Effects of Gonadotropin Treatment and Withdrawal on Follicular ...

BIOLOGY OF REPRODUCTION 55, 693-702 (1996)

Effects of Gonadotropin Treatment and Withdrawal on Follicular Growth, Cell Proliferation, and Atresia in Ewes' Albina Jablonka-Shariff, Lawrence P. Reynolds, and Dale A. Redmer 2 Department of Animal and Range Sciences, North Dakota State University, Fargo, North Dakota 58105 ABSTRACT To determine the effects of FSH-P treatment and subsequent withdrawal on follicular growth, cell proliferation, and atresia, ewes (n = 4 ewes/treatment group) received twice daily injections of saline or FSH-P beginning on Day 13 of the estrous cycle (length of the estrous cycle = 16.5 days) and were slaughtered after 0, 48, or 72 h of treatment (i.e., on Days 13, 15, or 16). Some treatment groups received FSH-P from Day 13 until slaughter (FSH-P-treated), whereas some received FSH-P for 2448 h followed by saline for 24-48 h (FSH-P withdrawal). All ewes received an i.v. injection of bromodeoxyuridine (BrdU, a thymidine analogue) 1 h before slaughter. For both ovaries from each ewe, the number and surface diameter of all visible antral follicles were determined, and antral follicles were classified as small ( 3 mm), medium (> 3 mm to 6 mm), or large (> 6 mm). As an index of the rate of cell proliferation, BrdU was immunolocalized in paraffin-embedded tissue sections, and the labeling index (LI; BrdU-labeled nuclei as a percentage of total nuclei) was determined for granulosa and thecal cells of nonatretic and early atretic antral follicles of known diameter. Follicular status (atretic vs. nonatretic) was evaluated morphologically. Moreover, the presence of apoptosis was detected in situ by using the terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling method. For untreated and saline-treated ewes, the number of small follicles per ewe increased (p < 0.01) from Day 13 to Day 15, then decreased again on Day 16, whereas numbers of medium and large follicles did not differ across days. Compared with saline-treated ewes, ewes receiving FSH-P from Day 13 until slaughter had fewer (p < 0.05) small but more (p < 0.05) medium and large follicles. Compared with FSH-Ptreated ewes, FSH-P withdrawal resulted in fewer (p < 0.05) medium and large but more (p < 0.05) small follicles. Across all follicular size classes, granulosa and thecal cell LI of nonatretic follicles was decreased (p < 0.05) by FSH-P withdrawal compared with FSH-P treatment. Additionally, across all follicular size classes, FSH-P withdrawal increased (p < 0.01) the percentage of follicles that were atretic compared with saline or FSH-P treatment. Histochemical staining of early and advanced atretic follicles showed that granulosa cells are the predominant site of cell death (apoptosis) during follicular atresia. Thus, compared with continuous FSH-P treatment, withdrawal of FSH-P resulted in decreased numbers of medium and large follicles, decreased proliferation of follicular cells, and an increased incidence of atresia associated with granulosa cell death. This model should prove useful for studying the mechanisms regulating follicular growth and atresia in ewes. INTRODUCTION In sheep, as in other mammals, preovulatory follicular development begins with the recruitment and growth of a cohort of primary follicles within the ovary [1-3]. SucAccepted May 8, 1996. Received November 6, 1995. 'This work was supported, in part, by USDA competitive grant 9337203-9721 (to D.A.R. and L.P.R.). 2 Correspondence. FAX: (701) 231-7590; e-mail: [email protected] 693

cessful development to the preovulatory stage requires both proliferation and differentiation of follicular granulosa and thecal cells [4-6]. The vast majority of ovarian follicles never ovulate or form corpora lutea but rather become atretic at different stages of follicular development [7-9]. Studies in several species have demonstrated that follicular growth and development is mediated by gonadotropins [10-14]. Administration of exogenous gonadotropic preparations containing FSH-like activity has been shown to change the balance between healthy and atretic follicles by preventing or delaying atresia [9, 14-16]. A previous study from our laboratory demonstrated that treatment of ewes with FSH-P for 24, 48, or 72 h increased the number of medium and large antral follicles and decreased the incidence of atresia compared with saline treatment [17]. In addition, ewes receiving FSH-P for 24 h showed increased proliferation of follicular granulosa cells. We therefore postulated that FSH-P treatment of ewes not only induced recruitment and growth of follicles but also decreased the incidence of atresia [17]. Several models for atresia have been developed in an attempt to define when atresia begins and to identify the mechanism(s) by which gonadotropins modulate the incidence of atresia [9, 18-20]. However, despite these and other studies, the mechanisms by which FSH influences follicular growth and atresia in ewes are still unclear. Recent studies have suggested that programmed cell death (apoptosis) is the primary degenerative process that occurs during follicular atresia in chickens [21], sows [22, 23], cows [24], and rats [23, 25, 26]. Moreover, a recently developed method (3'-end labeling of fragmented DNA in situ) to localize apoptosis on histological sections could provide further information concerning the underlying mechanisms of follicular atresia [26-28]. The objective of this study, therefore, was to determine the effects of FSH-P treatment and subsequent withdrawal on follicular growth, cell proliferation, and atresia in ewes. MATERIALS AND METHODS Animals and Tissue Preparation Thirty-two ewes of mixed breeding (Western range ewes) that had exhibited an estrous cycle of normal duration (16.5 + 0.3 days) immediately preceding the treatment cycle were used for this study. Day 0 of the estrous cycle (standing estrus) was determined by using vasectomized rams. Beginning on Day 13 after estrus, ewes received no hormone or twice daily i.m. injections of saline or FSH-P (decreasing daily doses of 5, 4, and 3 mg on Days 13, 14, and 15, respectively; Schering, Kenilworth, NJ) diluted in saline, as we have described previously [17]. Ewes were assigned randomly for slaughter on Days 13, 15, or 16 (i.e., after 0, 48, or 72 h of saline or FSH-P treatment and subsequent withdrawal). The specific treatments are summarized on the left-hand side of Table 1. All ewes received an intravenous injection of 5-bromo-

JABLONKA-SHARIFF ET AL.

