1202

Contraction of the uterine muscle is pivotal in labor,menstruation, dysmenorrhea, and conception.1-5 Thebasic mechanisms underlying the control of myometrialcontractility are not well understood. Peptide hormones,such as oxytocin and vasopressin, are recognized as pro-moters of uterine contractions, but there is insufficientevidence to implicate these hormones as the primary ini-tiators of human parturition.6-8 Thus, a number of otherendocrine and paracrine factors are believed to haveroles in the control of uterine contractility and quies-cence. Among these, prostaglandins are believed to be of

key importance in the regulation of myometrial contrac-tions and cervical ripening.4,5 Prostaglandin E2α andprostaglandin F2α are the eicosanoid compounds mostoften studied in connection with human labor. Both areused routinely in clinical practice to induce cervicalripening and labor. Prostaglandin E2α production hasbeen demonstrated in fetal membranes before the onsetof labor.4 The exact molecular mechanisms responsiblefor the activity of prostaglandins in both the cervix andthe myometrium are incompletely understood but arebelieved to be linked to the activation of specific cellmembrane–bound receptors. The cellular effects ofprostaglandin E2α are mediated through several types oftransmembrane receptors. These receptors have beenclassified into 4 major subtypes as EP1, EP2, EP3, andEP4.4 Among these proteins, the receptor subtypes EP2and EP4 are believed to have a relaxant effect on the uter-ine muscle.9 In contrast, EP1 and EP3 subtypes have beenshown to increase the contractility of the uterus.9

Alternate splicing is especially prevalent among thefamily of prostanoid receptors, both in humans and ani-mals.10 Among human prostaglandin E2α receptor sub-

From the Departments of Obstetrics and Gynecologya and Pathology,bKeck School of Medicine, University of Southern California, Los Angeles.Supported by Departments of Obstetrics and Gynecology and Pathology,Keck School of Medicine, University of Southern California, Los Angeles.Frank Lynch Memorial Prize Essay, presented at the 68th Annual Meet-ing of the Pacific Coast Obstetrical and Gynecological Society, Asland,Ore, October 3-7, 2001.Reprint requests: Juan C. Felix, MD, LAC + USC Medical Center, 1240N Mission Rd, Room 1L23, Los Angeles, CA 90033.© 2002, Mosby, Inc. All rights reserved.0002-9378/2002 $35.00 + 0 6/6/123746doi:10.1067/mob.2002.123746

The expression of EP3-6 and inducible nitric oxide synthasemessenger RNA are correlated in pregnant and misoprostol-treated but not in nongravid or menopausal myometrium

Nima Goharkhay, MD,a Deborah A. Wing, MD,a Vivien Pan, MD,a Vance McCausland, MD,a

Mary Hanna, MD,b Yathi M. Naidu, PhD,b and Juan C. Felix, MDa,b

Los Angeles, Calif

OBJECTIVE: Our purpose was to investigate the expression level of inducible nitric oxide synthase (iNOS) innongravid and gravid human myometrium and assess its relation to the expression of EP3 prostaglandin re-ceptor isoforms.STUDY DESIGN: Myometrial tissue from a cohort of gravid, nongravid, and menopausal subjects and fromnongravid subjects exposed to misoprostol was obtained and analyzed for iNOS messenger RNA (mRNA)expression and compared with previously determined mRNA levels for the EP3-6 prostaglandin receptorfrom the same samples by means of semiquantitative polymerase chain reaction.RESULTS: Significantly higher levels of iNOS mRNA expression were found in the gravid compared with pre-menopausal nongravid (P < .02) and menopausal (P < .05,) samples. Linear regression analysis of iNOSversus EP3-6 expression showed a positive correlation between all studied samples (n = 47, P < .0001, r = 0.815). Among individual groups, a significant linear correlation was found only in pregnant (n = 10, P < .0001, r = 0.785) and misoprostol-exposed (n = 7, P = .0256, r = 0.815) subjects. No correlation betweeniNOS and EP3-6 mRNA levels was noted in the nongravid premenopausal groups (n = 10, P = .205, r = 0.350), although a nonsignificant trend was found for the menopausal group (n = 10, P = .0535, r =0.624). When the data from pregnant patients were stratified, both laboring and nonlaboring women dis-played the observed correlation (n = 6, P < .0001, r = 0.993; and n = 13, P = .008, r = 0.697, respectively).CONCLUSION: iNOS and EP3-6 expression are strongly correlated in gravid and misoprostol-treated, non-gravid myometrium. (Am J Obstet Gynecol 2002;186:1202-6.)

