antinidatory effect of luteal phase administration of mifepristone (ru486) is associated with...

ORIGINAL RESEARCH ARTICLE

Antinidatory Effect of Luteal PhaseAdministration of Mifepristone (RU486)Is Associated With Changes inEndometrial Prostaglandins Duringthe Implantation WindowNihar Ranjan Nayak, Jayasree Sengupta, and Debabrata Ghosh

Luteal phase administration of mifepristone provides asignificant degree of pregnancy protection to monkeys andwomen. Among several proposed mediators of the antini-datory action of luteal phase mifepristone, prostaglandins(PG) at the endometrial level appear important, and wasexamined in the present study using the rhesus monkey asthe primate model. To this end, the concentrations of PGE2

and PGF2a in endometrium and the profiles of cyclooxy-genase (COX) and 15-hydroxy prostaglandin dehydroge-nase (PGDH) were examined in untreated control animals,in animals subjected to mifepristone treatment (2 mg/day)alone or along with diclofenac (25 mg/day), or along with aPGE1 analog (100 mg misoprostol), in animals subjected todiclofenac alone treatment, and in animals treated withmisoprostol alone on cycle days 16, 17, and 18. Tissuesamples were collected on day 20 of treatment cycles fromanimals with discernible corpora lutea. Early luteal phasetreatment with diclofenac did not result in any remarkablechange in endometrial prostaglandin concentrations, how-ever, there was an increase in the profile of COX. Animalsexposed to misoprostol in the prereceptive stage, on theother hand, exhibited decreased expression of endometrialCOX. The concentrations of PGF2a and PGE2, as well asthe ratios of PGF2a to PGE2 concentrations, were increasedalong with a decrease in COX and PGD in endometrialsamples following luteal phase mifepristone treatment.Although the underlying cellular mechanism of regulationof COX and PGDH in mifepistone-treated endometriumremains to be examined, the decrease in PG catabolismthrough low PGDH may contribute to the increased PGand high ratio of PGF2a to PGE2 in mifepristone-exposed

endometrium. It is plausible that mifepristone action onendometrial cells is mediated by an altered ratio of PGF2a

to PGE2. Furthermore, it appears that the regulation of PGmilieu by COX and PGDH activities in reproductive tis-sues is under complex regulatory mechanism and is tem-porarily correlated with specific developmental events.CONTRACEPTION 1998;58:111–117 © 1998 Elsevier ScienceInc. All rights reserved.

KEY WORDS: cyclooxygenase, endometrium, mifepristone,prostaglandins, prostaglandin dehydrogenase

Introduction

Progesterone plays a key role in the process ofimplantation and establishment pregnancy.Thus, the use of a potent antiprogestin like

mifepristone (RU486) has emerged as a new approachto fertility regulation.1 Additionally, it provides a toolfor investigating the mechanism of action of proges-terone toward endometrial receptivity. Luteal phaseadministration of RU486 has been reported to induceseveral changes in the progesterone-dominated endo-metrium, such as desynchronization of endometrialmaturation, a delay in endometrial glandular secre-tory differentiation, and changes in vascular compart-ments in women2–6 and in monkeys.7–9 Among sev-eral proposed mediators of the antinidatory action ofluteal phase mifepristone, prostaglandins (PG) at theendometrial level appear important. Prostaglandins(mainly PGE and PGF) at the endometrial level play acritical role in the process of implantation.10,11 Inseveral nonprimate mammalian species, treatmentwith PG biosynthesis inhibitor prevents or delaysimplantation.12 However, the role of PG in the peri-implantation stage endometrium in primates is notclear. Concentrations of endogenous PG rise in endo-metrium during the luteal phase of the menstrual

Department of Physiology, All India Institute of Medical Sciences, New Delhi,India

Name and address for correspondence: Dr. Debabrata Ghosh, Departmentof Physiology, All India Institute of Medical Sciences, New Delhi 110029, India;Tel.: 191-11-696-3826; Fax: 191-11-686-2663

Submitted for publication March 6, 1998Revised May 12, 1998Accepted for publication June 23, 1998

