New Horizons for DeepSubsurface MicrobiologySubsurface microorganisms may grow slowly, but their diversity andpersistence offer insights into life in extremis

T. C. Onstott, F. S. Colwell, T. L. Kieft, L. Murdoch, and T. J. Phelps

The U.S. Department of Energy (DOE)launched the modern era of subsur-face microbiology at its SavannahRiver Plant (SRP) in South Carolinain 1986. Those first efforts, involving

three 200-m-deep wells along with proceduresto monitor for drilling-related contaminants,uncovered abundant and diverse microbial com-munities in subsurface aquifers (Fig. 1).

Geomicrobiologists soon had seven fieldcampaigns going, through which they exploreda wide range of environments and rock types,taking core samples with a variety of fluids andtracers. Moreover, by piggybacking onto petro-leum exploratory drilling programs, they soonwere retrieving microbial samples from depthsas great as 2.7 km. The research teams were andremain interdisciplinary, including microbiolo-

gists, geochemists, geologists, and hydrologists,drawn from U.S. universities as well as DOElaboratories. This quest quickly expanded glo-bally with research groups in Europe and else-where, including the University of Bristol in theU.K., Gotenberg University in Sweden, the En-vironmental Institute in Denmark, and at sev-eral institutions in Russia and Japan.

By 1998, leaders in the young field of geomi-crobiology agreed that a substantial fraction ofliving carbon biomass consists of microorgan-isms residing beneath the seafloor and soilzones. Moreover, biodiversity within the sub-surface was striking, yielding many novel spe-cies of Firmicutes as well as some novel archaealspecies. The sheer numbers of species were sur-prising in the face of scant nutrient fluxes. Forinstance, we estimated subsurface bacterial cell

turnover times in terms of decades or centu-ries for species recovered from aquifers 200to 400 m beneath the surface. Despite thisseeming lethargy, however, the subsurfacebiosphere controls water chemistry and in-fluences mineral and organic changes in sed-imentary rocks.

Geomicrobiologists Probed Great

Depths and Extreme Environments

Over the past decade, geomicrobiologistsbegan probing the subsurface biosphere todepths greater than 3 km in South Africangold mines, at 120°C sites in China, in icesheets at �20°C about 1 km below thesurface, beneath 0.5 km of permafrost, andwithin and beneath gas hydrate deposits ofvarying depths. Deep, hot spreading centerscontaining marine chemoautotrophic eco-

Summary

• A substantial fraction of living planetary bio-mass consists of highly diverse microorganismsresiding beneath the seafloor and soil zones.

• Microbial life in subsurface zones persists inextreme environments where nutrients may bescarce, the pace of life slow, and the demand onDNA repair mechanisms intense.

• Key questions include how the same microbialspecies can occupy very distant subsurface sitesand what are the temperature extremes at whichlife cannot exist.

• Experiments being designed for the Deep Un-derground Science and Engineering Laboratorypromise insights into many subsurface micro-bial secrets, perhaps including clues to the ori-gins of life.

T. C. Onstott is aProfessor at Prince-ton University,Princeton, N.J.,F. S. Colwell is aProfessor at Ore-gon State Univer-sity, Corvallis, T. L.Kieft is a Professorat New Mexico In-stitute of Miningand Technology,Socorro, L. Mur-doch is a Professorat Clemson Univer-sity, Clemson, S.C.,and T. J. Phelps issenior research sci-entist at Oak RidgeNational Laboratory,Oak Ridge, Tenn.

Volume 4, Number 11, 2009 / Microbe Y 499

systems contrast with cooler, subseafloor sedi-ments, which provide glimpses of heterotrophicecosystems living on old, buried photosynthatematerials.

In contrast to the marine realm, the terrestrialsubsurface contains ecosystems whose chemo-autotrophic nature increases with depth. Therock-water interactions that fuel these continen-tal subsurface ecosystems depend upon re-gional, topographically driven meteoric fluidflow. For instance, in the elevated plateaus com-prising the Witwatersrand Basin and the Colum-bia River basaltic aquifer, groundwater can pen-etrate to depths of several kilometers. In thegranitic aquifers of the Fennoscandian Precam-brian shield beneath the Baltic Sea and in theAtlantic Coastal Plain aquifers, however, re-gional flow is slower and shallower.

