Catalytically functional flavocytochrome chimeras of P450 BM3 and nitric oxide synthase

P450 BM3 and the nitric oxide synthases are related classes of flavocytochrome mono-oxygenase enzymes, containing NADPH-dependent FAD- and FMN-containing oxidoreductase modules fused to heme b-containing oxygenase domains. Domain-swap hybrids of these two multi-domain enzymes were created by genetic engineering of different segments of reductase and heme domains from neuronal nitric oxide synthase and P450 BM3, as a means of investigating the catalytic competence and substrate-binding properties of the fusions and the influence of tetrahydrpbiopterin and calmodulin binding regions on the electron transfer kinetics of the chimeras. Despite marked differences in hybrid stability and solubility, four catalytically functional chimeras were created that retained good reductase activity and which could be expressed successfully in Escherichia coli and purified. All of the BM3 reductase domain chimeras (chimeras I-III) exhibited inefficient flavin-to-heme inter-domain electron transfer, diminishing their oxygenase activity. However, the chimera containing the neuronal nitric oxide synthase reductase domain (chimera IV) showed good oxygenase domain activity, indicating that the flavin-to-heme electron transfer reaction is relatively efficient in this case. The data reinforce the importance of the nature of inter-domain linker constitution in multi-domain enzymes, and the difficulties posed in attempts to create chimeric enzymes with enhanced catalytic properties.     

Oxidation of human cytochrome P450 1A2 substrates by Bacillus megaterium cytochrome P450 BM3

Cytochrome P450 enzymes (P450s or CYPs) are good candidates for biocatalysis in the production of fine chemicals, including pharmaceuticals. Despite the potential use of mammalian P450s in various fields of biotechnology, these enzymes are not suitable as biocatalysts due to their low stability, low catalytic activity, and limited availability. Recently, wild-type and mutant forms of bacterial P450 BM3 (CYP102A1) from Bacillus megaterium have been found to metabolize various. It has therefore been suggested that CYP102A1 may be used to generate the metabolites of drugs and drug candidates. In this report, we show that the oxidation reactions of typical human CYP1A2 substrates (phenacetin, ethoxyresorufin, and methoxyresorufin) are catalyzed by both wild-type and mutant forms of CYP102A1. In the case of phenacetin, CYP102A1 enzymes show only O-deethylation product, even though two major products are produced as a result of O-deethylation and 3-hydroxylation reactions by human CYP1A2. Formation of the metabolites was confirmed by HPLC analysis and LC–MS to compare the metabolites with the actual biological metabolites produced by human CYP1A2. The results demonstrate that CYP102A1 mutants can be used for cost-effective and scalable production of human CYP1A2 drug metabolites. Our computational findings suggest that a conformational change in the cavity size of the active sites of the mutants is dependent on activity change. The modeling results further suggest that the activity change results from the movement of several specific residues in the active sites of the mutants.     

Fungal denitrification and nitric oxide reductase cytochrome P450nor

We have shown that many fungi (eukaryotes) exhibit distinct denitrifying activities, although occurrence of denitrification was previously thought to be restricted to bacteria (prokaryotes), and have characterized the fungal denitrification system. It comprises NirK (copper-containing nitrite reductase) and P450nor (a cytochrome P450 nitric oxide (NO) reductase (Nor)) to reduce nitrite to nitrous oxide (N(2)O). The system is localized in mitochondria functioning during anaerobic respiration. Some fungal systems further contain and use dissimilatory and assimilatory nitrate reductases to denitrify nitrate. Phylogenetic analysis of nirK genes showed that the fungal-denitrifying system has the same ancestor as the bacterial counterpart and suggested a possibility of its proto-mitochondrial origin. By contrast, fungi that have acquired a P450 from bacteria by horizontal transfer of the gene, modulated its function to give a Nor activity replacing the original Nor with P450nor. P450nor receives electrons directly from nicotinamide adenine dinucleotide to reduce NO to N(2)O. The mechanism of this unprecedented electron transfer has been extensively studied and thoroughly elucidated. Fungal denitrification is often accompanied by a unique phenomenon, co-denitrification, in which a hybrid N(2) or N(2)O species is formed upon the combination of nitrogen atoms of nitrite with a nitrogen donor (amines and imines). Possible involvement of NirK and P450nor is suggested.     

