Carbon metabolism and biogeography of candidate phylum " Candidatus Bipolaricaulota" in geothermal environments of Biga Peninsula, Turkey

Abstract

Terrestrial hydrothermal springs and aquifers are excellent sites to study microbial biogeography because of their high physicochemical heterogeneity across relatively limited geographic regions. In this study, we performed 16S rRNA gene sequencing and metagenomic analyses of the microbial diversity of 11 different geothermal aquifers and springs across the tectonically active Biga Peninsula (Turkey). Across geothermal settings ranging in temperature from 43 to 79°C, one of the most highly represented groups in both 16S rRNA gene and metagenomic datasets was affiliated with the uncultivated phylum "Candidatus Bipolaricaulota" (former "Ca. Acetothermia" and OP1 division). The highest relative abundance of "Ca. Bipolaricaulota" was observed in a 68°C geothermal brine sediment, where it dominated the microbial community, representing 91% of all detectable 16S rRNA genes. Correlation analysis of "Ca. Bipolaricaulota" operational taxonomic units (OTUs) with physicochemical parameters indicated that salinity was the strongest environmental factor measured associated with the distribution of this novel group in geothermal fluids. Correspondingly, analysis of 23 metagenome-assembled genomes (MAGs) revealed two distinct groups of "Ca. Bipolaricaulota" MAGs based on the differences in carbon metabolism: one group encoding the bacterial Wood-Ljungdahl pathway (WLP) for H₂ dependent CO₂ fixation is selected for at lower salinities, and a second heterotrophic clade that lacks the WLP that was selected for under hypersaline conditions in the geothermal brine sediment. In conclusion, our results highlight that the biogeography of "Ca. Bipolaricaulota" taxa is strongly correlated with salinity in hydrothermal ecosystems, which coincides with key differences in carbon acquisition strategies. The exceptionally high relative abundance of apparently heterotrophic representatives of this novel candidate Phylum in geothermal brine sediment observed here may help to guide future enrichment experiments to obtain representatives in pure culture.

Keywords: carbon fixation; deep biosphere; hydrothermal; pangenomics; phylogenomics; “Candidatus Bipolaricaulota”.

FIGURE 1

Study sites and microbial community structure. (A) Map showing the location of the 11 geothermal sampling sites in Biga Peninsula, Turkey. The legend shows the names of the samples corresponding to the numbers on the map. (B) Microbial community structure of collected samples (filtered fluids, sediments, mud, and biofilm samples) from the studied geothermal represented as relative abundance of 16S rRNA gene sequences (y axis). These bars are only displayed for those samples where a phylum/class reached relative abundance in the 16S rRNA gene dataset >1% in any of the samples. Legends show corresponding groups (left) and type of samples (right). The map used in this figure was created by Maps-For-Free which is credited in the image (https://maps-for-free.com/#close).

FIGURE 2

Redundancy analysis (RDA) ordination plot for Biga Peninsula Phylum operational taxonomic units (OTUs) distribution constrained by six environmental variables. Environmental parameters (blue arrows) included: oxygen concentration (O₂), temperature (T), pH, salinity, silica (Si), and sulfate (SO₄^2–) concentration. Color codes of the different phylum are the same as in Figure 1, being “Ca. Bipolaricaulota” highlighted in red.

Figure 3

Isotopic composition of the geothermal waters. The δ²H (‰) and δ¹⁸O (‰) plot of thermal waters obtained in the study (blue), and other studies (Yalcin, 2007; Magnabosco et al., 2016). The GMWL and EMMWL stand for Global Meteoric Water Line (Craig, 1961) and East Mediterranean Water Line (Gat and Carmi, 1970). The hypersaline brine waters at the Tuzla site (Figure 1) that do not fall on the EMMWL or GMWL are highlighted in red circle.

FIGURE 4

The phylogenetic tree is an analysis of concatenated 69 ribosomal proteins extracted from 23 metagenome assembled genomes affiliated to ‘Ca. Bipolaricaulota’. Black circles at nodes represent bootstrap support of 90%, gray circles represent bootstrap support from 70 to 90%, and white circles represent bootstrap support from 70 to 50%. Blue boxes represent the MAGs with no Wood-Ljungdahl pathway (WLP) encoded. The green stars show the MAGs containing the WLP. Bold and colored fonts represent the MAGs assembled from the geothermal sites in this study (dark blue and red). The bubble plot shows the relative abundance of the “Ca. Bipolaricaulota” MAGs assembled in this study corresponding to their position in the phylogenetic tree. The histograms above the bubble plot show the relative abundance of ‘Ca. Bipolaricaulota’ from 16S rRNA gene datasets for each sampling location. The map represents the sampling locations, which are also shown in Figure 1 and Supplementary Figure 1.

FIGURE 5

Comparative genomic analysis of gene clusters encoding Wood-Ljungdahl pathway (WLP) (top row), and those lacking the WLP (bottom row). The metagenome-assembled genomes (MAGs) recovered in this study are shown with bold letters and bigger font size; the MAGs taken from the public databases are displayed in black with relatively smaller font size. The heatmap shows average nucleotide identity between MAGs across the entire core genome. Values were calculated in the anvio v7 using pyANI (Pritchard et al., 2016). The colored squares represent the enzymes catalyzing the WLP.