Rhodanobacter soli sp. nov., isolated from soil of a ginseng field

The genus Rhodanobacter (family Xanthomonadaceae, order Xanthomonadales) was first proposed by Nalin et al. (1999) with the description of Rhodanobacter lindaniclasticus. However, the type strain of R. lindaniclasticus, LMG 18385^T, no longer exists (Mergaert et al., 2002). Members of the genus Rhodanobacter have been isolated from various habitats, but mainly from soil. At the time of writing, the genus Rhodanobacter includes a further six physiologically versatile species with validly published names: Rhodanobacter fulvus (Im et al., 2004), Rhodanobacter thiooxydans (Lee et al., 2007), Rhodanobacter spathiphylli (De Clercq et al., 2006), Rhodanobacter ginsengisoli, Rhodanobacter terrae (Weon et al., 2007) and Rhodanobacter ginsenosidimutans (An et al., 2009).

By using many different culture methods, we have attempted to isolate micro-organisms from soil in order to investigate community relations. In this study, one of these strains, DCY45^T, was characterized by a polyphasic approach, including phylogenetic analysis based on the 16S rRNA gene sequence and determination of genomic relatedness and chemotaxonomic, morphological and physiological properties, in order to define its precise taxonomic position. These results obtained in this study indicated that strain DCY45^T is a member of the genus Rhodanobacter and is clearly distinguished from Rhodanobacter species with validly published names.

Soil from a ginseng field was shaken vigorously in pure water. Strain DCY45^T was isolated via direct smearing of the stepwise solution on 10-fold-diluted R2A agar (Difco). Single colonies chosen on these plates were picked up and transferred onto new plates for isolation. After 3 days of incubation at 30 °C, the purified colony was identified by using partial 16S rRNA gene sequences and culturing was continued on R2A agar (Difco) for routine work; the strain was stored as a 25 % (w/v) glycerol suspension at −70 °C.

Cell morphology and motility were observed with a Nikon light microscope (×1000 magnification) with cells grown on R2A agar for 2 days at 30 °C. The Gram reaction was examined according to the staining method, as described by Bartholomew & Finkelstein (1958). After growth on an R2A plate at 30 °C for 24 h, cells were resuspended in a drop of deionized water and placed on a carbon- and Formvar-coated nickel grid for 30 s. Grids were floated on a drop of 0.1 % (w/v) aqueous uranyl acetate, blotted dry and then observed with an electron microscope (Carl Zeiss LEO912AB) at 100 kV under standard conditions (Supplementary Fig. S1, available in IJSEM Online). Oxidase activity was evaluated by the oxidation of 1 % p-aminodimethylaniline oxalate. Catalase activity was determined by observation of bubble production after adding a drop of 3 % (v/v) H₂O₂ solution. Growth at 4, 10, 20, 25, 30, 37 and 42 °C in R2A broth and at pH 4.0–10.0 (in increments of 0.5 pH units) in Luria–Bertani (LB) broth was assessed after 7 days of incubation; the pH was adjusted with 1 M NaOH or HCl. Salt tolerance was tested on R2A supplemented with 0–10 % (w/v) NaCl after 7 days of incubation. Degradation of DNA was detected on DNase test agar (Scharlau) by flooding with 1 M HCl. Hydrolysis of skimmed milk, starch, pectin and CM-cellulose (Kasana et al., 2008) was assessed as described by Atlas (1993) and evaluated after 5 days. Growth on nutrient agar, trypticase soy agar, R2A agar and LB agar (all from BD) was also evaluated at 30 °C. Utilization of substrates and enzyme activities were determined by using the API 20NE, API ZYM and API ID 32GN microtest systems according to the recommendations of the manufacturer (bioMérieux). Anaerobic growth was checked by using the BD GasPak EZ Gas Generating system for 10 days. The sachets provided gas generation (CO₂) for this system. H₂S production was examined on triple-sugar-iron agar (BBL).

Isoprenoid quinones were extracted with chloroform/methanol (2 : 1, v/v), concentrated with a vacuum rotary evaporator at 50 °C and extracted with 10 ml hexane. After using Sep-Pak for purification, the crude solution was analysed by HPLC as described by Hiraishi et al. (1996). Quantitative analysis of whole-cell fatty acids was performed after growth on tryptic soy agar (Difco) for 2 days at 28 °C. Fatty acid methyl esters were prepared, separated and identified with the Sherlock Microbial Identification System (MIDI, Inc.) (Sasser, 1990); results of two separate repeated experiments were collected.

For G+C content analysis, genomic DNA of strain DCY45^T was extracted and purified with the Qiagen Genomic-tip system 100/G. The DNA was then degraded enzymically into nucleosides by using P1 nuclease and the G+C content was determined by the method of Mesbah et al. (1989).

