Marine Genomics 17 (2014) 53–62

Contents lists available at ScienceDirect

Marine Genomics CIESM 2013

Archaeal populations in two distinct sedimentary facies of the subsurface of the Dead Sea C. Thomas a,⁎, D. Ionescu b,c, D. Ariztegui a, the DSDDP Scientific Team 1 a b c

Department of Earth Sciences, University of Geneva, Switzerland Leibniz Institute for Freshwater Ecology and Inland Fisheries, Stechlin, Germany Max Planck Institute for Marine Microbiology, Bremen, Germany

a r t i c l e

i n f o

Article history: Received 9 May 2014 Received in revised form 2 September 2014 Accepted 2 September 2014 Available online 16 September 2014 Keywords: Hypersaline Metagenomics Methanogenesis Halobacteria Dead Sea Subsurface

a b s t r a c t Archaeal metabolism was studied in aragonitic and gypsum facies of the Dead Sea subsurface using highthroughput DNA sequencing. We show that the communities are well adapted to the peculiar environment of the Dead Sea subsurface. They harbor the necessary genes to deal with osmotic pressure using high- and lowsalt-in strategies, and to cope with unusually high concentrations of heavy metals. Methanogenesis was identified for the first time in the Dead Sea and appears to be an important metabolism in the aragonite sediment. Fermentation of residual organic matter, probably performed by some members of the Halobacteria class is common to both types of sediments. The latter group represents more than 95% of the taxonomically identifiable Archaea in the metagenome of the gypsum sediment. The potential for sulfur reduction has also been revealed and is associated in the sediment with EPS degradation and Fe–S mineralization as revealed by SEM imaging. Overall, we show that distinct communities of Archaea are associated with the two different facies of the Dead Sea, and are adapted to the harsh chemistry of its subsurface, in different ways. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The Dead Sea as we know it today is a remnant water body of the previous Neogene ingression of the Mediterranean Sea into the Jordan-Arava Rift Valley (Zak, 1967). The so-called Sedom lagoon yielded numerous evaporite deposits in the area and finally disconnected from the sea. Waters originating from this lagoon have evolved within this framework, towards today's well known Mg–Ca–Cl brines of the Dead Sea, due to interactions with the surrounding geology, and through various dissolution and evaporation processes (Zak, 1967; Stein et al., 2000; Katz and Starinsky, 2009). The Dead Sea is now lying 427 m below sea level (2013) and its continuous retreat since 1950 (Katz and Starinsky, 2009) has led to an extreme salinity (TDS) of 348 g·L−1 (Oren and Gunde-Cimerman, 2012). In addition, the high concentrations of chlorine and divalent cations in the lake (~6.1 M Cl−, ~2 M Mg2+ and ~0.5 M·Ca2+; Ionescu et al., 2012) make it harsher for microbes to develop (Oren, 2001), approaching MgCl concentrations of 2.3 M which are thought to be the upper limits for life (Hallsworth et al., 2007). While the Dead Sea chemistry is heading towards unfavorable conditions for life (Oren, 2010a), its history is different and life has

⁎ Corresponding author at: Department of Earth Sciences, University of Geneva, rue des Maraichers 13, 1205 Geneva, Switzerland. E-mail address: [email protected] (C. Thomas). 1 Complete list of Dead Sea Deep Drilling Project (DSDDP) scientists available at www.icdp-online.org.

http://dx.doi.org/10.1016/j.margen.2014.09.001 1874-7787/© 2014 Elsevier B.V. All rights reserved.

been observed to thrive when sufficient dilution of the water occurs. One such case is the outflow of submarine springs on the western shore of the lake, where a diverse microbial community has taken advantage of nutrient and salinity gradients to develop (Ionescu et al., 2012; Häusler et al., 2014). In other cases, during events of rainy winters, water dilution in a mixed upper layer has also led to blooms of the alga Dunaliella and subsequently of its archaeal degraders of the Halobacteria class (Oren and Shilo, 1982; Oren, 1983b, 2010a). During more arid periods, studies have shown that the Dead Sea water column biota consists of halophilic Archaea from the Euryarchaeota phylum (Oren, 1993; Oren and Ventosa, 1999; Bodaker et al., 2010). Variations in evaporation/ precipitation ratio impact the lake physics and chemistry, subsequently influencing the biodiversity of the lake, as well as its geological record, through changes in authigenic precipitation of minerals. These effects are currently investigated in depth within the project of building a water connection between the Red Sea and the Dead Sea (Oren et al., 2004; Bardavid et al., 2007; Abu Qdais, 2008). In the framework of the paleoenvironmental investigation carried out within the ICDP-sponsored Dead Sea Deep Drilling Project (DSDDP), we were interested in recovering geomicrobiological information on the lake's subsurface, particularly on the potential of microbes to interact with and influence geochemical proxies used for paleoenvironmental reconstructions. Emphasis was put on current subsurface communities, their adaptation to their surrounding and their potential metabolism in such hypersaline conditions. Investigations of Archaea communities in the water column and what influences their

54

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

development have been carried out previously (Bodaker et al., 2010; Oren, 1983b). In contrast, there is currently no data on the contribution of these Archaea to the subsurface of the Dead Sea. This is of particular interest since in the marine subsurface (Biddle et al., 2006; Lipp et al., 2008) and continental saline (Waldron et al., 2007) sediments, Archaea have been identified as major contributors to the biomass. Therefore, using metagenomic methods, we address here the metabolic potential and diversity of Archaea in the subsurface of the Dead Sea and their putative influence on two of the main sedimentary facies (Neugebauer et al., in review): aragonite-detritus and gypsum (hereafter AD and GY, respectively). These sedimentary facies are linked to major changes in lake physics and chemistry, triggered by climatic changes or important meteorological events. We thus approach our data with an eye on the mechanisms influencing changes in communities, and their response to the water column influence. 2. Methods The samples originate from ICDP core 5017-1A of the DSDDP expedition. This core was retrieved at coordinates N 31°30′28.98″, E 35°28′ 15.60″ (middle of the Dead Sea) from a depth of 297 m (one of the deepest point of the lake). Sample GY originates from a core catcher, at depth of 90.64 m below lake floor (Table 1). Sampling was done under sterile conditions using autoclaved pre-cut syringes (see Vuillemin et al. (2010) for more details), in a shore geobiologyspecialized lab set up especially for the drilling expedition (December 2010). Sample AD was taken from a core interval (2.74 m) during core opening party in June 2011 at GFZ Potsdam. Cores transported to the ICDP facilities at GFZ Potsdam were sawed into halves. One-half was dedicated to picture taking and non-destructive measurements, while the other was saved for immediate sampling in sterile conditions using similar pre-cut syringes. To prevent any sampling of contaminated or oxidized parts, the syringe minicore was taken from the middle of the liner, and the surface part was removed. DNA was then further extracted using a phenol–chloroform extraction protocol modified from (Ionescu et al., 2009). Cells were extracted from 0.5 g of sediment after multiple cycles of PBS rinsing, 30 s sonication and quick centrifugation. They were then incubated for 20 min at 95 °C in 0.5 mL lysis buffer of 0.1 M Tris, 50 mM EDTA, 100 mM NaCl and 1% SDS (pH 8). 250 μL of phenol:chloroform: isoamyalcohol was added and the samples centrifuged at maximum speed after incubation for 10 min at room temperature. Supernatant was extracted again with the same process and the upper phase collected using Phase Lock Gel TM (PLG) tubes (5 Prime). It was then cleaned twice with 0.5 mL 24:1 (v:v) chloroform: isoamylalcohol and DNA was precipitated overnight at −20 °C in 1 volume of isopropanol and 2% (final volume) of 3 M sodium acetate (pH 5.5). Pellets were washed with 0.5 mL of 70% ice cold ethanol after a 30 min centrifugation at maximum speed, dried and finally dissolved in 10 μL molecular grade water. The QIAEX II Gel Extraction Kit (Qiagen) was used to further remove salt and purify DNA fragments according to the manufacturer's instructions. DNA extracts obtained from samples AD (aragonite alternating with detritus sample) and GY (gypsum sample) were sent to MR DNA™ lab, at Shallowater, Texas for whole genome amplification (through MDA) and metagenomic sequencing. DNA quantification was realized with the Qubit® dsDNA HS Assay Kit (Life Technologies) and amplification performed using the

