High abundance of JS-1- and Chloroflexi-related Bacteria in deeply buried marine sediments revealed by quantitative, real-time PCR

Authors


  • Editor: Patricia Sobecky

Correspondence: Anna Blazejak, Bundesanstalt für Geowissenschaften und Rohstoffe, Stilleweg 2, 30655 Hannover, Germany. Tel.: +49 0 511 643 3103; fax: +49 0 511 643 2304; e-mail: Anna.Blazejak@gmx.de

Abstract

Sequences of members of the bacterial candidate division JS-1 and the classes Anaerolineae and Caldilineae of the phylum Chloroflexi are frequently found in 16S rRNA gene clone libraries obtained from marine sediments. Using a newly designed quantitative, real-time PCR assay, these bacterial groups were jointly quantified in samples from near-surface and deeply buried marine sediments from the Peru margin, the Black Sea, and a forearc basin off the island of Sumatra. In near-surface sediments, sequences of the JS-1 as well as Anaerolineae- and Caldilineae-related Bacteria were quantified with significantly lower 16S rRNA gene copy numbers than the sequences of total Bacteria. In contrast, in deeply buried sediments below approximately 1 m depth, similar quantities of the 16S rRNA gene copies of these specific groups and Bacteria were found. This finding indicates that JS-1 and Anaerolineae- and Caldilineae-related Bacteria might dominate the bacterial community in deeply buried marine sediments and thus seem to play an important ecological role in the deep biosphere.

Introduction

A particular interest of deep biosphere research is to gain insights into the microbial community structure of marine subsurface sediments. Previous studies have shown that microbial life in this environment is widespread and represents one-tenth to one-third of all prokaryotic biomass on Earth (Whitman et al., 1998; Jørgensen & Boetius, 2007; Fry et al., 2008). In sediments up to a depth of 1600 meters below seafloor (m b.s.f.), at least 105 intact microbial cells cm−3 (Parkes et al., 2000; D'Hondt et al., 2004; Schippers et al., 2005; Roussel et al., 2008) and intact polar lipids (Biddle et al., 2006; Lipp et al., 2008) were found, evidence for living prokaryotic communities. To investigate the microbial community structure, molecular techniques based on the 16S rRNA gene have been applied showing a diverse microbial community with members of distinct, uncultured bacterial and archaeal lineages (Teske, 2006; Fry et al., 2008).

A comparison of the 16S rRNA gene clone libraries obtained from deeply buried sediments shows that among bacterial 16S rRNA gene clone sequences, in particular, sequences belonging to the JS-1 group and the Chloroflexi were frequently found (Teske, 2006). Sequences of the JS-1 group were firstly identified in sediments from the Japan Sea, giving the group its name (Rochelle et al., 1994). Meanwhile, the JS-1 group has been found to occur widespread in different deeply buried and shallow marine sediments as well as brackish and freshwater sediments (Webster et al., 2004, 2007; Teske, 2006; Fry et al., 2008). However, representatives of this group have not been cultivated as yet, and thus their metabolism remains unexplored. So far, only indirect evidence from stable-isotope probing exists suggesting that members of this clade are able to incorporate acetate and glucose (or glucose metabolite) under anaerobic, sulfate-reducing conditions (Webster et al., 2006).

The phylum Chloroflexi, the other well-represented bacterial group in sediments, has been recognized as a typical bacterial cluster containing a number of diverse environmental clones (Hugenholtz et al., 1998). This phylum is a deep-branching lineage of the Bacteria and can be divided into at least six major classes: Chloroflexi, Thermomicrobia, Anaerolineae, Caldilineae, ‘Dehalococcoidetes’, and a clone cluster called subphylum IV [SAR202 cluster (Morris et al., 2004)] (Rappe & Giovannoni, 2003; Hugenholtz & Stackebrandt, 2004). The 16S rRNA gene sequences obtained from subsurface sediments are frequently present within the classes Anaerolineae, Caldilineae, ‘Dehalococcoidetes’, and the unclassified subphylum IV (Coolen et al., 2002; Reed et al., 2002; Inagaki et al., 2003; Kormas et al., 2003; Parkes et al., 2005; Inagaki et al., 2006). Within these classes, a few cultured representatives are present, and thus conclusions about the metabolic nature of the related sequences remain speculative because a close relationship based on the 16S rRNA gene does not necessarily mean the same metabolism.

