Inhibition of canine tracheal smooth muscle by mediators from cultured bronchial epithelial
cells
X.-Y. YU, W. HUBBARD, AND E. W. SPANNHAKE Department of Environmental Health Sciences, School of Hygiene and Public Health and Division of Clinical Immunology, Asthma and Allergy Center, The Johns Hopkins Medical Institutions, Baltimore, Maryland 21205 Yu, X.-Y., W. Hubbard, and E. W. Spannhake. Inhibition of canine tracheal smooth muscle by mediators from cultured bronchial epithelial cells. Am. J. Physiol. 262 (Lung Cell. Mol. Physiol. 6): L229-L234, 1992.-To determine the effect of mediators released from cultured canine bronchial epithelial cells on contraction of canine tracheal smooth muscle, we treated smooth muscle strips with piperazine-N,N’bis( 2-ethanesulfonic acid) buffer “conditioned” by 5 h incubation with cultures of 4- to &day-old cultured epithelial cells. Pretreatment of tracheal smooth muscle with conditioned buffer for 5 min resulted in a significant shift to the right of the contractile dose-response curve to histamine in the range of low8 to 5 x 10m4 M. In addition, conditioned buffer induced a dose-related relaxation of the muscle precontracted by histamine (5 PM). Relaxant activity was also evident against tissues precontracted by 5-hydroxytryptamine or methacholine. Lipid extraction of conditioned buffer reduced its activity to that of the fresh buffer control. Prior treatment of the cells in culture with the cyclooxygenase inhibitor, sodium meclofenamate (4 PM), markedly reduced the relaxant effect of the conditioned buffer, whereas prior treatment with MK-886, an inhibitor of 5-lipoxygenase, did not alter relaxant activity. Analysis of prostanoids released into the buffer by epithelial cells indicated the presence of prostaglandin EP (PGEJ and 6ketoprostaglandin F1,, the former in concentrations sufficient to account for the effect of conditioned buffer on precontracted tracheal muscle. Prostaglandin I2 appeared to have a synergistic effect on PGEz-induced relaxation. We conclude that canine bronchial epithelial cells exhibit baseline release of a relaxant lipid factor(s) that can both inhibit and reverse the contraction of tracheal smooth muscle by histamine. The data suggest that PGEz is the mediator principally responsible for this relaxant activity released spontaneously by bronchial epithelial cells in culture. epithelium-derived relaxing factor; prostaglandin E,; prostaglandin I,; histamine; 5-hydroxytryptamine; methacholine; cyclooxygenase inhibitor; 5-lipoxygenase inhibitor; pharmacology; airways; bronchodilation; relaxation; cell culture; tissue bath IS considerable evidence that the dysfunction of epithelial cells (14) and an increased presence of inflammatory mediators (4) are both involved in the pathophysiology of asthma. The absence of a structurally intact airway epithelium has been observed to augment the contractile responses to immunologic, chemical, and electrical stimulation in tissues from humans and experimental animals (1,5,10, 12, 20,27,28). Clear indication exists that, in some cases,the loss of an epithelial diffusion barrier (12, 27) plays a part in the observed augmentation. However, an increasing body of evidence indicates that the release of a relaxant factor or factors from epithelial cells constitutes a key mechanism in the modulation of airway smooth muscle by the epithelium (1% .
THERE
1040-0605/92
$2.00
Copyright
0 1992
Demonstration of the synthesis and release of a wide range of metabolites of arachidonic acid by epithelial cells (8, 13, 16, 29, 30) has focused attention on the potential contribution of specific eicosanoids to epithelium-derived relaxing factor activity. Whereas some reports indicate the presence of relaxant activity that is unaffected by inhibition of the cyclooxygenase pathway (5, lo), others indicate that prostanoids appear to be responsible for the activity to varying degrees (6, 7, 11, 26)
The present study was designed to directly determine the activity of relaxant mediators released under baseline conditions from canine epithelial cells grown in culture, using canine tracheal smooth muscle devoid of epithelium as a test system. Histamine, an important mast cell-derived mediator of antigen-associated bronchoconstriction, was used as the contractile agent. With this system, we also examined the nature of the factor(s) involved in modulating smooth muscle contractile activity. A preliminary report of these studies has been published (32). METHODS Culture of canine bronchiaL epithelial ceZZs.Canine bronchial epithelial cells were isolated and cultured by a modification of the method of Coleman et al. (9). Bronchial tissue was incubated at 4°C for 20-24 h in Hanks’ balanced salt solution (HBSS) without Ca2+ or Mg2’, but containing pronase (1 mg/ ml), followed by longitudinal dissection. Epithelial cells were dislodged from the bronchial surface by gentle washing with jets of HBSS delivered from a 3-ml syringe. Cells were centrifuged at 350 g for 10 min at 4OC, washed in 50 ml of Dulbecco’s modified Eagle’s medium (DMEM), centrifuged again, and resuspended in DMEM containing 5% fetal calf serum, penicillin-streptomycin (10 pg/ml), gentamicin (5 mg/lOO ml), Fungizone (1.25 pg/ml), and nonessential amino acids (0.1 ml/500 ml). Suspensions containing 1.5-3 x lo6 cells/ml were prepared, and 2 ml were plated in each well of six-well cluster tissue culture plates precoated with 2 ml of human placental collagen/ well (100 pg/ml). Cell cultures were incubated at 37°C in a 5% C02-95% O2 mixture. Beginning 24 h after plating, the medium was changed every day until the cells became confluent. Light microscopic examination of the confluent cells revealed the uniform cobblestone appearance typical of epithelial cells in culture. Vertical sections of representative cultures examined by electron microscopy showed them to consist of cuboidal epithelial cells, the media-facing surface of which supported numerous microvilli and/or cilia. The lateral membranes of these cells exhibited complex interdigitations and junctional complexes with those adjacent. Preparation of epithelial ceZZ-derived supernatants. On day 4 or 5, after washing three times with piperazine-N,N’-bis(2ethanesulfonic acid) (PIPES) buffer (25 mM, pH 7.2) containing 2 mM Ca2’, 1 mM Mg2’, and 1 mg/ml glucose, cell cultures the American
Physiological
Society
L229
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
L230
BRONCHIAL
EPITHELIUM
INHIBITS
were incubated with this buffer alone in a volume of 1 ml/well for 5 h (conditioned buffer) or in the presence of the cyclooxygenase inhibitor, sodium meclofenamate (4 PM) or the 5 lipoxygenase inhibitor, MK-886 (1 PM) (22). After collection, half of the conditioned buffer was passed, by syringe, through a Sep-Pak Cls cartridge (Waters) to remove nonpolar (lipid) components (31). (Cartridges were activated with methanol and washed with water before application of the samples.) Fresh PIPES buffer served as control. Concentrations of PGE2 and 6-ketoprostaglandin FI, (6-keto-PGF1,) (the stable metabolite of PGI& were assayed in duplicate in the above buffer incubates by enzyme immunoassay (EIA) (Advanced Magnetics, Cambridge, MA). Preliminary experiments indicated that neither meclofenamate nor MK-886 in the concentrations employed had any influence on this assay. Screening of prostanoids in untreated conditioned buffer was performed by gas chromatography/mass spectrometry using deuterated standards as previously described (15). Preparation of isolated tracheal smooth muscle. Tracheas of mongrel dogswere excisedimmediately after death under pentobarbital anesthesiaand openedlongitudinally along the ventral surface.Smooth musclestrips (0.3 cm x 1.2 cm) bordered by 0.5 cm of cartilage were prepared from the muscularis portion by transversesectioningof the tissuefollowing removal of epithelium, fascia, and fat. Eight to 12 strips were prepared from each trachea. The strips were mounted in musclebaths containing 10 ml of Krebs-Henseleit solution maintained at 37°C and gassedcontinuously with 95% O2and 5% CO,. Isometric tension was measuredwith Grass force displacement transducers(modelFT 03) and recorded on a GrassPolygraph (model7D or 79D). The strips were treated with sodiummeclofenamate (4.5 M) throughout the experiment to inhibit the endogenousproduction of cyclooxygenaseproducts. The tissues were allowed to equilibrate for 2 h at a resting tension of 3 g (previously determined to be optimum for this preparation) and were washed at 15-min intervals with Krebs-Henseleit solution. Experimental protocol. The study consisted of four experimental protocols. In protocol 1, after tracheal smooth muscle was pretreated with conditioned buffer (supernatant fraction of PIPES buffer incubated with cultured epithelial cells for 5 h) for 5 min, the contractile dose-responserelationship to histaminewasdeterminedand normalized to maximal contraction in the presence of carbachol (100 PM). In protocol 2, tracheal smoothmusclewasprecontracted by histamine (5 PM), and the responseto conditioned buffer was determined when addedto the bath as a 2ml bolus or in increments of 0.1, 0.2, 0.7, 0.3, and 0.7 ml over a 35-min period to generate a cumulative concentration-responsecurve. In the latter cases,relaxation was normalized to maximum relaxation with isoprotereno1(100 PM). In protocol 1, the effect of pretreatment with the epithelium-derived supernatant (conditioned buffer) was compared with that of pretreatment with fresh buffer alone on the dose-responseto histamine. In protocol 2, the effects of fresh buffer, conditioned buffer, conditioned buffer from meclofenamate-treated cells, conditioned buffer from MK-886-treated cells, and conditioned buffer following removal of lipid were comparedon musclestrips precontracted by histamine. In the experiments in which the concentration dependenceof relaxation to the conditioned buffer was investigated, the effect of time alonewas alsodetermined. In protocol 3, to determine the extent to which the cell culture-derived relaxant activity was agonist specific, the ability of conditioned buffer to reverse contraction inducedby 5-hydroxytryptamine (0.15 PM), methacholine (0.13 PM), and histamine (5 FM) were compared.In protocol 4 the relative relaxant activities of authentic PGE2and PGIZ , alone and in combination, were determined on tracheal smooth muscleprecontracted with histamine.
