Heterologous expression of the atypical tetracycline chelocardin reveals the full set of genes required for its biosynthesis.

Background: Chelocardin (CHD) exhibits a broad-spectrum antibiotic activity and showed promising results in a small phase II clinical study conducted on patients with urinary tract infections. Importantly, CHD was shown to be active also against tetracycline-resistant Gram-negative pathogens, which is gaining even more importance in today's antibiotic crisis. We have demonstrated that modifications of CHD through genetic engineering of its producer, the actinomycete Amycolatopsis sulphurea, are not only possible but yielded even more potent antibiotics than CHD itself, like 2-carboxamido-2-deacetyl-chelocardin (CD-CHD), which is currently in preclinical evaluation. A. sulphurea is difficult to genetically manipulate and therefore manipulation of the chd biosynthetic gene cluster in a genetically amenable heterologous host would be of high importance for further drug-discovery efforts.
Results: We report heterologous expression of the CHD biosynthetic gene cluster in the model organism Streptomyces albus del14 strain. Unexpectedly, we found that the originally defined CHD gene cluster fails to provide all genes required for CHD formation, including an additional cyclase and two regulatory genes. Overexpression of the putative pathway-specific streptomyces antibiotic regulatory protein chdB in A. sulphurea resulted in an increase of both, CHD and CD-CHD production. Applying a metabolic-engineering approach, it was also possible to generate the potent CHD analogue, CD-CHD in S. albus. Finally, an additional yield increase was achieved in S. albus del14 by in-trans overexpression of the chdR exporter gene, which provides resistance to CHD and CDCHD.
Conclusions: We identified previously unknown genes in the CHD cluster, which were shown to be essential for chelocardin biosynthesis by expression of the full biosynthetic gene cluster in S. albus as heterologous host. When comparing to oxytetracycline biosynthesis, we observed that the CHD gene cluster contains additional enzymes not found in gene clusters encoding the biosynthesis of typical tetracyclines (such as oxytetracycline). This finding probably explains the different chemistries and modes of action, which make CHD/CD-CHD valuable lead structures for clinical candidates. Even though the CHD genes are derived from a rare actinomycete A. sulphurea, the yield of CHD in the heterologous host was very good. The corrected nucleotide sequence of the CHD gene cluster now contains all gene products required for the production of CHD in a genetically amenable heterologous host, thus opening new possibilities towards production of novel and potent tetracycline analogues with a new mode of action.