Introduction
Polycystic ovary syndrome (PCOS) is a leading cause of anovulatory infertility (
Azziz et al., 2016). However, owing to the inability to trace the whole process, and a lack of knowledge regarding the factors that signal both oocyte maturation and ovulation, the etiology and pathophysiology of PCOS remain elusive after decades of research. Although genetic variations in nearly 20 genes have been linked to PCOS, a driver gene is yet to be established (
Franks and McCarthy, 2004;
Mykhalchenko et al., 2017). Endocrine and metabolic abnormalities, including hyperandrogenism and associated traits such as increased serum testosterone and androstenedione levels, high estrogen (
Robinson et al., 1992), altered gonadotropin secretion (
McCartney et al., 2002), obesity (
Legro, 2012), and symptoms of type 2 diabetes (
Pelusi et al., 2004), are common in PCOS patients, suggesting that PCOS is an endocrine and/or metabolic disease. However, attempts to induce an animal PCOS model using known hormones or metabolites have had limited success, suggesting that follicle development- and ovulation-driving molecules either fluctuate during follicle development or are distinct from known hormones or metabolites.
Folliculogenesis is promoted by follicle-stimulating hormone (FSH) and luteinizing hormone (LH)—two hormones secreted by the pituitary gland during the menstrual cycle. FSH and LH levels decline during and after ovulation, respectively, and are considered regulators of ovulation. However, it is unclear whether both or one of them plays key ovulation-regulatory roles. A follicle is composed of a multiple-layered wall with outer layers of theca cells (TC) that produce androgen and a much thicker multilayer of granulosa cells (GC). Ovarian steroidogenesis occurs through LH receptors on the theca to produce androgen and through FSH receptors on granulosa cells (
Raju et al., 2013) to convert androgen to estrogen, a hormone that is required for oocyte development and maturation (
Knobil and Neill, 1998). Ovulation requires a breakdown of the follicle wall, which can be achieved through either outside-in or inside-out manner. Outer TC may be less likely to initiate follicle rupture and ovulation because they are separated from oocytes by multiple layers of GC and may not receive oocyte maturation signals, a prerequisite for ovulation. In contrast, GC, which can communicate with oocytes through follicular fluid (FF), may be good candidates for the initiation of follicle rupture and ovulation because they can receive oocyte maturation signals in FF. Therefore, the apoptotic rate of GC is high in the dominant follicle selection and pre-ovulatory maturation stages of folliculogenesis (
Regan et al., 2018), and GC from patients with PCOS exhibit a lower apoptotic rate (
Das et al., 2008).
FF contains steroid hormones, proteins, metabolites, and antioxidants (
Hildpetito et al., 1991;
Selvam et al., 2019) and thus has physiological significance as a source of nutrients for both oocytes and GC. Moreover, FF metabolites may also transmit oocyte maturation and other signals to GC because they are signaling molecules (
Chantranupong et al., 2015;
Efeyan et al., 2015). To explore whether any molecules in FF may signal ovulation, we screened FF metabolite and found elevated glutamine in the FF of patients with PCOS. We further elucidated how the FSH-regulated FF glutamine controls ovulation by controlling ASK1-JNK-mediated GC apoptosis and proved activating ASK1-JNK pathway can be an intervening strategy for PCOS.
Results
GC apoptosis is associated with human PCOS and murine ovulation
To identify ovulation regulators, we compared follicular cell apoptosis in the follicles of C57BL/6 mice at various developmental stages. The apoptosis of GC, but not TC in follicles, increased from the preantral to antral to pre-ovulatory stages, and apoptosis signals were markedly accumulated in FF-exposed GC (Fig. 1A), suggesting that ovulation regulators are present in FF. In human FF samples, flow cytometry (Fig. 1B) and TUNEL assays (Fig. 1C) revealed lower percentages and numbers of floating apoptotic cells in patients with PCOS than those in normal FF, respectively. These results, together with the fact that only the apoptotic marker cleaved Caspase-3 (Cl-Caspase-3), but not the levels of necrosis (p-RIPK1 and pMLKL), ferroptosis (GPX4), and autophagy (LC3B) markers in the FF floating cells of PCOS patients were lower than those in control subjects (Figs. 1D, S1A and S1B), support the hypothesis that apoptosis of FF exposed-GC drives ovulation.
High glutamine levels in FF of patients with PCOS
To identify the possible ovulation signals in FF, an untargeted metabolomic survey was conducted on the metabolites in the FF of 35 patients with PCOS and 37 age-matched individuals as control (Table S1). Among the metabolites that differed in levels between PCOS and normal FF, several amino acids, including higher glutamine and lower lysine, arginine, histidine, and tyrosine, were found in the FF of patients with PCOS (Figs. 1E, S1C and Table S2). Levels of testosterone (T) (Fig. S1D) were elevated in the FF from PCOS patients, however, levels of estradiol (E2) (Fig. S1E) were comparable between the FF of control and PCOS groups. These results suggest that amino acid dysregulation may regulate ovulation.
