In muscle or adipose tissue (AT), AKT prompts glucose transporter type 4 (GLUT4) to translocate from the cytoplasm to the cytomembrane, thereby mediating insulin-stimulated glucose uptake. GLUT4 expression in the adipocytes is observed to be reduced in PCOS, and GLUT1 gene expression is not increased in compensation (47). Interestingly, metformin medication is found to have lasting effects in PCOS patients in terms of induction of GLUT4 mRNA expression in AT, even after its withdrawal (48).

High-mobility group box 1 (HMGB1) is a multi-functional small nuclear protein regulating gene transcription (91). It functions as a DAMP for activation of TLR4 signaling, with the help of MD-2, when it is released into the extracellular matrix (92). HMGB1 is required in order for LPS and nucleic acids to manifest their full toxicity (93). Circulating HMGB1 has been found to be increased in PCOS. HMGB1 has been linked to inflammation, impaired insulin sensitivity, and endothelial dysfunction in PCOS patients (94, 95). At the transcriptional level, HMGB1 is also the direct mRNA target of miR-129 (96), while lncRNA ZFAS1 could competitively bind to miR-129 to induce HMGB1 expression. Moreover, the abundance of miR-155 in the FF in PCOS is negatively correlated with HMGB1 concentration (97). The enhanced autophagy induced by increased HMGB1 content may be a contributing factor in IR in the GCs of PCOS patients, embodied by over-expression of ATG7 and a decrease in p62 levels (98). The release of HMGB1 must undergo acetylation and nuclear-to-cytoplasmic translocation processes (99). Thus, increasing the activity of Sirtuin1 (SIRT1), an NAD⁺ dependent deacetylase, can prevent HMGB1 from being released into extracellular space, thereby reducing the inflammatory response (100). Hence, the alleviation of the metabolic disorders seen in PCOS by treatment regimens such as melatonin, resveratrol, and metformin is at least partially mediated by the activation of SIRT1 (77, 101, 102). After treatment with myo-inositol (MYO) in combination with alpha-lipoic acid (ALA), levels of HMGB1 in adolescents with PCOS have been found to return to normal (25, 103). Cryptotanshinone (CRY) also exerts its therapeutic efficacy via HMGB1/TLR4/NF-κB signaling (104).

The inflammatory cytokines L-1β and IL-18 are found to be increased in PCOS patients; these are related to ovulation, and moreover, the structure of IL‐18 is similar to that of the IL‐1 family. Maturation of these two inflammatory cytokines occurs by means of the activation of inflammasomes (105). Inflammasomes are multiple-protein complexes that are assembled and initiate inflammation in response to harmful stimuli. Dysregulation of inflammasomes is closely linked to numerous human diseases, including neurodegenerative diseases, cancers, and vascular diseases. There are two categories of inflammasomes: canonical and non-canonical inflammasomes. Generally speaking, the canonical inflammasomes are comprised of three components: sensor proteins (such as NLRP1, NLRP3, NLRP6, NAIP/NLRC4, AIM2, and PYRIN), the adaptor ASC, and effectors (pro-caspase1). When DAMPs or PAMPs are recognized by the PRRs, ASC and pro-caspase-1 are recruited to the complex. Subsequently, pro-caspase-1 is activated to form caspase-1, and pro–IL-1β, pro–IL-18, and gasdermin D are then cleaved into active forms (106). Among these, NLRP3 and AIM2 inflammasomes have been detected in PCOS patients; their presence may be driven by hyperandrogenism and fatty acids (107–109). Acetate, a histone deacetylase inhibitor, can suppress the activation of the NLRP3 inflammasome and restore the overactive kisspeptin system in rat models (110). MiR-1224-5p may inhibit activation of the NLRP3 inflammasome by targeting FOXO1 (111). Pioglitazone and metformin dual therapy could mitigate the psychological distress of PCOS patients by reducing NLRP3 inflammasome activation (112, 113), while plumbagin has been found to reduce pyroptosis of GCs. The latter inhibits the activity of WTAP, a key regulator of the RNA N6-methylase complex, consequently destabilizing the ASC mRNA and inhibiting activation of the NLRP3 inflammasome (114).

