Epithelium-dependent responses of serotonin in a co-axial bioassay system

European Journal of Pharmacology, 236 (1993) 97-105 © 1993 Elsevier Science Publishers B.V. All rights reserved 0014-2999/93/$06.00

97

EJP 53051

Epithelium-dependent responses of serotonin in a co-axial bioassay system /

/

Iclal (~akici, B a h a r T u n ~ t a n , N u r e t t i n A b a c i o ~ l u a n d I l k e r K a n z i k

Department of Pharmacology, Faculty of Pharmacy, Universityof Gazi, 06330-Hipodrom, Ankara, Turkey Received 9 July 1992, revised MS received 28 January 1993, accepted 9 February 1993

Serotonin ( 1 0 - 6 - 1 0 -4 M) produced relaxations in a concentration-dependent (at 10 -6 and 10 -5 M concentrations) manner followed by a contraction (at 10 -4 M concentration) in a co-axial system, which consisted of guinea-pig trachea as a donor organ for epithelial derived-relaxing factor(s) and phenylephrine-precontracted rat anococcygeus muscle as assay tissue. Serotonin produced a concentration-dependent contraction only in precontracted rat anococc,ygeus muscle mounted alone or mounted co-axially within epithelium-denuded trachea. Indomethacin (10 -6 M) significantly inhibited the initial relaxations (from 25.1 + 7.8 to 7.8 + 5.0% and from 35.6 + 8.7 to 10.4 + 8.3% at 10 -6 and 10 -5 M concentrations of serotonin), but did not affect the contraction. Imipramine (10 -8 M) and hydrocortisone (3 × 10 -5 M) reduced the initial relaxations (from 20.5 + 1.6 to 3.8 + 1.5% and from 32.1 + 6.4 to 18.9 + 3.9% at 10 - 6 M and 10 -5 M concentrations of serotonin, respectively) and also converted the serotonin (10 -4 M)-induced contraction to a relaxation. In the co-axial system with trachea from guinea-pigs previously sensitized with i.p. injected egg-ovalbumin, the serotonin-induced biphasic response was converted to a contractile response only after ovalbumin challenge. Histopathologic changes were observed in the epithelium of challenged tracheas taken from sensitized guinea-pigs and alterations of serotonin-induced epithelium-dependent responses were attributed to the morphological and/or functional damage of tracheal epithelium caused by ovalbumin challenge. In the modified co-axial system, phenylephrine-induced contractions faded quickly when the rat anococcygeus muscle was mounted in epithelium-intact guinea-pig trachea, and the percentage fade was significantly higher (92.3 + 2.8%) than that obtained when the anococcygeus muscle was mounted in epithelium-denuded trachea (56.9 + 8.4%) or when it was mounted alone (44.6 + 7.7%). Our results suggest that guinea-pig tracheal epithelium is capable of modulating the responsiveness of rat anococcygeus muscle to serotonin by affecting the basal or stimulated release of some inhibitory mediators. Co-axial bioassay; EpDRF (epithelium-derived relaxing factor); Ovalbumin challenge; Trachea (guinea-pig); Anococcygeus muscle (rat); 5-HT (5-hydroxytryptamine, serotonin)

1. Introduction The epithelium not only acts as a physical barrier to many large molecules (Hogg, 1981) but is also important for mucociliary clearance (Sleigh et al., 1988) and ion transport (Hogg and Eggleston, 1984). The epithelium can also act as a metabolic sink for various agents (Advenier et al., 1988; Frossard et al., 1989; Naline et al., 1989). The epithelium also produces arachidonic acid-derived metabolites in response to a variety of stimuli, including exogenous arachidonic acid, the

Correspondence to: I. (~akici, Department of Pharmacology,Faculty of Pharmacy, University of Gazi, 06330-Hipodrom, Ankara, Turkey. Tel. 90-4-2126645/3382, fax 90-4-2235018. This study was supported by the Research Foundation of Gazi University, Ankara. Preliminary results of the present study were presented at the 2nd Annual Congress of European Respiratory Society, August 29-September 3, Vienna, Austria, 1992.

Ca2+-ionophore A23187, bradykinin, platelet activating factor, and eosinophil-derived major basic protein (Orehek et al., 1975; Butler et al., 1987; Barnett et al., 1988; Jacoby et al., 1988). More recently, the epithelium has been demonstrated to produce non-prostanoid inhibitory factor (Hay et al., 1987; Fernandes et al., 1989). Epithelial damage has been observed in bronchial biopsy samples taken from patients with clinically stable allergic asthma (Laitinen et al., 1985; Beasley et al., 1989; Jeffery et al., 1989). It is conceivable that damage to the airway epithelium may compromise many of the epithelial cell functions listed above. There is some evidence to support this concept. For example, epithelial denudation has been found to increase the responsiveness of isolated cartilaginous non-asthmatic human airways to acetylcholine, histamine, and prostaglandin F2,, (Raeburn et al., 1986; Aizawa et al., 1988; Naline et al., 1989; Fernandes et al., 1990). Similar results

