Introduction

Bacteria belonging to the genus Bradyrhizobium and able to fix N2 in symbiosis with leguminous plants form a monophyletic group within the α class of Proteobacteria together with oligotrophic soil and aquatic bacteria (Durán et al. 2012). The development and introduction of several molecular techniques to the taxonomic studies of bacteria helped to identify high genetic diversity among Bradyrhizobium strains. For the phylogenetic reconstructions of the genus Bradyrhizobium bacteria, the atpD, dnaK, gyrB, glnII, recA, and rpoB markers have been commonly used (Delamuta et al. 2012; Kalita and Małek 2017; Menna et al. 2009; Rivas et al. 2009; Stępkowski et al. 2003; Vinuesa et al. 2005). The ftsA gene sequences have not been used in the phylogenetic analysis of root nodule bacteria. The FtsA protein functions at the earliest stage of bacterial division, connecting FtsZ, the principal component of the division machinery, to the cell membrane, and forms a structure called the proto-ring at the division site (Fujita et al. 2014). FtsA belongs structurally to the actin/Hsp70/hexokinase superfamily and is widespread in bacteria (Busiek and Margolin 2015). The functional preservation and universal distribution among bacteria make the ftsA gene suitable for inferring phylogenetic relationships.

The aim of the present study was to estimate the degree of ftsA gene sequence conservation among the genus Bradyrhizobium strains and to determine whether the ftsA gene could be used as a new marker in the phylogenetic analysis of Bradyrhizobium species. The availability of several fully and partially genome-sequenced strains allowed us to address this issue using in silico analysis. The ftsA phylogeny of bradyrhizobia was discussed in comparison to the phylogenies of other chromosomal genes glnII, and recA.

Materials and methods

PCR and sequencing

The region of 1140 bp length of ftsA gene was amplified using primers ftsAF (5’-ATCGGYTACAGCCAGATCCAGT-3′) and ftsAR (5’-CCTCGCGTAGCCATCGTCCRA-3′). The PCR protocol was as follows: 3 min of initial denaturation carried out at 95 °C followed by 35 cycles of 1 min at 95 °C, 30 s at 58 °C, 1 min at 72 °C with the final elongation step of 7 min at 72 °C. The amplified products were sequenced in both directions with ftsAF/fstAR primers using BigDye Terminator Cycle sequencing kit and the 3500 Genetic Analyzer according to the manufacturer’s procedures (Thermo Fisher Scientific).

Sequence data analysis

The ftsA gene sequences obtained in this study were deposited in the GenBank database under the accession numbers provided in Supplementary Table S1. The ftsA, glnII, and recA gene sequences of completely or partially sequenced genomes of Bradyrhizobium strains were obtained from NCBI Genome database. The complete list of strains and their accession numbers is available in Table S1. All the phylogenetic analyses were conducted in MEGA 7 (Kumar et al. 2016). Sequence identity values for single genes were calculated using BioEdit software (Hall 2011). The 2-D matrices generated in BioEdit were edited manually and converted into tables in Excel (Supplementary Tables S2-S5).

Results and discussion

Phylogenetic analysis was carried out using 733 bp long fragments of the ftsA gene of 69 bradyrhizobial strains encompassing sequences of eight strains isolated from root nodules of four Genisteae tribe plants growing in Poland, and 61 sequences of bradyrhizobial strains affiliated to 28 species of the genus Bradyrhizobium of which 44 were retrieved from the NCBI genomic database and 17 sequences were generated during this study.

The ftsA sequences divided the analyzed strains into two distinct groups, as shown on the phylogenetic tree (Fig. 1). One group consisted of 15 bradyrhizobial species, among others, B. japonicum, B. diazoefficiens, B. canariense, B. yuanmingense, and B. liaoningense. All root nodule isolates of the Genisteae plant species were positioned with strains representing B. japonicum. B. elkanii, B. erythrophlei, B. valentinum, and B. lablabi were assigned to the other group. In a similar way, all bradyrhizobia were grouped in the phylograms of the glnII and recA genes (Supplementary Figs. S1 and S2), which are commonly used as phylogenetic markers in the studies of Bradyrhizobium bacteria (Kalita and Małek 2017; Menna et al. 2009; Rivas et al. 2009).

Fig. 1
figure 1

Maximum likelihood phylogenetic tree of ftsA gene sequences of Bradyrhizobium strains. Bootstrap values ≥ 70% are given at branching points. The scale bar indicates the number of substitution per site

To assess the resolving power of the ftsA marker at the species level, we analyzed several sequences retrieved from fully or partially sequenced genomes of strains belonging to different Bradyrhizobium species. As can be seen on the ftsA phylogenetic tree, strains representing B. canariense, B. diazoefficiens, B. liaoningense, and B. yuanmingense form very well-resolved clusters (Fig. 1). This clustering is supported by the glnII and recA phylogenies (Supplementary Figs. S1 and S2). The phylogenetic analysis of the ftsA sequences of 21 strains named Bradyrhizobium japonicum demonstrated that 14 of them were placed on the ftsA phylogram in a common cluster together with the type strain B. japonicum USDA 6T (Fig. 1). The other seven strains were grouped with other species of the genus Bradyrhizobium or were placed on separate branches (Fig. 1). This scattered position of the B. japonicum strains on the ftsA tree is supported by the glnII and recA phylogenies (Supplementary Figs. S1 and S2). This led us to a conclusion that the nomenclature of the strains currently named B. japonicum and found outside the B. japonicum species cluster requires revision.

