Localization of Inhibin/Activin Subunit mRNAs within the Primate Ovary
Ralph H. Schwall, Anthony J. Mason, Josiah N. Wilcox, Steven G. Bassett, and Anthony J. Zeleznik Departments of Pharmacological Sciences (R.H.S.), Developmental Biology (A.J.M.), and Cardiovascular Research (J.N.W.) Genentech, Inc. South San Francisco, California 94080 Departments of Physiology and Obstetrics/Gynecology (S.G.B., A.J.Z.) University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania 15213
with each other to form inhibin versus activin are not understood. The ovary is the principle source of circulating inhibin, as evidenced by the rapid loss of immunoactivity from the blood following ovariectomy (4). The ability of cultured granulosa cells to secrete inhibin bioactivity (5) suggests that within the ovary, granulosa cells produce inhibin. In agreement with these results, in situ hybridization studies have localized a, /3A. and /3B subunit mRNAs within the granulosa layer of developing follicles in the rat ovary (6,7). Similar results have been obtained in bovine follicles (8), although the B form of inhibin («/3B) has not yet been identified in bovine follicular fluid (see Ref. 3). The only data available concerning inhibin/activin expression in the primate gonad are those of Davis et al., who demonstrated, by Northern analysis, that a and /3A subunit mRNAs are present in the human corpus luteum (9). This finding was supported by the extraction of inhibin immunoactivity and bioactivity from this tissue (9). In the present report we describe the cellular localization of inhibin/activin subunit expression in the monkey ovary, as determined by in situ hybridization. The results indicate that the primate ovary differs from the rat and cow ovary in that small antral follilces in the monkey express mRNA for the /?B subunit in the absence of a or /?A subunit mRNA. This is the first example in which one of the 0 subunits has been found to be selectively expressed within the ovary, suggesting that the two /3subunits are differentially regulated. Moreover, these data suggest that activin B, the /3B-/3B dimer, may be important in follicular development and selection in the primate ovary.
In order to gain further understanding of the physiology of inhibin and activin in the primate, the expression of inhibin/activin subunit mRNAs in the monkey ovary was examined by in situ hybridization. Granulosa cells of small antral follicles were found to express mRNA for the 0B subunit, which decreased to undetectable levels in dominant follicles. In contrast, expression of a and /?A subunit mRNAs was detected in granulosa cells of dominant follicles and in corpora lutea, but not in small antral follicles. These results indicate that the expression of the /?A and £B subunits is differentially regulated during the growth and development of ovarian follicles in the monkey. (Molecular Endocrinology 4: 75-79, 1990)
INTRODUCTION
Normal ovarian function is dependent upon follicle-stimulating hormone (FSH) and luteinizing hormone (LH), and the secretion of the gonadotropic hormones is in turn regulated by a complex interaction of factors from the hypothalamus and gonad. Two of these factors are inhibin and activin, which are unique in their ability to preferentially alter the secretion of FSH (1-3). In addition, inhibin and activin are functional antagonists; inhibin suppresses, whereas activin stimulates FSH secretion (1-3). Interestingly, inhibin and activin are structurally related; inhibin is a 32K dimer containing a-and /3subunits, while activin is a 24K dimer of inhibin /3subunits (1 -3). There are two forms of the /3-subunit, which give rise to inhibin A (a-j8A) and inhibin B (a-/3B) as well as activin A (j8A-/?A) and activin AB (/3A-/3B). The factors that regulate the association of the subunits
RESULTS
0888-8809/90/0075-0079$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
The ovaries used in these studies were collected from fertile, healthy monkeys whose reproductive histories 75
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 June 2015. at 15:16 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 76
Vol 4 No. 1
were well documented. The circulating steroid levels at the time of ovariectomy are summarized in Table 1. In addition, Table 1 contains an estimate of the number of days between the day of ovariectomy and the day of ovulation, based on steroid levels and the animal's reproductive history. Ovaries were also obtained from two sexually immature monkeys. The results are described below and are summarized in Table 2.