694

TABLE 1. Effects of FSH-P treatment and withdrawal on numbers of surface follicles per ewe. Treatment Treatment group

13

1 2 3 4 5 6 7 8

X S F F S F F F

t

Day* 14 S S F S S F F

15 X X X S S S F

16

No. of ewes

X X X X

4 4 4 4 4 4 4 4

No. of follicles by size class Large Medium Small 14.8 2 8.5 2 2 .5 13.3 17.8 21.0 10.3 5.8

± 3.1

+ 4. 8 b ± 4. 7 b a

± + + +

2.7 3.1 5.4 b 3.8 1.1

2.5 3.5 6.3 17.8 2.8 7.5 11.0 11.8

± 0.9 a ± 0.61 ± 1.5 "b ± 3.9c ± 1.4, 2.7" ± 2.9 b' ± 4.6b,

1.5 1.0 1.5 5.5 0.8 2.5 1.5 7.0

+ 0.6 a ± 0.4 ± 0.3 d ± 1.6 ' ± 0.31 0.9' ± 0.9' ± 2.9'

* X = day of slaughter, S = saline treatment, and F = FSH-P treatment on Days 13, 14, 15, and 16. 3 mm; medium, >3 mm to -6 mm; large, >6 mm. t Follicular size class: small, SEM with different superscripts differ (p < 0.05). i,'- Within each follicular size class, means

2-deoxyuridine (BrdU, a thymidine analogue; 5 mg/kg BW; Aldrich, Milwaukee, WI) 1 h before slaughter (-0800 h), as we have described previously [17, 29, 30]. Jugular venous blood samples (10 ml) were collected into heparinized tubes (Vacutainer; Becton-Dickinson, Rutherford, NJ) immediately before BrdU injection, and plasma from these samples was stored at -70°C until assayed for estradiol and progesterone concentrations. At slaughter, reproductive tracts were collected, placed on ice, and immediately transported to the laboratory. For both ovaries from each ewe, the number and surface diameter of all visible follicles were recorded, and antral follicles were classified by size as fol3 mm; medium, > 3 mm to lows: small, - 1 mm to 6 mm; large, > 6 mm [17]. In addition, the location of all visible follicles within an ovary was diagrammed so that follicular size before fixation was known [17]. All follicles that were subsequently analyzed histologically were accounted for on these surface diagrams. To flush blood cells and dilate the ovarian vascular beds, each ovary was perfused via the main ovarian artery with PBS (0.01 M phosphate and 0.14 M NaCl, pH 7.3) containing 0.1% (v:v) lidocaine, followed by perfusion with Carnoy's solution or 10% (v:v) buffered neutral formalin, as described previously [17, 29]. For each ewe, one randomly chosen ovary was fixed with each fixative. After perfusion fixation, each ovary was sliced along its transverse axis (i.e., perpendicular to the ovarian hilus) into 25 pieces (-5 mm thick), each of which contained one face that included several follicles of known diameter. These ovarian pieces were then immersed in their respective fixative (Carnoy's solution for 4 h or buffered neutral formalin for 24 h) and subsequently were processed for histology as we have previously described [17, 29, 30]. Tissue sections fixed in Carnoy's solution were used for immunohistochemical localization of BrdU, whereas those fixed in buffered neutral formalin were used for histochemical localization of apoptosis (see below). It should be emphasized that maximization of the number of follicles of known diameter that would be present in each histological section was the only criterion used in selecting the ovarian pieces. In addition, all ovarian pieces from each ovary were used for subsequent histological analysis. Thus, each follicle of known surface diameter had an equal chance of being included in the histological evaluation of cell proliferation, atresia, or apoptosis. Histology and Histochemistry The proportions of the total small, medium, and large follicles of known surface diameter that were subsequently

analyzed by using the ovarian pieces described above were 49.6%, 72.6%, and 83.3%, respectively. For each ovarian piece, at least one randomly chosen cross section of each follicle of known surface diameter was analyzed. To visualize proliferating granulosa and thecal cells, BrdU was immunolocalized in randomly selected Carnoy'sfixed, paraffin-embedded tissue sections (5 pm) by using a specific anti-BrdU monoclonal antibody (Boehringer Mannheim, Indianapolis, IN) and the modified avidin-biotinylated peroxidase complex (ABC) method, as we have described previously [17, 29, 30]. Control staining consisted of replacing the primary antibody with the same dilution of mouse ascites fluid (ICN Biochemicals, Costa Mesa, CA). Sections were counterstained briefly (-10 sec) with Harris' hematoxylin to visualize unlabeled nuclei [17, 29, 30]. To confirm that follicular cells that appeared to be apoptotic on the basis of their morphology had the fragmented DNA characteristic of apoptosis, randomly selected formalin-fixed, paraffin-embedded tissue sections (5 Rm) were stained by using the TUNEL (terminal deoxynucleotidyl transferase [TdT]-mediated dUTP-biotin nick end-labeling) method, as we have reported previously [28]. This method is based upon the specific addition by TdT of dUTP-biotin to the 3'-OH ends of the fragmented nucleosomal DNA of apoptotic cells [27]. Briefly, rehydrated tissue sections were treated with 2 N HCL for 30 min to denature nuclear chromatin [29, 30], followed by 3% (v:v) hydrogen peroxide for 10 min to quench endogenous peroxidase activity. The sections were rinsed with double-distilled water and immersed in TdT buffer (30 mM Trizma base, 140 mM sodium cacodylate, 1 mM cobalt chloride, pH = 7.2) for 10 min. Sections then were incubated in TdT buffer containing TdT (0.22 e.u./pl; Promega, Madison, WI) and biotin-11dUTP (4 IM; Sigma, St. Louis, MO) for 2 h at 37°C in a humidified atmosphere. The reaction was terminated by transferring the slides to citrate buffer (300 mM sodium chloride, 30 mM sodium citrate) for 15 min at room temperature. The sections were rinsed with double-distilled water and incubated with an aqueous solution of BSA (2 plg/ ml; fraction V, Sigma) for 10 min. Control staining consisted of incubating tissue sections in buffer containing only TdT or only dUTP (i.e., without substrate or without enzyme, respectively). Nuclei exhibiting DNA fragmentation were visualized by using the modified ABC method with VIP (Vector Laboratories, Burlingame, CA) as the peroxidase substrate. Sections were counterstained briefly with Harris' hematoxylin to visualize unlabeled nuclei. Alternatively, apoptotic nuclei were visualized by using fluores-

