Protein dynamics in cytochrome P450 molecular recognition and substrate specificity using 2D IR vibrational echo spectroscopy

Cytochrome (cyt) P450s hydroxylate a variety of substrates that can differ widely in their chemical structure. The importance of these enzymes in drug metabolism and other biological processes has motivated the study of the factors that enable their activity on diverse classes of molecules. Protein dynamics have been implicated in cyt P450 substrate specificity. Here, 2D IR vibrational echo spectroscopy is employed to measure the dynamics of cyt P450(cam) from Pseudomonas putida on fast time scales using CO bound at the active site as a vibrational probe. The substrate-free enzyme and the enzyme bound to both its natural substrate, camphor, and a series of related substrates are investigated to explicate the role of dynamics in molecular recognition in cyt P450(cam) and to delineate how the motions may contribute to hydroxylation specificity. In substrate-free cyt P450(cam), three conformational states are populated, and the structural fluctuations within a conformational state are relatively slow. Substrate binding selectively stabilizes one conformational state, and the dynamics become faster. Correlations in the observed dynamics with the specificity of hydroxylation of the substrates, the binding affinity, and the substrates' molecular volume suggest that motions on the hundreds of picosecond time scale contribute to the variation in activity of cyt P450(cam) toward different substrates.     

Dynamic mechanism for the autophosphorylation of CheA histidine kinase: molecular dynamics simulations

The two-component system (TCS) is an important signal transduction component for most bacteria. This signaling pathway is mediated by histidine kinases via autophosphorylation between P1 and P4 domains. Taking chemotaxis protein CheA as a model of TCS, the autophosphorylation mechanism of the TCS histidine kinases has been investigated in this study by using a computational approach integrated homology modeling, ligand-protein docking, protein-protein docking, and molecular dynamics (MD) simulations. Four nanosecond-scale MD simulations were performed on the free P4 domain, P4-ATP, P4-TNPATP, and P1-P4-ATP complexes, respectively. Upon its binding to the binding pocket of P4 with a folded conformation, ATP gradually extends to an open state with help from a water molecule. Meanwhile, ATP forms two hydrogen bonds with His413 and Lys494 at this state. Because of the lower energy of the folded conformations, ATP shrinks back to its folded conformations, leading to the rupture of the hydrogen bond between ATP and Lys494. Consequently, Lys494 moves away from the pocket entrance, resulting in an open of the ATP lid of P4. It is the open state of P4 that can bind tightly to P1, where the His45 of P1 occupies a favorable position for its autophosphorylation from ATP. This indicates that ATP is not only a phosphoryl group donor but also an activator for CheA phosphorylation. Accordingly, a mechanism of the autophosphorylation of CheA is proposed as that the ATP conformational switch triggers the opening of the ATP lid of P4, leading to P1 binding tightly, and subsequently autophosphorylation from ATP to P1.     

Substrate binding favors enhanced NO binding to P450cam

Ferric cytochrome P450cam from Pseudomonas putida (P450cam) in buffer solution at physiological pH 7.4 reversibly binds NO to yield the nitrosyl complex P450cam(NO). The presence of 1R-camphor affects the dynamics of NO binding to P450cam and enhances the association and dissociation rate constants significantly. In the case of the substrate-free form of P450cam, subconformers are evident and the NO binding kinetics are much slower than in the presence of the substrate. The association and dissociation processes were investigated by both laser flash photolysis and stopped-flow techniques at ambient and high pressure. Large and positive values of S and V observed for NO binding to and release from the substrate-free P450cam complex are consistent with the operation of a limiting dissociative ligand substitution mechanism, where the lability of coordinated water dominates the reactivity of the iron(III)-heme center with NO. In contrast, NO binding to P450cam in the presence of camphor displays negative activation entropy and activation volume values that support a mechanism dominated by a bond formation process. Volume profiles for the binding of NO appear to be a valuable approach to explain the differences observed for P450cam in the absence and presence of the substrate and enable the clarification of the underlying reaction mechanisms at a molecular level. Changes in spin state of the iron center during the binding/release of NO contribute significantly to the observed volume effects. The results are discussed in terms of relevance for the biological function of cytochrome P450 and in context to other investigations of the related reactions between NO and imidazole- and thiolate-ligated iron(III) hemoproteins.     

