Background
Polycystic ovary syndrome (PCOS) is a reproductive hormonal abnormality and a systemic-metabolic disorder. This syndrome affects an estimated 6–21% of reproductive-age women and is one of the main cause of infertility [
1]. Although the pathogenesis mechanism has not been well defined, PCOS is frequently associated with insulin resistance, chronic inflammation, and oxidative stress (OS) [
2,
3]. Increased secretion of LH compared to FSH and hyperandrogenism are also classical features of PCOS that 70–80% of women with hyperandrogenism are diagnosed with PCOS [
4,
5]. Hyperandrogenism is correlated wıth inflammation in PCOS pathogenesis [
6]. Furthermore, recent studies demonstrated that inflammatory markers, such as C-reactive protein (CRP), tumor necrosis factor (TNF), interleukin-6 (IL-6), interleukin-18 (IL-18), monocyte chemotactic protein-1 (MCP-1), and acute phase serum amyloid A (APSAA) increase in women with PCOS [
7‐
10].
Classical morphology of PCOS includes ovarian cortical thickening, multiple tiny capsular follicular cysts, luteinized inner theca, stromal hyperplasia and multiple immature follicles, which indicate cessation of folliculogenesis [
11]. Anti-Mullerian hormone (AMH) is secreted by the granulosa cells of secondary follicles and has an important role in both ovarian primordial follicle recruitment and dominant follicle selection [
12]. Since it has been suggested that there is a correlation between antral follicle count and AMH, this hormone has been proposed as a marker for PCOS [
13].
Resveratrol (3,5,4-trihydroxystilbene) is a natural polyphenol, found in grapes, red wine, peanuts, and several medicinal plants [
14]. This agent may have antidiabetic, antioxidant and antiinflammatory actions [
15‐
17]. Depending on the cell type, resveratrol has been shown to induce both proapoptotic [
18,
19] and antiapoptotic [
20,
21] effects in in vitro and in vivo studies. Kong et al. suggested that resveratrol significantly increases the total number of oocytes, significantly decreases the atretic follicles and inhibits both the primordial-to-developing-follicle transition and apoptosis in different age groups of rats [
22]. In a recent study in rodents, resveratrol protected oocytes from methylglyoxal-induced cytotoxicity, and this effect was mediated by decreasing reactive oxygen species (ROS) production [
23]. However, at present, there is a paucity of information about potential beneficial actions of resveratrol on reproductive functions and ovary.
Resveratrol interacts with multiple cellular targets, but many effects of resveratrol are attributed to its activation of SIRT1 (silent information regulator 1) [
24]. SIRT1, the mammalian homologue of yeast Sir2, deacetylates several targets in mammalian cells, acting as a key regulator of energy homeostasis, gene silencing, metabolism, genomic stability, and cell survival [
24‐
26]. In ovary, SIRT1 is expressed in oocytes and nuclei of human granulosa cells at various stages of follicular development and known to exhibit suppressive effects on inflammation in various cell types. Besides, it is suggested that it defends oocytes from age-dependent deficits through cellular protection against OS [
27].
Metformin (1,1-dimethylbiguanide hydrochloride), the insulin sensitizer, is an oral antihyperglycemic drug widely used for the treatment of type 2 diabetes [
28] and PCOS [
29]. Metformin can cause reductions in body weight [
30] and restore ovulation [
31]. It is also suggested that metformin reduces androgen synthesis from ovarian theca cells and suppresses ovarian steroidogenesis [
32,
33]. Generally, its mechanism of action is through the activation of the AMP activated kinase (AMPK), which acts as an energy sensor by monitoring the AMP/ATP status of the cell [
34,
35]. Interestingly, many target proteins of AMPK are so-called longevity factors, e.g., SIRT1, p53, and FoxOs, which not only can increase the stress resistance and extend the lifespan of many organisms but also inhibit the inflammatory responses [
36]. Moreover, AMPK activation protects against arachidonic acid+iron-induced OS through the inhibition of mitochondrial impairment and ROS production [
37]. However, the exact mechanism by which metformin improves ovarian function remains unclear.
