Volume 154 Number 3

occurring in 46,XX individuals. Clin Obstet Gynecol 1972;15:157.

3. Simpson jL, Elias S, Malinak RL, Buttram VC. Heritable aspects of endometriosis. I. Genetic studies. AM j OSSTET GYNECOL 1980;137:327.

4. Malinak RL, Buttram VC, Elias S, Simpson jL. Her­itable aspects of endometriosis. II. Clinical character­istics of familial endometriosis. AM j OSSTET GYNECOL 1980; 137:332.

5. Kistner RW. Current concepts in endometriosis: an in­ternational review. j Reprod Med 1977; 19:299.

6. Dmowski WP, Radwanska E. Current concepts on pa­thology, histogenesis and etiology of endometriosis. Acta Obstet Gynecol Scand Suppl 1984;123:29.

7. Weed jC, Arquembourg PC. Endometriosis: can it pro­duce an autoimmune response resulting in infertility? Clin Obstet Gynecol 1980;23:885.

8. Badawy SZA, Cuenca V, Stitzel A, jacobs RDB, Tomar RH. Autoimmune phenomena in infertile patients with endometriosis. Obstet Gynecol 1984;63:271.

Familial tendencies to endometriosis

9. Coulam CB, Ryan RJ. Premature menopause. l. Etiology. AM j OSSTET GYNECOL 1979; 133:639.

10. Shands KN, Schlech WF, Hargrett NT, Dan BB, Schmid GP, BennettjV. Toxic shock syndrome: case-control stud­ies at the Centers for Disease Control. Ann Intern Med 1982;96:895.

II. Halgerson SD, Foster LR. Toxic shock syndrome in ore­gon: epidemiologic findings. Ann Intern Med 1982; 96:909.

12. Osterholm MT, Davis jP, Gibson RW, Forfang jC, Stolz Sj, VergerontjM. Toxic shock syndrome: relation to ca­tamenial products, personal health and hygiene, and sex­ual practices. Ann Intern Med 1982;96:954.

13. Shands KN, Schmid GP, Dan BB, et al. Toxic shock syn­drome in menstruating women: association with tampon use and staphyloccus au reus and clinical features in 52 cases. N Engl j Med 1980;303: 1436.

14. Houston DE. Evidence for the risk of pelvic endometriosis by age, race and socioeconomic status. Epidemiol Rev 1984;6:167.

a-Adrenergic receptors in human myometrium during

pregnancy

Goran Berg, M.D., Rolf G. G. Andersson, Ph.D., and Gunnar Ryden, M.D.

LinkOping, Sweden

The distribution and mechanisms of Ct-adrenergic receptors have been studied in myometrial preparations from women delivered by cesarean section at term. With the use of the radioligand tritiated dihydroergocryptine, the number of Ct-adrenergic receptors was 210 fmol/mg of protein in the uterine

fundus and 195 fmol/mg of protein in the lower uterine segment. Competition experiments showed that 60% of the Ct-adrenergic receptors had properties as Ct,-adrenergic receptors and 40% as Ct2-adrenergic

receptors. In vitro tension studies verified the existence of physiologically active Ct,- and Ct2-adrenergic receptors in the myometrial preparations. Ct2-Adrenergic receptor stimulation resulted in lowered levels of

intracellular cyclic adenosine monophosphate. The intracellular cyclic adenosine monophosphate was further reduced by additional Ct,-adrenergic receptor stimulation, probably secondary to an activation of calcium-calmodulin-dependent phosphodiesterase. (AM J OBSTET GYNECOL 1986;154:601-6.)

Key words: a-Adrenergic receptors, myometrium, labor, cyclic adenosine mono phosphate

There is considerable experimental evidence that the human myometrium contains both 13- and a-adrenergic receptors. I. 2 The mechanism of action of l3-adrenergic receptors in the myometrium has been thoroughly studied.' However, the distribution of a-adrenergic re­ceptors in human myometrium has just recently been investigated and only in nonpregnant women. 2 The

From the Department of Obstetrics and Gynecology and the Depart­ment of Pharmacology, Linkoping University.

