We report here that CFTR -/- mice develop an exacerbated airway constriction in response to LPS as compared to their corresponding littermates, which is in agreement with observations made in CF patients . In addition to exacerbated pulmonary inflammation, CF patients manifest airway obstruction and wheezing  and near 40-50% of these patients have airway constriction. This led us to explore the mediators involved in the induction of airway constriction in CFTR -/- mice. Our studies suggest that cPLA2α, which catalyzes the key step of AA release, plays a role in enhanced airway constriction observed in CFTR -/- mice. Indeed, we found an increased concentration of AA in BALF from CFTR -/- vs CFTR +/+ mice. Our findings are in agreement with the pioneer work of Uozumi et al. reporting that cPLA2α plays a key role in increased airway resistance in response to allergic challenge . Concerning the mechanism by which CFTR regulates cPLA2α, recent studies in our laboratory suggested that cPLA2α activity is inhibited by CFTR through a protein-protein interaction [, unpublished data (Dif F, Wu YZ)]. The absence of CFTR or its F508del mutation (known to promote CFTR degradation) may increase cPLA2α activity by the removal of the CFTR inhibitory effect.
In the present study both cPLA2α null mutation and pharmacological inhibition by the cPLA2α inhibitor, ATK, reduced LPS-induced airway constriction in CFTR -/- mice. This identifies cPLA2α as a key factor in LPS-induced airway constriction in CFTR -/- mice. However, we cannot exclude the contribution of another PLA2, iPLA2, to this process. Indeed, ATK has been shown to interfere with iPLA2 activity [22, 23].
Our findings that the COX metabolites of AA did not contribute to LPS-induced airway constriction in CF animal model are in agreement with the previous studies of Vincent et al  which showed that aspirin fails to interfere with LPS-induced airway constriction in guinea pig. Aspirin did not interfere with LPS-induced airway constriction in mice and blockade of COX-2 activity by the specific inhibitor NS-398 only delayed this airway constriction . In the present study we only investigated PGE2 levels since other studies have shown an increased production of various PGs including PGF1, PGF2α in BALF of LPS-treated CFTR -/- mice compared to their littermates . Because aspirin is known to suppress the production of all PGs produced either by COX-1 and COX-2 pathways, we can conclude that PGs are not involved in LPS-induced airway constriction in CFTR -/- mice.
On the other hand, our findings suggest that 5-LOX, a major LOX pathway of AA metabolism is unlikely to be involved in LPS-induced airway constriction. Indeed, the levels of LTB4 and cysteinyl-leukotrienes (LTC4/D4/E4), the products of AA via LOX, are similar in CFTR-/- compared to CFTR +/+ mice. However, we cannot exclude that other LOX-dependent metabolites such as those of 12-LOX can play a role in airway constriction. Indeed, the expression level of this LOX has been shown to increase in bronchial tissues of CF patients .
Our studies suggest that increased PGE2 production in CFTR -/- mice may result, at least in part, from the availability of higher concentrations of free AA. This is in agreement with previous studies reporting that epithelial cells from CF patients release more AA than control cells and express higher levels of cPLA2α activity [28–30]. The fact that CFTR -/- and CFTR +/+ mice produce comparable levels of LTB4 and cysteinyl-leukotrienes is paradoxical given the enhanced production of free AA in BALF of CFTR -/- mice. Although the reasons for this paradoxical observation are still unclear, we suggest that a metabolic deviation of AA in favor of COX pathways may occur in lung tissues of CFTR -/- mice. This might be due to an enhanced activity of COX enzymes in spite of similar expression levels in lungs from CFTR -/- and CFTR +/+ mice. It is also likely that the activity of PGE synthase (PGES), which produces PGE2 from PGH2, may increase in lungs of CFTR -/- mice. Thus, in addition to changes in AA levels and cPLAa activity, an increased PGES activity could be a possible interpretation of PGE2 elevation in CFTR -/- mice.
Failure to detect changes in leukotriene levels in CFTR -/- mice is also in disagreement with the previously reported high LTB4 levels in BALF of CF patients compared to healthy subjects . This discrepancy might be due to differences in the expression levels of LOX, COX and PGES in CF patients compared to CF mice. Previous findings  reported an exacerbated expression of COX-2 in epithelial cells and nasal polyps from CF patients as compared to the corresponding controls . Whether this discrepancy is due to differences in animal species or cell types involved in COX expression remains to be investigated. It remains also unclear whether COX-2 up-regulation observed in polyps from CF patients is a direct consequence of CFTR mutation and/or a secondary consequence of airway inflammation and infection inherent to CF disease.
Although the molecular mechanisms involved in cPLA2α-induced airway constriction in LPS-challenged mice are still unclear it is likely that airway smooth muscle cells (SMC) may play a role in this process. In the asthmatic airway, acute airway constriction is caused, in part, by the enhanced presence of mediators released from inflammatory cells that directly induce bronchoconstriction and enhance bronchoconstrictor responses to other agonists. Airway obstruction and airway constriction in CF patients coincide with those seen in asthma and suggest that airway SMC remodeling may contribute to lung pathology in CF . Recent studies reported that accumulation and/or hypertrophy of airway SMCs contribute to airway narrowing and airway constriction in CF patients [32, 33]. A previous study showed that bradykinin-induced contraction of airway SMC occurs, in part, via a process involving a rise of [Ca2+] and enhanced release of AA . More recently, it has been shown that the AA metabolite 20-HETE induces sustained contraction of isolated guinea pig airway SMC . Interestingly, a recent study demonstrated that CFTR is also expressed in tracheal SMC and may contribute to bronchodilation . Thus, it is likely that enhanced airway constriction in CFTR-/- mice might partially be due to the lack of bronchodilation function of CFTR in tracheal SMC. On the other hand, morphological analysis of the trachea and airway functional studies showed the presence of disrupted or incomplete cartilage rings in trachea of both adult and newborn CFTR -/- and F508del mice . Although the loss of tracheal cartilage may predispose to collapse of the airways, the possible relationship between congenital malformations in CF mice and airway constriction remain to be investigated.
Our studies showed that although LPS induces airway constriction in CFTR-/- and cPLA2α +/+ mice at different intensity as compared to CFTR+/+ cPLA2α -/- mice, respectively, all mouse strains develop a similar extent of lung inflammation in term of neutrophil influx and MIP-2 production. This can be explained by the fact that cPLA2α may not play a role in lung inflammation in LPS challenged mice. It is also likely that PGE2 plays a role in attenuating lung inflammation in CFTR -/- mice since this prostaglandin is well known to exert an anti-inflammatory effect in lungs . Thus, the enhanced production of PGE2 in CFTR -/- mice may explain, at least in part, why these mice do not exhibit exacerbated lung inflammation. The fact that airway constriction occurs independently from lung inflammation is in agreement with previous reports. Indeed, Lefort et al. showed that airway constriction occurs independently of pulmonary neutrophil recruitment or TNFα synthesis . A similar report showed that increased airway constriction induced by inhaled LPS in COX-1 -/- and COX-2 -/- mice is dissociated from airway inflammation .