Key words: EP3 receptor isoforms, messenger RNA expression, human myometrium inducible ni-tric synthase, prostaglandin E2

Volume 186, Number 6 Goharkhay et al 1203Am J Obstet Gynecol

types, only EP3 has been shown to undergo alternatesplicing, and 9 different isoforms of EP3 have been de-scribed.10-14 The complexity involved in the expression ofEP3 as compared with other prostaglandin E2α receptors,together with its described role as stimulator of uterinecontractions and cervical ripening, suggested to us a potential role for the EP3 receptor isoforms in the regu-lation of myometrial contractility. Recently, we character-ized the messenger RNA expression levels of variousprostaglandin EP3 receptor isoforms, a family of puta-tively contractile prostaglandin receptors, in myome-trium from women in various physiologic states usingsemiquantitative reverse transcriptase–polymerase chainreaction (PCR) and found specific down-regulation ofEP3-2 and up-regulation of EP3-6 in late gestation.15

Nitric oxide (NO) has been shown to promote uterinerelaxation and induce cervical ripening16; however, thereis no consensus on the abundance and the mechanismsof action of NO and inducible nitric oxide synthase(iNOS) in cervical and uterine tissues.17-23 Several studies have demonstrated interactions between theprostaglandin and NO systems in various aspects of uter-ine function, including myometrial contractility, in animals and humans.24-28 The study of reciprocal associa-tions between iNOS and individual EP3 receptor iso-forms may point to specific pathways involved. In thisinvestigation we sought to evaluate the relative expres-sion levels of iNOS in our various study groups and to as-sess the possible relationships between EP3 receptorisoform and iNOS mRNA expression in myometrium invarious physiologic states. We hypothesized that iNOS ex-pression would be increased in gravid myometrium be-fore the onset of labor.

Methods

Study population. Myometrial tissue was obtained from47 women at the time of cesarean delivery or hysterec-tomy performed for a nonmalignant condition. Of thesewomen, 10 were menopausal, 10 were premenopausaland nongravid, 19 were gravid, and 7 were pre-menopausal women who had received intravaginal treat-ment with misoprostol. The Investigational Review Boardof the Los Angeles County–University of Southern Cali-

fornia Medical Center approved the protocols for tissueharvesting. All pregnant women in this study were in thethird trimester of a pregnancy (9 singleton gestations and1 twin gestation) for which cesarean delivery wasplanned. Six gravid women were determined to be inlabor (3 preterm and 3 term, defined as 37 completedweeks of gestation after last menstrual period) and 13were nonlaboring (3 of who were preterm). In all women,tissue was obtained from the anterior lower uterine seg-ment. Treatment with misoprostol consisted of a 200-µgvaginal insert positioned in the posterior fornix at least 5hours before hysterectomy. All specimens were immedi-ately stored at –80°C until RNA extraction was per-formed.

Semiquantitative reverse transcriptase–polymerasechain reaction. Tissue samples collected after surgerywere immediately transported to our laboratory andstored at –80°C within 30 minutes of collection. Total-cellRNA extraction was performed as a single-step acidguanidinium thiocyanate-phenol-chloroform extraction(Molecular Research Center, Cincinnati, Ohio). Theamount of total RNA was quantitated spectrophotometri-cally by measuring the absorption at 260 nm. An equalamount of RNA from each tissue sample was subjected toreverse transcription for complementary DNA synthesiswith use of random hexamer primers and recombinantreverse-transcriptase enzyme (Life Technologies, Rock-ville, Md). Product yield curves were derived with a sub-population of our samples for the analyzed isoforms andfor β-actin to ensure amplification of the PCR productswithin the exponential range of the PCR. We used theiNOS mRNA primers published by Kitagawa et al.29 β-Actin was used as the reference for the semiquantitativeanalysis, with the assumption that expression rates wereequal among the different samples.30 The primers usedfor amplification of EP3-6 were designed to yield a 169–bp amplimer with a 5´ primer sequenceATCTTCAATCAGACATCAGTTGAGC and 3´ primer se-quence CCATCTCCTGGCAAAACTTTC. For the PCRexperiments, we used the Advantage 2 Polymerase en-zyme (Clontech, Palo Alto, Calif) and an annealing tem-perature of 64.5°C for both iNOS and β-actin. Knownvolumes of the PCR products for each reaction set were

