Distinct taxonomic and functional profiles of high Arctic and alpine permafrost-affected soil microbiomes

Abstract

Background: Global warming is affecting all cold environments, including the European Alps and Arctic regions. Here, permafrost may be considered a unique ecosystem harboring a distinct microbiome. The frequent freeze-thaw cycles occurring in permafrost-affected soils, and mainly in the seasonally active top layers, modify microbial communities and consequently ecosystem processes. Although taxonomic responses of the microbiomes in permafrost-affected soils have been widely documented, studies about how the microbial genetic potential, especially pathways involved in C and N cycling, changes between active-layer soils and permafrost soils are rare. Here, we used shotgun metagenomics to analyze the microbial and functional diversity and the metabolic potential of permafrost-affected soil collected from an alpine site (Val Lavirun, Engadin area, Switzerland) and a High Arctic site (Station Nord, Villum Research Station, Greenland). The main goal was to discover the key genes abundant in the active-layer and permafrost soils, with the purpose to highlight the potential role of the functional genes found.

Results: We observed differences between the alpine and High Arctic sites in alpha- and beta-diversity, and in EggNOG, CAZy, and NCyc datasets. In the High Arctic site, the metagenome in permafrost soil had an overrepresentation (relative to that in active-layer soil) of genes involved in lipid transport by fatty acid desaturate and ABC transporters, i.e. genes that are useful in preventing microorganisms from freezing by increasing membrane fluidity, and genes involved in cell defense mechanisms. The majority of CAZy and NCyc genes were overrepresented in permafrost soils relative to active-layer soils in both localities, with genes involved in the degradation of carbon substrates and in the degradation of N compounds indicating high microbial activity in permafrost in response to climate warming.

Conclusions: Our study on the functional characteristics of permafrost microbiomes underlines the remarkably high functional gene diversity of the High Arctic and temperate mountain permafrost, including a broad range of C- and N-cycling genes, and multiple survival and energetic metabolisms. Their metabolic versatility in using organic materials from ancient soils undergoing microbial degradation determine organic matter decomposition and greenhouse gas emissions upon permafrost thawing. Attention to their functional genes is therefore essential to predict potential soil-climate feedbacks to the future warmer climate.

Keywords: Active layer; European alps; Functionality; High Arctic; Metagenome; Permafrost.

Fig. 1

Sampling sites (A) Val Lavirun (LAV), eastern Swiss Alps. (B) Photograph of the permafrost regions below the Piz Lavirun (C) Villum Research Station (VRS), northern Greenland. (D) Photograph of the arctic tundra

Fig. 2

Soil bacterial and fungal diversity at the phylum and genus levels, according to amplicon sequencing (A) Bacterial phyla, (B) fungal phyla, (C) bacterial genera, (D) fungal genera. LAV, Val Lavirun; VRS, Villum Research Station BSC = biological soil crust; aL = active-layer soils. pF = permafrost soils. A = A horizon enriched with organic matter and thus darker than the underlying B horizon. B = B horizon. Bc = transitional layer with characteristics of both B and C horizons, with B horizon characteristics dominant; partly affected by cryoturbation, as manifested by a disrupted and broken horizon

Fig. 3

Non-metric multidimensional scaling (NMDS) plot of fungal communities of two sampling sites and soil profiles The envfit function was used to show the chemical and physical parameters affecting fungal communities. LAV, Val Lavirun; VRS, Villum Research Station Three replicates per soil depth are shown. BSC = biological soil crust; aL = active-layer soils. pF = permafrost soils. A = A horizon enriched with organic matter and thus darker than the underlying B horizon. B = B horizon. Bc = transitional layer with characteristics of both B and C horizons, with B horizon characteristics dominant; partly affected by cryoturbation, as manifested by a disrupted and broken horizon. GWC = gravimetric water content; TOC = total organic carbon; TON = total organic nitrogen; C14 = radiocarbon ¹⁴ C

Fig. 4

Differences in COG functional categories between the two sampling sites and soil profiles LAV, Val Lavirun; VRS, Villum Research Station; aL, active layer; pF, permafrost The log₂-fold change (LFC) value pF vs. aL in LAV (A) and in VRS (B) is the log₂ of (gene abundance of pF/gene abundance of aL). The LFC value LAV vs. VRS is the log₂ of (gene abundance of LAV_aL/gene abundance of VRS_aL) for aL (C) and the log₂ of (gene abundance of LAV_pF/gene abundance of VRS_pF) for pF (D). For each pairwise comparison, only genes with a base mean > 20, a relative abundance > 0.005, p < 0.01, and LFC > 2.5 were selected. A list of all selected genes with their relative abundances, eggNOG classifications, and associated processes are provided in Additional file 1, Tables S5-8

Fig. 5

Differences in COG genes abundance between the two sampling sites and soil profiles LAV, Val Lavirun; VRS, Villum Research Station; aL, active layer; pF, permafrost The log₂-fold change (LFC) value pF vs. aL in LAV (A) and in VRS (B) is the log₂ of (gene abundance of pF/gene abundance of aL). The LFC value LAV vs. VRS is the log₂ of (gene abundance of LAV_aL/gene abundance of VRS_aL) for aL (C) and the log₂ of (gene abundance of LAV_pF/gene abundance of VRS_pF) for pF (D). For each pairwise comparison, only genes with a base mean > 20, a relative abundance > 0.005, p < 0.01, and LFC > 2.5 were selected. A list of all selected genes with their relative abundances, eggNOG classifications and associated processes are provided in Additional file 1, Tables S9-12

Fig. 6

Differences in CAZy genes between the two sampling sites and soil profiles LAV, Val Lavirun; VRS, Villum Research Station; aL, active layer; pF, permafrost The log₂-fold change (LFC) value pF vs. aL in LAV (A) and in VRS (B) is the log₂ of (gene abundance of pF/gene abundance of aL). The LFC value LAV vs. VRS is the log₂ of (gene abundance of LAV_aL/gene abundance of VRS_aL) for aL (C) and the log₂ of (gene abundance of LAV_pF/gene abundance of VRS_pF) for pF (D). For each pairwise comparison, only genes with a base mean > 50, p < 0.01, and LFC > 2.5 were selected; the genes are sorted by their functions in the depolymerization of carbon substrates. Glycosyl transferases are not shown. A list of all selected genes with their relative abundances, CAZy classifications and associated enzymatic activities is provided in Additional file 1, Tables S13-16. [AA], auxiliary activities; [CBM], carbohydrate-binding modules; [CE], carbohydrate esterases; [GH], glycoside hydrolases; [PL], polysaccharide lyases

Fig. 7

Differences in N-cycling gene families between the two sampling sites and soil profiles LAV, Val Lavirun; VRS, Villum Research Station; aL, active layer; pF, permafrost The log₂-fold change (LFC) value pF vs. aL in LAV (A) and in VRS (B) is the log₂ of (gene abundance of pF/gene abundance of aL). The LFC value LAV vs. VRS is the log₂ of (gene abundance of LAV_aL/gene abundance of VRS_aL) for aL (C) and the log₂ of (gene abundance of LAV_pF/gene abundance of VRS_pF) for pF (D). For each pairwise comparison, only genes with a base mean > 50 and p < 0.01 were selected. A list of all selected genes with their relative abundances, NCyc classifications, and associated processes is provided in Additional file 1, Tables S17-20