Exosomes are cell-created extracellular vesicles with sizes in the range of 50–150nm, containing various bioactive molecules. They can be transferred from one cell to another, thus enabling signal transduction and intercellular communication. In recent years, researchers have identified the dual regulatory role played by exosomes, which can negatively or positively affect the activation of inflammasomes. The function of an exosome is dependent on which cargos it carries and where it originates (115). The S100‐A9 protein in exosomes from the follicle fluids of PCOS patients may activate the NF‐κB signaling pathway (116), while human umbilical cord mesenchymal-stem-cell-derived exosomes (hUC-MSC-exos) can inhibit NF-κB signaling via stimulation of the phosphorylation and degradation of Ik-B, increasing the expression of anti-inflammatory cytokines IL-10, and are capable of activating M2 macrophage polarization. As a consequence, inflammation in the ovary is ameliorated (117). Exosomal miR-323-3p derived from adipose mesenchymal stem cells (AMSCs) can protect cumulus cells from apoptosis via direct inhibition of the over-expression of PDCD4 (PMID: 31549864). Moreover, it has been reported that the miR-21-5p carried by AMSC-EXOs can migrate to the liver and activate the IRS1/AKT pathway, thereby mitigating hepatic IR (118). To summarize, these findings indicate that exosomes derived from stem cells can inhibit inflammation through a variety of mechanisms, with microRNAs in stem cells serving as the key players (119, 120). Additionally, hsa-miR-1299, hsa-miR-6818-5p, hsa-miR-192-5p, and hsa-miR-145-5p are highly expressed in the blood of PCOS patients; therefore, these exosomal miRNAs might be promising biomarkers for the early diagnosis of PCOS (121, 122).

Accumulating evidence has demonstrated the close relationship between the gut microbiome and PCOS. Various studies have indicated that the composition of the gut microbiota is altered in PCOS patients, with reduced alpha and beta diversity and decreased Lactobacilli, Bifidobacteria, and Prevotellaceae content, while the abundance of Bacteroides vulgatus is increased (123, 124). This disturbance of the intestinal flora disrupts the integrity of the gut barrier, increasing gut permeability to LPS, and the resultant activation of the TLR-4/NF-κB-mediated inflammatory response. Moreover, these changes are also influenced by HA. It has been proposed that prenatal androgen exposure could lead to dysbiosis of the gut microbiota. One study has revealed an inverse correlation between alpha diversity and circulating testosterone levels (125).

Supplementation with α-Linolenic acid extracted from flaxseed oil has exhibited strong potential in mitigating PCOS-related symptoms: it significantly reduces the levels of LPS and inflammatory factors, alters microbiota composition, and increases the abundance of short-chain fatty acids (125). Additionally, it has been indicated that regimens such as the Bu Shen Hua Zhuo formula and the Shaoyao-Gancao decoction relieve the symptoms of rats with PCOS by modifying the gut microbiota to suppress inflammation (126, 127). Furthermore, IL-22 links intestinal bacteria to IR and androgen excess. Serum IL-22 concentrations are significantly lower in PCOS individuals. In terms of mechanism, increased levels of Bacteroides vulgatus decrease the levels of conjugated bile acids. This results in downregulation of intestinal type 3 innate lymphoid cells (ILC3s) to produce IL-22 in a GATA3-dependent manner (128). However, treatment to increase IL-22 inhibits the release of inflammatory cytokines CCL2, CCL20, and IL-1β, ultimately improving ovarian function (124).

Many previous studies have demonstrated the complex interactions between the keap1/Nrf2 and NF-κB pathways in anti-inflammatory and proinflammatory responses (136). First, Nrf2 activation may inhibit the release of certain inflammatory cytokines, cell adhesion molecules, and matrix metalloproteinases (MMPs). As a result, NF-κB activation is suppressed. Additionally, Nrf2 may disrupt the NF-κB nuclear translocation process (137). On the transcriptional level, NF-κB competes with Nrf2 for the binding site of CBP, a co-activator of Nrf2. Additionally, p65 may recruit the co-repressor HDAC3, decreasing the expression of downstream ARE-driven genes. Moreover, Nrf2 may interfere with the assembly of the NLRP3 inflammasome. Therefore, Nrf2 inducers have strong potential for use in anti-inflammatory therapeutic strategies in PCOS (Figure 4).