98 have been reported for canine (Flavahan et al., 1985), bovine (Barnes et al., 1985) and guinea-pig airways (Goldie et al., 1986). The immunological sensitization of laboratory animals, in particular of the guinea-pig, has become a widely adopted model of antigen-induced bronchoconstruction (Hand et al., 1982; Selig et al., 1988). We therefore used the ovalbumin-induced active sensitization model in the experiments. In the present study we used the co-axial bioassay technique to evaluate, firstly, whether the epithelium is involved in serotonin-induced responses in this system and, secondly, if serotonin-induced epithelium-dependent responses change after in vitro antigen challenge. We also used a modified co-axial bioassay system to confirm the ability of phenylephrine to maintain a sustained plateau and to investigate if there is a significant basal or stimulated release of some inhibitory mediator(s) in the guinea-pig tracheal epithelium.

10 -6 M)were exposed to cumulative concentrations of serotonin (10-6-10 -4 M). Then the same rat anococcygeus muscle was passed co-axially through either epithelium-intact or epithelium-denuded guinea-pig trachea, as described before (Giic et al., 1988), and mounted under the same resting tension as described above; a 1-h equilibration period was allowed. Serotonin was retested on the co-axial system. At the end of each experiment, papaverine (10 -4 M) was added to confirm the ability of the muscle to relax. In some experiments the effects of various inhibitors on serotonin-induced responses were investigated. The compounds examined were indomethacin (10 -6 M, 10 -5 M), CGS 8515 (10 -6 M) (a specific 5-1ipoxygenase inhibitor), methysergide (3 x 10 -6 M), phentolamine (10-8-5 X 10 -7 M), imipramine (10 -8 M), and hydrocortisone (3 x 10 -5 M).

2.3. Histological study 2. Materials and methods

2.1. Materials The composition of Krebs-Henseleit (mmol/1) was NaC1 118.0, KCI 4.7, CaCI2.H20 2.5, KH2PO 4 1.2, MgSO4.7H20 1.2, NaHCO 3 25.0 and glucose 10.0. The drugs used were phenylephrine hydrochloride, 5hydroxytryptamine creatinine sulfate complex, indomethacin, papaverine hydrochloride, phentolamine hydrochloride, egg ovalbumin, hydroquinone and N '°nitro-L-arginine, and were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.). CGS 8515 (4-N-(2amino-methylbenzoate)-1,2-o-naphthoquinone) and imipramine hydrochloride were kind gifts from CibaGeigy Corporation (New Jersey, USA and Istanbul, Turkey), Methysergide bimaleate was obtained from Sandoz Res. Lab. (Basel, Switzerland) and hydrocortisone was from Medichemex Ltd. (I~ndon, England). All drug solution were prepared on the day of each experiment. Ovalbumin was dissolved in 0.9% NaCI.

2.2. Co-axial bioassay experiments Male rats (200-250 g) were killed by a sharp blow on the head and bled. Anococcygeus muscle was isolated and prepared according to Gillespie (1972). The muscles were suspended under 750-mg tension in Krebs-Henseleit solution, maintained at 37°C, aerated with 5% CO 2 in 0 2 and equilibrated for 1 h. Although rat anococcygeus muscle has no cholinergic innervation, atropine (10 -7 M)was present throughout these experiments to inhibit any possible basal or serotoninstimulated released of acetylcholine from cholinergic nerves in guinea-pig trachea. Rat anococcygeus muscle preparations precontracted with phenylephrine (3 x

The epithelium of trachea segments was denuded by inserting a moistened wooden applicator stick into the lumen and by gently rubbing it with a corkscrew motion 5-6 times (Holroyde, 1986). At the end of each experiment rubbed and unrubbed tracheas were fixed with 10% buffered formalin. One week later, tracheas were embedded in paraffin, and sections were stained with hematoxylin-eosin. Tracheas removed from sensitized guinea-pigs before or after ovalbumin challenge were treated similarly. A Olympus BHS/BHT photomicrograph was used to monitor and photograph the preparations.

2.4. Sensitization procedure Guinea-pigs of either sex, weighing 250-350 g, were actively sensitized by the i.p. injection of ovalbumin (10 mg-kg -1) on days 1, 3 and 5. In the control group, 0.9% NaCI solution was injected i.p. on the same days. All animals were given a standard diet and kept at room temperature (18-20°C) until the day of challenge. The animals were killed by a blow on the head and bled 21 days after the last injection, and tracheas were isolated. Antigen challenge was performed by adding cumulative concentrations of ovalbumin (0.1 /~g-0.1 mg. m1-1) into the tissue bath. The tracheas were exposed to all concentrations of ovalbumin for 35-45 min. The effects of serotonin on phenylephrineinduced contractions were tested before ovalbumin administration and 20 min after challenge with the last concentration of ovalbumin (0.1 mg-ml-1).

