BASIC SCIENCE SECTION Failure of ritodrine to prevent preterm labor in the sheep StephenJ. Lye, PhD,"·b.c Brian A. Dayes, BSc,c Christian L. Freitag, MSc,"'c Julie Brooks, PhD,c and Robert F. Casper, MDb Toronto and London, Ontario, Canada OBJECTIVES: The purpose of this study was to determine whether continuous infusion of ritodrine could prevent preterm delivery in sheep. STUDY DESIGN: Sheep in preterm labor induced by RU 486 (mifepristone) received infusions of either ritodrine (n = 5) or saline solution (n = 5), and the progress of labor was monitored. 132-Adrenergic receptor density and function (agonist-induced cyclic adenosine monophosphate production) was measured in myometrial samples from both groups. RESULTS: Ritodrine initially inhibited labor contractions. This inhibition was only maintained for 16 hours, after which both the amplitude and frequency of electromyographic bursts and contractions returned. The failure of the myometrium to respond to ritodrine (desensitization) was associated with significant reductions in agonist-induced cyclic adenosine monophosphate production and 132-adrenergic receptor concentration in myometrial tissue collected from these animals compared with the saline solution-treated controls. CONCLUSIONS: Continuous infusion of ritodrine to sheep in preterm labor produces only a transient inhibition of contractions. This desensitization is caused by a down-regulation of myometrial 132-adrenergic receptors. (AM J OSSTET GYNECOL 1992;167:1399-408.)
Key words: Preterm labor, tocolysis, 13"-adrenergic agonists, sheep, uterine contractions Although the most widespread treatment for preterm labor is the continuous infusion of a selective 132adrenergic agonist, the efficacy ofthis therapy has been questioned.' Clearly these agents are potent inhibitors of myometrial contractile activit y2. 3; however, their clinical use for the prevention of preterm delivery requires that they maintain their effectiveness for long periods to significantly improve neonatal outcome. Surprisingly, in vivo studies analyzing the continuous changes in myometrial contractile activity during 132-adrenergic agonist administration over long periods have not been conducted. Studies with short-duration (several hours) infusions of !32-adrenergic agonists have clearly shown an inhibitory effect on uterine contractions in both pregnant" and nonpregnant 4 women. Similar findings have been made in a number of animal species, in-
From the Division of Perinatology, Samuel Lunenfeld Research Institute, Mount Sinai Hospital'; the Department of Obstetrics and Gynaecology, University ofTorontob; and the Lawson Research Institute, St. Joseph's Hospital, University of Western Ontario.' Supported by funds from the Medical Research Council of Canada, the Physicians Services Inc. Foundation of Ontario, and the Hospital for Sick Children Foundation. SJ.L. is a Career Scientist of the Ontario Ministry of Health; B.A.D. received a studentship from the Easter Seals Research Foundation. Received for publication November 12, 1991; revised March 16, 1992; accepted March 31,1992. Reprint requests: Stephen J. Lye, PhD, Division of Perinatology, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Ave., Suite 775, Toronto, Ontario, Canada, M5G 1X5. 611 138215
cluding the sheep" and rat. 6 We and others have also demonstrated the acute inhibition of myometrial contractions by 132-adrenergic agonists in vitro in humans 3 and animals.' However, if exposure to the agonist is prolonged, we have observed in the human 3 and sheep' that contractions return even though the drug remains active; the myometrium has become desensitized to the 132-adrenergic agonist. We have obtained similar results in vivo during experiments in nonpregnant sheep.s In these studies a continuous infusion of the 132-adrenergic agonist isoproterenol was only able to maintain myometrial inhibition for about 45 minutes, after which both the frequency and amplitude of uterine contractions returned to preinfusion levels. There are no reports examining the ability of !32-adrenergic agonists to inhibit contractions during preterm labor in animal models. 132-Adrenergic agonists function by increasing intracellular levels of cyclic adenosine monophosphate (cAMP), which in turn is believed to inhibit the activity of regulatory contractile proteins (e.g., myosin lightchain kinase). We and others" 9 have shown that during in vitro desensitization the ability of the myometrium to generate cAMP in response to 132-adrenergic agonists is suppressed. In addition to concerns as to the efficacy of these drugs, a number of side effects associated with agonist therapy have been reported. Changes in the endocrine environment have also been observed during infusion of 132-adrenergic agonists. In nonpregnant'O and
