INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Anaerobic methane oxidation is a process that effectively controls emission of methane from many anaerobic environments into the atmosphere and thus plays an important role in the global methane budget (2, 35). Three lines of evidence support the hypothesis that biologically mediated methane oxidation occurs under anaerobic conditions: geochemical modeling, rate measurements obtained with radioactive tracers, and changes in stable carbon isotope ratios for methane and carbon dioxide. The geochemical models indicate that a methane-consuming process is required to explain the concave methane concentration profile found in anoxic layers of marine sediments (27, 34, 43). Radioisotope tracer experiments with ¹⁴CH₄ and ³⁵SO₄² have shown that maximum anaerobic methane oxidation rates coincide with local maximum rates of sulfate reduction (2, 7, 13, 16, 17, 19-21) according to the net chemical reaction equation:

(1)

Methane produced by biological methanogenesis is strongly enriched for the light ¹²C carbon isotope and depleted for the heavy ¹³C isotope. The stable carbon isotope composition of organic compounds in the methane-sulfate transition zone has been used to determine if the compounds were derived from oxidation of isotopically light methane (i.e., the ¹²C-enriched methane produced by methanogenic bacteria). Unusually light CO₂ was found in the transition zone together with light ¹³C-depleted lipids that were presumably also formed from light methane (1, 3, 5, 23, 31, 44).

Despite numerous isolation attempts by several scientific research groups, the organisms responsible for anaerobic methane oxidation have not been identified yet, and the mechanism remains unknown. Two scenarios for anaerobic methane oxidation have been proposed: (i) a single sulfate-reducing bacterium and (ii) a consortium of different bacteria. The consortium hypothesis assumes that methane is oxidized by an unknown bacterium in association with a sulfate reducer (7, 10, 13, 15, 17, 46). It is presumed that an unknown compound is used as an electron shuttle between the two organisms to carry electron equivalents from the methane-oxidizing organism to the sulfate-reducing organism. One potential carrier that has frequently been suggested is hydrogen. In this case the equations become:

(2)

(3)

The consortium hypothesis is supported by the results of several studies in which the effects of molybdate, 2-bromoethanesulfonic acid, and fluoroacetate, which inhibit specific groups of bacteria, were monitored (2, 13, 46). Cultures of methanogenic bacteria have been shown to oxidize trace amounts of methane during growth (14, 45). Furthermore, recent studies have shown that ¹³C in lipid biomarkers for methanogenic archaea was depleted (10, 15, 32). In addition, a molecular investigation using restriction fragment length polymorphism and phylogenetic analysis of 16S rRNA indicated that methane was consumed by a phylogenetically distinct group of archaea (15). For these reasons it has been suggested that methanogenic bacteria are part of the consortium. A recent study by Boetius et al. (4) provided microscopic evidence that a consortium of archaea and sulfate-reducing bacteria is present in gas-hydrate-rich sediment. Laboratory experiments have shown that sulfate-reducing bacterial cultures may cooxidize small amounts of methane (6, 18, 38). However, some of the anaerobic methane oxidation rates measured with ¹⁴CH₄ prepared biologically by using Methanobacterium thermoautotrophicum may not be accurate as they could include carbon monoxide oxidation rates (14). The problem occurs because M. thermoautotrophicum produces significant amounts of ¹⁴CO. If the radioactive methane was not purified and the measured rates were low, the ¹⁴CO₂ detected may simply have been generated by ¹⁴CO oxidation (14). Hopcalite can be used to purify ¹⁴CH₄ (14). Alternatively, a methanogenic strain that does not produce high concentrations of carbon monooxide could be used (14).

Regardless of possible methodological problems, it is evident that some sulfate-reducing bacteria may cooxidize small quantities of methane (14), but the pure cultures tested so far could not account for the anaerobic methane oxidation rates measured in various environments (18).