Glutamine starvation increased, and glutamine supplementation decreased the apoptosis of cultured human granulosa COV434 cells, as previously reported (
Zhang et al., 2000). Altering other amino acid levels had negligible effects on COV434 cell apoptosis (Fig. 1F), suggesting that glutamine possibly regulates ovulation by mediating GC apoptosis. This notion was supported by the fact that pre-ovulatory porcine FF contained lower glutamine (Fig. 1G) but higher Cl-Caspase-3 (Fig. 1H) levels.
FSH promotes GC glutamine synthesis and FF glutamine levels
While tracing the origin of FF glutamine, we found that both FSH and LH levels were higher in the FF of patients with PCOS (Fig. 2A). Moreover, FSH, but not LH, increased glutamine levels intracellularly and in the culture media of primary mice GCs, KGN cells, and COV434 cells (Fig. 2B). These results suggest that FSH drives glutamine synthesis and secretion. FSH upregulated glutamine synthetase (GS), an enzyme that synthesizes glutamine from glutamate, protein levels (Fig. 2C), via upregulating
GS transcription (Fig. 2D) in these cells. These, together with the epidermal growth factor receptor (EGFR) inhibitors erlotinib and gefitinib, which bind to the tyrosine kinase domain and stop the activity of EGFR (
Yang et al., 2017), inhibit the ability of FSH to upregulate
GS transcription (Fig. 2D) and protein (Fig. 2C) levels. Mimicking FSH decrease at the pre-ovulation stage with FSH receptor binding inhibitor hFSH-β-(33-53) (TFA) (
Santa-Coloma et al., 1992) decreased GS levels (Fig. 2E) and glutamine levels in both cells and culture media (Fig. 2F) of FSH-stimulated COV434 cells, confirming that FSH promotes glutamine synthesis by promoting GS expression.
Immunofluorescence staining of mouse ovaries revealed that in primary follicles, GS was expressed mainly in oocytes and cumulus oophorus cells; GS expression in GCs, which have the same origin as cumulus oophorus cells, emerged in secondary follicles and, remarkably, GS expression was concentrated in the FF-exposed GC, and this trend was maintained throughout follicle development until the follicle rupture, in which GS expression diminished in the apoptotic GCs (Fig. 2G). Moreover, subcutaneous FSH injection in female mice promoted GS expression in GCs (Figs. 2H and 2I), consistent with that the GS expression was dependent on FSH stimulation (Fig. 2F). Furthermore, GS-overexpressing and GS-knockdown rendered COV434 cells resistant and sensitive to metformin-induced apoptosis, respectively (Fig. 2J). These results verify the hypothesis that FSH, which enhances GS expression by activating EGFR-mediated transactivation, controls GC GS and FF glutamine levels during follicle development and ovulation (Fig. 2K).
Glutamine inhibits GC apoptosis via ASK1-JNK apoptotic pathway
To investigate how GC apoptosis is regulated by FF glutamine, we starved the primary cultured ovarian GCs for glutamine to mimic glutamine reduction in FF during ovulation, employing theca cells, another type of cell in the follicle wall, as a control. GCs were more sensitive to glutamine starvation than theca cells (Fig. 3A), further supporting that glutamine deprivation specifically induces GC apoptosis during ovulation. Consistent with this, glutamine starvation-induced apoptosis in COV434 (Figs. 3B and S2A) and KGN (Fig. S2B) cells, two widely used human granulosa cell lines (
Zhang et al., 2000;
Nishi et al., 2001), in a time-dependent manner. Moreover, Jun N-terminal protein kinase (JNK) and P38, downstream targets of ASK1 that are inactivated by glutamine (
Ko et al., 2001), were activated by glutamine starvation in a time-dependent manner, whereas BCL-2 and Bcl-xL, two key molecules of the intrinsic apoptotic pathway, remained unaffected by glutamine deprivation (Figs. 3C and S2C). These results suggest that the death receptor- and ASK1-mediated extrinsic apoptotic pathway, rather than the mitochondrial-mediated intrinsic apoptotic pathway (
Sasaki et al., 2009), governs the glutamine-regulated apoptosis of GCs. This hypothesis is further supported by the following observations: (i) glutamine supplementation inhibited the phosphorylation of ASK1, JNK, and P38, but had negligible effects on BCL-2 and Bcl-xL in both COV434 and KGN cells (Figs. 3D and S2D); (ii) in the FF floating cells from PCOS patients, which were immersed in higher glutamine environment, ASK1-JNK apoptotic signaling, but not BCL-2 and Bcl-xL, was significant attenuated (Fig. 3E); (iii) ASK1 knockdown using short hairpin RNA in COV434 cells (Fig. S2E) did not affect the phosphorylation levels of JNK and P38 (Fig. 3F); (iv) overexpression of ASK1 in COV434 cells (Fig. S2F), enhanced the ability of glutamine to regulate JNK and P38 phosphorylation (Figs. 3G and 3H). Furthermore, phosphorylation of JNK, which is co-localized with cleaved Caspase-3, was noticed in the inner layer of GCs exposed to FF (Fig. 3I).
