The effects of the antiprogesterone RU486 (Mifepristone)* on an endometrial secretory glycan: an immunocytochemical study

Transcription

1 FERTILITY AND STERILITY Copyright 1991 The American Fertility Society Printed on ocid-free paper in U.S.A. The effects of the antiprogesterone RU486 (Mifepristone)* on an endometrial secretory glycan: an immunocytochemical study Rosalind A_ Graham, RSc.t Tin-Chiu Li, M.R.C.O.G.:\: Mourad W. Seif, M.R.C.O.G.t John D. Aplin, Ph.D.t Ian D. Cooke, F.R.C.O.G.:\: University of Sheffield, Jessop Hospital for Women, Sheffield, and University of Manchester, Saint Mary's Hospital, Manchester, United Kingdom Objective: To examine the effects of progesterone (P) receptor blockade by RU486 (Mifepristone; Roussel-Uclaf, Paris, France) on a secretory endometrial glycan recognized by monoclonal antibody D9Bl. Design: Retrospective comparison of endometrial biopsies from treated and untreated omen from 2 to 8 days after the luteinizing peak (LH) peak. Setting: Infertility clinic, Jessop Hospital for Women, Sheffield. Patients: Tenty-to normal fertile omen received the RU486. A control group of 44 normal fertile omen ere also assessed. Interventions: RU486 as administered to 22 normal omen during the first half of the luteal phase and an endometrial biopsy examined 3 days later. Main Outcome Measures: Immunohistochemistry as used to assess the production and secretion of the D9B1 epitope. Results: When the drug as given 2 days after the LH peak, it prevented appearance of the epitope. When RU486 as administered 5 days after the LH peak, epitope already present in gland cells as subsequently secreted. Conclusions: These data suggest that production of the sialo-oligosaccharide is P-dependent, but secretion through established intracellular pathays is P-independent. Fertil Steril 55:1132, 1991 Recently e have developed a biochemical approach to the assessment of endometrial glandular secretory differentiation. 1-3 We have shon that a sialoglycan, recognized by monoclonal antibody D9Bl, appears in gland cells 2 to 3 days after ovulation and reaches a maximum in glandular secretions 6 to 7 days after the luteinizing hormone (LH) peak, suggesting a probable dependence on progesterone (P) for its production. In the present study, Received July 23, 1990; revised and accepted February 14, * Mifepristone, Roussel-Uelaf, Paris, France. t Department of Obstetrics and Gynaecology, University of Manchester, Saint Mary's Hospital. :j: Department of Obstetrics and Gynaecology, University of Sheffield, Jessop Hospital For Women. Reprint requests: J. D. Aplin, Ph.D., Department of Obstetrics and Gynaecology, University of Manchester, St. Mary's Hospital, Whitorth Park, Manchester, M13 OJH, United Kingdom Graham et ai. RU486 effects on endometrial secretion e have investigated the nature of this dependence in vivo using the antiprogestin RU486,4 hich has the ability to interrupt the hormonal support of the endometrium regardless of hether fertilization and implantation have occurred. 5 During the early luteal phase, plasma P levels increase steadily. Within 2 to 3 days of the LH surge, the endometrial gland cells have organized themselves for the production of secretory material.6 By 5 days after the LH peak, an extensive vesicular netork has been elaborated in the apical cytoplasm. The cells also contain many stacks of apical Golgi profiles at this stage. 6 Morphological evidence6-l0 suggests that a large increase then occurs in the secretion of glycogen and glycoprotein into the gland lumens, reaching maximal levels around the time of implantation. We no report the effects of treating omen ith RU486 beteen 2 and 5 days after the LH peak on Fertility and Sterility

