Protein engineering of the monooxygenase P450 BM3 toward improved performance in allylic hydroxylation

  • Protein-Engineering der Monooxygenase P450 BM3 hinsichtlich verbesserter Leistung in allylischer Hydroxylierung

Gärtner, Anna; Schwaneberg, Ulrich (Thesis advisor); Elling, Lothar (Thesis advisor)

Aachen (2020)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2020


Keto-isophorone is an intermediate for the synthesis of flavors, fragrances, pharmaceuticals and nutrients like vitamin E. Common chemical conversion is achieved over a low yielding isomerization step from α- to β-isophorone. Cytochrome P450 BM3 monooxygenase from Bacillus megaterium has been identified as an alternative biological catalyst for α-isophorone oxidation. The wild-type inserted a single oxygen atom from molecular dioxygen into a C-H bond resulting mainly in 4-hydroxy-isophorone, a precursor of keto-isophorone. Semi-rational protein engineering of P450 BM3 was tackled to overcome limitations in activity, regioselectivity and coupling efficiency during allylic oxidation of α-isophorone and to generate a robust biocatalyst for industrial application. For efficient identification of beneficial variants within the generated mutant libraries, a product-specific screening system was established. Multiplex capillary electrophoresis (MP-CE) allowed simultaneous detection and quantification of the target product and side products in 96-well format. Separation of the isophorone derivates was achieved in a micellar buffer system with a low standard deviation (12%) and a broad linear detection range (0.125 mM to at least 20 mM). The new MP-CE platform enabled the identification of robust variants with a 3.5-fold higher accuracy than the frequently conducted NADPH oxidation assay. Screening of (multi-) site-saturation libraries led to the identification of the beneficial substitutions R47S/Y51W/I401M (M2), V78N, F87V, and I263M for 4-hydroxy-isophorone formation. Subsequently, variant M2 was used as a template and the three other beneficial substitutions introduced in all possible combinations. Protein engineering led to generation of variant M2 V78N, which showed 3.82-fold increased 4-hydroxy-isophorone titers, 1.27-fold improved coupling efficiency, and 21.08-fold elevated NADPH oxidation rate compared to the WT. Catalytic performance of variant M2 V78N was then compared to variant RLYFIP, which was prior published as an effective catalyst for α-isophorone oxidation. In addition, two rational engineering strategies involving the increase of hydrophobicity in a protein channel of P450 BM3 and the modulation of the electron transfer pathway have been conducted as well as the introduction of inert decoy molecules for boosting of P450 BM3 performance studied Alternative strategies did not result in improvement of 4-hydroxy-isophorone formation by P450 BM3, however, led to a gain of information about prerequisites and necessary alterations for successful performance optimization. Over-oxidation of 4-hydroxy-isophorone by P450 BM3 variants to the corresponding ketone was partially observed, however, yields were low. Thus, the capability of a set of oxidoreductases, namely alcohol dehydrogenases and laccases, towards the formation of keto-isophorone was validated. Indeed, both approaches led to the identification of one enzyme from each studied enzyme class. Laccase from Trametes versicolor showed the best performance with N-hydroxyl group mediators and reached keto-isophorone yields up to 20% while ADH-R achieved 45.5% at pH 8.5 and 25.7% at pH 7.5. In combination with P450 BM3, yields of the double oxidation process starting from α-isophorone were rather low with 2.3% (laccase) and 8.8% (ADH-R), respectively. However, the laccase mediator system might present a new combination possibility for two-step processes with P450 BM3. Catalytic analysis of ADH-R revealed crucial points for reaction optimization to obtain an efficient system for one-step one-pot synthesis of keto-isophorone with simultaneous cofactor regeneration was shown.