Sphingopyxis marina sp. nov. and Sphingopyxis litoris sp. nov., isolated from seawater

Members of the genus Sphingopyxis are Gram-negative, non-fermentative, aerobic, non-spore-forming, yellow-pigmented or whitish-brown, motile and contain sphingoglycolipids. Takeuchi et al. (2001) proposed that Sphingomonas species should be divided into four different genera, namely Sphingobium, Novosphingobium, Sphingopyxis and Sphingomonas. The nomenclatural system of Takeuchi et al. (2001) has been generally accepted and used (Yabuuchi et al., 2002; Busse et al., 2003; Godoy et al., 2003; Yoon et al., 2005). In this study, we isolated two bacterial strains, designated FR1087^T and FR1093^T, from seawater and subjected them to a polyphasic analysis. The strains clearly differed from recognized Sphingopyxis species and from each other and therefore represent two novel species of the genus Sphingopyxis.

Strains FR1087^T and FR1093^T were isolated from a sample of coastal surface seawater collected off Jeju Island, Republic of Korea. The seawater sample was diluted with sterilized artificial seawater (Lyman & Fleming, 1940), spread onto a plate that contained marine agar 2216 (MA; Difco) and incubated at 30 °C. The isolates were routinely cultured on MA and maintained as a glycerol suspension (20 %, w/v) at −80 °C.

The 16S rRNA gene was amplified from a single colony by means of PCR with Taq polymerase (Takara). The PCR amplification and sequencing were performed as described previously (Chun & Goodfellow, 1995). The pairwise sequence similarity values were determined using the EzTaxon server (Chun et al., 2007). The regions available for all sequences (positions 47–1437; Escherichia coli numbering system) were used to construct the phylogenetic trees after the exclusion of ambiguous positions (positions 67–84). Phylogenetic analyses were performed by following the same procedure as that described previously by Yi et al. (2005), using jPHYDIT (Jeon et al., 2005; available at ) and paup 4.0 (Swofford, 1998). Phylogenetic trees were inferred using the maximum-likelihood (Felsenstein, 1981), maximum-parsimony (Fitch, 1971) and neighbour-joining (Saitou & Nei, 1987) methods. The resultant tree topologies were evaluated by means of bootstrap analyses (Felsenstein, 1985) based on 1000 resamplings.

Sequence comparisons with 16S rRNA gene sequences held in GenBank indicated that the two isolates were closely related to the genus Sphingopyxis. Strains FR1087^T and FR1093^T had a sequence similarity of 98.5 %. Sphingopyxis flavimaris showed the highest level of 16S rRNA gene sequence similarity with respect to strains FR1087^T and FR1093^T (97.9 % for both). Out of five signature nucleotides defined for the genus Sphingopyxis (Takeuchi et al., 2001), three were found in the two isolates. They, along with Sphingopyxis flavimaris, lacked the remaining two signature nucleotides (Table 1⇓).

The levels of sequence similarity between strain FR1087^T and other Sphingopyxis species were in the range 92.8–93.7 % and the corresponding values for strain FR1093^T were 91.8–93.1 %. The neighbour-joining tree showed that the two isolates formed a monophyletic clade with the type strain of Sphingopyxis flavimaris (Fig. 1⇓), with 100 % bootstrap support. This relationship was confirmed by all other tree-inferring methods used in this study.

The close taxonomic relationship, i.e. the high levels of 16S rRNA gene sequence similarity among strains FR1087^T and FR1093^T and Sphingopyxis flavimaris SW-151^T, led us to employ DNA–DNA relatedness experiments to obtain further genomic comparisons (Stackebrandt & Goebel, 1994). Genomic relatedness was determined using a Cary 300 Bio model UV/VIS spectrophotometer equipped with a Peltier-thermostatted 6×6 multicell changer and a temperature controller (Varian). The experimental details for the DNA–DNA hybridization studies were as described elsewhere (Yi & Chun, 2006). The level of genomic relatedness observed between Sphingopyxis flavimaris and strain FR1087^T was 21.2 %, while that between Sphingopyxis flavimaris and strain FR1093^T was 36.3 %. The two isolates shared a low DNA–DNA relatedness value (20.7 %). All of these values are below the threshold (70 %) for determining bacterial species (Wayne et al., 1987); the finding strongly suggested that the two isolates belonged to novel genomic species in the genus Sphingopyxis.

Growth under anaerobic conditions was checked in an anaerobic chamber on MA and on nitrate-supplemented MA, both of which had been prepared anaerobically under nitrogen gas. Growth at various concentrations of NaCl and sea salts was investigated using sea-salt-free ZoBell's agar [ZoBell, 1941; 15 g Bacto agar (Difco), 5 g Bacto peptone (Difco), 1 g yeast extract, 0.1 g ferric citrate in 1 l distilled water]. The temperature range for growth was determined optically using a temperature gradient incubator (TVS 126MA; Advantec) with marine broth 2216 (Difco) and temperatures of 4–45 °C (specifically at 4.0, 10.0, 15.1, 18.3, 21.2, 23.7, 26.5, 29.0, 31.8, 34.4, 37.3, 40.7 and 45 °C). Biochemical tests were performed using API 20NE and API ZYM kits (bioMérieux): the strips were each inoculated with a heavy bacterial suspension in artificial seawater or AUX medium (bioMérieux) supplemented with 2 % (w/v) sea salts. Hydrolysis of casein, starch, Tween 80, hypoxanthine, tyrosine and xanthine was tested on MA using the substrate concentrations described by Cowan & Steel (1965). The two isolates showed differences in terms of the hydrolysis of gelatin and the utilization of glucose. Both isolates utilized urea, unlike other phylogenetic neighbours. The results of the biochemical and physiological tests are given in the species description and are shown in Table 1⇑.

Cellular fatty acids from the test strains were analysed by GLC according to the instructions of the Microbial Identification System (MIDI). For analysis of the cellular fatty acid methyl esters, cells were grown on MA at 30 °C for 6 days. The two strains differed substantially in terms of their major fatty acids: details are given in the species description and are shown in Table 1⇑. Polar lipids were extracted using the procedures described by Minnikin et al. (1984) and were identified by using two-dimensional TLC followed by spraying with the appropriate detection reagents (Komagata & Suzuki, 1987). Sphingoglycolipid was found as a major polar lipid in both isolates. Polyamines were extracted and analysed according to the methods of Busse et al. (1989) and Flores & Galston (1982). Strain FR1087^T contained spermidine [8.6 μmol (g dry wt)⁻¹] and agmatine [4 μmol (g dry wt)⁻¹] as major components, whereas strain FR1093^T contained only spermidine [17.78 μmol (g dry wt)⁻¹] as the major component.

On the basis of the data from the polyphasic study presented here, it is evident that strains FR1087^T and FR1093^T represent separate novel species in the genus Sphingopyxis, for which the names Sphingopyxis marina sp. nov. and Sphingopyxis litoris sp. nov., respectively, are proposed.