© 1998 Elsevier Science Inc. All rights reserved. ISSN 0010-7824/98/$19.00655 Avenue of the Americas, New York, NY 10010 PII S0010-7824(98)00068-7

cycle when progesterone concentrations are high.13,14

However, contradictory reports are also available. Invitro exposure of antiprogestin to human endometrialcells stimulated endogenous PG production and in-hibited PG catabolism.15,16 On the contrary, Gem-zell-Danielsson and Hamberg17 have reported thatearly luteal phase administration of RU486 inhibitsthe release of endometrial PGF. We have, however,observed that the antinidatory effect, as well as thechanges induced by mifepristone in the luteal phasemonkey endometrium, are not altered by administra-tion of PG synthesis inhibitor or a PGE analog.9,18

The aim of the present study was to understand theinvolvement of PG in the process of implantation andtheir role in mediating the anti-implantation activityof RU486 in the rhesus monkey. To this end, theconcentrations of PGE and PGF in endometrium andthe profiles of enzymes responsible for their synthesis(cyclooxygenase) and degradation (15-hydroxy prosta-glandin dehydrogenase) were examined after treat-ment with different combinations of mifepristone, aPG synthesis inhibitor (diclofenac), and a PGE analog(misoprostol) in the rhesus monkey.

Materials and MethodsAnimalsHealthy, adult female and male rhesus monkeys(Macaca mulatta) of proven fertility were used in thisstudy. The details of animal selection and manage-ment have been described elsewhere.7–9,18,19 Mon-keys were individually housed in cages under semi-natural conditions in the Primate Research Facility ofthe All India Institute of Medical Sciences (NewDelhi, India). Monkeys showing at least two consec-utive cycles of normal length (28–32 days) were usedin this study. Daily blood samples were collected todetermine the peripheral serum levels of estrogen andprogesterone by radioimmunoassay methods as de-scribed previously,7–9,18,19 using antisera, chemicalsand protocol obtained from the World Health Organi-zation (WHO) Matched Assay Reagents Programme.20

The study was performed with the approval of theEthics Committee on the Use of Non-Human Pri-mates in Biomedical Research, All India Institute ofMedical Sciences.

Treatment GroupsA total of 49 female monkeys were used in this study.Twenty-four monkeys (n 5 24) were randomly as-signed to six different treatment groups (groups 1–6),each group having four animals for the measurementof concentrations of PG in endometrium. Twenty-five monkeys (n 5 25) were selected for immunohis-

tochemical studies in six treatment groups (groups1–6), groups 1–4 and 6 had four animals each, andgroup 5 had five animals. Monkeys in group 1 werenot subjected to any treatment and were used as thenormal control group. In group 2, animals were in-jected subcutaneously with RU486 (2 mg/day/animal)in 1 mL vehicle (1:4, benzyl benzoate:olive oil, v/v) oncycle days 16, 17, and 18. Animals in group 3 weretreated with diclofenac sodium (25 mg/day/animal,intramuscularly; Biochem Pharmaceutical Industries,India), a cyclooxygenase inhibitor on cycle days 16,17, and 18. In group 4, animals received a PGE1 analog(100 mg misoprostol, Laboratoires Searle, Boulogne-Billancourt, France) by gavage on cycle days 16, 17,and 18. In group 5, animals were subjected to treat-ment with RU486 (2 mg/day, subcutaneously) alongwith diclofenac (25 mg/day, intramuscularly) on cycledays 16, 17, and 18. Animals in group 6 also receivedRU486 (2 mg/day, subcutaneously) on cycle days 16,17 and 18, along with misoprostol (100 mg/day, orally).

Tissue samples were collected on day 20 of treat-ment cycles from animals with discernible corporalutea. Retrospective analyses based on steroid hor-mone concentrations in peripheral circulation re-vealed that all animals in six treatment groups hadovulations between days 9 and 14.

Radioimmunoassays of PGF2a and PGE2in the EndometriumEndometrial tissue samples were collected on matedmenstrual cycle day 20 of treatment cycles by per-forming laparotomy and fundal hysterotomy follow-ing ketamine (12 mg/kg, Ketlar, Parke-Davis, Mum-bai, India) anesthesia and were used for theestimation of PG. The presence of a corpus luteumwas also checked before tissue collection. The proce-dural details of tissue collection have been describedelsewhere.7–9,19 Extraction of PG from tissue sampleswere performed following the method described byWheeler et al.21 Briefly, tissue samples were washedquickly in ice-cold phosphate buffered saline (PBS),weighed, transferred to 5 mL of prechilled (220°C)absolute ethanol, and homogenized in an all-glasshomogenizer. After centrifugation at 4°C for 10 min,the supernatant was collected. The precipitate waswashed with 5 mL of prechilled absolute ethanol,centrifuged, and supernatant combined with the orig-inal extract, evaporated to dryness under reducedpressure at 45°C. Residue was dissolved in 15 mL ofwater (pH 4.5). Prostaglandins were extracted by par-titioning three times with 30 mL of redistilled ethylacetate, evaporated to dryness, residue redissolved in10 mL of redistilled ethyl acetate and stored at 220°Cuntil assayed. The concentrations of PGF2a and PGE2

