Antiestrogen- and tamoxifen-induced effects on calcium-activated chloride currents in epithelial cells carrying the ∆F508-CFTR point mutation

Cystic fibrosis (CF) is the most common fatal genetic disease in the white population. It is caused by mutations in a gene encoding the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR), a protein that functions mainly as a chloride channel. A mutation of the CFTR gene usually produces abnormal proteins that do not transport chloride ions and water properly, or are not transported to the apical membrane [1‐3]. More than 2000 genetic CFTR variants are known, the most frequent being the F508del. Most mutations of the CFTR gene are missense alterations, but frameshifts, splicing, nonsense mutations, and in-frame deletions and insertions have been described. About 15% of the genetic variants that have been identified are not associated with the disease [3] . The CFTR channel defect is mainly in chloride and bicarbonate transport. Interactions of CFTR and other ion channels, particularly the epithelial sodium channel, and interactions of CFTR with cellular pathways related to inflammation (inflammasome) might be important in the pathophysiology of CF [4]. The importance of understanding the pathophysiology of this disease in the first few years of life has been underscored by recent studies showing that, by the age of 3 years, almost a third of children with CF have computed tomographic evidence of mucus obstruction, bronchiectasis, and inflammation driven by neutrophils, neutrophil elastase, and recurrent episodes of infection [4, 5]. The primary hypothesis to explain these clinical features is that impaired mucociliary clearance caused by abnormal hydration of airway surface liquid is the key underlying defect [4, 5]. In newborn pigs with CF it has been observed that mucus fails to detach from submucosal gland ducts and accumulates in pulmonary airways, thus hindering mucociliary transport, an abnormality which, at the origin of the disease, is not dependent on infection or inflammation [6]. With progression of the disease, advancing infection and bronchiectasis further disrupt mucociliary transport, which, in turn, impairs bacterial clearance and promotes resistance to antibacterial defenses [4].

Although CF is not sex-linked, females with this disease experience a more rapid decline in lung function, have more pulmonary exacerbations and have a shorter life span compared with males with CF [7, 8]. Several lines of evidence indicate that the female sex hormone estrogen plays a relevant role. In vitro studies have shown that estrogen receptors ERα and ERβ are expressed in normal lung tissue [9, 10] and that ERα are expressed in cell cultures from non-CF and CF patients, at similar levels in males and females [10]. Choi and colleagues [11] have shown that 17β-estradiol, by interacting with ERα, up-regulates MUC5B gene expression and increases the production of MUC5B, one of the major mucins in the human airway submucosal glands [12], thus favoring mucus plugging. Pretreatment with the estrogen receptor antagonist ICI182,780 suppressed 17β-estradiol-induced MUC5B expression [11]. Inflammation is another important mechanism of lung injury and lung function decline in CF patients. A recent investigation demonstrated that exogenous administration of 17β-estradiol to male CF mice increases the severity of P. aeruginosa pneumonia by increasing inflammation and the production of mucin and inflammatory cytokines [13]. Furthermore, circulating levels of estradiol correlated with infective exacerbations among menstruating women with CF, and the exacerbation rate per year was significantly lower in women receiving oral contraceptives than in CF females with a regular or irregular menstrual cycle [14]. Moreover, exposure of P. aeruginosa to estradiol resulted in an increase of alginate, the predominant polysaccharide produced by P. aeruginosa and implicated in its pathogenicity, and enhanced the conversion from the non-mucoid to the mucoid morphology [14].

Another recent investigation showed that 17ß-estradiol, by reducing the ATP-induced increase in intracellular Ca²⁺, inhibits the activity of calcium-activated chloride channel (CaCC) [10]. Treatment with tamoxifen citrate (TMX) restored the ATP-induced increase in intracellular Ca²⁺ despite the presence of 17ß-estradiol, thus increasing airway surface liquid production. TMX was more potent than the pure ER antagonist ICI182780. However, direct measurements of the effects of TMX on CaCC currents are not available.

Our experiments were based on the hypothesis that stimulation of CaCC currents could restore the permeability to chloride ions that is downregulated by the malfunctioning of the CFTR ion channel. Homozygosity for the 508 mutation in CFBE cells reduces the chloride flux through the plasma membrane almost to zero, thereby preventing the movement of water. Comparing control 16HBE epithelial cells with the CFBE clone (see methods), the absence of the cAMP-activated chloride permeability in the cells carrying the F508^−/− point mutation is evident. Perforated patch current recordings obtained in both types of cells are shown in Fig. 1. The effect of perfusion of 10 μM FSK, causing a cytoplasmic increase of cAMP, is evident in 16HBE but not in CFBE cells. In addition, the increase of membrane current in 16HBE cells was totally inhibited by 10 μM Inh-172, a specific CFTR channel blocker. The experiment on CFBE cells, whose results are shown in Fig. 1, demonstrates the absence of CFTR channel activity in the mutated clone [17].