REPLI-g Midi Kit (Qiagen). Enzymatic fragmentation was carried out for 150 ng of each sample using the Ion Xpress Plus gDNA Fragment Library Preparation Kit (Life Technologies). Fragments of 200–300 bp were obtained from the Ion Shear reaction after 8 min in a 37 °C bath. After purification, Ion Xpress Barcode Adapters (Life Technologies) 13 and 14 were ligated to samples AD and GY respectively and underwent nick-repair following the manufacturer's instructions. Size selection at approximately 330 bp was done with the E-gel SizeSelect 2% Agarose Gel (Invitrogen). After new Qubit® determination, fragmented libraries were pooled to an equal DNA amount for each sample to create the final library. The latter was finally diluted to a concentration of approximately 78 pM, bound to Ion Sphere particles using the Ion OneTouchTM 200 Template Kit v2 DL resulting in 16.38% unenriched templated ISPs. After Ion OneTouch ES enrichment, template ISPs were sequenced on an Ion 318 chip using the Ion Torrent PGM (Life Technologies). Raw sequences were submitted to the MetaGenome Rapid Annotation using Subsystem Technology (MG-RAST) server for annotation and statistical analysis (Meyer et al., 2008). Annotations were made using the M5NR integrative database for proteins and M5RNA for 16S rRNA gene. Classification was made under MG-RAST default settings. Maximum e-value cut-off was 10−5, minimum identity cut-off at 60% and minimum alignment length cut-off of 15. Functions were classified using the Subsystem approach supported by the SEED environment (Overbeek et al., 2005). Statistical values regarding sequence length were obtained from that platform, and by using FastQC (Andrews, n.d.). The metagenomes are publicly available on MG-RAST under reference IDs 4561562.3 and 4561566.3 for GY and AD respectively. Diversity indexes were calculated using PAST (Hammer et al., 2001) at the class and genus levels, by excluding unclassified reads within the various phyla. Dominance D is measured as D = Σ((ni/n)2) where ni is number of individuals of taxon i. Shannon–Weiner Index H is H = Σ((ni/n)ln(ni/n)) and evenness is taken as Buzas and Gibson's evenness = eH/S, with S number of taxa. Scanning Electron Microscope investigation was done on samples after regular and critical point drying on a Jeol® JSM-7001 FA at the University of Geneva. Energy Dispersive X-ray spectroscopy was carried out on the same device. Samples were mounted on an aluminium stub with double-sided conductive carbon tape. An ultra-thin coating (15 nm) of gold was then deposited on the samples by low vacuum sputter coating. 3. Results Over the two samples analyzed, 2 147 029 quality-controlled sequences were obtained. Among them, 1 608 197 sequences encoded for 1 008583 peptides of which to 165 818 at least one annotation was assigned. The sequences had a mean length of 168 ± 63 bp and 171 ± 58 bp (for AD and GY respectively), with modal classes (of over 400 000 sequences for AD and over 900 000 for GY) of 230–239 bp for both samples. Overall, we obtained 128 259 sequences attributed to the domain Archaea. The remaining sequences, (bacterial part) will be discussed elsewhere. Sample GY displayed higher number of sequences phylogenetically related to Archaea (77647) but smaller archaeal proportion than AD (20.9% compared with 24.9%, i.e. 50 612 sequences for AD; Table 2; Table S1 in supp. material). Only few archaeal ribosomal rRNA gene sequences were obtained, therefore, classification using the M5RNA database produced few

Table 1 Sample description and principal geochemical characteristics. Data for sample AD have been retrieved from Nissenbaum (1975) as expected to be similar to the deep water mass before the 1979 turnover evidenced by the beginning of halite precipitation in the top meters of the core. Sample

Depth (m)

AD

2.74

GY

90.88

Lithology

Depositional environment

Salinity (%)

Na+ (mM)

Ca2+ (mM)

Mg2+ (mM)

Cl− (mM)

Alternating aragonite and mud laminae Gypsum

Startified Holocene Dead Sea

33.21

1726

429

1745

6177

End of Pleistocene non stratified Lake Lisan

24.90

1795

74

321

5438

SO2− (mM) 4 4.2 ~23

C. Thomas et al. / Marine Genomics 17 (2014) 53–62 Table 2 Sequence information and diversity indexes for both samples, calculated at the class level, for annotations against the M5RN database. Assemblage

AD

GY

Archaeal sequences Archaeal proportion (%) Mean sequence length Taxa Individuals Dominance Shannon–Weiner Evenness

50 612 20.9 168 ± 63 bp 9 44 218 0.1553 1.992 0.8144

77 647 24.9 171 ± 58 bp 9 81 061 0.95 0.1616 0.1306

results. In sample AD the genera Methanobacteria, Methanomicrobia and Thermococci from the Euryarchaeota phylum, and Nitrosopumilus from the Thaumarchaeota phylum were detected. No archaeal SSU rRNA sequence could be retrieved from sample GY (Table 3). Therefore, hereafter, taxonomy for both samples was obtained from the closest relative in the database of annotated protein coding sequences as suggested by the MG-RAST server, using default MG-RAST parameters:maximum e-value cutoff of 10−5, minimum identity cutoff of 60% and minimum alignment length cutoff of 15. Both samples had a similar number of archaeal classes; however, diversity was much higher in sample AD compared to GY at the class level (Shannon index of 1.99 and 0.16 respectively). Sequences putatively affiliated to all of the archaeal phyla (Euryarchaeota, Crenarchaeota, Korarchaeota, Nanoarchaeota and Thaumarchaeota) were identified (Table 2). A small fraction of the sequences (0.01% for GY, and 3.1% for AD) could not be attributed to any specific phylum but were attributed to the domain Archaea. In both samples, members of the Euryarchaeota phylum were dominant, with 88.6% and 99.1% of the AD and GY Archaea sequences, respectively. In the latter sample 96% of the Euryarchaeota sequences were associated to the Halobacteria class, resulting in a strong dominance index for GY (0.95). In the AD sample, no class was strongly dominating and the sequences were distributed between Halobacteria (25%), Methanomicrobia (20%), Methanobacteria (11%), Methanococci (10%), Thermococci (16%), Archaeoglobi (8%), Thermoplasmata (2%) and Methanopyri (3%) (Fig. 1), resulting in a larger evenness index (0.81 compared to 0.13 for GY; Table 3). Thermoprotei of the Crenarchaeota phylum covered 6.7% and 7.1% of the AD and GY sequences respectively. Unclassified Korarchaeota and Thaumarchaeota related sequences represented around 0.8% and 0.7% in the AD sample and 0.1 and 0.04% in GY, respectively. No Nanoarchaeota-related sequences were detected in the GY sample while they formed the remaining 0.07% of the assigned AD population. Classification extends only to the class level due to low coverage and is presented in the Supplementary material S1. Function subsystems were annotated against the redundant M5NR database using the SEED classification. They corresponded to three levels of “functional roles” which classify the final function, larger metabolic processes, or structural complexes (Overbeek et al., 2005). Level-one subsystems have been put in quotes for clarity. This approach shows that all archaeal classes in sample AD with the exception of Nanoarchaeota appear to possess similar functions (Fig. 2). Variations were found in “iron acquisition”, “metabolism of aromatic compounds”, “motility and chemotaxis”, “nitrogen metabolism”, “phosphorus metabolism”, “potassium metabolism” and “sulfur metabolism”, “stress

