Differential effects of epithelium removal on the responsiveness of guinea-pig tracheal smooth muscle to bronchoconstrictors

Contractile responses of the rat vas deferens after epithelium removal

The purpose of the present investigation was to verify the role of the epithelium in the functional response of the rat vas deferens. Our results showed that the contractile effect of cumulative doses of clonidine (3.10(-5)-3.10(-3)) was increased after the removal of the epithelium. The effect of clonidine in epithelium-free vas deferens returned to normal values when an isolated epithelium from another vas deferens was added to the organ bath, showing that the epithelium is responsible for this increase of maximum effect for clonidine. Drugs functionally or structurally related to clonidine, such as oxymetazoline, alpha-methylnorepinephrine and moxonidine, did not have their dose-response curves altered. The curves for other contractile agents, such as noradrenaline, acetylcholine, ATP, 5HT, bradykinin and histamine, or the relaxation induced by isoprenaline and forskolin were also not modified. Electrically-induced contractions at frequencies from 0.1 to 20 Hz and the mechanism of negative feed-back, brought about by clonidine (10(-10)-10(-8) M) through pre-synaptic alpha2-adrenoceptors, were not changed after the removal of epithelium. In conclusion, a significant function of the epithelium in the contractility of the rat vas deferens was demonstrated for clonidine, but not for other agonists.

Influence of the epithelium on acetylcholine release from parasympathetic nerves of the rat trachea

1. The present study was undertaken to investigate the influence of the airway epithelium on the release of acetylcholine (ACh) from parasympathetic nerves of the rat trachea. Epithelium-intact and epithelium-denuded preparations of rat trachea were incubated with [3H]-choline to incorporate [3H]-ACh into the cholinergic transmitter stores. Release of radiolabelled transmitter ACh was evoked by electrical field stimulation (60 s trains of 1 ms pulses, 5 Hz, 15 V). 2. Field stimulation both of epithelium-intact and epithelium-denuded radiolabelled tracheal preparations evoked an increase in the efflux of radioactivity; however, the mean stimulation-induced (S-I) efflux from epithelium-denuded preparations (2932 +/- 190 d.p.m., n = 9) was approximately 60% of that from epithelium-intact preparations (4802 +/- 820 d.p.m., n = 11). We have shown previously that, in epithelium-intact (but not epithelium-denuded) tracheal preparations, a substantial proportion of the S-I efflux is resistant to tetrodotoxin (1 microM) and to the removal of extracellular Ca2+, indicating that much of the S-I efflux is not caused by exocytotic release of neuronal [3H]-ACh. In epithelium-denuded tracheal preparations, superfused individually, phosphorylcholine (1 and 100 microM) did not alter S-I efflux. In epithelium-intact tracheal preparations, both in the absence and in the presence of atropine (1 microM), neither N(G)-nitro-L-arginine (100 microM), superoxide dismutase (100 units ml(-1)), indomethacin (10 microM), capsaicin (30 microM) nor alpha-chymotrypsin (1 unit ml(-1)) altered S-I efflux. 3. Experiments were also performed using two tracheal preparations superfused in series. When unlabelled epithelium-intact preparations were present in the upper chamber (superfused first), the S-I efflux from radiolabelled epithelium-denuded preparations in the lower chamber (superfused second) did not differ significantly from radiolabelled epithelium-denuded preparations superfused individually. Moreover, there was no significant difference in the S-I efflux from radiolabelled epithelium-denuded preparations in the lower chamber between experiments in which the upper chamber contained epithelium-intact or epithelium-denuded preparations. 4. Field stimulation of epithelium-intact tracheal preparations in the upper chamber with 90, 120 and 300-s periods (trains of 1 ms pulses, 5 Hz, 15 V) did not significantly alter the S-I efflux from radiolabelled epithelium-denuded tracheal preparations in the lower chamber. 5. When introduced into the upper (unlabelled epithelium-intact) and subsequently allowed to superfuse the lower (radiolabelled epithelium-denuded) tracheal preparations, the stable cholinomimetic carbachol (3 microM) markedly reduced the S-I efflux whereas ACh (0.1 and 1 microM) had no significant effect. However, in the presence of the anti-cholinesterase neostigmine (1 microM), ACh (1 microM) significantly reduced S-I efflux, indicating that ACh is subject to rapid hydrolysis by cholinesterase enzymes. When atropine (10 microM) was only exposed to radiolabelled epithelium-denuded preparations in the lower chamber, the inhibitory effects of ACh (1 microM) and carbachol (3 microM) on S-I efflux were prevented. 6. In conclusion, the findings of the present study do not support the notion that the airway epithelium exerts an inhibitory influence on ACh release from parasympathetic nerves of the rat trachea. Alternatively, if epithelium-dependent modulation of cholinergic transmission does occur in the rat trachea, then the mechanism does not appear to involve phosphorylcholine, nitric oxide, superoxide radicals, cyclo-oxygenase products of arachadonic acid, capsaicin-sensitive neuropeptides or vasoactive intestinal peptide. Moreover, the inhibitory effect of carbachol and ACh on transmitter ACh release in the rat trachea appears to be due solely to activation of prejunctional inhibitory muscarinic cholinoceptors on parasympathetic nerves and does not involve the liberation of a putative epithelium-derived inhibitory factor.

Bradykinin-induced responses in a coaxial bioassay system composed of rat anococcygeus muscle and guinea pig trachea

1. Epithelium-dependent effects of bradykinin (BK) were investigated in a coaxial bioassay system which consisted of guinea pig trachea as donor organ and rat anococcygeus muscle as test tissue. 2. BK (10(-9) to 10(-5) M) produced concentration-dependent relaxations on the phenylephrine (3 x 10(-6) M)-precontracted rat anococcygeus muscle mounted alone. Relaxations decreased significantly when muscle was mounted in epithelium-intact trachea. There was also a significant difference between the relaxations obtained in the muscle within epithelium-intact and epithelium-denuded trachea (at 10(-7) to 10(-5) M concentrations). 3. Capsaicin (10(-5) M) pretreatment did not change BK (10(-9) to 10(-5) M)-induced relaxations in each preparation compared with vehicle pretreatment. Indomethacin (10(-6) M) in combination with thiorphan (10(-5) M) and atropine (10(-6) M) did not affect the BK-induced relaxations of the muscle within capsaicin-pretreated epithelium-intact or denuded trachea. 4. CGS 8515 (a specific 5-lipoxygenase inhibitor, 10(-6) M) did not change BK (10(-5) M)-induced relaxation on the muscle alone, and caused an increase of BK-induced relaxation on the muscle within epithelium-intact trachea compared with that obtained without CGS 8515. 5. Results showed that epithelial or nonepithelial factors were capable of modulating the responsiveness of rat anococcygeus muscle to BK. The decreased relaxation by BK in anococcygeus muscle did not occur by the release of cyclooxygenase products or tachykinins from tracheal epithelium, but it may have occurred by the contractile action of lipoxygenase product secreted by nonepithelial sources. In addition, BK might stimulate the secretion of an epithelium-derived inhibitory factor from the trachea.

