Use of 16S rRNA-targeted oligonucleotide probes to investigate function and phylogeny of sulphate-reducing bacteria and methanogenic archaea in a UK estuary

Authors

  • K.J Purdy,

    Corresponding author
    1. Department of Biological Sciences, University of Essex, Colchester, Essex CO4 3SQ, UK
      *Corresponding author. Present address: School of Animal and Microbial Sciences, University of Reading, Whiteknights, P.O. Box 228, Reading RG6 6AJ, UK. Tel.: +44 (118) 3788892; Fax: +44 (118) 9316671. E-mail address: k.j.purdy@reading.ac.uk
    Search for more papers by this author
  • M.A Munson,

    1. Department of Biological Sciences, University of Essex, Colchester, Essex CO4 3SQ, UK
    Search for more papers by this author
    • 1

      Department of Oral Medicine and Pathology, Oral Microbiology Unit, King's College London, Guy's Tower, London SE1 9RT, UK.

  • T Cresswell-Maynard,

    1. Department of Biological Sciences, University of Essex, Colchester, Essex CO4 3SQ, UK
    Search for more papers by this author
  • D.B Nedwell,

    1. Department of Biological Sciences, University of Essex, Colchester, Essex CO4 3SQ, UK
    Search for more papers by this author
  • T.M Embley

    1. Molecular Biology Unit, Department of Zoology, Natural History Museum, London SW7 5BD, UK
    Search for more papers by this author

*Corresponding author. Present address: School of Animal and Microbial Sciences, University of Reading, Whiteknights, P.O. Box 228, Reading RG6 6AJ, UK. Tel.: +44 (118) 3788892; Fax: +44 (118) 9316671. E-mail address: k.j.purdy@reading.ac.uk

Abstract

Sulphate-reducing bacteria (SRB) and methanogenic archaea (MA) are important anaerobic terminal oxidisers of organic matter. However, we have little knowledge about the distribution and types of SRB and MA in the environment or the functional role they play in situ. Here we have utilised sediment slurry microcosms amended with ecologically significant substrates, including acetate and hydrogen, and specific functional inhibitors, to identify the important SRB and MA groups in two contrasting sites on a UK estuary. Substrate and inhibitor additions had significant effects on methane production and on acetate and sulphate consumption in the slurries. By using specific 16S-targeted oligonucleotide probes we were able to link specific SRB and MA groups to the use of the added substrates. Acetate consumption in the freshwater-dominated sediments was mediated by Methanosarcinales under low-sulphate conditions and Desulfobacter under the high-sulphate conditions that simulated a tidal incursion. In the marine-dominated sediments, acetate consumption was linked to Desulfobacter. Addition of trimethylamine, a non-competitive substrate for methanogenesis, led to a large increase in Methanosarcinales signal in marine slurries. Desulfobulbus was linked to non-sulphate-dependent H2 consumption in the freshwater sediments. The addition of sulphate to freshwater sediments inhibited methane production and reduced signal from probes targeted to Methanosarcinales and Methanomicrobiales, while the addition of molybdate to marine sediments inhibited Desulfobulbus and Desulfobacterium. These data complement our understanding of the ecophysiology of the organisms detected and make a firm connection between the capabilities of species, as observed in the laboratory, to their roles in the environment.

1Introduction

Estuarine and coastal marine sediments are of great biogeochemical importance [1–3] and thus their microbial communities have been studied in detail for a number of years. It has been shown that short-chain fatty acids, primarily acetate and propionate, and hydrogen represent the most important substrates in the terminal anaerobic oxidation processes in these sediments [4–9]. Traditional microbial ecological and newer molecular studies have informed us of the structure of the microbial communities in estuarine and coastal marine sediments [10–14]. However, there are few data that directly link the structure of the microbial community to activity that can be detected in situ. The result is that the majority of our understanding of the dynamics of the microbial communities in estuarine and coastal marine sediments is limited to extrapolating from pure culture studies with little solid evidence to support the suppositions put forward. If we are to begin to comprehend the dynamics and ecology of microbial communities in situ it is essential that the use of specific substrates be linked to the activity of specific organisms.

Sulphate-reducing bacteria (SRB) and methanogenic archaea (MA) are anaerobic terminal oxidisers of organic matter and can be seen as ecological equivalents [15,16], mineralising organic matter to CO2 or CO2 and CH4, respectively. SRB tend to outcompete MA when sulphate is freely available (i.e. in marine sediments) as they have higher specific affinities for the main substrates used by the two groups, acetate and H2[6,17–19]. However, even in marine sediments methanogenesis still occurs [1], primarily via the oxidation of C1-compounds, such as trimethylamine, that SRB cannot use [9]. In freshwater environments, where available sulphate is usually limited, sulphate reduction does still occur but methanogenesis dominates [20]. Up to 50% of organic matter degradation in freshwater and coastal marine sediments proceeds via methanogenesis or sulphate reduction, respectively [18,21].

