Decoding androgen excess in polycystic ovary syndrome: Roles of insulin resistance and other key intraovarian and systemic factors

Abstract

Recent studies have potentiated the essential role of androgens in normal folliculogenesis and, therefore, female fertility. Contrastingly, excess androgen levels, i.e., hyperandrogenism (HA), a hallmark characteristic of polycystic ovary syndrome, overrides the delicate balance of folliculogenesis, leading to follicular arrest and ovulatory issues. Insulin resistance (IR) has a profound effect on elevating androgen secretion and is considered one of the primary factors driving both ovarian androgen production and metabolic dysfunction in polycystic ovary syndrome. Together with IR, disruptions in key intraovarian and systemic factors, including activin, inhibin, follistatin, anti-Mullerian hormone, bone morphogenetic proteins, growth differentiation factor-9 and Kit ligand, as well as dysregulation in both the insulin and the transforming growth factor-β superfamily signaling pathway, contribute to follicular arrest, elevated androgen levels and metabolic dysfunction, exacerbating HA. Additionally, suppression of sex hormone-binding globulin, disrupted adipose-neuroendocrine signaling and altered microRNA expression heighten HA, with IR serving as the fundamental contributor. Emerging evidence implicates impaired atresia together with non-apoptotic cell death, such as ferroptosis and pyroptosis, which have also been associated with ovarian dysfunction. A comprehensive understanding of the most significant factors, particularly IR, which amplifies androgen production through hyperinsulinemia-mediated stimulation of theca cells, is essential for identifying targeted therapeutic strategies.

Key Words: Polycystic ovary syndrome; Hyperandrogenism; Insulin resistance; Oxidative stress; Obesity; Folliculogenesis; Ovary

Core Tip: Polycystic ovary syndrome is a multifactorial disorder characterized by excess ovarian androgen production, frequently amplified by insulin resistance. This review explores how insulin resistance intersects with systemic and intraovarian factors such as activin, inhibin, follistatin, anti-Mullerian hormone, bone morphogenetic proteins, growth differentiation factor-9 and Kit ligand, as well as the dysregulation in both the insulin and the transforming growth factor-β superfamily signaling pathway, to create a vicious cycle of reproductive and metabolic dysfunction. A deeper understanding of these factors and pathways emphasizes the need for individualized, multi-targeted therapeutic approaches beyond conventional treatments.

INTRODUCTION

Polycystic ovary syndrome (PCOS) is a common endocrine disorder that affects women of reproductive age[1]. The key pathophysiological features of PCOS include hyperandrogenism (HA), chronic anovulation (oligomenorrhea or amenorrhea), and polycystic ovarian morphology (PCOM). While traditionally viewed as a multifactorial condition, insulin resistance (IR) plays a central role as one of the primary factors in the development of PCOS[2]. Elevated insulin levels due to IR drive increased ovarian androgen production, which contributes to the hallmark symptoms of PCOS, including hirsutism, acne, and irregular menstrual cycles[3]. Additionally, PCOS is commonly accompanied by subclinical inflammation, metabolic abnormalities, oxidative stress, and obesity[2,4].

The role of androgen in the process of ovulation is perplexing at best. On the one hand, androgen plays an imperative role in the process of normal folliculogenesis, and on the other hand, excess androgen is central to the pathophysiology of PCOS. Under normal conditions, folliculogenesis is tightly regulated by the hypothalamic-pituitary-ovarian (HPO) axis, where gonadotropin-releasing hormone (GnRH) pulsatility controls the production of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from the anterior pituitary gland[5]. Rapid GnRH pulse frequency favors LH secretion, while slower pulses enhance FSH secretion[5]. A balanced LH/FSH ratio is critical for coordinated follicular development, i.e., FSH promotes granulosa cell proliferation, estradiol synthesis, and follicle maturation, while LH stimulates theca cells to produce androgens, which serve as substrates for aromatization[6]. At normal physiological levels, androgens enhance FSH receptor (FSHR) expression in granulosa cells, supporting follicle growth and selection[7].

In PCOS, this balance is dysregulated. Driven by HA, increased GnRH pulse frequency elevates LH secretion over FSH, contributing to the characteristic elevated LH/FSH ratio[8]. The excess LH promotes overproduction of androgens by the theca cells, while the decreased FSH levels impair aromatization of these androgens to estrogens in granulosa cells[9]. This hormonal imbalance disrupts normal folliculogenesis, leading to arrested follicle development and anovulation. The premature and heightened expression of androgen receptors (ARs) in the ovarian follicles maybe a significant factor that enables androgen excess to alter GnRH pulsatility, which disrupts the normal regulation of LH and FSH secretion, further promoting the androgenic states implicated in PCOS[10].

Concurrently, IR also affects the neuroendocrine axis[8]. Elevated insulin enhances the release of kisspeptin, a neuropeptide that stimulates GnRH-LH hypersecretion, which in turn, amplifies androgen synthesis, worsening HA[11]. Additionally, IR decreases the synthesis of hepatic sex hormone-binding globulin (SHBG), increasing free androgen levels[9,12]. Thus, this heralds a vicious cycle where both metabolic and reproductive issues persist, progressively worsening. Furthermore, members of the transforming growth factor-β (TGF-β) superfamily, including activin, inhibin, follistatin, anti-Mullerian hormone (AMH), bone morphogenetic proteins (BMPs), growth differentiation factor-9 (GDF-9) and Kit ligand (KL), play crucial roles in granulosa cell differentiation, oocyte maturation, and follicular selection[6]. Their dysregulation, together with the impairment of the insulin and the TGF-β superfamily signaling pathways, primarily influenced by excessive insulin, may impair these intra-ovarian regulatory mechanisms, perpetuating follicular arrest and contributing to the hyperandrogenic state of PCOS. Accordingly, this review aimed to shed light on these mechanisms to improve our understanding of the underlying pathophysiology and steer the advancement of targeted diagnostic and therapeutic strategies.

PREVALENCE OF PCOS

Recent epidemiological studies show the growing global burden in the prevalence of PCOS[13]. A 2025 systematic review involving over 561000 women reported global PCOS prevalence estimates ranging from approximately 6.6% to 10.9%, depending on the diagnostic criteria used[14]. Notably, there was a significant lack of data from the African region, highlighting the need for more research in these areas[14]. Complementing this, a comparative study showed that the global age-standardized point prevalence and annual incidence rates for PCOS have increased by 30.4% and 29.5% since 1990, respectively, with sub-Saharan Africa having lower-than-expected burdens across the measurement period. However, it has been noted that 68%-75% of patients with PCOS remain undiagnosed even after visiting many medical institutions, indicating that the incidence of this condition is probably underestimated[15,16]. Additionally, the lower detection rates may result from poor healthcare infrastructure. For example, the limited availability of ultrasound imaging equipment in Africa[17]. Furthermore, the diverse clinical manifestations, unknown etiology, and complicated pathophysiology have amplified the poor diagnosis of PCOS[18].

Furthermore, an analysis of Global Burden of Disease data from 1990 to 2021 showed that the number of women with PCOS nearly doubled globally, from 34.81 million to 65.77 million, with corresponding increases in incidence and disability-adjusted life years[19]. Of concern is the significant gap in public awareness and understanding of PCOS. This was reflected in a 2023 cross-sectional study in the United Arab Emirates, which found that 84.3% of women were familiar with the term PCOS, but only 21.7% demonstrated sufficient awareness of its causes, symptoms and complications[20]. Similarly, in Malaysia, a study conducted with women in the Klang Valley region indicated that nearly half of the respondents had poor knowledge and health-related practices concerning PCOS[21]. These findings suggest the need for targeted educational initiatives to enhance PCOS awareness and promote early diagnosis and management.

Additionally, as the symptoms of the syndrome worsen and intensify, it sets in motion a significant deterioration in quality of life, an increase in stress, negatively affecting the psychological and emotional well-being of women with PCOS[22,23]. The implications of this have been exemplified by Yadav et al[24] by indicating that the healthcare-related economic burden of PCOS exceeds 15 billion dollars yearly in the United States, considering the costs of PCOS diagnosis, costs related to PCOS-associated mental health, reproductive, vascular, and metabolic disorders. Given that PCOS shows a global prevalence, the economic burden it imposes is substantial, highlighting the need for greater attention from both scientific and policy-making sectors[24].