Despite the worldwide recognition of Bacteria from the JS-1 group and the phylum Chloroflexi in subsurface sediments, nothing is known about their abundance and therefore their potential ecological role in subsurface organic matter degradation. In this study, we have developed a new quantitative real-time PCR (Q-PCR) assay to quantify members of the JS-1 group and Chloroflexi-related Bacteria of the classes Anaerolineae and Caldilineae in marine sediments based on the detection of their 16S rRNA genes. Using this technique, we analyzed sediments from three different locations: near-surface and deeply buried sediments up to 121 m b.s.f. from the Peru continental margin and sediments with depths up to 12 m b.s.f. from the Black Sea and from a forearc basin off the Indonesian island Sumatra.

Materials and methods

Sample collection

Sediment samples were collected during several expeditions: on the Peru margin in June 2000 [Site 2, cruise SO147 of R/V Sonne (Kudrass, 2000)] and in March 2002 [Site 1227, Ocean Drilling Program (ODP) Leg 201 (D'Hondt et al., 2003; Jørgensen et al., 2005)], in the Black Sea in May 2007 [Site 20, cruise M72-5 of R/V Meteor (2007)], and in a forearc basin off the south-west coast of the Indonesian island Sumatra in September 2006 [Sites 3 and 6, cruise SO189-2 of R/V Sonne (Wiedicke-Hombach, 2006)] (Table 1). The Peruvian Sites 2 and 147 were not close to each other, but both were located on the upper slope of the continental margin in the coastal upwelling area with high input of organic carbon and sedimentation rates, thus providing similar conditions for microbial life (Schippers & Neretin, 2006). From all sites, samples for molecular analysis were taken aseptically from the center of the cores at all sites and were stored at least at −20 °C until further processing in the laboratory.

Table 1.   Technical data of the sampling sites from the Peru margin (ODP Leg 201 and SO147), a forearc basin off Sumatra (SO189-2), and the Black Sea (M72-5)
Cruise/legSitePositionWater depth (m)Sampling depth
(m b.s.f.)
Number of
analyzed samples
  1. MC, multicorer; PC, piston corer; GC, gravity corer.

ODP Leg 20112278°59.5′S 79°57.4′W42715122
SO1472 MC11°35.0′S 77°33.1′W860.3419
SO189-23 MC4°41.871′S 101°57.432′E17070.4812
6 PC4°41.859′S 101°57.479′E17101225
M72-520 GC43°57.25′N 35°38.46′E20485.819

For the recovery of deeply buried sediments from Site 1227 on the Peru margin, seawater-based drilling fluid was used. Thus, a potential contamination with seawater microorganisms was routinely checked by application of fluorescent beads of prokaryotic cell size and a chemical tracer (D'Hondt et al., 2003). Only uncontaminated samples were used for further analysis.

DNA extraction

DNA was isolated from 0.5 to 4 g sediment of various depths using a FastDNA® Spin for Soil Kit (MP Biomedicals, Solon, OH) with the following modification: to increase the yield of isolated DNA from clayish sediments, 200 μg polyadenylic acid (Roche Diagnostics GmbH, Mannheim, Germany) dissolved in sterile water was added to the sample in the first step of the extraction procedure (Webster et al., 2003). DNA extracts from blank tubes (no sediment added) were used as a procedural contamination control in later PCR analyses. Isolated DNA was stored in aliquots to avoid multiple defrosting and freezing and was thawed for Q-PCR measurements not more than twice.

