For strain 3841, mutation of flaE and flaH resulted in a reduction in swimming motility,

suggesting that these subunits probably contribute to the flagellar filament. However, FlaE and FlaH peptides were not detected in the wildtype flagellar preparations, indicating Liproxstatin-1 ic50 that these peptides may not be stable under the conditions used. Glycosylation of flagellin subunits We observed that for strain 3841, both the upper and the lower bands on the protein gel contained the same set of flagellin subunits (FlaA, FlaB, and FlaC) (Table 3). The molecular masses (around 35kDa; Additional file 3) of the bands observed on the gel also appeared to be higher than the predicted molecular masses (31kDa) for FlaA and FlaB. This suggests that at least FlaA and FlaB may have undergone post-translational modification, resulting in a higher molecular weight and subsequently slower Protein Tyrosine Kinase inhibitor migration in the protein gel. Analysis

of the flagellin amino acid sequences of R. leguminosarum (Fig. 1 &2) revealed the presence of two to four putative glycosylation signals (N-X-S/T, where X is any amino acid except proline) [55]. The MS/MS spectral data for the identified peptides containing the glyosylation signal were also analyzed for the presence of glycosylation, based on the presence of peaks (m/z) corresponding to Temozolomide in vivo different types of glycosylation (Additional file 4 shows a sample of a MS/MS spectrum). However, we have not identified any potential glycosylation for these peptides which may be attributed to the lability of this modification

[56, 57]. Also, sequence coverage only ranged from 18% to 46% (Fig. 1 and 2) and peptides at the C-termini of the flagellin subunits were not detected. The C-terminus contains a common glycosylation 6-phosphogluconolactonase site for the R. leguminosarum flagellin subunits but these glycosylations were not detected in the MS/MS analysis, which could be due to the above reason. Thus, we performed glycoprotein staining to determine if the flagellins are post-translationally modified by glycosylation. We observed positive staining for the flagellins of both VF39SM and 3841 suggesting that these flagellins are glycosylated (Fig. 6). We were unable to determine which flagellins are glycosylated because the seven flagellins were not separated on the protein gel. Glycosylation of flagellins has been reported in a number of animal and plant pathogens including Campylobacter jejuni [56, 57], Helicobacter pylori [57, 58], Pseudomonas aeruginosa [59, 60], Pseudomonas syringae [61, 62], Listeria monocytogenes [63, 64], A. tumefaciens [6], Acidovorax avenae [65], as well as in the nitrogen-fixing bacterium Azosprillum brasilense [66]. It has been suggested that glycosylation may play a role in flagellar filament assembly and in pathogenesis [67, 68].