Effects of Metformin on Reproductive, Endocrine, and Metabolic Characteristics of Female Offspring in a Rat Model of Letrozole-induced Polycystic Ovarian Syndrome With Insulin Resistant


 Background: Polycystic ovary syndrome (PCOS) is a complex and heterogeneous endocrine disorder that has many characteristic features including hyperandrogenemia, insulin resistance and obesity. The beneficial effects of metformin on reproduction, metabolism and endocrine, especially with the capacity to ameliorate insulin resistance (IR) in polycystic ovary syndrome (PCOS), place it as a good alternative for its widely prescribed, but its association with PCOS offspring remains uncertain. We aim to investigate the impact of metformin on reproductive, endocrine and metabolic characteristics in female offspring born to letrozole-induced PCOS-IR rats.Methods: 45 female wistar rats were implanted with letrozole-continuous-release pellets or placebo, subsequently treated with metformin or vehicle control, then mated with healthy male wistar rats. Estrous cycle, endocrine hormone profile, fasting insulin measurements and glucose tolerance test have been investigated and the expression of INSR in pancreas, FSHR and LHCGR in ovaries have been analyzed with Quantitative Real-time Polymerase Chain Reaction and Western Blotting assay.Results: Decreased conception rate and increased multiple pregnancy rates were found in PCOS-IR rats, which significantly improved after metformin treatment. Metformin significantly decreased the risk of body weight gain and increased INSR expression of pancreas in female F1 offspring in PCOS-IR rats. Decreased FSHR and increased LHCGR expressions in ovary were observed in female F1 rats of PCOS-IR and PCOS-IR+Met group. Nevertheless, there were no significant differences of INSR, FSHR and LHCGR expressions in female F2 offspring of PCOS-IR rats, as well as other PCOS phenotypes.Conclusions: The current study indicates that widespread reproductive, endocrine and metabolic changes in letrozole induced PCOS-IR rat model, but those PCOS phenotypes could not be inherit to offspring generations stably. Metformin may contribute to improve obesity, hyperinsulinemia and insulin resistance of female F1 offspring in PCOS-IR. The results of this study can be used as a theoretical basis for supporting metformin-using in the treatment of PCOS-IR patients.

as development of gestational diabetes (GDM) and hypertensive disorders (5,6). Infants born to women with PCOS are also predisposed to many adverse health outcomes. Accumulating evidence indicates that offspring of PCOS mothers is at higher risk for preterm birth, perinatal mortality, congenital abnormalities, increased hospitalization (6)(7)(8). In addition, these infants are more likely to be born by caesarean section and have lower birth weight (7). Obesity, metabolic dysfunction, comorbility and pregnancy complications of PCOS are likely to provide suboptimal intrauterine environment, which may have a detrimental impact on health of infant and prepubertal children born to women with PCOS, which may contribute to increased risk for their female offspring to also develop PCOS (9,10). PCOS has been reported to occur in 35% of premenopausal mothers of patients with PCOS (11). Underlying pathophysiological mechanisms of pregnancy complications along with its association with health of offspring remain uncertain.
Regarding the ethical limitations in human studies, animal models of PCOS provide a valuable tool to study the developmental origin, pathogenesis, mechanisms, long-term metabolic and endocrine consequences of PCOS, aid in detection of therapeutic strategies for PCOS patients, and also in uence of early-life environment on offspring characteristics in later life. Various strategies have been used to induce rat model of PCOS including exposure to androgens, estrogens, antiprogesterone agents, and constant light and genetic modi cations during the critical period of pre/postnatal life (12). These models have many features of PCOS, but few have both metabolic and endocrine abnormalities.
However, letrozole, a nonsteroidal aromatase inhibitor, blocks the conversion of androgens to estrogen and thus increases androgen level. In female rats, letrozole disruption shows disastrous cyclicity, elevated LH and T levels, anovulation, absence of corpus luteum, thickened theca cell layers, thin granulosa cell layers and increased ovarian weight. Continuous administration of letrozole to female rats starting before puberty induces both endocrine/reproductive and metabolic abnormalities similar to those in women with PCOS (13). Speci cally, these rats had insulin resistance, an increased subcutaneous fat mass, enlarged adipocytes in both subcutaneous and visceral adipose tissue, anovulation, and polycystic ovarian morphology, as well as increased secretion of LH, decreased secretion of FSH, and a high ovarian expression of mRNA of Cyp17a1 (13,14). Therefore, those letrozole induced rats are appropriate for studies of the mechanisms, consequences, and treatment of PCOS.