Oligonucleotide probes which specifically target SRB and MA groups and genera [22,23] have previously been used to investigate the structure of SRB and MA communities in sediment [13,24–27]. However, as stated above, few studies have attempted to link in situ SRB and MA community structure to utilisation of particular substrates. In sediment slurry microcosms from a Japanese estuary acetate utilisation at a marine-dominated site was linked to Desulfobacter and propionate degradation at both marine and freshwater-dominated sites to Desulfobulbus using 16S rRNA-targeted oligonucleotide probes [8]. The results of these slurry microcosm experiments were subsequently supported by probing of directly extracted RNA, which indicated that Desulfobacter were the dominant SRB at the marine site and that Desulfobulbus were active at both freshwater and marine ends of the estuary [27].

The use of slurry microcosms enables controlled experiments and hypothesis testing to be performed under conditions that resemble those in situ, and allows the investigation of the response of a natural, mixed community to a specific environmental perturbation. In slurry microcosms from a Japanese estuary that were amended with short-chain fatty acids a five-fold increase in relative oligonucleotide probe signal was detected within 2 days [8]. Such a rapid response was most likely due to the capacity of the standing populations and is strong support for the ecological relevance of this methodology. Here, we describe the use of anaerobic slurry microcosm experiments using sediment from two contrasting estuarine sites, to test hypotheses of substrate utilisation by MA and SRB. We amended the slurries with a range of ecologically important substrates including acetate, the primary substrate for anaerobic terminal oxidation, as well as hydrogen and a non-competitive substrate, trimethylamine. Oligonucleotide probes to different SRB and MA genotypes were then used to monitor changes in the microbial community that occurred in response to substrate addition. Our aim was to determine the SRB and MA genotypes that could metabolise these key environmental substrates in contrasting estuarine sediments.

2Materials and methods

2.1Study sites and sampling

Sediment cores (30 cm×10 cm deep) were collected from two sites on the River Colne, Essex, UK. Colne Point, the site at the mouth of the estuary, is marine-dominated whereas site 2 at East Hill Bridge, some 10 miles upstream of the mouth of the estuary, is predominately freshwater with occasional incursions of seawater, and is thus a dynamic estuarine environment. Sediment cores were returned to the laboratory and slurries prepared within a few hours of sampling.

2.2Experiments to link utilisation of key substrates to phylogenetic groups in two contrasting estuarine sites

The East Hill Bridge experiments were amended with substrates to investigate the competition between SRB and MA under conditions relevant to this site (see Table 1), and to determine which phylotypes appeared to be linked to substrate utilisation. The focus was on the competitive effects of sulphate as a model of a tidal incursion into this site. Thus, acetate and hydrogen, the two main competitive substrates for SRB and MA, were added with, and without, sulphate to simulate an incursion of high-sulphate seawater. The Colne Point experiments were designed to identify the dominant organisms involved in acetate, hydrogen and C1-compound (trimethylamine) metabolism in this permanently high-sulphate environment, with a major focus on the role of H2. In these slurries we used inhibitors of SRB (molybdate [28]) and MA (chloroform [29]) to investigate the competition between these two groups in marine-dominated sediments (Table 1), and to identify which particular SRB or MA were involved in the competition.

Table 1.  List of treatments used in sediment slurries from two sites on the Colne estuary
SiteTreatmentRationaleLabel
East Hill BridgeControl (unamended)ControlCEHB
 Sulphate (20 mM)General SRB substrateSEHB
 Acetate (20 mM)For freshwater acetate-usersAEHB
 Acetate+sulphate (20 mM each)For SRB acetate-usersASEHB
 H2/CO2 (80%:20%)For freshwater H2-usersHEHB
 H2/CO2+sulphate (20 mM)For SRB H2-usersHSEHB
Colne PointControl (unamended)ControlCCP
 20 mM molybdateInhibit SRBCMCP
 5 mM chloroformInhibit MACCCP
 H2/CO2For H2-usersHCP
 H2/CO2+20 mM molybdateFor H2-users/Inhibit SRBHMCP
 H2/CO2+5 mM chloroformFor H2-users/Inhibit MAHCCP
 Acetate (20 mM)For acetate-usersACP
 Trimethylamine (20 mM)For MA able to use non-competitive substratesTMACP

Turnover of acetate in sediments has been measured at 0.5 μmol ml−1 day−1, which is the equivalent to a concentration of 1 mM in the porewater [5]. Our previous work suggests that 20 mM acetate can be utilised within 8 days with no lag-phase in estuarine sediment slurries [8]. Therefore, in each slurry, sufficient substrate was added to maintain activity for 10–20 days (Table 1). These amendments are substantial compared to the in situ concentrations. However, the in situ concentration of a substrate is a measure of the concentration of that substrate in an homogenised sample and can not take into account localised elevated concentrations that will be associated with particles of decaying organic matter. It would seem reasonable to assume that organisms can experience substrate concentrations far in excess of that seen in the general porewater and that may well represent a significant, if transient, excess of available substrate. Thus, our experiments on a sediment model system represent a reasonable perturbation of the system and should allow us to extrapolate the results to the environment.