Cryoreduction EPR and 13C, 19F ENDOR study of substrate-bound substates and solvent kinetic isotope effects in the catalytic cycle of cytochrome P450cam and its T252A mutant

We recently used cryoreduction EPR/ENDOR techniques to show that a substrate can modulate the properties of both the monooxygenase active-oxygen intermediates and of the proton-delivery network which encompasses them. In the present report we use Q-band pulsed 19F ENDOR (Mims 3-pulse sequence) to examine the substrate binding geometries of camphor, through use of the 5,5'--difluorocamphor, and 13C ENDOR to examine the binding of 5-methylenyl camphor labeled with 13C at C11. These probes are examined in multiple states of the catalytic cycle of P450cam and its T252A mutant. As part of this investigation we further report a new cryoreduction reaction, the reduction of a ferroheme to the EPR-visible Fe(I) state, and use it to probe the substrate binding to the EPR-silent ferroheme state. Finally we report the solvent kinetic isotope effect on the decay of the camphor complex of the hydroperoxo-ferric intermediate, the first such measurement on an individual step within the P450cam reaction cycle. Following reduction of oxyferrous-P450cam, this step is the rate-limiting step in camphor hydroxylation, and its solv-KIE of 1.8 at 190 K establishes that it involves activation of the hydroperoxo moiety by transfer of the 'second' proton of catalysis. We suggest that the finding that the heme pocket can exist in multiple substates, including multiple substrate binding locations, even in P450cam, along with the established possibility that the hydroperoxo-ferriheme intermediate can react with substrate, may explain the formation of multiple products by P450s.     

Molecular recognition in (+)-alpha-pinene oxidation by cytochrome P450cam

Oxygenated derivatives of the monoterpene (+)-alpha-pinene are found in plant essential oils and used as fragrances and flavorings. (+)-alpha-Pinene is structurally related to (+)-camphor, the natural substrate of the heme monooxygenase cytochrome P450(cam) from Pseudomonas putida. The aim of the present work was to apply the current understanding of P450 substrate binding and catalysis to engineer P450(cam) for the selective oxidation of (+)-alpha-pinene. Consideration of the structures of (+)-camphor and (+)-alpha-pinene lead to active-site mutants containing combinations of the Y96F, F87A, F87L, F87W, and V247L mutations. All mutants showed greatly enhanced binding and rate of oxidation of (+)-alpha-pinene. Some mutants had tighter (+)-alpha-pinene binding than camphor binding by the wild-type. The most active was the Y96F/V247L mutant, with a (+)-alpha-pinene oxidation rate of 270 nmol (nmol of P450(cam))(-)(1) min(-)(1), which was 70% of the rate of camphor oxidation by wild-type P450(cam). Camphor is oxidized by wild-type P450(cam) exclusively to 5-exo-hydroxycamphor. If the gem dimethyl groups of (+)-alpha-pinene occupied similar positions to those found for camphor in the wild-type structure, (+)-cis-verbenol would be the dominant product. All P450(cam) enzymes studied gave (+)-cis-verbenol as the major product but with much reduced selectivity compared to camphor oxidation by the wild-type. (+)-Verbenone, (+)-myrtenol, and the (+)-alpha-pinene epoxides were among the minor products. The crystal structure of the Y96F/F87W/V247L mutant, the most selective of the P450(cam) mutants initially examined, was determined to provide further insight into P450(cam) substrate binding and catalysis. (+)-alpha-Pinene was bound in two orientations which were related by rotation of the molecule. One orientation was similar to that of camphor in the wild-type enzyme while the other was significantly different. Analysis of the enzyme/substrate contacts suggested rationalizations of the product distribution. In particular competition rather than cooperativity between the F87W and V247L mutations and substrate movement during catalysis were proposed to be major factors. The crystal structure lead to the introduction of the L244A mutation to increase the selectivity of pinene oxidation by further biasing the binding orientation toward that of camphor in the wild-type structure. The F87W/Y96F/L244A mutant gave 86% (+)-cis-verbenol and 5% (+)-verbenone. The Y96F/L244A/V247L mutant gave 55% (+)-cis-verbenol but interestingly also 32% (+)-verbenone, suggesting that it may be possible to engineer a P450(cam) mutant that could oxidize (+)-alpha-pinene directly to (+)-verbenone. Verbenol, verbenone, and myrtenol are naturally occurring plant fragrance and flavorings. The preparation of these compounds by selective enzymatic oxidation of (+)-alpha-pinene, which is readily available in large quantities, could have applications in synthesis. The results also show that the protein engineering of P450(cam) for high selectivity of substrate oxidation is more difficult than achieving high substrate turnover rates because of the subtle and dynamic nature of enzyme-substrate interactions.     