Whether it is 1-million-year-old ice, tens-of-million-years-old saline fracture water, or Per-mian-age salt, in environments where fluid fluxis negligible, viable cells with intact genomessurvive by respiring ever so slowly, while fastid-iously repairing radiation-damaged DNA. Onehypothesis for this mystery is that radiolyticprocesses provide a renewable source of reduc-tants and oxidants to these trapped terrestrialand marine subsurface chemoautotrophic mi-

crobial ecosystems. A complicating fac-tor in evaluating biodiversity is thatDNA fragments of deceased microor-ganisms possibly formed a pool of an-cient DNA, and differences found in16S rRNA sequences from disparatesites thus might reflect evolutionarytrends rather than geographical differ-ences.

Many Remaining Microbial

Mysteries of the Deep

One key mystery concerning the deepsubsurface biosphere is how the samebacterial species can be found in oilreservoir fluids that are thousands ofkilometers apart. This continuity maybe reasonable for marine environmentswhere cold ocean currents could dis-tribute thermophiles, or for surface con-tinental sites where the atmosphere isthe principal transport vector. In othersubsurface environments, such as shal-low aquifers and seafloor spreading cen-

ters, recently termed the “subseafloor ocean,”fluid movement may also be capable of broadlydisseminating microbes from a cosmopolitansource.

However, in deep subsurface environments,particularly petroleum-trapping ones, wherethe rock is impermeable or fluid flow is ex-tremely slow, barriers blocking subsurface mi-croorganisms appear to be substantial. Yet,finding the same microbial species in distant oilreservoirs implies either high microbial trans-port rates beneath the continental surfaces orslow transport rates combined with exception-ally slow mutation rates.

With increasing subsurface residence times,geological and climate histories may influencebiogeographical distinctions in the terrestrialsubsurface. If geographic isolation is impor-tant, then geochemically similar subsurfaceenvironments presumably should yield differ-ent microbial species. To address this issue, itwill be helpful to obtain genome sequences ofsubsurface microbial species that are sepa-rated by hundreds of kilometers to determinewhether genes that are present in one, but notthe other, reflect their different environments.If so, were those genes acquired by horizontalgene transfer locally? Further, if low-diversity

F I G U R E 1

Drilling rig at the C10 drill site at the U.S. Forest Service land several miles east of theSavannah River Plant in 1988. Coring at this site exceeded 407 m below surface.

500 Y Microbe / Volume 4, Number 11, 2009

microbial ecosystems are truly iso-lated from a much larger genetic pool,how do subsurface microorganismsevolve as their environments change?The discovery of viruses and CRISPRsequences in deep-fracture water envi-ronments raises the possibility thattransduction may account for somegenetic exchange.

What is the relationship between theplanktonic and sessile microbial com-munities in the fractured, low-porosityrock strata that prevail in the deep-continental and marine crust (Fig. 2)?Cores from the Atlantic Coastal Plainaquifers contain higher microbial abun-dance and metabolic diversity thandoes groundwater containing onlyplanktonic cells from the same aqui-fers. Surface-attached microbes fromshallow, fractured-rock environmentsare phylogenetically distinct from theplanktonic cells. Thus, variations inphysical substrates provide conditionsthat select for specific types of microor-ganisms. Additionally, solid-liquid in-terfaces can generate steep chemicalgradients, resulting in more diverse eco-logical niches than those offered by the bulkwater phase. These factors may contribute to thegenerally higher diversity and higher activityin the surface-attached microbial communitiesthan among planktonic cells.

These sessile communities are likely to bedirectly involved in altering and precipitatingdiagenetic mineral phases, and in changing thechemical and physical properties of rock. Al-though sampling planktonic microbial samplesfrom vent fluids and boreholes is relativelystraightforward, characterizing sessile commu-nities from fracture surfaces has been stymiedbecause the fracture faces are flushed with dril-ling fluids, which typically are heavily contami-nated. New approaches are required to addressthis issue.

Another Unsettled Issue: Upper

Temperature Limits of Subsurface

Microbes

Another mystery concerns the maximum tem-perature of subsurface microbial activity. Anal-yses of activity in oil reservoirs indicate that

85°C appears to mark the upper temperaturelimit of microbial life there, even though hyper-thermophiles that can live at temperatures up to115°C were recovered from oil wells during the1990s.

Mesocosm experiments indicate that at tem-peratures up to 90°C, acetogens rather thanthermogenesis produce acetate, and that theseacetogens may be fueled by H2 generated byoxidizing Fe2� in magnetite to Fe3�, which, inturn, oxidizes H2S to sulfate. This thermallycatalyzed cycling of Fe and S could also stimu-late autotrophic Fe3�- and sulfate-reducing bac-teria.

