Conditional loss of Brca1 in oocytes causes reduced litter size, ovarian reserve depletion and impaired oocyte in vitro maturation with advanced reproductive age in mice

Abstract

Background: An estimated 1 in 350 women carry germline BRCA1/2 mutations, which confer an increased risk of developing breast and ovarian cancer, and may also contribute to subfertility. All mature, sex steroid-producing ovarian follicles are drawn from the pool of non-renewable primordial follicles, termed the 'ovarian reserve'. The clinical implications of early ovarian reserve exhaustion extend beyond infertility, to include the long-term adverse health consequences of loss of endocrine function and premature menopause. We aimed to determine whether conditional loss of Brca1 in oocytes impacts ovarian follicle numbers, oocyte quality and fertility in mice with advancing maternal age. We also aimed to determine the utility of AMH as a marker of ovarian function, by assessing circulating AMH levels in mice and women with BRCA1/2 mutations, and correlating this with ovarian follicle counts.

Methods: In this study, we addressed a longstanding question in the field regarding the functional consequences of BRCA1 inactivation in oocytes. To recapitulate loss of BRCA1 protein function in oocytes, we generated mice with conditional gene deletion of Brca1 in oocytes using Gdf9-Cre recombinase (WT: Brca1^fl/flGdf9^+/+; cKO: Brca1^fl/flGdf9^cre/+).

Findings: While the length of the fertile lifespan was not altered between groups after a comprehensive breeding trial, conditional loss of Brca1 in oocytes led to reduced litter size in female mice. Brca1 cKO animals had a reduced ovarian reserve and oocyte maturation was impaired with advanced maternal age at postnatal day (PN)300, compared to WT animals. Serum anti-Müllerian hormone (AMH) concentrations (the gold-standard indirect marker of the ovarian reserve used in clinical practice) were not predictive of reduced primordial follicle number in Brca1 cKO mice versus WT. Furthermore, we found no correlation between follicle number or density and serum AMH concentrations in matched samples from a small cohort of premenopausal women with BRCA1/2 mutations.

Interpretation: Together, our data demonstrate that BRCA1 is a key regulator of oocyte number and quality in females and suggest that caution should be used in relying on AMH as a reliable marker of the ovarian reserve in this context.

Funding: This work was made possible through Victorian State Government Operational Infrastructure Support and Australian Government NHMRC IRIISS. This work was supported by funding from the Australian Research Council (ALW - DE21010037 and KJH - FT190100265), as well as the National Breast Cancer Foundation (IIRS-22-092) awarded to ALW and KJH. LRA, YML, LT, EOKS and MG were supported by Australian Government Research Training Program Scholarships. LRA, YML and LT were also supported by a Monash Graduate Excellence Scholarship. YC, SG and XC were supported by Monash Biomedicine Discovery Institute PhD Scholarships. LRA was also supported by a Monash University ECPF24-6809920940 Fellowship. JMS was supported by NHMRC funding (2011299). MH was supported by an NHMRC Investigator Grant (1193838).

Keywords: AMH; Ageing; BRCA1; DNA repair; Oocyte; Primordial follicle.

Fig. 1

(a) Schematic representation of the fertile lifespan breeding trial of wild-type (WT: Brca1^fl/flGdf9^+/+) and oocyte conditional Brca1 knockout (cKO: Brca1^fl/flGdf9^cre/+) female mice (n = 9/genotype) mated with proven C57BL6/J WT male studs. Females were kept for breeding for their entire fertile lifespan, until no litters had been produced for ≥2 months. (b) Time to first vaginal plug, (c) time to first litter and (d) age at last litter were recorded. (e) Maternal serum AMH concentrations (ng/mL) were measured at necropsy. (f) Total number of litters per female, (g) average number of pups per female across all litters (h) size of each litter and (i) first litter size were recorded. Data are mean ± SEM; unpaired t-test; ∗p < 0.05, ∗∗p < 0.01. (j) Total number of male and female offspring and (k) offspring weights per sex at postnatal day (PN)5 and PN20 were recorded. (l) Offspring weight at PN20 was adjusted for litter size for each genotype by Spearman's rank correlation test. Data are mean ± SEM; Kruskal–Wallis test; ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