Q-PCR measurements

Q-PCR measurements were run on an ABI Prism 7000 detection system (Applied Biosystems, Foster City, CA). Quantification of Bacteria in total was performed using a previously described Q-PCR assay based on the detection of the 16S rRNA gene (Nadkarni et al., 2002). This assay was carried out using the TaqMan® PCR Master Mix (Applied Biosystems). To quantify JS-1 and Chloroflexi-related Bacteria, a new Q-PCR assay was developed. Because our attempts to develop a Q-PCR assay covering specifically only all known 16S rRNA genes of JS-1 were not successful, we decided to develop a joint Q-PCR assay for both important bacterial phyla in marine sediment: JS-1 and Chloroflexi. For amplification of the 16S rRNA gene of these two bacterial groups, the forward primer 519F 5′-CAG CAG CCG CGG TAA YAC-3′, designed with the arb software (http://www.arb-home.de), and the reverse primer 655R (Webster et al., 2004) were used. The forward primer was developed to obtain a short PCR product of approximately 165 bp ensuring high amplification efficiency of the DNA that is required for the Q-PCR method. This primer matches perfectly to the 16S rRNA gene of these two bacterial groups, but is not specific to them. The specificity of the amplification is achieved by the reverse primer that is specific to the JS-1 Bacteria (Webster et al., 2004) and to members of Chloroflexi-related Bacteria of the classes Anaerolineae and Caldilineae. Prior quantifications, optimal primer concentration (in a range from 50 to 500 nM), and optimal template concentration were determined. The Q-PCR assay was performed using qPCR Master Mix for SYBR® Green I (Eurogentec, Seraing, Belgium) or Platinum® SYBR® Green qPCR SuperMix-UDG with ROX (Invitrogen, Carlsberg, CA), a primer concentration of 400 nM, and the following amplification conditions: 95 °C for 10 min and 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Two microliters of sample DNA was added to a PCR reaction with a total volume of 25 μL. Melting-curve analyses were run after each assay to check PCR specificity. For amplification standards, DNA was extracted, amplified, and purified from minipreps of cloned 16S rRNA gene sequences of the JS-1 group with the accession numbers AB177301.1 and AB177250.

Real-time data analysis

Relative standards were prepared by serial dilution (1 : 10) of the PCR product. For each standard, the concentration was plotted against the cycle number at which the fluorescence signal increased above the background or the cycle threshold (Ct value). The slope of each calibration curve was included in the following equation to determine the efficiency of the PCR reaction: efficiency=10(−1/slope)−1. According to this formula, an efficiency of 100% means a doubling of the product in each cycle. Data evaluation was performed using the software stepone v2.0 (Applied Biosystems).

Specificity tests with pure cultures

Specific amplification of the used primer pair to the 16S rRNA gene of JS-1 and Chloroflexi-related Bacteria was checked with DNA isolated from Bacteria in a pure culture of different phyla. DNA was isolated from Escherichia coli and Acidithiobacillus caldus DSM 9466 of Gammaproteobacteria, Leptospirillum ferrooxidans DSM 9468 of the phylum Nitrospira, and Sulfobacillus acidophilus DSM 10332 of the phylum Firmicutes. Amplification of the 16S rRNA gene was carried out under the same PCR conditions as those used for the newly developed Q-PCR assay. Further amplification tests with Bacteria of other phyla were also performed previously (Webster et al., 2004).

PCR amplification, cloning, and sequencing from environmental samples

DNA was isolated from sediment samples of the Peru margin from four different depths, 0.04 and 0.34 m b.s.f. (Site 2, SO147) as well as 3.6 and 6.6 m b.s.f. (Site 1227, ODP Leg 201). Amplification of the 16S rRNA gene was carried out under the same conditions as those for the newly developed Q-PCR assay. PCR products obtained from each depth were purified and subsequently cloned using the pGEM®-T Easy vector system (Promega, Madison, WI) and TOP10 chemo-competent cells (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Clones were randomly picked, resuspended in PCR-grade water, and selected for the correct insert size by PCR with vector primers. Approximately 55 positive clones per depth were sequenced with the vector primer M13 Forward. Sequencing reactions were run using ABI BigDye on an ABI Prism 3100 genetic analyzer (Applied Biosystems).

Phylogenetic analysis

Sequences were checked with blast (Altschul et al., 1997; Tatusova & Madden, 1999) for similarity searches. For sequence alignment and phylogenetic tree reconstruction, sequences were analyzed using the bioedit program (http://www.mbio.ncsu.edu/BioEdit/bioedit.html), the software arb (http://www.arb-home.de), and the rRNA database SILVA SSU Ref, release_94 July 2008. Briefly, after cutting of the vector sequence, sequences were aligned and added to a neighbor-joining 16S rRNA gene tree with Jukes Cantor and Felsenstein correction.

Nucleotide accession numbers

The rRNA gene sequences obtained in this study were submitted to the DDBJ/EMBL/GenBank nucleotide databases under the accession numbers GQ403530GQ403614.

