OUP user menu

Identity of active methanotrophs in landfill cover soil as revealed by DNA-stable isotope probing

Aurélie Cébron , Levente Bodrossy , Yin Chen , Andrew C. Singer , Ian P. Thompson , James I. Prosser , J. Colin Murrell
DOI: http://dx.doi.org/10.1111/j.1574-6941.2007.00368.x 12-23 First published online: 1 October 2007

Abstract

A considerable amount of methane produced during decomposition of landfill waste can be oxidized in landfill cover soil by methane-oxidizing bacteria (methanotrophs) thus reducing greenhouse gas emissions to the atmosphere. The identity of active methanotrophs in Roscommon landfill cover soil, a slightly acidic peat soil, was assessed by DNA-stable isotope probing (SIP). Landfill cover soil slurries were incubated with 13C-labelled methane and under either nutrient-rich nitrate mineral salt medium or water. The identity of active methanotrophs was revealed by analysis of 13C-labelled DNA fractions. The diversity of functional genes (pmoA and mmoX) and 16S rRNA genes was analyzed using clone libraries, microarrays and denaturing gradient gel electrophoresis. 16S rRNA gene analysis revealed that the cover soil was mainly dominated by Type II methanotrophs closely related to the genera Methylocella and Methylocapsa and to Methylocystis species. These results were supported by analysis of mmoX genes in 13C-DNA. Analysis of pmoA gene diversity indicated that a significant proportion of active bacteria were also closely related to the Type I methanotrophs, Methylobacter and Methylomonas species. Environmental conditions in the slightly acidic peat soil from Roscommon landfill cover allow establishment of both Type I and Type II methanotrophs.

Keywords
  • methane-oxidizing bacteria
  • landfill cover soil
  • DNA-stable isotope probing

Introduction

Methane is an important greenhouse gas with a high global warming potential. Its current atmospheric concentration is 1.7 parts (million vol)−1. The degradation of organic matter in landfill sites releases 36–73 Tg of methane annually (Howeling et al., 1999), which is about 6–12% of global methane emissions. Much of the biogas released into the atmosphere is from older and smaller landfill sites. Landfill gas migrates through the layers of the landfill cap where it may be subjected to oxidation by methanotrophic bacteria (Whalen et al., 1990; Jones & Nedwell, 1993) and landfill site cover soils have some of the highest aerobic methane oxidation capacity rates reported (Börjesson et al., 1998). However, methane removal rates reported for different landfill sites vary from 10% to 100% (Hilger & Humer, 2003). Considering the large amounts of methane produced in landfill sites, a deeper understanding of the methanotroph community structure directly controlling methane oxidizing activity is crucial for the control of greenhouse gas emissions.

The aerobic methane oxidizing bacteria use methane as sole carbon source. They are ubiquitous in nature and represent the largest biological sink for methane. They oxidize methane via methanol and formaldehyde to CO2 and incorporate carbon from methane into cell biomass. The first step in the pathway is catalyzed by methane mono-oxygenase (MMO). The soluble, cytoplasmic MMO (sMMO) is found in only some methanotrophs, whereas the particulate, membrane-bound MMO (pMMO) is present in all methanotrophs (Hanson & Hanson, 1996) except Methylocella silvestris (Dedysh et al., 2000). Methanotrophs are classified into two taxonomic groups, based on their 16S rRNA gene phylogeny, carbon assimilation pathways, phospholipid fatty acid (PLFA) profiles and intracellular membranes. Type I methanotrophs include Methylomonas, Methylobacter, Methylomicrobium, Methylosarcina, Methylosphaera, Methylococcus, Methylocaldum and Methylothermus. Type II methanotrophs include Methylosinus and Methylocystis (Hanson & Hanson, 1996; Bowman, 1999). Two recently described genera: Methylocapsa and Methylocella, form a tight taxonomic group with Beijerinckia indica and are related to Type II methanotrophs, based on their 16S rRNA gene sequences. Competition, mainly for methane, oxygen and nitrogen, might favour growth of different types of methanotroph. A slightly acidic landfill cover soil was shown to contain mainly Methylocystis and Methylosinus (Wise et al., 1999), whereas both Type I and Type II methanotrophs were detected in a landfill biofilter (Gebert et al., 2004) and cover soil (Börjesson et al., 1998; Uz et al., 2003). These studies provided information on the composition of methanotroph communities in different landfill covers, but do not give an insight into the active members of the community. The present study assesses this by stable isotope probing of 13C-enriched DNA (DNA-SIP), which is used to identify the active microorganisms in environmental samples. When using growth substrates enriched in a stable isotope such as 13C, the active microorganisms incorporate this isotope into their cellular components such as DNA, RNA or PLFAs (Radajewski et al., 2000; Dumont & Murrell, 2005; McDonald et al., 2005; Friedrich, 2006). Using DNA-SIP in combination with analysis of 16S rRNA genes and methanotroph-specific functional gene markers (pmoA, mmoX), we identified the active methanotroph population in a landfill cover peat soil.

