Identification of metabolically active proteobacterial and archaeal ...

Journal of Applied Microbiology ISSN 1364-5072

ORIGINAL ARTICLE

Identification of metabolically active proteobacterial and archaeal communities in the rumen by DNA- and RNA-derived 16S rRNA gene S.H. Kang1, P. Evans1, M. Morrison1,2 and C. McSweeney1 1 CSIRO Animal, Food and Health Sciences, Queensland Bioscience Precinct, St. Lucia, Brisbane, Qld, Australia 2 The Ohio State University, Columbus, OH, USA

Keywords 16S rRNA gene, mcrA, metabolic activity, methanogen, RNA, rumen. Correspondence Seungha Kang, CSIRO CAFHS, Queensland Bioscience Precinct, 306 Carmody Road, Brisbane, Qld 4067, Australia. E-mail: [email protected] 2012/2079: received 25 November 2012, revised 14 May 2013 and accepted 25 May 2013 doi:10.1111/jam.12270

Abstract Aims: To gain new insights into the metabolic contribution of bacterial group in the rumen. Methods and Results: Both DNA- and RNA-derived bacterial 16S ribosomal materials from bovine rumen contents were used as the template for bacterial community and analyse microbiota by three methods namely custom phylogenetic microarray, quantitative real-time PCR and denaturing gradient gel electrophoresis techniques. Bacterial analysis showed that genera affiliating with the Proteobacteria apparently made a greater metabolic contribution to rumen function than their population sizes indicated. Analysis of another rumen microbial group, the methanogens, using clone libraries for the expressed methyl coenzyme reductase subunit A (mcrA) revealed that an uncultivated methanogen clade contributes one-third of RNA-derived mcrA sequences based on a limited number of clones analysed. These uncultivated methanogen species were not observed in the mcrA gene library based on the DNA-derived sequences. Conclusions: The comparison of results obtained from DNA- and RNAderived materials suggests that some of the Proteobacteria and novel methanogen species appeared to be low in abundance in the rumen maintained on grain-based diets might play a greater role in rumen metabolism. Significance and Impact of the Study: These studies provide the first report to compare high-throughput analysis of bacterial 16S rRNA genes from DNAand RNA-derived materials to indicate differences that species make to community structure and metabolic activity.

Introduction The rumen microbial ecosystem is complex and plays an important role in the supply of nutrients to the host animal through the bioconversion of various plant materials into volatile fatty acids and microbial protein (Van 1994). Traditionally, metabolic processes in the rumen ecosystem have been derived from the study of axenic microbial cultures of species isolated as a means to identify metabolic capabilities (Hungate and Macy, 1973). More recently, molecular approaches that study 16S 644

rRNA gene sequences have identified a greater number of species not previously recognized by culture studies of rumen microbes (Stahl et al. 1988; Forster et al. 1997; Whitford et al. 1998; Wood et al. 1998; Kobayashi et al. 2000; Denman and Mcsweeney 2006; Mcsweeney et al. 2007; Kim et al. 2011). However, despite these advances, the 16S rDNA-based microbial profiles only provide a structural analysis of the ecosystem, while the viability and metabolic state of community members is not apparent (Sessitsch et al. 2002). On this basis, recent studies have noted that RNA-based techniques could provide

Journal of Applied Microbiology 115, 644--653 © 2013 The Society for Applied Microbiology