AIRWAY
SMOOTH
MUSCLE
Reagents. Sterile HBSS containing 0.35 g/l sodiumbicarbonate but without magnesium,calcium, or phenol red was prepared by Flow Laboratories, Irvine, Scotland. Krebs-Henseleit was prepared to contain (in mM) 117.6 NaCl, 5.4 KCl, 1.0 NaH2P04, 1.2 MgSO,, 2.5 CaC12,25.0 NaHC03, and 11.1 glucose.All smoothmuscleagonistsand metabolic pathway inhibitors were prepared fresh on the day of use. Carbachol, histamine, 5-hydroxytryptamine, and methacholine (Sigma Chemical, St. Louis, MO) were dissolvedin Krebs-Henseleit solution and kept on ice. Isoproterenol (Sigma)was dissolvedin KrebsHenseleit solution with the addition of ascorbic acid (10B5M) for preservation and kept on ice. MK-886 (generouslydonated by Merck Frosst, Quebec, Canada) was dissolved in PIPES buffer and sodiummeclofenamate(generouslydonatedby Warner Lambert, Ann Arbor, MI) wasdissolvedin Krebs-Henseleit solution or PIPES buffer and eachbrought to final concentration just prior to use.Prostaglandin E2 and the sodiumsalt of PG12(Upjohn) were dissolvedin ethanol at a concentration of I mg/ml and diluted cold in Krebs-Henseleitsolution just prior to use. Data analysis. Values for changesin smooth muscletension were expressedas a percentage of the maximal contractile or relaxant effects elicited by carbachol or isoproterenol, respectively. The resultswere expressedasthe meant SE. Statistical comparisonof meanswasmadeusing Student’s t test for group data. A P value ~0.05 was consideredsignificant. Animal care and use. Animals were handled in accordance with the standardsestablishedby the Animal Welfare Act and set forth in the Department of Health, Education, and Welfare (National Institutes of Health) guidelines.Experimental protocols were reviewed and approved by this institution’s Committee on Animal Care and Use. RESULTS
Treatment of tracheal smooth muscle with conditioned buffer for 5 min resulted in a significant rightward sh.ift of the contractile dose-response curve to histamine in the range of 10 t 8 to 5 X low4 M. Responsesto histamine at the upper end of the curve were unaffected (Fig. 1). To determine whether the relaxant activity of conditioned buffer exhibited dose dependence, tracheal smooth muscle precontracted with 5 PM histamine, a concentration near the mean contractile effective con-
z
n
100
. . . . . . . a......
Fresh
Buffer (n=8)
-
Cond.
Buffer (n=6)
1
* I
g
I
1()-7 1 O-8 HISTAMINE
I
l(+
I
I(+
I
1O-4
1 1O-3
M
CONCENTRATION
Fig. 1. Summary data from 8 experiments showing inhibitory effect of 5 min pretreatment of tracheal smooth muscle with conditioned buffer on dose-response curve to histamine over range of 1 X 10-’ to 5 X 10V4 M. Comparison is made to the dose-response curve generated following pretreatment with fresh buffer. Data normalized to maximal contraction induced by 100 PM carbachol and expressed as means t SE; n represents no. of tracheal smooth muscle strips. *P < 0.05, conditioned vs. fresh buffer.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
BRONCHIAL
EPITHELIUM
INHIBITS
centration (E&J, was challenged with volumes of conditioned buffer accumulating to a total of 2 ml. Figure 2 depicts a tracing from a typical experiment following this protocol comparing the effects of fresh and conditioned buffer. In Figure 3, data from 7-9 experiments are summarized. Compared with fresh buffer, conditioned buffer caused significant relaxation of tracheal smooth muscle at all but the 0.1 ml addition, reaching a maximum of more than 90% at 2.0 ml. Relaxation produced by fresh buffer was not significantly different from that produced by time alone. To determine the relative effectiveness of conditioned buffer against different contractile stimuli, tracheal smooth muscle was precontracted with &hydroxytryptamine (5HT, 0.15 PM), methacholine (0.13 PM), or histamine (5 PM) and challenged with Z-ml aliquots of conditioned buffer. These concentrations were deterHist
(5uM)
N
10 9
(ml)
0.1t
0.2t
t Oa7
1 0
FRESH
t 03.
‘)
t 0.7
BUFFER
AIRWAY
SMOOTH
L231
MUSCLE
mined, in preliminary experiments, to represent the EC50 for each agonist in this preparation. Figure 4 indicates that reversal of 5-HT-induced contraction was similar to, although statistically less than, that of histamineinduced contraction. In contrast, methacholine-induced contraction was much more resistant to the relaxant activity present in conditioned buffer. To investigate the nature of the relaxant factor(s) present in conditioned buffer, the effects of various buffer preparations on tracheal smooth muscle precontracted by histamine (5 ,uM) were compared (Fig. 5). The addition of 2 ml of conditioned buffer reversed the contraction by 83 t 4% compared with the effect of fresh buffer (14 t 3%). Removal of lipids from conditioned buffer by Sep-Pak Cl8 extraction markedly reduced the relaxant effect to a level (22 t 4%) not statistically different from fresh buffer. Fresh buffer, itself, caused an 18% reversal of the contracted state. The cause of this effect is undetermined, but could not be explained on the basis of dilution of the agonist or a shift in bath pH or temperature. To determine the involvement of eicosanoids in the lipid-associated relaxation of histamine-contracted tis-
Is0
q q
20 10
t
(ml)
E -
1
5-HT (n=5) Meth (n=6) Histamine (n=i 1)
0.1
r
9 0 1
CONDITIONED .
*
BUFFER
*
’
2
=
*
Om3
10
6
07.
I Is0
min
Fig. 2. Tracings from a typical experiment showing cumulative effect of various volumes of fresh or conditioned buffer, up to a total volume of 2 ml, on tracheal smooth muscle precontracted by 5 PM histamine. Iso, isoproterenol, 100 PM, added to induce maximal relaxation of tissues. . .. .. .. 0 . .. .. .
Untreated
(n=9)
TV
Fresh
Buffer
(n=7)
-
Cond.
Buffer
(n=8)
Fig. 4. Relative effectiveness of conditioned buffer in reversing contraction of tracheal smooth muscle induced by 0.15 FM 5-hydroxytryptamine (5-HT), 0.13 PM methacholine (Meth) or 5 PM histamine. Data normalized and expressed as in Fig. 3; n represents no. of strips. *p c 0.05, compared with histamine.
0 0
2 0 -20 Ia x
20
2 -40 ii! i
z
0 i= 0 3
-60
Fresh
Buffer (n=9)
Cond.
Buffer
- lipid (n=8)
Cond.
Buffer
(n=8)
Meclofenamate
1
(n=8)
MK 886 (n=6)
4 -20
x
W a
s -80
-40 i z -60 2 = -80
1.0
0.1
TOTAL
AC%bMULATED
1.3
DOSE
2.0
(ml)
Fig. 3. Summary data from a series of experiments, as depicted in Fig. 2, showing cumulative relaxant effect of various volumes of fresh or conditioned buffer on smooth muscle precontracted by 5 PM histamine. Effect of time alone (untreated) on contractile state was determined by measuring level of contraction in untreated tissues at times corresponding to those at which measurements were made in buffer-treated tissues. Data normalized to maximal relaxation to 100 PM isoproterenol and expressed as means & SE. Abscissa, accumulated volume added on a log scale. n represents number of strips. *P < 0.05, conditioned vs. fresh buffer.
