Can granulosa cells produce progesterone (P) in response to FSH stimulation? FSH actively promotes P synthesis and output from granulosa cells without luteinization by up-regulating the expression and increasing enzymatic activity of 3β-hydroxysteriod dehydrogenoase (3β-HSD), which converts pregnenolone to P. Serum P level may rise prematurely prior to ovulation trigger in stimulated IVF cycles and adversely affect implantation and clinical pregnancy rates by impairing endometrial receptivity. A translational research study. Human ovarian cortical samples (n = 15) and non-luteinizing FSH-responsive human mitotic granulosa cell line (HGrC1) were stimulated with rec-FSH at 12.5, 25 and 50 mIU/ml concentrations for 24 and 48 h. FSH receptor expression was knocked-down and up-regulated in the granulosa cells using short hairpin RNA (shRNA) technology and activin-A administration, respectively. The expressions of the steroidogenic enzymes were analyzed at mRNA level by real-time quantitative RT-PCR, and protein level by western blot and immunoprecipitation assay. The enzymatic activity of 3β-HSD was measured using a spectrophotometric method. In vitro estradiol (E₂) and P productions of the cells before and after FSH stimulation were measured by electro-chemiluminescence immunoassay method. Stimulation of the HGrC1 cells with FSH resulted in a dose-dependent increase in the mRNA and protein level of 3β-HSD. Overall, when all time points and FSH doses were analyzed collectively, FSH significantly up-regulated the mRNA expression of its own receptor (3.73 ± 0.06-fold, P < 0.001), steroidogenic acute regulatory protein (stAR, 1.7 ± 0.03-fold, P < 0.01), side-chain cleavage enzyme (SCC, 1.75 ± 0.03-fold, P < 0.01), aromatase (4.49 ± 0.08-fold, P < 0.001), 3β-HSD (1.68 ± 0.02-fold, P < 0.01) and 17β-hydroxy steroid dehydrogenase (17β-HSD, 2.16 ± 0.02-fold, P < 0.01) in the granulosa cells. Expression of 17α-hydroxylase (17α-OH, 1.03 ± 0.01-fold P > 0.05) did not significantly change. Similar changes were observed in the protein expression analysis of these enzymes on western blotting after FSH stimulation. FSH significantly increased 3β-HSD, 17β-HSD and aromatase in a dose-dependent manner but did not affect 17α-OH. Protein expression of P was increased along with 3β-HSD after FSH stimulation, which was further evidenced by immunoprecipitation assay. Enzymatic activity of 3β-HSD was significantly enhanced by FSH administration in the HGrC1 cells in a dose-dependent manner. In line with these findings P output (1.05 ± 0.3 vs. 0.2 ± 0.1 ng/ml, respectively, P < 0.001) from the samples stimulated with FSH were significantly increased along with E₂ (1918 ± 203 vs. 932 ± 102 pg/ml, respectively, P < 0.001) compared to unstimulated controls. FSH-induced increase in 3β-HSD expression was amplified and reversed in the HGrC1 cells when FSH receptor expression was up-regulated by activin-A and down-regulated with shRNA, respectively. As only the effect of FSH was studied we cannot extrapolate our findings to the potential effects of HMG and recombinant LH. This data provides a molecular explanation for the largely unexplained phenomenon of P rise during the follicular phase of gonadotropin stimulated IVF cycles. Our findings may progress the research to uncover potential mechanisms for preventing premature P rise that appears to be associated with inferior outcomes in women undergoing IVF. Funded by the School of Medicine and the Graduate School of Health Sciences of Koc University. All authors declare no conflict of interest. None.

