Molecular and Cellular Endocrinology, 84 (1992) R7-R13 0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00

MOLCEL

R7

02727

Rapid Paper Immunohistochemical localization of transforming growth factor-p, and -pZ during follicular development in the adult rat ovary Katja J. Teerds a and Jennifer H. Dorrington b aDepartment of Cell Biology and Histology, Faculty of Veterinary Medicine, University of Utrecht, Utrecht, Netherlands, and b Banting & Best Department of Medical Research, lJnioersi@ of Toronto, Toronto, Canada (Received

&J words:

Transforming

growth

factor-p,,

9 December

-&; Granulosa

1991; accepted

17 December

cell; Thecal-interstitial

1991)

cell; Luteal

cell; Follicular

development

Summary The transforming growth factors-p (TGF-/?) affect the metabolic activities of each of the cell types in the ovary. In vitro studies using immature rat ovaries have shown the expression of TGF-/3, and/or TGF-& mRNA in thecal/interstitial cells and in granulosa cells (Mulheron and Schomberg, 1990; Mulheron et al., 1991). To obtain information on the localization of TGF-@, and TGF-& in the rat ovary in vivo, we have examined the immunohistochemical staining using antibodies specific for either TGF-Pi or TGF-&. In the adult ovary the immunostaining for TGF-/3, was intense, whereas the staining for TGF-P, was faint. The pattern of immunostaining for TGF-/?, and TGF-& remained constant in the interstitial cell compartment and was not affected by the stage of the oestrous cycle. Since the interstitium surrounds follicles at all stages of development we conclude that TGF-P is not actively involved in regulating the progression of follicles at discrete stages. At the time of antrum formation in the follicle, intense staining for TGF-/3, was observed in thecal cells. Around the preovulatory stage of development, TGF-P, and TGF-& immunoreactivity was also found in the granulosa cells. In the corpus luteum, intense staining for TGF-P, was found in some areas, whereas other areas were negative. Weak to moderate staining for TGF-& was observed. In conclusion, all major cell types show strong immunostaining for TGF-P, and less intense staining for TGF-&, confirming an autocrine/paracrine role for TGF-Ps in the regulation of ovarian cell growth and function.

Introduction Even though luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are essential

Correspondence to: Katja J. Teerds, PhD, Department of Cell Biology and Histology, Faculty of Veterinary Medicine, University of Utrecht, P.O. Box 80.157, 3508 TD Utrecht, Netherlands. Tel. 31-30-535440; Fax 31-30-516853.

for the completion of follicular development, there are other factors synthesized by individual follicles that are able to regulate the growth and differentiation of the follicular cell populations. It has been proposed that these intrafollicular factors dictate the developmental fate of a particular follicle (Dorrington et al., 1987; Bendell and Dorrington, 1990). The transforming growth factors-p (TGF-/3) are polypeptide growth factors that affect the

R8

metabolic activities of each of the cell types in the ovary, i.e. the thecal/ interstitial cells, granulosa cells and oocyte (Dodson and Schomberg, 1987; Dorrington et al., 1988; Feng et al., 1988; Kim and Schomberg, 1989; Tsafriri et al., 1989; Hernandez et al., 1990). TGF+ stimulates or inhibits the growth of granulosa cells depending on the species and the stage of differentiation of the cells (Skinner et al., 1987; Bendell and Dorrington, 1988; May et al., 1989; Gangrade and May, 1990). For example, TGF-P plus FSH promote DNA synthesis in immature rat granulosa cells (Dorrington et al., 1988), whereas TGF-/3 inhibits the growth of bovine granulosa cells in the presence or absence of EGF (Skinner et al., 1987). In addition to its effects on growth, TGF-@ also has pronounced effects on the differentiation of granulosa cells; it stimulates FSH-induced aromatase activity, increases the number of EGF receptors on the cell surface, augments the induction of LH receptors by FSH, and stimulates progesterone production by rat granulosa cells (Feng et al., 1986; Dodson and Schomberg, 1987; Dorrington et al., 1988). Evidence for the local production of TGF-/3 in the follicle has been found in a number of species. TGF-P is secreted by thecal/interstitial cells of rats, pigs and cows (Skinner et al., 1987; Bendell and Dorrington, 1988; Gangrade and May, 1990; Hernandez et al., 1990), and rat granulosa cells (Kim and Schomberg, 1989; Mulheron and Schomberg, 1990; Bendell and Dorrington, 1991). In the DES-treated immature rat ovary, thecal/ interstitial cells express TGF-P, and TGF-& mRNA but only type 2 is inhibited by LH ~Mulheron et al., 1991). Rat granulosa cells from DES-treated immature rats express TGF-0, but not TGF-0, mRNA, and type 2 mRNA levels are inhibited by FSH (Mulheron and Schomberg, 1990). Clearly, the cells of the ovary have the potential in culture to secrete TGF-P, and/or TGF-/3,. The response of the cells depends upon the combined concentrations of TGF-P, and TGF-& in the vicinity of the receptors on the plasma membranes of these cells. To obtain information on the localization of TGF-P, and TGF-P, in specific areas of the ovary in vivo, we have examined the immunohistochemical staining in serial sec-

