Phyllobacterium loti sp. nov. isolated from nodules of Lotus corniculatus

Abstract

Strain S658^T was isolated from a Lotus corniculatus nodule in a soil sample obtained in Uruguay. Phylogenetic analysis of the 16S rRNA gene and atpD gene showed that this strain clustered within the genus Phyllobacterium. The closest related species was, in both cases, Phyllobacterium trifolii PETP02^T with 99.8 % sequence similarity in the 16S rRNA gene and 96.1 % in the atpD gene. The 16S rRNA gene contains an insert at the beginning of the sequence that has no similarities with other inserts present in the same gene in described rhizobial species. Ubiquinone Q-10 was the only quinone detected. Strain S658^T differed from its closest relatives through its growth in diverse culture conditions and in the assimilation of several carbon sources. It was not able to reproduce nodules in Lotus corniculatus. The results of DNA–DNA hybridization, phenotypic tests and fatty acid analyses confirmed that this strain should be classified as a representative of a novel species of the genus Phyllobacterium, for which the name Phyllobacterium loti sp. nov. is proposed. The type strain is S658^T( = LMG 27289^T = CECT 8230^T).

Lotus corniculatus L. (bird’s-foot trefoil) is the most important and widely distributed crop from the Lotus genus (Díaz et al., 2005). Although the species originated in temperate areas of Europe and North Africa, it has spread throughout the world as a result of cultivation (Díaz et al., 2005). Nowadays Lotus corniculatus is the most widely used legume in sown pastures in Uruguay, where it establishes symbiosis with Mesorhizobium (Sotelo et al., 2011).

In the present study, we characterized a novel strain, designated S658^T, which was isolated from Lotus corniculatus nodules, and showed that it represents a novel species of the genus Phyllobacterium. At the time of writing, the genus Phyllobacterium, described by Knösel (1962), contains nine species with validly published names (Flores-Félix et al., 2013), Phyllobacterium myrsinacearum, isolated from leaf nodules of Myrsinaceae (Mergaert et al., 2002), Phyllobacterium catacumbae, isolated from a volcanic rock (Jurado et al., 2005), Phyllobacterium trifolii, isolated from Trifolium nodules (Valverde et al., 2005), four species isolated from the rhizoplane of different plants, P. brassicacearum, P. bourgognense, P. leguminum and P. ifriqiyense (Mantelin et al., 2006), and P. endophyticum, a recently described species isolated from nodules of Phaseolus vulgarishttp://dx.doi.org/10.1601/nm.3268 (Flores-Félix et al., 2013).

Strain S658^T was isolated from a nodule of Lotus corniculatus in Uruguay during a study of bacteria nodulating this legume. The root nodules were washed several times with sterile distilled water and surface-sterilized for 1 min in 90 % (v/v) ethanol followed by 3 min in 4 % (v/v) NaClO, then washed three times in sterile distilled water, as previously described (Sotelo et al., 2011). The nodules were then crushed using a sterile glass rod and the homogenized nodule tissue was inoculated on modified yeast mannitol agar (YMA; Vincent, 1970) (containing l⁻¹: 10 g mannitol, 1 g yeast extract, 0.2 g K₂HPO₄, 0.2 g MgSO₄ . 7H₂O, 0.5 g NaCl, 20 g agar) and incubated at 28 °C for 4 days. The cultures used in further phenotypic and molecular studies were purified from a single colony after incubation for 4 days at 28 °C on YMA. Colonies were white, mucoid, translucent and convex on this medium. Reinfection experiments were performed in Lotus corniculatus, as previously described (Sotelo et al., 2011). Strain S658^T was unable to reproduce nodules in Lotus corniculatus and we failed to amplify the nodC gene with the primers described by Laguerre et al. (2001) that allowed the amplification of nodC genes of Mesorhizobium strains from Lotus corniculatus isolated in the same region as our strain (Sotelo et al., 2011).

Strain S658^T was grown in nutrient broth (Difco, Becton Dickinson, BBL) for 48 h at 22 °C to check for motility by phase-contrast microscopy using the hanging-drop method. Gram staining was carried out by the procedure described by Doetsch (1981) after 24 h incubation at 28 °C. The flagellation type was determined by electron microscopy after incubation of strain S658^T for 48 h in nutrient agar (Difco, Becton Dickinson, BBL) at 22 °C, as previously described (Rivas et al., 2007). Cells of strain S658^T were Gram-stain-negative, rod-shaped, non-sporulating, motile by peritrichous flagella and commonly observed as single cells (Fig. S1, available in the online Supplementary Material).

