The upper oceanic crust is mainly composed of basaltic lava. It has been continuously created on Earth for about 3.8 billion years. Basaltic lava is erupted and solidified at mid-ocean ridges where high-temperature basalt-seawater reactions provide substantial energy for sustaining chemosynthetic life. On ridge flanks, circulation of crustal fluid is hydrothermally driven within the basaltic lava overburdened with sediments. The portion of basaltic lava beneath sediment cover is referred to as basaltic basement. Previous studies at 3.5- and 8-million-year-old ridge-flank systems demonstrated that these young crustal aquifers, respectively, harbor anaerobic thermophiles and aerobic mesophiles that contribute to hydrogen, carbon, and sulphur cycling. After rock fractures are filled with secondary minerals, intensities of fluid circulation and basalt-seawater reactions sharply decline with increasing crustal age; with most crustal oxidation occurring in the first 10 million years after crust formation.
More than 90% of Earth’s ocean lithosphere is older than 10 million years and long past its early stage of relatively high crustal oxidation rate. Despite its vast areal extent, the nature and extent of life in this old crust is previously unknown, in part because of the technological and analytical challenges of exploring the igneous rock habitat through scientific drilling. Alteration textures suggestive of biological activity have been observed in oceanic crust as old as 3500 million years. However, the role of microbial activities in creating these textures and the age of the crust at the time of texture formation remain unknown.
In a paper published in the journal Communications Biology on 2 April 2020, Yohey Suzuki, Seiya Yamashita, Mariko Kouduka, and Yutaro Ao of the Department of Earth and Planetary Science at the University of Tokyo, Hiroki Mukai, also of the Department of Earth and Planetary Science at the University of Tokyo, and of the Mantle Drilling Promotion Office at the Japan Agency for Marine-Earth Science and Technology, Satoshi Mitsunobu of the Department of Environmental Conservation at Ehime University, Hiroyuki Kagi of the Geochemical Research Center at the University of Tokyo, Steven D’Hondt of the Graduate School of Oceanography at the University of Rhode Island, Fumio Inagaki of the Kochi Institute for Core Sample Research and Mantle Drilling Promotion Office at the Japan Agency for Marine-Earth Science and Technology, and Yuki Morono, Tatsuhiko Hoshino, Naotaka Tomioka, and Motoo Ito, also of the Kochi Institute for Core Sample Research at the Japan Agency for Marine-Earth Science and Technology, investigate the occurrence of microbial communities in subseafloor basaltic lava older than 10 million years, recovered by Integrated Ocean Drilling Program Expedition 329 in the South Pacific Gyre. The presence of microbial cells in the iron-rich smectite on old subseafloor basaltic rock was revealed by nanoscale solid characterizations. Analysis of their lipid profiles and DNA sequences reveals the dominance of heterotrophic Bacteria, suggesting the presence of organic matter resources in the subseafloor basalt.
Within the South Pacific Gyre, extremely low sedimentation rates lead to burial of sediments nearly depleted in organic matter. In this ultraoligotrophic environment, dissolved oxygen penetrates from the ocean floor to the basaltic basement and sustains aerobic microbes throughout the sediment column. During Expedition 329, using the drilling vessel JOIDES Resolution, core samples were obtained from basaltic basement at Sites U1365, U1367, and U1368 with crustal ages of 104 million years, 33.5 million years, and 13.5 million years.
Mineral characterisations were conducted for core samples with fractures/veins to clarify the presence of clay minerals typically produced by low-temperature rock–water interactions (weathering). X-ray diffraction analysis revealed the presence of iron-rich smectite in 33.5-million-year-old and 104-million-year-old core samples but not in 13.5-million-year-old core samples. Thin sections were prepared from the 33.5-million-year-old and 104-million-year-old core samples with sample codes: U1367F-6R1, U1365E-8R4, and U1365E-12R2 at depths of 51, 109.6, and 121.8m below the seafloor and observed by scanning and transmission electron microscopies coupled to energy-dispersive X-ray spectroscopic analysis. Iron-rich smectite was found at the rims of fractures and veins mainly filled with celadonite and iron oxyhydroxides in U1365E-8R4 and U1365E-12R2, respectively, whereas veins are filled with Iron-rich smectite in U1367F-6R1. Two types of compositionally distinct iron-rich smectite veins were observed in U1367F-6R1: one is similar to those found in U1365E-8R4 and U1365E-12R2 with high magnesium and potasium contents; the other is characterised by high iron content, as typically observed in iron-rich smectite from deep-sea hydrothermal mounds.