Formation of transient oxygen complexes of cytochrome p450 BM3 and nitric oxide synthase under high pressure

The kinetics of formation and transformation of oxygen complexes of two heme-thiolate proteins (the F393H mutant of cytochrome P450 BM3 and the oxygenase domain of endothelial nitric oxide synthase, eNOS) were studied under high pressure. For BM3, oxygen-binding characteristics (rate and activation volume) matched those measured for CO-binding. In contrast, pressure revealed a different CO- and oxygen-binding mechanism for eNOS, suggesting that it is hazardous to take CO-binding as a model for oxygen-binding. With eNOS, a ferric NO complex is formed as an intermediate in the second reaction cycle. Here we report the pressure stability of this compound. Furthermore, in the presence of 4-amino-tetrahydrobiopterin (ABH(4)), an analog to the natural second electron donor tetrahydrobiopterin (BH(4)), biphasic pressure profiles of the oxygen-binding rates were observed, both in the first and the second reaction cycles, indicative of the formation of an additional reaction intermediate. This was confirmed by experiments where ABH(4) was replaced by ABH(2), a cofactor which cannot deliver an electron. Altogether, high pressure appears to be a useful tool to characterize elementary steps in the reaction cycle of heme-thiolate proteins.     

Thermal inactivation of the reductase domain of cytochrome P450 BM3

Although the reductase domain of cytochrome P450 BM3 (BMR) catalyzes the reduction of cytochrome c and 2,6-dichlorophenolindophenol, we observed a catalytically independent loss of activity. By varying the incubation time for the enzyme prior to reaction initiation, we measured an inactivation rate of 0.22 min(-1). We hypothesized that either an active BMR dimer dissociates to an inactive monomer or BMR undergoes denaturation. We were not able to trap or destabilize a dimer, and BMR inactivation proved to be irreversible. Addition of excess FMN only slightly decreased the rate of inactivation from 0.22 to 0.13 min(-1), indicating inactivation likely does not reflect loss of flavin. When inactivation rates as a function of temperature were fit to the Arrhenius equation, the energy required to inactivate BMR was 9.9 kcal mol(-1)--equivalent to a few hydrogen bonds. The potential instability of BMR under certain conditions raises concerns for the use of BMR as a model or surrogate P450 reductase in other systems.     

Structural rationalization of novel drug metabolizing mutants of cytochrome P450 BM3

Three newly discovered drug metabolizing mutants of cytochrome P450 BM3 (van Vugt-Lussenburg et al., Identification of critical residues in novel drug metabolizing mutants of Cytochrome P450 BM3 using random mutagenesis, J Med Chem 2007;50:455-461) have been studied at an atomistic level to provide structural explanations for a number of their characteristics. In this study, computational methods are combined with experimental techniques. Molecular dynamics simulations, resonance Raman and UV-VIS spectroscopy, as well as coupling efficiency and substrate-binding experiments, have been performed. The computational findings, supported by the experimental results, enable structural rationalizations of the mutants. The substrates used in this study are known to be metabolized by human cytochrome P450 2D6. Interestingly, the major metabolites formed by the P450 BM3 mutants differ from those formed by human cytochrome P450 2D6. The computational findings, supported by resonance Raman data, suggest a conformational change of one of the heme propionate groups. The modeling results furthermore suggest that this conformational change allows for an interaction between the negatively charged carboxylate of the heme substituent and the positively charged nitrogen of the substrates. This allows for an orientation of the substrates favorable for formation of the major metabolite by P450 BM3.     

