Abstract
Cough is the most common reason to visit a primary care physician, yet it remains an unmet medical need. Fatty acid amide hydrolase (FAAH) is an enzyme that breaks down endocannabinoids, and inhibition of FAAH produces analgesic and anti-inflammatory effects. Cannabinoids inhibit vagal sensory nerve activation and the cough reflex, so it was hypothesised that FAAH inhibition would produce antitussive activity via elevation of endocannabinoids.
Primary vagal ganglia neurons, tissue bioassay, in vivo electrophysiology and a conscious guinea pig cough model were utilised to investigate a role for fatty acid amides in modulating sensory nerve activation in vagal afferents.
FAAH inhibition produced antitussive activity in guinea pigs with concomitant plasma elevation of the fatty acid amides N-arachidonoylethanolamide (anandamide), palmitoylethanolamide, N-oleoylethanolamide and linoleoylethanolamide. Palmitoylethanolamide inhibited tussive stimulus-induced activation of guinea pig airway innervating vagal ganglia neurons, depolarisation of guinea pig and human vagus, and firing of C-fibre afferents. These effects were mediated via a cannabinoid CB2/Gi/o-coupled pathway and activation of protein phosphatase 2A, resulting in increased calcium sensitivity of calcium-activated potassium channels.
These findings identify FAAH inhibition as a target for the development of novel, antitussive agents without the undesirable side-effects of direct cannabinoid receptor agonists.
Abstract
Fatty acid amide hydrolase inhibition as a target for the development of novel, safe antitussive therapy http://ow.ly/l4ZE30dbbB1
Introduction
Chronic cough severe enough to interfere with the normal activities of daily life is thought to affect ∼7% of the population [1]. It is often associated with inflammatory airway diseases including chronic obstructive pulmonary disease, asthma, idiopathic pulmonary fibrosis, lung cancer or conditions outside the lung such as gastro-oesophageal reflux and rhinosinusitis, and is also a side-effect of drug treatments such as angiotensin-converting enzyme inhibitors [2]. Treatment-resistant cough can also be idiopathic in origin and can account for 18–42% of patients at specialist cough clinics [3]. Despite its importance, the treatment options for cough are limited and recent evidence suggests over-the-counter cough medications are ineffective [4].
The cough reflex is regulated by vagal, sensory afferent nerves which innervate the airway [2, 5]. The cell bodies for airway sensory nerves are mostly housed in the nodose and jugular ganglia. There are several known sensory nerve subtypes present in the lung: some are more mechanically sensitive, e.g. the rapidly adapting receptors, slowly adapting receptors and the subtype known as the “cough” receptor, and some are more chemosensitive, e.g. C-fibres and Aδ nociceptors. Ion channels present on these vagal nerve termini can be activated by a wide variety of stimuli to elicit cough and other reflexes. Transient receptor potential (TRP) channels are the main family of ion channels implicated in the initiation of sensory reflexes.
Previous studies have highlighted the role of cannabinoid CB2 receptors in the modulation of airway sensory nerve activity and the cough reflex [6, 7]. However, directly administered cannabinoids are not devoid of undesirable side-effects [8]. The enzyme fatty acid amide hydrolase (FAAH) is an integral membrane protein found within the nervous system, and is responsible for the hydrolysis of the endocannabinoid N-arachidonoylethanolamide (AEA; anandamide) and other related amidated signalling lipids, such as palmitoylethanolamide (PEA), N-oleoylethanolamide (OEA) and linoleoylethanolamide (LEA). These signalling lipids exert their effects through various different receptors such as cannabinoid CB1/CB2 receptors and also via noncannabinoid receptors, including TRPV1, peroxisome proliferator-activated receptors (PPARs) and opioid receptors [9–11]. Elevating endogenous levels of AEA and other bioactive fatty acid amides (FAAs) through genetic deletion or pharmacological inhibition of FAAH elicits analgesic effects in pre-clinical models of pain and inflammation [12]. These lipid signalling molecules are synthesised and released on demand and act locally, as they are rapidly inactivated. This suggests that there may be therapeutic potential in modulating the levels of these ligands, thereby reducing the potential for side-effects that result from widespread systemic cannabinoid receptor activation.