Do different chemotherapy drugs exert the same magnitude of cytotoxicity on dormant primordial follicles and the growing follicle fraction in the ovary in vivo and on mitotic non-luteinized and non-mitotic luteinized granulosa cells in vitro? Cyclophosphamide (alkylating agent) and cisplatin (alkylating like) impacted both primordial and pre-antral/antral follicles and both mitotic and non-mitotic granulosa cells, whereas the anti-metabolite cancer drug gemcitabine was detrimental only to pre-antral/antral follicles and mitotic non-luteinized granulosa cells. It is already known that anti-metabolite cancer drugs are less detrimental to the ovary than alkylating and alkylating like agents, such as cyclophosphamide and cisplatin. This assumption is largely based on the results of clinical reports showing lower rates of amenorrhea in women receiving anti-metabolite agent-based regimens compared with those treated with the protocols containing an alkylating drug or a platinum compound. But a quantitative comparison of gonadotoxicity with a histomorphometric proof of evidence has not been available for many chemotherapy drugs. Therefore, we combined in this study in vivo and in vitro models of human and rat origin that allows a comparative analysis of the impact of different chemotherapy agents on the ovary and granulosa cells using real-time quantitative cell indices, histomorphometry, steroidogenesis assays, and DNA damage and cell death/viability markers. We also aimed to investigate if there is a difference between mitotic and non-mitotic granulosa cells in terms of their sensitivity to the cytotoxic actions of chemotherapy drugs with different mechanisms of action. This issue has not been addressed previously. This translational research study involved in vivo analyses of ovaries in rats and in vitro analyses of granulosa cells of human and rat origin. For the in vivo assays, 54 4- to 6-week old Sprague-Dawley young female rats were randomly allocated into four groups of 13 to receive a single IP injection of: saline (control), gemcitabine (200 mg/kg), cisplatin (50 mg/kg) or cyclophosphamide (200 mg/kg). The animals were euthanized 72 h later. Follicle counts and serum AMH levels were compared between the groups. In vitro cytotoxicity studies were performed using mitotic non-luteinized rat (SIGC) and human (COV434, HGrC1) granulosa cells, and non-mitotic luteinized human (HLGC) granulosa cells. The cells were plated at a density of 5000 cells/well using DMEM-F12 culture media supplemented with 10% FBS. Chemotherapy agents were used at their therapeutic blood concentrations. The growth of mitotic granulosa cells was monitored real-time using xCelligence system. Live/dead cell and apoptosis assays were also carried out using intravital Yo-Pro-1 staining and cleaved caspase-3 expression, respectively. Estradiol (E2), progesterone (P) and anti-Mullerian hormone (AMH) levels were assayed with ELISA. Cyclophosphamide and cisplatin caused massive atresia of both primordials and growing follicles in the rat ovary whereas gemcitabine impacted pre-antral/antral follicles only. Cyclophosphamide and cisplatin induced apoptosis of both mitotic non-luteinized and non-mitotic luteinized granulosa cells in vitro. By contrast, cytotoxicity of gemcitabine was confined to mitotic non-luteinized granulosa cells. This study tested only three chemotherapeutic agents. The experimental methodology described here could be applied to other drugs for detailed analysis of their ovarian cytotoxicity. These findings indicate that in vivo and in vitro cytotoxic actions of chemotherapy drugs on the ovarian follicles and granulosa cells vary depending upon the their mechanism of action and the nature of the granulosa cells. The School of Medicine and the Graduate School of Health Sciences of Koc University, and American Hospital Women's Health Center, Comprehensive Cancer Care and Fertility Preservation Programs, Istanbul Turkey. All authors declare no conflict of interest.

Autophagy is an evolutionarily conserved process that aims to maintain the energy homeostasis of the cell by recycling long-lived proteins and organelles. Previous studies documented the role of autophagy in sex steroid hormone biosynthesis in different animal models and human testis. Here we demonstrate in this study that sex steroid hormones estrogen and progesterone are produced through the same autophagy-mediated mechanism in the human ovary in addition to the human testis. In brief, pharmacological inhibition and genetic interruption of autophagy through silencing of autophagy genes (Beclin1 and ATG5) via siRNA and shRNA technologies significantly reduced basal and gonadotropin-stimulated estradiol (E₂), progesterone (P₄) and testosterone (T) production in the ex vivo explant tissue culture of ovary and testis and primary and immortalized granulosa cells. Consistent with the findings of the previous works, we observed that lipophagy, a special form of autophagy, mediates the association of the lipid droplets (LD)s with lysosome to deliver the lipid cargo within the LDs to lysosomes for degradation in order to release free cholesterol required for steroid synthesis. Gonadotropin hormones are likely to augment the production of sex steroid hormones by upregulating the expression of autophagy genes, accelerating autophagic flux and promoting the association of LDs with autophagosome and lysosome. Moreover, we detected some aberrations at different steps of lipophagy-mediated P₄ production in the luteinized GCs of women with defective ovarian luteal function. The progression of autophagy and the fusion of the LDs with lysosome are markedly defective, along with reduced P₄ production in these patients. Our data, together with the findings of the previous works, may have significant clinical implications by opening a new avenue in understanding and treatment of a wide range of diseases, from reproductive disorders to sex steroid-producing neoplasms, sex steroid-dependent malignancies (breast, endometrium, prostate) and benign disorders (endometriosis).

DIRAS3 is an imprinted tumor suppressor gene that encodes a 26 kDa GTPase with 60% amino acid homology to RAS, but with a distinctive 34 amino acid N-terminal extension required to block RAS function. DIRAS3 is maternally imprinted and expressed only from the paternal allele in normal cells. Loss of expression can occur in a single “hit” through multiple mechanisms. Downregulation of DIRAS3 occurs in cancers of the ovary, breast, lung, prostate, colon, brain, and thyroid. Reexpression of DIRAS3 inhibits signaling through PI3 kinase/AKT, JAK/STAT, and RAS/MAPK, blocking malignant transformation, inhibiting cancer cell growth and motility, and preventing angiogenesis. DIRAS3 is a unique endogenous RAS inhibitor that binds directly to RAS, disrupting RAS dimers and clusters, and preventing RAS-induced transformation. DIRAS3 is essential for autophagy and triggers this process through multiple mechanisms. Reexpression of DIRAS3 induces dormancy in a nu/nu mouse xenograft model of ovarian cancer, inhibiting cancer cell growth and angiogenesis. DIRAS3-mediated induction of autophagy facilitates the survival of dormant cancer cells in a nutrient-poor environment. DIRAS3 expression in dormant, drug-resistant autophagic cancer cells can serve as a biomarker and as a target for novel therapy to eliminate the residual disease that remains after conventional therapy.