.O
Fig. 5. Effect of single 2.0-ml additions of fresh buffer, conditioned buffer (Cond Buffer), conditioned buffer with lipids removed by SepPak extraction (Cond Buffer-lipid), conditioned buffer from meclofenamate-preincubated cells to inhibit prostaglandin production (Meclofenamate), and conditioned buffer from MK-886-preincubated cells to inhibit leukotriene production (MK-886) on tracheal smooth muscle precontracted with 5 PM histamine. Data normalized and expressed as in Fig. 3; n represents no. of strips. *‘P < 0.05, compared with fresh buffer.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
L232
BRONCHIAL
EPITHELIUM
INHIBITS
sues, inhibitors of prostaglandin and leukotriene synthesis were added to canine bronchial epithelial cell cultures during conditioning of the buffer. The relaxant activity of buffer harvested from canine bronchial epithelial cells in the presence of the cyclooxygenase inhibitor, sodium meclofenamate, was significantly inhibited (83 t 4 to 31 t 3%), almost to the level of the lipid-free conditioned buffer. In contrast, treatment with MK-886, a selective inhibitor of &lipoxygenase, had no effect on epithelial cell-derived relaxant activity (Fig. 5). Two random samples of buffer conditioned by epithelial cells in culture were subjected to analysis in duplicate by GC/MS to determine the spectrum of prostanoids present. Results indicated the presence of PGF2, (243 t 40 pg/ml); PGE, (4,787 t 411 pg/ml); 6-keto-PGF1, (1,394 t 496 pg/ml), whereas TXB2, PGDB, and its principal metabolite, 9&-PGFz, were undetectable. Because PGE2 and PG12 are the principal prostanoid metabolites reported to induce relaxation of canine tracheal smooth muscle (19), PGE2 and 6-keto-PGF1, were assayed in the various buffer preparations described above. The relative amounts of PGE2 and 6-keto-PGF1, followed the same pattern and were distributed in the following order: MK-886-conditioned buffer > conditioned buffer > meclofenamate-conditioned buffer > lipid-extracted buffer (Fig. 6). In preparations from cells with uninhibited prostaglandin production, PGEz concentrations were approximately twice those of 6-ketoPGF1, (Fig. 6). To confirm that PGEz and PGIB behaved as relaxant factors in the concentrations present in conditioned buffer, both prostanoids were added, singly and in combination, to precontracted tracheal smooth muscle in concentrations equivalent to one-fourth to one-third of those in 2 ml of conditioned buffer (Fig. 7). PGEg reversed histamine-induced contraction of tracheal smooth muscle by 49 t 4%, whereas PGIz was without significant effect. In the presence of these concentrations of PGIZ, relaxation by PGEz increased to 65 t 5%, a level statistically greater than that of PGEz alone (Fig. 7). DISCUSSION
The results of the present study indicate that canine bronchial epithelial cells in culture exhibit baseline release of a factor or factors that downregulate contractile (n=6)
g
6ooo- 111 fxE2 5ooo _ q 6-keto-PGF1,
-s
e
.
4000
-
g 3000 z zv) 2000
-
(n-12)
-
0
g
1000
-
(““‘)(n=lO)
Cond. Buffer
Meclofenamate
C.B.-lipid
MK 886
Fig. 6. Concentrations of PGEg and 6-keto-PGF,, in various buffer preparations described in Fig. 5, as determined by enzyme immunoassay. Data are means t SE, n represents no. of experiments.
AIRWAY
SMOOTH
q 20
-80
1
!
MUSCLE PG12 (n=13) IFEz
q
(n=ll)
PGIz + PGEz
(n=13)
*
d
Fig. 7. Effect of PGEz (1.7-3 ng), PGI, (1 ng), and PGEz (1.7-3 ng) in presence of PG12 (1 ng) on tracheal smooth muscle precontracted by histamine (5 PM). Data normalized and expressed as in Fig. 3; n represents no. of strips. *Fp c 0.05, PGEz vs. PGE, + PG12.
responses of tracheal smooth muscle. We have demonstrated the presence of this activity by the inhibition of histamine-induced contraction, as well by the dose-dependent reversal of contraction in histamine-treated tissues. Barnett et al. (6) reported similar inhibitory activity in supernatants from cultured canine tracheal epithelial cells that they ascribed to PGE2. In their studies, however, inhibition of the effects of electrical field stimulation, presumably at a prejunctional site, was seen only with supernatants from bradykinin-stimulated cells and not with those from unstimulated cells. It is interesting to note that concentrations of PGEz in supernatants from unstimulated cells reported by Barnett and colleagues (6) are similar to\ those found in the present . study. If this prostanoid is the factor principally responsible for the inhibitory effect in both studies, the lower concentrations required to elicit an effect in the present study may be related to a greater potency at postjunctional, as opposed to prejunctional, sites. Although the mechanism of the relaxant activity on smooth muscle was not directly investigated in the present study, experiments carried out in tissues contracted by 5-hydroxytryptamine and methacholine indicate that the effect is not histamine specific. Thus the effect appears to be exerted, albeit to a varying extent, at a point in the contractile process common to these three smooth muscle contractile agonists. It is known that histamine, 5-hydroxytryptamine, and muscarinic agonists, such as methacholine, stimulate the hydrolysis of polyphosphatidylinositols, the production of inositol phosphates and diacylglycerols, and increase intracellular concentration of free Ca2+, leading to smooth muscle contraction (21). ,&Agonists and PGE2 and PGI2 are known to activate adenylate cyclase, increase CAMP levels, and inhibit contraction (18, 19, 21). The involvement of this inhibitory pathway in the activity of cellderived supernatants in the present study is possible and is consistent with the observation that supernatants were less effective in reversing contraction induced by muscarinic stimulation, which has been shown to inhibit adenylate cyclase activity (21). Lipid extraction of the conditioned buffer resulted in the elimination of the inhibitory activity, consistent with the notion that relaxation was related to the presence of eicosanoid mediators. Support for the principal role of prostanoids in reversing the histamin.e-induced contrac-
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
BRONCHIAL
EPITHELIUM
INHIBITS
tion was provided by experiments in which relaxant activity was reduced to that of lipid-extracted buffer by pretreatment of the epithelial cells with sodium meclofenamate. The lack of an effect of pretreatment of the cells with the lipoxygenase inhibitor, MK-886, suggests that products of this pathway did not contribute to the lipid-associated inhibition. Analysis of buffer preparations indicated the presence of PGEz and 6-keto-PGF1,, the PG12 metabolite, in conditioned buffer. Cultured epithelial cells from a varietv of species, including the dog; are well known to synthesize PGEz (6, 8, 30). In intact airways, the principal source of PGI, appears to be the smooth muscle (25); however, 6-keto-PGF1, is also produced by airway epithelial cells (30). Although Shore et al. (23) have demonstrated the ability of PGIz to relax tracheal smooth muscle contracted by histamine, its activity is less than that of PGEZ in this system (2). Consistent with these observations, results of the present study indicate that exogenous PGEZ, in concentrations similar to those present in conditioned buffer, could account for the observed relaxation. Although PGIz concentrations were estimated to be only 60% less than those of PGEz, exogenous PGIz induced only minimal relaxation of histamine-contracted tracheal smooth muscle. The apparent synergism observed when both PGEz and PG12 were added to the tracheal preparations may have been due to the combined abilities of these mediators to reduce cakium influx (3) and stimulate adenylate cyclase (17, 24) and suggests that these two prostanoids may normally act in concert to mediate epithelium-derived modulation of tracheal contraction. Although the role of prostaglandins in the activity of epithelium-derived relaxant factor in vivo remains controversial, the present study demonstrates that canine bronchial epithelial cells in culture exhibit baseline release of prostanoids in concentrations sufficient to modulate the contraction of airwav smooth muscle. Based on our current knowledge of the array of cyclooxygenase products released from these cells and their effects on tracheal smooth muscle, the results of the present study are consistent with PGEZ being the principal factor responsible for this activity. The authors thank William Dunlap III for excellent technical support, Brian Schofield for expertise in preparing and evaluating microscopic samples, and Dr. Brad Undem for comments and suggestions. This study was supported by National Institutes of Health Grants HL-30195, ES-03505, and AI-20136. Address for reprint requets: E. W. Spannhake, Div. of Physiology, Dept. of Environmental Health Sciences, The Johns Hopkins School of Hygiene and Public Health, 615 N. Wolfe St., Baltimore, MD 21205.
