Paracoccus fistulariae sp. nov., a lipolytic bacterium isolated from bluespotted cornetfish, Fistularia commersonii

The genus Paracoccus was first described by Davis et al. (1969) to accommodate a Gram-negative, catalase- and oxidase-positive, coccoid-shaped bacterium, Paracoccus denitrificans. Subsequently, at least 26 species with validly published names have been proposed as members of the genus Paracoccus (Euzéby, 2009). Phylogenetic analyses based on 16S rRNA gene sequences have shown that the genus Paracoccus falls within the class Alphaproteobacteria (Anzai et al., 2000; Lee et al., 2005). Members of the genus Paracoccus are chemotaxonomically characterized as having ubiquinone-10 as the predominant isoprenoid quinone and C_18 : 1 as the major fatty acid (Kelly et al., 2006; Roh et al., 2009). Early described species of the genus Paracoccus were isolated from various terrestrial environments (Kelly et al., 2006) and some recently described species have been isolated from marine environments, including intertidal sediment and seawater (Khan et al., 2008; Kim et al., 2006; Lee et al., 2004; Liu et al., 2008; Pukall et al., 2003; Roh et al., 2009). In this study, we describe a Paracoccus-like bacterial strain, 22-5^T, which was isolated from the intestine of a bluespotted cornetfish, Fistularia commersonii.

Strain 22-5^T was isolated by means of the standard dilution-plating technique on marine agar 2216 (MA; Difco) at 25 °C. The type strains of Paracoccus homiensis and Paracoccus zeaxanthinifaciens were used as reference strains for phenotypic characterization, fatty acid analysis and DNA–DNA hybridization: P. homiensis KACC 11518^T was obtained from the Korean Agricultural Culture Collection and P. zeaxanthinifaciens KCTC 22688^T was obtained from the Korean Collection for Type Cultures. For DNA and isoprenoid quinone extraction, cultures were prepared in marine broth 2216 (MB; Difco) at 30 °C for 3 days. For fatty acid extraction and morphological, physiological and biochemical characterization, cultures were prepared on MA at 30 °C for 3 days.

Cell morphology was examined by light microscopy (Nikon E600) and transmission electron microscopy. Flagellation was determined by using a Philips CM-20 transmission electron microscope with cells from an exponentially growing culture. For this purpose, the cells were negatively stained with 1 % (w/v) phosphotungstic acid and the grids were examined after being air-dried. The Gram reaction was determined by using a Gram-stain kit (bioMérieux), according to the manufacturer's instructions. Growth at 4, 10, 20, 25, 28, 30, 35, 37, 40 and 45 °C was determined on MA. Growth with 0, 0.5, 1.0, 2.0 and 3.0 % (w/v) NaCl was investigated in trypticase soy broth prepared according to the formula of the Difco medium except for the exclusion of NaCl and inclusion of 0.45 % (w/v) MgCl₂ . 6H₂O or 0.06 % (w/v) KCl. Growth with 2.0–15.0 % (w/v) NaCl (in increments of 1.0 %) was investigated in MB. The pH range for growth was determined in MB adjusted to pH 4.5–9.5 (in increments of 0.5 pH unit) with sodium acetate/acetic acid and Na₂CO₃ buffers. Growth under anaerobic conditions was determined after incubation in a Forma anaerobic chamber on MA and on MA supplemented with potassium nitrate (0.1 %, w/v), both prepared anaerobically under a nitrogen atmosphere. Catalase and oxidase activities were determined as described by Cowan & Steel (1965). Lipolytic activity was investigated on tricaprylin-MA, which was prepared as follows: 5 ml tricaprylin was emulsified in 45 ml gum arabic solution (200 mM NaCl, 10 mM CaCl₂ and 5 %, w/v, gum arabic) for 2 min using a Waring blender and 50 ml tricaprylin emulsion was mixed with 450 ml MA. Hydrolysis of casein, starch, hypoxanthine, tyrosine and xanthine was tested on MA using the substrate concentrations described by Cowan & Steel (1965). Nitrate reduction and hydrolysis of aesculin, gelatin, urea and Tweens 20, 40, 60 and 80 were investigated as described by Lanyí (1987) with the modification that artificial seawater was used for preparation of media. The artificial seawater contained (l⁻¹ distilled water) 23.6 g NaCl, 0.64 g KCl, 4.53 g MgCl₂ . 6H₂O, 5.94 g MgSO₄ . 7H₂O and 1.3 g CaCl₂ . 2H₂O (Bruns et al., 2001). H₂S production was tested as described by Bruns et al. (2001). Acid production from carbohydrates was determined as described by Leifson (1963). Utilization of substrates as sole carbon and energy sources was tested as described by Baumann & Baumann (1981) using supplementation with 2 % (v/v) Hutner's mineral base (Cohen-Bazire et al., 1957) and 1 % (v/v) vitamin solution (Staley, 1968). Susceptibility to antibiotics was tested on MA plates using antibiotic discs containing the following (μg per disc unless otherwise stated): polymyxin B (100 U), streptomycin (50), penicillin G (20 U), chloramphenicol (100), ampicillin (10), cephalothin (30), gentamicin (30), novobiocin (5), tetracycline (30), kanamycin (30), lincomycin (15), oleandomycin (15), neomycin (30) and carbenicillin (100). Other physiological and biochemical tests were performed with the API 20E and API ZYM systems (bioMérieux). The morphological, cultural, physiological and biochemical characteristics of strain 22-5^T are given in the species description and Table 1⇓.

