New insights into the archaeal diversity of a hypersaline microbial mat obtained by a metagenomic approach

Systematic and Applied Microbiology 36 (2013) 205–214

Contents lists available at SciVerse ScienceDirect

Systematic and Applied Microbiology journal homepage: www.elsevier.de/syapm

New insights into the archaeal diversity of a hypersaline microbial mat obtained by a metagenomic approach ˜ c , J. Tamames d , R. Rosselló-Móra a A. López-López a,∗ , M. Richter b , A. Pena a

Marine Microbiology Group, Dptm. Recursos Naturals, Institut Mediterrani d’Estudis Avanc¸ats, IMEDEA-CSIC, C/Miquel Marqués 21, 07190 Esporles, Illes Balears, Spain Max Planck Institute for Marine Microbiology, Microbial Genomics and Bioinformatics Research Group, Bremen, Germany Área de Microbiología, Departamento de Biología, Universitat Illes Balears, Crta. Valledemossa, km 7.5, Palma de Mallorca, Illes Balears, Spain d Systems Biology Programme, Centro Nacional de Biotecnología CNB-CSIC, C/Darwin 3, 28049 Madrid, Spain b c

a r t i c l e

i n f o

Article history: Received 30 August 2012 Received in revised form 20 November 2012 Accepted 20 November 2012 Keywords: Archaea Methanogens Hypersaline habitats Microbial mats Metagenomic approach End-sequencing analyses

a b s t r a c t A metagenomic approach was carried out in order to study the genetic pool of a hypersaline microbial mat, paying more attention to the archaeal community and, specifically, to the putatively methanogenic members. The main aim of the work was to expand the knowledge of a likely ecologically important archaeal lineage, candidate division MSBL1, which is probably involved in methanogenesis at very high salinities. The results obtained in this study were in accordance with our previous report on the bacterial diversity encountered by using a number of molecular techniques, but remarkable differences were found in the archaeal diversity retrieval by each of the procedures used (metagenomics and 16S rRNA-based methods). The lack of synteny for most of the metagenomic fragments with known genomes, together with the low degree of similarity of the annotated open reading frames (ORFs) with the sequences in the databases, reflected the high degree of novelty in the mat community studied. Linking the sequenced clones with representatives of division MSBL1 was not possible because of the lack of additional information concerning this archaeal group in the public gene repositories. However, given the high abundance of representatives of this division in the 16S rRNA clone libraries and the low identity of the archaeal clones with known genomes, it was hypothesized that some of them could arise from MSBL1 genomes. In addition, other prokaryotic groups known to be relevant in organic matter mineralization at high salinities were detected. © 2013 Elsevier GmbH. All rights reserved.

Introduction Microbial mats are one of the oldest, most diverse and highly productive microbial ecosystems known [38]. The prokaryotic communities thriving in these habitats are fueled by oxygenic photosynthesis within the surface mat layers that support anaerobic assemblages in the underlying organic horizons [38]. Although largely recognized for their potential contribution to early oxygenation of the Earth, Archean microbial mats were also a likely source of reduction gases (including H2 , CH4 , CO and methyl sulphides), as demonstrated by a number of studies [3,6,15,20,56]. In a previous study, we investigated the prokaryotic abundance and diversity of a microbial mat formed at the bottom of a crystallizer in a multi-pond solar saltern in the south of Mallorca Island, and a highly diverse and dense benthic prokaryotic community was found to be thriving in such salt-saturated sediments [28]. One of

∗ Corresponding author. Tel.: +34 971611955; fax: +34 971611761. E-mail address: [email protected] (A. López-López). 0723-2020/$ – see front matter © 2013 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.syapm.2012.11.008

the most remarkable findings was the higher abundance and diversity of bacterial types compared to their archaeal counterparts. In fact, all the 16S rDNA archaeal clones retrieved affiliated with a single phylogenetic group, the candidate division MSBL1. This division was proposed as comprising the most abundant clones retrieved in a Mediterranean deep-sea hypersaline anoxic basin [55], for which no cultured representatives had ever been obtained. In a similar way, when molecular techniques were applied in order to explore the autochthonous microbiota of the brines of the crystallizers, the dominance of a phylotype was reported that had never been obtained previously by culturing techniques [4]. One decade later, Haloquadratum walsbyi was isolated, brought into pure culture, and described as one of the most ecologically relevant microorganisms in its habitat [5,25]. On the basis of the phylogenetic relatedness of MSBL1 to methanogens and the lack of other detected groups that might be responsible for methane production in the Mediterranean deep-sea brines, it was speculated that MSBL1 members might be involved in methanogenesis at very high salinities [55]. The 16S rRNA signature from very similar organisms was also found in the anoxic hypolimnion of a shallow hypersaline Solar Lake

206

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

in Egypt [8], and sediments of hypersaline Lake Chaka in China [19]. Methane is a major end product of anaerobic biomass degradation in anoxic environments where the concentrations of sulphate, nitrate, Mn(IV) or Fe(III) are low, since in their presence methanogenesis is normally outcompeted by anaerobic respiration, mainly for thermodynamic reasons [52]. Most of the methanogens live in environments of low ionic strength, or in marine biotopes, but some genera are able to grow in salt concentration s of up to 4 M NaCl [27]. In the upper layer of marine sediments, where the sulphate concentration is usually high, methanogenesis is restricted to substrates that cannot be metabolized by sulphate-reducing bacteria (SRB), such as the methylamines originating from the breakdown of osmoregulatory amines. In hypersaline ecosystems, the important source of these noncompetitive substrates is the N-trimethylated amino acid glycine betaine, which is an osmoregulator widespread among eukaryotes and prokaryotes living in hypersaline conditions [54]. Upon release from the cell, fermentative microorganisms can convert glycine betaine to methylamines, which provides an energy rich compound that can be used by many halophilic and halotolerant methanogens, but not by sulphate-reducing bacteria. Hydrogenotrophic methanogenesis could also be an important process in hypersaline environments since, contrarily to marine sediments, there are unusually high hydrogen partial pressures in hypersaline mats within the photosynthetically active surface layers of the mat [15,16]. Such conditions create favorable environments for hydrogenotrophic methanogenesis, since it alleviates the competition of methanogens with hydrogen-scavenging sulphatereducing bacteria [6,15]. Despite their relevance in carbon cycling within hypersaline habitats, few halophilic methanogens have been isolated recently, and none has been taxonomically described in the last decade. However, although recent studies highlight the presence of numerous Euryarchaeota that may be halophilic methanogens, they still have not been cultivated [32]. To understand carbon cycling better in extremely saline environments, as well as to understand how methanogenic microorganisms cope with desiccation in these habitats, it is important to study the diversity of halophilic methanogenic representatives, and attempt to culture representatives of this group. The lack of information about archaeal diversity in salt-saturated sediments, and especially about hypersaline methanogens, helped form the basis of this study that aimed to extend our knowledge of the genetic and metabolic diversity of the archaeal assemblage in a highly saline microbial mat of a Mediterranean multi-pond solar saltern (Mallorca, Spain), as well as to attempt to culture methane-producing members of the community. In addition, other prokaryotic groups known to be relevant in organic matter mineralization at high salinities (i.e. Firmicutes and sulphate-reducing bacteria [36]) were also considered.