Given CB2 receptor agonists have been shown to produce antitussive activity in pre-clinical models of cough [6, 7], we hypothesised that inhibition of FAAH may represent a target for the development of novel antitussive agents. In these studies we utilised the tussive agents commonly used for clinical cough challenge studies, i.e. capsaicin and citric acid (which represents a low pH solution) [2].
Material and methods
Animals
In vivo and ex vivo experiments were conducted in male Dunkin–Hartley guinea pigs (300–500 g; 400–800 g for single-fibre in vivo studies) (Harlan, Bicester, UK and B&K Universal, Hull, UK) housed in a temperature-controlled (21°C) room with food and water freely available for at least 1 week before commencing experimentation. The experiments were performed in accordance with the UK Home Office guidelines for animal welfare based on the Animals (Scientific Procedures) act 1986 and the ARRIVE guidelines [13].
Measurement of PF-04862853 and FAA levels
In stand-alone pharmacokinetic characterisation studies, blood samples (saphenous vein bleeds) were taken at 0, 30, 60 120, 180, 240, 300 and 360 min after oral administration. Plasma levels of the novel FAAH inhibitor PF-04862853 were measured in plasma by liquid chromatography-mass spectrometry (LC-MS) as described previously [14, 15]. Plasma was prepared as described previously [14, 15] in separate studies to assess AEA, PEA, OEA and LEA levels, also measured by LC-MS.
Conscious guinea pig cough counting
PF-04862853 (1 mg·kg−1) or vehicle (0.5% methylcellulose/0.1% Tween 80 in saline) was administered per os 4 h prior to cough recording. The studies were blinded and oral dosing was performed by an independent operator. In all experiments the operator was blinded to the treatment groups. Conscious guinea pigs were placed in individual double-chamber plethysmography boxes (Buxco, Wilmington, NC, USA) and cough assessed to an aerosol of citric acid (aerosol of 0.3 M in saline for 10 min) as described previously [7, 16]. Animals were sacrificed at the end of the experiment.
Guinea pig and human isolated vagus nerve recording
Recording of depolarisation of isolated guinea pig and human vagus was carried out as described previously [2, 6, 7, 16]. Briefly, nerves were stimulated twice with capsaicin (1 µM) or low pH (pH 5) perfused for 2 min (internal control) before vehicle/FAA was perfused for 10 min. Stimulation was then re-assessed in the presence of vehicle or FAA. After 5–10 min wash, nerves were stimulated at the end of each experiment to confirm viability. For relevant experiments, vehicle/antagonist/potassium channel blocker was perfused for 10 min prior to and in the presence of vehicle/FAA. For experiments to examine the role of Gi/o subunits, vehicle (saline) or pertussis toxin (PTX) was administered (25 µg, intraperitoneally) 3 days prior to removal of the vagus nerves for the isolated vagal nerve recording experiments.
The vagus nerve was dissected from human donor en bloc lungs unsuitable for transplantation (obtained from the International Institute for the Advancement of Medicine, Edison, NJ, USA), except in one case where the vagus nerve was dissected from lung tissue of a lung transplant recipient (obtained from the Royal Brompton & Harefield NHS Foundation Trust, London, UK). Samples were from nine patients (seven males and two females; median (range) age 51 (27–67) years) with no history of respiratory disease, except the recipient sample that was from a cystic fibrosis patient. In all cases tissue was consented for use in scientific research. Ethics approval was obtained from the Royal Brompton & Harefield NHS Foundation Trust.
In vivo single-fibre recording
Guinea pigs were anesthetised with urethane (1.5 g·kg−1, i.p.) supplemented with additional urethane as required. The trachea was cannulated to maintain blood gases and pH at physiological levels by means of artificial ventilation, with tidal volume 10 mL·kg−1 and 50–60 breaths·min–1 of laboratory air. A cannula was connected to an air pressure transducer (SenSym 647; Farnell, Leeds, UK) with a side-arm to a monitor to measure tracheal pressure. The right jugular vein and carotid artery were cannulated for injecting drugs and measuring systemic arterial blood pressure, respectively. Animals were paralysed by (intravenous) administration of vecuronium bromide; initially at 100 μg·kg−1, i.v., and maintained with 50 μg·kg−1 every 20 min. A cervical incision was made to locate both vagus nerves, which were cleared from the carotid artery and cut at the central end. The left vagus nerve was teased down until a single-fibre unit firing could be recorded using platinum electrodes. C-fibres used for these experiments were identified by the lack of spontaneous discharge and responsiveness to hyperinflation and deflation, responsiveness to capsaicin/citric acid administration, and possessing a conduction velocity <1 m·s−1 [2, 17]. Upon identification of a C-fibre, the experimental protocol was as follows: stimuli (capsaicin 100 μM for 15 s or citric acid 300 mM for 60 s) were administered as aerosols to provoke action potential firing, with time between stimulations to allow recovery to baseline activity levels. PEA (10 mg·kg−1, i.p.) was then administered and then 30 min later the same stimuli were administered to evaluate the effect of PEA on action potential firing.
