Metaproteomics reveals functional partitioning and vegetational variation among permafrost-affected Arctic soil bacterial communities

Abstract

Microbial activity in Arctic soils controls the cycling of significant stores of organic carbon and nutrients. We studied in situ processes in Alaskan soils using original metaproteomic methods in order to relate important heterotrophic functions to microbial taxa and to understand the microbial response to Arctic greening. Major bacterial groups show strong metabolic specialization in organic topsoils. α-/β-/γ-Proteobacteria specialized in the acquisition of small, soluble compounds, whereas Acidobacteria, Actinobacteria, and other detritosphere groups specialized in the degradation of plant-derived polymers. α-/β-/γ-Proteobacteria dominated the expression of transporters for common root exudates and limiting nitrogenous compounds, supporting an ecological model of dependence upon plants for carbon and competition with plants for nitrogen. Detritosphere groups specialized in distinct substrates, with Acidobacteria producing the most enzymes for hemicellulose depolymerization. Acidobacteria was the most active group across the three plant ecotypes sampled-the largely nonvascular, lower biomass intertussock and the largely vascular, higher biomass tussock and shrub. Functional partitioning among bacterial groups was stable between plant ecotypes, but certain functions associated with α-/β-/γ-Proteobacteria were more strongly expressed in higher biomass ecotypes. We show that refined metaproteomic approaches can elucidate soil microbial ecology as well as biogeochemical trajectories of major carbon stocks. IMPORTANCE The Arctic is warming twice as fast as the rest of the planet, and Arctic soils currently store twice as much carbon as the entire atmosphere-two facts that make understanding how Arctic soil microbial communities are responding to climate change particularly urgent. Greening of vegetation cover across the Arctic landscape is one of the most prominent climate-driven shifts in Arctic terrestrial ecology, with potentially profound effects on biogeochemical cycling by the soil microbiome. Here we use metaproteomics to document microbial metabolic functions that drive soil carbon and nutrient cycling processes in an Arctic tundra landscape. We identify functional roles among bacterial taxonomic groups that are largely stable across vegetation types, with certain functions strongly expressed by rhizosphere groups reflecting a community metabolic response to greening.

Keywords: arctic; greening; metabolism; metaproteomics; microbial ecology; permafrost; soil; warming.

Fig 1

Cell growth-related Functional Group Φ_bin values (heatmap) and average overall relative abundances (bars) across organic soil samples. Columns are ordered by cluster assignment, then by Functional Group category (bar color), then by overall relative abundance (NSAF). Φ_bin values are column-normalized to between 0 and 1 based on Φ_bin,max for that column.

Fig 2

Carbon-related Functional Group Φ_bin values (heatmap) and average overall relative abundances (bars), across organic soil samples. Columns are ordered by cluster assignment, then by Functional Group category (bar color), then by overall relative abundance (NSAF). Φ_bin values are column-normalized to between 0 and 1 based on Φ_bin,max for that column.

Fig 3

Nutrient-related Functional Group Φ_bin values (heatmap) and average overall relative abundances (bars), across organic soil samples. Columns are ordered by cluster assignment, then by Functional Group category (bar color), then by overall relative abundance (NSAF). Φ_bin values are column-normalized to between 0 and 1 based on Φ_bin,max for that column.

Fig 4

Linear discriminant analyses for organic soil samples between plant ecotypes of (A) Φ_bin values, with each data point representing one of 12 taxonomic bins in a given ecotype and (B) Φ_{Functional Group} values, with each data point representing one of 141 Functional Groups in a given ecotype. The percentage is the proportion of separation due to the discriminant function. ( C ) Principal component analyses of Functional Group Φ_bin values, showing principal components 1 and 2 (percentage indicates the proportion of variance explained by each PC), with lines connecting points representing the same taxonomic group in different plant ecotypes.

Fig 5

Summary of resource partitioning by major taxonomic groups in organic soils, as determined from Φ_bin data.