Process intensification for cytochrome P450 BM3-catalyzed oxy-functionalization of dodecanoic acid

Selective oxy-functionalization of nonactivated C-H bonds is a long-standing “dream reaction” of organic synthesis for which chemical methodology is not well developed. Mono-oxygenase enzymes are promising catalysts for such oxy-functionalization to establish. Limitation on their applicability arises from low reaction output. Here, we showed an integrated approach of process engineering to the intensification of the cytochrome P450 BM3-catalyzed hydroxylation of dodecanoic acid (C12:0). Using P450 BM3 together with glucose dehydrogenase for regeneration of nicotinamide adenine dinucleotide phosphate (NADPH), we compared soluble and co-immobilized enzymes in O₂-gassed and pH-controlled conversions at high final substrate concentrations (≥40mM). We identified the main engineering parameters of process output (i.e., O₂ supply; mixing correlated with immobilized enzyme stability; foam control correlated with product isolation; substrate solubilization) and succeeded in disentangling their complex interrelationship for systematic process optimization. Running the reaction at O₂-limited conditions at up to 500-ml scale (10% dimethyl sulfoxide; silicone antifoam), we developed a substrate feeding strategy based on O₂ feedback control. Thus, we achieved high reaction rates of 1.86g·L⁻¹·hr⁻¹ and near complete conversion (≥90%) of 80mM (16g/L) C12:0 with good selectivity (≤5% overoxidation). We showed that “uncoupled reaction” of the P450 BM3 (~95% utilization of NADPH and O₂ not leading to hydroxylation) with the C12:0 hydroxylated product limited the process efficiency at high product concentration. Hydroxylated product (~7g; ≥92% purity) was recovered from 500ml reaction in 82% yield using ethyl-acetate extraction. Collectively, these results demonstrate key engineering parameters for the biocatalytic oxy-functionalization and show their integration into a coherent strategy for process intensification.     

Cobaltocene-mediated catalytic monooxygenation using holo and heme domain cytochrome P450 BM3

The feasibility of replacing NADPH with 1,1'-dicarboxycobaltocene in the catalytic cycle of cytochrome P450 BM3 has been explored. Using the holoprotein, the surrogate mediator was observed to reduce both the FAD and FMN in the reductase domain, as well as the iron in the heme domain. In an electrochemical system, the mediator was able to support lauric acid hydroxylation at a rate of 16.5 nmol product/nmol enzyme/minute. Similar electron transfer and catalysis were observed for the heme domain alone in the presence of the metallocene; the turnover rate in this case was 1.8 nmol product/nmol enzyme/minute. Parallel studies under the same conditions using a previously reported cobalt sepulchrate mediator showed that the two systems give similar results for both the holoenzyme and the heme domain.     

4-cyanopyridine, a versatile spectroscopic probe for cytochrome P450 BM3

The nitrogenous pi -acceptor ligand 4-cyanopyridine (4CNPy) exhibits reversible ligation to ferrous heme in the flavocytochrome P450 BM3 (Kd=1.8 microm for wild type P450 BM3) via its pyridine ring nitrogen. The reduced P450-4CNPy adduct displays unusual spectral properties that provide a useful spectroscopic handle to probe particular aspects of this P450. 4CNPy is competitively displaced upon substrate binding, allowing a convenient route to the determination of substrate dissociation constants for ferrous P450 highlighting an increase in P450 substrate affinity on heme reduction. For wild type P450 BM3, Kd(red)(laurate)=82.4 microm (cf. Kd(ox)=364 microm). In addition, an unusual spectral feature in the red region of the absorption spectrum of the reduced P450-4CNPy adduct is observed that can be assigned as a metal-to-ligand charge transfer (MLCT). It was discovered that the energy of this MLCT varies linearly with respect to the P450 heme reduction potential. By studying the energy of this MLCT for a series of BM3 active site mutants with differing reduction potential (Em), the relationship EMLCT + (3.53 x = Em 17,005 cm)(-1) was derived. The use of this ligand thus provides a quick and accurate method for predicting the heme reduction potentials of a series of P450 BM3 mutations using visible spectroscopy, without the requirement for redox potentiometry.     