2.3Preparation and sampling of slurry experiments

Anaerobic slurries (50% v/v with site bottom water) of the top 5 cm of the sediment cores from each station were prepared and dispensed under anaerobic conditions [8]. Slurries (50 ml in 100 ml flasks for East Hill Bridge and 100 ml in 250 ml conical flasks for Colne Point, sealed with suba seals) were then amended (with triplicate flasks for each treatment) with the substrates and inhibitors listed in Table 1.

The slurries were incubated at 20°C in the dark for 10 days in the case of the East Hill Bridge slurries and for 20 days for the Colne Point slurries, with shaking on an orbital shaker at a speed just sufficient to maintain the sediment in suspension. The incubation temperature was selected as being typical of the in situ summer temperature for Colne River sediments.

Headspace gas was analysed for methane every 1–4 days. Samples of slurry (5 ml) were taken every 1–4 days, immediately centrifuged at 8000×g for 10 min, the supernatant removed, filtered through a glass fibre filter (GF/F, Whatman International, Maidstone, UK) to remove particulates and stored at −20°C prior to analysis for residual sulphate and acetate. Measurements of acetate concentrations in East Hill Bridge slurries were limited to the endpoint samples. Slurry samples were taken at the end of each experiment. The slurry was centrifuged at 8000×g for 10 min, the pellet was washed with 120 mM sodium phosphate, pH 8.0, to remove any extracellular nucleic acid [30] and stored at −20°C until nucleic acid extraction.

2.4Analysis of methane, sulphate and acetate in slurries

Headspace methane was analysed by gas chromatography (GC) with a flame ionisation detector using oxygen-free N2 as the carrier gas [31]. Differences between the final concentrations of methane in the different treatments were tested for statistical significance using a one-way analysis of variance (ANOVA) with a post hoc Tukey test after log n+1 transformation of the data [32].

Sulphate concentrations in the supernatants from the slurry samples were determined by ion chromatography (Dionex, Sunnydale, CA, USA). Acetate was analysed either by ion chromatography (Dionex) for East Hill Bridge slurries or by GC with FID detection after acid extraction into diethyl ether for Colne Point slurries [6].

2.5Extraction and quantification of nucleic acids from slurry samples, pure culture controls and preparation of in vitro transcribed 16S rRNA

DNA and RNA were extracted separately from samples of the stored sediments by the hydroxyapatite spin-column method [33]. Total rRNA was extracted from pure cultures (see Table 2) as described previously [8]. As active cultures of the methanogens studied in this work were not available (see Table 2) 16S rRNA was transcribed in vitro from representative 16S rDNA clones [34].

Table 2.  16S rRNA-targeted oligonucleotide probes and pure culture controls used in probing experiments on RNA extracted from Colne Estuary sediment slurries
  1. SRB cultures were grown according to Widdel and Bak [37]. B. subtilis was grown in nutrient broth.

ProbeTarget groupPure culture controlsReference
P338 (S-D-Bact-0338-a-A-18)Most bacteriaSRBs, B. subtilis NCMB 3610[53]
P129 (S-G-Dsb-0129-a-A-18)DesulfobacterDesulfobacter latus ATCC 43918[22]
P221 (S-G-Dsbm-0221-a-A-20)DesulfobacteriumDesulfobacterium autotrophicum DSM 3382[22]
P660 (S-G-Dsbb-0660-a-A-20)DesulfobulbusDesulfobulbus propionicus DSM 2056[22]
P915 (S-D-Arch-0915-a-A-18)All archaeaAll methanogens (in vitro RNA)[54]
MSMX860 (S-O-Msar-0860-a-A-21)MethanosarcinalesMethanosaeta (in vitro RNA)[23]
MG1200 (S-O-Mmic-1200-a-A-21)MethanomicrobialesMethanogenium (in vitro RNA)[23]

Total RNA extracted from sediment was quantified by UV spectrophotometry. Accurate quantification of oligonucleotide probe response to total RNA extracted from sediment is dependent on accurate quantification of the amount of 16S rRNA from pure culture controls. It cannot be assumed that 16S and 23S rRNA will be in a 1:1 ratio in pure culture extracts. Therefore the 16S rRNA from Escherichia coli (Sigma-Aldrich) was gel-purified and its concentration determined by UV spectrophotometry. Known volumes of total RNA extracted from pure cultures and in vitro transcribed RNA along with a standard curve of the pure E. coli 16S rRNA standards were run on an ethidium bromide-stained agarose gel. Band intensity of 16S rRNA was measured using the public domain NIH Image programme (developed at the U.S. National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/) and the concentration of the samples determined by comparison to the E. coli 16S rRNA standard curve.