In addition to molecular-stability constraints,the ultimate temperature limit for life in thesubsurface depends upon the balance betweenmaintenance energy demand rate, which ap-pears to increase with temperature, and the en-ergy supply rate from a mix of biochemicaland purely chemical reactions. This balance dic-tates physicochemical limits of subsurface lifeand suggests that salinity, redox gradients, ther-mal gradients, radiation, fluid flow, and eventectonic activity contribute to sustaining life.

F I G U R E 2

SEM image of cells attached to muscovite flakes within 2.9 Ga quartzite collected from2 km depth in South Africa. (Photo is courtesy of M. Davidson, Princeton University, andG. Southam, University of Western Ontario.)

Volume 4, Number 11, 2009 / Microbe Y 501

Does subsurface hyperthermophilicity occuronly where fluid flow rates and hence nutrientsupply rates are elevated? Addressing this ques-tion will require more in situ observations, in-cluding drilling into hot rock at varying depths,as well as extended-scale mesocosm heating ex-periments.

We need to sequence genomes of subsurfaceisolates as well as metagenomes and to conductfunctional analyses of environmental samplesfrom great depths, not only to learn what sub-surface microorganisms are capable of doing,but what they are doing in situ. The glacial paceof in situ microbial respiration hampers mi-croarray analyses of mRNA. Detecting such sig-nals may require altering subsurface envi-ronments, followed by long term monitoring ofcommunity responses.

In The Deep Hot Biosphere, the late astro-physicist Thomas Gold speculated that life orig-inated beneath the surface of the earth, a hy-pothesis that some geomicrobiologists areexploring as a way of understanding how life onEarth recovered from the sterilizing effects of the3.9 Ga, late-heavy bombardment, or lunar cat-aclysm, on its surface. Bench-top experimentsinvoke geochemical conditions that could occurin a subsurface setting. Does the equivalent ofDarwin’s “warm, little pond” exist in the sub-surface and support prebiotic chemistry? Watervugs—some isolated for millions of years—ex-ist, and some are associated with hydrothermaldeposits. Aseptic sampling of these natural,rock-encased water-storage sites could providedata with which to evaluate parts of Gold’shypothesis.

Plans for Studying Microbial Life in Deep

Underground Laboratories

The International Continental Drilling Programand the Integrated Ocean Drilling Program(IODP) held independent workshops in Septem-ber, both of which considered questions relatedto the deep biosphere. Bringing these two groupstogether would present new opportunities forland-sea transect expeditions and for exploringlife underground.

Useful though that approach could be, ana-lyzing core samples provides only snapshots ofthe microbial biosphere. A fuller analysis re-quires programs that cover a very broad rangeof space and time. For instance, the scheduled

South Pacific Gyre, North Pond, and Juan deFuca IODP drilling proposals are dedicated mi-crobiology expeditions, and the latter two aredesigned to capture kilometer-scale hydrologiccirculation cells to evaluate their energy andmicrobial inputs and outputs. These drillingprojects coupled with long-term observations atborehole installations should help to delineate a4-dimensional picture of the subseafloor oceanbiosphere.

Research activities to explore deep subsurfacelife are coalescing. The Dark Energy BiosphereInstitute (DEBI), newly funded by NSF as aresearch coordination network, will offer op-portunities to plan expeditions to study subsur-face life. DEBI workshops will focus on obser-vatory science, microbiology of sediments andthe crust, biogeochemistry, and an integratedunderstanding of marine and continental sub-surface biological systems. The workshops allaim to bring young and established scientiststogether to expand ways that we study andconceptualize life in the deep earth.

The only comparable effort in the continentalsubsurface is the Aspo Hard Rock Laboratorybelow the east coast of Sweden, where KarstenPedersen and his collaborators study microbialcommunities within fracture fluids. Their ap-proach involves circulating water from bore-holes into microcosms using high-pressure tub-ing, pumps, and anaerobic chambers. Plans forsimilar types of experiments are being developedfor several other underground research labora-tories.

Meanwhile, ambitious plans exist for the pro-posed Deep Underground Science and Engineer-ing Laboratory, or DUSEL, which takes advan-tage of the deepest (2.4 km) and largest (morethan 6 km horizontally) underground mine inthe United States, formerly Homestake goldmine in the Black Hills of South Dakota (Fig.3a). This program, to be led by Kevin Lesko, aphysicist, and William Roggenthen, a geophysi-cist, will provide the largest and deepest labora-tory for high sensitivity detectors for dark mat-ter, nucleon decay, double � decay and aspectrum of neutrinos, including a beam of neu-trinos from Fermilab, which is about 1,300 kmaway.

