Introduction

High salinity is toxic to most cells but many extremely halophilic archaea, i.e. haloarchaea, have been isolated in hypersaline environments (Grant 2004). Well known hypersaline environments for the isolation of haloarchaea are soda lakes, salt lakes and solar salterns, but some haloarchaea have been isolated from a salt-fermented seafood made from shrimp (Roh et al. 2007a, b, 2009; Roh and Bae 2009). All haloarchaea are classified within the family Halobacteriaceae in the order Halobacteriales of the phylum Euryarchaeota. This family currently contains 40 genera based on the List of Prokaryotic Names with Standing in Nomenclature (Euzéby 1997). In particular, the genus Halorubrum, which was proposed formally by McGenity and Grant (1995), currently includes 25 validly named species: Halorubrum (Hrr.) aidingense (Cui et al. 2006), Hrr. alkaliphilum (Feng et al. 2005), Hrr. aquaticum (Gutiérrez et al. 2011), Hrr. arcis (Xu et al. 2007), Hrr. californiense (Pesenti et al. 2008), Hrr. chaoviator (Mancinelli et al. 2009), Hrr. cibi (Roh and Bae 2009), Hrr. coriense (Oren and Ventosa 1996), Hrr. distributum (Oren and Ventosa 1996), Hrr. ejinorense (Castillo et al. 2007), Hrr. ezzenmoulense (Kharroub et al. 2006), Hrr. kocurii (Gutiérrez et al. 2008), Hrr. lacusprofundi (McGenity and Grant 1995), Hrr. lipolyticum (Cui et al. 2006), Hrr. litoreum (Cui et al. 2007), Hrr. luteum (Hu et al. 2008), Hrr. orientale (Castillo et al. 2006), Hrr. saccharovorum (Tomlinson and Hochstein 1976; McGenity and Grant 1995), Hrr. sodomense (McGenity and Grant 1995), Hrr. tebenquichense (Lizama et al. 2002), Hrr. terrestre (Ventosa et al. 2004), Hrr. tibetense (Fan et al. 2004), Hrr. trapanicum (McGenity and Grant 1995), Hrr. vacuolatum (Mwatha and Grant 1993; Kamekura et al. 1997) and Hrr. xinjiangense (Feng et al. 2004). Cells of the genus Halorubrum are rods or irregular-shaped, motile or non-motile, strictly aerobic and oxidase and catalase-positive. The major polar lipids are phosphatidylglycerol (PG), phosphatidylglycerol phosphate methyl ester (PGP-Me), phosphatidylglycerol sulfate (PGS) and/or sulfated mannosyl glucosyl diether (S-DGD-3) (McGenity and Grant 1995). The G+C content of the genomic DNA is in the range of 61.7–71.2 mol% (McGenity and Grant 2001; Roh and Bae 2009). The present study characterized a new haloarchaeon, strain B8T and determined the taxonomic position of this strain based on phenotypic, phylogenetic and chemotaxonomic analyses, according to the proposed minimal standards for the description of new taxa in the order Halobacteriales (Oren et al. 1997). Based on this polyphasic taxonomic study, strain B8T is considered to represent a new species in the genus Halorubrum, for which the name Hrr. halophilum sp. nov. is proposed here.

Materials and methods

Archaeal strain and culture conditions

A strain, designated B8T, was isolated from a salt-fermented seafood made from shrimp. A sample was serially diluted and spread onto a complex medium (DSM medium no. 954), which was adjusted to pH 7.0 and contained the following (l−1): 5 g casamino acids (BD), 5 g yeast extract (BD), 20 g MgCl2·6H2O, 2 g KCl, 12 g Tris, 0.2 g CaCl2·2H2O, and 200 g NaCl. A solid medium was prepared by adding 2 % (w/v) agar. A single colony was streaked repeatedly to obtain a pure culture at 37 °C. For long-term preservation, strain B8T was frozen at −80 °C in medium 954 supplemented with 5 % (v/v) dimethyl sulfoxide (DMSO). Hrr. lipolyticum JCM 13559T, Hrr. kocurii JCM 14978T, Hrr. saccharovorum DSM 1137T, Hrr. lacusprofundi DSM 5036T Hrr. aidingense JCM 13560T, Hrr. arcis JCM 13916T and Hrr. aquaticum JCM 14031T were obtained from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ) or Japan Collection of Microorganisms (JCM) and used in the comparative taxonomic analyses.