The free energy available for a consortium performing anaerobic methane oxidation at in situ concentrations of CO₂, sulfide, methane, and sulfate is 22.35 kJ per mol of methane oxidized (14, 45). Such a low energy yield approaches the minimum requirement for even the most frugal bacteria if two species have to share it. Either the bacteria must be well adapted to an extremely penurious existence, or an unidentified energy-yielding reaction is coupled to the anaerobic oxidation of methane (14).

We describe here a study of the anaerobic methane oxidation in a marine sediment from Aarhus Bay, Denmark. The biogeochemical concentration profiles and rate measurements obtained strongly support the hypothesis that anaerobic methane oxidation occurs in a narrow zone about 150 cm below the sediment surface. We tried to evaluate the relative abundance of participating organisms by using minimum cycles for detectable products PCR (MCDP-PCR) and to identify the organisms based on a phylogenetic analysis of molecular sequences retrieved at different depths. We investigated the occurrence of sulfate-reducing bacteria and a distinct group of archaea that was recently proposed to be associated with anaerobic oxidation of methane (15).

The sulfate-reducing bacteria are a polyphyletic group, which limits the usefulness of 16S rRNA-based approaches because physiological inferences can be made only if a molecular isolate is very closely related to known pure cultures of sulfate-reducing bacteria. We instead used a primer set designed by Wagner et al. (42) that targets a key enzyme in sulfate respiration, the dissimilatory sulfate reductase (DSR). This primer set amplifies a 1.9-kb fragment of the - and -subunits of the phylogenetically conserved DSR gene. We also designed a primer set which amplifies an 817-bp fragment of the 16S rRNA gene of the specific group of archaea that have been proposed to be involved in anaerobic methane oxidation by Hinrichs et al. (15).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Sediment sampling and handling. Sediment cores were collected in June 1999 at station 6 in Aarhus Bay, a semienclosed embayment off the east coast of Jutland, Denmark (56°09'30"N, 10°19'15"E). The salinity of the water just above the sediment at a depth of 16 m was 31, and the temperature was 5.4°C. The sediment consisted of very fine sand, silt, and clay, and the sediment was permanently reduced. Cores 7 cm wide and 6 m long were obtained with a piston corer with a removable polyvinyl chloride tube as a liner (20). The cores and polyvinyl chloride tubes were cut into four 1.5-m sections on the ship. The cores were stored in sealed plastic bags and brought to the laboratory. Core A was analyzed within 24 h after sampling. It was split longitudinally, and subsamples were taken at 20-cm intervals to determine the sulfate-methane transition zone. Density and porosity were also measured. Core B was stored at in situ temperature until it was analyzed 1 week after sampling. Ten-centimeter sections were cut off the end of the core, and subsamples were immediately taken from the two exposed surfaces by subcoring with syringes. The core was sampled at 4- to 6-cm intervals from the surface to a depth of 190.5 cm. Sediment from core B was used to quantify sulfate reduction and methane oxidation rates and for DNA extraction (as described below) and also to obtain the same measurements that were obtained with core A.

Porewater analyses. (i) Sulfate and chlorinity. The porewater used to determine sulfate and chloride concentrations was collected by centrifuging 10 cm³ of sediment and was preserved in 20% (wt/vol) zinc acetate. Porewater samples were filtered, and concentrations of sulfate and chloride were determined by ion chromatography (Sykam, Gilching, Germany). The eluent contained 7.5 mM Na₂CO₃, 5% ethanol, and 50 mg of 4-hydroxybenzonitrile liter¹ (13).

(ii) Methane. Ten cubic centimeters of sediment from each depth was incubated in a 100-ml infusion bottle containing 10 ml of 1 M NaOH to trap CO₂ and to stop methanogenesis. The bottles were capped immediately with bromobutyl rubber stoppers and shaken vigorously for 1 min. Gas samples were injected into a gas chromatograph equipped with a flame ionization detector (Packard) after separation on a Poropack Q column (13). The methane peak was recorded on a strip chart recorder and quantified by comparison with standards.

(iii) Density and porosity. The density and porosity of the sediment were measured for every third depth by using 50-cm³ sediment samples. Porosity was determined from the density and the water content; the latter was measured by drying 10 g of sediment overnight at 110°C.