2 the appearance and distribution of the D9B1 epitope. Using this approach, e have been able to examine the effect of P receptor blockade on to specific aspects of endometrial epithelial function: production of a stage-specific secretory glycan and its subsequent secretion. Subjects MATERIALS AND METHODS The study involved a retrospective analysis of material obtained in a previous prospective study.lo Thirty normal, healthy, fertile volunteers ho ere requesting tubal sterilization, or ho had previously been sterilized and ere seeking reversal of tubal sterilization, took part in the original study. Patients ere informed of the nature of the study and possible side effects of the drug and gave ritten consent. The criteria for selection of patients to obtain a homogeneous population have been previously described.lo Briefly, all omen ere of knon fertility, they had regular menstrual cycles of beteen 25 and 35 days, there as no demonstrated uterine pathology, and no oral contraceptive, hormone treatment, or intrauterine contraceptive device had been used for at least 3 months. Endometrial Biopsy and Administration of RU486 From the midfollicular phase, daily LH monitoring by radioimmunoassay on plasma samples or enzyme immunoassay on early morning urine samples as performed,s so that all endometrial biopsies could be timed accurately from the LH peak, a method shon to improve the precision of endometrial dating.s Initially, a baseline endometrial biopsy (biopsy 1) as performed beteen 2 and 6 days after the LH peak (LH + 2 to LH + 6) as an outpatient procedure. Endometrial structure as examined by light microscopy. Any patients ho had abnormal histology as assessed by the traditional dating criteria 7 ere eliminated from the study, thus maintaining a homogeneous population. Because the remaining patients ere normal as defined by the above criteria, their baseline biopsies ere used along ith others to create a control group of 44 omen to define the normal range of expression of the D9B1 epitope. 3 The ages of the control group ranged from 24 to 39 years ith a mean of 33.8 years and SD of 3.3 years. The patients treated ith RU486 ere also beteen 24 and 39 years of age ith the average age being 33.2 years and the SD 3.6 years. A single dose of RU486 as administered to the omen immediately after biopsy 1, beteen 2 and 6 days after the LH surge. The range of dosage as from 5 to 200 mglo (200 mg, n = 1; 150 mg, n = 5; 100 mg, n = 2; 75 mg, n = 2; 50 mg, n = 6; 25, 10, and 5 mg, n = 2 per group). Three days later, beteen LH + 5 and LH + 9, a second endometrial biopsy (biopsy 2) as performed. Although the numbers of patients at each individual combination of dose and time of biopsy ere insufficient to allo a statistical analysis, the results on a given day ere essentially similar throughout the dose range. Data from all treated patients biopsied on a given day ere therefore pooled for analysis. In a control experimentlo involving four subjects, e investigated the possibility that hen performing to endometrial biopsies ithin 3 days of each other, the trauma of the first biopsy could have an effect on further differentiation of endometrium that may be apparent in the second biopsy. To biopsies ere taken 3 days apart in the first half of the luteal phase (as in the study group) ithout the administration of RU486. The histologic dating of the second biopsies all agreed ith the chronological dating, suggesting the first biopsy had no significant effect on the second. Tissue Processing All endometrial biopsies ere pinned out on a piece of dental ax, cut up in a standardized manner into 5 pieces,6 and fixed immediately in 2% glutaraldehyde in sodium cacodylate buffer for 4 hours. To larger pieces of tissue ere processed for light microscopy and embedded in JB4 resin (Polysciences, Warrington, United Kingdom). Thin (2 m) sections ere cut using a glass knife on an AS500 microtome (Anglia Scientific, Cambridge, United Kingdom) and used either for histologic analysis using the dating criteria of Noyes et al. 7 and morphometric analysis 9 or for the immunocytochemical study. Immunocytochemical Analysis The indirect immunoperoxidase technique used has been previously described. 3 Briefly, it involved an overnight incubation of tissue sections in the mouse monoclonal antibody D9B1, folloed by buffer ashes and incubation in peroxidase-conjugated rabbit antimouse immunoglobulins (Dako, High Wycombe, United Kingdom). The peroxidase label as developed ith 3,3' -diaminobenzidine tetrahydrochloride to give an insoluble, bron polymeric product at the site of the reaction. Under in- Graham et al. RU486 effects on endometrial secretion 1133