There is a controversy in literature as to whether c-Abl is crucial for the induction of TAp63-mediated apoptosis and whether that inhibition of c-Abl with imatinib, which was designed to inhibit the oncogenic kinase BCR-ABL and c-kit, protects oocytes from chemotherapy-induced apoptosis in mice. No human data are available on this issue. We therefore aimed to explore whether genomic damage induced by chemotherapy drug cisplatin activates c-Abl along with TAp63 and the inhibition of c-Abl with imatinib prevents cisplatin-induced oocyte death and follicle loss in human ovary. Exposure to cisplatin induced DNA damage, activated TAp63 and SAPK/JNK pathway, and triggered apoptosis in the oocytes and granulosa cells. However, TAp63 activation after cisplatin was not associated with any increase in the expression of c-Abl. Imatinib did not prevent cisplatin-induced apoptosis of the granulosa cells or oocytes. Moreover, treatment with this drug resulted in the formation of bizarre shaped follicles lacking oocytes and increased follicular atresia by inducing apoptosis of granulosa cells and oocytes. Similar toxic effects were observed when ovarian tissue samples were incubated with a c-kit antagonist drug anti-CD117, but not with another c-Abl tyrosine kinase inhibitor GNF-2, which lacks an inhibitory action on c-kit. Intraperitoneal administration of imatinib to the xenografted animals produced similar histomorphological abnormalities in the follicles in human ovarian grafts and did not prevent cisplatin-induced follicle loss when co-administered with cisplatin. Our findings provide, for the first time, a molecular evidence for ovarian toxicity of this drug in human. Furthermore, this study together with two previous case reports of a severely compromised ovarian response to gonadotropin stimulation and premature ovarian failure in patients, while receiving imatinib, further heighten the concerns about its potential gonadotoxicity on human ovary and urge caution in its use in young female patients.

Is there any in vitro evidence for or against ovarian protection by co-administration of a GnRH agonist with chemotherapy in human? The co-administration of GnRH agonist leuprolide acetate with cytotoxic chemotherapy agents does not preserve ovarian reserve in vitro. Randomized controlled trials of the co-administration of gonadotrophin-releasing hormone (GnRH) agonists with adjuvant chemotherapy to preserve ovarian function have shown contradictory results. This fact, together with the lack of a proven molecular mechanism of action for ovarian protection with GnRH agonist (GnRHa) places this approach as a fertility preservation strategy under scrutiny. We therefore aimed in this study to provide in vitro evidence for or against the role of GnRHa in the prevention of chemotherapy-induced damage in human ovary. This translational research study of ex vivo and in vitro models of human ovary and granulosa cells was conducted in a university hospital between 2013 and 2015. Ovarian cortical pieces (n = 15, age 14–37) and mitotic non-luteinized (COV434 and HGrC1) and non-mitotic luteinized human granulosa cells (HLGC) expressing GnRH receptor were used for the experiments. The samples were treated with cyclophosphamide, cisplatin, paclitaxel, 5-FU, or TAC combination regimen (docetaxel, adriamycin and cyclophosphamide) with and without GnRHa leuprolide acetate for 24 h. DNA damage, apoptosis, follicle reserve, hormone markers of ovarian function and reserve (estradiol (E2), progesterone (P) and anti-mullerian hormone (AMH)) and the expression of anti-apoptotic genes (bcl-2, bcl-xL, bcl-2L2, Mcl-1, BIRC-2 and XIAP) were compared among control, chemotherapy and chemotherapy + GnRHa groups. The greatest magnitude of cytotoxicity was observed in the samples treated with cyclophosphamide, cisplatin and TAC regimen. Exposure to these drugs resulted in DNA damage, apoptosis and massive follicle loss along with a concurrent decline in the steroidogenic activity of the samples. GnRHa co-administered with chemotherapy agents stimulated its receptors and raised intracellular cAMP levels. But it neither activated anti-apoptotic pathways nor prevented follicle loss, DNA damage and apoptosis induced by these drugs. Our findings do not conclusively rule out the possibility that GnRHa may offer protection, if any, through some other mechanisms in vivo. GnRH agonist treatment with chemotherapy does not prevent or ameliorate ovarian damage and follicle loss in vitro. These data can be useful when consulting a young patient who may wish to receive GnRH treatment with chemotherapy to protect her ovaries from chemotherapy-induced damage. This study was funded by the School of Medicine and the Graduate School of Health Sciences of Koc University, and the American Hospital Women's Health Center, Comprehensive Cancer Care and Fertility Preservation Programs, Istanbul, Turkey. The authors declare no competing interests.