Fluorescence microscopy observations of the thin sections reveal that SYBR Green I-stained cell-like fluorescence signals are extensive along the rims of the rock fractures/veins associated with irom-rich smectite in U1365E-8R4 and U1365E-12R2. Although iron-rich smectite with high magnesium and potassium contents in U1367F-6R1 is correlated with fluorescence signals, fluorescence signals were not detected from veins filled with iron-rich smectite with high iron content in U1367F-6R1.
To confirm that these greenish signals originate from microbial cells rather than from autofluorescent materials, roughly 10 × 10-μm² sections with a thickness of about 3 μm were fabricated by focused ion beam, and element-mapping images were obtained using nanoscale secondary ion mass spectrometry. Focused ion beam nanoscale secondary ion mass spectrometry analysis of U1365E-8R4 revealed overlapping signals of cyano radicals, phosphorus ions, and sulphur ions on the dense spots stained with SYBR Green I, indicating that those greenish signals are derived from microbial cells. The microbial cells are localized in the proximity of microscale voids and enrobed within iron-rich smectite.
The same result was obtained by focused ion beam nanoscale secondary ion mass spectrometry of U1365E-12R2. Element mapping using scanning transmission electron microscopy equipped with energy dispersive spectroscopy showed that the microbial cells are spatially associated with laths of iron-rich smectite. Given this association and the large compositional difference between iron-rich smectite and the bentonite clay used for drilling mud, the microbial cells were not introduced from the drilling mud. These results indicate that the detected signatures along the mineral-filled fractures/veins are derived from indigenous microbial communities in the deep crustal biosphere beneath the oceanic and sedimentary biospheres.
Core samples were evaluated for contamination using fluorescence microspheres (0.5 μm in diameter) that mimic microbial cells introduced from drilling fluid. Microscopic counting of microspheres in subsamples before and after cleaning steps such as washing with 3% sodium chloride solution and flaming the exterior showed that untreated exteriors of core samples contained detectable microspheres, but most post-treatment sample interiors contained no detectable microspheres. These results clarify that the contamination evaluation was properly conducted to show the level of drilling contamination for DNA analysis. 16S ribosomal RNA gene sequences were obtained from the V4 to V6 regions by tagsequencing from four core samples with no detected microspheres (U1365E-8R4, and -12R2 and U1367F-4R1 and U1368F-4R2), one microsphere-detected sample (U1368F-7R3), drilling fluid used at Site U1365, and a DNA extraction blank. To identify potentially contaminant operational taxonomic units from drilling and subsequent laboratory manipulations, the highly contaminated core from U1368F-7R3, the drilling fluid from U1365E, and the negative control were compared to the microsphere-undetected samples.
Because α- and β-proteobacterial operational taxonomic units were identical among the contaminated sources such as the drilling fluid and the DNA extraction blank, the operational taxonomic units detected from the contaminated sources were removed from the microsphere-undetected core samples. In addition, operational taxonomic units obtained from the highly contaminated core (U1368F-7R3) were excluded for detailed analyses of indigenous microbial communities. According to phylogenetic affiliation based on 16S rRNA gene sequences, three types of microbial communities were identified.
Type SPG-I (relatively young crustal community: 13.5 million years old). At Site U1368, γ- and ε-proteobacterial sequences were proportionally abundant and included strains related to the genera Arcobacter, Thioreductor, Sulfurimonas, and Sulfurovum known as deep-sea sulphur- and/or hydrogen-oxidising chemolithoautotrophs and the genus Alteromonas globally distributed in deep-sea aquatic habitats with aerobic heterotrophy.
Type SPG-II (aged crustal communities: 33.5–104 million years old). At Sites U1365 and U1367, β-proteobacterial sequences were predominant and closely related to aerobic organotrophs, such as Roseateles depolymerans isolated from pumice-bearing lake sediment.
Type SPG-III was only observed in U1365E-12R2 (a depth of 122 meters below sea floor), in which γ-proteobacterial sequences affiliated within the family Methylococaceae were predominantly detected. In general, these crustal microbial communities were comprised of Methylococaceae members typically found in methane-rich fluids emitted from the deep ocean floor, including cold seeps and hydrothermal vents.