DIAGNOSIS OF PCOS

The heterogenous nature of the syndrome has produced considerable scientific debate and as yet, no conclusive clinical definition of PCOS has been reached[25]. Most commonly used are the Rotterdam diagnostic criteria[26], which defines PCOS when at least two of the following three features are present: Oligo or anovulation, clinical and/or biochemical signs of HA, and polycystic ovaries[27,28]. To have a consensus, the International Guidelines for the Assessment and Management of PCOS[29] integrates the criteria from the different scientific bodies, i.e., National Institute of Health 1990, Rotterdam 2003, and Androgen Excess and PCOS Society[30]. Thus, the consolidated definition has established four clinical phenotypes of PCOS, with different metabolic repercussions for each criterion (Table 1)[31]. Patients diagnosed with fullblown PCOS (phenotype A) are at higher risk of adverse metabolic and cardiovascular outcomes, while phenotype D is the least severe phenotype[32]. The correct diagnosis and phenotyping facilitates a more effective treatment strategy for the cardiometabolic risk of patients[31].

Table 1 Phenotypes for the classification and diagnosis of polycystic ovary syndrome.

Phenotype	Synonym	Diagnostic criteria
Phenotype A	Classic phenotype (full-blown PCOS)	Hyperandrogenism + oligo-ovulation + polycystic ovaries
Phenotype B	Classic phenotype (non-polycystic ovaries)	Hyperandrogenism + oligo-ovulation
Phenotype C	Ovulatory phenotype	Hyperandrogenism + polycystic ovaries
Phenotype D	Non-hyperandrogenic phenotype	Oligo-ovulation + polycystic ovaries

PCOS: Polycystic ovary syndrome.

Emerging advancements in molecular and omics-based technologies have identified several novel biomarkers that may enhance diagnosis. For instance, circular RNAs (circRNAs), a class of non-coding RNAs, have been found to be differentially expressed in PCOS patients and may influence processes like cell proliferation, apoptosis, and steroidogenesis[33]. Growing evidence indicates that circRNAs are involved in insulin secretion and pancreatic β-cell function and aberrations in circRNAs may reflect disruptions in insulin-related signaling pathways[34]. The strong association between circRNAs and IR may provide therapeutic targets for managing IR in PCOS[33]. Additionally, circRNAs are present in the follicular fluid of women with PCOS, indicating their potential as diagnostic biomarkers[33]. Similarly, lipidomic and proteomic studies of follicular fluid have identified novel metabolic signatures in women with PCOS, for example, Wnt1-inducible signaling pathway protein 1 as a downstream effector of Wnt/β-catenin signaling pathway[35]. Wnt1-inducible signaling pathway protein 1 may function in the interaction between IR, inflammation, and obesity and can be a therapeutic target for PCOS patients[35].

REMEDIES AND COMMON PHARMACEUTICAL TREATMENTS

Lifestyle and dietary modifications

Lifestyle and dietary modifications that include a balanced diet and exercise are the cornerstone of the initial treatment of PCOS before pharmacological intervention[36]. This holistic approach plays a key role in alleviating symptoms associated with the metabolic profile of the patient[37]. Unfortunately, lifestyle interventions are mostly ineffective due to the lack of compliance. Furthermore, the heterogeneous nature of the pathophysiology of PCOS makes it difficult to manage. Thus, various pharmacological treatment modalities are available to address the intricacies associated with PCOS. Given the different phenotypes of PCOS, the management of PCOS centers around these areas: HA, metabolic aberrations, and anovulation[37].

Pharmacological drugs

Pharmacological remedies currently used to manage PCOS are presented in Table 2[2,12,37-48]. The table compares these pharmacological drugs based on clinical benefits, mechanisms, adverse effects, and safety considerations. Additionally, comparative information, such as between letrozole and clomiphene citrate, is summarized in the table and accompanying Figure 1. The table provides a practical summary of treatment options and may serve as a patient-specific guide.

Figure 1 Illustration depicting the mode of action of clomiphene citrate in comparison to letrozole in ovulation induction. Clomiphene citrate acts as a non-steroidal selective estrogen receptor modulator, i.e., it binds to estrogen receptors in the hypothalamus, thereby inhibiting the negative estrogen feedback to the hypothalamus. In turn, this leads to a change in the pulsatile release of gonadotropin-releasing hormone that drives an increase in the secretion of follicle-stimulating hormone, promoting follicle development. Letrozole, an aromatase inhibitor, reduces estrogen production by blocking the conversion of androgens to estrogens, i.e., letrozole inhibits the cytochrome P450 aromatase enzyme complex. This creates negative feedback to the hypothalamus, stimulating the pituitary gland to produce more follicle-stimulating hormone, leading to folliculogenesis. Letrozole has no adverse effects on the endometrium and cervical mucus and is preferred to clomiphene citrate to induce ovulation. LH: Luteinizing hormone; FSH: Follicle-stimulating hormone.

Table 2 Pharmacological interventions for polycystic ovary syndrome: Efficacy, side effects and safety.

Drug class and treatment	Clinical benefits	Mechanism of action	Adverse effects	Safety considerations	Ref.
Hormonal regulators (combined oral contraceptives): Ethinylestradiol + drospirenone	Menstrual regulation, androgen suppression, reduces hirsutism and acne	Inhibits ovarian androgen production by suppressing gonadotropin, primarily, LH	Nausea, bloating, and mood changes	Weight gain, alterations in cardiometabolic parameters and venous thromboembolism risk	[38,39]
Insulin sensitizers (biguanides): Metformin	Improves insulin sensitivity, menstrual regularity, and ovulation	Inhibits hepatic gluconeogenesis and improves peripheral glucose uptake	GI upset	Caution with persistent use, increases homocysteine levels and increase risk of CVD	[12,37,40]
Insulin sensitizers (thiazolidinediones): Pioglitazone	Improves IR, ovulation, and excess androgen regulation	Decreases hepatic and peripheral IR through the activation of the nuclear hormone receptor (PPARα)	Weight gain and peripheral edema	Liver toxicity, not a first-line treatment option due to cardiovascular concerns	[37,41]
GLP-1 agonists: Liraglutide	Improves metabolic profile, including weight loss	Binds to insulin receptors on beta cells, stimulates insulin secretion, reduces glucagon secretion, inhibits hunger centers and delays gastric emptying	Nausea and emesis	Avoid during pregnancy; effective contraception is required and a washout period prior to pregnancy	[2,42,43]
SGLT2 inhibitors: Empagliflozin	Weight loss, improves insulin sensitivity, and reduces androgen levels	Renal glucose reabsorption is inhibited, promoting glycosuria	Increased urination, urinary tract infections, and dehydration	Monitor hydration and renal function; limited long-term data in PCOS	[44,45]
Anti-androgens: Spironolactone	Reduces hirsutism and acne	Reduces androgen production by blocking AR	May cause menstrual irregularity	Teratogenic; requires effective contraception; potassium monitoring advised due to hyperkalemia risk	[46]
Ovulation inducers: Clomiphene citrate	Ovulation induction	Selective estrogen receptor modulator	Mood swings, hot flashes	Risk of multiple pregnancies, endometrial thinning and thickening of cervical mucus	[47]
Ovulation inducers: Letrozole	Enhances ovulation and pregnancy rates	Aromatase inhibitor, reduces estrogen feedback to increase FSH	Fatigue, dizziness, hot flashes	Possibly teratogenic effects	[48]

LH: Luteinizing hormone; GI: Gastrointestinal; CVD: Cardiovascular disease; IR: Insulin resistance; PPARα: Peroxisome proliferator-activated receptor alpha; GLP-1: Glucagon-like peptide-1; SGLT2: Sodium-glucose cotransporter 2; PCOS: Polycystic ovary syndrome; AR: Androgen receptor; FSH: Follicle-stimulating hormone.

Emerging clinical evidence continues to enhance the therapeutic framework of PCOS. A meta-analysis by Zeng et al[49] demonstrated that combining metformin with spironolactone resulted in greater improvements in body mass index (BMI), fasting glucose and androgen levels than metformin alone. A meta-analysis of randomized controlled trials showed the efficacy and safety of glucagon-like peptide-1 agonists in PCOS women living with obesity by promoting weight loss and hormonal regulation by improving BMI, waist circumference, and androgen levels[50]. Interestingly, a recent network meta-analysis of non-pharmacological strategies, including electroacupuncture, diet, and exercise, yielded significant improvements in androgen-related symptoms[51]. Vitamin D supplementation was also associated with improved endometrial thickness, inflammatory markers and androgen profiles[52]. These findings endorse the growing trend of integrative and metabolic-targeted interventions in PCOS management.

ROLES OF HORMONES IN FOLLICULOGENESIS

The binding of LH to respective LH receptors on theca interna cells leads to an elevation in intracellular 3’,5’-cyclic adenosine monophosphate levels and subsequent activation of protein kinase A[53]. This elevation, in turn, promotes the activity of steroidogenesis by increasing the expression of steroidogenic acute regulatory protein (StAR) (Figure 2). Thus, the two-cell, two-gonadotropin model compartmentalizes steroidogenesis, where two types of cells (granulosa cells and theca cells) together with two gonadotropins (LH and FSH) function together in estrogen synthesis[6]. In this process, LH-influenced theca cells produce androgens that are aromatized to estrogen in the granulosa cells by the cytochrome P450 19A1 (CYP19A1) aromatase enzyme via the FSH signaling pathway (Figure 2)[54].