It has been demonstrated that, as a upstream modulator, AMPK activates NRF2 by promoting Nrf2 phosphorylation at serines 374, 408, and 433 (138). However, the interesting aspect of this is that, when AMPK kinase is inhibited by compound C, the basal level of Nrf2 is unaffected (139). AMPK may also regulate Nrf2 activity via an indirect pathway: it may deactivate GSK3β via PTEN–PI3K–AKT–GSK-3β cascade, thereby promoting Nrf2 ubiquitylation and degradation. Sulforaphane (SFN), considered to be a potent Nrf2 inducer, has extensive presence in certain cruciferous vegetables, including broccoli and kale (140). Numerous studies have proven its anti-aging, anticancer, and antioxidant properties via various mechanisms (141, 142). In the ovary, SFN reduces tumor invasion and angiogenesis through inhibition of the HIF-1 pathway (143). SFN has also been shown to exert a protective effect on GCs in PCOS patients via activation of the AMPK/Nrf2 pathway (144). Additionally, Ji and colleagues found that salidroside may reduce ROS production and apoptosis of GCs in PCOS patients through activation of AMPK/Nrf2 signaling (145).

NOX4 belongs to the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (NOX) family, which is a primary participant in endogenous ROS production (146). One publication has reported the involvement of NOX4 in inducing ROS production and impairing skeletal insulin sensitivity in testosterone-induced models of PCOS (147). Researchers have also demonstrated that NOX and Keap1/Nrf2 play interacting roles in the control of ROS levels. NOX4 deficiency has been shown to ease the symptoms of rats with PCOS (148). Intriguingly, over-expression of NOX4 may in turn activate expression of Nrf2, thus exerting a protective effect on cardiomyocytes under chronic conditions of overload (149). However, similar phenomena have not been observed in the case of lung or kidney injuries (150).

SIRTs are a family of NAD⁺-dependent histone deacetylases that regulate a variety of physiological functions. To date, seven members of the family (SIRT1–SIRT7) have been identified. SIRT1 can deacetylate LKB1, an upstream regulator of AMPK, increasing the cytoplasmic localization of LKB1 and thus promoting AMPK activation. In turn, AMPK enhances SIRT1 activity by boosting cellular NAD+ levels. Loss of SIRT1 leads to a larger population of early-stage follicles and to ovulation (158). A number of reports have demonstrated that increasing SIRT1 activity can alleviate PCOS (159). Additionally, SIRT1 helps with scavenging of advanced glycation end products (AGEs) through upregulation of glyoxalases, a key component of the anti-glycation defense system (160). As a mitochondrial protein, SIRT3 is found to be decreased in PCOS, and this deficiency contributes to the over-production of ROS and decreased mitochondrial membrane potential (161). Proliferator-activated receptor gamma coactivator-1alpha (PGC-1alpha) acts on the murine SIRT3 promoter; the resulting increase in expression of SIRT3 protects the mitochondria from OS damage (162). Interestingly, SIRT3 levels are also elevated in DHT-induced models of PCOS, supporting the concept of adaptive response (160). The increased expression of SIRT1 and SIRT3 induced by low-level oxidative stress may be the result of a compensatory defense, thus favoring autophagy, which is characterized by increased levels of autophagy marker LC3-II and decreased P62 levels in GCs; consequently, the effect is to promote ovarian health. Studies have shown that androgen can induce autophagy in the GCs in PCOS in a dose-dependent pattern; elevated serum homocysteinem levels are also involved (163, 164). Notably, a harsh environment might even lead to degradation of proteasome-mediated SIRTs and inhibition of autophagy.

In theory, restoration of AMPK expression is expected to be a potential therapeutic strategy.