Thermodynamic and kinetic studies on the binding of nitric oxide to a new enzyme mimic of cytochrome p450

A new model for the P450 enzyme carrying a SO(3)(-) ligand coordinated to iron(III) (complex 2) reversibly binds NO to yield the nitrosyl adduct. The rate constant for NO binding to 2 in toluene is of the same order of magnitude as that found for the nitrosylation of the native, substrate-bound form of P450(cam) (E.S-P450(cam)). Large and negative activation entropy and activation volume values for the binding of NO to complex 2 support a mechanism that is dominated by bond formation with concomitant iron spin change from S = (5)/(2) to S = 0, as proposed for the reaction between NO and E.S-P450(cam). In contrast, the dissociation of NO from 2(NO) was found to be several orders of magnitude faster than the corresponding reaction for the E.S-P450(cam)/NO system. In a coordinating solvent such as methanol, the alcohol coordinates to iron(III) of 2 at the distal position, generating a six-coordinate, high-spin species 5. The reaction of NO with 5 in methanol was found to be much slower in comparison to the nitrosylation reaction of 2 in toluene. This behavior can be explained in terms of a mechanism in which methanol must be displaced during Fe-NO bond formation. The thermodynamic and kinetic data for NO binding to the new model complexes of P450 (2 and 5) are discussed in reference to earlier results obtained for closely related nitrosylation reactions of cytochrome P450(cam) (in the presence and in the absence of the substrate) and a thiolate-ligated iron(III) model complex.     

Fluorescent probes for cytochrome p450 structural characterization and inhibitor screening

We have synthesized two luminescent probes (D-4-Ad and D-8-Ad) that target cytochrome P450cam. D-4-Ad luminescence is quenched by F?rster energy transfer upon binding (Kd = 0.83 muM) but is restored when the probe is displaced from the active site by camphor. In contrast, D-8-Ad (Kd approximately 0.02 muM) is not displaced from the enzyme, even in the presence of a large excess of camphor. The 2.2 A resolution crystal structure of the D-8-Ad:P450cam complex reveals extensive hydrophobic contacts between the probe and the enzyme, which result from the conformational flexibility of the B', F, and G helices. Probes with properties similar to those of D-4-Ad potentially could be useful for screening P450 inhibitors.     

Dissecting the thermodynamics and cooperativity of ligand binding in cytochrome P450eryF

Conformational flexibility and cooperativity in ligand recognition are two key aspects of the catalytic diversity of cytochrome P450 enzymes. In this study, we dissect the ligand binding stoichiometry and energetics of the soluble bacterial P450eryF by isothermal titration calorimetry (ITC) using three allosteric and two non-allosteric ligands of diverse chemistry. Complementary spectral binding studies and sequential, two-ligand docking simulations were performed to help assign the binding sites. Binding of 4-phenylpyridine (4-PP) or 4-(4-chlorophenyl)imidazole (4-CPI) showed 1:1 stoichiometry in ITC, consistent with the lack of cooperativity observed in spectral binding studies. The larger ligands 9-aminophenanthrene (9-AP), 1-pyrenebutanol (1-PB), and alpha-naphthoflavone (ANF) show cooperative spectral binding and yielded 2:1 stoichiometry. The associated thermodynamic parameters for the sites were calculated using a sequential binding mechanism. The binding constant (KD) for the first site was almost two times lower than that of the second site for all three compounds. Ligand binding at site 1 was entropically favored, whereas binding at site 2 was weak and entropically unfavorable. Simulations showed that two molecules of 9-AP, ANF or 1-PB can be adequately docked to two individual sub-sites within a large binding pocket. The absence of hydrophobic tethering and ligand stacking are consistent with the single low affinity binding site observed for 4-CPI and 4-PP. Competitive binding studies with P450eryF preloaded with either 1-PB or ANF showed a decrease in the affinities for 9-AP at both the sites, with large entropy-enthalpy compensation, indicating the ability of the binding pocket to accommodate two ligands of diverse chemistry and enable cooperativity.     