1400 Lye et al.
pregnant" (not in labor) sheep plasma concentrations of stimulatory prostaglandins were elevated during infusion of isoproterenol. Although prostaglandins have potent effects on the contractile activity of both vascular and myometrial smooth muscle, the consequences of their elevation during 132-adrenergic agonist infusion is not known. There is little information concerning the effects of long-term 132-adrenergic agonist administration on the fetus. In view of the controversy surrounding the use of 132-adrenergic agonists, we conducted our study to determine the effects of prolonged infusions on both uterine contractile activity and on maternal and fetal cardiovascular parameters. Specifically, we asked these questions: (1) Will the clinically used 132-adrenergic agonist ritodrine produce long-term inhibition of uterine contractile activity in sheep in preterm labor? (2) If not, what are the biochemical mechanisms associated with this desensitization? (3) What is the effect of prolonged 132-adrenergic agonist administration on maternal and fetal heart rate and fetal breathing movements? (4) Is 132-adrenergic agonist administration associated with changes in maternal or fetal plasma concentrations of prostaglandins? Material and methods
Animals. Surgery was conducted on 10 sheep of mixed breeds at 113 to 118 days of gestation. With the sheep under general anesthesia vascular catheters (model Vll, Bolab, Lake Havasu City, Ariz.) were placed into the fetal carotid artery and jugular vein and into the fetal trachea (to record fetal breathing movements). Vascular catheters (model V4, Bolab) were also inserted into the maternal femoral artery and vein. Uterine contractile activity (intrauterine pressure) was monitored continuously by means of amniotic catheters (Bolab); multifilament stainless steel electrodes (Cooner Wire Corp., Chatsworth, Calif.) sewn into the myometrium were used to monitor electromyographic activity. Analgesics and prophylactic antibiotics were administered after surgery, and the animals were allowed ~7 days to recover before experiments were begun. Protocol. Preterm labor was induced in all 10 sheep by the administration of the progesterone receptor antagonist RU 486 (mifepristone) 10 mg/kg subcutaneously (Roussel Vclaf). At this dosage we were able to induce labor (defined as contractions of frequency >25 per 2 hours, duration of < 1 minute, and amplitude of >5 mm Hg) in 44.2 ± 4.2 hours. The pattern of intrauterine pressure and electromyographic activity in animals in which labor was induced with RV 486 was identical to that seen in spontaneous term or corticotropin-induced preterm labor in sheep. Once preterm labor was established, the sheep were divided at random into two groups to receive either (1) saline (control)
November 1992 Am J Obstet Gynecol
1 mllhour or (2) ritodrine 3 J-Lg/kg/min (i.e., approximately 10 mg / hr to the pregnant ewe) via the maternal femoral vein catheter. The infusions were continued until high-amplitude uterine contractions with signs of maternal discomfort (curling of the lip or superimposed abdominal straining movements) were observed. These symptoms are normal features of advanced labor in sheep and are usually an indication that delivery is imminent. To confirm that the return of labor contractions in ritodrine-treated animals was the result of myometrial desensitization the infusion rate of the agonist was doubled at 20- to 30-minute intervals up to 80 mg/hr and the effects on electromyographic and intrauterine pressure activity monitored. Animals were then killed (overdose of pentobarbital sodium). Delivery was not used as an end point in this study for two reasons: (1) Although RU 486 reliably induces the highactivity pattern of uterine activity indicative of labor, the cervices of these animals undergo variable degrees of softening and dilatation, with some animals showing only a small change (thus impeding delivery of the lamb); (2) we wished to collect myometrial tissue for biochemical analysis of the functional integrity of the 132-adrenergic receptor system. Blood samples (10 ml) were withdrawn from the maternal femoral artery and femoral vein catheters at 30minute intervals from 60 minutes before to 180 minutes after commencement of either ritodrine or saline infusions. Samples were also