Identification of an Exopolysaccharide Biosynthesis Gene in Bradyrhizobium diazoefficiens USDA110

Exopolysaccharides (EPS) play critical roles in rhizobium-plant interactions. However, the EPS biosynthesis pathway in Bradyrhizobium diazoefficiens USDA110 remains elusive. Here we used transposon (Tn) mutagenesis with the aim to identify genetic elements required for EPS biosynthesis in B. diazoefficiens USDA110. Phenotypic screening of Tn5 insertion mutants grown on agar plates led to the identification of a mutant with a transposon insertion site in the blr2358 gene. This gene is predicted to encode a phosphor-glycosyltransferase that transfers a phosphosugar onto a polyprenol phosphate substrate. The disruption of the blr2358 gene resulted in defective EPS synthesis. Accordingly, the blr2358 mutant showed a reduced capacity to induce nodules and stimulate the growth of soybean plants. Glycosyltransferase genes related to blr2358 were found to be well conserved and widely distributed among strains of the Bradyrhizobium genus. In conclusion, our study resulted in identification of a gene involved in EPS biosynthesis and highlights the importance of EPS in the symbiotic interaction between USDA110 and soybeans.


Introduction
Establishment of an effective nodule symbiosis between nitrogen-fixing rhizobia and host plants depends on reciprocal complex molecular dialogues. In many interactions, nodule formation requires rhizobial exopolysaccharides (EPS) or EPS-derived oligosaccharides [1][2][3][4][5][6][7]. For example, the EPS of Bradyrhizobium diazoefficiens (B. japonicum) USDA110 was found to play an important role in the symbiosis with soybean plants [8][9][10]. However, EPS is dispensable for the symbiosis between Sinorhizobium (Ensifer) fredii HH103 and soybeans [11]. In the interaction between Mesorhizobium loti and Lotus japonicus, EPS is conditionally required during different symbiotic stages [12]. Likewise, EPS oligosaccharides produced by Sinorhizobium sp. NGR234 are essential for the nodulation of various legumes but dispensable for the nodulation of cowpea [13]. The absence of EPS or the synthesis of altered oligosaccharides can impair the rhizobial root hair infection process, suggesting that EPS contributes to the suppression of plant defense reactions in compatible interactions [14]. EPS can protect rhizobia and other bacteria from biotic and abiotic stresses by influencing the formation of bacterial biofilms, adhesion to root surfaces, and other mechanisms [2,3]. Moreover, EPS or oligosaccharides derived from EPS possess a signaling role during rhizobial infection. The LysM receptor kinase EPR3 in L. japonicus has been identified as a first EPS receptor, suggesting that EPS affects symbiotic signaling in host plants [12].

Tn5 Library Screening
The Tn5 mutant library stored at −80 • C was recovered on AG agar plates at 30 • C. The parent strain USDA110 and the Tn5 mutant colony were detected on AG agar plates containing 60 mg/L of Congo red at 30 • C. The bacterial morphology was photographed 5 days post inoculation.

Whole-Genome Short Read Re-Sequencing
One µg of microbial genomic DNA was sonicated to an average size of~350 bp by a Covaris-S220 ultrasonicator (Covaris, Woburn, MA, USA). Illumina DNA fragment library preparation was performed following the manual of the Next-Generation Sequencing DNA library preparation kit (Novagen, Tianjin, China). Briefly, the sonicated DNA products were end-repaired and ligated with an adapter. Adapter-ligated products were purified using AMPure XP beads (Agencourt-Berkman Coulter, IN, USA) and enriched through PCR amplification using the custom adapter-specific primers. The obtained unbiased short read library was further cleaned up with AMPure XP beads. Pair end Illumina HiSeq PE150 sequencing was performed with an Illumina NovaSeq 6000 sequencing system.

Used Software
The following software was used for sequence analysis.