Although severe adverse effects are rare, up to 30% of patients report gastrointestinal symptoms, including diarrhea, cramps, nausea, and vomiting, which can cause severe discomfort and lead to discontinuation of the drug [
38]. Finding alternative strategies better than or at least comparable with metformin’s efficacy for the treatment of PCOS is a useful approach for managing the adverse events of metformin. Since resveratrol, as an antioxidant and antiinflammatory agent is known to have some beneficial effects on the ovary and with the given information above about the AMPK-SIRT1 pathway [
27,
36,
37], in this study we hypothesise that combined therapy of resveratrol and metformin might have beneficial effects on PCOS via SIRT1 and AMPK activation. To test this hypothesis, in addition to the detection of TNF-α, MDA and serum hormone levels to measure inflammation, OS and hyperandrogenism, respectively, we investigated the number of follicles, the apoptotic index of the follicular cells, SIRT1 and AMPK immunreactivity in the follicles and the ultrastructure of the oocytes and the granulosa-theca cells in an experimental PCOS model. Because PCOS is the most common endocrine-metabolic disorder and causes infertility among women, focusing on the resveratrol and metformin’s potential therapeutic effects on ovary will be useful for public’s health.
Discussion
PCOS has been regarded as a chronic systemic disease, therefore in vitro studies are not enough to explain its pathogenesis mechanism exactly and should be supported by in vivo studies. For this reason we established our study as an in vivo rat modeling. We aimed to investigate the potential therapeutic effects of metformin and resveratrol against the PCOS-induced ovarian damage in terms of biochemical, histological, immunohistochemical, and ultrastructural aspects.
Various treatments such as androgens, estrogens, aromatase inhibitors, antiprogestins, changes in light exposure, and genetic manipulations have been described to establish rodent PCOS models. However, hyperandrogenism is the most consistent PCOS feature and hence most recent studies have focused on using androgens to induce PCOS in rodent models and they have clearly showed that excess androgen can induce many reproductive and metabolic features of human PCOS [
55]. DHEA is considered to be the key agent in androgen biosynthesis and known to effect polycystic changes in the rat ovary [
56]. Studies with DHEA-treated postnatal mice and rats resulted in anovulation, follicular cysts with a thin granulosa cell layer, increased numbers of atretic follicles, increase in fat and stroma in ovaries, altered ovarian steroidogenesis with elevated serum levels of androgens, estrogens, P, and prostaglandin [
55]. In other studies, it was found that DHEA-induced hyperandrogenism increases ovarian lipid peroxidation and decreases catalase activity and glutathione content whereas prepubertal hyperandrogenism increases serum TNF-α levels and consequently increasing lipid peroxidation and impairing ovarian function [
57]. Although the effects of exogenous androgens on ovarian morphology may vary depending on the period and duration of the androgen administration and the amount of hormone administered, we established our PCOS model with DHEA based on the previous studies.
Since hyperandrogenism plays a major role in the pathogenesis of PCOS, we used DHEA to induce PCOS modelling [
55,
56]. To prove the anovulation; we monitored the rats’s estrus cycle and make sure that they lost regular estrous cycle as shown previously [
49]. Increased LH/FSH ratio and AMH levels in PCOS group also supported our successful modelling.
In this study, metformin alone or combined with resveratrol reduced the augmented body and ovary weights in the PCOS group. Metformin’s weight reducing effects were reported previously and associated with its inducing effect on intracellular AMPK. AMPK reduces the activity of acetyl CoA carboxylase enzyme (ACC) and low activity of ACC decreases the biosynthesis of fatty acids and then increases β-oxidation, which can promote weight loss [
58,
59]. Our AMPK immunostaining data support this relation between AMPK and weight reducing effect of metformin. Although resveratrol could not significantly reduce body weight alone, based on SIRT1 immunostaining findings we suggest that its ovarian weight reducing effect might be mediated by SIRT1 activation. SIRT1 is a target which acts through resveratrol and helps in increased oxidative phosphorylation, fatty acid oxidation, reduction in fatty acid synthesis and stimulate lipolysis [
60].