Materials and methods

Sample description and potential methane oxidation rates

Roscommon landfill cover soil was sampled from a 10-year-old area of the remediated landfill in Ireland. The soil cover was a well-structured and established soil with a microbial community adapted to the landfill conditions, which vary considerably in terms of temperature, soil moisture and gas concentrations. The soil cap was a slightly acidic peat (pH 6.2) from a neighbouring bog; the sampling area was covered in grass that had vegetated naturally over 10 years. Soil was sampled in February 2005 from 20 to 40 cm depth of a 40-cm cover layer. Methane oxidation rates were measured in triplicate on 1 g of fresh soil (corresponding to 0.524 g of dry soil); the water content of this soil was 47.6%. Soil was placed in a sealed 120-mL crimp-top serum vial containing a head-space of 0.4% (v/v) methane, which was achieved by injecting 0.5 mL of 13CH4 using a gas-tight syringe. Samples were incubated in the dark at 20°C. Headspace CH4 concentrations were measured daily by sampling 0.2 mL of the headspace gas and manual injection on a gas chromatography equipped with a flame ionization detector (Cébron et al., 2007). The maximum potential methane oxidation rate determined by this method using saturating (high) concentrations of methane was 10.6±1.7 μmol CH4 (g dry soil)−1 day−1.

Stable isotope probing incubations

The SIP experiment was carried out using the methods described by Radajewski & Murrell (2001) and Morris et al. (2002). As previously shown (Cébron et al., 2007), the addition of nitrogen to soil can alter the active methanotroph community. Therefore, two different SIP incubations were carried out at 20°C on 5 g of fresh Roscommon soil placed in a 120-mL crimp-top serum vial. Both incubations were made into slurries, one with 5 mL of 1/10th strength nitrate mineral salts (NMS) medium (Whittenbury et al., 1970) and the second with the same volume of distilled water. Five mL of 13CH4 (>99% pure, Linde Gases) was injected into each soil slurry vial and methane concentration was measured as described above. After >90% of the 13CH4 had been consumed, the vials were flushed with air to remove any 13CO2 and to ensure that the slurries remained aerobic; another 5 mL of 13CH4 was then added. The vials were incubated until 0.6 mmol of 13CH4 (three injections of 5 mL of CH4) was consumed. Slurries were then centrifuged (25 000 g, 15 min) and soil pellets were stored at −80°C before DNA extraction.

DNA extraction and separation of 13C-DNA

DNA was extracted from both soil slurry pellets using a CO2-cooled bead beating method combined with the Bio 101 FastDNA spin kit for soil (QBiogene) as described by Cébron et al. (2007) and adapted from the method described by Radajewski & Murrell (2001) and Morris et al. (2002). DNA was eluted in 250 μL TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0).

Heavy (13C-labelled) and light (12C-labelled) DNA fractions from total DNA extracts were separated as described previously (Cébron et al., 2007). Ethidium bromide was removed from DNA fractions by butanol extraction and dialysis. After ethanol precipitation to remove CsCl, the four purified DNA fractions (heavy and light DNA from NMS and water supplemented slurries) were resuspended in 100 μL TE buffer and stored at −20°C.