Materials and methods Sample site and sample collection Sediment samples were taken from the Mediterranean S’Avall solar salterns located on the south-east coast of Mallorca Island at 39◦ 32 N; 002◦ 99 E. The sampling site, which had been previously characterized and named as Station E [28], is located in a crystallizer pond from which the salt is periodically harvested. Between 5 cm and 10 cm of the soft sediment were sampled in June 2009, prior to salt harvesting. At this time, the development of a microbial mat in the salt crust overlaying the black sediment was observable (Fig. 1). Thus, the microbial mat in the salt crust was taken as the superficial sample (sample ERV), and the underlying sediment as

Fig. 1. Extracted core from station E, showing the well-developed microbial mat at the sampling time (early summer).

the second subset sample (sample EN). Samples were extracted in methacrylate cores and, once homogenized, subsamples were taken with a sterile plastic syringe. Retrieval of microbial biomass Once taken, samples were immediately transported to the laboratory for processing. Due to sample manipulation difficulties encountered during DNA extractions, the cells were recovered from the solid phase. Briefly, homogenized subsamples (20 mL) were introduced into 50 mL Falcon tubes and mixed with 20 mL of PBS 4X by vigorous shaking. Cells were detached from sediments by immersing the tubes in an ultrasound bath (Selecta) for 15 min, mixing them every 3–4 min. The suspensions were centrifuged for 1 min at 2500 rpm and 14 ◦ C. The washing/ultrasound step was repeated four times (or until a clean supernatant was obtained). Supernatants were recovered and kept on ice in 50 mL Falcon tubes. A centrifugation step of 2 min at 2500 rpm was performed to remove any remaining sediment particles. The supernatant was transferred to a new 50 mL Falcon tube and centrifuged for 10 min at 10,000 rpm at 14 ◦ C. The segregation of cells from sediment does not significantly reduce the recovery of biomass, although it decreases the background and any contaminants [28]. The pellet was stored at −20 ◦ C for subsequent DNA extractions. Highly purified suspensions of microorganisms were obtained by density gradient centrifugation with Nycodenz (Axis-Shield PoC), as described previously [13]. Nucleic acid extractions The pellets obtained were resuspended in 5 mL saline buffer (1 M NaCl; 50 mM Tris–HCl), and the suspension was divided into 1 mL aliquots and introduced into 1.5 mL Eppendorf tubes. Tubes

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

were incubated for 1 h at 37 ◦ C with lysozyme (final concentration 1 mg mL−1 ), 30 min at 37 ◦ C with RNAse (final concentration 1 mg mL−1 ), 1 h at 50 ◦ C with proteinase K (final concentration 1 mg mL−1 ) and SDS (final concentration 1%). Incubations were performed in an orbital shaker (Thermo Electron Corp.) at 100 rpm. Lysates were extracted with phenol–chloroform–isoamylalcohol, as previously described [30]. The nucleic acids obtained were resuspended in 50 ␮L of sterile, nuclease-free water (Sigma). Fosmid library construction The environmental genomic libraries were constructed separately for samples ERV and EN using the CopyControl Fosmid Library Production kit (Epicentre, Madison, WI, USA), following the manufacturer’s instructions. Initial amounts of approximately 15 ␮g of DNA from both subsamples were used, as measured with a Nanodrop spectrophotometer (ND-1000 v3.5, Nanodrop Technologies, Inc.). Size selection of the genomic DNA was not necessary since most of the fragments obtained were in the 25–40 kb size range, as observed by pulsed-field gel electrophoresis (data not shown). The extracted DNAs were cloned into the 8139 bpPCC1FOS vector (Epicentre). The ligated fragments were then packaged into phage particles and used to transfect Escherichia coli EPI300-T1R (Epicentre). A total of 900 clones from sample EN and 700 clones from sample ERV were obtained (i.e. an initial library of 30–35 Mb for EN and 25–30 Mb for ERV, assuming an average insert size of 35–40 kb). Libraries were arranged in 250 ␮L 96-well plates. Complete sequencing of fosmids and annotation The selected fosmids were pooled in groups of 10 and pyrosequenced with 454 GS FLX Titanium technology (Roche) by the commercial sequencing service GATC Biotech (Konstanz, Germany). Potential protein-coding genes were identified by using Glimmer3 [10]. Transfer RNA genes were identified using tRNAScan-SE [29] and ribosomal RNA genes were searched with meta-rna 1.0 [17]. Annotation was performed by using the GenDB, version 2.2 system [34], supplemented by the tool JCoast version 1.7 [45]. For each predicted open reading frame (ORF), observations were collected from similarity searches against the sequence databases NCBI-nr, Swiss-Prot, KEGG and genomes DB [45], and the protein family databases Pfam [2] and InterPro [35]. SignalP was used for signal peptide predictions [3] and TMHMM for transmembrane-helix analysis [22]. Predicted protein coding sequences were automatically annotated by the in-house software MicHanThi [42]. The MicHanThi software predicts gene functions based on similarity searches using the NCBI-nr (including Swiss-Prot) and InterPro databases. The annotation of proteins highlighted within the scope of this study was subjected to manual inspection. For all observations regarding putative protein functions, an e-value cut off of 10−5 was considered. End-sequencing of fosmid clones Three 96-well plates were selected from sample EN and one from sample ERV (384 clones in total, both ends) for endsequencing. Induction of the fosmid clones to high-copy number clones was performed as recommended by Epicentre Biotechnologies (Madison, WI). The selected high-copy number clones were grown in 250 ␮L 96-well plates containing solid media and a selective antibiotic (LB with 12.5 ␮g mL−1 chloramphenicol), and they were sent to GATC (Konstanz, Germany) for end-sequencing. The primers used for sequencing were those included in the Epicentre kit (FP and RP). Fosmid end-sequences were revised and cleaned of vector contaminant sequences using Sequencher 4.1.4 software (Gene Codes Corp.).

207

For the taxonomic affiliation of the end-sequences, all ORFs were searched by BLASTx against the National Center for Biotechnology Information (NCBI) non-redundant protein sequences database [1]. BLASTx searches were used with expectation cut-off values of ≤1 × 10−10 in order to differentiate between archaeal and bacterial reads and, when possible, to bin the sequences within known taxa at a lower taxonomic level. Discordances in the affiliations obtained with each end of a single clone were manually inspected and corrected according to the following criteria: when each end matched different domains, the clone was not affiliated; when there were discordances at a lower taxonomic level (for example, one end affiliated with Deltaproteobacteria and the other to Gammaproteobacteria) the common affiliation was taken into account at the higher taxonomic level (Proteobacteria). All predicted ORFs were also searched for similarity against the cluster of orthologous groups (COG) [51] and the Kyoto Encyclopedia of Genes and Genomes (KEGG) [21] databases by using JCoast version 1.7 [45]. A match was counted if the similarity search resulted in an e value below 1 × 10−10 . Synteny analyses and taxonomic affiliation of the completely sequenced fosmids The Artemis Comparison Tool [7] was used to compare the sequences in this study with complete genomes deposited in the NCBI database. The comparison data were generated by performing BLASTn searches of the metagenomic sequences against the Methanohalobium evestigatum Z-7303 (CP002069), Aciduliprofundum boonei T469 (CP001941), Salinibacter ruber DSM 13855 (CP000159), H. walsbyi DSM16790 (AM1800888), and Halothermothrix orenii (CP001098) genomes, which allowed any regions of similarity to be identified. Taxonomic profiling based on environmental gene tag (EGT) analysis employing the WebCARMA software [12] was applied to the large contigs obtained as a first approach in order to differentiate between archaeal and bacterial fragments. Additionally, and in order to classify the contigs at lower taxonomic levels, the taxonomic ascription of the completely sequenced fosmids by phylogenetic inference was carried out with a set of genes that accomplished the following criteria: (1) minimum length of 100 amino acids (ranging from 100 to 1091 aa), (2) minimum identity of 30% with sequences deposited in databases (ranging from 30 to 88.5%), and (3) an e-value ≤ 1 × 10−10 . Those ORFs that did not show a phylogenetic signal (i.e. similar tree topologies to those obtained by using the 16S rRNA gene) were not taken into account for classification. For each fosmid, a minimum of two and a maximum of six protein coding genes were selected for phylogenetic inferences. Protein alignments were performed with Clustal W [24] and phylogenetic trees were constructed by the maximum likelihood method using the PHYLIP software [11]. As most of the metagenomic fragments analyzed in this study were only distantly related to sequenced genomes, the use of homology-based binning tools to affiliate them could have had limited accuracy [33]. In order to assess their relatedness further, a classification was performed based on genome sequence composition by calculating the tetranucleotide frequencies [40], and then comparing them with reference genomes. Tetranucleotide frequencies were calculated with JSpecies software [46]. Results End-sequencing analyses To gain information on the overall composition of the constructed metagenomic libraries and to search for archaeal reads,

