Identification of Active Methylotroph Populations in an Acidic Forest Soil

Home , Cytophagales, Methylocapsa acidiphila, Methylocella palustris, Methylotroph, Rhodoblastus acidophilus

Microbiology (2002), 148, 2331–2342 Printed in Great Britain

Identiﬁcation of active methylotroph populations in an acidic forest soil by stable- isotope probing

Stefan Radajewski,1 Gordon Webster,2† David S. Reay,3‡ Samantha A. Morris,1 Philip Ineson,4 David B. Nedwell,3 James I. Prosser2 and J. Colin Murrell1

Author for correspondence: J. Colin Murrell. Tel: j44 24 7652 2553. Fax: j44 24 7652 3568. e-mail: cmurrell!bio.warwick.ac.uk

1 Department of Biological Stable-isotope probing (SIP) is a culture-independent technique that enables Sciences, University of the isolation of DNA from micro-organisms that are actively involved in a Warwick, Coventry CV4 7AL, UK speciﬁc metabolic process. In this study, SIP was used to characterize the active methylotroph populations in forest soil (pH 35) microcosms that were exposed 2 Department of Molecular 13 13 13 13 and Cell Biology, to CH3OH or CH4. Distinct C-labelled DNA ( C-DNA) fractions were resolved University of Aberdeen, from total community DNA by CsCl density-gradient centrifugation. Analysis of Institute of Medical 16S rDNA sequences ampliﬁed from the 13C-DNA revealed that bacteria related Sciences, Foresterhill, Aberdeen AB25 2ZD, UK to the genera Methylocella, Methylocapsa, Methylocystis and Rhodoblastus had assimilated the 13C-labelled substrates, which suggested that moderately 3 Department of Biological Sciences, University of acidophilic methylotroph populations were active in the microcosms.

Essex, Wivenhoe Park, Enrichments targeted towards the active proteobacterial CH3OH utilizers were Colchester, Essex CO4 3SQ, successful, although none of these bacteria were isolated into pure culture. A UK parallel analysis of genes encoding the key enzymes methanol dehydrogenase 4 Department of Biology, and particulate methane monooxygenase reflected the 16S rDNA analysis, but University of York, PO Box 373, YO10 5YW, UK unexpectedly revealed sequences related to the ammonia monooxygenase of ammonia-oxidizing bacteria (AOB) from the β-subclass of the Proteobacteria. Analysis of AOB-selective 16S rDNA amplification products identified Nitrosomonas and Nitrosospira sequences in the 13C-DNA fractions, suggesting 13 certain AOB assimilated a significant proportion of CO2, possibly through a close physical and/or nutritional association with the active methylotrophs. Other sequences retrieved from the 13C-DNA were related to the 16S rDNA sequences of members of the Acidobacterium division, the β-Proteobacteria and the order Cytophagales, which implicated these bacteria in the assimilation of reduced one-carbon compounds or in the assimilation of the by- 13 products of methylotrophic carbon metabolism. Results from the CH3OH and 13 CH4 SIP experiments thus provide a rational basis for further investigations into the ecology of methylotroph populations in situ.

"$ Keywords: methanotrophs, community structure, culture-independent techniques, C, functional genes

...... INTRODUCTION † Present address: Cardiff School of Biosciences, Cardiff University, Main Building, Cardiff CF10 3TL, UK. Methylotrophs are micro-organisms that can use re- ‡ Present address: Institute of Ecology and Resource Management, duced one-carbon compounds as a sole source of carbon University of Edinburgh, Darwin Building, Edinburgh EH9 3JU, UK. and energy (Lidstrom, 1992). Two important methylo- Abbreviations: AOB, ammonia-oxidizing bacteria; DGGE, denaturing gradient gel electrophoresis; g.d.w., g dry weight; OTU, operational trophic substrates in the terrestrial environment are taxonomic unit; SIP, stable-isotope probing. methanol (CH$OH) and methane (CH%). CH$OH is The GenBank accession numbers for the sequences reported in this paper produced during the decomposition of plant polymers are AY080911–AY080961. and is probably excreted from living leaves and by CH%-

0002-5503 # 2002 SGM 2331 S. Radajewski and others oxidizing bacteria (Anthony, 1982), whereas CH% is Difficulties associated with the isolation of the vast mainly produced by methanogenic Archaea in anoxic majority of micro-organisms into laboratory culture environments. Methylotrophs that grow on CH$OH have hampered the identification of microbial popula- and CH% have been isolated from many environments tions involved in a specific process (Gray & Head, 2001; and the characterization of pure cultures has provided Hugenholtz et al., 1998). Several recently described information on the taxonomy, physiology, genetics and techniques can, however, establish this link between ecology of these physiologically specialized micro- identity and function under conditions approaching organisms. those in situ and without prior knowledge of the micro- organisms involved (Boschker et al., 1998; Lee et al., Most aerobic methylotrophs that are in laboratory 1999; Orphan et al., 2001; Ouverney & Fuhrman, 1999; culture belong to the class Proteobacteria. All of these Radajewski et al., 2000). Certain phospholipid-fatty- methylotrophs use the enzyme methanol dehydrogenase acid- and DNA-labelling techniques have been devel- to oxidize CH OH to formaldehyde, which is the central $ oped using one-carbon substrates enriched with radio- intermediate of one-carbon metabolism. Formaldehyde "% "$ active-carbon ( C) or stable-carbon ( C) isotopes; can subsequently be assimilated into cell carbon (Lid- these have been subsequently applied to identify the strom, 1992) or can be oxidized further to formate and active methylotroph communities in environmental CO to gain energy and to regenerate reducing equiva- # samples directly (Boschker et al., 1998; Radajewski et lents (reviewed by Vorholt et al., 1999). CH -oxidizing % al., 2000; Roslev & Iversen, 1999). bacteria (methanotrophs) are a subgroup of methylotrophs that have been divided into types I and II (γ- and In a preliminary report, we described the development α-Proteobacteria, respectively) based on phenotypic and and application of stable-isotope probing (SIP) (Rada- phylogenetic characteristics (Hanson & Hanson, 1996). jewski et al., 2000). This technique exploits the fact that "$ "& Methanotrophs oxidize CH% to CH$OH using the certain stable isotopes (e.g. C and N) occur at low enzyme methane monooxygenase, which occurs in two abundance in nature but can be highly enriched (i.e. distinct forms that have been reviewed in detail by "99%) in commercially available compounds. As Murrell et al. (2000b). A particulate, membrane-bound demonstrated by Meselson & Stahl (1958) with Escheri- "& + methane monooxygenase has been reported in all chia coli grown on NH%, DNA synthesized during methanotrophs except Methylocella palustris (Dedysh microbial growth on a substrate enriched with a heavy, et al., 2000), whereas only certain strains contain a stable isotope becomes labelled sufficiently to be re- soluble, cytoplasmic methane monooxygenase. The solved from unlabelled DNA by density-gradient centri- particulate methane monooxygenase has several simi- fugation. This principle has been demonstrated with "$ larities to the ammonia monooxygenase of ammonia- methylotroph and AOB cultures grown on CH OH "$ $ oxidizing bacteria (AOB), and it has been suggested that (Radajewski et al., 2000) and CO# (Whitby et al., 2001) these two enzymes may be evolutionarily related as a carbon source, respectively. Although the buoyant (Holmes et al., 1995). Although some AOB can oxidize density of DNA varies with its GjC content, the CH , none are known to use it as an energy source incorporation of a high proportion of a naturally rare, % # "& "$ (Be! dard & Knowles, 1989). stable isotope ( H, Nor C) into DNA enhances Functional genes encoding key enzymes in the methylo- greatly the density difference between labelled and trophic pathways of several Proteobacteria have been unlabelled DNA fractions (Rolfe & Meselson, 1959; identified. The large subunit of methanol dehydro- Vinograd, 1963). Indeed, SIP has been applied suc- cessfully to identify active CH OH utilizers in an oak genase, which contains the active site of the enzyme, is $ "$ encoded by mxaF (Lidstrom et al., 1994; Nunn & forest soil (Radajewski et al., 2000). Isolation of the C- Lidstrom, 1986). The putative active site subunit of labelled DNA fraction simultaneously established the particulate methane monooxygenase is encoded by function and activity of the micro-organisms and pmoA, with the homologous subunit in ammonia sequence analysis subsequently determined their iden- monooxygenase encoded by amoA (reviewed by Murrell tity. The aim of the present study was to characterize the active aerobic methylotroph populations in an acidic et al., 2000a). Conserved domains within the amino-acid "$ "$ sequences have enabled the design of PCR primers that oak forest soil exposed to CH$OH or CH% and to isolate certain methylotrophs detected in our prelimin- specifically amplify methane monooxygenase or meth- "$ anol dehydrogenase genes from methylotroph, enrich- ary study using CH$OH. ment culture and environmental DNA samples (Dedysh et al., 2002; Dunfield et al., 1999; Henckel et al., 1999; METHODS Holmes et al., 1999; McDonald & Murrell, 1997). In cultivated strains, the phylogenetic relationships be- Soil and microcosm experimental conditions. Microcosm tween the derived amino-acid sequences of pmoA and experiments were used to characterize the active methylotroph amoA, and to a lesser extent mxaF, reflect those obtained communities in an oak (Quercus petraea) forest soil (0–5 cm depth sample; pH 3n5) collected from the Gisburn Forest with 16S rDNA sequences (Holmes et al., 1995; Experiment, which was planted in 1955 on former sheep McDonald & Murrell, 1997). Therefore, in many cases, grazings (Brown & Iles, 1991). Microcosms consisted of a soil the identity of methylotrophs can be inferred from sample (see below) in a glass serum vial (125 ml) that was pmoA and mxaF sequences retrieved from environ- crimp-sealed with a butyl rubber stopper and incubated in the mental DNA samples. dark at 22–25 mC. Each experiment used two microcosms: one