The mechanism of the relaxing effect of ascorbic acid in guinea pig isolated tracheal muscle

1. The effect of ascorbic acid was studied in the guinea pig isolated tracheal muscle. 2. Ascorbic acid with relatively higher concentrations produced a dose-dependent relaxation in tracheal muscle submaximally precontracted with KCl, histamine, and carbachol. 3. Removing the epithelium did not significantly alter the relaxing effect of ascorbic acid in histamine- and KCl-precontracted strips. 4. The relaxing effect of ascorbic acid is stronger in carbachol-precontracted epithelium-denuded strips than in epithelium-intact strips. 5. Indomethacin, but not L-NAME, partially inhibited the relaxing effect of ascorbic acid. 6. These results indicate that the relaxation induced by ascorbic acid in guinea pig isolated tracheal muscle does not fully depend on the presence of epithelium but is partially mediated by the production of prostanoids from smooth muscle.

K+ channel openers produce epithelium-dependent relaxation of the guinea-pig trachea

The relaxant effects of the K+ channel openers, NIP-121, (+)-7,8-dihydro-6,6-dimethyl-7-hydroxy-8-(2-oxo-piperidin-1-yl)-6H - pyrano[2,3-f]benz-2,1,3-oxadiazole, and cromakalim, were investigated in epithelium-intact and -denuded tracheal spirals isolated from guinea-pigs. In the presence of 5 microM indomethacin, NIP-121 (0.01-1 microM) and cromakalim (0.1-10 microM) relaxed, in a concentration-dependent manner, epithelium-intact and -denuded trachea precontracted with a thromboxane A2 mimetic, U46619, 9,11-dideoxy-9 alpha, 11 alpha-methanoepoxy-prostaglandin F2 alpha (30 nM). The relaxations of epithelium-denuded trachea were significantly decreased as compared with those of epithelium-intact trachea. The relaxations induced by salbutamol or aminophylline were not affected by epithelium removal. In epithelium-intact trachea, the NIP-121- and cromakalim-induced relaxations were not modulated by the neutral endopeptidase inhibitor, phosphoramidon (10 microM), or the nitric oxide synthesis inhibitor, N omega-nitro-L-arginine (100 microM). However, the guanylate cyclase inhibitor, methylene blue (100 microM), significantly reduced NIP-121- and cromakalim-induced relaxation of epithelium-intact trachea. Methylene blue also reduced sodium nitroprusside-induced relaxation but did not affect isoprenaline-induced relaxation. These findings suggest that the K+ channel openers, NIP-121 and cromakalim, may induce, at least in part, epithelium-dependent and methylene blue-sensitive relaxation of the guinea-pig isolated trachea.

Characterization of ETB receptors mediating contractions induced by endothelin-1 or IRL 1620 in guinea-pig isolated airways: effects of BQ-123, FR139317 or PD 145065

1. We have characterized the receptors mediating contractions to endothelin-1 (ET-1) or IRL 1620, an ETB receptor selective agonist, in isolated strips of tissue prepared from different parts of the guinea-pig airways. We used as antagonists BQ-123 and FR139317 (ETA receptor-selective) and PD 145065 (ETA/ETB receptor non-selective). 2. ET-1 and IRL 1620 (10(-10) M to 10(-6) M) caused similar concentration-dependent contractions of strips of guinea-pig trachea and upper bronchus. In the guinea-pig trachea without epithelium or lung parenchyma, IRL 1620 was less potent than ET-1. 3. In the trachea, contraction to ET-1 (< 10(-8) M) was preceded by a transient relaxation which was inhibited by BQ-123 (10(-5) M) or FR 139317 (10(-5) M) or by the removal of the epithelium. The concentration-response curve to ET-1 in the trachea was shifted to the right by PD 145065 (10(-5) M to 10(-4) M). PD 145065 (10(-4) M) also inhibited the response to ET-1 (3 x 10(-7) M) by 55%. Contractions induced by IRL 1620 were not affected by BQ-123 (10(-6) M) or FR139317 (10(-6) M) but were significantly attenuated by 10(-5) M of either antagonist. PD 145065 at 10(-6) M strongly attenuated and at 10(-5) M abolished contractions induced by IRL 1620. 4. In the trachea, removal of the epithelium potentiated the effects of both agonists. BQ-123 (10-5 M)had no effect on contractions of the trachea without epithelium induced by ET-1, but FR139317 (10-5 M)caused a significant inhibition. PD 145065 (10-5 M to 10-4 M) caused a shift to the right of the ET-1 concentration-response curve without affecting the contractile effect at 3 x 10-7 M. All three antagonists inhibited contractions induced by IRL 1620.5. In the upper bronchus, BQ-123 (10-5 M) did not affect contractions induced by ET-1, whileFR139317 (10-5 M) attenuated (20-26%) only contractions induced by 1-3 x 10-7 M ET-1. PD 145065(10-5 M to 10-4 M) caused a shift to the right of the ET-1 concentration-response curve. The contractions induced by IRL 1620 were inhibited by BQ-123 or FR139317 (10-5M to 10-4 M). PD 145065(10-6 M) strongly inhibited contractions induced by IRL 1620 and PD 145065 (10-5 M) totally abolished them.6. The contractile action of ET-1 in the lung parenchyma was significantly and similarly attenuated by BQ-123 (10-5 M) or indomethacin (10-5 M), while FRI39317 (10-5 M) was less effective. PD 145065(10-6 to 10-5 M) inhibited contractions to ET-1. IRL 1620, which is less potent than ET-1 in this preparation, was antagonized by PD 145065 (10-5 to 10-6 M) but unaffected by BQ-123 (10-6 M to10-5M) or FR139317 (10-6 M).7. Thus, ETB receptors mediate contractions to ET-1 in all four guinea-pig airway preparations. In addition, contractions to ET-1 in the trachea and lung parenchyma are mediated in part by ETA receptors. In the latter tissue, these ETA receptors mediate contraction through the release of cyclooxygenase metabolites. Similarly, ETA receptors located on the epithelial cells also mediate the release of prostanoids in the trachea with epithelium but they are responsible for transient relaxations. Interestingly,contractions induced by IRL 1620 were more susceptible to inhibition by the different antagonists,most probably because it binds to the endothelin receptors in a reversible manner. High concentrations(10-5 M) of ETA-selective antagonists also inhibit responses to IRL 1620, most probably by an effect at ETB receptors in both the trachea and the upper bronchus.