This study was financially supported by Ostergotlands ltins landsting, the Expressen Perinatal Research Foundation, the Swedish Medical Society, Tore Nilsson's Fund, and the Medical Research Council (Grant 04P-6155).

Received for publication june 7, 1985; revised November 21, 1985; accepted November 25, 1985.

Reprint requests: Goran Berg, M.D., Department of Obstetrics and Gynecology, Linkoping Unversity, S-581 85 Linkoping, Sweden.

aim of this study was, therefore, to evaluate the a-ad­renergic receptor distribution and function in myo­metrium from pregnant women and to study in a conventional in vitro system the effect of different pharmacologic a-adrenergic receptor blockers on spon­taneously contracting myometrium and in receptor­binding assays in order to classify the subtypes of a­adrenergic receptors and to look for possible new means to improve today's tocolytic treatment.

Material and methods

Patients. Myometrial biopsy specimens were taken during cesarean section from the lower uterine seg­ment and/or the uterine fundus. The reason for ce­sarean section was related to obstetrics in all cases. The gestational ages were between 37 and 41 weeks and

601

602 Berg, Andersson, and Ryden

active labor had not started in any patients at the time of operative delivery. About two thirds of the women were anesthesized with an epidural block, and the re­mainder received the conventional intravenous anes­thetics thiopental sodium and leptanale. The myome­trial preparations were immediately immersed in Ring­er's solution and transported to the laboratory. Biopsy specimens from a total of 48 women were studied.

Studies of contractility in vitro. The myometrial preparations were mounted in special holders for registration of isometric tension by FT03 transducers on a Grass polygraph. The myometrial strips were suspended in a Krebs buffer solution at 37° C with the following composition: sodium chloride, 118.8 mmol/L; potassium chloride, 4.7 mmollL; calcium chlo­ride, 2.2 mmollL; magnesium chloride, 1.2 mmollL; sodium bicarbonate, 23.8 mmollL; potassium phos­phate, 1.2 mmol/L; glucose, 5.5 mmol/L. The solution was aerated with a gas mixture containing 95% oxygen and 5% carbon dioxide. All myometrial preparations were either untreated or pretreated with (1) phentol­amine (ai-and a,-adrenergic receptor blocker), (2) yo­himbine (a,-adrenergic receptor blocker), and (3) pra­lOsin (ai-adrenergic receptor blocker). All preparations were then treated with increasing concentrations of norepinephrine and the tension was measured.

Preparation of myometrial membranes. The myo­metrial tissue was cut into small pieces and homoge­nized in 10 volumes of buffer solution (sucrose, 0.25 mollL; Tris, 5 mmollL; magnesium chloride, 1 mmollL; pH 7.4) with the use of a Polytron PT 10. During the preparation procedure, the myometrial tis­sue was kept in ice-cold buffer solution. The homog­enate was filtered through a cheese-cloth and thereafter centrifugated at 400 x g for 10 minutes at 4° C. The supernatant was centrifugated at 28,000 x g for 10 minutes. The resulting pellets were washed and recen­trifuged twice at 28,000 x g. The washed pellets were suspended in three volumes of incubation buffer (Tris, 50 mmollL; magnesium chloride, 10 mmollL; pH 7.5) and used for the determination of tritiated dihydro­ergocryptine binding.