Results and discussion

In this study, a new Q-PCR assay for the quantification of Bacteria belonging to the JS-1 group and Chloroflexi-related Bacteria of the classes Anaerolineae and Caldilineae was developed and applied to samples from different marine sediment sites including samples from the deep biosphere. Results for the evaluation of the assay including specificity tests are presented before quantitative data for the environmental samples are shown.

Evaluation of the Q-PCR assay

Amplification quantities of the standard (PCR product of the 16S rRNA gene from JS-1 clone sequence AB177301.1) ranged from 2.28 × 103 to 2.28 × 108 molecules, with a correlation coefficient of 0.999 (Supporting Information, Fig. S1a). The efficiency of the PCR reactions was 95%. Detection of contaminant DNA was observed with quantities of up to 10 molecules per reaction (Fig. S1a). Such a contamination is probably caused by contaminant genomic DNA that is usually present in polymerase preparations and can lead to significant fluorescence in the no-template controls after about 35 cycles (Labrenz et al., 2004). In our experiments, the detection limit was set to quantities of 2.28 × 103 molecules; however, this can be reduced to quantities of 2.28 × 102, still ensuring reliable detection values due to an order of magnitude higher copy numbers than compared with the detection of the contaminated DNA. In general, detection limits for gene quantification by PCR can range between 10 and 200 copies per reaction. For example, Baldwin et al. (2003) detected oxygenase genes down to a detection limit of 200 copies using a SybrGreen assay, and Kolb et al. (2003) detected 10–100 copies of the pmoA gene from methanotrophs in soil samples.

Depending on which of the two master mixes, qPCR Master Mix for SYBR® Green I (MM1, Eurogentec) or Platinum® SYBR® Green qPCR SuperMix-UDG with ROX (MM2, Invitrogen, Calsberg, CA), was used for the Q-PCR assay, differences regarding the sensitivity of the DNA detection and the maximum temperature of the melting curve were observed. The observed difference of the melting temperature using different master mixes is not unusual. The melting temperature is based on the length and the GC content of the sequence and the concentrations of salt and the DNA intercalating fluorophore such as SybrGreen. Consequently, different chemical compositions of the master mixes can explain a shift in the melting temperature of up to a few degrees. The master mix MM1 was one threshold cycle (Ct value) more sensitive than the master mix MM2. The maximum temperature of the melting-curve was 78 °C for the master mix MM1 and 81 °C for the master mix MM2 (Fig. S1b). However, independent of the type of master mix, melting-curve analysis that was run after each assay always showed a single peak, proving the specificity of the PCR amplification.

The specificity tests in pure cultures, using the primer pair 519F/655R of the Q-PCR assay, did not show any amplification of DNA from cultures of E. coli and A. caldus DSM 9466 of the Gammaproteobacteria, L. ferrooxidans DSM 9468 of the phylum Nitrospira, and S. acidophilus DSM 10332 of the phylum Firmicutes.

For further specificity tests, environmental samples were analyzed for the sequence of the amplification product. Amplification of a single PCR product with an expected length of about 165 bp was obtained from DNA extracts of four sediment samples taken on the Peru margin from depths of 0.04, 0.34, 3.6, and 6.6 m b.s.f. To test the specificity of the DNA amplification using the primer pair 519F/655R, cloning of the PCR products was performed. Up to 54 positive clones per particular depth were analyzed. Because of the short length of the sequences (165 bp), similarity searches with blast (Altschul et al., 1997; Tatusova & Madden, 1999) were not reliable. However, their explicit phylogenetic relationship could be determined using the arb software platform. Two hundred and thirteen of 218 isolated sequences were assigned either to Bacteria of the JS-1 group (Fig. 1a) or to the classes Anaerolineae and Caldilineae of the phylum Chloroflexi (Fig. 1b). To show the relationship between these bacterial groups with other bacterial groups, an overview tree is demonstrated in Fig. 1c. Only five clones obtained for the surface sample were considered as other bacterial sequences (affiliated to the class Fusobacteria). Sequences of the JS-1 group were found in 4–7% of the screened clones, whereas sequences of Anaerolineae and Caldilineae represented up to 96% of all clones (Table 2). This phylotype distribution of the 16S rRNA gene sequences indicated that members of the Anaerolineae and Caldilineae from the phylum Chloroflexi appeared to be more abundant than Bacteria of the JS-1 group in the analyzed sediments from the Peru margin. The classes Anaerolineae and Caldilineae contain a vast number of environmental 16S rRNA gene sequences found frequently within various habitats, which argues for broad physiological properties of these organisms. Although these microorganisms have been thought to be ubiquitous and seem to play an important role in these ecosystems, only six closely related strains from the class Anaerolineae and one strain from the class Caldilineae could be isolated so far (Sekiguchi et al., 2003; Yamada et al., 2006, 2007). All these strains have in common that they are fermentative heterotrophs. However, it is questionable whether the related Bacteria whose sequences were isolated from different environments have the same metabolism because a close relationship based on the 16S rRNA gene does not necessarily mean similar metabolic properties.

