Epothilone C

Background: Epothilone B is a natural product that stabilizes microtubules, similar to paclitaxel (Taxol); therefore, epothilone B and several derivatives have shown obvious antitumour activities. Some of these products are in clinical trials, and one (ixabepilone, BMS) is already on the market, having been approved by the FDA in 2007. The terminal step in epothilone B biosynthesis is catalysed by the cytochrome P450 enzyme EpoK (CYP167A1), which catalyses the epoxidation of the C12-C13 double bond (in epothilone C and D) to form epothilone A and B, respectively. Although redox partners from different sources support the catalytic activity of EpoK in vitro, the conversion rates are low, and these redox partners are not applied to produce epothilone B in heterologous hosts. Results: Schlegelella brevitalea DSM 7029 contains electron transport partners that efficiently support the catalytic activity of EpoK. We screened and identified one ferredoxin, Fdx_0135, by overexpressing putative ferredoxin genes in vivo and identified two ferredoxin reductases, FdR_0130 and FdR_7100, by whole-cell biotransformation of epothilone C to effectively support the catalytic activity of EpoK. In addition, we obtained strain H7029-3, with a high epothilone B yield and found that the proportion of epothilone A + B produced by this strain was 90.93%. Moreover, the whole-cell bioconversion strain 7029-10 was obtained; this strain exhibited an epothilone C conversion rate of 100% in 12 h. Further RT-qPCR experiments were performed to analyse the overexpression levels of the target genes. Gene knock-out experiments showed that the selected ferredoxin (Fdx_0135) and its reductases (FdR_0130 and FdR_7100) might participate in critical physiological processes in DSM 7029. Conclusion: Gene overexpression and whole-cell biotransformation were effective methods for identifying the electron transport partners of the P450 enzyme EpoK. In addition, we obtained an epothilone B high-yield strain and developed a robust whole-cell biotransformation system. This strain and system hold promise for the industrial production of epothilone B and its derivatives.

Epothilone A, a microtubule-stabilizing agent used as therapeutics for the treatment of cancers, was biotransformed into three metabolites using Nocardia sp. CS692 and recombinant Nocardia overexpressing a cytochrome P450 from Streptomyces venezuelae (PikC). Among three metabolites produced in the biotransformation reaction mixtures, ESI/MS2 analysis predicted two metabolites (M1 and M2) as novel hydroxylated derivatives (M1 is hydroxylated at the C-8 position and M2 is hydroxylated at C-10 position), each with an opened-epoxide ring in their structure. Interestingly, metabolite M3 lacks an epoxide ring and is known as deoxyepothilone A, which is also called epothilone C. Metabolite M1 was produced only in PikC overexpressing strain. The endogenous enzymes of Nocardia sp. catalyzed hydroxylation of epothilone A to produce metabolite M2 and removed epoxide ring to produce metabolite M3. All the metabolites were identified based on UV-vis analysis and rigorous ESI/MS2 fragmentation based on epothilone A standard. The newly produced metabolites are anticipated to display novel cytotoxic effects and could be subjects of further pharmacological studies.

Background: Polyketide synthases (PKSs) include ketone synthase (KS), acyltransferase (AT) and acyl carrier protein (ACP) domains to catalyse the elongation of polyketide chains. Some PKSs also contain ketoreductase (KR), dehydratase (DH) and enoylreductase (ER) domains as modification domains. Insertion, deletion or substitution of the catalytic domains may lead to the production of novel polyketide derivatives or to the accumulation of desired products. Epothilones are 16-membered macrolides that have been used as anticancer drugs. The substrate promiscuity of the module 4 AT domain of the epothilone PKS (EPOAT4) results in production of epothilone mixtures; substitution of this domain may change the ratios of epothilones. In addition, there are two dormant domains in module 9 of the epothilone PKS. Removing these redundant domains to generate a simpler and more efficient assembly line is a desirable goal. Results: The substitution of module 4 drastically diminished the activity of epothilone PKS. However, with careful design of the KS-AT linker and the post-AT linker, replacing EPOAT4 with EPOAT2, EPOAT6, EPOAT7 or EPOAT8 (specifically incorporating methylmalonyl-CoA (MMCoA)) significantly increased the ratio of epothilone D (4) to epothilone C (3) (the highest ratio of 4:3 = 4.6:1), whereas the ratio of 4:3 in the parental strain Schlegelella brevitalea 104-1 was 1.4:1. We also obtained three strains by swapping EPOAT4 with EPOAT3, EPOAT5, or EPOAT9, which specifically incorporate malonyl-CoA (MCoA). These strains produced only epothilone C, and the yield was increased by a factor of 1.8 compared to that of parental strain 104-1. Furthermore, mutations of five residues in the AT domain identified Ser310 as the critical factor for MMCoA recognition in EPOAT4. Then, the mutation of His308 to valine or tyrosine combined with the mutation of Phe310 to serine further altered the product ratios. At the same time, we successfully deleted the inactive module 9 DH and ER domains and fused the ΨKR domain with the KR domain through an ~ 25-residue linker to generate a productive and simplified epothilone PKS. Conclusions: These results suggested that the substitution and deletion of catalytic domains effectively produces desirable compounds and that selection of the linkers between domains is crucial for maintaining intact PKS catalytic activity.