Follicle Size Class Contributes to Distinct Secretion Patterns of Inhibin Isoforms during the Rat Estrous Cycle
http://www.100md.com
《内分泌学杂志》
Department of Neurobiology and Physiology (H.A.K., T.K.W.) and Robert H. Lurie Comprehensive Cancer Center (T.K.W.), Northwestern University, Evanston, Illinois 60208
Department of Medicine, Northwestern University Medical School (T.K.W.), Chicago, Illinois 60611
Abstract
The differential production of inhibins must be exquisitely controlled at the cellular level to ensure the secretion of the appropriate ligand at specific times during the reproductive cycle. The mechanisms underlying inhibin dimer assembly, processing and secretion are not well understood. Here we verify that the secretion of inhibin A and inhibin B from the granulosa cell is discordant during the estrous cycle: discordant production or secretion of the inhibins was not observed during the pregnant mare serum gonadotropin-induced cycle. We correlated the discordant production and secretion of inhibin A and inhibin B into the serum with distinct patterns of inhibin - and -subunit colocalization during the cycle in granulosa cells. We determined that the discordant pattern of inhibin A and inhibin B during the rat estrous cycle is due to independent populations of antral follicles making inhibin B (small antral follicles) or inhibin A (large antral follicles).
Introduction
INHIBIN IS AN endocrine hormone that is released from the ovaries into the circulation. Inhibin acts on the anterior pituitary to suppress FSH secretion (1) and on the ovary to potentiate granulosa cell differentiation. There are two known forms of inhibin: inhibin A and inhibin B. The inhibins share a common -subunit that assembles with one of two highly related -subunits (A or B) to form the active hormones. All three subunits are initially synthesized as preprohormones. The biologically active form of the inhibins is a dimer of the C-region of the -subunit disulfide linked to a mature -subunit (2, 3). The -subunits of inhibin can also homo- or heterodimerize to form activin A, activin B, or activin AB, which are antagonists of inhibin action. Activins are antagonists of inhibin, adding to the complexity and intricacy of the granulosa cell biosynthetic machinery.
Inhibin A and inhibin B synthesis increases in the follicles recruited by the secondary FSH surge of the rat estrous cycle. Increasing inhibin levels in turn suppress FSH levels during the early part of the estrous cycle. Despite the shared subunit structure and initial induction pattern, the two hormones circulate in an independent manner. Inhibin B rises quickly in the early estrous cycle and reaches maximal levels during the early follicular phase (metestrus and diestrus), and is negatively correlated with FSH during the estrous cycle. Inhibin A rises more slowly and closely parallels the rise in estradiol secretion from Graafian follicles. Synthesis of all three inhibin subunits is inhibited in the Graafian follicles by the preovulatory surges of the gonadotropins, FSH and LH. LH also induces follicle degradation before ovulation (4). Changes in serum levels of inhibin A and inhibin B are very similar to ovarian A- and B-subunit mRNA levels, respectively, with a several-hour delay (5), and the half-lives of the inhibin subunit mRNA and protein are relatively short. Thus, after ovulation, inhibin A and inhibin B mRNA and serum levels fall (5, 6, 7, 8). The pituitary is therefore relieved from inhibin repression, and elevated FSH levels persist in the form of the secondary FSH surge that leads to the recruitment of a new cohort of follicles into the next cycle.
Much of our understanding regarding inhibin synthesis in granulosa cells of follicles relies on in situ hybridization studies (1, 9). The -subunit mRNA levels are considerably higher than A- and B-subunit mRNA levels throughout the rat estrous cycle, especially at the time of Graafian follicle development (5), thus favoring inhibin production over that of the highly related protein, activin. In addition, the inhibin/activin subunits are localized to different subpopulations of granulosa cells in the Siberian hamster (10). Early studies hinted at but did not comment on the spatially restricted localization of subunit mRNA and protein in a variety of species (9, 11, 12, 13, 14, 15).
We believe that differential localization, assembly, processing, and secretion of the inhibin -, A-, and B-subunits from granulosa cells of developing follicles contribute to the different levels of inhibin dimers detected during the rat estrous cycle. In this study, we analyzed the location of inhibin/activin subunit proteins in different follicle populations and in different compartments of the individual follicles. This was done in concert with measurements of both ovarian and serum levels of inhibin A and inhibin B to further understand how follicle anatomy may contribute to the pattern of inhibin A and inhibin B detected in circulation. We analyzed two well-established animal models, the normal cycling rat and the immature pregnant mare serum gonadotropin (PMSG)-induced rat. We examined the secretion pattern of inhibins during the estrous and induced cycle, inhibin levels in ovaries during the estrous and induced cycles, follicular fluid levels of inhibin subunits during the estrous cycle, and inhibin subunit protein compartmentalization in subpopulations of granulosa cells during the estrous and induced cycle.
Materials and Methods
Animal and tissue preparation
Adult female Sprague Dawley rats (Charles River Laboratories, Inc., Wilmington, MA) were housed in a temperature-controlled room, with lights on from 0500–1900 h, at Northwestern University (Evanston, IL). Animals were housed three to four per cage and were provided with food and water ad libitum. Two groups of animals were analyzed: a normal estrous cycle group and induced cycle group. Estrous cyclicity was monitored by daily examination of vaginal cytology. Animals exhibiting at least three consecutive 4-d estrous cycles were used in these experiments. The induced cycle animals were 21 d old and were injected with 10 IU PMSG (Sigma-Aldrich Corp., St. Louis, MO). Animals were killed at 24 or 48 h after injection. Additional groups of animals were injected with 10 IU human chorionic gonadotropin (hCG) (Sigma-Aldrich Corp.) Forty-eight hours after receiving PMSG, and were killed 6, 12, 24, and 48 h later. One female rat from each time point was deeply anesthetized and perfused transcardially with 4% paraformaldehyde, and ovaries were removed and after fixed overnight at 4 C. All other animals were asphyxiated in a carbon dioxide-chamber and decapitated. Trunk blood was collected and allowed to coagulate overnight at 4 C, and serum was collected and stored at –80 C. Ovaries were either snap-frozen on dry ice, fixed in 4% paraformaldehyde overnight, or dissected to collect follicular fluid. All fixed ovaries were washed in a series of ethanol gradients and stored in 70% ethanol for less than 1 wk until embedded in paraffin. Four-micrometer ovary sections were cut on a microtome and collected on Superfrost Plus slides (Fisher Scientific, Pittsburgh, PA). Ovary samples for hormone assays were obtained by homogenizing one whole organ in 0.5–1 ml 0.85% NaCl (wt/vol) as described previously (16). Individual follicles were dissected from freshly removed ovaries and at M1000 (400 μm or larger in diameter) and P1600 (550 μm or larger in diameter) during the rat estrous cycle. The follicles collected represent the largest follicles in the ovary at the particular time point. Follicular fluid was collected from 10 follicles from five animals at each time point using a 22-gauge needle. The follicular fluid was immediately assayed for inhibin A or inhibin B content using specific inhibin ELISAs. All animals were treated in full accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Hormone assays
Serum LH, FSH, estradiol, and progesterone were measured by RIA (Center for Research in Reproduction, University of Virginia, Charlottesville, VA). The FSH RIA had a detection range of 0.6–18 ng/ml and an intraassay variation of 2.5–21.7%. The LH RIA had a detection range of 0.07–37.4 ng/ml and an intraassay variation of 3.4–7.1%. Serum and ovary, inhibin A and inhibin B, were measured by ELISA (Diagnostic Systems Laboratories, Webster, TX). The serum levels of inhibin were previously reported by Chapman and Woodruff (17). The inhibin assays had a sensitivity range of 10–1000 pg/ml. Intraassay variations for inhibin A and inhibin B were 5.12% and 4.67%, respectively. Interassay variations for inhibin A and inhibin B were 21.33% and 22.9%, respectively. The inhibin assays were validated using recombinant human-inhibin A (WHO 91/624) and -inhibin B (National Institute for Biological Standards and Controls, Potters Bar, Hertfordshire, UK). Whole ovaries were homogenized, then diluted 1:10,000 in 0.85% NaCl for ELISA analysis. Follicular fluid was diluted in a series from 1:10 to 1:1000 for ELISA analysis. No cross-reactivity between heterologous ligands occurred using purified recombinant ligands for any of the assays. Native ovarian inhibin A and inhibin B ligands diluted linearly and in parallel to the standard curve (data not shown).
Follicle classification
Follicles were classified according to the following criteria: primordial, oocyte less than 25 μm in diameter with few squamous granulosa cells; primary, oocyte with presence of cuboidal granulosa cells to complete single layer of surrounding cuboidal granulosa cells; secondary, multiple layers of granulosa cells without formation of an antrum; early antral, less than 300 μm follicle diameter with a visibly forming antrum; antral, greater than 400 μm follicle diameter with a distinct cumulus granulosa cell layer surrounding oocyte; atretic, evidence of cellular pyknosis and disorganization to shedding of the granulosa cell layer (these follicles were excluded from analysis).
