PROSTAGLANDINSLEUKOTRIENES ANDESSENTIALFATTYACIDS Prostaglandins Leukomenes and Essentnl D Longman Group UK Ltd 1992

Fatty Aads


(1992) 46. 223-229

Immunohistochemical Localization of PGF2, Receptor in the Rat Ovary D. J. Orlicky, L. Fisher, N. Dunscomb

and G. J. Miller

Department qf Pathology, University of Colorado Health Sciences Center and University of Colorado Cancer Center, 4200 East Ninth Avenue, Denver, Colorado 80262. USA (Reprint requests to DJO) ABSTRACT. As a step towards understanding the role of prostaglandin F,, (PGF,,) in ovarian function, a rabbit antiserum against purified PGF,, receptor (PGF,,-R) was produced. This report details the use of this antiserum in immunohistocbemical staining of ovaries of non-pregnant and pregnant rats to ascertain which cell types, in vivo, possess PGF,,-R. In non-pregnant rats, three ovarian cell subpopulations contain immunoreactive PGF2,-R. These include: a subpopulation of the cells found in corpora lutea, a subpopulation of the thecal cells surrounding secondary and mature (Graafian) follicles, and a subpopulation of primary and secondary interstitial cells. The ovarian tissues and cell types in which immunoreactive PGF,,-R cannot be demonstrated include: the serosa bverlying the ovary and its vessels, the coelomic epithelium and its underlying cortical stroma, medullary stroma and vessels, granulosa cells of primary, secondary and mature follicles, the oocyte, and the blood vessels and stroma within corpora lutea. PGF2,- R immunohistochemical staining of corpora lutea from non-pregnant animals was examined both prior to the start of luteolysis and during luteolysis. During luteolysis, cells undergoing apoptosis stained for the presence of PGF2,-R. PGF2,-R immunohistochemical staining was also examined in corpora lutea during pregnancy and until 4 days postpartum. The major findings here were the apparent large increase in staining intensity of granulosa-lutein cells during pregnancy, and the loss of PGF,,-R immunopositivity of the granulosa-lutein cells during the postpartum period. In summary, three ovarian cell subpopulations, all of which can secrete steroids, possess immunoreactlve PGF2,-R.

INTRODUCTION It has been suggested that prostaglandin F,, (PGF,,) may be important in normal ovarian physiology for induction of ovulation (possibly by inducing follicular contractility) and luteolysis (l-lo). The mechanism of luteolysis may involve changes in ovarian blood flow (1 l-13), reduction in corpora lutea luteinizing hormone (LH) receptors (14), inhibition or uncoupling of LHstimulated CAMP formation and protein kinase A activation (15-19), or direct cytotoxicity (20). The aforementioned studies have either been performed in vivo without knowledge of whether a direct or indirect action of PGF,, was being observed, or with isolated cultured cells no longer in the confines of their natural milieu but with the hope that they would respond to PGF,, in culture as in vivo. Understanding the role of PGF,, would be greatly enhanced by clear delineation of the cell types possessing the PGFza receptor (PGF,,-R). In previous studies, the

Date received 8 November 199 1 Date accepted 73 December I99 1

PGF,,-R was isolated and purified and a polyclonal antiserum against the receptor was produced (21, 22). The present study describes the immunohistochemical localization of this receptor in various cells of the ovary. The advantage of this approach over that of previous studies using labeled ligand binding (23, 24) is that the higher intensity of the signal with immunohistochemistry makes the localization much less ambiguous.



Pregnant and non-pregnant, sexually mature, F344 (Fisher) rats, were obtained from the Harlan SpragueDawley Animal Facility in Fredrick. MD. Animals were killed by CO, asphyxiation. Tissues were immediately excised and fixed in Bouin‘s fixative, with agitation, for 2.5 h at room temperature. They were sequentially washed free of picric acid in 90% ethanol. 70% ethanol, 30% ethanol, and water for 1 h each. They were then post-fixed in phosphate buffered 10% formal in before routine paraffin processing. Fetuses of pregnant animals were checked for stage of development to insure proper dating of pregnancy. Sections (5 pu) were cut and

Fig. 1 Identification of anti-PGF,-R immunopositive rat ovarian ceils. Ovaries from non-pregnant and pregnant rats were collected, fixed, sectioned and stained all as described in Materials and Methods. Preimmune (B, D, F, H, J, L) and immune (A, C, E, G, I, K) stained sections of non-pregnant (A, B, E-L) and day 18 pregnant (C, D) ovaries are shown. Magnification of sections A-D is approximately 65X and of sections E-L is aproximately 650X. Abbreviations: A, cl = corpus luteum, C, cl = corpus luteum, af = atretic follicle, is = interstitial tissue; E, S = overlying serosa, ce = coelomic epithelium g = granulosa cells; G, v = vein, a = arteriole; I, ms = medullary stromal, g = granulosa cells.

