Leave no stone unturned: individually adapted xerotolerant Thaumarchaeota sheltered below the boulders of the Atacama Desert hyperarid core

Abstract

Background: The hyperarid core of the Atacama Desert is an extremely harsh environment thought to be colonized by only a few heterotrophic bacterial species. Current concepts for understanding this extreme ecosystem are mainly based on the diversity of these few species, yet a substantial area of the Atacama Desert hyperarid topsoil is covered by expansive boulder accumulations, whose underlying microbiomes have not been investigated so far. With the hypothesis that these sheltered soils harbor uniquely adapted microbiomes, we compared metagenomes and geochemistry between soils below and beside boulders across three distantly located boulder accumulations in the Atacama Desert hyperarid core.

Results: Genome-resolved metagenomics of eleven samples revealed substantially different microbial communities in soils below and beside boulders, despite the presence of shared species. Archaea were found in significantly higher relative abundance below the boulders across all samples within distances of up to 205 km. These key taxa belong to a novel genus of ammonia-oxidizing Thaumarchaeota, Candidatus Nitrosodeserticola. We resolved eight mid-to-high quality genomes of this genus and used comparative genomics to analyze its pangenome and site-specific adaptations. Ca. Nitrosodeserticola genomes contain genes for ammonia oxidation, the 3-hydroxypropionate/4-hydroxybutyrate carbon fixation pathway, and acetate utilization indicating a chemolithoautotrophic and mixotrophic lifestyle. They also possess the capacity for tolerating extreme environmental conditions as highlighted by the presence of genes against oxidative stress and DNA damage. Site-specific adaptations of the genomes included the presence of additional genes for heavy metal transporters, multiple types of ATP synthases, and divergent genes for aquaporins.

Conclusion: We provide the first genomic characterization of hyperarid soil microbiomes below the boulders in the Atacama Desert, and report abundant and highly adapted Thaumarchaeaota with ammonia oxidation and carbon fixation potential. Ca. Nitrosodeserticola genomes provide the first metabolic and physiological insight into a thaumarchaeal lineage found in globally distributed terrestrial habitats characterized by various environmental stresses. We consequently expand not only the known genetic repertoire of Thaumarchaeota but also the diversity and microbiome functioning in hyperarid ecosystems. Video Abstract.

Keywords: Archaea; Atacama; Hyperaridity; Soil microbiome; Xerotolerance.

Fig. 1

Sampling location and soil geochemistry. a Location of three sampling sites and their abbreviations in parentheses. Locations of previously studied boulder fields are mapped and their references are shown in the legend. Top-right corner maps the arid and hyperarid climate regions across South America [11]. b Non-metric multidimensional scaling (NMDS) ordination (stress = 0.024) of anion and cation concentrations in each soil sample. Different colors represent different sampling sites (green = L, red = M, yellow = Y). Filled vs unfilled data points correspond to the sample type information (below (B) or beside boulder (C)). Blue vectors represent fitted ion species onto the ordination with adjusted p-value < 0.05. c An example drone image used to map individual boulders at high resolution; raw data collected by Sager et al. [33] and reanalyzed in this study. Image frame corresponds to 15 x 15 m. d Corresponding mapping of individual boulders imaged in c, same scale. e Satellite image distinguishing “densely” and “loosely” packed boulder accumulations

Fig. 2

Phylogenetic tree of 30 most abundant taxa (rpS3 clusters) out of 147 and their normalized abundances across all samples. Filled stars represent the number of rpS3 sequences in the cluster that were successfully binned in mid-to-high quality genomes and gray bars indicate the average iRep values calculated for the mid-to-high quality genomes. Strongly supported branches as described in the “Materials and Methods” section are indicated with black dots

Fig. 3

Metabolic potential prediction across samples and mid-to-high quality genomes. a Relative abundances of chemoautolithotrophic marker genes predicted using METABOLIC for each sample. b Presence (blue) and absence (white) of chemoautolithotropic marker genes in mid-to-high quality genomes. Genomes are clustered based on taxa and the number of genomes in each cluster is shown in parentheses in the row names

Fig. 4

Phylogenomic placement of ABT genomes using 37 housekeeping single-copy genes. a Phylogenetic tree of 298 NCBI genomes annotated as Thaumarchaeota and eight ABT genomes. Aigarchaeota were identified and used as the outgroup. Black, brown and blue ranges distinguish whether organisms are ammonia-oxidizing Archaea (AOA) and their typical habitats (terrestrial vs marine). Strongly supported branches as described in the Materials and Methods section are indicated with black dots. b Magnified view of the branches placing the ABT (Ca. Nitrosodeserticola) genomes and its sister group Ca. Nitrosocosmicus. Strongly supported branches as described in the Materials and Methods section are indicated with black dots. c Lower-left (blue) triangle of the matrix corresponds to FastANI between genomes, where gray values indicate below calculation threshold (80% identity). Upper-right (red) triangle of the matrix corresponds to 16S rRNA gene identity values, where gray values are used for genomic bins without a 16S rRNA gene. d Lower-left (blue) triangle corresponds to the amino acid identity (AAI) and upper-right (red) triangle corresponds to the Orthologous Fraction (OF) between a pair of compared genomes

Fig. 5

Shared and auxiliary protein clusters of ABT and its sister genus. a Shared orthologous protein clusters (including singletons) across six genomes (three Ca. Nitrosocosmicus, three Ca. Nitrosodeserticola [ABT]). b Number of proteins in each genome. c Number of orthologous protein clusters (excluding singletons) shared across x number of genomes