Antibodies
Rabbit polyclonal antibodies against the -, A-, and B-subunits of inhibin (a gift from Wylie Vale, The Salk Institute, La Jolla, CA) were used at a final concentration of 2 μg/ml (anti--subunit), 3 μg/ml (anti-A-subunit) and 4 μg/ml (anti-B-subunit). Biotinylated goat antirabbit secondary antibody was used for general localization analysis and fluorescein-conjugated goat antirabbit antibody was used for colocalization analysis (Vector Laboratories, Inc., Burlingame, CA) at a final concentration of 5 μg/ml. For colocalization statistical analysis, inhibin subunit primary antibodies were labeled with fluorescent dyes (Alexa Fluor 488 and 555) using the Zenon Rabbit IgG Labeling Kit purchased from Molecular Probes, Inc. (Eugene, OR).
Immunofluorescence
Slides were deparaffinized in xylenes and rehydrated through a graded ethanol series. Antigen retrieval was performed in 0.01 M sodium citrate (Sigma-Aldrich Corp.) in a microwave for 2 min at high power and for 7 min at medium-low power. After cooling in antigen retrieval solution, slides were washed in Tris-buffered saline [TBS; 500 mM NaCl and 20 mM Tris (pH 7.6)] and permeabilized in TBS containing 0.1% Tween 20 (TBS-T). For Alexa Fluor colocalization experiments, antibodies were labeled with dyes and sections incubated according to instructions. For all other immunofluorescence experiments, slides were incubated in 3% H2O2 to quench endogenous peroxidase activity, rinsed in TBS, and placed into staining racks (Vector Laboratories, Inc.). Endogenous biotin and avidin binding was blocked with the Biotin-Avidin Blocking kit (Vector Laboratories, Inc.). Nonspecific binding was blocked by incubating slides in TBS containing 0.1% Triton-X and 10% serum from the host species of the secondary antibody. Serial sections were incubated with the primary antibody at 4 C for 15 h, washed in TBS-T, and incubated with biotin-labeled or fluorescein-labeled secondary antibody at room temperature for 1 h. Slides were washed in TBS-T and coverslipped or incubated in 5 mg/ml streptavidin-fluorescein conjugated tertiary antibody (Vector Laboratories, Inc.) for immunofluorescence imaging, then washed and coverslipped. Mounting media containing 4',6-diamidino-2-phenylindole (Vector Laboratories, Inc.) was used in counterstaining and coverslipping. Background staining was determined by replacing the primary antibody with buffer.
Analysis of immunofluorescence
The immunofluorescence was imaged with two-photon excitation laser scanning microscopy using Zeiss 510 LSM (upright configuration) (Carl Zeiss, Inc., Jena, Germany). The excitation beam produced by the femtosecond pulsed Ti:sapphire laser (Tsunami, Spectra-Physics; 8 W Millenia pump) was tuned (pumping power 6.5 W with 0.5 W entering the Zeiss AOM), was passed through an LSM 510 microscope with different dichroics (Carl Zeiss, Inc.), and focused onto coverslip adherent ovarian tissue using a x63 oil-immersion objective (Carl Zeiss, Inc.). The NLO META scan head (Carl Zeiss, Inc.) allowed data collection. Twenty sections (every fourth slide) for each ovary and three ovaries were analyzed for total number of secondary follicles with positive staining. Five sections (every 10th slide, approximately every 30th section) for each ovary and three ovaries were analyzed for total number of antral follicles with positive staining. For antral follicle colocalization analysis, ten individual follicles from each of three ovaries were analyzed for overall trend of subunit localization in positions 1–4 in antral follicles illustrated in Fig. 1. Position 1 is the mural granulosa cells connected to cumulus granulosa cells in follicle. Position 2 is the cumulus-oocyte complex and cumulus bridge connection to mural granulosa cells in follicle. Position 3 is the mural granulosa cells located a 90-degree angle from cumulus-oocyte complex in follicle. Position 4 is the mural granulosa cells located opposite cumulus-oocyte complex in follicle. A follicle was reported as positive when staining was detected by confocal microscopy. Simultaneous multicolor imaging was conducted to analyze colocalization of inhibin subunits in dominant antral follicle populations for each group analyzed. Ten antral follicles from three different ovaries were analyzed for colocalization. The experiment used Zenon-labeled primary antibodies as described above in immunofluorescence section. Simultaneous multicolor imaging was conducted at positions 1, 3, and 4 of antral follicles. The -subunit was more abundant than the -subunits. As such, we reported on a pixel by pixel basis the proportion (percent) of -subunit that colocalized with either -subunit using LSM 5.10 software. The average percent of colocalization in three different areas of each position for ten different follicles was analyzed.
Statistical analysis
The Pearson product moment correlation coefficient, R, was used to analyze the extent of a linear relationship between hormone profiles across the individual cycles. The t test for zero correlation was used to test the significance of R. An independent sample t test was used to determine significance of the mean percent of -, A-, or B-subunit positive secondary follicles across the estrous and induced cycles. An independent sample t test was used to determine significance of the mean percent of colocalization of subunits across the estrous and induced cycles. In all cases, P < 0.05 was considered significant. Values are reported as ± SEM.
Results
Serum and ovarian inhibin A and inhibin B profiles during the rat estrous and induced cycles
The concentration of inhibin A and inhibin B in serum, ovarian homogenates and follicular fluid during the rat estrous cycle was investigated. Consistent with previous studies, inhibin A and inhibin B serum concentrations were discordant during the estrous cycle (Fig. 2A). Similarly, inhibin A and inhibin B concentrations in the ovary did not correlate, particularly in the latter half of the cycle (Fig. 2B). The ovarian concentrations of inhibin A increased from M1000 to a peak at P1000, fell to a nadir at P1600, increased rapidly by E1400 and declined gradually until M1000. The ovarian concentrations of inhibin B increased from M1000 to a peak at P1000, rapidly declined to a nadir at E0600, increased rapidly by E1000, and fell by E1400 to levels that remained constant until M1000. Serum inhibin A and ovarian inhibin A levels were positively correlated throughout the estrous cycle (Fig. 2C), whereas serum inhibin B and ovarian inhibin B levels were not (Fig. 2D). The follicular fluid levels of inhibin A increased significantly (P = 0.0003) from M1000 to P1600, 465 ± 116 pg/ml to 1853 ± 278 pg/ml respectively. The follicular fluid levels of inhibin B increased but not significantly from M1000 to P1600, 767 ± 150 pg/ml to 1124 ± 160 pg/ml.
The concentrations of estradiol, FSH, LH, and progesterone in serum during the estrous cycle were investigated and previously reported by Chapman and Woodruff (17). As expected, serum inhibin A and estradiol concentrations were positively correlated, whereas serum inhibin B and FSH concentrations were negatively correlated during the estrous cycle. Serum LH concentrations surged beginning at P1600, peaking at P1830 and falling by P2400. Serum progesterone concentrations surged beginning at P1600, peaking at P1830 and falling by E0400.
The concentrations of inhibin A and inhibin B in serum and ovarian homogenates from gonadotropin-treated immature rats were investigated. As expected, the serum concentrations of inhibin A and inhibin B were positively correlated during the induced cycle (Fig. 3A). The ovarian concentrations of inhibin A and inhibin B were also positively correlated during the induced cycle (Fig. 3B). Ovarian concentrations of inhibin A and inhibin B rose with PMSG treatment to a peak by 48 h after PMSG injection, fell to a nadir by 12 h after hCG injection and began to rise by 24 h after hCG injection (Fig. 3B). Serum and ovarian concentrations of each inhibin were positively correlated during the induced cycle (Fig. 3, C and D).
The concentrations of estradiol and progesterone in serum during the induced cycle were investigated (Fig. 3, E–G). Both serum inhibin A and serum inhibin B concentrations were positively correlated with estradiol during the induced cycle (Fig. 3, E and F). Serum progesterone concentrations were constant during first 48 h after PMSG injection, peaked at 6 h after hCG injection and fell by 12 h after hCG injection to a level higher than before the peak (Fig. 3G).
Inhibin subunit localization in the ovary during the estrous and induced cycles
The pattern of the inhibin subunits in follicles present in ovaries collected during the estrous cycle was investigated. Inhibin subunits were mainly expressed in granulosa cells during the estrous cycle. The total number of follicles positive for individual subunits was analyzed. Of all follicles positive for the inhibin -subunit, at M1000, 74%; at D1000, 81%; and at P1600, 73% of all follicles positive for the -subunit were positive for A-subunit. At M1000, 88%; at D1000, 86%; and at P1600, 72% were positive for B-subunit. Corpora lutea did not express any of the inhibin/activin subunits at any time during the estrous cycle.
All inhibin subunits were expressed in secondary follicles during the estrous cycle (Fig. 4, A–L). At M1000 the inhibin -, A- and B-subunits were expressed in 91%, 85% and 88% of secondary follicles, respectively (Fig. 4, A–C and G–I). At P1600, the inhibin -, A-, and B-subunits were expressed in 89%, 67%, and 90% of secondary follicles, respectively (Fig. 4, D–F and G–I). Fewer secondary follicles expressed the A-subunit when compared with the - and B-subunits at P1600 (P < 0.01).
The pattern of the inhibin subunits in ovaries during the induced cycle was similarly investigated. Inhibin subunits were detected mainly in granulosa cells during the induced cycle. Of all follicles positive for the inhibin -subunit, at 12 h after PMSG injection 78% and 48 h after PMSG injection 91% of follicles were also positive for -subunits. The A- and B-subunits were produced in same follicles.