tions through the adult rat ovary using antibodies that are highly specific for TGF-/3, or TGF-&. Materials and methods Antibodies. The TGF-& polyclonal antibody (anti-LC-(l-30)) was raised against a synthetic peptide corresponding to the first 30 amino acids of mature TGF-/3, and was purified as described by Flanders et al. (1989). The TGF-P, antibody principally recognizes intracellular TGF-P, at the site of synthesis, and does not show any cross-reactivity with TGF-/3, (Flanders et al., 1989; Thompson et al., 1989). The TGF-& polyclonal antibody was raised against a synthetic peptide corresponding to the first 29 N-terminal amino acid residues of TGF-& as described by Van den Eijnden-Van Raaij et al. (1990). This anti-TGF-P, peptide antiserum does not show any cross-reactivity with TGF-Pi and completely neutralizes the growth inhibitory effect of TGF-& on mink-lung carcinoma (ML-CC164) cells. It is specific for TGF-J?, in several immunological assays, including enzyme-linked immunoabsorbent assays, immunoblotting and immunofluorescent experiments. Biotinylated horse anti-rabbit immunoglobulin G was used as a second antibody and was obtained from Vector Laboratories (Burlingame, CA, USA). Animals. Adult female Wistar rats were used. These rats had delivered a litter 3-4 weeks previously. All animals were killed by CO, asphyxiation without any treatment. Immunohistochemical staining. Ovarian tissue was immersion fixed in 4% paraformaldehyde in PBS for 8-10 h at room temperature, foliowed by post-ovation in Bouin’s solution consisting of 0.9% picric acid, 9% formaldehyde and 5% acetic acid. The tissue was dehydrated and embedded in paraffin. The immunostaining technique was performed on 5 pm sections according to Flanders et al. (1989) with minor modifications. Sections were deparaffinized and endogenous peroxidase was blocked with 1% HZO, in methanol for 30 min. The slides were subsequently washed in 0.01 M Tris-buffered-saline (TBS) (pH 7.4), incubated with 0.1 M glycine in TBS for 30 mitt, and rinsed with TBS. The slides were incubated with hyaluronidase (1 mg/ml, Sigma Chemical Co.,

R9

Poole, Dorset, UK) in Na-acetate buffer, 0.1 M, pH 5.5 for 30 min at 37°C. The slides were subsequently washed with TBS, blocked with 5% normal horse serum for 30 min and then incubated overnight at 4°C with the polyclonal antibody against TGF-/3, (gift from Dr. K.C. Flanders) or the polyclonal antibody against TGF-P, (gift from Dr. A.J.M. Van den Eijnden-Van Raaij) at a dilution of 1: 90 and 1: 50, respectively, in TBS with 1% BSA and 0.2% Tween 20. Dose response studies indicated that these dilutions of the antibodies gave optimal labelling results. Following this incubation the slides were rinsed with TBS and then treated for 60 min with a horse anti-rabbit polyclonal antibody (Vector Labs ABC-peroxidase staining kit Elite) diluted 1: 100 in TBS containing 1% BSA and 0.2% Tween 20. Slides were again washed in TBS and subsequently incubated for at least 60 min with components A and B of the ABC staining kit (Vector Labs). Both components (A and B) were diluted 1: 250 and prepared at least 15 min before addition to the sections to allow the complex to form. Slides were again washed in TBS and bound antibody was visualized with a 0.6 mg/ml solution of 3,3’-diaminobenzidine tetrachloride (Sigma Chemical Co.) in TBS, 0.03% H,O, for 4 min (TGF-/3,) or 10 min (TGF-&). The slides were subsequently stained with haematoxylin. In the controls the TGF-Pi and TGF-P, antibodies were replaced by preimmune serum to check the specificity of these antibodies, or the first antibodies were omitted in order to check the specificity of the second antibody. Results Specificity of the antibodies and immunoreactivity. The specificity of the immunohistochemical