Serum testosterone levels were higher in DHEA-induced PCOS rats but these levels were significantly decreased in all treatment groups compared to PCOS group. This decrease might be a result of anti-proliferative or pro-apoptotic effects of resveratrol on theca-interstitial cells as shown by previous studies [
61,
62], although our TUNEL assay results regarding theca cells are not in accordance with these data. However, our ultrastructural findings related to theca cells exhibited that mitochondrial damage and expanded sER were ameliorated in the resveratrol group. Resveratrol alone completely reversed the testesterone levels to the control levels, whereas testosterone levels in the metformin alone and combined therapy group were higher compared to the control group. Thus, it can be inferred from our results that resveratrol plays a more effective role in reducing testosterone levels. In parallel to our findings, it has been reported that serine-threonine kinase/protein kinase B pathway is a cell-signaling pathway included in ovarian steroidogenesis and its activity was decreased by resveratrol administration [
63]. On the other hand, metformin’s suppressive effects on ovarian steroidogenesis were described previously [
32,
33]. We suggest that resveratrol and metformin may also regulate ovulation by acting on gonadotropin hormones, since LH levels were significantly lower in the treatment groups than in the PCOS group. As a phytoestrogen, resveratrol may regulate HPG axis via affecting gene transcription in estrogen-sensitive tissues [
64].
DHEA-induced PCOS diminished the primordial follicle pool, however, primary follicles were not as much as expected in the PCOS group. Additionally, secondary and atretic follicles were increased while Graafian follicles were decreased in the PCOS group which is attributed to the cessation of folliculogenesis [
11]. Since secondary and atretic follicle numbers were elevated in the PCOS group, we suggest that some of the primary follicles have turned into secondary follicles and some have suffered from atresia. In parallel to the increase in secondary follicle numbers, higher AMH levels were detected in the PCOS group. In a previous study high serum and intrafollicular levels of AMH in PCOS patients has been reported due to increased number of small antral follicles. This excess of AMH plays a role in the characteristic follicular arrest of PCOS via inhibition of aromatase expression and FSH action [
65]. Since estrogen is known to be a survival factor for the maintainance of granulosa cells, inhibition of aromatase results in apoptosis of granulosa cells due to lower conversion of androgen to estrogen [
66]. Therefore, we suggest that AMH induces the follicular atresia via indirectly inhibiting estrogen secretion. This hypothesis was supported by increased number of TUNEL (+) granulosa cells and atretic follicles in the PCOS group in our study. In this study, treatment with resveratrol and/or metformin ameliorated the follicle numbers, which indicates the effect of the treatments on the maintenance of folliculogenesis. Decreasing number of secondary follicles in the treatment groups are consistent with decreasing levels of AMH. We demonstrated that higher plasma AMH levels in PCOS rats were improved by resveratrol, but not metformin. This reducing effect of resveratrol has been demonstrated before by Ergenoglu et al. [
67], however they have administered a different dose of resveratrol to DHT-induced PCOS rats. Resveratrol, acting through its AMH decreasing effect, may activate aromatase expression and eventually support follicullogenesis.
OS is considered as an inducing factor in the pathogenesis of PCOS and circulating markers of oxidative status, such as MDA, superoxide dismutase (SOD), and glutathione peroxidase (GPx) were demonstrated to be altered in patients with PCOS [
2,
68]. On the other hand, antioxidant effects of resveratrol both in ovary and other tissues have been reported previously [
69,
70]. In our study, high levels of MDA, the indicator of lipid peroxidation, observed in the PCOS group were recovered by resveratrol alone and combined therapy, but not by metformin alone which suggests that the improvement of serum MDA levels are due to the antioxidant effect of resveratrol which is attributed to its ROS production reducing properties [
23].