112 Nayak et al. Contraception1998;58:111–117

were estimated using Biotrak radioimmunoassay sys-tem kits and the standards and the protocols obtainedfrom Amersham International (Amersham, Bucking-hamshire, England). The concentrations of bothPGF2a and PGE2 were expressed as pg/100 mg wettissue. The assay for all samples were performed inthe same batch. Inasmuch as PGE is more stable in itsmethyl oximate form, the assay system for PGE2 wasdesigned for estimation of the methyl oximate deriv-ative of PGE2. Within-assay precision for PGF2a andPGE2 were 7.8% and 5.6%, respectively.

Immunohistochemistry of Cyclooxygenase (COX)and 15-Hydroxy Prostaglandin Dehydrogenase(PGDH)Endometrial tissue samples were collected on cycleday 20 from all monkeys as described previously. Theprocedural details of tissue processing and immuno-histochemistry have been described earlier.7–9,19

Briefly, tissue samples were fixed in phosphate-buff-ered neutral formaldehyde (4%) and embedded inparaffin wax. Tissue sections (5 mm) were deparaf-finized and hydrated through graded alcohol to phos-phate-buffered saline (PBS). Endogenous peroxidaseactivity was then quenched with 0.3% hydrogenperoxide in methanol. Sections were incubated over-night at 4°C with the primary antibody. Primaryantibodies for COX and PGDH were raised in rabbitsand were obtained from Cayman Chemical Company(Ann Arbor, MI). Sensitivity (COX, 1:50, v/v; PGDH,1:500, v/v) was precalibrated by diluting the stock(1 mg/mL) and performing 3–5 points titration, basedon the information provided by the manufacturer.Final visualization was achieved using the ABC kit(Vector Laboratories, Burlinghame, CA) and freshlymade diaminobenzidine hydrochloride (Sigma Chem-ical, St. Louis, MO) with hydrogen peroxide as de-scribed previously.7–9 Specificity of antibody ligand-ing and visualization were assessed by omitting

primary antibodies, replacing primary antibodieswith unrelated immunoglobulins from same species,omitting secondary antibodies, and replacing labeledsecondary antibody with unrelated immunoglobulinsfrom the same and other species. The reagents werepurchased from Vector Laboratories (Burlinghame,CA). All immunostaining procedures were performedin a single run. Duplicate sections were lightly coun-terstained with hematoxylin to facilitate the identi-fication of cellular elements. The immunohisto-chemically stained sections were analyzedmicroscopically to estimate morphometrically theareas of immunoprecipitation in glandular and stro-mal compartments in functional zones using a Leicamicroscope and a precalibrated computer-assistedvideo image analysis system (QWIN-Quantimet500C1, Leica, Cambridge, England). The details ofhistometric measurements are given elsewhere.7–9

Briefly, glandular and stromal compartments at 325were detected using an interactive planimeter ana-lyzer only in cases in which discernibility of thesestructures was distinct, and immunopositive areaswere measured in a particular compartment (segment)by detecting positive profiles in digitized images basedon an optimized grey level threshold after shadingcorrection and pixel calibration against the standardprovided by the manufacturer.

Statistical AnalysisStatistical analysis of quantitative measurementswere performed using the Kruskal-Wallis test fol-lowed by multiple comparision test.22 The probabil-ity level of p 5 0.05 was taken as the limit ofsignificance. The data are shown as means 6 SEM.