2.5. Modified co-axial system The co-axial system was modified by fixing a glass tube to the organ holder and the system was prepared

99 Q

in this glass tube. With this modification we were able to apply the drugs intraluminally and we could also prevent the serosal diffusion of possible relaxing factors. In this system, the percentage fade was calculated as shown below: % fade = [(h - r ) / h ] x 100, where h is the height of phenylephrine-induced contraction (mm), and r is the height of the plateau phase of the phenylephrine-induced contraction after it has faded (mm)

~HT~ ~

/ P~e

2.6. Statistical analysis

5 IHTi

/ plhe

i

~

ii

b

Responses to serotonin were calculated as percentages of the phenylephrine-induced precontraction. All concentrations are given as the final concentration in the bath solution. Results are expressed as the means _ S.E.M. and were analyzed statistically by using Student's t-test for paired or unpaired samples, where appropriate. The null hypothesis was rejected if the probability was less than or equal to 0.05.

s-HTt ~ ,I

S-HTt S 4

i

5-HT

ii

5 4 iii

Fig. 1. Responses to serotonin in (a) phenylephrine (Phe)-precontracted rat anococcygeus muscle: (i) control; (ii) effect of methysergide (3 x 10 -6 M); (b) basal tone of rat anococcygeus muscle: (i) control; effects of (ii) phentolamine (10 - 8 M), (iii) phentolamine (5 x 10 - 7 M).

tracted by serotonin (8.1 + 1.2, 31.6 + 3.2, and 41.1 + 4.0% at 10 -6, 10 -5 and 10 -4 M, respectively) (n = 6) (fig. 2). The same preparation in a co-axial system with donor epithelium-intact guinea-pig trachea tissue relaxed in response to initial concentrations of serotonin ( - 1 7 . 7 + 2.7 and - 2 4 . 3 + 3.1% at 10 -6 and 10 -5 M concentrations, respectively) (n = 10) and then gave a contraction (19.7 + 3.0%) which was smaller than the phenylephrine-induced contraction even at the highest concentration tested (10 -4 M) (fig. 2). Relaxation was not observed if the epithelium of guinea-pig trachea was removed mechanically. Contractions were 8.1 + 2.1, 24.6 + 4.0 and 31.6 + 5.6% at 10 -6, 10 -5 and 10 -4 M, respectively (n = 10) (fig. 2).The serotonin-induced responses in the co-axial system with epithelium-intact guinea-pig trachea were not influenced by CGS 8515 (10 -6 M) (n = 3), but indomethacin (10 -6 M) (n = 5) partly inhibited the initial relaxations. At a higher

3. Results

In preliminary experiments we demonstrated that serotonin-induced contractions of precontracted rat anococcygeus muscle were antagonized by methysergide (3 × 10 -6 M). In addition, serotonin-induced contractions of the rat anococcygeus muscle in resting tone were partly antagonized by 10 -8 M and completely by 5 x 10 -7 M phentolamine, suggesting that serotonin-induced responses were mediated by noradrenaline release a n d / o r that phentolamine produced an antiserotonin effect (fig. 1).

3.1. Responses of the co-axial bioassay preparation to serotonin Rat anococcygeus muscles precontracted with phenylephrine (3 x 10 -6 M) were additionally conTABLE 1

Effect of indomethacin or CGS 8515 on serotonin (5-HT)-induced responses of phenylephrine-precontracted rat anococcygeus muscle in the co-axial bioassay system. Responses were calculated as the percentage of the phenylephrine-induced precontraction. Data are expressed as means___ S.E.M. M: Molar concentration, n: N u m b e r of experiments, a Significantly different from its own control (P < 0.05). Preparation

n

Control (epithelium-intact) Indomethacin (10 - 6 M)

5 5

Control (epithelium-denuded) Indomethacin (10 - 6 M)

4 4

Control (epithelium-intact) Indomethacin (10 -5 M)

6 6

Control (epithelium-intact) CGS 8515 (10 - 6 M)

3 3

5-HT ( - log M) 6

5

-25.1+ -7.8+

7.8 5.0 a

4.8 + 2.6 5.7+ 0.8

4

-35.6+ -10.4+

8.7 8.3 a

20.1+ 3.6 16.6+ 5.3

22.3 + 7.3 20.2+ 6.8

30.9 + 10.9 16.9+ 5.9

-19.9-t- 6.2 - 4 . 0 + 2.5 a

- 2 0 . 1 + 8.1 -21.5+10.5

19.6+ 7.7 12.1+ 8.9

-19.5+11.4 - 1 6 . 9 + 1.1

- 2 9 . 4 + 7.9 -25.3-]- 2.5

18.6+ 4.9 8.4+ 1.1

100

60

Z o I-

-6 10

-5 10

-4 10 M

6

10

10

l o- 6 7

n z 0

40

16 5

1154 M

7

7

n

40-

I-o

o

rr l-Z 0 0

rr Z 0 0

20

U..l Z

LU

Z

"I"

o. w ..1 >. z w z a.

