309A exhibited altered product ratios and/or changes in the rates of product formation. F304A and A305V also displayed altered flavonoid stimulation that resulted in product ratios dramatically different from wild-type 3A4. Therefore, the kinetics of progesterone hydroxylation of these mutants and the wild-type enzyme were further assessed, and the data were analyzed with the Hill equation. Results with wild-type 3A4 and F304A indicated that at high progesterone concentrations, hydroxylation rates and product ratios are independent of the presence of alpha-NF. This suggests that progesterone may be equivalent to alpha-NF as an activator. In contrast, A305V exhibited autoactivation by progesterone but inhibition by alpha-NF.

Structural determinants of progesterone hydroxylation by cytochrome P450 2B5: the role of nonsubstrate recognition site residues

The highly related rabbit cytochromes P450 2B4 and 2B5 differ in only 12 amino acid positions, but only 2B5 has activity toward progesterone. Previously, simultaneous site-directed mutagenesis of four key substrate recognition site (SRS) residues (114, 294, 363, and 367) was shown to result in interconversion of the androstenedione hydroxylase specificities of cytochrome P450 2B4 and 2B5. However, the progesterone metabolite profiles of the 2B4 quadruple mutant or of a quintuple mutant in which residue 370 was also mutated to the 2B5 residue were not identical to that of P450 2B5. Therefore, single mutants of P450 2B5 at the remaining seven positions were constructed, expressed in Escherichia coli, and studied with progesterone as the substrate. The single mutants at positions 120 and 221, which are outside any known SRS, exhibited a significant alteration in progesterone hydroxylation. Based on these results, Ile-114, Arg-120, Ser-221, Ser-294, Ile-363, and Val-367 in cytochrome P450 2B4 were replaced simultaneously with Phe, His, Pro, Thr, Val, and Ala, respectively, from 2B5. This yielded a mutant with a very similar progesterone metabolite profile to that of 2B5, although the total activity was lower. An additional substitution at residue 370 produced a multiple mutant P450 2B4 I114F-R120H-S221P-S294T-I363V-V367A- T370M with very similar or identical substrate specificity, regio- and stereospecificity and kinetic properties to that of P450 2B5 wild type.

Use of homology modeling in conjunction with site-directed mutagenesis for analysis of structure-function relationships of mammalian cytochromes P450

In recent years, homology modeling has become an important tool to study cytochrome P450 function, especially in conjunction with experimental approaches such as site-directed mutagenesis. Molecular models of mammalian P450s can be constructed based on crystal structures of four bacterial enzymes, P450cam, P450 BM-3, P450terp and P450eryF, using molecular replacement or consensus methods. In a model built by molecular replacement, the coordinates are copied from those of a given template protein, while consensus methods utilize more then one protein as a template and are based on distance geometry calculations. The models can be used to identify or confirm key residues, evaluate enzyme-substrate interactions and explain changes in protein stability and/or regio- and stereospecificity of substrate oxidation upon residue substitution by site-directed mutagenesis. P450 models have also been utilized to analyze binding of inhibitors or activators, as well as alterations in inhibition and activation due to residue replacement.

Comparisons of catalytic selectivity of cytochrome P450 subfamily enzymes from different species

Historically there has been considerable interest in comparing patterns of biotransformation of xenobiotic chemicals in experimental animal models and humans, e.g. in areas such as drug metabolism and chemical carcinogenesis. With the availability of more basic knowledge it has become possible to attribute the oxidation of selected chemicals to individual cytochrome P450 (P450) enzymes in animals and humans. Further, these P450s can be characterized by their classification into distinct subfamilies, which are defined as having > 59% amino acid sequence identity. Questions arise about how similar these enzymes are with regard to structure and function. More practically, how much can be predicted about reaction specificity and catalysis? In order to address these issues, we need to consider not only the relatedness of P450s from different species but also (i) functional similarity within P450 subfamilies and (ii) the effects of small changes imposed by site-directed mutagenesis. Relationships in the P450 1A, 2A, 2B, 2C, 2D, 2E, 3A, and 17A subfamilies are briefly reviewed. Overall functional similarity is generally seen in subfamily enzymes but many examples exist of important changes in catalysis due to very small differences, even a single conservative amino acid substitution. Some general conclusions are presented about predictability within various P450 subfamilies.

