Pleiotropic control of secondary metabolism and morphological development by KsbC, a butyrolactone autoregulator receptor homologue in Kitasatospora setae

Abstract

The γ-butyrolactone autoregulator signaling cascades have been shown to control secondary metabolism and/or morphological development among many Streptomyces species. However, the conservation and variation of the regulatory systems among actinomycetes remain to be clarified. The genome sequence of Kitasatospora setae, which also belongs to the family Streptomycetaceae containing the genus Streptomyces, has revealed the presence of three homologues of the autoregulator receptor: KsbA, which has previously been confirmed to be involved only in secondary metabolism; KsbB; and KsbC. We describe here the characterization of ksbC, whose regulatory cluster closely resembles the Streptomyces virginiae barA locus responsible for the autoregulator signaling cascade. Deletion of the gene ksbC resulted in lowered production of bafilomycin and a defect of aerial mycelium formation, together with the early and enhanced production of a novel β-carboline alkaloid named kitasetaline. A putative kitasetaline biosynthetic gene cluster was identified, and its expression in a heterologous host led to the production of kitasetaline together with JBIR-133, the production of which is also detected in the ksbC disruptant, and JBIR-134 as novel β-carboline alkaloids, indicating that these genes were biosynthetic genes for β-carboline alkaloid and thus are the first such genes to be discovered in bacteria.

Fig 1

Chemical structures of metabolites of K. setae (A) and organization of the ksbC locus in K. setae (B). (A) Structures of bafilomycin B1 (compound 1) and kitasetaline (compound 2). (B) Gray arrows indicate putative regulatory genes, and white arrows indicate putative autoregulator biosynthetic genes.

Fig 2

Bafilomycin production in the ksbC disruptant. The amount of bafilomycin was the total amount of bafilomycin derivatives (bafilomycin A1, B1, and C1). Error bars represent standard deviations from triplicate experiments. WT, wild-type strain; ΔksbC/−, ksbC disruptant; ΔksbC/pSET152, ΔksbC strain carrying pSET152; ΔksbC/ksbC, ksbC-complemented ΔksbC strain; WT/pSET152, the wild-type strain carrying pSET152. (A) Effect of ksbC disruption on bafilomycin production. (B) Complementation of the ksbC disruptant and effect of the pSET152 integration into the genome of K. setae on bafilomycin production.

Fig 3

Kitasetaline production in the ksbC disruptant (A) and the wild-type strain (B). HPLC chromatograms of methanol extracts from each strain cultivated at the indicated time. mAU, milliabsorbance units at 276 nm. The peaks of kitasetaline and JBIR-133 are indicated as a black arrow and a gray arrow.

Fig 4

Chemical structures of JBIR-133 (A; compound 3) and JBIR-134 (B; compound 4). Lower column shows COSY and HMBC analysis. Bold lines represent ¹H-¹H correlations, and arrows indicate key HMBC correlations (¹H ↔ ¹³C).

Fig 5

Heterologous expression of the β-carboline gene cluster in S. avermitilis SUKA22 and determination of the minimal gene cluster. (A) Schematic representation of the fragments used in this experiment. I, pLT437 containing genes from kse_70630 to kse_70700; II, pLT438 containing genes from kse_70630 to kse_70690; III, pLT439 containing genes from kse_70630 to kse_70670; IV, pLT440 containing genes from kse_70660 to kse_70700. The genes indicated by arrows are completely included in each fragment. (B) HPLC analyses (276 nm) of the metabolite profiles of S. avermitilis SUKA22 carrying pLT437 (I), pLT438 (II), pLT439 (III), pLT440 (IV), and pKU492Aaac(3)IV as an empty vector, respectively. JBIR-133 and kitasetaline are eluted at 26.7 and 28.0 min, respectively.

Fig 6

Binding of KsbC to the intergenic region between kse_44590 and ksbS4 genes. (A) Location of probes used for the gel shift assay. The probes E-1 to E-3 used in the present study are shown, and their lengths are indicated on the left. The location of the putative 26-bp ARE sequence, situated 108 bp upstream of the ksbS4 gene, is indicated by a dark gray box. (B) Gel shift assay for the binding of purified His-tagged KsbC (rKsbC) to probes containing a plausible KsbC-binding site. (C) Comparison of the putative KsbC-binding sequence (ksbS4-ARE) with the ARE sequences located upstream of the afsA family genes (upper panel) and a logo-representation for conserved nucleotides using computational methods (lower panel). The genes in the list are afsA of S. griseus, farX of S. lavendulae FRI-5, and scbA of S. coelicolor A3(2). The consensus nucleotides in the ARE sequences are highlighted by black boxes. The logo representation was created using WebLogo analysis (http://weblogo.berkeley.edu/logo.cgi) based on the 26-bp ARE sequences. The relative sizes of the letters indicate their likelihood at the particular position. (D) Gene expression analysis of the ksbS4 and ksbA genes by qRT-PCR. WT/pSET152, the wild-type strain carrying pSET152; ΔksbC/pSET152, the ΔksbC strain carrying pSET152. The fold change is relative to the expression of each gene in the wild-type strain carrying pSET152. Error bars, SD from triplicate experiments.