Meta-omics reveals role of photosynthesis in microbially induced carbonate precipitation at a CO2-rich geyser

Abstract

Microbially induced carbonate precipitation (MICP) is a natural process with potential biotechnological applications to address both carbon sequestration and sustainable construction needs. However, our understanding of the microbial processes involved in MICP is limited to a few well-researched pathways such as ureolytic hydrolysis. To expand our knowledge of MICP, we conducted an omics-based study on sedimentary communities from travertine around the CO₂-driven Crystal Geyser near Green River, Utah. Using metagenomics and metaproteomics, we identified the community members and potential metabolic pathways involved in MICP. We found variations in microbial community composition between the two sites we sampled, but Rhodobacterales were consistently the most abundant order, including both chemoheterotrophs and anoxygenic phototrophs. We also identified several highly abundant genera of Cyanobacteriales. The dominance of these community members across both sites and the abundant presence of photosynthesis-related proteins suggest that photosynthesis could play a role in MICP at Crystal Geyser. We also found abundant bacterial proteins involved in phosphorous starvation response at both sites suggesting that P-limitation shapes both composition and function of the microbial community driving MICP.

Keywords: MICP; mass spectrometry; metagenomics; metaproteomics; nutrient limitation; rubisco.

Figure 1

Images of the crystal geyser (CG) field site and CaCO₃ crystals. A) Overview of the CG sampling site. Arrow CG-0 marks the geyser borehole, arrow CG-1 marks sampling site 1 m from borehole and arrow CG-10 marks sampling site 10 m from borehole. B) CG borehole surrounded by a rimpool of geyser water. Green River in the background. C) Flow path of the geyser waters to Green River. D) SEM image of travertine collected at CG-1. E) SEM image of travertine collected at CG-10.

Figure 2

Diversity of bins recovered from CG travertine. The phylogenomic tree was generated based on 16 concatenated ribosomal proteins from 339 medium to high-quality bins using GToTree. Bins were reconstructed for 81 different order-level lineages (107 bins had <40% alignment coverage and are thus not displayed). Taxonomic classifications are according to GTDB-Tk. The scale corresponds to the average amino acid substitution over the alignment.

Figure 3

Taxonomy, metagenomics-based, and metaproteomics-based abundance of taxa (bins) in microbial communities recovered from CG travertine. Each bubble represents one bin/organism. Only organisms contributing more than 1% of relative bin abundance based on metagenomic read mapping and/or metaproteomics-based proteinaceous biomass assessment are displayed. Proteinaceous biomass was averaged across three replicates respectively. Empty columns in the proteinaceous biomass panel are organisms that passed the abundance threshold for read-based bin abundance but were not detected at all in the proteinaceous biomass calculation. Unassigned bacteria represent bins that could not be classified with GTDB-Tk. Unbinned prokaryotes are prokaryotic contigs that were classified as such by Whokaryote but could not be assigned to a bin by MetaBat2. Unassigned contigs are short contigs that were filtered out by Whokaryote and were manually recovered.

Figure 4

Schematic depiction of the CO2-concentrating mechanism (CCM) linked to MICP in a cyanobacterial cell. Modified from Kamennaya et al., 2012 [10].

Figure 5

Abundance trends of detected proteins in CG travertine and inference of MICP mechanism. A) Overview of the metabolic categories of the 100 most abundant proteins, B) abundances of proteins for the uptake of specific organic substrates, C) stable carbon isotope analysis of CG travertine and water [62] and comparison samples from microbialite nodules from pavilion Lake, B.C. Canada [63]. Relative protein abundances are averages across three replicates respectively. Averages were summed up per category to show the average contribution of each specific metabolic category at each site respectively. Error bars represent average standard deviation from the mean within each category. SBP: Substrate-binding protein.