Novosphingobium acidiphilum sp. nov., an acidophilic salt-sensitive bacterium isolated from the humic acid-rich Lake Grosse Fuchskuhle

The genus Novosphingobium was proposed as a result of the dissection of the genus Sphingomonas (Yabuuchi et al., 1990) based on phylogenetic and chemotaxonomic analysis with Novosphingobium capsulatum as the type species (Takeuchi et al., 2001). The genus Novosphingobium currently comprises 14 recognized species. Members of the genus Novosphingobium are rather heterogeneous and originate from a variety of habitats including soil, coastal or freshwater sediments (Balkwill et al., 1997; Sohn et al., 2004; Liu et al., 2005; Suzuki & Hiraishi 2007), activated sludge/wastewater treatment plants (Neef et al., 1999; Fujii et al., 2003) and a contaminated groundwater bioremediation reactor (Tiirola et al., 2002). In 16S rRNA gene clone libraries of subsurface water samples of the acidic and humic acid-rich Lake Grosse Fuchskuhle (Mecklenburg-Brandenburg lake district, north-eastern Germany), a 16S rRNA gene sequence frequently appeared that was affiliated to the genus Novosphingobium (Glöckner et al., 2000; Burkert et al., 2003; Allgaier & Grossart, 2006). In contrast to most of the recognized species of this genus, 21 bp were lacking in the 16S rRNA gene sequence of this uncultured Novosphingobium species. The same 21 bp gap has recently been found in the 16S rRNA gene sequence of Novosphingobium nitrogenifigens isolated from a paper mill in New Zealand (Addison et al., 2007).

During field studies in July 2006, a yellow- to orange-pigmented bacterium (strain FSW06-204d^T) was isolated from subsurface water samples of the south-west (SW) basin of Lake Grosse Fuchskuhle. At the time of isolation, the pH of the SW basin was 4.7. The 16S rRNA gene sequence of strain FSW06-204d^T showed 99 % similarity to the cloned 16S rRNA gene sequences obtained from the SW basin of Lake Grosse Fuchskuhle (Glöckner et al., 2000; Burkert et al., 2003; Allgaier & Grossart, 2006). Strain FSW06-204d^T was enriched in a most-probable-number (MPN) dilution series of water samples performed in microtitre plates using synthetic freshwater (Bartscht et al., 1999) containing 10 μM NH₄Cl, 10 μM KH₂PO₄, 100 μM KNO₃, 200 μM MgSO₄ . 7H₂O, 100 μM CaCl₂ . 2H₂O and 250 μM CaCO₃, 1 ml l⁻¹ of SL 10 trace elements (Tschech & Pfennig, 1984) and 10 ml l⁻¹ vitamin solution (Balch et al., 1979) buffered with 10 mM HEPES (pH 7.2) and supplemented with 0.001 % (v/v) Tween 20, a mixture of six amino acids (l-cysteine, l-alanine, l-isoleucine, l-leucine, l-asparagine and l-arginine, each 200 μM), a fatty acid mixture of formate, acetate and propionate (each 200 μM), as well as with glucose, pyruvate, citrate, 2-oxoglutarate and succinate (each 200 μM) (Jaspers et al., 2001). After 8 weeks incubation at room temperature in the dark, the novel strain was isolated from one of the highest positive dilution steps by streaking subsamples on YPG agar plates [1.2 % agar supplemented with a 0.22 μm filter-sterilized concentrate of yeast extract, peptone and glucose with final concentrations of 0.075, 0.15 and 0.075 % (w/v), respectively]. After 4 weeks incubation at room temperature in the dark, orange-pigmented colonies were visible. Purification of single colonies was performed in YPG top agar dilution series (0.8 % agar) and the strain was preserved at −80 °C in 7 % DMSO.

