The protection of cells in the upper intestine against digestion by pancreatic trypsin depends on the prostanoid prostaglandin E2 (PGE2) and is mediated by protease-activated receptors in the epithelium. As the airway epithelium is morphologically similar and also expresses one of these receptors, PAR2, and is a major source of PGE2, we reasoned that bronchial epithelial PAR2 might also participate in prostanoid-dependent cytoprotection in the airways. Here we show that activation of PAR2, which co-localizes immunohistochemically with trypsin(ogen) in airway epithelium, causes the relaxation of airway preparations from mouse, rat, guinea-pig and humans by the release of a cyclooxygenase product from the epithelium. This physiological protective response in isolated airways also occurred in anaesthetized rats, where activation of PAR2 caused a marked and prolonged inhibition of bronchoconstriction. After desensitization of PAR2, the response to trypsin recovered rapidly by mechanisms dependent on de novo synthesis and trafficking of proteins. Our results indicate that trypsin released from the epithelium can initiate powerful bronchoprotection in the airways by activation of epithelial PAR2.

1 Responses of human bronchial strip preparations to contractile and relaxant agonists were measured in preparations from non-diseased and from asthmatic lung obtained 3-15 h post-mortem. 2 The potencies of carbachol and histamine were approximately two times less in asthmatic than in non-diseased bronchi. This was statistically significant for carbachol (P < 0.05), but not for histamine (P > 0.05). These results clearly indicate that the bronchial hyperreactivity to airway spasmogens observed in asthma is exclusively an in vivo phenomenon not involving increasing sensitivity of bronchial smooth muscle. 3 The potencies of the ,-adrenoceptor agonists isoprenaline, fenoterol and terbutaline were significantly reduced by 4-5 fold in asthmatic bronchi compared with non-diseased airways. In contrast, theophylline was equipotent in the two populations of airway preparations. Thus, it appears that severe asthma is associated with decreased bronchial smooth muscle P2-adrenoceptor function.

Interleukin (IL)-13 is a central mediator of the processes underlying the induction of airways hyperreactivity (AHR) in the allergic lung. However, the mechanisms by which IL-13 induces AHR and the associated role of inflammatory infiltrates as effector cells has not been fully elucidated. In this investigation, we show that intratracheal administration of IL-13 induces AHR in the presence and absence of inflammation. The initial AHR response (peak, 6 to 24 h; preinflammatory phase [PIP]) was dissociated from inflammation (eosinophilia) and mucus hypersecretion but was critically regulated by signaling through the IL-4 receptor alpha chain (IL-4Ralpha) and signal transducers and activators of transcription (STAT)-6. The second response (> 24 h, inflammatory phase [IP]) was characterized by an amplified AHR, eosinophil accumulation, and mucus hypersecretion. These features of the IP were not observed in IL-4Ralpha- or STAT-6-deficient mice. To determine the role of eosinophils in the induction of IP AHR and mucus hypersecretion, we administered IL-13 to IL-5-, eotaxin-, and IL-5/eotaxin- deficient mice. IL-13-mediated eosinophil accumulation was significantly attenuated (but not ablated) in IL-5-, eotaxin-, or IL-5/eotaxin-deficient mice. However, IL-13-induced AHR and mucus secretion occurred independently of IL-5 and/or eotaxin. These findings demonstrate that IL-13 can induce AHR independently of these eosinophil regulatory cytokines and mucus hypersecretion. Furthermore, IL-13-induced AHR, eosinophilia, and mucus production are critically dependent on the IL-4Ralpha chain and STAT-6.

