3.1 Identification of a CYP catalyzing HLDOA
Previously, from the G. lucidum genome, we narrowed down to 82 candidate genes which may be responsible for GA biosynthesis (Wang et al., 2018). In order to extend the biosynthetic step from HLDOA to a further GA, here we re-screened those genes by using the previously developed yeast chassis. Both plasmids pRS426-CYPx (Ura3 selection marker) and pRS425-CYP5150L8-iGLCPR (Leu selection marker) were co-transformed into S. cerevisiae YL-T3 to yield various engineered strains (Fig. 1a, Table S3). Considering that the biosynthetic genes of secondary metabolites may be arranged in cluster, we first examined whether the CYP genes located in the same gene cluster as CYP5150L8, i.e., CYP5150L6, CYP5150L9, CYP5150K5, CYP5150J5 and CYP5150J10, could catalyze HLDOA. The results showed that they could not catalyze, as no new peak appeared (Fig. 1b), suggesting that the GA biosynthetic genes might not be simply clustered together.
From the fermentation samples of the CYP5139G1 and CYP5150L8 (L8) co-overexpressed strain, generation of a new peak in the HPLC chromatogram was found (Fig. 1c). The peak with retention time (Rt) of 17.5 min could not be detected in either CYP5150L8 or CYP5139G1 single gene expressed strain, not to mention in the control with void plasmid.
By LC-MS analysis, compound A was detected with m/z at 455.3513 and 437.3410 (Fig. 1d), which was consistent with the molecular weight of [C30H48O4-H2O+H]+455.3525 and [C30H48O4-2H2O+H]+437.3420, respectively. Compared to HLDOA (C30H48O3), it was supposed that hydroxylation of HLDOA to compound A might have occurred, and cyp5139G1 was considered responsible for the hydroxylation of HLDOA.