Strain FSW06-204d^T formed small yellow colonies (<0.5 mm) with a smooth surface after 48 h at 25 °C on YPG agar plates. Colonies turned orange after a few more days of incubation. Cell morphology was investigated by phase-contrast microscopy (Fig. 1a⇓) for cells grown in K7 medium (0.1 % yeast extract, peptone and glucose) adjusted to pH 5.5 at 25 °C. During exponential growth, the rod-shaped cells of strain FSW06-204d^T were 1.4±0.4 μm long and 0.6±0.1 μm wide. Cells were slightly elongated in the stationary phase (1.6±0.3 and 0.6±0.1 μm), motile in the early exponential phase and of shorter appearance when grown in K7 medium without glucose. Electron microscopy of ultrathin cell sections stained with 1 % uranyl acetate showed a typical Gram-negative cell wall and no intracytoplasmic membrane systems (Fig. 1b and c⇓). In some cells, single electron-dense polar granules were observed which were similar to those observed for Sandarakinorhabdus limnophila so42^T, also isolated from a freshwater lake (Gich & Overmann, 2006). Polyhydroxyalkanoate granules were not observed in cells of the novel strain in contrast to observations of N. nitrogenifigens DSM 19370^T (Addison et al., 2007). Cells of strain FSW06-204d^T grown on YPG agar plates adjusted to pH 5.5 stained Gram-negative and were positive for cytochrome oxidase in a delayed reaction as determined by using an oxidase test (Merck). Since gas was not produced after the addition of 3 % H₂O₂, strain FSW06-204d^T was deemed to be catalase-negative. Endospores were not detected after staining with malachite green and capsules were lacking as determined by contrasting a bacterial suspension with china ink. Obligate aerobic to microaerophilic growth was indicated by growth at the surface and in a zone approximately 1 cm below the surface of YPG soft agar tubes (0.16 % agar, pH 5.5).

Temperature, salinity and pH-dependent growth were examined in K7 medium buffered with 10 mM filter-sterilized MES. The pH values of the K7 medium and MES buffer were adjusted using KOH and HCl. Growth experiments were performed in a total volume of 6 ml in 25 ml screw cap tubes under continuous shaking at 220 r.p.m. resulting in an oxygen concentration of 4–6 mg l⁻¹ as measured with an oxygen sensor (OxyScan Oxygen sensor 501, Aquatech). Optical density was monitored at 600 nm and all experiments were performed in duplicate and repeated twice. In MES-buffered K7 medium, growth occurred between pH 4.5 and 7.0, but not at pH 3.5 or 7.5. Highest growth rates were observed at pH 5.5 and 6.0. On YPG agar plates, growth was observed at between 4 and 32 °C, but not at 37 or 42 °C. In liquid medium, a minimum doubling time of 5.3 h was determined at 25 °C. Salinity-dependent growth was examined by the addition of NaCl (w/v) in 0.25 % (w/v) steps. Highest growth rates were observed when no or 0.25 % (w/v) NaCl were added, but decreased in the presence of 0.5 % (w/v) NaCl and growth was not observed after the addition of 0.75 % (w/v) NaCl. In summary, the optimal growth conditions for strain FSW06-204d^T were pH 5.5 to 6.0, 25 °C and with no addition of NaCl.

For phylogenetic characterization, chromosomal DNA was extracted as described by Pitcher et al. (1989) and a 1376 bp fragment of the 16S rRNA gene was amplified using primers targeting conserved regions of the gene, 8F (8–27, Escherichia coli numbering, Brosius et al., 1981) and 1492R (1492–1510) (Lane, 1991). Sequencing of the DNA fragment was performed with primers 8F, 1492R and two additional reverse primers, 907R (907–926) and 1100R (1100–1115) (Lane, 1991), using the Big Dye terminator cycle sequencing reaction kit and an ABI Prism 310 Genetic Analyzer (Applied Biosystems). DNA sequences were processed in mega (molecular evolutionary genetics analysis) version 3 software (Kumar et al., 2004) and phylogenetic analysis was performed with arb software (Ludwig et al., 2004). A pairwise distance matrix was calculated and phylogenetic trees were constructed using the neighbour-joining and maximum-parsimony methods including bootstrap values based on 1000 resampled datasets. Because strain FSW06-204d^T lacked a 21 bp sequence between positions 1256 and 1278 (E. coli numbering, Brosius et al., 1981), the respective 21 bp were manually removed from the sequences of the reference strains included in the calculations. Similarity values and phylogenetic trees were calculated with gaps included and after excluding gap positions.