2332 Stable-isotope probing to identify methylotrophs

"# incubated with C-labelled substrate and the other incubated "$ with C-labelled substrate. The CH OH experiments used "# "$$ CH OH (99 9% pure; BDH) or CH OH (99 4% pure, $ "$ n $ "#n 99 2% C; Isotec) and the CH experiments used CH (98% n "$ % "$ % pure; Linde Gas) or CH% (99% pure, 99% C; Linde Gas). The disappearance of CH$OH and CH% was monitored at 2–7 day intervals using a GC apparatus (Pye Unicam GCD 3500) equipped with a Poropak Q column (1 mi4 mm i.d.) and a flame-ionization detector. The first SIP experiment, which has been described in a preliminary report (Radajewski et al., 2000), used soil collected in November 1998 to identify the active CH$OH- assimilating methylotrophs. In brief, the microcosms contained 10 g of soil and were incubated with a starting CH$OH concentration of 0n5% (v\w). Twice during each incubation the microcosms were flushed with air (250 ml), using a 50 ml syringe to ensure that they remained aerobic. After two "$ additions of CH OH [day 1 and day 18 ( CH OH) or day 24 "# $ $ ( CH$OH)] a total of 2n4 mmol CH$OH had been completely oxidized (44 days). The active CH%-assimilating methylotrophs were charac- terized in three separate experiments using soil collected in November 1998 or January 1999. All soils were dried to a water content that was optimal for CH% oxidation (40–50% water-holding capacity) using a pressure-plate apparatus (Reay et al., 2001b). The dried soils were stored at 4 mC prior to use. In the first two CH% SIP experiments, soil samples (10 g) were added to the serum vials and CH% [0n2 mmol (5 ml), ...... November 1998 soil; 0n4 mmol (10 ml), January 1999 soil] was Fig. 1. Equilibrium centrifugation of DNA extracted from soil that oxidized 12C- or 13C-labelled (a) CH OH or (b) CH . injected into the headspace of the microcosms. After "95% 3 4 CsCl/ethidium bromide density gradients were centrifuged at of the initial CH% concentration had been oxidized the vials 265000 g for 12–16 h at 20 mC and visualized with UV light. The were opened, flushed with air (500 ml) to remove any position of the 13C-DNA fraction collected for analysis is accumulated CO#, resealed and the initial CH% concentration indicated by an arrow. Bar, 1 cm. was re-established. The incubation was continued in the same manner until approximately 2 mmol of CH% had been oxidized [10 0 2 mmol CH (69 days) or 5 0 4 mmol CH (54 days)]. i n % "$ i n % Despite CH% oxidation, no C-labelled fraction was observed redissolved in ultrahigh-purity water (50 µl), transferred to a following ultracentrifugation of DNA extracted from 10 g of 6i4 mm tin capsule containing 5 mg of a nitrogen-free "$ electrofocussing agent (Ultrodex, Pharmacia Biotech) and soil from either of the CH% microcosms. It was hypothesized "% "& that nutrient limitation (possibly N or P) may have prevented dried at 75 mC. The N\ N ratio was determined with an "$ C assimilation into DNA, possibly via the uncoupling of elemental analyser-isotope ratio mass spectrometer (Barrie & CH oxidation from cell growth and\or DNA replication Prosser, 1996). % "$ (Amaral & Knowles, 1995) or due to the assimilation of C DNA extraction and ultracentrifugation. DNA was extracted into a storage product (e.g. poly-3-hydroxybutyrate; Wend- from soil by bead beating. Soil from the CH OH microcosms "$ "# $ landt et al., 2001). Therefore, a third CH% SIP experiment was (1 g from C; 0n3 g from C) was processed in 0n3 g samples established with slurry microcosms that consisted of soil (4 g, using a DNA extraction kit (FastPrep; Bio 101) and a Ribolyser January 1999), dilute ammonia\nitrate mineral salts (ANMS) cell disrupter (Hybaid), as described previously (Yeates & medium (10 ml, 0n1iANMS; Whittenbury et al., 1970) Gillings, 1998). For the CH% microcosms (10 g soil or 10 ml + − "$ "# containing NH% (0n9 mM), NO$ (0n5 mM), phosphate (0n4 slurry from C; 3 g soil or 3 ml slurry from C), this method mM) (pH 5n5) and CH% (0n4 mmol). After repeated CH% was adapted to process larger samples (3 g soil or 3 ml slurry) additions (5i0n4 mmol, as described earlier) and incubation with a CO#-cooled bead beater (Braun), as described by with shaking at 80 r.p.m. for 76 days, approximately 2 mmol Morris et al. (2002). CsCl was added at 1 g (ml DNA −" −" CH% had been oxidized. solution) . Ethidium bromide (100 µl, 10 mg ml ) was added CH% oxidation potentials were determined for each soil as to the DNA solution, which was then transferred to a described previously (Reay et al., 2001a). Ammonium oxi- polyallomer ultracentrifuge tube (13i51 mm; Beckman). "% "& dation potentials were determined from the ratio of N\ N– DNA fractions were resolved following centrifugation at − NO$ present in the soils using the same incubation conditions 265000 g (55000 r.p.m. using a Beckman VTi 65 rotor) for as described previously (Reay et al., 2001a), except that 5 mM 12–16 h at 20 mC in a CsCl density gradient (Sambrook et al., "& NH Cl (Sigma) was supplied as a 10% solution. Nitrate was 1989) and visualized with UV light (365 nm). % "# "# extracted from the soil samples by shaking them with 500 ml DNA from each C-microcosm ( C-DNA) and the most "$ ultrahigh-purity water. The sample was then bound to an dense fraction (approx. 0 5 ml) from each C-microcosm "$ n anion exchange resin (Amberlite IRA–400, 16–50 mesh; ( C-DNA) was withdrawn gently from the primary gradients Sigma) and stored at 4 C until analysed. Nitrate was eluted (Fig. 1) using a 1 ml syringe and hypodermic needle (19 gauge). m "$ from the resin by an overnight incubation with KCl (0 80 M, Care was taken during collection of the C-DNA fraction to n "# "$ 25 ml). It was then concentrated by steam distillation (Bremner avoid co-extraction of the C-DNA band. The C-DNA & Keeney, 1966) and evaporated to dryness. The residue was fraction was added to a new CsCl\ethidium bromide gradient