staining for TGF-P, has been demonstrated in several organs of mice (Thompson et al., 1989) and in the thecal-interstitial cells of immature rat ovaries (Hernandez et al., 1990). The specificity of the immunohistochemical staining for TGF-& has been demonstrated by Van den Eijnden-Van Raaij et al. (1990) in P19 embryonal carcinoma cells, and the TGF-& producing African green monkey epithelial cell line BSC-1. In control experiments with ovarian sections in which the

polyclonal antibodies were either omitted from the procedure (data not shown) or when preimmune rabbit serum was employed (Fig. lA1, no staining occurred in the interstitial and follicular tissue. Some nonspecific staining could be detected in the blood vessels and the cytoplasm of the oocytes. Aspects of TGF-j3, and TGF-p, Immunoreactivity in the developing follicle. During the early

stages of follicular development the single layer of flattened granulosa cells surrounding the oocyte in the primordial follicle becomes cuboidal in shape to form the primary follicle. No detectable immunostaining for TGF-/?, and for TGF-P, was found in these primordial and primary follicles (Fig. 1B). In the next developmental stage when the granulosa membrana consists of two or three layers of cells, the interstitial cells of the ovarian stroma restructure to form a layer of theta cells around the periphery of the follicle. At this stage of development some patchy staining for TGF-Pi, but not for TGF-&, was detected in the granulosa cells. TGF-fi, and TGF-& immunoreactivity was not detected in the theta of preantral follicles. At this stage and later stages of follicular development the staining of the cytoplasm of the oocytes was more intense than the background levels, for both TGF-@, and TGF-P, (Fig. lC, D). Positive TGF-P, staining in thecal cells was found at the time of the formation of the follicular antrum (Fig. 10 In these follicles TGF-& staining was very faint or absent in the thecal layer (Fig. 1D). As the follicles reach the preovnlatory stage, the staining for TGF-Pi in the granulosa cells gradually became more uniform and all granulosa cells became positive for TGFpi (Fig. lE, G). D uring the latter stage the granulosa cells located in the vicinity of the basal lamina also became positive for TGF-&, and some faint staining could be detected in the inner layers of granulosa cells (Fig. lF, H). Localization of TGF-p, and TGF-/3, immunoreactivity in the corpus luteum. Immuno-

staining for TGF-Pi and TGF-P, was always present in corpora lutea independent of the stage of development; however, the pattern of staining for the growth factors was different. TGF-P, immunoreactivity was found in many of the large cells of the corpus luteum but in some cells the

- --.

--_

-_

-

_--

Rll

Fig. 2. Immunohistochemical localization of TGF-pi and TGF-& in the corpus luteum. A and B: Sections of corpus luteum incubated with TGF-/3r antibody. Small luteal cells (arrows), large luteal cells (arrowheads). C: Section of corpus luteum incubated with TGF-P, antibody. Magnification A 118 X, B 475 x and C 235 x

staining was considerably less or absent (Fig. 2A, B). In contrast to TGF-P,, TGF-& irnmunoreactivity was homogeneous in the cells of the corpus luteum (Fig. 2C). TGF-p, and TGF-p, irnmunolocalization in the interstitial compartment. The interstitial cells of the adult ovary showed strong immunoreactivity for TGF-/3,, whereas TGF-& staining was faint (Fig. 1). The s t aining pattern of the interstitial cells was constant throughout the ovary and was not affected by the presence of follicles in different stages of development. Discussion

The concepts that have developed the roles of TGF-P in the rat ovary generated for the most part from ovarian cell populations. In the case

concerning have been cultures of of thecal/

interstitial cell cultures the populations of cells were heterogeneous. In both granulosa cell cultures and thecal/ interstitial cell cultures prior exposure to DES, the isolation procedures and/or the culture conditions may have influenced the findings. In this novel study we have established that both TGF-P, and TGF-& are present in vivo in follicles at different stages of development, in corpora lutea and in the interstitium. In the interstitium the immunostaining for TGF-P, was intense, whereas the staining for TGF-P, was faint. It is unlikely that the differences in intensity of immunoreactivity were due to differences in the dilutions of the antibodies, since the conditions were optimized (see Materials and methods). Moreover, under the same conditions, adult rat testes served as a positive control, displaying a strong signal for TGF-& (unpublished data). The presence of two types of