Table 1. Reproductive Status of Monkeys at Ovariectomy Follicular phase Animal
Serum Estradiol
Follicle Diameter
Estimated Days before Ovulation
1711
75 pg/ml
5-6
5-6
105pg/ml
mm 8-9 mm
3-4
1602
Immature Follicles No hybridization with any of the riboprobes was detected in primordial or preantral follicles. Small antral follicles failed to hybridize with the a or 0A specific probes, but hybridized intensely with the a or /?A specific probe (Fig. 1, A-C). Hybridization with the /?B probe was distributed throughout the granulosa layer, including the cumulos oophorus (not shown), and was limited to the granulosa layer. This pattern was observed in small antral follicles in all ovaries examined, independent of whether they were collected during the follicular or luteal phase of the menstrual cycle. In addition, the same pattern of hybridization was observed in small antral follicles within each nondominant ovary (the ovary contralateral to the ovary containing the dominant follicle or corpus luteum), and in each of the ovaries collected from sexually immature monkeys. Hybridization was not detected in small antral follicles that exhibited gross morphological evidence of atresia.
Luteal Phase Animal
1885 1898 1894 1895
Serum Progesterone
1.8 1.9 3.7 3.5
Day of Luteal Phase
ng/ml ng/ml ng/ml ng/ml
Dominant Follicles
5 5 11 12
Table 2. Summary of Hybridization of Subunit Specific Probes in Monkey Ovaries Intensity of Hybridization «
/3A
Follicular Phase Ovaries Primordial Follicles Preantral Follicles Small Antral Follicles Dominant Follicles ++ ++ Luteal Phase Ovaries Primordial Follicles Preantral Follicles Small Antral Follicles Corpora Lutea ++ ++ Nondominant Ovaries Primordial Follicles Preantral Follicles Small Antral Follicles Ovaries from Sexually Immature Animals Primordial Follicles Preantral Follicles Small Antral Follicles -
/3B
+++ +/-" +++ +/-" +++ +++
Hybridization was scored on a scale o f " - " (no silver grains detected above background) to " + + + " (highest density of silver grains). a One follicle (monkey 1711) hybridized weakly, while the other (monkey no. 1602) displayed no hybridization above background. "Corpora lutea were negative except for a few regions in some sections in which hybridization appeared to be slightly greater than background.
Two pairs of ovaries were collected from animals during the mid to late follicular phase of the menstrual cycle. One ovary of each pair contained a large antral follicle which, based on its cellular morphology as well as peripheral serum estrogen concentrations at the time of ovariectomy, was presumed to be the dominant follicle destined for ovulation. In one of these follicles (monkey #1602) granulosa cells exhibited hybridization with the a and /3A probes (Fig. 1, D and E) but not with the 0B probe (Fig. 1F). The failure of the probe to hybridize was not due to a technical problem because small antral follicles within the same section exhibited the strong 0B hybridization described above. The second dominant follicle (monkey #1702), which was smaller and predicted to be a few days farther from ovulation (Table 1), hybridized to all 3 riboprobes, although the intensity of hybridization with the /3B probe was much lower than in small antral follicles within the same section. In both of the dominant follicles, specific hybridization was limited to the granulosa layer. The /3A hybridization in both dominant follicles was focal in nature, with small areas of intense hybridization interspersed with regions of little hybridization (Fig. 1E). The reason for the punctate pattern is not known, but raises the possibility that there may be heterogeneity in the a:PA ratio in different granulosa cells. Corpora Lutea Specific hybridization of the a and /?A probes was observed in all four corpora lutea examined (Fig. 1, G and H). Hybridization of the a riboprobe was distributed evenly throughout the luteal tissue and its intensity was similar to that observed in the dominant follicle. Hybridization of the /3A probe within the corpus luteum was focal (Fig. 1H), as it was in the dominant follicle. There were regions in some corpora lutea in which hybridiza-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 June 2015. at 15:16 For personal use only. No other uses without permission. . All rights reserved.
77
Inhibin/Activin Expression in the Primate Ovary
Fig. 1. Hybridization of Inhibin/Activin Subunits within the Primate Ovary Shown are representative sections hybridized with each of the subunit specific riboprobes. Photomicrographs were taken under epiluminescence, in which silver grains appear white. A-C, Small antral follicle hybridized with riboprobes for« (A), (3/k (B), and /3B (C) subunits (350 x) D-F, portion of the wall of preovulatory follicle from monkey no. 1602 hybridized with riboprobes for a (D), /3A (E), and j3B (F) subunits. Antrum (a) is at the top of each panel, granulosa layer (g) is in the center, and thecal layer (t) is at the bottom (850 x). G-l, Corpus luteum hybridized with riboprobes for a (H), 0A (I), and /3B (J) subunits. Luteal tissue (lu) is on the left and ovarian stromas is on the right (850 x).
tion with the /3B probe appeared to be slightly above the back-ground, but in most luteal tissues no hybridization was detected (Figure 11), even though small antral follicles within the same sections hybridized intensely.