FSH-P WITHDRAWAL AND FOLLICULAR ATRESIA ceinated avidin D (avidin-fluorescein isothiocyanate; Vector), and tissue sections were incubated for 45 min at room temperature in the dark. Fluorescence was examined by using a Nikon microscope (Microphot FX; Nikon, Garden City, NY). Labeling Index (LI) Granulosa and thecal cell LI was estimated for nonatretic and early atretic follicles (follicles in advanced atresia did not contain BrdU-labeled cells) by using computerized image analysis, as we have described previously [17, 29, 30]. The LI was expressed as the area of BrdU-labeled nuclei as a percentage of the total nuclear area. The mean LI was based upon three estimates for each cell type from each follicle, each of which was taken from a different area within the cross-section of that follicle. Thus, for each follicle, the LI for each cell type was estimated from several hundred nuclei, and, for each ewe, the LI for each cell type was based upon counts of an average of -7000 nuclei for small follicles, 3000 nuclei for medium follicles, and 1000 nuclei for large follicles. The coefficient of variation for the triplicate counts taken from different areas of the same fol1.4% for licle was 20.6 + 2.0% for granulosa and 23.4 thecal cells across all follicles. We have previously shown that this method of estimating the LI compares favorably with manual morphometric counting procedures [30]. Follicular Status For each of the randomly selected tissue sections that were used for histochemistry, follicular status of all antral follicles of known surface diameter (see above) was evaluated morphologically, and follicles were classified as nonatretic (healthy), early atretic, or advanced atretic, as we and others have described before [17, 31]. Follicles classified as nonatretic were characterized by the presence of a well-organized granulosa cell layer, an absence of pyknotic nuclei, and BrdU-labeled granulosa and thecal cell nuclei. Early atretic follicles exhibited loosening of the granulosa cells lining the antrum, a few pyknotic granulosa cell nuclei, and a relatively low rate of follicular cell proliferation. Follicles in advanced atresia exhibited an abundance of pyknotic nuclei, clumps of fragmented chromatin within the granulosa cell layer or follicular antrum, and no BrdU-labeled granulosa or thecal cells. Estradiol and Progesterone RIA Estradiol concentrations were measured in benzene extracts of plasma as previously validated in our laboratory [32]. The assay sensitivity was 1 pg/tube. All samples were assayed in a single assay, and the intraassay coefficient of variation was 3.4%. The recovery of [3 H]estradiol in ex1.5%, and concentrations of tracted samples was 93.5 estradiol were adjusted for recovery. Progesterone concentrations were measured in benzene: hexane (1:2) extracts of plasma as previously described [32]. The assay sensitivity was 25 pg/tube. All samples were assayed in a single assay, and the intraassay coefficient of variation was 2.6%. The recovery of 2.1%, [3 H]progesterone in extracted samples was 87.6 and concentrations of progesterone were adjusted for recovery. Statistical Analyses Number of surface follicles, LI of granulosa and thecal cells, and proportion of atretic follicles were analyzed as a

695

split-plot design by using the general linear models (GLM) procedure of SAS, with treatment group, follicular size class and/or cell type, and their interactions included in the model [33, 34]. For treatment group, animal within group was used as the error term [33, 34]. For estradiol and progesterone concentrations in plasma, the data were analyzed for treatment group effects by using GLM. To meet the assumption of homogeneity of variances [34], the data were transformed before statistical analysis by using a square root (LI) or log (number of follicles, estradiol, and progesterone concentrations) transformation; for the proportion of atretic follicles, the data were first subjected to the arcsin transformation for proportional data [34]. When an F test was significant (p < 0.05), differences between specific means were evaluated by using Bonferroni's t-test [34], including preplanned comparisons of saline-treated (groups 2 and 5) vs. FSH-treated (groups 4 and 8) vs. FSH withdrawal (groups 3, 6, and 7) ewes. Data are reported as means + SEM. RESULTS Number of Surface Follicles Across all treatment groups, the mean number of antral (small, medium, and large) surface follicles was 27.3 ± 2.2 per ewe. Across all treatments groups, the mean number of antral follicles decreased (p < 0.01) as follicular size increased (16.8 ± 2.5, 7.9 1.8, and 2.7 0.8/ewe for small, medium, and large follicles, respectively). As shown in Table 1, for untreated and saline-treated ewes, the number of small follicles per ewe increased (p < 0.05) from Day 13 to Day 15 (groups 1 and 2, respectively), then decreased on Day 16 (group 5); in contrast, the number of medium and large follicles did not differ across days, and 0.2/ewe, respectively. 0.3 and 1.1 averaged 2.9 Compared with their respective saline-treated controls, ewes receiving FSH-P from Day 13 until slaughter on Day 15 (group 2 vs. group 4) or on Day 16 (group 5 vs. group 8) had fewer (p < 0.05) small but more (p < 0.05) medium and large follicles (Table 1). Compared with ewes receiving continuous FSH-P treatment, FSH-P treatment and subsequent withdrawal did not change the total number of antral 2.6/ewe for FSH withdrawal follicles per ewe (28.0 [groups 3, 6, and 7] vs. 30.5 ± 6.0/ewe for continuous FSH [groups 4 and 8]). However, withdrawal of FSH-P altered the number of antral follicles within specific follicular size classes. Compared with FSH-P-treated ewes, those in which FSH was withdrawn had more (p < 0.05) small (group 3 vs. 4, and group 6 vs. 8) but fewer (p < 0.05) medium (group 3 vs. 4) and large (group 3 vs. 4, and groups 6 and 7 vs. 8) follicles (Table 1). Interestingly, ewes receiving two days of FSH and only one day of saline treatment (group 7) had an intermediate number of small follicles compared with their saline-treated (group 6) and FSH-P-treated (group 8) counterparts. Moreover, although group 7 ewes had a similar number of medium follicles, they had fewer large follicles compared with group 8 (Table 1). BrdU Immunohistochemistry Consistent with our previous report [17], immunohistochemical staining for BrdU was highly specific, as shown by its nuclear localization, paucity of staining in ovarian stroma, and lack of staining in control sections (Fig. 1, A and B). Immunolocalization of BrdU was restricted to the nuclei of granulosa, thecal, and vascular endothelial cells

