Immunocytochemical localization of angiotensin II immunoreactivity and demonstration of angiotensin II binding in the rat ovary

Immunocytochemical localization of angiotensin II immunoreactivity and demonstration of angiotensin II binding in the rat ovary

Kim et al. and forms of vitamin B 12 in milk. Am J Clin Nutr 1981;34:1717. 13. Nex!ll E, Hansen M, Poulsen SS, Olsen PS. Characterization and immunoh...

2MB Sizes 0 Downloads 139 Views

Kim et al.

and forms of vitamin B 12 in milk. Am J Clin Nutr 1981;34:1717. 13. Nex!ll E, Hansen M, Poulsen SS, Olsen PS. Characterization and immunohistochemical localization of rat salivary cobalamin-binding proteins, and comparison with human salivary haptocorrin. Biochim Biophys Acta 1985;838: 264. 14. HurlimanJ, Zuben C. Vitamin B 12 binders in human body

August 1988 Am J Obstet Gynecol

fluids. II. Synthesis in vitro. Clin Exp Immunol 1969; 4:141. 15. Oram JD, Reiter B. Inhibition of bacteria by lactoferrin and other iron-chelating agents. Biochim Biophys Acta 1968; 170:351. 16. Mason DY, Taylor CR. Distribution of transferrin, ferritin and lactoferrin in human tissues. J Clin Pathol 1978; 31:316.

Immunocytochemical localization of angiotensin II immunoreactivity and demonstration of angiotensin II binding in the rat ovary Abraham Lightman, MD, MSc," Carla L. Jones, MD, Neil J. MacLusky, PhD, Angela Palumbo, MD," Alan H. DeCherney, MD, and Frederick Naftolin, MD, DPhil

New Haven, Connecticut The cellular localization of angiotensin II immunoreactivity and the presence of angiotensin II binding in rat ovaries were studied. Angiotensin II immunohistochemical staining was demonstrated throughout the corpora lutea of gonadotropin-stimulated immature rats and pseudopregnant adult rats, as well as in some stromal and thecal cells surrounding large antral follicles. No immunostaining was observed in granulosa cells of preantral or antral rat follicles or in ovaries from unstimulated immature rats. With in vitro autoradiography, specific, saralasin-suppressible 125 1-angiotensin II binding was demonstrated in normal cycling rat ovaries: diestrus > proestrus > estrus. The combined findings of angiotensin II immunostaining in ovarian follicles and corpora lutea and of cycle-related angiotensin II binding support the hypothesis of a functional role for the ovarian renin-angiotensin system. (AM J OBSTET GYNECOL 1988;159:526-30.)

Key words: Angiotensin II, ovarian renin-angiotensin system, immunohistochemistry, ovarian peptides The renin-angiotensin system, a well-known regulator of blood pressure and volume homeostasis, is now recognized to have widespread distribution. The presence of extrarenal renin-angiotensin systems has been established in numerous tissues including the brain, adrenal glands, and reproductive organs. 1· 2 Several lines of evidence support the existence of an intrinsic ovarian renin-angiotensin system. High levels of prorenin, 3 renin-like activity, and angiotensin II immunoreactivity4 have been demonstrated in preovula-

From the Department of Obstetrics and Gynecology, Yale University School of Medicine and Yale University Center for Research in Reproductive Biology. Supported by the Andrew W. Mellon Foundation and National Institutes of Health Grant No. HD22970. Received for publication July 27, 1987; revised December 29, 1987; accepted March 5, 1988. Reprint requests: Dr. Frederick Naftolin, Yale University School of Medicine, 333 Cedar St., P. 0. Box 3333, New Haven, CT 06510-8063. "Lalor Foundation Fellow and Bruner Foundation Fellow. 'Lalor Foundation Fellow.