Resonance Raman scattering of cytochrome P450 BM3 and effect of imidazole inhibitors

Resonance Raman scattering from cytochrome P450 BM3 is obtained with a Raman microprobe using 406-nm excitation with an accumulation time of a few seconds. The small sample size and rapid measurement time make the routine characterization of P450 systems by resonance Raman spectroscopy easier. Addition of imidazole and imidazole derivatives as inhibitors causes the appearance of additional peaks due to vinyl modes, increases the relative intensity of symmetric modes that would be A(1g) in D(4h) symmetry, and causes a large drop in the intensity of nu(11). This information indicates that the ligation of imidazoles to the heme iron causes the alignment of the vinyl modes with the plane of the heme ring and reduces the out of plane distortion of the ring. The effect of both inhibitors is similar but there is a subtle difference in the extent of the reduction in the intensity of nu(11), which suggests that steric effects within the pocket are having some effect.     

Fatty acid-induced alteration of the porphyrin macrocycle of cytochrome P450 BM3.

Surface-enhanced resonance Raman scattering (SERRS) of substrate-free and substrate-bound forms of the P450 domain of cytochrome P450 BM3 are reported and assigned. Substrate-free P450 yields mixed spin heme species in which the pentacoordinate high-spin arrangement is dominant. The addition of laurate or palmitate leads to an increase in high spin content and to an allosteric activation of heme mode v29, which is sensitive to peripheral heme/protein interactions. Differences between laurate and palmitate binding are observed in the relative intensities of a number of bands and the splitting of the heme vinyl modes. Laurate binding to P450 results in different protein environments being experienced by each vinyl mode, whereas palmitate binding produces a smaller difference. The results demonstrate the ability of SERRS to probe substrate/prosthetic group interactions within an active site, at low protein concentrations.     

Resonance Raman studies of cytochrome P450BM3 and its complexes with exogenous ligands.

Resonance Raman spectra are reported for both the heme domain and holoenzyme of cytochrome P450BM3 in the resting state and for the ferric NO, ferrous CO, and ferrous NO adducts in the absence and presence of the substrate, palmitate. Comparison of the spectrum of the palmitate-bound form of the heme domain with that of the holoenzyme indicates that the presence of the flavin reductase domain alters the structure of the heme domain in such a way that water accessibility to the distal pocket is greater for the holoenzyme, a result that is consistent with analogous studies of cytochrome P450cam. The data for the exogenous ligand adducts are compared to those previously reported for corresponding derivatives of cytochrome P450cam and document significant and important differences for the two proteins. Specifically, while the binding of substrate induces relatively dramatic changes in the nu(Fe-XY) modes of the ferrous CO, ferric NO, and ferrous NO derivatives of cytochrome P450cam, no significant changes are observed for the corresponding derivatives of cytochrome P450BM3 upon binding of palmitate. In fact, the spectral data for substrate-free cytochrome P450BM3 provide evidence for distortion of the Fe-XY fragment, even in the absence of substrate. This apparent distortion, which is nonexistent in the case of substrate-free cytochrome P450cam, is most reasonably attributed to interaction of the Fe-XY fragment with the F87 phenylalanine side chain. This residue is known to lie very close to the heme iron in the substrate-free derivative of cytochrome P450BM3 and has been suggested to prevent hydroxylation of the terminal, omega, position of long-chain fatty acids.     

The role of Thr268 and Phe393 in cytochrome P450 BM3

In flavocytochrome P450 BM3 there are several active site residues that are highly conserved throughout the P450 superfamily. Of these, a phenylalanine (Phe393) has been shown to modulate heme reduction potential through interactions with the implicitly conserved heme-ligand cysteine. In addition, a distal threonine (Thr268) has been implicated in a variety of roles including proton donation, oxygen activation and substrate recognition. Substrate binding in P450 BM3 causes a shift in the spin state from low- to high-spin. This change in spin-state is accompanied by a positive shift in the reduction potential (DeltaE(m) [WT+arachidonate (120 microM)]=+138 mV). Substitution of Thr268 by an alanine or asparagine residue causes a significant decrease in the ability of the enzyme to generate the high-spin complex via substrate binding and consequently leads to a decrease in the substrate-induced potential shift (DeltaE(m) [T268A+arachidonate (120 microM)]=+73 mV, DeltaE(m) [T268N+arachidonate (120 microM)]=+9 mV). Rate constants for the first electron transfer and for oxy-ferrous decay were measured by pre-steady-state stopped-flow kinetics and found to be almost entirely dependant on the heme reduction potential. More positive reduction potentials lead to enhanced rate constants for heme reduction and more stable oxy-ferrous species. In addition, substitutions of the threonine lead to an increase in the production of hydrogen peroxide in preference to hydroxylated product. These results suggest an important role for this active site threonine in substrate recognition and in maintaining an efficiently functioning enzyme. However, the dependence of the rate constants for oxy-ferrous decay on reduction potential raises some questions as to the importance of Thr268 in iron-oxo stabilisation.     