696 FIG. 1. Immunohistochemical localization of BrdU in sections of nonatretic follicles of ewes. A) A section of a follicle from a ewe treated with FSH-P. B)A section of a follicle from a ewe after FSH-P withdrawal. C) A section, from an FSH-Ptreated ewe, that was stained with ascites fluid in place of the primary antibody as a control. In A, note many BrdU-labeled granulosa and thecal cells, indicating a high rate of cell proliferation. In B, note fewer BrdU-labeled cells compared with A. AF, antrum of follicle; G, granulosa cell layer; T,thecal cell layer. Large arrows indicate some of the BrdU-labeled nuclei. Small arrows indicate the follicular basement membrane (granulosa-thecal border). Bar = 50 Lrm.

JABLONKA-SHARIFF ET AL.

FSH-P WITHDRAWAL AND FOLLICULAR ATRESIA

697 FIG. 2. Immunohistochemical localization of BrdU in sections of atretic follicles of ewes. A and B)Sections of early atretic follicles from ewes in which FSH-P was withdrawn. C) A section of an advanced atretic follicle from a ewe receiving continuous FSH-P treatment. Note BrdU-labeled cells (arrows) in the early atretic follicles (A and B)but none in the advanced atretic follicle (C). Also note the pyknotic nuclei (arrowheads) in all of the atretic follicles (A, B, and C). AF, antrum of follicle; G, granulosa cell layer; T,thecal cell layer; 0, oocyte. Bar = 50 pIm.

JABLONKA-SHARIFF ET AL.

698

of healthy (nonatretic) and early atretic follicles (Fig. 1, A and B; Fig. 2, A and B). In ewes receiving continuous FSH-P treatment, healthy follicles exhibited intensely labeled granulosa and thecal cells (Fig. 1A). Healthy follicles from ewes in which FSH-P was withdrawn showed fewer BrdU-labeled nuclei compared with those from FSH-Ptreated ewes (compare Fig. lB with Fig. 1A). Early atretic follicles with only a few pyknotic nuclei and little cellular debris in the follicular fluid showed drastically reduced BrdU labeling in their granulosa, thecal, and cumulus cells (Fig. 2, A and B). Atretic follicles in advanced stages of atresia did not contain BrdU-labeled nuclei (Fig. 2C). Rate of Cell Proliferation (LI) For nonatretic follicles across all groups and follicular size classes, the LI of granulosa cells was greater (p < 0.01) than that of thecal cells (16.70 + 0.45 vs. 13.54 ± 0.36%, respectively). In addition, for nonatretic follicles across all groups, the LI of granulosa cells decreased (p < 0.01) as follicular size increased (21.73 + 0.62%, 15.65 + 0.65%, and 8.93 0.50% for small, medium, and large follicles, respectively; Fig. 3). Similarly, for nonatretic follicles across all groups, the LI of thecal cells decreased (p < 0.01) as follicular size increased (16.86 + 0.50%, 12.49 + 0.58%, and 9.21 + 0.43% for small, medium, and large follicles, respectively; Fig. 4). For nonatretic follicles across all follicular size classes, the LI of granulosa cells was 12.27 + 0.64% for FSH-P withdrawal, 19.14 0.95% for saline treatment, and 19.37 + 0.67% for continuous FSH-P treatment. Likewise, for nonatretic follicles across all follicular size classes, the overall LI of thecal cells was 10.39 + 0.50% for FSH-P withdrawal compared with 13.24 + 0.74% for saline treatment and 17.03 0.52% for continuous FSH-P treatment. Compared with saline-treated controls, continuous FSH-P treatment increased (p < 0.01) the LI of granulosa cells from medium and large but not small follicles (Fig. 3). In contrast, continuous FSH-P treatment increased (p < 0.01) the LI of thecal cells across all follicular size classes (Fig. 4). In addition, compared with FSH-P-treated ewes, those in which FSH-P was withdrawn showed a reduced (p < 0.01) LI of granulosa and thecal cells across all follicular size classes (Figs. 3 and 4).

