OUP user menu

Functional genes as markers for sulfur cycling and CO2 fixation in microbial communities of hydrothermal vents of the Logatchev field

Michael Hügler, Andrea Gärtner, Johannes F. Imhoff
DOI: http://dx.doi.org/10.1111/j.1574-6941.2010.00919.x 526-537 First published online: 1 September 2010


Life at deep-sea hydrothermal vents depends on chemolithoautotrophic microorganisms as primary producers mediating the transfer of energy from hydrothermal fluids to higher trophic levels. A comprehensive molecular survey was performed with microbial communities in a mussel patch at the Irina II site of the Logatchev hydrothermal field by combining the analysis of 16S rRNA gene sequences with studies of functional key genes involved in biochemical pathways of sulfur oxidation–reduction (soxB, aprA) and autotrophic carbon fixation (aclB, cbbM, cbbL). Most significantly, major groups of chemoautotrophic sulfur oxidizers in the diffuse fluids differed in their biosynthetic pathways of both carbon fixation and sulfur oxidation. One important component of the community, the Epsilonproteobacteria, has the potential to grow chemoautotrophically by means of the reductive tricarboxylic acid cycle and to gain energy through the oxidation of reduced sulfur compounds using the Sox pathway. The majority of soxB and all retrieved aclB gene sequences were assigned to this group. Another important group in this habitat, the Gammaproteobacteria, may use the adenosine 5′-phosphosulfate pathway and the Calvin–Benson–Bassham cycle, deduced from the presence of aprA and cbbM genes. Hence, two important groups of primary producers at the investigated site might use different pathways for sulfur oxidation and carbon fixation.

  • hydrothermal vent
  • CO2 fixation
  • APS reductase
  • sox enzyme system
  • ATP citrate lyase
  • RubisCO


Deep-sea hydrothermal vent environments represent highly productive ecosystems, fueled solely by a number of reduced inorganic substances (e.g. reduced sulfur compounds, hydrogen or methane) contained in the hydrothermal fluids. Through the oxidation of such compounds, chemolithoautotrophic microorganisms gain energy, which can be used for the fixation of inorganic carbon. Thereby, these microorganisms mediate the transfer of energy from the geothermal source to higher trophic levels and thus form the basis of the unique food chains existing in these environments (for recent reviews, see Kelley et al., 2002; Huber & Holden, 2008). Hydrogen sulfide represents one of the major energy sources used by chemolithoautotrophs and is regarded as a key feature for the development of hydrothermal vent communities (Kelley et al., 2002; Shock & Holland, 2004; Nakagawa & Takai, 2008; Sievert et al., 2008a).

The Logatchev hydrothermal field is an ultramafic-hosted system on the Mid-Atlantic Ridge (MAR), with the Irina II complex as the main structure consisting of a large mound with several black smoker chimneys at the top and mussel fields surrounding the base of the chimneys (Petersen et al., 2009). Previous studies at the Irina II complex examined the microbial community structure of high-temperature fluids and chimney structures based on 16S rRNA gene sequences (Perner et al., 2007a; Voordeckers et al., 2008). Venting of diffuse fluids is an important process at this site. The fluids that reach the seafloor with moderate temperatures provide energy for free-living and symbiotic chemolithoautotrophic bacteria, nourishing the dense mussel and shrimp populations. Furthermore, these fluids provide a window into the subseafloor biosphere that cannot be easily assessed directly (Huber & Holden, 2008).

Microbial communities in vent fluids at different hydrothermal habitats have been studied previously (Reysenbach et al., 2000; Huber et al., 2003, 2007; Brazelton et al., 2006; Perner et al., 2007a, b). However, most of these studies used 16S rRNA gene sequences and therefore could not conclude on the physiological properties and specific biochemical pathways acting in the studied communities. In order to gain an insight into the metabolic potential of vent communities, functional genes need to be studied. Although specific genes involved in nitrogen fixation, methanogenesis, sulfate reduction and carbon fixation have been analyzed (Dhillon et al., 2003, 2005; Teske et al., 2003; Campbell & Cary, 2004; Nakagawa et al., 2004a, b; Nercessian et al., 2005; Moussard et al., 2006; Perner et al., 2007b; Voordeckers et al., 2008), functional gene studies of energy-yielding processes in hydrothermal vent communities remain scarce (Nercessian et al., 2005; Elsaied et al., 2007). Only recently a GeoChip-based study provided deeper insight into the metabolic diversity of microbial communities in vent chimneys (Wang et al., 2009).