Figure 1.

Figure 1.

 Phylogenetic relationship of the 16S rRNA gene sequences for the bacterial candidate division JS-1 (a) and the phylum Chloroflexi (b) derived from Peru continental margin sediments as determined by neighbor-joining analysis. Sequences obtained in this study are marked in bold. Numbers within the clusters represent the number of isolated sequences belonging to the same cluster. Bar, five nucleotide substitutions per 100 nucleotides. Numbers in brackets to the right demonstrate class-level lineages in the phylum Chloroflexi as suggested by Yamada et al. (2006). An overview tree of Bacteria showing the relationship between the bacterial group JS-1 and phylum Chloroflexi and other bacterial groups. Bar, 10 nucleotide substitutions per 100 nucleotides (c).

Figure 1.

Figure 1.

 Phylogenetic relationship of the 16S rRNA gene sequences for the bacterial candidate division JS-1 (a) and the phylum Chloroflexi (b) derived from Peru continental margin sediments as determined by neighbor-joining analysis. Sequences obtained in this study are marked in bold. Numbers within the clusters represent the number of isolated sequences belonging to the same cluster. Bar, five nucleotide substitutions per 100 nucleotides. Numbers in brackets to the right demonstrate class-level lineages in the phylum Chloroflexi as suggested by Yamada et al. (2006). An overview tree of Bacteria showing the relationship between the bacterial group JS-1 and phylum Chloroflexi and other bacterial groups. Bar, 10 nucleotide substitutions per 100 nucleotides (c).

Table 2.   Summary of 16S rRNA gene clone library data for four sediment samples
Station/depth
(m b.s.f.)
Number of clones (% of total)
TotalJS-1
group
Anaerolineae
and Caldilineae
Other bacterial
sequences
Site 2, SO147
 0.04 m b.s.f.54049 (91%)5 (9%)
 0.34 m b.s.f.532 (4%)51 (96%)0
Site 1227, ODP Leg 201
 3.6 m b.s.f.544 (7%)50 (93%)0
 6.6 m b.s.f.523 (6%)49 (94%)0

In summary, the PCR amplification tests with the primer pair 519F/655R on DNA extracts from pure cultures of Bacteria of different phyla and from environmental sediment samples show that they can be considered specific to the 16S rRNA gene of the JS-1 and Chloroflexi-related Bacteria of the classes Anaerolineae and Caldilineae.

Quantification of the JS-1 and Chloroflexi-related Bacteria of the classes Anaerolineae and Caldilineae in environmental samples

The newly developed real-time PCR assay was applied to sediment samples taken from near-surface and deeply buried sediments. The efficiency of the PCR reactions was 95%. Melting-curve analysis, run after each assay, revealed a unique peak, ensuring the specificity of the PCR amplification. The abundance of the JS-1 and Anaerolineae and Caldilineae-related Bacteria ranged from 103 to 109 gene copies per gram sediment in marine sediments from the Peru continental margin, in the forearc basin off Sumatra, and in the Black Sea (Fig. 2).

Figure 2.

 DNA copy numbers of the 16S rRNA genes from the total Bacteria and JS-1-, Anaerolineae-, and Caldilineae-related Bacteria in sediments from the Peru margin (ODP Leg 201 and SO147), a forearc basin off Sumatra (SO189-2), and the Black Sea (M72-5). ▪, Bacteria; × , JS-1-, Anaerolineae-, and Caldilineae-related Bacteria.

