Expanded Genomic Sampling Refines Current Understanding of the Distribution and Evolution of Sulfur Metabolisms in the Desulfobulbales

Abstract

The reconstruction of modern and paleo-sulfur cycling relies on understanding the long-term relative contribution of its main actors; these include microbial sulfate reduction (MSR) and microbial sulfur disproportionation (MSD). However, a unifying theory is lacking for how MSR and MSD, with the same enzyme machinery and intimately linked evolutionary histories, perform two drastically different metabolisms. Here, we aim at shedding some light on the distribution, diversity, and evolutionary histories of MSR and MSD, with a focus on the Desulfobulbales as a test case. The Desulfobulbales is a diverse and widespread order of bacteria in the Desulfobacterota (formerly Deltaproteobacteria) phylum primarily composed of sulfate reducing bacteria. Recent culture- and sequence-based approaches have revealed an expanded diversity of organisms and metabolisms within this clade, including the presence of obligate and facultative sulfur disproportionators. Here, we present draft genomes of previously unsequenced species of Desulfobulbales, substantially expanding the available genomic diversity of this clade. We leverage this expanded genomic sampling to perform phylogenetic analyses, revealing an evolutionary history defined by vertical inheritance of sulfur metabolism genes with numerous convergent instances of transition from sulfate reduction to sulfur disproportionation.

Keywords: Desulfobulbaceae; comparative genomics; convergent evolution; dissimilatory sulfur metabolism; sulfur.

FIGURE 1

Sedimentary biogeochemical sulfur cycle. The sediment-water interface is indicated at the “Sediment surface.” First, seawater sulfate (SO₄^2– ) diffuses into the sediments and enters the reductive sulfur cycle promoted by microbial sulfate reduction (MSR) (indicated with red arrows), which couples the reduction of sulfate to sulfide (HS^– ) to the oxidation of organic matter (OM). A fraction of the produced sulfide precipitates with iron to form pyrite (FeS₂) and is ultimately buried (yellow arrow). A portion of biogenic sulfide is oxidized – either biotically through sulfide oxidation (SO, shown with a blue arrow) or abiotically – using common oxidants – oxygen (O₂) or nitrate (NO₃^2– ) – to yield intermediate sulfur species (S_int). These are then disproportionated via microbial sulfur disproportionation (MSD) to release sulfate and sulfide (depicted with a purple arrow).

FIGURE 2

(A) Tree of Life built with concatenated ribosomal proteins following Hug et al. (2016) collapsed at the phylum level as classified by GTDB-Tk showing the relationship of Desulfobacterota relative to Proteobacteria and other major bacterial groups. (B) Concatenated ribosomal protein phylogeny of the Desulfobacterota binned at the family (Desulfobulbales) or class (all other lineages) levels, labeled with taxonomic assignments from GTDB-Tk, showing the placement of and relationships within the Desulfobulbales. The number of genomes from each clade used in the construction of the tree noted in parentheses after taxonomy label. Nodes are labeled with TBE support values.

FIGURE 3

Phylogenetic tree showing the genomes of isolated and well-characterized members of the Desulfobulbales, including the families Desulfobulbaceae, Desulfocapsaceae, and Desulfurivibrionaceae. Nodes are labeled with TBE support values. Species names are highlighted with colors corresponding to the taxonomic family to which they are assigned. On the right, the characterized capacity for performing sulfur metabolisms is indicated.

FIGURE 4

Heatmap of metabolic functions produced by the KEGG-decoder of the members of the Desulfobulbales sequenced here. The color gradient reflects the fractional abundance of genes associated with a pathway encoded by a particular genome. In other words, white implies no genes associated with a pathway of interest are found in the genome and thus that said pathway is not constituted. Conversely, dark red indicates all genes required to perform the pathway of interest are found and that said metabolism is fully constituted in the genome. Implications for the presence or absence of metabolic pathways of interest in each genome are discussed in the text.