Figure 2 Steroid biosynthesis in the ovary shows the two-cell, two-gonadotropin model, where androgens synthesized in the theca cells are aromatized to estrogens in the granulosa cells, under normal physiological conditions. In the initial synthesis of androgen, under the control of luteinizing hormone, the side chain of cholesterol is cleaved by the P450 side chain cleavage enzyme, cytochrome P450 17A1, to pregnenolone, a precursor to all steroid hormones. Testosterone is a precursor converted to dihydrotestosterone or estradiol in target tissue by 5α-reductase and the aromatase enzyme, respectively. LH: Luteinizing hormone; LHCGR: Luteinizing hormone/chorionic gonadotropin receptor; cAMP: Cyclic adenosine monophosphate; PKA: Protein kinase A; StAR: Steroidogenic acute regulatory protein; CYP17A1: Cytochrome P450 17A1; 3β-HSD: 3β-hydroxysteroid dehydrogenase; 17β-HSD: 17β-hydroxysteroid dehydrogenase; CYP19A1: Cytochrome P450 19A1; FSHR: Follicle-stimulating hormone receptor; FSH: Follicle-stimulating hormone; DHT: 5α-dihydrotestosterone.

ROLE OF ANDROGENS IN NORMAL FOLLICULOGENESIS

Androgens play critical roles in female reproductive physiology, especially in the regulation of normal folliculogenesis. Their actions are mediated primarily through ARs expressed in ovarian theca and granulosa cells.

Primordial to primary follicle transition

During the early phase of folliculogenesis, androgens appear to support the activation of dormant primordial follicles[6]. Animal studies suggest that, despite the absence of AR expression in the primordial follicles, androgens may enhance the initial recruitment of follicles by upregulating intraovarian growth factors such as insulin-like growth factor 1 (IGF-1) and KL, promoting the transition from primordial to primary follicles (Figure 3)[55,56]. Notably, the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (Akt) pathway is one of the intracellular signal transduction pathways involved in primordial follicle activation in the ovary. Phosphatase and tensin homolog (PTEN), a key negative regulator of the PI3K-Akt pathway, acts to preserve the dormancy of primordial follicles by inhibiting premature activation[57,58]. Conversely, androgens may enhance PI3K-Akt signaling indirectly via stimulation of local growth factors, thereby promoting follicle activation[6]. This suggests a functional opposition: While PTEN maintains dormancy, androgens will encourage the balance toward activation. Disruption in this regulatory balance, especially excessive androgens, may contribute to early follicle recruitment with implications for disorders such as PCOS[59].

Figure 3 Different stages of normal folliculogenesis from the primordial follicle to the dominant (Graafian) follicle and ovulation. Gonadotropin-releasing hormone pulsatility controls the production of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from the anterior pituitary gland. Rapid gonadotropin-releasing hormone pulse frequency favors LH secretion, while slower pulses enhance FSH secretion. A balanced LH/FSH ratio is critical for coordinated follicular development, i.e., FSH promotes granulosa cell proliferation, estradiol synthesis, and follicle maturation, while LH stimulates theca cells to produce androgens, which serve as substrates for aromatization. Androgens begin to exert effects from the primordial stage by supporting the initial activation of dormant primordial follicles. Androgens exert their most prominent effects during the preantral and early antral stages, enhancing granulosa cell proliferation, FSH receptor expression and aromatase activity for estrogen biosynthesis. In the antral and dominant (Graafian) stages, androgens support follicle maturation and upregulate LH receptor expression. While androgens do not directly mediate ovulation, their prior actions facilitate the follicle’s response to the LH surge. Follicular atresia, which eliminates non-dominant follicles, is indirectly influenced by androgen. HPO: Hypothalamic-pituitary-ovarian; GnRH: Gonadotropin-releasing hormone; LH: Luteinizing hormone; FSH: Follicle-stimulating hormone; A: Atresia.

Preantral and early antral follicle growth

Androgens exert their most prominent effects during the preantral and early antral stages. Here, they act synergistically with FSH to promote granulosa cell mitosis, induction of FSHRs and stimulate estradiol production via increased aromatase activity[7]. This androgen-mediated sensitization to FSH is critical for promoting healthy follicle growth and selection as well as making the follicle more resistant to atresia[7,60].

Dominant follicle selection and maturation

In later stages, appropriate androgen signaling has been implicated in regulating the final stages of follicle maturation and ovulation[60]. However, androgen levels must be tightly controlled, i.e., while physiological concentrations are beneficial, excessive androgens, as seen in PCOS, disrupt the follicular environment, inhibit aromatase expression and impair granulosa cell function[10]. This imbalance contributes to follicular arrest, anovulation, and the development of PCOM[61].

Ovulation and luteinization

Although androgens are less dominant during ovulation, their earlier effects set the stage for successful follicular rupture and corpus luteum formation[62]. Indirectly, by supporting FSH-driven estradiol production and granulosa cell differentiation, androgens facilitate the LH surge necessary for ovulation. Proper androgen balance may also contribute to luteal phase adequacy, although further research is needed in this area[10].

Androgens and follicular atresia

Follicular atresia, the degenerative process that eliminates non-dominant follicles, is indirectly influenced by androgen signaling[63]. Physiological androgen levels can exert anti-apoptotic effects by increasing FSHRs in preantral and early antral follicles, enhancing follicular survival during early development[7]. However, excessive androgen exposure activates pro-apoptotic signaling in granulosa cells of larger follicles, thereby accelerating atresia[7]. In PCOM ovaries contain an abnormally large population of small antral follicles, but it is not static, as the follicles show high rates of atresia, with constant replacement from newer developing follicle cohorts[64,65]. Thus, in normal folliculogenesis, androgens are essential in early follicle activation, granulosa cell proliferation, and FSH responsiveness[7]. Androgens either support survival at critical periods of folliculogenesis or, in excess, induce degeneration[60]. This duality extends to the role of androgen in follicular survival vs atresia and demonstrates the importance of androgen balance for reproductive health. Disruption of this homeostasis, as in PCOS, leads to aberrant follicle development and anovulation.

ANDROGEN REGULATION THROUGH RECEPTOR BINDING

The biological activity of androgens occurs predominantly through the binding and activation of their AR, a member of the nuclear receptor superfamily that functions as a ligand-activated transcription factor[66]. Within ovarian follicles, AR is localized in the granulosa cells, stromal cells and oocytes, with expression highest in the granulosa cells[60]. It has been shown that the inactivation of AR results in premature ovarian failure, alluding to the crucial role of AR-mediated androgen action in normal folliculogenesis[67]. In studies using granulosa cell-specific AR knockout mice, it was revealed that selective loss of granulosa cell AR actions during preantral and antral developmental stages resulted in subfertility, defective follicle dynamics, altered ovulation rates, follicle depletion, increased follicle atresia, and reduced embryo viability[68]. Thus, the granulosa cells of preantral and antral follicles are essential sites for AR-mediated actions involved in maintaining follicle and embryo survival and, ultimately, optimal female fertility[68]. Contrary to previous assumptions, which have implicated androgens to inhibit antral follicle development, recent findings reveal that androgens have a growth-promoting effect on the early stages of follicle development[6]. This highlights the complex and intricate nature of androgens and their impact on folliculogenesis.

Several studies have shown that expression of AR decreases as the follicle develops[60], functioning predominantly as a substrate of estrogen synthesis. Thus, the decrease in AR expression coincides with rising granulosa cells CYP aromatase expression, allowing androgens produced by LH-stimulated theca cells to undergo FSH-stimulated aromatization to estrogen (Figure 2)[69]. Additionally, it has been postulated that androgens indirectly modulate P450 aromatase activity by increasing FSHR expression and FSH activity[70]. Likewise, Fujibe et al[71] demonstrated in their investigations that androgens could support follicle development during the FSH-dependent preantral stage by enhancing FSH action via increased expression of FSHR mRNA levels in prenatal follicles of mice. These findings indicate the indispensable role of androgen in early to preantral follicle development.

FACTORS THAT AFFECT ANDROGEN PRODUCTION IN THE OVARY

HA is often exacerbated by IR, a common but variable metabolic feature linked to PCOS[12]. The co-occurrence of HA and IR creates a pathological loop that worsens ovarian dysfunction, metabolic imbalance, and systemic inflammation[2]. These disturbances are mediated by the dysregulation of several intraovarian factors, molecular pathways, and hormonal systems, many of which are intricately involved in follicular growth, steroidogenesis, apoptosis, and cellular signaling. The sections below explore IR, key intraovarian and systemic factors, and their role in hyperandrogenic PCOS states by highlighting their regular physiological roles and altered states in PCOS.