Metformin is a well-known AMPK activator: it phosphorylates AMPK at the T172 site, thus blocking the expression of matrix metalloproteinase-2 (MMP2) and MMP-9, which may be related to abnormal follicular atresia (165, 166). A glucagon-like peptide 1 (GLP-1) receptor agonist can also be prescribed in order to lower glucose levels. Exenatide and liraglutide have both exhibited beneficial effects in PCOS patients when used alone or in combination with metformin, working via the AMPK/SIRT1 pathway (102, 167). Myo-inositol (MYO) plays a role in improving insulin sensitivity by activating AMPK and increasing GLUT-4 levels. Previous clinical trials have demonstrated that MYO treatment improves outcomes for PCOS patients who are receiving treatment with assisted reproductive technology (168, 169). Moreover, calcitriol supplementation elicits a cardioprotective effect by blunting PCOS-related cardiac remodeling (170). Considering the decreased NAD+ levels occurring in the GCs of PCOS patients, nicotinamide (NAM) and its metabolite N1-Methylnicotinamide (MNAM) have been shown to have therapeutic potential in PCOS (171, 172). In terms of the mechanism, low-dose MNAM supplementation triggers transient ROS generation induced by increased aldehyde oxidase 1 (AOX1) activity. In addition, these agents alleviate hyperandrogenism through inhibition of AMPK-activation-mediated expression of CYP17A1 (173).

Quercetin (QUR), a natural phytochemical with a high flavonol content, can help to ameliorate metabolic and endocrine abnormalities in PCOS, such as by improving IR, inflammation, and hyperandrogenism (174). Upon treatment with QUR, activity of AMPK and SIRT1 is increased, reducing fat accumulation by modulating the secretion of adipocytokines (158, 175). Other compounds, including fisetin and baicalin, have both exhibited beneficial effects via targeting of AMPK (176, 177).

miR-520h has been found to directly target IL-6R, thereby inhibiting the development of PCOS (207). Tofacitinib, one type of JAK inhibitor, was initially developed for the treatment of autoimmune diseases, such as rheumatoid arthritis (RA) (208). It is of interest that inhibition of miR-199a-5p reduces apoptosis in GCs of PCOS patients via activation of JAK/STAT3 pathway. However, this protective effect is counteracted by treatment with tofacitinib (209). Moreover, the flavonoids extracted from Nervilia fordii have shown potent effectiveness in reversing PCOS-related traits via regulation of IL-6 through the JAK2/STAT3 pathway (210). A troxerutin regimen is reported to decrease IR through activation of JAK1/STAT3 signaling in beta cells (211); this decreases the abundance of Bifidobacterium and alters the bile acid profile, ultimately boosting serum IL-22 levels (Figure 5).

Compelling evidence has revealed that YAP1/TAZ activity can be modulated by metabolism (229). AMPK is a key energy sensor; it negatively regulates YAP1 activity by activating LATS1/2 through AMOTL1 or by directly phosphorylating YAP1 at Ser 94. Subsequently, YAP1 is sequestered in the cytoplasm and inhibits downstream gene transcription. In concert with prior studies, metformin, a well-known AMPK activator, has been reported to target YAP1 at S127 (230). Findings also indicate that organ or tissue size is precisely controlled and coordinated with energy status. When there is limited energy availability, the Hippo pathway is activated to support basic survival rather than cellular proliferation (231). Under conditions of sufficient nutrients, however, the high level of YAP activity enhances glycolysis and activates mTOR, leading to excessive activation of dormant primordial follicles (154).

Wedge resection and laparoscopic ovarian drilling (LOD) have been successfully used to treat PCOS by normalizing ovulation (229). It has been found that LOD may loosen cortical layers via alteration of ovarian mechanical forces and destruction of hyperplastic stromal tissue through thermal effects (232), thus decreasing androgen synthesis and normalizing the HPO axis. In vitro activation (IVA) with the administration of actin polymerization—μM jasplakinolide (JASP) or sphingosine-1-phosphate (S1P)—could be a novel therapy for PCOS patients with fewer operative complications. This can convert G-actin into F-actin, increasing YAP nuclear translocation and thereby causing the disrupted follicular growth in PCOS patients to be resumed.