Enhanced electron transfer and lauric acid hydroxylation by site-directed mutagenesis of CYP119

CYP119, a cytochrome P450 from a thermophilic organism for which a crystal structure is available, is shown here to hydroxylate lauric acid in a reaction supported by putidaredoxin and putidaredoxin reductase. This fatty acid hydroxylation activity is increased 15-fold by T214V and D77R mutations. The T214V mutation increases the rate by facilitating substrate binding and enhancing the associated spin state change, whereas the D77R mutation improves binding of the heterologous redox partner putidaredoxin to CYP119 and the rate of electron transfer from it to the heme group. A sequence alignment with P450(cam) can, therefore, be used to identify a part of the binding site for putidaredoxin on an unrelated P450 enzyme. This information can be used to engineer by mutagenesis an improved complementarity of the protein-protein interface that results in improved electron transfer from putidaredoxin to the P450 enzyme. As a result, the catalytic activity of the thermo- and barostable CYP119 has been incorporated into a catalytic system that hydroxylates fatty acids.     

Thermodynamics of binding of 2-methoxy-3-isopropylpyrazine and 2-methoxy-3-isobutylpyrazine to the major urinary protein

In the present study we examine the thermodynamics of binding of two related pyrazine-derived ligands to the major urinary protein, MUP-I, using a combination of isothermal titration calorimetry (ITC), X-ray crystallography, and NMR backbone (15)N and methyl side-chain (2)H relaxation measurements. Global thermodynamics data derived from ITC indicate that binding is driven by favorable enthalpic contributions, rather than the classical entropy-driven hydrophobic effect. Unfavorable entropic contributions from the protein backbone and side-chain residues in the vicinity of the binding pocket are partially offset by favorable entropic contributions at adjacent positions, suggesting a "conformational relay" mechanism whereby increased rigidity of residues on ligand binding are accompanied by increased conformational freedom of side chains in adjacent positions. The principal driving force governing ligand affinity and specificity can be attributed to solvent-driven enthalpic effects from desolvation of the protein binding pocket.     

Molecular dynamics simulations of ligand-induced backbone conformational changes in the binding site of the periplasmic lysine-, arginine-, ornithine-binding protein

The periplasmic lysine-, arginine-, ornithine-binding protein (LAOBP) traps its ligands by a large hinge bending movement between two globular domains. The overall geometry of the binding site remains largely unchanged between the open (unliganded) and closed (liganded) forms, with only a small number of residues exhibiting limited movement of their side chains. However, in the case of the ornithine-bound structure, the backbone peptide bond between Asp11 and Thr12 undergoes a large rotation. Molecular dynamics simulations have been used to investigate the origin and mechanism of this backbone movement. Simulations allowing flexibility of a limited region and of the whole binding site, with and without bound ligands, suggest that this conformational change is induced by the binding of ornithine, leading to the stabilisation of an energetically favourable alternative conformation.     

Molecular dynamics simulation of the human adenosine A3 receptor: agonist induced conformational changes of Trp243

The adenosine A(3) receptor together with rhodopsin belongs to Class A of the G-protein coupled receptors. As the crystal structure of bovine rhodopsin represents the dark (inactive) state of the receptor, the details of GPCR activation are still unknown. In this molecular dynamics study we investigate how the homology model of the human adenosine A(3) receptor responds to ligand exposure. To this end we placed the homology model in a POPC membrane model. After equilibrating for 13 ns an agonist (Cl-IB-MECA) and an inverse agonist (PSB-10) were placed inside the putative binding pocket. In the following 10 ns molecular dynamics simulation we observed a different behaviour of the side-chain torsions of Trp243(6.48), depending on the presence or absence of the agonist or inverse agonist. This conformational change of Trp243 correlates with the assumed influence of ligands on receptor activation. Other predicted conformational changes of the receptor could not be observed yet. So Trp243 may represent the first switch in receptor activation.     