AIRWAY
4. 5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21. REFERENCES 1. Aizawa, H., N. Miyazaki, N. Shigematsu, and M. Tomooka. A possible role of airway epithelium in modulating hyperresponsiveness. Br. J. Pharmacol. 93: 139-145, 1988. 2. Anderson, W. H., J. J. Krazanowski, J. B. Polson, and A. Szentivanyi. Characteristics of histamine tachyphylaxis in canine tracheal smooth muscle. Naunyn-Schmeidebergs - Arch. Pharmacol. 308: 117-125,1979. 3, Anderson, W. H., J. J. Krazanowski, J. B. Polson, and A. Szentivanyi. The effect of prostaglandin E2 on histamine-stimulated calcium mobilization as a possible explanation for histamine
22.
23.
24.
SMOOTH
MUSCLE
L233
tachyphylaxis in canine tracheal smooth muscle. Nuunyn-Schmeidebergs Arch. Pharmacol. 322: 72-77, 1983. Barnes, P. J. Inflammatory mediator receptors and asthma. Am. Rev. Respir. Dis. 135: S26-S31, 1987. Barnes, P. J., F. M. Cuss, and J. B. Palmer. The effect of airway epithelium on smooth muscle contractility in bovine trachea. Br. J. Pharmacol. 86: 685-691, 1986. Barnett, K., D. B. Jacoby, J. A. Nadel, and S. C. Lazarus. The effects of epithelial cell supernatant on contractions of isolated canine tracheal smooth muscle. Am. Rev. Respir. Dis. 138: 780783,1988. Brauntein G., C. Labat, S. Brunelleschi, J. Benveniste, J. Marsac, and C. Brink. Evidence that the histamine sensitivity and responsiveness of guinea-pig isolated trachea are modulated by epithelial prostaglandin Ez production. Br. J. PharmacoZ. 95: 300-308.1988. Churchill, L., F. H. Chilton, J. H. Resau, R. Bascom, W. C. Hubbard, and D. Proud. Cyclooxygenase metabolism of endogenous arachidonic acid by cultured human tracheal epithelial cells. Am. Rev. Respir. Dis. 140: 449-459, 1989. Coleman, D. L., I. K. Tuet, and J. H. Widdicombe. Electrical properties of dog tracheal epithelial cells grown in monolayer culture. Am. J. Physiol. 246 (Cell Physiol. 15): C355-C359, 1984. Flavahan, N. A., L. L. Aarhus, T. J. Rimele, and P. M. Vanhoutte. The respiratory epithelium inhibits bronchial smooth muscle tone. J. Appl. Physiol. 58: 834-838, 1985. Hay, D. W. P., S. G. Farmer, D. Raeburn, V. A. Robinson, W. W. Fleming, and J. S. Fedan. Airway epithelium modulates the reactivity of guinea-pig respiratory smooth muscle. Eur. J. Pharmacol. 129: 11-18, 1986. Holroyde, M. C. The influence of epithelium on the responsiveness of guinea-pig isolated trachea. Br. J. Pharmacol. 87: 501-507, 1986. Holtzman, M. J., H. Aizawa, J. A. Nadel, and E. J. Goetzl. Selective generation of leukotriene B4 by tracheal epithelial cells from dogs. Biochem. Biophys. Res. Commun. 114: 1071-1076,1983. Laitinen, L. A., M. Heino, A. Laitinen, T. Kava, and T. Haahtels. Damage of the airway epithelium and bronchial reactivity in patients with asthma. Am. Rev. Respir. Dis. 131: 599-606, 1985. Liu, M. C., E. R. Bleecker, D. Proud, T. L. McLemore, and W. Hubbard. Profiling of bisenoic prostaglandins and thromboxane Bz in bronchoalveolar fluid from the lower respiratory tract of human subjects by combined capillary gas chromatography-mass spectrometry. Prostaglandins 35: 67-79, 1988. Leikauf, G. D., I. F. Ueki, J. A. Nadel, and J. H. Widdicombe. Bradykinin stimulates chloride secretion and prostaglandin E2 release by canine tracheal epithelium. Am. J. Physiol. 248 (Renal Fluid Electrolyte Physiol. 17): F48-F55, 1985. Littner, M., G. Kazmi, F. Lott, M. Golub, and M. Berger. Cyclooxygenase products and cyclic nucleotide levels with infusion of arachidonic acid, PG12, PGEZ, PGF2,, and 6-keto-PGF,, in an isolated dog lung. Am. Rev. Respir. Dis. 128: 868-874, 1983. Morrison, K. J., Y.-S. Gao, and P. M. Vanhoutte. Epithelial modulation of airway smooth muscle. Am. J. Physiol. 258 (Lung CelZ. MOL. Physiol. 2): L254-L262, 1990. Madison, J. M., C. A. Jones, R. M. Sankary, and J. K. Brown. Differential effects of prostaglandin E2 on contractions of airway smooth muscle. J. Appl. Physiol. 66: 1397-1407, 1989. Munakata, M., W. Mitzner, and H. Menkes. Osmotic stimuli induce epithelial-dependent relaxation in the guinea pig trachea. J. Appl. Physiol. 64: 466-471, 1988. Rasmussen, H., 6. Kelley, and J. S. Douglas. Interactions between Ca2+ and CAMP messenger system in regulation of airway smooth muscle contraction. Am. J. Physiol. 258 (Lung Cell. Mol. Physiol. 2): L279-L288, 1990. Rouzer C. A., A. W. Ford-Hutchinson, H. E. Morton, and J. W. Gillard. MK886, a potent and specific leukotriene biosynthesis inhibitor blocks and reverses the membrane association of 5lipoxygenase in ionophore-challenged leukocytes. J. Biol. Chem. 265: 1436-1442,199O. Shore, S. A., W. S. Powell, and J. G. Martin. Endogenous prostaglandins modulate histamine-induced contraction in canine tracheal smooth muscle. J. Appl. Physiol. 58: 859-868, 1985. Smith, J. J., J. D. McCann, and M. J. Welseh. Bradykinin
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
L234
25.
26.
27.
28. 29.
BRONCHIAL
EPITHELIUM
INHIBITS
stimulates airway epithelial Cl-secretion via two second messenger pathways. Am. J. Physiol. 258 (Lung Cell. Mol. Physiol. 2): 13691377,199o. Stuart-Smith, K., and P. M. Vanhoutte. Arachidonic acid evokes epithelium-dependent relaxations in canine airways. J. Appl. Physiol. 65: 2170-2180, 1988. Tschirhart, E., N. Frossard, C. Bertrand, and Y. Landry. Arachidonic acid metabolites and airway epithelium-dependent relaxant factor. J. PharmacoZ. Exp. Ther. 243: 310-316, 1987. Undem, B. J., D. G. Raible, N. F. J. Adkinson, and G. K. Adams III. Effect of removal of epithelium on antigen-induced smooth muscle contraction and mediator release from guinea pig isolated trachea. J. Pharmacol. Exp. Ther. 244: 659-665, 1988. Vanhoutte, P. M. Airway epithelium and bronchial reactivity. Can. J. Physiol. Pharmacol. 65: 448-450, 1987. Widdicombe, J. G., I. F. Ueki, D. Enery, D. Margolskee, J.