Binding of tritiated dihydroergocryptine. Binding of tritiated dihydroergocryptine to human myometrial membranes was assayed by filtration of the prepara­tions on glass fiber filters.' Membrane proteins were incubated with the indicated concentrations of tritiated dihydroergocryptine and incubation buffer in a total volume of 0.15 ml for 20 minutes at 37° C. At the end of incubation the samples were diluted in 2 ml of ice­

cold incubation buffer, followed by rapid vacuum fil­tration through Whatman glass fiber filters on a Millipore Sampling Manifold. The filters were then rapidly washed six times with 3 ml of the ice-cold in-

March, 1986 Am J Obstet Gynecol

cubation buffer. After drying, radioactivity bound to the filters was determined on a Packard liquid scintil­lation spectrometer. Specific binding was determined by subtracting nonspecific binding from total binding. In all figures and tables, bound tritiated dihydroer­gocryptine refers to specific binding. Nonspecific bind­ing was determined in the presence of phentolamine, 10 fJ..mollL. Protein was determined according to Lowry et al. 4

Determination of cyclic adenosine monophosphate in myometrial tissue. Stimulation with norepinephrine (1 fJ..moIlL) of the myometrial strips was performed in an organ bath with or without prazosin (1 fJ..moIlL) or yohimbine (1 fJ..moIlL) and in the same manner with myometrial preparations pretreated with propranolol (1 fJ..moIlL). After treatment with norepinephrine for 3 minutes in the organ bath, the strips were dismounted and frozen in Frigen 12 and solid carbon dioxide at -70° C.

The frozen preparations were homogenized in per­chloric acid and the nucleotides were separated on col­umns of Ag-1-X-8 (200 to 400 mesh, formate form). After elution with formic acid the eluates were lyoph­ilized. The cyclic adenosine monophosphate content was determined according to Gilman.5

Determination of phosphodiesterase activity. The activity of phosphodiesterase was determined by the method described by Thompson and Appleman6 with minor modifications. After 30,000 x g centrifugation of the myometrial homogenate for 10 minutes, the su­pernatant was used for analysis. The assay was per­formed in a solution containing Tris hydrochloride, 40 mmollL, pH 7.5; magnesium chloride, 5 mmollL; 20 fJ..g of 5-nucleotidase (snake venom), and tritiated cyclic adenosine monophosphate, 1 fJ..moIlL, containing more than 100,000 cpm, in a total volume of 0.1 ml. After 3 to 5 minutes at 37° C the reaction was terminated by heating the reaction mixture at 90° C for 1 minute, and 0.4 ml of distilled water was then added to the incu­bation tube. The entire contents of the tubes were transferred to prewashed AG 1 x 8 columns (C 1 form, 0.5 by 1.0 cm) followed by addition of 1 ml of distilled water for sampling of fractions containing the tritiated adenosine. The activity was measured in a liquid scin­tillation counter. Calcium-dependent phosphodiester­ase activity was measured in six strips. Ionized calcium was added in increasing concentrations to an enzyme preparation dialyzed in Tris buffer containing ethyl­eneglycol-bis(l3-amino-ethyl ether)n,n'tetraacetic acid (EGTA), and phosphodiesterase activity was measured

as described above. In 12 myometrial preparations tri­fluoperazine was added in increasing concentrations (0.1 fJ..mollL to 1 mmol/L of trifluoperazine) (see Fig.

5, Aand B).

Volume 154 Number 3

~ .. .2

J 'Ii

I .. .!

101 [nond"RlIiIll] M

Fig.!. Tests of antagonists on the contractile effects of nor­epinephrine (noradrenaline) in human myometrial prepara­tions pretreated with propranolol (10- 6 moI/L). In the tests of the antagonists, the myometrium was pre incubated with pra­zosin (10- 6 mollL) (0), yohimbine (10- 6 mollL) (+), or phen­tolamine (10- 6 mollL) (0) for 10 minutes before the norepi­nephrine tests were performed. Values are means ± SEM.

Pharmacologic agents. Tritiated dihydro-a-ergo­cryptine (50 Ci/mmol) (New England Nuclear, Dreieich, West Germany), L-norepinephrine bitartrate (Sigma Chemical Co., St. Louis, Missouri), phentol­amine methansulfon (Ciba Geigy AG, Basel, Switzer­land), prazosin hydrochloride (Pfizer, Inc., New York, New York), DL-propranolol hydrochloride (ICI, Mac­clesfield, Cheshire, England), trifluoperazine dihydro­chloride (Sigma Chemical Co.), yohimbine hydrochlo­ride (Sigma Chemical Co.).