ResultsAs shown in Table 1, PG concentrations were in-creased following mifepristone (group 2) and mifepris-

Table 1. Concentrations (mean 6 SEM, pg/100 mg of wet tissue) of endometrial PGF2a and PGE2, and their ratio(PGF2a/PGE2) in the endometrium on cycle day 20 in different treatment groups*

Treatment group(n 5 4)

Concentrationof PGF2a

Concentrationof PGE2 PGF2a/PGE2

Control (group 1) 871 6 1.49 185 6 1.44 4.76 6 1.11‡Mifepristone (RU486, group 2) 6412 6 1.27† 684 6 1.36‡ 9.38 6 1.26Diclofenac (group 3) 1659 6 1.39 225 6 1.23 7.33 6 1.14Misoprostol (group 4) 1197 6 1.38 268 6 1.19 4.44 6 1.19§Mifepristone plus diclofenac (group 5) 1023 6 1.57 254 6 1.46 4.05 6 1.30§Mifepristone plus misoprostol (group 6) 6237 6 1.24† 719 6 1.26‡ 8.67 6 1.21

*Data log transformed before analysis.†, significantly (p ,0.01) different from group 1, 3, 4, and 5.‡, significantly (p ,0.05) different from group 1, 3, 4, and 5§, significantly (p ,0.05) different from group 2, 3, and 6.

113Contraception Luteal Phase RU486 and Endometrial PG1998;58:111–117

tone plus misoprostol (group 6) treatments as com-pared to control (group 1), diclofenac (group 3),misoprostol (group 4), and diclofenac plus mifepris-tone (group 5) treatments. However, no significantdifferences were observed in the concentrations ofPGF2a and PGE2 among control (group 1), diclofenac(group 3), misoprostol (group 4), and mifepristone plusdiclofenac (group 5) treated endometrial samples. Al-though no significant differences in the ratio of endo-metrial PGF2a to PGE2 concentrations were notedamong treatment groups 1, 4, and 5 or among treat-ment groups 2, 3, and 6, the ratios were significantly(p ,0.05) increased following mifepristone (group 2),diclofenac (group 3), and mifepristone plus misopros-tol (group 6) treatments as compared with that in thecontrol group (group 1). The ratio was also signifi-cantly (p ,0.05) higher in the mifepristone (group 2)and mifepristone plus misoprostol (group 6) treatedendometrium compared with the misoprostol (group4) and mifepristone plus diclofenac (group 5) treat-ment groups.

Table 2 provides morphometric analyses of immu-nohistochemical staining of COX and PGDH in en-dometrial glands and stroma on cycle day 20 in thevarious treatment groups. Cyclooxygenase was im-munolocalized mainly in the cytoplasm of glandularcells. A higher (p ,0.05) degree of COX was observedin the diclofenac treatment group (group 3) comparedwith the control (group 1) and misoprostol (group 4)treatment groups. It was reduced in glandular epithe-lium in the mifepristone treatment groups (groups 2,5, and 6) as compared with the control (group 1) anddiclofenac (group 3) treatment groups, as well as ingroups treated with mifepristone with or withoutmisoprostol (groups 2 and 6) as compared with thegroup treated with misoprostol alone (group 4).

Immunopositive PGDH was observed in both glan-dular and stromal cells in endometrium. A higher(p ,0.05) degree of PGDH was observed in glandularcells following diclofenac treatment (group 3) com-

pared with the mifepristone plus diclofenac (group 5)and the mifepristone plus misoprostol (group 6) treat-ment groups. In stromal cells, it was decreased(p ,0.05) after treatment with mifepristone, with andwithout diclofenac or misoprostol (groups 2, 5, and 6)compared with the control (group 1) and misoprostol(group 4) treated endometrium, and also after mife-pristone with and without diclofenac or misoprostoltreatments (groups 2, 5, and 6) compared with thecontrol (group 1) and misoprostol (group 4) treatmentgroups.

DiscussionLuteal phase administration of a potent antiprogestinsuch as mifepristone (RU486) can render endome-trium hostile to blastocyst implantation2–9 and canprovide a significant degree of pregnancy protectionto monkeys18,23 and women.24 A mifepristone-medi-ated antinidatory effect could suggestively involvePG in implantation stage endometrium. Several ob-servations support this notion. Treatment with a PGsynthesis inhibitor such as diclofenac has been shownto antagonize antiprogestin induced preterm deliveryin rats.25 Antiprogestin has also been shown to stim-ulate PGF production by human endometrial stromalcells in vitro,15 by glandular cells of decidua invitro,16 and by decidual cells in vivo.26 The release ofPGF into luminal uterine fluid has been shown todecrease as a result of antiprogestin treatment towomen during early luteal phase.17 It has been shownearlier that luteal phase treatment with PG synthesisinhibitor (diclofenac) induces marginal repressivechanges in endometrial glands, resulting in a partialinhibition (about 42%) or delay of implantation.18