Chromosomal DNA was isolated and purified according to the method described previously (Yoon et al., 1996), with the exception that RNase T1 was used in combination with RNase A to minimize contamination with RNA. The 16S rRNA gene was amplified by PCR using two universal primers as described previously (Yoon et al., 1998). Sequencing of the amplification products and phylogenetic analysis were performed as described previously (Yoon et al., 2003). The identification of phylogenetic neighbours was achieved using the EzTaxon server (; Chun et al., 2007). The almost-complete 16S rRNA gene sequence of strain 22-5^T determined in this study comprised 1383 nt. In the neighbour-joining phylogenetic tree based on 16S rRNA gene sequences, strain 22-5^T fell within the clade comprising species of the genus Paracoccus and formed a coherent cluster with P. homiensis DD-R11^T and P. zeaxanthinifaciens ATCC 21588^T (Fig. 1⇓). The same cluster was found in the trees reconstructed using the maximum-likelihood and maximum-parsimony algorithms (Fig. 1⇓). Strain 22-5^T exhibited 97.4 and 96.6 % 16S rRNA gene sequence similarity with P. homiensis DD-R11^T and P. zeaxanthinifaciens ATCC 21588^T, respectively, and 94.0–96.6 % with the type strains of the other species of the genus Paracoccus. The DNA G+C content was determined by the method of Tamaoka & Komagata (1984) with the modification that DNA was hydrolysed and the resultant nucleotides were analysed by reversed-phase HPLC. The DNA G+C content of strain 22-5^T was 63.6 mol%.

Isoprenoid quinones were extracted according to the method of Komagata & Suzuki (1987) and analysed using reversed-phase HPLC and a YMC ODS-A (250×4.6 mm) column. The predominant isoprenoid quinone detected in strain 22-5^T was ubiquinone-10 (Q-10), which was in line with those for other members of the genus Paracoccus (Kelly et al., 2006; Roh et al., 2009). In this study, P. zeaxanthinifaciens KCTC 22688^T was also found to have Q-10 as the predominant isoprenoid quinone. The fatty acid methyl esters were extracted and prepared according to the standard protocol of the Hewlett Packard Microbial Identification System (MIDI; Sasser, 1990). The fatty acid profile of strain 22-5^T is shown in Table 2⇓ with those of P. homiensis KACC 11518^T and P. zeaxanthinifaciens KCTC 22688^T also analysed in this study. Strain 22-5^T contained large amounts of unsaturated, straight-chain, hydroxy and 11-methyl fatty acids; the major components (>10 % of total fatty acids) were C_18 : 1ω7c and C_18 : 0. The fatty acid profiles of strain 22-5^T, P. homiensis KACC 11518^T and P. zeaxanthinifaciens KCTC 22688^T were essentially similar in that C_18 : 1ω7c was the predominant fatty acid, although there were differences in the proportions of some fatty acids. The results obtained from chemotaxonomic analyses confirmed the result of phylogenetic analysis, i.e. that strain 22-5^T belongs to the genus Paracoccus (Kelly et al., 2006; Roh et al., 2009; Tables 1⇑ and 2⇓).

DNA–DNA hybridization was performed fluorometrically by the method of Ezaki et al. (1989) using photobiotin-labelled DNA probes and microdilution wells. Hybridization was performed with five replications for each sample. The highest and lowest values obtained for each sample were excluded and the means of the remaining three values were quoted as DNA–DNA relatedness values. Strain 22-5^T exhibited 44 and 32 % mean DNA–DNA relatedness with P. homiensis KACC 11518^T and P. zeaxanthinifaciens KCTC 22688^T, respectively, which indicated that it differed genetically from the two reference strains (Wayne et al., 1987).

Strain 22-5^T was distinguishable from the two reference strains by differences in several phenotypic characteristics, most of which were determined in this study using the same conditions and methods (Table 1⇑). The phylogenetic and genetic distinctiveness and the phenotypic properties of strain 22-5^T are sufficient to show that this strain is separate from recognized species of the genus Paracoccus (Stackebrandt & Goebel, 1994). Therefore, strain 22-5^T is considered to represent a novel species of the genus Paracoccus, for which the name Paracoccus fistulariae sp. nov. is proposed.