Table. Patient characteristics

Misoprostol Premenopausal nongravid Menopausal Gravid Total (n = 7) (n = 10) (n + 10) (n = 17) (n = 46)

Age (y) 42 (36-48) 43.1 (39-47) 57.2 (50-65) 29.7 (20-40) 40.5 (20-65)Ethnicity

Hispanic 5 (71%) 9 (90%) 9 (90%) 8 (80%) 40 (87%)Other 2 (29%) 1 (10%) 1 (10%) 2 (10%) 6 (13%)

Gravidity 3.5 (1-6) 4.2 (2-7) 5.6 (3-9) 3.5 (1-7) 3.5 (1-6)Parity 3 (1-5) 3.4 (2-6) 4.2 (1-9) 1.7 (0-5) 2.8 (0-9)Abortus 0.4 (0-1) 0.8 (0-3) 1.4 (0-4) 0.8 (0-5) 0.9 (0-5)

1204 Goharkhay et al June 2002Am J Obstet Gynecol

electrophoresed in agarose gels, stained with ethidiumbromide, and photographed with a digital camera. Theoptical density for each sample was measured by usingcomputerized image analysis software (Eastman Kodak,Rochester, NY). The relative abundance of the iNOS mes-sage was calculated as the ratio of iNOS PCR productamount divided by the amount of the β-actin PCR prod-uct for each sample.

Statistical analysis. The expression levels of measuredparameters are expressed as means ± SEM. Statisticalanalyses were performed by using analysis of variance, theFisher least significant difference test, and linear regres-sion analysis as appropriate. All tests were 2 sided, and a Pvalue of .05 was considered statistically significant.

Results

The demographics of the study population are pre-sented in the Table. We demonstrated the presence ofiNOS mRNA in all analyzed samples. The highest expres-sion levels for iNOS were observed in tissue from gravidwomen, followed by tissue from nongravid misoprostol-treated women, menopausal women, and premenopausalnongravid women (Fig 1). Overall, the difference amongthe 4 study groups was statistically significant (P < .05,analysis of variance). Specifically, a significantly higherlevel of iNOS mRNA expression was found in the myo-metrium of gravid women when compared with the sam-ples obtained from premenopausal nongravid women (P < .02) and menopausal women (P < .05, Fisher leastsignificant difference test). The highest variation in iNOSlevels was found in the samples from pregnant women,followed by samples from misoprostol-treated women.

The result of the linear regression analysis for the cor-relation between iNOS expression and EP3-6 expressionis shown in Fig 2. There is a significant correlation be-tween iNOS and EP3-6 expression levels in all groupscombined (n = 47, P < .0001, r = 0.815). Analysis of this re-lationship individually for each study group reveals that astatistically significant correlation was found only in thegravid (n = 10, P < .0001, r = 0.785) and misoprostol-ex-

posed (n = 7, P = .0256, r = 0.815) subjects. No correlationwas found between iNOS and EP3-6 mRNA levels for ei-ther the nongravid premenopausal group (n = 10, P = .205, r = 0.350) or the menopausal group (n = 10, P = .535, r = 0624), whereas a nonsignificant trend waspresent for the menopausal group (n = 10, P = .0535, r = 0.624). When the data from pregnant patients werefurther stratified, both laboring and nonlaboring womendisplayed the observed correlation (n = 6, P < .0001, r = 0.993; and n = 13, P = .008, r = 0.697, respectively), al-though the correlation was stronger for the laboringgroup than for the nonlaboring group (95% CI: 1.269-1.718 and 0.170-0.903, respectively).

