Insights into microbial sampling of ultra-low biomass, ultra-deep, hypersaline fluids fluids in Otaniemi, Finland

In the efforts to find new, environmentally friendly means for district heat production, two ultra-deep production wells were drilled in Otaniemi, Finland, reaching depths of approximately 6 km. At this depth, the fluids are saline, reaching 200 g L-1 TDS, anoxic, approximately 100°C hot and experiencing a hydrostatic pressure of 500 bar. The fluids are slightly alkaline and oligotrophic but contain CH4 and H2.
In an effort to study the microbiological composition of these fluids, samples were collected using a positive displacement sampler operated by ICDP’s Operational Support Group (OSG) in September, 2024. The sample volumes were ~600 mL/sample. The sampler was cleaned with 70% ethanol and rinsed with sterile MilliQ water. Contamination control samples were collected by filling the samplers with sterile water and collecting and treating the control sample in the same way as the actual samples. Contamination control samples were also collected from the water-glycol solution and mineral oil used for the sampler operation. The samples were studied by epifluorescence microscopy, DNA was extracted by first collecting the biomass on 0.1 𝜇m pore-size filters and the filtrate was additionally precipitated with PEG. The microbial communities were characterized by qPCR, amplicon sequencing and metagenomics. In addition, cultures targeting heterotrophic and autotrophic thermophiles were set up from all samples in anaerobic infusion bottles.
The first results indicate that even low-level contamination from reagents, equipment and sample handling is detrimental to the study. The amount of indigenous microorganisms was < 5 bacterial 16S rRNA gene copies mL-1 measured from the 0.1 𝜇m pore-size filter DNA extraction, whereas the contamination control (MilliQ water) contained between 160 – 380 gene copies mL-1. The filtrate, i.e. <0.1 𝜇m sized cells, the copy numbers were 17 – 86 mL-1, whereas the MilliQ contamination control samples contained 15 – 21 bacterial 16S rRNA gene copies mL-1. Epifluorescence microscopy corroborated the results. Amplicon sequencing showed a high proportion of contamination from sampling equipment, sample handling and laboratory reagents. Stringent curation of the data by removing all typical human contaminants and all taxa present also in the contamination controls, revealed a bacterial community mostly consisting of members of the Patescibacteria phylum. These bacteria are ultra-small and are often found in deep groundwater environments. Our study is the first to show Patescibacteria as the dominating component of the community in ultra-deep, hot, saline fluids, although their numbers are low. Culture based and metagenomic analyses are ongoing to reveal the first clues to the metabolic capacities of this phylum.