Bacterial sulfur oxidation pathways have been studied in a variety of sulfur-oxidizing bacteria over the last few years and the biochemistry behind these pathways is quite complex (for a recent review, see Ghosh & Dam, 2009). Neutrophilic sulfur-oxidizing bacteria use two types of sulfur oxidation pathways: one involving a multienzyme complex catalyzing the complete oxidation of reduced sulfur compounds to sulfate (Sox pathway) (Kelly et al., 1997; Friedrich et al., 2001) and another implementing elemental sulfur and sulfite as intermediates (Pott & Dahl, 1998; Kappler & Dahl, 2001; Ghosh & Dam, 2009). Components of a fully functional Sox complex involve SoxB, SoxXA, SoxYZ and SoxCD. In earlier studies, the SoxB-encoding gene has been used as a marker gene and its presence has been demonstrated in reference strains from the Alpha-, Beta-, Gammaproteobacteria and Chlorobi (Petri et al., 2001; Meyer et al., 2007). The sox genes have also been found within the genomes of certain Epsilonproteobacteria, strongly indicating that they use the Sox pathway for sulfur oxidation (Nakagawa et al., 2007; Sievert et al., 2008b). Only recently initial biochemical evidence for the operation of the Sox system has been reported for the epsilonproteobacterium Sulfurovum spp. NBC37-1 (Yamamoto et al., 2010). The alternative pathway of sulfur oxidation starts with the transformation of sulfide to polysulfide. Sulfur globules are formed and again remobilized by the dissimilatory sulfite reductase, yielding sulfite (Pott & Dahl, 1998). Subsequently, sulfite is either oxidized to sulfate by sulfite acceptor oxidoreductase or adenosine 5′-phosphosulfate (APS) is formed as an intermediate (APS pathway). The APS pathway involves the enzymes APS reductase and ATP sulfurylase (Kappler & Dahl, 2001). Dissimilatory APS reductase (AprBA) is also a key enzyme of the dissimilatory sulfate-reduction pathway in sulfate-reducing prokaryotes (Meyer & Kuever, 2007b). However, the enzyme is found in a wide variety of sulfur-oxidizing prokaryotes, where it is most likely involved in the reverse process, the transformation of sulfite to APS (Meyer & Kuever, 2007a). Primers designed for the amplification of a fragment of the aprA gene allowed the positive aprA amplification in sulfate-reducing as well as sulfur-oxidizing prokaryotes (Blazejak et al., 2006; Meyer & Kuever, 2007a, b).

Because chemolithoautotrophs are the primary producers at deep-sea vent habitats, a few studies have focused on the autotrophic potential of these biotopes (Nakagawa & Takai, 2008 and references therein). Genes encoding key enzymes of two important carbon fixation pathways, namely the Calvin–Benson–Bassham (CBB) cycle [key enzyme: ribulose 1,5-bisphosphate carboxylase/oxygenase (RubisCO) form I and II, genes: cbbL and cbbM, respectively] and the reverse tricarboxylic acid (TCA) cycle (key enzyme: ATP citrate lyase, genes aclA and aclB), have been studied to demonstrate the autotrophic potential of deep-sea vent communities (Campbell et al., 2003; Campbell & Cary, 2004; Moussard et al., 2006; Elsaied et al., 2007; Perner et al., 2007b; Voordeckers et al., 2008).

In this study, comprehensive analyses of the microbial community within hydrothermal fluids have been performed studying 16S rRNA gene sequences as well as sequences of genes for different pathways of carbon fixation (aclB, cbbM, cbbL) and sulfur metabolism (soxB, aprA) within one and the same sample.

Materials and methods

Sampling site, sample collection and fluid characteristics

The Irina II chimney complex in the Logatchev hydrothermal vent field is located at 14°45′N, 44°58′W on the MAR at water depths of about 3000 m (Petersen et al., 2009). About 600 mL of diffuse hydrothermal fluids with a temperature range of 24–43 °C were retrieved from a mussel bed during a dive with the remotely operated vehicle (ROV) Jason II (Woods Hole Oceanographic Institution, MA) during the cruise MSM 04 with R/V Maria S. Merian (Borowski et al., 2007). On board, the fluids were immediately processed. Sulfide and oxygen measurements resulted in 12 and 60 μM, respectively (see Borowski et al., 2007 for further details). As a reference sample, 2 L of seawater was retrieved using a rosette sampler attached to a CTD probe (location: 13°30′N, 45°00′W; depth: 2660 m).

