Report

The Enigma of Prokaryotic Life in Deep Hypersaline Anoxic Basins

See allHide authors and affiliations

Science  07 Jan 2005:
Vol. 307, Issue 5706, pp. 121-123
DOI: 10.1126/science.1103569

Abstract

Deep hypersaline anoxic basins in the Mediterranean Sea are a legacy of dissolution of ancient subterranean salt deposits from the Miocene period. Our study revealed that these hypersaline basins are not biogeochemical dead ends, but support in situ sulfate reduction, methanogenesis, and heterotrophic activity. A wide diversity of prokaryotes was observed, including a new, abundant, deeply branching order within the Euryarchaeota. Furthermore, we demonstrated the presence of a unique, metabolically active microbial community in the Discovery basin, which is one of the most extreme terrestrial saline environments known, as it is almost saturated with MgCl2 (5 M).

Deep hypersaline anoxic basins (hypersaline basins) found in the Eastern Mediterranean Sea (15) probably resulted from the dissolution of subterranean Miocene salt deposits that became exposed to seawater after tectonic activity (1). Brines enclosed in these basins are characterized by anoxic conditions, high pressure (around 35 MPa), and almost saturated salt concentrations (15). The high densities of the hypersaline brines limit mixing with overlaying oxic seawater and result in a sharp chemocline of 1 to 3 m. It is clear from this and previous studies that each hypersaline basin is geochemically distinct (table S1) (1, 310). L'Atalante, Bannock, and Urania brines have similar dominant ion compositions, but the overall salinity of Urania is lower, whereas concentrations of sulfide and methane are considerably higher (table S1). The most striking difference between the geochemistry of Discovery brine compared with the other three is the extremely high concentration of Mg2+ and low concentration of Na+ (table S1). The physical separation of the basins from each other, as well as their existence for thousands of years, may have resulted in the evolution of specific microbial communities in each hypersaline basin. To date, only a few Bacteria, typically found in oxic seawater, have been isolated from the chemocline (1114), but it remains unknown whether they are active in the hypersaline basins. The Discovery Basin contains a brine that has the highest concentration of MgCl2 (∼5 M) found thus far in a marine environment (5, 8); such concentrations are considered anathema to life.

We embarked on a detailed study of four different hypersaline basins: L'Atalante, Bannock, Discovery, and Urania, to investigate their geochemistry, uncultivated microbiota, and in situ microbial metabolic activities. Our aims were to determine the extent to which geochemical conditions influence the evolution of brine communities and to study whether life is possible under the hostile conditions of Discovery brine.

Microbial cells stained with 4′,6-diamidino-2-phenylindole (DAPI) were observed in the four brines with numbers ranging from 1.9 × 104 ml–1 in Discoveryto1.5 × 105 ml–1 in Urania (table S1) (15). The Bacteria/Archaea ratio, based on fluorescence in situ hybridization, indicated that Bacteria dominated the Discovery basin and were slightly more abundant in L'Atalante and Bannock basin, whereas Archaea dominated the Urania basin. In all four hypersaline basins, bacterial diversity was higher than archaeal diversity (table S1), with Urania basin showing lowest overall diversity. All of the hypersaline basins, including Discovery basin, showed evidence of sulfate reduction, methanogenesis, and heterotrophic activity (table S1). Analysis of the 16S ribosomal RNA (rRNA) gene sequence showed that high percentages of clone sequences obtained from the basins belonged to γ-, δ-, and e-Proteobacteria; Sphingobacteria; candidate division KB1 (16); Halobacteria; and a new division which we named MSBL1 [Mediterranean Sea Brine Lakes group 1 (table S1)]. In contrast, seawater above these hypersaline basins showed a very different community structure (table S1). Typical activities of representatives of most of the phylogenetic groups found in the hypersaline basins suggest that they are responsible for the observed sulfate reduction and heterotrophic activity. Thus, deep hypersaline anoxic basins of the Mediterranean are not biological dead-ends but contain active microbial communities that contribute to biogeochemical cycling of carbon and sulfur, as has been observed in other anoxic, highly saline ecosystems (17). Methane was produced in all four hypersaline basins (table S1), but few 16S rRNA gene sequences related to known methanogenic Archaea were found. The majority of the archaeal 16S rRNA gene sequences discovered belonged to a new sub-division that branched deeply within the Euryarchaeota, candidate division MSBL1 (Fig. 1). This candidate division is equivalent in genetic depth and breadth to, for example, Halobacteriales and represents a new order of yet-to-be-cultivated Archaea. On the basis of phylogenetic relatedness of the MSBL1 Archaea to methanogens and the lack of any other group detected that might be responsible for the methane production, it is reasonable to speculate that MSBL1-related Archaea are involved in methanogenesis at high salinity.

Fig. 1.

Phylogenetic tree showing the most dominant archaeal sequences belonging to the new subdivision MSBL1 from the four different deep hypersaline anoxic basins. Trees were constructed with the neighbor-joining method using 100 bootstrap replicates. The values between parentheses show the number of the specific sequence found compared with the total number of archaeal sequences analyzed. Bootstrap values above 50% are shown. Scale bar, 5% of evolutionary distance.