In the case of Wnt/β-catenin signaling (the canonical pathway), when the ligands (such as Wnt1, Wnt2, Wnt3, Wnt8a, and Wnt10b) bind to the Frizzled (FZD) receptors and LRP5/6 co-receptors, signaling is activated, leading to the recruitment of axis inhibition protein (AXIN1/2) and Dishevelled (DVL), which can inhibit the β-catenin destruction complex (consisting of AXIN, GSK3β, CK1, and APC) (233). The result is that unphosphorylated β-catenin is accumulated and then translocates to the nucleus, where it interacts with T cell-specific factor (TCF)/the lymphoid enhancer factor (LEF) family and recruits co-activators, including CBP/p300 and Bcl-9, thereby activating the expression of WNT target genes such as c-Myc, cyclin D1, and CDKN1A. In the absence of WNT ligands, the signaling pathway is OFF, and the destruction complex phosphorylates β-catenin, leading to ubiquitin–proteasomal degradation (235). Moreover, some antagonists, such as dickkopf-1 (DKK1), secreted Frizzled-related proteins (SFRPs), and WNT modulator in surface ectoderm (WISE), are extracellular modulators of Wnt signaling, inhibiting signaling at the ligand or receptor levels (236).

The roles of the canonical Wnt signaling pathway during follicle formation and maintenance have been widely studied (237). One study has verified the involvement of Wnt/β-catenin signaling in granulosa cell apoptosis in North Chinese women with PCOS. In this investigation, transcriptional levels of WNT1, WNT3, and WNT4 were found to be higher in the PCOS group, whereas levels of SFRP4 were lower. Moreover, survivin and BMP4 (downstream survival-related factors) and β-catenin were dramatically decreased. The level of Bax, an apoptosis-related protein, was significantly higher in the PCOS group than in the control group (238). However, these findings were inconsistent with research by Bicer. In this study, serum SFRP4 levels were found to be elevated in women with PCOS; more specifically, the higher a woman’s level of SFRP4, the greater her risk of developing PCOS (239). As an inflammatory adipocytokine, SFRP4 is associated with the onset of diabetes, both type 1 and type 2. It can decrease the influx of Ca²⁺ into β pancreatic cells, thereby suppressing the movement of insulin vesicles toward the membrane and impeding insulin release (240). The dynamic changes occurring during culture in vitro and differentiation of GCs could be a reasonable explanation (241). In rat models of PCOS, miR-324-3p inhibits the proliferation of and promotes apoptosis of ovarian granulosa cells by directly targeting WNT2B (242). Similarly, Sanchez and colleagues have also suggested the involvement of the WNT/β-catenin signaling pathway in the luteinized granulosa cell atresia observed in women with endometriosis (243). Another study has investigated the expression profiles of Wnt/β-catenin signaling genes in human oocytes, finding that the expression of WNT1 and GSK3β genes was elevated in PCOS patients, while there was no significant difference between the PCOS group and the control group in terms of APC and β-catenin expression levels. Interestingly, this study did not detect the expression of AXIN2, FZD4, TCF4, WNT5A, WNT3, WNT4, or WNT7A genes in the ovaries of either PCOS patients or healthy women (57). One study has identified the role of FZD3e in FSH-mediated estrogen production (244). Finally, significantly increased FZD3 expression in the cumulus cells of PCOS patients has been found to lead to over-activation of signaling and subsequently more production of stable β-catenin, thus disrupting the recruitment of CYP19A1 promoter (245).

Similar to Wnt/β-catenin signaling, the β-catenin-independent pathway is activated by WNT ligands (such as WNT11 and WNT5A) binding to FZD receptors and ROR1/2 or RYK co-receptors, resulting in activation of downstream effectors (246). One study has observed proinflammatory effects of wnt5a in the GCs of PCOS patients’ ovaries via an interaction between the noncanonical-Ror2 and PI3K/AKT/NF-κB signaling pathways (247). SFRP5 can combine with WNT5a to inhibit the inflammatory process (248). Moreover, inhibition of the NF-κB pathway may disrupt the activation of WNT5A signaling components ROR2 and JNK (249). Similar effects have been thoroughly demonstrated in many studies. For example, WNT-5A/B ligand expression has been found to be increased in the lung fibroblasts of COPD patients, inducing the production of pro-inflammatory cytokines IL-6 and CXCL8 through noncanonical FZD2 signaling (250). The WNT5A ligand is not limited to the noncanonical signaling pathway. Studies have indicated that a switch from canonical to noncanonical WNT signaling can occur. WNT5A can inhibit canonical Wnt signaling activity by competing with the receptor complex, thereby suppressing the secretion of steroid hormones (251, 252).