2333 S. Radajewski and others and centrifuged as described for the primary gradients. No RFLP analyses, and patterns similar to those of the 16S rDNA "# "$ C-DNA fractions were observed in either of the secondary and mxaF clones from the CH OH SIP experiment were "$ $ purification gradients (data not shown). The C-DNA frac- identified. A library of 15 clones was constructed for each PCR tion was re-collected in the minimum volume possible, which product from one tertiary enrichment (M1-N; pH 3 5) and "# n further removed any small amounts of C-DNA that may analysed as described for the CH$OH SIP experiment. have been co-extracted during the primary collection, thereby Complete mxaF and partial 16S rDNA sequence data (posi- increasing the isotopic purity of the DNA from the active tions 375–1060; E. coli numbering) were obtained for one methylotrophs. Ethidium bromide was extracted from the clone from each OTU. DNA with an equal volume of 1-butanol saturated with TE Neither B. indica nor Acidobacterium capsulatum has been (10 mM Tris, 1 mM EDTA; pH 8n0) buffer. Following three reported to grow on CH$OH (Becking, 1984; Kishimoto et al., extractions, the DNA was dialysed against a large volume of 1991). However, as these bacteria were closely related to "$ TE buffer, precipitated with ethanol overnight at k20 mC organisms detected in the CH$OH microcosm, B. indica (Sambrook et al., 1989) and dissolved in 100 µl TE buffer. subsp. indica (NCIMB 8712) and A. capsulatum (DSM 11244) "# Characterization of the 12C- and 13C-DNA fractions. The C- were incubated at 30 C using the respective basal medium "$ m and C-DNA fractions were used as templates for PCR. recommended by the DSM (glucose omitted) with CH$OH Primers specific for bacterial (Eubac27F\1492R), archaeal (12n5 mM) as the sole carbon source. DSM, Deutsche Samm- (Arch21F\1492R) and eukaryotic (EukF\EukR) small-subunit lung von Mikroorganismen; NCIMB, National Collection of rRNA genes (1450–1500 bp product) were used to characterize Industrial, Food and Marine Bacteria. the active population at the domain level (DeLong, 1992). The AOB-specific analysis of DNA fractions. As amoA sequences "$ "$ functional genes mxaF and pmoA were also targeted using were detected in C-DNA fractions from both the CH OH "$ $ primers mxa f1003\mxa r1561 (approx. 555 bp product; and CH% microcosms, a more detailed analysis of these DNA McDonald & Murrell, 1997) and A189\A682 (approx. 525 bp fractions was undertaken using primers selective for AOB product; Holmes et al., 1995), respectively. Reactions (50 µl) belonging to the β-subclass of the Proteobacteria. PCR primers contained 1 µl of template (approx. 5–50 ng) and were selective for 16S rDNA, βamof\βamor (McCaig et al., 1994) performed as described previously for each primer set. and CTO189fGC\CTO654r (Kowalchuk et al., 1997), were Amplification products of the correct size were cloned using used in a nested PCR approach that yielded a 465 bp product the TOPO TA cloning kit (Invitrogen) and libraries of 100 (Webster et al., 2002). Primers specific for the functional clones (bacterial 16S rDNA) or 50 clones (mxaF and pmoA) amoA gene, amoA-1F\amoA-2R (approx. 490 bp product), were constructed. Plasmid inserts were screened by digestion were also used (Rotthauwe et al., 1997). Each microcosm with restriction endonucleases: EcoRI\RsaI for 16S rDNA DNA sample was amplified in duplicate. Ammonia oxidizer "$ from the CH OH microcosm; EcoRI\RsaI and EcoRI\ 16S rDNA PCR products were analysed by denaturing $ "$ Sau3AI for 16S rDNA from the CH% microcosm; EcoRI\ gradient gel electrophoresis (DGGE) as described previously HincII for mxaF; and EcoRI\RsaI and EcoRI\HincII\PvuII (Webster et al., 2002) and the major bands were sequenced to for pmoA. DNA fragments were resolved by electrophoresis confirm identity. Representative clones or pure cultures for through a 2% agarose gel and each clone was assigned to an each 16S rDNA cluster of β-subgroup AOB, as designated by operational taxonomic unit (OTU) that represented a unique Stephen et al. (1996), were included as cluster controls for the RFLP. Complete sequence data (between the amplification 16S rRNA gene DGGE analysis. Bands were excised, resus- primers) were obtained for one clone from each OTU by pended in sterile distilled water (20 µl) and re-amplified using automated sequencing (model ABI 377, PE Biosystems). At 1 µl as template. Re-amplified PCR products were purified least 10% of clones within each OTU were partially sequenced with a centrifugal filter (Microcon YM–100; Millipore) and (" 500 bp), to verify that the restriction pattern represented a resuspended in sterile distilled water (20 µl) at a final DNA −" single sequence type. concentration of 15–20 ng µl . Sequencing reactions were Targeted enrichment of methylotrophs. In an attempt to performed with primer CTO189f. isolate the active proteobacterial methylotrophs from the Phylogenetic analysis. The  program (http:\\www. CH$OH experiment (represented by OTUs UP1–UP3), an mikro.biologie.tu-muenchen.de) was used for sequence align- enrichment strategy was formulated from the physiology of ments and phylogenetic analyses. All 16S rDNA sequences closely related strains [Methylocella palustris (Dedysh et al., were aligned using the automatic alignment tool of  2000) and Beijerinckia indica (Becking, 1984)] and the in- (, version 2.0), and the alignments were corrected cubation conditions used for the microcosm. Basal medium manually according to the constraints imposed by the sec- M1 (Dedysh et al., 1998) was used with CH$OH (0n2%, v\v) ondary structure of 16S rRNA. To identify close relatives of as the sole carbon source. KNO$ was omitted from some the clone sequences in the  database, all aligned sequences media (M1-N), with N# as the only nitrogen source. The initial were added to a tree of sequences ("1400 bp) by using a pH of the medium was adjusted to 5n0 with phosphoric acid. maximum-parsimony tool within . Analyses used the Soil (5 g, November 1998) was added to 0n5 mM potassium sequences of several relatives identified in  and  phosphate buffer (5 ml, pH 6n8) and the sample was gently (Altschul et al., 1990) searches of GenBank, as well as near agitated on a wrist-action shaker (Stuart Scientific) for 2 h at full-length sequences ("1350 bp) of extant methylotrophs, 4 mC. The slurry (5 µl) was added to 10 ml of medium in a methanotrophs and AOB in GenBank. A filter was generated glass serum vial (125 ml), which was crimp-sealed with a butyl that omitted alignment positions of sequence ambiguity (N) rubber stopper and incubated (20–25 mC) with shaking (80 and where sequence data were not available for all near full- r.p.m.). Enrichments were subcultured twice at 4–10 week length sequences. intervals into the same medium, which was adjusted to pH 5n0 To evaluate the tree topology, phylogenies were reconstructed or pH 3n5. with various data subsets using evolutionary distance (Jukes DNA was extracted from 1n5 ml of each enrichment (Marmur, and Cantor model), maximum-parsimony ( and - 1961) and used as a template for PCR with the bacterial 16S ) and maximum-likelihood (default parameters for  rDNA, mxaF and pmoA primers. PCR products were digested and ) analyses of the aligned near full-length with the same restriction endonucleases as used previously for ("1350 bp) sequences (Ludwig et al., 1998). Aligned partial

