Halococcus salifodinae BIp^T DSM 8989^T was obtained as a viable isolate from Permian rock salt deposits of a mine in Bad Ischl, Austria [1,2]. The strain grew optimally at a salinity of 20%–25%, a pH value of 7.4 and at 40 °C. Subsequently, several halococcal strains were isolated from similar sites in England and Germany, which had identical 16S rRNA gene sequences and numerous similar properties as the Bad Ischl strain BIp^T [3].

The genus Halococcus [4], emended by Oren et al. [5] currently comprises seven formally described species, which are listed here with their sites of isolation and reference in brackets: Hcc. morrhuae (seawater, saline lakes, salterns and salted products, [6]), Hcc. saccharolyticus (marine salterns, [7]), Hcc. salifodinae (rock salt from mines in Germany and Austria, also from brine in a salt mine in England, [1]), Hcc. dombrowskii (bore core from a salt mine in Austria, [8]), Hcc. hamelinensis (stromatolites of Shark Bay, Hamelin Pool in Western Australia, [9]), Hcc. qingdaonensis (crude sea-salt sample collected near Qingdao in Eastern China, [10]) and Hcc. thailandensis (fermented fish sauce produced in Thailand) [11]. Thus, two species—Hcc. salifodinae and Hcc. dombrowskii - were isolated from Permo-Triassic salt sediments, whereas the other five species can be regarded as inhabitants of hypersaline surface waters or heavily salted products.

A study by Wright [12] using 16S rRNA gene sequences of 61 haloarchaeal taxa, revealed that the mean genetic divergence over all possible pairs of halophilic archaeal 16S rRNA gene sequences was 12.4 ± 0.38%, indicating close relatedness. In comparison, the greatest genetic divergence within methanogenic archaea was 34.2% [12]. Within the halophilic archaea, Halococcus species form an even closer related group (see Figure 1), with 16S rRNA gene sequence similarities of 98.2%–98.7% between Hcc. thailandensis, Hcc. morrhuae, Hcc. qingdaonensis and Hcc. dombrowskii, and somewhat lower similarities of 93.7%–94.1% between Hcc. hamelinensis, Hcc. saccharolyticus and Hcc. salifodinae [11].

Our notions on prokaryotic evolution and evolution in general have been shaped by the concept of a molecular clock, which suggests an approximately uniform rate of molecular evolution among species and duplicated proteins over time [14]. Although subject to various criticisms, molecular-clock techniques still remain the only way to infer the timing of gene duplications and speciation events in the absence of fossil or biogeographical records [14]. The concept was applied previously to date the sequence divergences of halophilic archaeal protein-encoding genes, compared to the divergence of homologous non-halophilic eubacterial protein-encoding genes, assuming a point of haloarchaeal species diversion of 600 million years before present [15]. However, modern results from genome sequences revealed a much more complex history of life than can be depicted in bifurcating trees [16]. Widespread horizontal gene transfer—although occurring to different extents—, endosymbioses, gene losses and other processes cause the presence of different molecules with different histories in a species, and members of the same species were found to differ dramatically in gene content, leading to the suggestion of a fuzzy species concept in prokaryotes [16].

Some of these problems and uncertainties might be resolvable when viable microorganisms from well-dated ancient geological sites would be compared on a molecular basis with contemporary species. A crucial issue is the proof that microorganisms from ancient materials, like million year old deep subseafloor sediments, or Permian salt evaporites, are as old as the geological sites from which they were obtained (see [17,18,19] for discussions). The determination of the age of a single average bacterium is not possible with currently available methods, since its mass is only about a picogram. Thus, claims of ancient microorganisms were often dismissed as being due to laboratory contaminations.

Recently, small particles of about 0.4 μm in diameter were imaged by microscopy directly within fluid inclusions of 22,000–34,000 year old salt bore cores and, following successful culturing, identified as haloarchaea [18,20]. Embedding of halophilic microorganisms in fluid inclusions upon formation of salt crystals is well known, and fluid inclusions have been suggested as sites for preservation of microbial life [21,22,23]. In addition, Gramain et al. [24] reported isolation of haloarchaea from well-dated salt bore cores of Pliocene age (5.3 to 1.8 million years). Thus there is a growing body of evidence that haloarchaea survive for great lengths of time [24].

Here we review the properties of coccoid haloarchaea isolated from Permo-Triassic salt sediments, and relate them to those of halococci, which were isolated from surface waters. In addition, new data on Halococcus salifodinae concerning the chemical composition of its cell wall are included as well as DNA-DNA hybridization experiments between several strains of the species. Recently, the first genome sequence of a halococcus, Hcc. hamelinensis 100A6^T, became available [25] and therefore information for several genes (phaC synthases; subunit A of the rotary A-ATPase) is examined here for their potential use in delineating the evolution of haloarchaeal cocci.