Molecular analysis

PCR amplifications with each primer set were performed in a final volume of 50 μL containing 1 μL of 13C- or 12C-labelled template DNA at a concentration of 2 ng μL−1 as previously described (Cébron et al., 2007). The active methanotroph population from the two SIP microcosms was characterized at the domain level by cloning 1450-bp PCR products generated with the bacterial-specific 16S rRNA gene primer set 27F/1492R (Lane, 1991). The 12C- and 13C-DNA fractions from the two SIP microcosms were also characterized at a functional level with the primer set A189/mb661 (Costello & Lidstrom, 1999; Bourne et al., 2001) that amplified fragments of the pmoA gene. The heavy DNA fraction from the NMS-supplemented SIP microcosm was analyzed with primer sets 206F/886R (Lin et al., 2004) and mmoA/mmoB (Auman & Lidstrom, 2002) that amplified fragments of the mmoX gene. After PCR amplification with these four sets of primers, amplicons were purified with the Qiaquick PCR Purification kit (Qiagen). Clone libraries were generated according to the manufacturer's instructions by ligation of the purified PCR products into pCR 2.1 vector supplied with a TOPO TA cloning kit (Invitrogen, San Diego, CA). Colonies containing inserts of the correct length were randomly selected and cultivated in nutrient broth supplemented with ampicillin (50 mg mL−1). Twenty four clones were randomly selected from 16S rRNA and pmoA gene clone libraries and 20 clones for the mmoX libraries. Restriction fragment length polymorphism (RFLP) analysis was performed on plasmid mini-preps using a combination of EcoRI and RsaI restriction enzymes. RFLP patterns were resolved on 2% (w/v) agarose gels and grouped manually into operational taxonomic units (OTUs). One clone from each OTU was partially sequenced with the primer 357f (Lane, 1991) for the 16S rRNA gene clone libraries, and M13F for the pmoA and mmoX clone libraries.

16S rRNA gene PCR products, 550 bp in length, generated with the primers 341f with a GC clamp and 907r (Muyzer et al., 1998), were analyzed by denaturing gradient gel electrophoresis (DGGE) on 6% (w/v) polyacrylamide gels with a 30–70% denaturant gradient (urea and formamide). Gels were run at a constant voltage of 85 V for 16 h at 60°C in 1 × TAE buffer (40 mM Tris, 20 mM acetic acid, 1 mM EDTA; pH 8.3). Following electrophoresis, the gels were stained with SYBR Green I and scanned. Scans of DGGE gels were analyzed to determine the numbers of bands per profile. Distinct individual bands were excised, left overnight in 25 μL MilliQ water, re-amplified and run again on the DGGE system to ensure purity and correct mobility within the gel. PCR products yielding the correct pure bands were purified using the Qiaquick PCR Purification kit (Qiagen) and sequenced.

For microarray analysis, pmoA genes were amplified using the primers A189 and T7-mb661, the reverse primers containing the T7 promoter at the 5′ end. These primers enabled T7 RNA polymerase-mediated in vitro transcription using the PCR products as templates, to generate the RNA template for the hybridization on microarrays. RNA was used to challenge a microarray which contains a set of 59 oligonucleotides covering the whole known diversity of pmoA genes of methanotrophs. This pmoA microarray has been optimized for application in analysis of methanotrophic communities in landfill soils under different plant covers and has been extensively tested and validated with known pmoA genes and pmoA genes isolated directly from the environment (Bodrossy et al., 2003; Stralis-Pavese et al., 2004; L Bodrossy, JC Murrell et al., unpublished). The pmoA microarray was used exactly as described in Stralis-Pavese et al. (2004).

DNA sequencing and phylogenetic analysis

DNA sequencing was performed at the University of Warwick Central Molecular Biology Services Laboratory by cycle sequencing with the BigDye Terminator kit (Applied Biosystems, Warrington, UK) and ABI3100 capillary DNA sequencers. Phylogenetic analyses were carried out on 16S rRNA, pmoA and mmoX gene sequences from clone libraries and also from DGGE bands. Related environmental gene sequences from reference strains and clones were obtained from GenBank using the blastn program in blast version 2.1 (http://www.ncbi.nlm.nih.gov/BLAST), and then aligned with these new sequences using the clustalx program. Alignment analyses were performed on 450 bp of the 16S rRNA gene sequences, on 500 bp of the pmoA sequences and on 675 bp of the mmoX sequences. Dendrograms were constructed using the programs dnadist, neighbor, seqboot, and consense from phylip version 3.65 (Feselstein, 1989), and phylogenetic tree files were analyzed using tree view software as previously described (Cébron et al., 2007).

Nucleotide sequence accession numbers

The GenBank accession numbers for the 16S rRNA, pmoA and mmoX gene fragments sequenced in this study are EF625904EF525940, EF625940EF625960 and EF633468EF633474, respectively.

Results

Methane consumption during SIP incubations

The composition of active methane-oxidizing bacterial communities in Roscommon Landfill cover soil was studied by DNA-SIP following incubation with 13C-labelled methane and either nitrogen-rich NMS or water. Potential methane oxidation rates were different under these two contrasting conditions (Fig. 1) and were almost two-fold greater during incubation with NMS solution (14.0 μmol CH4 (g dry soil)−1 day−1) than with water (8.4 μmol CH4 (g dry soil)−1 day−1). The potential methane oxidation rate of the original soil (10.6±1.7 μmol CH4 (g dry soil)−1 day−1), determined immediately after sampling with nonlabelled methane, was of the same order of magnitude.