DNA extraction and amplification

For bacterial community analyses, 400 mL of hydrothermal fluids were concentrated on polycarbonate filters (type: GTTP, pore size 0.1 μm, Millipore, Eschborn, Germany) and kept at −20 °C. DNA was extracted from filters using the UltraClean Soil DNA Isolation Kit (MoBio, Solana, CA) according to the manufacturer's instructions. Bacterial 16S rRNA genes were PCR-amplified with PuReTaq Ready-To-Go-PCR Beads (GE Healthcare, Munich, Germany) using the primer pair 27F and 1387R (Lane, 1991; Marchesi et al., 1998). Primers (10 pmol μL−1), 10–100 ng of DNA template and sterile water were added to PuReTaq Ready-To-Go-PCR Beads (GE Healthcare) to a total volume of 25 μL. An initial denaturation step (92 °C for 2 min) was followed by 30 cycles of 92 °C for 40 s, 50 °C for 45 s and 72 °C for 60 s. For amplification of gene fragments of the SoxB component of the periplasmic thiosulfate-oxidizing Sox enzyme complex (soxB) and the α subunit of the APS reductase (aprA), primer sets soxB432F/soxB1446B and aps1F/aps4R and PCR conditions were used as described previously (Petri et al., 2001; Meyer & Kuever, 2007b). A fragment of the β subunit of the ATP citrate lyase genes (aclB) was amplified using the primer set 892F/1204R, applying the conditions as described before (Campbell et al., 2003). For the amplification of fragments of the genes coding for the large subunit of RubisCO cbbL and cbbM (RubisCO form I and form II, respectively), the primer sets cbbLF/cbbLR and cbbMF/cbbMR were applied using the conditions according to Campbell & Cary (2004). Each functional gene fragment was amplified in five parallel PCR reactions, which were subsequently pooled for the construction of the gene libraries.

Cloning and sequencing

The amplified and pooled PCR products were gel-purified using the Qiagen QIAquick gel extraction kit (Qiagen, Hilden, Germany) and cloned into pCR4-TOPO plasmid vectors using the TOPO-TA cloning kit (Invitrogen, Carlsbad, CA) as described by the manufacturer. An environmental clone library for each gene was constructed. From each library, clones were randomly chosen and analyzed for the insert-containing plasmid by direct PCR with the vector primers M13F and M13R, followed by gel electrophoreses of the amplified products. PCR products of the correct size were sequenced using the M13 primer set (see Table 1 for the number of sequenced clones for each library). Sequencing was performed using the bigdye terminator v1.1 sequencing kit in a 3730xl DNA Analyzer (Applied Biosystems, Carlsbad, CA) as specified by the manufacturer, resulting in sequence lengths of ∼1350 bp (16S rRNA genes), ∼1050 bp (soxB), ∼390 bp (aprA), ∼330 bp (cbbM) and ∼340 bp (aclB).

View this table:

Number of 16S rRNA gene and functional gene sequences associated with distinct phylogenetic groups (see Materials and methods for details)

Gene libraryClones sequenced/OTUs (iden.)Total ProteobacteriaOthers
16S rRNA geneClones11150191140
OTUs (97%)531861127
OTUs (99%)672781130
OTUs (94%)10721
OTUs (97%)2612734
OTUs (95.5%)7412
OTUs (97%)1111
  • * Uncertain, no clear assignment could be made.

