Polycystic ovary syndrome (PCOS) is an endocrine disorder affecting 5–15% of reproductive-age women. It is characterized by hyperandrogenemia, ovulatory dysfunction, and polycystic ovarian morphology, and is frequently associated with insulin resistance (IR) and metabolic dysregulation [1]. The traditional mechanistic paradigm attributes PCOS pathogenesis to dysregulation of the hypothalamic-pituitary-ovarian axis (HPOA). Abnormal gonadotropin-releasing hormone (GnRH) pulsatility disrupts the luteinizing hormone-to-follicle-stimulating hormone (LH/FSH) ratio, driving excessive ovarian androgen production and anovulation [2]. However, the systemic metabolic disturbances observed in PCOS suggest a more complex pathophysiology involving bidirectional hypothalamic-adipose tissue crosstalk. Emerging evidence demonstrates that hypothalamic nuclei regulate adipose tissue metabolism through sympathetic neural efferents, while adipose-derived factors reciprocally regulate hypothalamic function, forming an integrated neuroendocrine network [3].

In PCOS pathophysiology, hyperandrogenism and sympathetic overactivation synergistically disrupt hypothalamic-adipose signaling: leptin resistance and reduced adiponectin levels impair hypothalamic energy sensing and integration [4,5,6]. Concurrently, an imbalance between neuropeptide Y (NPY) and pro-opiomelanocortin (POMC) neuronal activity induces appetite dysregulation, exacerbating obesity and insulin resistance. Although androgen excess suppresses adipose tissue lipolysis in PCOS [7] elevated non-esterified fatty acids (NEFA) and chronic low-grade inflammation—mediated by cytokines such as tumor necrosis factor-alpha (TNF-α)—synergistically exacerbate hypothalamic dysfunction and ovarian androgen biosynthesis [8]. These interactions establish a self-reinforcing cycle of metabolic and reproductive dysfunction, highlighting the central role of dysregulated neuro-metabolic-endocrine crosstalk in PCOS pathogenesis. Collectively, these findings establish the HSF axis as a novel mechanistic framework. In this model, bidirectional neuroendocrine signaling coordinates metabolic and reproductive homeostasis, providing a paradigm shift in understanding PCOS pathogenesis.

The hypothalamus, situated ventrally at the base of the brain below the thalamus [9] functions as a pivotal neuroendocrine center. It regulates adipose tissue metabolism via sympathetic neural pathways. Specific nuclei—such as the arcuate nucleus (ARC), lateral hypothalamic area (LHA), and paraventricular nucleus (PVN)—regulate sympathetic output to enhance brown adipose tissue (BAT) thermogenesis and white adipose tissue (WAT) lipolysis, thereby maintaining systemic energy homeostasis [10,11,12,13]. In PCOS, however, this regulatory equilibrium is disrupted. A distinct dichotomy in sympathetic regulation is observed: diminished sympathetic excitation to BAT reduces thermogenic capacity, whereas sympathetic overactivation in WAT promotes lipometabolic dysfunction and proinflammatory responses [14]. Retrograde tracing studies have identified key hypothalamic nuclei projecting to BAT, including the ARC, PVN, LHA, dorsomedial hypothalamus (DMH), and ventromedial hypothalamus (VMH) [15]. Similarly, hypothalamic regulation of WAT primarily involves the suprachiasmatic nucleus (SCN), LHA, DMH, and PVN [16]. Notably, overlapping neural pathways (e.g., ARC, PVN, DMH, LHA) govern both BAT and WAT metabolism. Given the pivotal role of these nuclei in PCOS-related metabolic disturbances, elucidating their neural projection networks may provide mechanistic insights into adipose tissue dysfunction in PCOS pathogenesis [Supplemental Figure].

Specific projections and functional differentiation of neural nucleii

The ARC regulates BAT function through direct projections to the DMH and PVN [11]. POMC neurons in the ARC release α-melanocyte-stimulating hormone (α-MSH), which activates melanocortin 3/4 receptors (MC3/4R) in the DMH and PVN to enhance sympathetic drive to BAT [17]. Animal model studies have shown that inhibition of DMN or PVN significantly reduces ARC-mediated thermogenesis effects [11]. Conversely, ARC-derived NPY neurons inhibit sympathetic output via Y-receptor signaling in sympathetic preganglionic neurons or upstream regulatory nuclei such as PVN or LH, triggering inhibitory pathways [18]. Furthermore, NPY inhibits α-MSH release from ARC neurons, thereby reducing sympathetic activation [19] [Supplemental Figure S1].