Reduced conception rates associated with PCOS may be related to hyperandrogenism, obesity and insulin resistance. The prevalence of insulin resistance (IR) in the general population is 10 ~ 25%, whereas in PCOS patients, it is approximately 60 ~ 70% (15). Increased insulin resistance and compensatory high insulin concentrations (hyperinsulinemia) play a key role in the development of PCOS (16). IR, a prominent feature but not a criteria for PCOS, affects the majority of women with the syndrome, and play a role in the development of PCOS. Progressively, IR can lead to glucose intolerance, which occurs in 40% of women with PCOS after the age of 40, and half of these women become diabetic within six years (17). Pathophysiology of IR could be due to a defect in signal transduction of insulin (18).
Metformin is an insulin sensitizer and is predominantly used in the treatment of type 2 diabetes (19,20). Metformin inhibits hepatic gluconeogenesis and reduces the action of glucagon, resulting in a reduction in circulating insulin and glucose. Metformin is known to exert its effect on liver, adipose tissues and ovary. Few studies have reported anthropometric and endocrine differences in PCOS women with and without IR (15). As IR and resulting hyperinsulinemia are key metabolic features in women with PCOS, their amelioration through metformin could improve PCOS-associated symptoms and conception rates (21). Lovvik et al reported a reduction in the risk of late miscarriage and preterm birth in those who received metformin throughout pregnancy (22). Less is known about the effect of metformin on offspring of maternal PCOS with IR (PCOS-IR).
In the present study, we aimed to explore the impact of metformin treatment on the reproductive alterations, and developmental, endocrine, and metabolic characteristics in the rst-and secondgeneration female offspring born to letrozole-induced PCOS-IR rats.
Material And Methods

Animals and experimental design
The study was approved by the Animal Ethics Committee of West China Second University Hospital of Sichuan University. A total of 45 female Wistar rats (F0) were included in the study. The rats were purchased from Chengdu Dashuo Experimental Animals Limited Company (Chengdu, Sichuan, China), housed ve per cage under standard conditions (12:12-h of light-dark cycle; at 23 ± 2℃; 55-65% humidity), with ad libitum access to food and tap water. At the age of 21 days, the F0 rats were randomly divided into two experimental groups, control group (n = 15), and letrozole group (n = 30). letrozole rats were implanted with 90 days letrozole-continuous-release pellets randomly (Innovative Research, USA.), containing 36 mg of letrozole (daily dose, 400 µg daily). Control rats were implanted with placebo lacking the bioactive molecule (Innovative Research, USA.). Body weight of each rat was recorded weekly from 21d of age.
Two PCOS rats presented glucose repair, which had been excluded from the following research. PCOS-IR model rats were characterized as PCOS rats with insulin resistant but without glucose repair or diabetes.
Twenty adult F0 female PCOS-IR rats were selected and randomly divided into PCOS-IR group (n = 10) and PCOS-IR + Met group (n = 10). Ten rats were selected from controls as control group. Rats from PCOS-IR + Met group were treated 20 days' metformin treatment. Rats from PCOS-IR group and control were treated without nothing but with same feed and rear environment. For offspring acquisition, control ovarian stimulation (COS) was induced as previously described by Dowland et al (23). Brie y, 20 IU of pregnant mare serum gonadotropin (PMSG) (Jianglai Biological, Shanghai, China) was injected intraperitoneally, followed by 48 h later intraperitoneal injection of 20 IU human chorionic gonadotrophin (hCG) (Jianglai Biological, Shanghai, China). COS induction was given to all three groups. All rats were mated with healthy male wistar rats. The nal pregnant rat number of each group was recorded after picking up of the vaginal plug, then record the successful delivery number, the number of offspring and birth-weight in each group. Indexes of glucose metabolism were measured in sexual development period and after sexual maturity. The level of serum testosterone(T) was measured after sexual maturity. 10 female rats were selected from each rst generation (F1) group randomly, 1:1 caged mating (to avoid inbreeding) with healthy male wistar rats. The remaining female F1 rats were killed for insulin receptor (INSR) expression testing in the pancreatic tissues, follicle stimulating hormone receptor (FSHR), and lutropin-choriogonadotropic hormone receptor (LHCGR) expressions in the ovarian tissues. Upon the birth of the second generation (F2), birth weight was measured in each group, and the rest tests were same as F1. The experimental design and grouping were shown in Fig. 1.