[1 Saline ·

As mentioned previously, immunohistochemical staining showed that early atretic follicles typically contained some BrdU-labeled granulosa and/or thecal cells. Most of the early atretic follicles were found in the ewes subjected to FSH withdrawal, although a few early atretic follicles were found in saline-treated ewes. In contrast, FSH-P-treated ewes (groups 4 and 8) typically did not have early atretic follicles; rather, their atretic follicles were usually in advanced stages of atresia and contained no proliferating cells. Therefore, the LI for early atretic follicles was analyzed across groups (i.e., only follicular size class was included in the statistical model). In contrast to healthy follicles, in which LI was greater in granulosa compared with thecal cells, the LI was similar in granulosa and thecal cells of early atretic follicles (3.51 + 0.22% and 4.19 + 0.28%, respectively). In addition, for early atretic follicles the LI was similar for small, medium, and large size classes (4.23 + 0.32%, 3.66 + 0.30%, and 3.83 + 0.27%, respectively). Across all follicular size classes and cell types, the LI of early atretic follicles was low (p < 0.01) compared with that of nonatretic follicles (3.87 0.18% vs. 15.12 + 0.29%, respectively). Apoptosis Histochemical staining of apoptotic nuclei using peroxidase (Fig. 5) or fluoresceinated avidin (data not shown) detection demonstrated that the more antrally located granulosa cells of early and advanced atretic follicles are the predominant site of apoptosis. However, a few apoptotic cells were also observed in the thecal cell layers (Fig. 5, A and B). The nuclei of the degenerating granulosa cells and the apoptotic bodies were heavily stained, confirming the presence of substantial DNA fragmentation (Fig. 5, A and B). In addition, many of the positively stained cells observed within the granulosa layer and the antral cavity appeared to have a low cytoplasm-to-nucleus ratio. This staining was specific, as shown by the lack of staining in control sections (Fig. 5C). In addition, no staining for apoptotic cells was ever observed for morphologically healthy follicles.

L

FSH-P [ FSH-P withdrawal

.,>

Saline

FSH-P

FSH-P H withdrawal

30

J

b

30

- 25

c 25

: 20

.c 20

0)

= 15 0)

.,

_ 15

-0 _

g 10

10

a

0)

5 O 0

7 0 bVI/ALL

iUIV

IV

LAKUt

Follicular size class FIG. 3. BrdU labeling index (%) of granulosa cells from nonatretic follicles across groups of ewes treated with saline (groups 2 and 5), or continuous FSH-P (groups 4 and 8), or in which FSH-P was withdrawn (groups 3, 6, and 7). Numbers within the bars denote the total number of follicles analyzed for LI. Within each follicular size class, means + SEM with different superscripts differ (p < 0.05).

IVI/ALL

IVitIUMIV

LAKrt

Follicular size class FIG. 4. BrdU labeling index (%) of thecal cells from nonatretic follicles across groups of ewes treated with saline (groups 2 and 5), or continuous FSH-P (groups 4 and 8), or in which FSH-P was withdrawn (groups 3, 6, and 7). Numbers within the bars denote total number of follicles analyzed for LI. Within each follicular size class, means + SEM with different superscripts differ (p < 0.05).

FSH-P WITHDRAWAL AND FOLLICULAR ATRESIA

699 FIG. 5. In situ 3'-end labeling of fragmented DNA (apoptosis) in sections of atretic follicles of ewes by using the TUNEL method and indirect peroxidase detection. A and B) Sections from atretic follicles stained by using the TUNEL method. C) From a section adjacent to that in A, but in which biotinylated dUTP was omitted as a control. In A and B, note labeled nuclei of apoptotic cells (arrowheads), as well as apoptotic bodies (arrows). AF, antrum of follicle; G, granulosa cell layer; T, thecal cell layer. Bar = 50 am.

700

JABLONKA-SHARIFF ET AL.

Follicular Status Across all treatment groups and size classes of follicles, the percentage of follicles that were atretic (early + advanced) was increased (p < 0.01) by FSH-P withdrawal (60 7% for groups 3, 6, and 7) compared with salinetreated (36 + 4% for groups 2 and 5) or FSH-P-treated (10 ± 0.2% for groups 4 and 8) ewes. The incidence of atresia did not differ (p > 0.38), however, among follicular size classes, and no interaction between treatment groups and size classes was observed (p > 0.31). Compared with saline and FSH-P treatment, the incidence of atresia for ewes in which FSH-P was withdrawn was 52% and 18% vs. 62%, respectively, for small follicles, 32% and 4% vs. 55% for medium follicles, and 33% and 12% vs. 59% for large follicles. Plasma Estradiol and Progesterone For ewes in which FSH-P was withdrawn (groups 3, 6, and 7), plasma estradiol (6.2 1.4 pg/ml) was less (p < 0.05) than that of FSH-P-treated ewes (14.4 4.6 pg/ml for groups 4 and 8) but was similar (p > 0.05) to that of saline-treated ewes (5.5 + 1.0 pg/ml for groups 2 and 5) and Day 13 untreated ewes (4.2 + 0.6 pg/ml). Plasma concentrations of progesterone were greatest (p < 0.01) for Day 13 untreated ewes (4.4 ± 0.6 ng/ml for group 1) compared with all of the other treatment groups (0.7 0.2 ng/ml for groups 2-7), which were similar (p > 0.05). DISCUSSION Models using a number of experimental procedures have been developed to study atresia in rodents [19]. These models of atresia consistently show results that are comparable to the morphological and hormonal changes taking place during spontaneous atresia [19]. In the present experiment, a model was developed in which morphological and biochemical alterations associated with atresia can be studied in ewes. In the present study, FSH-P withdrawal influenced the number of follicles in all size classes compared with continuous FSH-P treatment. The decrease in the number of medium and large follicles after FSH-P withdrawal may have been due to the loss of gonadotropic support necessary to stimulate or maintain the growth of small and medium follicles, as evidenced by the decrease in granulosa and(or) thecal cell proliferation in these follicles. In support of this contention, FSH withdrawal was associated not only with decreased numbers of medium and large follicles but also with an increased number of small follicles. Alternatively, part of the response may have been due to a loss of medium and large follicles after withdrawal of FSH-P. Results from the present study also demonstrated that the LI of granulosa and thecal cells decreased after FSH-P withdrawal in the small and(or) medium follicular size classes. Although the large follicles may also have shown decreased cell proliferation if greater numbers of ewes had been used for these studies, the similarity in granulosa and thecal LI between saline-treated and FSH-withdrawal ewes indicates that the larger follicles are not as responsive to FSH withdrawal as the smaller follicles. This suggestion agrees with our previous observations that cell proliferation in large antral follicles of sheep is lower and less responsive to FSH compared with that in small or medium follicles [17]. These observations are consistent with previous studies in hamsters, in which the in vivo induction of atresia