55

response” and “virulence, disease and defense” subsystems (9 out of 27 obtained subsystems). For sample GY, functional information was unevenly distributed and was mainly held by Halobacteria. Highest sequence percentages were attributed to “amino acids and derivatives”, “carbohydrates”, “protein metabolism” and “clustering-based subsystems” (Fig. 2). The latter consists of genes that have been evidenced to belong together based on functional coupling, although little is known on their function. In the “carbohydrates” subsystem, reads related to di- and oligosaccharides, sugars and alcohols and CO2 fixation were present. We identified a large number of genes that were related to “stress response”, being either oxidative (309 reads for AD, 234 for GY; hereafter AD/GY), osmotic (149/17), nitrosative (151/0), carbon starvation (0/19), detoxification (6/0) and heat shock (239/0). Genes related to synthesis and use of osmoprotectants such as cholinebetaine uptake and biosynthesis were more numerous in AD than GY (113/17). This was also the case for trehalose uptake and utilization (76/11) and for biosynthesis of galacoglycans (278/73). Moreover, sequences related to glycerol uptake and utilization were found in AD (133), but not in GY. “Potassium homeostasis” related sequences were obtained from both facies (279/85; Fig. 3A). In the “virulence, disease and defense” subsystem, most sequences were related to heavy metal resistance. Among them, AD had a larger number of arsenic resistance and copper homeostasis related sequences (149/73 and 600/143 respectively), while cobalt–zinc–cadmium resistance and mercury resistance dominated in GY (20/111 and 0/14) (Fig. 3B). The main difference between the two samples was found in genes related to methanogenesis. Except for 2 sequences related to methanopterin biosynthesis (no methanogenesis related sequences were found for GY (as compared to 60 for AD)). In contrast, sequences from AD were related to “methanogenesis strays” (according to the MG-RAST system these genes are involved in methanogenesis but not fully characterized) (21), one-carbon metabolism reads associated with methanogenesis (58) and methanogenesis from methylated compounds (8, related to Methanomicrobia; Fig. 4A). Genes related to acetogenesis from pyruvate were present in both samples (500/132). Similarly, genes involved in fermentation were found as well in both samples (381 sequences for AD and 108 for GY), with specific lysine fermentation associated to the “amino acids and derivatives” subsystem (102/62) and anaerobic degradation of aromatic compounds (4/13; Fig. 4B). Nutrient-specific metabolisms were also of importance in our metagenomes. Namely, numerous reads related to “phosphorus metabolism” were obtained from sample AD and GY (221 and 1082 respectively). These were related to alkylphosphonate utilization (1/12), high (85/1068; AD/GY) and low (30/2) affinity phosphate transporters and pyrophosphate-energized proton pump (101/0). “Nitrogen metabolism” related genes were also detected for both AD (422) and GY (362). Ammonia assimilation dominates the N related genes in metagenome AD (371) and was the only N metabolic pathway associated to GY (362). AD metagenome also comprises N fixation (42) and denitrification related genes (9) (Fig. 4C). Numerous genes from both metagenomes could be assigned to “sulfur metabolism”. For AD, 211 reads related to inorganic sulfur assimilation were recovered, and 559 for GY. Among the sequences from the aragonitic sample, 18 were related to an enzyme involved in the dissimilatory pathway of sulfate

Table 3 M5RNA database (LSU and SSU rRNA fragments) annotations obtained for Archaea from sample AD. Domain

Phylum

Class

Order

Family

Genus

Abundance

Avg e-value

Avg % ident

Archaea Archaea Archaea Archaea Archaea

Euryarchaeota Euryarchaeota Euryarchaeota Euryarchaeota Thaumarchaeota

Methanobacteria Methanomicrobia Methanomicrobia Thermococci Unclassified Thaumarchaeota

Methanobacteriales Methanomicrobiales Methanosarcinales Thermococcales Nitrosopumilales

Methanobacteriaceae Methanospirillaceae Methanosarcinaceae Thermococcaceae Nitrosopumilaceae

Methanothermobacter Methanospirillum Methanosarcina Thermococcus Nitrosopumilus

5 1 1 4 4

−24 −27 −36 −31 −34

100 100 100 100 100

56

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

Fig. 1. Archaeal classes obtained from the AD and GY samples based on the M5NR database.

reduction (sulfate adenylyltransferase, dissimilatory-type, EC 2.7.7.4), and were present either in Thermoplasmata, Thermoprotei or Thermococci members. Sample GY also contained sequences related to organic sulfur assimilation (16) while none were found for AD (Fig. 4D). All subsystem-based annotations associated to their archaeal classes can be found in Table S2 of the Supplementary material. 4. Discussion In order to appropriately interpret the metagenomic data resulting from the subsurface environment of the Dead Sea, we first discuss the issues pertaining to the method itself. We then highlight the main metabolic features by focusing on stress response and metabolic potential of the communities, with an emphasis on major issues faced by hypersaline communities in subsurface environments. 4.1. Metagenomics in poorly characterized environments Archaea from the sample GY are largely dominated by Halobacteria. Their abundance increases by a 7-fold factor compared with the AD sample, and encompasses almost 97% of the complete metagenome. Unfortunately, low coverage and copy number probably precluded recuperation of 16S rRNA gene sequences. Regarding the functions, GY subsystems were very poorly represented by other classes than Halobacteria. While coverage is likely to be incomplete in such environments, there is little chance that Archaea other than Halobacteria are truly present and active in the gypsum layer. Members of this class are thus expected to take part in most of the metabolic activity there. These types of analyses emphasize the issue of DNA origin, which is often raised when working with sedimentary environments. In such

settings, DNA could originate from active sedimentary communities, as well as from dormant or dead ones. When it comes to the Dead Sea, it is even more relevant since high salinities tend to slow down DNA degradation by decreasing depurination rates (Lindahl, 1993). Transport of microbes from the surrounding environment is probably an issue, and non-extreme-halotolerant organisms will probably die when in contact with the Dead Sea water. The remaining organic material is thus very likely to end up sedimenting, and the genetic material it carries will probably be mixed with that of autochthonous organisms. Since we had little possibilities for deciphering autochthonous from allochthonous sequences, we chose for a first approach to focus on Archaea. Secondly, we took additional care in interpreting functional information when it relates to phylogenetic classes not known to present adaptation for hypersaline environments. Thus, the presence of sequences linked to genes for osmolyte uptake and synthesis was a key point in our data analysis, and allowed us to better constrain the organisms that are potentially active in the Dead Sea subsurface. Regarding activity, classic geomicrobiology methods tested under the Dead Sea chemistry failed. Non-specific binding of fluorochromes used for FISH techniques have precluded its efficient use (Thomas, unpublished), and ATP-tester tested in the field and shown to work in sedimentary environment (Vuillemin et al., 2010) were inefficient at such salinity. Assumptions on the activity of microbes in the sediment could thus only be made through remnant of their activity such as tentative EPS production and degradation observed using SEM (Fig. 6A–B), or occurrence of authigenic iron sulfide minerals, H2S gas escape and sulfur concretions (Fig. 5A–C). While such data is mere suggestive, it provides a platform for hypothesis generation on the metabolic potential in the subsurface of the Dead Sea. Another potential hint for an active community was the lack of genes related to dormancy or sporulation. But its absence could also well

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

57

Fig. 2. Abundance plot of level 1 subsystems of the SEED classification for the various classes of the two samples based on annotations against the M5NR database using MG-RAST. Abbreviations: CV cofactors, vitamins, prosthetic groups, pigments; PP phages, prophages, transposable elements, plasmids.

be related to poor database coverage of those functions, or to poorlysuited methods for DNA extraction from spore-forming microbes. Obviously, the Dead Sea subsurface environment is poorly known and does present a vast amount of novel genomic material, which precludes precise binning, or at least lowers its quality. This has already been raised with respect to hypersaline environments by López-López et al. (2013)

who could not recover from metagenomes the taxonomic assemblages previously found in anoxic mats of a solar saltern using clone libraries (López-López et al., 2010). Approaching these metagenomes with customized pipelines, for example by using differential coverage binning to unravel rare taxa (Albertsen et al., 2013) is currently investigated. However, constant effort for obtaining whole genomes of species thriving in

Fig. 3. Abundance plot of level 3 subsystems for AD (left) and GY (right) related to A. osmotic adaptation and B. heavy metal tolerance annotated against the M5NR database. In brackets are always the first, and when existing, the second subsystem level name. Abbreviations: CBS clustering-based subsystems, CH carbohydrates, do di- and oligosaccharides, MT membrane transport, os osmotic stress, PM potassium metabolism, Res Resistance to antibiotics and toxic compounds, sa sugar alcohols, Se stress response, VD virulence, disease and defense.