A comparison of the 16S rRNA gene sequences of strain FSW06-204d^T and the type strains of all the recognized species of the genus Novosphingobium revealed that the novel strain showed the highest similarity to Novosphingobium hassiacum (GenBank accession no. AJ416411, 96.7 %), followed by N. capsulatum (D16147, 96.6 %), Novosphingobium taihuense (AY500142, 96.5 %) and N. nitrogenifigens (DQ448852, 96.4 %). The 16S rRNA gene sequence similarity to other species of the genus Novosphingobium was in a range from 94.2 to 96.4 %. All signature nucleotides characteristic for the genus Novosphingobium as reported by Takeuchi et al. (2001) were present in the 16S rRNA gene sequence of strain FSW06-204d^T. A 21 bp gap already described for N. nitrogenifigens (Addison et al., 2007) was present in the 16S rRNA gene sequence of strain FSW06-204d^T between base positions 1256 and 1278 (E. coli numbering, Brosius et al., 1981). This 21 bp gap was not present in other recognized species of the family Sphingomonadaceae but occurred frequently in other members of the class Alphaproteobacteria, for example, in Filomicrobium fusiforme DSM 5304^T and the closely related Rhodobium marinum DSM 2698^T, but not in other closely related species (Rainey et al., 1998). In contrast, the gap seemed to be a common feature only within the family Rhodobacteraceae, as observed by searching within the arb database (Ludwig et al., 2004). The presence of this 21 bp signature gap distinguished strain FSW06-204d^T and N. nitrogenifigens DSM 19370^T from other species of the genus Novosphingobium. These two strains formed a distinct cluster within the currently recognized members of the genus Novosphingobium as shown by the calculation of phylogenetic trees when the gap region was excluded from the analysis (Fig. 2⇓ and Supplementary Fig S1 available in IJSEM Online). When the region of the 21 bp gap was included in the same calculations, no differences were revealed in the phylogenetic relationships of strain FSW06-204d^T. The gene sequence similarity values of strain FSW06-204d^T to other Novosphingobium species were only slightly different in this analysis. For N. hassiacum (GenBank accession no. AJ416411), similarity values were 97.2 % instead of 96.7 % and for Novosphingobium stygium (U20775), 95.6 % instead of 95.5 %. Therefore, the cluster formed by strains FSW06-204d^T and N. nitrogenifigens DSM 19370^T was independent of the 21 bp gap.

DNA–DNA hybridization experiments with strains FSW06-204d^T and N. nitrogenifigens DSM 19370^T were performed as described previously (Kämpfer et al., 2002) and revealed a low DNA–DNA relatedness between the strains. Cross-hybridization of genomic DNA from FSW06-204d^T to N. nitrogenifigens DSM 19370^T yielded a relatedness value of 39 %, whereas the reciprocal hybridization gave a lower value of 24 %.