The relaxant action of osthole isolated from Angelica pubescens in guinea-pig trachea

The effect of osthole, isolated from Angelica pubescens, on the contraction of guinea-pig trachea was studied. Osthole (25-100 mumol/l), theophylline (10-1000 mumol/l) and higher concentrations of nifedipine (0.1-100 mumol/l) suppressed the contraction response curves of tracheal smooth muscle caused by carbachol, prostaglandin F2 alpha (PGF2 alpha), U46619 (thromboxane A2 analogue) and leukotriene C4 (LTC4) in a concentration-dependent manner. The contraction caused by high K+ (120 mmol/l) and cumulative concentrations of CaCl2 (0.03-3 mmol/l) was also inhibited concentration-dependently by osthole (25-100 mumol/l), theophylline (10-1000 mumol/l) and lower concentrations of nifedipine (0.01-0.1 mumol/l). The relaxant actions of osthole were not affected by propranolol (1 mumol/l), glibenclamide (10 mumol/l) or removal of tracheal epithelium. Osthole (100 mumol/l) was still effective in causing tracheal relaxation in the presence of nifedipine (1 mumol/l). In Ca(2+)-free- and EGTA (0.2 mmol/l)-containing medium, the relaxing effect of osthole was more potent than in normal Krebs solution. Osthole (25 and 50 mumol/l) caused 2.9 and 6.5, or 3.0 and 5.6 fold, respectively, increase in potency of forskolin or sodium nitroprusside in causing tracheal relaxation but did not affect that by cromakalim. Osthole (50 mumol/l) enhanced the increase in tissue cAMP and cGMP levels induced by forskolin and sodium nitroprusside, respectively, and in higher concentrations (100 and 250 mumol/l), itself increased markedly tissue cAMP and cGMP contents. Osthole (10-250 mol/l) inhibited the activity of cAMP and cGMP phosphodiesterases in a concentration-dependent manner. It is concluded that osthole exerts a non-specific relaxant effect on the trachealis by inhibiting the cAMP and cGMP phosphodiesterases.

Independent modulation of horse airway smooth muscle by epithelium and prostanoids

The effects of epithelial removal and cyclooxygenase inhibition on contractions induced by exogenous acetylcholine (ACh) and electrical field stimulation (EFS) were evaluated in horse tracheal strips and bronchial rings. Epithelial removal potentiated the response to ACh but had no influence on the response to EFS. The effect of epithelial removal was not altered by pretreating the tissues with meclofenamate, a cyclooxygenase inhibitor. In trachealis strips, meclofenamate augmented contractions induced by EFS but not by ACh. In bronchial rings, meclofenamate augmented EFS-induced contraction to a greater extent than ACh-induced contraction. These effects of meclofenamate were epithelium-independent. We conclude that horse airway epithelium produces a relaxant factor that is not a prostanoid. Endogenous prostanoids originating from non-epithelial sites inhibit only cholinergic nerves in the trachea but both parasympathetic nerves and smooth muscle in the bronchi.

Influence of epithelium on the inhibition of melittin-induced contraction of guinea-pig isolated trachea by the potassium channel opener NIP-121

1. We have investigated the effect of the potassium channel opener, NIP-121, on contraction elicited by melittin (a phospholipase A2 activator) in epithelium-intact and epithelium-denuded trachea isolated from guinea-pigs. The effects of NIP-121 were compared with those of isoprenaline, aminophylline and hydrocortisone. 2. In the presence of the cyclo-oxygenase inhibitor, indomethacin (5 microM), melittin (3 micrograms ml-1) caused time-dependent contraction. The melittin-induced contractile response was significantly augmented by removal of the epithelium and was concentration-dependently and completely inhibited by the phospholipase A2 (PLA2) inhibitor, p-bromophenacyl bromide (BPB 10-100 microM), and the lipoxygenase inhibitor, phenidone (10-300 microM). Similar inhibition of the melittin response by BPB (10 microM) and phenidone (10 microM) was observed in the epithelium-denuded trachea. 3. Concentration-dependent inhibition of the melittin response was induced by preincubation with NIP-121 (0.03 and 0.1 microM), isoprenaline (0.001 and 0.01 microM), aminophylline (30 and 100 microM) and hydrocortisone (100 and 300 microM). The effect of NIP-121 was abolished by glibenclamide (1 microM). 4. The inhibitory effect of NIP-121 on the melittin response was greatly reduced by removing the epithelium while that of the isoprenaline, aminophylline or hydrocortisone was not changed. 5. The inhibition of the melittin response by these drugs was similar to their inhibition of the contraction elicited by a low concentration (3 nM) of leukotriene D4 (LTD4) in the epithelium-intact trachea. Inhibition of the LTD4 response by NIP-121 was not observed in the epithelium-denuded trachea. However, higher concentrations of NIP-121 (0.3 and 1 microM) did inhibit LTD4-induced contractions of epithelium-denuded trachea.6. These findings suggest that melittin causes epithelium-dependent contraction of the guinea-pig isolated trachea which is mediated by products of lipoxygenase activity. NIP-121 may inhibit the melittin response by activating glibenclamide-sensitive potassium channels, which appear to be epithelium-dependent (an indirect effect of NIP-121 apart from its direct effect on the airway smooth muscle) while isoprenaline, aminophylline and hydrocortisone act directly to relax the trachealis smooth muscle.