Imaging of intracellular calcium in isolated airway neurons
Dissection of jugular ganglia, dissociation of neurons from tissue, identification of airway terminating neurons and assessment of intracellular free calcium [Ca2+]i levels using the fluorophore Fura-2 AM were carried out as described previously [2, 18]. 50 mM potassium chloride solution was applied at the start and end of each experiment for 10 s to assess cell viability and normalise responses. Neurons were stimulated twice with low pH (pH 6) or capsaicin (100 nM) for 1 min before vehicle or PEA (1 nM) was applied for 10 min. Stimulation was then re-assessed in the presence of vehicle or PEA. The criterion for a “responsive cell” was judged as an increase in [Ca2+]i of ≥10% of the response to the 50 mM potassium chloride solution.
Compounds, materials and solutions
PF-04862853 (7-azaspiro[3.5]nonane urea) [14] was kindly supplied by Pfizer Global R&D (Cambridge, UK). SR141716A and SR144528 were purchased from Tocris Bioscience (Bristol, UK), and LEA was purchased from Cambridge Bioscience (Cambridge, UK). Fura-2 AM was purchased from Invitrogen (Paisley, UK). L-15 and Hanks' balanced salt solution were purchased from Gibco/Invitrogen (Carlsbad, CA, USA). All reagents not mentioned above were purchased from Sigma-Aldrich (St Louis, MO, USA).
For calcium imaging experiments, 50 mM potassium solution contained 50 mM KCl, 91.4 mM NaCl, 1 mM MgCl2, 2.5 mM CaCl2, 0.33 mM NaH2PO4, 10 mM glucose, 10 mM HEPES; pH adjusted to 7.4 at 37°C using KOH. For in vitro whole vagus and calcium imaging experiments, capsaicin and PEA were dissolved in 100% dimethylsulfoxide (DMSO) and solutions were diluted 1:1000 in the perfusing solution. This was Krebs–Henseleit solution [16] for vagus experiments and extracellular solution (5.4 mM KCl, 136 mM NaCl, 1 mM MgCl2, 2.5 mM CaCl2, 0.33 mM NaH2PO4, 10 mM glucose, 10 mM HEPES; pH adjusted to 7.4 at 37°C using NaOH) for calcium imaging experiments. Final vehicle concentration for an individual drug was 0.1%, or up to 0.3% when combinations of drugs were perfused together.
Data analysis and statistics
A p-value of <0.05 was considered significant. Significance was assessed by the paired t-test (where comparisons are made in the presence of drug and compared with initial control responses on the same preparation), or by either the Mann–Whitney U-test or Kruskal–Wallis with Dunn's post hoc analysis (depending on the number of treatment columns), as indicated in the figure legends.
Results
PF-04862853 inhibits FAAH, elevates FAAs and inhibits cough in a conscious guinea pig model
PF-04862853 is known to achieve peak drug levels at 4 h post-dose, and demonstrates good oral bioavailability of 53% and 33% in rats and dogs, respectively [14, 15]. Our results in guinea pig paralleled these findings, with peak levels of PF-04682853 (1 mg·kg−1, p.o.) occurring 2–4 h post-dose (figure 1a); we chose the 4 h time-point for further evaluations. PF-04862853 (1 mg·kg−1, p.o.) significantly inhibited citric acid-evoked cough (figure 1b). This dose caused robust inhibition of FAAH activity in the same animals (∼84%, similar to previous studies in rats and dogs [15]) with concomitant elevations in plasma AEA, PEA, OEA and LEA (∼4–7-fold; figure 1c). These data indicate that the antitussive activity of the FAAH inhibitor is associated with peak PF-04862853 levels, FAAH enzyme inhibition and elevation of FAAs.