The Klotho gene is widely recognized to be involved in the anti-aging process in mammals, and under-expression of Klotho is linked to many diseases, including hypertension, diabetes, and chronic kidney disease. The Klotho family includes α-, β-, and γ-Klotho isoforms, which are highly expressed in the kidney but are also found in other tissues such as the liver and AT. Membrane Klotho serves as a co-receptor for many fibroblast growth factors (FGFs) (253). Although exogenous supplementation of Klotho represents a promising treatment for diverse disorders, there is also growing evidence indicating that the over-expression of Klotho and PCOS are closely interrelated. One study measured circulating FGF19, FGF21, and β-Klotho levels in both in PCOS patients and healthy women, finding that β-Klotho levels were significantly higher in PCOS patients; this can be regarded as a strong indicator of PCOS diagnosis (254). In Klotho knock-out mice, the expression level of Wnt1 is increased and the initially inhibited IRS/Akt pathway is restored (255). As mentioned above, these two signaling pathways are both negatively associated with GCs apoptosis. Furthermore, the increased Klotho expression observed in PCOS patients is accompanied by hyperandrogenism. Mechanistically, androgen recruits AR through indirect binding to Klotho promoters at the transcriptional level (256). Meanwhile, other studies have verified that Klotho is expressed in the hypothalamus and the pituitary gland, regulating GH secretion. Of interest, Klotho expression is relatively low in premature ovarian failure (POF), and this upregulates the TGF-β/Smads signaling pathway, which inhibits autophagy. As a consequence, excess ROS disrupts the maturation of oocytes. Hence, it is speculated that the complicated roles played by Klotho are dependent on the signaling pathways in which it participates (257).

Administration of metformin reverses the low levels of SFRP-5 observed in PCOS patients, and this is expected to be a sensitive indicator for PCOS diagnosis (258). It has reported that CangFu Daotan Decoction normalizes aberrant metabolic features in a rat model of PCOS by suppressing the expression of WNT1 and β-Catenin via modification of m6A methylation (259). Many small molecules or antibody medications have also been found to function as Wnt signaling cascade inhibitors, such as vantictumab (OMP-18R5) and BMD4503-2, which are inhibitors of the Wnt receptor complex; however, development of these focuses primarily on tumor therapy (234, 260). It is believed that, in the near future, more therapeutic strategies targeting Wnt pathway signaling will be developed for PCOS.

Angiogenesis in the ovary is critical for follicular growth, ovulation, and regression of the corpus luteum (262). Several abnormalities in ovarian angiogenesis have been identified in women with PCOS, including increased ovarian vascularization and blood flow. The dysregulation of some angiogenic factors, such as vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), transforming growth factor-β (TGFβ), and basic fibroblast growth factor (bFGF), may be partially responsible for ovulation and aberrant cysts in PCOS (263). The VEGF family was the first and is now the most widely studied set of angiogenic factors in the ovaries. Studies have shown that VEGF is positively correlated with ovarian blood flow, and there is an increase in the density of blood vessels in the cortical stroma of the ovaries in PCOS. In contrast, notch ligand DLL4, mainly expressed on the endothelial cells, is a negative regulator of VEGF-mediated microvascular growth and branching via the prevention of excessive branching (264). Using an anti-Dll4 monoclonal antibody in vivo, Fraser and colleagues showed that this resulted in increased luteal angiogenesis and microvascular density; however, these corpora lutea were less functional and regressed earlier (265). Integrated in silico analysis has identified the Notch pathway as being associated with angiogenesis in PCOS, while the PI3K/Akt signaling pathway is the most enriched pathway (266).