We previously found that glutamine suppresses ASK1-mediated apoptosis through glutaminyl-tRNA synthetase (QARS)-catalyzed ASK1 K688 glutaminylation (
He et al., 2018). To investigate whether FF glutamine depletion induces GC apoptosis via the same mechanism, we tested the effects of glutamine on QARS knockdown (Fig. S3A) and QARS overexpressing (Fig. S3B) COV434 cells. QARS knockdown reduced the ability of glutamine to suppress ASK1 signaling in a dose-dependent manner (Fig. S3C), and potentiated glutamine deprivation to activate ASK1 signaling (Fig. S3D), whereas QARS overexpression increased the ability of glutamine to suppress ASK1 signaling (Fig. S3E). These results are consistent with the hypothesis that FF glutamine depletion initiates GC apoptosis by activating ASK1 apoptotic signaling through QARS-mediated glutamine signaling.
To further address the upstream receptor that transduces death signaling of glutamine starvation to the extrinsic ASK1-mediated apoptotic pathway, we knocked down the two main death receptors using siRNA or blocked their functions with specific inhibitors. Unlike the TNFR antagonist R7050, KR33493, an inhibitor of ASK1-mediated apoptosis through Fas-signaling inhibition (
Jeong et al., 2016), inhibited JNK and P38 phosphorylation (Figs. 3J and S3F) and apoptosis (Figs. 3K and S3G) induced by glutamine deprivation in COV434 and KGN cells. Consistent with this, knockdown of Fas, but not TNFR, inhibited the ASK1-mediated extrinsic apoptotic pathway and apoptosis (Figs. 3L and 3M). Together, these results suggest that GC apoptosis is specifically mediated by ASK1-JNK signaling.
Glutamine and FSH regulates ovulation in vitro
We tested their efficacies of FSH and glutamine on isolated murine follicles. Glutamine dose-dependently inhibited follicular rupture of cultured murine follicles (Figs. 4A and 4B). Moreover, FSH administration decreased the apoptosis of granulosa cells with glutamine starvation (Fig. 4C) and inhibited follicular rupture of cultured murine follicles (Figs. 4D and 4E). Furthermore, FSH treatment, which stimulates glutamine synthesis, weakens the ASK1-JNK apoptotic signaling in cultured granulosa cells (Fig. 4F). These solidify that FSH and it-promoted FF glutamine directly regulate ovulation rather than regulating ovulation through alternative mechanisms such as changing other ovulation-relative hormones/signaling.
High glutamine induced PCOS traits
Next, we investigated whether high FF glutamine levels, but not that of other amino acids such as threonine, are a causal factor of PCOS. Either high-glutamine or high-threonine chows were fed to 5-week-old C57BL/6 mice (designated as GLN mice and THR mice, respectively), which increased ovary and serum glutamine levels by approximately 30%–40% that mimicked the elevated FF glutamine levels found in PCOS, and a similar degree of threonine elevation, respectively, compared to those of normal chow-fed C57BL/6 mice (designated as CON mice) after 3 months of treatment (Fig. 5A).
Physiological markers and PCOS traits were compared among mice fed with different chow (Fig. S4A). Both GLN and THR mice consumed less (Fig. S4B), and exhibited lower body weights (Fig. S4C) than that of CON mice, which is consistent with the findings of a previous study showing that a high-protein diet decreases body weight (
Westerterp-Plantenga et al., 2004). The GLN and THR mice also consumed more water (Fig. S4D) and excreted more urine than that observed for CON mice (Fig. S4E), which is likely because high glutamine and threonine diets produce more ammonia that must be secreted in the form of urine.
GLN mouse ovaries phenocopied the ovaries from estrogen receptor 1 knockout (EsR1-KO) mice and dihydrotestosterone (DHT) induced PCOS mice, two known PCOS mice models (
Lee et al., 2009;
Rodriguez Paris et al., 2020) (Figs. 5B and S5A), to exhibit an increased number of atretic (Fig. 5C), early and antral follicles, and fewer pre-ovulatory follicles and corpora lutea (Fig. 5D) than those of CON mice, whereas THR mouse ovaries were indistinguishable from those of CON mice (Figs. 5B and 5D). Notably, reduced GC apoptosis, and thicker and more condensed GC layers were observed in the follicles of GLN mice (Fig. 5E), which is consistent with the hypothesis that high glutamine-induced insufficient GC apoptosis may promote PCOS traits.
