Predicting the chemical space of fungal polyketides by phylogeny-based bioinformatics analysis of polyketide synthase-nonribosomal peptide synthetase and its modification enzymes

Abstract

Fungal polyketide synthase (PKS)-nonribosomal peptide synthetase (NRPS) hybrids are key enzymes for synthesizing structurally diverse hybrid natural products (NPs) with characteristic biological activities. Predicting their chemical space is of particular importance in the field of natural product chemistry. However, the unexplored programming rule of the PKS module has prevented prediction of its chemical structure based on amino acid sequences. Here, we conducted a phylogenetic analysis of 884 PKS-NRPS hybrids and a modification enzyme analysis of the corresponding biosynthetic gene cluster, revealing a hidden relationship between its genealogy and core structures. This unexpected result allowed us to predict 18 biosynthetic gene cluster (BGC) groups producing known carbon skeletons (number of BGCs; 489) and 11 uncharacterized BGC groups (171). The limited number of carbon skeletons suggests that fungi tend to select PK skeletons for survival during their evolution. The possible involvement of a horizontal gene transfer event leading to the diverse distribution of PKS-NRPS genes among fungal species is also proposed. This study provides insight into the chemical space of fungal PKs and the distribution of their biosynthetic gene clusters.

Scheme 1

(A) Late biosynthetic steps to synthesize a β-ketoamide. (B) Chain elongation mechanism to synthesize a skipped polyene intermediate. Dotted lines show the optional pathways in the biosynthesis of most hybrid NPs.

Figure 1

(A) Phylogenetic tree of fungal PKS-NRPS hybrids; (B) summary of BGC classification by searching for the key modification enzymes. The four major clades, Ia, Ib, II, and III, include NP clades. The number of BGCs classified into each NP clade is denoted in parentheses.

Figure 2

Enlarged view of the phylogenetic tree around hybrids for synthesizing 2-pyridone containing hybrid NPs (Clade Ia). Functionally characterized hybrids are marked by red circles. The modification enzyme genes located adjacent to each hybrid gene are described in the right side of the phylogenetic tree. Hybrid genes lacking enough sequence data are highlighted in grey colour.

Scheme 2

(A) Proposed biosynthetic pathways of hybrid NPs produced by the pyridone clade_BGCs. The common biosynthetic intermediate is the 2-pyridone derivative biosynthesized by the action of expandase; (B) summary of key modification enzymes to categorize three NP clades; tenellin, leporin, and illicicolin. Detailed biosynthetic schemes are summarized in Supplementary Information.

Scheme 3

Proposed biosynthetic pathways of hybrid NPs produced by four NP clades; pyranonigrin, xyllorin, cyclopiazonic acid, and pyridone. The key modification enzymes involved in synthesizing each core are denoted in boldface. Detailed biosynthetic schemes are summarized in Supplementary Information.

Figure 3

Phylogenetic tree of DAases. Two major DA clades, I and II, are further divided into subclades, Iα, Iβ, IIα, IIβ-1, and IIβ-2, which correlate with the function of the characterized DAases. The accession numbers of the DAases used in this phylogenetic analysis are summarized in Table S5.

Scheme 4

Proposed biosynthetic pathways of hybrid NPs produced by BGCs of four NP clades; burnettramic acid, Sch210972, equisetin, and fusaridione. Detailed biosynthetic schemes are summarized in Supplementary Information.

Scheme 5

Proposed biosynthetic pathways of hybrid NPs produced by BGCs of NP clades of (A) pyrrolinone containing hybrid NPs and (B) tetronic acid derivatives. Asterisk indicates tentative BGC names for chaetoglobosin (chgg) and flavipucine (fla). Detailed biosynthetic schemes are summarized in Supplementary Information.

Scheme 6

Two alternative modifications of the β-ketoamide intermediate leading to two different cores, I and II.