Inhibin subunits were expressed in granulosa cells of secondary follicles during the induced cycle (Fig. 5, A–L). At 12 h after PMSG injection, the inhibin -, A-, and B-subunits were expressed in 87%, 71%, and 90% of secondary follicles, respectively (Fig. 5, A–C and G–I). Fewer secondary follicles expressed the A-subunit when compared with - and B-subunits at 12 h after PMSG injection (P < 0.01). At 48 h after PMSG injection the inhibin -, A- and B-subunits were expressed in 92%, 87%, and 91% of secondary follicles, respectively (Fig. 5, D–F and J–L).
Inhibin subunit colocalization in antral follicles during the estrous cycle
Colocalization of the inhibin subunits in antral follicles during the estrous cycle was investigated. The schematic in Fig. 1 describes the four different areas of granulosa cells analyzed in the colocalization studies. Representative immunoreactivity staining of the inhibin - and A-subunits in follicles from M1000 and P1600 during the rat estrous cycle is shown in Fig. 6. The proportion of -subunit pixels colocalized with A-subunit pixels is reported in Fig. 7A. Mural granulosa cell staining for the - and A-subunits was higher than staining in cumulus granulosa cells (Fig. 6, A–X). At M1000 both the - and A-subunit staining were more intense in the basement membrane-facing mural granulosa cells, especially at positions closest to the outside of ovary (position 4 in this reference follicle) (Fig. 6, S–U). At P1600 the - and A-subunits staining was localized throughout the mural granulosa cells with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions 3 and 4, closest to the outside of the entire ovary (Fig. 6, P–R and V–X). The - and A-subunits were colocalized in a higher number of mural granulosa cells on P1600 than on M1000 (Fig. 6, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with A-subunit pixels was found to be significantly higher at P1600 in positions 3 and 4 when compared with M1000 (Fig. 7A).
Colocalization of the inhibin - and B-subunits in antral follicles during the estrous cycle was similarly investigated. Representative immunoreactivity staining in follicles from M1000 and P1600 during the rat estrous cycle is shown in Fig. 8. The proportion of -subunit pixels colocalized with B-subunit pixels is reported in Fig. 7B. Mural granulosa cell staining for the - and B-subunits was greater than staining in cumulus granulosa cells (Fig. 8, A–X). The -subunit staining was relatively uniform throughout the mural granulosa (Fig. 8, M and S). The B-subunit staining was relatively uniform throughout the mural granulosa cells at M1000 (Fig. 8B, N and T). At P1600, the -subunit was localized throughout the mural granulosa cells, with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions located closest to the outside of ovary (positions 3 and 4 in the representative follicle) (Fig. 8, P and V). In contrast, the B-subunit staining at P1600 was more intense in antral-facing mural granulosa cells and in most outward basement membrane-facing mural granulosa cells, which was most apparent at positions closest to the outside of entire ovary (positions 3 and 4 in the representative follicle) (Fig. 8, Q and W). The - and B-subunits were colocalized in a greater proportion of mural granulosa cells on M1000 than on P1600 (Fig. 8, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with B-subunit pixels was found to be significantly greater at M1000 in positions 1, 3, and 4 when compared with P1600 (Fig. 8B).
Inhibin subunit colocalization in antral follicles during the induced cycle
Finally, colocalization of the inhibin - and -subunits in antral follicles was investigated during the induced cycle. Representative immunoreactivity staining of - and A subunits in follicles from 12 and 48 h after PMSG injection are shown in Fig. 9. The proportion of -subunit colocalized with A-subunit is reported in Fig. 10A. The - and A-subunits were localized to all granulosa cells at 12 and 48 h after PMSG injection during the induced cycle (Fig. 9, A–X). Mural granulosa cell staining for the - and A-subunits was greater than staining in cumulus granulosa cells (Fig. 9, G–L). Both the - and A-subunit showed relatively uniform staining throughout the mural granulosa cells at 12 h after PMSG injection (Fig. 9, A–C, M–O, and S–U). At 48 h after PMSG injection, -subunit staining was localized throughout the mural granulosa cells with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions located closest to the outside of the entire ovary (position 3 and 4 in the representative follicle) (Fig. 9, P and V). In contrast, A-subunit staining was localized relatively uniformly throughout mural granulosa cells (Fig. 9, Q and W). The - and A-subunits were colocalized in a greater proportion of mural granulosa cells at 48 h after PMSG injection than at 12 h after PMSG injection (Fig. 9, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with A-subunit pixels was found to be significantly greater at 48 h after PMSG injection in positions 3 and 4 when compared with 12 h after PMSG injection (Fig. 10A).
Representative immunoreactivity staining of the inhibin - and B-subunits in follicles from 12 and 48 h after PMSG injection during the rat induced cycle is shown in Fig. 11. The proportion of -subunit colocalized with B-subunit is reported in Fig. 10B. The - and B-subunit were localized to all granulosa cells at 12 and 48 h after PMSG injection during the induced cycle (Fig. 11, A–X). Mural granulosa cell staining for the - and B-subunits was greater than staining in cumulus granulosa cells. Both the - and B-subunits staining are relatively uniform throughout the mural granulosa cells at 12 h after PMSG injection (Fig. 11, A–C, M–O, and S–U). At 48 h after PMSG injection, the -subunit staining was localized throughout the mural granulosa cells with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions located closest to the outside of the entire ovary (positions 3 and 4 in this representative follicle) (Fig. 11, P and V). In contrast, the B-subunit was localized throughout the mural granulosa cells with an intense increase in staining at the very antral edge of antral-facing mural granulosa cells (Fig. 11, Q and W). The - and B-subunits were colocalized in a greater proportion of mural granulosa cells at 12 h after PMSG injection than at 48 h after PMSG injection (Fig. 11, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with B-subunit pixels was found to be significantly greater at 12 h after PMSG injection in positions 3 and 4 when compared with 48 h after PMSG injection (Fig. 10B).
Discussion
In previous work, we observed differential compartmentalization of the mRNAs encoding the inhibin -, A-, and B-subunits in the Siberian hamster (10). Specifically, in Graafian follicles of long photoperiod female hamsters the B-subunit was detected in the antral-facing mural granulosa cell layer, the -subunit was relatively uniform throughout mural granulosa cell layer and the A-subunit was uniformly distributed with increased expression in basement membrane-facing mural granulosa cell layer. Early studies hinted at but did not comment on the spatially restricted localization of inhibin subunit mRNA and protein in a variety of species (9, 11, 12, 13, 14, 15). The intent of this study was to investigate the type of follicles producing the inhibin subunits, and specifically the granulosa cells within these follicles, to investigate the differential compartmentalization of inhibin subunits in the rat ovary to gain insight into the difference between inhibin A and inhibin B production and function. We analyzed the estrous cycle and PMSG-induced cycle models to further understand how different follicle class sizes are related to the differential secretion of inhibin A and inhibin B. The induced cycle model was used as a positive control. If a heterogeneous population of granulosa cells within a single follicle is necessary to obtain discordant inhibin A and inhibin B secretion during the estrous cycle, then this relationship would be totally abolished during the induced cycle, where the follicle selection, growth and ovulation are synchronized.
As expected from previous studies, inhibin A and inhibin B were differentially secreted during the estrous cycle (5, 7). In addition, we found that early on estrus inhibin B levels rose more quickly and to a higher concentration, compared with inhibin A. The inhibin A ovarian and serum concentrations were highly correlated, whereas inhibin B concentrations in the serum and ovary throughout the cycle did not show a strong positive correlation. One explanation of these data is that the time between inhibin B production and secretion is much shorter at 1000 h on metestrus than at 1600 h on proestrus. In addition, we found that inhibin A levels were significantly greater in the follicular fluid of dominant follicles when compared with smaller antral follicles, and although the trend was similar for inhibin B it was not significant. These results are consistent with previous measurement of inhibin in human follicular fluid (18). Inhibin B may be retained in the ovary within individual granulosa cells or in the follicular fluid of small follicles on proestrus. The physiological significance of the rise in inhibin B on the morning of estrus is to cause the immediate inhibition of FSH at a time when neither estradiol nor inhibin A have been stimulated. As the cycle progresses, both inhibin A and estradiol are secreted, reinforcing the negative feedback of the ovary to the pituitary.
The second model in this study was the induced cycle in immature female rats treated with PMSG and hCG. Sequential administration of these hormones induces a synchronized selection of follicles into the growing pool and subsequent ovulation. During the induced cycle, there was a strong positive correlation between serum inhibin A and inhibin B levels, ovarian inhibin A and inhibin B levels, serum and ovarian inhibin A levels, and serum and ovarian inhibin B levels, which suggest that exogenous FSH and LH overrides the control mechanisms that limit inhibin B secretion to discrete times during the estrous cycle.