IR and the influence of members of the TGF-β superfamily in hyperandrogenic PCOS

Interplay between IR, activin, and inhibin in PCOS: Under normal physiological conditions, two glycoproteins, activin and inhibin, members of the TGF-β superfamily, modulate theca cell steroidogenesis in the ovary[72]. Several studies have confirmed that the biological effects of activin on theca cell androgen production are inhibitory. In contrast, inhibin acts in an opposing manner by causing an increase in the production of androgens[73]. More specifically, in the granulosa cells, activin appears to elevate the levels of CYP aromatase, the FSHR, and estradiol. Additionally, it has been shown that deleting the activin receptors leads to FSH deficiency, with female mice becoming acyclic and sterile[74]. Thus, favoring the hypothesis that activin is an FSH-stimulating agent[75]. Contrastingly, in the theca cells, activin seems to decrease the levels of StAR, which is responsible for transporting cholesterol esters into the mitochondria to initiate steroidogenesis, thereby downregulating androgen production[72,73].

Inversely, inhibin, which is produced by the granulosa cells, modulates activin bioactivity[72]. Inhibin complexed with betaglycan (an inhibin co-receptor) functions to antagonize activin activity by competitively binding to activin receptors, thereby blocking the effects of activin (Figure 4). Another glycoprotein, follistatin, although not classified as a member of the TGF-β superfamily, interacts and regulates the activity of activin (Figure 4). Ultimately, follistatin and inhibin downregulate activin activity[75]. The activin-follistatin-inhibin system is considered a key regulator of the HPO axis[76].

Figure 4 Mechanism diagram of activin, transforming growth factor-β signaling regulation, and suppressor of mothers against decapentaplegic pathway activation. A: The modulation of activin signaling by inhibin and follistatin. Activin promotes follicle-stimulating hormone synthesis and ovarian function through suppressor of mothers against decapentaplegic signaling pathways. Inhibin acts as an antagonist by binding to activin receptors via its co-receptor betaglycan, preventing activin from signaling. Follistatin neutralizes activin by directly binding to it, blocking its interaction with receptors; B: Transforming growth factor-β (TGF-β) is produced in a latent (inactive) form, bound to latent TGF-β binding proteins anchored in the extracellular matrix. When needed, TGF-β is activated and released from the extracellular matrix to bind its receptors on the cell surface, initiating signaling pathways. This system ensures that TGF-β acts only when required. TGF-β: Transforming growth factor-β; SMAD: Suppressor of mothers against decapentaplegic.

In PCOS, this balance is disrupted, particularly due to the effects of IR[77-79]. Chronic hyperinsulinemia, central to IR, not only stimulates ovarian theca cells directly to increase androgen production but also upregulates follistatin levels, thereby suppressing activin’s beneficial role in FSH support and androgen suppression. Interestingly, Sylow et al[80] showed in their study that elevated circulating follistatin, but not activin A, was strongly associated with IR in patients with type 2 diabetes (T2D). In the mouse model, follistatin was found to mediate diabetes by fostering IR specifically in white adipose tissue[81].

Similarly, an impaired activin A and inhibin B ratio is associated with PCOS, which may further dampen FSH signaling and contribute to arrested follicular development[72]. Furthermore, Walton et al[82] generated a mouse model with inhibin loss of function and their analyses revealed that complete inactivation of inhibin resulted in an elevation in circulating FSH, which triggered ovarian overstimulation and pregnancy loss. However, a partial reduction in inhibin levels enhanced female fecundity. Their conclusions were further supported by Brûlé et al[83]. These studies garnered confirmatory evidence regarding the delicate balance of the activin, inhibin, and follistatin necessary for the physiology of the female reproductive system, and an imbalance in their signaling can trigger pathologies such as PCOS[72].

Crosstalk between IR and AMH in PCOS: In typical physiological conditions, the glycoprotein AMH is secreted by the granulosa cells from the primary follicle stage and peaks in the pre-antral stage and early antral stage in human follicles[84]. The secretion starts to decrease as the follicle becomes FSH-dependent. In folliculogenesis, AMH has two windows of action. Firstly, AMH exerts an inhibitory effect on the initial recruitment of resting primordial follicles[85], possibly as an adaptive measure to prevent excessive recruitment, thereby maintaining homeostasis in the ovary[86]. Secondly, AMH negatively affects preantral and small antral follicle growth by attenuating their responsiveness to FSH[87]. In human studies, AMH was shown to reduce FSH-stimulated aromatase expression in granulosa cells and also decrease FSHR mRNA expression[88].

As an inhibitor of aromatase activity, AMH may drive an elevated androgenized state, leading to HA, oligo/anovulation, and infertility that are symptomatic of patients with PCOS[87]. Elevated serum AMH expression may play a central role in the follicular arrest characteristically observed in anovulatory PCOS (Figure 5)[89]. Recent studies have indicated a complex and, to some extent, inconsistent relationship between AMH levels and IR in women with PCOS. Some studies report a positive correlation between AMH and IR markers. For instance, a 2023 retrospective cohort study in Chinese women with PCOS found that serum AMH levels were significantly higher in those with IR, with positive correlations noted between AMH, homeostasis model assessment of IR, fasting insulin, androgens, and LH/FSH ratio, suggesting that elevated AMH may be linked to increased IR in this population[90]. Similarly, a study in adolescents with PCOS identified AMH as an independent determinant of IR, indicating that AMH may play a role in the early development of metabolic dysfunction in this group[91].

Figure 5 Intraovarian and systemic factors play a role in disrupting folliculogenesis in polycystic ovary syndrome, resulting in altered steroidogenesis. Heightened luteinizing hormone, insulin resistance, and obesity stimulate theca cell androgen production, while granulosa cell dysfunction impairs aromatization of androgens to estrogens. This hormonal imbalance leads to irregular or absent selection of a dominant follicle, resulting in the persistence of small antral follicles, leading to the characteristic polycystic ovarian morphology. Hyperandrogenism further amplifies the arrest of follicular development, resulting in a vicious cycle of anovulation and hormonal imbalance. LH: Luteinizing hormone; FSH: Follicle-stimulating hormone; PCOM: Polycystic ovarian morphology.

However, a systematic review and meta-analysis of 22 studies concluded that the overall association between AMH and IR in PCOS is weak and statistically insignificant (pooled correlation coefficient = 0.089), highlighting considerable heterogeneity across studies[92]. This variability may be attributed to several confounding factors, such as different PCOS phenotypes, different BMI classifications, variation in environmental factors and genetics across regions, and various age groups[92]. Together, these factors demonstrate the complexity of the AMH-IR relationship and suggest the need for well-controlled, phenotype-specific investigations to elucidate underlying mechanisms.

Interactions between IR, GDFs, and BMPs in PCOS: Under conditions of physiological equilibrium, BMPs and GDFs have been identified to play essential roles in the regulation of folliculogenesis by paracrine/autocrine mechanisms[93]. In GDF-deficient mouse models, there was anomalous follicular growth beyond the primary stage[93,94]. It was confirmed in initial studies by Orisaka et al[6] that GDF-9 plays an essential role in promoting follicle growth by up-regulating follicular androgen biosynthesis. Conversely, BMPs, specifically BMP4, have been shown to have potent suppressive action on androgen production via the theca cells[95]. The knockdown models of BMP4 showed increased androgen levels and decreased estrogen levels[96]. When ligands of BMP4 bind to their respective receptors on the cell surface, they initiate signal transduction that leads to the activation of suppressor of mothers against decapentaplegic proteins[97]. Mechanistically, the impaired suppressor of mothers against decapentaplegic proteins expression negatively regulates BMP4 signaling, in turn impacting steroid hormone formation[98]. Liu et al[98] showed that ovarian BMP4 levels significantly decreased in HA. Conversely, BMP4 treatment downregulated the androgen production by modulating the expression of steroidogenic enzymes (Figure 5).

Sproul et al[99] showed in their study that BMP15 polymorphisms were linked to anovulation and infertility in PCOS. Additionally, variants in GDF-9 have been associated with impaired reproductive outcomes, such as reduced parity, possibly due to altered oocyte development or early pregnancy loss[99]. These genetic disruptions may not directly cause PCOS but could contribute to fertility issues in affected individuals. Furthermore, in the context of IR, the dysregulation of these oocyte-derived factors could exacerbate follicular arrest and anovulation, emphasizing the multifactorial nature of PCOS-related infertility.