The PVN acts as a critical integrative hub, receiving ARC-derived signals to activate sympathetic pathways. It simultaneously sends glutamatergic projections to the rostral ventrolateral medulla (RVLM) and nucleus tractus solitarius (NTS), amplifying sympathetic excitation in adipose tissue [20]. Additional PVN-derived factors, including brain-derived neurotrophic factor (BDNF) [15] thyrotropin-releasing hormone (TRH) and ceramides, further potentiate sympathetic outflow. Similarly, the DMH modulates sympathetic activity through dual mechanisms: glutamatergic neurons stimulate sympathetic excitation, while NPYergic neurons exert inhibitory effects [13, 21].

Other hypothalamic nuclei, such as the LHA and VMH, also regulate adipose tissue via sympathetic pathways. These neural nuclei integrate signals from hypothalamic and peripherally derived sources to orchestrate adaptive metabolic responses; however, significant interspecies divergences exist between animals and humans, necessitating expanded human-centric investigations to elucidate mechanisms.

Hypothalamic secretory factors regulating adipocyte function

The LHA modulates adipose tissue metabolism via endocrine actions of the neurohormone orexin (OX) [22]. OX directly promotes differentiation of in vitro-cultured preadipocytes into brown adipocytes, with clinical studies revealing reduced OX levels in adipose tissue of obese individuals [23]. Additionally, hypothalamic-derived oxytocin exerts direct regulatory effects on adipocytes through widely expressed oxytocin receptors (OXTRs). Oxytocin activates OXTR, upregulating adipose triglyceride lipase (ATGL) and hormone-sensitive lipase (HSL) to enhance triglyceride hydrolysis. [Supplemental Table] [24]These findings underscore the dual neuroendocrine regulation of adipose metabolism through orexinergic and oxytocinergic pathways. However, systematic characterization of receptor distribution patterns and downstream signaling cascades is required to fully elucidate hypothalamic-adipose cross-talk.

Neural integration of temperature signals

Based on animal experiments, thermoregulatory responses are initiated by cold stimuli detected at cutaneous sensory endings. These stimuli are transmitted via primary sensory neurons within dorsal root ganglia (DRG) to distinct populations of secondary glutamatergic neurons in the spinal cord dorsal horn(DH).These signals subsequently activate tertiary neurons in the lateral parabrachial nucleus (LPBN) of the hypothalamus. The pathway further converges on the median preoptic subnucleus (MnPOA) within the preoptic area (POA), where GABAergic interneurons are recruited to inhibit warm-sensitive neurons in the medial preoptic area (MPA). Notably, GABAergic neurons in the DMH exert tonic suppression on local glutamatergic neurons. Experimental studies demonstrate that GABAergic neurons in the prefrontal POA release this inhibition during cold exposure, thereby amplifying sympathetic output [25,26,27,28].

Activated DMH glutamatergic neurons subsequently drive sympathetic premotor neurons in the rostral raphe pallidus nucleus (rRPa) and parapyramidal region (PaPy), ultimately enhancing sympathetic activity through spinal preganglionic neurons. This hierarchical three-neuron pathway (DRG, DH, LPBN, POA) enables cold-sensitive neurons in the POA to activate the sympathetic nervous system (SNS), stimulating BAT thermogenesis and promoting WAT browning.

Sympathetic neural circuitry activates adipose lipid mobilization

The hypothalamic-sympathetic regulation of adipose tissue is orchestrated through integration of thermal and metabolic signals. After metabolic signaling or cold activation of cutaneous thermoreceptors, hypothalamic efferent neurons coordinate signal transmission. The signal travels through spinal thoracolumbar sympathetic premotor neurons to postganglionic sympathetic neurons, forming a sequential neural pathway [29]. Norepinephrine (NE) released by sympathetic neurons binds to β3-adrenergic receptors (β3-ARs) on both BAT and WAT. This signaling selectively stimulates lipolysis while concomitantly inhibiting lipoprotein lipase (LPL) activity, thus restricting triglyceride uptake [30,31,32]. [Supplemental Table] UCP1, a mitochondrial uncoupling protein, drives non-shivering thermogenesis by dissipating the proton gradient across the inner mitochondrial membrane. Liberated fatty acids enter mitochondria and undergo β-oxidation in the tricarboxylic acid (TCA) cycle, generating a proton gradient. Activated UCP1 binds fatty acids to form proton channels, shuttling protons back into the mitochondrial matrix and bypassing ATP synthesis to release energy as heat [32, 33] [Supplemental Figure S1].

Sympathetic activation elevates intracellular AMP: ATP ratios, triggering AMPK-dependent phosphorylation of PGC-1α that enhances its transcriptional coactivator function. This cascade potentiates thermogenic output in BAT and promotes adipose tissue browning in WAT [34]. [Supplemental Table S1] Chronic cold adaptation induces adipose thermogenic plasticity and sympathetic remodeling. PR domain-containing 16 (PRDM16) serves as a master regulator of thermogenic gene programming in brown/beige adipocytes. This protein suppresses fibrogenic pathways via interaction with transcription factors such as GTF2IRD1. Simultaneously, PRDM16 promotes neurotrophic factor secretion (e.g., S100b), sustaining dense sympathetic innervation critical for sustained thermogenic capacity [35].