Morphological, physiological and biochemical characterization

All of the phenotypic tests were performed using medium 954 or a halophile medium (HMD) that contained (l−1): 20 g MgCl2·6H2O, 5 g K2SO4, 0.1 g CaCl2·2H2O, 0.1 g yeast extract, 0.5 g NH4Cl, 0.05 g KH2PO4, 0.5 g casamino acids as a carbon source and 180 g NaCl (Savage et al. 2007), unless indicated otherwise. The cell morphology and size were determined using a transmission electron microscope (SUPRA 55 VP; Carl Zeiss). Gram staining was performed using the published method for haloarchaea (Dussault 1955). Cell lysis in distilled water was tested by incubating cells in distilled water for 1 week before transfer to medium 954. The optimal conditions for growth in medium 954 with various NaCl concentrations were tested using 0–30 % (w/v) NaCl at intervals of 5 %. The optimal pH range for growth was assayed from pH 5.0–11.0 at intervals of 1.0 using HMD with the following buffers: 10 mM 2-(N-morpholino) ethanesulfonic acid for pH 5.0 and 6.0, 10 mM bis–Tris propane for pH 7.0–9.0 and 10 mM 3-(cyclohexylamino)-1-propanesulfonic acid for pH 10.0 and 11.0. Growth was also tested at different temperatures, i.e. 5–60 °C at intervals of 5 °C. The standard phenotypic tests for nitrate and nitrite reduction in aerobic conditions, indole formation, urease activity and the hydrolysis of casein and starch were performed as described by Benson (2002) using medium 954 as the basal medium. The hydrolysis of Tween 20, 40 and 80 were tested as described by González et al. (1978) and the hydrolysis of gelatin was tested according to Smibert and Krieg (1994) using medium 954 as the basal medium. Anaerobic growth was tested in filled, stoppered tubes using medium 954 in the presence of 30 mM nitrate, 5 g l-arginine, 5 g DMSO or 5 g trimethylamine N-oxide (TMAO) at 37 °C in an anaerobic chamber (Coy), where the atmosphere comprised N2/CO2/H2 (90:5:5, by vol.). H2S formation was tested according to Cui et al. (2007). Acid production from d-glucose was tested by growing strain B8T in HMD supplemented with 1 % d-glucose. Methyl red and Voges–Proskauer tests were determined using MR-VP Broth (BD) and Simmon’s citrate test was performed using Simmons citrate agar (DB), for which each medium was supplemented with 20 % (w/v) NaCl. Arginine, lysine and ornithine decarboxylases tests were carried out using Moeller Decarboxylase Broth Base (BD) containing 20 % (w/v) NaCl as the basal medium. To assess the utilization of sole carbon and energy sources, HMD was supplemented with 10 mM bis–Tris propane and 1 % of the following substrates: acetate, l-alanine, l-arginine, L-aspartate, citrate, d-fructose, fumarate, d-galactose, d-glucose, l-glutamate, glycerol, glycine, dl-lactate, lactose, l-lysine, l-malate, maltose, mannitol, d-mannose, l-ornithine, pyruvate, d-ribose, sorbitol, l-sorbose, starch, succinate, sucrose or d-xylose. For testing antibiotic sensitivity, strain B8T was inoculated on agar medium plates using antibiotic discs with the following amounts (μg per disc, unless indicated): ampicillin (20), bacitracin (0.1 IU), chloramphenicol (50), ciprofloxacin (10), erythromycin (25), neomycin (50), norfloxacin (20), novobiocin (50), penicillin G (20 IU) and rifampin (10). The strain was incubated for 2 weeks at 37 °C.