Electron-transfer reactions and functionalization of cytochrome P450cam monooxygenase system in reverse micelles

Enzyme-based electron-transfer reactions involved in the cytochrome P450 monooxygenase system were investigated in nanostructural reverse micelles. A bacterial flavoprotein, putidaredoxin reductase (PdR), was activated and shown to be capable of catalyzing the electron transport from NADH to electron-carrier proteins such as cytochrome b5 (tCyt-b5) and putidaredoxin (Pdx) in reverse micelles. Ferric tCyt-b5 in reverse micelles was effectively converted to its ferrous form by the exogenous addition of separately prepared reverse micellar solution harboring PdR and NADH. The fact that direct interactions of macromolecular proteins should be possible in the reverse micellar system encouraged us to functionalize a multicomponent monooxygenase system composed of the bacterial cytochrome P450cam (P450cam), putidaredoxin (Pdx), and PdR in reverse micelles. The successful camphor hydroxylation reaction catalyzed by P450cam was significantly dependent on the coexistence of Pdx, PdR, and NADH but not H2O2, suggesting that the oxygen-transfer reactions proceeded via a "monooxygenation" mechanism. This is the first report of a multicomponent cytochrome P450 system exhibiting enzymatic activity in organic media.     

Computation of binding free energy with molecular dynamics and grand canonical Monte Carlo simulations

The binding of a ligand to a receptor is often associated with the displacement of a number of bound water molecules. When the binding site is exposed to the bulk region, this process may be sampled adequately by standard unbiased molecular dynamics trajectories. However, when the binding site is deeply buried and the exchange of water molecules with the bulk region may be difficult to sample, the convergence and accuracy in free energy perturbation (FEP) calculations can be severely compromised. These problems are further compounded when a reduced system including only the region surrounding the binding site is simulated. To address these issues, we couple molecular dynamics (MD) with grand canonical Monte Carlo (GCMC) simulations to allow the number of water to fluctuate during an alchemical FEP calculation. The atoms in a spherical inner region around the binding pocket are treated explicitly while the influence of the outer region is approximated using the generalized solvent boundary potential (GSBP). At each step during thermodynamic integration, the number of water in the inner region is equilibrated with GCMC and energy data generated with MD is collected. Free energy calculations on camphor binding to a deeply buried pocket in cytochrome P450cam, which causes about seven water molecules to be expelled, are used to test the method. It concluded that solvation free energy calculations with the GCMC/MD method can greatly improve the accuracy of the computed binding free energy compared to simulations with fixed number of water.     

Biotransformation of the sesquiterpene (+)-valencene by cytochrome P450cam and P450BM-3

The sesquiterpenoids are a large class of naturally occurring compounds with biological functions and desirable properties. Oxidation of the sesquiterpene (+)-valencene by wild type and mutants of P450cam from Pseudomonas putida, and of P450BM-3 from Bacillus megaterium, have been investigated as a potential route to (+)-nootkatone, a fine fragrance. Wild type P450cam did not oxidise (+)-valencene but the mutants showed activities up to 9.8 nmol (nmol P450)(-1) min(-1), with (+)-trans-nootkatol and (+)-nootkatone constituting >85% of the products. Wild type P450BM-3 and mutants had higher activities (up to 43 min(-1)) than P450cam but were much less selective. Of the many products, cis- and trans-(+)-nootkatol, (+)-nootkatone, cis-(+)-valencene-1,10-epoxide, trans-(+)-nootkaton-9-ol, and (+)-nootkatone-13S,14-epoxide were isolated from whole-cell reactions and characterised. The selectivity patterns suggest that (+)-valencene has one binding orientation in P450cam but multiple orientations in P450BM-3.     

Cytochrome P450cam-monoterpene interactions

The use of the molecular mechanics AMBER force field (FF) to predict product profiles for the hydroxylation of the monoterpenes 1R-camphor, 1S-camphor, 1R-norcamphor, 1S-norcamphor and camphane by the enzyme cytochrome P450cam from the soil bacterium Pseudomonas putida was investigated. Predictions were carried out by applying multiple substrate (starting) orientations in the enzyme pocket in two procedures: a procedure based on molecular dynamics (MD) and a procedure based on short MD simulations followed by geometry optimisations. The latter (GO) procedure is faster and enabled the use of more monoterpene starting orientations. Monoterpene orientations were transformed into product profiles by applying both energetic and geometrical criteria appropriate for the (monoterpene) hydrogen abstraction reaction. Good predictions compared to experimental data were obtained for most compounds in both the MD and GO procedures. Prior to the product profile calculations, the FF was calibrated by reproducing the experimental data for the binding energy of 1R-camphor and 1S-camphor to P450cam and the energy of vaporisation of water. Focus of the calibration was on the value for the scaling factor for the electrostatic interactions.     

