Strain and culture condition
The sample was collected from coastal waters, near the town Chongwu in Southeast China (118.545685 °E, 24.53178 °N). Sample was brought to the lab and stored at 4 oC (refrigerator) for up to 2 days before being processed. Strain HICWT was isolated and purified over five times using the seawater double-layer agar plating method as previously described with Vibrio alginolyticus LF TCBS 15 (=MCCC 1K03520) as the prey bacterium (Schoeffield and Williams 1990; Ye et al. 2019). Cells from the plaque were examined by light microscopy (CX22RFS1, OLYMPUS) with 1% (w/v) crystal violet staining. A single-plaque (with slender rod cells of strain HICWT and a few residual prey cells) on the double-layer plate was picked into a 20 ml tube-type bottle with 2 ml 1/40 (v/v) marine broth 2216E (MB, peptone 5 g, yeast extract 1 g, seawater 1 L, pH 7.2–7.6) and incubated at 28 °C with 200 r.p.m for 2–3 days. Small amounts of cells (each around 1×105–6 cell ml–1) of strain HICWT and prey strain LF TCBS 15 were detected in 1/40 MB culture. The axenic independent strain HICWT was purified from the 1/40 MB co-culture by the standard dilution plating on marine agar 2216E (MA, pH 7.2–7.6), after incubation at 28 °C for 6-7 days. A yellow colony different from the prey strain was picked, checked by light microscopy (CX22RFS1, OLYMPUS) with 1% (w/v) crystal violet staining, and then purified by streaking three times on MA. The strain was maintained in MB at 28 °C for 24 h and preserved in MB supplemented with 20% (v/v) glycerol at -20 °C and -80 °C. For long-term storage, the cultures of strain HICWT were lyophilized in 10% (w/v) skim milk, and then deposited at the Marine Culture Collection of China (MCCC) and Japan Collection of Microorganisms (JCM). M. aquimarina JCM11811T and M. ruestringensis DSM13258T were obtained from Marine Culture Collection of China and cultivated as reference strains under identical conditions.
Phenotypic and biochemical characterisation
Cell morphology were observed by light microscopy (CX22RFS1, OLYMPUS) and transmission electron microscopy (TEM, HT7800, Hitachi). For negative stains, cells from 24 h cultures on MA were resuspended with 0.1 mol/L phosphate buffer (pH7.4), then a 400-mesh grid was inverted over a drop of cell suspensions for 1 min. The grid was then washed on 2 drops of water and the cells were stained with 2.0% (w/v) uranyl acetate for 10 s.
For predatory characteristic detection, strain HICWT was cultivated either in double-layer agar plate or in seawater with washed prey cells (around 1×109 cell ml–1) (Ye et al. 2019). Light microscopy (CX22RFS1, OLYMPUS) and transmission electron microscopy (TEM, HT7800, Hitachi) were used to assess the cell-to-cell contact with attachment to the prey. For negative stains, a 400-mesh grid was inverted over a drop of 24 h co-cultures seawater, washed on 1 drop of water, and the cells were stained as mentioned above.
Gram staining was performed using a Gram Stain kit (QingDao Hopebio-Technology Co., Ltd) according to the instructions of the manufacturer. Growth temperature and pH values were assessed in MB at different temperatures (15, 20, 25, 28, 30, 35, 37, 40 and 45 oC) and pH values (3.0–10.0 at 1.0 unit intervals) for 42 h. MB with different pH values were prepared using the following biological buffers: citrate/phosphate (pH 3.0–7.0), Tris/HCl (pH 7.0–9.0) and sodium carbonate/sodium bicarbonate (pH 9.0–10.0). Growth at various NaCl concentrations (0, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0 and 10.0% (w/v)) was investigated using MB with different NaCl concentrations prepared in the laboratory according to the formula of the medium. Gliding motility was examined on MA with 1.0% agar (w/v) and using the hanging-drop method after growing cells in MB broth for 48 h at 28°C (Bernardet et al. 2002). The aerobic condition of the strain HICWT was determined by a semi-solid (0.5%) stab culture (Dong and Cai 2001). The presence of flexirubin-type pigments were investigated as described by Reichenbach (1989) and Bernardet (2002). An acetone: methanol (7:2, v/v) mixture was used to extract bacterial carotenoids pigments, and the whole-cell spectrum for carotenoid pigments was detected using the UV–visible spectrophotometer (AOE Instruments (Shanghai) Co., Ltd.) according to Hameed et al. (2011). Oxidase reagent (Hangzhou Microbial Reagent Co., Ltd) was used for testing oxidase activity, and catalase activity was evaluated by the production of oxygen bubbles in 3% (v/v) aqueous hydrogen peroxide solution according to the method of Dong & Cai (2001). Additional enzymic activities and biochemical features of strain HICWT and the reference strains were determined by the API ZYM and API 20NE kits (bioMérieux) according to the manufacturer’s instructions, except that inocula were prepared by suspending cells in artificial seawater (Yang et al. 2013). Antibiotic susceptibility tests were performed using 6 mm filter-paper discs (Hangzhou Microbial Reagent Co., Ltd) with antibiotics added at the following concentrations (µg per disc unless stated otherwise): penicillin (10 U), erythromycin (15), neomycin (30), gentamicin (10), tetracycline (30), doxycycline (30), minocycline (30), kanamycin (30), amikacin (30), oxacillin (1), ampicillin (10), carbenicillin (100), piperacillin (100), cefradine (30), cephalexin (30), cefazoline (30), cefuroxime (30), ceftazidime (30), ceftriaxone (30), cefoperazone (75).