Growth of microorganisms: Halococcus salifodinae strains Blp^T (DSM 8989^T), BG2/2 (DSM 13045), Br3 (DSM 13046), N1 (DSM 13070), H2 (DSM 13071), Hcc. morrhuae DSM 1307^T, Hcc. saccharolyticus DSM 5350^T and Hcc. dombrowskii DSM 14522^T were grown in modified M2 medium as described by Denner et al. [1], which contained (g/l): yeast extract (5.0), casamino acids (5.0), NaCl (200.0), Tris/HCl (12.1), MgCl₂ x 6 H₂O (20.0), CaCl₂ x 2 H₂O (2.0) and KCl (0.2) at pH 7.4. For testing the presence of polyhydroxyalkanoates, strains were also grown in synthetic medium with 1% (w/v) glucose [64], except that KBr was used instead of NaBr. Hcc. hamelinensis JCM 12892^T and Hcc. qingdaonensis JCM 13587^T were grown in complex medium (DSM no. 372, http://www.dsmz.de/microorganisms/medium/pdf/DSMZ_Medium372.pdf). All cultures were incubated at 37 °C with shaking at 180 rpm in an Innova 4080 incubator (New Brunswick Scientific).

Preparation and analysis of cell walls: For preparation of cell walls, cultures were harvested by centrifugation at 8000 rpm for 20 min. The cell pellet was washed three times with deionized water and cells were disrupted with glass beads (Ø 0.5–0.7 mm) in a Braun cell homogenisator (model MKS) for 20 min. The crude cell wall preparation was incubated overnight with trypsin (0.5 mg/ml; Merck) in a 0.05 M (pH 7.8) potassium phosphate buffer at 37 °C. Cell wall preparations were washed four times with deionized water and freeze-dried (Lyovac, Leybold-Heraeus). Neutral sugars and uronic acids were released from cell walls under vacuum with 2 M HCl at 100 °C for 2 h and 3 h, respectively; amino sugars with 4 M HCl at 100 °C for 16 h. After removal of the acid the aqueous uronic acid solution was adjusted to pH 9 with an ammonia solution (1 M) and incubated for 2 h at room tem­perature to delactonize. The compounds were identified and quantified by HPLC using a CarboPac^®PA1 column (Dionex) and pulsed amperometric detection. Monosaccharides and amino sugars were separated with a gradient reaching from 16 to 300 mM NaOH; for separation of uronic acids a gradient consisting of (A) 100 mM NaOH and (B) a solution of 100 mM NaOH with 1 M sodium acetate was used. Acetate [65], sulfate [66] and phosphate [67] were determined by the described methods.

Enzyme tests: The API ZYM system (bioMerieux) was used for the identification of 19 enzymatic activities [68]. Test strips were inoculated with cells in the exponential phase of growth and were incubated at 37–39 °C for up to 24 h [8]. Standard tests (oxidase, catalase, nitrate reduction, gelatin liquefaction, hydrolysis of starch, casein and Tween) were performed as described previously [8] or as recommended by Oren et al. [69]. Tests were performed at least in triplicate.

DNA-DNA hybridization. DNA was isolated as described by Cashion et al. [70]. Levels of DNA-DNA hybridization were determined spectrophotometrically by the renaturation method of De Ley et al. [71], with the modifications by [72] and [73]. Renaturation rates were computed by the program TRANSFER.BAS [74]. These experiments were carried out by the Identification Service of the DSMZ, Braunschweig, Germany.

Other methods: For comparative phylogenetic analyses 16S rRNA gene sequences from validly described Halococcus spp. were obtained from the European Molecular Biology Laboratory (EMBL) web interface and fitted in a subset of aligned archaeal sequences obtained from the Ribosomal Database Project II [75]. Phylogenetic relationships of the sequences were constructed by using distance-matrix methods (corrections as in Jukes and Cantor [76]). The web-based software MEGA 2 (http://www.megasoftware.net; [77]) and Clustal X [78] were used for sequence analysis and for construction of the phylogenetic tree, including maximum-likelihood and maximum parsimony methods. Confidence of the branching patterns was assessed by bootstrap analysis (1000 replicates). Scanning and transmission electron microscopy was performed as described previously [1,3].

From considerations of close genetic relatedness and origin from ancient geological materials it is concluded that Halococcus salifodinae strains can be considered as living fossils and constitute a promising source of novel evolutionary information.