S. Kang et al.

information in addition to community structure that attempts to explain the metabolic activity of microbial populations in their natural environment (Wagner 1994; Keer and Birch 2003; Kang et al. 2009; Revetta et al. 2011). In this regard, methanogenic archaea are one group of particular interest in the rumen as they form the potent greenhouse gas methane during enteric fermentation (Cordruwisch et al. 1988). Numerically dominant methanogen species are believed to be responsible for the majority of methane formed. However, given that increased methane production correlates with increased transcripts of the methanogenesis-specific methyl coenzyme reductase subunit A (mcrA) gene (Freitag and Prosser 2009); those species expressing mcrA genes at a greater proportion than their densities suggest may contribute more to total methane than previously thought (Freitag and Prosser 2009). For this reason, studies of expressed mcrA genes in the rumen ecosystem are of interest to identify contributors to methane formation in the rumen. To give a more complete picture of the microbial metabolic profile in the rumen, a high-throughput system that allows species or groups to be analysed quickly and efficiently would be of value. For this reason, we recently have developed a RNA extraction method optimized for rumen contents, to identify which microbial species make contributions to rumen metabolism greater than previously attributed by DNA-based analyses (Kang et al. 2009). In this study, we compare a selected rumen bacterial taxonomy using genomic DNA- and RNA-derived materials through hybridization to a phylogenetic custom microarray that targets c. 500 species. Results from the custom microarray were then validated using denaturing gradient gel electrophoresis (DGGE) and qRT-PCR techniques. These studies provide the first report to compare high-throughput analysis of bacterial 16S rRNA genes from DNA- and RNA-derived materials to indicate differences that species make to community structure and metabolic activity. We also investigate the methanogenic archaea in the bovine rumen by comparing mcrA gene clone libraries from DNA- and RNA-derived materials. Materials and methods Collection and processing of rumen content samples from Brahman steers Fresh rumen fluid collected from three fistulated adult Brahman-cross steers maintained on grain-based diets in the South East Queensland region (Mount Cotton, Brisbane) was strained through cheesecloth. Samples were stored either by adding 02 ml aliquot of combined rumen fluid with an equal volume of phosphate-buffered

Metabolically active proteobacterial and archaeal communities

saline (PBS, pH 74) and then stored frozen at 80°C, or by adding an equal volume of phenol–ethanol (PE) solution at 4°C (5% water-saturated phenol and 95% ethanol) to prevent RNA degradation (Papenfort et al. 2008). The samples preserved by the respective methods were then processed by high-speed centrifugation at 10 000 g for 1 min at room temperature, and the pellet was resuspended in dissociation solution (DSS; 01% Tween 80, 1% methanol, 1% tertiary butanol adjusted to pH 2 with HCl) (Kang et al. 2009). Following a low-speed centrifugation at 200 g for 3 min at room temperature, DNA from the supernatant fraction was extracted using the bead-beating method of Denman and McSweeney (2006), with total RNA extracted using the methods of Kang et al. (2009). Only one technical replicate for each animal was extracted and used in the downstream measurements of microbial diversity. RNA degradation in extracted material was inhibited by the addition of 1 ll RNAseOut (Invitrogen, Carlsbad, CA, USA). Once extracted, the absence of DNA contamination in RNA samples was confirmed by an inability to generate PCR products from the RNA extracts using universal bacterial 16S rRNA gene primers, 27F (5′-AGAGTTTGATCMTGGCTCAG-3′) and a 1492R primer (5′-GGYTACCTTGTTACGACTT-3′). Extracts of DNA and RNA were stored at 80°C and thawed on ice as required. PCR amplification of the 16S ribosomal nucleic acids for hybridization to custom phylogenetic microarray The development of the custom microarray and its associated methods used in this study is described in detail by Kang et al. (2010) with the relevant microarray probe information and other methodological details also accessible at the GEO database (http://www.ncbi.nlm.nih.gov/ geo/query/acc.cgi?acc=GPL9353). 16S ribosomal materials for custom microarray hybridization were generated directly from DNA extracts or cDNA synthesized from extracted RNA using the primer set F27 (5′-AGAGTTTGATCMTGGCTCAG-3′) and a T7/ R1492 fusion primer (5′-TCTAATACGACTCACTATAG GGGGYTACCTTGTTACGACTT-3′) for bacteria (the underlined region is the T7 primer). cDNA was synthesized from 500 ng of RNA extracts by reverse transcriptase PCR using the ThermoScript RT-PCR polymerase enzyme (Invitrogen) with random hexamer primers according to the manufacturer’s instructions. The PCR cycling conditions to generate the DNAderived and RNA-derived 16S rRNA genes for custom microarray hybridization consisted of initial denaturation for 5 min at 95°C, 30 cycles of amplification (30 s at 94°C, 30 s at 58°C and 90 s at 72°C) and final elongation at 72°C for 5 min. PCR amplicons were purified with the


