The rkp-1 cluster is required for secretion of Kdo homopolymeric capsular polysaccharide in Sinorhizobium meliloti strain Rm1021

Abstract

Under conditions of nitrogen stress, leguminous plants form symbioses with soil bacteria called rhizobia. This partnership results in the development of structures called root nodules, in which differentiated endosymbiotic bacteria reduce molecular dinitrogen for the host. The establishment of rhizobium-legume symbioses requires the bacterial synthesis of oligosaccharides, exopolysaccharides, and capsular polysaccharides. Previous studies suggested that the 3-deoxy-D-manno-oct-2-ulopyranosonic acid (Kdo) homopolymeric capsular polysaccharide produced by strain Sinorhizobium meliloti Rm1021 contributes to symbiosis with Medicago sativa under some conditions. However, a conclusive symbiotic role for this polysaccharide could not be determined due to a lack of mutants affecting its synthesis. In this study, we have further characterized the synthesis, secretion, and symbiotic function of the Kdo homopolymeric capsule. We showed that mutants lacking the enigmatic rkp-1 gene cluster fail to display the Kdo capsule on the cell surface but accumulate an intracellular polysaccharide of unusually high M(r). In addition, we have demonstrated that mutations in kdsB2, smb20804, and smb20805 affect the polymerization of the Kdo homopolymeric capsule. Our studies also suggest a role for the capsular polysaccharide in symbiosis. Previous reports have shown that the overexpression of rkpZ from strain Rm41 allows for the symbiosis of exoY mutants of Rm1021 that are unable to produce the exopolysaccharide succinoglycan. Our results demonstrate that mutations in the rkp-1 cluster prevent this phenotypic suppression of exoY mutants, although mutations in kdsB2, smb20804, and smb20805 have no effect.

FIG. 1.

Analysis of capsular polysaccharides from wild-type S. meliloti Rm1021 and mutants affecting the rkp-1 gene cluster. Wild-type and mutant strains were cultured in the presence of [U-¹⁴C]glucose, and cell surface polysaccharides were isolated via EDTA-TEA extraction, fractionated using Tris-Tricine SDS-PAGE, and detected by staining with Alcian blue (A) or by phosphorimaging (B) as described in Materials and Methods. Lane 1, strain MGM044 (wild-type Rm1021/pTE3 [WT/pTE3]); lane 2, MGM312 (WT/pMW23); lane 3, MGM050 (rkpA::Nm/pTE3); lane 4, MGM314 (rkpA::Nm/pMW23); lane 5, MGM048 (rkpG::pDW33/pTE3); lane 6, MGM388 (rkpG::pDW33/pMW23); lane 7, MGM454 (rkpH::pDW33/pTE3); lane 8, MGM455 (rkpH::pDW33/pMW23); lane 9, MGM423 (rkpI::pDW33/pTE3); lane 10, MGM424 (rkpI::pDW33 pMW23); lane 11, MGM425 (rkpJ::pDW33/pTE3); and lane 12, MGM426 (rkpJ::pDW33/pMW23). The presence of the plasmid containing rkpZ (pMW23) is indicated beneath each lane with a plus, while the presence of the control plasmid (pTE3) is indicated with a minus. The open triangle refers to the interface between the stacking and separating gel. (C) Organization of the rkp-1 gene cluster in S. meliloti Rm1021. Genes are represented as arrows, with the length of the intergenic regions depicted in base pairs. HMW, high molecular weight; LMW, low molecular weight.

FIG. 2.