Stoichiometry of 7-ethoxycoumarin metabolism by cytochrome P450 2B1 wild-type and five active-site mutants

Recombinant P450 2B1 wild-type and the active-site mutants I114V, F206L, V363A, V363L, and G478S were purified and studied. The efficiency of coupling of reducing equivalents to 7-hydroxycoumarin formation was decreased for all the mutants except I114V. Uncoupling to H2O was increased for F206L, V363A, and G478S, decreased for V363L, and unchanged for I114V. Uncoupling to H2O2 was increased for V363L and decreased for I114V, F206L, and V363A. The findings from this study provide firm biochemical evidence that residues 206, 363, and 478 comprise part of the substrate binding site of P450 2B1.

Significance of glycine 478 in the metabolism of N-benzyl-1-aminobenzotriazole to reactive intermediates by cytochrome P450 2B1

The effect of mutating Gly 478 to Ala in rat cytochrome P450 2B1 on the metabolism of N-benzyl-1-aminobenzotriazole was investigated. The 7-ethoxy-4-(trifluoromethyl)coumarin O-deethylation activity of the wild-type enzyme was completely inactivated by incubating with 1 microM BBT. The G478A mutant, however, was not inactivated by incubating with up to 10 microM BBT. Whereas metabolism of BBT by the wild-type 2B1 resulted in the formation of benzaldehyde, benzotriazole, aminobenzotriazole, and a new metabolite, the G478A mutant generated only the later. This metabolite was found by NMR, IR, and mass spectrometry to be a dimeric product formed from the reaction of two BBT molecules. Two spectral binding constants, a high-affinity constant that was the same for both enzymes (30-39 microM) and a low-affinity constant that was 5-fold lower for the mutant enzyme (0.3 mM vs 1.4 mM), were observed with BBT. The apparent Km and kcat values for the G478A mutant with BBT were 0.3 mM and 12 nmol (nmol of P450)-1 min-1, respectively. Molecular modeling studies of BBT bound in the active site of P450 2B1 suggested that a mutation of Gly 478 to Ala would result in steric hindrance and suppress oxidation of BBT at the 1-amino nitrogen. When BBT was oriented in the 2B1 active site such that oxidation at the 7-benzyl carbon could occur, no steric overlap between Ala 478 and the substrate was observed. Thus, this orientation of BBT would be preferred by the mutant leading to oxidation at the 7-benzyl carbon and subsequent dimer formation. These findings indicate that a glycine 478 to alanine substitution in P450 2B1 altered the binding of BBT such that inactivating BBT metabolites were no longer generated.

Reassessment of cytochrome P450 2B2: catalytic specificity and identification of four active site residues

Cytochromes P450 2B metabolize a variety of compounds and have provided an excellent framework for identifying determinants of substrate specificity. Among the rat 2B enzymes, a puzzling difference has emerged between the reported substrate specificity of purified hepatic 2B2 and that of certain 2B1 mutants containing 2B1 --> 2B2 substitutions. To address these discrepancies, we have characterized two 2B2 variants. A cDNA clone designated 2B2FF was obtained from phenobarbital-induced Lewis rats and, like some previously isolated variants, was found to contain phenylalanine at positions 58 and 114. A second 2B2 clone was generated by restoring Leu and Ile, respectively, at these positions. These enzymes were expressed in Escherichia coli and analyzed with androstenedione, testosterone, progesterone, ethoxycoumarin, benzyloxyresorufin, and pentoxyresorufin. The expressed 2B2 variants metabolized most substrates at rates that were 1-9% of those of 2B1. When steroid regio- and stereospecificity was examined, the metabolite profiles of expressed 2B2 and 2B2FF conflicted with the 16beta- and 16alpha-hydroxylation observed for purified hepatic 2B2 from Sprague-Dawley rats. These and other results suggested that the purified hepatic 2B2 contained a small percent of the 2B1 enzyme. Masses of tryptic peptides were consistent with identity between purified hepatic 2B2 and 2B2FF. On the basis of a three-dimensional homology model and the construction and analysis of 2B2 mutants, residues 114, 363, 367, and 478 were identified as determinants of substrate specificity. In addition, 2B1 and the expressed 2B2 variants showed differential susceptibility to the mechanism-based inactivators chloramphenicol and N-(2-p-nitrophenethyl)chlorofluoroacetamide.

