Gastro-oesophageal reflux is a common clinical disorder associated with a variety of respiratory symptoms, including chronic cough and exacerbation of asthma. In this study, the potential role of acid-induced tachykinin release was examined in guinea pigs and rabbits, by examining the effects of the tachykinin NK1 and NK3 receptors antagonists (SR 140333 and SR 142801, respectively) (1–10 mg·kg−1) on plasma protein extravasation induced in airways by hydrochloric acid (HCl) infusion in the oesophagus.
Guinea pigs were anaesthetised with urethane, while rabbits were subject to neuroleptoanalgesia with hypnorm. Airway vascular leakage was evaluated by measuring extravasation of Evans blue dye. All animals were pretreated with atropine (1 mg·kg−1 i.p.), propranolol (1 mg·kg−1 i.p.), phosphoramidon (2.5 mg·kg−1 i.v.) and saline or tachykinin receptor antagonists (1–10 mg·kg−1 i.p.).
Infusion of 1 N HCl into the oesophagus led to a three- and five-fold increase in plasma extravasation in the main bronchi and trachea, respectively. This increase was largely prevented by the tachykinin NK1 and NK3 receptor antagonists SR 140333 and SR 142801 (1–10 mg·kg−1).
These results suggest that protein extravasation in the airways, as induced by intra-oesophageal HCl infusion, is mainly dependent on the release of tachykinins, and that both NK1 and NK3 tachykinin receptors are involved. The results suggest that HCl-induced sensory nerve stimulation may act in the periphery on intermediate neurons and/or ganglia where NK3 receptors have been shown to play an important role.
Gastro-oesophageal reflux is a common clinical disorder associated with a variety of respiratory symptoms, including chronic cough and bronchoconstriction 1–4. It has been shown to be more frequent in people with asthma 4. Experimental studies have shown that oesophageal hydrochloric acid (HCl) perfusion induces a slight bronchoconstriction in dogs 5, cats 6 and in asthmatic patients 3, 4, and that it potentiates bronchoconstriction induced by isocapnic hyperventilation of dry air or metacholine in asthmatic subjects 4.
The mechanism by which gastro-oesophageal reflux might aggravate asthmatic symptoms remains unclear. Although instillation of acid in airways produces bronchospasm in animals 6–7, pulmonary aspiration of refluxed stomach content has not been convincingly demonstrated in clinical or scintigraphical studies of large groups of asthmatic patients with gastro-oesophageal reflux 8, 9.
Many reports have suggested that gastro-oesophageal reflux could worsen bronchoconstriction in asthmatics via a vagally-mediated reflex triggered by stimulation of the oesophageal acid-sensitive receptors 4, 10, 11. However, the involvement of C‐fibres and tachykinins has recently been reported in a model of airway neurogenic inflammation, induced by intra-oesophageal HCl stimulation in guinea pigs 12.
In the airways, the activation of C‐fibre afferent nerves leads to local release of tachykinins, a group of neuropeptides, including substance P (SP), neurokinin A (NKA) and neurokinin B (NKB). Tachykinins are present in several types of peripheral tissues, including gastro-intestinal tract, lung and bladder 13. They are particularly responsible for several biological effects in the airways, including: bronchospasm, increase in microvascular permeability, vasodilatation, stimulation of glandular secretions, facilitation of cholinergic neurotransmission, and the recruitment and activation of inflammatory cells. Moreover, sensory nerves also mediate respiratory defence reflexes, such as coughing and sneezing 13–15.
The biological actions of tachykinins are mediated via three types of receptors, denoted tachykinin NK1, NK2 and NK3, which have the highest affinity for SP, NKA and NKB, respectively 16. This receptor classification has been established from receptor-binding and functional studies using selective agonists or antagonists for tachykinin receptors 16. It has now been recognised that the expression of the tachykinin NK3 receptor is confined mainly to the central and peripheral nervous system, whilst tachykinin NK1 and NK2 receptors are expressed both in the nervous system and target organs, including airways 17, 18.