Statistical methods. All values are given as mean val­ues ± SEM. Levels of significance were tested by Stu­dent's t test with the use of paired and unpaired data. The study was approved by the Ethical Committee, Link6ping University.

Results Contractility. Norepinephrine in a concentration of

10- 7 to 10-4 mol/L increased the tension in a dose­dependent manner in spontaneously contracting myo­metrial strips from the fundus uteri as well as from the lower uterine segment (Fig. 1). The ECoo of the con­tractile effect of norepinephrine was 5 X 10- 7 mollL. The dose-effect curve was markedly shifted to the right when the strips were pretreated with phentolamine, Ij.LmollL (Fig. 1). The ECoo value in the presence of phentolamine was about 10- 4 mollL.

When the myometrial strips were pretreated with either prazosin, 1 j.Lmol/L, or yohimbine, 1 j.LmollL, be­fore norepinephrine stimulation, a shift to the right of the dose-response curve was registered in both groups. The EC50 (the concentration that increases the tension

Myometrial a-adrenergic receptors during pregnancy 603

300 o isthmus

- • fundus .. .. . ~ .. ,. f! 200 0 .. ... ... .. = E ell ...... I 'i 100

'Z" E .!:!..- -....

2 • 10 14 18

Fig. 2. Number of a-adrenergic receptor binding sites in the lower uterine segment (isthmus) and uterine fundus of un­treated pregnant women. Values are means ± SEM.

50%) values were 1.7 X 10- 5 and 2.5 x 10-' mollL in the presence of prazosin and yohimbine, respec­tively, indicating the presence of physiologically active 0.,- and a,-adrenergic receptors in the myometrium of pregnancy.

Tritiated dihydroergocryptine binding. Binding of tritiated dihydroergocryptine to human myometrial membrane associated with pregnancy is shown in Fig. 2, where data represent specific binding (total minus nonspecific binding). Nonspecific binding was mea­sured in the presence of phentolamine, 10 j.LmollL. The total numbers of specific binding sites were about 195 and 210 fmol/mg of protein in the lower uterine seg­ment and uterine fundus, respectively (Fig. 2). The affinity of the receptors was comparable in the two locations as the equilibrium dissociation constants (K,,) of tritiated dihydroergocryptine were almost equal, 6 and 6.2 nmollL, respectively. a-Adrenergic receptor antagonists potently competed for the binding sites. The competitive antagonist phentolamine competed with a K" of 28 nmol/L (Fig. 3, A). Competition ex­periments with the ai-adrenergic receptor antagonist prazosin and the a,-adrenergic receptor antagonist yohimbine (Fig. 3, B) gave clear support for the notion that tritiated dihydroergocryptine bound to two sites. The estimated Kd value was 0.5 nmollL for the high­affinity site of prazosin and 1 nmol/L for yohimbine. The proportions of 0.,- and a 2-adrenergic receptor sites were 60% and 40%, respectively, of the total tritiated dihydroergocryptine binding capacity.

Cyclic adenosine monophosphate. In both un­treated and propranolol-pretreated myometrial biopsy specimens, cyclic adenosine monophosphate produc­tion was significantly reduced when stimulated with norepinephrine, 1 j.LmollL (Fig. 4, A and B). Prazosin and yohimbine partly inhibited the norepinephrine­induced reduction of the cyclic adenosine mono-

604 Berg, Andersson, and Ryden

III 100 ..., .. .. .. • 10 • .. ...

z co ,

20 '"Z .!!......

2

-I -& -3 -4

I.. [ •• " ••• il'] M

March, 1986 Am J Obstet Gynecol

Fig. 3. Inhibition of tritiated dihydroergocryptine ([3 HJ-DHE) binding to a-adrenergic receptors by a-adrenergic receptor antagonists. A: Phentolamine. B: Prazosin and yohimbine.