208

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

four 96-well microplates (three from sample EN and one from sample ERV) were screened by end-sequencing. Sequences were regularly obtained of over 700 nucleotides that had a success rate of 87%. In total, 666 individual reads were obtained with an average size of 750 bp. After quality control, 565 sequences were kept for further analyses. Nucleotide sequences are deposited in Genbank under the accession numbers JY252052-JY252597. Taxonomic affiliation of the end-sequences BLASTx searches with a cut-off value of ≤1 × 10−10 were used to distinguish between bacterial and archaeal reads and, when possible, to bin the end-sequences within known taxa at a lower taxonomic level. In this way, 82% of the clones affiliated at a domain level and bacterial sequences were shown to be dominant in both sample layers (88% of bacterial vs 12% of archaeal clones in layer ERV, and 79% of bacterial vs 21% of archaeal clones in layer EN). At a lower taxonomic level, it was possible to affiliate approximately half of the clones of each library (47% in ERV and 54% in EN). Therefore, this sequence subset could not be representative enough for an accurate estimation of the prokaryotic diversity in the samples. However, the bacterial taxa detected in the analyzed fosmids were roughly the same as those found by 16S rRNA gene cloning and sequencing (although in very different proportions in layer ERV). On the other hand, the differences found in the archaeal diversity retrieval by each of the procedures were more remarkable (Fig. 2). According to sequence-end taxon binning, the ERV

metagenomic library was largely dominated by Bacteria, as reads affiliating with Archaea only represented 10% of the total. Among the Bacteria, Firmicutes (mainly the Halanaerobiaceae family) was the most represented bacterial group (42%), followed by Deltaproteobacteria (16%, mainly the Desulfohalobiaceae family), Bacteroidetes/Chlorobi (16%), and Gammaproteobacteria (11%). Groups with less representation were Alphaproteobacteria and Chlamydiae/Verrucomicrobia, which each accounted for 5%. Regarding Archaea, fosmids were affiliated with Halobacteriales (43%) and Methanomicrobia (14%), or were categorized as “Unclassified Euryarchaeota” (43%). In agreement with the scarcity of archaeal reads found in this sample, 16S rRNA genes could not be amplified with archaeal-specific primers in the extracted DNA of this superficial layer [28]. The EN metagenomic library was also dominated by Bacteria but clones associated with archaeal phylotypes were found in higher proportion (21% of the reads), in accordance with the previously reported abundances obtained by fluorescence in situ hybridization [28]. As in the 16S rRNA gene library, the most represented bacterial groups were found in similar proportions: Firmicutes (21%), Gammaproteobacteria (20%), Bacteroidetes/Chlorobi (18%), and Deltaproteobacteria (17%). Sequences with lower representation affiliated with Actinobacteria (10%), Alphaproteobacteria (6%), Chlamydiae/Verrucomicrobia (3%), and Planctomycetes (3%). Among archaeal clones, most sequences affiliated with Halobacteriales (76%), a group that comprises the majority of the hyperhalophilic archaeal members present in the brines overlying the sampled

Fig. 2. Taxonomic affiliations obtained in each layer for Bacteria (A) and Archaea (B) with 16S rRNA gene libraries and metagenomic end-sequence analyses.

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

hypersaline sediments [37]. Only a minor fraction of the assigned reads affiliated with methanogens of the Methanomicrobia, Methanobacteria, or Methanococci classes (6%), Thermoplasmata (5%), and Thermoprotei (4%). The rest of the clones (9%) were categorized as “Unclassified Euryarchaeota”. Functional classification of the end-sequences End-reads were functionally classified according to COG and KEGG categories [21,51]. Using a cut-off e-value of 10−5 , 45% of the ORFs were classified into COGs, with E (amino acid metabolism and transport, 11%) and C (energy production and conversion, 10%) being the dominant categories, followed by L (replication, recombination and repair), O (posttranslational modification, protein turnover and chaperones), T (signal transduction mechanisms), and P (inorganic ion transport and metabolism), which each accounted for 6%. No significant differences were observed at this level when comparing ERV and EN layers. With the applied cut-off value, 42% of the ORFs were assigned to KEGG categories (195 out of 463). Carbohydrate degradation pathways (glycolysis, citrate cycle, and pyruvate metabolism) were the most represented metabolic pathways in both layers (27% of the metabolic matches), followed by amino acid metabolism (20% of the metabolic matches). In addition, genes directly involved in carbon fixation (4.5%) and methane metabolism (1.2%) were detected. Chaperons, mainly heat shock proteins, and two-component signal transduction systems were well-represented in both layers (8 and 23 matches in the ERV and EN layers, respectively). Also important was the number of ABC transporters (21 matches), which act in the active transport of monosaccharides, amino acids, peptides, antibiotics, inorganic ions and osmoprotectans (such as glycine and betaine transporters). Cell motility and chemotaxisrelated proteins were only detected in the EN layer (25 matches). Completely sequenced fosmids As we were mostly interested in the archaeal assemblage of the studied environment, and specifically the methanogenic Archaea, the fosmids with at least one end matching with methanogenic Archaea (29 fosmids), Thermoplasmata (7 fosmids) and Thermoprotei (4 fosmids) were selected from the end-sequence dataset for complete sequencing by pyrosequencing. After sequencing and assembly, 36 large contigs were obtained with an average size of 33,352 bp (1.2 Mb in total) from which 30 were kept for further analyses (after removing chimeras and E. coli fragments). Nucleotide sequences of the completely sequenced fosmids were deposited in GenBank under the accession numbers JX684073–JX684102. Taxonomic classification of fosmids The taxonomic profiling based on EGT analysis employing WebCARMA was applied as a first approach to differentiate between archaeal and bacterial metagenomic fragments (Table S1), and, when possible, the position of the fosmids at lower taxonomic levels was assessed better by constructing phylogenetic trees of selected genes that accomplished the criteria described above. Table S1 lists the ORFs used for analyses of each fosmid, with their respective expectation values, the closest relative after phylogenetic reconstruction and, in the case that the selected ORF had no phylogenetic signal, the best match retrieved by BLASTp. By combining the results obtained with both phylogenetic inference and WebCARMA software, 14 fosmids could be affiliated to Archaea and 16 to Bacteria, and most of them were further classified at lower taxonomic ranks (Table S1).

