Structural Heterogeneity of Terminal Glycans in Campylobacter jejuni Lipooligosaccharides

Abstract

Lipooligosaccharides of the gastrointestinal pathogen Campylobacter jejuni are regarded as a major virulence factor and are implicated in the production of cross-reactive antibodies against host gangliosides, which leads to the development of autoimmune neuropathies such as Guillain-Barré and Fisher Syndromes. C. jejuni strains are known to produce diverse LOS structures encoded by more than 19 types of LOS biosynthesis clusters. This study demonstrates that the final C. jejuni LOS structure cannot always be predicted from the genetic composition of the LOS biosynthesis cluster, as determined by novel lectin array analysis of the terminal LOS glycans. The differences were shown to be partially facilitated by the differential on/off status of three genes wlaN , cst and cj1144-45. The on/off status of these genes was also analysed in C. jejuni strains grown in vitro and in vivo, isolated directly from the host animal without passaging, using immunoseparation. Importantly, C. jejuni strains 331, 421 and 520 encoding cluster type C were shown to produce different LOS, mimicking asialo GM 1 , asialo GM 2 and a heterogeneous mix of gangliosides and other glycoconjugates respectively. In addition, individual C. jejuni colonies were shown to consistently produce heterogeneous LOS structures, irrespective of the cluster type and the status of phase variable genes. Furthermore we describe C. jejuni strains (351 and 375) with LOS clusters that do not match any of the previously described LOS clusters, yet are able to produce LOS with asialo GM 2 -like mimicries. The LOS biosynthesis clusters of these strains are likely to contain genes that code for LOS biosynthesis machinery previously not identified, yet capable of synthesising LOS mimicking gangliosides.

Glycan:glycan interactions: High affinity biomolecular interactions that can mediate binding of pathogenic bacteria to host cells

“…3, Tables S1-S4, and Dataset S2). Interactions with K D values less than 50 μM are comparable to typical values observed for binding between lectins or antibodies and glycans (2,3,24). For S. flexneri, wild-type 2a LPS, the highest affinity interaction was observed with the A blood group antigen with a K D of 0.81 μM, whereas C. jejuni 11168 LOS bound with the highest affinity to the B blood group antigen (K D = 0.14 μM; Fig.…”

“…As well as the O-blood group antigen, Caco-2 cells also express GAGs (53), which would provide binding sites for 11168 ΔcgtA/neuB with higher affinity than the wild-type C. jejuni 11168. Unfortunately, direct comparison between the C. jejuni 81-176 ΔcgtA mutant and the C. jejuni 11168 ΔcgtA/neuB mutant used in this study is not possible, because the cgtA/neuB in C. jejuni 11168 is a bifunctional enzyme involved in the synthesis of CMP-Neu5Ac and the translocation of GalNAc (24). From the results obtained in this study for C. jejuni, it appears that the terminal galactose and N-acetylgalactosamine are required for high-affinity interactions with blood group and Lewis antigens, but actually inhibit interactions with GAGs (Table S1).…”

“…Arrays were performed with 1 μg of LOS and rough mutant LPS or 10 μg of LPS with arrays performed as described in Semchenko et al (24). Scanning was performed by using a Proscan four laser Microarray scanner and the Scanarray express software.…”

“…SPR was performed by using the Biacore T100 system and series S L1 sensor chips. LOS/LPS from isolated from wild-type and rough mutant strains of S. flexneri, C. jejuni, H. influenzae, and S. typhimurium were captured onto the L1 chip at concentrations of 0.75 mg/mL and using unglycosylated lipid A from E. coli as the negative control on flow cell 1 as described in Semchenko et al (24). The running buffer used was 1× PBS (containing 2 mM MgCl 2 , 2 mM CaCl 2 ) at a flow for analysis of 30 μL/min.…”

Deciphering the bacterial glycocode: recent advances in bacterial glycoproteomics

“…In their landmark work, Mahal and Hsu employed lectin arrays to distinguish strains of Escherichia coli and to observe temporal changes in glycosylation states across various growth phases[38,39]. More recently, lectin arrays have been used by Gao et al to study the effect of growth media on the glycosylation profile of E. coli [40], while Semchenko et al used lectin-based arrays to elucidate structural aspects of lipooligosaccharide in C. jejuni [41]. …”