Multiple copies of the oxytetracycline gene cluster in selected Streptomyces rimosus strains can provide significantly increased titers

BACKGROUND: Natural products are a valuable source of biologically active compounds that have applications in medicine and agriculture. One disadvantage with natural products is the slow, time-consuming strain improvement regimes that are necessary to ensure sufficient quantities of target compounds for commercial production. Although great efforts have been invested in strain selection methods, many of these technologies have not been improved in decades, which might pose a serious threat to the economic and industrial viability of such important bioprocesses. RESULTS: In recent years, introduction of extra copies of an entire biosynthetic pathway that encodes a target product in a single microbial host has become a technically feasible approach. However, this often results in minor to moderate increases in target titers. Strain stability and process reproducibility are the other critical factors in the industrial setting. Industrial Streptomyces rimosus strains for production of oxytetracycline are one of the most economically efficient strains ever developed, and thus these represent a very good industrial case. To evaluate the applicability of amplification of an entire gene cluster in a single host strain, we developed and evaluated various gene tools to introduce multiple copies of the entire oxytetracycline gene cluster into three different Streptomyces rimosus strains: wild-type, and medium and high oxytetracycline-producing strains. We evaluated the production levels of these engineered S. rimosus strains with extra copies of the oxytetracycline gene cluster and their stability, and the oxytetracycline gene cluster expression profiles; we also identified the chromosomal integration sites. CONCLUSIONS: This study shows that stable and reproducible increases in target secondary metabolite titers can be achieved in wild-type and in high oxytetracycline-producing strains, which always reflects the metabolic background of each independent S. rimosus strain. Although this approach is technically very demanding and requires systematic effort, when combined with modern strain selection methods, it might constitute a very valuable approach in industrial process development.