We investigated the compartmental localization of inhibin subunits in the ovary to further elucidate the differential secretion of inhibin A and inhibin B during the estrous cycle. Distribution of inhibin -, A-, and B-subunits in different ovarian compartments was studied using immunofluorescence. The proportion of -subunit pixels that colocalized with the A-subunit pixels was significantly greater in positions 3 and 4 of mural granulosa cells at 1600 h on proestrus than at 1000 h on metestrus. The opposite pattern of colocalization was observed between the - and B-subunit. The proportion of -subunit pixels that colocalized with the B-subunit pixels was significantly greater at positions 1, 3, and 4 of mural granulosa cells at 1000 h on metestrus than at 1600 h on proestrus. All mural granulosa cells have the potential to produce inhibin A and inhibin B at 1000 h on metestrus and 1600 h on proestrus. However, antral follicles show greater inhibin B colocalization at 1000 h on metestrus, and greater A colocalization at 1600 h on proestrus. The overall trend of the immunostaining experiments suggests greater inhibin A assembly with the highest potential for secretion toward the basement membrane at 1600 h on proestrus. In contrast, inhibin B assembly was greater in metestrus, whereas at 1600 h on proestrus the potential for secretion was toward the antrum and the thin line of mural granulosa cells lining the basement membrane at positions closest to the outside of the ovary. Thus, the distinct compartmentalization of each inhibin isoform, the increased colocalization of - and B-subunits at 1000 h on metestrus, and the increased colocalization of the - and A-subunits at 1600 h on proestrus may contribute to the discordance in serum levels of inhibin A and inhibin B observed throughout the normal estrous cycle. Inhibin A is secreted as it is made in increasing concentrations by the growing population of granulosa cells during metestrus to proestrus, whereas inhibin B is secreted during estrus/early metestrus and accumulates in ovary during proestrus.
We investigated variations in the localization of production of the inhibin subunits in the large pool of growing antral follicles during the induced cycle. Distribution of inhibin -, A-, and B-subunits in different ovarian compartments was studied using immunofluorescence at two particular time points, 12 and 48 h after PMSG treatment. Interestingly, the - and A-subunits were colocalized in a greater proportion of mural granulosa cells at 48 h after PMSG injection than at 12 h after PMSG injection, particularly at positions 3 and 4. The - and B-subunits had the opposite pattern, colocalizing in a greater proportion of granulosa cells at 12 h after PMSG injection than at 48 h after PMSG injection, and again, colocalization occurred most frequently at positions 3 and 4. Thus, the subunit colocalization patterns at 12 and 48 h after PMSG treatment were consistent with those seen in the dominant antral follicle populations at 1000 h on metestrus and 1600 h on proestrus. These results confirm that inhibin subunit compartmentalization in antral follicles during folliculogenesis occurs, and that inhibin B is likely produced in smaller selected antral follicles that switch to inhibin A production once they develop into Graafian follicles. The induced cycle model shows a synchronized switch from inhibin B to inhibin A assembly as follicle size increases, with no difference in inhibin secretion between the two isoforms. Synchronization of follicle development with exogenous gonadotropins overrides the granulosa cell heterogeneity that underlies the discordant secretion of inhibin A and inhibin B during the cycle.
The discordance of inhibin A and inhibin B secretion during the estrous cycle was originally thought to be due to different populations of follicles making inhibin A and inhibin B. This study reveals that all follicles are capable of producing both inhibin isoforms, and that heterogeneity in a single follicle’s granulosa cell population is the basis for the temporal differences in inhibin A and inhibin B secretion into the blood. The secretion rate may be determined by the location of assembled dimers within the follicles (i.e. antrum-facing vs. basement membrane-facing granulosa cells). Little is known about the intracellular mechanisms responsible for controlling secretion of the inhibins by a granulosa cell. Spatially, the granulosa cells located near the edge of individual follicles would have the greatest access to the vasculature. The patterns of inhibin assembly and release are consistent with the local, concentration-dependent role of inhibin A in granulosa cell growth and development and an endocrine, time-dependent role of inhibin B on inhibition of pituitary FSH expression. The mechanisms controlling the exquisitely timed assembly and release of each inhibin isoform should be further examined. Little is known about the ability to process, store, degrade, and secrete the dimeric inhibins in vivo. A deeper understanding of inhibin biosynthesis and secretion is crucial to understanding mechanisms of inhibin action within the reproductive axis. Future studies will investigate the role of disulfide bond formation, processing, and cleavage involved in the secretion of inhibin A and inhibin B.
Acknowledgments
We thank Dr. W. W. Vale for the inhibin and activin subunit polyclonal antibodies. We also thank the Diagnostic Systems Laboratories for the ELISA kits. In addition, we thank Dr. Alfred Rademaker for his expertise in statistical analysis. Finally, we would like to express sincere appreciation for the participation of Dr. Stacey Chapman-Tobin and Dr. Signe Kilen in this study.
Footnotes
This work was supported by National Institutes of Health (NIH) Grant HD-37096 (to T.K.W.) and by National Institute of Child Health and Human Development/NIH through Cooperative Agreement U54-HD-28934 as part of the Specialized Cooperative Centers Program in Reproductive Research.
First Published Online September 29, 2005
Abbreviations: hCG, Human chorionic gonadotropin; PMSG, pregnant mare serum gonadotropin.
Accepted for publication September 19, 2005.
References
Woodruff TK, D’Agostino J, Schwartz NB, Mayo KE 1988 Dynamic changes in inhibin messenger RNAs in rat ovarian follicles during the reproductive cycle. Science 239:1296–1299
Hsueh AJ, Dahl KD, Vaughan J, Tucker E, Rivier J, Bardin CW, Vale W 1987 Heterodimers and homodimers of inhibin subunits have different paracrine action in the modulation of luteinizing hormone-stimulated androgen biosynthesis. Proc Natl Acad Sci USA 84:5082–5086
Rivier C, Vale W, Rivier J 1987 Studies of the inhibin family of hormones: a review. Horm Res 28:104–118
Spaczynski RZ, Tilly JL, Mansour A, Duleba AJ 2005 Insulin and insulin-like growth factors inhibit and luteinizing hormone augments ovarian theca-interstitial cell apoptosis. Mol Hum Reprod 11:319–324
Arai KY, Ohshima K, Watanabe G, Arai K, Uehara K, Taya K 2002 Dynamics of messenger RNAs encoding inhibin/activin subunits and follistatin in the ovary during the rat estrous cycle. Biol Reprod 66:1119–1126
Bernard DJ, Woodruff TK 2001 Inhibin binding protein in rats: alternative transcripts and regulation in the pituitary across the estrous cycle. Mol Endocrinol 15:654–667
Woodruff TK, Besecke LM, Groome N, Draper LB, Schwartz NB, Weiss J 1996 Inhibin A and inhibin B are inversely correlated to follicle-stimulating hormone, yet are discordant during the follicular phase of the rat estrous cycle, and inhibin A is expressed in a sexually dimorphic manner. Endocrinology 137:5463–5467
Woodruff TK, Krummen LA, Chen S, Chen S, DeGuzman G, Lyon R, Baly DL, Allison DE, Garg S, Wong WL, Hebert N, Mather JP, Cossum P 1993 Pharmacokinetic profile of recombinant human (rh) inhibin A and activin A in the immature rat. I. Serum profile of rh-inhibin A and rh-activin A in the immature female rat. Endocrinology 132:715–724
Meunier H, Cajander SB, Roberts VJ, Rivier C, Sawchenko PE, Hsueh AJ, Vale W 1988 Rapid changes in the expression of inhibin -, A-, and B-subunits in ovarian cell types during the rat estrous cycle. Mol Endocrinol 2:1352–1363
Kenny HA, Bernard DJ, Horton TH, Woodruff TK 2002 Photoperiod-dependent regulation of inhibin in Siberian hamsters: I. Ovarian inhibin production and secretion. J Endocrinol 174:71–83
Meunier H, Roberts VJ, Sawchenko PE, Cajander SB, Hsueh AJ, Vale W 1989 Periovulatory changes in the expression of inhibin -, A-, and B-subunits in hormonally induced immature female rats. Mol Endocrinol 3:2062–2069
van den Hurk R, Van de Pavert SA 2001 Localization of an activin/activin receptor system in the porcine ovary. Mol Reprod Dev 60:463–471
Nagamine N, Nambo Y, Nagata S, Nagaoka K, Tsunoda N, Taniyama H, Tanaka Y, Tohei A, Watanabe G, Taya K 1998 Inhibin secretion in the mare: localization of inhibin , A, and B subunits in the ovary. Biol Reprod 59:1392–1398
Tisdall DJ, Hudson N, Smith P, McNatty KP 1994 Localization of ovine follistatin and and A inhibin mRNA in the sheep ovary during the oestrous cycle. J Mol Endocrinol 12:181–193
Silva JR, van den Hurk R, van Tol HT, Roelen BA, Figueiredo JR 2004 Gene expression and protein localization for activin-A, follistatin and activin receptors in goat ovaries. J Endocrinol 183:405–415
Herath CB, Yamashita M, Watanabe G, Jin W, Tangtrongsup S, Kojima A, Groome NP, Suzuki AK, Taya K 2001 Regulation of follicle-stimulating hormone secretion by estradiol and dimeric inhibins in the infantile female rat. Biol Reprod 65:1623–1633
Chapman SC, Woodruff TK 2003 Betaglycan localization in the female rat pituitary: implications for the regulation of follicle-stimulating hormone by inhibin. Endocrinology 144:5640–5649
Schneyer AL, Fujiwara T, Fox J, Welt CK, Adams J, Messerlian GM, Taylor AE 2000 Dynamic changes in the intrafollicular inhibin/activin/follistatin axis during human follicular development: relationship to circulating hormone concentrations. J Clin Endocrinol Metab 85:3319–3330(Hilary A. Kenny and Teresa K. Woodruff)
Department of Medicine, Northwestern University Medical School (T.K.W.), Chicago, Illinois 60611
Abstract
The differential production of inhibins must be exquisitely controlled at the cellular level to ensure the secretion of the appropriate ligand at specific times during the reproductive cycle. The mechanisms underlying inhibin dimer assembly, processing and secretion are not well understood. Here we verify that the secretion of inhibin A and inhibin B from the granulosa cell is discordant during the estrous cycle: discordant production or secretion of the inhibins was not observed during the pregnant mare serum gonadotropin-induced cycle. We correlated the discordant production and secretion of inhibin A and inhibin B into the serum with distinct patterns of inhibin - and -subunit colocalization during the cycle in granulosa cells. We determined that the discordant pattern of inhibin A and inhibin B during the rat estrous cycle is due to independent populations of antral follicles making inhibin B (small antral follicles) or inhibin A (large antral follicles).