Phylogenetic analysis

All sequences were edited with chromaspro c.c1.33 and compared with the NCBI database using blast (Altschul et al., 1990). The 16S rRNA gene sequences were aligned with the arb software (http://www.arb-home.de) using the arb fastaligner utility (Ludwig et al., 2004). If not already present, the closest relatives according to blast search were also added to the arb database. The sequence alignment was adjusted manually, taking known secondary structures into account. Amino acid sequences of functional genes were aligned using clustal x (Thompson et al., 1997) and adjusted manually using bioedit (Hall, 1999). Neighbor-joining trees were calculated using paup, version 4.0b10. Maximum-likelihood based trees were constructed using phyml (Guindon & Gascuel, 2003). Bootstrapping included 1000 replicates for neighbor-joining-based trees and 100 replicates for maximum-likelihood-based trees. Calculated 16S rRNA gene trees were reimported into arb and shorter sequences were added using the ‘quick add marked’ tool. Operational taxonomic units (OTUs) were defined based on 97% nucleotide sequence identity for the 16S rRNA gene sequences. OTUs based on 99% nucleotide sequence identity were also calculated (Table 1). In order to define OTUs (referred to as sequence types) for the functional genes (soxB, aprA, cbbM, aclB), inferred amino acid sequences and 16S rRNA gene sequences from isolates possessing the respective genes were used to produce matrices of pairwise distance values (as described by Weber & King, 2010) (see Supporting Information). As a result, sequence types based on 97% (aprA, aclB), 95.5% (cbbM) and 94% (soxB) amino acid sequence identity were defined for the respective functional genes. Rarefaction and rank abundance curves were plotted (see Supporting Information, Figs S1 and S2).

Nucleotide sequence accession numbers

Sequence data have been submitted to EMBL/GenBank/DDBJ databases under accession numbers FN562806FN562916 (16S rRNA genes), FN562666FN562697 (aclB), FN562917FN562931 (cbbM), FN562762FN562805 (soxB) and FN56 2698FN562761 (aprA).


Phylogenetic analyses of the bacterial community

Diffuse fluid from a mussel field at the base of the Irina II structure of the Logatchev hydrothermal field showed a maximum temperature of 43.3 °C at the point of discharge. Sulfide and oxygen concentrations were 12 and 60 μM, respectively (Borowski et al., 2007). The bacterial 16S rRNA gene was amplified and a clone library was constructed to analyze the microbial community structure in these fluids. The majority of the 53 detected OTUs were related to Epsilonproteobacteria (34%), most of which clustered within the marine group 1, also called group F (Table 1, Figs 1 and 2). So far, only a few strains of this highly diverse group have been cultured. The closest cultured relative to our sequences was Sulfurovum lithotrophicum (Inagaki et al., 2004). Three OTUs fell into group D (Nautilia/Caminibacter group) and one into group B (Sulfurimonas group) (classification according to Corre et al., 2001).


Percentage composition of 16S rRNA gene and functional gene (soxB, aprA, cbbM, aclB) libraries. Note that a bacterial 16S rRNA gene library was constructed from DNA isolated from the fluid sample, as well as from DNA isolated from a deep-sea water sample without hydrothermal influence (reference sample).


Phylogenetic trees based on 16S rRNA gene sequences of Proteobacteria. The tree was calculated using the maximum-likelihood method. Bootstrap values are shown as percentages of 100 bootstrap replicates. Sequences obtained in this study are indicated in bold.

Sequences of Gammaproteobacteria (17% of all retrieved sequences) were distributed in six different OTUs (Fig. 2). The most frequently found phylotype showed >99% identity to the 16S rRNA gene sequence of the intranuclear bacterium Candidatus Endonucleobacter bathymodioli of Bathymodiolus spp., also occurring at the Irina II site (Zielinski et al., 2009). This bacterium has been proposed to represent an intranuclear parasite of these mussels (Zielinski et al., 2009). Other sequences cluster with heterotrophic Gammaproteobacteria of the genus Psychromonas or show highest similarities to methanotrophic or methylotrophic isolates, respectively.

Further groups present in the clone library included Bacteroidetes (12 OTUs), Fusobacteria (one OTU), Deltaproteobacteria (one OTU), Alphaproteobacteria (one OTU), Verrucomicrobia (two OTUs), Spirochaeta (one OTU) Mollicutes (one OTU) and Caldithrix (one OTU), as well as candidate divisions RE-1 (three OTUs), SR1 (two OTUs) and OD1 (three OTUs).

A reference clone library of bacterial 16S rRNA gene fragments from the water column nearby showed a significantly different composition. It completely lacked Epsilonproteobacteria and was dominated by Gammaproteobacteria (48%, sequences related to the genera Alteromonas, Halomonas, Chromohalobacter, Marinobacter, Acinetobacter, Alcanivorax) and Alphaproteobacteria (30%, sequences related to the genera Erythrobacter, Sulfitobacter, Citromicrobium, Leisingera, Hirschia), followed by Bacteroidetes (16%) and 5% other groups (data not shown).

