Roflumilast Partially Reverses Smoke-induced Mucociliary Dysfunction

Andreas Schmid; Nathalie Baumlin; Pedro Ivonnet; John S. Dennis; Michael Campos; Stefanie Krick; Matthias Salathe


Respiratory Research. 2015;16(135) 

In This Article


Our results show that inhibition of PDE4 with roflumilast improves parameters of mucociliary clearance in NHBE cells. It increases ASL volume by enhancing apical chloride efflux through CFTR. It also stimulates CBF via an indirect effect on ASL volume and a direct effect on cilia. These changes reverse some of the negative effects of cigarette smoke on MCC and provide further mechanistic evidence that may explain the beneficial effects of this drug on reducing COPD exacerbations.

Following observations that PDE4 inhibition in rodents reduced tobacco smoke-mediated neutrophil influx in BAL, reduced lung parenchyma infiltration of neutrophils, macrophages and lymphocytes, and inhibited endothelial-neutrophil cell interactions,[22,23] initial studies of roflumilast in populations with COPD focused on its anti-inflammatory effects. Roflumilast has been found to reduce sputum neutrophils, eosinophils, and soluble markers of neutrophilic and eosinophilic inflammation compared with placebo.[24] It also decreases other inflammatory molecules such as metalloproteinases and TGF-ß.[25–27] Clinically, this translates not only in a decrease in the rate of mild to moderate exacerbations and a prolongation of the time to the next exacerbation, but also into improved spirometric values and patient-oriented outcomes such as improvements in dyspnea scores.[28,29]

MCC is a key mechanism to protect the airways from inhaled particles and infectious agents. Major components of this apparatus include: 1) effectively beating cilia that move mucus out of the airways and 2) an adequate ASL volume allowing cilia to beat efficiently and hydrating mucus optimally. If MCC fails, patients can develop chronic bronchitis, COPD, bronchiectasis and are prone to pulmonary infections.[30] Our studies may provide some understanding on basic non-inflammatory related mechanisms as to why roflumilast positively influences COPD in patients, namely its effects on ASL volume and ciliary beating.

In our Ussing chamber experiments, 100 nM roflumilast alone did not increase Isc (Fig. 2a). Since phosphodiesterase inhibitors do not directly increase [cAMP]i, but decrease its break down, this finding is not surprising, at least for an acute situation. Chronically, it's possible that roflumilast can elevate cAMP through intrinsic activities of adenylyl cyclases via some basic stimulation by adenosine for instance. This theory is supported by our findings that roflumilast did not change baseline CBF in control cells, but increased it in cells exposed to control airflow (Fig. 5a). Airflow increases ATP release through pannexins[31] and the resulting adenosine increase could stimulate adenylyl cyclase,[32] allowing roflumilast to have a significant effect.

In contradiction to our findings, Lambert et al. published a study where the sole addition of 1 nM roflumilast increased Isc close to 10 μA/cm2.[33] Additionally, roflumilast had an EC50 of 2.9 nM with a maximal effect of about 40 μA/cm2. In order to achieve these high responses with such doses of roflumilast, the authors highlighted the use of a low extracellular chloride concentration (higher driving force). We repeated these experiments but did not observe the Isc stimulations described by Lambert. In addition and in contrast to these authors, we saw a significant effect of 10 μM forskolin addition on Isc after exposure to roflumilast. It is unclear why similar experiments can lead to such different results. One possible explanation is that we used fully differentiated NHBE cells whereas the cited study used primary bronchial epithelial cells grown in submerged monolayers or Calu3 cells. In addition, others have published increases in CBF with sole addition of roflumilast N-oxide,[34] an effect only seen upon smoke exposure in our culture system.

Our study showed beneficial effects of roflumilast on NHBE cells exposed to cigarette smoke. Interestingly, smoke did not decrease [cAMP]i responses as measured by FRET, even though it might lower baseline [cAMP], consistent with smoke-induced decreases in CFTR function and ASL volume. Baseline FRET measurements, cannot be calibrated, and therefore these results have to be interpreted with caution. On the other hand, roflumilast rescued decreased apical CFTR conductance and ASL volume by simultaneously increasing [cAMP]i. A decrease of cAMP in cigarette smoke extract has previously also been described in bronchial fibroblasts.[35]

It has already been known that cigarette smoke reduces CFTR function[7–9] and that this negative effect can be rescued by roflumilast.[33] However, our study goes further by demonstrating that the roflumilast-induced improvement of CFTR function leads to a rescue of ASL volume and a complex improvement of CBF (direct and indirect effect via ASL volume) as well.

The effect of the [cAMP]i on CBF has been well documented in mammals.[36–38] Here we show that CBF increases more than 50 % in smoke-exposed cells in the presence of roflumilast and an additional 25 % after addition of forskolin. Roflumilast also increased CBF by about 50 % in air-exposed cultures. This finding is exciting as the control cultures exposed to airflow represent the airway epithelium in non-smoking individuals simulating air movements during inhalation and exhalation. Furthermore, formoterol, a long acting ß2 -adrenergic agonist and cornerstone of the treatment of symptomatic COPD,[21] enhanced roflumilast's ability to stimulate CBF in cigarette smoke exposed cells as shown in Fig. 6b.

While alterations in one parameter of mucociliary clearance not always translate into a complementary change of all parameters, it has been shown that a 16 % change in CBF can be associated with a 56 % improvement in mucociliary clearance.[39] Thus, the observed differences with respect to CBF could translate into significant improvements of mucociliary clearance.

Our data contain some unexpected findings. The ASL volume increased initially upon smoke exposure, likely due to mucus secretion. Those initial changes were not necessarily reflected in CBF, indicating that additional mechanisms must be at work that could include surface viscosity. We also showed that airway epithelial cell CFTR mRNA expression is increased in subjects who were active smokers. However, CFTR function and protein expression have been shown to be depressed not only here but also in patients who actively smoke.[40] Thus, mRNA levels of CFTR are not good predictors of the channel's apical function. Our results show a decreased expression of CFTR in ALI cultures compared to brushes from airways, a finding that has been previously reported.[41]

The pathophysiology of COPD is complex. Tobacco smoke induces many anatomical changes including mucus cell hyperplasia and a lower number of ciliated cells with shortened cilia[42] as well as peri-bronchiolar fibrotic changes. Besides its effects on inflammatory cells, PDE-4 inhibition has been shown to decrease EGF-induced MUC5AC expression in human airway epithelial cells,[43] to reduce airway mucous metaplasia via its anti-inflammatory properties,[44] and to exhibit antifibrotic effects by targeting fibroblasts.[45] Together, these observations along with our findings of additional positive effects on smoke-induced impairment of MCC, provide a more comprehensive mechanistic understanding of the beneficial effects of roflumilast in COPD, in particular for those with chronic bronchitis. While the effect of roflumilast seems clear for improving ciliary function, we do not suggest that this is the only way roflumilast might be beneficial. The effect on ASL volume will affect also mucus hydration and the two effects together will likely have the most significant effect in disease.