DISCUSSION
The results presented here demonstrate, in the monkey ovary, that none of the three inhibin/activin subunit mRNAs is expressed at detectable levels in primordial and preantral follicles, that small antral follicles express j8B but not a or 0A mRNA, and that, conversely, dominant follicles and corpora lutea express a and /3A mRNAs but little if any |8B mRNA. The appearance of the a subunit mRNA in the dominant follicles and corpora lutea appears to coincide with the elevation in circulating levels of inhibin during the mid to late follicular through the luteal phase of the menstrual cycle
(10). Our finding that the monkey corpus luteum expresses a and j8A mRNAs is consistent with results obtained by Northern analysis of human corpora lutea (9). However, there appear to be species differences since expression of inhibin/activin subunits has not been detected in the corpus luteum of the rat (6,7), cow (8,11) or sheep (11). The expression of inhibin/activin subunit mRNAs in the monkey ovary differs from that in other species in several additional ways. In the rat ovary, « subunit expression has been detected in follicles as small as the primordial stage (7), whereas we were only able to detect a subunit expression in dominant follicles. In addition, a-subunit mRNA and immunoactivity has been detected in theca interna and interstitial cells in the rat, although only during late proestrus and early estrus (7). Similarly, large antral follicles in the cow express a subunit in the theca, as well as granulosa layer (8). We have not observed a subunit mRNA in theca cells or interstitial tissue in the monkey ovary, but we cannot exclude the possibility that such expression may occur transiently near the time of ovulation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 June 2015. at 15:16 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 78
The most striking difference between the monkey ovary and that of other species is that in the monkey, small antral follicles express I3B mRNA in the absence of any detectable a or /3A mRNA. To our knowledge this is the first example in which selective expression of one of the /3 subunits has been detected in the ovary. Moreover, the expression of /3B mRNA was markedly diminished in dominant follicles. The reduction in /3B expression during follicular development may explain why the /3B subunit has not been identified in the bovine ovary, even though its coding sequence is contained within the bovine genome (see Ref. 3). Bovine inhibin was purified from folicular fluid of large antral follicles (12), and the bovine ovarian cDNA libraries that have been used in cloning have been prepared from granulosa cells of large antral follicles (13). We would predict that if the cow is similar to the monkey, such large follicles would express little /3B subunit and that the small antral follicles that express /?B would be too small (
Ralph H. Schwall, Anthony J. Mason, Josiah N. Wilcox, Steven G. Bassett, and Anthony J. Zeleznik Departments of Pharmacological Sciences (R.H.S.), Developmental Biology (A.J.M.), and Cardiovascular Research (J.N.W.) Genentech, Inc. South San Francisco, California 94080 Departments of Physiology and Obstetrics/Gynecology (S.G.B., A.J.Z.) University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania 15213
with each other to form inhibin versus activin are not understood. The ovary is the principle source of circulating inhibin, as evidenced by the rapid loss of immunoactivity from the blood following ovariectomy (4). The ability of cultured granulosa cells to secrete inhibin bioactivity (5) suggests that within the ovary, granulosa cells produce inhibin. In agreement with these results, in situ hybridization studies have localized a, /3A. and /3B subunit mRNAs within the granulosa layer of developing follicles in the rat ovary (6,7). Similar results have been obtained in bovine follicles (8), although the B form of inhibin («/3B) has not yet been identified in bovine follicular fluid (see Ref. 3). The only data available concerning inhibin/activin expression in the primate gonad are those of Davis et al., who demonstrated, by Northern analysis, that a and /3A subunit mRNAs are present in the human corpus luteum (9). This finding was supported by the extraction of inhibin immunoactivity and bioactivity from this tissue (9). In the present report we describe the cellular localization of inhibin/activin subunit expression in the monkey ovary, as determined by in situ hybridization. The results indicate that the primate ovary differs from the rat and cow ovary in that small antral follilces in the monkey express mRNA for the /?B subunit in the absence of a or /?A subunit mRNA. This is the first example in which one of the 0 subunits has been found to be selectively expressed within the ovary, suggesting that the two /3subunits are differentially regulated. Moreover, these data suggest that activin B, the /3B-/3B dimer, may be important in follicular development and selection in the primate ovary.