Hamamoto et al. 12 have shown that HCl infusion in guinea-pig oesophagus causes plasma extravasation in the airways. This effect is abolished in animals treated with capsaicin, a drug that induces a depletion of tachykinins from sensory nerves, or by the tachykinin NK1 receptor antagonist FK 888. As these effects were observed in animals pretreated with atropine and propranolol and were abolished by bilateral vagotomy, Hamamoto et al. 12 suggested that HCl effects in that airway involve nerve pathways in the periphery.
It has been suggested that tachykinin NK3 receptors play an important role in parasympathetic ganglia and nerve transmission in the periphery 18–20. Thus, the current authors postulated that the tachykinin NK3 receptor antagonist SR 142801 21 might prevent HCl-induced neurogenic inflammation in the airways. To test this hypothesis, the airway plasma extravasation in the guinea pig was examined by stimulating the oesophagus using HCl, in the presence of SR 142801 or SR 140333, an NK1 receptor antagonist 22. To extend these observations to another species, a similar set of experiments were performed on rabbits.
Materials and methods
Oesophageal stimulation by hydrochloric acid
New Zealand white rabbits (2.5–2.8 kg) or tricolour guinea pigs (300–400 g) of either sex were used throughout the study. The guinea pigs were anaesthetised with urethane (1.25 g·kg−1 i.p.; Prolabo, Paris, France). Rabbits were subjected to neuroleptoanalgesia. They were premedicated with an intraperitoneal injection of diazepam (5 mg·mL−1, 5 mg·kg−1; Roussel Pharma SpA, Milan, Italy) and subsequently administered hypnorm (0.4 mL·kg−1; a mixture of fentanyl citrate 0.315 mg·mL−1 and fluanisone 10 mg·mL−1 i.m.; Janssen Pharmaceutical Ltd, Grove, Oxfordshire, UK). Guinea pig jugular veins or rabbit-ear marginal veins were cannulated for injection of drugs. Thirty min before experimentation, all animals were pretreated with atropine (1 mg·kg−1 i.p.; Sigma, St Louis, MO, USA) and propranolol (1 mg·kg−1 i.p.; Sigma) to block muscarinic and β‐adrenergic receptors, and phosphoramidon (2.5 mg·kg−1 i.v., 10 min before exposure to HCl (Prolabo); Sigma) to inhibit tachykinin metabolism.
The oesophageal wall was partly sectioned at the level of the 3rd–5th tracheal cartilage ring. In guinea pigs, a catheter was placed in the midoesophagus. Thus, the oesophagus was ligated at the upper portion to inhibit HCl leakage. The lower end of the oesophagus from the abdomen was then exposed and ligated to block communication between the oesophagus and the stomach. In rabbits, a second catheter with an attached latex balloon was inflated to block communication between the oesophagus and the stomach. Thus, there was no leakage of fluid from the oesophagus to the stomach.
Measurement of airway microvascular leakage
The vascular permeability was quantified by the extravasation of Evans blue dye 23. Evans blue dye (30 mg·kg−1) was injected into the rabbit's ear marginal vein or the guinea pig's jugular vein, followed 1‐min later by intra-oesophageal HCl 1 N or saline (0.4 mL). After the induction of leakage (10 min after completion of the intra-oesophageal HCl infusion), the thorax was opened and a blunt-ended 13–14-gauge needle was passed through a left ventriculotomy into the aorta. The ventricule was cross clamped and blood was expelled through an incision into the right atrium at 80 mmHg pressure with ∼100 mL (guinea pig) or ∼700 mL (rabbit) of phosphate buffer. The lung was then removed. The connective tissues, vasculature and parenchyma were gently scraped away and the airways were divided into four components: trachea, main bronchi and proximal and distal intrapulmonary airways 23. The tissues were blotted dry, placed in preweighed tubes and reweighed, and their dye content was extracted in formamide at 37°C for 18 h. Dye concentration was quantified by light absorbance at 620 nm (DCP spectrophotometer; Vital, Dieren, the Netherlands), and tissue content (ng dye·mg wet weight tissue−1) was calculated from a standard curve of dye concentrations in the 0.5–10 µg·mL−1 range.
Pretreatment with drugs
Guinea pigs or rabbits received a single dose (1–10 mg·kg−1 i.p.) of the NK3 (SR 142801; Sanofi Recherche, Montpellier, France) or NK1 (SR 140333; Sanofi Recherche) receptor antagonist, or vehicle 30 min before exposure to HCl.