Induction by endothelin-1 of epithelium-dependent relaxation of guinea-pig trachea in vitro: role for nitric oxide

1. The purpose of the present experiments was to study the underlying mechanisms responsible for the relaxant action of endothelin-1 (ET-1) in the guinea-pig trachea in vitro. 2. In tracheal strips precontracted (60-70% of the maximum) with carbachol, ET-1 (1-100 nM) evoked slowly developing concentration-dependent relaxations. Preincubation of the tissues with the thromboxane A2/prostaglandin H2 receptor antagonist, BM 13505 (5 microM) significantly potentiated the relaxant response to ET-1. 3. Removal of the epithelium changed the response of precontracted tracheal preparations to ET-1 from a relaxation to a sustained contraction. 4. ET-1-induced relaxations were abolished by methylene blue (10 microM) and were almost completely attenuated by oxyhaemoglobin (5 microM) and NG-monomethyl-L-arginine (L-NMMA, 100 microM), an inhibitor of nitric oxide synthesis, but were not altered by indomethacin (10 microM). 5. In tracheal strips under passive tension, ET-1 (1-100 nM) elicited dose-dependent contractions. The sensitivity of tissues to ET-1 was significantly enhanced by removal of the epithelium (apparent EC50 values were 28.1 +/- 4.1 and 12.5 +/- 0.8 nM in intact and rubbed trachea, respectively, n = 7, P < 0.01). 6. Preincubation of intact tracheal strips with methylene blue, oxyhaemoglobin or L-NMMA did not mimic the effect of epithelium removal on ET-1-induced contractions. 7. There was a concentration-dependent increase in thromboxane A2 but not in PGE2 and prostacyclin release from intact tracheal strips following stimulation with ET-1 (5-100 nM). 8. These results show that ET-1 exerts a dual action on guinea-pig isolated trachea: it evokes contractions at low resting tone, whereas it induces relaxations at higher resting tone. The relaxant action of ET-1 may be mediated by nitric oxide released from epithelial cells and resultant activation of smooth muscle guanylate cyclase.

Inhaled A23187 produces a preferential sensitization to substance P

We examined the effect of A23187-induced lung injury on airway responses to a variety of bronchoconstrictive aerosols in conscious guinea pigs. Guinea pigs were exposed to aerosolized A23187 or vehicle for 12 min or until labored breathing began. Animals were allowed to recover for 24 h, and then they were challenged with inhaled histamine, leukotriene D4 (LTD4), platelet-activating factor (PAF), or substance P. Eight minutes after start of the bronchoprovocative aerosol, the guinea pigs were killed and excised lung gas volume (ELGV) measurements were used as an index of in vivo airway obstruction. No differences in ELGV dose-response curves to LTD4 were seen in A23187- and vehicle-exposed animals. A23187 exposure produced small increases in both histamine and PAF sensitivity. However, A23187 caused a much more pronounced leftward shift in the dose-response to substance P. Coadministration of the neutral endopeptidase inhibitor, thiorphan, did not reduce the A23187-related airway responses to substance P. Histologic evaluation of A23187-treated lungs revealed peribronchiolar inflammation, bronchiolar epithelial injury, and mild alveolitis. We conclude that A23187 treatment produces differential airway responses to bronchoactive agents, with a preferential sensitization to substance P.

Potassium channel openers, NIP-121 and cromakalim, enhance the relaxation induced by sodium nitroprusside in the guinea-pig isolated trachea

1. The effect of the potassium channel openers, NIP-121 and cromakalim, on agonist-induced relaxation of the guinea-pig isolated trachea was investigated and the results were compared with those in the epithelium-denuded trachea. 2. Tracheal strips were incubated with a potassium channel opener or vehicle for 30 min in the presence of 5 microM indomethacin and then contracted with 30 nM leukotriene D4 (LTD4). Relaxant agents were added to the organ bath after the LTD4-elicited contraction had reached a plateau. 3. In epithelium-intact trachea, NIP-121 0.1 microM and cromakalim 1 microM, which did not modify the LTD4 (30 nM)-induced contraction, significantly enhanced the sodium nitroprusside (SNP)-induced relaxation. This enhancement of relaxation was not seen in the case of relaxation induced by the cyclic AMP-dependent bronchodilators isoprenaline, vasoactive intestinal peptide or prostaglandin E2. The enhancement of SNP-induced relaxation by NIP-121 and cromakalim was abolished in the presence of the ATP-sensitive potassium channel blocker, glibenclamide (1 microM). NIP-121 and cromakalim did not produce any significant changes in the relaxation induced by 8-bromoguanosine-cyclic monophosphate (8-Br-cyclic GMP), a cyclic GMP analogue. 4. In epithelium-denuded trachea, SNP-induced relaxation alone was significantly enhanced but that induced by 8-Br-cyclic GMP was not changed. Neither NIP-121 nor cromakalim enhanced SNP-induced relaxation in denuded trachea.5. These results suggest that in the presence of an intact epithelium the enhancement by NIP-121 and cromakalim of the relaxation of guinea-pig tracheal smooth muscle induced by SNP may be associated with the opening of glibenclamide-sensitive potassium channels.

The effect of SK&F 95654, a novel phosphodiesterase inhibitor, on cardiovascular, respiratory and platelet function

1. SK&F 95654 inhibited the guanosine 3':5'-cyclic monophosphate (cyclic GMP)-inhibited phosphodiesterase (cGI-PDE) with an IC50 value of 0.7 microM. The IC50 values were greater than 100 microM for the other four phosphodiesterase isoenzymes tested. The R-enantiomer of SK&F 95654 (IC50 = 0.35 microM) was a more potent inhibitor of cGI-PDE than was the S-enantiomer (IC50 = 5.3 microM). 2. In the guinea-pig working heart, SK&F 95654 produced a positive inotropic response without altering heart rate. 3. Oral administration of SK&F 95654 to conscious dogs caused dose-dependent increases in left ventricular dp/dtmax in the range 10-50 micrograms kg-1. These positive inotropic responses were maintained for 3 h without simultaneous changes in heart rate or blood pressure. The peak effects on left ventricular dp/dtmax were similar for orally and intravenously administered compound, indicating good oral bioavailability. 4. SK&F 95654 caused a potent inhibition of U46619-induced aggregation in both a human washed platelet suspension (WPS) (IC50 = 70 nM) and in human platelet-rich plasma (PRP) (IC50 = 60 nM), indicating that the compound shows negligible plasma binding. 5. The R-enantiomer of SK&F 95654 was twenty fold more potent as an inhibitor of platelet aggregation than was the S-enantiomer. The similarity of this ratio to that obtained on the cGI-PDE suggests that SK&F 95654 inhibits platelet aggregation via its effects on cGI-PDE. This was also indicated by studies which showed that SK&F 95654 increased adenosine 3':5'-cyclic monophosphate (cyclic AMP) levels and activated cyclic AMP-dependent protein kinase in human platelets. 6. Collagen-induced aggregation of rat PRP was also inhibited by SK&F 95654 (ICso = 65 nM). The effects of SK&F 95654, administered intravenously, on ex vivo platelet aggregation were studied in the conscious rat. At 1 mg kg-', SK&F 95654 inhibited aggregation for at least 4 h post dose and was more potent than the two other cGI-PDE inhibitors studied (siguazodan and SK&F 94120).7. In contrast to its potent effects on heart and platelets, SK&F 95654 caused only a modest relaxation of histamine- or U46619-induced bronchoconstriction in the anaesthetized, ventilated guinea-pig.8. Taken together, these results indicate that SK&F 95654 may be a suitable agent for the treatment of congestive heart failure.