Introduction
INHIBIN IS AN endocrine hormone that is released from the ovaries into the circulation. Inhibin acts on the anterior pituitary to suppress FSH secretion (1) and on the ovary to potentiate granulosa cell differentiation. There are two known forms of inhibin: inhibin A and inhibin B. The inhibins share a common -subunit that assembles with one of two highly related -subunits (A or B) to form the active hormones. All three subunits are initially synthesized as preprohormones. The biologically active form of the inhibins is a dimer of the C-region of the -subunit disulfide linked to a mature -subunit (2, 3). The -subunits of inhibin can also homo- or heterodimerize to form activin A, activin B, or activin AB, which are antagonists of inhibin action. Activins are antagonists of inhibin, adding to the complexity and intricacy of the granulosa cell biosynthetic machinery.
Inhibin A and inhibin B synthesis increases in the follicles recruited by the secondary FSH surge of the rat estrous cycle. Increasing inhibin levels in turn suppress FSH levels during the early part of the estrous cycle. Despite the shared subunit structure and initial induction pattern, the two hormones circulate in an independent manner. Inhibin B rises quickly in the early estrous cycle and reaches maximal levels during the early follicular phase (metestrus and diestrus), and is negatively correlated with FSH during the estrous cycle. Inhibin A rises more slowly and closely parallels the rise in estradiol secretion from Graafian follicles. Synthesis of all three inhibin subunits is inhibited in the Graafian follicles by the preovulatory surges of the gonadotropins, FSH and LH. LH also induces follicle degradation before ovulation (4). Changes in serum levels of inhibin A and inhibin B are very similar to ovarian A- and B-subunit mRNA levels, respectively, with a several-hour delay (5), and the half-lives of the inhibin subunit mRNA and protein are relatively short. Thus, after ovulation, inhibin A and inhibin B mRNA and serum levels fall (5, 6, 7, 8). The pituitary is therefore relieved from inhibin repression, and elevated FSH levels persist in the form of the secondary FSH surge that leads to the recruitment of a new cohort of follicles into the next cycle.
Much of our understanding regarding inhibin synthesis in granulosa cells of follicles relies on in situ hybridization studies (1, 9). The -subunit mRNA levels are considerably higher than A- and B-subunit mRNA levels throughout the rat estrous cycle, especially at the time of Graafian follicle development (5), thus favoring inhibin production over that of the highly related protein, activin. In addition, the inhibin/activin subunits are localized to different subpopulations of granulosa cells in the Siberian hamster (10). Early studies hinted at but did not comment on the spatially restricted localization of subunit mRNA and protein in a variety of species (9, 11, 12, 13, 14, 15).
We believe that differential localization, assembly, processing, and secretion of the inhibin -, A-, and B-subunits from granulosa cells of developing follicles contribute to the different levels of inhibin dimers detected during the rat estrous cycle. In this study, we analyzed the location of inhibin/activin subunit proteins in different follicle populations and in different compartments of the individual follicles. This was done in concert with measurements of both ovarian and serum levels of inhibin A and inhibin B to further understand how follicle anatomy may contribute to the pattern of inhibin A and inhibin B detected in circulation. We analyzed two well-established animal models, the normal cycling rat and the immature pregnant mare serum gonadotropin (PMSG)-induced rat. We examined the secretion pattern of inhibins during the estrous and induced cycle, inhibin levels in ovaries during the estrous and induced cycles, follicular fluid levels of inhibin subunits during the estrous cycle, and inhibin subunit protein compartmentalization in subpopulations of granulosa cells during the estrous and induced cycle.
Materials and Methods
Animal and tissue preparation
Adult female Sprague Dawley rats (Charles River Laboratories, Inc., Wilmington, MA) were housed in a temperature-controlled room, with lights on from 0500–1900 h, at Northwestern University (Evanston, IL). Animals were housed three to four per cage and were provided with food and water ad libitum. Two groups of animals were analyzed: a normal estrous cycle group and induced cycle group. Estrous cyclicity was monitored by daily examination of vaginal cytology. Animals exhibiting at least three consecutive 4-d estrous cycles were used in these experiments. The induced cycle animals were 21 d old and were injected with 10 IU PMSG (Sigma-Aldrich Corp., St. Louis, MO). Animals were killed at 24 or 48 h after injection. Additional groups of animals were injected with 10 IU human chorionic gonadotropin (hCG) (Sigma-Aldrich Corp.) Forty-eight hours after receiving PMSG, and were killed 6, 12, 24, and 48 h later. One female rat from each time point was deeply anesthetized and perfused transcardially with 4% paraformaldehyde, and ovaries were removed and after fixed overnight at 4 C. All other animals were asphyxiated in a carbon dioxide-chamber and decapitated. Trunk blood was collected and allowed to coagulate overnight at 4 C, and serum was collected and stored at –80 C. Ovaries were either snap-frozen on dry ice, fixed in 4% paraformaldehyde overnight, or dissected to collect follicular fluid. All fixed ovaries were washed in a series of ethanol gradients and stored in 70% ethanol for less than 1 wk until embedded in paraffin. Four-micrometer ovary sections were cut on a microtome and collected on Superfrost Plus slides (Fisher Scientific, Pittsburgh, PA). Ovary samples for hormone assays were obtained by homogenizing one whole organ in 0.5–1 ml 0.85% NaCl (wt/vol) as described previously (16). Individual follicles were dissected from freshly removed ovaries and at M1000 (400 μm or larger in diameter) and P1600 (550 μm or larger in diameter) during the rat estrous cycle. The follicles collected represent the largest follicles in the ovary at the particular time point. Follicular fluid was collected from 10 follicles from five animals at each time point using a 22-gauge needle. The follicular fluid was immediately assayed for inhibin A or inhibin B content using specific inhibin ELISAs. All animals were treated in full accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Hormone assays
Serum LH, FSH, estradiol, and progesterone were measured by RIA (Center for Research in Reproduction, University of Virginia, Charlottesville, VA). The FSH RIA had a detection range of 0.6–18 ng/ml and an intraassay variation of 2.5–21.7%. The LH RIA had a detection range of 0.07–37.4 ng/ml and an intraassay variation of 3.4–7.1%. Serum and ovary, inhibin A and inhibin B, were measured by ELISA (Diagnostic Systems Laboratories, Webster, TX). The serum levels of inhibin were previously reported by Chapman and Woodruff (17). The inhibin assays had a sensitivity range of 10–1000 pg/ml. Intraassay variations for inhibin A and inhibin B were 5.12% and 4.67%, respectively. Interassay variations for inhibin A and inhibin B were 21.33% and 22.9%, respectively. The inhibin assays were validated using recombinant human-inhibin A (WHO 91/624) and -inhibin B (National Institute for Biological Standards and Controls, Potters Bar, Hertfordshire, UK). Whole ovaries were homogenized, then diluted 1:10,000 in 0.85% NaCl for ELISA analysis. Follicular fluid was diluted in a series from 1:10 to 1:1000 for ELISA analysis. No cross-reactivity between heterologous ligands occurred using purified recombinant ligands for any of the assays. Native ovarian inhibin A and inhibin B ligands diluted linearly and in parallel to the standard curve (data not shown).
Follicle classification
Follicles were classified according to the following criteria: primordial, oocyte less than 25 μm in diameter with few squamous granulosa cells; primary, oocyte with presence of cuboidal granulosa cells to complete single layer of surrounding cuboidal granulosa cells; secondary, multiple layers of granulosa cells without formation of an antrum; early antral, less than 300 μm follicle diameter with a visibly forming antrum; antral, greater than 400 μm follicle diameter with a distinct cumulus granulosa cell layer surrounding oocyte; atretic, evidence of cellular pyknosis and disorganization to shedding of the granulosa cell layer (these follicles were excluded from analysis).
Antibodies
Rabbit polyclonal antibodies against the -, A-, and B-subunits of inhibin (a gift from Wylie Vale, The Salk Institute, La Jolla, CA) were used at a final concentration of 2 μg/ml (anti--subunit), 3 μg/ml (anti-A-subunit) and 4 μg/ml (anti-B-subunit). Biotinylated goat antirabbit secondary antibody was used for general localization analysis and fluorescein-conjugated goat antirabbit antibody was used for colocalization analysis (Vector Laboratories, Inc., Burlingame, CA) at a final concentration of 5 μg/ml. For colocalization statistical analysis, inhibin subunit primary antibodies were labeled with fluorescent dyes (Alexa Fluor 488 and 555) using the Zenon Rabbit IgG Labeling Kit purchased from Molecular Probes, Inc. (Eugene, OR).