2334 Stable-isotope probing to identify methylotrophs

"$ "$ "$ 16S rDNA sequences from the CH$OH enrichments and the However, when C-DNA from the CH$OH or CH% AOB-specific analyses, and selected sequences from GenBank microcosm was used as a template, an amplification (Z73369, Z88588, AF145871 and AF358018) were inserted product was only obtained with the primer set that was "$ into each tree using a parsimony tool available within  and specific for Bacteria, indicating that the C-DNA the filter generated with complete sequence data. As topologies fractions contained a restricted microbial community. obtained from the different analyses were generally similar, "$ RFLP analysis of 100 16S rDNA clones from each C- the trees presented are based on the maximum-likelihood "$ analysis, with the confidence of branch points determined DNA fraction identified six OTUs in the CH OH "$ $ using a strict consensus rule applied to the results of the three library and eight OTUs in the CH% library. Phylo- analysis methods. Multifurcations indicate points at which the genetic analysis of the 16S rDNA sequences from the "$ branching order was collapsed until it was supported by all CH$OH library identified three OTUs (UP1–UP3) that three methods of analysis. clustered with members of the α-subclass of the Proteo- Functional-gene sequences (pmoA and mxaF) were aligned bacteria (Fig. 2). A further three OTUs (UA1–UA3; each manually to related sequences extracted from GenBank. A consisting of one clone) were most closely related filter was generated from the derived amino-acid sequences that excluded alignment positions containing amino-acid (95n6–98n0% identity) to 16S rDNA sequences of mem- ambiguity (X) and missing sequence data. Phylogenies were bers of the Acidobacterium division. The 16S rDNA reconstructed with evolutionary distance (Dayhoff PAM sequence of UP1 (49 clones) was most closely related to model), maximum-parsimony ( and ) and maxi- that of UP8 (99n5% identity) from the CH% library. The mum-likelihood (default parameters for  and ProteinIML) sequence of UP2 (47 clones) was 97n4% similar to the analyses, and tree topologies were evaluated as described for 16S rDNA sequence of Rhodoblastus acidophilus (for- the 16S rDNA analysis. merly Rhodopseudomonas acidophila; Imhoff, 2001). Environmental 16S rDNA sequences were inspected for The sequence of UP3 (one clone) was most similar to the potential chimeras using the program   (version 16S rDNA sequences of B. indica, Methylocella palu- 2.7; http:\\rdp.cme.msu.edu). Functional-gene DNA se- stris, Methylocapsa acidiphila and clone UP1 (95n6– quences were inspected for chimeras by searching for large 97 0% identity). Phylogenetic analysis of the 16S rDNA n "$ regions of unexpected nucleotide changes when compared sequences from the CH library identified seven OTUs with sequences from reference organisms. One potential 16S % (UP4–UP9) that clustered in the α- and β-subclasses of rDNA chimera was identified from the CH$OH enrichment and it was not included in the analyses. It was also necessary the Proteobacteria and one OTU (UC1; one clone) that to add a single nucleotide ambiguity (N) to the DNA sequence clustered with the order Cytophagales (Fig. 2). The of the functional-gene clones P12.7, P12.9, P13.7 and M12.1 to sequences of UP4 (92 clones), UP5 (two clones), UP6 and maintain the correct reading frame for each derived amino- UP7 (each one clone) had 99n4–99n7 % identity with each acid sequence. Identity between sequences was determined other and 96n1–97n4% identity with 16S rDNA se- using the similarity matrix option available within . quences from the methanotrophs Methylosinus trichosporium and Methylocystis parvus. The sequence of RESULTS UP8 (two clones) was most closely related to those of Soil microcosms and CsCl density gradients UP1 and UP3 and to the 16S rDNA sequences of Methylocella palustris and Methylocapsa acidiphila. Low-affinity (5%, v\v) CH% oxidation potentials of 0n10 −" −" The sequence of UP9 (one clone) was most closely and 0n44 µmol (g.d.w. soil) h were observed for the related to that of clone LO13.5 (96n8% identity) and to soils collected in November 1998 and January 1999, the 16S rDNA sequences of Herbaspirillum seropedicae, respectively. Ammonium oxidation potentials for the −" −" Janthinobacterium lividum and Paucimonas lemoignei soil samples were 3n88 and 4n15 nmol (g.d.w. soil) h , (95n7–96n0% identity). The sequence of UC1 exhibited respectively. most similarity to that of clone LO13.4 (99n6% identity). "$ As reported previously (Radajewski et al., 2000), a C- DNA fraction was extracted from 1 g of soil from the Functional-gene libraries microcosm (10 g soil) that had oxidized 2 4 mmol of "$ "$ n The mxaF and pmoA primer sets amplified specific CH OH (Fig. 1a). However, no C-DNA band could "# "$ $ products of the correct size from both the C- and C- be observed from 10 g of soil from two independent DNA fractions extracted from the CH OH and CH microcosms that had oxidized approximately 2 0 mmol $ % "$ n microcosms. RFLP analysis of mxaF clones from the of CH . Consequently, the CH slurry experiment was "$ % % "$ C-DNA fractions identified three OTUs (M13.1, 48 established, in which a distinct and intense C-DNA clones; M13.2 and M13.3, each one clone) in the fraction was observed (Fig. 1b). The DNA fractions "$ CH OH library and four OTUs (M13.4, 41 analysed were collected from the soil microcosms $ clones; M13.5 and M13.6, each four clones; M13.7, one exposed to CH OH and the soil slurry microcosms "$ $ clone) in the CH library. Similar mxaF sequences and exposed to CH . % % proportions of clones within each OTU were identified "# "# in analyses of C-DNA from the CH OH (M12.1, 48 Identity of methylotrophs in the 13C-DNA $ "# clones; M12.2 and M12.3, each one clone) and CH "# "# "# % Using C-DNA from the CH$OH or CH% microcosm (M12.4, 42 clones; M12.5, five clones; M12.6, three as a template, specific amplification products were clones) microcosms. Phylogenetic analysis of derived obtained with PCR primers that targeted the small- MxaF sequences from the CH$OH libraries (M13.1– subunit rRNA genes of Bacteria, Archaea and Eukarya. M13.3 and M12.1–M12.3), using the XoxF-like meth-

2335 S. Radajewski and others

Nitrosospira briensis (L35505) Nitrosospira multiformis (L35509) Nitrosospira sp. AHB1 (X90820) UBP12OH1 UBP12CH1 UBP12OH2 UBP13CH2 Uncultured Nitrosomonas sp. (AJ224410) UBP13CH1 Nitrosomonas europaea (AF037106) β-Proteobacteria UBP13OH1 Nitrosococcus mobilis (AF037105) Burkholderia graminis (U96941) EOH2 (5) Burkholderia phenazinium (U96936) UP9 (1) Janthinobacterium lividum (Y08846) Herbaspirillum seropedicae (Y10146) Uncultured bacterium LO13.5 (AF358019) Paucimonas lemoignei (X92554) Methylophilus methylotrophus (L15475) EOH3 (1) Uncultured bacterium (AJ318146) Frateuria sp. NO-16 (AF376025) Rhodanobacter lindanoclasticus (AF039167) Methylobacter psychrophilus (AF152597) γ-Proteobacteria Methylosarcina fibrata (AF177296) Methylomonas methanica (AF150806) Methylococcus capsulatus (X72771) Methylocaldum szegediense (U89300)

EOH1 (9) Methylocella palustris (Y17144) Methylocapsa acidiphila B2 (AJ278726) Uncultured bacterium LO13.10 (AF358002) UP3 (1) (AF200697) Afipia genospecies 6 (U87772) UP2 (47) (AF200695) Rhodoblastus acidophilus (M34128) Beijerinckia indica (M59060) UP8 (2) UP1 (49) (AF200694) Methylosinus sp. LW2 (AF150786) α-Proteobacteria Methylosinus trichosporium (AF150804) Methylosinus sp. PW1 (AF150802) Methylocystis sp. M (U81595) Methylocystis sp. LW5 (AF150790) Methylocystis parvus (AF150805) UP5 (2) UP4 (92) UP6 (1) UP7 (1) Methanotroph AML–A3 (AF177298) Methanotroph AML–A6 (AF177299) Methylobacterium extorquens (AF267912) Hyphomicrobium methylovorum (Y14307) Acidomonas methanolica (X77468) UC1 (1) Uncultured bacterium LO13.4 (AF358018) Uncultured bacterium (Z88588) Uncultured bacterium ZD0403 (AJ400347) Cytophagales Flavobacterium columnare (AB015481) Capnocytophaga ochracea (X67610) UA2 (1) (AF200698) UA1 (1) (AF200696) Acidobacterium capsulatum (D26171) Uncultured bacterium WD247 (AJ292581) Acidobacterium division Uncultured bacterium (Z73369) UA3 (1) (AF200699) Uncultured bacterium K20–83 (AF145871) Holophaga foetida (X77215) Thermotoga maritima (M21774)

...... Fig. 2. For legend see facing page.

