Man Is not a rodent: aquaporins in the airways.

Abstract

The physiological significance of fluid homeostasis in the respiratory tract is readily apparent. The nasopharynx and upper airways must humidify the inspired airstream, as well as regulate the volume and composition of the airway surface layers. The distal lung must mobilize fluid at the time of birth in preparation for the transition to ex utero life and must handle a variety of challenges to fluid balance that could interfere with gas exchange throughout life. Disruption in water flux at these sites may contribute to the pathogenesis of rhinnorrhea, impaired mucociliary transport, exerciseor cold-induced asthma, and pulmonary edema. The molecular determinants of these processes in the respiratory tract therefore continue to be the focus of intense investigation. In this issue, Gynn and colleagues provide important new information about the expression of aquaporin water channel proteins in the human airway epithelium (1). Aquaporins (AQPs) are membrane channel proteins that are highly and, in most cases, specifically permeable to water. Ten mammalian AQPs have been identified to date, and homologues have been demonstrated at all levels of life, including bacteria, yeast, and plants (2). Several AQPs have been demonstrated to have permeabilities in addition to water. AQP3, AQP7, and AQP9 are permeable to small solutes (for example, glycerol), an observation whose functional significance is still undefined. Studies in Xenopus oocytes suggest that AQP1 is permeated by CO 2 (3). This finding was confirmed in proteoliposomes reconstituted with AQP1 (4), but was not observed in erythrocytes from AQP1-null mice lacking the protein (5). The magnitude of the CO 2 permeability is low compared to that of water, and the physiological significance is not yet clear. Studies of distribution and ontogeny in the rat respiratory tract established a network of four AQPs with nonoverlapping distribution (6–8). AQP1 is present in both the apical and basolateral membrane of microvascular endothelial cells and fibroblasts, while AQP3, AQP4, and AQP5 polarize to the apical or basolateral membrane at different sites in the respiratory epithelium. Curiously absent from the AQP network in the rat respiratory tract are AQPs on the apical membrane of nasopharyngeal or airway epithelium, or on the basolateral membrane of type I pneumocytes. Several explanations for these findings have been suggested: an unidentified AQP is expressed in those locations; water transport across the apical membrane occurs by non-AQP mediated mechanisms; or unilateral expression of an AQP in the epithelium suggests an alternative function besides transcellular water movement, for example cell volume regulation (8–10). Gynn and colleagues indicate that in the human respiratory tract there is an alternative explanation. The authors observed many similarities between the human and rat respiratory tract in the distribution of AQP3, AQP4, and AQP5, including the apical expression of AQP5 in secretory cells of subepithelial glands, the basolateral expression of AQP4 in superficial epithelium, the presence of AQP3 in basal cells of the nasopharyngeal and upper airway epithelium, and the expression of AQP5 in type I pneumocytes. The most notable findings in the study by Gynn and colleagues, however, are the differences between the two species. In contrast to the rat, AQPs are present in the apical membrane of airway epithelium in the human respiratory tract. AQP5 was detected by in situ hybridization in the superficial epithelium of nasopharyngeal and bronchial epithelium, as well as in subepithelial glands. Immunofluorescence confirmed the expression of AQP5 protein in the apical membrane in nasopharynx and subepithelial glands; however, AQP5 protein “was not routinely detected” in the bronchial epithelium, a discrepancy with the results of in situ hybridization that warrants further evaluation. At the level of the bronchioles, AQP3 was expressed not only in basal cells, but also in the apical membrane of columnar cells in the bronchiolar epithelium. In addition to being distinct from the rat respiratory tract, this is also the first example of apical trafficking of AQP3; in kidney, as well as in other, tissues, AQP3 has been localized exclusively to the basolateral membrane of different epithelia (11). Finally, the authors suggest that AQP3 is present in type II pneumocytes and that AQP4 is present in alveolar epithelial cells. Higher resolution imaging studies will be necessary to confirm the presence and distribution of these water channels in alveolar epithelium. The details of water transport in the respiratory tract continue to be a source of discussion, if not frank controversy. How is water supplied to the airway for humidification of the inspired airstream? Gynn’s observations, coupled with prior descriptions of AQP1 in the subepithelial ( Received in original form January 26, 2001 )