Molecular modeling of cytochrome P450 3A4

The three-dimensional structure of human cytochrome P450 3A4 was modeled based on crystallographic coordinates of four bacterial P450s: P450 BM-3, P450cam, P450terp, and P450eryF. The P450 3A4 sequence was aligned to those of the known proteins using a structure-based alignment of P450 BM-3, P450cam, P450terp, and P450eryF. The coordinates of the model were then calculated using a consensus strategy, and the final structure was optimized in the presence of water. The P450 3A4 model resembles P450 BM-3 the most, but the B helix is similar to that of P450eryF, which leads to an enlarged active site when compared with P450 BM-3, P450cam, and P450terp. The 3A4 residues equivalent to known substrate contact residues of the bacterial proteins and key residues of rat P450 2B1 are located in the active site or the substrate access channel. Docking of progesterone into the P450 3A4 model demonstrated that the substrate bound in a 6-orientation can interact with a number of active site residues, such as 114, 119, 301, 304, 305, 309, 370, 373, and 479, through hydrophobic interactions. The active site of the enzyme can also accommodate erythromycin, which, in addition to the residues listed for progesterone, also contacts residues 101, 104, 105, 214, 215, 217, 218, 374, and 478. The majority of 3A4 residues which interact with progesterone and/or erythromycin possess their equivalents in key residues of P450 2B enzymes, except for residues 297, 480 and 482, which do not contact either substrate in P450 3A4. The results from docking of progesterone and erythromycin into the enzyme model make it possible to pinpoint residues which may be important for 3A4 function and to target them for site-directed mutagenesis.     

Cytochrome P450cam–monoterpene interactions

The use of the molecular mechanics AMBER force field (FF) to predict product profiles for the hydroxylation of the monoterpenes 1R-camphor, 1S-camphor, 1R-norcamphor, 1S-norcamphor and camphane by the enzyme cytochrome P450cam from the soil bacterium Pseudomonas putida was investigated. Predictions were carried out by applying multiple substrate (starting) orientations in the enzyme pocket in two procedures: a procedure based on molecular dynamics (MD) and a procedure based on short MD simulations followed by geometry optimisations. The latter (GO) procedure is faster and enabled the use of more monoterpene starting orientations. Monoterpene orientations were transformed into product profiles by applying both energetic and geometrical criteria appropriate for the (monoterpene) hydrogen abstraction reaction. Good predictions compared to experimental data were obtained for most compounds in both the MD and GO procedures. Prior to the product profile calculations, the FF was calibrated by reproducing the experimental data for the binding energy of 1R-camphor and 1S-camphor to P450cam and the energy of vaporisation of water. Focus of the calibration was on the value for the scaling factor for the electrostatic interactions.     

The origin of the low-spin character of the resting state of cytochrome P450cam investigated by means of active site analogues

Crown-capped iron(S-) porphyrins 1 x H2O and 2 x H2O and their corresponding Ba2+ complexes have been prepared as active site analogues of the resting state of cytochrome P450cam. cw-EPR studies and electronic structure calculations at the density functional theory (DFT) level of model systems suggest a functional role of the water cluster of P450cam.     

Heteronuclear NMR investigation on the structure and dynamics of the chromophore binding pocket of the cyanobacterial phytochrome Cph1

Structural changes of the chromophore in phytochrome proteins associated with its photocycle are still not fully understood. We use heteronuclear NMR to investigate the conformation and dynamics of the chromophore in the binding pocket of the cyanobacterial phytochrome Cph1. On the basis of distance information obtained from three-dimensional nuclear Overhauser enhancement (3D-NOESY) spectra using the photochemically intact photosensory module of Cph1 we demonstrate that the chromophore is in the ZZZssa form in the P(r) (red absorbing form) state and the ZZEssa form in the P(fr) (far-red absorbing form) state of the protein. While ZZZssa for the P(r) state is in agreement with a recently determined X-ray structure, no comparable information for the P(fr) state of photochemically intact phytochrome has been available up to now. In addition, the chromophore in the binding pocket of Cph1 exhibits a notable mobility, which is distinctly different in the two photostates.