A group of rabbits was treated with capsaicin (total dose 80 mg·kg−1 s.c.; Pierce, Rockford IL, USA) or vehicle (10% ethanol, 10% tween 80 and 80% NaCl 0.9%) administered over a 3‐day period (5 mg·kg−1 s.c. on day 1, 50 mg·kg−1 s.c. on day 2 and 25 mg·kg−1 s.c. on day 3). The volume of injection was 1 mL·kg−1.
Statistical analysis of results
Data are expressed as mean±sem. Statistical analysis of the results was performed using the unpaired t‐test. p‐Values of <0.05 were considered significant.
Effect of phosphoramidon on airway plasma extravasation induced by oesophageal stimulation with hydrochloric acid in guinea pigs and rabbits
There were no significant differences at any airway level in extravasation of Evans blue dye after sham stimulation with intra-oesophageal saline (0.9%) infusion in guinea pigs pretreated with either saline or phosphoramidon (fig. 1⇓). In both species, infusion of HCl (1 N) into the oesophagus without phosphoramidon significantly increased (p<0.01) plasma extravasation in the trachea and the main bronchi, and this effect was potentiated by phosphoramidon (p<0.001) (fig. 1⇓). No effects were observed in the proximal and distal intrapulmonary airways (results not shown).
Influence of hydrochloric acid concentration on airway plasma extravasation in guinea pigs
Only infusion of HCl at 1 N into the oesophagus in the presence of phosphoramidon induced a significant plasma extravasation in guinea pigs pretreated with atropine, propranolol and phosphoramidon, whereas 0.3 and 3 N HCl did not (fig. 2⇓).
Effects of SR 140333 and SR 142801 on plasma extravasation in guinea-pig airways
The plasma extravasation was significantly inhibited in guinea pigs by the tachykinin receptor antagonists SR 140333 and SR 142801 (3–10 mg·kg−1) (fig. 3⇓). When the two tachykinin receptor antagonists were administered together (3 mg·kg−1 of each), the inhibition of HCl-induced plasma extravasation in the airways was also significant (p<0.001) (fig. 4⇓), but no additional effect or synergy was observed.
Effects of SR 140333 and SR 142801 on plasma extravasation in rabbit airways
Figure 5⇓ shows that both SR 140333 and SR 142801 (1 mg·kg−1) induced a strong inhibition (p<0.01–p<0.001) of plasma extravasation in rabbit airways following HCl instillation in the oesophagus. The effects were more pronounced in trachea than in main bronchi.
Effects of systemic capsaicin treatment on hydrochloric acid-induced airway plasma extravasation in rabbits
The plasma extravasation in airways induced by HCl was significantly abolished in the capsaicin-treated rabbits (fig. 6⇓).
In the present study, it has been shown that in both guinea pigs and rabbits pretreated with atropine and propranolol to block muscarinic and adrenergic receptors, oesophageal stimulation by HCl causes neurogenic inflammation in the trachea and main bronchi. Airway plasma extravasation was abolished by capsaicin in rabbits, in agreement with previous results obtained in guinea pigs by Hamamoto et al. 12. These authors also demonstrated that this effect was inhibited in the guinea pig by the tachykinin NK1 receptor antagonist, FK 888. The current authors provide evidence that, in addition to the tachykinin NK1 receptor antagonist SR 140333, the NK3 receptor antagonist SR 142801 was able to prevent neurogenic inflammation in airways both in guinea pigs and rabbits. These results suggest that nonadrenergic, noncholinergic (NANC) nerves, communicating between the oesophagus and airways, are implied in this response and that, in addition to NK1 receptors, NK3 receptors are involved in transmission of the neurogenic response.
Neurogenic inflammation, induced by oesophageal instillation of HCl in the airways of guinea pigs or rabbits, is not associated with leakage from intra-oesophageal HCl 12, but involves nerve pathways communicating between the oesophagus and airways. A vagally-mediated autonomic pathway may be involved in the experimental models described here, as Hamamoto et al. 12 have shown that airway leakage induced by HCl was abolished in guinea pigs by bilateral vagotomy.