Genes involved in sulfur cycling

The genetic potential of microbial sulfur metabolism was analyzed by amplifying key genes of known sulfur oxidation pathways. A gene encoding an essential component of the Sox enzyme system, namely soxB (SoxB, sulfate thiohydrolase), was chosen along with the gene encoding the α subunit of APS reductase (aprA).

The majority of soxB sequences identified (77%) retrieved from the fluid sample were affiliated to those of Epsilonproteobacteria (Figs 1 and 3a) and formed seven distinct sequence types. Among the cultured strains, soxB of Sulfurovum spp. NBC-37-1 was most similar to the majority of these sequences. One sequence type (soxE8) was related to soxB of Sulfurimonas denitrificans and another one (soxG10) to the soxB gene of Arcobacter spp. Notably, the soxB genes of Arcobacter spp. and soxG10 formed a distinct cluster separate from all other soxB sequences of Epsilonproteobacteria (Fig. 3a). Apart from epsilonproteobacterial sequences, two other, novel clusters of soxB genes were identified. One cluster (represented by soxB8) was distantly affiliated with genes of Gammaproteobacteria, and the other (represented by soxB3) with Alphaproteobacteria. In addition, a single sequence (soxF6) was included in the soxB cluster of the Rhodobacter/Roseobacter group of the Alphaproteobacteria.


Phylogenetic trees based on the amino acid sequences of the soxB (a) and aprA (b) genes. The trees were calculated using the maximum-likelihood method. Bootstrap values are shown as percentages of 100 bootstrap replicates. Sequences obtained in this study are indicated in bold.

Based on the topology of the tree, the aprBA genes of sulfur-oxidizing and sulfate-reducing prokaryotes form two phylogenetic lineages (Meyer & Kuever, 2007a). The aprA sequences obtained in this study were highly diverse and sequences of both lineages were found in the aprA clone library (Fig. 3b). Because sulfate-reducing as well as sulfur-oxidizing prokaryotes use APS reductase, the genes of this reversible enzyme can be found in both groups. Yet, the majority of the obtained sequences (77%, 15 sequence types) could be clearly attributed to sulfur-oxidizing bacteria, while 11 sequence types (23% of the obtained sequences) clustered with the aprA gene of sulfate-reducing prokaryotes. Among the sequences assigned to sulfate-reducing prokaryotes, the majority were related to aprA sequences from Deltaproteobacteria (Fig. 3b). Others (aprE7, aprB2, aprA3) showed high similarities to aprA genes of the deeply rooting Thermodesulfobacteria and to an archaeal aprA gene (aprC11). Among aprA sequences attributed to sulfur-oxidizing bacteria, >80% of the sequence types could be assigned to gammaproteobacterial groups. The others clustered with aprA of Pelagibacter spp. (SAR11 group, Alphaproteobacteria). Despite the existence of a large database of aprA sequences from gammaproteobacterial isolates, almost all sequences from the Irina II fluids formed new lineages within the Gammaproteobacteria distinct from those of cultured strains (Fig. 3b). Quite interestingly, these sequences were most similar to a number of aprA sequences of epi- and endosymbionts from vent mussels (Bathymodiolus spp.), vent crabs (Kiwa hirsuta), vent shrimp (Rimicaris exoculata) and oligochaetes (e.g. Olavius algarvensis) (Blazejak et al., 2006; Meyer & Kuever, 2007a; Goffredi et al., 2008; Zbinden et al., 2008).

Genes involved in autotrophic CO2 fixation

In order to gain insights into the potential of the Irina II fluid bacterial community to fix CO2 via the CBB cycle or the reductive TCA cycle, genes encoding key enzymes of these two autotrophic carbon fixation pathways were amplified, including cbbL and cbbM, encoding the large subunit of RubisCO form I and form II, respectively, and the gene for the β subunit of ATP citrate lyase aclB.