Furthermore, a mifepristone-induced anti-implanta-tion effect on endometrium could not be reversed bytreatment with either a PG synthesis inhibitor or aPG analog.18 Thus, the role of PG at the endometriallevel during blastocyst implantation and the physio-

Table 2. Morphometric analysis (mean 6 SEM) of immunohistochemical staining of cycloxygenase (COX) and 15-hydroxyprostaglandin dehydrogenase (PGDH) in endometrial glands and stroma on cycle day 20 after different treatments*

Index Group 1 Group 2 Group 3 Group 4 Group 5 Group 6

Immunopositive COX(total stained area, %)

Gland cells 47.1Bc 6 5.8 11.6CD 6 4.5 64.2dEF 6 5.3 37.5F 6 6.7 28.8aF 6 6.6 7.4A 6 2.6Stromal cells ND ND ND ND ND ND

Immunopositive PGDH(total stained area, %)

Gland cells 24.3 6 5.7 16.3 6 4.1 30.6ef 6 5.8 21.5 6 4.8 12.7 6 4.4 10.3 6 4.7Stromal cells 28.5b 6 3.9 12.4d 6 6.4 27.0 6 5.4 31.1ef 6 4.3 15.0a 6 4.3 16.5a 6 3.9

*Control (group 1), mifepristone (RU486, group 2), diclofenac (group 3), misoprostol (group 4), mifepristone plus diclofenac (group 5), and mifepristone plusmisoprostol (group 6); a, b, c, d, e, and f denote significant (p ,0.05) difference from group 1, group 2, group 3, group 4, group 5, and group 6, respectively;corresponding block letters denote the level of significance at 1% (p ,0.01).


logical basis of PG involvement in the mediation ofmifepristone-induced luteal phase contraceptive ac-tion remains unclear.

In the present study, the concentrations of PGF2a

and PGE2 in the endometrium, as well the concentra-tion ratio of PGF2a to PGE2 were found to be in-creased after the luteal phase administration of mife-pristone in the rhesus monkey. It has been shownearlier that the ratio of PGF2a to PGE2 was higher inthe rat uterus after mifepristone treatment.27 Evi-dence from studies involving various nonprimatespecies also suggests that the PG milieu plays acritical role in the regulation of endometrial vascu-larity, implantation, and decidualization.10–12 Thus,it appears feasible that marked changes in the PGmilieu of implantation stage endometrium after mife-pristone treatment may be responsible for implanta-tion failure in rhesus monkeys. However, endome-trial PG physiology at the time of implantationappears complex. For example, misoprostol treatmentduring the early luteal phase did not induce anychange in the morphology of the endometrium9 or theimplantation rate,18 or in the PG milieu, as observedin the present study. On the other hand, treatmentwith diclofenac induced marked changes in PG mi-lieu (increased ratio of PGF2a to PGE2), associatedwith slight repressive changes in a subset of glands,and marginal inhibition of implantation, and with noadditional effect after increasing the dose of diclofe-nac treatment.9,18 Coadministration of diclofenacwith mifepristone resulted in a decrease in concen-trations of PGF2a and PGE2, and their ratio in endo-metrium, along with decrease in the degree of extrav-asation and leukocytic infiltration, and increasedtissue yield;9 yet the anti-implantation efficacy ofthis antiprogestin remained unaffected.18 It has beensuggested that increased production of PG by endo-metrium and decidua is not the only factor underly-ing increased uterine contractility observed aftermifepristone administration for termination of earlypregnancy.28 Increased uterine contractions occureven if decidual prostaglandin production is sup-pressed by indomethacin.26 Thus, the antinidatoryaction of mifepristone is more likely to be mediatedthrough a multifactorial mechanism and the local PGmilieu could be a single module in this complexmechanism.

To investigate further the state of PG metabolismin implantation stage endometrium after early lutealphase treatment with mifepristone, diclofenac, miso-prostol, mifepristone plus diclofenac, and mifepris-tone plus misoprostol, the immunopositive precipita-tion areas for cyclooxygenase (COX) and 15-hydroxyprostaglandin dehydrogenase (PGDH) were assessedin the glandular and stromal compartments of endo-

metrium. Similar to previous reports from humanstudies,11,12 in the present study immunopositiveCOX was observed in the endometrial surface andglandular epithelium of luteal phase endometrium inthe rhesus monkey; PGDH was detected both inglandular epithelia and in stromal cells.

