Refinement of 3D models of horseradish peroxidase isoenzyme C: predictions of 2D NMR assignments and substrate binding sites

Metabolism of Bergamottin by Cytochromes P450 2B6 and 3A5

Cytochromes P450 (P450) 2B6 and 3A5 are inactivated by bergamottin (BG). P450 2B6 metabolized BG primarily to M3 and M4 and one minor metabolite (M1). The metabolites were analyzed, and the data indicated that M1 was bergaptol, M3 was 5'-OH-BG, and M4 was a mixture of 6'- and 7'-OH-BG. Because 6'- and 7'-OH-BG were the primary metabolites, it suggested that P450 2B6 preferentially oxidized the geranyloxy chain of BG. Metabolism of BG by P450 3A5 resulted in three major metabolites: [bergaptol, M3 (5'-OH-BG), and M5 (2'-OH-BG)], and two minor metabolites [M2 (6',7'-dihydroxy-BG) and M4 (6'- and 7'-OH-BG)]. Because bergaptol was the most abundant metabolite formed, it suggested that P450 3A5 metabolized BG mainly by cleaving the geranyl-oxy chain. Molecular modeling studies confirmed that docking of BG in the P450 2B6 active site favors oxidation in the terminal region of the geranyl-oxy chain, whereas positioning the 2'-carbon of BG nearest the heme iron is preferred by P450 3A5. Glutathione (GSH)-BG conjugates were formed by both P450. Each enzyme predominantly formed conjugates with m/z values of 662. Tandem mass spectrometry analysis of the GSH conjugates indicated that the oxidation forming a reactive intermediate occurred on the furan moiety of BG, presumably through the initial formation of an epoxide at the furan double bond. The data indicate that oxidation of the geranyl-oxy chain resulted in the formation of stable metabolites of BG, whereas oxidation of the furan ring produced reactive intermediates that may be responsible for binding to and inactivating P450 2B6 and 3A4.

Theoretical study of the ligand-CYP2B4 complexes: Effect of structure on binding free energies and heme spin state

The molecular origins of temperature-dependent ligand-binding affinities and ligand-induced heme spin state conversion have been investigated using free energy analysis and DFT calculations for substrates and inhibitors of cytochrome P450 2B4 (CYP2B4), employing models of CYP2B4 based on CYP2C5(3LVdH)/CYP2C9 crystal structures, and the results compared with experiment. DFT calculations indicate that large heme-ligand interactions (ca. -15 kcal/mol) are required for inducing a high to low spin heme transition, which is correlated with large molecular electrostatic potentials (approximately -45 kcal/mol) at the ligand heteroatom. While type II ligands often contain oxygen and nitrogen heteroatoms that ligate heme iron, DFT results indicate that BP and MF heme complexes, with weak substrate-heme interactions (ca. -2 kcal/mol), and modest MEPS minima (>-35 kcal/mol) are high spin. In contrast, heme complexes of the CYP2B4 inhibitor, 4PI, the product of benzphetamine metabolism, DMBP, and water are low spin, have substantial heme-ligand interaction energies (<-15 kcal/mol) and deep MEPS minima (<-45 kcal/mol) near their heteroatoms. MMPBSA analysis of MD trajectories were made to estimate binding free energies of these ligands at the heme binding site of CYP2B4. In order to initially assess the realism of this approach, the binding free energy of 4PI inhibitor was computed and found to be a reasonable agreement with experiment: -7.7 kcal/mol [-7.2 kcal/mol (experiment)]. BP was determined to be a good substrate [-6.3 kcal/mol (with heme-ligand water), -7.3 kcal/mol (without ligand water)/-5.8 kcal/mol (experiment)], whereas the binding of MF was negligible, with only marginal binding binding free energy of -1.7 kcal/mol with 2-MF bound [-3.8 kcal/mol (experiment)], both with and without retained heme-ligand water. Analysis of the free energy components reveal that hydrophobic/nonpolar contributions account for approximately 90% of the total binding free energy of these substrates and are the source of their differential and temperature-dependent CYP2B4 binding. The results indicate the underlying origins of the experimentally observed differential binding affinities of BP and MF, and indicate the plausibility of the use of models derived from moderate sequence identity templates in conjunction with approximate free energy methods in the estimation of ligand-P450 binding affinities.