Cortisol in PCOS and its role in BAT

Cortisol can affect the regulation of lipid metabolism in the sympathetic nervous system of PCOS patients, and the specific neuro-endocrine mechanism is not fully understood. Some research results have laid a foundation for further exploration in this field.During sympathetic excitation, postganglionic neurons potentiate CRH secretion. This potentiation is achieved through the release of NE, which facilitates the enhancement of corticotropin-releasing hormone secretion by these neurons. This neuronal activity thereby activates the HPA axis.

Concurrently, postganglionic neurons (chromaffin cells) within the adrenal medulla release epinephrine into the systemic circulation. This hormone acts endocrinally to bind β-adrenergic receptors on adrenocortical cells, amplifying their responsiveness to adrenocorticotropic hormone (ACTH) stimulation and thereby facilitating cortisol biosynthesis. 17α-Hydroxylase catalyzes the conversion of pregnenolone to 17α-hydroxypregnenolone; subsequently, 17,20-lyase further converts 17α-hydroxypregnenolone to dehydroepiandrosterone (DHEA), the precursor of androgen. In PCOS patients, elevated cortisol-induced hyperinsulinemia upregulates the enzymatic activity of 17α-hydroxylase and 17,20-lyase. Therefore, augmenting androgen synthesis in both ovarian theca cells and adrenal zona reticularis [36]. In addition, abnormal cortisol levels may inhibit the activity of aromatase, reduce the conversion of androgens to estrogen, lead to the accumulation of androgens in the body, and the relative decrease of estrogen levels, which in turn affects the feedback regulation of the HPO and aggravates the reproductive endocrine disorders in patients with PCOS.

Cortisol can affect insulin sensitivity through a variety of mechanisms, which in turn can affect the onset of PCOS. On one hand, cortisol activates glucocorticoid receptors in hepatocytes to stimulate gluconeogenesis, provoking fasting hyperglycemia. This metabolic shift triggers compensatory hypersecretion of insulin to maintain glucose homeostasis. However, chronic cortisol exposure induces systemic insulin resistance through downregulation of insulin signaling pathways [37]. On the other hand, persistently elevated cortisol liberates excessive free fatty acids (FFA) from adipose tissue, which ectopically accumulate in skeletal muscle and hepatic tissues to exacerbate insulin resistance [38]. In women with PCOS, this insulin resistance frequently culminates in hyperinsulinemia, which disrupts ovarian function through multiple mechanisms. Firstly, hyperinsulinemia directly stimulates insulin receptors in ovarian theca cells, augmenting androgen synthesis [39]. Concurrently, this hyperandrogenic milieu suppresses dominant follicle recruitment and ovulation. Secondly, hyperinsulinemia may indirectly modulate steroidogenic enzyme activity in adipose tissue, perturbing cortisol and androgen metabolism. This promotes adipose tissue redistribution with preferential abdominal fat accumulation, further exacerbating insulin resistance. Collectively, these interactions establish a vicious cycle: glucocorticoid excess leads to dysregulated lipid metabolism, which in turn contributes to insulin resistance. Insulin resistance then induces a hyperandrogenic status, ultimately resulting in ovulation disorders that exacerbate the clinical manifestations of PCOS [39].

Sympathetic enhancement drives PCOS inflammation and insulin resistance

Women with PCOS typically exhibit autonomic nervous system imbalances, marked by sympathetic hyperactivity and reduced parasympathetic activity. These imbalances lead to increased lipolysis, heightened insulin resistance, and impaired ovulation [40]. From rodent models, we know that ovarian function and lipid metabolism disorders can be improved by reducing sympathetic drive [41]. Therefore, we can infer that improving autonomic balance is a strategy to further improve PCOS, but there are few studies in this area.

Adipose tissue is not merely a passive energy reservoir but a functionally diverse and highly heterogeneous organ system [42]. Based on structural, functional, and anatomical distinctions, it is classified into three major subtypes: WAT, BAT, and beige adipose tissue. In PCOS, increased sympathetic nervous system activity affects different fat tissues in opposing ways. Paradoxically, BAT thermogenesis remains unchanged despite overall sympathetic overactivity. In contrast, heightened sympathetic activity in WAT worsens inflammation but fails to stimulate browning [43,44,45].