In an estradiol valerate-induced polycystic ovary mouse model, folliculogenesis was disrupted, with follicles arrested at the preantral-to-antral transition[100]. This coincided with the reduced expression of GDF-9 and BMP receptor II, a granulosa cell-expressed receptor crucial for BMP15 and GDF-9[100]. Since GDF-9 and BMP15 are oocyte-derived growth factors essential for granulosa cell proliferation and follicle maturation, their dysregulation may underlie ovulatory dysfunction. This is supported by prior studies showing that GDF-9-null mice fail to develop healthy follicles beyond the primary stage[93], and GDF-9 variants in women are associated with subfertility, including reduced parity and potential early pregnancy loss[99]. Thus, continued research is needed to clarify the consequences of GDF-9 and BMP15 polymorphisms, particularly in metabolically compromised PCOS phenotypes.

TGF-β superfamily signaling pathway and IR: The TGF-β superfamily signaling pathway is intricately involved in the regulation of follicular growth and development[101,102]. In addition to the above-mentioned ligands (i.e., inhibin, activin, AMH, GDFs, and BMPs), the TGF-β family consists of three primary isoforms: TGF-β1, TGF-β2, and TGF-β3[102]. The activity of the TGF-β superfamily is regulated by an extracellular matrix (ECM) protein known as fibrillin (FBN) (Figure 4). The ligands of the TGF-β superfamily are sequestered in an inactive latent form bound to latent TGF-β binding proteins[103]. These then bind to FBN and are stored in their latent form in the ECM until required[104].

Thus, FBN regulates TGF-β signaling by managing its activation and availability in the ECM and its dysfunction may contribute to PCOS disease mechanisms[102]. Genetic studies have identified an association between the FBN3 gene variant D19S884 A8 and metabolic disturbances in women with PCOS[102]. Notably, women with PCOS who have one or two copies of the D19S884 A8 allele exhibit significantly elevated fasting insulin levels, and increased homeostasis model assessment-IR scores compared to those with other genotypes, indicating the presence of a more pronounced IR[105]. This polymorphism has also been linked with β-cell dysfunction and dysregulation in basal glucose homeostasis, further linking FBN3 to the metabolic phenotype of PCOS[102]. These findings suggest that genetic variation in FBN3 may impair the delicate balance of TGF-β signaling, possibly contributing to both androgen excess and IR in PCOS. Considering that multiple ECM-binding proteins, including follistatin, influence the TGF-β pathway, it is likely that FBN3 and follistatin variants function together to disrupt TGF-β-dependent regulation of metabolic and reproductive function in PCOS[102]. The regulation of the FBN-TGF-β interactions offers a novel and promising strategy to manage HA, IR, and anovulation in PCOS[104]. Integrating such approaches with existing therapies could significantly improve outcomes for individuals with PCOS.

KL system and IR in PCOS

The KL system, also called stem cell factor or steel factor, is a granulosa cell-derived factor that regulates theca cell function[106]. The membrane receptor of KL, c-Kit, is a receptor tyrosine kinase protein. C-Kit is expressed on the secondary and antral theca cells and may be involved in androgen production in the theca cells[107]. Thus, KL impacts the crosstalk between granulosa cells and theca cells, which is crucial for androgen and estrogen production. Tuck et al[108] observed a markedly increased KL in follicles at all stages of development within the examined PCOS ovaries. The authors postulated that the increased levels of KL observed in PCOS ovaries may underlie several abnormalities of PCOS, such as increased ovarian reserve due to the diverse roles KL has in animal models (Figure 5)[108]. Thus, the overexpression of KL or c-Kit signaling in PCOS may amplify HA by increasing theca cell activity. This impairs granulosa cell function, including a decrease in aromatase activity, leading to the reduced conversion of androgens to estrogens.

In the context of PCOS, where both IR and hyperinsulinemia are prominent features, the interplay between c-Kit signaling and the insulin receptor pathway may have important implications. C-Kit is not only active in pancreatic β-cells but also in ovarian cells[109]. In β-cells, c-Kit activation promotes insulin secretion by upregulating insulin receptor and insulin receptor substrate 1 (IRS-1), supporting PI3K/Akt pathway[108,109]. However, prolonged c-Kit activation results in negative feedback through serine phosphorylation of IRS-1, inhibiting insulin receptor signaling and reducing insulin output[109]. This may indicate that dysregulated c-Kit expression could contribute to the metabolic infertility phenotype in PCOS. Chronic c-Kit-mediated hyperinsulinemia may exacerbate ovarian androgen production, while long-term impairment of insulin signaling could worsen metabolic dysfunction. Thus, tight regulation of c-Kit-IR interactions may be critical in managing the metabolic aspects of PCOS. Hence, targeting c-Kit activation may offer potential therapeutic targets to address abnormal follicular development and IR[109].

IGF-1 and IR in PCOS

Essential for physiological androgen steroidogenesis are IGF-1 that function to enhance FSH function in granulosa cells by stimulating the differentiation and proliferation of theca cells and granulosa cells[110]. The synergistic action of IGF-1 and LH stimulates androgen production from the theca cells[111]. IGF-1 contributes to steroidogenesis by mimicking and amplifying LH actions[112]. Furthermore, the stimulatory effects of FSH on aromatase expression depend on IGF-1 action and the expression and activation of the IGF-1 receptor (IGF-1R)[113-115]. The inactivation of IGF-1R resulted in completely sterile mice, with only 10% to 60% surviving to adulthood[116]. More adversely, IGF-1R knockout mice die at birth, indicating that several other critical pathways may become dysfunctional in the absence of IGF-1R[116].

Interestingly, the action of insulin on the production of ovarian androgens is believed to be through IGF-1R on theca cells[117]. High levels of insulin and IGF-1 amplify the effect of LH on granulosa cells, which disrupts follicular development and ovulation, leading to anovulation[118]. A recent meta-analysis showed that women with PCOS had significantly higher insulin levels, which reduces the hepatic production of IGF-binding protein 1[119]. Decreased levels of IGF-binding protein 1 result in increased bioavailability of free IGF-1, further triggering androgen production and contributing to the cycle of HA and IR[120]. It has been shown that IGF-1R is necessary to induce the PI3K/Akt pathway[121]. The IGF-1/PI3K pathway functions as a crucial downstream signaling mechanism activated by the binding of IGF-1 to IGF-1R.

IGF-1/PI3K pathway and the IR link

Under normal conditions, when IGF-1 binds to its receptor, IGF-1R, it activates the intracellular signaling cascade, particularly the PI3K/Akt pathway[112]. There is evidence emerging that the IGF-1R/PI3K pathway may be central in the pathogenesis of PCOS. Thus, the increase in androgen levels characteristic of PCOS patients is more related to dysfunction in the IGF-1R/PI3K pathway[122]. Two other role players in the pathway are the IRS-1 and IRS-2 proteins that contribute to the binding and subsequent activation of the PI3K pathway[122,123]. He et al[122] concluded from their studies that the IGF-1R/PI3K pathway is differently expressed in PCOS granulosa cells compared with controls, with IGF-1R, IRS-1 and IRS-2 significantly increased while PTEN decreased. PTEN plays an important regulatory role. The selective deletion of PTEN in mice resulted in the overactivation of PI3K signaling in the ovaries[59]. The mice exhibited elevated androgen levels, ovary enlargement, antral follicle accumulation, and early fertility loss, hallmark features of human PCOS[59]. Thus, it can be concluded that dysfunctional PTEN may be involved in PCOS pathogenesis[122]. Another earlier study by Nahum et al[124] showed the “progonadotrophic” roles for insulin and IGF-1 in regulating normal ovarian androgen production. Their study also revealed the effects of insulin in driving the etiology of HA, interestingly, both with and without hyperinsulinemia in PCOS[124]. In line with this, recent studies have found comparable evidence[125,126]. Targeting the IGF-1/PI3K pathway could provide therapeutic advantages by addressing ovarian and metabolic anomalies in PCOS.

PI3K/Akt-forkhead box O signaling and IR

Under normal conditions, a critical downstream effector of PI3K is the serine/threonine kinase Akt[58]. The activation of the PI3K/Akt pathway promotes insulin-mediated glucose uptake in target tissues by facilitating the transport of glucose transporter 4 to the plasmalemma[127]. It has been shown that the PI3K/Akt signaling pathway plays a crucial role in the ovary. The PI3K/Akt pathway exerts an anti-apoptotic effect, partly by phosphorylating the forkhead box O (FOXO1/3/4/6) transcription factors[128]. Phosphorylation is a regulatory mechanism that controls the activity of FOXO proteins. Thus, in the presence of insulin, FOXO transcription factors are phosphorylated in response to the activation of the PI3K/Akt pathway[129]. Their transcriptional functions are turned off following nuclear exclusion and sequestered into the cytoplasm[129]. In the cytoplasm, FOXO is degraded by proteasomes, which accompanies transcriptional activity loss, thereby preventing it from re-entering the nucleus and activating its target genes[130]. FOXO signaling can induce apoptosis in oocytes and granulosa cells and inhibit the activation of primordial follicles[130]. In particular, FSH is a protective factor that promotes the survival of antral follicles because it antagonizes apoptosis in granulosa cells. Thus, an inverse relationship exists between FOXO and FSH, where FOXO activity is negatively influenced by FSH[131].

