REPORTS 18. S. H. Naik et al., Nat. Immunol. 7, 663 (2006). 25. C. Jakubzick et al., J. Immunol. 180, 3019 Supporting Online Material 19. H. Karsunky, M. Merad, A. Cozzio, I. L. Weissman, (2008). www.sciencemag.org/cgi/content/full/1170540/DC1 M. G. Manz, J. Exp. Med. 198, 305 (2003). 26. The authors thank S. Bhuvanendran for two-photon Materials and Methods 20. C. B. M. Jakubzick, A. J. Bonito, E. L. Kuan, M. Merad, microscopy, K. Velinzon and T. Shengelia for cell sorting, Figs. S1 to S11 G. J. Randolph, J. Exp. Med. 205, 2839 (2008). and R. Steinman and A. Silva for discussions and critical References 21. M. D. Catalina et al., J. Exp. Med. 184, 2341 (1996). reading of the manuscript. K.L. was supported by a Movies S1 to S8 22. R. L. Lindquist et al., Nat. Immunol. 5, 1243 (2004). C. H. Li Memorial Scholarship Award from the Rockefeller 23. J. M. Kim, J. P. Rasmussen, A. Y. Rudensky, Nat. Immunol. University. T.A.S. was supported by the Schering 6 January 2009; accepted 4 March 2009 8, 191 (2007). Foundation. This work was supported in part by Published online 12 March 2009; 24. R. Tussiwand, N. Onai, L. Mazzucchelli, M. G. Manz, grants from the NIH to M.C.N. M.C.N. is an HHMI 10.1126/science.1170540 J. Immunol. 175, 3674 (2005). investigator. Include this information when citing this paper. 34 range = 20.8 to 21.7‰ for SSO4;0.08‰ and ‰ 33 A Contemporary Microbially range = 0.06 to 0.09 for SSO4; n =6)were 2– similar to measurements of seawater SO4 from the past 15 My measured in marine barites (11). Maintained Subglacial 18 In contrast, values of d OSO4 (3.3‰; range = 2.7 to 4.9‰; n =6)wereupto7‰ depleted Ferrous “Ocean” d18 compared with that of seawater OSO4 from T ‰ 1 1 2 3 4 marine barites over the Pliocene [10.4 1.6 Jill A. Mikucki, * Ann Pearson, David T. Johnston, Alexandra V. Turchyn, James Farquhar, d34 d18 1 5 6 7 (12)]. Because values of both Sand Oin Daniel P. Schrag, Ariel D. Anbar, John C. Priscu, Peter A. Lee 2– SO4 are influenced independently by microbial sulfur metabolism, including sulfate reduction, An active microbial assemblage cycles sulfur in a sulfate-rich, ancient marine brine beneath Taylor disproportionation, and reoxidation reactions Glacier, an outlet glacier of the East Antarctic Ice Sheet, with Fe(III) serving as the terminal electron (13, 14), we expect that both oxygen and sulfur acceptor. Isotopic measurements of sulfate, water, carbonate, and ferrous iron and functional isotopes would be affected. For example, dissim- ′ gene analyses of adenosine 5 -phosphosulfate reductase imply that a microbial consortium facilitates ilatory sulfate reduction to sulfide would cause a catalytic sulfur cycle. These metabolic pathways result from a limited organic carbon supply 34 2– on April 30, 2009 preferential enrichment of S (SO4 → H2S, because of the absence of contemporary photosynthesis, yielding a subglacial ferrous brine that is fractionation factor (e) = 20 to 40‰ for natural anoxic but not sulfidic. Coupled biogeochemical processes below the glacier enable subglacial populations) (15) because 34S-depleted S2– is microbes to grow in extended isolation, demonstrating how analogous organic-starved systems, such sequestered in iron sulfides. This would increase as Neoproterozoic oceans, accumulated Fe(II) despite the presence of an active sulfur cycle. 34 the d SSO4 value of the remaining sulfate pool. Our isotope data indicate that incorporation of ubglacial environments represent a largely present is anoxic and highly ferrous and the pH is 18O-depleted brine water oxygen into sulfate has unexplored component of Earth’sbio- circumneutral (Table 1), activity and DNA se- occurred (Fig. 1B and table S3). During glacial sphere (1). In the McMurdo Dry Valleys, quence data reveal that it supports a metabolically advancement, meltwater mixing with the remain- S 18 www.sciencemag.org Antarctica, an iron-rich subglacial outflow (Blood active, largely marine microbial assemblage (7). ing seawater decreased the d O value of the Falls) flows from the Taylor Glacier (Fig. 1A), Taylor Glacier is frozen to its bed, and brine from a marine value to its current com- 18 providing unique access to a subglacial ecosys- surface-derived water does not penetrate to the position (d H2OBrine = –39.5‰; Table 1). The 18 tem. The likely fluid source to Blood Falls is a base (8). Poorly understood hydrologic controls depleted value of d OSO4 cannot be explained pool of marine brine of unknown depth trapped result in episodic release of brine. The data in this by abiotic oxygen isotope exchange between 2– underneath the glacier ~4 km from the glacier study are from one of these active discharge SO4 and water. Such equilibration would take snout where the overlying ice is ~400 m thick (2). events. Salts and the iron mineral goethite rapidly tens of millions of