Fig. 2 Immunohistochemical staining of PGF,-R in corpora lutea of non-pregnant, pregnant and postpartum rats. Ovaries from non-pregnant, pregnant and postpartum rats were collected, fixed, sectioned and stained all as described in Materials and Methods. Preimmune (B, D, F, H) and immune (A, C, E, G) stained sections of non-pregnant (A, B), day 13 pregnant (C, D), day 18 pregnant (E, F) and 4 day postpartum (G, H) ovaries are shown. Magnification of sections A-H is approximately 650X.




and Essential Fatty Acids

stained by the peroxidase-antiperoxidase method of Stemberger (25) using a 1:200 dilution of preimmune or immune sera (22), 1:60 dilution of swine antirabbit immunoglobulins (Dako Corporation, Carpinteria, CA) and a 1: 100 dilution of rabbit peroxidase-antiperoxidase conjugate (ICN Immunobiologicals, Lisle, IL). The staining was developed with diaminobenzidine and H202, then the section was very lightly counterstained with hematoxylin. Prior to studies on immunohistochemical localization of the PGF,,-R. the method of fixation was optimized. Four different fixatives were examined including: 10% phosphate buffered formalin, 1% phosphate buffered glutaraldehyde, Bouin’s, and periodate-lysineparaformaldehyde. Use of Bouin’s provided the strongest signal and least non-specific staining.

RESULTS Figure 1 shows PGF,,-R immunohistochemical staining of the non-pregnant and day 18 pregnant ovary. In A and B are sections of non-pregnant (normally cycling) ovary and in C and D are sections of late pregnant ovary (day 18; normal gestation time is 2 l-22 days) all presented at low magnification. At this magnification it is seen that a subset of cells within the corpora lutea are positive (brown) as are a type of cell in both primary and secondary interstitial tissues. The atretic follicle seen in Figure 1C was judged so by the presence of pyknotic nuclei (26. 27). In the sections stained with preimmune serum (B, D). only incompletely blocked endogenous peroxidase in red blood cells was detected. Sections E and F show the non-pregnant serosa surrounding the ovary. a vessel in that serosa. the ovarian coelomic epithelium, cortical stroma beneath the coelomic epithelium and a portion of a developing follicle in the cortex of the ovary. In these sections only a subpopulation of thecal cells (theta intemae) are seen to stain for the presence of PGF>,-R. The developing follicle’s granulosa cells and stroma outside the theta do not appear to contain PGF,,-R. Sections G and H are from the non-pregnant central. medullary region of the ovary and show blood vessels and the stroma between these vessels. All of the cells in these sections appear to lack PGF2,-R by this method. Sections I and J are from the medullary area in the non-pregnant ovary and show medullary stroma as well as a portion of a developing follicle. As above, the medullary stroma is immunohistochemically negative while a subpopulation of the theta is positive. Lastly. sections K and L show at higher magnification an area of the non-pregnant ovary which contains secondary interstitial cells. Therefore, primary and secondary interstitial tissue in the ovaries from pregnant and nonpregnant animals is seen to stain for the presence of PGF,,-R (Figure 1C. 1K and 1A, respectively). Figure 2 shows the immunohistochemical staining for PGF,,-R in rat corpora lutea during the non-pregnant.