2.6Immobilisation of extracted RNA onto nylon membranes and hybridisation conditions

RNA (either 100 ng or 1000 ng total NA/slot) was slot-blotted onto nylon membrane (Hybond-N, Dupont, Stevenage, UK) after denaturation using 2% w/v glutaraldehyde [8] along with known amounts of rRNA from pure culture controls and in vitro transcribed RNA. Oligonucleotide probes (Table 2) were synthesised commercially (MWG, Germany) and end-labelled with γ-32P-ATP (NEN-Dupont, Hounslow, Middlesex, UK [35]). Hybridisations were performed and membranes processed as described previously [8].

2.7Scanning densitometry of autoradiographs and determination of signal levels

Autoradiographs were quantified using laser scanning densitometry (Personal Densitometer, Molecular Dynamics, Kemsing, UK). Densitometrically measured signals were converted to an amount of rRNA for each sample by comparison to a pure culture control standard curve [8]. The standardised results are expressed as a percentage of the aggregate signal from general bacterial (p338) and archaeal (p915) probes. The standardised results are a composite measure of both the population size (number of cells) and the activity of individual cells as expressed by intracellular 16S rRNA [36]. Means and standard deviations were determined after arcsine transformation of the percentage data and then back-transformed to give the quoted results [32]. Differences between signals for each probe were tested for statistical significance by a one-way ANOVA with a post hoc Tukey test after log n+1 transformation of the data [32].

3Results

3.1Effects of substrate addition in freshwater-dominated East Hill Bridge slurries

In East Hill Bridge slurries the unamended control (CEHB) produced methane at a constant rate of approximately 250 nmol ml−1 day−1 for the 10 days of the experiment (Fig. 1a). Methane production increased significantly (F1,5, P<0.0002) in the acetate-amended slurries (AEHB) where methane production exceeded that in the control slurries by 2.6-fold by day 10 (Fig. 1a). Signal from the Methanosarcinales-targeted probe was significantly greater in these slurries than in the controls (P<0.0001; Fig. 2a), reaching 4.5% (S.E.M.=0.4% (n=3 in all cases)) of the total prokaryotic signal compared to 1.3% (S.E.M.=0.1%) in the controls.

Figure 1.

a: Headspace methane accumulation. b: Sulphate concentrations in East Hill Bridge slurries.

Figure 2.

Results of hybridisation of SRB-targeted oligonucleotide probes to rRNA extracted from East Hill Bridge sediment slurries. Data expressed relative to the combined signal from general bacterial (p338) and archaeal (p915) probes. Signal level means were determined after arcsine transformation of the data, then back-transformed to give the quoted results and statistically analysed using a one-way (ANOVA) with a post hoc Tukey test after log transformation of the data [32]. Error bars=1 S.E.M. (n=3).

The addition of sulphate alone (SEHB) and H2/CO2+sulphate (HSEHB) immediately reduced methane production to 2% and 8% of the control slurries, respectively (P<0.00005; Fig. 1a), although methane production continued throughout the experiment. Sulphate was consumed in the SEHB slurries after a 2-day lag period, falling to 10 mM by day 10, while sulphate consumption began immediately in the HSEHB slurries, falling to 1.6 mM by day 10 (Fig. 1b). Signals from both the Methanosarcinales- and Methanomicrobiales-targeted probes were significantly lower in both these sulphate-amended slurries than in the control slurries (P<0.005 and 0.05, respectively; Fig. 2a,b). Signal from Desulfobacter (p129) did increase in the SEHB slurries but this was not significant at the 95% level (P=0.11; Fig. 2c).

The addition of H2/CO2 (HEHB) and acetate+sulphate (ASEHB) did not significantly affect methane production compared to the unamended control (P>0.05). No significant change was detected in signal from either a Methanomicrobiales (MG1200) or Methanosarcinales (MSMX860) probe in the HEHB slurries compared to the control slurries (P>0.05; Fig. 2a,b). However, signal from the Methanomicrobiales probe, but not the Methanosarcinales probe, was significantly lower in the ASEHB slurries compared to the control slurries (P<0.05). Day 10 acetate concentrations in the CEHB and HEHB slurries were very similar (see Table 3).