209

The classification obtained by calculating the tetranucleotide frequencies based on genome sequence composition was highly congruent with the aforementioned analyses in most of the cases (Fig. 4). Briefly, six well-differentiated clusters of similarity were obtained: Cluster 1, embracing the two deltaproteobacterial reference genomes included in the analyses and fosmid FLSS-21, that was ascribed to the same taxonomic group by both WebCARMA and phylogenetic analyses; Cluster 2, containing nine of the fosmids ascribed to Firmicutes plus two reference genomes belonging to Firmicutes and one to methanogens; Cluster 3, composed of the only fosmid that affiliated to S. ruber and reference genomes of Bacteroidetes; Cluster 4, comprising seven out of the eight fosmids affiliated to methanogens, and the majority of the reference genomes of this archaeal group included in the analyses. Surprisingly, two of the Firmicutes reference genomes (Acetohalobium arabaticum and Halanaerobium praevalens) fell within this cluster (mostly composed of methanogens) with three of the fosmids that were ascribed to Firmicutes by WebCARMA and phylogenetic inference; Cluster 5 was composed of fosmids that were ascribed to different taxonomic groups (FLSS-3 to Firmicutes, FLSS19 to Halobacteria, FLSS-9 to Methanobacteria/Methanomicrobia, and FLSS-16 to Deltaproteobacteria) that seemed to be distantly related to the reference genomes used for the calculations; and, lastly, cluster 6 was composed of the two reference genomes belonging to haloarchaea and three out of the five fosmids that we had previously ascribed to this group. Although the clustering obtained by the distance in tetranucleotide frequencies was highly congruent with our affiliations and the 16S rRNA-based phylogeny, some misclassifications were detected (underlined in Fig. 4). Main features of the archaeal fosmids Half of the archaeal fosmids obtained affiliated with methanogens (8 out of 14) but only one showed synteny with the existing reference genome of a known microorganism, M. evestigatum [58]. In this regard, FLSS-30 (40,084-bp long containing 44 predicted ORFs) appeared completely syntenic to the corresponding section of the M. evestigatum Z-7303 genome (Fig. 3), with an average nucleotide identity of over 99.5%. Most of the predicted ORFs were shown to be involved in aminoacyl-tRNA biosynthesis and the metabolism of cofactors and vitamins. The different subunits of a V-type ATPase, most likely involved in methane metabolism, were also codified in this fosmid (Fig. 3). Of the remaining 13 fosmids affiliated with Archaea, five were classified within Halobacteria and one remained unclassified. These archaeal fosmids did not show synteny with any sequenced genome of known species and, moreover, the similarity of the annotated ORFs with deposited sequences in the databases was, in most cases, relatively low (approximately 30% amino acid identity). The phylogenetic analyses clearly affiliated FLSS-2 (34,456-bp long) with Methanomicrobia (Table S1). However, we could not be confident with the result since both WebCARMA and tetranucleotide frequency calculation classified it as belonging to Bacteria and Firmicutes, respectively. Among the annotated ORFs, a number of genes involved in amino acid metabolism were detected, as well as an ORF coding for carbon monoxide dehydrogenase (EC 1.2.99.2). Interestingly, this is a key enzyme in methane metabolism that forms part of a membrane-bound multienzyme complex which catalyses the overall reaction: CO + H2 O = CO2 + H2 . This enzyme may also be present in acetogenic bacteria, which utilize a reductive one-carbon pathway for the synthesis of acetyl-CoA, a metabolic precursor of both acetate and biomass [9]. As two of the three methods employed for classification placed this fragment within Firmicutes and, moreover, phylogenetic reconstruction of the ORF coding for carbon monoxide dehydrogenase indicated a close relationship with this same group (data not shown), this classification

210

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

Fig. 3. Synteny diagram of fosmid FLSS-30 and the corresponding section of the Methanohalobium evestigatum Z-7303 chromosome (CP002069). Colors correspond to the different COGs (clusters of orthologous groups of proteins).

was kept as valid. A similar problem occurred with fosmid FLSS25 (34,902 bp) that was affiliated with methanogens and clustered within the methanogens group in the tetranucleotide analyses. However, it was classified by WebCARMA as belonging to the Bacteria domain. In this case, the classification obtained with the first two methods (Methanopyri/Methanomicrobia/Methanococci) was kept as valid. The fosmid contained ORFs coding for proteins involved in pentose, fatty acid, and pyrimidine metabolism and transport, as well as components of ABC transporters and a P-type ATPase potassium ion transporter. The existence of an ORF coding for a Fic/DOC family protein, involved in cell division and the curation of prophages, was remarkable. FLSS-6 (38,601 bp) affiliated with methanogens, although a number of ORFs (coding for a DNA helicase, an ATPase, a translation elongation factor, and several transcriptional regulators) were closely related to A. boonei [44]. To assess its phylogenetic position better, phylogenetic trees of all ORFs coded in this fosmid were reconstructed and, interestingly, most of them indicated that this metagenomic fragment belonged to a methanogenic archaeon loosely related to the reference species available in the databases (as examples see Figs. S1, S2, and S3). Given the relative high abundance of MSBL1 16S rRNA gene sequences in the clone libraries [28], the putative archaeal nature of this fosmid (revealed by both phylogenetic and tetranucleotide analyses), and the low relatedness with known archaeal genomes, it could be hypothesized that this fragment belonged to the hitherto uncultured MSLB1 methanogenic archaea. Unfortunately, the lack of genetic information on MSBL1 members in public repositories did not allow such a hypothesis to be confirmed or rejected. FLSS-7 (43,201 bp) also affiliated with Archaea (by WebCARMA) and to methanogens (Methanopyri) by phylogenetic inference and tetranucleotide analyses. This fosmid contained, among others, ORFs coding for enzymes involved in the regeneration of coenzymes M and B after the action of coenzyme-B sulfoethylthiotransferase, which is used as the terminal methyl carrier in methanogenesis. FLSS-8 (35,887 bp) and FLSS-9 (26,405 bp) were ascribed to Methanobacteria/Methanomicrobia. The most remarkable finding in both fosmids was the existence of ORFs coding for formylmethanofuran dehydrogenase (EC 1.2.99.5), which catalyzes a methanofuran-dependent exchange between CO2 and the formyl group of formylmethanofuran, an intermediate in methanogenesis from CO2 . In addition, a number of chaperones (DnaJ, DnaK, GrpE, Hsp20), as well as a transcriptional activator of heat shock protein coding genes, were found in fosmid FLSS-8. DnaK, DnaJ and GrpE form a cellular chaperone machinery capable of repairing heat-induced protein damage in E. coli cells [49], and therefore they