Previous studies have indicated that certain miRNAs, such as let-7a, miR-221/miR-222, and miR-92b, are involved in follicular development and follicular atresia. Xu et al. identified miR-483-5p as the regulator of Notch3 expression in human cumulus GC (267). Moreover, LINC00173 has been reported to upregulate JAG1 expression and to be involved in the development of PCOS via downregulation of miR-124-3p (268). Considering the easy access to some miRNAs from blood or fluid samples, it is possible that miRNAs could be used as effective tools for PCOS diagnosis at a very early stage (269).

Research has confirmed that the Notch pathway is also essential for steroid hormone synthesis. Studies have found that, when Notch signaling is inhibited in small preantral follicles, the expression of genes in the steroid biosynthetic pathway is upregulated (270). This may have the ability to regulate GATA4-dependent promoters through the notch downstream effectors HEY1, HEY2, and HEYL. GATA4 is a crucial transcription factor for steroidogenesis (271, 272).

Notch proteins and ligands are abundant in the hippocampus. It has been found that Notch regulates the functions of learning and memory formation. Disruption of Notch signaling has been implicated as a cause of several human diseases, for example, Alzheimer’s disease (AD) (273). The ligand DLL1 has been shown to be upregulated in the brains of AD patients. Liraglutide (LIR) treatment ameliorates cognitive memory impairment in rats with PCOS via downregulation of Notch overexpression, with the mechanism taking the form of increased acetylcholine levels and decreased Aβ aggregation. Additionally, LIR exerts an anti-inflammatory effect through repression of the activation of the NF-κB pathway (274).

TGF‐β has three isoforms (TGF‐β1, β2, and β3), with TGF‐β1 being the most common. Many studies have indicated that disruption of the TGF-β1/Smads signaling pathway plays an important role in tissue fibrosis, including renal fibrosis, cardiac fibrosis, and hepatic fibrosis. The mechanism may be due to the transformation of fibroblasts into myofibroblasts, the excessive synthesis of extracellular matrix (ECM), and the inhibition of ECM degradation caused by activation of the TGF-β1/Smads signaling pathway. Moreover, many downstream proteins of TGF-β, such as tissue inhibitor of metalloproteinases (TIMPs), matrix metalloproteinases (MMPs), α-smooth muscle actin (α-SMA), and connective tissue growth factor (CTGF), contribute to organ fibrosis (276). To date, however, few studies have focused on ovarian fibrosis. Recent studies have linked this condition to many ovarian diseases, including PCOS and POI (282). The common features of ovarian fibrosis in PCOS patients are a thickened tunica albuginea, due to increased collagen deposition, and increased cortical stroma. Therefore, the aberrant theca cells produce more steroid hormones (283). TGF-β1 also inhibits the activity of P450 aromatase, an enzyme that converts androgen to estrogen (240, 281). TGF-β1 gene polymorphisms are associated with the development of PCOS and characteristics of women with PCOS in the Korean and Han Chinese population. A study has confirmed that ER stress, a potential determinant of pro-fibrotic remodeling during tissue fibrosis, is activated in the granulosa cells of the ovary in PCOS via induction of TGF-β1 expression (284). Over-activation of the NLRP3 inflammasome, induced by excess androgen, results in significantly increased levels of fibrotic factors, such as TGF-β, CTGF, α-SMA, β-catenin, and the environment becomes more collagenous, with increased expression of collagen I and collagen IV (107). However, it remains unclear whether ovarian fibrosis is the cause of PCOS or the result.

The Fibrillin 3 gene (allele 8 of D19S884) appears to be a susceptibility gene for PCOS and is associated with impaired glucose homeostasis (276). Fibrillin 3 regulates the bioactivity of TGF-β by binding to the TGF-β superfamily ligands. Hence, some variants of the Fibrillin gene are thought to alter the normal function of TGF-β signaling and contribute to the pathogenesis of PCOS (285). Elevated follistatin levels have been detected in PCOS patients, regardless of weight. As this is an activin-binding protein, increased follistatin levels result in the neutralization of more activin, which is capable of stimulating FSH secretion. Accordingly, this leads to more androgen production by theca cells and disrupts follicular development (286, 287).