Table 3.  Residual acetate concentrations in East Hill Bridge slurries at the end of the experiment (day 10)
TreatmentAcetate concentration mean (μM) [S.E.M. (n=3)]
CEHB (control)52 [5.3]
SEHB (sulphate)15 [2.5]
AEHB (acetate)87 [1.5]
ASEHB (acetate+sulphate)2050 [180]
HEHB (H2/CO2)42 [4.2]
HSEHB (H2/CO2+sulphate)17 [1.8]

Sulphate concentrations fell more rapidly in the ASEHB slurries (after a lag of 1 day) than in the SEHB slurries (Fig. 1b). That the addition of acetate stimulated the utilisation of sulphate in these slurries suggests that the SRB in the sediment were substrate-limited. There was simultaneously a highly significant (P<0.0001) increase in signal from a Desulfobacter-targeted probe (p129), which was undetectable in the control slurries (CEHB) but was 21% of the total prokaryotic signal in the ASEHB slurries (Fig. 2c). Acetate concentrations in the ASEHB slurries were much higher (2 mM) than those in the other slurries, including the AEHB slurries (87 μM; see Table 3). In both these cases (AEHB and ASEHB) these results indicated significant consumption of added acetate. The higher residual acetate in the AS slurries, where methane production was significantly lower (P<0.0005) than in the AEHB slurries, suggests that acetate consumption was more effective when the acetoclastic methanogen community was not in direct competition with SRB.

One other result of note in these slurries was the enrichment of Desulfobulbus in the HEHB slurries (Fig. 2e). Although the increase in signal in these slurries was small (from 0.3% (S.E.M.=0.03%) in the control slurries to 1.1% (S.E.M.=0.1%)) in the HEHB slurries, it was significant (P<0.005). Signal from the Desulfobulbus-targeted probe p660 in the HEHB slurries was also significantly greater than that in the HSEHB slurries (P<0.05). Isolates of Desulfobulbus are known to be able to utilise H2/CO2 during sulphate reduction [37] but it was surprising to find that signal from this group was greater in the non-sulphate-amended HEHB slurries than in the sulphate-amended HSEHB slurries.

No significant changes in signal were detected with the Desulfobacterium-targeted probe (p221) despite the fact that this group represented a substantial proportion of the standing SRB signal in all of the slurries (3–5%; Fig. 2e).

3.2Effect of substrate addition and inhibitors in marine-dominated Colne Point slurries

The addition of H2/CO2 (HCP), H2/CO2+molybdate (HMCP), acetate (ACP) and trimethylamine (TMCP) to slurries from the marine-dominated Colne Point all caused substantial increases in methane production (Fig. 3a). Methane production in the TMCP slurries increased 1560-fold compared to the control slurries (Fig. 3a; F1,7, P<0.0005). The addition of trimethylamine had no effect on sulphate removal in these slurries (data not shown in Fig. 3b). There was, however, a highly significant increase in signal from the Methanosarcinales probe which, while undetected in the control slurries, comprised 16.8% (S.E.M.=1.0%) of the total prokaryotic signal in the TMCP slurries (Fig. 4b; P<0.0001), by far the largest response detected for an archaeal group in these experiments.

Figure 3.

a: Headspace methane accumulation. b: Sulphate concentrations. c: Acetate concentration in Colne Point slurries. Sulphate data from CMCP, CCCP, HMCP and TMCP slurries were essentially identical to that of the control slurries (CCP) and are not shown.

Figure 4.

Results of hybridisation of SRB- and MA-targeted oligonucleotide probes to rRNA extracted from Colne Point sediment slurries. Data expressed relative to the combined signal from general bacterial (p338) and archaeal (p915) probes. Signal level means were determined after arcsine transformation of the data, then back-transformed to give the quoted results and statistically analysed using a one-way (ANOVA) with a post hoc Tukey test after log transformation of the data [32]. Error bars=1 S.E.M. (n=3).

Methane production in the H2/CO2-amended slurries (HCP and HMCP) also increased significantly, by 1320-fold and 3410-fold, respectively (Fig. 3a; P<0.0005). However, in the HCP slurries the production of methane was inconsistent between the replicates with one slurry beginning methane production on day 5 while the other two did not produce substantial amounts of methane until day 14. The three slurries also produced widely varying amounts of methane (range 2170–12 840 nmol ml−1), which explains the large errors associated with this treatment. Interestingly, by day 14 sulphate concentration in these slurries was about 1.7 mM (Fig. 3b), a concentration at which sulphate reduction has been reported to become sulphate-limited in marine sediments [38,39]. No such variation was seen in the HMCP slurries, where large amounts of methane accumulated and no sulphate was consumed (Fig. 3a,b). However, no significant changes in signal from the two methanogen-targeted probes used were detected (Fig. 4a,b). In addition, despite the fact that sulphate was consumed in the HCP slurries, the only significant change was a reduction in signal from the Desulfobacterium probe (Fig. 4e; P<0.05) compared to the control slurries. In the HMCP slurries signal from both Desulfobulbus and Desulfobacterium decreased significantly (Fig. 4d,e; P<0.05 in each case) compared to the control slurries.