The 16S rRNA gene was amplified and sequenced as described by Rivas et al. (2007) and the atpD gene using the primers atpD-273F and atpD-771R, as described by Weir et al. (2004). The sequences obtained were compared with those from GenBank using the blastn program (Altschul et al., 1990) and the 16S rRNA gene sequences were also compared with those from the EzTaxon-e server (Kim et al., 2012). Sequences were aligned using clustal x software (Thompson et al., 1997). The distances were calculated according to Kimura’s two-parameter model (Kimura, 1980). Phylogenetic trees of 16S rRNA and atpD genes were inferred using the neighbour-joining model (Saitou & Nei, 1987), and in the case of 16S rRNA gene, the maximum-likelihood model (Rogers and Swofford, 1998) was also used. mega5 software (Tamura et al., 2011) was used for all analyses.

The resulting neighbour-joining tree from 16S rRNA gene analysis is shown in Fig. 1, and the results were congruent with those from maximum-likelihood modelling (data not shown) showing that strain S658^T belongs to the genus Phyllobacterium. Comparison of the 16S rRNA gene sequence of strain S658^T against those in the EzTaxon-e database showed that its closest relative was P. trifolii PETP02^T with 99.79 % sequence similarity without gaps, but strain S658^T contains an insert of 85 bp located between nucleotides 70 and 155 of the rrs gene sequence, which is absent in all species of the genus Phyllobacterium described to date.

Fig. 1.

Neighbour-joining tree based on nearly complete 16S rRNA gene sequences of Phyllobacterium loti sp. nov. S658^T, species of the genus Phyllobacterium and those from the closest genus Mesorhizobium. The significance of each branch is indicated by a percentage bootstrap value calculated for 1000 subsets. Bar, 1 nt substitutions per 100 nt.

Phylogenetic analysis of the atpD (ATPase synthase β subunit) gene (Fig. 2) showed that strain S658^T clustered within the Phyllobacterium clade, forming a separate branch within this genus, and related to a group formed by P. brassicacearum LMG 22836^T (STM 196^T) and P. trifolii PETP02^T, with the closest strain, P. trifolii PETP02^T, having 96.4 % sequence similarity, which was the same as that found between P. brassicacearum LMG 22836^T and P. trifolii PETP02^T. This suggested that the novel strain could also belong to a different species of the genus Phyllobacterium.

Fig. 2.

Neighbour-joining tree based on partial atpD gene sequences of Phyllobacterium loti sp. nov. S658^T and related members of the family Phyllobacteriaceae. The significance of each branch is indicated by a percentage bootstrap value calculated for 1000 subsets. Bar, 2 nt substitutions per 100 nt.

DNA–DNA hybridization experiments were carried out in triplicate using the method of Ezaki et al. (1989), following the recommendations of Willems et al. (2001). Strain S658^T was hybridized with P. trifolii PETP02^T showing 47 % (±8 %) DNA–DNA relatedness. This value is lower than the threshold value of 70 % DNA–DNA relatedness for definition of bacterial species (Wayne et al., 1987), indicating that the strain isolated in this study belongs to a new species of the genus Phyllobacterium.

DNA for analysis of DNA base composition was prepared according to the method of Chun & Goodfellow (1995). The DNA G+C content was determined using the thermal denaturation method (Mandel & Marmur, 1968). The DNA G+C content of strain S658^T was 57.8 mol%. This value is within the range of DNA G+C content reported for members of the genus Phyllobacterium (51–58 mol%; Mantelin et al., 2006).

Cellular fatty acids were analysed using the Microbial Identification System (midi; Microbial ID) Sherlock 6.1 and the RTSBA6 library, according to the technical instructions provided (Sasser, 1990). P. trifolii PETP02^T was included in this analysis together with strain S658^T. The strains were grown on TSA plates (Tryptic Soy Agar, Becton Dickinson; BBL) for 48 h (late-exponential growth phase) at 28 °C. The results of the fatty acid analysis are shown in Table 1. The major fatty acids (>10 %) present in strain S658^T were C_16 : 0 (20.7 %), C_18 : 1ω7c 11-methyl (16.5 %) and C_18 : 1ω6c/C_18 : 1ω7c in summed feature 8 (48.5 %). Strain S658^T differed from the remaining tested strains in the amount of C_18 : 1ω7c 11-methyl, which was higher in the novel strain, in the amount of C_19 : 0 cyclo ω8c, which was lower, and in the minor presence of C_17 : 0. It also differed from its closest relative P. trifolii PETP02^T in the amounts of C_16 : 0 and C_18 : 1ω6c C_18 : 1ω7c (in summed feature 8), in the presence of C_18 : 1 2-OH and in the absence of C_16 : 0 3-OH.

Table 1. Cellular fatty acid composition of strain S658^T and type strains of the closest related species of the genus Phyllobacterium.

Strains: 1, S658^T; 2, P. trifolii LMG PETP02^T; 3, P. endophyticum PEPV15^T; 4, P. brassicacearum LMG 22836^T; 5, P. bourgognense LMG 22837^T. All data are from this study. nd, not detected.