pregnant and postpartum time periods. Sections A and B are of a non-pregnant rat corpus luteum prior to breakdown (luteolysis is addressed in Fig. 3). In the nonpregnant corpus luteum. luteal cells appear to contain immunoreactive PGF,,-R although in a heterogeneous fashion, and stromal fibroblasts and vessels in between the luteal cells do not contain immunohistochemically reactive PGF?,-R. Sections C and D are of a midpregnant, day 13 corpus luteum. Here the luteal cells appear much more uniformly stained and are much larger. Vessels and stroma still fail to stain for the presence of PGF,,-R during pregnancy. Sections E and F are of a late pregnant, day 18 corpus luteum. Here again the luteal cells stain positive for the presence of PGF,,-R although not as uniformly as the mid-pregnant luteal cells stain. Sections G and H are of a 4-day postpartum corpus luteum. The immunopositive luteal cells seen here are noticeably smaller than during pregnancy. and are now interspersed with larger vessels. Vessel endothelium and stromal fibroblasts do not contain immunoreactive PGF,,-R. Figure 3 details the breakdown (luteolysis) and fibrotic replacement (corpora albicantia formation). rcspectively, of non-pregnant corpora lutea and pregnant corpora lutea. Sections A and B show apoptotic luteal cells possessing the characteristic shrunken morphology and fragmented nuclei with lack of nuclear envelope (28). These cells appear to stain much more intensely for PGF,,-R than do the surrounding morphologically normal luteal cells. The apoptotic luteal cells appear to be randomly distributed throughout the central portion of the corpus luteum beginning luteolysis. Sections C and D are later in the luteolytic sequence and show a central portion of largely fragmented cells interspersed with a few PGF,,-R immunopositive luteal cells. The core is surrounded by a rim of intensely immunopositive cells lying at the edge of the corpus luteum near where the theta internae would have been prior to ovulation. As an internal control to show that all dying cells do not stain immunopositive for PGF?,-R, sections E and F are presented which contain dying granulosa cells in an atretic follicle. In this case both the granulosa cells and dying cell show an absence of PGF,,,-R immunopositivity. In contrast to the apoptosis observed in the breakdown of corpora lutea of non-pregnant animals, the breakdown of corpora lutea associated with pregnancy appears to correlate with a loss of PGFZtx-R immunopositivity without apoptosis. Beginning in late pregnancy (day 1X of pregnancy shown here. sections G and H) the initial phase is fibrotic replacement (or simple loss of PGF?,,R) in the center of a corpus luteum. The resulting tissue appears to have more extracellular connective tissue and a decreased density of nuclei. Neither apoptotic cells nor polymorphonuclear inflammatory cells are apparent. Sections I and J are of a 4-day postpartum corpus luteum to show the extensive fibrotic replacement of the corpus luteum. Here again. the fibrotic appearing central


tissue in the postpartum corpus luteum is negative for PGF,,-R staining. It is of note that, although the central portion of the corpus luteum appears to be replaced by a different cell type, no cells are seen in ‘the process of dying either by apoptosis or pyknosis. Lastly, sections K and L are presented to show at high power this area where the luteal cells are ‘losing’ their PGF*,-R immunopositivity.

DISCUSSION In an attempt to understand the normal physiologic role(s) of PGF2, in the ovary, the PGF,,-R has been purified and a rabbit polyclonal antibody against it produced for use in immunohistochemical studies. It is the intent of the studies described herein to identify which cells in the ovary possess PGF:,-R and could, therefore, respond directly to PGF,,. Those cells which do not possess PGF,,-R and yet are believed to respond to PGF?, may do so through an indirect mechanism. The antibody used in these studies indicates that three main cell subpopulations contain PGF,,-R. These are: a subpopulation of cells in the corpus luteum, a subpopulation of thecal cells surrounding developing (secondary and mature) follicles and a subpopulation of cells in primary and secondary interstitial tissue. The ovarian tissues and cell types which do not stain positively for the presence of PGF,,-R include: the overlying serosa and serosa vessels, the coelomic epithelium and its underlying (cortical) stroma, medullary stroma, medullary vessels, all stages of developing and dying granulosa cells, the oocyte, and corpora luteal vessels and stroma. From these results, a direct effect of PGFza in two of the roles hypothesized for PGF?, is suggested. First of all, PGF,, has been postulated to cause luteolysis either by a reduction of corpora lutea LH receptors (14) inhibition or uncoupling of LH stimulation of CAMP synthesis (15-19) or inhibition of LH stimulation of protein kinase A activity (15, 18). All of these postulated mechanisms of luteolysis involve a target luteal cell which responds to LH. LH receptors are believed to be present primarily on luteal cells derived from theta internae (32, for a review see 33), and possibly on the large luteal cells which come from small (stem) luteal cells (for a review see 33). PGF?,-R was observed on both large and small luteal cell subpopulations suggesting PGF,, may act directly on these LH responsive cells during luteolysis. In an alternative mechanism, PGF,, has been suggested to cause luteolysis of large luteal cells by a direct cytotoxic effect (20). Here, PGF,,-R has been documented on large luteal cells, towards the center of corpora lutea (and probably of granulosa cell origin; for a review see 2), that are undergoing apoptosis. This result is consistent with the possibility of a direct PGF,, cytotoxic effect, but cannot distinguish this cytotoxic mechanism from the above mentioned


