Microbial phytosterol degradation is accompanied by the formation of steroid pathway intermediates, which are potential precursors in the synthesis of bioactive steroids. Degradation of these steroid intermediates is initiated by ⌬ 1 -dehydrogenation of the steroid ring structure. Characterization of a 2.9-kb DNA fragment of Rhodococcus erythropolis SQ1 revealed an open reading frame (kstD) showing similarity with known 3-ketosteroid ⌬ 1 -dehydrogenase genes. Heterologous expression of kstD yielded 3-ketosteroid ⌬ 1 -dehydrogenase (KSTD) activity under the control of the lac promoter in Escherichia coli. Targeted disruption of the kstD gene in R. erythropolis SQ1 was achieved, resulting in loss of more than 99% of the KSTD activity. However, growth on the steroid substrate 4-androstene-3,17-dione or 9␣-hydroxy-4-androstene-3,17-dione was not abolished by the kstD gene disruption. Bioconversion of phytosterols was also not blocked at the level of ⌬ 1 -dehydrogenation in the kstD mutant strain, since no accumulation of steroid pathway intermediates was observed. Thus, inactivation of kstD is not sufficient for inactivation of the ⌬ 1 -dehydrogenase activity. Native polyacrylamide gel electrophoresis of cell extracts stained for KSTD activity showed that R. erythropolis SQ1 in fact harbors two activity bands, one of which is absent in the kstD mutant strain.Rhodococcus species are well known for their catabolic potential (5, 40). Several Rhodococcus species degrade natural phytosterols. Microbial phytosterol degradation proceeds via the formation of steroids as pathway intermediates (16,21,22), i.e., 4-androstene-3,17-dione, 1,4-androstadiene-3,17-dione, and 9␣-hydroxy-4-androstene-3,17-dione (Fig.

Summary9a-Hydroxylation of 4-androstene-3,17-dione (AD) and 1,4-androstadiene-3,17-dione (ADD) is catalysed by 3-ketosteroid 9a-hydroxylase (KSH), a key enzyme in microbial steroid catabolism. Very limited knowledge is presently available on the KSH enzyme. Here, we report for the first time the identification and molecular characterization of genes encoding KSH activity. The kshA and kshB genes, encoding KSH in Rhodococcus erythropolis strain SQ1, were cloned by functional complementation of mutant strains blocked in AD(D) 9a-hydroxylation. Analysis of the deduced amino acid sequences of kshA and kshB showed that they contain domains typically conserved in class IA terminal oxygenases and class IA oxygenase reductases respectively. By definition, class IA oxygenases are made up of two components, thus classifying the KSH enzyme system in R. erythropolis strain SQ1 as a two-component class IA monooxygenase composed of KshA and KshB. Unmarked in frame gene deletion mutants of parent strain R. erythropolis SQ1, designated strains RG2 (kshA mutant) and RG4 (kshB mutant), were unable to grow on steroid substrates AD(D), whereas growth on 9a-hydroxy-4-androstene-3,17-dione (9OHAD) was not affected. Incubation of these mutant strains with AD resulted in the accumulation of ADD (30-50% conversion), confirming the involvement of KshA and KshB in AD(D) 9a-hydroxylation. Strain RG4 was also impaired in sterol degradation, suggesting a dual role for KshB in both sterol and steroid degradation.

This paper reports the first method for the construction of unmarked gene deletion mutants in the genus Rhodococcus. Unmarked deletion of the kstD gene, encoding 3‐ketosteroid Δ1‐dehydrogenase (KSTD1) in Rhodococcus erythropolis SQ1, was achieved using the sacB counter‐selection system. Conjugative mobilization of the mutagenic plasmid from Escherichia coli S17‐1 to R. erythropolis strain SQ1 was used to avoid its random genomic integration. The kstD gene deletion mutant, designated strain RG1, still possessed about 10% of the KSTD enzyme activity of wild‐type and was not affected in its ability to grow on the steroid substrates 4‐androstene‐3,17‐dione (AD) and 9α‐hydroxy‐4‐androstene‐3,17‐dione (9OHAD). Biochemical evidence subsequently was obtained for the presence of a second KSTD enzyme (KSTD2) in R. erythropolis SQ1. UV mutants of strain RG1 unable to grow on AD were isolated. One of these mutants, strain RG1‐UV29, had lost all KSTD enzyme activity and was also unable to grow on 9OHAD. It stoichiometrically converted AD into 9OHAD in concentrations as high as 20 g l−1. The two KSTD enzymes apparently both function in AD and 9OHAD catabolism. These isoenzymes have been inactivated in strain RG1 (KSTD1 negative) and strain RG1‐UV29 (KSTD1 and KSTD2 negative), respectively.