The ovarian reserve is depleted during puberty in a hormonally driven process dependent on the pro-apoptotic protein BMF

Abstract

In females, germ cells are maintained in ovarian structures called primordial follicles. The number of primordial follicles in the ovarian reserve is a critical determinant of the length of the fertile lifespan. Despite this significance, knowledge of the precise physiological mechanisms that regulate primordial follicle number is lacking. In this study we show that a wave of primordial follicle depletion occurs during the transition to adulthood in mice. We demonstrate that this sudden and dramatic loss of primordial follicles is hormonally triggered and identify the pro-apoptotic BH3-only protein, BCL-2 modifying factor (BMF), as essential for this process, implicating the intrinsic apoptotic pathway as a key mechanism. The elimination of primordial follicles during puberty is not only a striking developmental event, it is also physiologically important because it ultimately reduces the availability of primordial follicles and determines the duration of fertility. Collectively, these findings show that puberty is a critical developmental window for the regulation of the size of ovarian reserve, impacting on female fertility and reproductive longevity.

Figure 1

Follicle numbers in ovaries from WT and Bmf^−/− female mice. Primordial (a), primary (b), secondary and antral follicles (c), and total follicles (d) were counted in the ovaries of WT and Bmf^−/− mice at PN20, 30, 40 and 50 (n=6/age/genotype). Data are expressed as mean±S.E.M. *P<0.05 for comparison (two-tailed unpaired t-test) of follicle numbers between PN40 and 50 WT female mice. ^#P<0.05 for comparisons (two-tailed unpaired student’s t-test) between Bmf^−/− and WT at PN50. For clarity, only statistical significance of select comparisons are shown. See Supplementary Tables 1 and 2 for statistical significance of all pairwise comparisons within a genotype. (e) Representative images of primordial follicles in PAS-stained ovarian sections from WT and Bmf^−/− mice at PN50. Sections are 20 μM thick enabling primordial follicles to be clearly identified by focussing up and down in the z-axis. Black inset boxes in top images represent area shown below at higher magnification. Black arrow heads indicate primordial follicles. White arrow heads indicate empty follicles. Scale bars: Top images=100 μm, Bottom images=20 μm. (f) Representative images of secondary follicles in PAS-stained ovarian sections from WT and Bmf^−/− mice at PN50. Black inset boxes in top images represent area shown below at higher magnification. Black arrow heads indicate secondary follicles. White arrow heads indicate atretic follicles. Scale bars: Top images=100 μm, Bottom images=20 μm

Figure 2

Follicular atresia in ovaries from WT and Bmf^−/− female mice. The numbers of atretic follicles were determined in ovaries from WT and Bmf^−/− mice at PN20, 30 40 and 50 (n=6/age/genotype). (a) Representative images of atretic preantral (secondary) and antral follicles in PAS-stained ovarian sections from WT and Bmf^−/− mice at PN50. Atretic preantral and antral follicles were characterised by the presence of degrading oocytes and/pyknotic granulosa cells. Scale bars=50 μm. (b) Representative images of TUNEL-positive atretic follicles (brown staining) in secondary follicles in ovarian sections from WT and Bmf^−/− mice at PN50. Scale bars=50 μm. (c) Data are expressed as mean±S.E.M. ^#P<0.05 for comparison of Bmf^−/− versus WT females at each age P<0.05 (two-tailed unpaired student’s t-test). For clarity, only pairwise comparisons between WT and Bmf^−/− females at each age are shown. See Supplementary Tables 1 and 2 for statistical significance of all pairwise comparisons within a genotype

Figure 3

Onset of puberty in WT and Bmf^−/− female mice. (a) The percentages of WT and Bmf^−/− mice with at least one corpus luteum present in their ovaries were determined at PN20, 30, 40 and 50 (n=6/age/genotype). The numbers of mice analysed are shown above each bar. (b) The average numbers of corpora lutea present per ovary in WT and Bmf^−/− mice (only mice with at least one corpora lutea were included in the analysis) were determined at PN20, 30, 40 and 50. Data are expressed as mean±S.E.M. (where n=3 or more). ^#P<0.05 for comparison of Bmf^−/− versus WT at PN50 (two-tailed unpaired student’s t-test). (c) Representative images of PAS-stained ovarian sections from WT and Bmf^−/− mice at PN20, 30, 40 and 50. The notation ‘CL’ indicate the location of each corpus luteum. Scale bars=200 μm

Figure 4

Follicle loss as female mice approach adulthood is hormonally driven. Mice were treated with saline (controls) or Cetrorelix daily beginning at PN25 (Day 1 of the study) and finishing at PN50 (Day 25 of the study) (n=6/treatment group) (a–f). From the time of vaginal opening, mice were smeared daily to monitor the onset of estrous cyclicity. (a) Examples of estrus cycling pattern in saline- or Cetrorelix-treated mice. 0= not cycling, vagina not yet opened, 1= metestrus, 2=estrus, 3= proestrus and 4= diestrus. All saline-treated mice displayed onset of estrous cyclicity, whereas mice treated with Cetrorelix either did not cycle or had very disrupted cycles. (b) Age at vaginal opening. (c) The average number of times mice in saline- and Cetrorelix-treated groups entered diestrus during the 25-day study. Data are expressed as mean±S.E.M. *P<0.05 (two-tailed unpaired student’s t-test). (d) Ovarian volume. *P<0.05 (two-tailed unpaired student’s t-test). (e) Representative images of PAS-stained ovarian sections from saline- and Cetrorelix-treated mice. The notation ‘CL’ indicates the location of each corpus luteum. Black arrow heads indicate primordial follicles. Scale bars: Top images=200 μm, Bottom images=40 μm. (f) Primordial follicles were counted in ovaries of saline- and Cetrorelix-treated mice at PN50 (grey bars to the right of the dotted vertical line). For comparative purposes, these data are presented alongside the number of primordial follicles in normal WT ovaries at PN40 and PN50 (data from Figure 1a) (black bars to the left of the dotted vertical line). Data are expressed as mean±S.E.M. *P<0.05 (two-tailed unpaired student’s t-test). (g) Twenty-day-old sexually immature female mice were treated with saline (controls) or equine chorionic gonadotropin (eCG, 5IU), followed 44–48 h later by human chorionic gonadotrophin (hCG, 5IU) (n=6/treatment group). Ten days later (PN30), the ovaries harvested and the primordial follicles counted. Data are expressed as mean±S.E.M. *P<0.05 (two-tailed unpaired student’s t-test)