(iv) Steady-state sulfate reduction rates, methane oxidation rates, and methane production rates estimated from sulfate and methane concentration profiles. Net production and consumption rates were estimated from the curvature of in situ concentration profiles. For the sake of simplicity all metabolic rates pertaining to methane and sulfate were considered positive if the compound was being produced and negative if it was being consumed; i.e., production rates [P(z)] are positive rates, and consumption rates [R(z)] are negative rates, and the resulting net rate of metabolism of methane or sulfate [M(z)] is M(z) = P(z) + R(z). Using this convention, Fick's second law of one-dimensional diffusion becomes:

(4)

where C(z,t) is the concentration at depth z and time t, D_s is the diffusivity, and M(z) is the net metabolic rate of production and consumption for either methane or sulfate. An assumption of this formula is that D_s does not change in the depth range being investigated. D_s for marine sediments can be estimated from the equation given by Ullman and Aller (41):

(5)

where D₀ is the molecular diffusion coefficient and  is the porosity. The exponent m ranges from 2 to 3 in different sediments depending on porosity. In our study we used a value of 2.5, which is typical for fine-grained silty sediments with a porosity greater than 0.71 and less than 0.86 (41). The molecular diffusion coefficients for methane and sulfate in seawater at 5°C were 0.95 × 10⁵ and 0.58 × 10⁵ cm² s¹, respectively (26). Under steady-state conditions equation 4 becomes:

(6)

where C(z) is the concentration at depth z.

Rate measurements. (i) Sulfate reduction measurement. Approximately 10 µl of radiolabelled sulfate (37 kBq of ³⁵SO₄²/µl) was injected into 10-cm³ subsamples of sediment. The subsamples were incubated in an anaerobe jar at 5°C for 67 h. Incubation was stopped by mixing the sediment with 10 ml of 20% (wt/vol) zinc acetate. Reduced sulfur compounds (H₂S, FeS, FeS₂, and S⁰) were stripped from the sediment as H₂S by single-step chromium distillation (12). A flow of N₂ was used to carry the H₂S from the reaction vessel to tubes containing 10 ml of 2% (wt/vol) zinc acetate. Hydrogen sulfide was trapped as ZnS. A scintillation counter (Tricarb 2200 CA; Packard) was used to measure the radioactivity of the reduced sulfur compounds in the zinc acetate trap and the radioactivity of the nonreduced ³⁵SO₄² after addition of scintillation liquid (Ecoscint A; Packard). The sulfate reduction rate was calculated as described by Fossing and Jørgensen (12).

(ii) Methane oxidation rates. Radioisotope-labelled methane was biosynthesized by Methanococcus deltae and was purified with Hopcalite as described by Harder (14). Approximately 20 µl of radiolabelled methane (0.6 kBq of ¹⁴CH₄/µl) was injected into 10-cm³ subsamples of sediment, which were incubated at 5°C for 67 h under anaerobic conditions. Incubations was stopped by mixing the sediment with 10 ml of 1 M NaOH. Radiolabelled ¹⁴CO₂ was collected from the samples as described by Hansen et al. (13). Methane radioactivity was measured by injecting samples of headspace gas into a gas chromatograph equipped with a flame ionization detector. ¹⁴CO₂ was then produced by combustion and trapped in a scintillation vial containing 1 ml of 1 M NaOH. Finally, 10 ml of HiOnic Fluor scintillation cocktail (Packard) was added to the trap. Methane oxidation rates were calculated from the amount of ¹⁴CO₂ formed and from the concentration and radioactivity of methane as described by Iversen and Blackburn (19).