Construction and assessment of models of CYP2E1: predictions of metabolism from docking, molecular dynamics, and density functional theoretical calculations

3D models of CYP2E1 were constructed for the purpose of structure-based prediction of 2E1 metabolism of diverse substrates based on configuration sampling of ligand-atom-oxyferryl center distances and quantum chemical criteria. Models were constructed on the basis of sequence alignments of 2E1 with templates of known structure, including rabbit CYP2C5 (3LVdH) and bacterial CYP450s. Following geometric and energetic assessments, the utility of the model was tested in structure-based predictions of metabolism. Autodock was used to dock chlorzoxazone, p-nitrophenol, N-nitrosodimethylamine, acetominophen, caffeine, theophylline, and methoxyflurane into the model CYP2E1 employing a model oxyferryl heme with charges based on density functional theoretical parametrization. In all cases, the lowest energy bound docked configurations corresponded to ones with the substrate intimately associated with the oxyferryl center. Configurations among the lowest energy docked forms of each of the ligands had orientations relative to the oxyferryl center consistent with the experimentally observed metabolites. Docking of long-chain dialkylnitrosoamines revealed no heme binding site bound configurations, in agreement with the negligible metabolism of these ligands. The lowest energy docked configurations of chlorzoxazone, p-nitrophenol, and N-nitrosodimethylamine, high-affinity substrates of 2E1, were used to initiate 300 ps molecular dynamics (MD) trajectories. The MD-sampled ligand-oxyferryl heme reactant configurations were in good accord with density functional theoretical (DFT) optimized oxyferryl-heme-ligand geometries. Analysis of the MD-sampled ligand-2E1 configurations from multiple docked orientations indicates the configurations with closest exposure of reactive centers to the oxyferryl heme to be correlated with observed metabolites with proper consideration of H-abstraction energetics. DFT assessment of relative radical energetics is directly compared with differential H-abstraction activation energetics by compound I and by a p-nitrosophenoxy radical compound I surrogate for the specific case of methoxyflurane and is shown to be in good agreement.

Controlled layer-by-layer immobilization of horseradish peroxidase

Horseradish peroxidase (HRP) was biotinylated with biotinamidocaproate N-hydroxysuccinimide ester (BcapNHS) in a controlled manner to obtain biotinylated horseradish peroxidase (Bcap-HRP) with two biotin moieties per enzyme molecule. Avidin-mediated immobilization of HRP was achieved by first coupling avidin on carboxy-derivatized polystyrene beads using a carbodiimide, followed by the attachment of the disubstituted biotinylated horseradish peroxidase from one of the two biotin moieties through the avidin-biotin interaction (controlled immobilization). Another layer of avidin can be attached to the second biotin on Bcap-HRP, which can serve as a protein linker with additional Bcap-HRP, leading to a layer-by-layer protein assembly of the enzyme. Horseradish peroxidase was also immobilized directly on carboxy-derivatized polystyrene beads by carbodiimide chemistry (conventional method). The reaction kinetics of the native horseradish peroxidase, immobilized horseradish peroxidase (conventional method), controlled immobilized biotinylated horseradish peroxidase on avidin-coated beads, and biotinylated horseradish peroxidase crosslinked to avidin-coated polystyrene beads were all compared. It was observed that in solution the biotinylated horseradish peroxidase retained 81% of the unconjugated enzyme's activity. Also, in solution, horseradish peroxidase and Bcap-HRP were inhibited by high concentrations of the substrate hydrogen peroxide. The controlled immobilized horseradish peroxidase could tolerate much higher concentrations of hydrogen peroxide and, thus, it demonstrates reduced substrate inhibition. Because of this, the activity of controlled immobilized horseradish peroxidase was higher than the activity of Bcap-HRP in solution. It is shown that a layer-by-layer assembly of the immobilized enzyme yields HRP of higher activity per unit surface area of the immobilization support compared to conventionally immobilized enzyme.

Homology modeling and substrate binding study of human CYP2C18 and CYP2C19 enzymes

It is well established that the variable binding-site architecture and composition of the P450 metabolizing heme proteins are major modulators of substrate and product specificity. Even the three closely related human liver isozymes, CYP2C9, CYP2C18, and CYP2C19, do not share all substrates and do not always lead to the same preferred hydroxylation products. The lack of knowledge of their three-dimensional (3D) structures has hindered efforts to understand the differences in their specificities. Building on previous work for the CYP2C9 enzyme, 3D models of CYP2C18 and 2C19 have been constructed and validated by computational methods developed and tested in our laboratory. They were used to characterize explicit enzyme-substrate complexes using the isoform-specific substrates progesterone and (S)-mephenytoin for 2C19 and 2-[2,3-dichloro-4-(3-hydroxypropyloxy)benzoyl]thiophene for 2C18. The results allowed both common and unique binding-site residues to be identified in each model. The calculated preferred hydroxylation site was obtained for each substrate and was found to be consistent with experimental observation. Comparisons were made among the 2C9, 2C18, and 2C19 model binding sites to investigate the subtle differences among them. These models can be used as structure-based guides for mutagenesis studies and screening of potential pharmaceuticals or toxins.