Determination of the 16S rRNA and RNA polymerase subunit B′ (rpoB′) gene sequences and phylogenetic analysis

To obtain chromosomal DNA, cells of strain B8T were harvested, extracted and purified using a DNA extraction kit (G-spin™; iNtRON Biotechnology). The 16S rRNA gene of strain B8T was amplified using a PCR pre-Mix (iNtRON Biotechnology) and the primer set Arch21F and 1492R (DeLong 1992), as described previously (Roh et al. 2008). To check the heterogeneous 16S rRNA gene sequences, PCR products were ligated and transformed using an All-in PCR cloning kit (BioFact) according to the manufacturer’s protocol. Multiple clones were picked randomly and then sequenced. PCR-mediated amplification and sequencing of the rpoB′ genes were carried out according to Minegishi et al. (2010). The almost full-length 16S rRNA and rpoB′ gene sequences were determined using SeqMan (DNASTAR). The phylogenetic neighbours and the pairwise sequence similarities were determined using EzTaxon-e (Kim et al. 2012). The 16S rRNA gene sequences of strain B8T and validly named related species were aligned using the SILVA Incremental Aligner (Pruesse et al. 2012). The rpoB′ gene sequences of the related taxa were obtained from GenBank (http://www.ncbi.nlm.nih.gov) and multiple sequence alignment was performed using the Clustal_W program. The phylogenetic trees were constructed using the neighbor-joining (NJ) (Saitou and Nei 1987), minimum-evolution (ME) (Nei et al. 1998) and maximum-likelihood (ML) (Felsenstein 1981) algorithms based on Kimura’s two-parameter model (Kimura 1980) with 1,000 randomly selected bootstrap replicates using MEGA5 (Tamura et al. 2011).

Determination of the DNA–DNA hybridization (DDH), DNA G+C content, quinones and polar lipid analysis

The DDH experiments were performed using the fluorometric method with photobiotin-labeled DNA probes and microwell plates (MaxiSorp, FluoroNunc), as described by Ezaki et al. (1989). The genomic DNA G+C content was determined as described by González and Saiz-Jimenez (2002).

Polar lipids were extracted and detected using thin layer chromatography on a silica gel 60 F254 plate (Merck), according to the method of Minnikin et al. (1984). The compositions of the polar lipid spots were determined by spraying each plate with specific detection reagents as follows: sulfuric acid–ethanol (1:2, by vol) for total lipids, molybdenum blue for phospholipids and α-naphthol-sulfuric acid for glycolipids. The quinones of strain B8T were analyzed using an HPLC system (UltiMate 3000; Dionex) coupled to a diode array detector and a single quadrupole mass spectrometer (HCT Ion-Trap MS; Bruker).

Results and discussion

Cells of strain B8T were observed to be Gram-stain negative, pleomorphic, short rod or oval shaped, 0.5–0.6 μm in width and 0.9–1.1 μm in length (Supplementary Fig. S1). The colonies were red with a smooth and rounded shape on solid medium. Strain B8T was found to grow in the presence of 15–30 % (w/v) NaCl, at 20–45 °C and at pH 7.0–9.0, with optimum growth in the presence of 25 % NaCl, at 35–40 °C and pH 8.0. The cells of strain B8T lysed in distilled water. Strain B8T was found to reduce nitrate in aerobic conditions and to be positive for acid production from d-glucose, catalase and oxidase. The novel strain was found to be able to hydrolyze Tween 20, 40 and 80, but not casein, gelatin and starch. H2S was not produced from Na2SO3. Indole was not formed and no urease activity was observed. Strain B8T could not grow in anaerobic conditions using nitrate, l-arginine, DMSO or TMAO. The Methyl red, Voges-prokauer and Simmon’s citrate tests were negative. Strain B8T did not produce arginine dihydrolase, lysine decarboxylase or ornithine decarboxylase. Strain B8T was found to be sensitive to novobiocin, bacitracin, erythromycin, rifampin, neomycin and ciprofloxacin, but resistant to penicillin G, ampicillin, chloramphenicol and norfloxacin. Detailed results of the phenotypic tests and the nutritional features of this strain are presented in the species description. Table 1 shows the different characteristics of strain B8T compared with those of closely related type strains of members of the genus Halorubrum, which indicates that strain B8T can be distinguished from closely related members of the genus Halorubrum.