The pathogenesis mechanism and etiology of PCOS are complex and it was suggested that inflammation may also be involved in the development of metabolic aberrations and ovarian dysfunction [
71]. Previous studies have demonstrated high levels of inflammation markers such as TNF-α and IL-6 in patients with PCOS [
72]. Our finding regarding TNF-α is in accordance with the previous data. Resveratrol has been reported as a potent antiinflammatory compound activating AMPK-SIRT1 pathway [
36] and it has been suggested to supress TNF-α-induced IL-8 release from the endometrial stromal cells via SIRT1 pathway in endometriosis [
73]. In parallel with this data, our results showed for the first time that resveratrol treatment decreased plasma and tissue TNF-ɑ levels in rats with DHEA-induced PCOS. We suggest from our findings that resveratrol may reduce inflammation via SIRT1 activation, since increased SIRT1 immunreactivity was detected in the resveratrol and combined treatment groups. Metformin also seems to have anti-inflammatory effects on women with PCOS as evidenced by a significant decrease in circulating levels of C-reactive protein and white blood cells [
74]. Our data also supported this effect of metformin with the finding of reduced TNF-α levels in Metformin-treated rats compared to PCOS group and in addition suggests that this effect may be mediated through AMPK activation since AMPK immunreactivity was higher in the metformin and combined treatment groups compared to the PCOS group. Since resveratrol and metformin are activators of SIRT1 and AMPK, respectively, and since we didn’t set a waiting period after the treatments (sacrification was done just after the treatments), overexpression of SIRT1 and AMPK at the end of the experiment in the treatment groups compared to the control group is acceptable. Based on previous reports regarding the effect of resveratrol and/or metformin on PCOS [
14,
67] and biochemical, morphological and ultrastructural results of this study, we suggest that resveratrol and/or metformin may be used to improve ovarian functions in PCOS. Physiological safety of overexpression of SIRT1 and AMPK may be assessed with a multidisciplinary approach if considered necessary.
In vitro studies on the apoptotic effects of resveratrol on different cell types in the ovary have yielded contradictory results. Specifically, in granulosa cells, resveratrol decreased the activity of caspases 3/7 and slightly increased the cell number in vitro [
75], however in theca-interstitial cells, it increased the activity of caspases 3/7, inhibited DNA synthesis and decreased the number of viable cells in theca-interstitial cell culture [
61]. In our study, in vivo, we observed that resveratrol decreased the apoptotic index of the granulosa cells, but not the theca cells, supporting the previous data.
There is lack of information concerning ultrastructural examination of PCOS. In this study, mitochondrial disruption with degenerated cristae, expanded sER, and high lipid droplet content were observed in the granulosa and theca cells of follicles in the PCOS rats as shown by previous studies [
54,
76]. Since mitochondria are responsible for oxidant-antioxidant balance in the cell, degeneration of this organelle indicates an oxidative stress in our PCOS model, which was supported by the increase in serum MDA levels. Moreover, based on the known relation with apoptotic processes, degeneration of the mitochondria suggests the mitochondria-dependent pathway for the apoptosis displayed by TUNEL findings in the PCOS rats. Impaired sER and large amount of lipid droplets observed in this group might be related to hormonal abnormality. It has been already known that sER is responsible for testosterone synthesis in the cell and as a result of AMH-induced impaired aromatase expression, androgen accumulation may occur in follicular cells in PCOS [
77]. So, this accumulation may disrupt sER structure and cause dilatations in the organelle. Increased number of autophagosomes in the granulosa cells was another ultrastructural observation in the PCOS group in this study. Recent studies demonstrate that autophagy was enhanced in the ovarian tissue of both humans and rats with PCOS. Consistent with this, granulosa cells from PCOS rats showed increases in the autophagy marker [
78]. All ultrastructural findings were ameliorated by the treatments. In addition to inhibitory effect on steroidogenesis [
63], antioxidant capacity of resveratrol has a major role in this improvoment. It has been known that resveratrol effectively scavenges hydroxyls and superoxides and protects against lipid peroxidation in cell membranes and DNA damage caused by ROS [
79]. It has been reported that SIRT1 activation inhibits hyperglycemia-induced apoptosis by reducing OS and mitochondrial dysfunction in human endothelial cells [
80]. As a result of these data, we infer that resveratrol improved mitochondrial ROS production via SIRT1 activation. Ultrastructural findings also supported the potential therapeutic effects of resveratrol and metformin on granulosa and theca cells of follicles in PCOS group.