Role of toll-like receptors in host responses to mucosal bacterial infections

Sammanfattning: The first contact between invading bacteria and the host occurs at the mucosal surface. Interestingly, the mucosal linings of different organs are exposed to bacteria to various degrees. The intestinal epithelium is continuously exposed to a vast number of commensal bacteria that must be tolerated, while an immune response must be initiated once pathogens enter the gastro-intestinal tract. The lung and the urinary tract are considered as sterile compartments, and as such they consider all bacteria as pathogens. In my thesis I have studied by which mechanisms human epithelial cells from different organs detect microbial infections. We have primarily compared epithelial cells from the bladder, which occasionally are exposed to bacteria, with renal epithelial cells that should remain sterile. Renal epithelial cells responded to bacterial attachment, while bladder epithelial cells responded to lipopolysaccharide (LPS). We could correlate the different LPS- responsiveness to whether the organs expressed the LPS receptor Toll-like receptor (TLR) 4. LPS signaling in bladder epithelial cells was indeed found to be TLR4-dependent, but the signaling also required the coreceptor sCD14. Expression of TLR4 allows rapid microbial detection but may also confer unwanted inflammatory responses in some organs e.g. the kidney. Therefore, all epithelial cells do not express TLR4. Toll was originally identified in Drosophila where it binds the cytokine- like peptide Spaetzle rather than a microbial product. Because Spaetzle is processed by a serine protease, we investigated whether proteases are important for LPS signaling in human cells as well. Three serine protease inhibitors were found to prevent LPS signaling, however, they also inhibited Il- 1 beta induced signaling, indicating an intracellular target. We identified the target for these inhibitors to be the proteasome, which is required for IkappaB degradation. The endotoxic features of LPS are in large attributed to lipid A and its acylation-state. We found that human bladder epithelial cells only responded to stimulation by hexa-acylated Escherichia coli LPS, and that penta-acylated LPS inhibited this signaling, probably by competing for binding to CD14. Bacterial pathogens that cause chronic infections may avoid recognition by the innate immune system by altering the acylation-state of lipid A. This is exemplified by Helicobacter pylori whose tetra-acylated lipid A is not detected by TLR4. Moreover, the induction of pro- inflammatory genes by H. pylori was strictly dependent on the presence of the cag pathogenicity island. We found that H. pylori induced a different cytokine response in primary cells than in gastric cell lines. While gastric cell lines exclusively produced IL-8 in response to H. pylori infection, primary cells also produced IL-6 and TNF alpha. Primary cells were able to regulate their cytokine production in response to infection. This may be an important trait to limit the inflammatory response to avoid excessive tissue damage. In my thesis I have shown that epithelial cells use different mechanisms to detect microbial infections, and that bacteria may alter their surface structures in order to modulate host recognition.