°

!

LU _I >Z LU 'I" Q. -20

-20

-40

-40

2. Responses to serotonin of phenylephrine (3 x 10-6 M)precontracted rat anococcygeus muscle. [] Rat anococcygeus muscle alone; • mounted in epithelium-intact guinea-pig trachea; [] mounted in epithelium-denuded trachea. Values are means + S.E.M. n = number of experiments. * Significant when compared with rat anococcygeus muscle mounted alone. ** Significant when compared with rat anococcygeus muscle within epithelium-intact trachea.

Fig.

concentration (10 -5 M), indomethacin had a similar effect on both the relaxant and contractile phase of the serotonin-induced response. After epithelium removal, indomethacin (10 -6 M) did not affect the serotonin-induced contractions (table 1). Hydrocortisone (3 x 10 -5 M) significantly reduced the serotonin-induced responses of the rat anococcygeus muscle. In the co-axial system, hydrocortisone did not produce a significant inhibition of the relaxant, but potentiated the contractile phase of the serotonin-

Fig. 3. Responses to serotonin of phenylephrine ( 3 x 10 -6 M)-precontracted rat anococcygeus muscle. [] Mounted in epithelium-intact trachea from unsensitized guinea-pig (after ovalbumin challenge); • mounted in epithelium-intact trachea from sensitized guinea-pig (before ovalbumin challenge); • mounted in epitheliumintact trachea from sensitized guinea-pig (after ovalbumin challenge). Values are means + S.E.M. n = number of experiments. * Significant when compared with rat anococcygeus muscle within epithelium-intact trachea from unsensitized guinea-pig (after ovalbumin challenge) and sensitized guinea-pig (before ovalbumin challenge).

induced response. In the co-axial system, imipramine (10 -8 M) in combination with hydrocortisone (3 x 10 -5 M) significantly reduced the serotonin-induced relaxant phase and converted the serotonin-induced contraction to a relaxation (table 2). Indomethacin (10 -6 M) added to this combination did not produce an additional significant inhibition (table 2).

TABLE 2 Effect of hydrocortisone and imipramine alone and in combination with indomethacin on the serotonin (5-HT)-induced responses of phenylephrine-precontracted rat anococcygeus muscle (RAM) mounted alone or in the co-axial bioassay system. Responses were calculated as a percentage of the phenylephrine-induced precontraction. Data expressed as means+S.E.M. M: Molar concentration, n: Number of experiments, a Significantly different from its own control (P < 0.05). Preparation

n

5-HT ( - log M) 6

5

4

RAM alone Control Hydrocortisone ( 3 × 10 - s M)

7 7

5.5 ___1.3 0.3__.0.2 a

40.2 + 4.3 23.2+ 2.7 a

29.6 + 0.8 8.3+ 1.9 a

Coaxial system Control (epithelium-intact) Hydrocortisone (3 x 10-5 M) a

4 4

- 17.9 + 6.2 - 8.7 -1-4.2

- 26.8 + 12.6 - 10.1 + 9.3

20.5 + 1.4 53.8+ 7.9 a

Control(epithelium-intact) Hydrocortisone (3 x 10 -5 M) + imipramine (10 -6 M) a Hydrocortisone (3 × 10-5 M) + imipramine (10 -5 M) + indomethacin (10-6 M ) a

4

-20.5+1.6

~32.1 + 6.4

19.3 + 5.4

4

- 3.8 + 1.5

- 18.9 + 3.9 a

- 2 5 . 6 + 13.6 a

4

- 2.1 + 2.1

- 10.5 + 1.4

-

17.2+ 8.8

101 a

Fig. 4. Histological sections (hematoxylin and eosin staining: × 100 magnification) of guinea-pig trachea. (a) Epithelium-intact trachea from unsensitized guinea-pigs, (b) epithelium-denuded trachea from unsensitized guinea-pigs; epithelium-intact trachea from sensitized guinea-pigs, (c) before, (d) after ovalbumin challenge. L: lumen, E: epithelium, C: cartilage, Inf: inflammation.

When the rat anococcygeus muscle was mounted co-axially within epithelium-intact trachea obtained from sensitized guinea-pigs, serotonin caused a biphasic effect similar to the response obtained in the co-axial system with unsensitized guinea-pig trachea before ovalbumin challenge ( - 23.7 + 5.8, - 23.0 + 10.5 and 19.5 + 4.6% at 10 -6, 10 -5 and 10 -4 M concentrations, respectively) (n = 7) (fig. 3). However, serotonin-induced relaxations reversed to contractions after ovalbumin challenge (1.3 + 5.7, 15.8 + 8.3 and 18.5 + 10.2% at 10 -6, 10 -5 and 10 -4 M, respectively) (n = 7) (fig. 3).