Table 1 Differential characteristics of strain B8T and closely related species in the genus Halorubrum
Full size table

Sequence comparisons indicated that strain B8T has three rRNA genes, rrnA, rrnB and rrnC. The almost complete 16S rRNA of the three rRNA genes and the rpoB′ gene sequences of strain B8T were obtained (1,430, 1,450, 1,450 and 1,830 bp, respectively; GenBank accession numbers EF077637, KF848218, KF848217 and KF700332, respectively). Similarities between the three 16S rRNA gene sequences were 99.0–99.8 %. A comparison with related sequences showed that strain B8T (based on the 16S rRNA rrnA gene sequence) shares the highest levels of similarity with the following strains with validly published names: Hrr. lipolyticum JCM 13559T (99.0 % 16S rRNA gene sequence similarity), Hrr. saccharovorum DSM 1137T (99.0 %), Hrr. kocurii JCM 14978T (98.8 %), Hrr. lacusprofundi DSM 5036T (98.3 %), Hrr. arcis JCM 13916T (98.0 %), Hrr. aidingense JCM 13560T (97.7 %), Hrr. aquaticum JCM 14031T (97.0 %) and other type strains in the genus Halorubrum (93.7–96.5 %). Strain B8T clustered with Hrr. lipolyticum, Hrr. saccharovorum Hrr. kocurii, Hrr. lacusprofundi and Hrr. aidingense in the phylogenetic trees based on the 16S rRNA gene sequences (Fig. 1). The phylogenetic analyses based on the rpoB′ gene sequences showed that strain B8T is closely related to the following members: Hrr. kocurii JCM 14978T (97.2 % rpoB′ gene sequence similarity), Hrr. saccharovorum DSM 1137T (96.8 %), Hrr. lipolyticum JCM 13559T (96.3 %), Hrr. lacusprofundi DSM 5036T (94.9 %), Hrr. arcis JCM 13916T (94.2 %), Hrr. aidingense JCM 13560T (93.2 %) and Hrr. aquaticum JCM 14031T (92.2 %). The phylogenetic trees based on the rpoB′ gene sequences showed a similar topology compared with trees of the 16S rRNA gene (Fig. 1b).

Fig. 1
figure 1figure 1

Phylogenetic tree based on the neighbour-joining (NJ) algorithm for the 16S rRNA (a) and rpoB′ (b) gene sequences of strain B8T and closely related taxa. The numbers on the nodes indicate the bootstrap values (>70 %), which were calculated using the NJ/minimum-evolution (ME)/maximum-likelihood (ML) probabilities. The closed circles represent the nodes obtained using both the ME and ML methods, whereas the open circles indicate nodes recovered using either the ME or ML method. Methanococcus vannielii DSM 1224T and Haloquadratum walsbyi DSM 16790 were used as the outgroup for the phylogenetic trees based on the 16S rRNA and rpoB′ gene sequences, respectively. Bar, 0.02 (a) and 0.05 (b) accumulated changes per nucleotide