58

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

Fig. 4. Abundance plot of level 3 subsystems for AD (left) and GY (right when present) related to A. methanogenesis, B. fermentation, C nitrogen metabolism and D. sulfur metabolism. Annotation against the M5NR database. For A and B, in brackets are the first, and when existing, the second subsystem level names. For C, nitrogen metabolism is level 1 subsystem, line headers are level 3 subsystems. For D, sulfur assimilation encompasses functions of inorganic sulfur assimilation except for organic sulfur assimilation (level 3 subsystem). Level 2 subsystems are either identical to level 3 or absent in the M5NR classification. Abbreviations: 1C one-carbon metabolism; AA amino-acids and derivatives; CH carbohydrates; cc central carbohydrate metabolism; CV cofactors, vitamins, prosthetic groups, pigment; Fe fermentation; MAC metabolism of aromatic compounds; pm2 pyruvate metabolism II: acetyl-CoA, acetogenesis from pyruvate. **function level.

hypersaline anoxic sediments or water column such as that of the Mediterranean deep hypersaline anoxic lakes must be sustained. The task is rendered difficult by the fact that some of the dominating classes in these environments have been proven to have acquired specific metabolism through lateral gene transfer (LGT). This is the case for Halobacterium, which may have acquired a subset of aerobic respiration genes of bacterial origin through several events of LGT (Kennedy et al., 2001). This type of event seems to be particularly important in the Dead Sea, where lateral gene transfer from a halophilic Archaea and thermophilic Bacteria of the Thermotoga class has been found (Rhodes et al., 2010). This work emphasizes the fact that LGT events may occur between organisms from different extreme environments, giving a potential explanation to the numerous sequences in our metagenomes that were related to thermophiles. Abundant reads associated with the

crenarchaeotal Thermoprotei, Korarchaeota, and from euryarchaeotal Thermococci, Thermoplasmata and Archaeoglobi classes suggest the presence of thermophilic to hyperthermophilic organisms in the Dead Sea subsurface environment. While water temperature is lukewarm, and may exceed 30 °C at the surface, it rarely reaches 25 °C at the bottom, and thus does not make a suitable set for the establishment of thermophilic species. The occurrence of LGT events may explain some of it. In their work, Rhodes et al. (2010) also stress the potential for recovering rare, unexpected but functioning organisms in the Dead Sea environments. External contribution, notably from deep hydrothermal springs related to the faulting of the Dead Sea Basin, that has been evidenced through seismic data (Niemi and Ben-Avraham, 1997), is not known to date. Hence, it may be that the obtained proteins share similarities with that of species of thermophilic classes, without truly belonging to

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

59

high salt-in strategy used by members of the Halobacteria group (Oren, 2001) can thus be also employed in both the AD and GY facies. Potassium uptake and efflux systems have been acknowledged to be key systems in Halobacterium species NRC-1, used as a model for haloarchaea (Ng et al., 2000). The overview of subsystem functions suggests that although more susceptible to osmotic stress, the aragonite community may be able to use various osmolites to cope with salinity, making the low salt-in strategy a dominant one for its community (Oren, 1999). However, Halobacteria communities, known as halophiles “par excellence” (Oren, 2002) may use the most energy-efficient strategy (high salt-in) in this sedimentary facies. In sample GY, dominated by Halobacteria-related sequences, K+ homeostasis will probably be preferred although sugar and choline uptake may be possible.

those thermophilic species. Moreover, since not all members of these particular classes are necessarily thermophilic, the attribution of this lifestyle to these organisms may be imprecise.

4.2.2. Toxicity of metals Several genes related to Halobacteria in sample GY, and more generally to Euryarchaeota in sample AD point towards the high adaptation of the archaeal community to heavy metals, which for most of them, are moderately (Cd, Zn) to highly (Co, Cu) enriched in the meromictic Dead Sea water column and its interstitial water (Nissenbaum, 1977). Arsenic, cobalt, zinc, cadmium and mercury AD are dealt with in sample AD using a subset of resistance proteins, pump driving and transporting ATPases and ion reductases, which are part of the larger system for heavy metal resistance in prokaryotes, and particularly in Archaea (Nies, 2003). Sequences related to copper homeostasis, found for unspecified Euryarcheota in sample AD and in Halobacteria-related sequences in sample GY, highlight the importance of dealing with this metal in the Dead Sea chemistry. While copper is an essential metal for certain enzymes, concentrations in the Dead Sea have been shown to be very high and to vary greatly with depth, location and season (Nissenbaum, 1977). The system for copper homeostasis mostly relies on copper translocating P-type ATPase which has been shown to be an efficient transporter of copper across the cytoplasmic membrane in Escherichia coli (Rensing and Grass, 2003) and is present in several archaeal extremophilic copper-resistant Ferroplasma genomes (BakerAustin et al., 2005). It is. Overall, Halobacteria of sample GY seem to harbor a large number of genes for adaptation to the metals enriched in the Dead Sea sediment. At shallower depths, in the aragonitic community, these adaptations are less frequent and are specialized on arsenic and copper.

4.2. Coping with the harsh Dead Sea conditions

4.3. Coping with deep sedimentary environments

4.2.1. Salinity The two metagenomes present a significant number of sequences related to stress response. However, samples seem to differ in the responses that can be given to this stress. The aragonitic sample (AD) has a higher number of sequences (relative and absolute as compared with GY) directly related to “osmotic stress”, probably inferring that in the environment representative of the aragonite sediment, specialization towards high salinity is not dominant, as compared with the gypsum sample, whose organisms are better adapted to high salinity. The presence of higher number of sequences related to choline and betaine uptake and biosynthesis indicates that the low-salt-in strategy for osmotic adaptation (Oren, 2008) is prevalent in the AD sample. This is confirmed also by an abundance of features related to biosynthesis of galactoglycans and related lipopolysaccharides and trehalose uptake and utilization (predicted ABC transporters among others). Additionally, Thermoprotei and Thermococci classes exhibit functions related to glycerol uptake and utilization which are not found in the gypsum sample (GY), suggesting that in the aragonitic facies a large array of solutes is at the disposition of Archaea for osmotic equilibration (Roesser and Müller, 2001). While low salt-in strategy genes were present in both samples, one could also find a high number of sequences related to “potassium metabolism” in the two metagenomes, indicating the potential for K+-homeostasis both in gypsum and aragonitic facies. The