Dysregulation of the PI3K/Akt and FOXO signaling pathways in PCOS is pivotal in linking IR with HA and follicular dysfunction[132]. Concurrently, the impairment of the PI3K/Akt pathway, characteristic in most PCOS patients, may lead to the activation and overexpression of FOXO3 in granulosa cells, resulting in higher apoptosis levels in these cells[128]. When hormones and/or growth factors are depleted, FOXO3 dephosphorylates and translocates into the nucleus, resulting in transcription of pro-apoptotic factors and loss of mitochondrial membrane potential[128]. Furthermore, it has been evidenced that FOXO1, which has been extensively studied, mediates gluconeogenesis by upregulating several key genes, including phosphoenolpyruvate carboxykinase and glucose 6-phosphatase[133,134]. Interestingly, during IR, FOXO1 becomes less phosphorylated and therefore more active to promote expression of gluconeogenic genes, which may promote a hyperglycemic state[133]. Some studies have identified aberrant FOXO1-mediated signaling to contribute to the development of low-grade chronic inflammation in the body, leading to HA in PCOS patients[135,136]. Thus, the aberrant activation of the PI3K/Akt and FOXO pathways can contribute to the pathophysiology of PCOS. Hence, modulation of these pathways may provide therapeutic interventions that can address the hormonal and metabolic dysfunctions associated with PCOS.

Mitogen-activated protein kinase/extracellular signal-regulated kinase cascade and IR in PCOS

Under normal physiological conditions, insulin and growth factors such as IGF-1 activate both the PI3K/Akt and mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) pathways in a synchronized way, maintaining proper regulation of metabolism, cell growth, and ovarian function. Both these pathways are associated with LH-induced changes in steroid biosynthesis in ovarian theca cells[137]. Although the PI3K/Akt and MAPK/ERK pathways operate in a balanced and complementary manner, disruption of their coordination can impair PI3K/Akt signaling while relatively amplifying the MAPK/ERK activity, contributing to the development of IR[138].

Recent evidence has implicated the dysfunctional activation of the MAPK/ERK cascade in PCOS, leading to excess androgen biosynthesis and suppressed activity of IRS, thus promoting IR development[139]. The link between IR and activation of the MAPK pathway was demonstrated by Zhang et al[140], where the authors assessed the effects of berberine on MAPK-related protein expression and observed a significant berberine-mediated suppression of p38, ERK, and c-Jun N-terminal kinases. These molecules can provoke the abnormal activation of serine/threonine kinase signaling pathways[141]. Thus, their downregulation was accompanied by reduced PCOS symptoms and IR values in the berberine-treated rat model[140]. These findings were further supported by another study, which demonstrated that ovaries from PCOS rats exhibited decreased PI3K/Akt signaling alongside increased MAPK/ERK pathway activation, suggesting a significant disruption in ovarian glucose metabolism[138].

Furthermore, studies have indicated that impairment in the MAPK/ERK signaling pathway can influence steroidogenic enzyme expression and androgen production[142]. For instance, research suggests that in PCOS theca cells, reduced phosphorylation of ERK1/2 correlated with increased CYP17A1 mRNA expression and increased androgen biosynthesis, implying that decreased ERK1/2 activity may contribute to HA in PCOS[142]. Interestingly, Fukuda et al[143] concluded that LH stimulates CYP17A1 mRNA expression and androgen production in theca cells via activation of the PI3K/Akt pathway, with MAPK/ERK signaling facilitating this process. Insulin appears to act synergistically with LH to trigger StAR protein, stimulating steroidogenic action, with IR, possibly fuelling the impaired MAPK/ERK signaling pathway[144]. IR may play a prominent role in causing gonadotropin dysfunction and steroid hormone imbalance among individuals with PCOS[144]. Tailored strategies to address IR may constitute the cornerstone of remedying both the symptoms and root causes of PCOS.

SHBG and IR connection

In normal homeostatic conditions, the SHBG, a glycoprotein secreted by the liver, plays a critical role as a gatekeeper by binding androgen with high affinity and specificity[145]. Frequently, PCOS patients have abnormally low serum SHBG levels, contributing to a hyperandrogenic state[146]. Total testosterone and SHBG concentrations are inversely related, where a suppressed SHBG results in an increase of bioavailable androgen, thereby leading to a rise in peripheral androgen activity[146]. Since the hepatocytes are the primary site of SHBG synthesis, the nuclear transcription factor, hepatocyte nuclear factor 4 alpha, is the primary determinant of hepatic SHBG expression, thereby controlling SHBG activity[147]. Hyperinsulinemia, a consequence of IR, directly inhibits hepatic SHBG synthesis by possibly downregulating the hepatocyte nuclear factor 4 alpha expression, resulting in elevated free androgen levels in the body[145]. IR and decreased SHBG levels result in a vicious cycle that exacerbates HA and metabolic disturbances in PCOS.

Recent studies have supported the inverse correlation between SHBG levels and markers of IR. A study involving European women with PCOS found that SHBG levels below 40 nmol/L indicated metabolic dysregulation, with waist circumference significantly predicting SHBG levels[148]. Waist circumference is an essential predictor of SHBG levels, particularly in IR states[148]. A larger waist circumference may indicate more abdominal fat, which is metabolically active and contributes to IR by promoting the release of inflammatory molecules and free fatty acids into the bloodstream, resulting in decreased circulating SHBG levels. Waist circumference has been proposed as a better biomarker for glucose and lipid metabolism abnormalities than PCOS status alone, emphasizing its importance in assessing metabolic risk. Targeted treatments addressing IR could help restore SHBG and alleviate symptoms of PCOS.

Neuroendocrine regulation and the IR link in PCOS

Aberrant neuroendocrine signaling has been proposed to play a fundamental role in the development of PCOS. In the neuroendocrine system, the hierarchical arrangement of the hypothalamus, pituitary, and ovary forms the HPO axis that produces hormones in response to signals from the nervous system[149]. In PCOS, dysregulation often occurs in the HPO axis, resulting in a change in the GnRH pulsatile pattern (Figure 6), favoring the release of LH. The elevated LH concentrations can aggravate the hyperandrogenic PCOS state[150]. The exact mechanism driving the disruptions in GnRH pulsatility in PCOS is not fully understood. However, studies have implicated imbalances in the neurotransmitters and neurohormones that regulate GnRH secretions[151].

Figure 6 Schematic illustration of the role of kisspeptin/neurokinin B/dynorphin neurons in regulating the rhythmic release of gonadotropin-releasing hormone, which in turn triggers the secretion of luteinizing hormone and follicle-stimulating hormone. KNDy: Kisspeptin/neurokinin B/dynorphin; GnRH: Gonadotropin-releasing hormone; FSH: Follicle-stimulating hormone; LH: Luteinizing hormone; ARC-ME: Arcuate-median eminence complex.

Of particular interest are the kisspeptin/neurokinin B/dynorphin (KNDy) neurons that are recognized for their co-expression of KNDy, which are involved in orchestrating the rhythmic release of GnRH, which in turn triggers the secretion of LH and FSH[8] (Figure 6). Recent studies have indicated the potential link between KNDy, PCOS and IR[11]. IR in PCOS can disrupt the finely tuned balance of KNDy neurons, causing hormonal dysregulation and reproductive disturbances[152]. Due to the complexity of these interactions, further research is required to understand the specific mechanisms underlying insulin’s effects on KNDy neurons in the context of PCOS.

Additionally, the disruptions in the neural circuits of the neuron-glia network in the arcuate nucleus and median eminence may advance the neuroendocrine disturbances characteristic of PCOS[153]. Glial cells, including astrocytes and tanycytes, play a pivotal role in modulating the secretions of the GnRH by regulating synaptic inputs to GnRH neurons and structural changes in response to the fluctuations in the steroid hormone feedback (such as estrogen)[153]. Thus, the dysregulation in the interaction between neurons and glial cells may contribute to the amplified androgen production observed in PCOS[153].

