Media and strain studies for the scaled production of cis-enone resorcylic acid lactones as feedstocks for semisynthesis

Abstract

Resorcylic acid lactones (RALs) with a cis-enone moiety, represented by hypothemycin (1) and (5Z)-7-oxozeaenol (2), are fungal secondary metabolites with irreversible inhibitory activity against protein kinases, with particularly selective activity for inhibition of TAK1 (transforming growth factor beta-activated kinase 1). Gram-scale quantities of these compounds were needed as feedstock for semi-synthesizing RAL-analogues in a step-economical fashion. To do so, this study had three primary goals: identifying fungi that biosynthesized 1 and 2, enhancing their production by optimizing the fermentation conditions on the lab scale, and developing straight forward purification processes. After evaluating 536 fungal extracts via an in-house dereplication protocol, three strains were identified as producing cis-enone RALs (i.e., MSX78495, MSX63935, MSX45109). Screening these fungal strains on three grain-based media revealed enhanced production of 1 by strain MSX78495 on oatmeal medium, while rice medium increased the biosynthesis of 2 by strain MSX63935. Furthermore, the purification processes were improved, moving away from HPLC purification to utilizing two to four cycles of resuspension and centrifugation in small volumes of organic solvents, generating gram-scale quantities of these metabolites readily. In addition, studying the chemistry profiles of strains MSX78495 and MSX63935 resulted in the isolation of ten other RALs (3-12), two radicinin analogues (13-14), and six benzopyranones (15-20), with 19 and 20 being newly described chlorinated benzopyranones.

Fig. 1

Structures of hypothemycin (1) and (5Z)-7-oxozeaenol (2)

Fig. 2

a, b show the absolute amounts of hypothemycin (1) and (5Z)-7-oxozeaenol (2), respectively, across cultures grown on rice, oatmeal, and Cheerios media. An asterisk Indicates significantly higher productions of hypothemycin (1) by strain MSX78495 in panel (a) or (5Z)-7-oxozeaenol (2) by strain MSX63935 in panel (b), compared to other fungal strains (p < 0.05) as demonstrated in Tables S1 and S2. Data are presented as mean ± SD. One of the three biological replicates of strain MSX63935 on rice medium did not show optimal growth, which caused the large error bars in (b) for the absolute amount of 2 from strain MSX63935 grown on rice

Fig. 3

Illustration of the isolation and purification protocol of hypothemycin (1), showing how 1 can be generated both from a column fraction and via centrifugation. Over a gram of 1 was generated from nine flasks: three fermented on rice, three fermented on oatmeal, and three fermented on Cheerios. We predict that the yield would be even higher by ~20% if all nine flasks were grown on oatmeal

Fig. 4

Illustration of the isolation and purification protocol of (5Z)-7-oxozeaenol (2), showing how 2 can be generated both from a column fraction and via centrifugation. As 760 mg of 2 were generated from two flasks fermented on rice, we predict that over a gram of 2 could be generated with this procedure from three flasks of strain MSX63935 grown on rice

Fig. 5

Fungal secondary metabolites produced by strains MSX78495 and MSX63935