Nodulation Tests with Soybean Plants
A Leonard jar assembly [32] was used to assess nodule formation and plant growthpromoting effects caused by USDA110 and mutant bacteria. The upper jar was filled with sterilized soil (1/3) and sterilized vermiculite (2/3), while the lower jar contained nitrogenfree nutrient solution (Fåhraeus medium) [33]. The whole jar units were autoclaved before use. Soybean seeds were surface sterilized with 75% ethanol for 10 min and then washed with sterilized H 2 O three times. Seeds were further treated with a 5% sodium hypochlorite solution for 5 min and washed with H 2 O at least ten times. Sterilized seeds were soaked with H 2 O for 24 h and subsequently germinated on 0.2% (w/v) agar plates at 22 • C. Germinated seeds were then transferred into the upper jars (1 plant per jar).
For inoculation with bradyrhizobia, 1 mL of a bacterial suspension suspended in 10 mM of MgSO 4 (bacterial growth reaching late logarithmic phase; OD 600 ≈ 1.5) was added to each jar. Plants were cultivated at 24 ± 2 • C under a 16 h light/8 h dark cycle at a light intensity of 18,000 lux. More than three experiments (biologically independent repeats) were performed, and each experiment (6 plants per strain) contained at least six technical repeats.

Determination of Carbohydrate Contents
Carbohydrate contents of ethanol precipitates were quantified with the anthrone method as described previously [35]. Samples were dissolved in H 2 O and aliquots (30 µL) were mixed with 83 µL of anthrone reagent (0.01 g anthrone, 0.5 mL ethyl acetate, and 5 mL concentrated H 2 SO 4 ). Samples were incubated in a water bath at 100 • C for 4 min and then cooled on ice for 10 min. The optical density of the samples was measured at 620 nm with a microplate reader (Synergy H1MF, BioTek Instruments, Winooski, VT, USA). Known glucose concentrations served as a reference.

Preparation of Oligosaccharides from Ethanol Precipitated Polysaccharide
Ethanol precipitated material was air-dried and then dissolved in a volatile buffer (100 mM pyridine/acetic acid buffer, 8.07 mL pyridine, 5.724 mL acetate, and 986.206 mL H 2 O; pH 5.0). Subsequently, the sample was passed through a Detoxi-Gel TM Endotoxin Removing Gel column (Thermofisher, IL, USA) to remove lipophilic compounds. The sample was then loaded on a Sephadex G-75 gel filtration column (GE Healthcare Life Sciences, Marlborough, UK) equilibrated with the same volatile buffer. The high-molecular faction was collected and lyophilized to remove the volatile pyridine/acetic buffer. The sample (≈50 µg) was then dissolved in 200 µL of H 2 O and mixed with 200 µL of 4 M trifluoroacetic acid (TFA). After incubation in a boiling water bath for 3.5 h, the hydrolyzed mixture was cooled down on ice and the pH was adjusted to 5 with NaOH. The aliquot was then passed through a Dowex 50 W × 4-100 column (H + form; treated with 2 M HCl for 3 h and equilibrated with H 2 O). Finally, the oligosaccharides were eluted with H 2 O and purified on a Sephacryl S-100 gel filtration column (GE Healthcare Life Sciences, Marlborough, UK) by using an AKTA protein purification system (GE Healthcare Life Sciences, Marlborough, UK). Pyridine/acetic acid buffer served as the running solvent. The low-molecular fraction containing oligosaccharides was identified with the anthrone method, lyophilized, and subjected to mass spectrometry analysis.

Mass Spectrometry Analysis of Oligosaccharides
Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) was used to analyze the oligosaccharides obtained by TFA hydrolysis. To prepare the matrix, 1 mg/mL of 2,3-dihydroxybenzoic acid (Sigma-Aldrich, St. Louis, MO, USA) was added to 50% acetonitrile supplemented with 0.1% (w/v) trifluoroacetic acid. The oligosaccharide sample (≥20 µg) was dissolved in H 2 O and mixed with the same volume of matrix. The test mixture (2 µL) was applied to an MTP 384 ground steel plate and air-dried. MALDI-TOF MS analysis was performed with an Ultraflex III MALDI-TOF mass spectrometer (Bruker Daltonics, Billerica, MA, USA) in the positive ionization mode. The accelerating voltage of the ion source was 2.5 kV and the accelerating voltage of the reflector was 26.3 kV. The sample was ionized with a Smartbeam 1 ultraviolet laser (λ = 355 nm) [36].