3 cident illumination the polymer reflected light, the amount of hich could be measured in particular regions of endometrial glands, i.e., the cell base, apex, and gland lumen. 3 This immunocytochemical analysis as performed on the second endometrial biopsy taken from 22 of the original 30 patients, beteen LH + 5 and LH + 8. The other 8 patients had endometrial biopsies taken on LH + 9, after RU486 administration on LH + 6. This time frame is not appropriate for examination of the D9Bl epitope, most of hich is already secreted by LH To blocks of tissue ere examined per patient, and 25 glands ere examined per block. Data from the treatment group ere compared ith the normal fertile group using Student's t-tests and a to-ay ANOVA. RESULTS Figure 1 shos the mean reflected light measurements for both the control and post-ru486 endometrial biopsies in the cell base, apex, and gland lumen. The control endometrium shoed intracellular staining rising to a maximum 3 days after the LH peak (LH + 3) in the cell base and 5 days after the LH peak (LH + 5) in the apical cytoplasm. Secretory staining began to rise at LH + 4 and as maximal at LH + 6 to 7. A dramatic reduction in the amount of epitope as seen hen RU486 as administered on LH + 2 and the endometrial biopsy examined 3 days later on LH + 5. Intracellular staining as minimal, and any epitope present as found ithin the gland lumen (Fig. 2A). Normally on LH + 5, intracellular staining is present in both basal and apical areas of the cytoplasm (Fig. 2B). Student's t-tests ere performed on data from the to sets of biopsies taken on day LH + 5. Significant differences are observed in the amount of epitope present in all three regions of the gland (cell base: t = 2.24, df = 9, P < 0.01; cell apex: t = 5.28, df = 9, P < 0.001; gland lumen: t = 5.11, df = 9, P < 0.001). When RU486 as given on LH + 3 and the biopsy examined on LH + 6, a significant reduction in the amount of epitope compared ith normal could be seen (Figs. IB and C) in the cell apex (t = 3.79, df = 17, P < 0.002) and gland lumen (t = 11.92, df = 17, P < 0.001) but not in the subnuclear region (Fig. la). The majority of staining present as in the gland lumens, although some distinct apical deposits ere still detectable. In post-ru486 biopsies performed on day LH + 7 (RU486 administered on LH + 4), the intracellular staining as minimal, ith the epitope detectable 1134 Graham et ai. RU486 effects on endometrial secretion u.j u.j :I: c.!j Z d: 3 2 ( A - CELL BASE 0 CHRONOLOGICAL DATE (LH+) I 6 :I: c.!j 4 Z d: B - CELL APEX ---+/ A" " 2 t. ".. 1 CHRONOL OG I CAL C - GLAND LUMEN (12) (5) 20 DATl (LH+) ( ) (4) /' -... :I:, c.!j (5) (5) /1" Z 10 r/ (7) (4) d: (6) (2) 0 CHRONOLOGICAL DATE (LH+) Figure 1 Mean reflectance measurements over (A), the cell base, (B), the cell apex, and (e), the gland lumen on different days in the secretory phase of fertile (.. _) and RU486-treated (- - -) omen. Bars at each point indicate SEs. Numbers in brackets represent the number of subjects in each group on each day. in the gland lumen. This pattern is similar to that expected in normal LH + 7 endometrium. Hoever, a significant difference from a normal LH + 7 biopsy as seen in the gland lumen (t = 2.37, df = 7, P < 0.05). Less epitope appeared to have accumulated in the lumens of RU486-treated biopsies. Finally, hen RU486 as administered on LH + 5 and the endometrium examined on LH + 8, a similar picture to LH + 7 as seen. Figure 3 illustrates the staining patterns seen in a typical post RU486 gland and a normal endometrial gland at LH + 8. No significant difference as seen in the gland lumen nor the cell base, but a significant diminution in apical epitope as detected (t = 2.73, df = 6, P < 0.05). Fertility and Sterility