3.2. Results of histological examination Rubbing the luminal surface of the trachea removed the epithelial layer completely without damaging the underlying tissues (fig. 4b). No significant light microscopic alterations were seen in sensitized but not in antigen (ovalbumin)-challenged tracheas. Basement membrane, ciliated epithelium, and goblet cells were observed (fig. 4c). However, in the challenged tracheas taken from sensitized guinea-pigs epithelial desquamation was common and goblet cells were not observed. The epithelium appeared to be

swollen and intracellular spaces were widened, giving the tissue a fragile appearance. The lamina propria was edematous and lymphatic edema was also seen. Hemorrhagic loci were noted in the capillary lumen and tracheal adventitia. Recruitment of eosinophils and inflammation were observed in the vasculature, connective tissue, and smooth muscle (fig. 4d).

3.3. Responses of the modified co-axial bioassay preparation to phenylephrine When the rat anococcygeus muscle was mounted in epithelium-intact guinea-pig trachea in the modified co-axial system, each concentration of phenylephrine a

b

c

lgT_~

Stain

Fig. 5. Responses of the rat anococcygeus muscle to phenylephrine

(e 3x 10.6 M) in the modified co-axial bioassay system. (a) Alone, (b) within epithelium-intact trachea, (c) within epithelium-denuded trachea of guinea-pig, in the glass tube.

102 TABLE 3 Fade of the phenylephrine (3 × 10 - 6 M)-induced contractions in the modified co-axial system. Responses were calculated as the percentage of phenylephrine-induced precontraction. Data expressed as means+ S.E.M. M: Molar concentration. n: Number of experiments, a Significantlydifferent from rat anococcygeusmuscle (RAM) alone and guinea-pig trachea (GPT) (epitheliumdenuded) + RAM (P < 0.05). Compound

RAM alone

GPT (epithelium-intact) + RAM 92.3 + 2.8 a

Propranolol (10-6 M)

44.6 + 7.7 (n = 4) -

Control

GPT (epithelium-denuded) + RAM

Treatment _ (n = 13)

93.1 + 4.3

93.4+ 1.7

56.9 + 8.4 (n = 13) -

(n = 3) Imipramine (10-6 M)

-

97.4 + 1.9

97.1 + 1.5

-

93.8 + 3.3

-

90.3 + 1.2

-

89.2 + 6.8

-

93.8 + 2.1

-

(n = 4) Indomethacin (10 - 6

M)

-

97.8 + 1.4 (n = 8)

CGS 8515 (10-6 M)

-

95.0 + 0.8 (n = 3)

H y d r o q u i n o n e (10 - 6 M )

-

96.8 + 1.3 (n = 4)

N~-Nitro-L-arginine (10-4 M)

-

90.2 + 2.8 (n = 6)

produced a concentration-dependent contraction. This contraction faded quickly when phenylephrine was applied intraluminally (data not shown). At the precontraction concentration ( 3 × 10 -6 M), the phenylephrine-induced contraction faded when the rat anococcygeus muscle was mounted in epithelium-intact or epithelium-denuded guinea-pig trachea (fig. 5). In epithelium-intact guinea-pig trachea, the percentage fade was significantly greater (92.3 + 2.8%) than that obtained with epithelium-denuded guinea-pig trachea in the modified co-axial bioassay system (56.9 + 8.4%) (table 3). When the rat anococcygeus muscle was mounted alone in the glass tube of the modified co-axial bioassay system, the intraluminal phenylephrine (3 × 10 -6 M)-induced contraction faded less (44.6 + 7.7%) than in epithelium-intact guinea-pig trachea (fig. 5). There was no statistically significant difference between the responses of rat anococcygeus muscle alone and within epithelium-denuded guinea-pig trachea in the modified co-axial system (table 3). None of the antagonists (propranolol, 10 -6 M; imipramine, 10 -6 M; indomethacin, 10 -6 M; CGS 8515, 10 -6 M; hydroquinone, 10 -6 M and N'°-nitro-Larginine, 10 -4 M) influenced the percentage fade of the phenylephrine-induced contractions (table 3).

4. Discussion The increased sensitivity of airway smooth muscles of several species to bronchoconstrictors and bronchodilators following mechanical removal of the epithelium is attributed to three main reasons. (1) Since the epithelium imposes a physical layer between the

underlying airway smooth muscle and the external environment, the hyperresponsiveness observed on removal of the epithelium may be due to the simple disappearance of a diffusion barrier (Holroyde, 1986; U n d e m et al., 1988). (2) Removal of the epithelium may reduce the enzymatic breakdown of contractile agonists and thus potentiate their effects (Devillier et al., 1988; Fine et al., 1989; Frossard et al., 1989; Advenier et al., 1988; Farmer et al., 1986). (3) An epithelium-derived relaxing factor could exist which modulates the responsiveness of airway smooth muscle (Flavahan et al., 1985; Hay et al., 1987; Fernandes et al., 1989). In the co-axial system, the role of epithelium acting as a diffusion barrier is unlikely to be a problem as in this system agonist reaches the assay tissue (rat anococcygeus muscle) without having to cross the epithelium. Antagonism of serotonin-induced contractions of rat anococcygeus muscle by methysergide (a nonspecific 5-HT 1 and 5-HT 2 antagonist) or phentolamine (a-receptor blocker) showed that responses were mediated by specific serotonin receptors (subtypes were not investigated) (fig. 1). In the presence of epithelium-intact guinea-pig trachea, conversion of serotonin-induced contractions to relaxations (at the initial concentrations) implied that the alterations might be due to either the smooth muscle or the epithelium of guinea-pig trachea. After mechanical removal of the epithelium of guinea-pig trachea, only a contractile response, similar to that noted with rat anococcygeus muscle alone, was observed, indicating that this relaxation was epitheliumdependent (fig. 2). Therefore, the possibility that there is airway epithelium-dependent release of an inhibitory