Analysis of Blr2358 Homologs
The amino acid sequences of Blr2358 (NP_768998) were used to blast the genomes of the Bradyrhizobium genus in the IMG database (https://img.jgi.doe.gov/, assessed 5 October 2021) using default parameters (BLASTP, E-value: 1 × 10 −5 ). A protein domain search was performed for the sugar transferase domain (pfam02397) in the IMG database.

Phylogenetic Tree Analysis
The 16S ribosomal RNA gene sequences of representative strains of the Bradyrhizobium genus were extracted from the IMG database, assessed 7 October 2021. A phylogenetic tree of 16S rDNA sequences was constructed by MEGA6 using the neighbor-joining method with 1000 bootstrap replications.

Phenotypic Screening of a B. diazoefficiens USDA110 Transposon Mutant Library
To identify genetic elements of the entire genome involved in polysaccharide biosynthesis, we conducted a Tn5 mutagenesis screen in B. diazoefficiens strain USDA110. We took advantage of a previously created Tn5-based random insertion USDA110 mutant library [37]. We first conducted a preliminary search for suitable media and found that AG medium supplemented with Congo red was the most suitable for displaying EPS production. In our screening assay, wild-type USDA110 colonies exhibiting a mucoid morphology served as a control. Tn5-inserted colonies showing a non-mucoid phenotype and a reduced colony size were considered as potential EPS mutants. A total of 50,000 Tn5 mutant colonies were screened, and 20 colony candidates with a dry, non-mucoid phenotype were isolated. Three colonies with the strongest non-mucoid phenotype were chosen for further investigation.

Identification of Tn5 Insertion Sites
We next sought to identify the precise Tn5 insertion sites in the obtained mutant strain candidates. Instead of the conventional HiTAIL-PCR method [38], our study used a whole genome short-read re-sequencing (WGS) method to determine Tn5 localizations. Single or multiple Tn5 insertion sites in those mutant colonies were identified by a customized bioinformatics pipeline and Tn5 flanking sequences were then confirmed by Sanger sequencing. Three mutants (BJ#25, BJ#47, and BJ#57) were found to bear a Tn5 insertion in gene blr2358, and both BJ#47 and BJ#57 contained another Tn5 insertion in a leuC gene (bl10416) ( Table 1 and Table S1). As a Tn5 insertion in gene blr2358 was common in these Microorganisms 2021, 9, 2490 6 of 12 three mutants, we inferred that gene blr2358 is most likely responsible for the non-mucoid phenotypes observed in the mutant colonies. The blr2358 gene is predicted to encode a phosphor-glycosyltransferase family protein (WP_011085145.1). The following experiments were therefore focused on the blr2358 gene (colony BJ#25).

