Taxonomic and biosynthetic diversity of the marine actinomycete Salinispora across spatial scales

Abstract

The spatial scales of bacterial taxonomic and natural product biosynthetic diversity remain poorly understood. This is especially true at the population level, where contrasts between small and large-scale biogeographical patterns are seldom reported. To address these unknowns for the marine actinomycete genus Salinispora, we sequenced the genomes of 99 strains cultured from sediments collected within a 1 m² plot (microscale strains). Ninety-six of the microscale strains were identified as S. arenicola, suggesting that this is the most abundant species in the sediments sampled. These strains were assigned to 2 of the 11 populations identified based on 99% ANI among 61 public genomes obtained from 10 global collection sites (global strains). The populations showed evidence of geographic isolation, suggesting that barriers to dispersal or ecological contingencies limit distributions across large spatial scales. An assessment of S. arenicola biosynthetic gene diversity among 157 (combined microscale and global) genomes revealed 100 gene cluster families (GCFs), of which one-third were detected in either one or all strains. Sixty-seven percent of the global GCFs were detected among the microscale strains, indicating that deep sampling from a single location recovered a large percentage of the global biosynthetic diversity. Paired genomic and metabolomic analyses of the microscale strains linked compounds to an orphan PKS-NRPS GCF, while the metabolites ikarugamycin and fridamycin E were identified for the first time from Salinispora. This study provides insight into the diversity and biosynthetic potential of Salinispora at various spatial scales while expanding the collection of natural products reported from the genus.

Importance: The marine actinomycete genus Salinispora has become a model organism for natural product discovery and to address actinomycete diversity and distributions in marine systems. While biogeographic patterns have been reported at global scales, contrasts have yet to be made with the species diversity that can be recovered from a single location. Here we sequenced the genomes of 96 S. arenicola strains cultured from marine sediments collected within a 1 m² plot and compared the diversity detected to public genomes obtained from global collection sites. The results provide evidence of geographic isolation among S. arenicola populations and biosynthetic genes that are mobilized across population boundaries. Multi-omic analyses linked compounds to their respective biosynthetic genes and revealed compounds not previously reported from the genus. This study adds to our growing understanding of Salinispora diversity and biosynthetic potential.

Keywords: Salinispora; biogeography; diversity; natural products.

Fig 1

ANI dendrogram and sources of S. arenicola strains. (A) The inner purple line demarcates 11 color-coded 99% ANI populations. Strain origin (location) is indicated by the second color-coded circle. The outer circle (black) indicates the microscale strains. (B) Geographic distributions of the 11 (99%) ANI populations. Pie graphs indicate the proportion of each population isolated from that location. The map was made using the maps package in RStudio.

Fig 2

S. arenicola biosynthetic potential. (A) Relative diversity and abundance of biosynthetic classes across 157 S. arenicola genomes. Diversity is expressed as the percentage of all GCFs in each biosynthetic class. Abundance is expressed as the percentage of all BGCs in each biosynthetic class. (B) Number of strains in which each GCF was observed. Eighteen GCFs were observed in only one strain, while another 18 were observed in all 157 strains. (C) GCF distributions between the microscale and global genomes. (D) GCF rarefaction curves for the microscale and global genomes. Average y-axis values (black or orange circles) and standard deviation are plotted. Numbers in parentheses indicate total number of genomes followed by total number of GCFs. Numbers in brackets indicate Chao1 diversity estimates.

Fig 3

S. arenicola gene cluster family (GCF) network. Each node represents a BGC and each cluster of nodes a GCF. GCFs are categorized as shared (present in both data sets), global (global-specific), or microscale (microscale-specific) and color-coded by biosynthetic class. Clusters are composed of n + 1 nodes (where n = number of BGCs). As such, singletons are represented by two nodes. GCF annotations are provided for experimentally validated Salinispora products or MIBiG BGC matches (in quotations).

Fig 4

GCF distributions in S. arenicola. Columns indicate presence (blue) of 100 GCFs across 157 S. arenicola genomes (y-axis). The three matrices describe shared, global-specific, or microscale-specific GCFs. For each matrix, the columns are arranged from left to right according to GCF abundance. Annotated GCFs are indicated by compound names and highlighted according to confidence level (blue = validated, orange = 56–91% MIBiG match, green = 12–40% MIBiG match to enediyne BGCs, and pink = select MIBiG matches with 25–32% similarity). The first column of the y-axis is colored according to geographic origin and ordered by 99% ANI populations (Fig. 1), which are numbered (microscale shown in bold red) and demarcated by gray horizontal lines. The second column delineates microscale (black) and global strains (white). See Data S2 for strain designations.

Fig 5

Paired omic analysis of microscale strains. Presence-absence table of select ions linked to validated or putative GCFs (see Data S3 for full data set). Columns represent ions (with red, yellow, and green gradients representing molecular weight from light to heavy) ordered by retention times (t_R). Compound names linked to ions are listed. Rows delineated into microscale strain populations 1 and 5 based on 99% ANI groupings. Ions and GCFs shown in this figure (colored boxes) were only observed in population 5. Pink: ion detected, Blue: GCF present.