Isolation and Culture Conditions
The marine sediment samples were collected from the natural saline-alkali wetland soil of Binhai new district, Tianjin, China (38°46′N, 117°13′E), and were transferred to the laboratory with ice. Then they were purified into a single colony on the modified R2A (MR2A) medium using the traditional dilution inoculation method, for 1 liter medium containing the following ingredients: 0.5 g casamino acids (Difo), 0.5 g of yeast extract (Difo), 0.5 g of sodium pyruvate, 0.5 g of peptone, 0.5 g of glucose, 3.0 g of trisodium citrate, 2.0 g of KCl, 0.3 g of K2HPO4, 0.5 g of CaCl2, 20.0 g of MgSO4∙7H2O, 180 g of NaCl, pH 7.0-7.5, agar 15.0 g, 121 ºC, autoclaving 20 min (Reasoner and Geldreich 1985). The purified strain was preserved in cryotubes at -80 ºC by adding 250 μL glycerol/SW 30 (80:20, v/v) to 750 μL fresh culture (OD600 0.8-1.0) (Dyall-Smith, 2009) for further characterization. Halorubrum saccharovorum JCM 8865T, Halorubrum persicum JCM 30541T and Halorubrum kocurii JCM 14978T were incubated on MR2A medium with 18.0 % (w/v) NaCl and used as the reference strains for all experiments.
Morphological, Physiological and Biochemical Characterization
Cell size, morphology, and motility of strain WN019T were established by using a Leica microscope equipped with phase–contrast optics (Leica DM 6000 B) during exponential growth phase. Cell morphology was also assessed by transmission electron microscopy (TEM), i.e., cells were harvested from exponentially growing culture, and the cells were negatively stained with 0.5 % uranyl acetate and the grids were examined at the microscope (Tecnai Spirit, FEI, Hillsboro, OR, USA). Gram staining was performed using BD Gram staining kits according to the manufacturer’s instructions. Oxidase activity was tested using the oxidase reagent kit (bioMérieux) according to the manufacturer’s instructions. Catalase activity was determined by pouring a 3.0 % H2O2 solution onto bacterial colonies and observing bubble production. The optimal growth temperature of strain WN019T was determined after incubation on MR2A medium at 4, 10, 15, 20, 25, 30, 33, 37, 40, 45, and 50 °C (at pH 7.5). NaCl tolerance was tested in MR2A medium amended with 0.0-25.0 % NaCl (w/v) at intervals of 1.0 %. The pH range for growth was measured by adjusting the final pH value to 4.0-13.0 at intervals of 0.5 (at 18.0 % NaCl, w/v, 37 °C) with the appropriate buffers (Na2HPO4/NaH2PO4 for pH 5.0-7.0 and Na2CO3/NaHCO3 for pH 8.0-12.0). Bacterial growth was measured as increase in turbidity at 600 nm, using a DU 800 spectrophotometer (Beckman Coulter). Anaerobic growth was determined through measuring the OD600 nm at 37 °C with 18.0 % NaCl (w/v) in the tubes with the butyl rubber stopper and screw cap.
For all physiological experiments, we selected Halorubrum saccharovorum JCM 8865T, Halorubrum persicum JCM 30541T, and the closely related strain Halorubrum kocurii JCM 14978T as reference organisms. Unless otherwise stated, all the strains mentioned above were incubated at 37 °C in MR2A medium amended with final concentration of 18.0 % NaCl for strain WN019T and final concentration of 18.0 % NaCl for reference organisms.
Biochemical activities and use of organic substrates as sole carbon and energy sources were evaluated using API 20E, API 20NE kits (bioMérieux) and Biolog GENIII MicroPlates (bioMérieux) according to the manufacturer’s instructions, except for a salinity adjustment to 18.0 % (w/v) NaCl. This involved supplementing with (l-1) 2.0 g of KCl, 0.3 g of K2HPO4, 0.5 g of CaCl2 and 20.0 g of MgSO4∙7H2O (pH between 7.0-7.5). Susceptibility to antibiotics was assessed on MR2A (18.0 % NaCl) using the disc–diffusion plate method (Fraser and Jorgensen 1997) with discs containing ampicillin (10 μg), chloramphenicol (30 μg), erythromycin (15 μg), penicillin G (10 μg), streptomycin (10 μg), vancomycin (30 μg), gentamicin (10 μg), polymyxin B (30 μg), neomycin (30 μg), ciprofloxacin (5 μg), nalidixic acid (30 μg), aphidicolin (20 μg), norfloxacin (10 μg), nitrofurantoin (300 μg), trimethoprim (5 μg), mycostatin (100 μg), novobiocin (30 μg), and bacitracin (0.04 IU per disc). Antibiotics–containing discs were placed on MR2A (18.0% NaCl) plate surfaces, and the bacterial cultures (200 μL) that were spread on the plate were checked for clearing zones after 3 d at 37 °C.