Perforated patch experiments performed on CFBE cells in control conditions showed a current inversion at − 47.8 ± 6.5 mV, consistent with limited chloride permeability. The basic current of CFBE cells was inhibited by 300 μM NPPB (Fig. 2a). To assess the inhibitory action of the estrogen hormone, CFBE cells were exposed to 10 nM 17β-estradiol for 30 min. The presence of 17β-estradiol significantly reduced the chloride currents, both in basal conditions and in the presence of ATP (100 μM). Activation of the purinergic receptor P2Y₂ has the effect of CaCC agonism (Fig. 2b and d). However, stimulation with 10 μM TMX induces an increase of intracellular calcium [10]. Accordingly, TMX completely restored the currents abolished by 17β-estradiol, both in basal conditions and after stimulation with ATP (Fig. 2c and e).

The same experimental paradigm was applied to single IB3–1 cells: the findings are illustrated in Fig. 3. Cells with 508 mutation in heterozygosis showed a current reversal potential of − 57.8 ± 7.1 mV, consistent with greater permeability to chloride than that of CFBE cells. Figure 3a shows the effect of 300 μM NPPB on the basic membrane current of IB3–1 cells. IB3–1 cells show a similar behavior to that of CFBE cells after stimulation with 17β-estradiol and ATP (Fig. 3b and d) and in the presence of TMX and ATP simultaneously with estrogen stimulation.

Collectively, these findings indicate that the basic membrane currents recorded in CFBE and IB3–1 cells is mainly carried by chloride through the CaCC.

TMX had a direct effect on membrane ionic permeability of CFBE and IB3–1 cells. Figure 4 shows the results of parallel studies of CFBE and IB3–1 cells stimulated by TMX in perforated patch experiments. Figure 4a and b depict current/voltage relationships for CFBE and IB3–1 cells, respectively, in control conditions and after stimulation with 10 μM TMX.

The membrane current stimulation produced by TMX (Fig. 5a, circles) can be further enhanced by the addition of 100 μM ATP (Fig. 5a, triangles). Acute application of NPPB totally inhibits the whole cell current stimulated by TMX and ATP (Fig. 5b), confirming that the final targets of TMX and ATP are CaCC.

The in vitro efficacy of TMX on membrane chloride ionic currents in CFBE cells was compared with that of two recently discovered CFTR permeability enhancers. Ivacaftor (VX-770) and lumacaftor (VX-809) are two approved drugs (Orkambi^®, Vertex Pharmaceuticals) used in combination for the treatment of CF in patients with F508-del mutation in homozygosis. Using an Ussing chamber, Van Goor and colleagues showed that the two drugs increased the CFTR current in experiments performed on a cell monolayer. The procedure enabled the transepithelial current to be recorded in the voltage-clamp mode [18]. In our experiments, we monitored the change of chloride current density on single CFBE cells incubated for 24 h with 3 μM VX-809. Figure 6a shows a current/voltage plot on CFBE cells treated with VX-809 and acutely stimulated with 1 μM VX-770 immediately prior to electrophysiological measurements. At the end of each experiment we also added 10 μM FSK to the external solution. Membrane current amplitude was enhanced upon VX-809/VX-770 exposure; the subsequent application of FSK resulted in a trend towards an increase, although this was not statistically significant. The CFTR blocking agent Inh-172 completely inhibited the membrane current (Fig. 6a and b).

To compare the effects of TMX/ATP with those of VX-809/VX-770/FSK, average current/voltage relationships for both the experimental conditions were plotted (Fig. 7a) and compared with the membrane current in basal conditions. Both curves show a significant increase of the current density. At the 80 mV voltage step, TMX/ATP caused a mean increase in membrane permeability of 29% compared to that produced by VX-809/VX-770/FSK.

The combined effects of TMX/ATP and VX-809/VX-770/FSK were seen to be additive. CFBE cells were exposed to VXs/FSK and, after a steady current had been reached, to TMX/ATP. The current/voltage relationship shown in Fig. 8a depicts the membrane current elicited at the different voltages in the presence of VXs/FSK (squares), followed by addition of TMX/ATP (circles) and, finally, perfusion with CFTR and CaCC blockers (triangles). The histogram in Fig. 8b illustrates the different current density values at a membrane voltage of + 80 mV. The average TMX/ATP current value has been imported from Fig. 7. The light gray column on the right of Fig. 8b presents the plain sums of VXs/FSK plus TMX/ATP single averages.

Our study shows that TMX increases the CaCC currents by two mechanisms: 1) indirectly, as an antagonist of the negative effect of 17β-estradiol on currents, and 2) directly, independently from the antiestrogen effect, acting on the CaCC, the alternate chloride transport pathway (Fig. 9) The effect of TMX on CaCC develops soon after TMX treatment. The CaCC-dependent currents elicited by TMX are quantitatively superior to the CFTR-dependent chloride currents elicited by ivacaftor+lumacaftor. Of note, the combined administration of TMX/ATP and VXs/FSK had an additional effect on chloride currents.