might be highly useful for microorganisms thriving in an ecosystem with intense solar irradiation and drastic changes in temperature during the circadian cycle. FLSS-10 (28,704 bp) affiliated with Methanopyri/Methanobacteria and contained ORFs coding for archaeal-type flagella accessory protein D, isoprenoid biosynthesis, and a preprotein translocase SecD/F involved in the maintenance of a proton electrochemical gradient. ORFs codifying for different subunits of a V-type H+transporting ATPase, whose function could be the extrusion of Na+ ions coupled to the intrusion of H+ during methanogenesis, were detected. FLSS-23 (34,111 bp), affiliated with Methanobacteria/Methanomicrobia, contained ORFs coding for enzymes involved in amino acid metabolism and the de novo biosynthesis of purines and pyrimidines, and a high number of ORFs coding for hypothetical proteins encountered in different species of methanogens and other Archaea. FLSS-24 (36,116 bp) was phylogenetically affiliated with Methanosalsum zhilinae, an alkaliphilic, halophilic, methylotrophic species of methanogenic Archaea [31]. Although the phylogenetic analyses carried out with three different ORFs of the contig clearly placed this fragment into the class Methanomicrobia (Table S1), it was grouped within the Halanaerobiaceae family by both WebCARMA and tetranucleotide analyses (Table S1, Fig. 4), therefore this classification was kept as correct. As well as some ORFs involved in pyrimide and amino acid metabolism and transport, a number of ORFs coding for different types of transporters were found. Interestingly, an antibiotic resistance protein (transferase EC 2.7.1.95) and a transcriptional regulator of the TetR family were codified in this fragment. The transferase had been shown to confer resistance to various aminoglycosides, and the TetR family regulators are involved in the transcriptional control of multidrug efflux pumps, pathways for the biosynthesis of antibiotics, and in the response to osmotic stress and toxic chemicals [43]. Fosmids ascribed to Bacteria In addition, sixteen fosmids affiliated with Bacteria (Table S1) but none of them showed a high degree of synteny with reference genomes. Most of the bacterial-classified fosmids (12 out of 16) affiliated with Firmicutes (mainly Halanaerobiaceae and Clostridiaceae families). The calculated GC content of these fosmids consistently averaged 41% (ranging from 30 to 52%) and all but one (FLSS-3) fell within Firmicutes in the tetranucleotide clustering (Fig. 4). The remaining bacterial fosmids showed higher GC contents, ranging from 57 to 61%, and were ascribed to Deltaproteobacteria (two fosmids) and Bacteroidetes (one fosmid), whereas only one remained unclassified at a lower taxonomic level after

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

211

Fig. 4. Dendrogram showing the relationships of the fosmids with reference genomes based on tetranucleotide frequency calculations. Scale bar represents the percentage distance obtained from the calculated distance matrices using 256 tetranucleotide combinations.

the analyses (FLSS-18), since the three different methods used for classification yielded different results (Table S1, Fig. 4). Members of the detected groups are known to be important in organic matter mineralization at high salinities and their metabolic end products lead to interrelationships with methanogens in hypersaline habitats [36]. The possible interactions with the detected methanogenic archaea are discussed below. Discussion This study aimed at gaining first insights into the genomic repertoire of the archaeal candidate division MSBL1, a likely ecologically important lineage that has only been detected by molecular techniques in salt saturated environments, from both terrestrial [8,19,28] and deep-sea habitats [55]. In our previous work, all archaeal 16S rRNA gene sequences amplified from the anoxic layer of a hypersaline microbial mat matched with this group [28], and they were probably involved in methanogenesis at very high salinities [55]. Therefore, this habitat seemed to be highly suitable for expanding the knowledge of this phylogenetic group for which no cultured representatives had ever been obtained. To study the genetic pool of the crystallizer sediments, with the emphasis more on the archaeal community and, specifically, the putatively methanogenic members, a metagenomic approach was carried out in combination with culturing techniques (see Supplementary Material). To start building a picture of the archaeal

diversity and the functional potential of the environment, the high molecular weight DNA extracted from the microbial mat thriving in the intermediate salt layer flanking brines and anaerobic sediments (layer ERV; Fig. 1), and that extracted from the deeper part of the sediments (from 2 to 8 cm, layer EN; Fig. 1), were cloned separately. Overall, the diversity data obtained with metagenomics were in good agreement with that inferred from SSU rRNA gene libraries analyses in both layers [28]. At a domain level, the results obtained were in accordance with our previous report on the prokaryotic diversity encountered by a number of molecular techniques in these sediments, in which the bacterial population was shown to be more abundant and diverse than its archaeal counterparts [28]. However, although the results obtained with both approaches (metagenomics and 16S rRNA-based methods) were concordant with respect to the bacterial diversity in both the ERV and EN layers, remarkable differences were found in the archaeal diversity retrieval by each of the procedures. In the previous survey, we could not amplify 16S rRNA genes with archaeal-specific primers in an ERV subsample [28], which was in agreement with the scarcity of archaeal reads found in this layer. However, in layer EN all the archaeal 16S rRNA gene clones affiliated with the candidate division MSBL1, whereas by end-sequencing most of the clones were closely related to the order Halobacteriales (76%), which comprises the majority of the archaeal members present in the brines overlying the sampled hypersaline sediments [37]. In addition, a minor

212

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

fraction of the assigned reads affiliated with Methanomicrobia, Thermoplasmata, and Thermoprotei. The discrepancy between the two approaches in the archaeal diversity retrieval may be explained on the one hand by the scarcity of archaeal isolates, and hence of complete genomes occurring in environmental samples, and more specifically from hypersaline sediments. Despite the accumulated knowledge of several novel archaeal groups and their apparent abundance in both extreme and non-extreme habitats, few of these Archaea have been obtained in pure cultures in the laboratory, and the preliminary evidence for their specific physiologies mostly stems from environmental genomic studies [50]. On the other hand, assigning sequences of as-yet-uncultivated archaea to a specific lineage is not possible, as often nothing but the 16S rRNA gene sequence of the lineage is known. The existence of the candidate division MSBL1 was revealed during a survey in which 16S rRNA genes were directly amplified from the environment, and subsequent attempts to cultivate representatives of the group have failed [55]. As there is no clue about the genetic repertoire of MSBL1, it cannot be confirmed or discarded whether some of the fosmids in this study could have been fragments of their genome. Regarding the functional assignments, no significant differences were observed when comparing ERV and EN layers, most likely due to the underrepresentation of ERV reads rather than a similar functionality in both layers. Overall, carbohydrate degradation and amino acid metabolism were the most represented metabolic pathways, suggesting that: (i) heterotrophic metabolism of sugars is important in this environment, as shown previously in other hypersaline microbial mats [23], and (ii) amino acids are required in high amounts for growth and maintenance of the halophilic microbiota, possibly because most compatible solutes are based on amino acids and amino acid derivatives [47]. The high ratio of signal transduction genes found suggests a high cell-to-cell interaction level [41], agreeing with the idea that life in sediments occurs mostly associated to particles that generates microhabitats for the development of prokaryotic assemblages. The abundance of twocomponent signal transduction systems involved in chemotaxis was also remarkable, since they enable bacteria to sense, respond, and adapt to changes in their environment or their intracellular state, conferring adaptability to organisms, which is needed in an environment subjected to wider environmental fluctuations and gradients [23]. The presence of proteins related to cell motility is consistent with the requirement of the organisms living between oxic and anoxic zones to be motile and chemotactic, since they could accumulate substrates with high reductive potential in the anoxic zone, and then move to the oxic zone to harvest this potential by oxidation [23]. The fact that, in the oxic layer, no ORFs coding for proteins involved in motility were detected must also be due to the underrepresentation of reads from this sample. Chaperones, mainly heat shock proteins, were also detected in both layers, which can be associated with heat stress caused by the high level of sun irradiance in this habitat [48]. The taxonomic adscription of the 30 complete sequenced fosmids was carried out by three different approaches in order to increase the overall amount and accuracy of the assignments. The use of WebCARMA software allowed those ORFs encoding for known proteins to be detected and classified, but for reads containing more than one EGT the obtained classifications were merged into a single one (determined as the lowest common ancestor), thereby reducing not only potential misclassifications but also the number of actual classifications in the lower taxonomic ranks. In order to increase these, phylogenetic analyses were carried out for selected genes that allowed the metagenomic fragments to be classified more accurately. Lastly, as most of the fragments seemed to be only distantly related to sequenced genomes, their relatedness was assessed further by performing a clustering based on