FAAs inhibit guinea pig isolated vagus nerve depolarisation via the CB2 receptor
We examined the effects of PEA, AEA, OEA and LEA on tussive stimuli-induced depolarisation of guinea pig isolated vagus nerve. PEA (1 nM to 1 µM) caused a concentration-related inhibition of both low pH- and capsaicin-induced depolarisation of isolated guinea pig vagus nerve (figure 2a and b). At a concentration of 100 nM (submaximal for PEA), all four FAAs caused substantial inhibition (60–80%) of the depolarisation induced by low pH and capsaicin, which was significantly reversed by the CB2 receptor antagonist SR144528 but not by the CB1 receptor antagonist SR141716A (figure 2c–g).
PEA inhibits single-fibre firing in vivo and calcium influx in vitro in airway innervating nerves/neurons
PEA (10 mg·kg−1, i.p.) significantly reduced both total action potential firing and firing frequency of single airway innervating C-fibres in response to activation with either an aerosol of citric acid (0.3 M) or capsaicin (100 µM) (figure 3a–c).
We also utilised an in vitro preparation, whereby we examined intracellular calcium flux in isolated jugular ganglion neurons stained with the retrograde tracer dye DiI, administered intranasally in vivo, to identify airway terminating neutrons in vitro. PEA (1 nM) significantly abrogated the intracellular calcium signal induced by either low external pH (pH 6) or by capsaicin (1 µM) in vitro in jugular ganglia neurons stained with DiI (indicating airway innervating neurons) (figure 4a–d).
PEA inhibits guinea pig isolated vagus nerve depolarisation via the Gi/o signalling pathway and effects on calcium-activated potassium channels
PTX was utilised to block signal transduction via Gi/o subunits. In guinea pigs dosed with PTX, PEA inhibition of capsaicin-induced depolarisation was greatly abrogated compared with PEA inhibition in vagus nerves from vehicle-dosed control animals, indicating a role for Gi/o signalling (figure 5a). Cannabinoid effects are often suggested to be at least partially mediated by effects on potassium channels; therefore, we examined the effect of various potassium channel blockers on PEA inhibition of vagus nerve depolarisation. Clotrimazole, paxilline and glibenclamide, blockers of intermediate (IKCa) and large (BKCa) conductance calcium-activated and ATP-sensitive (KATP) potassium channels, respectively, had no significant effect on the inhibitory effects of PEA on guinea pig vagus nerve depolarisation. By contrast, apamin (1 µM), a blocker of small conductance (SKCa) calcium-activated potassium channels, almost abolished PEA inhibition (∼85–90%) of capsaicin- and low pH-induced depolarisation (figure 5b and c). Consistent with an activity on SKCa channels, PEA inhibited depolarisation of guinea pig vagus nerve evoked by agonists of multiple excitatory ion channels examined (supplementary figure S1), i.e. the TRPA1 agonist acrolein (300 μM), the synthetic TRPV4 agonist GSK1016790A (300 nM) or the endogenous TRPV4 agonist hypo-osmotic solution (−80 mOsm).
As the SKCa channel was implicated in the inhibitory actions of PEA, we examined the possible involvement of protein phosphatase 2A (PP2A), which forms a regulatory part of the SKCa channel multiprotein complex along with casein kinase II (CKII) and calmodulin (CaM). CaM is bound to the SKCa channel and translates calcium binding into channel opening; CKII phosphorylation of CaM decreases SKCa calcium sensitivity (and therefore open probability), whereas PP2A dephosphorylation of CaM increases calcium sensitivity [19]. A PP2A inhibitor (okadaic acid; 1 nM) blocked the inhibitory effects of PEA on vagus nerve depolarisation, whereas a CKII inhibitor (4,5,6,7-tetrabromobenzotriazole; 10 µM) inhibited depolarisation itself (figure 5d). These results suggest that PEA activates SKCa channels by activation of PP2A, which dephosphorylates SKCa-bound CaM to increase the channel's calcium sensitivity.
PEA inhibits human vagus nerve depolarisation via a CB2–PP2A–SKCa channel mechanism
To increase the translational nature of this study we performed key experiments in a unique human vagal sensory nerve assay assessing depolarisation in a similar fashion to the guinea pig. PEA (100 nM) inhibited both low pH- and capsaicin-induced depolarisation of human vagus nerves by ∼70–80% (figure 6). Furthermore, PEA inhibition of depolarisation was completely reversed by the CB2 receptor antagonist but not the CB1 receptor antagonist (figure 6b) and substantially reversed (∼85%) by the SKCa channel blocker (apamin; 1 μM) (figure 6c). In preliminary experiments (n=2), the PP2A inhibitor blocked the effects of PEA and an inhibitor of CKII mimicked the effects of PEA (figure 6d), suggesting PEA inhibition of human vagus nerve depolarisation is also mediated by the CB2–PP2A–SKCa channel pathway.