DUSEL also will provide a multilevel drillingplatform for observing the ecohydrology ofshallow to deep continental environmentswhere meteoric water flow is the primary mixing

502 Y Microbe / Volume 4, Number 11, 2009

mechanism. A high-capacity filtration plant togenerate water for neutrino detectors can alsosupply water for microbiology drilling projects,obviating the need for closed-loop mud tankswith their high concentrations of contaminants.Another advantage of underground coring com-pared to drilling from the surface is that drillingwater can be circulated down the outside of thecoring barrel and up the center. Thus, when thecoring bit intersects a water-filled fracture atpressure, the core is ejected with little contami-nation of the fracture surface.

DUSEL Expected To Provide Many

Microbial Insights

The DUSEL facility will have access to groundwater of diverse ages, according to modeling

studies (Fig. 3b). On the south side, water mayflow quickly downward from the ground sur-face to reach depths of 1 km in less than 1 year.In contrast, water reaching the lower depths ofthe north side of the mine comes from rockpores and may be thousands or more years old.The potential capture footprint extends out-ward from current mine workings for kilome-ters and could provide access to up to 100 km3

volume for microbial biogeographic studies(purple zone, Fig. 3b).

From the existing infrastructure at the2,438-m level where the rock temperature is55°C, boreholes (Fig. 3b, yellow lines) can beextended downward an additional 3–4 km toreach the 120°C isotherm and to explore theupper temperature limit of life. The boreholeswill be used to withdraw fluids and will become

F I G U R E 3

(a) Perspective of surface relief and workings of proposed DUSEL (former Homestake mine). (Image courtesy of Dr. Z. Hladysz of SDSMT.)(b) Perspective of mine workings (polygonal area at center of diagram) and vicinity showing simulated particle flow paths from surface todepth during the excavation period of the mine to present day. Color shading shows the water flux into the workings. The purple area boundsthe region where ground water has been captured by the mine, according to the simulation. Yellow lines are potential cored boreholes forfuture microbiology and hydrology experiments. North is to the left and south is to the right in these diagrams.

Volume 4, Number 11, 2009 / Microbe Y 503

experimental stations for conducting in situtranscriptomic and proteomic experiments, ei-ther with mobile underground laboratories(MULEs) as done at Aspo or via push-pull ex-periments within the packer-sealed fracturesthemselves.

DUSEL also offers the advantage of operatingin a facility for assay and acquisition of radio-pure materials (AARM) at the 1,474-m level,where background radiation is 10�5 times thatof typical surface counting facilities. The AARMlaboratory will facilitate the use of ultrapureradiolabeled compounds for quantifying veryslow metabolic or catabolic processes duringpush-pull experiments or to map these processesusing microautoradiographic analyses of micro-cosm experiments.

To date, three other microbially related fa-cilities are proposed for DUSEL: the coupledthermal-hydrological-mechanical-chemical-biological experimental (THMCB) facility,the fracture processes facility (FPF), and theDUSEL-CO2 facility. The THMCB experi-ments will be constructed at the 1,474-m levelusing borehole arrays to access intact volumesof rock. Plans call for experiments with mul-tiyear heating and cooling cycles, with thegoal of attaining 250°C at the center of theblock to simulate hydrothermal conditionsand induce convective circulation of fluidswithin fractures. Geomicrobiologists willmonitor how microbial communities withinfractures respond to increased thermal andfluid fluxes, changes in nutrient concen-trations, and how quickly microorganismscolonize heat-sterilized zones during the cool-ing cycle. The temperature ranges of theTHMCB experiments overlap those of the ul-tradeep boreholes from the 2,438-m level andprovide an experimental foundation for in situobservations.

The FPF will study the creation and reacti-vation, of faults and fractures in rock at the90-, 1,474-, and 2,250-m levels. Alternateheating and cooling of boreholes will expandand contract the rock mass, inducing displace-ments along preexisting fractures and creatingnew faults. Repetitive thermal cycling willlead to repetitive fracturing. Fractures createdduring this experiment will enhance perme-ability and could release H2, which in turncould enhance microbial activity. The FPF isnot only designed to determine what controlsrock strength, but it could also determine towhat extent subsurface chemoautotrophic ac-tivity is regulated by tectonic events.