4.3.1. Fermentation Fermentation is well represented in both samples and implies the use of various sources, from carbohydrates to amino acids and aromatic compounds (Fig. 4B). Hypersaline environments are known to accumulate numerous organic compounds (Ollivier et al., 1994). It appears that the metabolisms that survive in those extreme conditions have the potential for growing on multiple substrates, in order to take advantage of the variety and quality of organic matter available, which does not always correlate with its quantity (Glombitza et al., 2013). The specific archaeal enzyme involved in degradation of acetyl-CoA to acetate is abundant in sample AD, and only found once in sample GY. This ADPforming acetyl-CoA synthetase is specific to Archaea employing fermentation (Schäfer et al., 1993). While fermentative activity cannot be attributed to any specific phylum in sample AD it appears that Halobacteria are the only ones harboring fermentation related sequences in sample GY. Generally, they are considered as the main aerobic heterotrophs of hypersaline environments (Oren, 2010b), but members of this class are also known to degrade organic matter in anoxic settings, similar to the Dead Sea sediment. This is the case for Halorhabdus sp. (Wainø et al., 2000; Antunes et al., 2008) and Halobacterium sp. (Oren, 1983a). Here in sample GY, the Archaea are suggested to have the ability to perform anaerobic degradation or aromatic compounds and lysine. The latter is potentially also occurring in

Fig. 5. Possible indications of active S cycle in the sediment core. (A) Euhedral pyrite in an aragonite (needle) rich sediment (65 m below lake floor). Scale bar is 100 nm; (B) Fe–S globule in between aragonite needles and diatom fragments (338 m below lake floor). Scale bar is 1 μm; (C) Core picture of an interval of aragonite laminae alternating with detrital laminae (aragonite facies) with numerous sulfur concretions (white). The core is around 10 cm wide.

60

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

sample AD, but is not attributable to a specific class or even phylum. For both samples, fermentation of different products is possible. Halobacteria are major actors in both samples, although they would probably use different substrates: preferentially mixed acids in sample AD, and amino-acids, sugars and aromatic compounds in GY. In sample AD, they could be assisted in this degradation by Archaeoglobi, Thermococci and Thermoplasmata members mainly. 4.3.2. Methanogenesis We bring data that strongly suggest that methanogenesis is dominant in the aragonitic sample and is probably a major process in comparison to the gypsum sample. This is highlighted by various methanogenesis related C1 metabolism hits and a group of uncharacterized genes related to methanogenesis (classified as “methanogenesis strays” in the MG-RAST system) detected for AD. Additionally, pathways for methanopterin biosynthesis, a cofactor in methanogenesis (van Beelen et al., 1984) were detected. The number of sequences related to methanogenesis from methylated compounds suggests that this is an important metabolism in the aragonitic sediment (Fig. 4A). The substrate for methanogenesis under extreme salinities is more likely to be a methylated compound, as more energy can be derived from its degradation and thus be directed towards osmotic adaptation (Oren, 2010b). All the sequences related to methanogenesis from methylated compounds are attributed to putative members of the Thermoprotei and Methanomicrobia classes. Among the latter, Methanosarcina, which could be detected through their 16S rRNA gene signature, are notably able to perform methylotrophic methanogenesis (Garcia et al., 2000). This class also hosts the rare halophilic methanogenic genera Methanohalobium (Zhilina and Zavarzin, 1987) and Methanohalophilus (Paterek and Smith, 1988; Boone et al., 1993; Davidova et al., 1997). However, these were not identified in our data. An in-depth study of the genomes of these organisms will give more precise information concerning identity of methanogens, substrate use, and adaptation to high salinity. 4.3.3. Sulfur metabolism Sequences related to the assimilation of inorganic sulfur are found in both metagenomes, however they were more abundant in sample GY than AD. The potential for organic sulfur assimilation (alkanesulfonate assimilation), was exclusively detected in sample GY. The latter can be used as an additional source for sulfur, as it has been highlighted in microbial mats of Pozas Azules (Breitbart et al., 2009). Alkanesulfonate assimilation is known to be expressed in E. coli in situations of S starvation (Eichhorn et al., 2000). This could be interpreted as good adaptation of gypsum communities to nutrient starving ecosystems. In sample AD, most sulfur metabolism genes relate to inorganic sulfur assimilation. Alkyl hydroperoxide reductase protein C is believed to be induced in conditions of sulfate starvation. It is very likely than in such anoxic setting, and as highlighted by previous geochemical studies of the Dead Sea or Lake Lisan sediment, sulfur is in its reduced form in the sediment (Nissenbaum and Kaplan, 1976; Torfstein et al., 2005). While methanogenic Archaea are likely to be present in this sediment, the finding of F420-dependent sulfite reductase provides a way of accommodating sulfite to sulfide rich environments, and methanogenic activity. Indeed, this enzyme allows Methanocaldococcus jannaschii to cope with sulfite and use it for S assimilation (Johnson and Mukhopadhyay, 2005). It even allows Methanococcus maripaludis to grow and produce methane with sulfite as its only sulfur source (Johnson and Mukhopadhyay, 2008). Such potential is very interesting since the Dead Sea core is generally rich in reduced sulfur, as evidenced by H2S degassing and numerous sulfur concretions (Neugebauer et al., in press). Overall, the archaeal communities have the potential for assimilating sulfur either as sulfate, even when it is potentially lacking, or as sulfite while it may be toxic. This brings a novel insight into the adaptations displayed by these extreme subsurface communities. Additionally, sulfate adenylyltransferase of the dissimilatory type (18 in AD) is putatively indicative of the occurrence of sulfate reduction in the aragonite level. This enzyme was identified as

part of the dissimilatory system of Archaeoglobus fulgidus, a hyperthermophilic sulfate reducing Archaeon of the Archaeoglobi class (Klenk et al., 1997). This class has been identified in the AD metagenome (Fig. 1). Dissimilatory sulfate reduction has been only recently detected in the Dead Sea (Haeusler et al., in revision), supporting several geochemical analyses that have concluded its occurrence in the lake's hypolimnion in the past. Accordingly, sulfate and sulfide S isotopes from the lower water mass before the 1979 turnover (Nissenbaum and Kaplan, 1976) or from the outcropping laminated and disseminated gypsum of the Lisan Formation (Torfstein et al., 2005), show signature of dissimilatory sulfate reduction in the stratified paleo-Dead Sea. Other indications observed in the DSDDP core, like common euhedral pyrite minerals (Fig. 5A), iron–sulfur globules (Fig. 5B), H2S degassing or native S concretions (Fig. 5C) allow us to suggest the presence of an active S cycle in the aragonitic sediments. In sample AD, SEM investigation under secondary and backscattered electron, coupled with energy dispersive X-ray spectrometry shows the occurrence of Fe–S minerals, often embedded in a matrix which bares strong resemblance to microbial biofilms (Fig. 6A). The sulfide in Fe–S mineralization is likely to originate from the reduction of sulfate, available in the Dead Sea water (Nissenbaum, 1975). Cooccurrence with microbial biofilms suggests active degradation of the available organic matter, potentially by fermenters and sulfate reducing microbes (Dupraz et al., 2004), which could lead to the accumulation of Fe–S minerals within its matrix. Production and particularly degradation of EPS, suggested by microbes and gas vacuoles within the biofilm (Fig. 6B) outlines the in situ activity of the subsurface biosphere of the Dead Sea.