AIRWAY
SMOOTH
MUSCLE
Yerget, and J. A. Nadel. Release of cyclooxygenase products from primary cultures of tracheal epithelia of dog and human. Am. J. Physiol. 257 (Lung Cell. Mol. Physiol. 1): L361-L365, 1989. 30. Xu, G. L., K. Sivarajah, R. Wu, P. Nettesheim, and T. Eling. Biosynthesis of prostaglandins by isolated and cultured airway epithelial cells. Exp. Lung Res. 10: 101-114, 1986. 31. Yoss, E. B., E. W. Spannhake, J. T. Flynn, J. E. Fish, and S. P. Peters. Arachidonic acid metabolism in normal human alveolar macrophages: stimulus specificity for mediator release and phospholipid metabolism, and pharmacologic modulation in vitro and in vivo. Am. J. Respir. CeZZ Mol. Biol. 2: 69-80, 1990. 32. Yu, X. Y., and E. W. Spannhake. The comparison of the effects of histamine on PGEz release from canine tracheal epithelial (CTE) and bronchial epithelial (CBE) cells in culture (Abstract). Am. Reu. Respir. Dis. 141: A911, 1990.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
cells
X.-Y. YU, W. HUBBARD, AND E. W. SPANNHAKE Department of Environmental Health Sciences, School of Hygiene and Public Health and Division of Clinical Immunology, Asthma and Allergy Center, The Johns Hopkins Medical Institutions, Baltimore, Maryland 21205 Yu, X.-Y., W. Hubbard, and E. W. Spannhake. Inhibition of canine tracheal smooth muscle by mediators from cultured bronchial epithelial cells. Am. J. Physiol. 262 (Lung Cell. Mol. Physiol. 6): L229-L234, 1992.-To determine the effect of mediators released from cultured canine bronchial epithelial cells on contraction of canine tracheal smooth muscle, we treated smooth muscle strips with piperazine-N,N’bis( 2-ethanesulfonic acid) buffer “conditioned” by 5 h incubation with cultures of 4- to &day-old cultured epithelial cells. Pretreatment of tracheal smooth muscle with conditioned buffer for 5 min resulted in a significant shift to the right of the contractile dose-response curve to histamine in the range of low8 to 5 x 10m4 M. In addition, conditioned buffer induced a dose-related relaxation of the muscle precontracted by histamine (5 PM). Relaxant activity was also evident against tissues precontracted by 5-hydroxytryptamine or methacholine. Lipid extraction of conditioned buffer reduced its activity to that of the fresh buffer control. Prior treatment of the cells in culture with the cyclooxygenase inhibitor, sodium meclofenamate (4 PM), markedly reduced the relaxant effect of the conditioned buffer, whereas prior treatment with MK-886, an inhibitor of 5-lipoxygenase, did not alter relaxant activity. Analysis of prostanoids released into the buffer by epithelial cells indicated the presence of prostaglandin EP (PGEJ and 6ketoprostaglandin F1,, the former in concentrations sufficient to account for the effect of conditioned buffer on precontracted tracheal muscle. Prostaglandin I2 appeared to have a synergistic effect on PGEz-induced relaxation. We conclude that canine bronchial epithelial cells exhibit baseline release of a relaxant lipid factor(s) that can both inhibit and reverse the contraction of tracheal smooth muscle by histamine. The data suggest that PGEz is the mediator principally responsible for this relaxant activity released spontaneously by bronchial epithelial cells in culture. epithelium-derived relaxing factor; prostaglandin E,; prostaglandin I,; histamine; 5-hydroxytryptamine; methacholine; cyclooxygenase inhibitor; 5-lipoxygenase inhibitor; pharmacology; airways; bronchodilation; relaxation; cell culture; tissue bath IS considerable evidence that the dysfunction of epithelial cells (14) and an increased presence of inflammatory mediators (4) are both involved in the pathophysiology of asthma. The absence of a structurally intact airway epithelium has been observed to augment the contractile responses to immunologic, chemical, and electrical stimulation in tissues from humans and experimental animals (1,5,10, 12, 20,27,28). Clear indication exists that, in some cases,the loss of an epithelial diffusion barrier (12, 27) plays a part in the observed augmentation. However, an increasing body of evidence indicates that the release of a relaxant factor or factors from epithelial cells constitutes a key mechanism in the modulation of airway smooth muscle by the epithelium (1% .
THERE
1040-0605/92
$2.00
Copyright
0 1992
Demonstration of the synthesis and release of a wide range of metabolites of arachidonic acid by epithelial cells (8, 13, 16, 29, 30) has focused attention on the potential contribution of specific eicosanoids to epithelium-derived relaxing factor activity. Whereas some reports indicate the presence of relaxant activity that is unaffected by inhibition of the cyclooxygenase pathway (5, lo), others indicate that prostanoids appear to be responsible for the activity to varying degrees (6, 7, 11, 26)
The present study was designed to directly determine the activity of relaxant mediators released under baseline conditions from canine epithelial cells grown in culture, using canine tracheal smooth muscle devoid of epithelium as a test system. Histamine, an important mast cell-derived mediator of antigen-associated bronchoconstriction, was used as the contractile agent. With this system, we also examined the nature of the factor(s) involved in modulating smooth muscle contractile activity. A preliminary report of these studies has been published (32). METHODS Culture of canine bronchiaL epithelial ceZZs.Canine bronchial epithelial cells were isolated and cultured by a modification of the method of Coleman et al. (9). Bronchial tissue was incubated at 4°C for 20-24 h in Hanks’ balanced salt solution (HBSS) without Ca2+ or Mg2’, but containing pronase (1 mg/ ml), followed by longitudinal dissection. Epithelial cells were dislodged from the bronchial surface by gentle washing with jets of HBSS delivered from a 3-ml syringe. Cells were centrifuged at 350 g for 10 min at 4OC, washed in 50 ml of Dulbecco’s modified Eagle’s medium (DMEM), centrifuged again, and resuspended in DMEM containing 5% fetal calf serum, penicillin-streptomycin (10 pg/ml), gentamicin (5 mg/lOO ml), Fungizone (1.25 pg/ml), and nonessential amino acids (0.1 ml/500 ml). Suspensions containing 1.5-3 x lo6 cells/ml were prepared, and 2 ml were plated in each well of six-well cluster tissue culture plates precoated with 2 ml of human placental collagen/ well (100 pg/ml). Cell cultures were incubated at 37°C in a 5% C02-95% O2 mixture. Beginning 24 h after plating, the medium was changed every day until the cells became confluent. Light microscopic examination of the confluent cells revealed the uniform cobblestone appearance typical of epithelial cells in culture. Vertical sections of representative cultures examined by electron microscopy showed them to consist of cuboidal epithelial cells, the media-facing surface of which supported numerous microvilli and/or cilia. The lateral membranes of these cells exhibited complex interdigitations and junctional complexes with those adjacent. Preparation of epithelial ceZZ-derived supernatants. On day 4 or 5, after washing three times with piperazine-N,N’-bis(2ethanesulfonic acid) (PIPES) buffer (25 mM, pH 7.2) containing 2 mM Ca2’, 1 mM Mg2’, and 1 mg/ml glucose, cell cultures the American
Physiological
Society
L229
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
L230
BRONCHIAL
EPITHELIUM
INHIBITS
were incubated with this buffer alone in a volume of 1 ml/well for 5 h (conditioned buffer) or in the presence of the cyclooxygenase inhibitor, sodium meclofenamate (4 PM) or the 5 lipoxygenase inhibitor, MK-886 (1 PM) (22). After collection, half of the conditioned buffer was passed, by syringe, through a Sep-Pak Cls cartridge (Waters) to remove nonpolar (lipid) components (31). (Cartridges were activated with methanol and washed with water before application of the samples.) Fresh PIPES buffer served as control. Concentrations of PGE2 and 6-ketoprostaglandin FI, (6-keto-PGF1,) (the stable metabolite of PGI& were assayed in duplicate in the above buffer incubates by enzyme immunoassay (EIA) (Advanced Magnetics, Cambridge, MA). Preliminary experiments indicated that neither meclofenamate nor MK-886 in the concentrations employed had any influence on this assay. Screening of prostanoids in untreated conditioned buffer was performed by gas chromatography/mass spectrometry using deuterated standards as previously described (15). Preparation of isolated tracheal smooth muscle. Tracheas of mongrel dogswere excisedimmediately after death under pentobarbital anesthesiaand openedlongitudinally along the ventral surface.Smooth musclestrips (0.3 cm x 1.2 cm) bordered by 0.5 cm of cartilage were prepared from the muscularis portion by transversesectioningof the tissuefollowing removal of epithelium, fascia, and fat. Eight to 12 strips were prepared from each trachea. The strips were mounted in musclebaths containing 10 ml of Krebs-Henseleit solution maintained at 37°C and gassedcontinuously with 95% O2and 5% CO,. Isometric tension was measuredwith Grass force displacement transducers(modelFT 03) and recorded on a GrassPolygraph (model7D or 79D). The strips were treated with sodiummeclofenamate (4.5 M) throughout the experiment to inhibit the endogenousproduction of cyclooxygenaseproducts. The tissues were allowed to equilibrate for 2 h at a resting tension of 3 g (previously determined to be optimum for this preparation) and were washed at 15-min intervals with Krebs-Henseleit solution. Experimental protocol. The study consisted of four experimental protocols. In protocol 1, after tracheal smooth muscle was pretreated with conditioned buffer (supernatant fraction of PIPES buffer incubated with cultured epithelial cells for 5 h) for 5 min, the contractile dose-responserelationship to histaminewasdeterminedand normalized to maximal contraction in the presence of carbachol (100 PM). In protocol 2, tracheal smoothmusclewasprecontracted by histamine (5 PM), and the responseto conditioned buffer was determined when addedto the bath as a 2ml bolus or in increments of 0.1, 0.2, 0.7, 0.3, and 0.7 ml over a 35-min period to generate a cumulative concentration-responsecurve. In the latter cases,relaxation was normalized to maximum relaxation with isoprotereno1(100 PM). In protocol 1, the effect of pretreatment with the epithelium-derived supernatant (conditioned buffer) was compared with that of pretreatment with fresh buffer alone on the dose-responseto histamine. In protocol 2, the effects of fresh buffer, conditioned buffer, conditioned buffer from meclofenamate-treated cells, conditioned buffer from MK-886-treated cells, and conditioned buffer following removal of lipid were comparedon musclestrips precontracted by histamine. In the experiments in which the concentration dependenceof relaxation to the conditioned buffer was investigated, the effect of time alonewas alsodetermined. In protocol 3, to determine the extent to which the cell culture-derived relaxant activity was agonist specific, the ability of conditioned buffer to reverse contraction inducedby 5-hydroxytryptamine (0.15 PM), methacholine (0.13 PM), and histamine (5 FM) were compared.In protocol 4 the relative relaxant activities of authentic PGE2and PGIZ , alone and in combination, were determined on tracheal smooth muscleprecontracted with histamine.