Microbial communities have previously been observed in rock core and fluid samples from North Pond (North Atlantic International Oceanic Drilling Program Site; 8 million years old), where oxygenated cold fluid actively circulates in sediment-covered basaltic basement. For microbiological investigations, a Circulation Obviation Retrofit Kit was installed to collect fluid samples from the basaltic basement at North Pond. However, it is possible that microbial communities in fluid samples are distinct from those attached to adjacent rock surfaces. Microbial communities in the North Pond fluid samples are mainly comprised of members of Campylobacterales and Alteromonadales. In contrast, rock core samples were not dominantly colonised by Campylobacterales members but Alteromonadales members. Dominant microbial populations obtained in our rock sample from Site U1368 (13.5 million years old) were similar to those obtained from the North Pond fluid samples. Thus, the nature of the basement fluid may be very similar in relatively young (8 million years old and 13.5 million years old) basaltic basement in both the Atlantic and Pacific Oceans. These results also indicate that the crustal biosphere can be technically evaluated from rock cores, as well as from circulating fluid.
Basement at 13.5 million years old and 33.5 million years old is mainly composed of pillow lava covered with 12- to 17-m thick sediment. The deepest sediment at both sites contains similar concentrations of dissolved oxygen and dissolved nitrate. Although the crustal structure and the dissolved oxidant chemistry are fairly similar at both sites, microbial community composition differs notably between the 13.5- and 33.5-million year old basements. The clay minerals that form in fractures/veins by low-temperature rock-water interactions provide information that may explain the difference between these communities; the presence and absence of iron-rich smectite in fractures/veins at Sites U1367 and U1368 indicate that iron-rich smectite formation was inhibited by vigorous seawater circulation at U1368.
Although the basaltic basement at Site U1365 comprises lava flows where fluid flow is generally between sheeted layers rather than along chilled margins of pillow lava, its microbial community composition is similar to that found at Site U1367, which is consistent with the presence of iron-rich smectite at Sites U1365 and U1367. Seafloor heat flow at U1367 and U1365 is consistent with conduction as the dominant mode of heat transport, while heat flow at U1368 falls below the expected conduction-only level, consistent with apparent heat transport by fluid circulation within the rocky crust. This difference is consistent with the basement ages of the respective sites, as fluid circulation and advective heat transport are generally much more vigorous in relatively young, warm crust (such as the 13.5-million year old crust at U1368) than in much older and consequently cooler crust (such as the 33.5-million year old and 104-million year old crust at, respectively, U1367 and U1365). We suggest that the basement habitability is controlled by heat and fluid flows, which generally decrease over time, versus the primary structure of the crust (e.g. pillow basalt or flow basalt). In addition, the formation of iron-rich smectite in the basaltic basement appears to be correlated with the kinds of microorganisms in aged oceanic crust.
Observations of microbial cells in focused ion beam sections (10 μm × 10 μm × 3 μm) suggest a cell density range of 3 300 000 000 cells/cm³. In the focused ion beam sections, 15 and 2 cyanide-bearing spots derived from microbial cells are visualized in U1365E-8R4 and U1365E-12R2, which gives approximate cell numbers of 50 000 000 000 and 7 000 000 000 cells/cm³. This cell density is narrowly limited to the iron-rich smectite at the interface between basalt and alteration minerals. Within that interface, cell density is exceedingly high in comparison with cell density in the deepest sediment overlying the basaltic basements at Sites U1365 and U1367 (roughly 100 cells/cm³), and in comparison with low-temperature fluids collected from 8-million year old basalt basement at North Pond (roughly 10 000 cells/cm³). The range of cell density estimated for the iron-rich smectite of the basalt-water interface is nearly the same or higher as in organic-rich near-seafloor sediment deposited on continental margins.
To verify the cell density estimates in the two focused ion beam sections, μ-Raman spectroscopy was used to obtain a diagnostic spectrum from the microbe-smectite assemblage. The spectrum is composed of broad peaks at 1200–1600 cm⁻¹ attributed to amorphous organic matter and a slope increasing with Raman shift attributed to smectite. The fingerprint spectrum was obtained throughout the interface regions filled with iron-rich smectite with high magnesium and potassium contents in U1367F-6R1, U1365E-8R4, and U1365E-12R2, but not from that filled with iron-rich smectite with the high iron content in U1367F-6R1. The lack of the fingerprint spectrum from iron-rich smectite with high iron content may be due to its formation at a deep-sea hydrothermal mound near the mid-ocean ridge.