645


MinElute PCR purification kit (Qiagen Inc., Valencia, CA, USA), and cRNA was generated by in vitro transcription-based synthesis of single-stranded RNA (cRNA) using the MEGAscript T7 in vitro transcription kit (Ambion, Austin, TX, USA). After purification with a MEGAclear kit (Ambion), 1 lg of the sample cRNA and 140 ng of standard cRNA (internal control) were labelled at the same time using Label IT lArray Cy5 reagent (Mirus, Madison, WI, USA) for 1 h at 37°C and protected from light, and then, 01 volumes of the 109 stop reagent (Mirus) was added to terminate the labelling reaction. The labelled cRNA (25 ll) was fragmented using 59 fragmentation buffer (Mirus) at 94°C for 15 min. No further purification steps were performed, and a total of 6 ll (c. 120 ng) of the labelled cRNA samples was made up to a volume of 30 ll with hybridization solution and incubated at 42°C for 16 h with the custom microarray (manufactured by CustomArray, Mukilteo, WA, USA). Comparative phylogenetic microarray analysis from DNA and RNA extracts Nucleic acid-hybridized microarray slides were scanned using an Axon GenePix 4000A microarray scanner (Axon Instruments, Inc., Union City, CA, USA) at 100% laser power, 350–400 PMT (photomultiplier sensitivity) and 5-lm resolution. The resulting images were analysed using the GenePix Pro 6.0 software (Axon Instruments), where all probes that exhibited a signal below the background level of negative control probes were excluded from further analysis. The resulting GenePix Results (GPR) format files were exported to GeneSpring 7.3 software (Agilent Technologies, Santa Clara, CA, USA) where signal intensity was normalized using the standard ‘one colour’ option of the GeneSpring program, and the effect between rDNA and rRNA was identified by one-way analysis of variance and differences. Statistical analyses were performed using a GeneSpring software program, one-way ANOVA and pairwise comparisons based on Tukey’s honest significance differences method. Differences between data sets were deemed significant when a P ≤ 005 was obtained. Methyl coenzyme reductase subunit A (mcrA) gene clone library construction and analysis mcrA amplification products were generated by PCR from both DNA and cDNA using primers and methods of Luton et al. (2002). Also, archaeal 16S ribosomal amplification products from DNA and cDNA materials were generated using the primers set 4Fa (5′-TCCGGTTGATCCTGCCRG) (Hershberger et al. 1996) and 1492R (5′-GGMTACCTT646

S. Kang et al.

GTTACGACTT) (Salzman et al. 2002) using the PCR amplification conditions of Hershberger et al. (1996). Correctly sized PCR products were confirmed by separation on a 1% agarose gel, excised, purified and ligated into the pGEM-T vector following the protocol of the Promega TA Cloning kit (Madison, WI, USA). 16S ribosomal and mcrA insert positive clones were randomly selected and sequenced with an ABI3130xl automatic sequencer (Applied Biosystems, Foster City, CA, USA). Sequences were edited manually for ambiguous bases; chimeras were identified using the Bellerophon chimera identifier software (Huber et al. 2004) and removed from further analysis. Sequences of sufficient quality were imported into the ARB phylogeny package (Ludwig et al. 2004) and aligned with database sequences and similarity matrices constructed using the Kimura-2-correction parameter method (Kimura 1980). Aligned sequences were exported to the Cyberinfrastructure for Phylogenetic Research (CIPRES) phylogeny portal and a maximum-likelihood dendogram for mcrA protein sequences constructed using Randomized Axelerated Maximum Likelihood (RAxML) (Stamatakis et al. 2008) with 1000 bootstrap resamplings. Exploring the results obtained from the microarray using quantitative real-time PCR amplification Real-time PCR was performed with either extracted DNA or synthesized cDNA diluted 1 : 10 using primers specific for Desulfovibrio spp., Rhizobium etli., Clostridium leptum subgroup (Clostridial cluster IV), Oscillospira sp. (Clostridial cluster IV) and Eubacterium rectale (Clostridial cluster XIVa) described in Table 1 using the iCycler Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). Amplification reactions consisted of 125 ll of 29 iQ SYBR Green Supermix (Bio-Rad), 300 nmol l 1 each of forward and reverse primers from Table 1 with c. 100 ng template DNA or cDNA and sterile distilled water to a final volume of 25 ll. Reaction conditions for nucleic acid amplification were 95°C for 10 min, followed by 40 cycles of 95°C for 15 s, 56°C for 30 s and 72°C for 30 s. The abundance of target species from both DNA and cDNA was expressed as a proportion of the total bacterial 16S rRNA gene copy number according to the equation: relative proportion = 2 (Ct target Ct total bacteria), where Ct represents threshold cycle as described by Kang et al. (2010). Validation of Proteobacteria custom microarray results with PCR-DGGE A nested PCR approach for DGGE analysis was used to assess differences in b-proteobacterial species between DNA and cDNA materials. An initial PCR with b-Proteobacteria-specific primers (F27: 5′-AGAGTTTGATCMT