For the Peru margin, two sites with different sediment depths were investigated: near-surface sediments with a depth of 0–0.35 m b.s.f. (Site 2, SO147) and deeply buried sediments with a depth of 0–121.4 m b.s.f. (Site 1227, ODP Leg 201). These sites are located on the Peru continental shelf that is characterized by organic-rich ocean-margin sediments. Detailed geochemical and geological data are described by Böning et al. (2004) for Site 2 (SO147) and by D'Hondt et al. (2003, 2004) for Site 1227 (Leg 201). In the near-surface sediments, quantification of the 16S rRNA gene from the bacterial groups JS-1, Anaerolineae, Caldilineae, and for the Bacteria in total showed almost similar depth profiles, showing a decrease of the 16S rRNA gene copies with depth. The JS-1-, Anaerolineae-, and Caldilineae-related Bacteria were detected with 107–108 gene copies per gram sediment, which accounted for 5–10% of the total Bacteria. This significant difference in the measured quantities indicates that in upper sediment layers (top 35 cm) at this station, the bacterial groups JS-1, Anaerolineae, and Caldilineae were present in high numbers, but they were, however, not a dominant group within the total Bacteria. Other bacterial groups that were quantified previously in the sediments from the same site were sulfate-reducing Bacteria and Fe(III)- and Mn(IV)-reducing Bacteria of the family Geobacteraceae (Schippers & Neretin, 2006). Both bacterial groups were present between 0.5% and 2% of the total Bacteria, thus also reflecting only a minor part of the bacterial community. Apparently, bacterial groups other than sulfate-reducers, Fe(III)- and Mn(IV)-reducers of the family Geobacteraceae and Bacteria belonging to the JS-1, Anaerolineae, and Caldilineae might be dominant in the investigated near-surface sediment on the Peru margin. Another possibility could be that no bacterial group is dominant, but multiple bacterial groups can co-occur in similar quantities in these sediments.

In contrast to the near-surface sediments from the Peru margin, in the deeply buried sediments from the same area (Site 1227, ODP Leg 201), the gene copy numbers of JS-1-, Anaerolineae-, and Caldilineae-related Bacteria and the total Bacteria were detected in similar values with a nearly identical depth gradient (Fig. 2). Gene copy numbers of the bacterial groups JS-1, Anaerolineae, and Caldilineae and the total Bacteria decreased strongly within the upper 9 m b.s.f. of the sediment from 2.0 × 107 to 1.3 × 106 and 2.7 × 107 to 1.9 × 106 copies g−1, respectively. After a further decline to about 106 gene copies g−1 in the adjacent sediment layers (9–45 m b.s.f.), again a strong decrease between 45 and 55 m b.s.f. was noted. Detection of the total Bacteria declined to the order of 104 copies g−1 and remained within this range of magnitude up to 121 m b.s.f. Gene copy numbers of the bacterial groups JS-1, Anaerolineae, and Caldilineae also decreased to the order of 104 DNA copies g−1 at 55 m b.s.f., but declined further to the order of 103 gene copies g−1 at the depth of 75 m b.s.f. and were not detectable below 102 m b.s.f. Except for depths below 75 m b.s.f., the 16S rRNA gene from the bacterial groups JS-1, Anaerolineae, and Caldilineae was detected in most samples, with only slightly lower than or similar copy numbers to those of the total Bacteria, indicating that these specific groups might dominate these sediments. A high abundance of Bacteria affiliated to Chloroflexi including the classes Anaerolineae and Caldilineae and to the uncultured group JS-1 in subsurface sediments from the Peru margin (Site 1227) was also observed by 16S rRNA gene clone library analyses (Inagaki et al., 2006). In sediment layers up to 50 m b.s.f., nearly 90% of the clones were associated either with Chloroflexi or the uncultured group JS-1. Because no 16S rRNA gene clone library data for sediment layers below 50 m b.s.f. are available, it remained unknown whether both bacterial groups would also be frequently present in clone libraries from the deeper sediments. Fe(III)- and Mn(IV)-reducing Bacteria of the family Geobacteraceae and sulfate-reducing Bacteria, other bacterial groups that were quantified previously in the same sediments from the Peru margin (Site 1227), could be detected in only a limited number of samples (Schippers & Neretin, 2006). Their gene copy numbers showed a significantly lower abundance (approximately ≤1%) than the gene copy numbers of total Bacteria, suggesting that these bacterial groups are only a minor part of the bacterial community in the deeply buried sediment as well in the near-surface sediments from the Peru margin. Based on the molecular technique, Q-PCR, estimation of the 16S rRNA genes from the bacterial groups JS-1, Anaerolineae, and Caldilineae in this study argue for their dominance in these sediments.

