The 19-hydroxylation of cortexolone by the fungus Pellicularia filamentosa : a thesis presented in partial fulfilment for the degree of Doctor of Philosophy in Biotechnology at Massey University

Abstract

The microbiological 19-hydroxylation of steroids was studied to investigate the feasibility of a microbial process to produce 19-hydroxysteroids. These are important precursors of the valuable 19-norsteroids. At present, in industrial processes, 19-hydroxylation is performed by chemical synthesis. Fungi, selected from the genera Pestalotia (25 strains) and Pellicularia (5 strains), were screened for their steroid-hydroxylating activities. Thus, hydroxylation of the substrates progesterone, 4-androstene-3,17-dione, and cortexolone (17α ,21-dihydroxy-4-pregnene-3,20-dione) was studied. Of the organisms tested, only Pellicularia filamentosa f.sp. microsclerotia IFO 6298 and Pellicularia filamentosa f.sp. sasakii IFO 5254 were found to perform 19-hydroxylation. Thus, both fungi could produce 19-hydroxycortexolone from cortexolone, with the former organism the more active in this respect-Hydrocortisone (11β-hydroxycortexolone) was also produced by both organisms. Neither organism, however, could similarly hydroxylate progesterone nor 4-androstehe-3,17-dione. With these substrates, products other than the 19-hydroxylated derivatives were formed. Thus a degree of substrate specificity was recognised for steroidal-19-hydroxylation by these fungi. None of the Pestalotia species tested could 19-hydroxylate any of the three substrates, despite claims in the literature, but instead were very active in 11 α -hydroxylation. In particular, many species were able to 11 α-hydroxylate progesterone (0.5 g/l concentration) in greater than 90% (w/w) yield. Using P. filamentosa f.sp. microsclerotia IFO 6298, in batch fermentation, at a cortexolone concentration of 0.5 g/l yields of 19- and 11β-hydroxycortexolone totalled approximately 40% (w/w) of the consumed substrate. The ratio of the two products, typically, was approximately 1.2:1 (19:11β). Only small variations in this ratio were ever observed. The steroid losses which were observed did not proceed via the hydroxy products as intermediates, but via a degradation pathway, from cortexolone, parallel to the hydroxylation reactions. The 11β- and 19-hydroxylase enzyme-system of P. filamentosa f.sp. microsclerotia IFO 6298 was shown to be inducible by cortexolone. By using the protein synthesis inhibitor, cycloheximide, in fermenter culture the effects of dissolved oxygen tension (DOT) on enzyme induction and enzyme expression were separately investigated. For both hydroxylations, an optimum DOT for induction was shown at 15% of saturation, while the optimum for expression is at 30% of saturation. Thus, maximum rates of hydroxylation were achieved when induction was performed at low DOT, followed by elevation to ensure maximum expression. The effects of specific glucose consumption rate and specific growth rate were investigated using chemostat cultures, under automatic DOT control (at 30% of saturation). At a constant specific growth rate, the importance of glucose metabolism to the hydroxylation process was demonstrated. Thus, with glucose-limited cultures, decreasing specific hydroxylation rates were observed with decreasing specific glucose consumption rates, possibly as a result of the restricted availability of NADPH, which is required for hydroxylation to occur. Conversely, with nitrogen-limited cultures, it was observed that the hydroxylase system is subject to glucose repression. Thus, with high specific glucose consumption rates, cultures showed low levels of hydroxylation activity. Maximum activity was obtained at a point, which apparently represents a balance between sufficient glucose metabolism to maintain full expression of activity and a condition of excess glucose uptake resulting in repression of activity. This finding may be of considerable significance for fungal steroid-hydroxylation processes in general, since relief of repression, when it exists, could give several-fold increases in specific hydroxylation rates as observed in this study. Over the range of specific growth rates studied (0.028 h-l to 0.119 h-l) no significant effect on specific hydroxylation rates was observed.