In order to gain further understanding of the physiology of inhibin and activin in the primate, the expression of inhibin/activin subunit mRNAs in the monkey ovary was examined by in situ hybridization. Granulosa cells of small antral follicles were found to express mRNA for the 0B subunit, which decreased to undetectable levels in dominant follicles. In contrast, expression of a and /?A subunit mRNAs was detected in granulosa cells of dominant follicles and in corpora lutea, but not in small antral follicles. These results indicate that the expression of the /?A and £B subunits is differentially regulated during the growth and development of ovarian follicles in the monkey. (Molecular Endocrinology 4: 75-79, 1990)
INTRODUCTION
Normal ovarian function is dependent upon follicle-stimulating hormone (FSH) and luteinizing hormone (LH), and the secretion of the gonadotropic hormones is in turn regulated by a complex interaction of factors from the hypothalamus and gonad. Two of these factors are inhibin and activin, which are unique in their ability to preferentially alter the secretion of FSH (1-3). In addition, inhibin and activin are functional antagonists; inhibin suppresses, whereas activin stimulates FSH secretion (1-3). Interestingly, inhibin and activin are structurally related; inhibin is a 32K dimer containing a-and /3subunits, while activin is a 24K dimer of inhibin /3subunits (1 -3). There are two forms of the /3-subunit, which give rise to inhibin A (a-j8A) and inhibin B (a-/3B) as well as activin A (j8A-/?A) and activin AB (/3A-/3B). The factors that regulate the association of the subunits
RESULTS
0888-8809/90/0075-0079$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
The ovaries used in these studies were collected from fertile, healthy monkeys whose reproductive histories 75
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 June 2015. at 15:16 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 76
Vol 4 No. 1
were well documented. The circulating steroid levels at the time of ovariectomy are summarized in Table 1. In addition, Table 1 contains an estimate of the number of days between the day of ovariectomy and the day of ovulation, based on steroid levels and the animal's reproductive history. Ovaries were also obtained from two sexually immature monkeys. The results are described below and are summarized in Table 2.
Table 1. Reproductive Status of Monkeys at Ovariectomy Follicular phase Animal
Serum Estradiol
Follicle Diameter
Estimated Days before Ovulation
1711
75 pg/ml
5-6
5-6
105pg/ml
mm 8-9 mm
3-4
1602
Immature Follicles No hybridization with any of the riboprobes was detected in primordial or preantral follicles. Small antral follicles failed to hybridize with the a or 0A specific probes, but hybridized intensely with the a or /?A specific probe (Fig. 1, A-C). Hybridization with the /?B probe was distributed throughout the granulosa layer, including the cumulos oophorus (not shown), and was limited to the granulosa layer. This pattern was observed in small antral follicles in all ovaries examined, independent of whether they were collected during the follicular or luteal phase of the menstrual cycle. In addition, the same pattern of hybridization was observed in small antral follicles within each nondominant ovary (the ovary contralateral to the ovary containing the dominant follicle or corpus luteum), and in each of the ovaries collected from sexually immature monkeys. Hybridization was not detected in small antral follicles that exhibited gross morphological evidence of atresia.
Luteal Phase Animal
1885 1898 1894 1895
Serum Progesterone
1.8 1.9 3.7 3.5
Day of Luteal Phase
ng/ml ng/ml ng/ml ng/ml
Dominant Follicles
5 5 11 12
Table 2. Summary of Hybridization of Subunit Specific Probes in Monkey Ovaries Intensity of Hybridization «
/3A
Follicular Phase Ovaries Primordial Follicles Preantral Follicles Small Antral Follicles Dominant Follicles ++ ++ Luteal Phase Ovaries Primordial Follicles Preantral Follicles Small Antral Follicles Corpora Lutea ++ ++ Nondominant Ovaries Primordial Follicles Preantral Follicles Small Antral Follicles Ovaries from Sexually Immature Animals Primordial Follicles Preantral Follicles Small Antral Follicles -
/3B
+++ +/-" +++ +/-" +++ +++
Hybridization was scored on a scale o f " - " (no silver grains detected above background) to " + + + " (highest density of silver grains). a One follicle (monkey 1711) hybridized weakly, while the other (monkey no. 1602) displayed no hybridization above background. "Corpora lutea were negative except for a few regions in some sections in which hybridization appeared to be slightly greater than background.