526

tory human follicle fluid from both gonadotropinstimulated and normal cycles. In addition to a reninlike enzyme having been demonstrated in rat luteal tissue, 5 cultured luteal cells have been shown to produce renin, 6 and renin messenger ribonucleic acid (mRNA) has been identified in rat ovaries by northern blot techniques 7 and in rat corpus luteum by in situ hybridization." Furthermore, high levels of angiotensin II immunoreactivity have been found in rat ovaries, 9 ovarian angiotensin II receptors have been identified and characterized, 9 · 10 and angiotensin II has been shown to stimulate estrogen synthesis in vitro in rat ovaries. 10 Because the ovary is a complex structure, with its follicle components, corpus luteum, and stroma developing and interacting throughout the reproductive cycle, we have attempted to determine the presence of angiotensin II and angiotensin II binding in these temporal/ morphologic compartments. In this report we present evidence that angiotensin II immunoreactivity is present in luteal and thecal cells of gonadotropin-stimulated rat ovaries, and that cycle-related fluctuations in ovarian angiotensin II receptors occur in the adult rat.

Angiotensin II immunoreactivity and binding

Volume 159 Number 2

527

Fig. 1. Sections of pregnant mare serum gonadotropin/hCG-stimulated immature rat ovary immunohistochemically stained for angiotensin II. A, Positive staining in corpus luteum cells (arrow). B, Antral follicle with no immunostaining of granulosa (g), theca, or stroma cells. (DE antiserum, I : 3500; original magnification x 1000.)

Material and methods

Tissue preparation for angiotensin II immunohistochemistry. Four groups of Sprague-Dawley rats (Charles River Farms) were studied. In the first group of I 0 immature rats, 5 received 10 IU of pregnant mare serum gonadotropin on day 25 followed by 2 IU of human chorionic gonadotropin (hCG) 48 hours later, and 5 received 4 IU of pregnant mare serum gonadotropin followed by 1.2 IU of hCG. To ensure the formation of functional corpora lutea (pseudopregnancy), uterine cervical stimulation with a glass rod was also performed I2 to 14 hours after the hCG injection. In the second group of IO immature rats, 5 were given 4 IU and 5 were given IO IU of pregnant mare serum gonadotropin on day 25. All rats in the first two groups were killed I2 days after the pregnant mare serum gonadotropin injection (on day 37) and the ovaries were removed and fixed as described below. The third group of five immature rats did not receive any treatment. The fourth group consisted of five regularly cycling adult rats in which cervical stimulation was performed on the morning of estrus (8 AM) to induce pseudopregnancy. The ovaries were removed and fixed 10 days after the cervical stimulation, when pseudopregnancy was evident by daily vaginal smears showing sheets of epithelial cells and leukocytes. The kidneys from adult and immature animals were also removed and fixed to be used as positive controls for the immunostaining. As a further control, in two adult rats the median eminence was dissected and embedded after intracardial perfusion with Bouin's solution. Unless otherwise stated, ovaries and kidneys were immersion fixed in 4% paraformaldehyde (pH 6.5, O.I

mol/L acetate buffer) for 24 hours at 4° C. All tissues were then dehydrated and embedded in paraffin. Serial 30 f.1. sections were cut and placed on slides. Primary angiotensin II antisera. Rabbit antisera against angiotensin II were obtained from two wellestablished sources: One was a gift from Dr. D. Ganten (DE antisera) and the other was from Dr. A. NegroVilar (NV antisera). The characteristics of these two antisera have been described previously"; while being otherwise highly specific, both antisera cross-react completely with des-asp-angiotensin II (angiotensin III). For all our experiments, the DE antiserum was used at a dilution of I : 2500 to I : 5000, while the NV antiserum was diluted I : 1000 to I : 2000. Immunocytochemical staining. The general protocol was that of Hsu et al. 12 with the avidin-biotinperoxidase method (Vectastain ABC Kit, Vector Laboratories, Inc., Burlingame, Calif.). Tissue sections were incubated with the primary angiotensin II antiserum for 24 hours at 4° C. The slides were then washed with O.I mol/L phosphate-buffered saline solution at pH 7.6 and incubated for 30 minutes in a speciesspecific biotinylated antibody solution. The sections were then washed and incubated with the avidinbiotinylated horseradish peroxidase complex for 1 hour, followed by washing and 5 to 10 minutes of incubation in a solution of 3,3' -diaminobenzidine (O.I% in O.I mol/L Tris buffer at pH 7.2) and 0.02% hydrogen peroxide to develop the chromogen. The slides were dehydrated in graded alcohols, cleared in xylene, and mounted. To demonstrate the specificity of immunocytochemical staining, each of the studied ovarian tissues was examined after preabsorption of the antisera with var-

528

August 1988 Am J Obstet Gynecol

Lightman et al.