Relationship between histamine, lipoxygenase and cyclooxygenase products in antigen-induced contraction in guinea-pig tracheal tube preparations

We have used a tracheal tube preparation to study antigen-induced contraction in sensitized guinea pig airways. Treatment with both the cyclooxygenase inhibitor indomethacin and the lipoxygenase inhibitor MK-886 (L-663,536) affected this contraction in preparations with intact epithelium. Indomethacin potentiated and MK-886 inhibited part of the contraction. Leukotriene release from tracheal tubes was measured after antigen challenge, and was found to be significant in preparations with an intact epithelium. When the epithelium was removed, the histamine receptor antagonist mepyramine reduced antigen-induced contraction by 90%. Our results show that when the epithelium is absent, histamine is the most important mediator in the contraction. With the epithelium left intact, the contraction is more complex: both the cyclooxygenase and lipoxygenase pathways are involved, and our findings indicate that eicosanoid production is associated with the airway epithelium.

The effect of epithelium removal on leukotriene E4-induced histamine hyperresponsiveness in guinea-pig tracheal smooth muscle

1. Removal of the epithelium resulted in a threefold increase in guinea-pig tracheal sensitivity to histamine without increasing the maximal response. 2. Preincubation of epithelially-denuded guinea-pig tracheal smooth muscle with leukotriene E4 (LTE4) in vitro increased the subsequent maximal response of the tissues to histamine. The sensitivity of the tissues to histamine was unaffected by LTE4 pretreatment. 3. Pretreatment of the epithelially-denuded tissues with the LTE4-analogue, 20-COOH LTE4, did not affect the maximal response to histamine. 4. LTE4 pretreatment increased the maximal response of the epithelially-denuded tissues to substance P (SP) but did not affect the maximal response to carbachol, KCl nor to the beta-adrenoceptor agonist, isoprenaline. 5. LTE4-induced airway histamine hyperresponsiveness was blocked by indomethacin (5 microM), GR32191 (3 microM) and atropine (1 microM). 6. Both LTE4 and U46619 pretreatment increased the contractile response of tracheal smooth muscle to electrical field stimulation. 7. It is proposed that LTE4 induces an increased maximal response of epithelially-denuded guinea-pig airway smooth muscle to both histamine and substance P via a facilitation of cholinergic neurotransmission, which is dependent upon the secondary generation of prostanoid mediator(s) acting on TP-receptors situated on cholinergic nerve terminals. Further, it is suggested that the increased maximal response of the epithelially-intact tissues to both histamine and substance P, after LTE4 pretreatment, may be suppressed by an epithelially-derived factor.

Airway epithelial cells regulate membrane potential, neurotransmission and muscle tone of the dog airway smooth muscle

1. The effects of epithelial cells were investigated on resting membrane potential and neuro-effector transmission in smooth muscle cells of the dog tracheal and bronchiolar tissues. 2. The mean value of the resting membrane potential of the epithelium-intact bronchiolar smooth muscle cells of the dog was--70.0 +/- 1.1 mV (+/- S.D., n = 40) and mechanical denudation of the epithelial layer depolarized the membrane to -57.0 +/- 2.5 mV (+/- S.D., n = 40). Application of isolated and dispersed epithelial cells (greater than 2 x 10(5) cells/ml) to the perfusing solution repolarized the membrane of epithelium-denuded bronchiolar smooth muscle cells to -67.0 +/- 2.7 mV (+/- S.D., n = 20). The mean resting membrane potential of the mucosa-free tracheal smooth muscle cells was -59.1 +/- 1.4 mV (+/- S.D., n = 50), and application of isolated and dispersed cells (greater than 2 x 10(5) cells/ml) hyperpolarized the membrane to -67.2 +/- 1.8 mV (+/- S.D., n = 50). These repolarizing actions were not modified by indomethacin (10(-5) M). 3. In the epithelium-denuded bronchioles, ACh (greater than 10(-9) M) dose-dependently depolarized the smooth muscle cells, while in the epithelium-intact bronchioles, ACh (10(-11) - 10(-8) M) did not affect the resting membrane potential. At a concentration of 10(-7) M, ACh significantly depolarized the membrane. 4. Electrical field stimulation (EFS; 50 microseconds in duration and about 10-20 V in strength) applied to ring preparations of the bronchioles evoked twitch-like contractions (hereafter referred as twitch contraction), and size of the twitch contractions gradually and continuously decreased in the presence or absence of indomethacin (10(-5) M) and guanethidine (10(-6) M). When similar experiments were performed using epithelium-denuded bronchiolar ring preparations, in no case was there a prominent reduction in the amplitude of the twitch contractions in the presence of indomethacin and guanethidine. 5. The decremental response of the twitch contraction observed in the epithelium-intact bronchioles was overcome by application of the leukotriene synthesis inhibitor AA861 (10(-6) M) and the leukotriene antagonist ONO1078 (10(-5) M). 6. Leukotrienes C4 and D4 (LTC4 and LTD4, greater than 10(-8) M) evoked muscle contraction with a steady increase in muscle tone, up to a certain level. However, at 10(-9) M, LTC4 increased and LTD4 decreased the amplitude of the twitch contractions evoked by EFS in the epithelium-intact bronchioles.(ABSTRACT TRUNCATED AT 400 WORDS)

Potential sources of artefact in the co-axial bioassay

The apparent release of relaxant activity from airway epithelium (epithelium-derived relaxing factor, EpDRF) has been examined in a co-axial bioassay system. The endothelium-denuded rat aorta, placed inside either the epithelium-intact guinea-pig trachea or rabbit bronchus relaxed in response to acetylcholine. In a modification of the standard preparation, the airway was slit longitudinally and immobilised inside a silicone rubber tube. Under these conditions, the acetylcholine-induced relaxation was abolished. Under the conditions of the co-axial bioassay, the oxygen tension in the lumen of either airway tube was lower than that of the bathing fluid. Upon addition of acetylcholine at concentrations which caused relaxation in the co-axial bioassay, the oxygen tension inside the epithelium-intact, but not the epithelium-denuded guinea-pig trachea was depressed to levels which would have affected the contractile response of a rat aorta. We suggest that the assay of relaxant activity from airways using co-axial preparations may be complicated by changes in volume and oxygen tension in the lumen of the donor airway and discuss how such problems might be avoided.