Smectite is a fine-grained clay mineral, with a large surface area to adsorb organic matter. As dominant microbial communities detected from 33- and 104-million year old basaltic basements are heterotrophic, it is conceivable that organic matter bound to iron-rich smectite may help to sustain the high cell density at the basalt interface. Clay fractions were separated from the core samples and their organic carbon content was quantified. The clay fractions mainly composed of iron-rich smectite contained up to 22-fold higher organic carbon than the bulk core samples, supporting the inference that mineral-bound organic matter fuels heterotrophic activities of microorganisms at the basalt interface. Fourier transform infrared-ray spectra were obtained from the clay fractions to clarify the presence of lipids, based on the aliphatic CH₃/CH₂ absorbance ratios (R₃/₂). Given that the R₃/₂ values are domain-specific: Eukarya 0.3–0.5, Bacteria 0.6–0.7 and Archaea 0.8–1.0, the R₃/₂ ranges of the clay fractions from U1365E-8R4, U1365-12R2, and U1367F-6R1 were approximately in the Bacterial range, which agrees with the dominance of Bacteria indicated by 16S rRNA gene sequences from the corresponding core samples U1365E-8R4 and U1365-12R2 and from the other core sample collected from the same site(U1367F-4R1).
16S rRNA gene sequences related to aerobic and anaerobic methanotrophs were prominent among the 16S rRNA gene sequences detected from the mineral-filled fractures in 104-million year old basaltic basement. Almost half of the 16S rRNA gene sequences analyzed from U1365E-12R2 were closely related to Methyloprofundus sedimenti, an aerobic methanotrophic Bacterium isolated from a deep-sea sediment sample associated with a Whale fall. Additionally, anaerobic methane-oxidising Archaea subtype 1 (ANME-1) was detected from U1365E-8R4. As methane concentrations are below the detection limit (below 1.3 μM) in all sediment samples at Site U1365, methane bound to iron-rich smectite might be a source of energy for their persistence in situ.
The results of this study greatly extend understanding of bioenergetics and habitability in Earth’s upper oceanic crust. Previous studies of bioenergetics in subseafloor basalt have generally focused on chemoautotrophic mineral oxidation, which mostly occurs in crust younger than about 10 million years. Our results indicate that cells encased in iron-rich smectite densely coat rock surfaces of much older (33.5 million years and 104 million years) basalt and are largely sustained by aerobic heterotrophy and methanotrophy. Organic matter that may sustain these communities in the upper crustal aquifer includes (i) dissolved organic matter in the seawater that flows through the fractures and veins, and (ii) organic matter abiotically synthesized during rock weathering (e.g. Lost City where amino-acid production associated with formation of iron, magnesium-rich smectite in gabbroic basement at the Lost City hydrothermal field).
These results also have important implications for understanding the abundance and global distribution of microbial cells in the upper oceanic crust. Mineral (iron and sulphur) oxidation rates are highest in crust younger than about 10 million years. The number of cells that might be supported by aerobic iron oxidation in the upper marine crust has been estimated as 2 400 000 000 000 000 000 000 000 000 cells, potentially equivalent to 10% of total cell abundance in marine sediment. Because the abundant microbes reliant on aerobic heterotrophy and methanotrophy reside in much older crust (33.5 million years and 104 million years), inclusion of these heterotrophic and methanotrophic cells may substantialy increase estimate of total cell abundance in the upper oceanic crust.
The results of this study also have implications for the possibility of life on Mars and other planetary bodies. Basaltic crust is ubiquitous on other planets, such as Mars, as well on Earth. The Martian basaltic crust formed 4 billion years ago, to be followed by formation of iron, magnesium-rich smectite via hydrothermal alteration and weathering at the surface and in the subsurface until about 3 billion years ago. On modern Mars, the surface is cold and dry under high vacuum conditions, and methane is emitted from the subsurface into the atmosphere. Recently, the presence of subsurface liquid water has been indicated, which spurred international interest in the search for extraterrestrial life. Given the subsurface presence of methane and liquid water on Mars, the communities fueled by organic matter and methane in subseafloor basalt on Earth provide a clear model for extant life and/or biosignatures from past life in the subsurface of Mars and other planets.