sequence composition. For this purpose, the tetranucleotide frequencies were calculated, since it has been demonstrated that these frequencies carry an innate (although sometimes weak) phylogenetic signal that is independent of existing genomic data [40]. Some significant discrepancies were observed in the affiliations when the results obtained by homology- and sequence composition-based methods (Fig. 4) were compared but it was not possible to recognize whether such discrepancies reflected unequal evolutionary rates after divergence of the retrieved organisms from common ancestors [40], or were due to lateral gene transfer events [39]. It is well known that grouping prokaryotes based on sequence composition may result in relationships with important differences compared to those based on 16S rRNA phylogenies [40], but the high level of congruence between the two approaches (at least to differentiate Bacteria and Archaea) confirm the reliability of the affiliations obtained by the homology-based binning methods used. A total of 14 archaeal-related metagenomic fragments could be identified by contrasting the results obtained with the three approaches. Eight of these were confidentially ascribed to methanogens but it was only possible to identify the genus/species to which they belonged in one case. Although there is genomic information of some methanogenic genera that are usually found in hypersaline environments (Methanohalobium, Methanosarcina, Methanohalophilus and Methanosalsum [37]), only fosmid FLSS-30 showed synteny with the genome of a known methanogen (M. evestigatum Z-7303), with an average nucleotide identity of over 99.5%. M. evestigatum is an extremely halophilic microorganism with optimal growth at a salinity of 25% that was first isolated from sediments of a hypersaline lake. It is a methylotrophic methanogen that is able to use methanol and methylamines for growth, and it is found in a variety of hypersaline habitats [58]. The remainder of the archaeal fosmids (seven methanogen-related, five affiliating with Halobacteria, and one with unclassified Archaea) did not show synteny with any known genome-sequenced species. Aciduliprofundum boonei often appeared as the closest matching genome to the archaeal genes contained in our fosmid library, except for the haloarchaeal-related ones (data not shown). This microorganism is an anaerobic heterotrophic sulphur- and iron-reducing thermoacidophile belonging to the phylum Euryarchaeota that was isolated from a deep-sea hydrothermal vent [44]. Interestingly, 16S rRNA gene sequences of the candidate division MSBL1, the only phylogenetic group detected in our samples by 16S rRNA gene amplification and sequencing [28], were also retrieved in deep-sea waters with a high sulphide and methane content [55]. Taking into account that A. bonnei might have a similar life-style to MSBL-1 division representatives, it would not be surprising if both organisms shared a pool of homologous genes that allowed them to deal with the similar conditions of their particular environments. The bacterial fosmids found in our samples were ascribed to taxonomic groups with halophilic representatives (Halanaerobiaceae, SRBs, Bacteroidetes; see Table S1) that are known to be linked with methanogens at a trophic level. Some of them are able to produce methylamines (such as members of the Halanaerobiales) and/or acetate and trimethylamine from betaine (such as the halophilic homoacetogen A. arabaticum [59]), thus supplying the main substrates for methanogenesis in hypersaline environments. It is well known that methanogens may be outcompeted by sulphatereducing bacteria with increasing salinity (and, hence, sulphate availability) but, alternatively, methanogenesis may increase with salinity because of the use by other microorganisms of glycine betaine as a compatible solute to cope with high salinity. Additionally, competition may be diminished by sulphate being depleted, for example, at depth or by precipitation of sulphate minerals. It is clear that the relationship between methanogenesis and salinity is far from simple, and is dependent on the geological, historical and physicogeochemical characteristics of the environment, but

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

the interaction with other bacteria is crucial for the existence of methanogenesis at high salinities [32]. Despite the relative high diversity of methanogens revealed by the metagenomic approach, enrichment experiments designed for the isolation of methane producing microorganisms allowed only the cultivation of a single archaeal phylotype (Table S2). All the sequences obtained with the archaeal domain-specific primers were identical and showed 99% similarity with M. evestigatum DSM3721T , which is capable of disproportionating methanol (included in the culture medium) to methane and carbon dioxide [14]. Amplification products were also obtained from culture enrichments with bacterial domain-specific primers. The two bacterial phylotypes detected were closely related to sequences retrieved from a hypersaline microbial mat [18], with Halanaerobium hydrogeniformans being the most closely related cultured microorganism, but with only 90–92% sequence identity (Table S2). Assuming that for prokaryotic organisms similarity cut-offs of 93.2% and 87.7% for the 16S rRNA gene may be confidently used as genus and family boundaries, respectively [57], the bacterial sequences obtained probably belonged to a new genus of the family Halanaerobiaceae. Members of this family, included in the low G + C branch of the Firmicutes, are obligate anaerobes, and most live by fermentation of sugars or (in a few cases only) amino acids [37]. However, more work has to be undertaken in order to obtain pure cultures of the detected phylotypes so that their physiology and metabolism can be studied. In summary, the lack of synteny of most of the fragments in this study with any genome of a known microorganism, together with the low degree of similarity of the annotated ORFs with sequences in the databases, reflects the high novelty degree in the studied mat community, as already shown in similar environments [26]. It was not possible to link the sequenced clones with representatives of the division MSBL1, given the lack of additional information in the public gene repositories and the absence of 16S rRNA genes in our dataset. However, the high abundance of representatives of this division in the 16S rRNA clone libraries and the low identity of the putative archaeal clones with known genomes, led to the hypothesis that some of these fragments could arise from MSBL1 genomes. Nevertheless, it is clear that hypersaline habitats constitute a valuable source of new prokaryotic types and genes with novel activities and potential biotechnological applications, as pointed out by other authors [23,26,53]. Therefore, attempts to culture new methanogenic types remains as a research priority in order to understand the methanogen-specific phylogeny and the physiological features of their adaptation for living at extremely high salinities. Acknowledgments The research has been supported by the Consolider Ingenio CECSD2007-0005 and Spanish Plan Nacional CGL2009 12651-C02-02 projects, as well as by funds for competitive research groups from the Government of the Balearic Islands, which were all co-financed with FEDER support from the European Union. ALL was financed by a postdoctoral contract from the Balearic Islands University. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.syapm. 2012.11.008. References [1] Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402.