4.3.4. Nitrogen cycle Nitrogen cycling by microbial communities has been argued to be very limited in the Dead Sea (Stiller and Nissenbaum, 1999). Nevertheless, pathways towards the assimilation of ammonium, particularly abundant in the Dead Sea (Oren, 1983b), seem to be present in both metagenomes. Concomitantly, the recovery of Thaumarchaeota-related sequences, likely belonging to the Nitrosopumilus genus as highlighted by the SSU similarity (Table 3) supports the presence of nitrogen metabolism in this environment. However, this group is not known to be adapted to high salinity and anoxic conditions. Thus far members of this phylum were found in highly oligotrophic (Pelve et al., 2011), suboxic (Labrenz et al., 2010), and hyperthermal environments (Brochier-Armanet et al., 2012). Additionally, sequences related to nitrogen fixation and denitrification were uniquely detected in sample AD, and attributed to Crenarchaeotea of the Thermoprotei class. This is probably imparted to the fact that nitrate availability is limited in the Dead Sea (Stiller and Nissenbaum, 1999). Sequences related to these organisms had never been observed in the lake water community studies. The work by Bodaker et al. (2010) only identified one sequence related to Thaumarchaeota in the residual Dead Sea water community of 2007. The recovery of such genes in the sediment is thus unlikely to originate from the Dead Sea water column community. Direct transport and sedimentation is possible. These groups are active in soils (Leininger et al., 2006) and could have been carried along with floods. Crenarchaeota are also ubiquitously found at seawater salinities. Hypersaline conditions should then prevent their activity. Their occurrence in sedimentary environments, and in suboxic zones has been evidenced in several studies (Francis et al., 2007). Correlative peaks in archaeal amoA genes and Crenarchaeota 16S rRNA gene sequences within the suboxic zone of the Black Sea water column (Coolen et al., 2007) suggest a potential for activity in oxygen poor environments. It is clear that recent findings suggest that Crenarchaeota members exhibit large flexibility in their metabolic and ecological potential (Francis et al., 2007). Nevertheless, our data only suggest the presence of genes related to ammonia assimilation and N fixation in the sediment. Further work including monitoring of incubated sediments or direct measurements of N fluxes would be needed to document nitrogen cycling in the Dead Sea.

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

61

Fig. 6. SEM pictures of active organic matter degradation suggestive of sulfate reduction and fermentation in samples AD (A) Backscattered electron photograph of a biofilm-like structure in between euhedral halite and oxide minerals. Whitish spots are Fe–S mineralizations embedded within the microbial matrix. Their typical EDX spectrum is shown in the upper right corner. (B) Planktonic centric diatom exhibiting a biofilm-like structure in its central depression. Arrows show putative microbes. Note potential gas vacuoles formed in the EPS (white arrows). Scale bars are 1 μm.

5. Conclusion

References

Archaeal metagenomes recovered from two different sedimentary facies of the Dead Sea show the impact of salinity on the resident microbial communities. Dominance of Halobacteria (notably in the deepest sample) and presence of numerous genes related to osmotic adaptation, either from the high- or low-salt-in strategy, highlight the harshness of this environment. Metagenomes also reveal that these subsurface archaeal communities are adapted to high concentration of heavy metals, nutrient starvation and shortage of high quality organic matter. Differences in metabolic pathways employed in the two sedimentary facies have been found. The aragonitic sediment hosts communities that are probably capable of performing methanogenesis from methylated compounds. Methanogen-related DNA is for the first time shown to exist in the Dead Sea environment, thus getting closer to validating hypotheses of pioneering works that predicted the occurrence of methanogens in the lake (Marvin-DiPasquale et al., 1999). Additionally, fermentation and dissimilatory sulfate reduction are suggested by the aragonite metagenome and supported by Fe–S mineralizations and traces of EPS degradation observed under SEM. The 90 m gypsum sample is much less diverse. Retrieved DNA indicates that the archaeal community is mostly composed of members of the Halobacteria class, which have the potential for performing fermentation from various different substrates. Versatility more than specificity seems to be the key for survival at such depths. However, the complexity and poor knowledge of hypersaline subsurface environments is a clear barrier to the full understanding of the implication of these metagenomes. Complementing studies in this direction will help to better constrain the metabolism of such extreme assemblages. As an example, the complete assembly of the metagenomes of these samples, both from Archaea and Bacteria, will help unveil the specificity of such extreme environments, and the true metabolic potential of its inhabitants. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.margen.2014.09.001.

Abu Qdais, H., 2008. Environmental impacts of the mega desalination project: the Red– Dead Sea conveyor. Desalination 220, 16–23. Albertsen, M., Hugenholtz, P., Skarshewski, A., Nielsen, K.L., Tyson, G.W., Nielsen, P.H., 2013. Genome sequences of rare, uncultured bacteria obtained by differential coverage binning of multiple metagenomes. Nat. Biotechnol. 31, 533–538. Andrews, S., d. FastQC: a quality control tool for high throughput sequence data http:// www.bioinformatics.babraham.ac.uk/projects/fastqc. Antunes, A., Taborda, M., Huber, R., Moissl, C., Nobre, M.F., da Costa, M.S., 2008. Halorhabdus tiamatea sp. nov., a non-pigmented, extremely halophilic archaeon from a deep-sea, hypersaline anoxic basin of the Red Sea, and emended description of the genus Halorhabdus. Int. J. Syst. Evol. Microbiol. 58, 215–220. Baker-Austin, C., Dopson, M., Wexler, M., Sawers, R.G., Bond, P.L., 2005. Molecular insight into extreme copper resistance in the extremophilic archaeon “Ferroplasma acidarmanus” Fer1. Microbiology 2637–2646. Bardavid, R.E., Mana, L., Oren, A., 2007. Haloplanus natans gen. nov., sp. nov., an extremely halophilic, gas-vacuolate archaeon isolated from Dead Sea–Red Sea water mixtures in experimental outdoor ponds. Int. J. Syst. Evol. Microbiol. 57, 780–783. Biddle, J.F., Lipp, J.S., Lever, M. a, Lloyd, K.G., Sørensen, K.B., Anderson, R., Fredricks, H.F., Elvert, M., Kelly, T.J., Schrag, D.P., Sogin, M.L., Brenchley, J.E., Teske, A., House, C.H., Hinrichs, K.-U., 2006. Heterotrophic Archaea dominate sedimentary subsurface ecosystems off Peru. Proc. Natl. Acad. Sci. U. S. A. 103, 3846–3851. Bodaker, I., Sharon, I., Suzuki, M.T., Feingersch, R., Shmoish, M., Andreishcheva, E., Sogin, M.L., Rosenberg, M., Maguire, M.E., Belkin, S., Oren, A., Béjà, O., 2010. Comparative community genomics in the Dead Sea: an increasingly extreme environment. ISME J. 4, 399–407. Boone, D.R., Mathrani, I.M., Liu, Y., Menaia, J.A.G.F., Mah, R.A., Boone, J.E., 1993. Isolation and characterization of Methanohalophilus portucalensis sp. nov. and DNA reassociation study of the genus Methanohalophilus. Int. J. Syst. Bacteriol. 43, 430–437. Breitbart, M., Hoare, A., Nitti, A., Siefert, J., Haynes, M., Dinsdale, E., Edwards, R., Souza, V., Rohwer, F., Hollander, D., 2009. Metagenomic and stable isotopic analyses of modern freshwater microbialites in Cuatro Ciénegas,Mexico. Environ. Microbiol. 11, 16–34. Brochier-Armanet, C., Gribaldo, S., Forterre, P., 2012. Spotlight on the Thaumarchaeota. ISME J. 6, 227–230. Coolen, M., Abbas, B., van Bleiswijk, J., Hopmans, E., Kuypers, M., Wakeham, S., Sinninghe Damsté, J.S., 2007. Putative ammonia-oxidizing Crenarchaeota in suboxic waters of the Black Sea: a basin- wide ecological study using 16S ribosomal and functional genes and membrane lipids. Environ. Microbiol. 9, 1001–1016. Davidova, I., Harmsen, H.M., Stams, A.M., Belyaev, S., Zehnder, A.B., 1997. Taxonomic description of Methanococcoides euhalobius and its transfer to the Methanohalophilus genus. Antonie Van Leeuwenhoek 71, 313–318. Dupraz, C., Visscher, P.T., Baumgartner, L.K., Reid, R.P., 2004. Microbe–mineral interactions: early carbonate precipitation in a hypersaline lake (Eleuthera Island, Bahamas). Sedimentology 51, 745–765. Eichhorn, E., Van Der Ploeg, J.R., Leisinger, T., 2000. Deletion analysis of the Escherichia coli taurine and alkanesulfonate transport systems. J. Bacteriol. 182, 2687–2695. Francis, C., Beman, J.M., Kuypers, M., 2007. New processes and players in the nitrogen cycle: the microbial ecology of anaerobic and archaeal ammonia oxidation. ISME J. 19–27. Garcia, J.L., Patel, B.K., Ollivier, B., 2000. Taxonomic, phylogenetic, and ecological diversity of methanogenic Archaea. Anaerobe 6, 205–226. Glombitza, C., Stockhecke, M., Schubert, C.J., Vetter, A., Kallmeyer, J., 2013. Sulfate reduction controlled by organic matter availability in deep sediment cores from the saline, alkaline Lake Van (Eastern Anatolia, Turkey). Front. Microbiol. 4, 1–12. Hallsworth, J.E., Yakimov, M.M., Golyshin, P.N., Gillion, J.L.M., D'Auria, G., de Lima Alves, F., La Cono, V., Genovese, M., McKew, B. a, Hayes, S.L., Harris, G., Giuliano, L., Timmis, K.N., McGenity, T.J., 2007. Limits of life in MgCl2-containing environments: chaotropicity defines the window. Environ. Microbiol. 9, 801–813.