by antiserum to pregnant mare's serum decreased follicular LI and significantly reduced DNA synthesis in granulosa cells [20, 35]. Similarly, reduced FSH in rats was associated with a decrease in the number of granulosa cells present in large follicles [36]. Moreover, in rats it has been shown that within 6 h after hypophysectomy, fewer follicles responded to an ovulatory dose of hCG. By 12 h after hypophysectomy, all large and medium follicles exhibited morphological signs of atresia, and most failed to respond to hCG treatment [37]. An interesting finding in the present study concerned the pattern of BrdU labeling of granulosa and thecal cells after FSH-P withdrawal. Immunohistochemical evaluation showed that BrdU staining was present not only in nonatretic follicles but also in granulosa and/or thecal cells of early atretic follicles. However, even though early atretic follicles exhibited some BrdU staining, the LI of their granulosa and thecal cells was very low compared with that of granulosa and thecal cells in nonatretic follicles. In mice, Byskov [38] showed that [3 H]thymidine incorporation declined in early atretic follicles compared with nonatretic follicles and continued to decline as atresia progressed. Similarly, a decreased LI of granulosa cells was observed in early atretic follicles of rats [11], rabbits [39], and pigs [40]. The results of the present study indicate that continuous administration of FSH-P can maintain follicular cell proliferation and prevent atresia. Conversely, compared with continuous FSH treatment, withdrawal of FSH is associated with an increased prevalence of atresia in antral follicles. It has been hypothesized that larger antral follicles require gonadotropic support to sustain their growth and prevent follicular atresia in numerous mammalian species, including sheep [5, 13-16, 41, 42]. In the present experimental model, FSH-P withdrawal was associated with an increase in the percentage of follicles that were atretic compared with FSH-P or saline treatment. Thus, effects of FSH-P treatment and withdrawal were opposite to the effects of continuous FSH-P administration observed in our present as well as our previous [17] studies. An interesting observation from the present study was that FSH-P treatment was associated with increased proliferation of granulosa and thecal cells. Since thecal cells probably do not possess FSH receptors, however, a direct effect of FSH on thecal cell proliferation seems unlikely, and a paracrine interaction between the granulosa and thecal cells may be involved [5, 6, 19]. Overall, these results support the hypothesis that FSH is necessary for follicular growth and the prevention of atresia in several mammalian species. A fundamental enigma of reproductive biology concerns what differentiates a healthy from an atretic ovarian follicle. It is well known that as atresia of large follicles progresses, the architecture of the follicular cell layers changes markedly [18, 19, 31, 43]. One of the most notable features as atresia progresses is the presence of increasing numbers of pyknotic nuclei and atretic bodies in the membrana granulosa. Since pyknosis is a sign of cellular death and not necessarily of follicular dysfunction, follicular atresia may be more accurately reflected by using other histological parameters of atresia in addition to pyknosis. Because the atretic process involves both cell death and removal of expended cells, it has been related to a degenerative process known as apoptosis [44]. Internucleosomal DNA fragmentation is a hallmark of apoptosis and has been suggested to be one of its earliest events [45]. In the present study, the possibility that apoptosis is re-

FSH-P WITHDRAWAL AND FOLLICULAR ATRESIA

sponsible for cell death during ovarian follicular atresia in ewes was examined. In situ detection of DNA fragmentation in histological sections showed that staining of apoptotic nuclei and apoptotic cells was restricted exclusively to granulosa cells of atretic but not healthy follicles. Positive nuclear labeling was also found in a few thecal cells of some atretic follicles. In addition, apoptotic nuclei were observed in the antrum, suggesting that apoptotic bodies are released into the follicular fluid in atretic follicles. Moreover, for early atretic follicles, DNA fragmentation was also found in intact cells. Thus, it can be suggested that apoptosis may be one of the earliest processes of atresia. Similarly, apoptosis was detected in granulosa cells of atretic follicles from rats [26] and cows [24]. In one of these studies [24], it was suggested that granulosa cell apoptosis also may occur in healthy follicles and/or occur early in the atretic process before other morphological or biochemical signs of degeneration or dysfunction are detectable. Recent biochemical studies in follicles undergoing atresia have suggested that apoptotic cell death is the primary degenerative process that occurs during follicular atresia [21, 23, 25, 40]. Thus, along with a dramatic decrease in the rate of granulosa and thecal cell proliferation, the onset of apoptosis in granulosa cells appears to be an early event in the process of atresia in ewes. A change in the pattern of steroidogenesis has been associated with atretic follicles of many species [19]. Steroidogenic activity of follicles during experimentally induced atresia was inferred in the present study by measuring estradiol concentrations in ovine plasma. Decreased concentrations of plasma estradiol in ewes after FSH-P withdrawal was accompanied by decreased numbers of medium and large follicles, an increase in the percentage of follicles that were atretic, and a decreased LI of granulosa and thecal cells. Similarly, a decrease in estradiol concentrations during atresia has been observed in antral follicles of rodents [46-49], ewes [50-52], cows [53, 54], sows [55], mares [56], and humans [57]. Thus, in the present study, FSH-P treatment and subsequent withdrawal resulted in decreased numbers of medium and large follicles, decreased LI of granulosa and thecal cells of nonatretic follicles, and an increased percentage of follicles that were in various stages of atresia. In addition, FSH-P withdrawal reduced the systemic concentration of plasma estradiol. Moreover, apoptotic cell death within the membrana granulosa appeared to be an early event during follicular atresia. This model should be useful for studying the mechanisms of follicular growth and atresia in a widely used large animal species. ACKNOWLEDGMENTS We gratefully acknowledge J.D. Kirsch and K.C. Kraft for expert technical assistance and J. Berg for clerical assistance.