103

factor into the lumen which can modulate the responsiveness of rat anococcygeus muscle to serotonin was investigated. Earlier studies provided evidence that serotonin or acetylcholine stimulated the release of an inhibitory factor in some airway preparations (Flavahan et al., 1985; Barnes et al., 1985; Ilhan and ~ahin, 1986). In the present study, the results obtained with indomethacin suggest that serotonin caused the release of an inhibitory cyclooxygenase metabolite of arachidonic acid from the epithelium of guinea-pig trachea. After epithelium removal indomethacin did not affect the serotonin-induced contractions although it reduced the serotonin-induced relaxations significantly (at 10 -6 and 10 -5 M concentrations) when the epithelium was intact (table 1). The airway mucosa of the guinea-pig trachea has been shown to be an important source of cyclooxygenase metabolites (Orehek et al., 1975). These cyclooxygenase products are not prostaglandin F2~ or prostaglandin E2, because these arachidonic acid products have a direct effect on the rat anococcygeus muscle, causing a small increase in tone (Al Timimi et al., 1978). However, this does not preclude the involvement of other cyclooxygenase metabolites of arachidonic acid. Leukotrienes, generated by lipoxygenase metabolism of arachidonic acid, may also be involved in this modulatory role of the epithelium. However, the lack of effect of CGS 8515 (table 1) suggests that 5-1ipoxygenase products are not involved. Other studies with the guinea-pig trachea which have demonstrated that indomethacin treatment had no effect on epithelium-dependent changes in sensitivity to contractile agonists (Hay et al., 1987; Holroyde, 1986) or on the release of an epithelium-dependent vascular relaxant factor (Fernandes et al., 1989) would argue against such a contention. The inability of indomethacin to reverse the relaxations completely to contractions at the high dose (10 -5 ) used would suggest that in addition to the release of certain prostanoids, an epithelium-derived inhibitory mediator is also released which is not a prostanoid. The results obtained with hydrocortisone (extraneuronal uptake blocker) in the co-axial system imply that tracheal epithelial cells might take up amines but that the rat anococcygeus muscle does not (see the results on the rat anococcygeus muscle alone, table 2). The inhibition by hydrocortisone of serotonin-induced contractions of rat anococcygeus muscle alone might be attributed to the fact that inhibition of extraneuronal uptake enhances neuronal uptake (Langer, 1970; Hughes, 1971). It has also been suggested that neuronal uptake of amines is more prominent than tissue uptake in the rat anococcygeus muscle (Nash et al., 1974). The results obtained with imipramine in combination with hydrocortisone confirmed this suggestion. Thus, addition of imipramine to the bath in the presence of hydrocortisone converted the contraction in-

duced by a high concentration of serotonin to a relaxation, probably by inhibiting neuronal uptake and causing an enhancement on epithelium-dependent responses to serotonin. These results suggest that inhibition of uptake potentiated the ability of serotonin to release prostanoid a n d / o r especially nonprostanoid relaxing factors from the epithelium of guinea-pig trachea, because indomethacin added to imipramine and hydrocortisone in combination did not produce any significant inhibition of the response (table 2). In the co-axial system with epithelium-intact guineapig trachea isolated from sensitized guinea-pigs, epithelial damage (morphological and functional) due to ovalbumin challenge might have contributed to the reversal of the serotonin-induced contractions. Histological examination confirmed this suggestion. Alterations and damage were only seen in the epithelium of sensitized trachea after ovalbumin challenge (fig. 4d) but not in the epithelium of unchallenged trachea of sensitized guinea-pig (fig. 4c). These results are compatible with our earlier findings (~akici et al., 1990). In some animal models of anaphylaxis, airway edema, disruption of mucosiliary transport, decrease in airway caliber and compliance, and alterations in epithelium were also found in challenged airway preparations (Persson and Karlsson, 1987; Selig et al., 1988). In the modified co-axial system, the fading of the phenylephrine-induced contractions might be attributed to luminal hypoxia, as suggested by Gunn and Piper (1991), or to the secretion of an inhibitory mediator. The former notion cannot be rejected, but there should not be hypoxic changes caused by constriction of the tracheal lumen in this system because the agonist we used (phenylephrine) has no contractile effect on the tracheal smooth muscle of guinea-pig. In addition, the results obtained in the presence of epithelium-intact or epithelium-denuded guinea-pig tracheas also confirmed the involvement of the epithelium (table 3, fig. 5). Thus, the percentage fade observed in the presence of guinea-pig trachea with an intact epithelium was greater than that obtained in the rat anococcygeus muscle alone or in the presence of epitheliumdenuded guinea-pig trachea. Since the percentage fade was not affected by indomethacin or hydroquinone and N°'-nitro-L-arginine, we suppose that the basal release of some inhibitory mediators (not prostanoids or nitric oxide) might be responsible for the fading of the phenylephrine-induced contractile response (table 3). In conclusion, the results of the present study provide evidence for a role of prostanoid and non-prostanoid epithelium-derived inhibitory factors in the relaxant responses to serotonin in the co-axial system. In addition, because of epithelial damage in the tracheas from previous sensitized guinea-pigs challenged with ovalbumin, there might be a loss of serotonin receptor binding sites on the epithelium or serotonin could bind