2336 Stable-isotope probing to identify methylotrophs

M12.1 (48) M13.1 (48) (AF200700) As many methanotrophs can also use CH$OH as a M12.3 (1) M12.2 (1) carbon source, the pmoA primer set A189\A682 was M13.2 (1) (AF200701) MEOH2 (1) used with template DNA from both the CH% and M13.3 (1) (AF200702) MEOH1 (13) CH$OH microcosms. RFLP analysis of pmoA clones MEOH3 (1) "$ Methylobacterium extorquens (M31108) from the C-DNA fractions identified five OTUs in Methylocapsa acidiphila B2 (AJ278730) "$ both the CH OH (P13.1, 46 clones; P13.2–P13.5, each Methylocella palustris (AJ278731) $ "$ Methylococcus capsulatus (U70511) one clone) and CH microcosm (P13.6, 30 Methylomonas methanica (U70512) % M13.4 (41) clones; P13.7, 13 clones; P13.8, four clones; P13.9, two M12.6 (3) M12.4 (42) clones; P13.10, one clone). In contrast to the mxaF Unidentified peat clone 1 (U70518) M13.7 (1) libraries, several different sequences and OTUs (Fig. 4) M12.5 (5) M13.5 (4) were identified from the corresponding microcosms Unidentified peat clone 5 (U70521) "# M13.6 (4) exposed to CH$OH (P12.1, 21 clones; P12.2, 13 clones; Unidentified peat clone 3 (U70519) Methylosinus trichosporium (U70516) P12.3, seven clones; P12.4, five clones; P12.5, two Methylocystis parvus (U70515) "# Methylocystis sp. M (U70517) clones; P12.6 and P12.7, one clone each) or CH% Hyphomicrobium methylovorum (AB004097) Methylophilus methylotrophus (U41040) (P12.8, 35 clones; P12.9, eight clones; P12.10, four Methylovorus sp. SS1 (AF184915) clones; P12.11, two clones; P12.12, one clone). Phylo- Methylobacterium extorquens (XoxF) (U72662) genetic analysis of derived amino-acid sequences from "$ ...... the CH$OH library revealed that four OTUs (49 Fig. 3. Maximum-likelihood tree of derived MxaF sequences clones) were closely related to the AmoA of Nitroso- detected in the 13C-DNA fraction of microcosms exposed to 13 13 monas europaea (Fig. 4) and only one OTU (one clone) CH3OH (#)or CH4 ($) and from CH3OH enrichments (*). The tree was constructed with the MxaF homologue (XoxF) of was similar to the PmoA of Methylosinus trichosporium. Phylogenetic analysis of the derived PmoA\ Methylobacterium extorquens (U72662) as the outgroup and a "$ filter (170 aa) that excluded alignment positions with sequence AmoA sequences from the CH% library identified four ambiguity or missing data. The confidence of branch points was groups of sequences. The derived PmoA sequences of determined by three separate analyses (maximum-likelihood, P13.6 and P13.10 were most closely related to that of evolutionary distance and maximum-parsimony), with multi- "$ furcations indicating branch points that were collapsed until clone LOPA13.1, from a CH% SIP experiment with supported in all three analyses using a strict consensus rule. peat soil (Morris et al., 2002), and were distinct from the Prefixes indicate sequences from the 13C-DNA (M13) and PmoA of type II methanotrophs and the environmental 12C-DNA (M12) fractions or from the CH OH enrichment 3 clone RA14 (Holmes et al., 1999). The derived amino- (MEOH). The number of clones assigned to each OTU by RFLP analysis is shown in parentheses. Accession numbers for the acid sequence of P13.7 clustered with the PmoA of sequences are also shown (numbers prefixed by one or two Methylocystis spp., whereas that of P13.8 grouped with letters in parentheses). Known methylotrophs and clones the AmoA of Nitrosomonas spp. OTU P13.9 had a 13 retrieved from the C-DNA fractions of CH3OH or CH4 SIP derived amino-acid sequence that was weakly related experiments are shown in bold text. (58 8–61 3% aa identity) to the AmoA of β-proteobac- n n "# terial AOB. Clones from the C-DNA libraries con- anol dehydrogenase large-subunit homologue of Meth- tained derived amino-acid sequences that grouped with the PmoA- and AmoA-containing clones from the ylobacterium extorquens as the outgroup (Chistoserdova "$ & Lidstrom, 1997), placed these sequences in a cluster corresponding C-DNA libraries, as well as with the that is distinct from all previously described methylo- PmoA of type I methanotrophs (P12.6) and distant homologues of PmoA\AmoA (P12.3–P12.5 and P12.10). troph MxaF sequences. The most closely related MxaF "# "$ sequences (Fig. 3) were of Methylobacterium extorquens The differences between the C- and C-DNA pmoA libraries further demonstrated the restricted nature of (72n8–80n1% aa identity) and the moderately acidophilic "$ methanotrophs Methylocella palustris and Methylo- the population within the C-DNA fractions. capsa acidiphila (73n8–81n0% aa identity). Phylogenetic analysis of the derived MxaF sequences from the CH% AOB-specific analysis of DNA fractions libraries (M13.4–M13.7 and M12.4–M12.6) identified a group of sequences most similar to the MxaF sequences To investigate the unexpected observation of amoA "$ of the methanotrophic genera Methylosinus and Methy- sequences in the C-DNA fractions, DNA from both locystis (Fig. 3). microcosms was used as a template for PCR amplifica-

Fig. 2. Maximum-likelihood tree of 16S rDNA sequences amplified from the 13C-DNA fraction of microcosms exposed to 13 13 CH3OH (#)or CH4 ($) and from CH3OH enrichments (*). Partial sequences (!1350 bp; dashed line) were added to the tree using a maximum-parsimony option within . The tree was constructed with the 16S rDNA sequence of Thermotoga maritima as the outgroup and a filter (1160 nt) that excluded alignment positions with sequence ambiguity or missing data. The confidence of branch points was determined by three separate analyses (maximum-likelihood, evolutionary distance and maximum-parsimony), with multifurcations indicating branch points that were collapsed until supported in all three analyses using a strict consensus rule. Prefixes indicate sequences of uncultivated organisms of the Acidobacterium division (UA), the Proteobacteria (UP), the Cytophagales (UC) or the β-Proteobacteria (UBP), or the sequences of clones from the CH3OH enrichment (EOH). The major phylogenetic groups are indicated at the right of the figure. The number of clones assigned to each OTU by RFLP analysis is shown in parentheses. Accession numbers for the sequences are also shown (numbers prefixed by one or two letters in parentheses). Known methylotrophs and clones 13 retrieved from the C-DNA fractions of CH3OH or CH4 SIP experiments are shown in bold text.

2337 S. Radajewski and others

P12.7 (1) ...... Methanotroph AML–A2 (AF177327) Methylocystis parvus (U31657) Fig. 4. Maximum-likelihood tree of derived P13.7 (13) PmoA sequences detected in the 13C-DNA P12.9 (8) 13 P13.5 (1) fraction of microcosms exposed to CH3OH Methylosinus trichosporium (AF186586) (#)or13CH ($). The tree was constructed P13.6 (30) 4 P12.8 (35) with the derived amino-acid sequence of P12.12 (1) clone RA21 as the outgroup and a filter P13.10 (1) LOPA13.1 (AF358046) (161 aa) that excluded alignment positions Uncultured bacterium Bakchar31 (AJ278729) with sequence ambiguity or missing data. Methylocapsa acidiphila B2 (AJ278727) Uncultured bacterium RA14 (AF148521) The confidence of branch points was Methylococcus capsulatus (L40804) determined by three separate analyses Methylocaldum szegediense (U89303) P12.6 (1) (maximum-likelihood, evolutionary distance Methylomonas methanica (U31653) and maximum-parsimony), with multifurca- Nitrosococcus oceani (AF047705) tions indicating branch points that were P12.1 (21) P13.1 (46) collapsed until supported in all three P13.2 (1) analyses using a strict consensus rule. Prefixes Nitrosomonas europaea (L08050) 13 P13.8 (4) indicate sequences from the C-DNA (P13) Nitrosomonas eutropha (U51630) or 12C-DNA (P12) fractions. The number of P12.2 (13) P13.3 (1) clones assigned to each OTU by RFLP analysis P13.4 (1) is shown in parentheses. Accession numbers P12.11 (2) Nitrosospira multiformis (U15733) for the sequences are also shown (numbers Nitrosospira tenuis (U76552) prefixed by one or two letters in paren- P12.4 (5) P13.9 (2) theses). Known methylotrophs and clones P12.5 (2) retrieved from the 13C-DNA fractions of P12.3 (7) CH OH or CH SIP experiments are shown in P12.10 (4) 3 4 Uncultured bacterium RA21 (AF148522) bold text.

...... Fig. 5. DGGE of ammonia-oxidizer-selective 16S rDNA PCR products. AOB cluster controls correspond to the AOB cluster designation of Stephen et al. (1996). Lanes (left panel): 1–4, Nitrosospira clusters 1–4; 5–7, Nitrosomonas clusters 5–7. Duplicate amplifications (right panel; A and B) were performed with the microcosm DNA samples. Lanes: 1, 13C-DNA 12 13 12 (CH3OH); 2, C-DNA (CH3OH); 3, C-DNA (CH4); 4, C-DNA (CH4). DGGE bands that were sequenced are marked with an 13 12 i. The prefixes 13OH or 12OH correspond to the C- or C-DNA fractions, respectively, from the CH3OH microcosm; the 13 12 prefixes 13CH or 12CH correspond to the C- or C-DNA fractions, respectively, from the CH4 slurry microcosm. tion with AOB-specific primers. Nested PCR of the 16S belonging to Nitrosospira clusters 1 (UBP12OH1) and 3 "# rDNA of AOB within the β-subclass Proteobacteria (UBP12OH2) in the C-DNA. In the CH microcosms, "# "$ % using C- and C-DNA fractions from both the the dominant DGGE bands were identified as sequences CH OH and CH SIP experiments generated specific belonging to Nitrosomonas cluster 7 (UBP13CH1) and $ % "$ products that were resolved by DGGE (Fig. 5). In the Nitrosospira cluster 3 (UBP13CH2) in the C-DNA, CH OH microcosms, analysis of the dominant DGGE and as sequences belonging to Nitrosospira cluster 2 $ "# bands identified AOB sequences belonging to Nitroso- (UBP12CH1) in the C-DNA. No PCR amplification monas cluster 5 (UBP13OH1; cluster designation of product was obtained with the amoA-specific primer "$ Stephen et al., 1996) in the C-DNA and AOB sequences set.