Full size image

The DDH values of strain B8T with the type strains of Hrr. lipolyticum JCM 13559T, Hrr. saccharovorum DSM 1137T, Hrr. kocurii JCM 14978T, Hrr. lacusprofundi DSM 5036T, Hrr. arcis JCM 13916T, Hrr. aidingense JCM 13560T and Hrr. aquaticum JCM 14031T were 50 ± 9, 48 ± 11, 23 ± 2, 17 ± 3, 23 ± 4, 12 ± 2 and 14 ± 6 %, respectively. DDH values of <70 % represent distinct species according to current cut-offs employed in prokaryotic systematics (Wayne et al. 1987; Stackebrandt and Goebel 1994), so strain B8T can be considered a distinct genospecies in the genus Halorubrum. The DNA G+C content of strain B8T was determined to be 64.6 mol% (T m), which is similar to the levels found in the closely related strain Hrr. lipolyticum JCM 13559T (65.9 mol%) (Cui et al. 2006). The polar lipids detected in strain B8T comprised PG, PGP-Me, PGS, S-DGD-3 and an unidentified phospholipid (Supplementary Fig. S2). The major polar lipid profile of strain B8T resembles that of the closely related Halorubrum species Hrr. lipolyticum, Hrr. saccharovorum, Hrr. lacusprofundi and Hrr. arcis which contain PG, PGP-Me, PGS and S-DGD-3 (McGenity and Grant 1995; Cui et al. 2006; Xu et al. 2007). However, the presence of PGS and S-DGD-3 in strain B8T help distinguish the strain from Hrr. kocurii and Hrr. aidingense, respectively (Cui et al. 2006; Gutiérrez et al. 2008). The major isoprenoid quinone detected in strain B8T was menaquinone (MK)-8 and minor ones were MK-7(H4) and MK-8(H2).

In conclusion, the results of the phenotypic, phylogenetic and chemotaxonomic analyses showed that the haloarchaeal strain B8T belongs to the genus Halorubrum. Table 1 shows that strain B8T exhibits some differences compared with closely related type strains in the genus Halorubrum. Thus, based on this polyphasic taxonomic study, strain B8T is considered to represent a novel species in the genus Halorubrum, for which the name Hrr. halophilum nov. is proposed.

Description of Halorubrum halophilum sp. nov.

Halorubrum halophilum (ha.lo’phi.lum. Gr. n. hals, halos, salt; Gr. adj. philos, loving; N.L. neut. adj. halophilum salt-loving)

Cells are Gram-stain negative, pleomorphic, short rod or oval shaped and 0.5–0.6 μm in width and 0.9–1.1 μm in length. Colonies are red, smooth and round in shape. Cell lysis occurs in distilled water. Growth occurs in the presence of 15–30 % (w/v) NaCl (optimum, 25 %), at 20–45 °C (optimum, 35–40 °C) and pH 7.0–9.0 (optimum, pH 8.0). Anaerobic growth does not occur in the presence of nitrate, l-arginine, DMSO or TMAO. Cells are positive for nitrate reduction in aerobic conditions, catalase and oxidase activity, acid production from d-glucose and the hydrolysis of Tween 20, 40 and 80, but negative for nitrite reduction in aerobic conditions, indole formation, H2S production, urease activity, Methyl red, Voges–Proskauer, Simmon’s citrate and the hydrolysis of casein, gelatin or starch. Arginine dihydrolase, lysine decarboxylase and ornithine decarboxylase are not produced. Acetate, l-arginine, fumarate, d-galactose, d-glucose, l-glutamate, dl-lactate, d-mannose, l-ornithine, pyruvate and succinate are utilized as sole carbon and energy sources, whereas l-alanine, l-aspartate, citrate, d-fructose, glycerol, glycine, lactose, l-lysine, l-malate, maltose, mannitol, d-ribose, sorbitol, l-sorbose, starch, sucrose and d-xylose are not utilised. The polar lipids are phosphatidylglycerol, phosphatidylglycerol phosphate methyl ester, phosphatidylglycerol sulfate, sulfated mannosyl glucosyl diether and an unidentified phospholipid. The major isoprenoid quinone is MK-8. The DNA G+C content of the type strain is 64.6 mol%.

The type strain, B8T (=JCM 18963T = CECT 8278T), was isolated from a salt-fermented seafood made from shrimp. The GenBank/EMBL/DDBJ accession number for the 16S rRNA rrnA, rrnB and rrnC and rpoB′ gene sequences of strain B8T are EF077637, KF848218, KF848217 and KF700332, respectively.