S. Kang et al.


Table 1 Real-time PCR and DGGE primers used in the present study Length (bp)

Bacteria

Forward primer (5′–3′)

Reverse primer (3′–5′)

References

Real-time PCR Universal

CGGCAACGAGCGCAACCC (1114f)

CCATTGTAGCACGTGTGTAGCC (1275r) ACATCTAGCATCCATCGTTTACAGC (826r) ATTCAAGGGGTACCGTCTTC (466r)

Denman and Mcsweeney 2006 Fite 2004

130

Mackie 2003

391

CCTAGTATTCATCGTTTACGGCGTG (825r) GGGCTGGTAAGGTTCTGCGCGTT (967r) CTTCCTCCGTTTTGTCAA (1163r)

Schwiertz 2002

347

This study

225

Matsuki 2004

239

ATTACCGCGGCTGCTGG (534R)

Muyzer 1993

–

Desulfovibrio spp. (d-Proteobacteria) Oscillospira sp. (cluster IV) Eubacterium rectale (cluster XIVa) Rhizobium etli (a-Proteobacteria) Clostridium leptum subgroup (cluster IV) DGGE

CCGTAGATATCTGGAGGAACATCAG (691f) AAGGAGTTTTCGGACAACGG (75f) CGGTACCTGACTAAGAAGC (477f) TCCATTACTGACGCTGAGGTGCGAAGC (742f) GCACAAGCAGTGGAGT (932f) CGCCCGCCGCGCGCGGCGGGCGGGGCGGGG GCACGGGGGGCCTACGGGAGGCAGCAG (GC-341F)

135

DGGE, denaturing gradient gel electrophoresis. The underlined sequence corresponds to the GC clamp.

GGCTCAG-3′ and R680: 5′- TCACTGCTACACGYG -3′) that amplify the V3 16S region was employed using the methods of Fierer et al. (2005). PCR conditions were 15 min at 95°C, followed by 35 cycles of 95°C for 1 min, 30 s at 60°C and 72°C for 1 min. b-Proteobacteria amplicons were then PCR amplified with the GC clamp DGGE primer set for total bacteria (see Table 1) using the methods of Muyzer et al. (1993) on a GeneAmp PCR system 9700 (Applied Biosystems). PCRDGGE products were separated on an 8% (w/v) polyacrylamide gel with a denaturing gradient from 30% to 60% using a Dcode TM DGGE system (Bio-Rad Laboratories). Electrophoresis was performed at 80 V for 12 h at constant temperature of 60°C. The amplicon bands were stained using the AgNO3 method of Sanguinetti et al. (Sanguinetti et al. 1994) after completion of electrophoresis and images visualized using the Quantity One Image Capture software (Bio-Rad Laboratories). Bands of interest were excised from the gel, placed into 30 ll of Tris–EDTA buffer, freeze–thawed at 80°C for 10 min and 65°C for 10 min for three cycles to facilitate nucleic acid diffusion from the gel fragments and centrifuged at 13 000 g for 5 min. One microlitre of the supernatant was used as template to re-amplify using PCR with respective primers and conditions described previously. The PCR products were checked for correct size on an agarose gel (15%) and then purified using a MinElute PCR purification kit (Qiagen Inc.) before cloning into an appropriate vector. Inserts were cloned, sequenced and analysed as described for mcrA and archaeal 16S clone libraries.