years at subglacial temper- Pliocene surface uplift of the Taylor Valley floor, precipitate upon contact of the outflow with the atures and pH (16) (Table 1). However, oxygen Downloaded from 2– and the associated recession of the Ross Sea oxidizing atmosphere (5). Although regelation as isotope exchange between sulfite (SO3 )and Embayment, isolated this pocket of brine (3). the glacier slides over the bedrock and brine may water occurs rapidly [50% exchange in <5 min 2– 2– Before isolation from direct contact with the at- add trace amounts of meteoric gases including O2 (17)], and the reduction of SO4 to SO3 is mosphere, the brine was cryoconcentrated (4, 5), (9), multiple geochemical measurements reveal biologically mediated (7). Complete equilibration – 2– 2– resulting in hypersalinity (~1375 mM Cl ). This no quantitatively significant contributions. Radio- of SO4 via cycling through SO3 and exchang- brine has been isolated for at least 1.5 million carbon data confirm that dissolved inorganic ing all oxygen atoms with water would result in a 14 years (My), when the Taylor Glacier last ad- carbon (DIC) is old [D CDIC = –993 T 1(SE) 30.9‰ offset from brine water or a value for 18 vanced over the area (6). Although the brine at per mil (‰)(10)]. Presumably, interactions with d OSO4 of –8.6‰ (Fig.1B).Thisoffsetrepre- air at the point of outflow collection are sents the temperature-adjusted equilibrium frac- 14 18 2– 1 responsible for the small amount of CDIC in tionation factor of OinSO3 relative to brine Department of Earth and Planetary Sciences, Harvard Uni- 14 versity, Cambridge, MA 02138 USA. 2Department of Or- our samples ( C-free would be –1000‰). No H2O(e =30.9‰ at –5.2°C). Given the measured 18 ganismic and Evolutionary Biology, Harvard University, dissolved O2 was detected in the brine [reduction values of d OSO4 (2.7‰ to 4.9‰), isotopic mass Cambridge MA, 02138 USA. 3Department of Earth Sciences, 2– 4 potential (Eh) = 90 mV], and iron was 97% Fe(II), balance requires that 30 to 40% of the SO4 pool University of Cambridge, Cambridge CB2 3EQ, UK. Depart- indicating that inputs from ice-bound atmospher- has exchanged its O atoms with water, likely ment of Geology, University of Maryland, College Park, MD 2– 20742, USA. 5School of Earth and Space Exploration and ic O2 are minimal. through equilibration with SO3 . Department of Chemistry and Biochemistry, Arizona State Heterotrophic activity was measured by using Characterization of the Blood Falls microbial University, Tempe, AZ 85287, USA. 6Department of Land 3H-thymidine incorporation (Table 1) (10). Al- assemblage has revealed taxa that could partici- Resources and Environmental Sciences, Montana State 7 though the most abundant electron acceptor in pate in active sulfur cycling, including autotrophs University, Bozeman, MT 59717, USA. Hollings Marine Lab- 2– oratory, College of Charleston, Charleston, SC 29412, USA. the brine is SO4 , isotope data suggest that and heterotrophs (table S1). Sulfate is biological- 2– 2– *Present address: Department of Earth Sciences, Dartmouth SO4 is not terminally reduced in quantities lyreducedtoSO3 by using (phospho-)adenosine College,Hanover,NH03755,USA.Towhomcorrespondence sufficient to affect isotopes of sulfur. Values of 5′-phosphosulfate–reductase by assimilatory [3′- 34 33 should be addressed. E-mail: [email protected] d SSO4 and D SSO4 in the brine (21.0‰ and phosphoadenosine 5′-phosphosulfate (PAPS)] www.sciencemag.org SCIENCE VOL 324 17 APRIL 2009 397 REPORTS Fig. 1. (A)BloodFallsatthesnoutoftheTaylorGlacier(77°72′S 162°27′E). (Inset) Conceptual model for the possible modes of redox cycling of iron, sulfur, and organic material in the Blood Falls brine based on data from this study. Red arrows indicate assimilatory pathways (via the cell), and blue arrows indicate an alternate pathway via catalytic sulfur cycling mediated by APS reductase. (B)Valuesof 18 18 on April 30, 2009 d OSO4 relative to dH2 O, given different reaction scenarios. Sulfate and water will have equilibrated by no more than 1‰ within 50 My, therefore no net 18 18 2– 2– change would be expected in the d OSO4 of a seawater brine mixed with O-depleted glacier water (gray diamond). Where SO4 reduction occurs, SO3 18 equilibrates with the in situ water, e =25to30.5‰,resultingind OSO4 predicted by the shaded gray line (15, 24). In marine sediments (32)(blackdiamond), 2– 2– 2– complete equilibration of SO3 with in situ water is observed, indicating quantitative reduction of SO4 .IfSO4 reduction occurred to completion in Blood 18 Falls, values of d OSO4 between –8and–10‰ (25 to 30.5‰ isotopically heavier than the in situ water) would be expected. Because the data (3.3‰,black 2– circle) plot above this line of complete equilibration, only a portion (30 to 40%) of the total SO4 pool has equilibrated.