Strategies to improve the efficiency of oxidation reactions of the enzyme Cytochrome P450Bm3

Abstract

Selective and efficient C-H bond oxidation using conventional organic methods is a challenging task. As biocatalysts, cytochrome P450 enzymes provide an enzymatic route for the oxidation of unreactive carbon-hydrogen (C-H) bonds. CYP102A1 (P450Bm3) from the Bacillus megaterium bacterium is one of the most studied members of the P450 superfamily because of its solubility and high activity towards fatty acids (C12-15). P450Bm3 is a self-sufficient enzyme and only needs NADPH (as a source of electrons) and O2 for oxidation reactions. The primary subject of this thesis is exploring the regio- and stereoselectivity of four rate accelerating variants of P450Bm3: KT2 (A191T/N239H/I259V/A276T/L353I), R19 (R47L/Y51F/H171L/Q307H/N319Y), RLYFIP (R47L/Y51F/I401P) and RLYFAP (R47L/Y51F/A330P), in the oxidation of substituted benzenes that contain different alkyl and vinyl groups. Whereas wild type (WT) P450Bm3 showed low activity for the oxidation of these substrates, the mutated variants showed high product formation rates and maintained the regioselectivity of WT P450Bm3. However, the RLYFAP variant, which contains an alanine330 (A330) to proline mutation (A330P) resulted in different regioselectivity. For example, oxidation of 3-ethyltoluene with RLYFAP occurred at the benzene ring and generated a phenol product (2-ethyl-4-methylphenol, 92 %) in contrast to the WT, R19, KT2 and RLYFIP variants, which oxidised the alkyl side chain and the aromatic ring with equal yields. The effects of polyfluorinated fatty acids (PFCs) on the activity and selectivity of P450Bm3 variants were also studied. Decoy molecules are dummy substrates with a similar structure to fatty acids. Therefore, the carboxylate group interacts with key residues of P450Bm3 (Arg47 and Tyr51) and places the enzyme in a catalytically ready state while leaving enough space for non-native species to get close to the haem centre. The addition of decoy molecules to the variants increased the range of the substrates along with providing higher productivity. The regioselectivity of the enzyme that catalysed oxidations was maintained and the stereoselectivity slightly improved. For example, the addition of PFC10 to the WT Bm3 in the oxidation of ethylbenzene enhanced the enantiomeric excess from 48 % to 68% (R). 4-Hydroxyisophorone is a flavour and fragrance compound and the selective oxidation of isophorone to (R)-4-hydroxyisophorone with RLYFIP, R19, WT and GVQ (A74G/F87V/L188Q) variants was studied. GVQ and RLYFIP showed a 280-fold increase in oxidation activity over WT. Addition of decoy molecules further increased oxidation activity with the majority of the variants. However, decoy molecules reduced the activity of the GVQ variant for the reaction. The product distributions with the variants were almost identical and (R)-4-hydroxyisophorone was the main product with > 90 % ee. The decoy molecules (PFC3-PFC10) were employed in the hydroxylation of cycloalkanes (C5-C10) to investigate how the combination of different sizes of substrates and decoy molecules changes the oxidation activity of the WT P450Bm3 and its variants. The highest product formation was observed with cyclooctane. The combination of decoy molecules with WT P450Bm3 and other variants increased productivity. By increasing the size of the decoy molecule, higher activity was achieved, and PFC10 was the best decoy molecules in all variants, excluding the GVQ variant, which showed lower activity with decoy molecules. Second generation decoy molecules, which are a combination of first-generation decoy molecules (PFC9) and amino acids were also employed in the hydroxylation of cycloalkanes as well and significantly improved the activity of P450Bm3WT in the oxidation of smaller substrates (cyclopentane and cyclohexane) to the corresponding cyclic alcohol (cyclopentanol and cyclohexanol). It is hypothesised that the incorporation of amino acid into the decoy molecules better-allowed interaction with the active site of the P450Bm3. A single mutation of threonine 268, which is involved in dioxygen activation, to glutamic acid (T268E) was made to the haem domain of P450Bm3 WT. This mutant was found to convert WT Bm3 into an H2O2-dependent variant (Bm3T268E or Bm3TE). The activity of Bm3TE in the oxidation of styrene, ethylbenzene and methylthiobenzene in the presence of H2O2 was tested and compared with the holoenzymes (WT and R19 variants). While no activity was observed with WT and R19 variants, Bm3TE successfully oxidised the substrates above. The addition of the second generation decoy molecules to holoenzymes enhanced the oxidation activity, but they did not significantly improve the activity of Bm3TE. The regioselectivity was maintained with the Bm3TE variant although stereoselectivity of ethylbenzene oxidation changed and was more in favour of the (S) enantiomer. Finally, encapsulation of Bm3TE in metal-organic frameworks (MOFs) was studied. Since Zeolitic imidazolate frameworks (ZIFs) can be synthesised under mild conditions, ZIF-8 and ZIF-90 crystals were the candidates MOFs for in-situ encapsulation of Bm3TE to increase thermal and chemical stability. It was found that Bm3TE lost almost all its activity in the process of encapsulation. Therefore, ZIF-8 and ZIF-90 crystals were pre-synthesised to study immobilisation of Bm3TE on the surface. But again, no activity was observed. Given that organic ligands and MOF surface were a challenge during the immobilisation of Bm3TE, bacterial compartments were used for encapsulation of P450Bm3. Two encapsulin compartments from Thermotoga maritima (Tm) and Myxococcus xanthus (Mx) were employed for capsid formation around the P450Bm3 R19 variant. Bm3 R19 was used as a cargo protein after being tagged with short targeting peptides (TPs) of the natural cargo of Tm and Mx encapsulins. Encapsulated P450Bm3 R19 in Tm encapsulin showed almost no activity but low level of product formation were observed with the P450Bm3 R19 in the Mx encapsulin. Future studies will include modification of the encapsulin plasmids and conditions of protein expression for more efficient packaging of P450Bm3 R19 and other P450 systems.