32 [A] 00 .. ~ 24 IL -E

...... 16 .. E IL

a. 8 ::Ii c u

NA - + + + - + + +

PHAZ + - - - + -YO - + +

Fig. 4. Effects of norepinephrine (NA) (10- 6 mollL) on the cyclic adenosine monophosphate (cAMP) levels of human myometrium in the absence or presence of prazosin (PRAZ) (10- 6 mollL) or yoh­imbine (YO) (10- 6 moIlL). A: Preincubation in normal Krebs buffer. B: Preincubation in Krebs buffer containing propranolol, 10- 6 mollL. Values are means ± SEM. The significance of the changes from control values are indicated by * = P < 0.05 and ** = p < 0.01.

phosphate content in the propranolol-pretreated prep­arations (Fig. 4). In the absence of propranolol, yohimbine almost completely reversed the norepi­nephrine response while prazosin had no significant effect (Fig. 4, A).

Phosphodiesterase. The lowering effect on cyclic adenosine monophosphate seems to be mainly due to an activation of a 2-adrenergic receptors, while a,-ad­renergic receptors were involved to a minor extent. Changes in ionized calcium are suggested to play an important role in the a,-adrenergic receptor response.8

Calcium ions also influence the cyclic adenosine mono­phosphate-phosphodiesterase activity, at least in the brain.9 We therefore studied the influence of ionized calcium ions on the phosphodiesterase activity obtained from human myometrium and found that increasing concentrations of calcium (10- 6 to 10- 3 mollL) caused a slight but significant increase in phosphodiesterase activity (Fig. 5, A). When the calmodulin-inhibitory drug trifluoperazine was tested, a remarkable inhibition of the phosphodiesterase activity was noticed in the concentration interval 10- 6 to 10- 3 moUL (Fig. 5, B).

Comment

In the present study we have investigated a-adren­ergic receptors of the human myometrium in preg­nancy with two conventional in vitro systems, tension recordings and receptor-binding assays.

In the tension studies, both a,- and a 2-adrenergic receptors seemed to be involved in the smooth muscle contraction mechanism. The a,- and a 2-adrenergic re­ceptor blocking agent phentolamine was approximately doubly effective in preventing increased tension after norepinephrine stimulation when compared with ei­ther the a,-blocker prazosin or the a 2-blocker yohim­bine (Fig. 1).

In the receptor-binding assays we chose the radioli­gand tritiated dihydroergocryptine and discriminated a,- and a 2-adrenergic receptors by studying the dis­placement of tritiated dihydroergocryptine by selective a-antagonists as described by Hoffman et al.lO Our re­sults indicate that both a,- and a 2-adrenergic receptors are present in myometrium from pregnant women in the proportion 60: 40 and that the distribution of a­

adrenergic receptors is comparable in the lower uterine

Volume 154 Number 3

segment and uterine fundus (Fig. 2). Wikland et al." recently demonstrated a marked reduction in fluores­cent nerve fibers in myometrium at term compared with myometrium from nonpregnant women. They concluded that neuronal factors of maternal origin are unlikely to be involved in the mvometrial contractile response in pregnant uteri at term. The a l - and a 2 -

