Cystic fibrosis mice rehabilitated for studies of airway gland dysfunction

Cystic fibrosis, the most common inherited lethal disease in Caucasians, is caused by mutations in the CF gene. The gene product, the CF transmembrane conductance regulator (CFTR), is a regulated chloride channel expressed in the apical membrane of secretory and absorptive epithelia of affected organs. Lung disease is the prime cause of morbidity and mortality among patients with CF, but, despite intensive efforts, its pathogenesis remains ill-understood.

Numerous mechanisms have been proposed to link the defect in CFTR expression or function to CF airway disease, including abnormal airway surface liquid composition, hyperabsorption of airway fluid, defective submucosal gland secretion, defective antimicrobial function, excessive inflammatory responses, and defective macrophage function (Verkman et al. 2003; Di et al. 2006). In humans, the abundance of glands in the large airways (trachea, bronchi), their massive contributions to secreted mucins and antimicrobials, and the high expression level of CFTR in serous cells suggest a key role in CF pathophysiology (Wine & Joo, 2004).

By optical monitoring of the volume expansion of mucus droplets accumulating under mineral oil at the orifice of individual submucosal glands, Joo et al. (2002) showed increased fluid secretion in response to both cholinergic, Ca²⁺-linked agonists (carbachol) and cAMP agonists (VIP, forskolin) in human trachea. In CF glands, carbachol-induced secretion was maintained (albeit reduced when corrected for the hypertrophied state), but the VIP/forskolin-provoked fluid secretion was completely lost. According to their model, serous cell secretion of electrolytes and water in response to both cAMP- and Ca²⁺-linked agonists is completely defective in CF, whereas fluid secretion by mucous cells involves Ca²⁺-sensitive chloride channels that are not affected, or even hyperexpressed, in CF. Consequently, not the production and composition of mucus but its dilution by serous cell fluid is impaired in CF, in line with direct measurements of mucus viscosity through photobleaching of mucus droplets injected with fluorescent indicators (Jayaraman et al. 2001).

Since the trachea of CF patients suffers from end-stage lung disease, confirmation of this concept in an animal model would be greatly desirable. CFTR null mice do not have overt lung disease and their gland phenotype is a direct consequence of CFTR deficiency. However, CF mouse tracheae were believed not to qualify as a model for studying CFTR dysfunction in view of major anatomical and physiological differences with human trachea, including the preponderance of Clara cells over ciliated cells, a lack of Na⁺ hyperabsorption, a virtually normal Cl⁻ secretory response to both Ca²⁺ ionophore and forskolin/cAMP, little or no CFTR expression, and the misconception that mice lack airway glands (Grubb & Boucher, 1999).

Now, a study of submucosal gland secretion by the mucus droplet approach in the trachea of CFTR–/– mice by Ianowski et al. (2007) published in this issue of The Journal of Physiology has broken down this doctrine. First, they report that upper trachea from congenic C57BL/6J mice contains 15–20 airway glands in wild-type mice and even larger numbers in CFTR–/– animals. Furthermore, gland fluid secretion could be stimulated by cholinergic and central (electrical field) stimuli in both WT and CF mice, implying that it was CFTR independent. In contrast, gland secretion in response to VIP and forskolin (amounting to ∼8% of the carbachol response) was completely lost in CF mice. Even more interestingly, local stimulation of sensory neurons in the surface epithelium by capsaicin (the active ingredient in chilli pepper oil) mimicked the effect of VIP on fluid secretion in WT mice and likewise failed to provoke secretion in CF mice. This pattern of gland secretion in the mouse qualitatively reproduces the CF defect in human trachea and suggests the existence of two different modes of gland secretion: a CFTR-dependent, low capacity secretion stimulated by local irritants (viruses, particles, bacteria) and mediated by local reflex pathways; and a (largely) CFTR-independent, high capacity secretion driven by strong vagal input as caused by vigorous exercise, cough induction and hypotonic saline, all currently applied as CF therapies.

Thus this work has both physiological and clinical significance, yet questions remain. For example, is the viscosity and acidity of the mucus secreted by CF mouse trachea in the high capacity mode normal or elevated, like in CF patients? Is the loss of the low capacity ‘house-keeping’ secretion in tracheal glands of the C57BL/6J CF mice causally related to the development of spontaneous and progressive lung disease of early onset, manifest mainly in the distal airways of this particular mouse strain (Kent et al. 1997; Durie et al. 2004)? Does a similar secretory defect exist in tracheal glands of ΔF508-CFTR mutant mice, and is it correctable by pharmacological chaperones emerging from high-throughput screening of chemical libraries (Van Goor et al. 2006)?

It will also be of interest to define the physiological importance and possible dysfunction in CF of the often neglected mode of local CFTR regulation involving cGMP signalling. This pathway is triggered from the luminal side by endo- or paracrinic release of the bioactive peptide guanylin, and by local production of nitric oxide. Cyclic GMP regulation of CFTR is highly prominent in intestinal crypts (Vaandrager et al. 2005) and in airway Clara cells (Kulaksiz et al. 2002), but its impact on the submucosal glands of the airways has not been explored yet. The option to exploit CF mouse models for answering these questions is exciting in the light of the growing notion that defective serous gland function is a primary manifestation of CF lung disease.