2338 Stable-isotope probing to identify methylotrophs

Growth of enrichments and pure cultures on CH OH sociation with methylotrophs, would be sufficiently 3 "$ labelled with C to have been collected in the discrete Pure cultures of B. indica subsp. indica and A. capsula- "$ C-DNA fractions. tum grew on the recommended heterotrophic medium, but no growth was observed after 14 days incubation Characterization of the microbial community within the "$ with CH$OH (12n5 mM) as the sole carbon source. No C-DNA fractions by analysis of 16S rDNA sequences amplification product was obtained from either strain identified distinct bacterial populations involved in the with the PCR primers that target mxaF. assimilation of CH$OH and CH% in the soil microcosms (Fig. 2). However, the abundance of clones in the DNA from all of the CH OH enrichments generated an $ libraries must be interpreted with caution, as limitations amplification product with the 16S rDNA and mxaF inherent in SIP may be compounded by biases introduced PCR primers, but not with the pmoA primers (A189\ by PCR amplification (Wintzingerode et al., 1997). In A682). RFLPs similar to those obtained from the "$ particular, clones occurring at low frequency may CH OH microcosm were identified in one tertiary $ represent bacteria that grow relatively slowly on the enrichment (M1-N; pH 3 5). Both the 16S rDNA library "$ n C-labelled substrate, bacteria which are cross-feeding and the mxaF library (15 clones each) contained three "$ on non-primary C-labelled compounds or sequences OTUs. The sequence of EOH1 (nine clones) was similar "# amplified from traces of C-DNA that may have been to the 16S rDNA sequence of Methylocella palustris and "$ co-extracted with the C-DNA fraction. We noted the sequences of clones UP1–UP3 (Fig. 2). The sequences previously that the 16S rDNA sequences recovered from of EOH2 (five clones) and EOH3 (one clone) clustered "$ the CH OH microcosm were, surprisingly, confined to with the 16S rDNA sequences of members of the genera $ only two phylogenetic groups and most closely related Burkholderia and Frateuria, respectively (Fig. 2). De- to genera that are rarely reported as being involved in rived MxaF sequences (MEOH1, 13 clones; MEOH2 CH OH utilization (Radajewski et al., 2000). The and MEOH3, each one clone) were most similar to those $ closely related strains Beijerinckia indica and A. capsu- of M13.1–M13.3 (Fig. 3). latum did not grow on CH$OH; however, both of these strains, as well as the methanotrophs Methylocella DISCUSSION palustris and Methylocapsa acidiphila and the photo- autotroph R. acidophilus, grow at acidic pH in culture. All of the SIP experiments involved long periods (40 "$ "$ Although strains closely related by 16S rRNA phylogeny days) of continuous exposure to CH OH or CH , "$ $ % do not necessarily share physiological similarities, the which enabled the incorporation of C into the DNA of acidity of the forest soil in conjunction with these active methylotrophs. Non-methylotrophic micro-or- "$ phylogenetic relationships suggests that UP1–UP3 and ganisms could have also assimilated the C-labelled "$ UA1–UA3 may represent moderately acidophilic bac- intermediates and by-products (e.g. CO ) of methylo- # teria. Balancing the caveats associated with clones troph metabolism during the relatively long incubation occurring at low frequency with the fact that the library period. However, to observe a ‘heavy’ DNA fraction the "$ contained three distinct, but related, OTUs in each major C isotope within DNA must be the C that was phylogenetic group (i.e. UP1–UP3 and UA1–UA3) also added as the labelled substrate. Thus, the turnover of "$ suggests that these bacteria may have grown directly on C due to cross-feeding can be assessed qualitatively "$ "$ CH OH at the relatively high concentration (0 5%, from the presence absence and the position of a C- $ n \ v w; approximately 300 mM at 40% water-holding DNA fraction in a CsCl gradient. For example, the \ "$ capacity) used in the microcosm. absence of any visible C-DNA fraction from two of the "$ three CH microcosms, despite all three oxidizing Most of the 16S rDNA, pmoA and mxaF clones retrieved % "$ "$ approximately 2 mmol of CH , suggests that the non- from the acidic (pH 3 4) CH slurry microcosm were "$ % n % primary C-labelled compounds were diluted substan- closely related to sequences of extant methanotrophs "# tially by the production of C-labelled substrates (e.g. belonging to the α-Proteobacteria (type II). A recent "# CO ) and\or by other trophic interactions within the study using fluorescent in situ hybridization and combi- "$ # C-microcosms. Similarly, the observation of very nations of highly specific oligonucleotide probes identi- "$ "# "$ discrete C- and C-DNA fractions in the CH OH fied Methylocystis spp. as the major component of the "$ $ soil and CH slurry microcosms rather than a smear of methanotroph population in acidic (pH 4 2) Sphagnum "$ % n C-DNA (e.g. Morris et al., 2002) suggests that most peat samples (Dedysh et al., 2001). Distinct mxaF cross-feeding micro-organisms did not incorporate a sequences (Fig. 3, peat clones) detected in an acidic peat "$ high proportion of C into their DNA. Since the (pH 3 6) have also been proposed to originate from a "$ "# n separation between the C- and C-DNA fractions of a population of acidophilic type II methanotrophs "$ "# pure culture grown on C- or C-labelled substrates (McDonald & Murrell, 1997). Therefore, certain 16S was approximately 1 cm in an equilibrium CsCl density rDNA (UP4–UP7) and derived MxaF (M13.4–M13.7) "$ gradient (Radajewski et al., 2000), it is estimated that the sequences retrieved from the CH slurry microcosm "$ % C-DNA fractions collected from the microcosms support the moderately acidophilic nature of some as yet "$ contained 65–100% C (Fig. 1). Therefore, only the uncultivated Methylocystis-like methanotrophs. "$ "$ DNA of micro-organisms that assimilated Casa Another OTU from the CH% slurry microcosm (UP8) primary carbon source, such as methylotrophs or grouped more closely to Methylocella (Dedysh et al., organisms in a close physical and\or nutritional as- 2000), Methylocapsa (Dedysh et al., 2002) and clones

2339 S. Radajewski and others

"$ retrieved from a CH% SIP experiment with a peat soil Holmes et al., 1999; Reay et al., 2001b). However, it was (LO13.10; Morris et al., 2002) than to other methano- unexpected that the derived amino-acid sequence of "$ trophs within the α-Proteobacteria. However, the se- almost all of the clones from the CH OH microcosm "$ $ quence of UP8 was most similar to the sequence of UP1, (P13.1–P13.4) and several from the CH% slurry micro- indicating that it may represent bacteria that assimilated cosm (P13.8) grouped with the AmoA sequences of CH OH produced by methanotrophs within the soil Nitrosomonas spp. (Fig. 4). DGGE of AOB-specific 16S $ "$ slurry microcosm. The combined results from these SIP rDNA amplification products from the C-DNA tem- "$ "$ experiments with CH$OH and CH% suggest that plates confirmed that sequences similar to Nitrosomonas methylotrophs closely related to Methylocella, Methylo- spp. (UBP13OH1 and UBP13CH1) and Nitrosospira capsa and Methylocystis are active and potentially spp. (UBP13CH2) were present in the same DNA widespread in moderately acidic environments. fractions that contained the amoA sequences (Fig. 2 and The physiology of the organisms represented by UP9 (β- Fig. 4). However, the lack of an amoA-specific PCR Proteobacteria) and UC1 (Cytophagales) is not clear, as product and the variability observed in DGGE for some they did not group near any known methylotrophs. It duplicate amplifications (e.g. Fig. 5, lanes 3A and 3B) suggested that the AOB template DNA was not abun- must also be stressed that these clones occurred at low "$ frequency and, in the absence of isolates or more dant in the C-DNA fractions. Since extant Nitroso- monas spp. are obligately autotrophic AOB and both definitive data, the physiology of the bacteria that are + soil samples had the capacity to oxidize NH% [approx. represented by these clones should be considered as −" −" speculative. It is, however, interesting to note that from 4 nmol (g.d.w. soil) h ], it seems likely that some "$ "$ a library of 100 16S rDNA clones in a CH% SIP AOB had assimilated a significant proportion of CO# experiment, Morris et al. (2002) identified six clones in the microcosms, possibly through a close physical (LO13.5) with 96n8% similarity to UP9, which is higher and\or nutritional association with active methylo- than the next nearest relative currently retrieved in a trophs. Interestingly, Nitrosomonas cluster 5 AOB  search. They also detected several Cytophaga sequences have not been previously reported in soil, nor sequences, including three clones (LO13.4) with 99n6% have representatives been cultivated. Therefore, it is also similarity to UC1. These culture-independent observa- possible that these AOB may have an unusual physiology "$ tions from two distinct CH% SIP experiments raise the when compared with other Nitrosomonas spp. Analysis possibility that bacteria not previously considered to be of larger DNA fragments or the isolation of represen- involved in CH% oxidation may derive a significant tative strains may help to resolve these curious observa- proportion of their carbon from products of methano- tions. troph metabolism, or possibly even from CH% itself. An enrichment strategy to target methylotrophs repre- One significant feature of the SIP approach used here is sented by OTUs UP1–UP3 was only partly successful, as "$ most mxaF and 16S rDNA clones were similar to, but that the C-DNA fractions contained the entire "$ genomes of the active methylotrophs, which enabled a distinct from, those retrieved from the CH$OH micro- parallel analysis of functional-gene sequences within cosm. Sequences detected in these enrichment libraries these populations. As the mxaF and pmoA PCR primers also suggested that acidophilic, N#-fixing, methylo- were designed from sequences of extant Proteobacteria, trophic Burkholderia spp. may exist. Indeed, symbiotic it is likely that most of the functional-gene sequences N#-fixing Burkholderia strains have recently been iso- "$ detected in the C-DNA fractions were present in one lated from the root nodules of tropical legumes (Moulin or more of the Proteobacteria identified from 16S rDNA et al., 2001) and a N#-fixing, CH$OH utilizer with a 16S sequences (UP1–UP9). Indeed, derived MxaF sequences rDNA sequence identical to EOH2 (Burkholderia sp.) "$ in the CH$OH microcosm were most similar to the was isolated from the enrichments and is being charac- MxaF sequences of Methylocella, Methylocapsa and terized further. Although the strain corresponding to "$ Methylobacterium, whereas in the CH% microcosm EOH1 could not be purified, these results demonstrate they were most similar to the MxaF sequences of the potential use of SIP for the enrichment and isolation Methylocystis and Methylosinus. Similarly, the majority of novel micro-organisms with specific metabolic capa- "$ of pmoA clones from the CH% slurry experiment were bilities. similar to the pmoA of extant type II methanotrophs. Despite the remarkable simplicity of the SIP technique, Dunfield et al. (2002) have recently identified certain its suitability for resolving certain structure–function Methylocystis and Methylosinus isolates that contain relationships may be limited (Radajewski et al., 2000). "$ two highly distinct pmoA-like sequences, which suggests In these experiments, it was critical that sufficient C that OTUs P13.6 and P13.10 may represent a second was incorporated into the DNA of the functionally "$ pmoA lineage possessed by some of the Methylocystis- active methylotrophs to permit collection of a C-DNA "$ like methanotrophs (UP4–UP7). Such potential diversity fraction. Factors including C turnover due to predation "$ of pmoA-like sequences within a single methanotroph or substrate transformation, C dilution due to the "# also excludes identification of the bacterium containing assimilation of other carbon substrates (i.e. C-labelled) "$ the sequence P13.9. or C assimilation without DNA replication, might The pmoA primers A189\A682 can amplify pmoA, have influenced the community composition within the "$ "$ amoA and homologous sequences of unknown function C-DNA fractions. In the two CH SIP experiments "$ % from environmental samples (Henckel et al., 2000; where a C-DNA band was not observed, consideration