AIRWAY
SMOOTH
MUSCLE
Reagents. Sterile HBSS containing 0.35 g/l sodiumbicarbonate but without magnesium,calcium, or phenol red was prepared by Flow Laboratories, Irvine, Scotland. Krebs-Henseleit was prepared to contain (in mM) 117.6 NaCl, 5.4 KCl, 1.0 NaH2P04, 1.2 MgSO,, 2.5 CaC12,25.0 NaHC03, and 11.1 glucose.All smoothmuscleagonistsand metabolic pathway inhibitors were prepared fresh on the day of use. Carbachol, histamine, 5-hydroxytryptamine, and methacholine (Sigma Chemical, St. Louis, MO) were dissolvedin Krebs-Henseleit solution and kept on ice. Isoproterenol (Sigma)was dissolvedin KrebsHenseleit solution with the addition of ascorbic acid (10B5M) for preservation and kept on ice. MK-886 (generouslydonated by Merck Frosst, Quebec, Canada) was dissolved in PIPES buffer and sodiummeclofenamate(generouslydonatedby Warner Lambert, Ann Arbor, MI) wasdissolvedin Krebs-Henseleit solution or PIPES buffer and eachbrought to final concentration just prior to use.Prostaglandin E2 and the sodiumsalt of PG12(Upjohn) were dissolvedin ethanol at a concentration of I mg/ml and diluted cold in Krebs-Henseleitsolution just prior to use. Data analysis. Values for changesin smooth muscletension were expressedas a percentage of the maximal contractile or relaxant effects elicited by carbachol or isoproterenol, respectively. The resultswere expressedasthe meant SE. Statistical comparisonof meanswasmadeusing Student’s t test for group data. A P value ~0.05 was consideredsignificant. Animal care and use. Animals were handled in accordance with the standardsestablishedby the Animal Welfare Act and set forth in the Department of Health, Education, and Welfare (National Institutes of Health) guidelines.Experimental protocols were reviewed and approved by this institution’s Committee on Animal Care and Use. RESULTS
Treatment of tracheal smooth muscle with conditioned buffer for 5 min resulted in a significant rightward sh.ift of the contractile dose-response curve to histamine in the range of 10 t 8 to 5 X low4 M. Responsesto histamine at the upper end of the curve were unaffected (Fig. 1). To determine whether the relaxant activity of conditioned buffer exhibited dose dependence, tracheal smooth muscle precontracted with 5 PM histamine, a concentration near the mean contractile effective con-
z
n
100
. . . . . . . a......
Fresh
Buffer (n=8)
-
Cond.
Buffer (n=6)
1
* I
g
I
1()-7 1 O-8 HISTAMINE
I
l(+
I
I(+
I
1O-4
1 1O-3
M
CONCENTRATION
Fig. 1. Summary data from 8 experiments showing inhibitory effect of 5 min pretreatment of tracheal smooth muscle with conditioned buffer on dose-response curve to histamine over range of 1 X 10-’ to 5 X 10V4 M. Comparison is made to the dose-response curve generated following pretreatment with fresh buffer. Data normalized to maximal contraction induced by 100 PM carbachol and expressed as means t SE; n represents no. of tracheal smooth muscle strips. *P < 0.05, conditioned vs. fresh buffer.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
BRONCHIAL
EPITHELIUM
INHIBITS
centration (E&J, was challenged with volumes of conditioned buffer accumulating to a total of 2 ml. Figure 2 depicts a tracing from a typical experiment following this protocol comparing the effects of fresh and conditioned buffer. In Figure 3, data from 7-9 experiments are summarized. Compared with fresh buffer, conditioned buffer caused significant relaxation of tracheal smooth muscle at all but the 0.1 ml addition, reaching a maximum of more than 90% at 2.0 ml. Relaxation produced by fresh buffer was not significantly different from that produced by time alone. To determine the relative effectiveness of conditioned buffer against different contractile stimuli, tracheal smooth muscle was precontracted with &hydroxytryptamine (5HT, 0.15 PM), methacholine (0.13 PM), or histamine (5 PM) and challenged with Z-ml aliquots of conditioned buffer. These concentrations were deterHist
(5uM)
N
10 9
(ml)
0.1t
0.2t
t Oa7
1 0
FRESH
t 03.
‘)
t 0.7
BUFFER
AIRWAY
SMOOTH
L231
MUSCLE
mined, in preliminary experiments, to represent the EC50 for each agonist in this preparation. Figure 4 indicates that reversal of 5-HT-induced contraction was similar to, although statistically less than, that of histamineinduced contraction. In contrast, methacholine-induced contraction was much more resistant to the relaxant activity present in conditioned buffer. To investigate the nature of the relaxant factor(s) present in conditioned buffer, the effects of various buffer preparations on tracheal smooth muscle precontracted by histamine (5 ,uM) were compared (Fig. 5). The addition of 2 ml of conditioned buffer reversed the contraction by 83 t 4% compared with the effect of fresh buffer (14 t 3%). Removal of lipids from conditioned buffer by Sep-Pak Cl8 extraction markedly reduced the relaxant effect to a level (22 t 4%) not statistically different from fresh buffer. Fresh buffer, itself, caused an 18% reversal of the contracted state. The cause of this effect is undetermined, but could not be explained on the basis of dilution of the agonist or a shift in bath pH or temperature. To determine the involvement of eicosanoids in the lipid-associated relaxation of histamine-contracted tis-
Is0
q q
20 10
t
(ml)
E -
1
5-HT (n=5) Meth (n=6) Histamine (n=i 1)
0.1
r
9 0 1
CONDITIONED .
*
BUFFER
*
’
2
=
*
Om3
10
6
07.
I Is0
min
Fig. 2. Tracings from a typical experiment showing cumulative effect of various volumes of fresh or conditioned buffer, up to a total volume of 2 ml, on tracheal smooth muscle precontracted by 5 PM histamine. Iso, isoproterenol, 100 PM, added to induce maximal relaxation of tissues. . .. .. .. 0 . .. .. .
Untreated
(n=9)
TV
Fresh
Buffer
(n=7)
-
Cond.
Buffer
(n=8)
Fig. 4. Relative effectiveness of conditioned buffer in reversing contraction of tracheal smooth muscle induced by 0.15 FM 5-hydroxytryptamine (5-HT), 0.13 PM methacholine (Meth) or 5 PM histamine. Data normalized and expressed as in Fig. 3; n represents no. of strips. *p c 0.05, compared with histamine.
0 0
2 0 -20 Ia x
20
2 -40 ii! i
z
0 i= 0 3
-60
Fresh
Buffer (n=9)
Cond.
Buffer
- lipid (n=8)
Cond.
Buffer
(n=8)
Meclofenamate
1
(n=8)
MK 886 (n=6)
4 -20
x
W a
s -80
-40 i z -60 2 = -80
1.0
0.1
TOTAL
AC%bMULATED
1.3
DOSE
2.0
(ml)
Fig. 3. Summary data from a series of experiments, as depicted in Fig. 2, showing cumulative relaxant effect of various volumes of fresh or conditioned buffer on smooth muscle precontracted by 5 PM histamine. Effect of time alone (untreated) on contractile state was determined by measuring level of contraction in untreated tissues at times corresponding to those at which measurements were made in buffer-treated tissues. Data normalized to maximal relaxation to 100 PM isoproterenol and expressed as means & SE. Abscissa, accumulated volume added on a log scale. n represents number of strips. *P < 0.05, conditioned vs. fresh buffer.