withdrawn at 60 minute intervals from the fetal carotid artery. The samples were collected into ice-cold, heparinized tubes and centrifuged (1500g for 10 minutes at 4° C); the plasma was removed and stored at - 20° C until assayed for 13,14dihydro-15-keto-prostaglandin F20 (PGFM), progesterone, and estradiol-1713 (maternal samples) or prostaglandin E2 (PGE 2) (fetal samples). Samples were also collected for determination of maternal and fetal blood gases (blood gas analyzer, ABL3, Corning, Medfield, Mass.). Polygraph monitoring. Intrauterine pressure, maternal and fetal blood pressure, and fetal breathing movements (tracheal pressure) were monitored by connection of the appropriate catheters to pressure transducers (model P23, Statham, Spectramed, Oxnard, Calif.) and the signals passed through a preamplifier (model 7Pl, Grass Instruments, Quincy, Mass.). Measurements of maternal and fetal heart rate were obtained by processing the preamplifier output through a tachograph (Grass). The myometrial electromyographic signal was processed through a wide-band AC preamplifier (model 7P5, Grass) with the use of a onehalf amplitude, low-frequency filter of 0.3 Hz and a onehalf amplitude, high-frequency filter of 10kHz. All signals were recorded on a polygraph (model 78, Grass). Data analysis. The frequency and mean maximum
Volume 167 Number 5
amplitude of electromyographic bursts and intrauterine pressure cycles (contractions) were calculated in each 2-hour period from 10 hours before the administration of saline solution-ritodrine until the completion of the experiment. Values for fetal and maternal blood pressure and heart rate were recorded at 5-minute intervals and the mean calculated over five 4-hour periods: (1) before administration of RU 486; (2) 4hour period before saline solution-ritodrine; (3) 4hour period after start of saline solution-ritodrine; (4) 4-hour period from + 8 to + 12 after start of saline solution-ritodrine; and (5) 4-hour period before the end of the experiment, when abdominal straining movements were present. Fetal breathing movements were defined as negative deflections in tracheal pressure of at least 2 mm Hg occurring in bursts of ::::30second durations. The mean percent time spent breathing was calculated for each of the five time periods. Myometrial cAMP generation. The integrity of 132adrenergic receptor function in the myometrium of saline- and ritodrine-infused animals was examined in vitro at the conclusion of the experiment. Samples of myometrial tissue were collected at autopsy into Krebs' Ringer bicarbonate buffer containing 4.6 mmol/L potassium chloride, 1.16 mmol/L magnesium sulfate heptahydrate, 1.16 mmol/L sodium phosphate monobasic, 2.5 mmol/L calcium chloride dihydrate, 115.5 mmol/L sodium chloride, 21.9 mmol/L sodium bicarbonate, and ILl mmol/L dextrose, pH 7.4. M yometrial strips (2 X 2 x 15 mm) were cut from the sample, and approximately 100 mg tissue was incubated in 5 ml of Krebs' at 37° C for 30 minutes to equilibrate. The strips were then placed in fresh Krebs' (5 ml) containing 0, 0.1, 1.0, 10, or 100 ILmol/ L isoproterenol. After 5 minutes the strips were rapidly frozen on dry ice and stored at -70° C. The frozen myometrial strips were subsequently weighed and minced in 1.5 ml ice-cold 5% trichloroacetic acid containing 0.5 mmol/L isobutylmethylxanthine (a phosphodiesterase inhibitor). The samples were homogenized (Brinkman Polytron, Sybron-Brinkman Canada, Rexdale, Ontario; setting 10, 10-second pulse x 4). Samples were spun at 1O,000g for 10 minutes at 4° C. The pellet was redissolved in 2 ml of O.IN sodium hydroxide for protein determination with bovine serum albumin (type IV, Sigma Chemical, St. Louis) as a standard. Aliquots of supernatant (3 x 100 ILl) were dried under air and each brought up in 0.5 ml of 50 mmol/L sodium acetate buffer, pH 6.2. Myometrial P2-adrenergic receptor analysis. Preparation of myometrial membrane fraction and subsequent analysis of ligand binding was carried out with a method similar to that of Dattel et al. 12 Myometrial membrane preparation. Frozen tissue (collected at autopsy) was pulverized on dry ice and
Failure of ritodrine to prevent preterm labor
1401
