Wild-type P450cam has been shown to oxidize ethylbenzene to 1-phenylethanol with some stereoselectivity but low activity and coupling (Scheme 4.3) . Substituting residues in the upper region of the substrate pocket (e.g. the T185L and V247M mutations, see Fig. 4.1) increased the coupling 2-fold, while similar mutations near the porphyrin plane (e.g. T101M and V295I) lowered the coupling efficiency. Bulky side-chains high in the pocket force substrates to bind closer to the heme and retard uncoupling. However, such substitutions close to the heme tend to increase uncoupling because there is less space in the vicinity of the heme and substrates may have to bind further away. The rate and coupling of alkylben-zene oxidation were increased by the T185L and T185F mutations but there was little correlation between side-chain properties and activity or product selectivity . 2-Phenylethanol is the more desirable fine chemical product from ethylben-zene oxidation but this compound was not formed. The enzymes instead attacked the more reactive benzylic C-H bond to give 1-phenylethanol, suggesting that the substrate is sufficiently mobile within the active site to allow the enzyme to sample different parts of the substrate and the most reactive C-H bond is attacked.
Styrene and naphthalene were oxidized by the Y96F mutant at rates 25 and 140 times those of wild-type P450cam, respectively (Scheme 4.4) [49, 50]. Both wildtype P450cam and the Y96F mutant oxidized tetralin stereospecifically to the 1-K alcohol (Scheme 4.4) [51, 52]. Styrene oxide, 1-naphthol and 1-tetralol are useful intermediates in synthesis. Both the NADH turnover rate and coupling were greatly improved for the Y96F mutant, demonstrating the value of increased active site hydrophobicity for the oxidation of hydrophobic molecules. The increased heme spin state shift induced by styrene binding to the Y96F mutant was accompanied by a more positive heme reduction potential . The activity of this mutant was further increased by incorporating the V247L mutation to improve the enzyme-substrate match . The Y96A mutation gave less dramatic increases because the larger active site resulted in substrate mobility and hence more uncoupling. However the Y96A mutant oxidized diphenylmethane, diphe-nylether and diphenylamine at the para position while the wild-type and Y96F mutant were inactive (Scheme 4.4) , demonstrating the utility of creating space in the active site for larger substrates. P450cam mutants with increased diphenylmethane oxidation activity over the wild-type enzyme have been identified by screening for colored indole oxidation products following site-saturation mutagenesis at Y96 and F98 . These active mutants also contain mutations that increase the active site volume by replacing Y96 and F98 with residues that have smaller side-chains.
The planarity and rigidity of polycyclic aromatic hydrocarbons (Scheme 4.5) impinge significantly on coupling efficiencies for their oxidation by P450cam . For example, the NADH turnover rates for pyrene oxidation by the Y96A and Y96F mutants were 3 times those of wild type, although the couplings were generally low.
Polychlorinated aromatics are hazardous environmental contaminants because of their lipid solubility, toxicity, and potential carcinogenicity. Many of these compounds are resistant to biodegradation, and the recalcitrance increases with the degree of chlorination. Chlorinated phenols, being more reactive, are more readily degraded. Biodegradation systems can therefore be generated by genetically augmenting chlorophenol-degrading organisms with P450 enzymes that oxidize polychlorinated benzenes to phenols (Scheme 4.6) . P450cam was engineered to oxidize polychlorinated benzenes by using the Y96F mutation to promote hydrophobic molecule binding and the F87W, Y96W, and V247L mutations to force the benzenes closer to the heme. The mutants were up to three orders of magnitude more active than the wild-type enzyme for oxidizing di-, tri-, and tetrachlorobenzenes, with coupling efficiencies as high as 95%. The turnover activity and coupling for pentachlorobenzene (PeCB) oxidation were low .
The crystal structure of the F87W/Y96F/V247L mutant complexed with 1,3,5-trichlorobenzene (TCB) showed multiple substrate binding orientations with different angles between the benzene ring and the porphyrin . The most productive orientation had TCB almost parallel to, and in van der Waals contact with, the heme (Fig. 4.3a). The structure showed that PeCB could not bind in this orientation because of steric interference with Leu244. The L244A mutation was therefore introduced and the F87W/Y96F/L244A/V247L mutant oxidized PeCB 45 times faster than the parent mutant, while hexachlorobenzene oxidation activity was increased 200-fold. Both compounds are oxidized to pentachlorophe-nol. The crystal structure of this mutant complexed with PeCB (Fig. 4.3b, unpublished results) strikingly confirmed that the substrate was bound in the parallel orientation, with a chlorine being accommodated in the space created by the L244A mutation. Introduction of a gene cassette encoding this mutant and the electron transfer proteins putidaredoxin reductase and putidaredoxin into a pen-tachlorophenol-degrading Sphingobium strain generated a novel microorganism capable of degrading hexachlorobenzene .
Substitutions at F87 in P450BM-3 have been used to alter the side-chain volume at this residue and increase the activity and alter the enantioselectivity of propyl-benzene (Scheme 4.7) and 3-chlorostyrene oxidation. In the case of 3-chlorosty-rene the F87G mutant gave (R)-chlorostyrene oxide with up to 96% ee . Experiments with styrene revealed that other positions in and outside of the binding site also affected the absolute configuration of the product and could even
1,3,5-Trichlorobenzene Scheme 4.6
1,3,5-Trichlorobenzene Scheme 4.6
complex shows that the L244 side-chain will sterically hinder the binding of PeCB. Introduction of the L244A mutation generates space to accommodate the extra chlorines in PeCB and results in a 45-fold increase in PeCB oxidation activity.
lead to inversion of the enantioselectivity. The enantiomeric excess of the reaction product of styrene oxidation ranged from 58% ee (S)-styrene oxide (A74E/F87V/P386S) via 49% (R)-styrene oxide (F87A) and 65% (R)-styrene oxide (A74G/F87V/L188Q) to 92% (R)-styrene oxide for the F87G mutant (Scheme 4.7) .
S-Styrene oxide Styrene
The A74G/F87V/L188Q mutant oxidized naphthalene and three-ring polyaro-matics such as fluorene and acenaphthene three orders of magnitude faster than the wild-type enzyme . A different approach combined the hydrophobic R47L/ Y51F mutations at the entrance to the substrate channel with the F87A, I263A, and A264G mutations in the active site to give similar orders of magnitude increases in activity for oxidizing polyaromatics with up to four rings . The A74G/F87V/L188Q mutant showed faster NADPH turnover but the R47/Y51F series of mutants had higher couplings, such that the overall substrate oxidation activities were comparable. Polychlorinated aromatic oxidation by CYP102 enzymes has been less studied, but the A74G/F87V/L188Q mutant has been shown to oxidize polychlorinated dioxins . The activity pattern was similar to that for chlorinated benzene oxidation by P450cam (i.e. dioxins with up to three chlorines were oxidized but 2,3,7,8-tetrachlorodibenzodioxin (TCDD) was not oxidized). The enzyme attacked at unsubstituted positions and NIH shifts involving chlorines were observed.
Was this article helpful?