Nucleic acid extraction. Nucleic acids were extracted by using a FastDNA spin kit for soil (Bio 101, Vista, Calif.) according to the manufacturer's instructions. This method relies on mechanical cell lysis by bead beating (FastPrep DNA extractor; Bio 101) followed by selective DNA adsorption to microporous silicate filters. The bound DNA is then washed with ethanol in the presence of chaotropic salts and finally eluted in a low-salt buffer. Nucleic acid extraction was evaluated on a 1% Seakem GTG agarose gel (FMC Bioproducts, Rockland, Maine) electrophoresed with TAE buffer (40 mM Tris, 20 mM acetic acid, 1 mM EDTA; pH 8.3). The gel was stained with SybrGold (100 ng/ml; Molecular Probes, Leiden, The Netherlands) for 20 min. For analysis and documentation a transilluminator and a digital camera (Gel Doc 2000; Bio-Rad, Hercules, Calif.) were used. Images were acquired and analyzed with the software Quantity One (Bio-Rad). Larger amounts of DNA were recovered from the samples closest to the sediment surface; however, the concentrations of recovered DNA were similar for all samples obtained at depths below 100 cm from the sediment surface. The standard error was less than 10% of the average DNA concentration in each case. To test the quality of the DNA recovered, PCR amplification of ribosomal DNA was performed with universal primers for 24 cycles. Similar amounts of PCR product were obtained with DNA extracts from all depths below 100 cm (standard error, <15%).

MCDP-PCR amplification. MCDP-PCR is a method that is used to evaluate the relative abundance of template genes present in a sample (9). The principle is to determine the minimum number of PCR cycles necessary to detect a faint amplificate (approximately 0.1 ng) on an agarose gel. The rationale behind the method is that product inhibition reduces amplification efficiency in the last cycles of a PCR when the product concentration is high. Thus, by reducing the number of cycles to the minimum number required for a detectable product, product interference is reduced as much as possible. The number of cycles required to produce a detectable product has been shown to indicate the relative abundance of template molecules (9, 11). An automated version of this technique is widely used in real-time PCR approaches developed by Hoffmann-La Roche, Bio-Rad, and Perkin-Elmer to quantify target DNA in diagnostic medical applications (28). An alternative approach to obtain a barely detectable product is to dilute the sample prior to amplification. However, dilution also reduces the concentration of PCR-inhibiting compounds (e.g., humic acids) present in the sample, which makes it difficult to compare the amplification results obtained for differently diluted samples. It is difficult to get an absolute estimate of gene copies by MCDP-PCR; however, it is possible to evaluate the relative abundance of genes in different samples. In this study, two primer pairs were used. The first primer set was designed by Wagner et al. (42) to specifically amplify a 1.9-kb fragment of the DSR gene; this set consisted of primers DSR1F (5'-ACSCACTGGAAGCACG-3') and DSR4R (5'-GTGTAGCAGTTACCGCA-3'). A different primer was designed to target the sequences of archaea that were proposed by Hinrichs et al. (15) to be anaerobic methane oxidizers; this primer was primer ANMEF (5'-GGCUCAGUAACACGUGGA-3'). When sequences were subjected to Probe Match analysis with small-subunit rRNAs of the Ribosomal Database Project (Center for Microbial Ecology, University of Michigan), only four sequences of uncultured archaea matched the probe sequence. All the matching sequences were sequences of tentative methane oxidizers from the Eel River Basin study. The highly specific primer was used together with a universal 16S rRNA primer (907R; 5'-CCGTCAATTCCTTTRAGTTT-3') (25) to detect these archaea in the sediment. The reaction mixture used for PCR amplification contained 36.5 µl of distilled H₂O, 5 µl of buffer (100 mM Tris-HCl, 750 mM KCl, 15 mM MgCl₂; pH 8.8), 5 µl of a 10× deoxynucleoside triphosphate (dNTP) mixture (containing each dNTP at a concentration of 125 µM), 1 µl of each primer (50 pmol/µl), and 0.5 µl of Taq polymerase (5,000 U/ml; Pharmacia). One microliter (approximately 10 ng) of diluted DNA extract was added. PCR amplification was carried out with a DNA Thermocycler (PT-200 Peltier thermal cycler; MJ Research). Each cycle in the PCR program for DSR gene amplification consisted of 1 min of denaturation at 94°C, 1 min of annealing at 54°C, and 3 min of extension at 72°C. The reaction was completed by a 10-min final extension step at 72°C. PCR with 28, 30, 32, and 34 cycles were performed with the DSR gene primers. The PCR program used for amplification of 16S ribosomal DNA with the ANME primer set consisted of 30 s of denaturation at 92°C, 1 min of annealing at 57°C, and 0.45 s (plus 1 s/cycle) of extension at 72°C. The reaction was completed by a 5-min final extension step at 72°C. With the ANME primer pair 25, 26, 27, 28, and 29 cycles were performed. The PCR products were loaded on a 2% Nusieve 3:1 agarose gel (FMC Bioproducts), and electrophoresis was performed with TAE buffer. The gels were stained with SybrGold (100 ng/ml; Molecular Probes), and the results were evaluated and documented as described above. The detection limit of SybrGold was less than 100 pg of double-stranded DNA per band in routine applications. Band intensities were estimated by using custom macros for the image analysis program NIH Image 1.62b7 written by Wayne Rasband (available from zippy.nimh.nih.gov). The intensity estimates were standardized by using a constant camera aperture and division by exposure time. The reproducibility of the estimates was evaluated by comparing band intensities in marker lanes of the gels analyzed.