Disruption of the Blr2358 Gene Impairs EPS Production
The blr2358 gene is located in a large gene cluster with coordinates 2,561,178 to 2,595,273 (blr2358-bll2384; [9,22] ( Figure 1A). The gene cluster contains predicted polysaccharide synthesis genes, namely blr2358 and five other glycosyltransferase genes (blr2374, bll2376, bll2377, bll2380, and bll2381), three acyltransferase genes (blr2360, blr2367, and blr2369 named exoZ in Kaneko et al. 2002), and two genes predicted to be involved in the polymerization and export of a polysaccharide, namely blr2362 (named exoP in Kaneko et al. 2002) and bll2379. In addition, the cluster contains lspL (blr2382; UDP-glucuronic acid epimerase gene) and ugdH (blr2383; UDP-glucose 6-dehydrogenase gene), which, when deleted, lead to the formation of EPS lacking galacturonic acid [9]. Bioinformatic analysis suggested that blr2358, blr2359 and blr2360 form an operon and that the expression of these genes is controlled by a promoter upstream of blr2358.
The colony morphology phenotype of the BJ#25 mutant supported the view that the blr2358 gene is implicated in EPS synthesis. Compared to the mucoid colony phenotype of the parent strain, the non-mucoid colonies of #BJ25 containing Tn5-inserted blr2358 were relatively small in size and dry on AG agar plates ( Figure 1B). Furthermore, the complementation of the Tn5-inserted blr2358 mutant by overexpression of an episomal copy of blr2358 restored polysaccharide production to a certain extent ( Figure 1B). When culture supernatants of the wild-type strain and the complementation strain (grown in defined GMS liquid medium) were precipitated with ethanol, a clear precipitate was formed, while nearly no material was precipitated from culture supernatants of the mutant (Supplementary Figure S1 in Supplementary Information). Anthrone test results indicated that the ethanol precipitate from the wild-type and the complementation strain contained considerable amounts of carbohydrates. In contrast, much less carbohydrates were quantified for the corresponding samples of the mutant strain, reaching a level close to control samples, i.e., GMS medium (without bacteria) treated with ethanol ( Figure 1C). These findings provided indications that the gene product of blr2358 is involved in the synthesis of an extracellular polysaccharide.
To confirm that the carbohydrate-rich ethanol precipitate from the wild-type culture supernatant is indeed EPS ( Supplementary Information, Figure S2A), we performed a mass spectrometry (MALDI-TOF MS) analysis for oligosaccharides obtained by the hydrolysis of EPS with TFA. We found that the obtained mass spectrometry signals largely correspond to the known EPS structure of strain USDA110 (Supplementary Information, Figure S2B and Table S2). These findings support the view that the Tn5-inserted blr2358 mutant is impaired in EPS synthesis.
To confirm that the carbohydrate-rich ethanol precipitate from the wild-type culture supernatant is indeed EPS ( Supplementary Information, Figure S2A), we performed a mass spectrometry (MALDI-TOF MS) analysis for oligosaccharides obtained by the hydrolysis of EPS with TFA. We found that the obtained mass spectrometry signals largely correspond to the known EPS structure of strain USDA110 (Supplementary Information, Figure S2B and Table S2). These findings support the view that the Tn5 -inserted blr2358 mutant is impaired in EPS synthesis. USDA110 carrying the empty pBBR1-MCS5 plasmid (W T + Ctrl) and the mutant carrying pBBR1-MCS5-blr2358 (blr2358::Tn + blr2358) are also shown. Equal amounts of bacteria we re spotted on an AG plate containing Congo red and colonies we re photographed after incubation at 30 °C for 5 days. (C) Quantification of carbohydrates in e thanol pre cipitates obtained from supe rnatants of ce ll culture s. WT, blr2358::Tn mutant, WT carrying an e mpty control plasmid, and blr2358::Tn mutant comple me nted with a blr2358 copy we re assayed. Ethanol pre cipitates from GMS me dium without bacte ria we re also analyzed. Data (means ± SD; n = 6; t-te st) we re e xpressed as μg pre cipitate d carbohydrate s (glucose e quivalents) pe r mL bacterial culture supernatant (from bacterial suspensions adjuste d to OD600 ≈ 1.00).