Recent research has increasingly centered on the pathophysiology of neuroinflammation, or neurodegeneration linked to gut dysbiosis[153]. Numerous studies have linked alterations in gut microbiomes in preclinical PCOS models and clinical PCOS patients[154-157]. Thus, the dysfunction of the gut-inflammation-brain axis can be a crucial link between the metabolic, hormonal, and neuroendocrine disturbances characteristic of PCOS. Targeting this axis can offer promising targets for discovering novel strategies to remedy PCOS[153].

Adipose-neuroendocrine-IR connection

Adipose tissue, mainly white adipose tissue, plays a pivotal role in the pathophysiology of PCOS because of its capacity to produce and metabolize steroid hormones[158]. In individuals with PCOS, alterations in insulin sensitivity and adipocytokine secretion occur. Adipose tissue expresses enzymes such as aldo-keto reductase type 1C3 (AKR1C3), which converts androstenedione to the biologically active androgen, testosterone[159]. The AKR1C3 expression has been shown to be regulated by insulin[159]. IR, an underlying feature of PCOS, leads to increased insulin levels, which in turn, may upregulate the adipose androgen production, thus exacerbating HA[159]. Studies have proposed that increased intra-adipose androgen production by AKR1C3 is influenced by IR and plays a critical role in propagating adipose tissue dysfunction and lipotoxicity in PCOS[159,160].

To further elucidate the neuroendocrine impacts and adipose irregularities, Cox et al[161] developed an adipocyte and brain-specific AR knockout model with a dihydrotestosterone-induced PCOS mouse model. The wildtype mice developed PCOS-like reproductive and metabolic traits[161]. Interestingly, these traits were absent in dihydrotestosterone-treated adipocyte and brain-specific AR knockout females, indicating the crucial role of adipose tissue and brain ARs in PCOS development[161]. Thus, the study suggested that adipose tissue and the brain are key sites for androgen-mediated actions involved in the developmental origins of PCOS[161]. Exploring a closely related aspect, microRNAs (miRNAs) were differentially expressed in adipose tissue in women with PCOS compared to BMI-matched controls[162,163]. This will be elaborated in the subsequent section.

Adipose tissue-miRNA-IR link

Recent studies have shown that miRNAs, i.e., small RNA molecules, can bind to mRNA of target genes, resulting in their degradation or inhibiting mRNA translation[164]. Furthermore, miRNAs have been implicated in altered steroidogenesis and may play an essential role in the pathogenesis of HA in PCOS[165,166]. Additionally, the variance in miRNA expression observed in insulin-sensitive tissues of obese or overweight individuals and patients with T2D hints at a plausible involvement of these small RNA molecules in the complications associated with metabolic syndrome and T2D[167]. For example, miRNA-423-3p may interact with adiponectin receptor 2, destroying adiponectin and/or adiponectin receptor 2. Adiponectin and its receptors have been shown to play a vital role in regulating IR and androgen levels, including the restoration of ovulation in PCOS mice[168]. Thus, the disruption of adiponectin and its receptors by miRNA-423-3p may directly affect normal steroidogenesis, leading to the upregulation of androgens, intensifying the pathogenicity of HA[165,166].

Conversely, PCOS patients often have a decreased expression of miRNA-592, which triggers the increased expression of mRNA for LH/chorionic gonadotropin receptor (LHCGR). The binding of LH or chorionic gonadotropin to the LHCGR, found on the cell surface of the ovaries, prompts various cellular responses, including the production of androgens, exacerbating HA in PCOS[165]. Overall, miRNAs act as crucial gene expression mediators in both the neuroendocrine and adipose tissue, and their dysregulation may contribute to the hormonal imbalances observed in PCOS[164].

Follicular atresia

It has been recognized that follicular atresia is an essential regulatory process in women to maintain a healthy reproductive system[169]. Ultimately, only about 400 follicles (< 1%) ovulate throughout the reproductive lifespan of females, and the majority (> 99%) undergo atresia[170]. The apoptosis of granulosa cells is considered the principal mechanism of atresia[169]. Initial studies in the pathophysiology of PCOS have touted androgen excess for retarding the follicular atresia process. However, the downregulation of FSH action by androgen is consistent with a role for androgen in follicular atresia. During cyclic recruitment, a cohort of antral follicles escapes apoptosis due to the protective action of FSH[171]. One of these follicles, the dominant follicle, grows faster than the rest of the cohort and secretes higher levels of estrogen and inhibin[172,173]. A negative feed loop ensues that suppresses pituitary FSH release, depriving subordinate follicles of adequate FSH required for survival, thus causing them to regress and eventually disintegrate[172].

Plenty of evidence shows that androgens modulate FSH responsiveness at intermediate stages of follicular development, therefore contributing to determining which follicles are “selected” to achieve full preovulatory maturity[71,174]. Słomczyńska et al[175] showed that testosterone acts synergistically with FSH to increase aromatase expression in the small porcine follicles. Other studies, such as those by Nielsen et al[176] and Dewailly et al[84], showed that AR mRNA and intrafollicular androgen levels positively correlated with FSHR mRNA levels. Thus, FSHR transcripts increase rapidly and achieve maximal levels at the early antral stage[177]. As folliculogenesis proceeds, FSHR expression declines in the mid-antral follicles and is replaced by LHCGR, signifying follicular dominance[177,178]. Thus, the switch from FSH to LH responsivity is a requirement to prepare the dominant follicle for ovulation and luteinization[63]. Consequently, the dominant follicle will survive the decrease in FSH levels by developing LH receptors, shielding it from atresia, attributed to LH’s proliferative and anti-apoptotic signals at the intracellular level[63].

In PCOS patients with PCOM, the follicles are arrested in development and do not show apparent signs of atresia. The expression of AR and FSHR may be amplified, as observed in granulosa cells from PCOS patients compared to controls[84]. Aberrant granulosa cell function has been observed, particularly in anovulatory women with PCOS, exhibiting premature responsiveness to LH, implying an early acquisition of LH receptors[179]. Consequently, this may culminate in the failure to select a dominant follicle and a lack of ovulation[64]. Furthermore, the elevated LH stimulation induces androgen production in ovarian theca cells by upregulating the key androgen-producing enzyme (CYP17A1). Thus, accelerating the hyperandrogenic state that may result in the premature stagnation of follicular growth or follicular “arrest”[64,180]. Consequently, the hyperandrogenic state may protect subordinate follicles from atresia, resulting in follicular persistence as seen in patients with PCOM[64]. However, prevailing conditions such as obesity need to be taken into consideration. Recent findings have shown suppressed antral follicle development in women with obesity, resulting in follicles being selected at a small diameter[181]. The smaller size of follicles at selection may reflect premature responsiveness to LH, which correlates with the anovulatory issues associated with PCOS[181].

Insulin and the IGF system may profoundly affect the hyperandrogenic state observed in PCOS[182-184]. Under normal physiological conditions, insulin enhances LH-induced androgen synthesis in the ovary. Thus, in PCOS anovulatory patients with IR and hyperinsulinemia, the synergistic effect of high insulin and LH triggers the HA state[2,144]. Ultimately, the enhanced LH, HA, and hyperinsulinemia may induce premature expression of LH receptors in small diameter follicles, leading to the stagnation of follicular growth, negatively impacting atresia. IR is a pivotal factor that perturbs the homeostasis towards impairment of follicular maturation by exacerbating granulosa cell dysfunction through pro-apoptotic and autophagy-related pathways, resulting in atresia becoming ineffective, culminating in follicular arrest rather than selection of a dominant follicle[185].

Recently identified forms of regulated cell death, notably pyroptosis and ferroptosis, have been recognized as essential contributors to the metabolic and reproductive irregularities noted in PCOS[186,187]. These processes, which are driven by oxidative stress, inflammation, and iron dysregulation, interconnect closely with insulin signaling pathways and may exacerbate IR[186,187]. The functional and regulatory roles of pyroptosis and ferroptosis in the context of PCOS are explored in the following sections.

Ferroptosis in follicular atresia: Ferroptosis is a form of cell death characterized by the accumulation of lipid peroxides and iron-dependent reactive oxygen species, which causes the disruption of cellular membranes and eventually cell death[188,189]. Recently, ferroptosis has increasingly been recognized for its role in ovarian granulosa cell dysfunction in PCOS[186]. Glutathione is a critical cofactor for glutathione peroxidase 4 (GPX4), an enzyme that prevents ferroptosis by reducing lipid peroxides[186,190]. Under normal conditions, GPX4 prevents cell damage by converting fatty acid hydroperoxides into non-toxic fatty acid alcohols[191]. Thus, inhibiting GPX4 can interfere with intracellular iron homeostasis and increase lipid peroxidation[192].