Further characterization of strain FSW06-204d^T was first performed using a substrate assimilation panel and enzyme tests with chromogenic substrates [p-nitrophenyl- (pNP) and p-nitroanilide- (pNA) linked substrates] (Kämpfer et al., 1991). In contrast to the inoculation procedure described by Kämpfer et al. (1991), strain FSW06-204d^T was not inoculated after resuspension in 0.9 % (w/v) NaCl because of its sensitivity to salt. Instead, a culture of strain FSW06-204d^T was resuspended in MES buffer (pH 5.5) to give a final concentration of 50 mM MES for incubation. The pH of the phosphate buffer was adjusted to the pH range for optimum growth for FSW06-204d^T. Microtitre plates were incubated for 7 to 14 days at room temperature in the dark. In contrast to all other species of the genus Novosphingobium examined using these tests (Table 1⇓, data from Kämpfer et al., 2002; Tiirola et al., 2005), growth of strain FSW06-204d^T and also of N. nitrogenifigens DSM 19370^T was only observed in the media containing the chromophoric substrates and was not observed in the substrate assimilation media (Kämpfer et al., 1991). Therefore substrate assimilation tests for strains FSW06-204d^T, N. nitrogenifigens DSM 19370^T and the closest related type strain, N. hassiacum W-51^T, were performed with a selection of substrates in twofold diluted, glucose-free K7 medium with 10 mM MES (pH 5.5). Substrates were added to a final concentration of 5 mM. Tests were performed in duplicate using two independent pre-cultures for inoculation. Substrate utilization was considered positive if the OD₆₀₀ value of the substrate-amended cultures increased by more than 0.05 OD units in comparison with the control cultures grown in parallel without the addition of a carbon substrate. The results for the substrate assimilation tests as compared with those from the literature for the recognized species of the genus Novosphingobium are presented in Table 1⇓. The fact that strains FSW06-204d^T and N. nitrogenifigens DSM 19370^T were not able to grow on the assimilation medium meant that these strains could be differentiated from other species of the genus Novosphingobium on the basis of their growth requirements. The results of the substrate assimilation tests for N. hassiacum W-51^T (Table 1⇓) were identical overall when the tests were performed with either our method or with the method of Kämpfer et al. (1991). Our results show that strain FSW06-204d^T can be distinguished from N. nitrogenifigens DSM 19370^T and N. hassiacum W-51^T by 12 and 14 traits, respectively, of 31 tested characteristics listed in Table 1⇓. Strain FSW06-204d^T can be separated from all other species of the genus Novosphingobium by several characteristics when compared with the data from the literature (Table 1⇓).

Fatty acids were analysed from biomass that was grown for 72 h on K7 agar plates at 22 °C (strain FSW06-204d^T and N. hassiacum W-51^T) and 28 °C (N. nitrogenifigens DSM 19370^T) as described by Kämpfer & Kroppenstedt (1996). The predominant fatty acids of strain FSW06-204d^T were C_18 : 1ω7c (43.3 %) and C_17 : 1ω6c (17.6 %). Two major 2-hydroxy fatty acids, C_14 : 0 2-OH (8.1 %) and C_15 : 0 2-OH (6.5 %), were also present, but no 3-hydroxy fatty acids were detected (Table 2⇓). The genus Novosphingobium was distinguished from the genera Sphingomonas sensu stricto, Sphingopyxis and Sphingobium by the absence of other major 2-hydroxy fatty acids in addition to C_14 : 0 2-OH (Takeuchi et al., 2001). However, strain FSW06-204d^T contained rather similar amounts of fatty acids C_14 : 0 2-OH and C_15 : 0 2-OH. This profile has also been shown for N. hassiacum W-51^T (Kämpfer et al., 2002). The detailed fatty acid profile for strain FSW06-204d^T is shown in Table 2⇓ and shows the affiliation of strain FSW06-204d^T to the genus Novosphingobium (Takeuchi et al., 2001) and the differences observed between this strain and the closest related type strains, N. hassiacum W-51^T and N. nitrogenifigens DSM 19370^T.