1

13C-CH4 oxidation by a 5-g Roscommon cover soil sample during two SIP experiments with NMS and water-supplemented slurries. Arrows indicate the three additions of 5 mL (4% v/v in headspace) of methane (at 0, 92 and 186 h for NMS slurry and at 0, 143 and 314 h for the water slurry). Error bars indicate the SD of triplicate methane concentration measurements.

pmoA sequence composition as revealed by clone library and microarray analyses

The composition of pmoA sequences was examined in both heavy and light DNA fractions recovered from the two contrasting SIP incubations by (1) construction of clone libraries coupled to RFLP analysis and phylogenetic affiliation of representative clone gene sequences (Fig. 2) and (2) microarray analyses (Fig. 3). In both SIP microcosms, the number of OTUs recovered from 13C-DNA was higher than that from 12C-DNA fractions, indicating a successful DNA-SIP experiment. For libraries constructed with 13C-DNA recovered from NMS- and water-supplemented slurries, six and seven pmoA OTUs, respectively, were found, against only four OTUs present in both pmoA libraries that had been constructed using 12C-DNA (Fig. 2).

2

Phylogenetic neighbour-joining tree of pmoA gene sequences from four clone libraries constructed using 12C and 13C-DNA fractions from the two SIP microcosm incubations. Names of sequences begin with ‘12’ or ‘13’ for the light and heavy DNA fractions, then ‘R’ for Roscommon soil and ‘NMS’ or ‘H2O’ respectively for NMS and water slurry incubations. ‘mb’ signifies the use of the mb661 primer and the last number is the clone number. Reference strains and clones sequences were taken from GenBank and are named by their accession numbers. The tree is rooted with the amoA gene sequence from Nitrosococcus oceanus. Bootstrap values greater than 70% derived from 1000 replicates are shown and were obtained using a distance matrix program neighbour-joining method within phylip 3.65. The bar represents 10% sequence divergence.

3

Microarray results showing the efficiency of hybridization of pmoA PCR products obtained on light (12C) and heavy (13C) DNA fractions from the NMS and water-supplemented slurries. For each probe spotted on the microarray, the hybridization efficiency relative to the proportion of PCR product hybridized on each probe is indicated by a grey grading scale, from white for no hybridization through grey to black for more than 50% of the PCR product hybridized (a value of 1 on the bar corresponds to black, the highest signal intensity obtained for the given probe during probe set validation). The corresponding OTUs (see Fig. 2) obtained from the pmoA clone libraries have been indicated. Ia=type Ia universal probes, II=type II universal probes. The sensitivity of the microarray allows the detection of methane oxidizing bacteria present at 5% or more of the total methanotroph community (i.e. the community targeted by the PCR).