2340 Stable-isotope probing to identify methylotrophs of such factors provides some insight into the CH%- Chistoserdova, L. & Lidstrom, M. E. (1997). Molecular and oxidizing population in this soil, even though the mutational analysis of a DNA region separating two methylo- functional component of the community was not identi- trophy gene clusters in Methylobacterium extorquens AM1. "$ fied. To obtain visible C-DNA fractions, it was thus Microbiology 143, 1729–1736. necessary to use incubation conditions that enriched for Dedysh, S. N., Panikov, N. S. & Tiedje, J. M. (1998). Acidophilic methylotrophs in a physically, chemically and bio- methanotrophic communities from Sphagnum peat bogs. Appl logicallycomplexenvironment,whichonlypartlyreflect- Environ Microbiol 64, 922–929. ed that occurring in situ. Future improvements, such as Dedysh, S. N., Liesack, W., Khmelenina, V. N., Suzina, N. E., the extraction of biomarkers that do not rely upon Trotsenko, Y. A., Semrau, J. D., Bares, A. M., Panikov, N. S. & replication of the chromosome for labelling (e.g. rRNA; Tiedje, J. M. (2000). Methylocella palustris gen. nov., sp. nov., a "$ Gray & Head, 2001), the cloning of large C-DNA new methane-oxidizing acidophilic bacterium from peat bogs, representing a novel subtype of serine-pathway methanotrophs. fragments to link phylogenetic and functional genes, the Int J Syst Evol Microbiol 50, 955–969. incubation of larger sample volumes and the use of time- course experiments, may permit a more sensitive res- Dedysh, S. N., Derakshani, M. & Liesack, W. (2001). Detection and enumeration of methanotrophs in acidic Sphagnum peat by 16S olution of the function and succession of methylotrophs rRNA fluorescence in situ hybridization, including the use of in the environment. Furthermore, as SIP is an effective newly developed oligonucleotide probes for Methylocella palu- tool for the isolation of the genome of micro-organisms stris. Appl Environ Microbiol 67, 4850–4857. with previously unknown metabolic capabilities, it can Dedysh, S. N., Khmelenina, V. N., Suzina, N. E., Trotsenko, Y. A., provide a rational basis for the application of molecular Semrau, J. D., Liesack, W. & Tiedje, J. M. (2002). Methylocapsa biological techniques to investigate these uncultivated acidiphila gen. nov., sp. nov., a novel methane-oxidizing and methylotroph populations in situ. dinitrogen-fixing acidophilic bacterium from Sphagnum bog. Int J Syst Evol Microbiol 52, 251–261. ACKNOWLEDGEMENTS DeLong, E. F. (1992). Archaea in coastal marine environments. Proc Natl Acad Sci U S A 89, 5685–5689. This study was funded by the UK Natural Environment Dunfield, P. F., Liesack, W., Henckel, T., Knowles, R. & Conrad, R. Research Council EDGE thematic programme (Award GST\ (1999). High-affinity methane oxidation by a soil enrichment 02\1864 to J.C.M., P.I. and D.B.N. and GST\02\1837 to culture containing a type II methanotroph. Appl Environ Micro- J.I.P.) with support from the European Union 4th Framework biol 65, 1009–1014. Programme (Grant QLRT-1999-31528 to J.C.M.). We thank Forest Enterprises for access to the study site and Peter Dunfield, P. F., Yimga, M. T., Dedysh, S. N., Berger, U., Liesack, W. Dunfield for sharing findings prior to their publication. & Heyer, J. (2002). Isolation of a Methylocystis strain containing a novel pmoA-like gene. FEMS Microbiol Ecol (in press). Gray, N. D. & Head, I. M. (2001). Linking genetic identity and REFERENCES function in communities of uncultured bacteria. Environ Micro- Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. biol 3, 481–492. (1990). Basic local alignment search tool. J Mol Biol 215, 403–410. Hanson, R. S. & Hanson, T. E. (1996). Methanotrophic bacteria. Amaral, J. A. & Knowles, R. (1995). Growth of methanotrophs in Microbiol Rev 60, 439–471. methane and oxygen counter gradients. FEMS Microbiol Lett Henckel, T., Friedrich, M. & Conrad, R. (1999). Molecular analyses 126, 215–220. of the methane-oxidizing microbial community in rice field soil by Anthony, C. (1982). The Biochemistry of Methylotrophs. London: targeting the genes of the 16S rRNA, particulate methane Academic Press. monooxygenase, and methanol dehydrogenase. Appl Environ Microbiol 65, 1980–1990. Barrie, A. & Prosser, S. J. (1996). Automated analysis of light- $ element stable isotopes by isotope ratio mass spectrometry. In Henckel, T., Jackel, U., Schnell, S. & Conrad, R. (2000). Molecular Mass Spectrometry of Soils, pp. 1–46. Edited by T. W. Boutton & analyses of novel methanotrophic communities in forest soils that S.-i Yamasaki. New York: Marcel Dekker. oxidize atmospheric methane. Appl Environ Microbiol 66, Becking, J.-H. (1984). Genus Beijerinckia Derx 1950. In Bergey’s 1801–1808. Manual of Systematic Bacteriology, vol. 1, pp. 311–321. Edited by Holmes, A. J., Costello, A., Lidstrom, M. E. & Murrell, J. C. (1995). N. R. Krieg. Baltimore: Williams & Wilkins. Evidence that particulate methane monooxygenase and ammonia ! Bedard, C. & Knowles, R. (1989). Physiology, biochemistry, and monooxygenase may be evolutionarily related. FEMS Microbiol + Lett 132, 203–308. specific inhibitors of CH%,NH%, and CO oxidation by methanotrophs and nitrifiers. Microbiol Rev 53, 68–84. Holmes, A. J., Roslev, P., McDonald, I. R., Iversen, N., Henriksen, Boschker, H. T. S., Nold, S. C., Wellsbury, P., Bos, D., de Graaf, W., K. & Murrell, J. C. (1999). Characterization of methanotrophic Pel, R., Parkes, R. J. & Cappenberg, T. E. (1998). Direct linking of bacterial populations in soils showing atmospheric methane microbial populations to specific biogeochemical processes by uptake. Appl Environ Microbiol 65, 3312–3318. "$ C-labelling of biomarkers. Nature 392, 801–805. Hugenholtz, P., Goebel, B. M. & Pace, N. R. (1998). Impact of Bremner, J. M. & Keeney, D. R. (1966). Determination and culture-independent studies on the emerging phylogenetic view of isotope-ratio analysis of different forms of nitrogen in soils: III. bacterial diversity. J Bacteriol 180, 4765–4774. Exchangable ammomium, nitrate, and nitrite by extraction- Imhoff, J. F. (2001). Transfer of Rhodopseudomonas acidophila distillation methods. Soil Sci Soc Amer Proc 30, 577–582. to the new genus Rhodoblastus as Rhodoblastus acidophilus gen. Brown, A. H. F. & Iles, M. A. (1991). Water chemistry profiles nov., comb. nov. Int J Syst Evol Microbiol 51, 1863–1866. under four tree species at Gisburn, NW England. Forestry 64, Kishimoto, N., Kosako, Y. & Tano, T. (1991). Acidobacterium 169–185. capsulatum gen. nov. sp. nov. an acidophilic chemoorganotrophic