A faint cbbM amplification product could be retrieved, while all attempts to amplify a cbbL gene fragment failed despite repeated trials using different DNA template concentrations. The cbbM sequences obtained were related to those of Alpha- and Gammaproteobacteria (Figs 1 and S3), with sequences most similar to those from Thiomicrospira pelophila and Thiomicrospira thermophila (cbbMA11 and cbbMB12, respectively) and the symbiont of Calyptogena magnifica, Candidatus Ruthia magnifica (cbbMB5). The most frequently found sequence type (cbbMA8) was related to cbbM sequences retrieved from tubeworm symbionts. Another sequence type (cbbMB10) was most similar to the cbbM gene of the alphaproteobacterium Rhodobacter sphaeroides. Two sequences (cbbMA10, cbbMB3) were not affiliated to any known cbbM gene of cultured bacterial strains. They form a distinct lineage together with cbbM sequences obtained from other environmental samples (Fig. S3).

A gene fragment encoding the small subunit of ATP citrate lyase (aclB) could be amplified successfully and the construction of the clone library yielded many positive clones. All sequences of the aclB gene were related to those of phylogenetically different groups of Epsilonproteobacteria (Figs 1 and S4). The majority of the aclB sequences (nine sequence types) were associated with several novel clusters, of which the closest related aclB gene retrieved from a pure culture is that of S. lithotrophicum (group F). Other aclB sequences were related to those of Caminibacter mediatlanticus of group D (aclBA6) or Thioreductor micantisoli of group G (aclBE3). The sequence aclBB4 forms a novel lineage together with environmental aclB sequences from high-temperature Irina II fluids and from chimney samples of vents in the Guaymas Basin (Campbell & Cary, 2004).


The microbial community structure within hydrothermal fluids from the Irina II structure of the Logatchev vent field was analyzed using genes involved in different pathways of sulfur metabolism and CO2 fixation as functional markers and the 16S rRNA gene as a phylogenetic marker. For the first time, different key genes of important energy-yielding processes were analyzed together with key genes of different carbon fixation pathways in order to reveal the genetic capabilities of the chemolithoautotrophic primary producers at this deep-sea vent site. The 16S rRNA gene sequences obtained clearly differ from those of a reference sample apart from the hydrothermal fluids and demonstrate overall similarities in the community composition to those of similar previously studied hydrothermal vent habitats. In fact, there is accumulating evidence that Epsilon- as well as Gammaproteobacteria constitute the major part of the bacterial community at the Irina II site and at other deep-sea hydrothermal vent habitats (e.g. Reysenbach et al., 2000; Corre et al., 2001; Campbell et al., 2006; Perner et al., 2007a; Nakagawa & Takai, 2008; Wang et al., 2009).

Epsilonproteobacteria as important primary producers

Phylogenetic analyses of the 16S rRNA gene, aclB and soxB genes demonstrated that Epsilonproteobacteria most likely are an important part of the microbial community at the Irina II site able to oxidize reduced sulfur compounds via the Sox pathway and fixing CO2 via the reductive TCA cycle. During the last few years, the great importance or even dominance of Epsilonproteobacteria has been pointed out in a number of studies using the 16S rRNA gene as a marker (for a review, see Campbell et al., 2006). These studies included free-living bacterial populations in vent fluids, on surfaces of vent structures, or in the shallow subsurface where mixing between vent fluid and ambient sea water occurs (Alain et al., 2004; Kormas et al., 2006; Moussard et al., 2006; Huber et al., 2007). Only a small number of epsilonproteobacterial isolates have been obtained so far. These are able to generate energy through the oxidation of reduced sulfur compounds or hydrogen (for an overview, see Takai et al., 2005; Nakagawa & Takai, 2008; Sievert et al., 2008a). Most of the isolates were chemolithoautotrophs using the reductive TCA cycle for carbon fixation (Hügler et al., 2005; Takai et al., 2005). The first whole-genome sequences of autotrophic free-living epsilonproteobacterial isolates recently became available and provide an insight into their metabolic capabilities (Nakagawa et al., 2007; Sievert et al., 2008b; Campbell et al., 2009).

