Protein glycosylation by NleB and the secretion of therapeutic nanobody fusions

Abstract

Escherichia coli and Salmonella enterica are Gram-negative bacteria that are among the leading causes of gastrointestinal tract infections. These bacteria interact with mammalian hosts by using a type III secretion system (T3SS) to secrete effectors into host cells. NleB and SseK are two E. coli and Salmonella effector orthologs that are unusual glycosyltransferases. They glycosylate host protein substrates on arginine residues with N-acetyl glucosamine (GlcNAc) to inhibit the function of host proteins involved in the innate immune response. Originally, it was thought that the effectors are inactive within the bacterium and fold into their active conformations only after being secreted. However, mass spectrometry experiments to identify glycosylation substrates of NleB orthologs challenged this dogma, providing the premise for the first part of this thesis. Mass spectrometry suggested that the septum site-determining protein used in cell division, MinC, and the peptidyl-prolyl cis-trans isomerase protein that promotes bacterial stress tolerance and virulence, FklB, were glycosylated by NleB on arginine residues R107 and R129, respectively. Data from in vitro assays suggested that NleB glycosylated the arginine residues of both bacterial substrates, but when tested further, MinC and FklB were not confirmed to be glycosylated in Citrobacter rodentium, S. enterica, enteropathogenic E. coli (EPEC), or enterohemorrhagic E. coli (EHEC). This could be due to the MinC and FklB bacterial proteins being false positives when the mass spectrometry experiments were completed. Despite this negative data, intra-bacterial glycosylation of the mammalian substrates, the fas-associated protein with death domain (FADD), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and tumor necrosis factor receptor type 1-associated death domain protein (TRADD), were then studied. Results detected that all three mammalian substrates were glycosylated by NleB on arginine residues R177, R197/R200, and R235, respectively, and that FADD was a stronger glycosylation target of NleB than GAPDH and TRADD. The second part of this thesis also explores secretion through the T3SS, but instead of secreting a whole effector, camelid nanobodies were fused with signal sequences 20 amino acids in length from effectors that are naturally secreted and examined. An attenuated bacterium with an engineered T3SS was created to secrete the nanobody fusions, and in vivo analyses before and after precipitation were completed using C. rodentium and E. coli. Nanobodies fused with effector signal sequences upstream from the camelid signal sequences were expressed, but not secreted through the T3SS. This concludes that a 20 amino acid signal sequence does not promote the secretion of the nanobodies through a T3SS.