Unlike THR mice, GLN mice recapitulated the phenotypes of androgen-induced PCOS mice (
Risal et al., 2019), resulting in higher serum testosterone levels (Fig. 5F) than in CON mice. Moreover, a longer anogenital distance in F1, F2, and F3 female mice (Fig. 5G) and more dead embryos were found in GLN mice (Fig. 5H), which further confirmed high testosterone exposure in GLN mice. Furthermore, only GLN mice simulated EsR1-KO mice and DHT mice to induce irregular estrous cycles (Figs. 5I and S5B), had a longer estrous cycle (Figs. 5J and S5C), took longer to become pregnant (Fig. 5K), and had smaller litter pup numbers (Fig. 5L). Finally, moderate impairment of glucose tolerance and insulin tolerance was observed in GLN mice according to the glucose (GTT) and insulin tolerance test (ITT) assays, respectively (Figs. 5M and 5N), which is consistent with the observed reduction in the systemic insulin sensitivity mediator adiponectin (Fig. S5D) and a relatively normal basal level of insulin (Fig. S5E) in GLN mice. Additionally, GLN mice exhibited a lower subcutaneous fat mass (Fig. S5F) and smaller adipocyte sizes (Fig. S5G), which was consistent with the lower lipid storage in the liver (Fig. S5H), thereby refuting the possibility that glutamine induces PCOS symptoms by increasing the body fat. Collectively, these results indicate that high ovarian glutamine levels may induce PCOS traits.
Notably, although GLN mice exhibited an increase in blood testosterone (Fig. 5F), no statistically significant differences were observed in the levels of E2 (Fig. S5I), ovulation-promoting LH (Fig. S5J), and follicle-stimulating FSH (Fig. S5K), all of which are dysregulated by testosterone, suggesting that glutamine may perform other actions on these hormones, in addition to increasing testosterone levels.
Alleviating PCOS through glutamine deprivation
To further confirm that FF glutamine signals oocyte maturation and promotes GC apoptosis and ovulation, we examined the effects of glutamine removal in GLN mice. Glutamine was removed from GLN mice by feeding the mice normal chow for one month, which reduced their ovarian glutamine levels to levels comparable to those of CON mice (Fig. 6A). Glutamine removal was accompanied by fewer antral follicles and more corpus luteum in the ovaries of GLN mice (Figs. 6B and 6C), suggesting that removal of high-glutamine chow promoted ovulation. This finding was further substantiated by the observation that the removal of high-glutamine chow restored the regular estrous cycle of GLN mice (Fig. 6D), shortened the time to get pregnant (Fig. 6E), and increased the litter pup number to levels comparable to those of CON mice (Fig. 6F). Moreover, the removal of glutamine chow decreased the blood testosterone levels of GLN mice (Fig. 6G), but had negligible effects on E2 (Fig. S5L), FSH (Fig. S5M), and LH (Fig. S5N), which was consistent with the finding that glutamine chow did not alter these hormones (Figs. S5I–K) and the hypothesis that glutamine may have additional effects on hormones other than testosterone. Furthermore, the removal of the high-glutamine diet decreased glucose tolerance in GLN mice (Fig. 6H) and increased their sensitivity to insulin (Fig. 6I). Together, these results confirm that dynamic regulation of glutamine levels is associated with ovulation and PCOS traits.
Targeting Fas-ASK1 apoptotic pathway to alleviate PCOS
To verify that Fas-ASK1-mediated apoptotic pathway plays a critical role in follicular maturation and ovulation, female mice were treated with compounds that inhibit the Fas-ASK1 signaling pathway. Oral gavage of KR33493 and GS4997, which inhibit Fas and ASK1, respectively, induced PCOS-like symptoms in mice, such as polycystic ovary morphology (Figs. 7A and 7B), and elevated levels of serum testosterone (Fig. S6A), but not of estrogen (Fig. S6B).
To further confirm that GC apoptosis drives ovulation, we treated C57BL/6 female mice with AT-101, a compound that activates the ASK1/JNK pathway-mediated apoptosis (
Zerp et al., 2009). AT-101 specifically induced GC apoptosis in GLN mouse follicles (Figs. 7C, S6C and S6D), confirming that GC apoptosis is specifically regulated by the ASK1-JNK apoptotic pathway. Moreover, oral gavage of AT-101 stimulated ovulation, as shown by the increased number of corpora lutea (Figs. 7D and S6G). AT-101 treatment consistently reduced the number of early antral and antral follicles, increased the number of pre-ovulatory follicles (Fig. 7E), restored regular estrous cycles in GLN mice (Figs. 7F and 7G), decreased serum testosterone (Fig. 7H) but not estradiol (Fig. S6I), and increased glucose tolerance (Figs. 7I, and S6J) and insulin sensitivity (Figs. 7J and S6K). Furthermore, AT-101 treatment increased the number of pups in GLN mice (Fig. 7K). Together, these results confirmed that AT-101 attenuates the PCOS-induced effects of glutamine.
If insufficient GC apoptosis is the cause of anovulation, AT-101 treatment should also alleviate PCOS traits induced through other mechanisms. We tested this hypothesis by applying AT-101 treatment to an EsR1-KO-induced PCOS mice model (
Walters et al., 2012). AT-101 treatment induced GC-specific apoptosis (Figs. 7C, S6E and S6F), increased the number of corpora lutea (Figs. 7D and S6H), decreased the number of early antral and antral follicles, increased the number of pre-ovulatory follicles (Fig. 7E), restored regular estrous cycles in EsR1-KO mice (Figs. 7F and 7G), decreased serum testosterone levels (Fig. 7H), and increased glucose tolerance (Figs. 7L and S6L) and insulin sensitivity (Figs. 7M and S6M). Notably, AT-101 decreased the number of hemorrhagic cysts in the ovaries of EsR1-KO mice (Fig. S6N). Given that EsR1-KO mice had the same normal serum glutamine levels as control mice (Fig. S6O), these results confirmed that the induction of GC apoptosis alleviated PCOS traits.