homogenized (2 X 80-second pulse at 8000 rpm; Ultra-Turrex T25 Polytron, Terochem Laboratories, Mississauga, Ontario) in 50 mmol/L Tris-I mmol/L ethylenediaminetetraacetic acid buffer (pH 8.0). Homogenized tissue was centrifuged at 1000g for 15 minutes at 4° C, and the supernatant retained on ice. The pellet was resuspended in Tris-ethylenediaminetetraacetic buffer and the homogenization repeated. The homogenates were combined and centrifuged at 40,000g for 30 minutes at 4° C. The pellet was retained and resuspended in 50 mmol/L Tris-4 mmol/L magnesium chloride buffer (pH 7.4), and the centrifugation step was repeated. The resulting pellet was suspended in I ml Tris-magnesium chloride buffer and frozen at - 70° C for future receptor assay. P2-Adrenergic receptor assay. Frozen membrane preparations were thawed and uniform suspensions obtained by brief homogenization with a Teflon wandglass dounce homogenizer. Protein content was determined as described above. Aliquots of the membrane preparation (20 to 30 ILg protein) were incubated in the presence of varying concentrations of the antagonist iodine 125-labeled cyanopindolol (6 to 200 pmol/L, Amersham, Oakville, Ontario) for 60 minutes at 30° C. Incubations were terminated via rapid vacuum filtration with cold Tris-magnesium chloride buffer through filters (Whatman GFIC, Whatman Limited, Maidstone, England). Radioactivity retained on the filters was counted (LKB 1282 Compugamma Universal 'V-counter, Fisher Scientific, Toronto). Nonspecific binding was defined as binding in the presence of 10 ILmol/ L isoproterenol. Radioimmunoassays. The arterial and venous concentrations of PGE z and PGFM in 500 ILl aliquots of plasma were estimated by specific radioimmunoassays that have been fully characterized and validated for sheep plasma. 13 In brief, plasma was acidified (to pH 3) with hydrochloric acid and the prostaglandins extracted with diethyl ether (8 volumes). The ether extract was evaporated to dryness under nitrogen, and the samples were reconstituted with phosphate-buffered saline solution and incubated with specific antisera and tritiated prostaglandins (PGE z , New England Nuclear, Lachine, Quebec; PGFM, Amersham). The interassay and intraassay coefficients of variation for each prostaglandin were
aw
rJJ 1 ° ° \T/o-o-o 1/
~
(0 -10
-5
T
1 T 1/·""'·
°
T .-.--.
.-.
T....../
T
• T/· I'·
• 0
T
Ritodrine
0-0 Saline
5
10
15
20
25
TIME FROM RITODRINE/SALINE TREATMENT (IN HRS) Fig. 1. Frequency (mean ± SEM) of contractions (intrauterine pressure cycles) during 10 hours before and 20 hours after start of ritodrine (solid circles) or saline (open squares) infusion to sheep (n = 5 for each group) in preterm labor.
cific antisera and tritiated steroids. The interassay and intraassay coefficients of variation for each prostaglandin were
Key words: Preterm labor, tocolysis, 13"-adrenergic agonists, sheep, uterine contractions Although the most widespread treatment for preterm labor is the continuous infusion of a selective 132adrenergic agonist, the efficacy ofthis therapy has been questioned.' Clearly these agents are potent inhibitors of myometrial contractile activit y2. 3; however, their clinical use for the prevention of preterm delivery requires that they maintain their effectiveness for long periods to significantly improve neonatal outcome. Surprisingly, in vivo studies analyzing the continuous changes in myometrial contractile activity during 132-adrenergic agonist administration over long periods have not been conducted. Studies with short-duration (several hours) infusions of !32-adrenergic agonists have clearly shown an inhibitory effect on uterine contractions in both pregnant" and nonpregnant 4 women. Similar findings have been made in a number of animal species, in-
From the Division of Perinatology, Samuel Lunenfeld Research Institute, Mount Sinai Hospital'; the Department of Obstetrics and Gynaecology, University ofTorontob; and the Lawson Research Institute, St. Joseph's Hospital, University of Western Ontario.' Supported by funds from the Medical Research Council of Canada, the Physicians Services Inc. Foundation of Ontario, and the Hospital for Sick Children Foundation. SJ.L. is a Career Scientist of the Ontario Ministry of Health; B.A.D. received a studentship from the Easter Seals Research Foundation. Received for publication November 12, 1991; revised March 16, 1992; accepted March 31,1992. Reprint requests: Stephen J. Lye, PhD, Division of Perinatology, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Ave., Suite 775, Toronto, Ontario, Canada, M5G 1X5. 611 138215