Expression of rkp-1 cluster genes. Transcriptional fusions between the rkpG, rkpH, rkpI, and rkpJ genes and uidA (encoding β-glucuronidase) were constructed using plasmid pDW33. The strains were grown to stationary phase (optical density at 600 nm of ∼2.5) and assayed for activity as described in Materials and Methods. Enzyme activity is indicated in Miller units. Error bars represent standard deviations from experiments carried out in triplicate. Lane 1, MGM050 (rkpA::Nm/pTE3); lane 2, MGM048 (rkpG::pDW33/pTE3); lane 3, MGM400 (rkpG::pDW33 rkpA::Nm/pTE3); lane 4, MGM454 (rkpH::pDW33/pTE3); lane 5, MGM465 (rkpH::pDW33 rkpA::Nm/pTE3); lane 6, MGM423 (rkpI::pDW33/pTE3); lane 7, MGM429 (rkpI::pDW33 rkpA::Nm/pTE3); lane 8, MGM425 (rkpJ::pDW33/pTE3); and lane 9, MGM431 (rkpJ::pDW33 rkpA::Nm/pTE3). Plasmid pTE3 had no effect on the expression of any of the rkp::uidA fusions (M. G. Müller et al., unpublished).

FIG. 3.

Mutants deficient in rkpA but harboring pMW23 produce a very-high-molecular-weight polysaccharide. (A) Wild-type and mutant strains, harboring the vector control or pMW23, were cultured in the presence of [U-¹⁴C]glucose. Polysaccharides then were extracted via phenol-water, fractionated by Tris-Tricine PAGE, and visualized by phosphorimaging. Lane 1, MGM044 (wild-type Rm1021/pTE3 [WT/pTE3]); lane 2, MGM312 (WT/pMW23); lane 3, MGM050 (rkpA::Nm/pTE3); lane 4, MGM314 (rkpA::Nm/pMW23); lane 5, MGM404 (kdsB2::pDW33/pTE3); lane 6, MGM405 (kdsB2::pDW33/pMW23); lane 7, MGM383 (smb20805::pDW33/pTE3); and lane 8, MGM385 (smb20805::pDW33 ::pDW33/pMW23). In panels A and B the presence of the plasmid containing rkpZ (pMW23) is indicated beneath each lane with a plus, and the presence of the control plasmid (pTE3) is indicated with minus. The open triangle refers to the interface between the stacking and separating gel. (B) Cells were treated with EDTA-TEA and polysaccharides from both the EDTA-TEA-soluble fractions (lanes 1 to 5) and EDTA-TEA-insoluble fractions (lanes 6 to 10) and then were extracted via phenol-water, fractionated by Tris-Tricine PAGE, and visualized by phosphorimaging. Lanes 1 and 6, MGM044; lanes 2 and 7, MGM312; lanes 3 and 8, MGM314; lanes 4 and 9, MGM405; and lanes 5 and 10, MGM385. HMW, high molecular weight; LMW, low molecular weight.

FIG. 4.

Cellular location of very-high-molecular-weight polysaccharides. (A) Analysis of periplasmic contents of wild-type and rkpA::Nm cells. Periplasmic polysaccharides were liberated from wild-type and mutant strains using osmotic shock as described in Materials and Methods. Polysaccharides recovered from both the soluble (lanes 1 to 4) and insoluble osmotic shock fractions (lanes 5 to 8) then were subjected to Tris-Tricine SDS-PAGE, and the polysaccharides were visualized by staining with Alcian blue. Lanes 1 and 5, strain MGM044 (wild-type Rm1021/pTE3 [WT/pTE3]); lanes 2 and 6, MGM312 (WT/pMW23); lanes 3 and 7, MGM050 (rkpA::Nm/pTE3); and lanes 4 and 8, MGM314 (rkpA::Nm/pMW23). The presence of the plasmid containing rkpZ (pMW23) is indicated beneath each lane with a plus, and the presence of the control plasmid (pTE3) is indicated with a minus. The open triangle refers to the interface between the stacking and separating gel. (B) Subcellular fractionation of very-high-molecular-weight capsular polysaccharide extracted from insoluble osmotic shock fractions. Wild-type and mutant cells were treated with osmotic shock. The insoluble pellets then were exposed to freeze-thaw followed by ultracentrifugation. Polysaccharides were extracted from the resulting pellet fraction via phenol-water extraction (lanes 5 to 8) or supernatants were left untreated (lanes 1 to 4). Finally, polysaccharides from both ultracentrifugation fractions were fractionated using Tris-Tricine SDS-PAGE and stained with Alcian blue. Lanes 1 and 5, strain MGM044; lanes 2 and 6, MGM312; lanes 3 and 7, MGM050; and lanes 4 and 8, MGM314. HMW, high molecular weight; LMW, low molecular weight.