of PGF2. Receptor in the Rat Ovary


inhibition/uncoupling LH mechanism. It is of note, however, that the non-pregnant corpora lutea show a subpopulation of luteal cells, towards the outer rim of the corpora lutea (and possibly of theta intemae origin; for a review see 2), which are also immunopositive for PGF*,-R yet which are not undergoing apoptosis. Besides these two direct effects of PGF,,, two indirect effects on ovarian physiology are suggested by data presented here. First, PGF,, is hypothesized to be important for the induction of ovulation possibly by inducing follicular contractility (for a review see 3,10). Although PGF,,-R was observed to be present in the theta, it was in the theta intemae, not the theta extemae where the smooth muscle which surrounds mature rat follicles is believed to reside (29). This suggests that the PGF?, induced follicular contraction is an indirect effect. Second, PGF?, is believed to play a role in ovarian vascular constriction resulting in luteolysis (1 l13). Although we have documented PGF,,-R vasoimmunoreactivity specifically in lung bronchial artery smooth muscle (30) (suggesting this antibody is able to recognize vascular muscle PGF,,-R), we have been unable to detect PGF1,-R in the vasculature of the ovary. This result suggests that PGFza causes ovarian vascular constriction through an indirect mechanism. Altematively, PGF?, may not induce a decrease in rat ovarian blood flow as it does in other species following injection of luteolytic doses of PGFza (3 1). The presence of PGF2,-R on cells of interstitial tissue has also been documented. Interstitial tissue is believed to be important in normal ovarian physiology due to its role in ovarian androgen/estrogen synthesis and follicular atresia (for review see 34, 35). More specifically, this subpopulation of cells is active in the conversion of pregnenolone to dehydroepiandrosterone and androstenedione (35). The presence of PGF?,-R on interstitial tissue as well as on granulosalutein cells and theta intemae cells (which also are important in metabolism/ secretion of steroid hormones) may suggest a common role of PGFza in modulation of steroid synthesis. With regard to the subcellular localization of PGF,,R, light microscopy reveals the presence of the antigen in the cytoplasm and on the plasma membrane. Further characterization will require immunohistochemical electron microscopy studies which are presently ongoing. A potential problem of the results presented here would occur with the presence of a second class of PGF>,-R which is non-immunoreactive with the antibody used here. We have no results which suggest PGF?,-R in addition to those which this antibody recognizes, however, work continues to explore this possibility. In summary, a polyclonal anti-PGF,,-R antiserum has been used to immunohistochemically localize which cells in the ovary have PGF,,-R. These results are a necessary first step in attempting to analyze the hypothesized roles for PGFza in the ovary. These results also suggest that this antiserum is a valuable tool with which to study PGF?, function in vivo.

Fig. 3 Luteolysis and fibrotic replacement of non-pregnant and pregnant corpora lutea, respectively. Ovaries from non-pregnant, pregnant and postpartum rats were collected, fixed, sectioned and stained all as described in Materials and Methods. Preimmune (B, D. F, H, J, L) and immune (A, C, E, G, I, K) stained sections of non-pregnant (A-F), day 18 pregnant (G and H) and 4 day-postpartum (I-L) ovaries are shown. Magnification of sections, A. B, E, F, K, L is approximately 1300X and of sections C, D, G, H, I, J is approximately 325X. Abbreviations: A, arrows point to apoptotic luteal cells; E, arrow points to apoptottc granulosa cell; I, fc = fibrotic center.


