In both bacteria, the flagellins are essential virulence factors glycosylated by pseudaminic acid (PA). We have identified and disrupted genes required for PA synthesis in both bacteria, and shown that this affects flagellin production. Further analysis and glycoprotein staining revealed that in H. pylori, the PA pathway is necessary for the glycosylation of proteins other than flagellins and for the synthesis of additional virulence factors, including LPS and urease. Enzymatic deglycosylation analysis uncovered a second set of H. pylori proteins glycosylated with an unknown sugar synthesized by a PA-independent pathway. We have identified one as a membrane-associated isoform of an immunodominant antigen that protects H. pylori from oxidative stress.

Numerous C. jejuni glycoproteins also possess a heptasaccharide containing diacetamidobacillosamine (DAB). Our DAB biosynthesis mutant was deficient in at least one glycoprotein and was avirulent in a chicken model, underscoring the role of this pathway in virulence. We have shown that both PA and DAB pathways are present in C. jejuni cell extracts and are investigating through activity-based assays and glycoprotein blotting how different growth conditions related to pathogenesis affect the relative activities of both pathways.

Understanding how each pathway affects virulence will reveal the best targets for the development of glycosylation inhibitors to treat these major infections.

Protein glycosylation, or the post-translational modification of proteins by sugar molecules, plays an important role in this bacterium's virulence including affecting cell-host interactions, immunogenicity and pathogenicity. C jejuni harbours numerous proteins glycosylated by a heptasaccharide containing diacetamidobacillosamine. Our laboratory investigates the enzymes required to make this sugar as well as the glycoproteins on which this glycan is found.

Analysis of C jejuni proteins by Western blot suggests that the molecular chaperone GroEL is post-translationally modified. A stress-induced protein as well as a major immunogenic antigen in C jejuni infections, GroEL prevents misfolding and aggregation of partially denatured proteins through an ATP-dependent process and may play an essential role in intestinal tract colonization and bacterial survival at high temperatures. GroEL and other chaperones have been shown to be glycosylated in other bacteria including another Campylobacter species. It has also been proposed that C jejuni GroEL is glycosylated based on its reactivity with a lectin.

The focus of this work is to determine if GroEL is really glycosylated, identify the glycan present on GroEL and determine the function of the glycosylation modification on GroEL.

We are using ATP affinity followed by anion exchange chromatography to purify GroEL and will determine its exact molecular weight and the glycan attachment site by mass spectrometry (MS). Initial enzymatic deglycosylation using five different enzymes suggests that the glycan present on the protein is not recognized by the enzymes available. The glycan present will be identified by ion-exchange chromatography, MS and NMR. The effect of the sugar on the activity of GroEL will be ascertained. We will also determine the role of the glycosylated form of GroEL and its glycan in the virulence of C jejuni by looking at the ability of mutants with defects in either the expression of the glycoprotein or its glycosylation to adhere and invade intestinal epithelial cells.

This information will permit us to infer what enzymes are required to make these sugars, allowing the enzymes to later be targeted to inhibit protein glycosylation should glycosylation be shown to be important for virulence.

METHODS: All enzymes were overexpressed and purified before monitoring their activity by capillary electrophoresis. The cj1121c and cj1294 genes were disrupted by a chloramphenicol resistance cassette and virulence-related phenotypes were investigated.

RESULTS: We determined that Cj1293 is a UDP-GlcNAc C6 dehydratase. It leads to the formation of 4-keto-arabino and 4-keto-gluco intermediates. We showed that Cj1294 and Cj1121c are aminotransferases that use the arabino and gluco intermediates generated by Cj1293 as substrates, respectively. Both activities are present in C. jejuni extracts, with preponderance of the Cj1121c activity. We showed that Cj1123c is a N-acetyltransferase that uses the Cj1121c reaction product as a substrate. The cj1294 and cj1121c mutants are non-motile but the cj1121c mutant produces normal flagellins and flagella, whereas the cj1294 mutant does not. cj1121c is important for interactions with Caco2 cells whereas cj1294 is not. Finally, cj1121c is essential for colonization of chicken intestine and for the glycosylation of proteins other than flagellins.

DISCUSSION: Our data indicate that Cj1293 and Cj1294 are involved in flagellin glycosylation via PA biosynthesis, and that Cj1121c and Cj1123c are involved in general protein glycosylation via DAB biosynthesis. The data demonstrate a cross-talk between both pathways, with a preponderant role of Cj1121c on virulence, hence identifying Cj1121c as a target for inhibitor development.