Molecular modelling of mammalian CYP2B isoforms and their interaction with substrates, inhibitors and redox partners

1. The construction of three-dimensional models of CYP2B isozymes from rat (CYP2B1), rabbit (CYP2B4) and man (CYP2B6), based on a multiple sequence alignment with CYP102, a unique eukaryotic-like bacterial P450 (in terms of possessing an NADPH-dependent FAD- and FMN-containing oxidoreductase redox partner) of known crystal structure, is reported. 2. The enzyme models described are shown to be consistent with experimental evidence from site-directed mutagenesis studies, antibody recognition sites and amino acid residues identified as being associated with redox partner interactions, together with the location of a key serine residue (Ser-128) likely to be involved in protein kinaseA-mediated phosphorylation. 3. A substantial number of known substrates and inhibitors of CYP2B isozymes are shown to fit the putative active sites of the enzyme models in agreement with their reported position of metabolism or mode of inhibition respectively. In particular, there is complementarity between the characteristic non-planar geometries of CYP2B substrates and key groups in the enzymes' active sites. 4. Molecular modelling of CYP2B isozymes appears to rationalize a number of the reported findings from quantitative structure-activity relationship investigations on series of CYP2B substrates and inhibitors.

Alanine-scanning mutagenesis of a putative substrate recognition site in human cytochrome P450 3A4. Role of residues 210 and 211 in flavonoid activation and substrate specificity

Alanine-scanning mutagenesis was performed on amino acid residues 210-216 of cytochrome P450 3A4, the major drug-metabolizing enzyme of human liver. Mutagenesis of this region, which has been proposed to align with the C-terminal ends of F-helices from cytochromes P450BM-3, P450terp, and P450cam, served as a test of the applicability of the substrate recognition site model of Gotoh (Gotoh, O. (1992) J. Biol. Chem. 267, 83-90) to P450 3A4. The results, using two steroid substrates, indicated that substitution of Ala for Leu210 altered the responsiveness to the effector alpha-naphthoflavone and the regioselectivity of testosterone hydroxylation. Replacement of Leu211 by Ala also decreased the stimulation by alpha-naphthoflavone, whereas mutations at residues 212-216 had little effect. The diminished flavonoid responses of the 210 and 211 mutants were observed over a wide range of progesterone and alpha-naphthoflavone concentrations. Further characterization was performed with the additional effectors beta-naphthoflavone, flavone, and 4-chromanone. The finding that P450 3A4 with one altered residue, Leu210 --> Ala, can have both an altered testosterone hydroxylation profile and response to flavonoid stimulation provides evidence that the substrate binding and effector sites are at least partially overlapping.

Secobarbital-mediated inactivation of cytochrome P450 2B1 and its active site mutants. Partitioning between heme and protein alkylation and epoxidation

Secobarbital (SB) is a relatively selective mechanism-based inactivator of cytochrome P450 2B1, that partitions between epoxidation and heme and protein modification during its enzyme inactivation. The SB-2B1 heme adduct formed in situ in a functionally reconstituted system has been spectrally documented and structurally characterized as N-(5-(2-hydroxypropyl)-5-(1-methylbutyl)barbituric acid)protoporphyrin IX. The SB-protein modification has been localized to 2B1 peptide 277-323 corresponding to the active site helix I of cytochrome P450 101. The targeting of heme and this active site peptide suggests that the 2B1 active site topology could influence the course of its inactivation. To explore this possibility, the individual SB epoxidation, heme and protein modification, and corresponding molar partition ratios of the wild type and seven structural 2B1 mutants, site-directed at specific substrate recognition sites, and known to influence 2B1 catalysis were examined after Escherichia coli expression. These studies reveal that Thr-302 is critical for SB-mediated heme N-alkylation, whereas Val-367 is a critical determinant of 2B1 protein modification, and Val-363 is important for SB epoxidation. SB docking into a refined 2B1 homology model coupled with molecular dynamics analyses provide a logical rationale for these findings.