In the present study, early luteal phase treatmentwith diclofenac did not result in any remarkablechange in endometrial prostaglandin concentrations.The observed failure of diclofenac to inhibit endome-trial PG production in this study might have resultedfrom an associated increase in the activity of COX.The highest activity of immunopositive COX wasdetected in gland cells of endometrium obtained fromdiclofenac-exposed monkeys. This may be attribut-able to a rebound response to an initial inhibition ofCOX by diclofenac. It has been demonstrated thatvascular cells have the ability to rapidly synthesizeCOX proteins to replace aspirin-inactivated COX.29

Animals exposed to misoprostol in the prereceptivestage, on the other hand, exhibited decreased (20%)expression of endometrial COX. Thus, the regulationof PG milieu by COX activity in reproductive tissuesis under complex regulatory mechanism and is tem-porarily correlated with specific developmentalevents.

The observation in the present study that theexpression of COX protein was found to be decreased,whereas the concentration of PG in endometriumwas increased after mifepristone treatment in thepresent study appears paradoxical. There are twoisoforms of cyclooxygenase, COX-1 (constitutive) andCOX-2 (inducible). COX-1 enzyme is expressed con-stitutively and is a “housekeeping gene” productinvolved in physiological production of PG necessaryfor normal cellular processes. In contrast, COX-2 isnearly undetectable in most tissues under normalphysiological conditions, but its expression increasesdramatically during inflammation or following expo-sure to mitogenic stimuli.30,31 Differential expressionof COX-1 and -2 proteins have been demonstrated inrat uterus with increased expression of COX-2 duringlabor, proestrus, and estrus stages, suggesting itsinvolvement in increased uterine contractility andcervical ripening.32 Therefore, stimulation of PG pro-duction, increased uterine contractility, and cervicalripening following mifepristone treatment may ensuefrom increased synthesis of COX-2 protein. Further-more, several studies indicate that glucocorticoids,such as dexamethasone can completely inhibit induc-tion of COX-2.33 Conceivably, RU486, which also hasantiglucocorticoid activity, could stimulate COX-2synthesis. Recently, it has also been reported thatleukocytic infiltration is significantly increased inendometrium of rhesus monkeys following mifepris-


tone treatment,7–9 which might help to contribute tothe increased PG production through synthesis ofCOX-2. However, the scenario of endometrial regula-tion of COX proteins seems to be much more com-plex, and thus the observed decrease in immunoposi-tive COX localization following mifepristonetreatment using a polyclonal antibody that can detectboth isoforms of COX remains enigmatic. Furtherstudies are needed to investigate the concentrationsand activities of COX-1 and COX-2, and the profilesof respective mRNAs after mifepristone treatment.

The antiprogestin, mifepristone, used in thepresent study, repressed stromal expression of immu-nodetectable PGDH. The observed increase in con-centration of PG after antiprogestin treatment mightbe caused from decreased catabolism of PG.15,16,26

Indeed, PGDH, which is the principal enzyme for PGcatabolism, is a progesterone-dependent enzyme.34 Ithas been observed in women that mifepristone treat-ment reduced PGDH in decidua and in secretoryphase human endometrium.35 Furthermore, it hasbeen reported that F-type PG are less readily oxidizedby PGDH than that of E-type PG, indicating a highercatabolism of PGE2 than PGF2a.36 It is thus likely thata decrease in PG catabolism through low PGDH,especially in the stromal compartment, may contrib-ute to the increased ratio of PGF2a to PGE2 in mife-pristone-exposed endometria of rhesus monkeys.However, the physiological significance of thischange remains to be elucidated. PGF2a is a potentvasoconstrictor.37 A major target of mifepristone ac-tion in progesterone-dominated endometrium, as itappears from earlier reports, is constriction and re-gression of endometrial vasculature.3,7–9 Collectively,it appears plausible that mifepristone action on endo-metrial vascular cells is mediated by an altered ratioof PGF2a to PGE2.

AcknowledgmentsThe research was conducted with funds received fromthe Rockefeller Foundation. We thank the SpecialProgramme of Research, Development and ResearchTraining in Human Reproduction, World Health Or-ganization, for providing the steroid radioimmunoas-say supplies. Dr. Nayak was supported by a fellow-ship from the Indian Council of Medical Research.