Two pairs of ovaries were collected from animals during the mid to late follicular phase of the menstrual cycle. One ovary of each pair contained a large antral follicle which, based on its cellular morphology as well as peripheral serum estrogen concentrations at the time of ovariectomy, was presumed to be the dominant follicle destined for ovulation. In one of these follicles (monkey #1602) granulosa cells exhibited hybridization with the a and /3A probes (Fig. 1, D and E) but not with the 0B probe (Fig. 1F). The failure of the probe to hybridize was not due to a technical problem because small antral follicles within the same section exhibited the strong 0B hybridization described above. The second dominant follicle (monkey #1702), which was smaller and predicted to be a few days farther from ovulation (Table 1), hybridized to all 3 riboprobes, although the intensity of hybridization with the /3B probe was much lower than in small antral follicles within the same section. In both of the dominant follicles, specific hybridization was limited to the granulosa layer. The /3A hybridization in both dominant follicles was focal in nature, with small areas of intense hybridization interspersed with regions of little hybridization (Fig. 1E). The reason for the punctate pattern is not known, but raises the possibility that there may be heterogeneity in the a:PA ratio in different granulosa cells. Corpora Lutea Specific hybridization of the a and /?A probes was observed in all four corpora lutea examined (Fig. 1, G and H). Hybridization of the a riboprobe was distributed evenly throughout the luteal tissue and its intensity was similar to that observed in the dominant follicle. Hybridization of the /3A probe within the corpus luteum was focal (Fig. 1H), as it was in the dominant follicle. There were regions in some corpora lutea in which hybridiza-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 June 2015. at 15:16 For personal use only. No other uses without permission. . All rights reserved.
77
Inhibin/Activin Expression in the Primate Ovary
Fig. 1. Hybridization of Inhibin/Activin Subunits within the Primate Ovary Shown are representative sections hybridized with each of the subunit specific riboprobes. Photomicrographs were taken under epiluminescence, in which silver grains appear white. A-C, Small antral follicle hybridized with riboprobes for« (A), (3/k (B), and /3B (C) subunits (350 x) D-F, portion of the wall of preovulatory follicle from monkey no. 1602 hybridized with riboprobes for a (D), /3A (E), and j3B (F) subunits. Antrum (a) is at the top of each panel, granulosa layer (g) is in the center, and thecal layer (t) is at the bottom (850 x). G-l, Corpus luteum hybridized with riboprobes for a (H), 0A (I), and /3B (J) subunits. Luteal tissue (lu) is on the left and ovarian stromas is on the right (850 x).
tion with the /3B probe appeared to be slightly above the back-ground, but in most luteal tissues no hybridization was detected (Figure 11), even though small antral follicles within the same sections hybridized intensely.
DISCUSSION
The results presented here demonstrate, in the monkey ovary, that none of the three inhibin/activin subunit mRNAs is expressed at detectable levels in primordial and preantral follicles, that small antral follicles express j8B but not a or 0A mRNA, and that, conversely, dominant follicles and corpora lutea express a and /3A mRNAs but little if any |8B mRNA. The appearance of the a subunit mRNA in the dominant follicles and corpora lutea appears to coincide with the elevation in circulating levels of inhibin during the mid to late follicular through the luteal phase of the menstrual cycle
(10). Our finding that the monkey corpus luteum expresses a and j8A mRNAs is consistent with results obtained by Northern analysis of human corpora lutea (9). However, there appear to be species differences since expression of inhibin/activin subunits has not been detected in the corpus luteum of the rat (6,7), cow (8,11) or sheep (11). The expression of inhibin/activin subunit mRNAs in the monkey ovary differs from that in other species in several additional ways. In the rat ovary, « subunit expression has been detected in follicles as small as the primordial stage (7), whereas we were only able to detect a subunit expression in dominant follicles. In addition, a-subunit mRNA and immunoactivity has been detected in theca interna and interstitial cells in the rat, although only during late proestrus and early estrus (7). Similarly, large antral follicles in the cow express a subunit in the theca, as well as granulosa layer (8). We have not observed a subunit mRNA in theca cells or interstitial tissue in the monkey ovary, but we cannot exclude the possibility that such expression may occur transiently near the time of ovulation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 June 2015. at 15:16 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 78
The most striking difference between the monkey ovary and that of other species is that in the monkey, small antral follicles express I3B mRNA in the absence of any detectable a or /3A mRNA. To our knowledge this is the first example in which selective expression of one of the /3 subunits has been detected in the ovary. Moreover, the expression of /3B mRNA was markedly diminished in dominant follicles. The reduction in /3B expression during follicular development may explain why the /3B subunit has not been identified in the bovine ovary, even though its coding sequence is contained within the bovine genome (see Ref. 3). Bovine inhibin was purified from folicular fluid of large antral follicles (12), and the bovine ovarian cDNA libraries that have been used in cloning have been prepared from granulosa cells of large antral follicles (13). We would predict that if the cow is similar to the monkey, such large follicles would express little /3B subunit and that the small antral follicles that express /?B would be too small (