.O
Fig. 5. Effect of single 2.0-ml additions of fresh buffer, conditioned buffer (Cond Buffer), conditioned buffer with lipids removed by SepPak extraction (Cond Buffer-lipid), conditioned buffer from meclofenamate-preincubated cells to inhibit prostaglandin production (Meclofenamate), and conditioned buffer from MK-886-preincubated cells to inhibit leukotriene production (MK-886) on tracheal smooth muscle precontracted with 5 PM histamine. Data normalized and expressed as in Fig. 3; n represents no. of strips. *‘P < 0.05, compared with fresh buffer.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
L232
BRONCHIAL
EPITHELIUM
INHIBITS
sues, inhibitors of prostaglandin and leukotriene synthesis were added to canine bronchial epithelial cell cultures during conditioning of the buffer. The relaxant activity of buffer harvested from canine bronchial epithelial cells in the presence of the cyclooxygenase inhibitor, sodium meclofenamate, was significantly inhibited (83 t 4 to 31 t 3%), almost to the level of the lipid-free conditioned buffer. In contrast, treatment with MK-886, a selective inhibitor of &lipoxygenase, had no effect on epithelial cell-derived relaxant activity (Fig. 5). Two random samples of buffer conditioned by epithelial cells in culture were subjected to analysis in duplicate by GC/MS to determine the spectrum of prostanoids present. Results indicated the presence of PGF2, (243 t 40 pg/ml); PGE, (4,787 t 411 pg/ml); 6-keto-PGF1, (1,394 t 496 pg/ml), whereas TXB2, PGDB, and its principal metabolite, 9&-PGFz, were undetectable. Because PGE2 and PG12 are the principal prostanoid metabolites reported to induce relaxation of canine tracheal smooth muscle (19), PGE2 and 6-keto-PGF1, were assayed in the various buffer preparations described above. The relative amounts of PGE2 and 6-keto-PGF1, followed the same pattern and were distributed in the following order: MK-886-conditioned buffer > conditioned buffer > meclofenamate-conditioned buffer > lipid-extracted buffer (Fig. 6). In preparations from cells with uninhibited prostaglandin production, PGEz concentrations were approximately twice those of 6-ketoPGF1, (Fig. 6). To confirm that PGEz and PGIB behaved as relaxant factors in the concentrations present in conditioned buffer, both prostanoids were added, singly and in combination, to precontracted tracheal smooth muscle in concentrations equivalent to one-fourth to one-third of those in 2 ml of conditioned buffer (Fig. 7). PGEg reversed histamine-induced contraction of tracheal smooth muscle by 49 t 4%, whereas PGIz was without significant effect. In the presence of these concentrations of PGIZ, relaxation by PGEz increased to 65 t 5%, a level statistically greater than that of PGEz alone (Fig. 7). DISCUSSION
The results of the present study indicate that canine bronchial epithelial cells in culture exhibit baseline release of a factor or factors that downregulate contractile (n=6)
g
6ooo- 111 fxE2 5ooo _ q 6-keto-PGF1,
-s
e
.
4000
-
g 3000 z zv) 2000
-
(n-12)
-
0
g
1000
-
(““‘)(n=lO)
Cond. Buffer
Meclofenamate
C.B.-lipid
MK 886
Fig. 6. Concentrations of PGEg and 6-keto-PGF,, in various buffer preparations described in Fig. 5, as determined by enzyme immunoassay. Data are means t SE, n represents no. of experiments.
AIRWAY
SMOOTH
q 20
-80
1
!
MUSCLE PG12 (n=13) IFEz
q
(n=ll)
PGIz + PGEz
(n=13)
*
d
Fig. 7. Effect of PGEz (1.7-3 ng), PGI, (1 ng), and PGEz (1.7-3 ng) in presence of PG12 (1 ng) on tracheal smooth muscle precontracted by histamine (5 PM). Data normalized and expressed as in Fig. 3; n represents no. of strips. *Fp c 0.05, PGEz vs. PGE, + PG12.
responses of tracheal smooth muscle. We have demonstrated the presence of this activity by the inhibition of histamine-induced contraction, as well by the dose-dependent reversal of contraction in histamine-treated tissues. Barnett et al. (6) reported similar inhibitory activity in supernatants from cultured canine tracheal epithelial cells that they ascribed to PGE2. In their studies, however, inhibition of the effects of electrical field stimulation, presumably at a prejunctional site, was seen only with supernatants from bradykinin-stimulated cells and not with those from unstimulated cells. It is interesting to note that concentrations of PGEz in supernatants from unstimulated cells reported by Barnett and colleagues (6) are similar to\ those found in the present . study. If this prostanoid is the factor principally responsible for the inhibitory effect in both studies, the lower concentrations required to elicit an effect in the present study may be related to a greater potency at postjunctional, as opposed to prejunctional, sites. Although the mechanism of the relaxant activity on smooth muscle was not directly investigated in the present study, experiments carried out in tissues contracted by 5-hydroxytryptamine and methacholine indicate that the effect is not histamine specific. Thus the effect appears to be exerted, albeit to a varying extent, at a point in the contractile process common to these three smooth muscle contractile agonists. It is known that histamine, 5-hydroxytryptamine, and muscarinic agonists, such as methacholine, stimulate the hydrolysis of polyphosphatidylinositols, the production of inositol phosphates and diacylglycerols, and increase intracellular concentration of free Ca2+, leading to smooth muscle contraction (21). ,&Agonists and PGE2 and PGI2 are known to activate adenylate cyclase, increase CAMP levels, and inhibit contraction (18, 19, 21). The involvement of this inhibitory pathway in the activity of cellderived supernatants in the present study is possible and is consistent with the observation that supernatants were less effective in reversing contraction induced by muscarinic stimulation, which has been shown to inhibit adenylate cyclase activity (21). Lipid extraction of the conditioned buffer resulted in the elimination of the inhibitory activity, consistent with the notion that relaxation was related to the presence of eicosanoid mediators. Support for the principal role of prostanoids in reversing the histamin.e-induced contrac-
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
BRONCHIAL
EPITHELIUM
INHIBITS
tion was provided by experiments in which relaxant activity was reduced to that of lipid-extracted buffer by pretreatment of the epithelial cells with sodium meclofenamate. The lack of an effect of pretreatment of the cells with the lipoxygenase inhibitor, MK-886, suggests that products of this pathway did not contribute to the lipid-associated inhibition. Analysis of buffer preparations indicated the presence of PGEz and 6-keto-PGF1,, the PG12 metabolite, in conditioned buffer. Cultured epithelial cells from a varietv of species, including the dog; are well known to synthesize PGEz (6, 8, 30). In intact airways, the principal source of PGI, appears to be the smooth muscle (25); however, 6-keto-PGF1, is also produced by airway epithelial cells (30). Although Shore et al. (23) have demonstrated the ability of PGIz to relax tracheal smooth muscle contracted by histamine, its activity is less than that of PGEZ in this system (2). Consistent with these observations, results of the present study indicate that exogenous PGEZ, in concentrations similar to those present in conditioned buffer, could account for the observed relaxation. Although PGIz concentrations were estimated to be only 60% less than those of PGEz, exogenous PGIz induced only minimal relaxation of histamine-contracted tracheal smooth muscle. The apparent synergism observed when both PGEz and PG12 were added to the tracheal preparations may have been due to the combined abilities of these mediators to reduce cakium influx (3) and stimulate adenylate cyclase (17, 24) and suggests that these two prostanoids may normally act in concert to mediate epithelium-derived modulation of tracheal contraction. Although the role of prostaglandins in the activity of epithelium-derived relaxant factor in vivo remains controversial, the present study demonstrates that canine bronchial epithelial cells in culture exhibit baseline release of prostanoids in concentrations sufficient to modulate the contraction of airwav smooth muscle. Based on our current knowledge of the array of cyclooxygenase products released from these cells and their effects on tracheal smooth muscle, the results of the present study are consistent with PGEZ being the principal factor responsible for this activity. The authors thank William Dunlap III for excellent technical support, Brian Schofield for expertise in preparing and evaluating microscopic samples, and Dr. Brad Undem for comments and suggestions. This study was supported by National Institutes of Health Grants HL-30195, ES-03505, and AI-20136. Address for reprint requets: E. W. Spannhake, Div. of Physiology, Dept. of Environmental Health Sciences, The Johns Hopkins School of Hygiene and Public Health, 615 N. Wolfe St., Baltimore, MD 21205.