Discussion
Recent evidence suggests that CB2 receptors mediate inhibition of airway sensory nerve activity and cough in animal models, highlighting the CB2 receptor as a potential target for the treatment of cough [6, 7]. However, cannabinoids are associated with a variety of central effects such as sedation, nausea and addiction, which limit their use as therapies [20, 21]. The pharmacological blockade of FAAH has emerged as a potentially attractive strategy by which endocannabinoid levels are elevated and retain the beneficial effects of cannabinoid receptor activation to relieve pain [22], while avoiding the undesirable effects of global cannabinoid receptor activation.
Over the past decade, there have been advances in the development of FAAH inhibitors and initial interest focused on developing FAAH inhibitors to augment the actions of FAAs and reduce pain. More recently, emerging literature has shown that these FAAs, through interactions with a variety of receptors (extracellular and intracellular), can induce a diverse array of effects that include appetite suppression, modulation of lipid and glucose metabolism, vasodilatation, cardiac function, and inflammation [22]. However, to the best of our knowledge, the effects of FAAH inhibitors have not been studied in the airways. We hypothesised that inhibition of FAAH by a selective inhibitor may have antitussive effects. The FAAH inhibitor used in this study (i.e. PF-04862853) is a recently developed compound shown to have excellent potency, selectivity, pharmacokinetic properties and in vivo efficacy in a model of inflammatory pain [15].
PF-04862853, examined 4 h post-dose, inhibited cough in guinea pigs and was associated with increases in the plasma levels of FAAs (PEA, AEA, OEA and LEA). Although there are several subclasses of FAAs, the endogenous FAAs identified in the present study belong to the N-acylethanolamide subclass, and have been identified in mammalian tissues in the brain and in peripheral tissues [23–25]. These FAAs are referred to as “endocannabinoids” because they can function as agonists of CB1 and CB2 receptors. Given that exogenous cannabinoids exert antitussive activity via action on CB2 receptors on sensory nerves [6, 7], we hypothesised that these FAAs mediate the antitussive effects of FAAH inhibition by likewise inhibiting sensory nerve activation via the CB2 receptor.
All the FAAs examined inhibited low pH- and capsaicin-evoked depolarisation of the guinea pig isolated vagus nerve, an effect reversed by the selective CB2 receptor antagonist, but not the CB1 receptor antagonist [26, 27]. Furthermore, PEA (used as the exemplar) also inhibited capsaicin- and citric acid-evoked action potential firing in guinea pig airway innervating C-fibres in vivo, and the intracellular calcium increases induced by capsaicin or low pH in isolated airway terminating guinea pig jugular neurons in vitro. The combination of these experiments strongly suggests that the FAAs elevated by FAAH inhibition act on CB2 receptors on sensory nerves to mediate the antitussive effects of PF-04862853. Furthermore, in vagus nerve tissue from PTX-treated animals, PEA inhibition was lost, suggesting engagement of Gi/o subunit signalling downstream of the CB2 receptor. Although our data utilising the vagus nerve preparation would indicate a peripheral mechanism of action, central effects of FAAs cannot be disregarded.
Although the FAAs examined have been reported to activate CB2 receptors [28–31], previous publications have also suggested effects on CB1 receptors [29], PPAR (PPAR-α, -β and -γ) receptors [10, 31–33], and TRPV1 [9, 34, 35] and TRPM8 [36] ion channels. However, a role for CB2 receptors has been clearly identified in these studies, which utilised highly selective pharmacological tools [26, 27], that is consistent with previous studies [6, 7]. However, it is still impossible to completely rule out a contribution from noncannabinoid receptors to the in vivo activity of FAAs.