The smallest increase in methane production (three-fold) was in the ACP slurries (Fig. 3a) and was limited to just 3 days (days 7–9) of the 14-day experiment. The methane produced represented <0.1% of the added acetate. Again no significant changes were detected in the methanogen community, but the addition of acetate did significantly affect the probe signal from the SRB community. Sulphate was consumed along with acetate in stoichiometric quantities (Fig. 3b,c) and signal from the Desulfobacter probe, p129, increased significantly from being undetectable in the unamended control slurries to 27% (S.E.M.=3.4%) of the total prokaryotic signal (Fig. 4c; P<0.0005) in the ACP slurries.

The addition of inhibitors had a substantial effect on methane production in the control slurries (control+molybdate (CMCP) and control+chloroform (CCCP)). The addition of molybdate (CMCP), a specific inhibitor of sulphate reduction [28], led to a large, 12-fold increase in methane production (P<0.00001). Signal from the Methanosarcinales probe (MSMX860, Fig. 4a) did increase in these slurries but this was not significantly different from the control slurries (P=0.11) and no significant increase was seen with the Methanomicrobiales probe (MG1200, Fig. 4b). The archaeal proportion (p915) of the total prokaryotic signal, however, did increase significantly (P<0.05), from 0.4% (S.E.M.=0.03) in the control to 5.3% (S.E.M.=1.1), indicating increased archaeal activity in these slurries. The inhibition of sulphate reduction in the CMCP slurries led to a significant reduction in signal from Desulfobulbus (Fig. 4d; P<0.05) compared to the control slurries (CCP), which mirrored the reduction in signal from Desulfobulbus in the molybdate-amended HMCP slurries. Signal from Desulfobacterium also decreased but this was not significant (Fig. 4e; P=0.21). No significant signal was detected from p129 (Desulfobacter, Fig. 4c) in any of the control slurries.

Although methanogenesis is not as important a process in the high-sulphate marine sediments of Colne Point as it is at East Hill Bridge, the addition of chloroform still caused a significant 10-fold reduction in methane production (CCCP and HCCP, Fig. 3a; P<0.001). However, there were no significant changes detected in signal for methanogen-targeted probes (Fig. 4a,b) during the experiment.

4Discussion

In most environmental systems, and particularly in sediments, carbon flow is limited primarily by the availability of electron donors and acceptors. Within dynamic systems such as estuaries the availability of labile organic matter and electron acceptors, such as sulphate, are the primary determining factors in the balance between sulphate reduction and methanogenesis. We currently have little understanding of the identity of the different groups of SRB and MA that are responsible for this activity in situ. However, linking in situ function to particular groups within microbial communities is extremely difficult. Here we have utilised model sediment slurry microcosms that can be easily manipulated in the laboratory to investigate aspects of carbon flow in estuarine sediments and to identify the SRB and MA involved in the oxidation of important sedimentary substrates.

In order to detect a change in the community that would allow us to link function and phylogeny it was necessary to perturb the system. A major concern when interpreting the data presented here is that the perturbations used could select for organisms that are best suited to high substrate concentrations at the expense of more ecologically important organisms. The rapidity of a metabolic response and the enrichment of pre-existing communities is strong evidence for the ecological relevance of this methodology. In only three slurries groups that are not detected in the control slurries are detected in amended slurries. These are Desulfobacter in the acetate-amended slurries ASEHB and ACP (Figs. 2c and 4c) and the Methanosarcinales in the TMCP slurries (Fig. 4a). In the case of Desulfobacter at East Hill Bridge signal from this probe was also detected in slurries that were amended only with sulphate (SEHB, Fig. 2c), indicating Desulfobacter do not require additional organic substrates to increase their activity. Desulfobacter have been previously detected at Colne Point [40]. Therefore it would appear that the detected Desulfobacter signal is from an active standing community. In the case of the TMCP slurries there is no lag in methane production, indicating the activity of a standing community and not enrichment of an otherwise dormant population. Thus, the data we present appear to be the amplified response of the natural communities to the increased availability of a substrate.

Results from probing RNA extracted from the slurries used in these experiments linked certain MA and SRB groups to acetate utilisation in situ. In the freshwater, low-sulphate conditions normally prevalent at East Hill Bridge the Methanosarcinales were associated with acetate consumption (AEHB slurries), but when we simulated the high-sulphate conditions that occur after a tidal incursion into East Hill Bridge (ASEHB slurries) Desulfobacter and not the Methanosarcinales was linked to acetate consumption. These results match our understanding of the ecophysiology of both the Methanosarcinales and Desulfobacter as acetoclasts [37,41]. The Methanosarcinales probe targets the acetoclastic genera Methanosaeta and Methanosarcina. We have previously hypothesised that genotypes related to Methanosaeta concilii are the predominant freshwater acetoclastic MA globally [42]. Directly linking the Methanosarcinales to in situ acetate consumption at East Hill Bridge, a predominantly freshwater estuarine site, is entirely consistent with this hypothesis.

