Iron-Fueled Life in the Continental Subsurface: Deep Mine Microbial Observatory, South Dakota, USA

Abstract

Iron-bearing minerals are key components of the Earth's crust and potentially critical energy sources for subsurface microbial life. The Deep Mine Microbial Observatory (DeMMO) is situated in a range of iron-rich lithologies, and fracture fluids here reach concentrations as high as 8.84 mg/liter. Iron cycling is likely an important process, given the high concentrations of iron in fracture fluids and detection of putative iron-cycling taxa via marker gene surveys. However, a previous metagenomic survey detected no iron cycling potential at two DeMMO localities. Here, we revisited the potential for iron cycling at DeMMO using a new metagenomic data set including all DeMMO sites and FeGenie, a new annotation pipeline that is optimized for the detection of iron cycling genes. We annotated functional genes from whole metagenomic assemblies and metagenome-assembled genomes and characterized putative iron cycling pathways and taxa in the context of local geochemical conditions and available metabolic energy estimated from thermodynamic models. We reannotated previous metagenomic data, revealing iron cycling potential that was previously missed. Across both metagenomic data sets, we found that not only is there genetic potential for iron cycling at DeMMO, but also, iron is likely an important source of energy across the system. In response to the dramatic differences we observed between annotation approaches, we recommend the use of optimized pipelines where the detection of iron cycling genes is a major goal. IMPORTANCE We investigated iron cycling potential among microbial communities inhabiting iron-rich fracture fluids to a depth of 1.5 km in the continental crust. A previous study found no iron cycling potential in the communities despite the iron-rich nature of the system. A new tool for detecting iron cycling genes was recently published, which we used on a new data set. We combined this with a number of other approaches to get a holistic view of metabolic strategies across the communities, revealing iron cycling to be an important process here. In addition, we used the tool on the data from the previous study, revealing previously missed iron cycling potential. Iron is common in continental crust; thus, our findings are likely not unique to our study site. Our new view of important metabolic strategies underscores the importance of choosing optimized tools for detecting the potential for metabolisms like iron cycling that may otherwise be missed.

Keywords: DeMMO; FeGenie; continental subsurface; deep subsurface; iron cycling.

FIG 1

Relative abundances of iron cycling genes annotated via FeGenie. Saturated cool colors denote genes for iron oxidation; warm colors show those for iron reduction. “Other” represents hypothetical proteins attributed to porins and cytochromes for iron reduction. Pastels denote housekeeping genes related to iron acquisition, storage, and regulation. Note that genes corresponding to “iron acquisition–siderophore synthesis” are not visible here due to their low relative abundance.

FIG 2

Heat map of genes corresponding to energy metabolisms with iron annotated via FeGenie with hierarchical clustering. Gene names corresponding to x axis gene IDs are provided in Table S4.

FIG 3

Energy metabolism pathways in DeMMO communities from combined FeGenie and METABOLIC annotations. Scale is log of relative abundances of genes per metagenome; gray indicates that the pathway was not detected via FeGenie or METABOLIC pipelines. “Oxygen reduction” includes “Oxidative phosphorylation” and “Oxygen Metabolism (Oxidative phosphorylation Complex IV)” categories from METABOLIC. “Hydrogen oxidation” includes the “Hydrogenases” category from METABOLIC. “Iron reduction” includes the “iron reduction” category from FeGenie and “Metal reduction” category from METABOLIC. “Iron oxidation” includes the “iron oxidation” category from FeGenie.

FIG 4

Relative abundances of iron cycling taxa annotated via FeGenie. Only the FeGentie categories “iron oxidation” and “iron reduction” are shown. The “Other” gene category represents hypothetical proteins attributed to porins and cytochromes for iron reduction.

FIG 5

Energy density of metabolic reactions with iron modeled using in situ geochemistry is scaled to limiting reactant availability. Reactions that do not appear on this figure are endergonic. Labels on the y axis denote the iron species and reaction number, where “pyr,” “sid,” “lep,” “goe,” “hem,” and “fer” correspond to pyrite, siderite, lepidocrocite, goethite, hematite, and ferrihydrite, respectively. Reaction numbers correspond to reactions in Table 2.