Fatty acid	1	2	3	4	5
Saturated straight-chain
C13 : 1 at 12-13	nd	nd	nd	nd	1.6
C14 : 0	0.79	nd	0.71	nd	1.4
C16 : 0	20.68	12.3	22.6	15.8	22.1
C17 : 0	0.83	nd	nd	nd	nd
C18 : 0	2.78	1.8	1.9	1.5	6.4
Hydroxy fatty acids
C16 : 0 3-OH	nd	1.3	1.6	2.3	1.6
C18 : 1 2-OH	1.03	nd	1.7	nd	nd
C18 : 0 3-OH	nd	nd	0.6	0.7	nd
Cyclopropane fatty acids
C17 : 0 cyclo	nd	nd	2.2	0.4	1.4
C19 : 0 cyclo ω8c	2.62	6.4	13.9	5.4	10.7
Unsaturated
C20 : 2ω6,9c	nd	0.7	nd	0.4	nd
C18 : 1ω7c 11-methyl	16.53	10.7	9.2	9.7	9.9
Summed feature*
2	0.97	1.5	3	3.1	1.4
3	5.73	2.5	3.6	2.6	4.3
8	48.05	62.9	39.1	58.3	39.3

^*Summed features represent groups of two or three fatty acids that could not be separated by the Microbial Identification System. Summed feature 2 contained C_14 : 13-OH/ C_16 : 1 iso 1; summed feature 3 contained C_16 : 1ω6c/C_16 : 1ω7c; summed feature 8 contained C_18 : 1ω6c/C_18 : 1ω7c.

Strain S658^T was cultivated for 48 h in TSA at 28 °C to obtain the cell mass required for quinone analysis. Quinone analysis was carried out by the Identification Service of the DSMZ (Braunschweig, Germany). Ubiquinone Q-10 was the unique respiratory quinone.

Strain S658^T was grown on TY plates (containing l⁻¹: 4 g tryptone, 3 g yeast extract, 0.9 g CaCl₂ and 20 g agar) for 24 h for catalase and oxidase production analysis. Catalase production was assayed by using 0.3 % hydrogen peroxide with one colony taken from TY plates. Oxidase activity was detected using N,N,N′,N′-tetramethyl-1,4-phenylenediamine dihydrochloride. The ability to grow at different temperatures (4, 6, 10, 28, 37 and 39 °C) and in the presence of different NaCl concentrations (1, 1.5, 2 and 2.5 %) was determined on the YMA modified medium described above. The ability to grow at different pH (4.5, 5, 6, 7 and 8) was determined in yeast mannitol broth (YMB; containing l⁻¹: 10 g mannitol, 1 g yeast extract, 0.2 g K₂HPO₄, 0.2 g MgSO₄ . 7H₂O, 0.5 g NaCl). PCA buffer (0.4 M Na₂HPO₄ and 0.2 M citric acid) was used to adjust the pH from 4.5 to 6.5 and TE buffer (0.2 M Tris adjusted to pH 8 with HCl plus 1 mM EDTA) was used for pH 8. Other physiological and biochemical tests were carried out using API 20NE and API ID32GN systems (bioMérieux) following the manufacturer’s instructions and the results were read after 72 h and 5 days incubation at 28 °C. To test natural antibiotic resistance the following antibiotics were used: ampicillin (2 µg), erythromycin (2 µg), ciprofloxacin (5 µg), penicillin (10 IU), polymyxin B (300 IU), cloxacillin (1 µg), oxytetracycline (30 µg), gentamicin (10 µg), cefuroxime (30 µg) and neomycin (5 µg), (Becton Dickinson, BBL). The disc diffusion method on YMA medium (Vincent, 1970) incubated at 28 °C was used. The results were read after 48 h incubation; the strain was considered to be sensitive if the inhibition zone was at least 1 cm in diameter. The type strains of species of the genus Phyllobacterium available in the LMG culture collection were included in the phenotypic study as reference strains. Phenotypic characteristics of the novel species are reported in the species description and the differences with respect to other species of the genus Phyllobacterium are recorded in Table 2. Our results were basically in agreement with those of Mantelin et al. (2006) for P. brassicacearum and P. bourgognense, although assimilation of gluconate by the type strains of these species was negative using the API 32GN system and that of l-alanine and maltose were negative for the type strains of P. bourgognense and P. leguminum, respectively. In the case of P. trifolii, the assimilation of sucrose was positive using the API 32GN system whereas it was not detected by Valverde et al. (2005) using the API 50CH system.

Table 2. Phenotypic differences among species of genus Phyllobacterium.