Heterotropic and homotropic cooperativity by a drug-metabolising mutant of cytochrome P450 BM3

Recently, we described a triple mutant of the bacterial cytochrome P450 BM3 as the first mutant with affinity for drug-like compounds. In this paper, we show that this mutant, but not wild-type BM3, is able to metabolise testosterone and several drug-like molecules such as amodiaquine, dextromethorphan, acetaminophen, and 3,4-methylenedioxymethylamphetamine that are known substrates of human P450s. Interestingly, the metabolism of 3,4-methylenedioxymethylamphetamine and acetaminophen could be stimulated up to 70-fold by the addition of caffeine, a known activator of rat P450 3A2. With testosterone metabolism, homotropic cooperativity was observed. This shows that heterotropic and homotropic cooperativity, known to occur in the P450 3A family, can also take place in BM3. BM3 therefore can be used as a model system to study atypical kinetics in mammalian P450s. Second, this study shows that BM3 can be engineered to a drug-metabolising enzyme, making it a promising candidate to use as biocatalyst in drug discovery and synthesis.     

Formation of nitric oxide by cytochrome P450-catalyzed oxidation of aromatic amidoximes.

Rat liver microsomes catalyze the oxidation of para-hexyloxy-benzamidoxime 1 to the corresponding arylamide 2 and NO2-, by NADPH and O2. Involvement of cytochromes P450 as catalysts of this reaction was shown by the strong inhibitory effects of CO and miconazole and the spectacular increase of the activity upon treatment of rats with dexamethasone, a specific inducer of cytochromes P450 of the 3A subfamily. Formation of NO during oxidation of 1 was shown by detection of the formation of cytochrome P450- and cytochrome P420-Fe(II)-NO complexes by visible and EPR spectroscopy. The formation of these complexes should be responsible, at least in part, for the fast decrease of the rate of microsomal oxidation of 1 with time. These results suggest that exogenous compounds containing amidine or amidoxime functions could act as precursors of NO in vivo after in situ oxidation by cytochromes P450.     

Characterization of the mechanism of cytochrome P450 reductase-cytochrome P450-mediated nitric oxide and nitrosothiol generation from organic nitrates

Mammalian cytochrome P450 reductase (CPR) and cytochrome P450 (CP) play important roles in organic nitrate bioactivation; however, the mechanism by which they convert organic nitrate to NO remains unknown. Questions remain regarding the initial precursor of NO that serves to link organic nitrate to the activation of soluble guanylyl cyclase (sGC). To characterize the mechanism of CPR-CP-mediated organic nitrate bioactivation, EPR, chemiluminescence NO analyzer, NO electrode, and immunoassay studies were performed. With rat hepatic microsomes or purified CPR, the presence of NADPH triggered organic nitrate reduction to NO2(-). The CPR flavin site inhibitor diphenyleneiodonium inhibited this NO2(-) generation, whereas the CP inhibitor clotrimazole did not. However, clotrimazole greatly inhibited NO2(-)-dependent NO generation. Therefore, CPR catalyzes organic nitrate reduction, producing nitrite, whereas CP can mediate further nitrite reduction to NO. Nitrite-dependent NO generation contributed <10% of the CPR-CP-mediated NO generation from organic nitrates; thus, NO2(-) is not the main precursor of NO. CPR-CP-mediated NO generation was largely thiol-dependent. Studies suggested that organic nitrite (R-O-NO) was produced from organic nitrate reduction by CPR. Further reaction of organic nitrite with free or microsome-associated thiols led to NO or nitrosothiol generation and thus stimulated the activation of sGC. Thus, organic nitrite is the initial product in the process of CRP-CP-mediated organic nitrate activation and is the precursor of NO and nitrosothiols, serving as the link between organic nitrate and sGC activation.     