Acknowledgements This work was supported by grants UCHSC BRSG-05357, UCHSC ACS #29, Milheim Cancer Foundation Grant #86-52, and NIH Grant HD25961-01, all to DJO and the Histopathology Core Laboratory of the University of Colorado Cancer Center supported by NC1 grant 2P30 CA46934-04. We also thank C. Mraz and N. Hart for their expert secretarial assistance and P. Shanley for editorial assistance.

References 1. Horton E W, Poyser N L. Uterine luteolytic hormone: a physiologic role for prostaglandin Fzu Physiological Reviews 56: 595-651.1976. 2. Nisewender G D, Nett T M. The corpus lutem and its control. p, 489-525 in The Physiology of Reproduction (E Knobil E, J Neil1 et al. eds) Raven Press Ltd. New York, 1988. 3. Lipner H. Mechanism of mammalian ovulation. p. 447488 in The Physiology of Reproduction (E Knobil, J Neil1 et al. eds) Raven Press Ltd. New York, 1988. 4. Freeman M E. The ovarian cycle of the rat. p. 1893-1926 in The Physiology of Reproduction (E Knobil. J Neil1 et al. eds) Raven Press Ltd. New York, 1988. 5. Orczyk G P, Behrman H R. Ovulation blockade by aspirin or indomethacin - in viva evidence for a role of prostaglandin in gonadotrophin secretion. Prostaglandins I : 3-20, 1972. 6. Mori T. Kohda H, Kinoshita Y, Ezaki T, Morimoto N, Nishimura T. Inhibition by indomethacin of ovulation induced by human chorionic gonadotrophin in immature rats primed with pregnant mare serum gonadotrophin. Journal of Endocrinology 84: 333-341, 1980. 7. Armstrong D T, Grinwich D L, Moon Y S. Zamecnik J. Inhibition of ovulation in rabbits by intrafolhcular injection of indomethacin and prostaglandin F antiserum. Life Sciences 14: 129-140. 1974. 8. Hamada Y, Wright K H, Wallach E E. In vitro reversal of indomethacin-blocked ovulation by prostaglandin Flu. Fertility and Sterility 30: 702-706. 1978. 9. Wallach E E, Bronson R. Hamada Y, Wright K H. Stevens V C. Effectiveness of prostaglandin Fza in restoration of HMG-HCG induced ovulation in indomethacin-treated rhesus monkeys. Prostaglandins IO: 129-138, 1975. 10. Jones R E. Orlicky D J. Austin H B. Rand M S, Lopez K H. Indomethacin inhibits ovarian PGE secretion and gonadotropin-induced ovulation in a reptile (Amolis carolinensis). Journal of Experimental Zoology 255: 5762. 1990. I 1. Nett T M, McClellan MC. Niswender G D. Effects of prostaglandins on the bovine corpus luteum: Blood flow. secretion of progesterone and morphology. Biology of Reproduction 15: 6678, 1976. 12. Nisewender G D, Reimers T I. Diekman M A. Nett T M. Blood flow: A mediator of ovarian function. Biology of Reproduction 14: 64-8 I. 1976. 13. Weston P G, Hixon J E. Effects of in vivo prostaglandin Fzu administration on in vitro progesterone synthesis by bovine corpora lutea. Biology of Reproduction 22: 259-268, 1980. 14. Behrman H, Grinwich D H. Hichens M, MacDonald G J. Effect of hypophysectomy. prolastin and prostaglandin Fz,, on gonadotropin binding in vivo and in vitro in the corpus luteum. Endocrinology 103: 349-357. 1978. 15. Khan M I. Rosberg S, Lahav M, Lamprecht S A, Selstam G, Herlitz H. AhrCn K. Studies on the mechanism of action of the inhibitory effect of prostaglandin Fza on cyclic AMP accumulation in rat corpora lutea of various ages. Biology of Reproduction 21: 1175-l 183. 1979. 16. Fletcher P W, Nisewender G D. Effects of PGF,, on progesterone secretion and adenylate cyclase activity in bovine luteal tissue. Prostaglandins 20: 803-818, 1982. 17. Rodgers R J, O’Shea J D. Findlay J K. More studies on granulosa cells. Journal of Reproduction and Fertility 75: 85-94. 1985.



