the ability to eliminate any gene in mice and the increasing ability to identify specific genes responsible for human single-gene diseases is providing a cornucopia of biological information on the relation of genotype to phenotype. Tracing the pleiotropic effects of genes is greatly facilitated when cross-species comparisons are possible. In that regard, and quite apart from its primary motivation of alleviating disease, the vigorous research program on cystic fibrosis (CF), which includes detailed studies of mice and humans who have various mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) anion channel, is illuminating a wide range of basic physiological functions, some expected and some not.
A fascinating example of an unexpected finding is the discovery of a progressive and eventually massive loss of olfactory neurons in cftr −/− mice (CF mice) reported in the recent AJP-Cell Physiology article by Grubb, Ostrowski and their colleagues (11). The signature features of CF arise primarily and possibly exclusively from disruption of ion-mediated fluid transport in epithelia. In humans, inadequate fluid secretion by the pancreas, liver, intestine, vas deferens, sinuses, airway surface cells, and airway submucosal glands contributes to impaction of the macromolecular secretions from those organs, while decreased salt absorption by the sweat ducts produces the salty sweat that is the most reliable phenotypic marker. In CF airways, decreased fluid secretion is further complicated by increased fluid absorption, and the combined transport defects cripple the mucosal innate defenses, so that bacterial and fungal infections of the airways usually start early and quickly become chronic (4). It is the unremitting airway infections that produce most of the suffering and premature death in people with CF.
In the CF mouse, the same organs tend to be affected, but the outcome is altered by differences in the relative roles of CFTR and other anion channels in epithelia, by differences related to the size of organs, and for other reasons that are still being determined. For example, potentially lethal impaction of the intestine (meconium ileus) occurs in about 10% of human CF neonates, but it is nearly universal in mice, while CF mouse airways are much less susceptible to chronic infections than human airways, and the mouse pancreas, vas deferens, and most other organs also show a less severe phenotype (6, 8).
Why do olfactory neurons degenerate in the CF mouse, and why is it of interest? The olfactory epithelium contains at least six cell types: the olfactory receptor neurons, with axons that enter the brain to synapse in the olfactory bulb; sustentacular cells, which are in close apposition to the receptor neurons (see Fig. 1); horizontal and globose basal cells, which are precursors to the other cell types; microvillar cells of unknown function; and Bowman's gland acinar and duct cells. CFTR is not expressed in the olfactory receptor neurons, but evidence from Grubb et al. (11) and from prior work (17) indicates that the sustentacular cells express both CFTR and ENaC (the amiloride-sensitive epithelial Na+ channel). In fact, both of these channels appear to be more heavily expressed in sustentacular cells than in typical ciliated epithelial cells of the mouse airways, and their contributions to epithelial ion transport are correspondingly more robust.
Consistent with effects seen in human respiratory epithelium, measurements of ion transport across olfactory epithelium indicate that CF mice do not secrete Cl− in response to the CFTR activator forskolin, and they show much larger than normal decreases in transport after exposure to amiloride. (Interestingly, the olfactory epithelium of wild-type mice showed almost no response to amiloride, indicating that ENaC-mediated transport was minimal in the isolated tissue.) In CF mice the olfactory receptor neurons, normal at birth, begin to show pronounced blebbing after about a month and then progressively disappear, as assayed both morphologically and functionally, while the sustentacular cells themselves remain intact and perhaps even increase in volume. The basal cells also seem to be unaffected. This pattern is quite different from the effects seen with olfactory bulb lesions, but does resemble normal aging. The tentative conclusion of the authors is that increased fluid absorption by the sustentacular epithelium thins the protective mucus layer and exposes the olfactory receptor neurons to damage via desiccation (11).
The olfactory neuroepithelium is of great interest to developmental neurobiologists. The olfactory receptor neurons were the first and for many years the only examples of mammalian neurons that are continuously renewed (10) and that also display vigorous regeneration when injured (9). What happens to this process in the CF mouse? Is the basic regenerative process negatively affected, is a compensatory increase following increased neuron death not recruited, or is it simply overwhelmed by a greatly increased rate of degeneration? The latter possibility seems less likely given the massive degeneration and robust regenerative response elicited by olfactory bulb lesions in wild-type mice (5, 9). Answers to these questions may be forthcoming, because studies of olfactory receptor regeneration after injury have outlined some of the pathways involved and have identified a subpopulation of pluripotent horizontal basal cells that can give rise to multiple glial and neuronal cell types (5, 12).
It will also be of great interest to elucidate more fully the role of sustentacular cells in fluid transport. While they express CFTR and ENaC, they also express a tetrodotoxin-sensitive Na+ channel (20), and they have a different pattern of aquaporin expression compared with other respiratory epithelial cells: they express only AQP3 (1), whereas gland cells have AQP5 in the apical membrane and AQP3 + AQP4 on the basolateral membrane (19) and non-olfactory surface epithelial cells in the mouse nose express abundant AQP4 (1). These findings provide numerous avenues to test the hypothesis that receptor neurons are lost because of aberrant fluid transport through the sustentacular cells. Would CF mice that lack AQP3 be protected from olfactory receptor neuron loss? What happens to the tight junctions between the degenerating receptors and the sustentacular cells? What role, if any, is played by Bowman's glands in maintaining the integrity of the olfactory epithelium? Would the same loss of neurons be observed in mice in which the sustentacular cells overexpress ENaC? Is the CF mouse olfactory epithelium more susceptible to infection, and if so, does that play a role in olfactory neuron loss?
The work of Grubb et al. (11) has created a new model for investigating how this complex neuroepithelium is maintained. Their finding is a continuation of a stream of discoveries occasioned by the cloning of the CF gene in 1989 (16) and the production of cftr −/− mice just 3 years later (18). Despite almost two decades of productive research, the trend indicates that young physiologists need have no fear that they have missed the gold rush. Insights from the CF mouse have not abated, as exemplified by recent demonstrations of defects in CF mouse salivary glands (3), alveolar macrophages (7), airway glands (13), and even the timing of puberty, which is also aberrant even in mice that are heterozygous for CFTR mutations (15).
Philip Ball, writing in Nature about the packing of DNA, noted: “We know about molecules; we know about cells and organelles; but the stuff in between is messy and mysterious.” (2), and he in turn quotes from François Jacob: “Each level of organization represents a threshold where objects, methods and conditions of observation suddenly change… Biology has then to articulate these levels two by two, to cross each threshold and unveil its peculiarities of integration and logic” (14). Working out how the loss of an anion channel in a “supporting cell” leads to the loss of adjacent neurons is likely to provide just the kind of physiological insight that bridges levels of organization and will thus be of interest well beyond the CF research community.
I thank Tony Nguyen for help in securing articles and copyright permissions and Dr. Mary Lucero for discussions about sustentacular cell electrophysiology.
- Copyright © 2007 the American Physiological Society