213

[2] Bateman, A., Coin, L., Durbin, R., Finn, R.D., Hollich, V., Griffiths-Jones, S., Khanna, A., Marshall, M., Moxon, S., Sonnhammer, E.L.L., Studholme, D.J., Yeats, C., Eddy, S.R. (2004) The Pfam protein families database. Nucleic Acids Res. 32, 138–141. [3] Bendtsen, J.D., Nielsen, H., von Heijne, G., Brunak, S. (2004) Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340, 783–795. [4] Benlloch, S., Martínez-Murcia, A., Rodriguez-Valera, F. (1995) Sequencing of bacterial and archaeal 16S rDNA genes directly amplified from a hypersaline environment. Syst. Appl. Microbiol. 18, 574–581. [5] Bolhuis, H., Poele, E.M., Rodriguez-Valera, F. (2004) Isolation and cultivation of Walsby’s square archaeon. Environ. Microbiol. 6, 1287–1289. [6] Buckley, D.H., Baumgartner, L.K., Visscher, P.T. (2008) Vertical distribution of methane metabolism in microbial mats of the Great Sippewissett Salt Marsh. Environ. Microbiol. 10, 967–977. [7] Carver, T.J., Rutherford, K.M., Berriman, M., Rajandream, M.A., Barrell, B.G., Parkhill, J. (2005) ACT: the Artemis comparison tool. Bioinformatics 21, 3422–3423. [8] Cytryn, E., Minz, D., Oremland, R.S., Cohen, Y. (2000) Distribution and diversity of archaea corresponding to the limnological cycle of a hypersaline stratified lake (Solar Lake, Sinai Egypt). Appl. Environ. Microbiol. 66, 3269–3276. [9] Drake, H.L., Kusel, K., Matthies, C. (2007) Acetogenic prokaryotes. Prokaryotes 2, 354–420. [10] Delcher, A.L., Bratke, K.A., Powers, E.C., Salzberg, S.L. (2007) Identifying bacterial genes and endosymbiont DNA with Glimmer. Bioinformatics 23, 673–679. [11] Felsenstein, J. 2005 PHYLIP (Phylogeny Inference Package) version 3.6. Distributed by the Author, Department of Genome Sciences, University of Washington, Seattle. [12] Gerlach, W., Jünemann, S., Tille, F., Goesmann, A., Stoye, J. (2009) WebCARMA: a Web application for the functional and taxonomic classification of unassembled metagenomic reads. BMC Bioinform. 10, 430–440. [13] Guazzaroni, M.E., Golyshin, P.N., Ferrer, M. (2010) Molecular methods to study complex microbial communities. In: Marco, D. (Ed.), Metagenomics: Theory, Methods and Applications, Caister Academic Press, Norfolk, UK, pp. 55–77. [14] Hinrichs, K.U., Hayes, J.M., Sylva, S.P., Brewer, P.G., DeLong, E.F. (1999) Methaneconsuming archaebacteria in marine sediments. Nature 398, 802–805. [15] Hoehler, T.M., Bebout, B.M., Des Marais, D.J. (2001) The role of microbial mats in the production of reduced gases on the early Earth. Nature 412, 324–327. [16] Hoehler, T.M., Albert, D.B., Alperin, M.J., Bebout, B.M., Martens, C.S., Des Marais, D.J. (2002) Comparative ecology of H2 cycling in sedimentary and phototrophic ecosystems. Anton. Leeuw. 81, 575–585. [17] Huang, Y., Gilna, P., Li, W. (2009) Identification of ribosomal RNA genes in metagenomic fragments. Bioinformatics 25, 1338–1340. [18] Isenbarger, T.A., Finney, M., Ríos-Velázquez, C., Handelsman, J., Ruvkun, G. (2008) Miniprimer PCR a new lens for viewing the microbial world. Appl. Environ. Microbiol. 74, 840–849. [19] Jiang, H., Dong, H., Zhang, G., Yu, B., Chapman, L.R., Fields, M.W. (2006) Microbial diversity in water and sediment of Lake Chaka: an athalassohaline lake in northwestern China. Appl. Environ. Microbiol. 72, 3832–3845. [20] Jonkers, H.M., Koopmans, G.F., Van Gemerden, H. (1998) Dynamics of dimethyl sulfide in a marine microbial mat. Microb. Ecol. 36, 93–100. [21] Kanehisa, M., Goto, S., Kawashima, S., Okuno, Y., Hattori, M. (2004) The KEGG resource for deciphering the genome. Nucleic Acids Res. 32, 277–280. [22] Krogh, A., Larsson, B., von Heijne, G., Sonnhammer, E.L. (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305, 567–580. [23] Kunin, V., Raes, J., Harris, K., Spear, J.R., Walker, J.J., Ivanova, N., von Mering, C., Bebout, B.M., Pace, N.R., Bork, P., Hugenholtz, P. (2008) Millimeter-scale genetic gradients and community-level molecular convergence in a hypersaline microbial mat. Mol. Syst. Biol. 4, 198. [24] Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J., Higgins, D.G. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947–2948. [25] Legault, B.A., López-López, A., Alba-Casado, J.C., Doolittle, W.F., Bolhuis, H., Rodriguez-Valera, F., Papke, R.T. (2006) Environmental genomics of “Haloquadratum walsbyi” in a saltern crystallizer indicates a large pool of accessory genes in an otherwise coherent species. BMC Genomics 7, 171–184. [26] Ley, R.E., Harris, J.K., Wilcox, J., Spear, J.R., Miller, S.R., Bebout, B.M., Maresca, J.A., Bryant, D.A., Sogin, M.L., Pace, N.R. (2006) Unexpected diversity and complexity of the Guerrero Negro hypersaline microbial mat. Appl. Environ. Microbiol. 72, 3685–3695. [27] Liu, Y., Whitman, B. (2008) Metabolic, phylogenetic, and ecological diversity of the methanogenic Archaea. Ann. N. Y. Acad. Sci. 1125, 171–189. [28] López-López, A., Yarza, P., Suárez-Suárez, A., Antón, J., Niemann, H., RossellóMóra, R. (2010) Extremely halophilic microbial communities in anaerobic sediments from a solar saltern. Environ. Microbiol. Rep. 2, 258–271. [29] Lowe, T.M., Eddy, S.R. (1997) tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Res. 25, 955–964. [30] Massana, R., Murray, A.E., Preston, C.M., DeLong, E.F. (1997) Vertical distribution and phylogenetic characterization of marine planktonic Archaea in the Santa Barbara Channel. Appl. Environ. Microbiol. 63, 50–56. [31] Mathrani, I.M., Boone, D.R., Mah, R.A., Fox, G.E., Lau, P.P. (1988) Methanohalophilus zhilinae sp. nov., an alkaliphilic, halophilic, methylotrophic methanogen. Int. J. Syst. Bacteriol. 38, 139–142. [32] McGenity, T.J. (2010) Methanogens and methanogenesis in hypersaline environments. In: Timmis, K.N. (Ed.), Handbook of Hydrocarbon and Lipid Microbiology, Springer-Verlag, Berlin, Heidelberg, pp. 666–676.