References1. Van Look PFA, von Hertzen H. Post-ovulatory methods

of fertility regulation: the emergence of antiprogester-ones. In: Van Look PFA, Perez-Palacios G, eds. Contra-ceptive Research and Development 1984 to 1994: TheRoad From Mexico City to Cairo and Beyond. Delhi:Oxford University Press, 1994:151–201.

2. Li TC, Dockery P, Thomas P, Rogers AW, Lenton EA ,Cooke ID. The effects of progesterone receptor blockadein the luteal phase of normal fertile women. Fertil Steril1988;50:732–42.

3. Johannisson E, Oberholzer M, Swahn ML, BygdemanM. Vascular changes in the human endometrium fol-lowing the administration of the progesterone antago-nist RU486. Contraception 1989;39:103–17.

4. Swahn ML, Bygdeman M, Xing S, Cekan S, Masironi B,Johannisson, E. The effect of RU486 administered dur-ing the early luteal phase on bleeding pattern, hormonalparameters and endometrium. Hum Reprod 1990;5:402–8.

5. Gemzell-Danielsson K, Svalander P, Swahn ML, Johan-nisson E, Bygdeman M. Effect of single post-ovulatorydose of RU486 on endometrial maturation in the im-plantation phase. Hum Reprod 1994;9:2398–404.

6. Dockery P, Ismail RMJ, Li TC, Warren MA, Cooke ID.The effect of a single dose of mifepristone (RU486) onthe fine structure of the human endometrium duringthe early luteal phase. Hum Reprod 1997;12:1778–84.

7. Ghosh D, Sengupta J, Hendrickx AG. Effect of single-dose early luteal phase administration of mifepristone(RU486) on implantation stage endometrium in therhesus monkey. Hum Reprod 1996;11:2026–35.

8. Ghosh D, Kumar PGLL, Sengupta J. Effect of earlyluteal phase administration of mifepristone (RU486) onleukaemia inhibitory factor, transforming growth factorb and vascular endometrium of the rhesus monkey. JEndocrinol 1998;157:115–25.

9. Nayak NR, Ghosh D, Sengupta, J. Effect of luteal phaseadministration of mifepristone (RU486) and prostaglan-din analogue or inhibitor on endometrium in the rhesusmonkey. Hum Reprod 1998;13:1047–56.

10. Kennedy TG. Involvement of local mediators in blasto-cyst implantation. In: Glasser SR, Mulholland J, Psy-choyos A, eds. Endocrinology of Embryo-EndometriumInteractions. New York: Plenum Press, 1994:183–94.

11. Smith SK. The role of prostaglandins in implantation.Bailliere’s Clin Obstet Gynecol 1991;5:73–93.

12. Ghosh D, Sengupta J. Prostaglandins and blastocystimplantation in primates. In: Ghosh D, Sengupta J, eds.Cellular and Molecular Signalling in Reproduction.New Delhi: New Age International, 1996:145–62.

13. Downie J, Poyser N, Wunderlick M. Levels of prosta-glandins in human endometrium during normal men-strual cycle. J Physiol 1974;236:465–72.

14. Ishihara O, Tsutsumi O, Mizuno M. Metabolism ofarachidonic acid and synthesis of prostanoids in humanendometrium and decidua. Prostaglan Leuket Med1986;24:93–102.

15. Kelly RW, Healy DL, Cameron MJ, Cameron IT, BairdDT. The stimulation of prostaglandin production bytwo antiprogesterone steroids in human endometrialcells. J Clin Endocrinol Metab 1986;62:1116–23.

16. Smith SK, Kelly RW. The effect of antiprogestinsRU486 and ZK98734 on the synthesis and metabolismof prostaglandins F2a and E2 in separated cells fromearly cell decidua. J Clin Endocrinol Metab 1987;65:527–34.

17. Gemzell-Danielsson K, Hamberg M. The effect of anti-progestin (RU486) and prostaglandin biosynthesis in-hibitor (naproxen) on uterine fluid prostaglandin F2concentrations. Hum Reprod 1994;9:1626–30.

18. Nayak NR, Ghosh D, Lasley BL, Sengupta J. Anti-


implantation activity of luteal phase mifepristone ad-ministration is not mimicked by prostaglandin synthe-sis inhibitor or prostaglandin analogue in the rhesusmonkey. Contraception 1997;55:103–14.