The most abundant methanotroph pmoA sequences were from Type I methanotrophs closely related to Methylobacter sp. (OTU1 to OTU6, Fig. 2). OTU1 and OTU2 were present in different proportions in the four DNA libraries. OTU1 was closely related to pmoA from Methylobacter sp. LW12 and Methylobacter sp. SV96 but formed a separate branch with environmental clones (Fig. 2). This branch is represented by the clones MLM2 and MLA39 and the DGGE band P2 recovered from a meromictic soda lake (Lin et al., 2005) and clone 1HA_9, which was found in soda lake sediment (Lin et al., 2004). The methanotroph pmoA sequences that grouped into OTU1 were more abundant in the NMS incubation than in the water incubation. OTU2 clones were closely related to the pmoA from Methylobacter psychrophilus (Fig. 2) and to environmental pmoA clones recovered from Lake Washington sediment (clone pLWPmoA-2, Nercessian et al., 2005; uncultured eubacterium pAMC503; Costello & Lidstrom, 1999). The methanotroph pmoA sequences that grouped into OTU2 were present in Roscommon landfill cover soil but did not seem to represent a large part of the active methanotroph community. OTU3 was closely related to pmoA from Methylobacter albus and Methylobacter sp. LW1 (Fig. 2). Clone sequences from OTU3 were found in libraries constructed using 12C-DNA fractions but were in low proportions and even absent from the libraries constructed with 13C-DNA fractions from water and NMS slurries, respectively. OTU4 was closely related to OTU1, and represented a small fraction of the pmoA genes recovered from the NMS slurry (Fig. 2). OTU5, OTU6 and OTU7 were only recovered in 13C-DNA fractions from both SIP microcosms (Fig. 2). This indicates that these OTUs represent major groups of active methanotrophs. The pmoA clone sequences from OTU5 were closely related to a pmoA cluster represented by these genes from Methylobacter LW14, Methylobacter HG-1, Methylobacter albus and Methylobacter LW1 but they were in a separate branch with no other representative. The pmoA clones from OTU6 grouped with pmoA genes from Methylobacter sp. LW12 and Methylobacter sp. SV96 branching in the same cluster as clone pLWPmoA-13 (Fig. 2) isolated from Lake Washington sediment (Nercessian et al., 2005). The pmoA sequences from OTU7 were closely related to pmoA from Methylomonas species in a cluster represented by Methylomonas methanica and uncultured bacterium clone D1a (Fig. 2) isolated from a Danish soil (Bourne et al., 2001). The recovery of Methylomonas-like pmoA clones in active fractions of both SIP treatments was confirmed also by microarray results (Fig. 3). Finally, only the pmoA clones that grouped into OTU8 belonged to Type II methanotrophs and these were closely related to pmoA from Methylocystis sp. These pmoA sequences were only recovered from the water slurry SIP incubation and were in a higher proportion in the active population (Fig. 2). Microarray analyses also suggested that the ratio of Type II to Type I methanotrophs was higher in the water-supplemented slurry than in the NMS slurry. However, in contrast with the clone library results, pmoA microarray analyses indicated that Type II methanotrophs, related to Methylocystis, were better represented than Type I (Fig. 3). The Methylomicrobium-related pmoA signal obtained with probe Mmb562 was possibly a false positive because there was no signal with the related probe Mmb303 (Fig. 3). Similarly, Methylocapsa-related pmoA signals obtained with probe B2all343 (a generic probe targeting pmoA of Methylocapsa and related environmental clones) is a false positive (Fig. 3) because it can be considered positive only if the corresponding specific probes (B2-400 or B2rel251) are positive.

Analysis of 16S rRNA gene clone libraries and DGGE

Four 16S rRNA gene libraries were constructed with 12C-DNA and 13C-DNA from the two SIP incubation experiments. The 48 clones randomly selected from the two libraries made with 12C-DNA fractions gave 48 different RFLP patterns, none of which was found in clones from 13C-DNA libraries. Thus, only 16S rRNA gene fragments from the heavy DNA clone library were sequenced and analyzed phylogenetically. Nine and 13 different OTUs (Fig. 4) were found for the 16S rRNA clone libraries constructed using the 13C-DNA from the NMS and water slurry SIP incubation experiments, respectively. Approximately the same proportion (around 55–63% of the clones) of 16S rRNA gene sequences belonging to methanotrophs was recovered from the two libraries (Fig. 4). However, the distribution of Type I and Type II 16S rRNA gene sequences was different (Fig. 4). The water slurry seemed to favour Type II methanotrophs, as previously shown by the pmoA microarray analysis. Type I methanotroph 16S rRNA gene sequences were separated into three OTUs (I, II and III, Fig. 4) that grouped within a branch represented by Methylobacter psychrophilus, Methylobacter sp. SV96, Methylobacter sp. T20 and uncultivated clones recovered from various soils. Type II methanotrophs were more frequently detected in 16S rRNA gene clone libraries than when targeting the pmoA gene. The methanotrophs that do not possess the pmoA gene or those with a pmoA gene that was not readily amplified with the applied pmoA specific PCR primers can then be detected by targeting 16S rRNA gene. Indeed, Methylocella and Methylocapsa acidophila-like clone sequences (OTUs V, VI and VII, Fig. 4) were recovered from both heavy DNA fractions and represented 6% and 42% of the 16S rRNA gene clones, respectively, from NMS and water-supplemented slurries. In addition, Methylocystis-related 16S rRNA gene clones (OTU IV) represented 17% of the active NMS slurry community and only 5% of the water slurry active community (as determined by examination of 13C-DNA).