4 DISCUSSION Figure 2 Expression ofthe D9B1 epitope in endometrial glands from biopsies taken 5 days after the LH surge from (A), a patient treated ith RU486 on LH + 2 (X165) and (B), a normal fertile oman (X330) Data ere examined using a to-ay ANOVA. Table 1 shos the results from the control group of normal omen and the patients treated ith RU486, at 5, 6, 7, and 8 days after the LH surge. The reflected light measurements obtained from the basal and apical regions of the gland cells sho a significant effect of both time, i.e., the chronological date ofthe biopsy in number of days from the LH surge and condition, i.e., control compared ith treatment group. A significant interaction beteen time and condition can also be seen. In the gland lumen, a significant effect of condition but not time as noted, and no significant interaction as seen. A comparison as also made beteen the postru486 LH + 5 biopsies obtained after administering the drug on LH + 2 and normal day LH + 2 biopsies. Significantly more epitope as present in the apical region of gland cells in the normal LH + 2 biopsies than in the post-ru486 LH + 5 biopsies (t = 3.81, df = 9, P < 0.01). No significant difference existed beteen the amount of epitope in the cell base and gland lumen. When the post-ru486 LH + 6 biopsies ere compared ith normal LH + 3 biopsies, significantly diminished staining as detected after treatment in all three regions of the gland. The earlier in the secretory phase that RU486 is administered, the more marked is its effect. On day LH + 2 of the normal cycle, the epitope is just beginning to be produced ithin gland cells,3 hereas on days LH + 3 to LH + 6, significant reserves of intracellular epitope are detected, indicating ongoing synthesis and intracellular mobilization. Throughout this period, significantly reduced intracellular immunostaining in the treated specimens strongly suggests that RU486 acts to block the production of the epitope, hich must therefore be regarded as a P-dependent process. Biochemical elucidation of the P-dependent step in sialoglycan production remains to be achieved; sialylation is a late step in glycoprotein biosynthesis, and so there are numerous possible control points at hich may exert an effect on the final product. Observations made on immunodeposits in gland lumens suggest a second important conclusion relating to tissue behavior on days LH + 5 onard, hen the sialoglycan is accumulating as a secretory product. Administration of RU486 at or after day LH + 5 allos an examination of the effect of P Figure 3 Expression of the D9B1 epitope in endometrial glands from biopsies taken on LH + 8 from (A), a patient treated ith RU486 on LH + 5 (X330) and (B), a normal fenile oman (X440). No difference is evident in the glandular staining pattern, suggesting that epitope present in gland cells on LH + 5 is secreted normally during drug exposure. Graham et al. RU486 effects on endometrial secretion 1135

5 Table 1 To-ay ANOV A Performed on Data From Normal Fertile Women and RU486-Treated Patients on Days 5, 6, 7, and 8 After the LH Surge Source of variation Base Apex Lumen Condition (df = 1,39)" 8.09 b d 12.12b " Numerator and denominator. b p < < P < d P < Time (df = 3,39)" 11.21b 4.84 b 2.17 Interaction (df = 3, 39)" 3.62< 4.67 b 0.89 receptor blockade on the ability of existing intracellular reserves of epitope to escape the cell. Although the quantity of product is reduced relative to the untreated group (thus indicating some synthetic activity continuing at this time), it is clear that material that has already been produced is indeed secreted. The period of major intracellular accumulation of the D9Bl epitope coincides ith the elaboration of an abundant Golgi apparatus and secretory netork ithin gland cells. It ould appear from the present data that, providing such a netork is in place at the time of drug administration, intracellular reserves of the D9Bl epitope are transported through and out of the cell