Discussion
In the current study, by metabolic, cell biologic,
in vitro and
in vivo approaches, we found that FSH, the hormone that promotes GC proliferation and initiates follicle development for oocyte maturation, plays an additional role in regulating GC glutamine synthesis for ovulation. Glutamine plays a plethora roles in oocyte/follicular development/maturation. First, stemness oocyte maturation inside a follicle prefers glutamine as a nutrient source (
Vardhana et al., 2019). FSH-facilitated glutamine synthesis thus supplies oocyte nutrients. Second, follicle development requires the GC wall to remain intact, and FSH-facilitated glutamine synthesis can stabilize the GC wall by preventing GC apoptosis, which initiates ovulation starting from FF-exposed GCs. Third, glutamine is an ideal ovulation signal because it can be removed at pre-ovulation, by both the decline of FSH levels and by the consumption of matured human oocytes, which are as large as 100–120 µm at maturation (
Bae and Foote, 1975) and are effective glutamine scavenger. Thus, glutamine connects hormone, metabolic, and oocyte/follicle maturation signals to regulate oocyte maturation and ovulation (Fig. 2K).
Previous observations support the role of FF glutamine in ovulation. LH promotes glutamine metabolism in oocytes (
Zuelke and Brackett, 1993) and may promote ovulation by facilitating FF glutamine depletion. Moreover, many PCOS-related traits are associated with altered glutamine metabolism. For example, androgen signaling promotes glutamine uptake (
White et al., 2017), dysregulated insulin signaling (
Vuguin et al., 1999) and IGF-I stimulates glutamine uptake in cells cultured in glutamine-starved environments (
Wasa et al., 2001). These glutamine uptake-enhancing factors may provide additional support for the hypothesis that high glutamine levels promote PCOS, which was confirmed in our mouse model, and by that although glutamine ameliorates inflammation in DHAE-induced PCOS rats, it failed to ameliorate PCOS histology in them (
Wu et al., 2020). Our findings also suggest that some PCOS traits may be more consequential than causal. For example, hyperandrogenemia in PCOS may be induced by the retention of theca cells, which synthesize androgen (
Lischinsky and Armstrong, 1983), on unruptured follicles; thus, decreased luteum, whose formation requires ovulation (
Devoto et al., 2009), is expected to be a consequence of anovulation.
We reveal that the extrinsic apoptotic pathway, i.e. ASK1-JNK apoptotic pathway, mediates ovulation, and inhibition in ASK1-JNK pathway, such as by elevated FF glutamine, promotes PCOS. This adds to the intrinsic apoptotic pathway functions in follicular atresia for dominant follicle selection (
Tilly, 2001) for apoptosis to program oocyte/follicle development/maturation. That ovulation starts from apoptosis of FF-exposed interior GC follicle wall and that GC apoptosis is specifically regulated by the ASK1-JNK apoptotic pathway provides translation values for treating anovulatory diseases such as PCOS. We verified this possibility in mice by showing that metformin, an apoptosis inducer via ASK1-JNK pathway (
Feng et al., 2014;
Ma et al., 2019), induces ovulation (
Barbieri, 2003;
Nestler, 2008) and alleviates PCOS as documented (
Wang et al., 2020), and these effects can be recaptured by ASK1-JNK apoptotic pathway inducer AT-101 in both glutamine- and EsR1-KO-induced PCOS mouse models. These highlight that GC-specific, potent, and less toxic ASK1-JNK apoptotic inducers can be explored to be novel anti-PCOS drugs.
Materials and methods
Antibodies and reagents
The antibodies used in this research were purchased from the following sources: ASK1 (Cell Signaling Technology, Cat. No. 3762), p-ASK1 (Cell Signaling Technology, Cat. No. 3765), JNK (Cell Signaling Technology, Cat. No. 9252), p-JNK (Cell Signaling Technology, Cat. No. 4668), P38 (Cell Signaling Technology, Cat. No. 9212), p-P38 (Cell Signaling Technology, Cat. No. 9211), Caspase3 (Cell Signaling Technology, Cat. No. 9662), Cleaved-Caspase3 (Cell Signaling Technology, Cat. No. 9661), Bcl-XL (Cell Signaling Technology, Cat. No. 2764), LC3B (Cell Signaling Technology, Cat. No 3868), Bcl-2 (ABclonal, Cat. No. A0208), p-RIPK1(S166) (ABclonal, Cat. No. AP1115), RIPK1 (ABclonal, Cat. No. A19580), pMLKL(S358) (ABclonal, Cat. No. AP1244), MLKL (ABclonal, Cat. No. A13451), RIPK3 (ABclonal, Cat. No. A5431), QARS (Proteintech, Cat. No. 12645-1-AP), HA (Abmart, Cat. No.M20003), ACTIN (GeneScript, Cat. No. A00702), GPX4 (Abcam, Cat. No. ab125066), GS (Proteintech, Cat. No. 11037-2-ap), and α-SMA (Cell Signaling Technology, Cat. No.19245).
