Abstract
The CFTR gene displays a tightly regulated tissue-specific and temporal expression. Mutations in this gene cause cystic fibrosis (CF). In this study we wanted to identify trans-regulatory elements responsible for CFTR differential expression in fetal and adult lung, and to determine the importance of inhibitory motifs in the CFTR-3′UTR with the aim of developing new tools for the correction of disease-causing mutations within CFTR.
We show that lung development-specific transcription factors (FOXA, C/EBP) and microRNAs (miR-101, miR-145, miR-384) regulate the switch from strong fetal to very low CFTR expression after birth. By using miRNome profiling and gene reporter assays, we found that miR-101 and miR-145 are specifically upregulated in adult lung and that miR-101 directly acts on its cognate site in the CFTR-3′UTR in combination with an overlapping AU-rich element. We then designed miRNA-binding blocker oligonucleotides (MBBOs) to prevent binding of several miRNAs to the CFTR-3′UTR and tested them in primary human nasal epithelial cells from healthy individuals and CF patients carrying the p.Phe508del CFTR mutation. These MBBOs rescued CFTR channel activity by increasing CFTR mRNA and protein levels.
Our data offer new understanding of the control of the CFTR gene regulation and new putative correctors for cystic fibrosis.
Abstract
Transcription factors/miRNAs that regulate fetal to adult CFTR expression change are new targets for CF treatment http://ow.ly/zEgYQ
Introduction
Most transcription factors are part of multi-protein complexes that recruit other transcription factors and cofactors to mediate local and long-range chromatin changes through physical modifications [1]. Within the transcriptional and post-transcriptional trans-regulatory elements that make up gene regulatory networks, transcription factors often interact with microRNAs (miRNAs) to control gene expression [2, 3]. miRNAs are post-transcriptional regulators that are expressed in a tissue-specific or developmental stage-specific manner, thereby greatly contributing to cell/tissue-specific protein expression profiles, including during lung organogenesis [4, 5].
Mutations in the cystic fibrosis transmembrane conductance regulator (CFTR; MIM# 602421) gene are responsible for cystic fibrosis (MIM# 219700), a common recessive lethal disorder. Chronic lung disease is the major cause of mortality and morbidity in patients with cystic fibrosis. CFTR gene expression is spatially and temporally regulated and several studies have demonstrated the differential use of transcription start sites, depending on the tissue type or the developmental stage [6–9]. In the lung, CFTR transcripts can be detected early during embryo development (12th week of pregnancy) and their level progressively increases up to the 24th week of pregnancy. Thereafter, CFTR expression in the airways decreases and is repressed until after birth and remains very low during adult life [10–14]. The changes in CFTR protein expression in human fetuses are consistent with CFTR mRNA temporal pattern [14]. This developmental regulated expression suggests that CFTR is involved in lung organogenesis, possibly by participating in the mechanico-sensory process that is essential for the regulation of Wnt/β-catenin signalling during lung development [15].
Several repressors and activators have been implicated in the regulation of CFTR transcription [16–22]. For instance, CCAAT-enhancer-binding protein (C/EBP)δ positively regulates CFTR transcription by binding to an inverted CCAAT element in the promoter [16]. In addition, C/EBPβ binds to a DNase I-hypersensitive site that is present only in tissues expressing CFTR [23] and we recently demonstrated its binding to CFTR minimal promoter [24]. Chromatin structures that facilitate long-range interactions between various regulatory elements cluster specifically to the CFTR promoter exclusively in CFTR expressing cells [25, 26]. Recently, it was reported that trans-interacting Forkhead Box A (FOXA) factors induce CFTR activation through binding to a key CFTR cis-regulatory element located in introns 10 and 11 [27–29]. CFTR expression is post-transcriptionally regulated also by miRNAs, such as miR-145 and miR-494 [30–32]. Indeed, several miRNAs, including miR-145, are expressed in primary adult human airway epithelial cells where CFTR expression is low [30]. Moreover, increased expression of some miRNAs has been reported in bronchial epithelium and primary culture of airway epithelia from patients with cystic fibrosis compared to healthy controls [33, 34].
Nevertheless, the developmental control of CFTR expression is still poorly understood in humans and the identification of cis- and trans-acting factors responsible for this complex spatio-temporal regulation is challenging. The aims of this study were 1) to identify trans-regulatory elements responsible for the differential expression of CFTR in fetal and adult lung; and 2) to determine the importance of inhibitory motifs in the 3′UTR of CFTR in order to develop new tools for CFTR mutation correction in patients with cystic fibrosis.