CFTR ion channels in healthy subjects ensure the correct hydration of the respiratory system airway surface. CFTR chloride flux induces an osmotic driven movement of water through the apical membrane of respiratory epithelia. Water release in the lumen dilutes the mucopolysaccharide layer on the luminal cell surface. Fluidity of mucus towards the external environment ensures correct removal of solid residues introduced during breathing [4]. Inflammation and host-defense defect also contribute to the pathophysiology of CF lung disease and lung function deterioration [4].

In CF, CaCC in airway epithelial cells contribute to the secretion of chloride and water. Although the role of CaCC in normal and CF airway epithelial cells has not been completely elucidated, several lines of evidence suggest that CaCC could have clinically relevant effects in CF patients. Mice homozygous for disruption of the CFTR genes, unlike their human counterparts, fail to show any gross lung pathology [19]. It is possible that CaCC are more expressed in mice, thus compensating for the lack of CFTR function. By contrast, mice knockout for the TMEM16A gene, the gene encoding a CaCC in different secretory epithelia [20‐22], exhibit phenotypic abnormalities consistent with a role in fluid and electrolyte homeostasis [23‐25]. Finally, in their pivotal study, Knowles and colleagues [26] showed that administration of the CaCC activating agents ATP and UTP in vivo and in vitro was more effective at inducing changes in nasal potential difference in CF patients than in normal controls, suggesting that the activity of CaCC is more pronounced in CF cells than in non-CF cells.

Coakley et al. [10] demonstrated in CF airway epithelial cells that the acute addition of 17ß-estradiol reduced the ATP-induced increase in intracellular Ca²⁺and the production of airway surface liquid. The presence of TMX restored the intracellular Ca²⁺ concentration and potentiated the ATP-induced airway surface liquid secretion, even in the presence of 17ß-estradiol. Our study confirms the results obtained by Coakley et al. [10] and gives new insights on the action of TMX. We explored the possibility of increasing chloride ion fluxes in single epithelial cell cultures, CFBE cells carrying the 508 mutation in homozygosis and IB3–1 cells with the same mutation but in heterozygosis, by stimulating the CaCC. Stimulation of CaCC could be an effective way of creating additional chloride flux in order to promote the movement of water. We found that TMX acts both indirectly, as an antagonist of the negative effect of 17β-estradiol on chloride efflux, and directly, independently of its antiestrogen effect, acting on the CaCC, the alternate chloride transport pathway [17].

TMX, counteracting the effect of 17β-estradiol and acting on the P2Y purinergic receptor, restores the intracellular Ca²⁺ concentration [10]. It is likely that the increase of intracellular Ca²⁺was responsible for the increment of chloride currents observed in our study, which was further amplified by the presence of ATP, a P2Y₂ activator (Figs. 2, 3, 4, and 5). The whole cell increase of chloride current density was relevant and greater than that elicited by the association of VX-809 + VX-770 plus FSK (Fig. 7). The direct effect of TMX was evident after a few seconds, suggesting that TMX, rather than its metabolites, is responsible for the activation of CaCC. Combined experiments using both VXs/FSK and TMX/ATP (Fig. 8) demonstrated an additional effect on membrane current. This is very important in the prospective of a combined therapeutic use of both treatments. However, the combined action of VXs/FSK and TMX/ATP produced a smaller increment of the membrane permeability than expected from the mathematical addition of the average single current increases due to each compound. There are several possible explanations for this. For example, intracellular calcium mobilization could act on different regulatory pathways. As a consequence, activation of chloride flux by one compound could be responsible for partial activation of other pathways.

To demonstrate that TMX by acting on CaCCs bypasses the CFTR defect, in our study we used two cell lines, CFBE cells, bearing the most frequent mutation F508^−/− (class II mutation, defective processing), and IB3–1 cells, bearing the less frequent mutation F508del/W1282X (class I mutation, premature stop codon). We used concentrations of TMX which can be attained in human airways in order to extrapolate our results to the clinical setting. Our findings confirm that TMX bypasses the CFTR defect and, therefore, it might be a curative, mutation-independent treatment applicable to all patients with CF. Adult premenopausal women with CF might gain further advantage from treatment with TMX during the ovulatory period (approximately 1 week per month), when estrogens inhibit CaCC currents [10].

A limit of our study was that it was performed in single cells, rather than in airway epithelial tissue. The effects of TMX on mucociliary transport and inflammation should be studied also in CF animal models. However, “in vitro” cell-based models, which are cost-effective and give preliminary reliable information, have been of immense value for the process of developing new strategies in CF.

Therapeutic interventions acting on channels alternative to CFTR (e.g., ENaC inhibitors) and a study evaluating the impact of sex hormones (estrogen and progesterone) on lung disease are ongoing [27, 28].