The observations that Discovery brine contains 1.9 × 104 microbial cells ml–1 and that most of the 16S rRNA gene sequences are related to phylogenetic groups not found in normal seawater indicate that a specific microbial community is present. Other environments with elevated concentrations of MgCl2 are the Dead Sea and Lake Bonney in the Antarctic (18, 19), but concentrations do not exceed 2 M. Haloarchaeal species capable of growth at 1 M MgCl2 have been isolated from the Dead Sea (20, 21). However, life has never been observed in, nor have prokaryotes been isolated, that can tolerate 5 M MgCl2, the concentration in Discovery hypersaline brine (table S1).

The bacterial community of the Discovery brine was unequivocally different from that of the Discovery interface and overlying sea-water, as indicated by automated ribosomal intergenic spacer analysis profiles (22). Similar results were obtained when Morisita indices of similarity (23) were calculated for the archaeal and bacterial operational taxonomic unit (OTU) distribution, based on 16S rRNA gene sequence analysis. The similarity indices of bacterial or archaeal OTU distribution between seawater and the seawater-brine interface were 0.001 and 0.01, respectively; these values were 0.48 for Bacteria and 0.89 for Archaea when OTU distributions between interface and brine were compared. Clearly, the seawater communities were different from the communities occurring in the Discovery interface or brine, whereas the brine communities did include microorganisms found at the interface. However, most bacterial sequences and part of the archaeal sequences were unique to the hypersaline Discovery brine. Most striking is the proportional increase, from seawater to brine, of clone sequences similar to Halorhabdus utahensis (99% similarity), an Archaeon that tolerates up to 0.8 M MgCl2 (24). It was absent in the archaeal clone library of seawater, but constituted 11% of the interface clone libraries and, strikingly, 33% of the brine clone libraries (table S1). Ectoenzymatic activities, glutamic acid uptake, sulfate reduction, and methane production rates were measured at in situ temperature in brine and interface samples (table S1). These results show that prokaryotes can be active under the extreme conditions of Discovery basin. The rates of ectoenzymatic activities were higher in Discovery compared with the three NaCl-dominated basins. In contrast, methane production and sulfate reduction rates were lower in Discovery than in the other basins, except for the sulfate reduction rate in Bannock. This might indicate that in Discovery brine heterotrophic prokaryotes are more important than those involved in methanogenesis and sulfate reduction compared with the other brines. This idea is supported by the observation that clone sequences of a phylotype like H. utahensis, a heterotrophic Archaeon capable of fermentative growth, is enriched relative to the MSBL1 group in Discovery when compared with L'Atalante, Bannock, and Urania brines. Results from our study show that life is possible in brines with very high concentrations of divalent cations and shed new light on the existence of microbial life in other types of brines. For example, it has been questioned whether life is possible in Don Juan Pond, an Antarctic brine of 3.2 M CaCl2 (19, 25, 26).

A cluster analysis of the combined archaeal and bacterial 16S rRNA gene data from the four hypersaline basins highlighted the differences between the microbial communities present (Fig. 2, left). The microbial community in Discovery basin is most different from the others, whereas Bannock and L'Atalante are most similar. Cluster analysis of the four basins based on (bio)-geochemical data from table S1 showed a pattern identical to that of the 16S rRNA gene data, indicating that microbial community structure and geochemical conditions are directly linked (Fig. 2, right). The high concentration of MgCl2 in the Discovery basin is likely to influence evolution of the microbial community in its brine, because most halophilic microorganisms studied can grow under high NaCl concentrations but not under high MgCl2 concentrations (27). The elevated concentrations of sulfide and methane in the Urania basin provide valuable sources of energy for many microorganisms, so the higher concentrations of these substrates might have selected a microbial community with high numbers of sulfide- and methane-oxidizing microbes, as well as sulfide-tolerant prokaryotes.

Fig. 2.

UPGMA (unweighted pair group method with arithmetic mean) cluster analysis based on the Morisita index of similarity between the operational taxonomic unit (>97% 16S rRNA gene sequence similarity) distribution of the four deep hypersaline anoxic basins (left), and Euclidean distances between the geochemical data of table S1 for the four hypersaline basins (right).

Evidence that microbial life is possible at 5 M MgCl2 widens the picture of microbial adaptation to salinity. It has been suggested that primordial life on earth started in hyper-saline water (28, 29); furthermore, extra-terrestrial objects are known to contain brines exposed to evaporation, which results in an increase of divalent cations (30, 31). Our results indicate that microbial metabolism can proceed at significant levels in some of the most extreme terrestrial hyper-saline environments and lend further support to the possibility of extraterrestrial life.

Supporting Online Material

www.sciencemag.org/cgi/content/full/307/5706/121/DC1

Materials and Methods

Table S1

References and Notes

References and Notes

View Abstract

Navigate This Article