104

to its receptor sites, but the epithelium might have a decreased ability to synthesize or release some mediators.

Acknowledgements The authors thank Ciba-Geigy Corporation (New Jersey, USA and Istanbul, Turkey) for their gifts of CGS 8515 and imipramine hydrochloride.

References Advenier, C., P. Devillier, R. Matran and E. Naline, 1988, Influence of epithelium on the responsiveness of guinea-pig isolated trachea to adenosine, Br. J. Pharmacol. 93, 295. Aizawa, H., N. Miyazaki, N. Shigematsu and M. Tomooka, 1988, A possible role of airway epithelium in modulating hyperresponsiveness, Br. J. Pharmacol. 93, 139. Al Timimi, K.S., J.R. Bedwani and A.W.B. Stanton, 1978, Effects of prostaglandin E 2 and a prostaglandin endoperoxide analogue on neuroeffector transmission in the rat anococcygeus muscle, Br. J. Pharmacol. 63, 167. Barnes, P.J., F.M. Cuss and J.B. Palmer, 1985, The effect of airway epithelium on smooth muscle contractility in bovine trachea, Br. J. Pharmacol. 86, 685. Barnett, K., D.B. Jacoby, J.A. Nadel and S.C. Lazarus, 1988, The effects of epithelial cell supernatant on contraction of isolated canine tracheal smooth muscle, Am. Rev. Respir. Dis. 138, 780. Beasley, R., W.R. Roche, J.A. Roberts and S.T. Holgate, 1989, Cellular events in the bronchi in mild asthma and after bronchial provocation, Am. Rev. Respir. Dis. 139, 806. Butler, G.B., K.B. Adler, J.N. Evans, D.W. Morgan and J.L. Szarek, 1987, Modulation of rabbit airway smooth muscle responsiveness by respiratory epithelium: involvement of an inhibitory metabolite of arachidonic acid, Am. Rev. Respir. Dis. 135, 1099. (~akici, I., I. Kanzik, N. Abacio~lu and H. Zengil, 1990, Effect of ovalbumin challenge on the responses to acetylcholine in a co-axial bioassay system, Int. Pharm. J. (Congress Suppl.) CS176 (Abstr.). Devillier, P., C. Advenier, G. Drapeau, J. Marsal and D. Regoli, 1988, Comparison of the effects of epithelium removal and of an enkephalinase inhibitor on the neurokinin-induced contractions of guinea-pig isolated tracheas, Br. J. Pharmacol. 94, 675. Farmer, S.G., J.C. Fedan, D.W.P. Hay and D. Raeburn, 1986, The effects of epithelium removal on the sensitivity of guinea-pig isolated trachealis to bronchodilator drugs, Br. J. Pharmacol. 89, 407. Fernandes, L.B., J.W. Paterson and R.G. Goldie, 1989, Co-axial bioassay of a smooth muscle relaxant factor released from guinea-pig tracheal epithelium, Br. J. Pharmacol. 96, 117. Fernandes, L.B., L.M.H. Preuss, J.W. Paterson and R.G. Goldie, 1990, Epithelium-derived inhibitory factor in human bronchus, Eur. J. Pharmacol. 187, 331. Fine, J.M., T. Gordon and D. Sheppard, 1989, Epithelium removal alters responsiveness of guineapig trachea to substance P, J. Appl. Physiol. 66, 232. Flavahan, N.A., L.L. Aarhus, T.J. Rimele and P.M. Vanhoutte, 1985, Respiratory epithelium inhibits bronchial smooth muscle tone, J. Appl. Physiol. 66, 834 Frossard, N., K.R. Rhoden and P.J. Barnes, 1989, Influence of epithelium on guinea-pig airway responses to tachykinins: role of