In PCOS, the dysregulation of GPX4 function may worsen the oxidative stress-induced inflammation, a key feature of PCOS, further advancing the androgen synthesis and secretion. Iron overload and oxidative stress can also impair insulin signaling pathways, contributing to IR[193]. Therefore, controlling ferroptosis may play an essential role in PCOS pathophysiology. The targeted regulation of ferroptosis, including restoring GPX4 function, regulating iron metabolism and managing oxidative stress, may help improve metabolic and reproductive outcomes in women with PCOS[194]. Interestingly, natural compounds like leonurine have been found to significantly improve ovarian function, hormone disorders, and IR, while reducing granulosa cell ferroptosis[186].

Pyroptosis in follicular atresia: Pyroptosis, an inflammatory type of programmed cell death, may contribute to the chronic inflammation characteristic of PCOS by the excessive release of proinflammatory cytokines, such as interleukin-1β, interleukin-18 and the formation of pyroptotic bodies, which may exacerbate ovarian dysregulation and hormonal imbalance[195]. Additionally, nucleotide-binding oligomerization domain, leucine-rich repeat, and NLR family pyrin domain-containing protein 3 (NLRP3) is a cytosolic sensor involved in activating the inflammasome complex[187]. Studies have implicated NLRP3 as a significant factor leading to inflammation and IR induced by obesity[196]. Yang et al[187] found that the expression levels of NLRP3 were positively correlated with those of IGF-1. This may indicate that the upregulation of NLRP3 in PCOS may reveal a complex interplay between the increased inflammatory and pyroptosis-related factors and IR[187].

Additionally, studies have indicated that persistent excessive inflammation can cause cell dysfunction and activate pyroptosis[195]. The amplified inflammatory response impairs ovarian function, exacerbating the development of PCOS, resulting in the abnormal overactivation of pyroptosis[195]. Targeted therapies investigating mechanisms that help suppress pyroptosis or regulate the relevant inflammatory pathways could help alleviate inflammation and improve both ovarian function and metabolic health in women with PCOS. For instance, mogroside V, an active ingredient found in the monk fruit, has been shown to inhibit NLRP3-mediated pyroptosis in granulosa cells, thereby improving insulin sensitivity and ovarian function in PCOS models[187].

LIMITATIONS

Despite accumulating findings linking HA and IR to disruptions in key signaling pathways in PCOS, several limitations must be acknowledged. First, many mechanistic insights are based on animal models and in vitro systems. While these models provide important findings, they may not accurately denote the intricate hormonal and metabolic environment of human PCOS. Interspecies differences in ovarian physiology and insulin signaling pathways hinder the direct translation of these results to clinical practice. Second, the inherent heterogeneity of PCOS, including variability in androgen excess, metabolic dysfunction, and ovulatory status, complicates the generalization of pathway-specific findings. Individual signaling pathways may vary across different PCOS phenotypes, making it difficult to identify universal therapeutic targets. Third, although pathways such as PI3K/Akt, MAPK/ERK, and miRNA networks have been studied individually, their interactions within the ovarian microenvironment remain poorly understood. A more integrative approach is needed to understand how these pathways influence each other in PCOS. Fourth, there is a lack of long-term effects and intervention studies specifically targeting the molecular pathways involved in PCOS. Current treatments, such as insulin sensitizers or antiandrogens, often have a widespread impact on the body, making it challenging to determine their exact influence on specific signaling mechanisms. Fifth, a significant limitation is the difficulty in accounting for the diverse environmental and lifestyle exposures, which can substantially influence disease pathogenesis. Emerging pollutants, including pharmaceuticals, personal care products, microplastics, and endocrine-disrupting chemicals, have been implicated in the etiology of PCOS[197]. However, the complexity and variability of these environmental factors, coupled with challenges in determining exposure levels and timing, complicate the establishment of defined cause-and-effect relationships. Additionally, most studies rely on animal models or in vitro systems that may not fully replicate human exposure, limiting the translation of the findings. Furthermore, the cumulative and synergistic effects of various environmental toxins, often encountered simultaneously, are poorly understood, necessitating more research to elucidate their role in PCOS pathophysiology[197].

In continuation of the above discussion, maternal health during pregnancy is crucially important in fetal programming and the future risk of PCOS in offspring. Animal models suggest that maternal HA at a critical period of fetal development may result in permanent changes in fetal physiology that can trigger the development of PCOS later in adult life[198].

Finally, developing areas of investigation, such as ferroptosis and pyroptosis, including the epigenetic regulation, are gaining attention. However, these mechanisms require further validation through studies involving human ovarian tissue and larger, well-defined patient cohorts to determine causality and therapeutic significance.

EMERGING RESEARCH PERSPECTIVE (TARGETING EPIGENETICS, GENETIC EXPRESSION, PHYTOCHEMICALS, AND THE GUT MICROBIOME IN PCOS)

Emerging insights suggest that the complex interplay between genetic expression and epigenetic regulation is a promising avenue in understanding and managing IR and HA in PCOS[199]. Epigenetic modifications include DNA methylation, histone modifications and non-coding RNA activity that can modulate genes, either by exacerbating or mitigating their expression[200]. For instance, aberrant methylation has been implicated in the downregulation of insulin-signaling genes, such as INSR and IRS1, aggravating IR in PCOS patients[201]. Likewise, hypomethylation of the LHCGR gene promoter has been linked with increased receptor expression, which may increase LH signaling and promote excessive steroidogenesis, impairing normal folliculogenesis in PCOS[202].

Plant-derived compounds such as curcumin have exhibited the ability to reverse these epigenetic abnormalities by modulating gene expression related to oxidative stress and metabolism, by focusing on two key genes: Sirtuin 1 and peroxisome proliferator-activated receptor gamma coactivator-1α[203]. The peroxisome proliferator-activated receptor gamma coactivator-1α acts to amplify the body’s antioxidant defense mechanisms by elevating detoxifying enzymes like superoxide dismutase and GPX[204]. Sirtuin 1 regulates insulin secretion, antioxidant action, apoptosis and protects DNA[205].

The gut microbiome has also emerged as a dynamic epigenetic regulator by producing metabolites like short-chain fatty acids, which can influence gene expression[206,207]. Dysbiosis commonly seen in PCOS may contribute to systemic IR and HA through epigenetic mechanisms. Additionally, dysbiosis can interfere with how the body processes estrogen, affecting the gut-HPO axis and triggering inflammation by allowing harmful substances like endotoxins to leak into the bloodstream[208,209]. These effects can disrupt the HPO axis and worsen IR and high androgen levels. These findings highlight the importance of integrated therapeutic approaches that combine phytochemicals, microbiome modulation, and epigenetics, aiming to target the multifaceted nature of PCOS.

CONCLUSION

The dysregulation caused by excess androgen production and action on follicle maturation and the development of PCOS is well established. However, recent research suggests that several ovarian and antecedent factors may contribute to HA. The most critical may be the dysfunction in the insulin signaling pathway and pathogenicity of IR in PCOS patients. The physiological abnormalities associated with IR appear to be a forerunner in the upsurge of the metabolic aberrations related to HA. Metabolic disturbances such as obesity, dyslipidemia, and T2D exacerbate IR, leading to a vicious cycle that may cause long-term health issues such as cardiovascular disease. Furthermore, IR negatively affects the HPO axis, resulting in anomalous gonadotropin release and impaired folliculogenesis. Further studies investigating the mechanisms that target the impact of IR on the HPO axis and GnRH secretions may help control the hormonal imbalances and PCOS.

A deep understanding of the molecular disruptions in PCOS, especially the interaction between IR, HA, and signaling pathways like PI3K/Akt and MAPK/ERK, offers potential for more targeted therapies. Enhancing PI3K/Akt activity or modulating MAPK/ERK signaling might improve insulin sensitivity and reduce androgen excess. Similarly, targeting the activin-inhibin-follistatin axis and restoring GDF-9 and Kit signaling suggests beneficial prospects of improving folliculogenesis and fertility. Heightening SHBG levels to reduce free androgen may contribute to lowering androgenic symptoms. Additionally, emerging research on miRNAs, ferroptosis, and pyroptosis holds promise for novel, molecular-guided treatments. These strategies highlight the shift toward personalized medicine in PCOS, where treatments are aligned with individual molecular profiles to enhance outcomes.

Furthermore, advancements in the PCOS landscape continues to unveil novel biomarkers and therapeutic targets for clinical application. Biomarkers associated with IR and HA offer potential for early and more precise diagnosis and guide more individualized treatment approaches. Molecules such as AMH, SHBG, adiponectin and emerging epigenetic regulators like certain circRNAs may help bridge the gap between metabolic and reproductive phenotypes. As research evolves, integrating these biomarkers into clinical practice could enhance metabolic outcomes and reproductive health in women with PCOS. Leveraging emerging treatment options may offer hope to women who suffer from PCOS, especially in terms of mental, reproductive, and metabolic health.