Comment

We have demonstrated the presence of iNOS mRNA inthe myometrium of all women studied. We have alsoshown variations in iNOS expression between women indifferent physiologic states. Our findings support previ-ous studies demonstrating increased levels of iNOS dur-ing late gestation.17,31-33

We have previously analyzed the expression of variousEP3 prostaglandin receptor isoforms in the same cohortof patients.15 We found elevated expression of the EP3-6receptor isoform in gravid myometrium. EP3-2 receptorisoform expression was significantly reduced, whereasEP3-3 was unaltered compared with nongravid samples.We have postulated that this variation in receptor isoformexpression during late gestation is involved in the regula-tion of myometrial contractility in labor.

Previous studies evaluating the effect of NO on myo-metrium point to its relaxant effect.17,18,22 For this rea-son, we hypothesized that we would observe a decrease inthe levels of iNOS during labor when myometrium is ac-tively contracting. However, our observations show the

Fig 1. Relative messenger RNA expression levels (means ± SEM)for inducible nitric oxide synthase (iNOS) by study groups. aSig-nificant differences between gravid and menopausal groups (P <.01). bSignificant difference between gravid and premenopausalnongravid groups (P < .02).

Fig 2. Correlation of inducible nitric oxide synthase (iNOS) mes-senger RNA expression with EP3-6 mRNA expression by studygroups.

Volume 186, Number 6 Goharkhay et al 1205Am J Obstet Gynecol

highest levels of expression in near-term myometrium,particularly in myometrium that has been in latent or ac-tive labor. Three possible explanations exist for these ob-servations. First, it is possible that NO does not play asignificant role in myometrial contractility duringlabor.18,21,23 Alternatively, NO could exert an influenceon myometrial contractility by regulation of uterine vas-cular tone.34 Finally, NO may act as a relaxant modulatorthat opposes contractile effectors, such as prostaglandins.In this instance, as the effect of contractile modulatorsincreases, NO would increase to neutralize their effect.Eventually, the contractile effectors would overcome therelaxant effect of NO, resulting in the myometrial con-tractions and labor. The latter 2 hypotheses would ac-count for the correlation between the expression of iNOSand EP3-6 seen in our study cohort.

The myometrium of misoprostol-treated, nongravidwomen demonstrated a significant correlation betweenmRNA expression levels of iNOS and EP3-6. When thisgroup is compared with the untreated nonpregnant pre-menopausal women, little differences in demographiccharacteristics are noted (Table). These findings suggestthat the increased expression of iNOS in this group maybe due to the interaction of misoprostol, a prostaglandinE1 analog and a known EP3/EP2 agonist, with prosta-glandin receptors, specifically with the EP3-6 receptorisoform.

Although it did not reach statistical significance, atrend was present between iNOS and EP3-6 expression inmenopausal subjects. Whether this relationship is signifi-cant remains to be determined in a larger cohort. If thisrelationship is confirmed, increased levels of iNOS inmenopausal myometrium would again be associated witha state of myometrial quiescence.

We believe that our findings showing an interaction be-tween the EP3-6 receptor isoform and iNOS point to an association between the actions of NO and prosta-glandins on human myometrium near the time of partu-rition. This association strongly suggests an associationbetween these two proteins in the regulation of myome-trial contractions during labor. At this point, we can onlyspeculate as to the mechanisms of this interaction. If ourhypothesis is correct, it is likely that the up-regulation ofiNOS and the concomitant increase in NO levels arecaused by downstream effector molecules, resulting fromthe contractile effects of EP3 receptor stimulation. Thiswould signify the presence of a potential feedback mech-anism between the contractile effect of prostaglandinsand the relaxant effect of NO. Alternatively, it is possiblethat a common inducer of both iNOS and EP3-6 may bethe reason for their concurrent increase in expression.Although this is theoretically possible, a common inducerwould strongly indicate a common effect of the involvedeffectors, that is, a contractile or relaxant effect of bothNO and EP3 receptor stimulation. Prevailing data at this

point in time strongly suggest that this is not the case.Our data clearly demonstrate a modulation of iNOS inmyometrium during various physiologic states. This mod-ulation is strongly correlated with the presence of an EP3receptor isoform. Although the mechanism of action iscurrently not understood, prostaglandins are likely to beinvolved, as suggested by the presence of increased iNOSexpression in our samples of misoprostol-treated non-gravid myometrium. We recognize the need for furtherstudies at the protein and second messenger levels to con-firm or refute this hypothesis.