Although the fractured, low-porosity rock ofHomestake mine would seem to preclude re-search into CO2 sequestration, the DUSEL CO2

facility will take advantage of existing, largediameter, subvertical boreholes that run fromthe 334- to 882-m level. They will be cased andfilled with sand and clay, enabling them to op-erate at pressures where supercritical CO2 isstable. They should provide a detailed picture ofhow CO2 transport processes are affected by thephase transition to gas or liquid. Smaller micro-cosm experiments attached to the main boreholeassembly will be used to investigate how micro-bial activity is affected during CO2 injection.This facility thus provides a pore-to-kilometer-scale platform for developing models for CO2

sequestration.These experiments represent a subset of those

planned for the first 5 years of DUSEL, andDUSEL is open to proposals for more experi-ments. With its extensive infrastructure, kilome-ter-scale spatial access, and a multidecade life-time, DUSEL will usher in the next generation ofdeep terrestrial biosphere studies. It promisesinsights into many of its well-hidden microbialsecrets, and perhaps the origin of life itself.

SUGGESTED READING

Biddle, J. F., S. Fitz-Gibbon, S. C. Schuster, J. E. Brenchley, and C. H. House. 2008. Metagenomic signatures of the PeruMargin subseafloor biosphere show a genetically distinct environment. Proc. Natl. Acad. Sci. U.S.A. 105:10583–10588.Chivian, D., E. Alm, E. Brodie, D. Culley, P. Dehal, T. DeSantis, T. Gihring, A. Lapidus, L.-H. Lin, S. Lowry, D. Moser, P.Richardson, G. Southam, G. Wanger, L. Pratt, G. Andersen, T. Hazen, F. Brockman, A. Arkin, and T. Onstott. 2008.Environmental genomics reveals a single species ecosystem deep within the Earth. Science 322:275–278.D’Hondt, S., B. B. Jorgensen, D. J. Miller, A. Batzke, R. Blake, B. A. Cragg, H. Cypionka, G. R. Dickens, T. Ferdelman, K. U.Hinrichs, N. G. Holm, R. Mitterer, A. Spivack, G. Z. Wang, B. Bekins, B. Engelen, K. Ford, G. Gettemy, S. D. Rutherford,H. Sass, C. G. Skilbeck, I. W. Aiello, G. Guerin, C. H. House, F. Inagaki, P. Meister, T. Naehr, S. Niitsuma, R. J. Parkes,A. Schippers, D. C. Smith, A. Teske, J. Wiegel, C. N. Padilla, and J. L. S. Acosta. 2004. Distributions of microbial activities indeep subseafloor sediments. Science 306: 2216–2221.Lin, L. H., J. Hall, J. Lippmann-Pipke, J. A. Ward, B. S. Lollar, M. DeFlaun, R. Rothmel, D. Moser, T. M. Gihring, B.

504 Y Microbe / Volume 4, Number 11, 2009

Mislowack, and T. C. Onstott. 2005. Radiolytic H-2 in continental crust: Nuclear power for deep subsurface microbialcommunities. Geochemistry Geophysics Geosystems 6:Q07003.Lipp, J. S., Y. Morono, F. Inagaki, and K.-U. Hinrichs. 2008. Significant contribution of Archaea to extant biomass in marinesubsurface sediments. Nature 454:991–994.Parkes, R. J., P. Wellsbury, I. D. Mather, S. J. Cobb, B. A. Cragg, E. R. C. Hornibrook, and B. Horsfield. 2007. Temperatureactivation of organic matter and minerals during burial has the potential to sustain the deep biosphere over geologicaltimescales. Organic Geochemistry 38:845–852.Phelps, T. J., E. M. Murphy, S. M. Pfiffner, and D. C. White. 1994. Comparison between geochemical and biological estimatesof subsurface microbial activities. Microbial Ecology 28:335–349.Sadoulet, B., E. Beier, C. Fairhurst, T.C. Onstott, R.G.H. Robertson and J. Tiedje, 2007. Deep Science: A Deep UndergroundScience and Engineering Initiative. The National Science Foundation, Washington, D.C.Wang, F., H. Zhou, J. Meng, X. Peng, L. Jiang, P. Sun, C. Zhang, J. D. V. Nostrand, Y. Deng, Z. He, L. Wu, J. Zhou, and X.Xiao. 2009. GeoChip-based analysis of metabolic diversity of microbial communities at the Juan de Fuca Ridge hydrothermalvent. Proc. Natl. Acad. Sci. U.S.A. 10.1073_pnas.0810418106.Zhang, G., H. Dong, Z. Xu, D. Zhao, and C. Zhang. 2005. Microbial Diversity in Ultra-High-Pressure Rocks and Fluids fromthe Chinese Continental Scientific Drilling Project in China. App. Env. Microbio. 71:3213–3227.

Volume 4, Number 11, 2009 / Microbe Y 505