Biomass for polyamine, quinone and polar lipid analyses was grown in K7 medium adjusted to pH 4.5. Polyamines were analysed as described by Busse & Auling (1988) and Stolz et al. (2007). Respiratory quinones and polar lipids were determined according to Tindall (1990a, b) and Altenburger et al. (1996). The polyamine pattern of strain FSW06-204d^T, with the predominant compound spermidine (31.8 μmol g⁻¹ dry weight) and small amounts of putrescine (0.1 μmol g⁻¹ dry weight) and spermine (0.2 μmol g⁻¹ dry weight), was in accordance with those of all other species of the genus Novosphingobium so far examined for this trait (Busse et al., 1999; Takeuchi et al., 2001; Kämpfer et al., 2002; Tiirola et al., 2005). Strain FSW06-204d^T exhibited a quinone system consisting of ubiquinone Q-10 (91 %) and Q-9 (9 %), which was in accordance with the characteristics of the genus (Busse et al., 1999; Takeuchi et al., 2001; Kämpfer et al., 2002; Tiirola et al., 2005). The polar lipid profile consisted of the major compounds phosphatidyldimethylethanolamine and phosphatidylethanolamine, moderate amounts of a highly hydrophobic yellow pigment stainable with α-naphthol, moderate amounts of diphosphatidylglycerol, phosphatidylmonomethylethanolamine, phosphatidylglycerol, sphingoglycolipid, an unknown highly hydrophilic aminophosphoglycolipid, two unknown polar lipids and minor to trace amounts of one unknown glycolipid, two unknown phospholipids, two unknown polar lipids and another highly hydrophobic yellow pigment (see Supplementary Fig. S2 in IJSEM Online). Overall, this profile showed a high similarity to those of other Novosphingobium species but characteristic features of strain FSW06-204d^T were the lack of phosphatidylcholine and a moderate content of sphingoglycolipids.

Pigment analysis was performed with cells of strain FSW06-204d^T that were in the exponential growth phase in K7 medium adjusted to pH 5.5 at 25 °C. Absorbance spectra of acetone/methanol (7/2, v/v) extracts were characterized by λ_max at 453 and 480 nm and cell-free extracts were characterized by λ_max at 432/433, 458 and 489/490 nm, indicating the presence of carotenoids (see Supplementary Fig. S3a in IJSEM Online). Similar absorbance spectra have been reported for other species of the genera Sphingomonas and Novosphingobium (Busse et al., 1999, 2003). In order to characterize the pigment composition of strain FSW06-204d^T in comparison to other species of the genera Novosphingobium, Sphingomonas and Sphingopyxis in greater detail, several strains were grown in K7 medium at 25 °C. Pigments extracted with methanol/acetone (7/2, v/v) were separated by reverse-phase HPLC (HP 1100 Series; Hewlett Packard) on a C18 silica column (Multophyp ODS 5 μm, 250×4.6 mm; CS Chromatographie Service) using a modification of method B as previously described by Airs et al. (2001). In brief, the mobile phase containing 5 % ammonium acetate, 80 % methanol and 15 % acetonitrile was changed to 19 % methanol, 1 % acetonitrile and 80 % ethylacetate, thereby eluting apolar compounds. The gradient program was shortened from 85 to 40 min using a flow rate of 0.5 ml min⁻¹, yielding sufficient separation of carotenoids (see Supplementary Fig. S3b). The most abundant carotenoid of strain FSW06-204d^T was also most abundant in all of the investigated species of the genus Novosphingobium and Sphingomonas parapaucimobilis DSM 7463^T, but was not present in Erythrobacter longus ATCC 33941^T (see Supplementary Fig. S3b). Comparison of absorption spectra and retention times in HPLC analysis (see Supplementary Fig. S3c and Supplementary Table S1) suggests that this carotenoid represents nostoxanthin, the major carotenoid in Sphingomonas paucimobilis (Jenkins et al., 1979) and N. capsulatum (Bally et al., 1990). Another pigment of strain FSW06-204d^T (Peak 3, Supplementary Fig. S3b) was also common in all investigated species (see Supplementary Table S1) and was identified as bacteriorubixanthinal by its presence in E. longus (Kim et al., 2007). As N. nitrogenifigens DSM 19370^T was slightly yellow-pigmented, only the major pigments were present in HPLC chromatograms (see Supplementary Table S1). Strain FSW06-204d^T could be distinguished from other Novosphingobium species by the presence of an unidentified carotenoid (Peak 8, Supplementary Fig. S3b) in relatively high amounts (13.1 %, Supplementary Table S1).