REFERENCES

1. Olson DM, Mijovic JE, Sadowsky DW. Control of human parturi-tion. Semin Perinatol 1995;19:52-63.

2. Cibils LA. Contractility of the nonpregnant human uterus. Ob-stet Gynecol 1967;30:441-61.

3. Akerlund M. Pathophysiology of dysmenorrhea. Acta Obstet Gy-necol Scand Suppl 1979;87:27-32.

4. Maggi M, Baldi E, Susini T. Hormonal and local regulation ofuterine activity during parturition, II: the prostaglandin andadrenergic systems. J Endocrinol Invest 1994;17:757-70.

5. Goharkhay N, Stanczyk FZ, Gentzschein E, Wing DA. Plasmaprostaglandin E2 metabolite levels during labor induction with asustained-release prostaglandin E(2) vaginal insert. J Soc Gy-necol Invest 2000;7:338-42.

6. Zeeman GG, Khan-Dawood FS, Dawood MY. Oxytocin and its re-ceptor in pregnancy and parturition: current concepts and clin-ical implications. Obstet Gynecol 1997;89:873-83.

7. Russell JA, Leng G. Sex, parturition and motherhood withoutoxytocin? J Endocrinol 1998;157:343-59.

8. Goharkhay N, Felix JC, Naidu Y, Hanna M, Kovacs BW, Wing DA.Expression of the oxytocin receptor (OTR) and the vasopressinV1aR receptor in human myometrium. J Soc Gynecol Invest2001;8(Suppl):126-7A.

9. Brodt-Eppley J, Myatt L. Prostaglandin receptors in lower seg-ment myometrium during gestation and labor. Obstet Gynecol1999;93:89-93.

10. Kotani M, Tanaka I, Ogawa Y, Usui T, Tamara N, Mori K, et al.Structural organization of the human prostaglandin EP3 recep-tor subtype gene (PTGER3). Genomics 1997;40:425-34.

11. Kotani M, Tanaka I, Ogawa Y, Usui T, Mori K, Ichikawa A, et al.Molecular cloning and expression of multiple isoforms of humanprostaglandin E receptor EP3 subtype generated by alternativemessenger RNA splicing: multiple second messenger systems andtissue-specific distributions. Mol Pharmacol 1995;48:869-79.

12. Regan JW, Bailey TJ, Donello JE, Pierce KL, Pepperl DJ, ZhangD, et al. Molecular cloning and expression of human EP3 recep-tors: evidence of three variants with differing carboxyl termini.Br J Pharmacol 1994;112:377-85.

13. Adam M, Boie Y, Rushmore TH, Muller G, Bastien L, Mckee KT,et al. Cloning and expression of three isoforms of the humanEP3 prostanoid receptor. FEBS Lett 1994;338:170-4.

14. Schmid A, Thierauch KH, Schleuning WD, Dinter H. Splice vari-ants of the human EP3 receptor for prostaglandin E2. Eur JBiochem 1995;228:23-30.

15. Goharkhay N, Felix JC, Naidu Y, Hanna M, Wing DA. The ex-pression of the EP3-2 prostaglandin receptor is selectively re-duced in gravid human myometrium [abstract]. J Soc GynecolInvest 2001;8(Suppl):126A.

16. Chwalisz K, Garfield RE. Role of nitric oxide in the uterus andcervix: implications for the management of labor. J Perinat Med1998;26:448-57.

17. Bansal RK, Goldsmith PC, He Y, Zaloudek CJ, Ecker JL, Riemer RK.A decline in myometrial nitric oxide synthase expression is asso-ciated with labor and delivery. J Clin Invest 1997;99:2502-8.

18. Bartlett SR, Bennett PR, Campa JS, Dennes WJ, Slater DM, Mann GE,et al. Expression of nitric oxide synthase isoforms in pregnanthuman myometrium. J Physiol 1999;521 Pt 3:705-16.

1206 Goharkhay et al June 2002Am J Obstet Gynecol

19. Yallampalli C, Dong YL, Gangula PR, Fang L. Role and regula-tion of nitric oxide in the uterus during pregnancy and parturi-tion. J Soc Gynecol Invest 1998;5:58-67.