in the normal ay. Thus the secretory process itself appears to be P-independent. In the morphometric study of Li et ai.,1o it as reported that P receptor blockade by RU486 reduces glandular secretory activity. Evidence for this came from reductions in the amount of secretory material in the gland lumen, the volume fraction of the gland occupied by the gland lumen, and the maximum diameter of the gland. Studies using lectins to detect endometrial glycans ll have demonstrated an overall rise in the production of a range of structures from a basal level in the proliferative phase to significantly higher levels in the secretory phase. Thus P can stimulate de novo glycan production (as e have demonstrated ith D9Bl) or elevate an already active glycosylation pathay. It appears that RU486 must block both of these effects. The effect of RU486 on endometrial glandular secretions as observed throughout the dose range (5 to 200 mg) studied, suggesting that the production of the D9Bl epitope is sensitive to lo levels of RU486. Timing of administration of RU486 is more important in its subsequent effect on the tissue; the most striking results are seen hen administration is on day LH + 2 or LH + 3, before most epitope production has begun. These results are consistent ith other studies,12 as ell as ith our earlier light microscopic observations1o of significant morphological changes throughout the dose range and the importance of timing the administration of the drug in relation to the LH peak. Our data suggest that the early secretory phase is a critical period of P dependent gene expression, beginning the tissue reorganization required for normal implantation and placentation. Other data obtained using antibody D9Bl13 indicate that adequate differentiation may only be achieved if the target tissue is suitably responsive to P stimulation and that an intrinsic endometrial dysfunction may interfere in some infertile omen. Ackrledgment. We gratefully acknoledge the support of Roussel-Uclaf, Paris, France. REFERENCES 1. SeifMW, Aplin JD, Foden LJ, Tindall VR: A novel approach for monitoring the endometrial cycle and detecting ovulation. Am J Obstet Gynec 160:357, Aplin JD, Hoadley ME, Seif MW: Hormonally regulated secretion of keratan sulphate by human endometrium epithelium. Biochem Soc Trans 17:136, Smith RA, Seif MW, Rogers A W, Li TC, Dockery P, Cooke ID, Aplin JD: The endometrial cycle: the expression of a secretory component correlated ith the luteinising hormone peak. Hum Reprod 4:236, Baulieu EE: Contragestion and other clinical applications of RU486, an antiprogesterone at the receptor. Science 245:1351, Baulieu EE, Ulmann A: Antiprogesterone activity of RU486 and its contragestive and other applications. Hum Reprod 1:107, Dockery P, Li T-C, Rogers AW, Cooke ID, Lenton EA: The ultrastructure of the glandular epithelium in the timed endometrial biopsy. Hum Reprod 3:826, Noyes RW, Hertig AT, Rock J: Dating the endometrial biopsy. Fertil Steril 1:3, Li T-C, Rogers AW, Lenton EA, Dockery P, Cooke 10: A comparison beteen to methods of chronological dating of human endometrial biopsies during the luteal phase, and their correlation ith histological dating. Fertil SteriI48:928, Li T-C, Rogers AW, Dockery P, Lenton EA, Cooke ID: A ne method of histological dating of human endometrium in the luteal phase. Fertil Steril 50:52, Li T -C, Dockery P, Thomas P, Rogers AW, Lenton EA, Cooke ID: The effects of progesterone receptor blockade in the luteal phase of normal fertile omen. Fertil Steril 50:732, Aplin JD: Glycans as biochemical markers of endometrial secretory differentiation. J Reprod Fertil. In press, Sahn ML, Bygdeman M, Cekan S, Xing S, Masironi B, Johannisson E: The effect ofru486 administered during the early luteal phase on bleeding pattern, hormonal parameters and endometrium. Hum Reprod 5:402, Graham RA, Seif MW, Aplin JD, Li TC, Cooke 10, Rogers A W, Dockery P: An endometrial factor in unexplained infertility. Br Med J 300:1428, Grahametal. RU486 effects on endometrial secretion