FIG. 5.

Compositional analysis of very high-molecular-weight (HMW) polysaccharide. Polysaccharide was prepared from the wild type and rkpA mutants harboring plasmid pMW23. (A) Fractionation of polysaccharide extract derived from Rm1021 containing pMW23 by size-exclusion chromatography. Chromatography was performed in the presence of detergent (dissociative conditions) on a column of Sephacryl S-400 HR. The high-molecular-weight polysaccharide (fractions 15 to 26) migrates at the column void volume, indicating an approximate molecular weight of ≥1 × 10⁶ to 2 × 10⁶. LPS migrated in fractions 37 to 53 as revealed by glycosyl composition analysis (M. G. Müller et al., unpublished). Abs, absorbance. (B) Compositional analysis of polysaccharide isolated from void volume (Vo). The molar ratios of glycosyl residues were normalized such that the amount of mannose was equivalent in both the Rm1021 and rkpA::Nm samples.

FIG. 6.

Analysis of capsular polysaccharides from wild-type S. meliloti and kdsB2, smb20804, and smb20805 mutant strains. Wild-type and mutant strains were cultured in liquid medium, cell surface polysaccharides then were isolated via EDTA-TEA extraction and fractionated using Tris-Tricine SDS-PAGE, and polysaccharides were visualized by staining with Alcian blue, as described in Materials and Methods. Lane 1, strain MGM044 (wild-type Rm1021/pTE3 [WT/pTE3]); lane 2, MGM312 (WT/pMW23); lane 3, MGM404 (kdsB2::pDW33/pTE3); lane 4, MGM405 (kdsB2::pDW33/pMW23); lane 5, MGM433 (smb20804::pDW33/pTE3); lane 6, MGM434 (smb20804::pDW33/pMW23); lane 7, MGM383 (smb20805::pDW33/pTE3); and lane 8, MGM385 (smb20805::pDW33 ::pDW33/pMW23). The presence of the plasmid containing rkpZ (pMW23) is indicated beneath each lane with a plus, and the presence of the control plasmid (pTE3) is indicated with a minus. HMW, high molecular weight; LMW, low molecular weight.

FIG. 7.

Nitrogen-fixing ability of wild-type and mutant strains affected in capsule biosynthesis. Sterile M. sativa seedlings were placed onto Jensen agar slants. Plants were inoculated 48 h after planting with wild-type S. meliloti or mutant strains harboring either vector control (pTE3) or pMW23. Plants were harvested 80 days postinoculation, and their dry weight was measured. Lane 1, strain MGM044 (wild-type Rm1021/pTE3 [WT/pTE3]); lane 2, MGM312 (WT/pMW23); lane 3, MGM050 (rkpA::Nm/pTE3); lane 4, MGM314 (rkpA::Nm/pMW23); lane 5, MGM404 (kdsB2::pDW33/pTE3); lane 6, MGM405 (kdsB2::pDW33/pMW23); lane 7, MGM359 (exoY::pDW33/pTE3); lane 8, MGM361 (exoY::pDW33/pMW23); lane 9, MGM355 (exoY::pDW33 rkpA::Nm/pTE3); lane 10, MGM357 (exoY::pDW33 rkpA::Nm/pMW23); lane 11, MGM352 (exoY::pVO155 kdsB2::pDW33/pTE3), lane 12, MGM354 (exoY::pVO155 kdsB2::pDW33/pMW23); and lane 13, sample mock inoculated with sterile 10 mM MgSO₄. The presence of the plasmid containing rkpZ (pMW23) is indicated beneath each lane with a plus, and the presence of the control plasmid (pTE3) is indicated with a minus. The exoY::pDW33 mutant and exoY::pVO155 mutant behave identically under all conditions tested (M. G. Müller et al., unpublished). Similar results were observed in an additional, independent experiment (M. G. Müller et al., unpublished).