Results Microarray and qRT-PCR analysis of bacterial profiles in rumen contents There is no relationship of storage method (PE or 80°C) with the DNA- and RNA-derived 16S rRNA gene sequences used for the microarray analysis, but microbial profiles showed differences when DNA- and RNA-derived materials pooled from three animals were compared. A total of 36 probes on the microarray had greater than a fivefold difference in signal intensity between the pooled DNA and RNA samples (P < 005) when analysed using the GeneSpring GX 7.3 software (Fig. 1a,b). The comparative microarray analysis indicated the RNA-derived signal was greater than the DNA-derived signal for species mostly belonging to the phylum Proteobacteria irrespective of the sample storage method (Fig. 1a). Of the upregulated proteobacterial species, Enterobacter sp., Desulfovibrio sp., Legionella sp., Escherichia sp. and Thiocapsa rosea. showed 10-fold greater signal than that of the DNA-derived material (Fig. 1b). Also, probes for members of the Firmicutes phylum such as Cl. leptum and Streptococcus spp. showed upscaled signal intensities in RNA-derived signal compared with the DNA-derived signal. Conversely, the highly abundant Firmicutes species (mainly those from the Clostridial cluster XIVa) produced a downscaled signal intensity for the DNA-derived samples (Fig. 1a). As a stand-alone result, the microarray data indicate significant differences (P < 005) between the quantity of


647


S. Kang et al.

Relative signal intensity (Log2)

(a)

(b) Enterobacter sp. Desulfovibrio sp. Legionella sp. Thiocapsa rosea Escherichia coli Acetobacter aceti Veillonella atypica Endosymbiont Desulfotomaculum spp. Desulfovibrio hydrothermalis Escherichia coli Atopobium rimae Helicobacter sp. Streptococcus mitis Treponema sp. Clostridium leptum Streptococcus pneumoniae Burkholderia cepacia Methylobacter sp. Methylocella palustris Thermodesulforhabdus norvegica Lactobacillus reuteri Dermatophilus congolensis Methylophaga sp. Desulfofrigus fragilis Butyrivibrio ruminis Ruminococcus obeum Butyrate producing bacteria Enterococcus gallinarum Ruminococcus products Clostridium clostridiiforme Methylocystis spp Clostridium sp. Ruminococcus productus Ruminococcus hansenii Clostridium difficile

5·0

1·0

0.3

PE (R)

–80 (R)

PE (D)

–80 (D)

*

Gammaproteobacteria Deltaproteobacteria Gammaproteobacteria Gammaproteobacteria Gammaproteobacteria Alphaproteobacteria Firmicutes Gammaproteobacterial Firmicutes Deltaproteobacteria Proteobacteria Actinobacteria Epsilonproteobacteria Firmicutes Spirocaea Firmicutes Firmicutes Betaproteobacteria Gammaproteobacteria Gammaproteobacteria Deltaproteobacteria Firmicutes Actinobacteria Proteobacteria Deltaproteobacteria Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Alphaproteobacteria Firmicutes Firmicutes Firmicutes Firmicutes

–15

Enterobacter sp. Desulfovibrio sp. Legionella sp. Thiocapsa rosea Escherichia coli Acetobacter aceti Veillonella atypica Endosymbiont Desulfotomaculum spp. Desulfovibrio hydrothermalis Escherichia coli Atopobium rimae Helicobacter sp. Streptococcus mitis Treponema sp. Clostridium leptum Streptococcus pneumoniae Burkholderia cepacia Methylobacter sp. Methylocella palustris Thermodesulforhabdus norvegica Lactobacillus reuteri Dermatophilus congolensis Methylophaga sp. Desulfofrigus fragilis Butyrivibrio ruminis Ruminococcus obeum Butyrate producing bacteria Enterococcus gallinarum Ruminococcus products Clostridium clostridiiforme Methylocystis spp Clostridium sp. Ruminococcus productus Ruminococcus hansenii Clostridium difficile

–10

–5

0

5

10

15

Folds

Figure 1 Differential signal intensity of probes on the custom microarray. (a) Selected probes representing those bacterial species show significant differences in relative abundance between DNA (D)- and RNA (R)-derived samples in the storage treatments of 80 ( 80°C degrees) and phenol/ethanol. The relative signal intensity between the DNA- and RNA-derived nucleic acids has been sorted by the fold change option of GeneSpring GX 7.3 analysis software. (b) Relative fold changes in bacterial species from DNA relative to RNA-derived materials from microarray analysis. Species showing greater than fivefold differences in probe signal intensity are presented. The bacterial names against probe identities are to the left and right of zero on the x-axis. *Butyrate-producing bacteria is belonging to clostridium cluster XIVa.