2341 S. Radajewski and others bacterium containing menaquinone from acidic mineral environ- radioisotope uptake by specific microbial cell types in situ. Appl ment. Curr Microbiol 22, 1–7. Environ Microbiol 65, 1746–1752. Kowalchuk, G. A., Stephen, J. R., De Boer, W., Prosser, J. I., Radajewski, S., Ineson, P., Parekh, N. R. & Murrell, J. C. (2000). Embley, T. M. & Woldendorp, J. W. (1997). Analysis of ammonia- Stable-isotope probing as a tool in microbial ecology. Nature 403, oxidizing bacteria of the β subdivision of the class Proteobacteria 646–649. in coastal sand dunes by denaturing gradient gel electrophoresis Reay, D. S., Nedwell, D. B. & McNamara, N. (2001a). Physical and sequencing of PCR-amplified 16S ribosomal DNA fragments. determinants of methane oxidation capacity in a temperate soil. Appl Environ Microbiol 63, 1489–1497. Water Air Soil Pollut 1, 401–414. Lee, N., Nielsen, P. H., Andreasen, K. H., Juretschko, S., Nielsen, Reay, D. S., Radajewski, S., Murrell, J. C., McNamara, N. & J. L., Schleifer, K.-H. & Wagner, M. (1999). Combination of Nedwell, D. B. (2001b). Effects of land-use on the activity and fluorescent in situ hybridization and microautoradiography – a diversity of methane oxidizing bacteria in forest soils. Soil Biol new tool for structure–function analyses in microbial ecology. Biochem 33, 1613–1623. Appl Environ Microbiol 65, 1289–1297. Rolfe, R. & Meselson, M. (1959). The relative homogeneity of Lidstrom, M. E. (1992). The aerobic methylotrophic bacteria. In microbial DNA. Proc Natl Acad Sci U S A 45, 1039–1043. The Prokaryotes, 2nd edn, vol. 1, pp. 431–445. Edited by A. Roslev, P. & Iversen, N. (1999). Radioactive fingerprinting of Balows, H. G. Tru$ per, M. Dworkin, W. Harder & K.-H. microorganisms that oxidize atmospheric methane in different Schleifer. New York: Springer-Verlag. soils. Appl Environ Microbiol 65, 4064–4070. Lidstrom, M. E., Anthony, C., Biville, F., Gasser, F., Goodwin, P., Rotthauwe, J.-H., Witzel, K.-P. & Liesack, W. (1997). The ammonia Hanson, R. S. & Harms, N. (1994). New unified nomenclature for monooxygenase structural gene amoA as a functional marker: genes involved in the oxidation of methanol in Gram-negative molecular fine-scale analysis of natural ammonia-oxidizing popu- bacteria. FEMS Microbiol Lett 117, 103–106. lations. Appl Environ Microbiol 63, 4704–4712. Ludwig, W., Strunk, O., Klugbauer, S., Klugbauer, N., Weizen- Sambrook, J., Fritsch, E. F. & Maniatis, T. (editors) (1989). Molecular Cloning: a Laboratory Manual, 2nd edn, vol. 1. Cold egger, M., Neumaier, J., Bachleitner, M. & Schleifer, K.-H. (1998). Spring Harbor, New York: Cold Spring Harbor Laboratory. Bacterial phylogeny based on comparative sequence analysis. Electrophoresis 19, 554–568. Stephen, J. R., McCaig, A. E., Smith, Z., Prosser, J. I. & Embley, T. M. (1996). Molecular diversity of soil and marine 16S rRNA Marmur, J. (1961). A procedure for the isolation of DNA from gene sequences related to β-subgroup ammonia-oxidizing microorganisms. J Mol Biol 3, 208–218. bacteria. Appl Environ Microbiol 62, 4147–4154. McCaig, A. E., Embley, T. M. & Prosser, J. I. (1994). Molecular Vinograd, J. (1963). Sedimentation equilibrium in a buoyant analysis of enrichment cultures of marine ammonia oxidisers. density gradient. In Methods in Enzymology, vol. VI, pp. 854–870. FEMS Microbiol Lett 120, 363–367. Edited by S. P. Colowicker & N. O. Kaplan. London: Academic McDonald, I. R. & Murrell, J. C. (1997). The methanol dehydro- Press. genase structural gene mxaF and its use as a functional gene probe Vorholt, J. A., Chistoserdova, L., Stolyar, S. M., Thauer, R. K. & for methanotrophs and methylotrophs. Appl Environ Microbiol Lidstrom, M. E. (1999). Distribution of tetrahydromethanopterin- 63, 3218–3224. dependent enzymes in methylotrophic bacteria and phylogeny of Meselson, M. & Stahl, F. W. (1958). The replication of DNA in methenyl tetrahydromethanopterin cyclohydrolases. J Bacteriol Escherichia coli. Proc Natl Acad Sci U S A 44, 671–682. 181, 5750–5757. Webster, G., Embley, T. M. & Prosser, J. I. (2002). Morris, S. A., Radajewski, S., Willison, T. W. & Murrell, J. C. Grassland management regimens reduce small-scale heterogeneity and (2002). Identification of the functionally active methanotroph species diversity of β-proteobacterial ammonia oxidizer popula- population in a peat soil microcosm by stable-isotope probing. tions. Appl Environ Microbiol 68, 20–30. Appl Environ Microbiol 68, 1446–1453. Wendlandt, K.-D., Jechorek, M., Helm, J. & Stottmeister, U. Moulin, L., Munive, A., Dreyfus, B. & Boivin-Masson, C. (2001). (2001). Producing poly-3-hydroxybutyrate with a high molecular Nodulation of legumes by members of the β-subclass of Proteo- mass from methane. J Biotechnol 86, 127–133. bacteria. Nature 411, 948–950. Whitby, C. B., Hall, G., Pickup, R., Saunders, J. R., Ineson, P., "$ Murrell, J. C., Gilbert, B. & McDonald, I. R. (2000a). Molecular Parekh, N. R. & McCarthy, A. (2001). C incorporation into DNA biology and regulation of methane monooxygenase. Arch Micro- as a means of identifying the active components of ammonia- biol 173, 325–332. oxidizer populations. Lett Appl Microbiol 32, 398–401. Murrell, J. C., McDonald, I. R. & Gilbert, B. (2000b). Regulation of Whittenbury, R., Phillips, K. C. & Wilkinson, J. F. (1970). En- expression of methane monooxygenases by copper ions. Trends richment, isolation and some properties of methane-utilizing Microbiol 8, 221–225. bacteria. J Gen Microbiol 61, 205–218. Nunn, D. N. & Lidstrom, M. E. (1986). Isolation and complementa- Wintzingerode, F. V., Gobel, U. B. & Stackebrandt, E. (1997). tion analysis of 10 methanol oxidation mutant classes and Determination of microbial diversity in environmental samples: identification of the methanol dehydrogenase structural gene of pitfalls of PCR-based rRNA analysis. FEMS Microbiol Rev 21, Methylobacterium sp. strain AM1. J Bacteriol 166, 581–590. 213–229. Orphan, V. J., House, C. H., Hinrichs, K.-U., McKeegan, K. D. & Yeates, C. & Gillings, M. R. (1998). Rapid purification of DNA DeLong, E. F. (2001). Methane-consuming archaea revealed by from soil for molecular biodiversity analysis. Lett Appl Microbiol directly coupled isotopic and phylogenetic analysis. Science 293, 27, 49–53.

484–487...... Ouverney, C. C. & Fuhrman, J. A. (1999). Combined microauto- Received 4 February 2002; revised 18 March 2002; accepted 15 April radiography–16S rRNA probe technique for determination of 2002.

2342