Orchestrated Biosynthesis of the Secondary Metabolite Cocktails Enables the Producing Fungus to Combat Diverse Bacteria

Abstract

Fungal secondary metabolites with antibiotic activities can promote fungal adaptation to diverse environments. Besides the global regulator, individual biosynthetic gene clusters (BGCs) usually contain a pathway-specific transcription factor for the tight regulation of fungal secondary metabolism. Here, we report the chemical biology mediated by a supercluster containing three BGCs in the entomopathogenic fungus Metarhizium robertsii. These clusters are jointly controlled by an embedded transcription factor that orchestrates the collective production of four classes of chemicals: ustilaginoidin, indigotide, pseurotin, and hydroxyl-ovalicin. The ustilaginoidin BGC is implicated as a late-acquired cluster in Metarhizium to produce both the bis-naphtho-γ-pyrones and the monomeric naphtho-γ-pyrone glycosides (i.e., indigotides). We found that the biosynthesis of indigotides additionally requires the functions of paired methylglucosylation genes located outside the supercluster. The pseurotin/ovalicin BGCs are blended and mesosyntenically conserved to the intertwined pseurotin/fumagillin BGCs of Aspergillus fumigatus. However, the former have lost a few genes, including a polyketide synthase gene responsible for the production of a pentaene chain used for assembly with ovalicin to form fumagillin, as observed in A. fumigatus. The collective production of chemical cocktails by this supercluster was dispensable for fungal virulence against insects and could enable the fungus to combat different bacteria better than the metabolite(s) produced by an individual BGC could. Thus, our results unveil a novel strategy employed by fungi to manage chemical ecology against diverse bacteria. IMPORTANCE Fungal chemical ecology is largely mediated by the metabolite(s) produced by individual biosynthetic gene clusters (BGCs) with antibiotic activities. We report a supercluster containing three BGCs that are jointly controlled by an embedded master regulator in the insect pathogen Metarhizium robertsii. Four classes of chemicals, namely, ustilaginoidin, indigotide, pseurotin, and hydroxyl-ovalicin, are collectively produced by these three BGCs along with the contributions of tailoring enzyme genes located outside the supercluster. The production of these metabolites is not required for the fungal infection of insect hosts, but it benefits the fungus to combat diverse bacteria. The findings reveal and advocate a "the-more-the-better" strategy employed by fungi to manage effective adaptations to diverse environments.

Keywords: Metarhizium; chemical ecology; master regulator; secondary metabolism; supercluster.

FIG 1

Different chemicals putatively biosynthesized by a supercluster of M. robertsii. (A) Different classes of chemicals associated with this study. (B) Comparative structural relationship of the biosynthetic gene clusters encoded in different fungi. The gene models labeled in the same color are putatively involved in the biosynthesis of the same class of compounds. The orthologous genes between fungal species are connected with gray lines. The genes of A. fumigatus framed in red boxes have no homologues in M. robertsii. The putative functions of these genes are listed in Table S1.