DNA- and RNA-derived materials for certain species, and on this basis, the custom microarray results were partially validated using real-time PCR (as a proportion of total bacteria). Desulfovibrio sp. that belongs to the d-Proteobacteria increased abundance in the RNAderived material for both the real-time PCR and microarray analyses of 7- (Fig. 2) and 10–14-fold (Fig. 1b) differences, respectively. Also, the Firmicutes species Cl. leptum that belongs to Clostridial cluster IV was identified as having almost fivefold difference by realtime PCR and fivefold higher for RNA-derived material over DNA-derived material (Fig. 1b). Other species that were not detected by the microarray but were assayed with real-time PCR showed Oscillospira sp. (belonging to the Clostridial cluster IV) increased by 25-fold in RNAderived material compared with DNA-derived material (Fig. 2), and Rh. etli (belonging to the a-Proteobacteria), only assayed by real-time PCR, showed high in RNAderived material as well (Fig. 2). Conversely Eu. rectale that belongs to the Clostridial cluster XIVa was approximately fivefold less in RNA-derived material at the realtime PCR (Fig. 2). These results show that although the exact fold change cannot be replicated in microarray 648

results, the direction of change is confirmed by the realtime PCR results. DGGE analysis of Proteobacteria As a further validation of the microarray analysis, DGGE targeting b-Proteobacteria species showed some different diversity between the DNA- and RNA-derived samples (Fig. 3). Several individual bands were observed to have greater intensity for the RNA-derived than DNA-derived materials (Fig. 3. arrows a and b). The sequence for one of the two bands excised and sequenced from the RNA-derived sample was most similar to the b-Proteobacterium Burkholderia cepacia with 98% identity (arrow A). The other excised band (arrow B) had 96% identity to an uncultured b-Proteobacterium clone TH_a192 belonging to a b-Proteobacteria group. Phylogenetic analysis of methanogen mcrA genes As an alternate to estimating diversity differences using microarray or real-time PCR methods as described previously, clone libraries were constructed from mcrA


S. Kang et al.


C.leptum subgroup Rhizobium etli Eubacterium rectale Oscillospira spp. Desulfovibrio genus –10

–5

0

5

10

15

20

25

30

Figure 2 Real-time PCR of selected bacterial species in DNA- and RNA-derived materials. Right bar (RNA) and left bar (DNA) on the y-axis represent bacterial species in higher abundance in the respective group.

M RNA DNA

Methanobrevibacter gottschalkii HOT species or most similar to the mcrA gene from a recently discovered methanogen, Methanomassiliicoccus luminyensis (Dridi et al. 2012) (Fig. 4). Interestingly, mcrA transcripts most similar to the M. luminyensis species were only identified from RNA-derived mcrA clones in 4 OTUs (12 clones), while the three remaining RNA-derived mcrA OTUs (20 clones) were most similar to either Mbb. ruminantium M1T species or Mbb. gottschalkii HOT. The DNA-derived mcrA sequence OTUs representing 15 clones were also most similar to the Mbb. ruminantium M1T and Mbb. gottschalkii HOT species. No archaeal 16S rRNA gene sequences were generated from DNA-derived materials for reasons unknown, but a total of 91 16S rRNA sequences from RNA-derived material were identified from the same cDNA material as the mcrA data, and all were most similar to the M. luminyensis species 16S rRNA gene (data not shown). These mcrA and 16S rRNA sequences of the M. luminyensis species affiliated sequences are similar to those previously identified from rumen samples of cattle and sheep (Tajima et al. 2001; Wright et al. 2004, 2006). Discussion

A B Figure 3 Denaturing gradient gel electrophoresis analysis PCR-amplified 16S rRNA gene fragment of beta-Proteobacteria-specific primer (27F/680R) set with an attached GC clamp. The arrow A: band (178 bp): Burkholderia cepacia (98% identity), the arrow B: band (195 bp):Uncultured beta-Proteobacterium clone TH_a192 (96% identity).

and 16S rRNA sequences to show differences in DNAand RNA-derived materials. In total, 47 mcrA gene sequences were generated from clone libraries, of which 15 of the 47 mcrA sequences were derived from genomic DNA and the remaining 32 sequences were obtained from cDNA synthesized from RNA-derived material. Analysis of the 47 mcrA clones found that they grouped with Methanobrevibacter ruminantium M1T species,

