Cholesterol efflux stimulates metalloproteinase-mediated cleavage of occludin and release of extracellular membrane particles containing its C-terminal fragments.

That changes in membrane lipid composition alter the barrier function of tight junctions illustrates the importance of the interactions between tetraspan integral tight junction proteins and lipids of the plasma membrane. Application of methyl-beta-cyclodextrin to both apical and basolateral surfaces of MDCK cell monolayers for 2 h, results in an approximately 80% decrease in cell cholesterol, a fall in transepithelial electrical resistance, and a 30% reduction in cell content of occludin, with a smaller reduction in levels of claudins-2, -3, and -7. There were negligible changes in levels of actin and the two non-tight junction membrane proteins GP-135 and caveolin-1. While in untreated control cells breakdown of occludin, and probably other tight junction proteins, is mediated by intracellular proteolysis, our current data suggest an alternative pathway whereby in a cholesterol-depleted membrane, levels of tight junction proteins are decreased via direct release into the intercellular space as components of membrane-bound particles. Occludin, along with two of its degradation products and several claudins, increases in the basolateral medium after incubation with methyl-beta-cyclodextrin for 30 min. In contrast caveolin-1 is detected only in the apical medium after adding methyl-beta-cyclodextrin. Release of occludin and its proteolytic fragments continues even after removal of methyl-beta-cyclodextrin. Sedimentation and ultrastructural studies indicate that the extracellular tight junction proteins are associated with the membrane-bound particles that accumulate between adjacent cells. Disruption of the actin filament network by cytochalasin D did not diminish methyl-beta-cyclodextrin-induced release of tight junction proteins into the medium, suggesting that the mechanism underlying their formation is not actin-dependent. The 41- and 48-kDa C-terminal occludin fragments formed during cholesterol depletion result from the action of a GM6001-sensitive metalloproteinase(s) at some point in the path leading to release of the membrane particles.

Knockdown of occludin expression leads to diverse phenotypic alterations in epithelial cells.

The function of occludin (Occ) in the tight junction is undefined. To gain insight into its role in epithelial cell biology, occludin levels in Madin-Darby canine kidney II cells were suppressed by stably expressing short interfering RNA. Suppression of occludin was associated with a decrease in claudins-1 and -7 and an increase in claudins-3 and -4. Claudin-2 levels were unaffected. The tight junction "fence" function was not impaired in suppressed Occ (Occ-) clones, as determined by BODIPY-sphingomyelin diffusion in the membrane. The most striking changes were those related to control of the cytoskeleton and the "gate" function of tight junctions. A reduced ability of Occ- clones to extrude apoptotic cells from the monolayers suggested that neighbors of apoptotic cells either failed to sense their presence or were unable to coordinate cytoskeletal activity necessary for their extrusion. To further test the extent to which actin cytoskeletal activity depends on the presence of occludin, Occ- and Occ+ monolayers were depleted of cholesterol. Previous studies showed that cholesterol depletion is associated with reorganization of the actin cytoskeleton and a fall in transepithelial electrical resistance. In contrast to control Occ (Occ+) cells, transepithelial electrical resistance did not fall significantly in cholesterol-depleted Occ- monolayers and they failed to generate Rho-GTP, one of the signaling molecules involved in regulating the actin cytoskeleton. While steady-state transepithelial electrical resistance was similar in all clones, tight junction permeability to mono- and divalent inorganic cations was increased in Occ- monolayers. In addition, there was a disproportionately large increase in permeability to monovalent organic cations, up to 6.96 A in diameter. Chloride permeability was unaffected and there was little change in mannitol flux. The data suggest that occludin transduces external (apoptotic cells) and intramembrane (rapid cholesterol depletion) signals via a Rho signaling pathway that, in turn, elicits reorganization of the actin cytoskeleton. Impaired signaling in the absence of occludin may also alter the dynamic behavior of tight junction strands, as reflected by an increase in permeability to large organic cations; the permeability of ion pores formed of claudins, however, is less affected.

Lung dendritic cells are primed by inhaled particulate antigens, and retain MHC class II/antigenic peptide complexes in hilar lymph nodes for a prolonged period of time.

Intratracheal (IT) administration of heat-killed Listeria monocytogenes (HKL) results in an influx of macrophage and dendritic cell (DC) precursors into the lung interstitium. Low-density, FcR+, interstitial lung cells isolated from rats instilled 24 hr before with HKL or vehicle alone, were > 90% Mar1+. After culturing with granulocyte-macrophage colony-stimulating factor (GM-CSF) for 3 days, up to 24% of the loosely adherent cells were DC that stimulated allogeneic T-cell proliferation in an mixed lymphocyte reaction (MLR) assay. After only an overnight incubation with GM-CSF, however, the capacity of interstitial Mar1+ cells to stimulate HKL immune T-cell proliferation without exogenous antigen was low. By contrast, when DC were isolated as major histocompatibility complex (MHC) class II+ cells from rat lungs at 1, 3, 7 and 14 days after HKL instillation and cultured overnight with GM-CSF, their antigen presentation capacity without added exogenous antigen was robust, but declined over the 2-week period. Interestingly, hilar lymph node DC maintained their HKL antigen-presenting capacity for up to 2 weeks after instillation of HKL. Following IT administration of PKH-26 labelled HKL, fluorescent or immunolabelled organisms were detected in OX62+ DC in airway epithelium, lung interstitium and hilar lymph nodes in situ and in MHC class II+ DC isolated from these sites. We conclude that newly immigrated Mar1+ lung DC precursors, while efficient in endocytosing particulate antigens, are incapable of eliciting a significant proliferative response from HKL-sensitized T cells. By contrast, MHC class II+ DC isolated from lungs and incubated overnight with GM-CSF induce vigorous antigen-specific T-cell proliferation. Antigen-loaded lung DC in hilar lymph nodes maintain their antigen presentation capacity for up to 2 weeks.

Matrix metalloproteinase-9-deficient dendritic cells have impaired migration through tracheal epithelial tight junctions.

When sampling inhaled antigens, dendritic cells (DC) must penetrate the tight junction (TJ) barrier while maintaining the TJ seal. In matrix metalloproteinase (MMP)-9-deficient mice, in vivo experiments suggest that migration of DC into air spaces is impaired. To examine the underlying mechanisms, we established a well-defined in vitro model using mouse tracheal epithelial cells and mouse bone marrow DC (BMDC). Transmigration was elicited with either macrophage inflammatory protein (MIP)-1alpha or MIP-3beta in a time-dependent manner. Control MMP-9(+/+) BMDC cultured with granulocyte macrophage-colony-stimulating factor for 7 d showed a 30-fold greater transepithelial migration toward MIP-3beta than MIP-1alpha, indicating a more mature DC phenotype. MMP-9(-/-) BMDC as well as MMP-9(+/+) BMDC in the presence of the MMP inhibitor GM6001, although showing a similar preference for MIP-3beta, were markedly impaired in their ability to traverse the epithelium. Expression levels of CCR5 and CCR7, however, were similar in both MMP-9(-/-) and MMP-9(+/+) BMDC. Expression of the integral TJ proteins, occludin and claudin-1, were examined in BMDC before and after transepithelial migration. Interestingly, occludin but not claudin-1 was degraded following transepithelial migration in both MMP-9(-/-) and control BMDC. In addition, there was a > 2-fold increase in claudin-1 expression in MMP-9(-/-) as compared with control BMDC. These observations indicate that occludin and claudin-1 are differentially regulated and suggest that the lack of MMP-9 may affect claudin-1 turnover.