Cloning of DSR and 16S rRNA genes. Fresh PCR amplificates obtained with primers DSR1F and DSR4R and with primers ANMEF and 907R were purified with a QIAquick PCR purification kit (Qiagen GmbH, Hilden, Germany). The amplificates were ligated into a pCR-XL-TOPO vector and transformed into ONE SHOT Escherichia coli cells by following the manufacturer's directions (TOPO XL PCR cloning; Invitrogen, Leek, The Netherlands). Each clone was screened for an insert of the correct length by direct PCR amplification with the original primers, using 16 cycles and the PCR programs mentioned above. Plasmids of randomly selected clones containing the correct-size insert were then recovered with a QIAprep spin miniprep kit (Qiagen GmbH).

Sequencing. Partial DNA sequences were obtained from extracted plasmid templates with Cy-5-labelled primers targeting the plasmid sequence surrounding the insert (VektorF [5'-TTTGGCCCTCTAGATG-3'] and VektorR [5'-CTATGCATCAAGCTTGG-3']) (T. R. Thomsen, M. Wagner, K. K. Brandt, K. Ingvorsen, and N. B. Ramsing, unpublished data), using a thermosequenase fluorescent cycle sequencing kit (Pharmacia Biotech, Uppsala, Sweden) and an ALFexpress DNA sequencer (Pharmacia Biotech). The following components were used for each sequencing reaction: 3.375 µl of distilled H₂O, 0.375 µl of dimethyl sulfoxide, 1.25 µl of primer (1.25 pmol/µl), 2 µl of a dNTP mixture, and 1 to 2 µl of template (isolated plasmids). Amplification was carried out for 40 cycles, with each cycle consisting of 30 s of denaturation at 94°C, 30 s of annealing at 57°C, and 5 min of extension at 72°C.

Phylogenetic analysis. DSR gene sequences were aligned and analyzed by using the ARB program package (available from www.mikro.biologie.tu-muenchen.de/Pub/ARB) (39). DNA sequences were translated into amino acid sequences and aligned manually with the Genetic Data Environment (GDE), version 2.2, sequence editor implemented in the ARB software environment. Nucleic acid sequences were subsequently aligned on the basis of the amino acid alignment. Trees based on aligned sequences were constructed by using the FITCH distance matrix program in Phylip 3.53. Only unambiguously aligned amino acid positions from the - and -subunits of the DSR gene found in all sequences were used. The final data set consisted of 264 amino acids. Both nucleic acid and amino acid alignments were also evaluated by using PAUP*, version 4.0 (40). The nucleic acid alignment was analyzed by distance matrix and maximum-likelihood approaches, whereas the amino acid alignment was evaluated by using parsimony- and distance matrix-based algorithms. For all types of phylogenetic analysis we used the default settings in PAUP*, version 4.0.

