Identification of a critical gene involved in the biosynthesis of the polyene macrolide lavencidin in Streptomyces lavendulae FRI-5 using the Target-AID (activation-induced cytidine deaminase) base editing technology

Abstract

Polyene macrolide antibiotics, produced mainly as secondary metabolites of streptomycetes, have distinct chemical structures and include clinically important antifungal drugs. We recently isolated the 28-membered polyene macrolide lavencidin from Streptomyces lavendulae FRI-5. Here, we identify and characterize the lavencidin biosynthetic (lad) gene cluster by combining a gene disruption system based on a base editing technology and in silico analysis. Sequence analysis of the draft genome of S. lavendulae FRI-5 revealed plausible lavencidin biosynthetic genes, which could be assigned roles in the biosynthesis of the polyketide backbone and the peripheral moiety, as well as in the regulation of lavencidin production. The introduction of a stop codon into the ladA5 polyketide synthase (PKS) gene by the base editing system resulted in a complete loss of lavencidin production, indicating that the type I modular PKS system is responsible for the biosynthesis of lavencidin.IMPORTANCEPolyene macrolide antibiotics display a unique mode of action among fungicides and exhibit potent fungicidal activity to which resistance does not readily develop. Deciphering the biosynthetic pathways of these fascinating compounds will provide chemical diversity for the development of industrially and clinically important agents. In this study, the Target-AID (activation-induced cytidine deaminase) system enabled us to identify the lad gene cluster involved in lavencidin biosynthesis, paving the way for the rational design of lavencidin derivatives with new or improved biological activity. Furthermore, this base editing system is capable of precisely and rapidly substituting the target nucleotide in several streptomycetes. Thus, our Target-AID system would be a powerful and versatile tool for the genetic engineering of streptomycetes as well as for analyzing the functions of uncharacterized genes, expanding the chemical diversity of useful bioactive compounds, and discovering novel natural products.

Keywords: Streptomyces; base editing; lavencidin; polyene macrolide antibiotics; target-AID.

Fig 1

Chemical structures of indigoidine (1), lavencidin (2), and RKGS-A2215A (3) produced by S. lavendulae FRI-5.

Fig 2

Target-AID base editing for actI-ORF2 mutagenesis in S. coelicolor A3(2). (A) Map of the Target-AID vector pLK101. ermE*P, a strong constitutive promoter; sgRNA, gRNA scaffold with a 20-nt targeting sequence; tipAp, a thiostrepton-inducible promoter; dCas9, the nuclease dCas9 (D10A and H840A) gene; PmCDA1_str, the cytidine deaminase PmCDA1 gene; UGI_str, the UGI gene; LVA_str, the protein degradation tag; tsr, a thiostrepton resistance gene; rep^pSG5, a temperature-sensitive replication origin from pSG5; oriT, the origin of the transfer of plasmid RK2; aac (3)IV, an apramycin resistance gene; ori, high-copy-number origin of replication from ColE1. (B) Sequence alignment of the actI-ORF2 locus edited by pTarget-ACT. The PAM sequence is shown in the box. The expected editing site is shown by the nucleotide number relative to the PAM sequence. The translated amino acid sequences are indicated above the corresponding nucleotide sequences. Mutated bases are shown in bold, and asterisks indicate stop codons. The numbers displayed to the left of the nucleotide sequence represent the ratio of the number of randomly selected exconjugants of stop codon-introduced clones to that of total sequenced clones. The right panel shows a representative Sanger sequencing chromatogram of the target region with the predicted C-to-T mutation (a mutated base is indicated by a black arrow). (C) Production of pigmented antibiotics in the mactI-ORF2 strains. WT, the wild-type strain; WT + pLK101, the WT carrying pLK101; mactI-ORF2, mactI-ORF2 strains. Each strain was streaked on R5⁻ agar medium and incubated for 4 days at 30°C. The plate was photographed from the bottom.

Fig 3

Indigoidine production in the mlbpA strain. WT, wild-type strain; mlbpA, mlbpA strains. The translated amino acid sequences are indicated above the corresponding nucleotide sequences. Mutated bases are shown in bold, and asterisks indicate stop codons. (A) Structural organization of LbpA and sequence alignment of the lbpA locus edited by pTarget-lbpA. The functional domains are shown: A, adenylation; Ox, oxidation; T, thiolation; TE, thioesterase. The PAM sequence is shown in the box. The expected editing site is shown by the nucleotide number relative to the PAM sequence. The translated amino acid sequences are indicated above the corresponding nucleotide sequences. A mutated base is shown in bold, and asterisks indicate stop codons. (B) Indigoidine production of the mlbpA strain grown on solid cultivation (top plate) and in liquid cultivation (bottom flask). The top panel shows a representative Sanger sequencing chromatogram of the target region with the predicted C-to-T mutation. A mutated base is indicated by a black arrow. For solid cultivation, each strain was cultivated on ISP medium two and incubated for 2 days at 28°C. The plate was photographed from the bottom. For liquid cultivation, each strain was inoculated into liquid medium B and incubated at 28°C for 7 h. IM-2-C₅ was added at the final concentration of 1.2-ng/mL medium, followed by incubation for an additional 2 h.

Fig 4

Genetic organization of the lavencidin biosynthetic gene cluster. Each arrow indicates the direction of transcription and the relative size of the gene. ORFs predicted to be involved in lavencidin biosynthesis are shaded. The proposed functions of each ORF are given here and summarized in Table 1.

Fig 5

Lavencidin production in the mladA5 strain. WT, wild-type strain; mladA5, mladA5 strain. The translated amino acid sequences are indicated above the corresponding nucleotide sequences. Mutated bases are shown in bold, and asterisks indicate stop codons. (A) Structural organization of LadA5 and sequence alignment of the ladA5 locus edited by pTarget-ladA5. The functional domains are shown: KS, ketosynthase; AT, acyltransferase; KR, ketoreductase; ACP, acyl carrier protein; TE, thioesterase. The PAM sequence is indicated by the black box, and the expected editing site is shown by the nucleotide number relative to the PAM sequence. (B) HPLC chromatograms of n-butanol extracts for analysis of lavencidin production. mAU, milliabsorbance units at 290 nm. Lavencidin (2) and RKGS-A2215A (3) were eluted at 13.7 and 13.0 min, respectively.

Fig 6

Predicted model for lavencidin biosynthesis. The circles represent enzymatic domains in the PKS polypeptide: KS, ketosynthase; AT, acyltransferase; KR, ketoreductase; ACP, acyl carrier protein; DH, dehydratase; ER, enoylreductase; TE, thioesterase. The presumed inactive KS and DH domains of the loading module and module seven are shaded in black.