While an SRB community (Desulfobacter) was stimulated by the addition of acetate and sulphate in the East Hill Bridge slurries (ASEHB), acetate consumption was more rapid in the methanogenic AEHB slurries. This suggests that while methanogenesis is the normal acetoclastic process at this site, acetoclastic sulphate reduction can become more important when sulphate concentrations increase. The effect of sulphate addition in these sediments was seen in all the sulphate-amended slurries (SEHB, ASEHB and HSEHB) which, in the case of SEHB and HSEHB, caused a significant decrease in methane production and reduced both residual acetate concentrations and signal from Methanosarcinales- and Methanomicrobiales-targeted probes. Thus the addition of sulphate fundamentally affected carbon flow in these sediments and enhanced sulphate reduction at the expense of methanogenesis. The SRB group responsible (Desulfobacter) was only unambiguously identified in the ASEHB slurries and while signal from Desulfobacter increased in the SEHB slurries it was not statistically significantly different from the control slurries (P=0.11). It appears that Desulfobacter are capable of a rapid response to an increase in available sulphate and acetate in these sediments. This ability to respond to elevated sulphate concentrations, such as those caused by the occasional tidal incursion into this predominantly freshwater site, is an example of an ’r’ survival strategy [43], which could confer a significant ecological advantage on Desulfobacter within a dynamic estuarine environment.

The addition of acetate to slurries from Colne Point (ACP) also led to a highly significant increase in signal from Desulfobacter, along with stoichiometric utilisation of sulphate and acetate. We have previously linked this genus to the degradation of acetate in marine-dominated estuarine sediments [8] which, in conjunction with these data, support an important role for Desulfobacter in the in situ degradation of acetate. In contrast, stable isotope probing of phospholipid fatty acids has linked acetate utilisation in other marine sediments to Desulfotomaculum and to Desulfofrigus, an acetate-utilising SRB related to Desulfococcus[44,45]. Further work, preferably using both approaches on the same samples, is necessary to explore the basis for this apparent incongruence and to determine if it is biological or methodological. Methane production did increase in these acetate-amended slurries (ACP) but was limited to only 3 days and accounted for less than 0.1% of the added acetate. Therefore acetate degradation in these marine sediments was an SRB-dominated process and the limited capacity of indigenous MA to utilise acetate required substantial lag periods (7 days in this case), and ceased when acetate concentrations fell below the threshold that can maintain methanogenesis.

These results indicate that the dynamics of sulphate reduction and methanogenesis in the Colne estuary sediments were entirely consistent with our understanding of balance between these two processes being driven primarily by the availability of sulphate. In the predominantly freshwater, low-sulphate sediments of East Hill Bridge acetate degradation, and therefore the majority of anaerobic terminal oxidation of organic matter, was dominated by methanogenesis by the Methanosarcinales. Sulphate reduction and Desulfobacter dominated at those times when sulphate concentrations were high, such as after an incursion of seawater. At Colne Point the permanently high sulphate concentrations of this marine-dominated site result in sulphate reduction by Desulfobacter dominating acetate consumption, with a negligible contribution from acetoclastic methanogenesis. These data suggest that the dominant SRB and MA in these sediments have already been cultured and have been studied in some detail. In a microbial world beset by uncultivated diversity it is perhaps comforting to have evidence which suggests that we do appear to have some environmentally important organisms in culture and under study.

While acetate is the primary substrate for MA and SRB in sediments, little acetate is available to MA in the marine-dominated sediments at Colne Point. Methanogenesis still occurs, primarily via the oxidation of C1-compounds that SRB cannot utilise [9]. The addition of trimethylamine (TMCP slurries), an important C1-compound in marine sediments [7,46] led to a highly significant increase in signal from the Methanosarcinales, linking this group to trimethylamine consumption in situ. The Methanosarcinales include C1-compound utilising MA from Methanosarcina, Methanolobus and Methanococcoides[41]. Clones related to the obligate C1-compound users of Methanolobus and Methanococcoides were detected in clone libraries from Colne Point [47], suggesting that these genera are the primary candidates for trimethylamine consumption in these marine-dominated sediments. As was seen with Desulfobacter in the acetate-amended slurries, the response of the Methanosarcinales to TMA addition was large and highly significant. This would suggest a community severely limited by the availability of a labile carbon source but capable of a rapid and substantial response to that substrate when it becomes available.