20. Ledingham MA, Thomson AJ, Young A, Macara LM, Greer IA,Norman JE. Changes in the expression of nitric oxide synthasein the human uterine cervix during pregnancy and parturition.Mol Hum Reprod 2000;6:1041-8.

21. Barber A, Robson SC, Lyall F. Hemoxygenase and nitric oxidesynthase do not maintain human uterine quiescence duringpregnancy. Am J Pathol 1999;155:831-40.

22. Buhimschi I, Yallampalli C, Dong YL, Garfield RE. Involvementof a nitric oxide-cyclic guanosine monophosphate pathway incontrol of human uterine contractility during pregnancy. Am JObstet Gynecol 1995;172:1577-84.

23. Dennes WJ, Slater DM, Poston L, Bennett PR. Myometrial nitricoxide synthase messenger ribonucleic acid expression does notchange throughout gestation or with the onset of labor. Am JObstet Gynecol 1999;180:387-92.

24. Ledingham MA, Denison FC, Kelly RW, Young A, Norman JE.Nitric oxide donors stimulate prostaglandin F(2alpha) and in-hibit thromboxane B(2) production in the human cervix duringthe first trimester of pregnancy. Mol Hum Reprod 1999;5:973-82.

25. Dong YL, Yallampalli C. Interaction between nitric oxide andprostaglandin E2 pathways in pregnant rat uteri. Am J Physiol1996;270:E471-6.

26. Chaud MA, Franchi AM, Beron de Astrada M, Gimeno MF. Roleof nitric oxide on oxytocin-evoked contractions and prosta-glandin synthesis in isolated pregnant rat uterus. ProstaglandinsLeukot Essent Fatty Acids 1997;57:323-9.

27. Chaud M, Franchi AM, Rettori V, McCann SM, Gimeno MF. Ni-tric oxide in the contractile action of bradykinin, oxytocin, andprostaglandin F2 alpha in the estrogenized rat uterus. Proc NatlAcad Sci U S A 1997;94:11049-54.

28. Chaud MA, Franchi AM, Beron de Astrada M, Gimeno MF. Ni-tric oxide mediates platelet-activating factor stimulatory actionon uterine prostaglandin production. Prostaglandins Leukot Es-sent Fatty Acids 1998;58:55-9.

29. Kitagawa M, Takahashi M, Yamaguchi S, Inoue M, Ogawa S, Hi-rokawa K, et al. Expression of inducible nitric oxide synthase(NOS) in bone marrow cells of myelodysplastic syndromes.Leukemia 1999;13:699-703.

30. Nickoloff BJ, Naidu Y. Perturbation of epidermal barrier func-tion correlates with initiation of cytokine cascade in human skin.J Am Acad Dermatol 1994;30:535-46.

31. Riemer RK, Buscher C, Bansal RK, Black SM, He Y, Natuzzi ES.Increased expression of nitric oxide synthase in the myome-trium of the pregnant rat uterus. Am J Physiol 1997;272:E1008-15.

32. Ekerhovd E, Weidegard B, Brannstrom M, Norstrom A. Nitricoxide-mediated effects on myometrial contractility at term dur-ing prelabor and labor. Obstet Gynecol 1999;93:987-94.

33. Nelson SH, Steinsland OS, Wang Y, Yallampalli C, Dong YL,Sanchez JM. Increased nitric oxide synthase activity and expres-sion in the human uterine artery during pregnancy. Circ Res2000;87:406-11.

34. Kublickiene KR, Nisell H, Poston L, Kruger K, Lindblom B.Modulation of vascular tone by nitric oxide and endothelin 1 inmyometrial resistance arteries from pregnant women at term.Am J Obstet Gynecol 2000;182:87-93.

Receive tables of contents by e-mail

To receive the tables of contents by e-mail, sign up through our Web site athttp://www.mosby.com/ajog. Choose E-mail Notification. Simply type your e-mail address in the box and click the Subscribe button. Alternatively, you may send an e-mail message to majordomo@mosby.com. Leave the subject line blank, and type the following as the body of yourmessage: Subscribe ajog_toc.

You will receive an e-mail to confirm that you have been added to the mailing list. Note thattable of contents e-mails will be sent out when a new issue is posted to the Web site.