adrenergic receptors demonstrated in our study are probably localized on the myometrial cells, since very few nerve fibers were present in term myometrium. Both a-adrenergic receptor subtypes seem to be of im­portance for eliciting the excitatory muscle response. The mechanisms believed to mediate the responses of the two receptors seem, however, to differ. Previous studies on a 2-adrenergic receptors in platelets, for in­stance, have shown that stimulation of a 2-adrenergic receptors lowers the intracellular cyclic adenosine monophosphate, probably through an inhibition of ad­enylate cyclase." The same mechanism seems to be ap­plicable in human myometrium in pregnancy (Fig. 4, A and B). However, an additional effect on the cyclic adenosine monophosphate metabolism was seen after stimulation of ai-adrenergic receptors, especially when the myometrial biopsy specimens were pretreated with a i3-blocker, propranolol. One explanation of this could be that ai-adrenergic receptor stimulation alters the intracellular ionized calcium content, as found in liver cells by Exton" and in fatty cells by Garcia Sainz,12 and thus secondarily reduces cyclic adenosine monophos­phate. In the present study we were able to show that an increase in ionized calcium caused a small but sig­nificant increase in cyclic adenosine monophosphate­dependent phosphodiesterase, which might in turn re­duce the cyclic adenosine monophosphate content. The calmodulin antagonist trifluoperazine reduced phos­phodiesterase activity in a dose-dependent manner (Fig. 5, A and B), indicating the presence of a calcium­calmodulin-dependent phosphodiesterase. This has previously been demonstrated in the brain."

Recent publications concerning preterm labor and 13-sympathomimetic treatment have suggested a role of endogenous a-adrenergic stimuhuioll ill explaining the often limited effect of i3-sympathollllllletics. In a study on human myometrium in pregnancy by Ke et al.,l:l 13-sympathomimetics were administered intermittently to the in vitro preparations and in this way the investi­gators could avoid down-regulation of i3-adrenergic re­ceptors. They also postulated that in clinical practice patients treated with i3-sympathomimetics showed a higher endogenous catecholamine level, which might stimulate the myometrium to contractions, at least when the i3-adrenergic receptor system is down-regu­lated, as has been demonstrated after terbutaline treatment. I

The clinical importance of the a-adrenergic receptor

Myometrial a-adrenergic receptors during pregnancy 605

60

.. 40

~

:i -u • 20 I ... Q Go

-6 -5 -4 -3 -2

Fig. 5. A: The effect of calcium on human myometrial phos­phodiesterase aClivit\ The enzyme preparation was dialyzed against Tris buller containing I mmollL of EGTA before the assay. Values are means ± SEM. B: The effect of dif­ferent concentrations of trifluoperazine on cyclic adenosine monophosphate-phosphodiesterase (PDE) activity. Values are means ± SEM.

system cannot be stated before further studies on the number of a-adrenergic receptors in preterm and term pregnanCies with and without labor have been per­formed.

The excellent assistance of Mrs. Helle Noorlind­Brage is greatly appreciated.

REFERENCES

I. Berg G, Andersson RGG, Ryden G. I3-Adrenergic recep­tors in human myometrium during pregnancy: Changes in the number of receptors after l3-mimetic treatment. AM J OBSTET GYNECOL 1985;151:392.

2. Bottari SP, Vokaer A, Kaivez E, Lescrainier JP, Vauquelin G. Identification and characterization of a 2-adrenergic re­ceptors in human myometrium by ['H]rauwolscine bind­ing. AM J OBSTET GYNECOL 1983; 146:639.

Berg, Andersson, and Ryden

3. Huszar G, Roberts JM. Biochemistry and pharmacology of the myometrium and labor: Regulation at the cellular and molecular levels. AM J OBSTET GV"ECOI. I 9S2; 142:225.

4. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Bioi Chern 1951; 193:265.

5. Gilman AG. A protein binding assay for adenosine 3'5' cyclic monophosphate. Proc Nat! Acad Sci USA 1970; 67:305.

6. Thompson WJ, Appleman MM. Multiple cyclic nucleotide phosphodiesterase activities from rat brain. Biochemistry 1971;10:311.

7. Williams LT, Mullikin D, Lefkowitz RJ. Identiffication of alpha-adrenergic receptors in uterine smooth muscle membranes by "H-dihydroergocryptine binding. J Bioi Chern 1976;251:6915.

S. Exton J. Molecular mechanism involved in alpha-adren­ergic responses. Mol Cell Endocrinol I 9S1 ;23:233.

March, 1986 Am J Obstet Gynecol

9. Kakiuchi S, Yamazaki R. Calcium dependent phospho­diesterase activity and its activating factor (PAS) from brain. Biochem Biophys Res Commun 1970;41: II 04.