Materials and methods
Gene reporter vectors, expression plasmid constructs and directed mutagenesis
The pGL3b-CFTR-WT vector has been previously described [35]. The 3′UTR of the CFTR gene (1.7 kb from the termination codon to the poly-adenylation signal) was subcloned in the pGL3-control vector (Promega, Charbonnieres, France) downstream of the Luciferase gene (pGL3C-CFTR-3UTR). The expression plasmids used are listed in supplementary table S1. Point mutations in cis-motifs to abolish binding were introduced by direct mutagenesis using the QuickChange®II site-directed mutagenesis kit (Stratagene, Massy, France). All constructs were verified by direct sequencing.
Short-interfering RNAs (siRNAs), miRNA precursors and miRNA-binding blocker oligonucleotides (MBBOs) are described in supplementary table S2.
Cell culture
The cell lines and culture media used for this work are described in table 1.
Nasal cells from healthy individuals or patients with cystic fibrosis and homozygous for the p.Phe508del CFTR mutation were obtained by scratching the inferior turbinate epithelium with ASI Rhino-Pro® curettes (Arlington Scientific, Springville, UT, USA). All signed informed written consent and this research project received the agreement by the French ethical research committee (N°ID-RCB 2011-A01520-41). Nasal cells were cultured in an air–liquid interface (ALI) culture system as previously described [37] with minor modifications. All the media used were supplemented with antibiotics and all the supports were coated with collagen I. After 3 weeks of growth in monolayer, 300 000 cells per well were plated in collagen I-coated 12 mm Transwell-Clear® supports, 0.4 μm pore size (Corning Inc., Corning, NY, USA). The ALI medium in the upper compartment was removed after confluence and the medium in the lower compartment was changed every 2–3 days. Experiments were performed when epithelial cells were well differentiated by visual inspection (at least 28 days).
Transient transfections
All the details of the transient transfection assays are in supplementary table S3.
RNA extraction and RT-qPCR
Total RNA was extracted, reverse transcribed and amplified as previously described [38]. miRNAs were purified with the miRNeasy Mini Kit and the RNeasy MinElute Cleanup kit (Qiagen, Courtaboeuf, France). Reverse transcription was performed using 40 ng of miRNA and the miRCURY LNA™ Universal cDNA Synthesis Kit (EXIQON, Vedbaek, Denmark) and qPCR was performed with a 1:10 dilution of first strand DNA and microRNA LNA™ primers specific for each miRNA (EXIQON). The relative expression levels were calculated using the comparative DDCt method with SNORD44 and SNORD48 small nucleolar RNAs as endogenous controls.
Quantitative chromatin immunoprecipitation assays
Quantitative chromatin immunoprecipitation (Q-ChIP) was carried out as previously described [24]. Purified cross-linked chromatin was immunoprecipitated with 3 μg of each antibody (Santa Cruz, Heidelberg, Germany; Clinisciences, Montrouge, France). As a control for nonspecific DNA binding, 3 μg of anti-immunoglobulin G antibodies were used (Santa Cruz, Clinisciences). Results were expressed relative to the input signal and to nonspecific immunoprecipitated chromatin.
Electromobility shift assay
Electromobility shift assay (EMSA) was performed as previously described [17].
Reporter assay
Cells were harvested 48 h after transfection and the activity of firefly luciferase and Renilla luciferase was measured using the Dual-Glo® Luciferase Assay System (Promega).
Labelling, detection and analysis for miRNA profiling
Total RNA (100 ng) from A549 and HBEpiC cells was labelled using the Agilent miRNA Complete Labeling and Hybridization Kit (Agilent Technologies, Massy, France) and then hybridised to the Agilent Human miRNA Microarray (V2, Agilent) that contains probes for 723 mature human miRNAs. Arrays were scanned using an Agilent scanner and features were extracted with the Agilent Feature Extraction software (version 10.5.1.1). Expression data were initially normalised to the 75th percentile and then averaged among the groups using the GeneSpring GX (Agilent) software. The Kruskall–Wallis test was used for group comparisons and the Benjamini–Hochberg correction was applied to adjust for multiple comparisons. Unsupervised hierarchical clustering was performed using the GeneSpring GX software.