Evaluation of alkoxyresorufins as fluorescent substrates for cytochrome P450 BM3 and site-directed mutants

In this study, the first fluorescent assay for bacterial cytochrome P450 BM3 (BM3) and mutants is described. BM3 mutants are potentially very versatile biocatalysts for the production of fine chemicals. A fluorescent assay would be very useful for the identification of nonnatural ligands in high-throughput inhibition assays. Because of the ease and sensitivity of alkoxyresorufin O-dealkylation assays, four different alkoxyresorufins were evaluated as substrates. Wild-type BM3 showed extremely low activity toward all four alkoxyresorufins tested. Five different BM3 mutants were constructed, carrying different combinations of mutations R47L, F87V, and L188Q, which were previously shown to increase activity toward nonnatural substrates. For all mutants, a high benzyloxyresorufin O-dealkylation (BROD) activity was found. The triple mutant of BM3, R47L/F87V/L188Q, showed the highest activity, increasing 900-fold compared to wild-type BM3. The BROD assay could also be applied in whole Escherichia coli cells; permeabilization by lipopolysaccharide deficiency strongly increased activity. To demonstrate the applicability of the BROD assay to screening for novel ligands of BM3 R47L/F87V/L188Q, a library of 45 drug-like compounds was tested for inhibition. Of these compounds, 8 showed strong inhibition of the BROD activity, demonstrating for the first time that drug-like molecules also can bind with high affinity to BM3 mutants.     

Spectrophotometric activity microassay for pure and recombinant cytochrome P450-type nitric oxide reductase

Nitric oxide reductase (NOR) of the P450 oxidoreductase family accepts electrons directly from its cofactor, NADH, to reduce two nitric oxide (NO) molecules to one nitrous oxide molecule and water. The enzyme plays a key role in the removal of radical NO produced during respiratory metabolism, and applications in bioremediation and biocatalysis have been identified. However, a rapid, accurate, and sensitive enzyme assay has not yet been developed for this enzyme family. In this study, we optimized reaction conditions for the development of a spectrophotometric NOR activity microassay using NOC-5 for the provision of NO in solution. We also demonstrate that the assay is suitable for the quantification and characterization of P450-type NOR. The K_m and k_cat kinetic constants obtained by this assay were comparable to the values determined by gas chromatography, but with improved convenience and cost efficiency, effectively by miniaturization. To our knowledge, this is the first study to present the quantification of NOR activity in a kinetic microassay format.     

Filling a hole in cytochrome P450 BM3 improves substrate binding and catalytic efficiency

Cytochrome P450BM3 (CYP102A1) from Bacillus megaterium, a fatty acid hydroxylase, is a member of a very large superfamily of monooxygenase enzymes. The available crystal structures of the enzyme show non-productive binding of substrates with their omega-end distant from the iron in a hydrophobic pocket at one side of the active site. We have constructed and characterised mutants in which this pocket is filled by large hydrophobic side-chains replacing alanine at position 82. The mutants having phenylalanine or tryptophan at this position have very much (approximately 800-fold) greater affinity for substrate, with a greater conversion of the haem iron to the high-spin state, and similarly increased catalytic efficiency. The enzyme as isolated contains bound palmitate, reflecting this much higher affinity. We have determined the crystal structure of the haem domain of the Ala82Phe mutant with bound palmitate; this shows that the substrate is binding differently from the wild-type enzyme but still distant from the haem iron. Detailed analysis of the structure indicates that the tighter binding in the mutant reflects a shift in the conformational equilibrium of the substrate-free enzyme towards the conformation seen in the substrate complex rather than differences in the enzyme-substrate interactions. On this basis, we outline a sequence of events for the initial stages of the catalytic cycle. The Ala82Phe and Ala82Trp mutants are also very much more effective catalysts of indole hydroxylation than the wild-type enzyme, suggesting that they will be valuable starting points for the design of mutants to catalyse synthetically useful hydroxylation reactions.