Immunofluorescence
Slides were deparaffinized in xylenes and rehydrated through a graded ethanol series. Antigen retrieval was performed in 0.01 M sodium citrate (Sigma-Aldrich Corp.) in a microwave for 2 min at high power and for 7 min at medium-low power. After cooling in antigen retrieval solution, slides were washed in Tris-buffered saline [TBS; 500 mM NaCl and 20 mM Tris (pH 7.6)] and permeabilized in TBS containing 0.1% Tween 20 (TBS-T). For Alexa Fluor colocalization experiments, antibodies were labeled with dyes and sections incubated according to instructions. For all other immunofluorescence experiments, slides were incubated in 3% H2O2 to quench endogenous peroxidase activity, rinsed in TBS, and placed into staining racks (Vector Laboratories, Inc.). Endogenous biotin and avidin binding was blocked with the Biotin-Avidin Blocking kit (Vector Laboratories, Inc.). Nonspecific binding was blocked by incubating slides in TBS containing 0.1% Triton-X and 10% serum from the host species of the secondary antibody. Serial sections were incubated with the primary antibody at 4 C for 15 h, washed in TBS-T, and incubated with biotin-labeled or fluorescein-labeled secondary antibody at room temperature for 1 h. Slides were washed in TBS-T and coverslipped or incubated in 5 mg/ml streptavidin-fluorescein conjugated tertiary antibody (Vector Laboratories, Inc.) for immunofluorescence imaging, then washed and coverslipped. Mounting media containing 4',6-diamidino-2-phenylindole (Vector Laboratories, Inc.) was used in counterstaining and coverslipping. Background staining was determined by replacing the primary antibody with buffer.
Analysis of immunofluorescence
The immunofluorescence was imaged with two-photon excitation laser scanning microscopy using Zeiss 510 LSM (upright configuration) (Carl Zeiss, Inc., Jena, Germany). The excitation beam produced by the femtosecond pulsed Ti:sapphire laser (Tsunami, Spectra-Physics; 8 W Millenia pump) was tuned (pumping power 6.5 W with 0.5 W entering the Zeiss AOM), was passed through an LSM 510 microscope with different dichroics (Carl Zeiss, Inc.), and focused onto coverslip adherent ovarian tissue using a x63 oil-immersion objective (Carl Zeiss, Inc.). The NLO META scan head (Carl Zeiss, Inc.) allowed data collection. Twenty sections (every fourth slide) for each ovary and three ovaries were analyzed for total number of secondary follicles with positive staining. Five sections (every 10th slide, approximately every 30th section) for each ovary and three ovaries were analyzed for total number of antral follicles with positive staining. For antral follicle colocalization analysis, ten individual follicles from each of three ovaries were analyzed for overall trend of subunit localization in positions 1–4 in antral follicles illustrated in Fig. 1. Position 1 is the mural granulosa cells connected to cumulus granulosa cells in follicle. Position 2 is the cumulus-oocyte complex and cumulus bridge connection to mural granulosa cells in follicle. Position 3 is the mural granulosa cells located a 90-degree angle from cumulus-oocyte complex in follicle. Position 4 is the mural granulosa cells located opposite cumulus-oocyte complex in follicle. A follicle was reported as positive when staining was detected by confocal microscopy. Simultaneous multicolor imaging was conducted to analyze colocalization of inhibin subunits in dominant antral follicle populations for each group analyzed. Ten antral follicles from three different ovaries were analyzed for colocalization. The experiment used Zenon-labeled primary antibodies as described above in immunofluorescence section. Simultaneous multicolor imaging was conducted at positions 1, 3, and 4 of antral follicles. The -subunit was more abundant than the -subunits. As such, we reported on a pixel by pixel basis the proportion (percent) of -subunit that colocalized with either -subunit using LSM 5.10 software. The average percent of colocalization in three different areas of each position for ten different follicles was analyzed.
Statistical analysis
The Pearson product moment correlation coefficient, R, was used to analyze the extent of a linear relationship between hormone profiles across the individual cycles. The t test for zero correlation was used to test the significance of R. An independent sample t test was used to determine significance of the mean percent of -, A-, or B-subunit positive secondary follicles across the estrous and induced cycles. An independent sample t test was used to determine significance of the mean percent of colocalization of subunits across the estrous and induced cycles. In all cases, P < 0.05 was considered significant. Values are reported as ± SEM.
Results
Serum and ovarian inhibin A and inhibin B profiles during the rat estrous and induced cycles
The concentration of inhibin A and inhibin B in serum, ovarian homogenates and follicular fluid during the rat estrous cycle was investigated. Consistent with previous studies, inhibin A and inhibin B serum concentrations were discordant during the estrous cycle (Fig. 2A). Similarly, inhibin A and inhibin B concentrations in the ovary did not correlate, particularly in the latter half of the cycle (Fig. 2B). The ovarian concentrations of inhibin A increased from M1000 to a peak at P1000, fell to a nadir at P1600, increased rapidly by E1400 and declined gradually until M1000. The ovarian concentrations of inhibin B increased from M1000 to a peak at P1000, rapidly declined to a nadir at E0600, increased rapidly by E1000, and fell by E1400 to levels that remained constant until M1000. Serum inhibin A and ovarian inhibin A levels were positively correlated throughout the estrous cycle (Fig. 2C), whereas serum inhibin B and ovarian inhibin B levels were not (Fig. 2D). The follicular fluid levels of inhibin A increased significantly (P = 0.0003) from M1000 to P1600, 465 ± 116 pg/ml to 1853 ± 278 pg/ml respectively. The follicular fluid levels of inhibin B increased but not significantly from M1000 to P1600, 767 ± 150 pg/ml to 1124 ± 160 pg/ml.
The concentrations of estradiol, FSH, LH, and progesterone in serum during the estrous cycle were investigated and previously reported by Chapman and Woodruff (17). As expected, serum inhibin A and estradiol concentrations were positively correlated, whereas serum inhibin B and FSH concentrations were negatively correlated during the estrous cycle. Serum LH concentrations surged beginning at P1600, peaking at P1830 and falling by P2400. Serum progesterone concentrations surged beginning at P1600, peaking at P1830 and falling by E0400.
The concentrations of inhibin A and inhibin B in serum and ovarian homogenates from gonadotropin-treated immature rats were investigated. As expected, the serum concentrations of inhibin A and inhibin B were positively correlated during the induced cycle (Fig. 3A). The ovarian concentrations of inhibin A and inhibin B were also positively correlated during the induced cycle (Fig. 3B). Ovarian concentrations of inhibin A and inhibin B rose with PMSG treatment to a peak by 48 h after PMSG injection, fell to a nadir by 12 h after hCG injection and began to rise by 24 h after hCG injection (Fig. 3B). Serum and ovarian concentrations of each inhibin were positively correlated during the induced cycle (Fig. 3, C and D).
The concentrations of estradiol and progesterone in serum during the induced cycle were investigated (Fig. 3, E–G). Both serum inhibin A and serum inhibin B concentrations were positively correlated with estradiol during the induced cycle (Fig. 3, E and F). Serum progesterone concentrations were constant during first 48 h after PMSG injection, peaked at 6 h after hCG injection and fell by 12 h after hCG injection to a level higher than before the peak (Fig. 3G).
Inhibin subunit localization in the ovary during the estrous and induced cycles
The pattern of the inhibin subunits in follicles present in ovaries collected during the estrous cycle was investigated. Inhibin subunits were mainly expressed in granulosa cells during the estrous cycle. The total number of follicles positive for individual subunits was analyzed. Of all follicles positive for the inhibin -subunit, at M1000, 74%; at D1000, 81%; and at P1600, 73% of all follicles positive for the -subunit were positive for A-subunit. At M1000, 88%; at D1000, 86%; and at P1600, 72% were positive for B-subunit. Corpora lutea did not express any of the inhibin/activin subunits at any time during the estrous cycle.
All inhibin subunits were expressed in secondary follicles during the estrous cycle (Fig. 4, A–L). At M1000 the inhibin -, A- and B-subunits were expressed in 91%, 85% and 88% of secondary follicles, respectively (Fig. 4, A–C and G–I). At P1600, the inhibin -, A-, and B-subunits were expressed in 89%, 67%, and 90% of secondary follicles, respectively (Fig. 4, D–F and G–I). Fewer secondary follicles expressed the A-subunit when compared with the - and B-subunits at P1600 (P < 0.01).
The pattern of the inhibin subunits in ovaries during the induced cycle was similarly investigated. Inhibin subunits were detected mainly in granulosa cells during the induced cycle. Of all follicles positive for the inhibin -subunit, at 12 h after PMSG injection 78% and 48 h after PMSG injection 91% of follicles were also positive for -subunits. The A- and B-subunits were produced in same follicles.