Western blotting
Whole proteins were extracted using 1X Laemmli buffer. Proteins were separated on 7% or 10% SDS-PAGE gels and transferred to PVDF membrane (Westran® Clear Signal Whatman®; Dominique Dutscher, Issy les Moulineaux, France). Antibodies and concentrations used are described in the captions to the figures.
CFTR activity
CFTR activity was assessed by iodide-mediated quenching of the halide-sensitive yellow fluorescent protein (YFP), as previously described [36], using the Premo Halide sensor technology (Invitrogen, Villebon sur Yvette, France). 40 h after incubation with MBBOs or negative control, CFTR conductance was stimulated with an agonist mixture (forskolin, 3-isobutyl-1-methylxanthine, apigenin) for 10 min. Then, CFTR-mediated iodide efflux was measured in each individual well by recording the fluorescence emission continuously (400 ms per point) for 2 s (baseline) and after addition of 50 μL of 140 mM iodide solution.
Statistical analysis
Q-ChIP, luciferase and RT-qPCR assays were performed at least three times and samples were analysed at least in triplicate. The t-test was employed for paired comparisons using InStat (GraphPad Software, version 3.0, Instat 3 folder). For assessing the effect of MBBOs, the Wilcoxon signed rank test was used with the R software to generate box plots with significances.
Results
Transcription factors involved in the regulation of CFTR temporal expression
To identify the regulatory elements that participate in CFTR downregulation in human lung after birth, we used primary human fetal bronchial epithelial cells (HBEpiC, three primary cultures from three different donors, gestational age in table 1) and whole lung cells as well as the A549 and Beas-2B cell lines (from adult human lung). CFTR transcript level in fetal HBEpiC and whole lung cells was comparable or even higher than in T84 cells, in which CFTR is constitutively expressed (fig. 1a) [39]. Moreover, CFTR expression in HBEpiC and whole lung cells was more than four-fold higher than in A549, Beas-2B cells and in adult lung RNA from healthy individuals. Based on the well-known CFTR expression profile in fetal and adult lung [12, 13], we considered that HBEpiC cells are a representative model of fetal lung cells and that CFTR expression in A549 and Beas-2B cells is comparable to that of adult lung.
To identify putative cis-regulatory motifs, by using open-source bioinformatics software we then performed an in silico analysis of the CFTR 5′UTR. We thus focused on the transcription factors FOXA and C/EBP, based on two criteria: 1) their high score in predicted transcription factor binding and 2) their involvement in lung morphogenesis. First, we confirmed FOXA and C/EBP binding to the CFTR minimal promoter by Q-ChIP (fig. 1b). Moreover, EMSA showed that C/EBPβ and FOXA2 bind directly to the targeted region (fig. 1c).
Next, we investigated the role of these transcription factors in the regulation of CFTR promoter activity. When overexpressed in A549 and Beas-2B cells (adult lung), FOXA1, FOXA2 and C/EBPα had a repressive effect on CFTR transcription (fig. 2a and b). In line with this result, FOXA1, FOXA2 and C/EBPα silencing using specific siRNAs increased CFTR expression (fig. 2a and b). As control, qPCR and western blot showed that endogenous transcription factor levels were strongly reduced in the presence of each specific siRNA (fig. S1). Conversely, in fetal HBEpiC cells, FOXA1 and FOXA2 did not have any effect, while C/EBPα strongly induced CFTR transcription (fig. 2c). Finally, C/EBPβ overexpression increased endogenous CFTR transcript level in A549 (2×) and Beas-2B (3×) cells and even more in HBEpiC cells (30 ×) (fig. 2a–c). Analysis of the endogenous level of each transcription factor showed a homogeneous expression in the different cell lines (fig. S2).
As FOXA and C/EBP factors do not act alone, we re-examined the in silico analysis data by using a lower cut-off in order to select other transcription factors that may contribute to CFTR regulation. Possible candidates of this transcription factor network included RREB-1 (the transcription factor with the highest score), several transcription factors with a known effect on CFTR gene expression, such as USF2, SRF and YY1 [17, 18, 35], and transcription factors involved in lung morphogenesis, but with a lower score than C/EBPs and FOXAs (SOX17, FOXF1 and NKX2.1) [40]. We then used reporter assays to investigate whether these regulatory elements could participate in the temporal regulation of CFTR expression. Co-transfection of the pGL3b-CFTR-WT reporter vector (wild-type minimal CFTR promoter) with the ubiquitously expressed USF2 and SRF or the developmental-specific NKX2.1 induced luciferase activity in adult and fetal lung cells (fig. 2d). A forced expression of YY1 protein caused a strong decrease in reporter activities (∼50% of the control luciferase value). Conversely, RREB-1 inhibited CFTR transcriptional activity in the adult pulmonary A549 and Beas-2B cell lines but not in primary fetal HBEpiC cells. Similarly, the activating effect of SOX17 and FOXF1 in fetal HBEpiC cells was reduced or even abolished in A549 and Beas-2B cells.
