Distinct and rich assemblages of giant viruses in Arctic and Antarctic lakes

Abstract

Giant viruses (GVs) are key players in ecosystem functioning, biogeochemistry, and eukaryotic genome evolution. GV diversity and abundance in aquatic systems can exceed that of prokaryotes, but their diversity and ecology in lakes, especially polar ones, remain poorly understood. We conducted a comprehensive survey and meta-analysis of GV diversity across 20 lakes, spanning polar to temperate regions, combining our extensive lake metagenome database from the Canadian Arctic and subarctic with publicly available datasets. Leveraging a novel GV genome identification tool, we identified 3304 GV metagenome-assembled genomes, revealing lakes as untapped GV reservoirs. Phylogenomic analysis highlighted their dispersion across all Nucleocytoviricota orders. Strong GV population endemism emerged between lakes from similar regions and biomes (Antarctic and Arctic), but a polar/temperate barrier in lacustrine GV populations and differences in their gene content could be observed. Our study establishes a robust genomic reference for future investigations into lacustrine GV ecology in fast changing polar environments.

Keywords: Last Ice Area margin; Nucleocytoviricota; lacustrine polar virology; meta-analysis; metagenomics.

Figure 1

Metagenomic composition of lacustrine giant virus diversity from pole to pole; (A) geographic location of all 20 lakes used in our meta-analysis; lakes from the same region are identified with a similar color palette, and groupings are highlighted by dashed ellipses; Group indicators (“F,” “M,” “HS,” and “E”) following lake names stand for freshwater, meromictic, hypersaline, and epishelf, respectively; (B) distribution and contribution of giant virus orders across different group of lakes (from top to bottom: LIM lakes, Arctic/subarctic lakes, temperate lakes, and Antarctic lakes); bubble sizes are in log for visualization purposes; (C) maximum-likelihood phylogenetic tree of the GVMAGs inferred from a concatenated protein alignment of seven core GVOGs; shades indicate similar family level; cultivated viruses are indicated in red; tree annotations from inside to the outside: (1) GVMAG origin and cryo-feature (we considered the lake from which the GVMAG was initially assembled as its origin), (2) order classification, (3) GC content, and (4) GVMAG length.

Figure 2

Distribution and similarities in giant virus communities across lakes and regions; (A) NMDS ordination of lake metagenomes; lakes from the same region are identified with common shapes and a similar color palette; group indicators (“F,” “M,” “HS,” and “E”) following lake names stand for freshwater, meromictic, hypersaline, and epishelf, respectively; sample conductivity is represented with red lines on the ordination diagram; (B) horizontal bar plot showing the distribution of all assembled GVMAGs across all regions (left) and the percentage of unique and shared GVMAGs across regions (right); only percentages of shared GVMAGs >0.1% between two regions are shown.

Figure 3

Structural differentiation of giant virus communities across and between LIM lakes and Arctic/subarctic lakes; location of LIM lakes and Arctic/subarctic lakes; the map shows the numbers of total, unique, and shared GVMAGs across the lakes; letters within the circles correspond to the initials of the lake names; the star indicates SAS2A as the subarctic outgroup (see Methods); the LIA is represented by a shading on the north coast of the archipelago and Ellesmere Island; the LIA is shown in a schematic style here to emphasize its proximity to the LIM lakes.

Figure 4

Giant virus community composition across LIM lakes and putative eukaryotic hosts; (A) relative abundance of giant virus orders and families in each LIM lake; (B) relative abundance of eukaryotic clades in each LIM lake; (C) network of predicted strength of association between giant virus families and eukaryotic clades in LIM lakes; the thickness of interaction links between groups is related to the decreasing Gini index; a thicker link stands for a higher score, which indicates greater importance of the variable in the model, and stronger predicted associative strength (see Methods for more details), circles represent giant virus families, and squares represent eukaryotic groups; the color legend is the same as in A and B.

Figure 5

Giant virus phylogeny and genetic makeup dissimilarities between polar and temperate lakes; (A) NMDS ordination of phylogenetic distances (Unifrac) between GVMAGs from all samples; lakes from the same region are identified with a common shape and color; (B) UpSet plot showing the numbers of unique polar protein clusters and shared between polar and temperate regions; bar plot on the right shows proportions of proteins clusters unique to a polar region, shared with other polar regions and shared with temperate and polar regions.