As mentioned above, GDF8 also belongs to the TGF-β superfamily and plays crucial roles in the regulation of folliculogenesis, in steroidogenesis, and in luteal function (288). As an intra-ovarian factor, a number of studies have linked dysregulated GDF8 to many types of reproductive disorders, such as ovarian hyperstimulation syndrome (OHSS) and PCOS. GDF-8 expression levels in follicular fluid and in serum tend to be elevated in PCOS patients, although it is noteworthy that this effect is only observed in obese PCOS patients (289). Moreover, elevated GDF8 levels in women with PCOS often function as a predictor of poorer pregnancy outcomes when women are undergoing IVF treatment (290). One study has revealed that GDF8 is the culprit for the abnormal glucose metabolism observed in PCOS as a result of its stimulation of the expression of SERPINE1 through the ALK5-mediated SMAD2/3-SMAD4 signaling pathway. The SERPINE1 gene encodes plasminogen activator inhibitor 1 (PAI-1), leading to hypofibrinolysis, and is closely associated with diabetes and IR (291, 292). Intriguingly, unlike GDF-8, there is no obvious difference between normal and polycystic ovaries in the expression of GDF11, which has a similar molecular structure and function to GDF8 (293).

One of the main characteristics of PCOS patients is HA, and the ovarian theca cells are thought to be one of the major sources of excess androgen in PCOS patients (311). Liu et al. have demonstrated that establishment of a theca cell lineage requires both granulosa cells and oocytes, with involvement through multicellular interactions via GDF9 and Hh signaling (309). Interestingly, HH ligands (DHH and IHH) are expressed in the granulosa cells of primary-to-antral-stage follicles, whereas the downstream effector Gli1 is found in the theca layer (312). Moreover, expression of Srd5a3 and Cyp17a1, which are steroidogenesis-related genes, has been found to be decreased in Ihh single-knockout mice. This implies a role for Ihh in the regulation of steroidogenesis in the ovary, and the same study also proved that Ihh has greater influence on the activation of the Hedgehog signaling pathway (313).

In the absence of Dhh and Ihh, the theca cell layer fails to form and preantral follicles are unable to develop. Li observed higher levels of Ihh and Ptch2 in PCOS patients based on RT-PCR analysis when compared to a control group, indicating over-expression of the Hh signaling pathway in PCOS patients. This is also related to the aberrant follicular growth observed in PCOS patients (314). Yi et al. have demonstrated that overactivation of the Hedgehog signaling pathway in the ovary inhibits ovulation. In their experiment, transgenic mice expressing a dominant active allele of the signal transducer Smoothened (SmoM2) were able to prevent the suppression of its activity by PTCH, thus leading to sustaining activation of Hedgehog signaling. The Amhr2cre/+SmoM2 mice exhibited extremely reduced ovulation. This may be attributable to impaired muscle cells or contractile cells in the theca layer, which are very important for ovulation (315, 316). Another study has demonstrated that the anovulatory follicles of mhr2cre/+SmoM2 mutant mice are caused by denser capillaries in the ovarian cortex and a lack of vascular smooth muscle in the theca cells of such mice (317). Therefore, it is very likely that disrupted SH signaling contributes to the development of anovulation in PCOS (311). High levels of GTPase immunity-associated protein (GIMAP) 7 expression is observed in the GCs of PCOS patients. In contrast, silencing of the GIMAP7 gene protects GCs from apoptosis and OS via upregulation of the SHH signaling pathway. The expression of downstream components SHH and SMO is increased, along with increased levels of the antioxidants SOD and GSH (318).

The Hh signaling pathway inhibitor cyclopamine can decrease apoptosis of ovarian granulosa cells in PCOS (314). Conversely, another study has revealed that administration of cyclopamine partially counteracts the protective effects of GIMAP7 knockout on rats with PCOS with increased ROS levels (318). Further studies focusing on the impact of different Hh ligands on PCOS development are warranted. Moreover, exposure to certain chemicals, such as acetazolamide and itraconazole, as well as maternal smoking, can disrupt HH signaling, thereby interfering with the development of follicles and contributed to infertility to some extent (319). More attention needs to be paid to the avoidance of exposure to these substances.