Homology modeling and substrate binding study of human CYP2C9 enzyme

The CYP2C subfamily of human liver P450 isozymes is of major importance in drug metabolism. The most abundant 2C isozyme, CYP2C9, regioselectively hydroxylates a wide variety of substrates. A major obstacle to understanding this specificity in human CYP2C9 is the absence of a 3D structure. A 3D model of CYP2C9 was built, assessed, and used to characterize explicit enzyme-substrate complexes using methods previously developed in our laboratory. The 3D model was assessed by determining its stability to unconstrained molecular dynamics and by comparison of specific properties with those of known protein structures. The CYP2C9 model was then used to characterize explicit enzyme complexes with three structurally and chemically diverse substrates: (S)-naproxen, phenytoin, and progesterone. Each substrate was found to bind to the enzyme with a favorable interaction energy and to remain in the binding site during unconstrained molecular dynamics. Moreover, the mode of binding of each substrate led to calculated preferred hydroxylation sites consistent with experiment. Binding-site residues identified for the models included Arg 105 and Arg97 as key cationic residues, as well as Asn 202, Asp 293, Pro 101, Leu 102, Gly 296, and Phe 476. Site-specific mutations are proposed for further integrated computational and experimental study.

Molecular dynamics simulations of P450 BM3--examination of substrate-induced conformational change

Cytochrome P450 BM3, of bacterial origin, is one of only five isozymes of the ubiquitous family of over 400 metabolizing heme proteins with a known crystal structure and only one of two with both substrate-free and substrate-bound forms determined. P450 BM3 is of particular interest since it has a similar function and similar substrates as mammalian P450s particularly of the 4A subfamily. Thus, the extent to which the substrate-free form of P450 BM3 undergoes a conformational change upon binding of a typical fatty acid substrate, palmitoleic acid, has been the subject of recent active experimental effort. Surprisingly, direct examination of the substrate-free (pdb2hpd.ent and pdb2bmh.ent) and substrate-bound (pdb1fag.ent) forms do not provide a clear answer to this question. The main reason for this ambiguity is that the two substrate-free monomers reported in the crystal structures themselves have significantly different conformations from each other, one with a more open substrate-access channel than the other. Since there is no way to tell to which substrate-free form the substrate binds, the effect of substrate binding cannot be deduced directly from comparisons of the experimental substrate-bound and substrate-free forms. The computational studies reported here have been designed to more robustly establish the effect of substrate binding on this isozyme. Specifically, molecular dynamics simulations were performed for each of the two substrate-free forms found in the asymmetric unit of the X-ray structure and for the two corresponding substrate-bound forms, constructed by docking palmitloeic acid into each of them. Comparisons of the results showed that palmitoleic acid binding had little effect on the conformation of the more closed substrate-free form of P450 BM3. By contrast, in the more open substrate-free form, this same substrate induced a closing of the entrance to the substrate-binding channel. The MD averaged structure of these two complexes obtained from docking of pamitoleic acid into the two asymmetric units of the substrate-free form were also compared to that obtained starting with the X-ray structure of the substrate-bound form. These results taken together led to the conclusion that, if indeed the substrate induces conformational changes in P450 BM3, the mouth of the substrate-access channel first closes down in response to the presence of the substrate, followed by rotation of the F-G domain to further optimize the P450 BM3-substrate interaction that would occur at a later stage.

Redox- and anion-linked protonation sites in horseradish peroxidase: analysis of distal haem pocket mutants

We have investigated the effects of mutations at residues His-42, Arg-38 and Phe-41 in the distal haem pocket of horseradish peroxidase on the changes in protonation state that accompany redox- and ligand-linked changes to the haem group. The mutations H42L and R38L result in the loss of a characteristic pH dependency in the visible spectrum of the ferrous form and a diminished dependency of the midpoint redox potential of the haem group on pH. The results support the view that His-42, with its pK probably modulated by Arg-38, provides the protonation site on the reduced enzyme that is responsible for these pH dependencies. The mutations H42L and R38L also have major effects on the binding of cyanide to the haem. We have already reported that binding of cyanide to the ferrous forms of these mutants becomes too weak to be measurable [Meunier, Rodriguez-Lopez, Smith, Thorneley and Rich (1995) Biochemistry 34, 14687-14692]. The pH dependency of the rate constants for binding of cyanide to the oxidized form of H42L suggests that CN- is the kinetically active species, in contrast with wild-type horseradish peroxidase, where HCN is the binding form. For the R38L variant, the pH dependency of cyanide binding suggests that the pK of His-42 in the absence of Arg-38 is raised to 7.5-8, in the oxidized form. In contrast with these changes, the mutant F41A exhibits cyanide-binding behaviour that is similar to that of the wild type, both in its oxidized and reduced forms. However, the rate constant for carbon monoxide recombination increases substantially, suggesting that the access route for carbon monoxide, but not for cyanide, is perturbed by this amino acid substitution.