Ocean Drilling Program (ODP) and Integrated Ocean Drilling Program (IODP) sites from which there is published data on functional genes are shown in Fig. 1. So far, these studies have resulted in publications on four different functional genes: the aforementioned mcrA, dsrAB, and fhs, as well as the gene for reductive dehalogenase (rdhA), an enzyme which catalyzes the reductive removal of halogen groups in halogenated organic compounds (Futagami et al., 2009; Table 1). All publications have used PCR-based methods to obtain sufficient amounts of gene product for DNA sequencing and for quantitative analyses via real-time PCR. The phylogenetic and metabolic diversity of organisms detected so far, the sites where they have been found, and information on their closest cultured relatives are reviewed in the following sections.
Methane-cycling archaeal communities have been phylogenetically characterized at five locations – all in the Pacific Ocean: the Nankai Trough (Newberry et al., 2004), the Peru Trench (Inagaki et al., 2006a), the Peru Margin (Webster et al., 2006), the Cascadia Margin (Yoshioka et al., 2010), and off Shimokita Peninsula (Imachi et al., 2011; Fig. 1, Table 1). In addition, mcrA genes have been quantified by real-time PCR in cores from Cascadia Margin (Colwell et al., 2008) and the Porcupine Seabight in the Atlantic Ocean (Webster et al., 2009), and two strains of hydrogenotrophic methanogens have been isolated from subseafloor sediment of the Nankai Trough (Mikucki et al., 2003; Kendall et al., 2006; Table 1). At most sites, mcrA genes were only detected and analyzed at few sampling depths. Therefore, detailed correlations between the community composition of methanogens and anaerobic methanotrophs and environmental variables, such as geochemical gradients or lithostratigraphy, are not possible, and it remains to be shown which variables drive the community composition of methane-cycling Archaea in subsurface habitats. The difficulty of detecting mcrA or 16S rRNA gene sequences of known methanogens or methanotrophs even at sites with high biogenic methane concentrations (Biddle et al., 2006) suggests that existing methods either underestimate the population size or that small populations of methanogens and methanotrophs produce and consume the vast biogenic methane reservoir found in subseafloor sediments (Inagaki et al., 2006a).
The published mcrA sequences show a considerable diversity of methane-cycling Archaea in subseafloor sediments, with genera belonging to the orders Methanosarcinales (Methanosarcina, Methanococcoides, Methanosaeta, ANME-2), Methanocellales (Methanocella), Methanomicrobiales (Methanoculleus), Methanobacteriales (Methanobrevibacter, Methanobacterium), Methanococcales (Methanococcus), ANME-1, and a novel order with one recent isolate (Methanomassiliicoccus; Fig. 2a, Table 1). The genera Methanosarcina, Methanobrevibacter, and Methanobacterium have been found in the greatest number of locations, with Methanosarcina and Methanobrevibacter in 3/5 (Newberry et al., 2004; Webster et al., 2006; Imachi et al., 2011) and Methanobacterium in 2/5 locations (Yoshioka et al., 2010; Imachi et al., 2011). In some cases, the degree of mcrA sequence similarity between sites is striking, with nearly identical phylotypes related to Methanosarcina barkeri detected in the Nankai Trough (Newberry et al., 2004) and Peru Margin (Webster et al., 2006) and equally similar mcrA sequences related to Methanobrevibacter arboriphilus detected in the Nankai Trough (Newberry et al., 2004), Peru Margin (Webster et al., 2006), and off Shimokita Peninsula (Imachi et al., 2011; Fig. 2a). All other genera have only been detected in single locations, suggesting that, despite overlaps, the community of methane-cycling Archaea varies widely between sites.