Fig. 2

(a) Numbers of ovulated oocytes harvested following exogenous hormonal stimulation from WT and Brca1 cKO mice at PN80 (WT n = 11; Brca1 cKO n = 15) and (b) PN200 (WT n = 8; Brca1 cKO n = 6). Data are presented as mean ± SEM; t-test. (c) Representative images of intact mature MII, immature MI, immature germinal vesicle (GV) or fragmented/dead oocytes obtained from control WT and Brca1 cKO animals at PN80; bar = 50 μm. (d) The top panel shows representative images of GV oocytes harvested from PN300 WT or Brca1 cKO animals before in vitro maturation (IVM) culture was performed for 16 h; bar = 50 μm. The lower panel shows representative images of either mature MII oocytes, or a combination of MII or immature or non-viable oocytes (arrows) following IVM from WT or Brca1 cKO PN300 mice; bar = 50 μm. (e) The maturation rate (%) from GV to MII oocytes for each genotype is depicted in black bars, and proportion (%) of immature or non-viable oocytes are depicted in grey bars. Proportions are calculated from oocytes derived from n = 8 WT and n = 9 Brca1 cKO mice across three independent experiments; Fisher's exact test, ∗∗∗∗p < 0.0001. (f) PN300 MII oocytes harvested following IVM were immunostained with αβ-tubulin (green) to label the meiotic spindle, and DAPI (blue) to label the DNA on the metaphase plate.

Fig. 3

(a) Schematic representation of animals used for ovarian follicle counts from WT and Brca1 cKO female mice aged PN5, PN20, PN50, PN200 and PN300 (n = 6–10/age/genotype). (b) Ovarian follicle endowment was assessed by quantifying total follicles in ovaries from mice aged PN5 (WT n = 10; Brca1 cKO n = 8). (c) Age at vaginal opening was assessed to determine puberty onset (WT n = 16; Brca1 cKO n = 12). (d) Representative photomicrographs of periodic acid Schiff (PAS) stained ovarian follicles: i. primordial; bar = 10 μm, ii. primary; bar = 10 μm, iii. secondary; bar = 100 μm, iv. antral; bar = 200 μm, v. atretic secondary; bar = 100 μm and vi. atretic antral; bar = 200 μm. (e) Primordial and (f) primary follicles were quantified (WT black squares: PN20 n = 7, PN50 n = 9, PN200 n = 5, PN300 n = 8; Brca1 cKO open circles: PN20 n = 7, PN50 n = 7, PN200 n = 7, PN300 n = 12). (g) Ovarian tissue sections were immunostained with c-Kit to label oocytes and γH2AX to detect the accumulation of endogenous DNA damage in WT and Brca1 cKO primordial follicle oocytes at age PN50. Representative images of γH2AX staining in primordial follicles are shown; specifically, (i-ii) negative and (iii) positive follicles from WT animals, as well as (iv) negative and (v-vi) positive follicles from Brca1 cKO animals; arrowheads = γH2AX foci; red = γH2AX, green = c-Kit; blue = Hoechst; bars = 10 μm. (h) Representative images of primordial follicles from (i) positive and (ii) negative controls are shown. (i) The proportion of γH2AX-positive primordial follicles were quantified in ovaries from WT and Brca1 cKO animals. A total n = 241 primordial follicles were analysed across n = 4 sections/ovary from n = 5 mice/genotype. Data are presented as mean ± SEM; unpaired t-test or Mann–Whitney test; ∗p < 0.05, ∗∗∗p < 0.001.

Fig. 4

(a) Healthy secondary and antral and (b) atretic secondary and antral follicles were quantified in ovaries from animals at ages PN20, PN50, PN200 and PN300 based on follicle classifications represented in Fig. 3d (WT black squares: PN20 n = 7, PN50 n = 9, PN200 n = 5, PN300 n = 8; Brca1 cKO open circles: PN20 n = 7, PN50 n = 7, PN200 n = 7, PN300 n = 12). (c) Serum AMH concentrations (ng/mL) from female WT and Brca1 cKO animals at ages PN20, PN50, PN200 and PN300 were measured (WT: PN20 n = 11, PN50 n = 11, PN200 n = 5, PN300 n = 3; Brca1 cKO: PN20 n = 6, PN50 n = 7, PN200 n = 6, PN300 n = 8). (d) Matched sample serum AMH concentrations (ng/mL) (dependant variable) were measured and correlated by Spearman's rank correlation test with total primordial follicle number (independent variable) in WT (n = 42) and Brca1 cKO mice (n = 50) across all ages. Data are presented as mean ± SEM; unpaired t-test; ∗p < 0.05, ∗∗∗p < 0.001.

Fig. 5

(a) Human cortical ovarian tissues from BRCA1 (n = 8) and BRCA2 (n = 10) mutation carriers (mc) were immunohistochemically stained for DDX4 as a marker of oocytes. Photomicrographs are representative of the classifications used to identify i. primordial follicles; bar = 100 μm, ii. primary follicles; bar = 100 μm, iii. secondary follicles; bar = 100 μm iv-v. abnormal follicles; bar = 100 μm vi. abnormal follicles; bar = 200 μm. (b) Total primordial follicle number, (c) primordial follicle density per tissue area (μm²) and (d) total healthy follicle density per tissue area (μm²) were enumerated. (e) Total abnormal follicle number and (f) abnormal follicle rate (%) were quantified. (g) Serum AMH levels were quantified by ELISA. Data are mean ± SEM; unpaired t-test.