1 The potential mediator role of the prostanoid PGE 2 in airway smooth muscle relaxations induced by peptidic and proteolytic activators of PAR-1, PAR-2, PAR-3 and PAR-4 was investigated in carbachol-precontracted mouse isolated tracheal segments. 2 The tethered ligand domain sequences of murine PAR-1 (SFFLRN-NH 2 ), PAR-2 (SLIGRL-NH 2 ) and PAR-4 (GYPGKF-NH 2 ), but not PAR-3 (SFNGGP-NH 2 ), induced smooth muscle relaxation that was abolished by the non-selective cyclo-oxygenase (COX) inhibitor, indomethacin. The relative order for mean peak relaxation was SLIGRL-NH 2 4GYPGKF-NH 2 & SFFLRN-NH 2 4SFNGGP-NH 2 . 3 SFFLRN-NH 2 , SLIGRL-NH 2 and GYPGKF-NH 2 , but not SFNGGP-NH 2 , induced signi®cant PGE 2 release that was abolished by indomethacin. Like that for relaxation, the relative order for mean PGE 2 release was SLIGRL-NH 2 4GYPGKF-NH 2 4SFFLRN-NH 2 4SFNGGP-NH 2 . 4 In dose-response studies, SLIGRL-NH 2 induced concentration-dependent increases in PGE 2 release (EC 50 =20.4 mM) and smooth muscle relaxation (EC 50 =15.8 mM). 5 The selective COX-2 inhibitor, nimesulide, but not the COX-1 inhibitor valeryl salicylate, signi®cantly attenuated SLIGRL-NH 2 -induced smooth muscle relaxation and PGE 2 release. 6 Exogenously applied PGE 2 induced potent smooth muscle relaxation (EC 50 =60.3 nM) that was inhibited by the mixed DP/EP 1 /EP 2 prostanoid receptor antagonist, AH6809. SLIGRL-NH 2 -induced relaxation was also signi®cantly inhibited by AH6809. 7 In summary, the results of this study strongly suggest that PAR-mediated relaxation in murine tracheal smooth muscle is dependent on the generation of the spasmolytic prostanoid, PGE 2 . PARstimulated PGE 2 release appears to be generated preferentially by COX-2 rather than COX-1, and induces relaxation via activation of the EP 2 receptor.

1 Quantitative autoradiographic, biochemical and functional studies were performed to investigate the endothelin receptor subtypes and signal transduction systems that mediate endothelin-I (ET-l)-induced contraction in rat isolated tracheal smooth muscle.2 Specific binding of 0.5 nM [251I]-ET-1 to tracheal smooth muscle was inhibited by at least 40% in the presence of either the ETA receptor selective ligand BQ-123 (1 gLM) or the ETB receptor-selective ligand sarafotoxin S6c (30 nM), indicating the presence of both ETA and ETB receptors in this tissue. 3 ET-1 and sarafotoxin S6c were both potent spasmogens of rat isolated tracheal smooth muscle preparations. Sarafotoxin S6c-induced contractions were unaffected in the presence of the ETA receptor antagonist BQ-123 (10 M), but were markedly attenuated in tissue previously exposed to lOOnM sarafotoxin S6c to induce ETB receptor desensitization. ET-1-induced contractions were, at most, only partially attenuated either by blocking the ETA receptor-effector system (with 10;LM or by desensitizing the ETB receptor-effector system with sarafotoxin S6c. However, ET-l-induced contractions were markedly attenuated by blocking both receptor-effector systems simultaneously. These findings suggest that ET-1 could induce contraction by stimulating either ETA or ETB receptors. 5 In Ca2'-free Krebs bicarbonate solution, 100 nM ET-1 induced a significantly larger contraction than that induced by 100 nM sarafotoxin S6c (46.6 ± 5.6% C,., versus 8.8 ± 2.8% Cmax, n = 5-7). This presumed intracellular Ca2+-dependent phase of contraction induced by ET-1 was significantly inhibited by 10 iLM BQ-123 (7.5 ± 1.0% C.). Subsequent addition of 2.5 mM Ca2+ induced a second phase of contraction. The extracellular Ca2+-dependent phase of contraction induced by ET-1 was similar in magnitude to that induced by sarafotoxin S6c (63.6 ± 4.5% C.. versus 58.0 ± 3.7% C.) and was not inhibited by BQ-123. Sarafotoxin S6c-induced contractions were not inhibited by the L-type Ca 2-channel antagonists, nicardipine or verapamil. 6 In summary, ETA and ETB receptors coexist in rat isolated tracheal smooth muscle and stimulation of both receptor subtypes contributes to ET-l-induced contraction in this tissue. However, stimulation of these receptor subtypes appears to induce contraction by activating different second messenger pathways; ETA receptor stimulation induces phosphoinositide turnover and subsequent release of intracellular Ca2+ whereas stimulation of ETB receptors facilitates the influx of extracellular Ca2+.