The level of ftsA gene sequence diversity was assessed by calculating the number and percentage of variable positions in the alignment. There were 286 variable characters in the ftsA alignment, which corresponds to 38.8% of all nucleotide positions included in the analysis. This value was higher than the number of variable positions estimated for the glnII (31.2%) and recA (31.4%) alignments, indicating that the ftsA gene sequence analysis yields more phylogenetic information. The interspecific level of ftsA sequence similarity ranged from 80 to 97.4% and was comparable with glnII (84.1–98.1%) and recA (88.5–97%) genes. The intraspecific ftsA sequence similarity ranged from 97.1 to 100%. A similar range of sequence variation at the species level was observed for glnII (96.6–100%) and recA (96–100%) genes (Supplementary Tables S2-S4). The highest values of interspecific and the lowest value of intraspecific sequence similarity overlap, which means that there are no gaps that allow distinguishing between different species.

Since the phylogenetic analysis of ftsA gene sequences has proven to be reliable in discrimination of closely related Bradyrhizobium strains, we decided to evaluate the usefulness of this marker in species identification. As the ftsA sequences of B. japonicum strains were most commonly represented in our analysis, we have carefully checked the ftsA sequence alignment to see whether there are any polymorphisms unique to this species. The single-nucleotide polymorphism (SNP) has been used earlier in identification of bacterial species. For example, the SNP analysis of 16S rRNA gene sequences was used for Bacillus cereus and Bacillus anthracis discrimination (Hakovirta et al. 2016). The SNP analysis of three genes was used for distinguishing closely related species of the genus Brucella (Scott et al. 2007). It was also demonstrated that a single-nucleotide polymorphism in the rpoB gene allows specific identification of Salmonella enterica serotype Typhimurium (Hernandez Guijarro et al. 2012).

All 22 B. japonicum strains including eight isolates from root nodules of the Genisteae plants have guanine at position 225 of the ftsA alignment, whereas other bradyrhizobia, including seven misnamed B. japonicum strains, have cytosine or thymine (Fig. 2). This single-nucleotide polymorphism corresponding to position 561 in the ftsA gene of Bradyrhizobium japonicum USDA 6T occurs at the third position of the 187th codon and results in nonsynonymous substitution.

Fig. 2
figure 2

Nucleotide sequence alignment of the ftsA gene fragments. Only differences relative to the top sequence (B. japonicum USDA 6T) are shown. The shaded nucleotide position 225 corresponds to the single-nucleotide polymorphism (guanine) identified in Bradyrhizobium japonicum strains

As the observed SNP appeared to be particularly promising for the identification of B. japonicum, we decided to broaden the analysis by comparison of the ftsA gene sequences retrieved from both complete and partially sequenced bradyrhizobial genomes available in the NCBI Genome database. As a result, 176 ftsA gene sequences obtained from 97 strains affiliated to 34 known bradyrhizobial species, 76 strains named Bradyrhizobium sp., and three Bosea reference sequences were aligned and checked for the SNP at position 225. Guanine at position 225 was observed in the ftsA sequence of Bradyrhizobium sp. G22 strain in addition to the 22 B. japonicum strains mentioned previously (Supplementary Fig. S3). The ftsA sequence of Bradyrhizobium sp. G22 was most similar to the ftsA gene sequences of B. japonicum (97–100%) (Supplementary Table S5). Jones et al. (2016) demonstrated that Bradyrhizobium sp. G22 is closely related to B. japonicum USDA 6T and B. japonicum E109. Bradyrhizobium sp. G22 is also positioned within the B. japonicum cluster on the ftsA phylogram reconstructed with 176 sequences (Supplementary Fig. S4).

In conclusion, we have demonstrated that the ftsA gene may serve as a useful molecular marker in phylogenetic and taxonomic studies of genus Bradyrhizobium bacteria. It holds enough phylogenetic information to distinguish closely related species and, at the same time, it is sufficiently conserved at the intraspecific level allowing correct clustering of strains belonging to a single species. The results of the comparative ftsA sequence analysis suggest that the presence of guanine at nucleotide position 561 of the full length gene sequence could be considered as a unique feature of Bradyrhizobium japonicum strains. More studies with more bradyrhizobial isolates should provide the evidence whether this SNP could be used as an additional marker for identification of B. japonicum bacteria within genus Bradyrhizobium populations.