A three-dimensional protein model for human cytochrome P450 2D6 based on the crystal structures of P450 101, P450 102, and P450 108

Cytochromes P450 (P450s) constitute a superfamily of phase I enzymes capable of oxidizing and reducing various substrates. P450 2D6 is a polymorphic enzyme, which is absent in 5-9% of the Caucasian population as a result of a recessive inheritance of gene mutations. This deficiency leads to impaired metabolism of a variety of drugs. All drugs metabolized by P450 2D6 contain a basic nitrogen atom, and a flat hydrophobic region coplanar to the oxidation site which is either 5 or 7 A away from the basic nitrogen atom. The aim of this study was to build a three-dimensional structure for the protein and more specifically for the active site of P450 2D6 in order to determine the amino acid residues possibly responsible for binding and/ or catalytic activity. Furthermore, the structural features of the active site can be implemented into the existing small molecule substrate model, thus enhancing its predictive value with respect to possible metabolism by P450 2D6. As no crystal structures are yet available for membrane-bound P450s (such as P450 2D6), the crystal structures of bacterial (soluble) P450 101 (P450cam), P450 102 (P450BM3), and P450 108 (P450terp) have been used to build a three-dimensional model for P450 2D6 with molecular modeling techniques. Several important P450 2D6 substrates were consecutively docked into the active site of the protein model. The energy optimized positions of the substrates in the protein agreed well with the original relative positions of the substrates within the substrate model. This confirms the usefulness of small molecule models in the absence of structural protein data. Furthermore, the derived protein model indicates new leads for experimental validation and extension of the substrate model.

Molecular modelling of CYP1A subfamily members based on an alignment with CYP102: rationalization of CYP1A substrate specificity in terms of active site amino acid residues

1. Using a novel amino acid sequence alignment, proteins of the CYP1A subfamily have been produced from the CYP102 crystal structure template via residue replacement and energy minimization procedures. 2. Known substrates and inhibitors of CYP1A1 and CYP1A2 are shown to fit their respective active sites via key interactions with complementary amino acid residues. Substrates used in the modelling studies include: caffeine, PhIP, oestradiol, 2,4- and 2,5-diaminotoluenes, Glu-P-1, phenacetin, acetanilide, 7-methoxy and 7-ethoxyresorufins, 11-methyl cyclopenta[a]phenanthren-17-one, 7-ethoxycoumarin, aflatoxin B1, benzo[a]pyrene, benzo[a]pyrene-7,8-diol and 1'-hydroxy 3-methylcholanthrene. 3. A number of aspects relating to CYP1A substrate specificity and metabolism can be explained in terms of the enzyme models, as it is found that key interactions with active site amino acid residues direct CYP1A-mediated metabolism in the known positions.

Identification of residues 99, 220, and 221 of human cytochrome P450 2C19 as key determinants of omeprazole activity