Data from these slurries also linked Desulfobulbus to non-sulphate-dependent hydrogen utilisation (HEHB slurries). This suggests that under the sulphate-limited conditions that usually prevail at East Hill Bridge Desulfobulbus utilise an alternative electron acceptor to sulphate, possibly nitrate or nitrite [37]. Water-column nitrate concentrations in the upper reaches of the Colne estuary are usually high (annual mean approximately 0.5 mM [48]) and thus nitrate may represent a viable alternative electron acceptor to sulphate for facultative SRB/nitrate respirers at East Hill Bridge. Desulfobulbus sp. are also able to ferment a wide range of compounds, grow mixotrophically on H2 and an organic substrate [37] and Desulfobulbus propionicus can also utilise H2 to reduce ethanol and acetate to propionate [10] which could also explain the increase in signal detected. What was particularly noteworthy about this result was that Desulfobulbus was not enriched in the (hydrogen+sulphate)-amended slurries (HSEHB). This would suggest that at East Hill Bridge Desulfobulbus preferentially utilise an alternative electron acceptor to sulphate even when sulphate it is freely available. Other sulphate reducers can respire nitrate [37] while Desulfovibrio desulfuricans has been reported to use nitrate in preference to sulphate [49]. The reduced signal from Desulfobulbus and Desulfobacterium in the Colne Point slurries amended with molybdate (CMCP and HMCP) suggests that both genera are dependent on sulphate reduction in these marine-dominated sediments despite both being capable of utilising alternative electron acceptors to sulphate [37]. Unlike at East Hill Bridge, however, nitrate concentrations in the water column at Colne Point reach a maximum of only about 20 μM and are even lower in the sediment porewater [48]. Therefore the lack of a ready supply of an alternative to sulphate at Colne Point may explain the dependence of these two genera on sulphate reduction. Alternatively there may be a difference between the physiology, and presumably the phylogeny, of freshwater and marine Desulfobulbus, with the former being more physiologically flexible than the latter. Such a difference would add more complexity to the study of SRB, and by extension microbial distribution in general. Physiological flexibility within a genus will make the assignment of a consistent ecological role to that genus extremely difficult.

We have successfully linked several groups of MA and SRB to acetate and hydrogen consumption in these experiments but we failed to determine the in situ substrate for several other important SRB and MA groups, or identify all primary SRB and MA that utilised hydrogen in both East Hill Bridge and Colne Point.

The Methanomicrobiales were inhibited in all the sulphate-amended East Hill Bridge slurries (SEHB, ASEHB and HSEHB). Cultured members of the Methanomicrobiales are not acetotrophic [41], which indicates that the oxidation of other substrates, presumably H2, was affected by the addition of sulphate in these slurries. The activation of sulphate reduction in these slurries significantly reduced in situ acetate concentrations and presumably reduced H2 partial pressure below the thermodynamic threshold for methanogenesis [50,51]. However, the lack of enrichment of the Methanomicrobiales in the HEHB slurries would perhaps suggest that they may not utilise H2 in situ. An alternative explanation is that H2 is not limiting in these sediments. This could be due to the hydrogenotrophic methanogens at East Hill Bridge being closely associated with indigenous sources of H2 (e.g. anaerobic ciliates [51]) and thus were not substantially affected by the addition of headspace H2.

Despite the rapid consumption of sulphate in the HSEHB slurries no increase in signal from any of the SRB probes used was detected in these slurries. This would suggest that a group we have not detected mediated sulphate-dependent H2 consumption in these slurries. The obvious candidate genus for this activity is Desulfovibrio, for which we could not probe successfully in these experiments. Alternative hydrogen sinks, such as acetogens, may also be active in the brackish sediments of East Hill Bridge. However, residual acetate concentrations in these slurries match those in the control (CEHB) and sulphate-amended (SEHB) slurries (Table 3) and no significant increase was detected in either the Methanosarcinales or the Desulfobacter signal, both of which have been shown to respond to acetate in these slurries. Thus there is no evidence of acetate production in the HEHB or HSEHB slurries and therefore the dominant H2 sink organisms in these slurries seem unlikely to be acetogenic and remain unidentified.

We have been unable to link Desulfobacterium to utilisation of any substrate within these or previous slurries [8]. Isolated strains of Desulfobacterium are known to be able to utilise a wide range of substrates, including short-chain fatty acids, hydrogen, alcohols and, in several cases, aromatic compounds, and are able to utilise a range of alternative electron acceptors to sulphate, including nitrate and nitrite. Therefore it is possible that this group of flexible generalists use minor components within the carbon pool in sediments rather than compete with more specialised organisms, such as Desulfobacter, for the more common substrates such as acetate or hydrogen. There is ample evidence to suggest that co-metabolism of a range of substrates can be competitively advantageous especially under nutrient-limited conditions (see [52] and references therein). Therefore it may be that versatile organisms like Desulfobacterium co-metabolise a range of substrates and that this explains why they represent a significant proportion of the standing community but do not appear to respond to the single substrate additions we have reported here.

Acknowledgements

This work was funded by NERC grants GR3/8958 and GR3/10908 to T.M.E. and D.B.N. We thank C. Johnston for technical support and the Electron Microscopy and Mineral Analysis laboratory, Natural History Museum, London for their help in acetate analysis.

Ancillary