Engineering of heme-dependent monooxygenases towards heterocycle conversion

de Almeida Santos, Gustavo; Schwaneberg, Ulrich (Thesis advisor); Blank, Lars M. (Thesis advisor)

Aachen (2020, 2021)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2020

Abstract

Aromatic oxygen and nitrogen-containing heterocycles (O- and N- heterocycles) are significantly abundant in nature as they are an important class of bioactive molecules and are involved in a variety of fundamental biological functions. Benzo-1,4-dioxane and indole are O- and N- heterocycles respectively, and their regioselective hydroxylation can produce small to macrocyclic building blocks with great importance for drugs with antimicrobial, antigrastic, spasmolytic, antipsychotic, anxiolytic and hepatoprotective. Additionaly, these heterocycles can be used for the production of pesticides and dyes. However, the traditional chemical syntheses of said derivatives requires at least one of the following: rare and costly catalyst, heating, active cooling and multi-step reactions. The usage of P450 monooxygeanses can be used to hydroxylate these heterocycles in a more efficient and sustainable way. Three different monooxygenases were selected to investigate the structural determinants of activity over the benzo-1,4-dioxane and indole heterocycles. Cytochrome P450 BM3 monooxygenase from Bacillus megaterium, P450 Cand_1 monooxygenase from Pseudomonas sp. 19-rlim and P450 Cand_10 from Phenylobacterium zucineum. Each P450 was subjected to different engineering strategies: P450 BM3 to epPCR, P450 Cand_1 to sequence saturation method (SeSaM) and P450 Cand_10 to site-saturation-mutagenesis (SSM) to improve their activity towards benzo-1,4-dioxane and/or indole. Product-specific screening systems were developed and applied to efficiently identify improved variants in all generated libraries. Significantly improved variants from P450 Cand_10 and P450 Cand_1 were not found, P450 Cand_1 exhibited poor and inconsistent expression in 2.2 mL deep-well plates and for P450 Cand_10 the determinants for activity over the tested substrates are likely to be present outside the active site, in the active site tunnel access and/or protein shell. Regarding P450 BM3, the phenol detection 4-AAP assay, as well as the developed multiplex capillary electrophoresis (MP-CE), enabled simultaneous detection and quantification of the target product and side products in a 96-well format. Employing both methods allowed for the identification of position R255, which when substituted by leucine in the P450 BM3 WT leads to ≈140-fold increase in the initial oxidation rate of nicotinamide adenine dinucleotide phosphate (NADPH) (WT: 8.3 ± 1.3 min−1; R255L: 1168 ± 163 min−1), ≈21-fold increase in total turnover number (TTN) (WT: 40 ± 3; R255L: 860 ± 15), and, ≈2.9-fold increase in coupling efficiency (WT: 8.8 ± 0.1%; R255L: 25.7 ± 1.0%). Computational analysis revealed that when R255 is substituted by leucine (substitution distant from the heme-cofactor) the previously existing salt-bridge between R255 and D217 (in WT) ceases to exist, introducing flexibility into the I-helix and rearranging the heme, thus allowing a more efficient hydroxylation. This improvement was not limited to hydroxylation of benzo-1,4-dioxane and ≈20-fold improvement in conversion of the O-heterocycles, phthalan, isochroman, 2,3-dihydrobenzofuran, benzofuran, and dibenzofuran was found. The improvement observed in variant R255L provides useful routes to produce pharmaceutical precursors in a selective and environmentally friendly way via late-stage hydroxylation.