Multiplex Base-Editing Enables Combinatorial Epigenetic Regulation for Genome Mining of Fungal Natural Products

Abstract

Genome mining of cryptic natural products (NPs) remains challenging, especially in filamentous fungi, owing to their complex genetic regulation. Increasing evidence indicates that several epigenetic modifications often act cooperatively to control fungal gene transcription, yet the ability to predictably manipulate multiple genes simultaneously is still largely limited. Here, we developed a multiplex base-editing (MBE) platform that significantly improves the capability and throughput of fungal genome manipulation, leading to the simultaneous inactivation of up to eight genes using a single transformation. We then employed MBE to inactivate three negative epigenetic regulators combinatorially in Aspergillus nidulans, enabling the activation of eight cryptic gene clusters compared to the wild-type strains. A group of novel NPs harboring unique cichorine and polyamine hybrid chemical scaffolds were identified, which were not reported previously. We envision that our scalable and efficient MBE platform can be readily applied in other filamentous fungi for the genome mining of novel NPs, providing a powerful approach for the exploitation of fungal chemical diversity.

Figure 1.

Design and validation of CBE system for premature stop codon introduction in A. nidulans. (a) C-to-T conversions of CBE and the representation of the possible stop codon introductions. nCas9 (Cas9 nuclease nickase) is an RNA-guided endonuclease that catalyzes site specific nicking of a single strand of double stranded DNA. A uracil DNA glycosylase inhibitor domain (UGI) was fused with the deaminase domain to improve the CBE efficiency. (b) Biosynthetic pathway of the yellow compound YWA1 and its green derivative. (c) Five protospacers were selected from yA or wA genes. The protospacer adjacent motif (PAM) sequences are shown in red. (d) C-to-T conversion of CBE on five editing sites. Results presented as mean ± s.d. of three biological replicates. (e) Phenotypes of the CBE mediated gene inactivation on yA and wA genes in A. nidulans.

Figure 2.

CBE mediated multiplex genome editing in A. nidulans. (a) The tRNA-gRNA arrays for multiple sgRNA generation. The five proto-spacers used here are listed in Figure 1c. (b) The cartoon representation of the tRNA processing system for multiple sgRNA generation. The tRNA-gRNA array can be processed by endogenous RNase P and RNase Z to release functional sgRNAs. (c) CBE mediated C-to-T conversion for editing 3-, 4-, and 5-loci simultaneously in A. nidulans. Results presented as the mean of three biological replicates. (d) Statistical analysis of normalized C-to-T editing frequencies shown in c. (e) The editing results for eight-gene inactivation by MBE in A. nidulans TN02A7. The eight protospacers used here are listed in Table S4.

Figure 3.

CBE mediated multiplex genome editing in A. oryzae and A. flavus. (a) C-to-T conversion rates at three target sites in A. oryzae. (b) C-to-T conversion rate on three target sites in A. flavus. (c) C-to-T conversion rates at four target sites in T. atroroseus. (d) The editing results for eight-gene inactivation by MBE in A. oryzae. The eight protospacers used here are listed in Table S4. Results presented as mean ± s.d. of three biological replicates.

Figure 4.

Metabolic profile analyses of the A. nidulans mutants. (a) Phenotypes of some A. nidulans mutants. (b) UV traces (254 nm) of combinational inactivation of CclA, ClrD, and HdaA in A. nidulans LO8030. (c) Extract ion chromatogram (EIC) traces of the LO8030 mutants at m/z 556.1. (d) UV traces (254 nm) of combinatorial inactivation of CclA, ClrD, and HdaA in A. nidulans TN02A7. Metabolites were extracted using small scale extraction (see Methods). The identities of the compounds were verified by NMR, UV, and/or HRMS (Figures S15–S17 and Scheme S1). All metabolic profiles were repeatable and based on analysis of at least three different colonies.

Figure 5.

The identification of new compounds and their proposed biosynthetic pathway. (a) Gene cluster for 6, 16, and 17 biosynthesis and verification by gene knockout. (b) Identification of novel NPs from histone modifier mutants. The absolute configuration of the chiral center on 17 was determined by comparing the experiment spectra with the calculated model molecules (Figure S22). The rotation value of 3 is nearly zero, indicated that it is a racemate. (c) Proposed pathway for the biosynthesis of 6 and 16. Detailed pathway from 11 to 31 can be found in Figure S23. Gene annotations: cicA, transporter; cicB, adenylate-forming reductase; cicC, oxidoreductase; cicD, regulatory protein; cicE, O-methyltransferase; cicF, polyketide synthase; cicG, function unknown; cicH, cytochrome P₄₅₀.