cluding the sheep" and rat. 6 We and others have also demonstrated the acute inhibition of myometrial contractions by 132-adrenergic agonists in vitro in humans 3 and animals.' However, if exposure to the agonist is prolonged, we have observed in the human 3 and sheep' that contractions return even though the drug remains active; the myometrium has become desensitized to the 132-adrenergic agonist. We have obtained similar results in vivo during experiments in nonpregnant sheep.s In these studies a continuous infusion of the 132-adrenergic agonist isoproterenol was only able to maintain myometrial inhibition for about 45 minutes, after which both the frequency and amplitude of uterine contractions returned to preinfusion levels. There are no reports examining the ability of !32-adrenergic agonists to inhibit contractions during preterm labor in animal models. 132-Adrenergic agonists function by increasing intracellular levels of cyclic adenosine monophosphate (cAMP), which in turn is believed to inhibit the activity of regulatory contractile proteins (e.g., myosin lightchain kinase). We and others" 9 have shown that during in vitro desensitization the ability of the myometrium to generate cAMP in response to 132-adrenergic agonists is suppressed. In addition to concerns as to the efficacy of these drugs, a number of side effects associated with agonist therapy have been reported. Changes in the endocrine environment have also been observed during infusion of 132-adrenergic agonists. In nonpregnant'O and
1400 Lye et al.
pregnant" (not in labor) sheep plasma concentrations of stimulatory prostaglandins were elevated during infusion of isoproterenol. Although prostaglandins have potent effects on the contractile activity of both vascular and myometrial smooth muscle, the consequences of their elevation during 132-adrenergic agonist infusion is not known. There is little information concerning the effects of long-term 132-adrenergic agonist administration on the fetus. In view of the controversy surrounding the use of 132-adrenergic agonists, we conducted our study to determine the effects of prolonged infusions on both uterine contractile activity and on maternal and fetal cardiovascular parameters. Specifically, we asked these questions: (1) Will the clinically used 132-adrenergic agonist ritodrine produce long-term inhibition of uterine contractile activity in sheep in preterm labor? (2) If not, what are the biochemical mechanisms associated with this desensitization? (3) What is the effect of prolonged 132-adrenergic agonist administration on maternal and fetal heart rate and fetal breathing movements? (4) Is 132-adrenergic agonist administration associated with changes in maternal or fetal plasma concentrations of prostaglandins? Material and methods
Animals. Surgery was conducted on 10 sheep of mixed breeds at 113 to 118 days of gestation. With the sheep under general anesthesia vascular catheters (model Vll, Bolab, Lake Havasu City, Ariz.) were placed into the fetal carotid artery and jugular vein and into the fetal trachea (to record fetal breathing movements). Vascular catheters (model V4, Bolab) were also inserted into the maternal femoral artery and vein. Uterine contractile activity (intrauterine pressure) was monitored continuously by means of amniotic catheters (Bolab); multifilament stainless steel electrodes (Cooner Wire Corp., Chatsworth, Calif.) sewn into the myometrium were used to monitor electromyographic activity. Analgesics and prophylactic antibiotics were administered after surgery, and the animals were allowed ~7 days to recover before experiments were begun. Protocol. Preterm labor was induced in all 10 sheep by the administration of the progesterone receptor antagonist RU 486 (mifepristone) 10 mg/kg subcutaneously (Roussel Vclaf). At this dosage we were able to induce labor (defined as contractions of frequency >25 per 2 hours, duration of < 1 minute, and amplitude of >5 mm Hg) in 44.2 ± 4.2 hours. The pattern of intrauterine pressure and electromyographic activity in animals in which labor was induced with RV 486 was identical to that seen in spontaneous term or corticotropin-induced preterm labor in sheep. Once preterm labor was established, the sheep were divided at random into two groups to receive either (1) saline (control)