FIG 2

Differential expression, transcription activation, and regulation control of UpmR in M. robertsii. (A) Differential expression of UpmR by M. robertsii after being grown in different conditions. OE-R, UpmR overexpression mutant. Samples include the conidia harvested from PDA and rice media, appressoria (AP) formed on the fly wings, hyphal body (HB) cells harvested from the wax moth hemolymph, and the mycelia harvested from the SDB, GMM, and CDB broths. Values are presented as mean ± SD. A one-way ANOVA was performed to determine the differences between samples: different capital letters labeled above columns, P < 0.01; different lowercase letters, P < 0.05. (B) Differential expressions of three core genes by the WT, UpmR deletion, and overexpression (OE-R) strains. (C) Differential expressions of three core genes by the WT, UpmR, and FapR overexpression mutants. The fungi were grown in GMM for 4 days prior to RNA extraction and gene expression analysis. In panels B and C, values are presented as mean ± SD, and two-tailed Student’s t tests were conducted to determine the differences between the WT and the individual mutants: **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant. (D) Confirmation of the UpmR transcription activation feature by yeast two-hybrid analysis. Similar to the positive control (BD_GAL4-AD_GAL4), BD_GAL4-UpmR has an activation activity to enable the yeast cells to grow on the synthetic dropout (SD) medium. Yeast cells transformed with only BD_GAL4 were used as a negative-control. The serially-diluted yeast cells were grown on SD media for 3 days. (E–G) Targeting of the three core gene promoters by UpmR via dual luciferase (LUC) activity assays in tobacco leaves. Column a represents spots co-infiltrated with the plasmids p1301-35S-Nos and pGreenII-0800-promoter-LUC. Column b represents spots co-infiltrated with the plasmids p1301-35S-UpmR-Nos and pGreenII-0800-promoter-LUC. The lower panels show the corresponding quantitative luminescence intensities (mean ± SD) of each treatment. There were three repeats for each treatment, made on different plants. The differences in signal intensity were compared using two-tailed Student's t tests: **, P < 0.01; *, P < 0.05.

FIG 3

Chemical analysis of the compounds produced by different strains of M. robertsii. (A) HPLC profiles showing the production or nonproduction of ustilaginoidin D (UD) by different strains. Purified UD was used as a standard. The fungi were grown in GMM for 7 days, and the mycelia were extracted for analysis. (B) HPLC profiles showing the production or nonproduction of indigotides by different strains. The fungi were grown in GMM broth for 9 days, and the culture filtrates were extracted for analysis. (C) HPLC profiles unveiling the genes involved in the production of pseurotins. Purified pseurotin A was used as a standard. The fungi were grown in GMM for 7 days, and the culture filtrates were extracted for analysis. (D) GC-MS analysis of yeast cells expressing the TC enzymes FmaA of A. fumigatus and mFmaA of M. robertsii. Mock yeasts were transformed with an empty vector. TIC, total ion chromatography. (E) Fragmented GC-MS spectra of the compound produced by the yeast cells expressing mFmaA. (F) LC-MS profiles confirming the functions of genes for mer-f3 production. The fungi were grown in GMM for 7 days, and the mycelia were harvested and extracted for analysis. EIC, extracted ion chromatography.

FIG 4

Cross inhibition assays of the WT and mutants of M. robertsii versus different bacteria. (A) Varied levels of the inhibition or noninhibition of different bacteria after being cocultured with the WT and mutant strains of M. robertsii. The cells of different bacteria were cocultured with fungal spores in LB broth for 12 h, and the bacterial cells were harvested for the measurement of their OD600 values (mean ± SD). There were five replicates for each treatment, and two-tailed Student’s t tests were conducted for each bacterial species between the WT and the individual mutants of M. robertsii: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (B) No obvious difference in spore germinations between the WT and the different mutants after growth in LB broth for 12 h. (C) Varied levels of spore germinations between the WT and the mutant strains of M. robertsii after challenges with different bacteria. The mixed cultures were incubated in LB broth for 12 h. A one-way ANOVA was conducted to determine the differences between strains. Bacterial species are indicated in each panel. Values are presented as mean ± SD. Within each panel, different letters are labeled above: capital letters, P < 0.01; lowercase letters, P < 0.05; ns, not significant.

FIG 5

Schematic biosynthesis of the four classes of compounds by the supercluster of M. robertsii. The PKS mUstP cluster can produce either the pigment ustilaginoidin D or the glycoside indigotides by means of the function of the MrGT1/MrMT1 genes located outside the supercluster. Either 2,3-unsaturated or 2,3-saturated naphtho-γ-pyrone can be used as the substrate of MrGT1/MrMT1. Indigotides I and J are novel compounds identified in this study. FPP, farnesyl-pyrophosphate. SAM, S-adenosyl methionine. The putative function of each gene is listed in Table S1.