Acknowledgment This research was funded by the Swiss National Science Foundation (projects 200021-132529 and 200020-149221/1). The authors wish to thank A. Vuillemin for his support on the field and in the lab, and A. Martignier for her expertise on SEM analysis as well as two anonymous reviewers who greatly improved the quality of the manuscript.

62

C. Thomas et al. / Marine Genomics 17 (2014) 53–62

Hammer, Ø., Harper, D., Ryan, P., 2001. PAST: paleontological statistics software package for education and data analysis. Palaeontol. Electron. 4, 9. Häusler, S., Noriega-Ortega, B.E., Polerecky, L., Meyer, V., de Beer, D., Ionescu, D., 2014. Microenvironments of reduced salinity harbour biofilms in Dead Sea underwater springs. Environ. Microbiol. Rep. 6, 152–158. Ionescu, D., Penno, S., Haimovich, M., Rihtman, B., Goodwin, A., Schwartz, D., Hazanov, L., Chernihovsky, M., Post, A.F., Oren, A., 2009. Archaea in the Gulf of Aqaba. FEMS Microbiol. Ecol. 69, 425–438. Ionescu, D., Siebert, C., Polerecky, L., Munwes, Y.Y., Lott, C., Häusler, S., Bižić-Ionescu, M., Quast, C., Peplies, J., Glöckner, F.O., Ramette, A., Rödiger, T., Dittmar, T., Oren, A., Geyer, S., Stärk, H.-J., Sauter, M., Licha, T., Laronne, J.B., de Beer, D., 2012. Microbial and chemical characterization of underwater fresh water springs in the Dead Sea. PLoS One 7, e38319. Johnson, E.F., Mukhopadhyay, B., 2005. A new type of sulfite reductase, a novel coenzyme F420-dependent enzyme, from the methanarchaeon Methanocaldococcus jannaschii. J. Biol. Chem. 280, 38776–38786. Johnson, E.F., Mukhopadhyay, B., 2008. Coenzyme F 420-dependent sulfite reductaseenabled sulfite detoxification and use of sulfite as a sole sulfur source by Methanococcus maripaludis. Appl. Environ. Microbiol. 74, 3591–3595. Katz, A., Starinsky, A., 2009. Geochemical history of the Dead Sea. Aquat. Geochem. 15, 159–194. Kennedy, S., Ng, W., Salzberg, S., Hood, L., DasSarma, S., 2001. Understanding the adaptation of Halobacterium species NRC-1 to its extreme environment through computational analysis of its genome sequence. Genome 11, 1641–1650. Klenk, H., Clayton, R.A., Tomb, J., Dodson, R.J., Gwinn, M., Hickey, E.K., Fleischmann, R.D., Quackenbush, J., Lee, N.H., Dougherty, B.A., Mckenney, K., Adams, M.D., 1997. The complete genome sequence of the hyperthermophilic, sulphate-reducing archaeon Archaeoglobus fulgidus. Nature 394, 6342–6349. Labrenz, M., Sintes, E., Toetzke, F., Zumsteg, A., Herndl, G.J., Seidler, M., Jürgens, K., 2010. Relevance of a crenarchaeotal subcluster related to Candidatus Nitrosopumilus maritimus to ammonia oxidation in the suboxic zone of the central Baltic Sea. ISME J. 4, 1496–1508. Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., Nicol, G.W., Prosser, J.I., Schuster, S.C., Schleper, C., 2006. Archaea predominate among ammonia-oxidizing prokaryotes in soils. Nature 442, 806–809. Lindahl, T., 1993. Instability and decay of the primary structure of DNA. Nature 362, 709–715. Lipp, J.S., Morono, Y., Inagaki, F., Hinrichs, K., Group, O.G., Group, G., 2008. Significant contribution of Archaea to extant biomass in marine subsurface sediments. Nature 454, 991–994. López-López, A., Yarza, P., Richter, M., 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. López-López, A., Richter, M., Peña, A., Tamames, J., Rosselló-Móra, R., 2013. New insights into the archaeal diversity of a hypersaline microbial mat obtained by a metagenomic approach. Syst. Appl. Microbiol. 36, 205–214. Marvin-DiPasquale, M., Oren, A., Cohen, Y., Oremland, R., 1999. Radiotracer studies of bacterial methanogenesis in sediments from the Dead Sea and Solar Lake (Sinai). In: Oren, A. (Ed.), Microbiology and Biogeochemistry of Hypersaline Environments. CRC Press, Boca Raton (FL), pp. 149–160. Meyer, F., Paarmann, D., D'Souza, M., Olson, R., Glass, E.M., Kubal, M., Paczian, T., Rodriguez, A., Stevens, R., Wilke, A., Wilkening, J., Edwards, R.A., 2008. The metagenomics RAST server — a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinforma. 9, 386. Neugebauer, I., Brauer, A., Schwab, M.J., Waldmann, N., Enzel, Y., Kitagawa, H., Torfstein, A., Frank, U., Dulski, P., Agnon, A., Ariztegui, D., Ben-Avraham, Z., Goldstein, S.L., Stein and the DSDDP Scientific Party, 2014. Lithology of the long sediment record recovered by the ICDP Dead Sea Deep Drilling Project (DSDDP). Quat. Sci. Rev. http://dx.doi.org/10. 1016/j.quascirev.2014.08.013 (in press). Ng, W.V., Kennedy, S.P., Mahairas, G.G., Berquist, B., Pan, M., Shukla, H.D., Lasky, S.R., Baliga, N.S., Thorsson, V., Sbrogna, J., Swartzell, S., Weir, D., Hall, J., Dahl, T. a, Welti, R., Goo, Y. a, Leithauser, B., Keller, K., Cruz, R., Danson, M.J., Hough, D.W., Maddocks, D.G., Jablonski, P.E., Krebs, M.P., Angevine, C.M., Dale, H., Isenbarger, T. a, Peck, R.F., Pohlschroder, M., Spudich, J.L., Jung, K.W., Alam, M., Freitas, T., Hou, S., Daniels, C.J., Dennis, P.P., Omer, a D., Ebhardt, H., Lowe, T.M., Liang, P., Riley, M., Hood, L., DasSarma, S., 2000. Genome sequence of Halobacterium species NRC-1. Proc. Natl. Acad. Sci. U. S. A. 97, 12176–12181. Niemi, T., Ben-Avraham, Z., 1997. Active tectonics in the Dead Sea Basin. In: Niemi, T., Ben-Avraham, Z., Gat, J. (Eds.), The Dead Sea: The Lake and Its Setting. Oxford University Press, Oxford-New-York, pp. 73–81. Nies, D.H., 2003. Efflux-mediated heavy metal resistance in prokaryotes. FEMS Microbiol. Rev. 27, 313–339. Nissenbaum, A., 1975. The microbiology and biogeochemistry of the Dead Sea. Microb. Ecol. 2, 139–161. Nissenbaum, A., 1977. Minor and trace elements in Dead Sea water. Chem. Geol. 19, 99–111. Nissenbaum, A., Kaplan, I.R., 1976. Sulfur and carbon isotopic evidence for biogeochemical processes in the Dead Sea. In: Nriagu, J. (Ed.), Environmental Biogeochemistry. Ann Arbor Science Publishers, Ann Arbor, pp. 309–325.