Nucleotide sequence accession numbers. The GenBank accession numbers for the small-subunit sequences are as follows: ANME1, AF314243; ANME7, AF314244; ANME8, AF314249; ANME11, AF314247; ANME12, AF314242; ANME14, AF314246; ANME16, AF314245; ANME17, AF314248; ANME19, AF314241; and ANME23, AF314250. The GenBank accession numbers for DSR gene sequences are as follows: a-a, AF316066; a-E, AF316067; a-F, AF316065; a-G, AF316050; a-20, AF316070; a-24, AF316044; a-54, AF316068; a-55, AF316063; a-70, AF316039; a-73, AF316062; a-75, AF316042; b-2, AF316051; b-4, AF316073; b-6, AF316071; b-15, AF316046; b-18, AF316053; b-20, AF316061; b-28, AF31606; b-30, AF316059; b-31, AF316047; b-37, AF316052; c-C, AF316054; c-D, AF316048; c-E, AF316049; c-F, AF316058; c-I, AF316056; c-J, AF316040; c-1, AF316045; c-18, AF316069; c-23, AF316064; c-31, AF316072; c-40, AF316057; c-47, AF316043; c-59, AF316055; c-63, AF316041; Desulfobacter halotolerans, AF521159; Desulfocella halophila, AF321158; KYF135, AF321149; KYF124, AF321156; KYF128, AF321154; KYF136, AF321155; KYF312, AF321153; KYF313, AF321152; KYF314, AF321157; KYF322, AF321151; and KYF 324, AF321150.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Biogeochemical analyses. The sediment density was about 1.30 g cm³ at all depths below 50 cm (data not shown). The porosity ranged from 0.71 to 0.86 ml cm³, as shown in Fig. 1. The sulfate concentration (Fig. 1) was highest at the sediment surface (18.7 mM in porewater) and decreased with depth to less than 2 mM below a depth of 136 cm. The methane concentration (Fig. 1) was less than 0.1 mM in the upper 150 cm but increased gradually below a depth of 150 cm to a maximum value of 1.5 mM at a depth of 281 cm. The cores contained numerous gas bubbles at depths below approximately 290 cm, which probably explains the large variations in the concentrations measured below this depth. Bubbles were formed when the methane core was brought to the surface as the in situ hydrostatic pressure was removed. The upper limit of bubble formation coincided with a methane concentration of about 1.5 mM, which corresponded to a partial pressure of 1 atm. The zone of bubble formation at the in situ hydrostatic pressure was likely to be considerably deeper in the sediment. The depth sounder employed during sampling showed that there was strong reflectance from a sediment layer approximately 3.5 m below the seafloor. The reflection (often referred to as the methane mirror) was most likely caused by small methane bubbles within the sediment. The sulfate-methane transition zone was at a depth of 140 to 160 cm. The methane profile was concave in this depth interval, indicating that there was net consumption. Activities were calculated from the core A profiles because the concentration measurements were made immediately after sampling. Sulfate reduction activity was greatest in the zone ranging from a depth of 138 cm to a depth of 159 cm, and the calculated rate was 1.74 nmol cm³ day¹ (Fig. 2A). The calculated maximum methane oxidation value was 0.59 nmol cm³ day at a depth of 140 to 169 cm (Fig. 2B). Net methane production began at a depth of 169 cm but reached a maximum value of 1.14 nmol cm³ day¹ at a depth of 215 cm (Fig. 2C). The narrow zone of high methane production was most likely an artifact as the inflexion point of the methane profile (Fig. 1) occurred where the partial pressure of methane was 1 atm. Thus, it is likely that methane was vented from the deeper parts of the profile by bubble formation. The measured rates of sulfate reduction and methane oxidation are also shown in Fig. 2A and B, respectively. The highest sulfate reduction rate occurred at the 140.5-cm depth (1.05 nmol cm³ day¹), whereas the highest methane oxidation rate occurred at the 150.5-cm depth (0.79 nmol cm³ day¹). The calculated cumulative sulfate reduction rate, based on the sulfate concentration profile, was 365.57 µmol m² day¹, whereas the measured rate, obtained by isotopic techniques, was only 84.16 µmol m² day¹. The calculated cumulative methane oxidation rate, based on the methane concentration profile, was 170.35 µmol m² day¹, whereas the measured rate, obtained by isotopic techniques, was 45.17 µmol m² day¹. Sulfate reduction values obtained by Jørgensen et al. (22) at station 6 were integrated to estimate that the sulfate reduction rate in the total core was 4,801.41 µmol m² day¹.