Figure 2. Phylogenetic distance trees of (a) mcrA, (b) dsrAB, and (c) fhs, including phylotypes obtained during ODP and IODP expeditions in bold. All trees are based on nucleotide sequences aligned in ARB (Ludwig et al., 2004) and were constructed using the neighbor-joining function in SeaView (http://pbil.univ-lyon1.fr/software/seaview.html). Bootstrap values (in %) were calculated from 1000 replicates each. Only values of ≥ 50% are shown.
Download figure to PowerPoint
Interestingly, none of the mcrA groups so far detected during drilling expeditions are unique to the deep biosphere. Several groups can be considered typical of marine environments, however. These include the putatively methanotrophic ANME-1 and ANME-2 Archaea, and the methanogenic Methanococcoides and Methanococcus genera. ANME-1 and ANME-2 Archaea have mainly been found in sulfate-rich habitats or within/near sulfate–methane transition zones (SMTZs), from estuarine and coastal to deep sea cold seep, CO2 lake, and hydrothermal sediments (e.g. Boetius et al., 2000; Thomsen et al., 2001; Michaelis et al., 2002; Teske et al., 2002; Inagaki et al., 2006b; Lloyd et al., 2011; Yanagawa et al., 2011; Biddle et al., 2012; for review see Knittel & Boetius, 2009). Similarly widespread in the marine environment are the Methanococcoides, a genus predominantly detected in and repeatedly isolated from marine sediment samples, that is comprised of obligately methylotrophic methanogens within the Methanosarcinales (Sowers & Ferry, 1983; Dhillon et al., 2005; Singh et al., 2005; Roussel et al., 2009a; Parkes et al., 2012; reviewed in Whitman et al., 2006). The H2/CO2 and formate-utilizing Methanococcus genus has mainly been detected in and isolated from salt marsh, estuarine and coastal sediment (Whitman et al., 1986; Franklin et al., 1998; Ward et al., 1989; reviewed in Whitman et al., 2006). Other genera, such as the Methanosarcina, Methanosaeta, Methanocella, and Methanoculleus, cannot be considered ‘typically marine’, as they are, apart from estuarine and marine sediment, frequent across a wide range of nonmarine anoxic habitats, including animal feces and sewage sludge, landfills, lake sediment and freshwater wetlands (e.g. Lanoil et al., 2001; Lueders et al., 2001; Luton et al., 2001; Castro et al., 2004; Dhillon et al., 2005; Liu & Whitman, 2008; Nunoura et al., 2008; Sakai et al., 2008; Zhang et al., 2008; Roussel et al., 2009a; Parkes et al., 2012; reviewed in Whitman et al., 2006 and Liu & Whitman, 2008). Perhaps the most surprising finding is the detection of the Methanobacterium and Methanobrevibacter genera. Traditionally, Methanobacterium species have been considered sensitive to salinity, with culture media (NaCl) in excess of 0.2 M (c. 40% of typical seawater salinity) inhibiting cell growth (Whitman et al., 2006). The recent isolation of several new species from marine or saline environments has changed this perception, however (Shlimon et al., 2004; Mori & Harayama, 2011). And the genus Methanobrevibacter is a typical inhabitant of animal intestines, sewage sludge, and decaying plant matter on land (Zeikus & Henning, 1975; Miller & Wolin, 1986; Ufnar et al., 2006; Whitman et al., 2006) – environments that differ strikingly from the energy-depleted deep subsurface biosphere!