November 1992 Am J Obstet Gynecol
1 mllhour or (2) ritodrine 3 J-Lg/kg/min (i.e., approximately 10 mg / hr to the pregnant ewe) via the maternal femoral vein catheter. The infusions were continued until high-amplitude uterine contractions with signs of maternal discomfort (curling of the lip or superimposed abdominal straining movements) were observed. These symptoms are normal features of advanced labor in sheep and are usually an indication that delivery is imminent. To confirm that the return of labor contractions in ritodrine-treated animals was the result of myometrial desensitization the infusion rate of the agonist was doubled at 20- to 30-minute intervals up to 80 mg/hr and the effects on electromyographic and intrauterine pressure activity monitored. Animals were then killed (overdose of pentobarbital sodium). Delivery was not used as an end point in this study for two reasons: (1) Although RU 486 reliably induces the highactivity pattern of uterine activity indicative of labor, the cervices of these animals undergo variable degrees of softening and dilatation, with some animals showing only a small change (thus impeding delivery of the lamb); (2) we wished to collect myometrial tissue for biochemical analysis of the functional integrity of the 132-adrenergic receptor system. Blood samples (10 ml) were withdrawn from the maternal femoral artery and femoral vein catheters at 30minute intervals from 60 minutes before to 180 minutes after commencement of either ritodrine or saline infusions. Samples were also withdrawn at 60 minute intervals from the fetal carotid artery. The samples were collected into ice-cold, heparinized tubes and centrifuged (1500g for 10 minutes at 4° C); the plasma was removed and stored at - 20° C until assayed for 13,14dihydro-15-keto-prostaglandin F20 (PGFM), progesterone, and estradiol-1713 (maternal samples) or prostaglandin E2 (PGE 2) (fetal samples). Samples were also collected for determination of maternal and fetal blood gases (blood gas analyzer, ABL3, Corning, Medfield, Mass.). Polygraph monitoring. Intrauterine pressure, maternal and fetal blood pressure, and fetal breathing movements (tracheal pressure) were monitored by connection of the appropriate catheters to pressure transducers (model P23, Statham, Spectramed, Oxnard, Calif.) and the signals passed through a preamplifier (model 7Pl, Grass Instruments, Quincy, Mass.). Measurements of maternal and fetal heart rate were obtained by processing the preamplifier output through a tachograph (Grass). The myometrial electromyographic signal was processed through a wide-band AC preamplifier (model 7P5, Grass) with the use of a onehalf amplitude, low-frequency filter of 0.3 Hz and a onehalf amplitude, high-frequency filter of 10kHz. All signals were recorded on a polygraph (model 78, Grass). Data analysis. The frequency and mean maximum
Volume 167 Number 5
amplitude of electromyographic bursts and intrauterine pressure cycles (contractions) were calculated in each 2-hour period from 10 hours before the administration of saline solution-ritodrine until the completion of the experiment. Values for fetal and maternal blood pressure and heart rate were recorded at 5-minute intervals and the mean calculated over five 4-hour periods: (1) before administration of RU 486; (2) 4hour period before saline solution-ritodrine; (3) 4hour period after start of saline solution-ritodrine; (4) 4-hour period from + 8 to + 12 after start of saline solution-ritodrine; and (5) 4-hour period before the end of the experiment, when abdominal straining movements were present. Fetal breathing movements were defined as negative deflections in tracheal pressure of at least 2 mm Hg occurring in bursts of ::::30second durations. The mean percent time spent breathing was calculated for each of the five time periods. Myometrial cAMP generation. The integrity of 132adrenergic receptor function in the myometrium of saline- and ritodrine-infused animals was examined in vitro at the conclusion of the experiment. Samples of myometrial tissue were collected at autopsy into Krebs' Ringer bicarbonate buffer containing 4.6 mmol/L potassium chloride, 1.16 mmol/L magnesium sulfate heptahydrate, 1.16 mmol/L