Human P450 2C19 is selective for 4'-hydroxylation of S-mephenytoin and 5-hydroxylation of omeprazole, while the structurally homologous P450 2C9 has low activity toward these substrates. To identify the critical amino acids that determine the specificity of human amino acids that determine the specificity of human P450 2C19, we constructed chimeras of p450 2C9 replacing various proposed substrate binding sites (SRS) with those of P450 2C19 and then replaced individual residues of P450 2C19 and then replaced individual residues of P450 2C9 by site-directed mutagenesis. The 339 NH2-terminal amino acid residues (SRS-1-SRS-4) and amino acids 160-383 (SRS-2-SRS-5) of P450 2C19 conferred omeprazole 5-hydroxylase activity to P450 2C9. In contract, the COOH terminus of P450 2C19 (residues 340-490 including SRS-5 and SRS-6), residues 228-339 (SRS-3 and SRS-4) and residues 292-383 (part of SRS-4 and SRS-5) conferred only modest increases in activity. A single mutation Ile99 --> His increased omeprazole 5-hydroxylase to approximately 51% of that of P450 2C19. A chimera spanning residues 160-227 of P450 2C19 also exhibited omeprazole 5-hydroxylase activity which was dramatically enhanced by the mutation Ile99 --> His. A combination of two mutations, Ile99 --> His and Ser200 --> Pro, converted P450 2C9 to an enzyme with a turnover number of omeprazole 5-hyrdroxylation, which resembled that of P450 /c19. Mutation of Pro221 --> Thr enhanced this activity. Residue 99 is within SRS-1, but amino acids 220 and 221 are in the F-G loop and outside any known SRS. Mutation of these three amino acids did not confer significant S-mephenytoin 4'-hydroxylase activity to P450 2C9, although chimeras containing SRS-1-SRS-4 and SRS-2-SRS-5 of P450 2C19 exhibited activity toward this substrate. Our results thus indicate that amino acids 99, 220, and 221 are key residues that determine the specificity of P450 2C19 for omeprazole.

Engineering a mineralocorticoid- to a glucocorticoid-synthesizing cytochrome P450

Site-directed mutagenesis of a domain (amino acids 299-338) aligning to the I-helix region of P450cam, P450BM3 and P450terp was used to investigate the different regioselectivities displayed in the hydroxylation reactions performed by human aldosterone synthase (P450aldo) and 11beta-hydroxylase (P45011beta). The two enzymes are 93% identical and are essential for the synthesis of mineralocorticoids and glucocorticoids in the human adrenal gland. Single replacement of P450aldo residues for P45011 beta-specific residues at positions 296, 301, 302, 320, and 335 only gave rise to slightly increased 11beta-hydroxylase activities. However, a L301P/A320V double substitution increased 11beta-hydroxylase activity to 60% as compared with that of P45011 beta. Additionally substituting Ala-320 for Val-320 of P45011 beta further enhanced this activity to 85%. The aldosterone synthase activities of the mutant P450aldo proteins were suppressed to a varying degree, with triple replacement mutant L301P/E302D/A320V retaining only 10% and double replacement mutant L301P/A320V retaining only 13% of the P450aldo wild type activity. These results demonstrate a switch in regio- and stereoselectivities of the engineered P450aldo enzyme due to manipulation of residues at three critical positions, and we attribute the determination of these features in P450aldo to the structure of a region analogous to the I-helix in P450cam.

Structural flexibility and functional versatility of cytochrome P450 and rapid evolution

P450 represents a large group of heme-thiolate enzymes that exhibit remarkably diverse activities for the metabolism of numerous endogenous and exogenous chemicals. Recent site-directed mutagenesis studies indicate that a single mutation at any of the key residues can be enough to alter the substrate and/or product specificities in the P450 activities. Molecular modeling predicts that these key residues are located within the substrate heme pocket. Structural elements involved in diversifying P450 activity appear to correspond to the B' helix, the F helix and the F/G interhelical loop in the bacterial P450s. Structures represented by these regions are extremely variable despite the fact that the core of the P450 substrate pocket is well conserved. A mutation within these regions may result in a significant geometrical alteration of the pocket and lead to diversify the P450 activity. Phylogenetical analysis shows a relatively high rate of nonsynonymous substitution within these substrate binding regions. The functional versatility of P450 can thus be largely accounted for in terms of pocket change brought about by rapid mutations.

The roles of individual amino acids in altering substrate specificity of the P450 2a4/2a5 enzymes

A single amino acid substitution is sufficient to alter substrate specificity of P450 enzymes. Mouse P450 2a5, for example, has its substrate specificity converted from coumarin 7- to testosterone 15 alpha-hydroxylase activity by the substitution of Phe at position 209 to Leu. Furthermore, placing Asn at this position confers a novel corticosterone 15 alpha-hydroxylase activity to this P450. Recent site-directed mutational studies show the presence of the topologically common residues, each of which can determine the specificities of various mammalian P450s. For instance, residue 209 (in 2a5) corresponds to a residue at position 206 in rat P4502B1 that regulates its steroid hydroxylase activity. High substrate specificity often observed in an individual P450, therefore, can be determined and altered by the identities of a few critical residues. The structural flexibility of the substrate-heme pocket may also provide P450 enzymes with the ability to display a broad range of substrate specificities. Understanding the underlying principles whereby the flexible pocket determines P450 activities may lead us to the prediction of P450 activities based on the identities of key amino acid residues.