REFERENCES 1. Cahill LP, Mauleon P. A study of the population of primordial and small follicles in the sheep. J Reprod Fertil 1981; 61:201-206. 2. McNatty KP, Gibb M, Dobson C, Ball K, Coster J, Heath DA, Thurley DC. Preovulatory follicular development in sheep treated with PMSG and/or prostaglandin. J Reprod Fertil 1982; 65:111-123. 3. Fortune JE. Ovarian follicular growth and development in mammals. Biol Reprod 1994; 50:225-232. 4. Roy SK, Greenwald GS. In vitro effects of follicle-stimulating hormone, luteinizing hormone, and prolactin on follicular deoxyribonucleic acid synthesis in the hamster. Endocrinology 1988; 122:952-958.

701

5. Mariana JC, Monniaux D, Driancourt MA, Mauleon P. Folliculogenesis. In: PT Cupps (ed.), Reproduction in Domestic Animals, 4th ed. San Diego, CA: Academic Press; 1991: 119-171. 6. Monniaux D, Pisselet C. Control of proliferation and differentiation of ovine granulosa cells by insulin-like growth factor-i and folliclestimulating hormone in vitro. Biol Reprod 1992; 46:109-119. 7. Tsafriri A, Braw RH. Experimental approaches to atresia in mammals. Oxf Rev Reprod 1984; 6:226-265. 8. Hirshfield AN. Development of follicles in the mammalian ovary. Int Rev Cytol 1991; 124:43-101. 9. Hubbard CJ, Oxberry B. Follicular atresia. In: Familiari G, Makabe S, Motta PM (eds.), Ultrastructure of the Ovary. Boston, MA: Kluwer Academic; 1991: 273-285. 10. Pedersen T. Cell population kinetics of the ovary of the immature mouse after FSH stimulation. In: Butt WR, Crooke AC, Ryle M (eds.), Gonadotrophin and Ovarian Development. Edinburgh, UK: Livingstone; 1972: 312-324. 11. Peluso JJ, Steger RW. Role of FSH in regulating granulosa cell division and follicular atresia in rats. J Reprod Fertil 1978; 54:275-278. 12. Rao MC, Midgley AR Jr, Richards JS. Hormonal regulation of ovarian cellular proliferation. Cell 1978; 14:71-78. 13. Monniaux D. Short-term effects of FSH in vitro on granulosa cells on individual sheep follicles. J Reprod Fertil 1987; 79:505-515. 14. McNatty KP, Hudson NL, Heath DA, Shaw L, Blay L, Berry L, Lun S. Effect of chronic FSH administration on ovarian follicular development, ovulation rate and corpora lutea formation in sheep. J Endocrinol 1993; 138:315-325. 15. Braw RH, Tsafriri A. Follicles explanted from pentobarbitone-treated rats provide a model for atresia. J Reprod Fertil 1980; 59:259-265. 16. NcNatty KP, Hudson NL, Gibb M, Ball K, Henderson KM, Heath DA, Lun 5, Kieboom LE. FSH influences follicle viability, oestradiol biosynthesis and ovulation rate in Romney ewes. J Reprod Fertil 1985; 75:121-131. 17. Jablonka-Shariff A, Fricke PM, Grazul-Bilska AT, Reynolds LP, Redmer DA. Size, number cellular proliferation, and atresia of gonadotropin-induced follicles in ewes. Biol Reprod 1994; 51:531-540. 18. Turnbull KE, Braden AH, Mattner PE. The pattern of follicular growth and atresia in the ovine ovary. Aust J Biol Sci 1977; 30:229-241. 19. Greenwald GS, Terranova PE Follicular selection and its control. In: Knobil E, Neill JD (eds.), The Physiology of Reproduction. New York: Raven Press; 1988: 1:387-445. 20. Greenwald GS. Temporal and topographic changes in DNA synthesis after induced follicular atresia. Biol Reprod 1989; 41:175-181. 21. Tilly JL, Kowalski KI, Johnson AL, Hsueh AJW. Involvement of apoptosis in ovarian follicular atresia and postovulatory regression. Endocrinology 1991; 129:2799-2801. 22. Hurwitz A, Adashi EY. Ovarian follicular atresia as a apoptotic process: a paradigm for programmed cell death in endocrine tissues. Mol Cell Endocrinol 1992; 84:C19-C23. 23. Tilly JL, Kowalski KI, Schomberg DW, Hsueh AJW. Apoptosis in atretic ovarian follicles is associated with selective decreases in messenger ribonucleic acid transcripts for gonadotropin receptors and cytochrome P 450 aromatase. Endocrinology 1992; 131:1670-1676. 24. Jolly PD, Tisdall DJ, Heath DA, Lun S, McNatty KR Apoptosis of bovine granulosa cells in relation to steroid synthesis, cyclic adenosine 3',5'-monophosphate response to follicle-stimulating hormone and luteinizing hormone, and follicular atresia. Biol Reprod 1994; 51:934944. 25. Hughes FM, Gorospe WC. Biochemical identification of apoptosis (programmed cell death) in granulosa cells: evidence for a potential mechanisms underlaying follicular atresia. Endocrinology 1991; 129: 2415-2422. 26. Billig H, Furuta I, Hsueh AJW. Gonadotropin-releasing hormone directly induces apoptotic cell death in the rat ovary: biochemical and in situ detection of deoxyribonucleic acid fragmentation in granulosa cells. Endocrinology 1994; 134:245-252. 27. Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 1992; 119:403-501. 28. Zheng J, Fricke PM, Reynolds LP, Redmer DA. Evaluation of growth, cell proliferation and cell death in bovine corpora lutea throughout the estrous cycle. Biol Reprod 1994; 51:623-632. 29. Zheng J, Johnson ML, Redmer DA, Reynolds LP. Estrogen and progesterone receptors, cell proliferation, and c-fos expression in the ovine uterus during early pregnancy. Endocrinology 1996; 137:340348. 30. Jablonka-Shariff A, Grazul-Bilska AT, Redmer DA, Reynolds LP.