10. Hoffman BB, DeLean A, Wood CL, Schocken DD, Lef­kowitz RJ. Alpha adrenergic receptor sUbtypes: quanti­tative assessment by ligand binding. Life Sci 1979; 24: 1739.

II. Wikland M, Lindblom B, Dahlstrom A, Haglid KG. Struc­tural and functional evidence for the denervation of hu­man myometrium during pregnancy. Obstet Gynecol 19S4;64:503.

12. Garcia-SainzJA, Hoffman BB, Li S-Y, Lefkowitz RJ, Fain JF. Role of alpha-I adrenoceptors in the turnover of phos­phatidylinositol and of alpha-2 adrenoceptors in the reg­ulation of cyclic AMP accumulation in hamster adipocytes. Life Sci 19S0;27:953.

13. Ke R, Vohra M, Caspar R. Prolonged inhibition of human myometrial contractility by intermittent isoproterenol. AM J OBSTET GVNECOL 19S4;149:S41.

Corticotropin-releasing factor can stimulate gonadotropin

secretion by human fetal pituitaries in superfusion

Zeev Blumenfeld,* M.D., Robert W. Kuhn, Ph.D., and Robert B. Jaffe, M.D.

San Francisco, California

In previous studies, corticotropin-releasing factor was found to elicit a rise in circulating adrenocorticotropic

hormone in human subjects and laboratory animals, but no stimulatory effect of corticotropin-releasing factor on other pituitary hormones was detected. Since stress may be associated with luteinizing hormone

changes as well as with those of corticotropin-releasing factor and adrenocorticotropic hormone, we quantified gonadotropin responses to corticotropin-releasing factor and arginine vasopressin in 11 human fetal pituitaries with use of both superfusions and static incubations. Exposure to corticotropin-releasing factor brought about a significant increase in adrenocorticotropic hormone and gonadotropin concentrations in the effluent media by means of the superfusion system. Similar concentrations of corticotropin-releasing factor significantly increased adrenocorticotropic hormone secretion into the medium by dispersed fetal pituitary cells cultured on an extracellular matrix but failed to increase luteinizing hormone and follicle-stimulating hormone secretion. Exposure to 3 mmol/L 8-bromo-cyclic adenosine

monophosphate caused an increase in all three peptides, both in superfusion and static incubations. Dose­response studies showed that the effect on gonadotropin secretion occurred at concentrations of 8-bromo­cyclic adenosine monophosphate two orders of magnitude lower than those affecting adrenocorticotropic hormone secretion. The purity of corticotropin-releasing factor and arginine vasopressin used in these studies was confirmed by high-performance liquid chromatography. These in vitro results are consistent with a paracrine effect of corticotropes acting on gonadotropes. The combination of static incubation and superfusion has proved useful in elucidating the effects of different secretagogues on pituitary cells. (AM J OSSTET GVNECOL 1986;154:606-12.)

Key words: Fetal pituitary, gonadotropin

From the Reproductive Endocrinology Center, Department of Ob­stetrics, Gynecology and Reproductive Sciences, University of Cal­ifornia, San Francisco.

Supported in part by National Institutes of Health Grants HD 08478 and HD 11979.

Received for publication june 9,1985; revised November 7,1985; accepted November 25, 1985.

Reprint requests: Robert B. jaffe, M.D., Department of Obstetrics, Gynecology and Reproductive Sciences, Room M -1490, University of California, San Francisco, CA 94143.

*Recipient of American Physician Fellowship.

606

Shortly after the elucidation of the structure of corticotropin-releasing factor by Vale et aI., I there were several studies in which its effects on the secretion of adrenocorticotropic hormone, growth hormone, thy­roid-stimulating hormone, luteinizing hormone, and prolactin in laboratory animals2

." and in normal hu­mans4

,5 were assessed. Whereas corticotropin-releasing factor was found to elicit a rise in circulating adreno-