These results show that FOXA1, FOXA2 (directly) and C/EBPα negatively regulate CFTR transcription in a specific manner in mature lung cells, while C/EBPβ induces CFTR transcription through direct binding to the promoter, regardless of the temporal stage. Other transcription factors, such as SOX17, RREB-1 and FOXF1, play also a role in the temporal regulation of CFTR expression in fetal and adult lung.
A complex pattern of cis- and trans-acting elements in the 3′UTR of CFTR is involved in the temporal regulation of its expression
To evaluate the effect of the 3′UTR on the post-transcriptional regulation of CFTR, we then transfected A549, Beas-2B and HBEpiC cells with the pGL3C-CFTR-3UTR reporter vector (CFTR 3′UTR) or vector alone. The 3′UTR of CFTR strongly repressed luciferase activity in all cell types, indicating that this region contains cis-repressive elements (fig. 3a). Using the bioinformatic tool AREsite (http://rna.tbi.univie.ac.at/cgi-bin/AREsite.cgi), we identified four new putative AU-rich elements (ARE) in the 3′UTR of the CFTR gene (ARE-4816, ARE-5533, ARE-5698 and ARE-6074) in addition to those previously described [41] and that we renamed ARE-4585, ARE-4760 and ARE-4891, according to their nucleotide position (fig. 3b). To determine the role of these motifs in the regulation of CFTR expression, we transfected A549, Beas-2B and HBEpiC cells with pGL3C-CFTR-3UTR reporter vectors in which each of these motifs was mutated and then measured luciferase activity. Only ARE-4760 appeared to be implicated in mRNA stabilisation because mutation of this motif was associated with a decrease in luciferase activity compared to cells transfected with pGL3C-CFTR-3UTR (wild-type sequence) (fig. 3b). ARE-4585, ARE-5533, ARE-5698 and ARE-6074 seemed to be involved in mRNA destabilisation in A549 and/or Beas-2B cells, whereas they had no significant effect in HBEpiC cells (fig. 3b). The strongest effect was obtained using ARE-5698, which in silico was identified as the most conserved ARE motif in the CFTR 3′UTR. Other cis-acting elements might explain the repressive activity of the 3′UTR of CFTR in adult cell lines. Computational predictions detected 13 putative miRNA-binding motifs in the CFTR 3′UTR (fig. 3c). Among the previously studied miRNAs, miR-145 has been involved in the regulation of CFTR expression in colonic and pancreatic cell lines [30]. We then assessed the role of miRNAs in the post-transcriptional control of CFTR in pulmonary cells by using luciferase reporter assays after transfection with miRNA precursors and the pGL3C-CFTR-3UTR reporter vector. MiR-942, miR-665, miR-383, miR-1290 and miR-1246 did not induce any significant effect in any cell type, whereas miR-600 reduced luciferase activity in all cell lines compared to control miRNA (fig. 3c). MiR-505, miR-943, miR-377, miR-145, miR-384 and miR-101 decreased luciferase activity in A549 and/or Beas-2B cells, but not in HBEpiC cells (fig. 3c).
As the strongest repressive effect on CFTR post-transcriptional regulation in A549 and Beas-2B cells was induced by miR-101, we next focused on this miRNA and confirmed its negative impact on endogenous CFTR transcript level after transfection in adult pulmonary cells (fig. 4a). After transfection of the miR-101 precursor, miR-101 overexpression was verified in the three cell lines (fig. 4b). We also confirmed the endogenous expression of miR-101 and its differential expression in A549, Beas-2B (adult) and HBEpiC cells (fetal) (fig. 4c).
Previous studies demonstrated that miRNA-mediated regulation might require the presence of an ARE sequence [42–44]. As the miR-101 and miR-600 binding sites overlap with the ARE-6074 motif and the miR-384 binding site overlaps with the ARE-5698 motif (fig. 4d), we asked whether the effect of these miRNAs following binding to the 3′UTR of CFTR is dependent on the integrity of the ARE motifs. To this aim we co-transfected the miR-101, miR-600 and miR-384 precursors with reporter vectors containing wild type or mutated CFTR 3′UTR. Only miR-101 lost its repressive effect on luciferase activity following the mutation of its binding site within the CFTR sequence or abrogation of the ARE-6074 (fig. 4d). Mutation of ARE-6074 and ARE-5698 did not affect the activity of miR-600 and miR-384, respectively.
