Reinvestigation into the role of lipopolysaccharide Glycosyltransferases in Helicobacter pylori protein glycosylation.

Abstract

Protein glycosylation has been considered as a fundamental phenomenon shared by all domains of life. In Helicobacter pylori, glycosylation of flagellins A and B with pseudaminic acid have been rigorously confirmed and shown to be essential for flagella assembly and bacterial colonization. In addition to flagellins, several other proteins including RecA, AlpA/B, and BabA/B in H. pylori have also been reported to be glycosylated and to be dependent on the lipopolysaccharide (LPS) biosynthetic pathway. However, these proteins have not been purified for sugar-specific staining or structural analysis to confirm the existence of carbohydrate motifs. Here, using a combined approach of genetics, protein purification, and sugar-specific staining, we demonstrate that RecA is not a glycoprotein. Moreover, using LPS-protein reconstitution experiments, we demonstrate that the presence of O-antigen containing full-length LPS interferes with the electrophoretic mobility of H. pylori RecA and many other proteins including AlpA/B on SDS-PAGE. Finally, we demonstrate that full-length LPS extracted from E. coli affects electrophoretic migration of H. pylori proteins, while full-length LPS extracted from H. pylori similarly influences the electrophoretic migration of E. coli proteins. The impact is more subtle with E. coli LPS compared to H. pylori LPS, indicating that the magnitude of effect of LPS effects on protein mobility is dependent on bacterial source of the LPS. These findings suggest that the effects of full-length LPS on protein electrophoresis may represent a more general phenomenon. As LPS is a unique component of virtually all Gram-negative bacteria, our data suggest that when observing protein electrophoretic mobility shifts between wild-type and LPS mutant strains or between subcellular fractionation samples, the influence of LPS on protein electrophoretic migration should be considered first, rather than interpreting it as potential protein glycosylation that is dependent upon LPS biosynthetic pathway.