Genomic DNA of strain DCY45^T for 16S rRNA gene sequencing was extracted and purified with the Genomic DNA isolation kit (Core Bio System). The 16S rRNA gene was amplified from chromosomal DNA using the universal bacterial primer set 9F/1512R (Weisburg et al., 1991) and the purified PCR products were sequenced by Genotech (Daejeon, Korea) (Kim et al., 2005). The partial sequence of the 16S rRNA gene was compiled with SeqMan software and edited using the BioEdit program (Hall, 1999). Multiple alignments were performed with the clustal_x program (Thompson et al., 1997) and Kimura's two-parameter model (Kimura, 1983). The neighbour-joining method was used to calculate evolutionary distances (Saitou & Nei, 1987) and a maximum-parsimony tree was constructed using the mega 4 program (Kumar et al., 2001). Bootstrap analysis with 1000 replicates was conducted in order to obtain confidence levels for the branches (Felsenstein, 1985). Type strains of all species of Rhodanobacter and closely related type strains were included in the phylogenetic tree.

Strain DCY45^T was cultured on R2A agar (Difco) at 30 °C for 2 days, yielding brownish-yellow, circular colonies. Cells of strain DCY45^T were Gram-negative, aerobic, motile rods. Strain DCY45^T was able to grow at 20–37 °C. Physiological and biological characteristics of strain DCY45^T are summarized in the species description and a comparison of selective characteristics with related type strains is shown in Table 1⇓.

The cellular fatty acid profiles of strain DCY45^T and the related type strains R. fulvus IAM 15025^T, R. thiooxydans DSM 18863^T, R. spathiphylli LMG 23181^T, R. terrae DSM 19241^T and R. ginsenosidimutans KCTC 22231^T are shown in Supplementary Table S1. As mentioned above, the type strain of the type species of the genus is no longer extant and could not be included in our studies. The major cellular fatty acids in strain DCY45^T were iso-C_16 : 0 (15.9 %), iso-C_17 : 1ω9c (23.8 %) and iso-C_15 : 0 (15.9 %). The amounts of iso-C_16 : 0 in R. fulvus IAM 15025^T, R. thiooxydans DSM 18863^T, R. spathiphylli LMG 23181^T, R. terrae DSM 19241^T and R. ginsenosidimutans KCTC 22231^T were 2.9, 8.1, 6.6, 19.8 and 1.8 % and the amounts of iso-C_15 : 0 in these strains were 18.3, 18.2, 25.8, 8.5 and 23.1 %, respectively. These data show considerable differences in fatty acid profiles between strain DCY45^T and other Rhodanobacter type strains (Supplementary Table S1). Strain DCY45^T contained ubiquinone with eight isoprenoid units (Q-8) as the predominant isoprenoid quinone; Q-8 has been reported in all members of the genus Rhodanobacter for which the quinone composition has been determined (Table 1⇑). The G+C content of genomic DNA of strain DCY45^T was 65.4±0.1 mol%.

The 16S rRNA gene sequence of strain DCY45^T was found to be a continuous stretch of 1475 nt. Strain DCY45^T was determined to belong to the class Gammaproteobacteria, order Xanthomonadales and family Xanthomonadaceae. The closest phylogenetic neighbours of strain DCY45^T were R. thiooxydans LCS2^T (98.4 % 16S rRNA gene sequence similarity), R. fulvus Jip2^T (98.4 %), R. spathiphylli B39^T (97.9 %), R. terrae GP18-1^T (98.0 %), R. ginsenosidimutans CSC17Ta-90^T (97.7 %), R. ginsengisoli GR17-7^T (96.4 %) and R. lindaniclasticus RP5557^T (96.0 %). In the phylogenetic tree (Fig. 1⇓), strain DCY45^T clearly belonged to the lineage of the family Xanthomonadaceae, as evidenced by the high bootstrap value.

Levels of DNA–DNA relatedness were determined from duplicate experiments with DNA of the novel strain and its five closest phylogenetic relatives. Strain DCY45^T, R. fulvus IAM 15025^T and R. thiooxydans DSM 18863^T were used as probes for reciprocal tests. Hybridization was measured as fluorescence intensity at 37 °C every 15 min until these values were over 10 000, and each point was read twice. Nucleic acid probes prepared with photobiotin label were detected using a microplate reader (Ezaki et al., 1989); each sample was placed in three duplicate wells. Values of DNA–DNA hybridization (mean±sd, n=3) between DCY45^T and R. fulvus IAM 15025^T, R. thiooxydans DSM 18863^T, R. terrae DSM 19241^T, R. spathiphylli LMG 23181^T and R. ginsenosidimutans KCTC 22231^T were 58.9±2.8, 40.2±2.1, 35.0±2.0, 31.5±3.9 and 18.8±1.5 %, respectively (Supplementary Table S2). These values are below the 70 % cut-off point recommended for species delineation (Wayne et al., 1987).

Our results support the assignment of strain DCY45^T to a separate, previously unrecognized species within the genus Rhodanobacter, for which the name Rhodanobacter soli sp. nov. is proposed.