4

Phylogenetic neighbour-joining tree of 16S rRNA gene sequences from two clone libraries and from DGGE bands (see Fig. 4) representing the active bacterial populations (13C-DNA fractions) from NMS and water slurry SIP microcosm incubations. Names of sequences begin with ‘R’ for Roscommon soil and then ‘NMS’ and ‘H2O’ respectively for NMS and water slurry incubations during SIP. The last number is the clone number and the DGGE band sequences are specified. Reference strains and clones sequences were taken from GenBank and are named by their accession numbers. The tree is rooted with the 16S rRNA gene sequence from Clostridium sp. 9B4. Bootstrap values greater than 70% derived from 1000 replicates are shown and were obtained using a distance matrix program neighbour-joining method within phylip 3.65. The bar represents 10% sequence divergence.

A high proportion of methylotrophs was also recovered in the heavy DNA fractions, representing 39% and 28% of 16S rRNA gene clones from NMS and water slurry SIP experiments, respectively (Fig. 4). Some methylotrophs (OTUs IX, X and XI) were closely related to Methylophilus sp. and the others (OTUs VIII and XIII) were closely related to Hyphomicrobium strains previously described as methylotrophs, growing on methanol (but not methane). Finally, the remainder of the 16S rRNA gene clones were from other bacteria closely related to the genera Clostridium, Planctomycetales and Bdellovibrio (Fig. 4), whose ability to grow on methane or other one carbon compounds has not been reported.

16S rRNA gene sequences were also studied by DGGE analyses (Fig. 5). The DGGE patterns obtained for the 12C-DNA fractions gave many bands because the entire bacterial community present in the soil matrix was targeted. It seems that the DGGE profiles between 12C- and 13C-DNA are similar in terms of numbers of bands but dominant bands in DGGE profiles with 13C-DNA differ from those with 12C-DNA. Individual bands were easily visualized on DGGE profiles corresponding to the dominant active bacteria in both SIP microcosms. The major bands (Fig. 5) were sequenced and subjected to phylogenetic analysis (Fig. 4).

5

DGGE pattern of 16S rRNA gene PCR products obtained from the 12C- and 13C-DNA fractions of the NMS and water slurry SIP microcosm incubations. Bands that were re-amplified, purified and sequenced (see phylogenetic affiliation on Fig. 4) are labelled with circles and phylogenetic affiliations of these 16S rRNA gene sequences, according to blastn, are indicated.

Two major 16S rRNA gene bands present in both SIP treatments (RNMS-5DGGE and -6DGGE, RH2O-1DGGE and -2DGGE, Fig. 5) were from Type I methanotrophs closely related to M. psychrophilus, Methylobacter sp. SV96 and Methylobacter sp. T20, and were also closely related to OTUs I, II and III from clone libraries (Fig. 4). For the water slurry incubation, a third Type I methanotroph 16S rRNA gene band was characterized (RH2O-3DGGE, Fig. 5), which was closely related to Methylobacter sp. 5FB and the environmental methanotroph 16S rRNA gene clone 5H_48 isolated from Transbaikal soda lake sediment (Lin et al., 2004). Type II methanotrophs were represented by

  1. one of the prominent DGGE bands (RNMS-9DGGE and RH2O-6DGGE) closely related to clone LO13.10 isolated from peat soil (Morris et al., 2002), which had a 16S rRNA gene sequence that belonged to the Methylocapsa and Methylocella branch,

  2. the DGGE bands RNMS-10DGGE and RH2O-8DGGE, which are closely related to 16S rRNA genes of Methylocystis genera, and

  3. the two bands RH2O-4DGGE and RH2O-7DGGE are related to uncultured methanotroph clones found in peat and forest soils, respectively (Fig. 4) and to Methylocystis sp. L32, WRS and IMET10486 respectively (Fig. 4).

The methylotroph-related DGGE band (RNMS-1DGGE closely related to Methylophilus species, Fig. 4) and few other bacterial genera bands (RNMS-3DGGE, RNMS-7DGGE and RNMS-8DGGE, Fig. 5) were detected only in the 13C-DNA fraction from the NMS slurry SIP incubation.

Analyses of mmoX clone libraries

The diversity of active methanotrophs possessing the sMMO was investigated by constructing clone libraries using two different primer sets: 206F/886R and mmoA/mmoB. Figure 6 shows the phylogenetic distribution of representative mmoX clone sequences. The majority of mmoX clones were closely related to Methylocystis sp., but the first primer set permitted the detection of mmoX clones (10% of libraries) that were closely related to mmoX from Methylomonas species and mmoX clones (15% of libraries), which were closely related to the mmoX from Methylocella species.