of PGFZ, Receptor in the Rat Ovary

18. Benhaim A. Bonnamy P J, Papadopoulos V, Mittre H, Leymaire P. In vitro action of on progesterone and CAMP synthesis in small bovine luteal cells. Prostaglandins 33: 227-239, 1987. 19. Lahav M, Davis J S, Rennert H. Mechanism of the luteolytic action of prostaglandin Fzn in the rat. Journal of Reproduction and Fertility 37: 233-240. 1989. 20. Silvia W J, Fitz T A. Mayan M H, Nisewender G D. Cellular and molecular mechanisms involved in luteolysis and maternal recognition of pregnancy in the ewe. Animal Reproductive Science 7: 57-74, 1984. 21. Orlicky D J. [3H]Prostaglandin Fza membrane binding reexamined. Prostaglandins, Leukotrienes and Essential Fatty Acids 40: 181-189, 1990. 22. Orlicky D J. Miller G J, Evans R M. Identification and purification of a bovine corpora luteal membrane glycoprotein with [3H]prostaglandin Fza binding properties, Prostaglandins. Leukotrienes and Essential Fatty Acids 41: 51-61. 1990. 23. Hofman G E. Rao Ch V, Barrow G H, Sanfilippo J S. Topography of human uterine prostaglandin E and Flu receptors and their profiles during pathological states. Journal of Clinical Endocrinology and Metabolism 57: 360-366, 1983. 24. Luborsky-Moore J L, Wright K. Behrman H B. Demonstration of luteal cell membrane receptors for prostaglandin Fza by ultrastructural and binding analysis. p. 633638 in Ovarian Follicular and Corpus Luteum Function (R P Channing, J M Marsh, WA Sadler eds) Plenum Press, 1979. 25. Stemberger L A, Hardy P H. Cuculis J I. Meyer H G. The unlabeled antibody enzyme method of immunohistochemistry. Preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. Journal of Histochemistry and Cytochemistry 18: 315-333. 1970. 26. Hirshfield A N. Size-frequency analysis of atresia in cycling rats. Biology of Reproduction 38: 1181-l 188, 1988. 27. Greenwald G S, Terranova P F. Follicular selection and its control. p. 387446 in The Physiology of Reproduction (E Knobil, J Neil1 et al. eds) Raven Press Ltd, New York. 1988. 28. Wyllie A H. Cell death: a new classification separating apoptosis from necrosis. p. l-34 in Cell Death in Biology and Pathology (I D Bowen. R A Lockshin eds) Chapman and Hill, New York, 198 1. 29. Osvaldo-Decima L. Smooth muscle in the ovary of the rat and monkey. Journal of Ultrastructural Research 29: 2 18-237, 1970. 30. Orlicky D J, Williams-Skipp C. Immunohistochemical localization of the PGFza receptor in non-ovarian tissues. Manuscript in preparation. 3 1, Behrman H R, Luborsky-Moore J L, Pang C Y. Wright K. Dorflinger L J. Mechanisms of PGF,, action in functional luteolysis. p. 557-571 in Ovarian Follicuhu and Corpus Luteum Function. Advances in Experimental Medicine and Biology Vol 112 (C P Channing. J Marsh, W A Sadler eds), Plenum Press. New York. 1979. 32. Fitz T A. Mayan M H. Sawyer H R, Niswender G D. Characterization of two steroidogenic cell types in the bovine corpus luteum. Biology of Reproduction 27: 703-7 11, 1982. 33. Weber D M, Fields P A, Romrell L J. Tumwasom S. Ball B A, Drost M. Fields M J. Functional differences between small and large luteal cells of the late-pregnant vs. nonpregnant cow. Biology of Reproduction 37: 685697. 1987. 34. Richardson G S. Ovarian Physiology. New England Journal of Medicine Medical Progress Series. Little, Brown and Company. Boston. 1967. 35. Erickson G F, Magoffin D A, Dyer C A, Hofeditz C. The ovarian androgen producing cells: A review of structure/ function relationships. Endocrine Reviews 6: 37 l-399, 1985.


Immunohistochemical localization of PGF2 alpha receptor in the rat ovary.

As a step towards understanding the role of prostaglandin F2 alpha (PGF2 alpha) in ovarian function, a rabbit antiserum against purified PGF2 alpha re...
11MB Sizes 0 Downloads 0 Views