()

.. )

c

Fig. 2. In vitro autoradiography of 1251-angiotensin II binding on serial sections of ovaries from regularly cycling virginal adult rats (A , diestrus day; B, proestrus day; C, estrus day). The dark circular 125 1-angiotensin II binding in the diestrus ovaries (A 2) was completely prevented by preincubation with saralasin (A 1) . Arrow in A 1 points to one of the unmarked ovarian sections. Note also that the 125 1-angiotensin II binds to the uterine tissue (B, an-ow).

ious concentrations of angiotensin II . The working dilutions (1: 3000 of DE antiserum and 1: 1500 of NV antiserum) were incubated with four doses of synthetic Ile 5 angiotensin II (25, 50, 500, and 1000 J.Lg/ml; Sigma Chemical, St. Louis, Mo.) for 24 hours at 4° C. In addition , as a control for each tissue tested, the primary antiserum was replaced by preimmune rabbit serum, which resulted in total absence of staining. In vitro autoradiography of angiotensin II binding. Two regularly cycling adult rats were studied on the morning of each of the following cycle days: estrus, diestrus, and proestrus. The protocol followed that of Israel et al. 13 After decapitation, the ovaries with the adjacent tube and the upper uterus were removed, mounted on cryostat chucks, and frozen in dry ice. Sections measuring 10 J.Lm were cut and thaw mounted onto subbed (1.0% gelatin, 0.05% chrome alum) glass slides. Sections were stored at -40° C until processing (within a few days). Angiotensin II binding sites were labeled in vitro by incubation of the sections (2 hours at 25° C) with 0 .25 nmol/L 125 I-angiotensin II alone or in the presence of unlabeled competitor (l0- 6 mol/L saralasin; Sigma Chemical) to determine the specificity of binding. After incubation, the slides were washed at 4° C, air dried, and placed in x-ray film cassettes with appropriate 125 1-labeled standards and exposed against Ultrafilm (LKB Industries, Md .) at 4° C for 3 days. After exposure the films were developed in Kodak D 19 developer, rinsed briefly in water, fixed, washed, and dried.

Results

Rat ovaries. A total of 30 rats were studied. All of the pregnant mare serum gonadotropin/hCG/cervical stimulation-treated rats' ovaries contained 5 to 10 corpora lutea and numerous follicles in various stages, as did those from adult cervical-stimulated rats. Very intense granular angiotensin II immunostaining was found in the corpus luteum cells (Fig. 1, A). Although most of the corpus luteum cells stained positively for angiotensin II, intensity varied from cell to cell. Less intense immunostaining (as compared with corpus luteum cells) was found in some of the thecal and stromal cells surrounding the corpus luteum and large antral follicles. Between the high- and low-dose pregnant mare serum gonadotropin/hCG/cervical stimulation treatments, no consistent difference was observed in the location or intensity of staining. There was no staining in the granulosa cells of preantral or antral follicles in any of the four groups of rats studied (Fig. l, .B). Ovaries from immature rats treated only with pregnant mare serum gonadotropin .contained follicles in all stages and usually no corpus luteum. Again, no immunostaining was seen in granulosa cells, but weak staining was present in some of the thecal and stromal cells surrounding large antral follicles. The untreated immature rats' ovaries did not contain activated follicles or corpora lutea and there was no angiotensin II immunostaining in these ovaries. In the immunohistochemical studies, all staining in the rat ovarian tissue was blocked by preabsorption of

Volume 159 Number 2

the primary angiotensin II antiserum with 50 fLg/ml of angiotensin II. As well, no positive staining was found without the use of the immune antiserum. It is important to note that the same results were obtained with both of the angiotensin II antisera, the only difference being in the working titer; whereas the optimal titer of DE antisera was l : 3000 to l :4000, the best results with NV antisera were found in titers of l: 1500 to l: 2000. The rat kidneys and hypothalamus used as positive controls showed, as expected, intense staining in juxtaglomerular cells and in the median eminence, respectively, that was abolished by preabsorption of the antiserum with angiotensin II. Autoradiography. Angiotensin II binding was demonstrated in diestrus (Fig. 2, A) and proestrus (Fig. 2, B) rat ovaries, but not in ovaries from estrus animals (Fig. 2, C). Binding, recognized by blackened silver grain deposition in the film, showed a heavy circular concentration of 125 I-angiotensin II in the cortical area where follicles are found in these virginal rats. Comparison of sections on the same film registering binding from both days showed denser radioactive marking in diestrus than in proestrus ovaries, but no formal quantification was performed. Preincubation with the competitive angiotensin II receptor inhibitor saralasin completely prevented 125 I-angiotensin II binding. As a further control, and in concordance with a previous report, 14 there was also saralasin-suppressible 125 I-angiotensin II binding present in the uterine tissue that had been included as part of the section (Fig. 2, B). Comment