Protease injury in airways disease

In summary, proteases are present in the airway in inflammatory airways disease. These enzymes can damage the airway epithelium. As a consequence, airway function can be altered, and long-term changes in airway anatomy can result. Although the exact cellular and biochemical mechanisms that lead to these changes are incompletely described, it seems likely that they will play important roles in clinical airways disease. As such, these pathways may represent novel opportunities for therapeutic intervention.

Is the airway epithelium responsible for histamine metabolism in the trachea of guinea pigs?

The aim of the study was to investigate whether or not the epithelium plays an active role in histamine metabolism, via the histaminase and/or methyltransferase pathways. Isolated tracheas from guinea pigs sensitized to egg albumin (EA) were used. The epithelium was either left intact or removed from the preparations. The tracheal tubes were mounted in a chamber, allowing estimation of smooth muscle tension, and perfused with buffer. In some experiments the perfusate was collected for determination of histamine and methylhistamine. Mepyramine was used to evaluate the contribution of histamine to the EA-induced contraction. Mepyramine reduced the contraction by 90% when the epithelium was removed; with intact epithelium the reduction was 47%. Aminoguanidine, a histaminase inhibitor, significantly potentiated the response to EA when the epithelium was left intact. Traces of methylhistamine were detected in tissue extracts and perfusates. We conclude that histaminase is present in the preparation and that it can contribute to the inhibitory effect of the epithelium by means of histamine degradation.

The release of a non-prostanoid inhibitory factor from rabbit bronchus detected by co-axial bioassay

1. Methacholine relaxed phenylephrine-contracted aorta of the rat with the endothelium intact. This effect was inhibited by haemoglobin, methylene blue, gossypol, phenidone and L-NG-nitroarginine methyl ester (L-NAME). Rat aorta denuded of endothelium failed to relax in response to methacholine, histamine and the peptidoleukotrienes C4, D4 and E4. 2. Methacholine and histamine but not leukotrienes C4, D4 and E4 relaxed phenylephrine-contracted rat aorta without endothelium when surrounded by rabbit epithelium-intact bronchus. The muscarinic antagonist atropine antagonized the methacholine-induced relaxation. 3. Removal of the epithelium either mechanically or chemically, abolished methacholine-induced relaxation of rat aorta in the co-axial bioassay. These data indicate that the epithelium is responsible for the observed relaxant effect to methacholine and histamine. 4. The cyclo-oxygenase inhibitor, indomethacin, the phospholipase A2 inhibitor, mepacrine and the lipoxygenase inhibitor, nordihydroguaiaretic acid (NDGA), failed to inhibit methacholine-induced relaxation of rat aorta in the co-axial bioassay. This indicates that the epithelium-derived inhibitory factor (EpDIF) is not a product of the cyclo-oxygenase or lipoxygenase pathway or a product derived from activation of phospholipase A2. 5. Haemoglobin, methylene blue, phenidone, gossypol and L-NAME failed to inhibit the relaxation of rat aorta in the co-axial bioassay. These results demonstrate that EpDIF detected in the co-axial bioassay is not endothelium-derived relaxing factor (EDRF) or nitric oxide. Similarly, catalase was without effect. 6. EpDIF is unlikely to be a peptide since papain and alpha-chymotrypsin failed to alter the methacholine-induced relaxation of rat aorta in the co-axial bioassay. Furthermore, thiorphan, captopril and aprotinin were also without effect, suggesting that EpDIF is not a substrate for airway peptidases. 7. The results presented in this paper demonstrate the release of a vasoactive epithelium-derived inhibitory factor (EpDIF) from rabbit intrapulmonary bronchi by use of a co-axial bioassay preparation.

Infection of cultured human tracheal epithelial cells by human parainfluenza virus types 2 and 3

Despite growing information of the effects of human respiratory virus infection on airway physiology, little information is available on the mechanisms of pathology and pathophysiology in these infections. The human respiratory pathogens, parainfluenza virus types 2 and 3 (hPIV2, hPIV3, respectively), clinically cause laryngotracheobronchitis (infection of the large proximal airways). In order to examine the pathobiology of these viruses in airway cells of human origin, we exposed primary cultures of human tracheal epithelial cells. Primary cultures of human tracheal epithelial cells were readily infected by these agents: cells exposed to hPIV2 and hPIV3 expressed viral antigens (demonstrated by indirect immunofluorescence assay), produced infectious virus, and demonstrated cytopathic effects (including early syncytium formation). Peak viral titers of 2 x 10(7) plaque-forming units per milliliter were obtained, similar to titers from permissive CV-1 cells. Trypan blue staining and direct cell counts demonstrated no difference in the viability of the control and infected cells until the infected cells began to detach from the culture substrate. However, infected cells release significantly more LDH than control cells by 48 h following infection at a multiplicity of infection of 1 virus/target cell. This system provides a model for studying the effects of infection of the human tracheal epithelium by human respiratory viral pathogens without confounding interactions with other cell and tissue types.

Lack of local neurogenic involvement in the hyperreactivity induced by epithelium removal in guinea pig trachea

Airway epithelium removal produces tracheobronchial smooth muscle hyperreactivity to different constrictor agonists. Aside from the loss of an epithelium-derived relaxant factor, other mechanisms such as production of local axon reflex could hypothetically be involved. In the present work we investigated the effect of nerve activity inhibition or ganglionic transmission blockade in the epithelium removal-induced hyperreactivity in guinea pig trachea. Tetrodotoxin, tetracaine or hexamethonium did not modify the enhanced sensitivity of denuded tracheas to histamine, suggesting that, at least under these experimental conditions, local axon or ganglionic reflexes are not involved.