As miRNAs have been previously described in lung development mainly in mice, we investigated their differential expression in adult human lung tissue and fetal primary whole lung cells using Agilent DNA microarrays. Analysis of the microarray data showed that 65 miRNAs had the highest expression variability between adult and fetal lung. Among the 30 probes with the strongest expression difference between adult and fetal lung (supplementary table S4), we found that miR-451, miR-150 and miR-145 were specifically upregulated in adult lung (fig. 5a). We confirmed the endogenous expression of miR-145 and its differential expression in A549, Beas-2B (adult) and HBEpiC cells (fetal) (fig. 5b).
These data demonstrate the implication of miRNAs in the tightly controlled developmental regulation of CFTR expression and, more particularly, of miR-101 and miR-145, the expression of which is higher in adult than in fetal lung. Moreover, they show that miR-101 directly acts on its cognate site in combination with an overlapping ARE motif.
From identifying crucial regulators of CFTR expression to testing new potential therapeutic tools for cystic fibrosis
The region encompassing the miR-101 binding site and ARE-6074 is critical for the miR-101 role in the regulation of CFTR expression. Based on this observation, we designed MBBOs to prevent binding of several miRNAs, including miR-101, miR-600, miR-145 and miR-384, to the 3′UTR of CFTR. Co-transfection of these MBBOs with the pGL3C-CFTR-3UTR reporter vector led to a 1.5- to 6-fold increase of luciferase activity in Beas-2B and A549 cells, respectively (fig. 6a). The positive effect on endogenous CFTR expression upon MBBO-1 transfection was confirmed in these cells (fig. 6b).
Next, we evaluated the effect of the MBBOs ex vivo because mutant mice do not develop the characteristic manifestations of human cystic fibrosis. To this aim, we added medium containing control oligonucleotide, MBBO-1 or MBBO-3 without any transfection reagent to the upper compartment of Transwell-Clear® supports in which reconstituted ALI epithelial cells obtained from human nasal cells of control individuals (n=8) and CF patients homozygous for the p.Phe508del mutation (n = 6) were cultured. After 2 h at 37°C, the medium was removed from the upper compartment to restore the ALI. Freshly prepared control oligonucleotide or MBBOs were added every 2 days and CFTR expression was assessed 24 h post-treatment. MBBO effect was even stronger in epithelial cells from control individuals. Indeed, MBBO-1 induced a 2- to 6-fold increase of the endogenous CFTR expression in the epithelium derived from healthy individuals (fig. 6c) compared to the 2- to 3-fold increase in cells from patients with CF (fig. 6d). This effect was not significantly improved by repeated incubation with MBBO-1.
MBBO-1 and MBBO-3 significantly increased CFTR mRNA (fig. 7a) and protein expression (fig. 7b) in cystic fibrosis epithelia compared to control oligonucleotide. We next investigated MBBO effect on CFTR channel activity by using a functional assay (iodide-mediated quenching of the halide-sensitive YFP variant) and the human bronchial epithelial cell lines CFBE41o- (derived from a patient with cystic fibrosis) and 16HBEo- (normal phenotype). We first confirmed the absence of CFTR-dependent anion transport in CFBE41o- cells (cystic fibrosis) compared to 16HBEo- (non-cystic fibrosis) cells in which iodide entered and quenched YFP fluorescence (fig. 7c). Addition of Inh-172 (a CFTR inhibitor) in non-cystic fibrosis cells led to results comparable to those obtained in cystic fibrosis cells, confirming that the assay measures CFTR-dependent anion transport. Incubation with MBBO-1 and MBBO-3 for 2 h, significantly increased anion transport in cystic fibrosis cells compared to untreated cells (fig. 7d) and fluorescence quenching was proportional to the amount of CFTR detected in the cells by immunoblotting.
These data support the importance of the regions encompassing the miR-101 and miR-145 binding sites in CFTR regulation and suggest that MBBOs could represent a new therapeutic option for CF.
