Genome Analysis of Streptomyces nojiriensis JCM 3382 and Distribution of Gene Clusters for Three Antibiotics and an Azasugar across the Genus Streptomyces

Abstract

Streptomyces spp. have been major contributors of novel natural products that are used in many application areas. We found that the nojirimycin (NJ) producer JCM 3382 has antimicrobial activity against Staphylococcus aureus via cellular degradation. Genome analysis revealed 30 biosynthetic gene clusters, including those responsible for producing antibiotics, including an azasugar NJ. In-depth MS/MS analysis confirmed the production of 1-deoxynojirimycin (DNJ) along with NJ. In addition, the production of tambromycins, setomimycin, and linearmycins was verified by spectroscopic analyses, including LC-MS and NMR. The distribution of the clusters of genes coding for antibiotics in 2061 Streptomyces genomes suggested potential producers of tambromycin, setomimycin, and linearmycin. For a DNJ gene cluster, homologs of gabT1 and gutB1 were commonly found; however, yktC1 was identified in only 112 genomes. The presence of several types of clusters suggests that different strains may produce different types of azasugars. Chemical-profile-inspired comparative genome analysis may facilitate a more accurate assessment of the biosynthetic potential to produce secondary metabolites.

Keywords: Streptomyces nojiriensis JCM 3382; antibiotics; biosynthetic gene cluster; comparative genomics; nojirimycin; secondary metabolism.

Figure 1

A circular diagram representing the genome of S. nojiriensis JCM 3382. From the outermost track to the center: (i) predicted genes (blue/red for forward/reverse strand and green for BGCs), (ii) 30 predicted BGC regions in blocks, (iii) GC content (blue/red for above/below average), and (iv) GC skew (green: >0 and yellow: <0).

Figure 2

HPLC chromatogram at 310 and 420 nm of JCM 3382 culture extract and metabolites with representative UV absorption spectra (A). Biosynthetic gene clusters (regions 6, 10, and 21) activated in this study are shown with their products (B).

Figure 3

Biosynthetic pathways of DNJ and DNJ intermediates identified using MS/MS. DNJ biosynthetic pathway (A) and product ion mass spectra in positive ion mode of molecular ions corresponding to NJ (B), NJ dehydrate (C), and DNJ (D) from the culture extract.

Figure 4

Distribution of proteins in gene clusters producing tambromycin (A), setomimycin (B), and linearmycin (C) in 2062 Streptomyces genome sequences. Protein sequences from each cluster were searched using TBLATN with an E-value cutoff of 1e−5. The dark navy-blue color shows the maximum bit score for each protein. The track outside the ideogram shows the sum of bit scores from BLASTN searches when the nucleotide sequences of each cluster were used as the query, with an E-value cutoff of 1e−20. The darker the color of a cell, the more homologous the corresponding sequence is. The strain JCM 3382 is marked by a red dot in each diagram.

Figure 5

Distribution of the gabT1-yktC1-gutB1 gene cluster over a phylogenomic tree of 2062 Streptomyces genomes. (A) Schematic diagrams describing the gene structure. gabT1, yktC1, and gutB1 are indicated by red, blue, and green, respectively. Gray indicates a gene other than these three. (B) Distribution of the gene cluster with the phylogenomic tree shown as a cladogram. A total of 110 genomes are indicated by colors in their terminal nodes: genomes containing (i) a canonical gene cluster (red), (ii) a canonical and a disordered cluster (pink), and (iii) a disordered cluster (green). The strain JCM 3382 is indicated with a star at the terminal node. A clade shaded in gray indicates 53 closely related genomes having a canonical cluster. The sequence homologies of GabT1, YktC1, and GutB1 are shown in color gradients. The more intense the color, the more homologous genes found in the corresponding genome. Homology was measured using the formula −log₁₀ (E-value). An E-value of 0 was replaced with 1e−200 to avoid infinity. The homology of the nucleotide sequences of the whole gene cluster was calculated as the sum of significant hits (E-value <1e−20) and is shown as a histogram. From the outermost track: the whole gene cluster, GutB1, YktC1, and GabT1.