AIRWAY
4. 5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21. REFERENCES 1. Aizawa, H., N. Miyazaki, N. Shigematsu, and M. Tomooka. A possible role of airway epithelium in modulating hyperresponsiveness. Br. J. Pharmacol. 93: 139-145, 1988. 2. Anderson, W. H., J. J. Krazanowski, J. B. Polson, and A. Szentivanyi. Characteristics of histamine tachyphylaxis in canine tracheal smooth muscle. Naunyn-Schmeidebergs - Arch. Pharmacol. 308: 117-125,1979. 3, Anderson, W. H., J. J. Krazanowski, J. B. Polson, and A. Szentivanyi. The effect of prostaglandin E2 on histamine-stimulated calcium mobilization as a possible explanation for histamine
22.
23.
24.
SMOOTH
MUSCLE
L233
tachyphylaxis in canine tracheal smooth muscle. Nuunyn-Schmeidebergs Arch. Pharmacol. 322: 72-77, 1983. Barnes, P. J. Inflammatory mediator receptors and asthma. Am. Rev. Respir. Dis. 135: S26-S31, 1987. Barnes, P. J., F. M. Cuss, and J. B. Palmer. The effect of airway epithelium on smooth muscle contractility in bovine trachea. Br. J. Pharmacol. 86: 685-691, 1986. Barnett, K., D. B. Jacoby, J. A. Nadel, and S. C. Lazarus. The effects of epithelial cell supernatant on contractions of isolated canine tracheal smooth muscle. Am. Rev. Respir. Dis. 138: 780783,1988. Brauntein G., C. Labat, S. Brunelleschi, J. Benveniste, J. Marsac, and C. Brink. Evidence that the histamine sensitivity and responsiveness of guinea-pig isolated trachea are modulated by epithelial prostaglandin Ez production. Br. J. PharmacoZ. 95: 300-308.1988. Churchill, L., F. H. Chilton, J. H. Resau, R. Bascom, W. C. Hubbard, and D. Proud. Cyclooxygenase metabolism of endogenous arachidonic acid by cultured human tracheal epithelial cells. Am. Rev. Respir. Dis. 140: 449-459, 1989. Coleman, D. L., I. K. Tuet, and J. H. Widdicombe. Electrical properties of dog tracheal epithelial cells grown in monolayer culture. Am. J. Physiol. 246 (Cell Physiol. 15): C355-C359, 1984. Flavahan, N. A., L. L. Aarhus, T. J. Rimele, and P. M. Vanhoutte. The respiratory epithelium inhibits bronchial smooth muscle tone. J. Appl. Physiol. 58: 834-838, 1985. Hay, D. W. P., S. G. Farmer, D. Raeburn, V. A. Robinson, W. W. Fleming, and J. S. Fedan. Airway epithelium modulates the reactivity of guinea-pig respiratory smooth muscle. Eur. J. Pharmacol. 129: 11-18, 1986. Holroyde, M. C. The influence of epithelium on the responsiveness of guinea-pig isolated trachea. Br. J. Pharmacol. 87: 501-507, 1986. Holtzman, M. J., H. Aizawa, J. A. Nadel, and E. J. Goetzl. Selective generation of leukotriene B4 by tracheal epithelial cells from dogs. Biochem. Biophys. Res. Commun. 114: 1071-1076,1983. Laitinen, L. A., M. Heino, A. Laitinen, T. Kava, and T. Haahtels. Damage of the airway epithelium and bronchial reactivity in patients with asthma. Am. Rev. Respir. Dis. 131: 599-606, 1985. Liu, M. C., E. R. Bleecker, D. Proud, T. L. McLemore, and W. Hubbard. Profiling of bisenoic prostaglandins and thromboxane Bz in bronchoalveolar fluid from the lower respiratory tract of human subjects by combined capillary gas chromatography-mass spectrometry. Prostaglandins 35: 67-79, 1988. Leikauf, G. D., I. F. Ueki, J. A. Nadel, and J. H. Widdicombe. Bradykinin stimulates chloride secretion and prostaglandin E2 release by canine tracheal epithelium. Am. J. Physiol. 248 (Renal Fluid Electrolyte Physiol. 17): F48-F55, 1985. Littner, M., G. Kazmi, F. Lott, M. Golub, and M. Berger. Cyclooxygenase products and cyclic nucleotide levels with infusion of arachidonic acid, PG12, PGEZ, PGF2,, and 6-keto-PGF,, in an isolated dog lung. Am. Rev. Respir. Dis. 128: 868-874, 1983. Morrison, K. J., Y.-S. Gao, and P. M. Vanhoutte. Epithelial modulation of airway smooth muscle. Am. J. Physiol. 258 (Lung CelZ. MOL. Physiol. 2): L254-L262, 1990. Madison, J. M., C. A. Jones, R. M. Sankary, and J. K. Brown. Differential effects of prostaglandin E2 on contractions of airway smooth muscle. J. Appl. Physiol. 66: 1397-1407, 1989. Munakata, M., W. Mitzner, and H. Menkes. Osmotic stimuli induce epithelial-dependent relaxation in the guinea pig trachea. J. Appl. Physiol. 64: 466-471, 1988. Rasmussen, H., 6. Kelley, and J. S. Douglas. Interactions between Ca2+ and CAMP messenger system in regulation of airway smooth muscle contraction. Am. J. Physiol. 258 (Lung Cell. Mol. Physiol. 2): L279-L288, 1990. Rouzer C. A., A. W. Ford-Hutchinson, H. E. Morton, and J. W. Gillard. MK886, a potent and specific leukotriene biosynthesis inhibitor blocks and reverses the membrane association of 5lipoxygenase in ionophore-challenged leukocytes. J. Biol. Chem. 265: 1436-1442,199O. Shore, S. A., W. S. Powell, and J. G. Martin. Endogenous prostaglandins modulate histamine-induced contraction in canine tracheal smooth muscle. J. Appl. Physiol. 58: 859-868, 1985. Smith, J. J., J. D. McCann, and M. J. Welseh. Bradykinin
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
L234
25.
26.
27.
28. 29.
BRONCHIAL
EPITHELIUM
INHIBITS
stimulates airway epithelial Cl-secretion via two second messenger pathways. Am. J. Physiol. 258 (Lung Cell. Mol. Physiol. 2): 13691377,199o. Stuart-Smith, K., and P. M. Vanhoutte. Arachidonic acid evokes epithelium-dependent relaxations in canine airways. J. Appl. Physiol. 65: 2170-2180, 1988. Tschirhart, E., N. Frossard, C. Bertrand, and Y. Landry. Arachidonic acid metabolites and airway epithelium-dependent relaxant factor. J. PharmacoZ. Exp. Ther. 243: 310-316, 1987. Undem, B. J., D. G. Raible, N. F. J. Adkinson, and G. K. Adams III. Effect of removal of epithelium on antigen-induced smooth muscle contraction and mediator release from guinea pig isolated trachea. J. Pharmacol. Exp. Ther. 244: 659-665, 1988. Vanhoutte, P. M. Airway epithelium and bronchial reactivity. Can. J. Physiol. Pharmacol. 65: 448-450, 1987. Widdicombe, J. G., I. F. Ueki, D. Enery, D. Margolskee, J.
AIRWAY
SMOOTH
MUSCLE
Yerget, and J. A. Nadel. Release of cyclooxygenase products from primary cultures of tracheal epithelia of dog and human. Am. J. Physiol. 257 (Lung Cell. Mol. Physiol. 1): L361-L365, 1989. 30. Xu, G. L., K. Sivarajah, R. Wu, P. Nettesheim, and T. Eling. Biosynthesis of prostaglandins by isolated and cultured airway epithelial cells. Exp. Lung Res. 10: 101-114, 1986. 31. Yoss, E. B., E. W. Spannhake, J. T. Flynn, J. E. Fish, and S. P. Peters. Arachidonic acid metabolism in normal human alveolar macrophages: stimulus specificity for mediator release and phospholipid metabolism, and pharmacologic modulation in vitro and in vivo. Am. J. Respir. CeZZ Mol. Biol. 2: 69-80, 1990. 32. Yu, X. Y., and E. W. Spannhake. The comparison of the effects of histamine on PGEz release from canine tracheal epithelial (CTE) and bronchial epithelial (CBE) cells in culture (Abstract). Am. Reu. Respir. Dis. 141: A911, 1990.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