According to thermodynamic competition theory, methanogenic Archaea using electron donors such as H2, formate and acetate are outcompeted by organisms performing energetically more favorable reactions, such as sulfate reduction, given the presence of electron acceptors required for the latter (e.g. Cappenberg, 1974; Cord-Ruwisch et al., 1988; Lovley & Goodwin, 1988). This principle has been shown to apply to coastal marine sediment (Hoehler et al., 1998). As a result, only minority populations of methanogens using so-called noncompetitive substrates, that is, C1 compounds such as methylamines, methyl sulfides, and methanol, not used by most sulfate reducers, can coexist with sulfate reducers in sulfate-rich sediment (e.g. Oremland & Polcin, 1982; King et al., 1983; Kiene et al., 1986). Based on closest cultured relatives, methanogens in the deep subsurface consume the full spectrum of known methanogenic substrates (H2/CO2, formate, acetate, C1 compounds; Table 1). The distribution only partially reflects the zonation observed elsewhere, however. While consistent with the notion of thermodynamics-driven competition, mcrA genes are below detection or in low numbers in sulfate-reducing sediment of the Peru Trench and Cascadia Margin, respectively, and increase within the methanogenesis zone (Inagaki et al., 2006a; Yoshioka et al., 2010); they are at peak abundance in sulfate-rich sediment above the SMTZ at IODP Sites 1244 and 1245 in the Cascadia Margin (Colwell et al., 2008). McrA with high sequence similarity to hydrogenotrophic M. arboriphilus are not only present but appear the most active in sulfate-rich surface sediments of ODP Site 1174 (Newberry et al., 2004). And the highest mcrA abundances detected so far in the deep subseafloor are from sulfate-rich sediments near the Porcupine Seabight Challenger Mound (IODP U1318), in depth layers with concomitantly high rates of hydrogenotrophic and aceticlastic methanogenesis (Webster et al., 2009).
So far, phylogenetic characterizations of sulfate (sulfite)-reducing microbes have been published from subseafloor sediment of the Peru Margin (ODP Site 1228; Webster et al., 2006) and rust deposits of a circulation obviation retrofit kit (CORK) at the seafloor on the Juan de Fuca Ridge Flank (ODP Site 1026; Nakagawa et al., 2006; Table 1). dsrA copy numbers have been quantified via real-time PCR in sediments of the Peru Margin (ODP Sites 1227, 1229, 1230; Schippers & Neretin, 2006), Porcupine Seabight (IODP Sites 1316–18; Webster et al., 2009), and the Gulf of Mexico continental slope (IODP Sites 1319–20, 1322, 1324; Nunoura et al., 2009).
Given that only one sampling depth from one subseafloor sediment column (48 mbsf, ODP Site 1228; Webster et al., 2006) has resulted in successful dsrA detection, general inferences regarding the ecology of sulfate-reducing microbes in this environment are yet to be made. The only phylotypes detected in this sample belong to a deeply branching, uncultivated dsrAB genetic cluster with unknown 16S rRNA gene identity that was first detected in hydrothermal sediment (Cluster IV; Dhillon et al., 2003; Fig. 2b). Under various monikers, this phylogenetically diverse group has since been reported from a wide range of estuarine to deep sea marine sediments and cold seeps (e.g. Bahr et al., 2005; Kaneko et al., 2007; Jiang et al., 2009; Leloup et al., 2009; Ye et al., 2009), as well as sulfidic springs, subsurface aquifers, and freshwater sediment on land (Elsahed et al., 2003; Bagwell et al., 2005; Pester et al., 2010). The repeated detection of Cluster IV well into sulfate-depleted marine sediment and freshwater sediment (Harrison et al., 2009; Leloup et al., 2009; Pester et al., 2010) suggests that it is capable of growth under low sulfate concentrations or even in the absence of sulfate.
