A standard histomorphometric approach has been used for nearly 40 years that identifies atretic (e.g., dying) follicles by counting the number of pyknotic granulosa cells (GC) in the largest follicle cross-section. This method holds that if one pyknotic granulosa nucleus is seen in the largest cross section of a primary follicle, or three pyknotic cells are found in a larger follicle, it should be categorized as atretic. Many studies have used these criteria to estimate the fraction of atretic follicles that result from genetic manipulation or environmental insult. During an analysis of follicle development in a mouse model of Fragile X premutation, we asked whether these ‘historical’ criteria could correctly identify follicles that were not growing (and could thus confirmed to be dying). Reasoning that the fraction of mitotic GC reveals whether the GC population was increasing at the time of sample fixation, we compared the number of pyknotic nuclei to the number of mitotic figures in follicles within a set of age-matched ovaries. We found that, by itself, pyknotic nuclei quantification resulted in high numbers of false positives (improperly categorized as atretic) and false negatives (improperly categorized intact). For preantral follicles, scoring mitotic and pyknotic GC nuclei allowed rapid, accurate identification of non-growing follicles with 98% accuracy. This method most often required the evaluation of one follicle section, and at most two serial follicle sections to correctly categorize follicle status. For antral follicles, we show that a rapid evaluation of follicle shape reveals which are intact and likely to survive to ovulation. Combined, these improved, non-arbitrary methods will greatly improve our ability to estimate the fractions of growing/intact and non-growing/atretic follicles in mouse ovaries.

In reproductive age women, the pool of primordial follicles is continuously depleted through the process of cyclic recruitment. Anti-Mullerian hormone (AMH) both inhibits the initial recruitment of primordial follicles into the growing pool and modulates the sensitivity of growing follicles to follicle stimulating hormone. Thus, AMH may be an important modulator of female infertility and ovarian reserve; however, the mechanisms regulating AMH remain unclear. To evaluate AMH levels in the absence of H19 lncRNA, H19 knockout (H19KO) mice were evaluated for analysis of ovarian AMH gene expression, protein production, and reproductive function, including assessment of follicle numbers and litter size analysis. To further investigate regulation of AMH by the H19/let-7 axis, let-7 binding sites on AMH were predicted, and in vitro studies of the effect of H19 knockdown/overexpression with let-7 rescue were performed. Lastly, response to superovulation was assessed via oocyte counts and estradiol measurements. The H19KO mouse demonstrates subfertility and accelerated follicular recruitment with increased spontaneous development of secondary, preantral, and antral follicles. Ovaries of H19KO mice have decreased AMH mRNA and protein, and AMH mRNA has a functional let-7 binding site, suggesting a plausible ncRNA-mediated mechanism for AMH regulation by H19/let-7. Lastly, in the absence of H19, superovulation results in higher estradiol and more oocytes, suggesting that H19 functions to limit the number of follicles that mature, produce estradiol, and ovulate. Thus, AMH’s inhibitory actions are regulated at least in part by H19, likely via let-7, marking this ncRNA pair as important regulators of the establishment and maintenance of the follicular pool.

High-grade serous ovarian cancer (HGSOC) is the deadliest ovarian cancer histotype due in-part to the lack of therapeutic options for chemotherapy-resistant disease. PARP inhibitors (PARPi) represent a targeted treatment. However, PARPi resistance is becoming a significant clinical challenge. There is an urgent need to overcome resistance mechanisms to extend disease-free intervals. We established isogeneic PARPi-sensitive and -resistant HGSOC cell lines. In three PARPi-resistant models, there is a significant increase in AP-1 transcriptional activity and DNA repair capacity. Using RNA-sequencing and an shRNA screen, we identified activating transcription factor 6 (ATF6) as a mediator of AP-1 activity, DNA damage response, and PARPi resistance. In publicly available datasets, ATF6 expression is elevated in HGSOC and portends a poorer recurrence-free survival. In a cohort of primary HGSOC tumors, higher ATF6 expression significantly correlated to PARPi resistance. In PARPi-resistant cell lines and a PDX model, inhibition of a known ATF6 regulator, p38, attenuated AP-1 activity and RAD51 foci formation, enhanced DNA damage, significantly inhibited tumor burden, and reduced accumulation of nuclear ATF6. Implications: This study highlights that a novel p38-ATF6-mediated AP-1 signaling axis contributes to PARPi resistance and provides a clinical rationale for combining PARPi and AP-1 signaling inhibitors.