Inhibin subunits were expressed in granulosa cells of secondary follicles during the induced cycle (Fig. 5, A–L). At 12 h after PMSG injection, the inhibin -, A-, and B-subunits were expressed in 87%, 71%, and 90% of secondary follicles, respectively (Fig. 5, A–C and G–I). Fewer secondary follicles expressed the A-subunit when compared with - and B-subunits at 12 h after PMSG injection (P < 0.01). At 48 h after PMSG injection the inhibin -, A- and B-subunits were expressed in 92%, 87%, and 91% of secondary follicles, respectively (Fig. 5, D–F and J–L).
Inhibin subunit colocalization in antral follicles during the estrous cycle
Colocalization of the inhibin subunits in antral follicles during the estrous cycle was investigated. The schematic in Fig. 1 describes the four different areas of granulosa cells analyzed in the colocalization studies. Representative immunoreactivity staining of the inhibin - and A-subunits in follicles from M1000 and P1600 during the rat estrous cycle is shown in Fig. 6. The proportion of -subunit pixels colocalized with A-subunit pixels is reported in Fig. 7A. Mural granulosa cell staining for the - and A-subunits was higher than staining in cumulus granulosa cells (Fig. 6, A–X). At M1000 both the - and A-subunit staining were more intense in the basement membrane-facing mural granulosa cells, especially at positions closest to the outside of ovary (position 4 in this reference follicle) (Fig. 6, S–U). At P1600 the - and A-subunits staining was localized throughout the mural granulosa cells with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions 3 and 4, closest to the outside of the entire ovary (Fig. 6, P–R and V–X). The - and A-subunits were colocalized in a higher number of mural granulosa cells on P1600 than on M1000 (Fig. 6, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with A-subunit pixels was found to be significantly higher at P1600 in positions 3 and 4 when compared with M1000 (Fig. 7A).
Colocalization of the inhibin - and B-subunits in antral follicles during the estrous cycle was similarly investigated. Representative immunoreactivity staining in follicles from M1000 and P1600 during the rat estrous cycle is shown in Fig. 8. The proportion of -subunit pixels colocalized with B-subunit pixels is reported in Fig. 7B. Mural granulosa cell staining for the - and B-subunits was greater than staining in cumulus granulosa cells (Fig. 8, A–X). The -subunit staining was relatively uniform throughout the mural granulosa (Fig. 8, M and S). The B-subunit staining was relatively uniform throughout the mural granulosa cells at M1000 (Fig. 8B, N and T). At P1600, the -subunit was localized throughout the mural granulosa cells, with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions located closest to the outside of ovary (positions 3 and 4 in the representative follicle) (Fig. 8, P and V). In contrast, the B-subunit staining at P1600 was more intense in antral-facing mural granulosa cells and in most outward basement membrane-facing mural granulosa cells, which was most apparent at positions closest to the outside of entire ovary (positions 3 and 4 in the representative follicle) (Fig. 8, Q and W). The - and B-subunits were colocalized in a greater proportion of mural granulosa cells on M1000 than on P1600 (Fig. 8, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with B-subunit pixels was found to be significantly greater at M1000 in positions 1, 3, and 4 when compared with P1600 (Fig. 8B).
Inhibin subunit colocalization in antral follicles during the induced cycle
Finally, colocalization of the inhibin - and -subunits in antral follicles was investigated during the induced cycle. Representative immunoreactivity staining of - and A subunits in follicles from 12 and 48 h after PMSG injection are shown in Fig. 9. The proportion of -subunit colocalized with A-subunit is reported in Fig. 10A. The - and A-subunits were localized to all granulosa cells at 12 and 48 h after PMSG injection during the induced cycle (Fig. 9, A–X). Mural granulosa cell staining for the - and A-subunits was greater than staining in cumulus granulosa cells (Fig. 9, G–L). Both the - and A-subunit showed relatively uniform staining throughout the mural granulosa cells at 12 h after PMSG injection (Fig. 9, A–C, M–O, and S–U). At 48 h after PMSG injection, -subunit staining was localized throughout the mural granulosa cells with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions located closest to the outside of the entire ovary (position 3 and 4 in the representative follicle) (Fig. 9, P and V). In contrast, A-subunit staining was localized relatively uniformly throughout mural granulosa cells (Fig. 9, Q and W). The - and A-subunits were colocalized in a greater proportion of mural granulosa cells at 48 h after PMSG injection than at 12 h after PMSG injection (Fig. 9, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with A-subunit pixels was found to be significantly greater at 48 h after PMSG injection in positions 3 and 4 when compared with 12 h after PMSG injection (Fig. 10A).
Representative immunoreactivity staining of the inhibin - and B-subunits in follicles from 12 and 48 h after PMSG injection during the rat induced cycle is shown in Fig. 11. The proportion of -subunit colocalized with B-subunit is reported in Fig. 10B. The - and B-subunit were localized to all granulosa cells at 12 and 48 h after PMSG injection during the induced cycle (Fig. 11, A–X). Mural granulosa cell staining for the - and B-subunits was greater than staining in cumulus granulosa cells. Both the - and B-subunits staining are relatively uniform throughout the mural granulosa cells at 12 h after PMSG injection (Fig. 11, A–C, M–O, and S–U). At 48 h after PMSG injection, the -subunit staining was localized throughout the mural granulosa cells with an increased intensity in areas of basement membrane-facing mural granulosa cells, most notably at positions located closest to the outside of the entire ovary (positions 3 and 4 in this representative follicle) (Fig. 11, P and V). In contrast, the B-subunit was localized throughout the mural granulosa cells with an intense increase in staining at the very antral edge of antral-facing mural granulosa cells (Fig. 11, Q and W). The - and B-subunits were colocalized in a greater proportion of mural granulosa cells at 12 h after PMSG injection than at 48 h after PMSG injection (Fig. 11, C, F, I, L, O, R, U, and X). The proportion of -subunit pixels colocalized with B-subunit pixels was found to be significantly greater at 12 h after PMSG injection in positions 3 and 4 when compared with 48 h after PMSG injection (Fig. 10B).
Discussion
In previous work, we observed differential compartmentalization of the mRNAs encoding the inhibin -, A-, and B-subunits in the Siberian hamster (10). Specifically, in Graafian follicles of long photoperiod female hamsters the B-subunit was detected in the antral-facing mural granulosa cell layer, the -subunit was relatively uniform throughout mural granulosa cell layer and the A-subunit was uniformly distributed with increased expression in basement membrane-facing mural granulosa cell layer. Early studies hinted at but did not comment on the spatially restricted localization of inhibin subunit mRNA and protein in a variety of species (9, 11, 12, 13, 14, 15). The intent of this study was to investigate the type of follicles producing the inhibin subunits, and specifically the granulosa cells within these follicles, to investigate the differential compartmentalization of inhibin subunits in the rat ovary to gain insight into the difference between inhibin A and inhibin B production and function. We analyzed the estrous cycle and PMSG-induced cycle models to further understand how different follicle class sizes are related to the differential secretion of inhibin A and inhibin B. The induced cycle model was used as a positive control. If a heterogeneous population of granulosa cells within a single follicle is necessary to obtain discordant inhibin A and inhibin B secretion during the estrous cycle, then this relationship would be totally abolished during the induced cycle, where the follicle selection, growth and ovulation are synchronized.
As expected from previous studies, inhibin A and inhibin B were differentially secreted during the estrous cycle (5, 7). In addition, we found that early on estrus inhibin B levels rose more quickly and to a higher concentration, compared with inhibin A. The inhibin A ovarian and serum concentrations were highly correlated, whereas inhibin B concentrations in the serum and ovary throughout the cycle did not show a strong positive correlation. One explanation of these data is that the time between inhibin B production and secretion is much shorter at 1000 h on metestrus than at 1600 h on proestrus. In addition, we found that inhibin A levels were significantly greater in the follicular fluid of dominant follicles when compared with smaller antral follicles, and although the trend was similar for inhibin B it was not significant. These results are consistent with previous measurement of inhibin in human follicular fluid (18). Inhibin B may be retained in the ovary within individual granulosa cells or in the follicular fluid of small follicles on proestrus. The physiological significance of the rise in inhibin B on the morning of estrus is to cause the immediate inhibition of FSH at a time when neither estradiol nor inhibin A have been stimulated. As the cycle progresses, both inhibin A and estradiol are secreted, reinforcing the negative feedback of the ovary to the pituitary.
The second model in this study was the induced cycle in immature female rats treated with PMSG and hCG. Sequential administration of these hormones induces a synchronized selection of follicles into the growing pool and subsequent ovulation. During the induced cycle, there was a strong positive correlation between serum inhibin A and inhibin B levels, ovarian inhibin A and inhibin B levels, serum and ovarian inhibin A levels, and serum and ovarian inhibin B levels, which suggest that exogenous FSH and LH overrides the control mechanisms that limit inhibin B secretion to discrete times during the estrous cycle.