Fig. 1. Immunohistochemical localization of TGF-P, and TGF-/3, in the adult rat ovary. A: Control section incubated with pre-immune serum instead of primary antibody. B, C, E, G: Sections incubated with TGF-& antibody. D, F, H: Sections incubated with TGF-& antibody. A and B show serial sections through a primary follicle (arrows), C and D an early antral follicle (arrows) and E, F, G and H a preovulatory follicle (arrows). Granulosa cells are indicated by arrowheads, thecal cells by arrows and interstitial cells by asterisks. Magnification A-D, G and H 475 X, E and F 97 x

R12

TGF-P in the same cell is not unique for the ovary and has been reported for other tissues (Roberts and Sporn, 1990). The pattern of immunostaining for TGF-Pi and TGF-P, remained constant in the interstitial cell compartment and was not affected by the stage of the oestrous cycle. This is consistent with the lack of response of TGF-@, gene expression to LH treatment (Mulheron et al., 1991). Since the interstitium surrounds follicles at all stages of development we conclude that this TGF-/3 is not actively involved in regulating the progression of follicles at discrete stages. Based on the finding that TGF-P inhibited androgen production in thecal/ interstitial cells (Magoffin et al., 1989; Hernandez et al., 19901 we speculate that the high levels of TGF-/3 in the interstitium may hold androgen production in this compartment in abeyance. Excessive androgen production in the ovary is detrimental to follicular development. At the time of antrum formation in the follicle, intense staining for TGF-Pi was observed in the thecal cells. This confirms previous findings which showed TGF-/3 immunoreactivity in the thecal cells of the immature rat ovary (Hernandez et al., 1990) and the adult mouse ovary (Thompson et al., 1989). Furthermore, cultures of homogeneous populations of thecal cells from bovine (Skinner et al., 1987; Lobb and Dorrington, 1989) and porcine (Gangrade and May, 1990) ovaries synthesize and secrete bioactive TGF-P. The appearance of TGF-P immunoreactivity in granulosa cells of antral follicles and the progressive increase as follicles approached the preovulatory stage is consistent with our recent finding that oestradiol-17fi stimulated the secretion of TGF-P by granolosa cells isolated from normal immature rat ovaries (Bendell and Dorrington, 1991). During the follicular phase of the oestrous cycle FSH induces aromatase activity in granulosa cells which converts androgens to oestradiol-17P (Dorrington et al., 1975). The demonstration that oestradiol-17P regulates the secretion of TGF-P by rat granulosa cells led us to postulate that TGF-P was the ‘missing link’ in the mitogenic action of oestrogen in the rat ovary. The appearance of TGF-/3, and TGF-& during the preovulatory period correlates with the ability of the follicle to synthesize oestrogen, providing ‘in vivo’

support for this hypothesis. The presence of TGF-Pi and TGF-& in the granulosa cells of the preovulatory follicle is also consistent with their proposed functional roles. The growth-promoting actions of FSH on granulosa cell proliferation, a prerequisite for follicular development to the preovulatory stage, are dependent on the presence of TGF-@ (Bendel and Dorrington, 1988; Dorrington et al., 1988). TGF-P also enhanced the ability of FSH to induce aromatase activity, an event which also proceeds during the preovulatory period (Bendell and Dorrington, 1988; Dorrington et al., 1988). Mulheron and Schomberg (1990) were not able to detect TGF-Pi mRNA expression in granulosa cells in vitro. An explanation for this discrepancy may be that these authors used immature DES treated rats in which no preovulatory follicles were present, while we only found significant TGF-P, immunoreactivity in granulosa cells from follicles approaching the preovulatory stage. In the corpus luteum, intense staining for TGF-/3, was found in some areas, whereas other areas were negative. Weak to moderate staining for TGF-/3, was observed. In vitro studies have indicated that TGF-P may be involved in the inhibition of progesterone metabolism in luteal cells by suppressing 20a-hydroxysteroid dehydrogenase activity (Matsuyama et al., 1990). In summary, we have shown that in the rat ovary all major cell types show strong immunostaining for TGF-/3, and less intense staining for TGF-&. The intense staining of TGF-/3, in the interstitium which surrounds follicles at all stages of development suggests that TGF-P, may not be actively involved in regulating the progression of follicles at distinct stages but more likely holds androgen production in abeyance. Moreover, these results confirm that TGF-Ps play an autocrine/paracrine role in the regulation of ovarian cell growth and function. Acknowledgements