endopeptidase and cyclooxygenase, J. Pharmacol. Exp. Ther. 248, 292. Gillespie, J.S., 1972, The rate anococcygeus muscles and its response to nerve stimulation and to some drugs, Br. J. Pharmacol. 45, 404. Goldie, R.G., J.M. Papadimitriou, J.W. Paterson, P.J. Rigby, H.M. Self and D. Spina, 1986, Influence of the epithelium on responsiveness of guinea-pig isolated trachea to contractile and relaxant agonists, Br. J. Pharmacol. 87, 5. Gunn, L.K. and P.J. Piper, 1991, Potential sources of artefact in the co-axial bioassay, Eur. J. Pharmacol. 203, 405. Giic, M.O., M. Ilhan and S.O. Kayaalp, 1988, The rat anococcygeus muscle is a convenient bioassay organ for the airway epitheliumderived relaxing factor, Eur. J. Pharmacol. 148, 405. Hand, J.M., J.A. Will and C.K. Buckner, 1982, Pharmacological alteration of antigen induced contraction of pulmonary arteries isolated from the actively sensitized guinea-pig, J. Pharmacol. Exp. Ther. 220, 526. Hay, D.W.P., R.M. Muccitelli, D.L. Horstemeyer, K.A. Wilson and D. Raeburn, 1987, Demonstration of the release of an epithelium-derived inhibitory factor from a novel preparation of guineapig trachea, Br. J. Pharmacol. 136, 247. Hogg, J.C., 1981, Bronchial mucosal permeability and its relationship to airways hyperreactivity, J. Allergy Clin. Immunol. 67, 421. Hogg, J.C. and P.A. Eggleston, 1984, Is asthma an epithelial disease, Am. Rev. Respir. Dis. 129, 207. Holroyde, M.C., 1986, The influence of epithelium on the responsiveness of guinea-pig trachea, Br. J. Pharmacol. 87, 501. Hughes, J., 1971, Evaluation of neuronal and extraneuronal uptake mechanism during adrenergic nerve stimulation, Br. J. Pharmacol. 42, 660P. Ilhan, M. and I. ~ahin, 1986, Tracheal epithelium release a vascular smooth muscle relaxant factor: demonstration by bioassay, Eur. J. Pharmacol. 131,293. Jacoby, D.B., I.F. Ueki, J.H. Widdicombe, D.A. Loegering, G.L Gleich and J.A. Nadel, 1988, Effect of human eosinophil major basic protein on ion transport in dog epithelium, Am. Rev. Respir. Dis. 137, 13. Jeffery, P.K., A.J. Wardlaw, F.C. Nelson, J.V. Collins and A.B. Kay, 1989, Bronchial biopsies in asthma: an ultrastructural, quantitative study and correlation with hyperreactivity, Am. Rev. Respir. Dis. 140, 1745. Laitinen, A., M. Heino, A. Laitinen, T. Kava and T. Haahtela, 1985, Damage of the airway epithelium and bronchial reactivity in patients with asthma, Am. Rev. Respir. Dis. 131, 599. Langer, S.Z., 1970, The metabolism of (3H)noradrenaline released by electrical stimulation from the isolated nictitating membrane of the cat and from the vas deferens of the rat, J. Physiol. London 208, 515. Naline, E., P. Devillier, G. Drapeau, L. Toty, H. Bakdach, D. Regoli and C. Advenier, 1989, Characterization of neurokinin effects and receptor selectivity in human isolated bronchi, Am. Rev. Respir. Dis. 140, 679. Nash, C.W., J.S. Gillespie and E.N. Robertson, 1974, Noradrenaline uptake properties of the anococcygeus muscle of the rat, Can. J. Physiol. Pharmacol. 52, 430. Orehek, J., J.S. Douglas and A. Bouhuys, 1975, Contractile responses of the guinea-pig trachea in vitro: modification by prostaglandin synthesis inhibiting drugs, J. Pharmacol. Exp. Ther. 194, 554. Persson, C.G.A. and J.A. Karlsson, 1987, In vitro response to bronchodilator drugs, in: Drug Therapy for Asthma-Lung Biology in Health and Disease, eds. J. Jenne, T. Murphy and C. Lenient (Marcel Dekker, New York) p. 129. Reaburn, D., D.W.P. Hay, S.G. Farmer and J.S. Fedan, 1986, Epithelium removal increases the reactivity of human isolated tracheal muscle to methacholine and reduces the effect of verapamil, Eur. J. Pharmacol. 123, 451.

105 Selig, W.M., A.M. Bendele and J.H. Fleisch, 1988, Antigen-induced edema formation, bronchoconstrietion, and pulmonary vasospasm in the isolated perfused guinea-pig lung, Am. Rev. Respir. Dis. 138, 552. Sleigh, M.A., J.R. Blake and N. Liron, 1988, The propulsion of mucus by cilia, Am. Rev. Respir. Dis. 137, 726.

Undem, B.J., D.J. Raible, N.F. Adkinson and G.K. Adams, 1988, Effect of removal of the epithelium on antigen-induced smooth muscle contraction and mediator release from guinea-pig isolated trachea, J. Pharmacol. Exp. Ther. 244, 659.