The Widely Conserved ebo Cluster Is Involved in Precursor Transport to the Periplasm during Scytonemin Synthesis in Nostoc punctiforme

Abstract

Scytonemin is a dimeric indole-phenol sunscreen synthesized by some cyanobacteria under conditions of exposure to UVA radiation. While its biosynthetic pathway has been elucidated only partially, comparative genomics reveals that the scytonemin operon often contains a cluster of five highly conserved genes (ebo cluster) of unknown function that is widespread and conserved among several bacterial and algal phyla. We sought to elucidate the function of the ebo cluster in the cyanobacterium Nostoc punctiforme by constructing and analyzing in-frame deletion mutants (one for each ebo gene and one for the entire cluster). Under conditions of UVA induction, all ebo mutants were scytoneminless, and all accumulated a single compound, the scytonemin monomer, clearly implicating all ebo genes in scytonemin production. We showed that the scytonemin monomer also accumulated in an induced deletion mutant of scyE, a non-ebo scytonemin gene whose product is demonstrably targeted to the periplasm. Confocal autofluorescence microscopy revealed that the accumulation was confined to the cytoplasm in all ebo mutants but that that was not the case in the scyE deletion, with an intact ebo cluster, where the scytonemin monomer was also excreted to the periplasm. The results implicate the ebo cluster in the export of the scytonemin monomer to the periplasm for final oxidative dimerization by ScyE. By extension, the ebo gene cluster may play similar roles in metabolite translocation across many bacterial phyla. We discuss potential mechanisms for such a role on the basis of structural and phylogenetic considerations of the ebo proteins.IMPORTANCE Elucidating the biochemical and genetic basis of scytonemin constitutes an interesting challenge because of its unique structure and the unusual fact that it is partially synthesized in the periplasmic space. Our work points to the ebo gene cluster, associated with the scytonemin operon of cyanobacteria, as being responsible for the excretion of scytonemin intermediates from the cytoplasm into the periplasm during biosynthesis. Few conserved systems have been described that facilitate the membrane translocation of small molecules. Because the ebo cluster is well conserved among a large diversity of bacteria and algae and yet insights into its potential function are lacking, our findings suggest that translocation of small molecules across the plasma membrane may be its generic role across microbes.

Keywords: alkaloids; cyanobacteria; ebo genes; excretion; lipid carriers; membrane transport; periplasm; scytonemin; secondary metabolism; sunscreens.

FIG 1

Genomic organization of the scytonemin operon in N. punctiforme and other cyanobacteria, including the ebo genes of unknown function found within the scy operon of most cyanobacteria (but distally in N. punctiforme). Triangles indicate the genes whose deletion mutants were examined in this study.

FIG 2

Structures of the scytonemin monomer and scytonemin. The likely final step in scytonemin biosynthesis involves oxidative dimerization of the scytonemin monomer to yield reduced scytonemin, which undergoes facile auto-oxidation to scytonemin proper. MW, molecular weight.

FIG 3

(A) Absorbance spectra of acetone cell extracts from wild-type (solid black), scyE mutant (solid gray), and ebo mutant (dotted black) strains after UVA induction of the scytonemin operon. The wild-type strain produced scytonemin, as indicated by a large absorbance maximum at 384 nm. (B) Absorbance spectra of acetone cell extracts of individual ebo gene deletion mutants after UVA induction of the scytonemin operon, all displaying a scytoneminless phenotype.

FIG 4

Separation and characterization of a compound accumulated after UVA induction by the ΔeboC strain. (A) HPLC chromatogram of acetone extract showing production of a novel compound eluting at 7.8 min and the absence of a scytonemin peak at 8 min. This pattern was found in all ebo mutants and in the ΔscyE strain (see Fig. S1 in the supplemental material), and none of the strains produced the compound without an induction of the syctonemin operon (see Fig. S1). The inlay shows the UV-visible light (UV-Vis) absorbance spectrum (solid line) and the fluorescence emission (Em) spectrum (dotted line) of the newly accumulated compound after collection from HPLC eluent. AU, absorbance units. (B) Chromatogram of wild-type extract after UVA induction, indicating the presence of scytonemin at 8 min and the absence of the 7.8-min peak. λ = 407 nm.

FIG 5

Confocal fluorescence imaging and quantification of the scytonemin monomer accumulation in vivo. (Left) Fluorescence images of the wild-type (top), Δebo (middle), and ΔscyE (bottom) strains with emission at 410 nm (to visualize the scytonemin monomer) and 665 nm (to visualize photopigments in the cytoplasm), under conditions of induction and without induction by UVA. (Right) Fluorescence intensity quantification within cells of the wild-type and mutant strains at 410 and 665 nm under inductive and noninductive conditions, respectively (n = 10; bars indicate standard errors of the means).

FIG 6

Intracellular localization of the scytonemin monomer in induced ΔscyE and ΔeboC cells. Overlay of the 665-nm images over the 410-nm images demonstrates the localization of the scytonemin monomer in the cytoplasm of ΔeboC cells (this was the case for each and all of the ebo mutants, as can be seen in Fig. S2). In ΔscyE cells; however, the scytonemin monomer accumulates in both the cytoplasm and the periplasm.

FIG 7

Partitioning of core biosynthetic proteins of the scytonemin operon between cytoplasm and periplasm by proteomic analyses of osmotic shock lysates. Target protein relative abundance ratios to three different known periplasm-targeted proteins are plotted against the strength of lysis buffer used. Proteins localized to the cytoplasm should show an increase in ratio with buffer strength, whereas ratios of proteins partitioning preferentially to the periplasm should remain invariant or decrease. All ratios were normalized to 1 at 0.3 M sucrose for ease of graphing, and P values are included for each target protein data set (α = 0.01).