According to the phylogeny of the 16S rRNA gene Epsilonproteobacteria in the Irina II fluids were members of the groups B, D and F, with the majority of the sequences related to group F. The trees of the functional genes (soxB, aclB) add up very well. The main fraction of amplified soxB genes could be assigned to Epsilonproteobacteria and most of these clustered with Sulfurovum spp. NBC37-1 (group F). A single sequence type of both 16S rRNA gene and soxB genes was affiliated with S. denitrificans (group B). These results demonstrate the good congruence between soxB phylogeny and 16S rRNA gene phylogeny. At the same time, they point toward the potential of Epsilonproteobacteria to oxidize sulfur via the Sox pathway, which so far has been suggested for individual strains only from genomic and initial enzymatic studies (Nakagawa et al., 2007; Sievert et al., 2008a, b; Yamamoto et al., 2010). In this context, it should be mentioned that Epsilonproteobacteria of group D cannot oxidize reduced sulfur compounds, but rely on hydrogen as an energy source. Consequently, their genomes lack the sox genes (Campbell et al., 2009).

Autotrophic Epsilonproteobacteria have been shown to use the reductive TCA cycle for CO2 fixation with a bona fide ATP citrate lyase as the key enzyme (Hügler et al., 2005; Takai et al., 2005; Voordeckers et al., 2008). In this study, various epsilonproteobacterial aclB sequences were found clustering mainly with sequences from groups F and D. This indicates the potential of Epsilonproteobacteria at the Irina II site to fix CO2 via the reductive TCA cycle.

Role of Gammaproteobacteria

Gammaproteobacteria were another abundant group in our clone libraries and represented major fractions of the sequences of 16S rRNA gene, cbbM and aprA genes (Fig. 1). Functional gene analyses revealed their potential to (1) oxidize sulfur mainly via the APS pathway and (2) to fix CO2 via the CBB cycle. A large variety of aprA genes could be assigned to Gammaproteobacteria (Fig. 3b). Interestingly, these genes clustered with aprA genes amplified from diverse epi- and endosymbionts, including mussel symbionts (Meyer & Kuever, 2007a), crab and shrimp epibionts (Goffredi et al., 2008; Zbinden et al., 2008) and symbionts of gutless marine worms (Blazejak et al., 2006). Our results suggest that at the Irina II site oxidation of sulfur via the APS pathway with sulfite as a free intermediate is performed by a variety of Gammaproteobacteria.

Sulfate-reducing bacteria

It is worth mentioning that deltaproteobacterial sequences related to the genus Desulfocapsa were found in the 16S rRNA gene as well as in the aprA gene library (Figs 2 and 3b). Desulfocapsa spp. can grow chemolithoautotrophically as sulfate reducers or by the disproportionation of thiosulfate, sulfite or elemental sulfur (Finster et al., 1998). APS reductase in these bacteria is involved in the dissimilatory sulfate reduction pathway (Meyer & Kuever, 2007b). Because of the different phylogenetic grouping of aprA from sulfur-oxidizing and sulfate-reducing bacteria, a clear distinction of the two groups can be made with respect to environmental samples. AprA genes of sulfur-oxidizing bacteria form APR lineage I, while sequences from sulfate-reducers form lineage II (Fig. 3b). However, the aprA genes of several sulfur-oxidizing bacteria, such as those of Beggiatoa spp., Thiobacillus spp. or the endosymbiont of Riftia pachyptila, cluster within lineage II, indicating lateral gene transfer of the apr genes (Meyer & Kuever, 2007a).

Analyses of the aprA gene fragments revealed that in addition to Deltaproteobacteria, other sulfate-reducing prokaryotes occur at the Irina II site, including thermophilic Thermodesulfobacteria and hyperthermophilic Archaeoglobales (Fig. 3b). As fluid temperatures of up to 43 °C at the point of discharge are unlikely to support the growth of these thermophilic microorganisms, it is anticipated that they thrive in the subsurface, where higher temperatures are reached, and are carried to the seafloor within the fluids.