Soybeans Inoculated with the Blr2358 Mutant Show Reduced Nodule Formation
Nodulation tests with soybean cv. Huaxia No. 3 were performed to investigate the symbiotic role of the blr2358 gene. Seedlings were inoculated with the USDA110 wildtype strain, the blr2358::Tn mutant, and the complementation strain (blr2358::Tn mutant with a functional copy of blr2358). The wild-type strain carrying the corresponding empty plasmid without blr2358 and mock-inoculated plants were included into the nodulation tests. At 8 weeks post inoculation, plants were harvested and phenotypically analyzed. As expected, the wild-type strain strongly promoted plant growth. Nodulated plants showed green leaves while leaves of mock-inoculated control plants were yellow . In comparison to the wild-type strain, inoculation with the blr2358::Tn mutant caused a significantly lower plant stem length and the number of formed nodules was reduced to half. In contrast, the complementation strain showed a symbiotic phenotype that was similar to USDA110 carrying the empty pBBR1-MCS5 plasmid (WT + Ctrl) and the mutant carrying pBBR1-MCS5-blr2358 (blr2358::Tn + blr2358) are also shown. Equal amounts of bacteria were spotted on an AG plate containing Congo red and colonies were photographed after incubation at 30 • C for 5 days. (C) Quantification of carbohydrates in ethanol precipitates obtained from supernatants of cell cultures. WT, blr2358::Tn mutant, WT carrying an empty control plasmid, and blr2358::Tn mutant complemented with a blr2358 copy were assayed. Ethanol precipitates from GMS medium without bacteria were also analyzed. Data (means ± SD; n = 6; t-test) were expressed as µg precipitated carbohydrates (glucose equivalents) per mL bacterial culture supernatant (from bacterial suspensions adjusted to OD 600 ≈ 1.00).

Soybeans Inoculated with the Blr2358 Mutant Show Reduced Nodule Formation
Nodulation tests with soybean cv. Huaxia No. 3 were performed to investigate the symbiotic role of the blr2358 gene. Seedlings were inoculated with the USDA110 wild-type strain, the blr2358::Tn mutant, and the complementation strain (blr2358::Tn mutant with a functional copy of blr2358). The wild-type strain carrying the corresponding empty plasmid without blr2358 and mock-inoculated plants were included into the nodulation tests. At 8 weeks post inoculation, plants were harvested and phenotypically analyzed. As expected, the wild-type strain strongly promoted plant growth. Nodulated plants showed green leaves while leaves of mock-inoculated control plants were yellow. In comparison to the wild-type strain, inoculation with the blr2358::Tn mutant caused a significantly lower plant stem length and the number of formed nodules was reduced to half. In contrast, the complementation strain showed a symbiotic phenotype that was similar to the wildtype (Figure 2). These findings indicate that blr2358 disruption in the USDA110 genome negatively affected nodulation and plant growth in the symbiotic interaction between USDA110 and soybeans. the wild-type (Figure 2). These findings indicate that blr2358 disruption in the USDA110 genome negatively affected nodulation and plant growth in the symbiotic interaction between USDA110 and soybeans.

Glycosyltransferase Genes Homologous to Blr2358 Are Conserved in the Bradyrhizobium Genus
The number of homologous genes in a bacterial genome can give indications of their functions. We searched for blr2358 gene homologs in the whole genome of B. diazoefficiens USDA110 and found two other glycosyltransferases, which share 31% and 36% amino acid sequence identity with Blr2358. Both proteins contain a sugar transferase domain (pfam02397) (Table S3) and may both function as phospho-glycosyltransferases. The search for Blr2358 homologs was further extended to the Bradyrhizobium genus. The results revealed that homologs are widely distributed within the 22 surveyed Bradyrhizobium strains (Figure 3). Amino acid sequence similarity ranged from 29% to 96%, with a minimum of two homologs found in Bradyrhizobium sp. KBS0725 and Bradyrhizobium sp. KBS0725, while a maximum of six homologs were identified in Bradyrhizobium sp. CCBAU 051011. Interestingly, a 16S rDNA-based phylogenetic tree analysis revealed that a small phylogroup containing four Bradyrhizobium strains is characterized by homologs with a relatively low similarity to Blr2358, i.e., ranging from only 31% to 50% (Figure 3 and Table  S3). These findings suggest that phospho-glycosyltransferases are associated with adaptive evolutionary scenarios in this clade. In general, the Blr2358 homologs were found to