6

Phylogenetic neighbour-joining tree of mmoX gene sequences from two clone libraries constructed using the 13C-DNA fraction from the NMS slurry SIP microcosm incubation. The sequences designated mmoX1 were obtained from a clone library constructed with 206F/886R primers and the sequences named mmoX2 were obtained from a clone library made with mmoA/mmoB primers. Reference strains and clones sequences were taken from GenBank and are named by their accession numbers. The tree is rooted with the partial butane monooxygenase gene sequence (BmoX) from Pseudomonas butanovorans. The bar represents 10% sequence divergence.

Discussion

Using several complementary molecular ecology techniques, we identified the active methanotroph population from the Roscommon landfill cover soil using DNA-SIP under two nutrient conditions to minimize the bias of species selection during incubation (Cébron et al., 2007). Recent studies on landfills have utilized various analytical approaches to investigate methane oxidation and methanotroph community composition in these complex environments (Wise et al., 1999; Uz et al., 2003; Börjesson et al., 2004; Crossman et al., 2004; Stralis-Pavese et al., 2004). Previously reported values for methane oxidation in landfill cover soils ranged from 0.93 to 1.62 μmol CH4 (g dry weight soil)−1 h−1 for mineral soils and from 8 to 25 μmol CH4 (g dry weight soil)−1 h−1 for organic matter rich cover soils (Börjesson et al., 2004). The latter activities are similar to those measured in Roscommon landfill cover soil at 10.6±1.7 μmol CH4 (g dry weight soil)−1 h−1, indicating that this landfill peat soil was a highly active methane oxidizing environment.

Aerobic methanotrophs are abundant in wetland soils (Segers, 1998) and in well-aerated upland soils where they consume atmospheric CH4 (Conrad, 1996). Landfill soils such as those from Roscommon appear to be intermediate, with spatial and seasonal variations in CH4 oxidation (Bogner et al., 1997; Börjesson et al., 1998). Methanotroph diversity depends on several environmental parameters: CH4 concentration and O2 availability (Amaral & Knowles, 1995), pH (Hilger et al., 2000), temperature (Börjesson et al., 2004), nitrogen source (Hanson & Hanson, 1996; Bodelier et al., 2000), soil structure and moisture content (Boeckx & Cleemput, 1996; Börjesson et al., 1998). No clear conclusions could be made about the niches occupied by Type I vs. Type II methanotroph species because of conflicting published results, with Type I and Type II methanotrophs each dominating in some ecosystems at either low or high CH4 concentrations (Hanson & Hanson, 1996; Henckel et al., 2000; Macalady et al., 2002). However, Type II methanotrophs typically dominate in low CH4 mixing ratio [<30 parts (million vol)−1] (Knief & Dunfield, 2005). In this study, both Type I and Type II methanotrophs were active in Roscommon landfill cover soil. Several other landfill studies showed the presence of both Type I and Type II methanotrophs, identified by 16S rRNA gene DGGE (Wise et al., 1999) and clone libraries (Uz et al., 2003), T-RFLP (Stralis-Pavese et al., 2006), PLFA (Börjesson et al., 1998; Crossman et al., 2004) and diagnostic pmoA microarray analyses (Stralis-Pavese et al., 2004). In some cases, Type II methanotrophs dominated the community, as revealed by PLFA analyses on biofilters charged with landfill gas (Gebert et al., 2004). Type I strains might be expected to dominate in nutrient-rich environments (Amaral & Knowles, 1995; Börjesson et al., 1998; Wise et al., 1999), which could explain why Type I methanotrophs were the most abundant in the SIP experiment carried out in NMS slurry. Indeed, results from clone library analyses indicated that clones closely related to pmoA from Methylobacter and Methylomonas represent a large proportion of active bacteria during SIP experiments and in situ (as found from mRNA analyses of pmoA genes in the same landfill soil sample (Chen et al., 2007). Moreover, NMS medium provides copper in its trace element solution, another key factor which may regulate growth of Type I vs. Type II methanotrophs (reviewed in Murrell et al., 2000; Lieberman et al., 2003). Under low copper conditions, as found in the water slurry, Type II methanotrophs and a few Type I genera such as Methylococcus and Methylomonas can produce sMMO (Miguez et al., 1997), which might allow them to out-compete methanotrophs with only pMMO (Hanson & Hanson, 1996; Murrell et al., 2000). These observations are in agreement with our results showing that (1) Type II methanotrophs closely related to Methylocystis were the most active bacteria according to pmoA microarray analyses for the water-supplemented SIP slurry, (2) pmoA clones closely related to Methylomonas also formed one dominant active clade, as indicated by pmoA library clone analysis, and (3) Type II methanotrophs related to Methylocystis and Methylocella dominate the active population when 16S rRNA gene diversity was analyzed. This last result was substantiated by mmoX clone library analyses where Methylocystis spp. were highly represented. The sMMO exhibits a broad substrate spectrum (Dalton & Stirling, 1982), which could confer higher co-metabolism turnover rates for bacteria such as Methylocystis and Methylocella, which were recovered in high proportions from the Roscommon landfill soil. These Type II methanotrophs were also found by Chen et al. (unpublished results) to be highly active in situ when using DGGE analyses of 16S rRNA gene transcripts. In situ, these sMMO-possessing bacteria might be more tolerant to the presence of pollutants in landfill soils, as such complex landfill environments can favour the maintenance of bacteria that are able to degrade methane and other carbon compounds (Kjeldsen et al., 1997; Hilger & Humer, 2003). The recently discovered acidophiles Methylocella palustris and M. acidophila are potentially important in terrestrial soils and are frequently found in mildly acidic environments such as the Roscommon landfill peat soil.