Ecological implications

By analyzing key genes involved in sulfur cycling and carbon fixation, we have gained an insight into the metabolic reactions driving the biogeochemical cycles in these fluids and the adjacent subsurface habitats and into the organisms involved in these reactions. Epsilonproteobacteria of group F appear to be the most abundant bacterial group in the investigated hydrothermal fluid. Their potential to gain energy from the oxidation of reduced sulfur compounds is supported by a variety of epsilonproteobacterial soxB gene fragments. In addition, as indicated by the diverse aclB gene sequences, most strains have the capability to grow autotrophically by means of the reductive TCA cycle, supporting previous studies at other hydrothermal vent sites (Campbell & Cary, 2004; Moussard et al., 2006; Voordeckers et al., 2008). Notably, investigations of high-temperature fluids of the Irina II site showed quite similar results, with Epsilonproteobacteria related to group F representing the most abundant bacterial clones (Perner et al., 2007a). Taken together, these results indicate that group F Epsilonproteobacteria might be important primary producers within the fluids and in the adjacent subsurface of the Irina II site. Furthermore, studies from other hydrothermal vent sites yielded similar results (e.g. Alain et al., 2004; Kormas et al., 2006; Huber et al., 2007). Why do these organisms succeed in various hydrothermal habitats? Compared with other epsilonproteobacterial isolates, members of the groups F and B can tolerate relatively high oxygen concentrations (Nakagawa et al., 2005), which might be beneficial in ecosystems where intensive mixing of hydrothermal fluid and seawater takes place, such as for example the shallow subsurface. In addition, their versatile metabolism, i.e. growing anaerobically and/or microaerobically with reduced sulfur species and/or hydrogen as energy sources (Takai et al., 2005; Campbell et al., 2006), together with the use of the energy-efficient reductive TCA cycle for autotrophic CO2 fixation (Hügler et al., 2005; Takai et al., 2005), certainly is a competitive advantage for this bacterial group. Consequently, they are perfectly adapted to the steep gradients of sulfide and oxygen found at the Irina II site and in other hydrothermal vent environments.

Although sulfur-oxidizing members of the Gammaproteobacteria have been isolated frequently from vent sites, our knowledge of their sulfur-oxidation pathways is rather limited. Some strains, for example Thiomicrospira crunogena XCL-2, seem to use the complete Sox enzyme system (Scott et al., 2006), while others such as Beggiatoa and the endosymbiont of the tubeworm R. pachyptila use alternative pathways (Hagen & Nelson, 1997; Markert et al., 2007). The huge variety of gammaproteobacterial aprA genes found in this study indicates that Gammaproteobacteria at the Irina II site mainly use the APS pathway for sulfur oxidation. At least some strains also have the capability to use the CBB cycle for carbon fixation because we found several gammaproteobacterial cbbM gene sequences, encoding RubisCO form II (Table 1, Fig. S3). It has been postulated that RubisCO form II is used in niches with high CO2 and low O2 levels, while RubisCO form I works best in a low-CO2 and high-O2 environment (Badger & Bek, 2008). In line with this hypothesis, we were only able to amplify the cbbM gene. Thus, the organisms present in the Irina II fluids seem to be well adapted to the prevailing environmental conditions.


Based on the results obtained in this study, we propose that Epsilonproteobacteria are of major importance as microbial players in sulfur oxidation and CO2 fixation in the shallow subsurface and hydrothermal fluids of the Irina II vent site. Using the energetically cheap reductive TCA cycle (less than half of the energy demand for CO2 fixation compared with the CBB cycle) as the major route of CO2 fixation and the Sox pathway for sulfur oxidation, they have a distinct advantage over Gammaproteobacteria, which use the CBB cycle for CO2 fixation and the APS pathway for the oxidation of reduced sulfur compounds. Notably, our results indicate that the two important groups of primary producers present at the investigated site, the Epsilon- and the Gammaproteobacteria, may use different pathways for sulfur oxidation and carbon fixation.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Fig. S1. Analytical rarefaction curve plotted for the 16S rRNA gene clone library.

Fig. S2. Rank abundance plots for 16S rRNA gene, soxB, aprA, cbbM and aclB gene libraries.

Fig. S3. Phylogenetic trees based on the amino acid sequences of the cbbM gene.

Fig. S4. Phylogenetic trees based on the amino acid sequences of the aclB gene.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


We would like to thank the captain and the crews of the R/V Maria S. Merian and the ROV Jason II (Woods Hole Oceanographic Institution, MA) for help with the sampling effort during the MSM 04/3 cruise (chief scientist C. Borowski). Special thanks are due to Mirjam Perner (IFM-GEOMAR, now at the University of Hamburg) for sample processing on board and DNA extraction. We thank two anonymous reviewers for their insightful suggestions. The work was supported by grants from the priority program 1144 ‘From Mantle to Ocean: Energy-, Material- and Life-cycles at Spreading Axes’ of the Deutsche Forschungsgemeinschaft.


  • Editor: Gary King


View Abstract