Glycosyltransferase Genes Homologous to Blr2358 Are Conserved in the Bradyrhizobium Genus
The number of homologous genes in a bacterial genome can give indications of their functions. We searched for blr2358 gene homologs in the whole genome of B. diazoefficiens USDA110 and found two other glycosyltransferases, which share 31% and 36% amino acid sequence identity with Blr2358. Both proteins contain a sugar transferase domain (pfam02397) (Table S3) and may both function as phospho-glycosyltransferases. The search for Blr2358 homologs was further extended to the Bradyrhizobium genus. The results revealed that homologs are widely distributed within the 22 surveyed Bradyrhizobium strains (Figure 3). Amino acid sequence similarity ranged from 29% to 96%, with a minimum of two homologs found in Bradyrhizobium sp. KBS0725 and Bradyrhizobium sp. KBS0725, while a maximum of six homologs were identified in Bradyrhizobium sp. CCBAU 051011. Interestingly, a 16S rDNA-based phylogenetic tree analysis revealed that a small phylogroup containing four Bradyrhizobium strains is characterized by homologs with a relatively low similarity to Blr2358, i.e., ranging from only 31% to 50% (Figure 3 and Table S3). These findings suggest that phospho-glycosyltransferases are associated with adaptive evolutionary scenarios in this clade. In general, the Blr2358 homologs were found to be conserved and broadly distributed across the Bradyrhizobium genus, suggesting that some of them may play a role in symbiotic adaptation to host plants. Microorga nisms 2021, 9, x FOR PEER REVIEW 9 of 12 be conserved and broadly distributed across the Bradyrhizobium genus, suggesting that some of them may play a role in symbiotic adaptation to host plants.

Figure 3.
Re pre sentative Blr2358 homologs in the Bradyrhizobium ge nus. A phyloge ne tic tree was made with MEGA6 software using the ne ighbor-joining me thod and 16S rDNA se que nces from re pre sentative Bradyrhizobium strains. The number of Blr2358 homologs in e ach strain is shown. The blue re ctangle indicates the clade of Bradyrhizobium strains that possess relatively low amino acid se que nce similarities to Blr2358.

Discussion
We found in this study that a Tn5 insertion in the blr2358 gene of B. diazoefficiens USDA110 correlates with a non-mucoid colony phenotype, suggesting that Blr2358 is involved in polysaccharide synthesis. Blr2358 and homologs are expected to be phosphoglycosyltransferases that transfer a phosphosugar onto a polyprenol phosphate substrate, i.e., Blr2358 is predicted to be responsible for the first step of EPS syn thesis. However, based on the EPS structure of B. diazoefficiens USDA110 [18,19] (Figure S2A), there are neither glucose nor galactose residues at the reducing end of the EPS subunit and future biochemical research is required to identify the phosphosugar preference of Blr2358. It is worth noting in this context that EPS synthesis was not fully restored by the complementation strain in the performed agar plate assay. This could be explained by the reduced growth of the complementation strain under the used test conditions (plasm id copy number effects) or the production of an altered EPS. Indeed, it is possible that the Tn5 insertion in blr2358 has certain polar effects on downstream genes, namely blr2359 and blr2360. The blr2359 gene encodes a putative glycosyl hydrolase which could eventually play a role in processing synthesized EPS subunits prior to polymerization. The blr2360 gene is predicted to encode an acyltransferase which could be involved in O-acetylation of the EPS subunit. These three genes are organized in the same transcriptional direction and may form an operon controlled by a promoter upstream of blr2358 ( Figure 1A). Future research will be required to characterize a complementation strain expressing all three genes.