The reagents employed in this study were purchased from the following companies: Glutamine (Sigma-Aldrich, Cat. No. E6627), FSH (Sigma, Cat. No. F4021 and Solarbio, Cat. No. F8470), LH (Sigma, Cat. No. L6420 and Solarbio, Cat. No. L8040), β-Estradiol (Sigma, Cat. No. E8875), KR33493 (MCE, Cat. No. HY-100755), R7050 (MCE, Cat. No. HY-110203), AT-101 (Selleck, Cat. No. S2812), Gefitinib (MCE, Cat. No. HY-50895), Erlotinib (MCE, Cat. No. HY-50896), hFSH-β-(33-53) (MCE, Cat. No. HY-P3343A), and 13C6-Glucose (Sigma, Cat. No. 389374).
Sample collection
This study was approved by the ethics committee of the Obstetrics & Gynecology Hospital of Fudan University and the Shanghai Ji’ai Genetics & IVF Institute (JIAI E2021-020).
Human samples
All patients provided written informed consent prior to participation. A total of 72 female patients aged between 23 and 41 (Materials and methods Table S1) were enrolled at the Shanghai Ji’ai Genetics & IVF Institute. Thirty-five of these women were clinically diagnosed as having PCOS and were assigned to the PCOS group, whereas 37 age-matched patients who exhibited normal ovarian function but required IVF treatment for other reasons, such as male partner infertility and uterine or cervical abnormalities, were assigned to the control (CON) group. The two groups were balanced with respect to their body mass index (CON vs. PCOS: 22.07 ± 0.47 vs. 22.73 ± 0.51). Ovarian stimulation and egg retrieval were conducted in accordance with standard IVF guidelines. First, the eggs were picked for IVF purposes, then, FF samples from the same patient were combined and centrifuged to separate the FF and floating granulosa cells before they were subjected to liquid chromatography with tandem mass spectrometry (LC–MS/MS) analysis and cell death assays, respectively.
Porcine FF collection
Porcine ovaries obtained from prepubertal gilts at a local slaughterhouse were transported to the laboratory in phosphate-buffered saline at 37°C within 1 h. The follicular fluid was aspirated from superficial follicles with an insulin syringe. FF samples from the same pig were categorized as two samples according to the diameter of the origin follicle (3–8 mm in diameter as antral follicles, larger than 8 mm as pre-ovulatory follicles). All samples were centrifuged to separate the FF and floating granulosa cells before they were subjected to LC–MS/MS analysis and Western blot analysis, respectively.
Unbiased metabolomic analysis
First, 50 μL of FF or serum was mixed with 200 μL of cold methanol, then centrifuged at 12,000 ×g for 30 min at 4°C. The supernatant was filtered with 0.22 µm of polytetrafluoroethylene membrane and transferred to a fresh glass vial for LC–MS/MS analysis. LC–MS/MS analyses were performed using an ultra-high-performance liquid chromatography system (Vanquish, Thermo Fisher Scientific) with an ACQUNITY UPLC BEH Amide column (1.7 µm, 2.1 mm × 100 mm) coupled to a Q Extractive HFX mass spectrometer (Qrbitrap MS, Thermo Fisher Scientific). The mobile phase consisted of (A) 0.3% formic acid and 2 mmol/L ammonium formate in water, and (B) 95% acetonitrile plus 5% water (containing 0.3% formic acid and 2 mmol/L ammonium formate). The analysis was performed by gradient elution at a speed of 0.4 mL/min. The injection volume was 1 μL and the column temperature was 40°C. The data were analyzed using Agilent MassHunter Profinder according to the metabolomic molecular weight, retention time, m/z, and peak area.
Murine follicle culture and in vitro ovulation
In vitro murine follicle culture and ovulation was carried out as described (
Skory et al., 2015;
Xu et al., 2015). Briefly, ovaries from 14- to 16-day-old mice were cut into four pieces and incubated in α-MEM medium containing 0.1% collagenase I and 0.02% DNase I for 30 min. Enzymatic digestion was quenched with L15 medium containing 1% FBS, and secondary follicles were collected and encapsulated in 0.5% (
w/
v) alginate. Follicles were grown in α-MEM medium supplemented with 10 mIU recombinant FSH, 3 mg/mL BSA, 1 mg/mL fetuin, 5 µg/mL insulin, 5 µg/mL transferrin, and 5 ng/mL disodium selenite for 6 days with half of the medium was exchanged every 2 days. Then, alginate was removed with 10 U/mL alginate lyse, and follicles were rinsed with L15 medium containing 1% FBS. The follicular rupture was induced by incubating follicles in glutamine-free α-MEM medium supplemented with 1.5 IU/mL hCG, 5 ng/mL EGF, 3 mg/mL BSA, 5 µg/mL insulin, 5 µg/mL transferrin and 5 ng/mL disodium selenite for 16 h. For glutamine inhibition, glutamine (0 mmol/L, 2 mmol/L, 4 mmol/L or 8 mmol/L) was added, and for FSH inhibition, 1 mmol/L glutamine and FSH (0 mIU, 10 mIU, 20 mIU or 30 mIU) were added.