Chemotaxonomic characterization
For analysis of the cell fatty acids, bacteria were cultured in MB at 28 oC for 24 h, the harvested cells were saponified, methylated and extracted using the standard MIDI (Sherlock Microbial Identification System, version 6.0B) protocol. The whole-cell fatty acid pattern were then analysed by gas chromatography (model 6850, Agilent Technologies) and identified using the TSBA6.0 database of the Microbial Identification System (Athalye et al. 1985; Sasser 1990). Polar lipids were extracted and examined using two-dimensional thin-layer chromatography according to Kates (1972). Isoprenoid quinones were extracted from freeze-dried cells with chloroform/methanol (2:1, v/v) and analysed by reversed-phase HPLC (Collins et al. 1984).
Phylogenetic and genomic analyses
The genomic DNA of strain HICWT was extracted using a Rapid Bacterial Genomic DNA Isolation Kit (B518225, Shanghai Sangon Biological Engineering Technology & Services Co., Ltd). The whole genome of strain HICWT was sequenced by the Guangdong Magigene Biotechnology Co., Ltd., using Solexa paired-end (150 bp library) sequencing technology protocol. SPAdes software (http://cab.spbu.ru/software/spades/) was used to do genome assembly with multiple-Kmer parameters (Bankevich et al. 2012). The draft genome data of HICWT has been deposited in GenBank with the accession number WYET00000000. The G+C contents of the genomic DNA were calculated from the sequenced genome (https://www.ezbiocloud.net/tools/ani). Open Reading Frames (ORFs) were predicted using Prodigal v2.6.3 (Hyatt et al. 2010) and the predicted protein coding sequences (CDS) were searched against the GenBank, Clusters of Orthologous Groups (COGs) and KEGG databases to analyse gene functions and metabolic pathways.
The partial 16S rRNA gene (around 1400 bp) was respectively amplified from the chromosomal DNA and plaques of the strain HICWT on the double-layer agar plate with V. alginolyticus LF TCBS 15 as the prey cells. The universal bacterial primers 27F and 1492R (Delong 1992) and the purified PCR product was sequenced by Xiamen Bioray Biotechnology Co., Ltd. The complete 16S rRNA gene sequence of the strain HICWT was obtained from its draft genome sequence. The 16S rRNA gene sequence analyses were carried out with the online tool EzBioCloud (http://eztaxon-e. ezbiocloud.net) (Kim et al. 2012). 16S rRNA gene sequences of related taxa were selected from the GenBank database. Phylogenetic trees were reconstructed using the MEGA software package version 7.0 (Kumar et al. 2016) with distance options according to the default parameter model and clustering with the neighbor-joining (NJ), maximum-likelihood (ML) and maximum-parsimony (MP) methods, supported using bootstrap values with 1000 replications.
The whole genome average nucleotide identity (gANI) was calculated using the algorithm as described by Yoon et al. (2017) with the web service of EzBioCloud (https://www.ezbiocloud.net/tools/ani) and the digital DNA-DNA hybridizations (dDDH) were determined online at http://ggdc.dsmz.de/ggdc.php# using the Genome-to-Genome Distance Calculation (GGDC) version 2.1 (Meier-Kolthoff et al. 2013). Genomic data of related species were downloaded from GenBank database.