Cytochrome P450

Since 1993, three new cytochrome P450 X-ray structures have been determined, giving a total of four known structures. Two of the new structures are in the substrate-free form and one is substrate-bound. These new structures, together with a wealth of mutagenesis studies on various P450s, have provided considerable information on what structural features control substrate specificity in P450s. In addition, some important insights into the catalytic mechanism have been made.

Altering the regiospecificity of androstenedione hydroxylase activity in P450s 2a-4/5 by a mutation of the residue at position 481

Mouse P450 2a-5 (coumarin 7-hydroxylase) acquires androstenedione (AD) hydroxylase activity by substituting Phe at position 209 with Asn. However, this mutant P450 2a-5 (F209N) and the corresponding mutant P450 2a-4 (L209N) exhibit different regiospecificites of androstenedione (AD) hydroxylase activity. While the former mutant catalyzes both AD 15 alpha- and 7 alpha-hydroxylase activities at similar rates, the latter mutant maintains the original high specificity of AD 15 alpha-hydroxylase activity. The AD hydroxylase activities in chimeric enzymes of the mutants L209N and F209N show that the regiospecificites are determined by the carboxy-terminal halves of the P450 molecules. Mutations at each of the four different residues within the carboxy-terminal halves indicate that the differences in regiospecificity are determined by the Val/Ala mutation at position 481. As the size of the hydrophobic amino acid at position 481 becomes larger (Ala < Val < Ile), the regiospecificities toward the C15 position of the AD molecule are dramatically increased. The regiospecificity is also increased by placing positively-charged Arg at position 481, although the remaining 15 alpha-hydroxylase activity in this mutant is considerably lower than the other P450s. The results indicate that the size of the residue at position 481 is a key factor in regulating the regiospecificity of AD hydroxylase activity in the P450s. Modeling AD in the substrate-heme pocket of bacterial P450 101A provided further support that residue 481 may reside near the steroid molecule so as to possibly affect the AD hydroxylase activity.

Three-dimensional models of human and other mammalian microsomal P450s constructed from an alignment with P450102 (P450bm3)

1. A novel modelling alignment for P450s, utilizing NADPH-P450 reductase for electron transfer, is proposed on the basis of analysis of their amino acid sequences. 2. Information used to facilitate the alignment process includes: the recent X-ray crystal structure of P450102 (P450bm3), site-directed mutagenesis experiments, chemical modification of specific residues, and antibody recognition studies. 3. The alignment has been used to construct a number of microsomal P450s of relevance to xenobiotic and endogenous metabolism.

Structure and function of cytochromes P450: a comparative analysis of three crystal structures

BACKGROUND

Cytochromes P450 catalyze the oxidation of a variety of hydrophobic substrates. Sequence identities between P450 families are generally low (10-30%), and consequently, the structure-function correlations among P450s are not clear. The crystal structures of P450terp and the hemoprotein domain of P450BM-3 were recently determined, and are compared here with the previously available structure of P450cam.

RESULTS

The topology of all three enzymes is quite similar. The heme-binding core structure is well conserved, except for local differences in the I helices. The greatest variation is observed in the substrate-binding regions. The structural superposition of the proteins permits an improved sequence alignment of other P450s. The charge distribution in the three structures is similarly asymmetric and defines a molecular dipole.

CONCLUSIONS

Based on this comparison we believe that all P450s will be found to possess the same tertiary structure. The ability to precisely predict other P450 substrate-contact residues is limited by the extreme structural heterogeneity in the substrate-recognition regions. The central I-helix structures of P450terp and P450BM-3 suggest a role for helix-associated solvent molecules as a source of catalytic protons, distinct from the mechanism for P450cam. We suggest that the P450 molecular dipole might aid in both redox-partner docking and proton recruitment for catalysis.