Modulation of bradykinin responses in airway smooth muscle by epithelial enzymes

We have studied the effect of epithelium removal on responses of guinea pig trachea to bradykinin (BK). BK (1 nM - 10 microM) gave a concentration-dependent relaxation when epithelium was present (E+: EC50 = 10 +/- 3 nM). Epithelium removal resulted in a biphasic response to BK with relaxation at low concentrations (E-: EC50 = 3.0 +/- 1.0 nM) and a recontraction to baseline at higher concentrations (EC50 = 2.0 +/- 1 microM). Phosphoramidon (10 microM), an inhibitor of neutral endopeptidase (NEP), which cleaves BK into inactive peptides, potentiated relaxation (EC50 = 1.0 +/- 0.9 nM in E+ and E respectively) and contraction in trachea with intact epithelium (EC50 = 0.08 +/- 0.03 microM). Inhibition of cyclooxygenase by indomethacin (5 microM), inhibited relaxation to BK in E+ tracheal segments, resulting in a slight contraction (EC50 = 1.0 microM), whereas a potent contractile response was observed in E- segments (EC50 1.6 microM, maximal contraction greater than 1 g). In the presence of both indomethacin and phosphormidon BK caused contraction, even in the presence of epithelium (EC50 = 0.2 +/- 0.11 microM), and the response in the absence of epithelium was similar to the response observed in trachea with intact epithelium (EC50 = 0.25 +/- 0.1 microM). The contractile effect of BK on airway smooth muscle may be inhibited by a protective role of epithelium, due to release of relaxant prostanoids and by degradation by epithelial NEP. In asthma, bronchoconstrictor responses to BK may be partly explained by loss of airway epithelium.

Release of substance P from guinea pig trachea leukotriene D4

Coordinated studies of leukotriene D4 (LTD4)-mediated contractile responses and LTD4-evoked release of the tachykinin substance P (SP) in both intact and epithelium abraded guinea pig tracheal smooth muscle preparations were performed. A partial contribution by axon reflex mechanisms to the magnitude of LTD4-induced tracheal contractions was suggested by a maximum inhibition of 21% and 28% by 5 x 10(-6) M tetrodotoxin (TTX) in abraded and intact preparations, respectively. SP-induced contractions were antagonized by the SP analog [DPro4DTrp7,9]-SP 4-11 in both types of preparation. The SP analog produced 58% and 72% inhibition of contractile responses to 10(-8) M LTD4 in abraded and intact preparations, respectively. Direct measurement of SP release by radioimmunoassay of the bathing medium showed TTX-sensitive LTD4-evoked release of SP. Inhibition by 5 x 10(-6) M TTX of LTD4-evoked SP release was 77%. The SP antagonist produced greater inhibition of LTD4-evoked contractions (58% in abraded, and 72% in intact preparations) than maximum TTX inhibition of LTD4-evoked contractions (21% in abraded, and 28% in intact). However, LTD4 (10(-8) M)-evoked SP release was at least 77% blocked by maximum doses of TTX. We therefore suggest that an additional agent, released by TTX-insensitive mechanisms, but whose contractile effects are also antagonized by [DPro4DTrp7,9]-SP 4-11, may participate in the LTD4 response.

Pharmacological evaluation of a guinea-pig tracheal epithelium-derived inhibitory factor (EpDIF)

1. An epithelium-derived inhibitory factor (EpDIF) released by guinea-pig tracheal epithelium was evaluated in a co-axial bioassay system consisting of an epithelium-intact guinea-pig tracheal tube surrounding endothelium-denuded rat aortic strip. 2. Histamine and several muscarinic agonists induced concentration-dependent relaxation of phenylephrine-contracted rat aorta via the release of EpDIF. However, several other agonists did not induce the release of EpDIF from guinea-pig trachea. These included the nicotinic cholinoceptor agonists nicotine (25 microM), 1,1-dimethyl-4-phenylpiperazinium (DMPP) (25 microM), calcium ionophore A23187 (0.5 microM), bradykinin (0.05-0.5 microM), substance P (5 microM), platelet activating factor (PAF, 1-100 nM), the leukotrienes (LT) LTC4, LTD4 and LTE4 (0.1-10 nM) as well as hyperosmotic stimuli. 3. Prostaglandin E2 (PGE2) induced concentration-dependent contraction of endothelium-denuded rat aortic preparations, indicating that this prostanoid could not be EpDIF. Furthermore, relaxation to histamine and methacholine, mediated via EpDIF, was not significantly altered in the presence of phenidone (50 microM) the cyclo-oxygenase/lipoxygenase inhibitor with radical scavenging properties or the cytochrome P-450 inhibitors metyrapone (1 mM) and SKF 525A (25 microM). This suggests that EpDIF is neither a prostanoid nor a cytochrome P-450 metabolite of arachidonic acid. 4. The soluble guanylate cyclase inhibitor, methylene blue (50 microM), caused small but significant increases in the potencies of both histamine and methacholine in co-axial assemblies, indicating that EpDIF did not activate this enzyme and therefore was not NO or a related substance. The beta-adrenoceptor antagonist, (-)-propranolol (1 microM), and the PAF-receptor antagonist, WEB 2086 (50 microM), also failed to alter significantly EpDIF-modulated relaxations. These data suggest that EpDIF is neither a stimulant of fiadrenoceptors nor of PAF receptors. 5. The present study provides some evidence that this vascular smooth muscle-sensitive EpDIF may not be related to the putative EpDIF previously hypothesized to modulate directly spasmogen-induced airway smooth muscle tone.

The role of the epithelium in modulating the responses of guinea-pig trachea induced by bradykinin in vitro

1. The effect of removing the epithelium on the responses of the guinea-pig isolated trachea (GPT) to bradykinin (BK) and prostaglandin E2 (PGE2) was investigated. 2. BK (3 pmol-10 nmol) induced dose-related relaxations of the intact (with epithelium), and contracted the rubbed (without epithelium) preparation of GPT. Similar responses were also obtained with PGE2 (0.3-3.0 nmol). 3. Indomethacin (1.4 microM) modified the BK-induced response of intact GPT, from a relaxation to a contraction, but inhibited the BK-induced contraction of the rubbed GPT. 4. There was a significant increase in PGE2 release from the intact GPT following stimulation with BK. 5. Removal of the epithelium from the GPT significantly reduced both basal and BK-induced generation of PGE2. 6. The induction of tone in the rubbed GPT by addition of acetylcholine (ACh) caused BK and PGE2 (0.3 nmol-3 nmol) to produce relaxations of the tissue. 7. Salbutamol (10(-8) M-10(-6) M) reduced the relaxations induced by BK on intact GPT, in a concentration-dependent manner. 8. These results suggest that both tone and an epithelial-dependent cyclo-oxygenase mechanism are important in modulating BK-induced responses of GPT.