View larger version (26K): [in this window] [in a new window]   FIG. 1.   Concentrations of sulfate and methane in sediment cores A and B. Porosity is indicated by the vertical line without symbols. Depths at which molecular investigations were done are indicated (a, 21.5 cm; b, 81.5 cm; c, 156.5 cm).

View larger version (27K): [in this window] [in a new window]   FIG. 2.   (A) Depth profiles for measured sulfate reduction rates obtained by tracer techniques and for rates calculated from the sulfate concentration profiles. (B) Depth profiles for measured methane oxidation rates obtained by tracer techniques and for rates calculated from the methane concentration profiles. (C) Depth profile for methane production rates based on the concentration profile for methane. The depths at which molecular investigations were done are indicated (a, 21.5 cm; b, 81.5 cm; c, 156.5 cm).

MCDP-PCR amplification. MCDP-PCR results obtained for different sampling depths with the ANMEF-907R and DSR gene primer sets are shown in Fig. 3A and C, respectively. The depth profiles for band intensity per gram of sediment analyzed are shown in Fig. 3B, D, and E. Amplifications in which different numbers of cycles were used were scaled to the lowest number of cycles employed by dividing the intensities by 2ⁿ where n is the number of additional cycles employed. When amplification results obtained with different numbers of PCR cycles with the ANME primer pair were compared (Fig. 3B), the scaled data curves were all similar and had a peak at and below the sulfate-methane transition zone. The similarity of the scaled curves obtained with different numbers of PCR cycles indicates that the assumed ideal amplification giving rise to a doubling of product for each cycle was reasonable.

View larger version (32K): [in this window] [in a new window]   FIG. 3.   (A) Agarose gel showing amplificates obtained with primers ANMEF and 907R after 27 PCR cycles. (B) Depth profile for relative gel band intensities of amplificates obtained with the ANMEF-907R primer set. Different symbols correspond to different numbers of PCR cycles, as follows: open circles, 29 cycles; gray circles, 28 cycles; solid circles, 27 cycles. (C) Gel showing amplificates obtained with the DSR gene primer set after 28 PCR cycles. (D and E) Depth profiles for gel band intensities of amplificates obtained with the DSR gene primer set by using different numbers of PCR cycles, as follows: solid circles, 28 cycles; light gray circles, 30 cycles; dark gray circles, 32 cycles; open circles, 34 cycles. The band intensities in panels B, D, and E were scaled by dividing by 2 for every cycle beyond the lowest number used with the primer set (e.g., the measured band intensities of DSR gene amplificates obtained after 30 cycles were divided by 4 to be equivalent to the intensities obtained after 28 cycles). For more information see Results. The samples picked for cloning and sequencing were samples from zone a (depth, 21.5 cm), zone b (depth, 81.5 cm), zone c (depth, 156.5 cm).

