Cell-matrix interactions in gland development in the lung.

The tubulo-acinar glands of the conducting airways participate in lung host defense and the pathogenesis of chronic inflammatory airway disease. Models of gland development may provide clues to mechanisms that promote hypersecretion in diseases such as chronic bronchitis and cystic fibrosis. The developmental anatomy of airway glands was described 30 years ago, but the regulation of gland formation remains a mystery. During the pseudoglandular stage of lung development, poorly differentiated surface epithelial cells grow radially from the lumen and invade the airway submucosa. The basal lamina is remodeled during this process and animal studies indicate a correlation between the expression of the 72-kD gelatinase (MMP-2) and gland morphogenesis. Cell culture models with lung epithelial and mesenchymal elements mimic early gland morphogenesis. Studies using this model have corroborated the involvement of 72-kD gelatinase in the formation of gland-like structures. While the secretion and activation of 72-kD gelatinase is correlated with epithelial matrix invasion, the presence of activated 72-kD gelatinase is not by itself sufficient to permit epithelial matrix invasion. Whether the 72-kD gelatinase participates in a more complex proteolytic cascade or whether appropriate localization of the enzyme is critical to airway epithelial morphogenesis is the focus of ongoing investigation.

Human fetal lung fibroblasts promote invasion of extracellular matrix by normal human tracheobronchial epithelial cells in vitro: a model of early airway gland development.

Epithelial invasion of extracellular matrix (ECM) is important during lung development and the pathogenesis of bronchogenic neoplasms. Airway submucosal gland development begins when clusters of surface epithelial cells invade the lamina propria between the tenth and thirtieth weeks of fetal life. The factors regulating this transient normal invasive behavior are unknown. We observed that normal human tracheobronchial epithelial (HTBE) cells from adult necropsy specimens penetrate collagen matrices when co-cultured with human fetal lung fibroblasts (HFLF). Invading clusters of epithelial cells resembled primordial glands, forming tubular structures and undergoing dichotomous branching. Using 48 different tracheobronchial specimens, we compared paired cultures of HTBE cells without (control) and with HFLF co-culture. Fixed, vertically sectioned culture substrates were examined, and invaginated epithelial cell clusters as well as total invading HTBE cells were counted. The co-cultured condition resulted in significantly more epithelial invaginations per two sections (4 +/- 1 versus 22 +/- 3, P < 0.0005) and more invading HTBE cells per two sections (31 +/- 9 versus 194 +/- 27, P < 0.0005) than controls. Epithelial invasion was noted by 36 h in culture and was greatest at HTBE cell/HFLF ratios near 1 and HFLF passages between 10 and 16. Epithelial cells co-cultured with a fibroblast cell line derived from the adult bronchiole showed no increase in ECM invasion compared with controls. These results demonstrate that fetal mesenchymal cells are capable of promoting invasive behavior in mature epithelial cells in vitro. Given the fibroblast type and passage specificity, this model should prove useful for investigating the cellular and molecular regulation of epithelial ECM invasion.

Induction of ICAM-1 expression on human airway epithelial cells by inflammatory cytokines: effects on neutrophil-epithelial cell adhesion.

Inflammation of the human airways in diseases such as chronic bronchitis, cystic fibrosis with Pseudomonas endobronchial infection, and possibly asthma during late-phase reactions involves a local influx of neutrophils (PMN) that may participate in airway epithelial injury. PMN-mediated cellular injury is most efficient under conditions of PMN-target cell adhesion. PMN express adhesive glycoproteins of the CD11/CD18 family that are counter-receptors for intercellular adhesion molecule-1 (ICAM-1), found on various cell types. We proposed that adherence by PMN to human airway epithelial cells via ICAM-1 might be an important mechanism in inflammatory airway diseases. We found that although PMN adhere poorly (less than 5%) to monolayers of human tracheal epithelial cells (TEC) in primary culture, they adhere readily (45 to 50%) to an SV40-immortalized line of human TEC, designated 9HTEo-. We also found 6-fold greater surface expression of ICAM-1 on 9HTEo- compared with primary TEC. Blocking surface ICAM-1 on 9HTEo- cells with specific monoclonal antibody inhibited PMN adherence by about 50%. Thus, ICAM-1 plays a major role in this adherence, although it is possible that other epithelial ligands contribute also. Antibodies to CD11a, CD11b, and CD18 on PMN also inhibited PMN-epithelial adherence. Treatment of primary TEC monolayers with the proinflammatory cytokines interleukin-1 (IL-1) or tumor necrosis factor-alpha (TNF-alpha) caused a 3- to 4-fold increase in both cell surface ICAM-1 expression and support of PMN adhesion.(ABSTRACT TRUNCATED AT 250 WORDS)

Human tracheobronchial epithelial cells direct migration of lung fibroblasts in three-dimensional collagen gels.

Normal airway morphogenesis and repair after injury depend in part on the interaction between the mesenchymal and epithelial cells in the tracheobronchial tree. We cultured human lung fibroblasts between layers of type I collagen gel and examined sections through these three-dimensional matrices to assess fibroblast migration. The migration assay used in these experiments allowed simultaneous assessment of directed and random fibroblast migration as well as cell number. We tested the hypothesis that human tracheobronchial epithelial (HTBE) cells direct the migration of fibroblasts. When fibroblasts were cultured alone, migration was nearly equivalent in the upper and lower collagen layers. When HTBE cells were plated on the upper collagen lattice, there was a net migration of fibroblasts toward the HTBE cells. The differential migration was evident early in culture but became maximal after 1 wk. Differences increased at higher HTBE cell inoculation densities. No epithelial chemokinetic or mitogenic influence was evident: total cell migration and total fibroblast number were not significantly different between the control and coculture sections. HTBE fibronectin production may contribute to directed migration because fibronectin, added to the upper lattice, reproduced a portion of the directed migration seen in coculture. Our data support the hypothesis that epithelial cells direct fibroblast migration.

Adenosine 3:5' cyclic monophosphate synthesis by human tracheal epithelial cells.

Intracellular cyclic AMP (cAMP) regulates many critical differentiated functions of tracheal epithelial cells. An in vitro model system for reliable study of cAMP metabolism in these cells has been developed. Viable tracheal epithelial cells could be recovered from greater than 50% of necropsy specimens. Culture success rate was not significantly affected by age of subject, endotracheal intubation, or time between death and autopsy, although most specimens were obtained within 24 h of death. Human tracheal epithelial cells grown in primary culture displayed a typical histologic epithelial appearance, and the ultrastructure showed microvilli, junctional complexes, and tonofilaments. The cells uniformly stained with fluorescent antibody to cytokeratin, and expressed receptors for isoproterenol and vasoactive intestinal peptide. Human tracheal epithelial cells grown serum-free in an equal volume mix of Ham's F12 medium and Dulbecco's minimal essential medium containing growth supplements (Medium A) and cholera toxin (CT) had higher basal cAMP levels and greater increase in intracellular cAMP in response to phosphodiesterase inhibition than cells grown in Medium A without CT. Cells grown in Medium A without CT had similar morphology and grew at a comparable rate but attached to the culture substratum less readily than cells grown in Medium A with CT. Cells grown in Medium A without CT had less cAMP response to phosphodiesterase inhibition, less rapid accumulation of cAMP, and greater proportional response to receptor-mediated stimulation of cAMP production compared to cells grown with CT, though the final cAMP levels achieved were comparable.(ABSTRACT TRUNCATED AT 250 WORDS)