Airway epithelium-derived inhibitory factor

Various bronchoactive agents can induce the release from the airway epithelium of an inhibitory substance that is able to relax certain tissues including rat aorta and possibly also airway smooth muscle. This substance, whose existence has recently been confirmed using a new bioassay system, is distinct from nitric oxide (EDRF) and is also known to be non-prostanoid in nature. Roy Goldie and colleagues describe the properties of this factor, and its potential clinical significance.

The effect of epithelium removal on human bronchial smooth muscle responsiveness to acetylcholine and histamine

The respiratory epithelium produces a variety of inflammatory mediators which may be influencing the bronchial hyperreactivity observed in patients with asthma. Animal studies have demonstrated that removal of the epithelium from tracheal and bronchial smooth muscle causes enhanced responses to cholinergic agonists and histamine (Hist). In this study the effect of epithelium removal on human bronchial smooth muscle response to acetylcholine (ACh) and Hist was assessed. Bronchial smooth muscle was obtained fresh from the operating theatre from 12 patients undergoing thoracotomy. Cumulative concentration effect curves (CCEC) for Hist and ACh were generated for epithelium intact and epithelium denuded muscle strips. All CCEC's were performed in duplicate and all denuded strips were obtained from the same airway immediately adjacent to the intact strip. The mean (+/- SEM) maximum response for Hist for the intact strip was 8.6 +/- 1.1 (grams/gram wet weight) and 12.0 +/- 1.4 (grams/gram wet weight) for the denuded strip (p less than 0.05). For ACh the values were 9.3 +/- 1.3 (g/g wet weight and 14.3 +/- 1.8 (g/g wet weight), respectively (p less than 0.05). The pD2 (-log EC50) for ACh was increased two-fold following epithelium removal (p less than 0.05). For Hist there was a similar increase in pD2 but this did not reach statistical significance. Thus removal of the epithelium from human isolated bronchial smooth muscle appears to modulate responsiveness to ACh and Hist. This enhanced responsiveness consequent to epithelium loss may prove important with respect to the development of worsening asthma.

Pharmacological profile of leukotrienes E4, N-acetyl E4 and of four of their novel omega- and beta-oxidative metabolites in airways of guinea-pig and man in vitro

1, The biological effects of metabolites of leukotriene E4 (LTE4) i.e. N-acetyl LTE4 (N-AcLTE4), 20-COOH-LTE4, 20-COOH-N-AcLTE4, as well as 18-COOH-19,20-dinor-LTE4 (dinor-LTE4) and 16-COOH-17,18,19,20-tetranor-14,15-dihydro-LTE4 (tetranor-LTE4) were investigated on superfused strips of guinea-pig trachea (GPT) and lung parenchyma (GPP) in vitro. 2. The actions of LTE4 were studied in isolated, superfused strips of human lung parenchyma (HP) and bronchus (HBr), in comparison with LTD4 and histamine. Effects of N-AcLTE4, the 20-carboxy metabolites, dinor-LTE4 and tetranor-LTE4 were also investigated in HBr. 3. N-AcLTE4 (0.1-10 nmol) induced dose-related contractions of GPT and was approximately 100 times less active than LTD4 (3-100 pmol). 4. In GPP, N-AcLTE4 (0.01-3 nmol) was equiactive with LTE4 (0.01-1 nmol) and approximately one order of magnitude less active than LTD4 (1-300 pmol). Contractions caused by N-AcLTE4 and LTE4 were very similar and approximately twice as sustained as those due to LTD4. 5. LTE4 (0.1-30 nmol) contracted strips of HP and HBr and was about 2-3 orders of magnitude less active than LTD4. As in GPP, the effect of LTE4 was more protracted than that of LTD4. Actions of N-AcLTE4 were similar to those of LTE4 in HBr. 6. 20-carboxy-LTE4, 20-carboxy-N-AcLTE4, dinor-LTE4 and tetranor-LTE4, all at 0.3-30 nmol, were inactive in GPT, GPP and HBr. 7. Indomethacin (2.8 microM) potentiated the effect of N-AcLTE4 in GPT, inhibited its contraction in GPP but did not affect that due to LTE4 in either HP or HBr. FPL 55712 (1.9 microM) antagonised leukotriene-induced contractions in GPT, GPP and HBr. 8. In conclusion, the metabolism of LTD4 to LTE4 or N-AcLTE4 may represent a detoxification but not an inactivation of cysteinyl-containing leukotrienes, since both metabolites still retain considerable biological activity in guinea-pig and human airways in vitro. However, further metabolism of LTE4 and N-AcLTE4 appears to result in inactivation of leukotrienes.

Epithelium derived relaxing factor: myth or reality?

Images

Is the epithelium-derived inhibitory factor in guinea-pig trachea a prostanoid?

To examine further the possible prostanoid involvement in the influence of the epithelium on guinea-pig tracheal smooth muscle responsiveness, we have analyzed the effects of LTD4, methacholine and histamine on the level of airway smooth muscle tone and on the amounts of PGE2, PGF2 alpha and PGI2 (determined by radioimmunoassay) in the presence and absence of the epithelium. Removal of the epithelium increased the sensitivity of guinea-pig trachea to the contractile effects of LTD4, methacholine and histamine. LTD4 (3-100 nM), methacholine (0.1-10 microM) or histamine (0.3-30 microM) did not increase prostanoid release above control values in either the presence or absence of the epithelium. The unstimulated release of PGE2 and PGF2 alpha, but not PGI2, was decreased in tissues lacking epithelium. Indomethacin (1 microM) reduced the baseline tone to a smaller extent in the absence of epithelium. In the presence but not the absence of the epithelium, indomethacin increased the sensitivity of preparations to the contractile effect of methacholine. The results support the postulate of an epithelium-derived inhibitory factor modulating guinea-pig tracheal smooth muscle responsiveness. The identity of this factor is not known but is not PGI2 and is unlikely to be PGF2 alpha or PGE2. However, the possibility remains that the basal release of PGE2 and/or PGF2 alpha derived from the epithelium may markedly affect the responsiveness of guinea-pig tracheal smooth muscle. Furthermore, the epithelium is a significant source of PGE2 and PGF2 alpha which may be involved in the maintenance of baseline tone.