Phylogenetic analysis. On the basis of the results obtained by MCDP-PCR we decided to clone DSR genes from three zones, zone a (depth, 21.5 cm), zone b, (81.5 cm), and zone c, (156.5 cm). We also cloned the amplificates acquired with the primers specific for 16S rRNA targeting ANME-like archaea in zone c to verify that these sequences were indeed related to the archaeal 16S rRNA targets reported by Hinrichs et al. (15). Phylogenetic trees for the DSR gene based on unambiguously aligned nucleic acid and amino acid data sets were estimated by using distance matrix, parsimony, and maximum-likelihood criteria. The three methods resulted in congruent tree topologies except for the phylogenetic position of Desulfobulbus propionicus. Maximum-likelihood and distance matrix approaches using nucleic acid sequences, as well as parsimony analysis using amino acid sequences, placed D. propionicus in the cluster of phylogenetically related incompletely oxidizing sulfate-reducing bacteria in the  subgroup of the class Proteobacteria. However, distance matrix analysis based on amino acid sequences placed D. propionicus among the low-G+C-content gram-positive bacteria with Desulfotomaculum as the closest relative. We chose to present distance matrix trees derived with PAUP*, version 4.0, using default settings.

View larger version (43K): [in this window] [in a new window]   FIG. 4.   Phylogenetic tree based on nucleic acids. Sequences obtained at different depths (zone a, 21.5 cm; zone b, 81.5 cm; zone c, 156.5 cm) of a station 6 sediment core were compared to pure cultures of sulfate-reducing bacteria. The tree was constructed by using default settings for the distance matrix algorithm in PAUP*, version 4.0, and bootstrapping with 100 replicates was performed. Only nodes supported by bootstrap values greater than 50% are shown. Scale bar = 0.05 substitution per nucleotide position.

View larger version (40K): [in this window] [in a new window]   FIG. 5.   Phylogenetic tree based on amino acids. Sequences obtained at different depths (zone a, 21.5 cm; zone b, 81.5 cm; zone c, 156.5 cm) of a station 6 sediment core were compared to pure cultures of sulfate-reducing bacteria. The tree was constructed by using default settings for distance matrix analysis in PAUP*, version 4.0. Scale bar = 0.05 substitution per nucleotide position.

View larger version (37K): [in this window] [in a new window]   FIG. 6.   Phylogenetic tree of archaeal rRNA sequences obtained from station 6 at a depth of 156.5 cm (zone c), sequences of pure cultures of archaea, and environment sequences. The tree was constructed by using default settings for distance matrix analysis in PAUP*, version 4.0, and bootstrapping with 100 replicates was performed. Only nodes supported by bootstrap values greater than 50% are shown. Scale bar = 0.05 substitutions per nucleotide position.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Biogeochemical studies. Based on concentration profiles and rate measurements, the sediment could be divided into three functional zones. The upper zone, zone I, from 0 to 140 cm, is the sulfate-containing upper zone; in this zone the maximum sulfate concentration, present at the surface (18.7 mM) decreases gradually. The sulfate-methane transition zone, zone II, from 140 to 160 cm, is the transition zone in which the two compounds are present simultaneously. Finally, the methanogenic zone, zone III, extends downwards from 160 cm. In this zone net methanogenesis increases and diffusion from below augments the methane concentration until it reaches a maximum value of 1.5 mM at 281 cm, where methane bubble formation starts. The slightly concave methane profile in zone II indicates that net methane oxidation occurs in this zone (27).

Molecular work. The MCDP-PCR profiles obtained with both the DSR and ANME primer pairs revealed a pronounced peak in the amount of amplificate produced from samples obtained in the sulfate-methane transition zone. The peak could have been due to PCR inhibitors present in surrounding strata. Nevertheless, the homogeneous nature of the sediment that far from the sediment surface makes this an unlikely explanation. Thus, the MCDP-PCR results support the possible occurrence of higher numbers of sulfate-reducing bacteria and special archaea in zone c (156.5 cm below the surface) than in zone b. Still, the largest DSR gene amplificate from sulfate-reducing bacteria was detected in the top sediment. However, the DSR gene of these sulfate-reducing bacteria was apparently phylogenetically different from the genes found in the deep sediment. Most sequences from zone a (depth, 21.5 cm) did cluster with environmental sequences from other estuarine sediments (i.e., Kysing Fjord), and the cluster (group I) was affiliated with DSR genes from known complete oxidizers belonging to the genera Desulfosarcina, Desulfococcus, and Desulfonema.