19. Ghosh D, Roy A, Sengupta J, Johannisson E. Morpho-logical characteristics of preimplantation stage endo-metrium in the rhesus monkey. Hum Reprod 1993;8:1579–87.

20. Sufi SB, Donaldson A, Jeffcoate SL. World Health Or-ganization Collaborating Centre for Research and Ref-erence Services in the Immunoassay of Hormones inHuman Reproduction—Method Manual. Geneva:World Health Organization, 1988.

21. Wheeler AG, Hurst PR, Poyser NL, Eckstein P. Uterinehistology and prostaglandin concentrations and utero-ovarian venous steroid and prostaglandin concentrationduring the luteal phase of the menstrual cycle inbaboons (Papio spp.) with or without an IUD. J ReprodFertil 1983;67:35–46.

22. Samuels ML. Statistics for the Life Sciences. San Fran-cisco: Dellen Publishing Co, 1994.

23. Ghosh D, Sengupta J. Anti-nidatory effect of a single,early post-ovulatory administration of mifepristone(RU486) in the rhesus monkey. Hum Reprod 1993;8:552–58.

24. Gemzell-Danielsson K, Swahn ML, Svalander P, Bygde-man, M. Early luteal phase treatment with mifepris-tone (RU486) for fertility regulation. Hum Reprod 1993;8:870–3.

25. Cabrol D, Carbonne B, Bienkiewiez A, Dallot A, Alj AE,Cedard L. Induction of labour and cervical maturationusing mifepristone (RU486) in the late pregnant rat:influence of a cyclooxygenase inhibitor (diclofenac).Prostaglandins 1991;42:71–9.

26. Norman JE, Wu WX, Kelly RW, Glasier AF, McNeillyAS, Baird DT. Effects of mifepristone in vivo on decid-ual prostaglandin synthesis and metabolism. Contra-ception 1991;44:89–98.

27. Peplow PV. Actions of cytokines in relation to arachi-

donic acid metabolism and eicosanoid production.Prostaglan Leuket Essen Fatty Acids 1996;54:303–17.

28. Swahn ML, Bygdeman M. The effect of antiprogestinRU486 on uterine contractility and sensitivity to pros-taglandin and oxytocin. Br J Obstet Gynecol 1988;95:126–34.

29. DeWitt DL. Prostaglandin endoperoxide synthase: reg-ulation of enzyme expression. Biochem Biophys Acta1991;1083:121–34.

30. DeWitt DL, El-Harith EA, Kraemer SA, Armstrong RL,Smith WL. The aspirin and heme-binding sites of ovineand murine prostaglandin endoperoxide synthases.J Biol Chem 1990;265:5192–8.

31. Masferrer JL, Zweifel BS, Manning PT, et al. Selectiveinhibition of inducible cyclooxygenase 2 in vivo isantiinflammatory and nonulcerogenic. Proc Natl AcadSci USA 1994;91:3228–32.

32. Dong YL, Gangula PRR, Fang L, Yallampalli C. Differ-ential expression of cycloxygenase-1 and -2 proteins inrat uterus and cervix during the oestrous cycle. Prosta-glandins 1996;52:13–34.

33. DeWitt DL, Meade EA, Smith WL. PGH synthaseisoenzyme selectivity: the potential for safer nonsteroi-dal antiinflammatory drugs. Am J Med 1993;95(suppl2A):40s–4s.

34. Casey ML, Hemsell DL, MacDonald PC, Johnston JM.NAD1-dependant 15-hydroxyprostaglandin dehydroge-nase activity in human endometrium. Prostaglandins1980;19:115–22.

35. Cheng L, Kelly RW, Thong KJ, Hume R, Baird DT. Theeffects of mifepristone (RU486) on prostaglandin dehy-drogenase in decidual and chorionic tissue in earlypregnancy. Hum Reprod 1993;8:705–9.

36. Flower JK. Drugs which inhibit prostaglandin biosyn-thesis. Pharmacol Rev 1974;26:33–67.

37. Myers A, Uotila P, Foegh ML, Ramwell P. The eico-sanoids: prostaglandins, thromboxane and leukotrines.In: DeGroot LJ, ed. Endocrinology, Vol. 3. Philadelphia:WB Saunders, 1989:2480–90.


antinidatory effect of luteal phase administration of mifepristone (ru486) is associated with...

Documents