This study demonstrates the presence of intense angiotensin II immunoreactivity in rat corpus luteum cells and of less intense immunostaining for angiotensin II in perifollicular stromal and thecal cells. Several models were used for this demonstration, including untreated immature rats (n = 5), pregnant mare serum gonadotropin-stimulated immature rats (n = 10, pregnant mare serum gonadotropin plus hCG and cervical-stimulated rats (n = 10), and adult rats in estrus receiving cervical stimulation (n = 5). All the positively stained cells were found in ovaries that had been stimulated with exogenous gonadotropins or by induced pseudopregnancy in adult rats. Because ovaries from untreated immature rats did not show staining despite the presence of preantral follicles, and because the development of follicles to the preantral stage is not dependent on gonadotropins, these findings constitute further indication of the stimulatory role for gonadotropins in the ovarian renin-angiotensin system. This role was previously evidenced by the fivefold rise of angiotensin II immunoreactivity and 10- to ISfold rise in renin-like activity in follicle fluid from gonadotropin-stimulated women and by the sharp

Angiotensin II immunoreactivity and binding

529

preovulatory rise of angiotensin II immunoreactivity and renin-like activity in the follicular fluid of regularly cycling women.< Conversely, corpora lutea formed after cervical stimulation displayed intense immunostaining. Thus cervical stimulation, which triggers the release of gonadotropins, prolactin, and oxytocin and is analogous to coitus, has no adverse effect on the ovarian renin-angiotensin system, and perhaps even supports its function in preparation for pregnancy. Neither the DE nor the NV antisera used to localize angiotensin II immunoreactivity shows cross-reactivity with peptides outside the angiotensin family. However, both cross-react completely with angiotensin III, 4% (NV antisera) and <0.1% (DE antisera) with angiotensin I, and about I% with tetradecapeptide renin substrate; therefore, it seems reasonable to assume that the antisera marked the presence of angiotensin II and/ or angiotensin III. 11 Conversely, although the immunostaining was blocked by preincubation of the antiserum with synthetic angiotensin II, it is not possible to determine with immunocytochemistry alone whether the angiotensin II immunoreactivity observed represents true angiotensin II or angiotensin III. The unlikely possibility that the staining results from other peptides of the renin-like activity (e.g., angiotensinogen) should be kept in mind. Chromatographic separations currently underway should resolve these issues. The question has previously been raised whether the angiotensin present in the ovary is produced locally or extracted from the circulation. 4 The rat ovary has been reported to contain renin-like enzymes in corpus luteum cells from the same animal model in which we found angiotensin II immunoreactivity, 5 and cultured rat luteal cells have been shown to secrete RLA. 6 Furthermore, angiotensin-converting enzyme, 15 angiotensinogen mRNA, 16 and renin mRNA 7 have also been reported in the whole rat ovary and we have localized renin mRNA to the corpus luteum by in situ hybridization techniques." Taken together, these data suggest that angiotensin II is generated in ovarian cells by a complete intraovarian renin-angiotensin system. With in vitro autoradiography, we found saralasinsuppressible angiotensin II binding in the rat ovary. The circular appearance of the binding marked on these virgin female rat ovaries indicates that angiotensin II receptors are located on ovarian follicles, in accordance with published data" localizing angiotensin II binding sites on rat granulosa cells. The demonstration of specific angiotensin II receptors on ovarian follicles supports a functional role for angiotensin II in follicle development or function. The finding that ovarian angiotensin II binding, and therefore the ovarian responsiveness to angiotensin II, displays cyclic variation is further indication that ovarian renin-angiotensin system function is closely related to the ovarian cycle. Thus it seems possible that there is a paracrine action

530

Lightman et al.

