A new Bacillus megaterium whole-cell catalyst for the hydroxylation of the pentacyclic triterpene 11-keto-ß-boswellic acid (KBA) based on a recombinant cytochrome P450 system.

The use of cytochromes P450 for the regio- and stereoselective hydroxylation of non-activated carbon atoms in biotechnological applications reflects an efficient and cost-effective alternative in comparison to classical organic chemistry. The prokaryotic cytochrome P450 CYP106A2 from Bacillus megaterium ATCC 13368 hydroxylates a variety of 3-oxo-Δ4 steroids and recently it was identified to carry out a one-step regioselective allylic hydroxylation of the diterpene abietic acid. The anti-inflammatory pentacyclic triterpene 11-keto-ß-boswellic acid (KBA) was found to be a further substrate of CYP106A2, being the first report of a pentacyclic triterpene conversion by a prokaryotic P450. The reaction products were analyzed by HPLC and the corresponding kinetic parameters were investigated. Structure determination of the main product by NMR revealed a 15α-hydroxylation of this substrate. In order to overcome the inability of a recombinant P450 whole-cell system in E. coli for the uptake of acids with terpene structure, we developed for the first time an expression system for cytochromes P450 in B. megaterium (strains MS941 and ATCC 13368). Interestingly, CYP106A2 was only successfully expressed in the plasmid-less B. megaterium strain MS941 but not in ATCC13368. This recombinant system, with the co-expressed heterologous redox chain of the P450, bovine adrenodoxin reductase (AdR), and bovine adrenodoxin (Adx), was applied for the whole-cell conversion of KBA. The formation of 15α-hydroxy-KBA was increased 15-fold in comparison with the naturally CYP106A2-expressing B. megaterium strain ATCC 13368.

Identification of CYP106A2 as a regioselective allylic bacterial diterpene hydroxylase.

The cytochrome P450 monooxygenase CYP106A2 from Bacillus megaterium ATCC 13368 catalyzes hydroxylations of a variety of 3-oxo-Δ(4) -steroids such as progesterone and deoxycorticosterone (DOC), mainly in the 15ß-position. We combined a high-throughput screening and a rational approach for identifying new substrates of CYP106A2. The diterpene resin acid abietic acid was found to be a substrate and was docked into the active site of a CYP106A2 homology model to provide further inside into the structural basis of the regioselectivity of hydroxylation. The products of the hydroxylation reaction were analyzed by HPLC and the V(max) and K(m) values were calculated. The corresponding reaction products were analyzed by NMR spectroscopy and identified as 12α- and 12ß-hydroxyabietic acid. CYP106A2 was therefore identified as the first reported bacterial cytochrome P450 diterpene hydroxylase. Furthermore, an effective whole-cell catalyst for the selective allylic 12α- and 12ß-hydroxylation was applied to produce the hydroxylated products.

Towards preparative scale steroid hydroxylation with cytochrome P450 monooxygenase CYP106A2.

Cytochrome P450 monooxygenases are of outstanding interest for the synthesis of pharmaceuticals and fine chemicals, due to their ability to hydroxylate C--H bonds mainly in a stereo- and regioselective manner. CYP106A2 from Bacillus megaterium ATCC 13368, one of only a few known bacterial steroid hydroxylases, enables the oxidation of 3-keto-4-ene steroids mainly at position 15. We expressed this enzyme together with the electron-transfer partners bovine adrenodoxin and adrenodoxin reductase in Escherichia coli. Additionally an enzyme-coupled cofactor regeneration system was implemented by expressing alcohol dehydrogenase from Lactobacillus brevis. By studying the conversion of progesterone and testosterone, the bottlenecks of these P450-catalyzed hydroxylations were identified. Substrate transport into the cell and substrate solubility turned out to be crucial for the overall performance. Based on these investigations we developed a new concept for CYP106A2-catalyzed steroid hydroxylations by which the productivity of progesterone and testosterone conversion could be increased up to 18-fold to yield an absolute productivity up to 5.5 g L(-1) d(-1). Product extraction with absorber resins allowed the recovery of quantitative amounts of 15beta-OH-progesterone and 15beta-OH-testosterone and also the reuse of the biocatalyst.