Ollivier, B., Caumette, P., Garcia, J.L., Mah, R.A., Caumettte, P., 1994. Anaerobic bacteria from hypersaline environments. Microbiol. Rev. 58, 27–38. Oren, A., 1983a. Halobacterium sodomense sp. nov., a Dead Sea Halobacterium with an extremely high magnesium requirement. Int. J. Syst. Bacteriol. 33, 381–386. Oren, A., 1983b. Population dynamics of halobacteria in the Dead Sea water column. Limnol. Oceanogr. 28, 1094–1103. Oren, A., 1993. The Dead Sea—alive again. Cell. Mol. Life Sci. 49, 518–522. Oren, A., 1999. Bioenergetic aspects of halophilism. Microbiol. Mol. Biol. Rev. 63, 334–348. Oren, A., 2001. The bioenergetic basis for the decrease in metabolic diversity at increasing salt concentrations: implications for the functioning of salt lake ecosystems. Hydrobiologia 466, 61–72. Oren, A., 2002. Diversity of halophilic microorganisms: environments, phylogeny, physiology, and applications. J. Ind. Microbiol. Biotechnol. 28, 56–63. Oren, A., 2008. Microbial life at high salt concentrations: phylogenetic and metabolic diversity. Saline Syst. 4, 2. Oren, A., 2010a. The dying Dead Sea: the microbiology of an increasingly extreme environment. Lakes Reserv. Res. Manag. 15, 215–222. Oren, A., 2010b. Thermodynamic limits to microbial life at high salt concentrations. Environ. Microbiol. 13, 1908–1923. Oren, A., Gunde-Cimerman, N., 2012. Fungal life in the Dead Sea. In: Raghukumar, C. (Ed.), Progress in Molecular and Subcellular Biology. Springer, Heidelberg, pp. 115–132. Oren, A., Shilo, M., 1982. Population dynamics of Dunaliella parva in the Dead Sea. Limnol. Oceanogr. 27, 201–211. Oren, a, Ventosa, a, 1999. Benjamin Elazari Volcani (1915–1999): sixty-three years of studies of the microbiology of the Dead Sea. Int. Microbiol. 2, 195–198. Oren, A., Gavrieli, I., Gavrieli, J., Kohen, M., Lati, J., Aharoni, M., 2004. Biological effects of dilution of Dead Sea brine with seawater: implications for the planning of the Red Sea–Dead Sea “Peace Conduit”. J. Mar. Syst. 46, 121–131. Overbeek, R., Begley, T., Butler, R.M., Choudhuri, J.V., Diaz, N., Chuang, H.-Y., Cohoon, M., de Crecy-Lagard, V., Disz, T., Edwards, R., Fonstein, M., Frank, E., Gerdes, S., Glass, E. M., Goesmann, A., Hanson, A., Iwata-Reuyl, D., Jensen, R., Jamshidi, N., Krause, L., Kubal, M., Larsen, N., Linke, B., McHardy, A.C., Meyer, F., Neuweger, H., Olsen, G., Olson, R., Osterman, A., Portnoy, V., Pusch, G., Rodionov, D., Rückert, C., Steiner, J., Stevens, R., Thiele, I., Vassieva, O., Ye, Y., Zagnitko, O., Vonstein, V., 2005. The subsystems approach to genome annotation and its use in the project to annotate 1000 genomes. Nucleic Acids Res. 33, 5691–5702. Paterek, J.R., Smith, P.H., 1988. Methanohalophilus mahii gen. nov., sp. nov., a methylotrophic halophilic methanogen. Int. J. Syst. Bacteriol. 38, 122–123. Pelve, E. a, Lindås, A.-C., Martens-Habbena, W., de la Torre, J.R., Stahl, D. a, Bernander, R., 2011. Cdv-based cell division and cell cycle organization in the thaumarchaeon Nitrosopumilus maritimus. Mol. Microbiol. 82, 555–566. Rensing, C., Grass, G., 2003. Escherichia coli mechanisms of copper homeostasis in a changing environment. FEMS Microbiol. Rev. 27, 197–213. Rhodes, M.E., Fitz-Gibbon, S.T., Oren, A., House, C.H., 2010. Amino acid signatures of salinity on an environmental scale with a focus on the Dead Sea. Environ. Microbiol. 12, 2613–2623. Roesser, M., Müller, V., 2001. Osmoadaptation in bacteria and archaea: common principles and differences. Environ. Microbiol. 3, 743–754. Schäfer, T., Selig, M., Schönheit, P., 1993. Acetyl-CoA synthetase (ADP forming) in archaea, a novel enzyme involved in acetate formation and ATP synthesis. Arch. Microbiol. 159, 72–83. Stein, M., Starinsky, A., Agnon, A., Katz, A., 2000. The impact of brine–rock interaction during marine evaporite formation on the isotopic Sr record in the oceans: evidence from Mt. Sedom, Israel. Geochim. Cosmochim. Acta 64, 2039–2053. Stiller, M., Nissenbaum, A., 1999. Geochemical investigation of phosphorus and nitrogen in the hypersaline Dead Sea. Geochim. Cosmochim. Acta 63, 3467–3475. Torfstein, A., Gavrieli, I., Stein, M., 2005. The sources and evolution of sulfur in the hypersaline Lake Lisan (paleo-Dead Sea). Earth Planet. Sci. Lett. 236, 61–77. Van Beelen, P., Labro, J.F., Keltjens, J.T., Geerts, W.J., Vogels, G.D., Laarhoven, W.H., Guijt, W., Haasnoot, C. a, 1984. Derivatives of methanopterin, a coenzyme involved in methanogenesis. Eur. J. Biochem. 139, 359–365. Vuillemin, A., Ariztegui, D., Vasconcelos, C., 2010. Establishing sampling procedures in lake cores for subsurface biosphere studies: assessing in situ microbial activity. Sci. Drill. 35–39. Wainø, M., Tindall, B.J., Ingvorsen, K., 2000. Halorhabdus utahensis gen. nov., sp. nov., an aerobic, extremely halophilic member of the Archaea from Great Salt Lake, Utah. Int. J. Syst. Evol. Microbiol. 50 (Pt 1), 183–190. Waldron, P.J., Petsch, S.T., Martini, A.M., Nüsslein, K., Nüslein, K., 2007. Salinity constraints on subsurface archaeal diversity and methanogenesis in sedimentary rock rich in organic matter. Appl. Environ. Microbiol. 73, 4171–4179. Zak, I., 1967. The geology of Mount SedomThesis Hebrew University of Jerusalem. Zhilina, T.N., Zavarzin, G.A., 1987. Methanohalobium evestigatus n. gen., n. sp. — an extremely halophilic methane-forming archebacterium (Methanohalobium evestigatus n. gen., n. sp.- Ekstremal'no-galofil'naia metanoobrazuiushchaia arkhebakteriia). Akademiia Nauk SSSR, Doklady, pp. 464–468.