Interpretation and experimental pitfalls of the DNA-SIP approach have been extensively discussed (Radajewski et al., 2000; Dumont & Murrell, 2005; McDonald et al., 2005; Neufeld et al., 2007) and the experimental artefacts created by in vitro SIP microcosms have been highlighted by Cébron et al. (2007) and, in this study, by the use of two different conditions for SIP incubations. The water-supplemented SIP slurry conditions were probably the closest to in situ conditions and should reveal representative methanotrophs active in situ. The techniques used here to describe populations of active methanotrophs have their strengths and limitations. Firstly, analyses of functional genes pmoA and mmoX allowed targeting of methanotrophs possessing pMMO and sMMO but limited the analyses to bacteria closely related to extant methanotrophs from which the primer sets were designed. However, one major advantage of DNA-SIP is that the diversity of active bacteria can be resolved by targeting the 16S rRNA genes, which allows analysis independent of prior knowledge. Results from pmoA clone libraries and microarrays showed that the diversity of pmoA was slightly higher in the active 13C-DNA than in the 12C-DNA fraction, which we interpret as meaning that the SIP incubation conditions were not too selective and allowed the growth of a variety of methanotrophs. Interestingly, according to microarrays results, the ratios of Type I to Type II methanotrophs were approximately the same for 13C and 12C DNA, suggesting that, for both SIP incubations, the composition of the active community seems to reflect that of the whole methanotrophic population without any selection or enrichment during incubation. Even if the same pmoA primer set was used for clone library and microarray studies, the clone library technique seems to overestimate the presence of Type I methanotroph. This could be due to the differences in experimental conditions (different primers or PCR conditions) or a better cloning yield of Type I vs. Type II pmoA PCR products. Finally, based on 16S rRNA gene analysis, both active bacteria that consumed 13CH4 and 13C-labelled derived compounds were targeted and the results obtained from clone libraries and sequencing of DGGE bands were highly congruent. By targeting the 16S rRNA gene, nonmethanotrophs may also have been detected, which had been assimilating 13C-labelled metabolites such as 13CH3OH and 13CO2 or 13C-labelled metabolites arising from primary consumers of methane (Lueders et al., 2006). Methylotrophs such as Methylophilus, Methylovorus, Aminomonas and Hyphomicrobium were found in relatively high proportions in 16S rRNA gene libraries and may have resulted from incorporation of carbon from the methanol produced by oxidation of 13CH4.

Conclusions

DNA-SIP was applied to a landfill cover soil in which the methanotrophic community was extensively examined through clone library, microarray and DGGE analyses on pmoA, mmoX and 16S rRNA genes. Type II methanotrophs closely related to the genera Methylocella-Methylocapsa dominated the active population, as indicated by 16S rRNA gene analyses. pmoA and mmoX analyses indicated that Methylocystis-related bacteria were the dominant Type II methanotrophs present in both samples, together with representatives of Type I methanotrophs closely related to both Methylobacter and Methylomonas.

Acknowledgements

We thank Dr Ruth Nevin (Environmental Microbiology Research Unit, Department of Microbiology NUI, Galway) for the collection of samples from Roscommon Landfill. This work was funded by the Natural Environment Research Council, UK, through grant NE/B505389/1.

Footnotes

  • Editor: Gary King

References

View Abstract