To identify bacterial or archaeal groups that were in low abundance (based on the proportion of the total community they make up) in an intestinal microbial community but were potentially contributing a significant portion of the community metabolic activity, we analysed 16S rRNA gene sequences and functional gene, mcrA, from DNAand RNA-derived rumen samples using phylogenetic microarray, DGGE and real-time PCR methods. The phenomenon of low abundance species contributing highly to an ecosystem is not unusual; studies of other environments such as the marine microbial community have brought new insights into the total bacterial profile through comparisons of DNA- and RNA-based clone libraries (Moeseneder et al. 2005). These authors suggested that previously undetected species at the DNA level can be detected using RNA-based methods, which have the potential to characterize distinct phylotypes from marine environment. From our results, the comparison of rumen contents of DNA- and RNA-derived materials showed significant changes for certain bacterial groups using our custom phylogentic microarray. These analyses suggested that although Proteobacteria are not regarded as a major bacterial phylum in the rumen, certain Proteobacteria, Enterobacter, Legionella, Desulfovibrio, Escherichia coli and Methylocella, may have a higher metabolic activity than absolute numbers suggest. This result was validated by DGGE analysis and confirmed results from a previous study where 28% of clones from


649


S. Kang et al.

Bovine rumen clone CLI5 (DQ192259) Bovine rumen clones (12 DNA and 18 RNA Clones) 0·01 Methanobrevibacter gottschalkii PG (EU919431) 53 Bovine rumen clone CLI02 (DQ192256) Methanobrevibacter smithii PS (ABU90120) 54 55 Macropod foregut clone TWM-MAY.01 (EU831308) Bovine rumen clone CLI15 (DQ192269) Methanobrevibacter olleyae ZA10 (EU9119431) Methanobreviacter oralis DSM7256 (DQ251045) Bovine rumen clones (3 DNA and 1 RNA Clones) 55 Bovine rumen clone CLI32 (EF379252) Methanobrevibacter ruminantium M1 (AAL29295) 60 Bovine rumen clone (1 RNA Clone) Methanobrevibacter wolinii GS (EU919432) Bovine rumen clone CLI17 (DQ192271) 98 Methanobrevibacter arboriphilus S1 (AAL29284) 53 Methanobacterium bryantii DSM863 (AAK16836) 61 Methanothermobacter wolfeii DSM2970 (BAF56659) Methanobacterium formicicum DSM1312 (AF414050) Methanobacterium formicicum DSM1312 (AF414051) 83 Methanothermobacter thermoautotrophicus ΔH (MTH1129) Macropod foregut clone TWM-MAY.10 (EU831377) 99 Methanosphaera stadtmanae MCB-3 (AAL29296) 92 Methanocaldococcus jannaschii (AAL29289) 80 65 Methanothermococcus thermolithotrophicus (AAL29297) Methanoculleus thermophilus (AAK16834) 58 Methanospirillum hungatei (AAL29287) 88 Methanosarcina barkeri FUSARO (AAZ69867) 88 M ethanocella paludicola SANAE1 (BAF56442) Bovine rumen clones (7 RNA Derived Clones) 55 Sludge digester clone HO4-2 (ABU90094) Bovine rumen clone CLI28 (EF379248) 88 Bovine rumen clone CLI25 (EF379248) 57 Sludge digester clone p24-2 (ABU90117) Bovine rumen clones (2 RNA Derived Clones) Bovine rumen clones (2 RNA Derived Clones) Bovine rumen clone (1 RNA Derived Clone) 64 Bovine rumen clone CLI9 (DQ192263) Sludge digester clone (ABU90121) 69 60 Sludge digester clone OS18 (AAL29263) Macropod foregut clone TWM-NOV.4 (EU831321) 99 Methanomasssilicoccus luminyensis B10 (AEO13319) Methanopyrus kandleri av19 (AAL29291)

Methanobacteriales

Methanococcales, Methanomicrobiales, Methanosarcinales, Methanocellales

Methanomassiliicoccusspecies affiliated clade

Methanopyrales

Figure 4 Phylogenetic dendogram of DNA- and RNA-derived mcrA clone library sequences with database mcrA sequences. Type in bold indicates those sequences identified by this study (grouped at a 97% similarity level) and the number of DNA- and RNA-derived sequences. Only bootstrapping values >50% are shown. Sulfolobus species are the designated outgroup.

rRNA-based clone libraries were affiliated with this phylum representing a much greater contribution to the microbial activity than expected (Kang et al. 2009). These results contrast with other studies that reported Proteobacteria to contribute