sodium phosphate monobasic, 2.5 mmol/L calcium chloride dihydrate, 115.5 mmol/L sodium chloride, 21.9 mmol/L sodium bicarbonate, and ILl mmol/L dextrose, pH 7.4. M yometrial strips (2 X 2 x 15 mm) were cut from the sample, and approximately 100 mg tissue was incubated in 5 ml of Krebs' at 37° C for 30 minutes to equilibrate. The strips were then placed in fresh Krebs' (5 ml) containing 0, 0.1, 1.0, 10, or 100 ILmol/ L isoproterenol. After 5 minutes the strips were rapidly frozen on dry ice and stored at -70° C. The frozen myometrial strips were subsequently weighed and minced in 1.5 ml ice-cold 5% trichloroacetic acid containing 0.5 mmol/L isobutylmethylxanthine (a phosphodiesterase inhibitor). The samples were homogenized (Brinkman Polytron, Sybron-Brinkman Canada, Rexdale, Ontario; setting 10, 10-second pulse x 4). Samples were spun at 1O,000g for 10 minutes at 4° C. The pellet was redissolved in 2 ml of O.IN sodium hydroxide for protein determination with bovine serum albumin (type IV, Sigma Chemical, St. Louis) as a standard. Aliquots of supernatant (3 x 100 ILl) were dried under air and each brought up in 0.5 ml of 50 mmol/L sodium acetate buffer, pH 6.2. Myometrial P2-adrenergic receptor analysis. Preparation of myometrial membrane fraction and subsequent analysis of ligand binding was carried out with a method similar to that of Dattel et al. 12 Myometrial membrane preparation. Frozen tissue (collected at autopsy) was pulverized on dry ice and
Failure of ritodrine to prevent preterm labor
1401
homogenized (2 X 80-second pulse at 8000 rpm; Ultra-Turrex T25 Polytron, Terochem Laboratories, Mississauga, Ontario) in 50 mmol/L Tris-I mmol/L ethylenediaminetetraacetic acid buffer (pH 8.0). Homogenized tissue was centrifuged at 1000g for 15 minutes at 4° C, and the supernatant retained on ice. The pellet was resuspended in Tris-ethylenediaminetetraacetic buffer and the homogenization repeated. The homogenates were combined and centrifuged at 40,000g for 30 minutes at 4° C. The pellet was retained and resuspended in 50 mmol/L Tris-4 mmol/L magnesium chloride buffer (pH 7.4), and the centrifugation step was repeated. The resulting pellet was suspended in I ml Tris-magnesium chloride buffer and frozen at - 70° C for future receptor assay. P2-Adrenergic receptor assay. Frozen membrane preparations were thawed and uniform suspensions obtained by brief homogenization with a Teflon wandglass dounce homogenizer. Protein content was determined as described above. Aliquots of the membrane preparation (20 to 30 ILg protein) were incubated in the presence of varying concentrations of the antagonist iodine 125-labeled cyanopindolol (6 to 200 pmol/L, Amersham, Oakville, Ontario) for 60 minutes at 30° C. Incubations were terminated via rapid vacuum filtration with cold Tris-magnesium chloride buffer through filters (Whatman GFIC, Whatman Limited, Maidstone, England). Radioactivity retained on the filters was counted (LKB 1282 Compugamma Universal 'V-counter, Fisher Scientific, Toronto). Nonspecific binding was defined as binding in the presence of 10 ILmol/ L isoproterenol. Radioimmunoassays. The arterial and venous concentrations of PGE z and PGFM in 500 ILl aliquots of plasma were estimated by specific radioimmunoassays that have been fully characterized and validated for sheep plasma. 13 In brief, plasma was acidified (to pH 3) with hydrochloric acid and the prostaglandins extracted with diethyl ether (8 volumes). The ether extract was evaporated to dryness under nitrogen, and the samples were reconstituted with phosphate-buffered saline solution and incubated with specific antisera and tritiated prostaglandins (PGE z , New England Nuclear, Lachine, Quebec; PGFM, Amersham). The interassay and intraassay coefficients of variation for each prostaglandin were
aw
rJJ 1 ° ° \T/o-o-o 1/
~
(0 -10
-5
T
1 T 1/·""'·
°
T .-.--.
.-.
T....../
T
• T/· I'·
• 0
T
Ritodrine
0-0 Saline
5
10
15
20
25
TIME FROM RITODRINE/SALINE TREATMENT (IN HRS) Fig. 1. Frequency (mean ± SEM) of contractions (intrauterine pressure cycles) during 10 hours before and 20 hours after start of ritodrine (solid circles) or saline (open squares) infusion to sheep (n = 5 for each group) in preterm labor.
cific antisera and tritiated steroids. The interassay and intraassay coefficients of variation for each prostaglandin were