We investigated the compartmental localization of inhibin subunits in the ovary to further elucidate the differential secretion of inhibin A and inhibin B during the estrous cycle. Distribution of inhibin -, A-, and B-subunits in different ovarian compartments was studied using immunofluorescence. The proportion of -subunit pixels that colocalized with the A-subunit pixels was significantly greater in positions 3 and 4 of mural granulosa cells at 1600 h on proestrus than at 1000 h on metestrus. The opposite pattern of colocalization was observed between the - and B-subunit. The proportion of -subunit pixels that colocalized with the B-subunit pixels was significantly greater at positions 1, 3, and 4 of mural granulosa cells at 1000 h on metestrus than at 1600 h on proestrus. All mural granulosa cells have the potential to produce inhibin A and inhibin B at 1000 h on metestrus and 1600 h on proestrus. However, antral follicles show greater inhibin B colocalization at 1000 h on metestrus, and greater A colocalization at 1600 h on proestrus. The overall trend of the immunostaining experiments suggests greater inhibin A assembly with the highest potential for secretion toward the basement membrane at 1600 h on proestrus. In contrast, inhibin B assembly was greater in metestrus, whereas at 1600 h on proestrus the potential for secretion was toward the antrum and the thin line of mural granulosa cells lining the basement membrane at positions closest to the outside of the ovary. Thus, the distinct compartmentalization of each inhibin isoform, the increased colocalization of - and B-subunits at 1000 h on metestrus, and the increased colocalization of the - and A-subunits at 1600 h on proestrus may contribute to the discordance in serum levels of inhibin A and inhibin B observed throughout the normal estrous cycle. Inhibin A is secreted as it is made in increasing concentrations by the growing population of granulosa cells during metestrus to proestrus, whereas inhibin B is secreted during estrus/early metestrus and accumulates in ovary during proestrus.
We investigated variations in the localization of production of the inhibin subunits in the large pool of growing antral follicles during the induced cycle. Distribution of inhibin -, A-, and B-subunits in different ovarian compartments was studied using immunofluorescence at two particular time points, 12 and 48 h after PMSG treatment. Interestingly, the - and A-subunits were colocalized in a greater proportion of mural granulosa cells at 48 h after PMSG injection than at 12 h after PMSG injection, particularly at positions 3 and 4. The - and B-subunits had the opposite pattern, colocalizing in a greater proportion of granulosa cells at 12 h after PMSG injection than at 48 h after PMSG injection, and again, colocalization occurred most frequently at positions 3 and 4. Thus, the subunit colocalization patterns at 12 and 48 h after PMSG treatment were consistent with those seen in the dominant antral follicle populations at 1000 h on metestrus and 1600 h on proestrus. These results confirm that inhibin subunit compartmentalization in antral follicles during folliculogenesis occurs, and that inhibin B is likely produced in smaller selected antral follicles that switch to inhibin A production once they develop into Graafian follicles. The induced cycle model shows a synchronized switch from inhibin B to inhibin A assembly as follicle size increases, with no difference in inhibin secretion between the two isoforms. Synchronization of follicle development with exogenous gonadotropins overrides the granulosa cell heterogeneity that underlies the discordant secretion of inhibin A and inhibin B during the cycle.
The discordance of inhibin A and inhibin B secretion during the estrous cycle was originally thought to be due to different populations of follicles making inhibin A and inhibin B. This study reveals that all follicles are capable of producing both inhibin isoforms, and that heterogeneity in a single follicle’s granulosa cell population is the basis for the temporal differences in inhibin A and inhibin B secretion into the blood. The secretion rate may be determined by the location of assembled dimers within the follicles (i.e. antrum-facing vs. basement membrane-facing granulosa cells). Little is known about the intracellular mechanisms responsible for controlling secretion of the inhibins by a granulosa cell. Spatially, the granulosa cells located near the edge of individual follicles would have the greatest access to the vasculature. The patterns of inhibin assembly and release are consistent with the local, concentration-dependent role of inhibin A in granulosa cell growth and development and an endocrine, time-dependent role of inhibin B on inhibition of pituitary FSH expression. The mechanisms controlling the exquisitely timed assembly and release of each inhibin isoform should be further examined. Little is known about the ability to process, store, degrade, and secrete the dimeric inhibins in vivo. A deeper understanding of inhibin biosynthesis and secretion is crucial to understanding mechanisms of inhibin action within the reproductive axis. Future studies will investigate the role of disulfide bond formation, processing, and cleavage involved in the secretion of inhibin A and inhibin B.
Acknowledgments
We thank Dr. W. W. Vale for the inhibin and activin subunit polyclonal antibodies. We also thank the Diagnostic Systems Laboratories for the ELISA kits. In addition, we thank Dr. Alfred Rademaker for his expertise in statistical analysis. Finally, we would like to express sincere appreciation for the participation of Dr. Stacey Chapman-Tobin and Dr. Signe Kilen in this study.
Footnotes
This work was supported by National Institutes of Health (NIH) Grant HD-37096 (to T.K.W.) and by National Institute of Child Health and Human Development/NIH through Cooperative Agreement U54-HD-28934 as part of the Specialized Cooperative Centers Program in Reproductive Research.
First Published Online September 29, 2005
Abbreviations: hCG, Human chorionic gonadotropin; PMSG, pregnant mare serum gonadotropin.
Accepted for publication September 19, 2005.
References
Woodruff TK, D’Agostino J, Schwartz NB, Mayo KE 1988 Dynamic changes in inhibin messenger RNAs in rat ovarian follicles during the reproductive cycle. Science 239:1296–1299
Hsueh AJ, Dahl KD, Vaughan J, Tucker E, Rivier J, Bardin CW, Vale W 1987 Heterodimers and homodimers of inhibin subunits have different paracrine action in the modulation of luteinizing hormone-stimulated androgen biosynthesis. Proc Natl Acad Sci USA 84:5082–5086
Rivier C, Vale W, Rivier J 1987 Studies of the inhibin family of hormones: a review. Horm Res 28:104–118
Spaczynski RZ, Tilly JL, Mansour A, Duleba AJ 2005 Insulin and insulin-like growth factors inhibit and luteinizing hormone augments ovarian theca-interstitial cell apoptosis. Mol Hum Reprod 11:319–324
Arai KY, Ohshima K, Watanabe G, Arai K, Uehara K, Taya K 2002 Dynamics of messenger RNAs encoding inhibin/activin subunits and follistatin in the ovary during the rat estrous cycle. Biol Reprod 66:1119–1126
Bernard DJ, Woodruff TK 2001 Inhibin binding protein in rats: alternative transcripts and regulation in the pituitary across the estrous cycle. Mol Endocrinol 15:654–667
Woodruff TK, Besecke LM, Groome N, Draper LB, Schwartz NB, Weiss J 1996 Inhibin A and inhibin B are inversely correlated to follicle-stimulating hormone, yet are discordant during the follicular phase of the rat estrous cycle, and inhibin A is expressed in a sexually dimorphic manner. Endocrinology 137:5463–5467
Woodruff TK, Krummen LA, Chen S, Chen S, DeGuzman G, Lyon R, Baly DL, Allison DE, Garg S, Wong WL, Hebert N, Mather JP, Cossum P 1993 Pharmacokinetic profile of recombinant human (rh) inhibin A and activin A in the immature rat. I. Serum profile of rh-inhibin A and rh-activin A in the immature female rat. Endocrinology 132:715–724
Meunier H, Cajander SB, Roberts VJ, Rivier C, Sawchenko PE, Hsueh AJ, Vale W 1988 Rapid changes in the expression of inhibin -, A-, and B-subunits in ovarian cell types during the rat estrous cycle. Mol Endocrinol 2:1352–1363
Kenny HA, Bernard DJ, Horton TH, Woodruff TK 2002 Photoperiod-dependent regulation of inhibin in Siberian hamsters: I. Ovarian inhibin production and secretion. J Endocrinol 174:71–83
Meunier H, Roberts VJ, Sawchenko PE, Cajander SB, Hsueh AJ, Vale W 1989 Periovulatory changes in the expression of inhibin -, A-, and B-subunits in hormonally induced immature female rats. Mol Endocrinol 3:2062–2069
van den Hurk R, Van de Pavert SA 2001 Localization of an activin/activin receptor system in the porcine ovary. Mol Reprod Dev 60:463–471
Nagamine N, Nambo Y, Nagata S, Nagaoka K, Tsunoda N, Taniyama H, Tanaka Y, Tohei A, Watanabe G, Taya K 1998 Inhibin secretion in the mare: localization of inhibin , A, and B subunits in the ovary. Biol Reprod 59:1392–1398
Tisdall DJ, Hudson N, Smith P, McNatty KP 1994 Localization of ovine follistatin and and A inhibin mRNA in the sheep ovary during the oestrous cycle. J Mol Endocrinol 12:181–193
Silva JR, van den Hurk R, van Tol HT, Roelen BA, Figueiredo JR 2004 Gene expression and protein localization for activin-A, follistatin and activin receptors in goat ovaries. J Endocrinol 183:405–415
Herath CB, Yamashita M, Watanabe G, Jin W, Tangtrongsup S, Kojima A, Groome NP, Suzuki AK, Taya K 2001 Regulation of follicle-stimulating hormone secretion by estradiol and dimeric inhibins in the infantile female rat. Biol Reprod 65:1623–1633
Chapman SC, Woodruff TK 2003 Betaglycan localization in the female rat pituitary: implications for the regulation of follicle-stimulating hormone by inhibin. Endocrinology 144:5640–5649
Schneyer AL, Fujiwara T, Fox J, Welt CK, Adams J, Messerlian GM, Taylor AE 2000 Dynamic changes in the intrafollicular inhibin/activin/follistatin axis during human follicular development: relationship to circulating hormone concentrations. J Clin Endocrinol Metab 85:3319–3330(Hilary A. Kenny and Teresa K. Woodruff)