Nerve pathways communicating between the oesophagus and airway may also be in the periphery, via autonomic ganglia 18, 19. Anatomical studies have demonstrated the presence of axon collaterals from visceral afferent fibres (gastro-intestinal tract, bladder, airways) in autonomic ganglia 19, 24, 25,. Furthermore, studies carried out in isolated ganglia have demonstrated effects of these afferent nerves on the electrophysiological properties of the autonomic ganglion neurons after selective stimulation of their sensory/afferent nerve terminals 19. Interestingly, Canning et al. 20 have reported that, in vitro, in a preparation of guinea-pig trachea and oesophagus with associated nerves, and in the presence of propranolol, atropine, indomethacin and the tachykinin NK2 receptor antagonist SR 48968, capsaicin elicited concentration-dependent relaxations of the trachea only when the adjacent oesophagus was intact. The effect of capsaicin was abolished by tetrodotoxin. On the basis of these results and of additional immunochemistry studies, the authors concluded that the NANC responses of the guinea pig trachea were mediated by nerve fibres emanating from parasympathetic ganglia that are intrinsic to the adjacent oesophagus 20, 26, 27. In a similar manner, the current authors suggest that stimulation by HCl of capsaicin-sensitive fibres in the oesophagus may elicit neurogenic inflammation in airways through intermediate neurons, including parasympathetic ganglion neurons.
The results from the current study demonstrate that, in rabbits as in guinea pigs, both tachykinin NK1 and NK3 receptors antagonists SR 140333 and SR 142801 were able to inhibit airway extravasation induced by HCl. The inhibitory effect of NK1 receptor antagonists has previously been shown by Hamamoto et al. 12 in guinea pigs with the compound FK 888. Inhibition by FK 888 or SR 140333 may be caused by direct inhibition of endogenous tachykinin effects on vascular cells, since NK1 receptors play a predominant role in microvascular leakage induced by tachykinins 23, 28, although a role in nerve transmission cannot be excluded.
In contrast, the inhibitory effect of the tachykinin NK3 receptor antagonist, SR 142801, was less anticipated, since the NK3 receptor does not seem to be involved in the impairment of vessels and endothelial cells leading to microvascular leakage 17. An effect of SR 142801 on NK1 receptors can also be excluded, since the selectivity of this compound has clearly been demonstrated by radioligand binding and well characterised in vitro by functional assays for tachykinin receptors 21, 29, 30. In vivo, the selectivity of SR 142801 is clearly suggested by two assays: in contrast to SR 48968, SR 142801 (1 mg·kg−1) did not inhibit bronchoconstriction induced by [Nle10] neurokinin A 4–10 in anaesthetised guinea-pigs 29; and unlike SR 140333, SR 142801 (1 mg·kg−1) failed to inhibit the hypotension induced by [Sar9, Met (O2)11] substance P in guinea pigs and dogs 22, 31. It has also been shown that SR 142801 failed to inhibit tachykinin and capsaicin-induced plasma extravasation in the ear skin of mice, an effect that could be suppressed by SR 140333 32.
These results are in agreement with those of Canning et al. 20 who demonstrated in their preparations of guinea pig trachea and oesophagus with associated nerves, that SR 142801 was able to inhibit tracheal responses to capsaicin. This suggests that, in the current in vivo model, HCl-induced sensory nerve stimulation may act on intermediate neurons and parasympathetic neurons where NK3 receptors play a predominant role 18, 19.
An effect of the NK3 receptor antagonist SR 142801 on nonadrenergic, noncholinergic nerves in periphery has also been reported in the guinea pig in other experimental models of airway hyperresponsiveness. Indeed, SR 142801 is able to prevent airway hyperresponsiveness to methacholine and the increased response to histamine on microvascular leakage in airways of guinea pigs exposed 24 h earlier to an aerosol of substance P or of citric acid 29, 33. In this paper, it has been shown that SR 142801 has additional potential in inhibiting airway responses linked to acidity and gastro-oesophageal reflux.
- Received June 6, 2001.
- Accepted January 16, 2002.
- © ERS Journals Ltd