1 Relaxant and contractile e ects of the tethered ligand domain sequences of murine PAR-1, PAR-2, PAR-3 and PAR-4, and of the proteases thrombin and trypsin were examined in mouse isolated tracheal preparations. The epithelium-and cyclo-oxygenase-dependence of these e ects and the potential modulatory e ects of respiratory tract viral infection were also investigated. 2 In carbachol-contracted preparations, trypsin, thrombin, and the tethered ligand domain sequences of murine PAR-1 (SFFLRN-NH 2 ), PAR-2 (SLIGRL-NH 2 ) and PAR-4 (GYPGKF-NH 2 ), but not PAR-3 (SFNGGP-NH 2 ), induced transient, relaxant responses that were abolished by the cyclo-oxygenase inhibitor indomethacin. 3 Repeated administration of SFFLRN-NH 2 , SLIGRL-NH 2 or GYPGKF-NH 2 (30 mM) was associated with markedly diminished relaxation responses (homologous desensitization), although there was no evidence of cross-desensitization between these peptides. 4 The tethered ligand domain sequences for PAR-1 and PAR-4 induced a rapid, transient contractile response that preceded the relaxant response. Contractions were not inhibited by indomethacin and were not induced by either thrombin or trypsin. 5 In¯uenza A virus infection did not signi®cantly a ect the responses induced by either the proteases or peptides. Furthermore, epithelial disruption caused by mechanical rubbing had no signi®cant e ect on responses to these PAR activators in preparations from either virus-or shaminfected mice. 6 In summary, the proteases trypsin and thrombin, and peptide activators of PAR-1, PAR-2 and PAR-4 induced relaxant responses of mouse isolated tracheal smooth muscle preparations, which were mediated by a prostanoid, probably PGE 2 . Interestingly, PAR-mediated relaxations were not signi®cantly diminished following acute damage to the epithelium caused by mechanical rubbing and/or the respiratory tract viral pathogen, in¯uenza A. British Journal of Pharmacology (2000) 129, 63 ± 70 Keywords: Protease-activated receptors; in¯uenza A; trachea; smooth muscle; trypsin; thrombin; epithelium; prostanoids Abbreviations: C max , contractile response to 10 mM carbachol; EC 40R , concentration of drug producing 40%R determined by interpolation; PAR, protease-activated receptor; PGE 2 , prostaglandin E 2 ; %C, drug-induced contractile response expressed as a percentage of the carbachol-induced pre-contraction; %R, drug-induced relaxant response expressed as a percentage of carbachol-induced pre-contraction

1Endothelin-1 (ET-1) binding site densities and constrictor activities were compared in airway smooth muscle preparations of human, guinea-pig, rat and mouse. 2 The mean contractile response to 0.3 pM ET-1 (measured as the % maximum response to 10puM carbachol, % C,.., + s.e.mean) and the mean concentration of ET-1 producing 30% C,.,. (95% confidence limits) were respectively; 85.9 + 5.4% and 3.4nM (2.4-5.0) for mouse trachea (n = 11), 88.8 + 4.7% and 18.2nM (11.2-25.2) for rat trachea (n = 6), 71.0 + 7.1% and 35.2 nm (5.4-231) for human bronchus (n = 3), and 32.3 + 3.0% and 241 nm (125-460) for guinea-pig trachea (n = 6). .5 amol mm-2) > mouse trachea (28.7 + 2.6 amol mm 2) > guinea-pig trachea (8.3 + 1.8 amol mm-2). 5 A positive relationship between [125I]-ET-1 binding site density and ET-1 constrictor activity was observed in airway smooth muscle preparations from rat, human and guinea-pig. The greater sensitivity of mouse trachea to the constrictor actions of ET-1 was not dependent on the release of cyclo-oxygenaseor epithelium-derived constrictor substances, but may have been due to an inter-species difference in the receptor-effector system for ET-1.

No abstract