To investigate the abundance of the bacterial groups JS-1, Anaerolineae, and Caldilineae in marine sediments from other locations, two further marine habitats, a forearc basin off Sumatra and the Black Sea, were investigated (Table 1). In the southern Bengkulu basin off Sumatra, a sediment core of approximately 12 m length was analyzed (R/V Sonne 189-2) (Wiedicke-Hombach, 2006). In these sediments, the quantification of the 16S rRNA gene from JS-1-, Anaerolineae-, and Caldilineae-related Bacteria and the Bacteria in total showed similar depth profiles (Fig. 2). Within the first 0.5 m b.s.f, a strong decrease from almost 109 to 107 gene copies g−1 for these groups was observed. The copy numbers remained around this number up to the depth of 12 m b.s.f. Except for the top 0.3 m b.s.f. and sediment layers between 6 and 9 m b.s.f., where JS-1-, Anaerolineae-, and Caldilineae-affiliated Bacteria were detected with copy numbers of up to an order of magnitude lower than the Bacteria, the gene copy numbers of the specific groups and the Bacteria occurred in almost identical amounts. Similar results were obtained for two additional sites in sediments off Sumatra (data not shown). A recently performed microbial diversity study using denaturing gradient gel electrophoresis of sediments from the northern Simeulue forearc basin off Sumatra showed that members of the JS-1 group were highly abundant in the bacterial community (M. Siegert, M. Krüger, B.M.A. Teichert & A. Schippers, unpublished data).

In the Black Sea, east of Crimea, a sediment core of about 6 m length was investigated, taken during the R/V Meteor cruise M72-5 (Hamburg University, 2007). At this station, similar depth profiles of gene copy numbers as for the other marine areas could be obtained showing a strong decrease from about 108 to 106 gene copies g−1 in upper sediment layers (0–0.5 m b.s.f.), which was followed by a slight decline to 105 gene copies g−1 in deeper sediments (Fig. 2). In the near-surface sediments, the JS-1-, Anaerolineae-, and Caldilineae-related Bacteria were detected with 16S rRNA gene copy numbers of up to one order of magnitude lower than the Bacteria in total. In sediments below 0.5 m b.s.f., no clear depth trend was noticeable. Both groups occurred in similar numbers and fluctuated between 105 and 107 gene copies g−1.

The present study demonstrates that quantification of the JS-1- Anaerolineae-, and Caldilineae-related Bacteria using a newly developed Q-PCR assay could be successfully applied to marine sediments from different areas. In near-surface sediments of the Peru margin, 16S rRNA gene copy numbers from these Bacteria were identified, with 5–10% of total Bacteria, indicating that they are present in high numbers, but do not dominate the bacterial community. In contrast, in deeply buried sediments JS-1-, Anaerolineae-, and Caldilineae-related Bacteria reached an abundance similar to the total Bacteria, which implied that these specific groups were the major component of the microbial community. A similar ratio of quantities was also observed in marine sediments from the Black Sea and forearc basins off Sumatra, suggesting that JS-1-, Anaerolineae-, and Caldilineae-related Bacteria could play a dominant role in the marine deep biosphere in general.

Acknowledgements

We thank the ODP Leg 201, R/V Meteor M72-5, R/V Sonne SO147, and SO189-2 personnel and especially Gerrit Köweker and Daniela Zoch for sediment sampling. Special thanks are due to Rudolf Amann for providing laboratory space for cloning and sequencing, to Takuro Nunoura for providing JS-1 clones, and to Christine Beardsley for critical comments on the manuscript. This research used samples and data provided by the ODP, which is sponsored by the US National Science Foundation (NSF) and participating countries under management of Joint Oceanographic Institutions (JOI) Inc. This work was supported by the German Research Foundation (DFG) priority program IODP/ODP, grant SCHI 535/5 to A.S.

Ancillary