Species: 1, Phyllobacterium loti sp. nov. (strain S658^T); 2, P. trifolii (Valverde et al., 2005), 3, P. endophyticum (Flores-Félix et al., 2013); 4, P. brassicacearum (Mantelin et al., 2006); 5, P. bourgognense (Mantelin et al., 2006); 6, P. leguminum (Mantelin et al., 2006); 7, P. ifriqiyense (Mantelin et al., 2006); 8, P. catacumbae (Jurado et al., 2005); 9, P. myrsinacearum (Knösel, 1984; Valverde et al., 2005; Mergaert & Swings, 2005; Mantelin et al., 2006). Data for P. ifriqiyense are from Mantelin et al. (2006) and those from the type strains of the remaining species are coincident with those from the original descriptions of each one and from Flores-Félix et al. (2013) except in the cases marked in this table. +, Positive; −, negative; w, weakly positive; v, variable; nd, no data.

Characteristic	1	2	3	4	5	6	7	8	9
Growth at/in:
37 °C	+	+	+	−	−	+	+	+	−
pH 5	+	−	+	+	+	w‡	+	+	+
2 % NaCl	w	+	+	w‡	−	+	+	−	+
Growth on:
Maltose*†	+	+	−	+	+	−‡	+	+	+
Gluconate†	−	+	−	−‡	−‡	+	+	+	+
Sucrose*	+	+§	−	+	+	−	nd	+	+
5-Ketogluconate*	w	+	+	+	+	−	+	+	+
d-Malate†	+	+	+	+	+	−	+	+	+
Malonate*	−	−	w	+	−	−	+	w	−
dl-Lactate*	−	+	+	+	+	+	+	+	+
Valerate*	−	+	−	+	+	−	v	+	+
l-Alanine*	−	w	+	+	−‡	−	+	+	+
4-Hydroxybenzoate*	−	−	−	+	−	−	−	−	+
3-Hydroxybutyrate*	+	+	+	+	+	−	+	+	+
2-Ketogluconate*	−	+	w	w	w	−	nd	+	+
Proline*	−	+	+	+	+	+	+	+	+
Resistance to:
Cefuroxime	+	+	+	+	−	+	nd	+	+
Gentamicin	+	−	−	−	−	+	nd	+	+
Polymyxin B	−	−	−	+	−	+	nd	+	+

^*Data from API ID32GN.

^†Data from API 20NE.

^‡Data differ from those of Mantelin et al. (2006).

^§Sucrose assimilation was not detected using API 50CH (Valverde et al., 2005).

In summary, the phylogenetic, chemotaxonomic and phenotypic characteristics of strain S658^T showed that it represents a novel species of the genus Phyllobacterium, for which the name Phyllobacterium loti sp. nov. is proposed.

Description of Phyllobacterium loti sp. nov.

Phyllobacterium loti (lo′ti. L. n. lotus the African lotus and also a botanical genus name; L. gen. n. loti of lotus, isolated from Lotus corniculatus)

Gram-stain-negative, aerobic rods as for other species of the genus. Colonies are small (<2 mm in diameter) and pearl white on YMA after 48 h incubation at 28 °C. The temperature range for growth is 6–39 °C (optimal growth occurs at 28 °C). The pH range for growth is pH 5–8 (optimal growth occurs at pH 7). Grows in the presence of up to 2 % (w/v) NaCl. The major cellular fatty acids are C_16 : 0, C_18 : 1ω7c 11-methyl and C_18 : 1ω6c/C_18 : 1ω7c (summed feature 8). The only quinone detected is Q-10. Catalase and oxidase activities are positive. Nitrate reduction and β-galactosidase activity are negative and aesculin is hydrolysed in the API 20NE system. Arginine dehydrolase, indole, gelatinase and urease are not produced. In the same system glucose, l-arabinose, mannose, mannitol, N-acetylglucosamine, maltose and d-malate are used as carbon sources. Assimilation of gluconate, caprate, adipate, citrate and phenylacetate is negative. In the API ID32GN system the assimilation of N-acetylglucosamine, d-ribose, inositol, sucrose, maltose, d-mannitol, d-glucose, l-fucose, d-sorbitol, l-arabinose, l-histidine and 3-hydroxybutyrate is positive, but assimilation of dl-lactate, l-alanine, itaconate, suberate, malonate, glycogen, 3-hydroxybenzoate, l-serine, salicin, melibiose, propionate, caprate, valerate, citrate, 2-ketogluconate, 4-hydroxybenzoate and l-proline is negative. Assimilation of l-rhamnose, acetate and 5-ketogluconate is weak. Sensitive to ciprofloxacin (5 µg), neomycin (5 µg), polymyxin B (300 IU) and oxytetracycline. Resistant to ampicillin (2 µg), penicillin (10 IU), cefuroxime (30 µg), cloxacillin (1 µg), gentamicin (10 µg) and erythromycin (2 µg).

The type strain is S658^T ( = LMG 27289^T = CECT 8230^T), and was isolated from a nodule of Lotus corniculatus in Uruguay. The DNA G+C content of the type strain is 57.8 mol%.