Cell apoptotic analysis
Granulosa cells from human FF and cell cultures were treated with a FITC Annexin V Apoptosis Detection Kit I (BD Biosciences, Cat. No. 556547) according to the manufacturer’s instructions. Briefly, granulosa cells were washed twice with phosphate-buffered saline (PBS), and re-suspended with 1 × binding buffer at a density of 1 × 106 cells/100 μL. Then, 5 μL of propidium iodide and 5 μL of Annexin V were added to the solution, which was kept in the dark for 15 min at room temperature. An additional 400 μL of 1× binding buffer was added to stop the reaction and the samples were analyzed using a flow cytometer (BD Biosciences).
For the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay, granulosa cells from human FF were washed twice with d2H2O and re-suspended in d2H2O at a density of 1 × 106 cells/mL. Then, 10 μL of cells was added and adhered on a poly-L-lysine coated slide. The cells or mouse ovary sections were then fixed with 4% paraformaldehyde solution for 15 min, followed by permeabilization with 0.2% Triton X-100 for 10 min at room temperature. Apoptotic cells were identified using the DeadEndTM Fluorometric TUNEL system (Promega, Madison, WI) according to the manufacturer’s instructions. Briefly, slides were incubated with the enzyme terminal deoxynucleotidyl transferase, recombinant at 37°C for 60 min, then incubated with 2 × SSC buffer for 15 min at room temperature to stop the reaction. Then, the nuclei were stained with DAPI and the slides were mounted with ProLong Glass Antifade Mountants (Thermo Fisher Scientific). Finally, mounted slides were subjected to confocal imaging using a Zeiss LSM710 system.
Effect of glutamine on apoptosis inhibition in granulosa cells
COV434 cells were cultured in RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (Gibco), KGN cells (the steroidogenic human ovarian granulosa-like tumor cell line) were cultured in DMEM/F12 medium (Hyclone) supplemented with 10% fetal bovine serum (Gibco). All cells were cultured in an incubator with 5% CO2 at 37°C. siRNA analysis and plasmid transfection were performed with Lipofectamine 3000 reagent (Thermo Fisher Scientific) according to the manufacturer’s instructions. siRNA was used at a concentration of 20 nmol/L, with 1 µg of plasmids for each well in a six-well plate.
Glutamine treatment in the cells was achieved by incubating cells in glutamine-free RPMI 1640 (Gibco) without serum, supplemented with different concentrations of glutamine, for 6 h before harvesting, as indicated in each experiment. For R7050 and KR33493 treatment, cells were incubated in glutamine-free RPMI 1640 without serum for 6 h before adding R7050 or KR33493, with the same concentration of R7050 or KR33493 added again after 1 h. Cells were then cultured for another hour and harvested for Western blot analysis.
Glutamine assays
For intracellular glutamine measurement, primary GCs, KGN, and COV434 cells were treated with 100 mIU/mL FSH, 100 mIU/mL LH, 1 µmol/L β-estradiol, or 100 mIU/mL FSH plus 1 µmol/L gefitinib, 1µmol/L erlotinib, and 200 ng/mL hFSH-β-(33–53) for 24 h, respectively. Cells were harvested and intracellular glutamine was determined using glutamine/glutamate-Glo assay (Promega, Cat. No. J8021). The data were normalized to the protein concentration, which was measured using BCA assay first, then normalized to the control group again to get relative glutamine levels.
To measure glutamine secreted to the culture media, primary GCs, KGN and COV434 cells were cultured in medium with 13C6-glucose for three generations, then cultured in regular medium with 100 mIU/mL FSH, 100 mIU/mL LH, 1 µmol/L β-estradiol, or 100 mIU/mL FSH plus 1 µmol/L gefitinib, 1µmol/L erlotinib, and 200 ng/mL hFSH-β-(33-53) for 24 h, respectively. The concentrations of 13C-labeled glutamine in the medium were qualified using LC/MS (Thermo Q Exactive HF). The data were normalized to the protein concentrations of the cells, which were measured using BCA assay first, then normalized to control group again to get relative glutamine levels.