Discussion
Expression studies carried out in humans, mice and goats have revealed that the CFTR gene is developmentally regulated [10, 12–14, 45]. The most well-known site of developmentally regulated CFTR expression is the airway surface epithelium, with relatively high expression during embryonic and fetal development, followed by a marked decrease in expression after birth [45]. Despite extensive studies, the mechanisms accounting for this switch in CFTR expression remain unknown.
In a critical region of the CFTR gene that contains several naturally occurring variants [18, 35, 46], we found many ubiquitous [17, 18] and tissue- or lung developmental-specific transcription factors [40] involved in coordinating the switch from strong to very low CFTR expression in lungs after birth. The specific occupancy of these factors on the promoter may, in interaction with others factors, including FOXA1 and C/EBPβ that bind to other part of the CFTR gene depending on the nucleosome positioning [28], influence the particular pattern of expression of this gene. We also show that miRNAs, including miR-101 and miR-145, negatively regulate the level of CFTR transcripts in adult lung cells, whilst having no effect in fetal lung cells. In addition to its specific role in mature lung cells, miR-101 decreases luciferase activity in an embryonic kidney cell line [31], whereas it does not affect CFTR mRNA stability in pancreatic cell lines [30], suggesting a potential role as a tissue-specific factor. We then demonstrate the implication of miRNAs in the tightly controlled developmental regulation of CFTR expression and more particularly we show that miR-101 acts on its cognate site in combination with an overlapping ARE motif.
Finally, we demonstrate the benefit of characterising regulatory factors to identify novel therapeutic targets. Early studies indicated that complementation of just 6–10% of CFTR transcripts generate enough CFTR levels to maintain normal chloride transport in epithelia [47]. These data are supported by findings that the presence of a naturally occurring sequence variation in the CFTR promoter, in cis of a severe mutation, increases transcription. This can allow the production of enough CFTR protein to reach the apical membrane cells and restore partial CFTR channel function, thus inducing a moderate cystic fibrosis phenotype despite the presence of a severe disease-causing mutation [35]. Similarly, stabilisation of p.Phe508del CFTR protein has been associated with increased p.Phe508del CFTR channel activity [48]. Recent work demonstrated that miR-138 mimics might restores CFTR-Phe508del expression and a functional chloride transport [32]. However, the authors underlined the fact that miR-138 mimics may have undesirable effects because miR-138 targets SIN3, a highly conserved transcriptional repressor which regulates many genes [32]. Over the past four decades, therapies for cystic fibrosis have focused entirely on symptoms to improve the patients' quality of life. The first treatment (VX-770) targeted the basic defect in p.Gly551Asp-CFTR (1.6% of patients with cystic fibrosis worldwide) [49]. The new molecule VX-809 has been investigated for patients carrying the p.Phe508del CFTR mutation; however, alone no clear improvement has been reported [50] and clinical trials testing the combination of different molecules are in progress. Herein, we tested a new putative therapeutic tool that specifically targets the CFTR gene [51]. Focusing on miR-101 and miR-145, we designed MBBOs that target the miRNA binding sites in the CFTR 3′UTR instead of the miRNA itself. This blockage led to the correction of CFTR channel activity through stabilisation of CFTR mRNA and increase in the protein level in nasal epithelial cells from patients homozygous for p.Phe508del, the most frequent CFTR mutation. As miR-101 and miR-145 knock-down is associated with deregulation of epigenetic pathways resulting in cancer progression [52] and lung cancer [53], our approach in which MBBOs block only their binding to their cognate CFTR mRNA motif may have therapeutic benefits by stabilising CFTR transcripts and ultimately providing enough functional protein to improve the patients' phenotype without disturbing other signalling cascades. These findings underline the importance in the continued understanding of pathways that are targeted in the lung after birth, which could ultimately lead to new targets in lung disorders, especially in cystic fibrosis.
Acknowledgements
This work was supported by grants from the French association Vaincre La Mucoviscidose, the CHU and INSERM. V. Viart and J. Bonini were supported by PhD studentships from Vaincre La Mucoviscidose. The authors thank Isabelle Vachier for her help in obtaining agreement from the French ethical research committee. The authors also thank the investigators mentioned in supplementary table S1 for the gifts of crucial reagents.
Footnotes
For editorial comments see Eur Respir J 2014; 45: 18–20 [DOI: 10.1183/09031936.00138914].
This article has online supplementary material available from erj.ersjournals.com
Conflict of interest: Disclosures can be found alongside the online version of this article at erj.ersjournals.com
- Received May 19, 2014.
- Accepted July 7, 2014.
- Copyright ©ERS 2015