Vaginal smears
The stage of cyclicity was determined by microscopic analysis of the predominant cell type in vaginal smears obtained daily from 11 weeks of age. The stage of the estrus cycle was determined by the main cell types in vaginal smears: proestrus (round, nucleated epithelial cells), estrus (corni ed squamous epithelial cells), metestrus (corni ed squamous epithelial cells and leukocytes), and diestrus (nucleated epithelial cells and leukocytes) (24).
Tissue sampling At 12 week of age (9 weeks after pellet implantation), blood samples were obtained for analyses of progesterone(P), 17β-estradiol(E2), testosterone(T), follicle stimulating hormone (FSH) and luteinizing hormone (LH). All blood samples were stored at -20 ℃. The rats were decapitated, and ovaries were excised, xed in neutral buffered 4% paraformaldehyde for 24 h, placed in 70% ethanol, dehydrated, and embedded in para n.

Morphological study
After processing of formalin-xed ovaries through para n embedding, they were longitudinally and serially sectioned into 5-µm thick slices, stained with hemotoxylin and eosin (H&E). The morphology of the ovarian tissues was evaluated under a light microscope by two persons blinded to the origin of the sections.
Endocrine hormone pro le, Fasting insulin measurements and Glucose tolerance test After rats were sacri ced, blood samples were obtained to analyze P, E2, T, FSH and LH. Plasma samples were stored at − 20 °C. The endogenous hormone levels were determined by radioimmunoassay kit according to the protocol provided. Fasting insulin (FINS) and fasting glucose (FPG) were measured by radioimmunoassay kit as previously mentioned after 8 hours of overnight fast. After that, rats were injected intraperitoneally with a bolus of 1 g/kg glucose in 0.9% NaCl. Blood glucose determination was assessed at 15, 30, 60, 120 minutes post injection. All kits contained standard samples for samples for quality control and were used in accordance with manufacturer instructions. Additionally, fasting insulin and glucose values were used to determine HOMA-IR (homeostasis model assessment of insulin resistance), which was calculated as: Fasting insulin (mIU/L) x Fasting glucose (mmol/L)/22.5 (25).

RNA isolation and Real-Time Quantitative-PCR (RT-qPCR)
Total RNA was extracted from tissues using TRIzol reagent (Life Technologies Inc., Carlsbad, CA, U.S.A.), according to the manufacturer's protocol. The quality and puri cation of RNA were analyzed using NanoVue Plus spectrophotometer (Health-care Bio-Science AB, Uppsala, Sweden). cDNA was synthesized from puri ed total RNA using a PrimeScript RT reagent kit (TaKaRa Biotechnology Co Ltd, Dalian, China). Primer sequences for INSR, FSHR, and LHCGR (Sango Biotech, Shanghai, China) were shown in Table 1. RT-qPCR was performed by using SYBR Green real-time PCR Master Mix (Toyobo, Osaka, Japan) and measured on an Applied Biosystems 7900 real-time PCR detection system (ABI, Foster City, CA, USA). The speci city of PCR products was con rmed by analysis of the dissociation curve. GAPDH was used as an endogenous control to normalize the target gene expression and the relative expression was calculated according to the formula 2 −ΔΔCt . All experiments were repeated 3 times. Table 1 Primer sequences of the qRT-PCR analysis Gene Forward primer (

Results
The establishment of PCOS-IR rat model After three weeks of pellet implantation, body weight in the PCOS group increased remarkably (Fig. 2C). The controls showed a normal estrous cycle of 4-5 days at both time points. The PCOS rats exposed to either dose of letrozole were completely acyclic, and vaginal smears showed that leukocytes were predominant, indicating pseudo-diestrus ( Fig. 2A). Ovaries from PCOS rats were enlarged and surrounded by fatty tissue (Fig. 2B). The morphology of ovary was observed by using hematoxylin-eosin staining.