Apamin, an inhibitor of SKCa channels, blocked the inhibition by PEA of depolarisation of the guinea pig isolated vagus nerve, confirming previous reports that CB1 and CB2 receptor agonists can activate potassium channels [37]. SKCa channels are one of a family of calcium-activated potassium channels and their activation causes potassium efflux from neurons, thereby decreasing their excitability. The SKCa channel is part of a multiprotein complex along with CaM, PP2A and CKII. CaM is bound to the SKCa channel and plays a key role in translating calcium binding into channel opening. The sensitivity of CaM, and therefore of the SKCa channel, to calcium is regulated by phosphorylation/dephosphorylation, with phosphorylation by CKII resulting in decreased calcium sensitivity and dephosphorylation by PP2A resulting in increased calcium sensitivity [19]. Our finding that okadaic acid, a PP2A inhibitor, blocks the inhibitory activity of PEA on vagus nerves suggests that PEA acts to increase the sensitivity to calcium (and therefore the activity) of the SKCa channel via activation of PP2A. Key datasets were repeated in human vagus nerve preparations with similar results obtained compared with guinea pig.
In conclusion, inhibition of FAAH produces antitussive activity in conscious guinea pigs. Our data indicate that this effect could be mediated by elevated FAAs, acting on cannabinoid (CB2) receptors on vagal sensory nerves, leading to activation of SKCa channels, an effect that translates to human vagus nerves. To date, interest in this area has focused on developing FAAH inhibitors to reduce pain, and early clinical studies have indicated that a closely related analogue (i.e. PF-04457845) to the compound investigated here was well tolerated with no effect on cognitive functions at doses exceeding those required for maximal inhibition of FAAH activity and elevation of FAAs [38]. These data suggest that inhibition of FAAH with consequent elevation of FAAs could be an interesting and novel therapeutic approach to treat chronic cough. Although efficacy in pre-clinical and clinical capsaicin challenge studies (e.g. utilising TRPV1 antagonists) has not translated to efficacy in chronic cough patients, the fact that FAA ligands inhibit depolarisation of the vagus evoked by agonists of multiple excitatory ion channels gives a broader chance of success [39–41]. With clinical-ready compounds available, proof-of-concept clinical studies could be undertaken by adopting a re-positioning approach and utilising compounds from the arena of pain therapeutics. This strategy was successful recently with the P2X3 inhibitor AF-219, which was found to demonstrate unprecedented efficacy in chronic cough in treatment-resistant patients [42].
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Figure S1. Palmitoylethanolamide (PEA) inhibits acrolein-, GSK1017690A- or hypoosmolar-induced vagus nerve depolarisation via activation of CB2 receptors. PEA (100 nM) inhibited depolarisation of the isolated guinea-pig vagus evoked by the TRPA1 agonist acrolein (300 μM), the synthetic TRPV4 agonist GSK1016790A (300 nM) or the endogenous TRPV4 agonist hypoosmotic solution (-80 mOsm). Data are presented as mean±SEM, n=3-4 per group. ERJ-00782-2017_Figure_S1
Disclosures
Supplementary Material
M.G. Belvisi ERJ-00782-2017_Belvisi
M.A. Birrell ERJ-00782-2017_Birrell
C. Perros-Huguet ERJ-00782-2017_Perros-Huguet
M. Papakosta ERJ-00782-2017_Papakosta
Acknowledgements
Author contributions: M.A. Wortley, J.J. Adcock, R. Kinloch, G. McMurray, C. Perros-Huguet, M. Papakosta, M.A. Birrell and M.G. Belvisi made substantial contributions to the conception or design of the work. M.A. Wortley, J.J. Adcock, E.D. Dubuis, S.A. Maher, S.J. Bonvini, I. Delescluse, M.A. Birrell and M.G. Belvisi made contributions to the acquisition, analysis or interpretation of data for the work. All authors participated in drafting the work or revising it critically for important intellectual content and in final approval of the version to be published, and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Footnotes
This article has supplementary material available from erj.ersjournals.com
Support statement: S.A. Maher and E.D. Dubuis were funded by a project grant from the Medical Research Council (MR/K020293/1). M.A. Wortley was also funded by the North West Lung Centre Charity. The human vagus experiments in this study were undertaken with the support of the NIHR Respiratory Disease Biomedical Research Unit at the Royal Brompton & Harefield NHS Foundation Trust and Imperial College London. Funding information for this article has been deposited with the Crossref Funder Registry.
Conflict of interest: Disclosures can be found alongside this article at erj.ersjournals.com
- Received April 13, 2017.
- Accepted June 21, 2017.
- Copyright ©ERS 2017