Western blot analysis
Western blot was performed following standard protocols. Briefly, cells were harvested with a loading buffer containing 50 mmol/L Tris–HCl pH 6.8, 10% glycerol (v/v), 2% SDS (w/v), 4% β-mercaptoethanol (v/v), and 0.0012% bromophenol blue (w/v). For the analysis, each sample was subjected to SDS-PAGE and transferred to nitrocellulose membranes (GE Healthcare Life Science). The membranes were blocked in 5% (w/v) skim milk in Tris-buffered saline with 0.1% (v/v) Tween-20 (TBST) for 1 h at room temperature, then probed overnight with primary antibodies in an antibody dilution buffer (QuickBlockTM, Beyotime) at 4°C. After incubation with horseradish peroxidase-conjugated secondary antibodies in TBST (containing 5% skim milk), membranes were developed using ECL-Plus (Thermo Fisher Scientific) and visualized using Typhoon (GE Healthcare Life Science). All the Western blot experiments were biologically repeated at least three times.
High-glutamine chow mouse PCOS model
All animal procedures were conducted in accordance with the animal care committee at Fudan University. Female C57BL/6J mice were obtained from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). Mice were housed in a specific pathogen-free facility at 20–22°C on a 12-h light/dark cycle with ad libitum access to food and water. Mice were randomly divided into three groups. One group was fed with standard chow (CON), and the other two groups were fed with 30% glutamine chow (GLN) or 30% threonine chow (THR), respectively, both of which were manufactured by FBSW Biotechnology Co., Ltd. The timeline for the mice experiments is illustrated in Fig. S4A. Estrogen receptor 1 knockout (EsR1-KO) mice were purchased from Shanghai Model Organisms.
Serum analysis
The blood samples were collected from the submandibular vein and centrifuged at 3,000 rpm at 4°C or 15 min and stored at −80°C for subsequent serum determination. The levels of testosterone, estradiol, luteinizing hormone, and follicle-stimulating hormone were determined by ELISA assay kits for mice.
ITT and GTT analysis
Assays were performed on female C57BL/6J mice. For GTT analysis, mice were intraperitoneally injected with glucose (1 g/kg) after 16 h of fasting, and the blood was sampled 0, 15, 30, 60, 90, and 120 min after glucose injection. For ITT analysis, mice were intraperitoneally injected with 0.4–0.5 units/kg of insulin after 6 h of fasting, and the blood was sampled 0, 15, 30, 60, 90, and 120 min after insulin injection.
Mice treatments
For FSH and hFSH-β-(33-53) treatments, 8 weeks-old female C57BL/6J mice were randomly divided into sub-groups; and received intraperitoneal injection of 1 Unit/mice of FSH, or 1 Unit/mice of FSH plus 100 µg/mice of hFSH-β-(33-53), mice injected with saline were used as controls. After 8 h of injection, mice were sacrificed and the ovaries were harvested for immunostaining.
For drug treatments, 6 weeks-old female C57BL/6J mice were randomly divided into sub-groups; and received oral gavage of AT-101 (10 mg/kg/day), or intraperitoneal injection of KR33493 (2.5 mg/kg/day), GS4997 (2.5 mg/kg/day), and R7050 (2.5 mg/kg/day) continuously for 2–3 weeks. The mice were then sacrificed, and the ovaries and blood samples were harvested for staining or histological/biochemical assays, respectively.
IHC staining
Mice ovaries were harvested and fixed in 4% paraformaldehyde in PBS for 4 h. For hematoxylin and eosin (H&E) staining, ovaries were dehydrated using an ethanol gradient from 50% to 100% and cleared in xylene, then embedded in paraffin. Paraffin-embedded ovaries were sectioned to a thickness of 4 µm on a microtome and picked on microscope slides. After drying overnight at 37°C, ovary sections were subjected to H&E staining using the standard protocol, and images were captured with a Leica microscope. For immunostaining, paraformaldehyde-fixed ovaries were immersed in 30% sucrose overnight, embedded in an OCT matrix (Leica), and cut into 5 µm-thick sections. Sections were fixed in 4% PFA at room temperature for 30 min and permeabilized with 0.1% Triton X-100 in PBS for 15 min. Sections were blocked with 5% normal goat serum in PBS for 1 h at room temperature, and incubated overnight with first antibody (1:50 dilution in 5% normal goat serum) at 4°C. After washing with PBS, sections were incubated with fluorescent conjugated secondary antibody (1:100) in the dark for 1 h at room temperature. Sections were washed and incubated with α-SMA antibody (1:300 dilution in 5% normal goat serum) in the dark for 2 h at room temperature, then incubated with Alex Flour 488 conjugated goat anti-rabbit (1:500) in the dark for 1 h at room temperature after washing with PBS. Sections were washed again and incubated with DAPI for 5 min to display the nuclei. Sections were then mounted in a fluorescent mounting medium and images were captured using a Zeiss LSM710 confocal laser scanning microscope (Carl Zeiss Microscopy, LLC).
Statistical analysis
Statistical analysis was performed using GraphPad Prism 8.0 (GraphPad Software, Inc.). The two-tailed Student’s t-test was performed for the two-group analysis. Comparisons between groups were performed by the unpaired two-tailed Student’s t-test. Pooled results were demonstrated as the means ± SEM Differences were considered statistically significant if the P value was less than 0.05. Significance was indicated as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001.
The Author(s) 2024. Published by Oxford University Press on behalf of Higher Education Press.
PDF
(98394KB)