To characterize respiratory quinones and polar lipids of strain WN019T and the reference strains, cells were grown in medium mentioned above (37 ℃) and harvested during late exponential growth phase. Respiratory quinones were extracted with chloroform/methanol (2:1) (v/v) from lyophilized cells (300 mg) and purified using high performance liquid chromatography (HPLC) (Minnikin DE 1984). Analysis of respiratory quinones were carried out by the Identification Service, DSMZ, Braunschweig. Germany. Polar lipids were extracted from 200 mg of freeze-dried cell material using a chloroform/methanol/0.3 % (w/v) aqueous NaCl mixture with the ratio of 1:2:0.8 (v/v/v), modified after Bligh and Dyer (1959), recovered into the chloroform phase by adjusting the mixture to a ratio of 1:1:0.9 (v/v/v), and separated by two-dimensional silica gel thin-layer chromatography (Qingdao. Haiyang Chemical Co.; silica gel GF254, 0.25-mm thick, China). The first dimension was developed in a chloroform/methanol/water (65:25:4, v/v/v) mixture and the second was in a chloroform/methanol/acetic acid/water (80:12:15:4, v/v/v/v) mixture. Total lipid materials were detected using molybdatophosphoric acid and specific functional groups were detected using spray reagents specific for defined functional groups. Polar lipid analysis was performed by the Identification Service, DSMZ, Braunschweig, Germany.
Genomic DNA was extracted using a commercial kit (TaKaRa MiniBEST Bacteria Genomic DNA Extraction Kit Ver. 3.0) based on the manufacture’s protocol. The amplification of the 16S rRNA fragment was done by PCR with 27F and 1492R as primers and genomic DNA as template. The genome was sequenced on the Illumina HiSeq2000 platform at Shanghai Personal Biotechnology Co., Ltd. China. Filtering and trimming of the genomic raw data were done with PRINSEQ v0.20.4 (Schmieder and Edwards 2011), and the trimmed reads were assembled using SOAPdenovo v.2.3 (Li et al. 2008, 2010) with default parameters. The genome completeness (100 %) was assessed using CheckM (version 1.03) (Parks et al. 2015). Protein-coding open reading frames were predicted by Glimmer (version 3.02) (Delcher et al. 2007). For RNA prediction, rRNAs were predicted by RNAmmer (version 1.2) (Lagesen et al. 2007), and tRNAs were predicted by tRNAscan-SE (version 1.21) (Lowe and Eddy 1997). The 16S rRNA gene sequence and the genome of strain WN019T was submitted to GenBank (https://www.ncbi.nlm.nih.gov/). Multiple sequence alignments of strain WN019T and the most closely related taxa were carried out using CLUSTALXv1.81 (Thompson et al. 1997). Phylogenetic trees were constructed with the maximum-likelihood (ML) method using MEGA v7.0 (Kumar et al. 2016), and neighbour-joining (NJ) and maximum parsimony (MP) phylogenetic trees were also constructed to confirm the phylogenetic position of the strain WN019T. The resultant tree topologies were evaluated by bootstrap analyses (1000 replications). Average nucleotide identity (ANI) values of the total genomic sequences shared between the genomic sequences of strain WN019T and closely related genomic sequences from GenBank were performed using the ANI-BLAST (ANIb) and ANI-MUMmer (ANIm) algorithms in JSpeciesWS (https ://jspecies.ribohost.com/jspeciesws/) (Richter et al. 2016). As a proposed complement to ANI values, digital DNA–DNA hybridization (dDDH) values were calculated using Genome-to-Genome Distance Calculator (GGDC2.1) (Meier-Kolthoff et al. 2013) using the BLAST+ method. Results were recommended based on the recommended formula 2 (identities/ HSP length), which was useful when dealing with incomplete draft genomes.