The area of the large follicle was greater in PCOS rats than controls. Letrozole-treated rats showed increases in ovary weight, area of the largest follicle, and number of cystic follicles as compared to control rats, and their ovaries contained atretic antral follicles and follicular cysts. An increased number of cystic follicles, granular cell layer thinning, and thickening of the theca cell layer was observed (Fig. 2E). Concentration of sex hormones in serum showed in Fig. 2D. Treatment with letrozole resulted in a rising of LH and T levels in PCOS group than that in controls. In contrary, the concentration of E2, P and FSH levels were reduced signi cantly in these rats. Increased fasting insulin (FINS) level and impaired insulin resistance, measured by HOMA-IR, were also observed in letrozole treated PCOS rats ( Table 2). These results indicated abnormal endocrine and metabolic changes in PCOS rats. Two rats, which presenting glucose intolerance, excluded from the following researches. A rat model of PCOS with insulin resistance (PCOS-IR) has been established. The effects of Metformin treatment on Pregnant outcomes in PCOS-IR rats Five rats were pregnant in PCOS-IR group, seven in PCOS-IR + Met group and nine in control group, conception rate of control rats was the highest. It was higher in PCOS-IR + Met group than PCOS-IR group (Table 3). Two rats in PCOS-IR group and one in PCOS-IR + Met group died in the late gestation for placental abruption. Therefore, three rats in PCOS-IR group were successfully delivered, six in PCOS-IR + Met group and nine in Control group. The average numbers of offspring in PCOS-IR and PCOS-IR-Met group were higher than in control group, the differences were statistically signi cant (Table 3). There were no signi cant changes of the ratio of male to female during those three groups. The effects of Metformin on Endocrine and Metabolic alters in F1 female offspring The birth weight of PCOS-IR + Met and PCOS-IR rats were lower than that in control group dur to the increased number of new-born rats. The body weight of the F1 female rats in PCOS-IR group was higher than PCOS-IR + Met and control group after 63 and 70 days in sexual maturity period, the differences were statistically signi cant (P < 0.05). There was no signi cant difference of body weight between PCOS-IR + Met and control group in sexual maturity period (Table 4). Glucometabolic index measured by IPGTT test of female F1 rats among all the three groups both in sexual development period and sexual maturity period showed no signi cant differences, and the same as FINS, HOMA-IR and T level (Table 4). The mRNA expression of INSR in pancreas of those three groups rats were analyzed by RT-qPCR. We found that INSR expressions of female F1 rats in PCOS-IR groups were signi cantly lower than that in control group (Fig. 3A). Furthermore, metformin increased the INSR expression in PCOS-IR rats (Fig. 3A). RT-qPCR was used to analyze the expression levels of endocrine related genes, including FSHR and LHCGR in ovary. The results showed that the expressions of FSHR in the ovary of PCOS-IR and PCOS-IR + Met group signi cantly decreased compared with control group; while there was no signi cant difference of its expression in female F1 between PCOS-IR + Met and PCOS-IR group (Fig. 3A). The results revealed that, LHCGR expression of PCOS-IR female F1 rats was signi cantly higher than that in control group, but there was no difference in LHCGR expression of F1 rats between PCOS-IR + Met and PCOS-IR group, as well as its expression between PCOS-IR + Met and control group (Fig. 3A). Western blotting was used to analyze related protein expression. The protein expressions of INSR in pancreas, FSHR and LHCGR in ovary have the same trend with their mRNA expressions analyzed by RT-qPCR (Fig. 3B).
Together, these ndings indicate high risk of PCOS development of female F1 rats with female ancestral exposure to excessive letrozole after mating with healthy males. Metformin treatment reduces its incident in female F1 offspring.

The effects of Metformin on Endocrine and Metabolic alters in PCOS-IR F2 female offspring
For the F2 generation, we did not detect signi cant difference of body weight in either group. Glucometabolic indexes of F2 female rats among all the three groups in both sexual development period and sexual maturity period showed no signi cant difference, as well as glucose levels of IPGTT, FINS, HOMA-IR and T level ( Table 5). The mRNA and protein expression pro les of INSR in pancreas tissues and FSHR and LHCGR in ovarian tissues collected from female F2 rats were similar during those three groups (Fig. 4).