214

A. López-López et al. / Systematic and Applied Microbiology 36 (2013) 205–214

[33] McHardy, A.C., Rigoutsos, I. (2007) What’s in the mix: phylogenetic classification of metagenome sequence samples. Curr. Opin. Microbiol. 10, 499–503. [34] Meyer, F., Goesmann, A., McHardy, A.C., Bartels, D., Bekel, T., Clausen, J., Kalinowski, J., Linke, B., Rupp, O., Giegerich, R., Pühler, A. (2003) GenDB, an open source genome annotation system for prokaryote genomes. Nucleic Acids Res. 31, 2187–2195. [35] Mulder, N.J., Apweiler, R., Attwood, T.K., Bairoch, A., Bateman, A., Binns, D., Bradley, P., Bork, P., Bucher, P., Cerutti, L., Copley, R., Courcelle, E., Das, U., Durbin, R., Fleischmann, W., Gough, J., Haft, D., Harte, N., Hulo, N., Kahn, D., Kanapin, A., Krestyaninova, M., Lonsdale, D., Lopez, R., Letunic, I., Madera, M., Maslen, J., McDowall, J., Mitchell, A., Nikolskaya, A.N., Orchard, S., Pagni, M., Ponting, C.P., Quevillon, E., Selengut, J., Sigrist, C.J., Silventoinen, V., Studholme, D.J., Vaughan, R., Wu, C.H. (2005) InterPro, progress and status in 2005. Nucleic Acids Res. 33, 201–205. [36] Ollivier, B., Caumette, P., Garcia, J.L., Mah, R.A. (1994) Anaerobic bacteria from hypersaline environments. Microbiol. Rev. 58, 27–38. [37] Oren, A. (2008) Microbial life at high salt concentrations: phylogenetic and metabolic diversity. Saline Syst. 4, 2. [38] Orphan, V.J., Jahnke, L.L., Embaye, T., Turk, K.A., Pernthaler, A., Summons, R.E., Des Marais, D.J. (2008) Characterization and spatial distribution of methanogens and methanogenic biosignatures in hypersaline microbial mats of Baja California. Geobiology 6, 376–393. [39] Pride, D.T., Blaser, M.J. (2002) Identification of horizontally acquired genetic elements in Helicobacter pylori and other prokaryotes using oligonucleotide difference analysis. Genome Lett. 1, 2–15. [40] Pride, D.T., Meinersmann, R.J., Wassenaar, T.M., Blaser, M.J. (2003) Evolutionary implications of microbial genome tetranucleotide frequency biases. Genome Res. 13, 145–158. [41] Quaiser, A., Zivanovic, Y., Moreira, D., López-García, P. (2011) Comparative metagenomics of bathypelagic plankton and bottom sediment from the Sea of Marmara. ISME J. 5, 285–304. [42] Quast, C. 2006 MicHanThi – design and implementation of a system for the prediction of gene functions in genome annotation projects. Master Thesis 2006, Max Planck Institute for Marine Microbiology, Microbial Genomics and Bioinformatics Research Group, Bremen, Germany, Available on request. [43] Ramos, J.L., Martínez-Bueno, M., Molina-Henares, A.J., Terán, W., Watanabe, K., Zhang, X., Gallegos, M.T., Brennan, R., Tobes, R. (2005) The TetR family of transcriptional repressors. Microbiol. Mol. Biol. Rev. 69, 326–356. [44] Reysenbach, A.L., Liu, Y., Banta, A.B., Beveridge, T.J., Kirshtein, J.D., Schouten, S., Tivey, M.K., Von Damm, K.L., Voytek, M.A. (2006) A ubiquitous thermoacidophilic archaeon from deep-sea hydrothermal vents. Nature 442, 444–447. [45] Richter, M., Lombardot, T., Kostadinov, I., Kottmann, R., Duhaime, M.B., Peplies, J., Glöckner, F.O. (2008) JCoast – a biologist-centric software tool for data mining and comparison of prokaryotic (meta)genomes. BMC Bioinform. 9, 177.

[46] Richter, M., Rosselló-Móra, R. (2009) Shifting the genomic gold standard for the prokaryotic species definition. Proc. Natl. Acad. Sci. U.S.A. 106, 19126–19131. [47] Roberts, M.F. (2005) Organic compatible solutes of halotolerant and halophilic microorganisms. Saline Syst. 1, 5. [48] Rodríguez-Valera, F., Ventosa, A., Juez, G., Imhoff, J.F. (1985) Variation of environmental features and microbial populations with salt concentrations in a multi-pond solar saltern. Microbiol. Ecol. 11, 105–107. [49] Schröder, H., Langer, T., Hartl, F.U., Bukau, B. (1993) DnaK, DnaJ and GrpE form a cellular chaperone machinery capable of repairing heat-induced protein damage. EMBO J. 12, 4137–4144. [50] Schleper, C., Jurgens, G., Jonuscheit, M. (2005) Genomic studies of uncultivated Archaea. Nat. Rev. 3, 479–488. [51] Tatusov, R.L., Fedorova, N.D., Jackson, J.D., Jacobs, A.R., Kiryutin, B., Koonin, E.V., Krylov, D.M., Mazumder, R., Mekhedov, S.L., Nikolskaya, A.N., Rao, B.S., Smirnov, S., Sverdlov, A.V., Vasudevan, S., Wolf, Y.I., Yin, J.J., Natale, D.A. (2003) The COG database: an updated version includes eukaryotes. BMC Bioinform. 4, 41. [52] Thauer, R.K., Kaster, A.K., Seedorf, H., Buckel, W., Hedderich, R. (2008) Methanogenic archaea: ecologically relevant differences in energy conservation. Nat. Rev. 6, 579–591. [53] Torres-Zapata, I., Gonzalez-Montalvo, A., Castro-Ruiz, C., Rios-Velazquez, C. (2010) Generation of large insert metagenomic libraries using indirect DNA extraction methods from benthic and ephemeral tropical hypersaline microbial mats. In: Méndez-Vilas, A. (Ed.), Current Research, Technology and Education Topics in Applied Microbiology and Microbial Biotechnology, Formatex Research Center, Spain, pp. 1569–1575. [54] Trüper, H.G., Galinski, E.A. (1989) Compatible solutes in haIophilic phototrophic procaryotes. In: Cohen, Y., Rosenberg, E. (Eds.), Microbial Mats. Physiological Ecology of Benthic Microbial Communities, American Society for Microbiology, Washington, DC, pp. 342–348. [55] van der Wielen, P.W., Bolhuis, H., Borin, S., Daffonchio, D., Corselli, C., Giuliano, L., D’Auria, G., de Lange, G.J., Huebner, A., Varnavas, S.P., Thomson, J., Tamburini, C., Marty, D., McGenity, T.J., Timmis, K.N. (2005) The enigma of prokaryotic life in deep hypersaline anoxic basins. Science 307, 121–123. [56] Visscher, P.T., Baumgartner, L.K., Buckley, D.H., Rogers, D.R., Hogan, M.E., Raleigh, C.D., Turk, K.A., Des Marais, D.J. (2003) Dimethyl sulphide and methanethiol formation in microbial mats: potential pathways for biogenic signatures. Environ. Microbiol. 5, 296–308. [57] Yarza, P., Ludwig, W., Euzeby, J., Amann, R., Schleifer, K.H., Glöckner, F.O., Rosselló-Móra, R. (2010) Update of the all-species living tree project based on 16S and 23S rRNA sequence analyses. Syst. Appl. Microbiol. 33, 291–299. [58] Zhilina, T.N., Zavarzin, G.A. (1987) Methanohalobium evestigatum gen. nov., sp. nov., extremely halophilic methane producing archaebacteria. Dokl. Akad. Nauk SSSR 293, 464–468. [59] Zhilina, T.N., Zavarzin, G.A. (1990) A new extremely halophile homoacetogen bacterium, Acetohalobium arabaticum gen. nov., sp. nov. Dokl. Akad. Nauk SSSR 311, 745–747.