The dsrAB community at ODP Site 1026 is from an artificial environment, a black rust deposit that was formed by the oxidation of a manmade steel structure during exposure to ridge flank crustal fluids and seawater (Nakagawa et al., 2006; Fig. 2b). Whether this community indeed mirrors sulfate-reducing communities in subseafloor basalt is thus to be determined. Two phylotypes share Ammonifex degensii, a sulfate and nitrate-reducing Chloroflexi species (Huber et al., 1996), as distant cultured relatives (76–78% sequence similarity; Table 1). Another phylotype, which was isolated and shown to reduce sulfate, is most closely related to Desulfotomaculum geothermicum, a Firmicute first isolated from geothermal groundwater (Daumas et al., 1988). The fact that relatives of A. degensii and a high diversity of Firmicutes were also detected in 16S rRNA gene-based clone libraries on fluids and colonization experiments from the same borehole (Cowen et al., 2003; Orcutt et al., 2010a) suggests that the rust deposit at ODP Site 1026 at least partially bears the signature of sulfate-reducing communities in basalt. The other two dsrAB phylotypes, strains Spi55 and Tc37, which were also isolated and shown to reduce sulfate, have Desulfovibrio salexigens and Desulfocella halophila as their closest cultured relatives, respectively (Table 1). The sequence similarity of Spi55 to D. salexigens is low (73%) thus leaving open whether Spi55 is indeed a member of the genus Desulfovibrio. No close relatives of Spi55 or Tc37 occurred in 16S rRNA gene clone libraries on crustal fluids or microcosms from the same borehole (Cowen et al., 2003; Orcutt et al., 2010a), leaving open the possibility of seawater or sedimentary origin.
All published real-time PCR quantifications of dsrA in subseafloor sediments have so far been obtained with the same primer pair (Kondo et al., 2004). Overall, dsrA copy numbers agree with sulfate profiles, showing an exponential decrease with depth on the Peru Margin (ODP Site 1227; Schippers & Neretin, 2006), and an overall decrease with depth across three sites in the Porcupine Seabight (IODP Sites U1316–18; Webster et al., 2009). At certain sites, that is, ODP Site 1229 on the Peru Margin, ODP Site 1230 in the Peru Trench, and IODP Sites 1319–20, 1322, and 1324 on the Gulf of Mexico continental slope, dsrA was detected at too few depths to identify clear depth-related trends (Schippers & Neretin, 2006; Nunoura et al., 2009). Perhaps surprisingly, considering the critical role of sulfate-reducing microbes in terminal organic matter remineralization in marine sediments (Jørgensen, 1982; D'Hondt et al., 2002, 2004) and high percentages of sulfate reducers reported from the coastal subsurface (Leloup et al., 2009), dsrA quantifications in sulfate-rich subseafloor sediments consistently suggest that sulfate reducers represent < 1% of total microbial populations (Schippers & Neretin, 2006; Nunoura et al., 2009; Webster et al., 2009).
The widespread detection of Chloroflexi in subseafloor sediments (Inagaki et al., 2003; Kormas et al., 2003; Parkes et al., 2005; Inagaki et al., 2006a; Reed et al., 2006; also see review by Teske, 2006) has raised questions regarding potential energy sources utilized by this phylum. Considerable 16S rRNA gene sequence similarity to members of the Dehalococcoides, a genus comprised of obligate dehalorespiring bacteria, has led to the profiling and intercomparison of rdhA gene phylogenetic composition across six drilling sites within the Pacific Ocean (Futagami et al., 2009). These include the oligotrophic ODP Site 1226 in the eastern equatorial Pacific, as well as the more energy-rich Site 1227 on the Peru Margin, Site 1230 in the Peru Trench, IODP Site 1301 on the Juan de Fuca Ridge Flank, and sites C9001 off Shimokita and C0002 in the Nankai Trough Forearc Basin (Futagami et al., 2009). Several primer pairs targeting rdhA were tested, with one specifically designed to target Dehalococcoides (Krajmalnik-Brown et al., 2004), leading to successful detections. No significant sequence similarity to cultured strains was found using the megablast function in blast (Table 1). However, all rdhA sequences shared known reductive dehalogenators (Dehalococcoides and Dehalogenimonas spp.) as their closest, albeit phylogenetically distant, cultured relatives based on the discontiguous megablast algorithm in blast. Given that known reductive dehalogenators possess many (up to 32) nonidentical copies of rdhA genes, the extent to which the diversity of subseafloor rdhA genes detected reflects diversity on the organismal level remains to be shown (Hölscher et al., 2004; Wagner et al., 2009).