Discussion
Given the complexity of human body, it is di culty to investigate the effect of the prenatal environment on subsequent generations in humans. High insulin resistance in PCOS patients is a hot topic in the eld of reproductive endocrinology and insulin resistance appears to be the fundamental key factor within the pathophysiology (26). Metformin treatment in PCOS patients still needed to deeper investigated, especially the effect of metformin on the offspring of PCOS-IR women has been rarely reported. Continuous administration of letrozole is appropriate to be used to replicate PCOS-like phenotypes in rodents. In the current study, rats with letrozole exposure showed PCOS-like reproductive and metabolic phenotypes compared with the controls. Consistently with previous study (27) A successful PCOS model was crucial for the continuation of the study. In our study, letrozole treated rats showed increased body weight, and PCOS-like reproductive and metabolic phenotypes, including signi cant alters in serum T, E2, LH, FSH and P, impaired insulin resistance and widespread metabolic abnormalities, as well as disrupted estrous cycles and poly ovaries. Compared to Maliqueo' study (13), we also investigated the glucose impair or diabetes in letrozole induced PCOS rat model. IPGTT tests reveal that there are no signi cantly changes in glucose levels in 0, 15, 30, 60, 120 min between PCOS group and controls, which is indicating that those letrozole treated rats had the PCOS phenotypes of obesity, hormone levels alter and insulin resistance, but without type 2 diabetes or glucose intolerance. Abnormal insulin tolerance has been observed in our study, but the glucose levels maintained normal. Glucose levels may remain normal in PCOS despite insulin resistance because of compensatory increased pancreatic β-cell insulin production resulting in hyperinsulinemia (30). A successful PCOS-IR female rat model has been induced, more importantly, the follow-up downregulation study that based on this model was effective and valuable.
In present study, conception rate of PCOS-IR rats decreased, and metformin could improve the conception rate of PCOS-IR rats, multiple pregnancy rates in PCOS-IR rats' groups signi cantly increased compared with controls, within PCOS-IR rats, it decreased after metformin treatment. Metformin increases the fetal concentration of sex hormone-binding globulin when taken during pregnancy (31) and reduces the secretion of in ammatory cytokines from trophoblast cells in vitro (32). Patients with PCOS have an increased risk for pregnancy complications (33). In present study, metformin has consistently been associated with improved ovulation and low multiple pregnancy rates. A Cochrane review on metformin concluded that ovulation and pregnancy rates were higher in women with PCOS taking metformin (34). A recent study found that in pregnant women with PCOS, metformin treatment from the late rst trimester until delivery might reduce the risk of late miscarriage and preterm birth, but no substantial difference in serious adverse events in either mothers or offspring (22). In present study, we found two cases of placenta abruption among PCOS-IR rats. Palomba et al (33) found a 3-to 4-fold increase in pregnancy induced hypertension (PIH) and preeclampsia, 3-fold increase in GDM, and a 2-fold higher chance for premature delivery among PCOS patients. Hyperandrogenism, obesity, insulin resistance, and other metabolic abnormalities, may contribute to the increased risk of obstetric and neonatal complications (33). Women with PCOS have shown placental in ammation, placental thrombosis and infarction during pregnancy, added to villous immaturity and nucleated fetal red blood cells (35), which may contribute to the placenta abruption of PCOS. More probable, maternal metabolic dysfunction in PCOS mothers compromises placental function of a female fetus with a genetic susceptibility to PCOS, promoting fetal hyperinsulinemia as cause for hyperandrogenism and altered folliculogenesis in utero (33,36).
In present study, the data revealed that the body weight of the F1 female rats in the group PCOS-IR was higher than both group PCOS-IR + Met and group Control after sexual development period, it showed that the female F1 offspring, which exposed to metformin, demonstrates higher weight gain compared to the offspring of mothers not treated with the drug. For F2 generation, no signi cant difference of body weight was detected in either group, which demonstrated that the phenotype of obesity is unstable to inherit to F2 generation in PCOS. Metformin use in PCOS pregnancies increases the risk of offspring overweight at 4 years of age compared to placebo-exposed children (37), but in their study, the clinical implications, body composition and metabolic health of these children are not known, which still should be subject to further investigation with long-term follew up of children. The T level was increased after letrozole administration, as usually observed in these murine models (13,38). Our data revealed that glucometabolic indexes of both F1 and F2 female rats among all the three groups in both sexual development period and sexual maturity period showed no signi cant difference, as well as the serum T level.