Discussion
We found in this study that a Tn5 insertion in the blr2358 gene of B. diazoefficiens USDA110 correlates with a non-mucoid colony phenotype, suggesting that Blr2358 is involved in polysaccharide synthesis. Blr2358 and homologs are expected to be phosphoglycosyltransferases that transfer a phosphosugar onto a polyprenol phosphate substrate, i.e., Blr2358 is predicted to be responsible for the first step of EPS synthesis. However, based on the EPS structure of B. diazoefficiens USDA110 [18,19] (Figure S2A), there are neither glucose nor galactose residues at the reducing end of the EPS subunit and future biochemical research is required to identify the phosphosugar preference of Blr2358. It is worth noting in this context that EPS synthesis was not fully restored by the complementation strain in the performed agar plate assay. This could be explained by the reduced growth of the complementation strain under the used test conditions (plasmid copy number effects) or the production of an altered EPS. Indeed, it is possible that the Tn5 insertion in blr2358 has certain polar effects on downstream genes, namely blr2359 and blr2360. The blr2359 gene encodes a putative glycosyl hydrolase which could eventually play a role in processing synthesized EPS subunits prior to polymerization. The blr2360 gene is predicted to encode an acyltransferase which could be involved in O-acetylation of the EPS subunit. These three genes are organized in the same transcriptional direction and may form an operon controlled by a promoter upstream of blr2358 ( Figure 1A). Future research will be required to characterize a complementation strain expressing all three genes. Structurally different EPS produced by different rhizobial strains may affect the nodulation of soybeans differently. In the case of S. fredii HH103, for example, an exoA mutant deficient in EPS synthesis induced effective nodules on soybean roots similar to the wild-type strain and even showed an increased competitive capacity to occupy nodules [11]. In contrast, EPS mutants of B. diazoefficiens USDA110 were found to be impaired in the symbiosis with soybeans [8][9][10]. In our study, soybean plants inoculated with the blr2358 mutant showed a significant decrease in nodule formation. Our findings support the view that EPS plays a positive role in the symbiotic relationship between USDA110 and soybeans. Our results also agree with a previous study that suggested a positive role of EPS produced by the B. japonicum strain USDA138 [39]. The distinct roles of different EPS could be explained by the discovery that host plants can sense EPS oligosaccharides produced by different rhizobial strains. The EPR3 receptor of L. japonicus can recognize EPS oligosaccharides of M. loti and that of other strains and hence activate host processes that control rhizobial root-hair infection [40][41][42].
Occurrence of genes closely related to blr2358 is rather common in bacterial genomes, including strain USDA110. It is expected that these genes encode phospho-glycosyltransferases. The number and variations of these genes in a genome may reflect the diversity of produced polysaccharides. In our study, searching for blr2358-encoding homologs in the USDA110 genome yielded two homologs (Table S3 and Figure 3). A certain redundancy of the EPS biosynthesis pathway has been observed for the exoB gene, which codes for UDP-glucose-4-epimerase and shares sequence similarity with the galE gene of B. japonicum 61A101C [43]. Despite possible functional redundancy of homologs, both an exoB mutant [8] and the blr2358 mutant of this study exhibited a clear phenotype, suggesting that homologs may be implicated in processes other than EPS synthesis. Further research on different Bradyrhizobium strains will be required to investigate the roles of Blr2358 and homologs in polysaccharide synthesis and other processes under free living conditions and in symbiosis with soybean.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/microorganisms9122490/s1, Figure S1. Photograph of ethanol precipitates obtained from culture supernatants of B. diazoefficiens USDA110, the blr2358 mutant and a complementation strain; Figure S2: MALDI-TOF spectrum of oligosaccharides of B. diazoefficiens USDA110 derived from the ethanol-precipitated polysaccharide; Table S1; Statistics of whole-genome short read resequencing results obtained from Tn5 mutant colonies. Table S2: Predicted and observed mass spectrometry signals obtained from the MALDI-TOF spectrum shown in Figure S2. Table S3: List of glycosyltransferase homologs of blr2358 in the Bradyrhizobium genus; Table S4: Oligonucleotides used in this study. Table S5: Bacterial strains and plasmids used in this study.
Author Contributions: Conceptualization and original draft preparation, C.X., H.R., W.C., C.S. and W.D.; investigation and visualization, C.X., H.R., W.C., C.S. and W.D.; writing, review, and editing, C.S. and W.D.; supervision, project administration, and funding acquisition, C.S. and W.D. All authors have read and agreed to the published version of the manuscript.