The authors thank Dr. KC. Flanders (Laboratory of Chemoprevention, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA) and Dr. A.J.M. Van den Eijnden-Van Raaij (Hubrecht Laboratory, Netherlands Institute for

R13

Developmental Biology, Utrecht, Netherlands) for the TGF-PI and TGF-& polyclonal antibodies. We are grateful for the skillful technical assistance of Mrs. M. de Boer-Brouwer, Mr. J.C. van Oudheusden, Mr. A. Tieleman, and to Mr. A.N. Van Rijn and Mr. R. Scriwanek (Department of Cell Biology, Medical School, University of Utrecht, Utrecht, Netherlands) for preparing the photographs. The research of Dr. Katja Teerds has been made possible by a fellowship of the Royal Netherlands Academy of Arts and Sciences. References Bendell, J.J. and Dorrington, J.H. (1988) Endocrinology 123, 941-948. Bendell, J.J. and Dorrington, J.H. (1990) Endocrinology 127, 533-540. Bendell, J.J. and Dorrington, J.H. (1991) Endocrinology 128, 2663-2665. Dodson, W.C. and Schomberg, D.W. (1987) Endocrinology 120,512-516. Dorrington, J.H., Moon, Y.S. and Armstrong, D.T. (1975) Endocrinology 97, 1328-1331. Dorrington, J.H., Bendell, J.J., Chuma, A. and Lobb, D.K. (1987) J. Steroid Biochem. 27, 405-411. Dorrington J., Chuma, A.V. and Bendell, J.J. (1988) Endocrinology 123, 353-359. Feng, P., Catt, K.J. and Knecht, M. (1986) J. Biol. Chem. 261, 14167-14170. Feng, P., Catt, K.J. and Knecht, M. (1988) Endocrinology 122, 181-188. Flanders, K.C., Thompson, N.L., Cissel, D.S., Van Obberberghen-Schilling, E., Baker, C.C., Kass, M.E., Ellings-

worth, L.R., Roberts, A.B. and Sporn, M.B. (1989) J. Cell Biol. 108, 653-660. Gangrade, B.K. and May, J.V. (1990) Endocrinology 127, 2372-2380. Hernandez, E.R., Hurwitz, A., Payne, D.W., Dharmarajan, A.M., Purchio, A.F. and Adashi, E.Y. (1990) Endocrinology 127, 2804-2811. Kim, I.C. and Schomberg, D.W. (1989) Endocrinology 124, 1345-1351. Lobb, D.K. and Dorrington, J.H. (1989) Growth Factors and the Ovary (Hirshfield, A.N., ed.), pp. 199-203, Plenum Publishing Corporation, New York. Magoffin, D.A., Gancedo, B. and Erickson, G.F. (1989) Endocrinology 125, 1951-1958. Matsuyama, S., Shiota, K. and Takahashi, M. (1990) Endocrinology 127, 1561-1567. May, J.V., Frost, J.P. and Schomberg, D.W. (1989) Endocrinology 123, 168-179. Mulheron, G.W. and Schomberg, D.W. (1990) Endocrinology 126, 1777-1779. Mulheron, G.W., Danielpour, D. and Schomberg, D.W. (1991) Endocrinology 129, 368-374. Roberts, A.B. and Sporn, M.B. (1990) in Peptide Growth Factors and Their Receptors, I (Sporn, M.B. and Roberts, A.B., eds.), pp. 419-472, Springer Verlag, Berlin - Heidelberg - New York. Skinner, M.K., Keski-Ojs, J., Osteen, K.G. and Moses, H.L. (1987) Endocrinology 121, 786-792. Thompson, N.L., Flanders, K.C., Smith, J.M., Ellingsworth, L.R., Roberts, A.B. and Sporn, M.B. (1989) J. Cell Biol. 108, 661-669. Tsafriri, A., Vale, W. and Hsueh, A.J.W. (1989) Endocrinology 125, 1857-1862. Van den Eijnden-Van Raaij, A.J.M., Koornneef, I., Slager, H.C., Mummery, C.L. and Van Zoelen, E.J.J. (1990) J. Immunol. Methods 133, 107-118