for ovarian angiotensin II made/concentrated in theca that functions via angiotensin II receptors in the granulosa cell layer. Some of the angiotensin II in the theca layer may also find its way into the follicle fluid and beyond. The original demonstration of gonadotropindependent renin-like activity and angiotensin 11/angiotensin III in ovarian follicle fluid raised the possibility of a functional renin-angiotensin system in the ovary. 4 The present findings of immunohistochemically demonstrable angiotensin II in corpora lutea and follicle cells and the presence of saralasin-suppressible, cycle-related angiotensin II binding in follicles indicate that the ovarian renin-angiotensin system may be a locally active autocrine I paracrine system with effects on specific ovarian components and their cellular contents. We thank Drs. D. Ganten and A. Negro-Vilar for the antisera against angiotensin II, Dr. M. Carcangiu for reviewing the ovarian histologic specimens, and Frances DeGrenier for superb manuscript preparation.

August 1988 Am J Obstet Gynecol

5. 6.

7.

8.

9.

10. 11.

12. 1. 2.

3. 4.

REFERENCES Naruse M, Naruse K, Kurimoto F, et al. Evidence for the existence of des-asp-angiotensin II in human uterine and adrenal tissues. J Clin Endocrinol Metab 1985;61 :480. Deschepper CF, Mellon SH, Cumin F, Baxter JD, Ganong WF. Analysis by immunocytochemistry and in situ hybridization of renin and its mRNA in kidney, testis, adrenal and pituitary of the rat. Proc Nat! Acad Sci USA 1986;83:7552. Glorioso N, Atlas SA, Laragh JH, Jewelewicz R, Sealey JE. Prorenin in high concentrations in human ovarian follicular fluid. Science 1986;23: 1422. Lightman A, Tarlatzis BC, Rzasa PJ, et al. The ovarian renin-angiotensin system: Renin-like activity and angiotensins II/III immunoreactivity in gonadotrophin stim-

ulated and unstimulated human follicular fluid. AMJ OR1987;156:808. Cabrera RR, Guardia DC, DeVito E. A renin-like enzyme in luteal tissue. Mol Cell Endocrinol 1986;47:269. Lightman A, Rzasa PJ, Jones C, et al. The ovarian reninangiotensin system: secretion of renin-like activity by cultured luteal cells [Abstract]. In: Proceedings of the sixtyninth annual meeting of the Endocrine Society. Indianapolis: Endocrine Society, 1987. Kim SJ, Shinjo M, Fukamizu A, et al. Identification of renin and renin messenger RNA sequence in rat ovary and uterus. Biochem Biophys Res Commun 1987;142: 169. Lightman A, Deschepper CF, Mellon SH, Ganong WF, Naftolin F. In situ hybridization identifies renin mRNA in the rat corpus luteum. Gynecol Endocrinol 1987; 1: 227. Husain A, Bumpus FM, De Silva P, Speth RC. Localization of angiotensin II receptors in ovarian follicles and the identification of angiotensin II in rat ovaries. Proc Nat! Acad Sci USA 1987;84:2489. Pucell AG, Bumpus FM, Husain A. Rat ovarian angiotensin II receptors. Characterization and coupling to estrogen secretion. J Bioi Chern 1987;262:7076. Brownfield MS, Reid IA, Ganten D, Ganong WF. Differential distribution immunoreactive angiotensin and angiotensin-converting enzyme in rat brain. Neuroscience 1983;7:1759. Hsu S, Raine L, Fanger H. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques. J Histochem Cytochem 1981;29:577. Israel A, Plunkett LM, Saavedra JM. Quantitative autoradiographic characterization of receptors for angiotensin and other neuropeptides in individual brain nuclei and peripheral tissues from single rats. Cell Mol Neurobiol 1985;5:211. Capponi AM, Catt KJ. Angiotensin II receptors in adrenal cortex and uterus. Binding and activation properties of angiotensin analogues. J Bioi Chern 1979;254:5120. Van Sande ME, Scharpe SL, Neels HM, Van Camp KO. Distribution of angiotensin converting enzyme in human tissues. Clin Chim Acta 1985;147:255. Ohkubo H, Nakayama K, Tanaka T, Nakanishi S. Tissue distribution of rat angiotensinogen mRNA and structural analysis of its heterogeneity. J Bioi Chern 1986;261:319.

STET GYNECOL

13.

14. 15. 16.