702

31.

32.

33. 34. 35.

36.

37.

38. 39.

40.

41.

42. 43. 44.

JABLONKA-SHARIFF ET AL.

Growth and cellular proliferation of ovine corpora lutea throughout the estrous cycle. Endocrinology 1993; 133:1871-1879. Brand A, deJong WHR. Qualitative and quantitative micromorphological investigations of the tertiary follicle population during the oestrous cycle in sheep. J Reprod Fertil 1973; 33:431-439. Redmer DA, Grazul-Bilska AT, Reynolds LP. Contact-dependent intercellular communication of bovine luteal cells in culture. Endocrinology 1991; 129:2757-2766. SAS. SAS User's Guide. Statistics, ed 5. Cary, NC: Statistical Analysis System Institute, Inc.; 1985. Kirk RE. Experimental Design: Procedures for the Behavioral Sciences, ed 2. Belmont, CA: Brooks/Cole; 1982: 106-109. Hubbard CJ, Greenwald GS. Changes in DNA, cyclic nucleotides and steroids during induced follicular atresia in the hamster. J Reprod Fertil 1981; 63:455-461. Chakravorty A, Mahesh VB, Mills TM. Control of peptides regulating mitosis of granulosa cells in immature rat ovary by oestrogen and gonadotrophin. J Reprod Fertil 1993; 97:91-100. Kaneko H, Taya K, Sasamato S. Changes in the secretion of inhibin and steroid hormones during induced follicular atresia after hypophysectomy in the rat. Life Sci 1987; 41:1823-1830. Byskov AG. Kinetic studies of follicular atresia in the mouse ovary. J Reprod Fertil 1974; 37:277-285. Kranzfelder D, Korr H, Mestwerdt W, Maurer-Schultze B. Follicle growth in the ovary of the rabbit after ovulation-inducing application of human chorionic gonadotropin. Cell Tissue Res 1984; 238:611620. Guthrie HD, Welch GR, Cooper BS, Zakaria AD, Johnson AL. Flow cytometric determination of degraded deoxyribonucleic acid in granulosa cells to identify atretic follicles during preovulatory maturation in the pig. Biol Reprod 1994; 50:1303-1311. Hirshfield AN. Effect of a low dose of mare's serum gonadotropin on follicular recruitment and atresia in cyclic rats. Biol Reprod 1986; 35: 113-118. McNeilly AS. The ovarian follicle and fertility. J Steroid Biochem Mol Biol 1991; 40:29-33. Hay ME Cran DG, Moor RM. Structural changes occurring during atresia in sheep ovarian follicles. Cell Tissue Res 1976; 169:515-529. Kerr JFR, Wyllie AH, Currie AR. Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 1972; 26:239-257.

45. Ardens MJ, Morris RG, Wyllie AH. Apoptosis: the role of the endonuclease. Am J Pathol 1990; 136:593-608. 46. Braw RH, Tsafriri A. Effect of PMSG on follicular atresia in the immature rat ovary. J Reprod Fertil 1980; 59:267-272. 47. Uilenbroek JTJ, Woutersen PJA, van der Schoot P. Atresia of preovulatory follicles: gonadotropin binding and steroidogenic activity. Biol Reprod 1980; 23:219-229. 48. Terranova PE Steroidogenesis in experimentally induced atretic follicles of the hamster: a shift from estradiol to progesterone synthesis. Endocrinology 1981; 108:1855-1890. 49. Tsafriri A, Eckstein B. Changes in follicular steroidogenic enzymes following the preovulatory surge of gonadotropins and experimentally induced atresia. Biol Reprod 1986; 34:783-787. 50. Moor RM, Hay ME Dott HM, Cran DG. Macroscopic identification and steroidogenic function of atretic follicles in sheep. J Endocrinol 1978; 77:309-318. 51. Carson RS, Findlay JK, Clarke IJ, Burger HG. Estradiol, testosterone and androstenedione in ovine follicular fluid during growth and atresia of ovarian follicles. Biol Reprod 1981; 24:105-113. 52. Taraska T, Reynolds LP, Redmer DA. In vitro secretion of angiogenic activity of ovine follicles. In: Hirshfield AN (ed.), Growth Factors and the Ovary. New York: Plenum Press; 1989: 267-272. 53. Grimes RW, Matton P, Ireland JJ. A comparison of histological and non-histological indices of atresia and follicular function. Biol Reprod 1987; 37:82-88. 54. Spicer LJ, Matton PJ, Echtemkamp SE, Convey EM, Tucker HA. Relationship between histological signs of atresia, steroids in follicular fluid, and gonadotropin binding in individual bovine antral follicles during postpartum anovulation. Biol Reprod 1987; 36:890-898. 55. Westhof G, Westhof KF, Braendle WL, diZerega GS. Differential steroids secretion and gonadotropin response by individual tertiary porcine follicles in vitro. Possible physiologic role of atretic follicles. Biol Reprod 1991; 44:461-468. 56. Condon WA, Ganjam VK, Kenney RM, Channing CP. Follicular atresia in the mare. In: Midgley AR, Sadler WA (eds.), Ovarian Follicular Development and Function. New York: Raven Press; 1979: 75-78. 57. Bomsel-Helmreich O, Gougeon A, Thebault A, Saltarelli D, Milgrom E, Frydman R, Papiernik E. Healthy and atretic human follicles in the preovulatory phase: differences in evolution of follicular morphology and steroid content of follicular fluid. J Clin Endocrinol & Metab 1979; 48:686-694.