Ecogenomics of the SAR11 clade

Abstract

Members of the SAR11 clade, despite their high abundance, are often poorly represented by metagenome-assembled genomes. This fact has hampered our knowledge about their ecology and genetic diversity. Here we examined 175 SAR11 genomes, including 47 new single-amplified genomes. The presence of the first genomes associated with subclade IV suggests that, in the same way as subclade V, they might be outside the proposed Pelagibacterales order. An expanded phylogenomic classification together with patterns of metagenomic recruitment at a global scale have allowed us to define new ecogenomic units of classification (genomospecies), appearing at different, and sometimes restricted, metagenomic data sets. We detected greater microdiversity across the water column at a single location than in samples collected from similar depth across the global ocean, suggesting little influence of biogeography. In addition, pangenome analysis revealed that the flexible genome was essential to shape genomospecies distribution. In one genomospecies preferentially found within the Mediterranean, a set of genes involved in phosphonate utilization was detected. While another, with a more cosmopolitan distribution, was unique in having an aerobic purine degradation pathway. Together, these results provide a glimpse of the enormous genomic diversity within this clade at a finer resolution than the currently defined clades.

Figure 1

Maximum likelihood phylogenomic tree of all SAR11 genomes available to date, together with those retrieved in this study (SAG‐MED, highlighted in red). Coloured dots next to the genome identifier indicate the origin of the genome, that is MAG (red), SAG (blue) or pure culture (yellow). Branches of the tree were coloured according to the previous classification (Giovannoni, 2017). Sequences were grouped within Subclades and genomospecies (black or white squares). Only subclades with at least 3 SAGs or pure culture genomes are shown.

Figure 2

A. Occurrence plot of SAR11 genomes within Tara stations. The horizontal axis stands for the number of metagenomic samples one genome recruited at least three RPKG (presence), while the y‐axis represents the average relative abundance (RPKG, semi‐log scale) of one genome within the samples where it is present. Genomes are coloured according to their subclade. Dashed lines correspond to delimited genomospecies. B. Relative abundance of SAR11 genomospecies in surface Tara Ocean metagenomes. For each sample, those genomospecies with less than 5% of relative abundance are included in ‘Others’. C. Linear Pearson's correlation coefficients between SAR11 genomospecies abundances. Only comparisons with p‐value ≤ 0.05 are shown.

Figure 3

A. Latitudinal transect following the GEOTRACES GA02 cruise. The first five boxes show the recruitment of different genomospecies by depth and latitude. Genomospecies that showed the same distribution were combined in the same figure. The last box shows the temperature profile. B. NMDS analysis of genomospecies according to Bray‐Curtis distance between GEOTRACES GA02 cruise samples. Only fitting statistically significant (p‐value ≤ 0.05) physicochemical parameters are shown. NMDS stress value: 0.116. C. Recruitment plots of one representative genome of IA.1/IV (found in cold latitudes) and IB.1/III (found in warm latitudes) in two metagenomes.

Figure 4

A. Pangenome analysis between genomes of genomospecies gMED and gWID. Functional characterization of the pangenome using SEED Subsystems database for the number of differential genes between genomospecies gMED and gWID. B. Boxplot indicating the recruitment values (x‐axis, log scale) of the phosphonate cluster at two depths [surface (SRF) and DCM] among Tara regions: IO, Indian Ocean; MS, Mediterranean Sea; NAO and SAO; North and South Atlantic Ocean; NPO and SPO, North and South Pacific Ocean; RS, Red Sea and SO, Southern Ocean. C. Longitudinal transect of the Mediterranean Sea showing the recruitment values of the phosphonate cluster.

Figure 5

Correlograms of SNPs position conservation along the reference genome of gMED (AG‐414‐O11) across the different metagenomes (microdiversity). Linear Pearson's correlation coefficients range from −1 to 1, taking 1 as the reference value of a metagenome against itself. (A) Horizontal and (B) vertical microdiversity through the water column in a single location. Map shows the location of the samples used in the analyses.