Role of human bronchial epithelium in eosinophilic inflammation and mucosal immunity : from basic to clinical aspects

Far from being a lining structure with a role of mechanical barrier, the human bronchial epithelium is a highly differentiated and diversified tissue interacting with the respiratory tract and the environment. Among the functions of the bronchial epithelium in diseased airways, its role as a first line protagonist in bronchial inflammation and mucosal immunity is well established. In particular, the airway epithelium plays a central role in the regulation of eosinophilic inflammation in the respiratory tract and is responsible for the production of SC and active transport of dIgA into the airways secretions as S-IgA. Our initial studies have first demonstrated that eosinophils are capable of adhesion to respiratory epithelial cells in vitro. These adhesive interactions are significantly increased by activation of eosinophils and by incubation of the epithelial cells with pro-inflammatory cytokines. The adhesive interaction involves the CD18/CD11B (Mac-1) integrin receptor on the eosinophil whose expression has been shown to be upregulated by eosinophil activation. However, although pro-inflammatory cytokines clearly upregulate ICAM-1 expression on epithelial cells, no significant inhibition of eosinophil adhesion was observed in our first sets of experiments, when using a blocking ICAM-1 antibody, R6.5. These data have been confirmed by other investigators. Our studies of neutrophil and eosinophil adhesion to respiratory epithelial cells demonstrate that RSV infection of epithelial cells increases neutrophils, but not eosinophil adhesion. Activated eosinophils demonstrated increased adhesion to epithelial cells; the adhesion of activated eosinophils was further enhanced by RSV infection of epithelial cells. Consistently with our first set of experiments, activated eosinophil adhesion was mediated by Mac-1 integrin receptor on eosinophils since activated eosinophil adhesion was blocked by anti-CD18, but not by anti-CD11a. In contrast, in our study of eosinophil adhesion to RSV infected respiratory epithelial cells, PMA-activated eosinophil adhesion to uninfected epithelial cells was significantly blocked by R6.5. Our conclusion that adhesion of activated eosinophils to RSV-infected repiratory epithelial cells was ICAM-1 and CD18/CD11b dependent demonstrate an important role for ICAM-1 in activated eosinophil adhesion to epithelial cells. There is thus an apparent discrepancy on the role of epithelial ICAM-1 between the results from our two suty of PMA-activated eosinophil adhesion to epithelial cells. However, one needs to consider the following information. First, the protocol of the two steps of experiments was not strictly identical. Indeed, the timing of addition of PMA, used as an eosinophil activator, was different. In one study of eosinophil adhesion to cytokine treated epithelial monolayer, PMA was added to the epithelial monolayer first, before addition of eosinophils, while in the other study, the eosinophils were added first, then the cell preparation was exposed to PMA (2, 4, and Stark JM – personal communication). PMA, a potent activator of protein kinase C, has thus been in close contact with the epithelial cells in the first set of experiments. Second, in contrast with HBEC which have constitutive expression of ICAN-1, there is virtually no ICAM-1 expression on untreated 1549 cells. It is thus difficult to understand how entire ICAM-1 Mab blocks eosinophil adhesion to epithelial cells that do not express ICAM-1. However, activated eosinophils express ICAM-1 (CD54) which might in turn interact with a counter ligand on epithelial cells. Third, previous studies if neutrophil adhesion had shown that fMLP treatment of neutrophils although inducing a 4- to 6-fold increase of CD11b/CD18, failed to increase neutrophil adhesion to respiratory epithelial monolayers treated with cytokines expressing high levels of ICAM-1. Their findings raised the possibility that respiratory epithelial cells express an unrecognized adhesion molecule capable of supporting neutrophil adhesion via LFA-1 (CD11a/CD18), since fMLP does not affect the expression of CD11a by neutrophils. Indeed, in our studies of eosinophil adhesion, the addition of PMA to epithelial cells is a likely mechanism of inducing the expression of an unrecognized epithelial adhesion molecule binding to CD18/CD11b. Fourth, the animal model of ICAM-1 knock-out mouse demonstrates little alteration in granulocyte migration in contrast with a dramatic decrease in eosinophil infiltration in the lung during allergen challenge. These data confirm a central role from ICAM-1 in eosinophil infiltration in the lung after allergen challenge, but suggest the development of alternate pathways of neutrophil adhesion to epithelial cells. Fifth, we have used an in vitro model of neutrophil adhesion to a respiratory epithelial cell line which by no means reflects the complexity of physiopathologic mechanisms in vivo. Thus, our data, consistently with some evidence form other studies, support the hypothesis that PMA-activated eosinophils are capable to interact with epithelial cells in both ICAM-1-dependent and independent pathways. The data raise the hypothesis that another epithelial receptor different from ICAM-1 is present or inducible on the cell surface. Whether this epithelial receptor corresponds to an already known molecule or a new adhesion molecule remains to be elucidated. SC production an dIgA ou IgM transcytosis are major properties of the bronchial epithelial cells, regulating the unidirectional transport of locally produced immunoglobulins without disrupting the epithelial barrier. We have demonstrated that human bronchial epithelial cells in primary culture were capable of SC productionThèse de doctorat en sciences biomédicales (immunologie) -- UCL, 199