Microorganisms with novel dissimilatory (bi)sulfite reductase genes are widespread and part of the core microbiota in low-sulfate peatlands

Abstract

Peatlands of the Lehstenbach catchment (Germany) house as-yet-unidentified microorganisms with phylogenetically novel variants of the dissimilatory (bi)sulfite reductase genes dsrAB. These genes are characteristic of microorganisms that reduce sulfate, sulfite, or some organosulfonates for energy conservation but can also be present in anaerobic syntrophs. However, nothing is currently known regarding the abundance, community dynamics, and biogeography of these dsrAB-carrying microorganisms in peatlands. To tackle these issues, soils from a Lehstenbach catchment site (Schlöppnerbrunnen II fen) from different depths were sampled at three time points over a 6-year period to analyze the diversity and distribution of dsrAB-containing microorganisms by a newly developed functional gene microarray and quantitative PCR assays. Members of novel, uncultivated dsrAB lineages (approximately representing species-level groups) (i) dominated a temporally stable but spatially structured dsrAB community and (ii) represented "core" members (up to 1% to 1.7% relative abundance) of the autochthonous microbial community in this fen. In addition, denaturing gradient gel electrophoresis (DGGE)- and clone library-based comparisons of the dsrAB diversity in soils from a wet meadow, three bogs, and five fens of various geographic locations (distance of ∼1 to 400 km) identified that one Syntrophobacter-related and nine novel dsrAB lineages are widespread in low-sulfate peatlands. Signatures of biogeography in dsrB-based DGGE data were not correlated with geographic distance but could be explained largely by soil pH and wetland type, implying that the distribution of dsrAB-carrying microorganisms in wetlands on the scale of a few hundred kilometers is not limited by dispersal but determined by local environmental conditions.

FIG. 1.

(A) Geographic locations of the investigated wetlands (filled circles). (B) Projection of the wetland soil samples on the ordination plane formed by PC1 and PC2 from dsrB DGGE-based community structure data and relationships of PC1 and PC2 to the pH and the sulfate and nitrate concentrations of soil water (indicated by arrows). For each wetland site, the positions of the three replicate soil cores (indicated by the letters A, B, and C) in the ordination biplot are depicted (positions of the Schremser Hochmoor cores are additionally indicated by arrows with dashed lines). DGGE data from the Schlöppnerbrunnen II fen site were not included in the PC analysis because the corresponding environmental parameters were not determined. The wetland type is indicated: B, bog; F, fen; WM, wet meadow. Gray shading additionally highlights the clustering of bog soil samples. PC1 was strongly negatively correlated with pH (correlation coefficient of −0.9963), whereas PC2 was moderately negatively correlated with sulfate (correlation coefficient of −0.7286) and nitrate (correlation coefficient of −0.6834).

FIG. 2.

Microarray-based dsrAB diversity analysis of Schlöppnerbrunnen II fen soils sampled from different depths in the years 2001, 2004, and 2007. (A) Results of microarray hybridizations are displayed as mean nSBRs of all probes within a probe-target group and are averaged between triplicate hybridizations (technical replicates for the year 2001 and biological replicates for the years 2004 and 2007) (results of individual replicate analyses are presented in Fig. S2 in the supplemental material). The color code translates into different mean nSBR values (right-axis legend). Sampling depth and year values are shown on the left axis. Probe-target groups are arranged phylogenetically on the upper horizontal axis according to their position in a schematic dsrAB neighbor-joining tree. The affiliation of probe-target groups to previously detected Schlöppnerbrunnen fen OTUs (51, 73) is indicated on the lower horizontal axis. Figure S1 and Table S4 in the supplemental material include further information on the phylogeny and GenBank accession numbers of the clones affiliated with the different probe-target groups and OTUs. (B) Multidimensional scaling plot based on Bray-Curtis similarities between microarray hybridization patterns shown in A. Minimal Bray-Curtis similarities of all samples in a given group of samples are indicated in color.

FIG. 3.

Absolute and relative abundances of three dsrAB OTUs at different depths of the Schlöppnerbrunnen II fen site in the years 2001, 2004, and 2007, as determined by qPCR. Error bars are the standard deviations of the means for the three replicates. Black circles indicate total numbers of bacterial and archaeal 16S rRNA genes. Copy numbers of dsrA genes of OTUs 1, 2, and 6 are displayed as white, light gray, and dark gray bars, respectively. Additionally, the relative abundance of each dsrAB OTU (given as a percentage of the total number of bacterial and archaeal 16S rRNA genes) and copy number ratios between individual dsrAB OTUs are shown for all years and depths.

FIG. 4.

DsrAB consensus tree showing the affiliation of selected dsrAB and dsrA clone sequences in comparison to dsrB DGGE sequences recovered from nine different wetlands: five fens (F), three bogs (B), and a wet meadow (WM). OTUs from the Rasner Möser fen and selected OTUs from the Schlöppnerbrunnen fen sites (51) were based on DsrAB sequences longer than 500 amino acids and are depicted in boldface and colored type. For each Rasner Möser OTU, a representative clone and the number of sequenced clones are given in parentheses and square brackets, respectively. DsrA sequences were added by using the ARB parsimony interactive tool without changing the overall tree topology and are indicated by dotted branches. Phylogenetic positions of DsrB DGGE sequences are indicated by asterisks (numbers behind asterisks indicate the dsrB OTU numbers) and were determined by adding the sequences to a separate DsrAB consensus tree using the ARB parsimony interactive tool. In addition, the presence or absence of each dsrB OTU in the nine wetlands was inferred from comparisons of DGGE banding patterns and is indicated by the presence or absence of the respective colored square. The scale bar indicates 10% estimated sequence divergence (distance matrix analysis).