Fetal nutritional and endocrine programming in utero may affect the neuroendocrine systems with longterm health consequences, such as hypertension, hypercholesterolemia, impaired glucose tolerance (39).
It is believed that both genetic and early-life environmental factors in the uterus may contribute to the development of PCOS (40). The effect of the insulin resistance of PCOS on fetal growth and in utero programming, independent of either obesity or gestational diabetes, has yet needed to be elucidated. Hyperinsulinemia in PCOS could possibly be due to defects in the expression and/or activity of insulin receptor (INSR) (15). In our study, drastic decreased expression of INSR in pancreas was observed in PCOS-IR F1 rats compared to controls. The number of INSR is a main determinant of cellular response to circulatory insulin, any decrease signi cantly reducing insulin sensitivity, and functional INSR are critical for insulin binding and signal transduction, any alterations in INSR markedly impairs insulin binding and subsequent signaling pathways (41). Presence of hyperinsulinemia in PCOS-IR has also been reported earlier, which may be due to some defect in downstream of INSR (42,43). The cellular insulin resistance in polycystic ovary syndrome has been further shown to involve a novel post-binding defect in insulin signal transduction, treatment of insulin resistance with a diabetes drug, such as metformin has become mainstream therapy in PCOS women (30). In present study, metformin could signi cantly increase the INSR expression of PCOS-IR F1 rats, which may contribute to improve hyperinsulinemia and insulin resistance, and reduce the incidence of PCOS-IR F1 correspondingly. Nevertheless, after mating with normal male rats, the INSR expression of F2 rats has no differences between PCOS-IR rats and controls, no matter with or without metformin treatment. However, metformin treatment can signi cantly reduce the expression level of INSR gene after sexual maturation of PCOS-IR rats' F1 female rats, which may reduce the incidence of PCOS correspondingly. Metformin has been shown to confer protective effects on mouse pancreas exposed to fatty-acid induced stress, and chronic high glucose exposure (44). Hence, the effects of metformin on metabolic risks in human pancreas remain to be determined.
Metformin treatment in PCOS has been associated with increased hepatic synthesis of sex hormonebinding globulin and decreased ovarian and adrenal androgens. FSH is an important endocrine hormone in regulating ovarian function, which is must bind to special receptor, FSHR, stimulates follicular development via FSHR activation and induces granule cell proliferation.

Conclusion
In conclusion, those ndings indicate a higher risk of reproductive, metabolic and endocrine abnormalities in PCOS and F1 offspring, and metformin treatment could improve pregnancy outcomes of PCOS. However, related data from F2 have no changes. In women with PCOS, a possible relationship with genetic, environmental, clinical and biochemical factor involved in this complex condition. The nongenetic factors play also an important role in the pathogenesis of PCOS. However, metformin treatment can signi cantly reduce the expression level of INSR gene after sexual maturation of PCOS-IR rats' F1 female rats, which may reduce the incidence of PCOS correspondingly. The results of this study can be used as a theoretical basis for supporting metformin-using in the treatment of PCOS-IR patients who are without impaired glucose tolerance or type two diabetes mellitus.  Schematic representation of experimental design and grouping. Female rats were implanted with letrozole-continuous-release pellets to establish PCOS-IR rat model. F0 female rats were randomly divided into PCOS-IR+Met, PCOS-IR and Control groups and crossed with healthy male rats to obtain F1 offspring. F1 female offspring crossed with healthy male rat to obtain F2 offspring. vs. control, t-test; E. Histology of Ovaries. Ovaries from PCOS groups rats contained increased number of cystic follicles, granular cell layer thinning, and thickening of the theca cell layer. Ovaries from controls showed corpora lutea and mature follicles with healthy oocytes and thick layer of granulosa cells (scale bar=200μm).