WAT serves as the primary energy reservoir in humans, storing excess energy in triglycerides. It also functions as an endocrine organ, secreting adipokines like adiponectin and leptin and proinflammatory cytokines such as TNF-α, IL-6 to regulate systemic metabolism. In PCOS, visceral adipose tissue (VAT) expansion manifests as central obesity. This expansion exacerbates lipolysis, promotes the release of inflammatory mediators, and is closely associated with IR and hyperandrogenism. Hypothalamic appetite dysregulation drives excessive energy intake, leading to lipid droplet accumulation and adipocyte hypertrophy. These pathological changes further aggravate metabolic dysfunction through adipose tissue remodeling and chronic low-grade inflammation [46].

Concurrently, IR promotes excessive VAT and subcutaneous adipose tissue (SAT) accumulation, with VAT expansion predominating [47]. Pathologic VAT remodeling—characterized by adipocyte hypertrophy, chronic inflammation, apoptosis, fibrosis, and NEFA release—creates a proinflammatory condition [48]. This cycle involves the abnormal deposition of FFA produced by lipolysis in non-adipose tissue organs such as the liver and muscle, thereby inhibiting insulin signaling [49, 50] [Supplemental Table], while insulin resistance in turn amplifies sympathetic nerve excitation [51]. Aberrant expression of transcriptional regulators PPARγ and C/EBPα in PCOS-WAT exacerbates adipocyte dysfunction, marked by hypoadiponectinemia and excessive FFA release [52]. These metabolic disturbances impair hypothalamic appetite regulation and heighten sympathetic output, perpetuating systemic dysregulation [53].

The hypothalamus enhances sympathetic outflow to WAT, increasing NE release. When norepinephrine binds to adipocytes, it inhibits the function of the insulin receptor substrate (IRS) through FFA [54] and upregulates monocyte chemoattractant protein-1 (MCP-1), which attracts macrophages [55]. [Supplemental Table] Infiltrating macrophages secrete TNF-α, which suppresses IRS-1 phosphorylation and GLUT4 translocation, intensifying IR [56]. In PCOS, chronic low-grade adipose inflammation synergizes with obesity to sustain this pathogenic loop.

Excess androgens lead to metabolic disorders in adipose tissue

BAT is brown in color due to its rich mitochondria. Its main function is to increase cellular heat production by non-shivering thermogenesis mediated by UCP1, and to consume energy in the form of heat to promote metabolism [57]. BAT also exerts anti-inflammatory effects via secreted factors such as fibroblast growth factor 21 (FGF21), which attenuates WAT-derived proinflammatory cytokine production and improves insulin sensitivity [58]. Consequently, higher BAT activity correlates with increased energy expenditure and reduced risks of obesity and metabolic disorders [59].

In PCOS, however, supraclavicular BAT activity is markedly diminished, as evidenced by lower skin temperatures in this region compared to healthy controls. This reduction inversely correlates with body mass index (BMI) and androgen levels [60]. Animal experiments have shown that excess androgen exacerbates BAT dysfunction because it down-regulates peroxisome proliferator-activated receptor γ coactivator 1-α (PGC1α), a major regulator of thermogenesis, thereby inhibiting energy release and exacerbating metabolic dysregulation [61]. Notably, a 2016 study by Chen Zijiang’s team demonstrated that exogenous BAT transplantation or pharmacological induction of WAT browning ameliorates both metabolic and reproductive abnormalities in PCOS models, suggesting therapeutic potential [62]. Xenotransplantation approaches further indicate that BAT activation elevates circulating adiponectin levels, partially restoring metabolic homeostasis in PCOS. However, clinical translation of these strategies requires rigorous investigation to address safety and efficacy in humans [63].

Beige adipose tissue (BeAT) exhibits hybrid characteristics of white and brown adipose tissues. Through a “browning” process triggered by cold exposure, exercise, or pharmacological stimuli, BeAT undergoes mitochondrial biogenesis and upregulates UCP1 expression. Predominantly localized in subcutaneous depots and interspersed with white adipocytes, BeAT enhances insulin sensitivity and lipid oxidation while secreting immunomodulatory and anti-inflammatory factors.

In PCOS, however, BeAT functionality appears impaired. Reduced expression of browning markers—notably PRDM16, a master regulator of white-to-beige adipocyte transdifferentiation—compromises thermogenic capacity and disrupts sympathetic innervation. However, anti-androgen therapies can reverse these defects by restoring PRDM16-mediated transcriptional programs, suggesting androgen excess directly suppresses BeAT plasticity in PCOS pathophysiology [64,65,66].

WAT in PCOS exhibits profound pathological alterations. Hyperandrogenemia directly activates androgen receptors (ARs) in WAT [67]driving premature differentiation of preadipocytes into dysfunctional mature adipocytes [68] [Supplemental Figure S1].

Pathologic remodeling of the fat-hypothalamic feedback circuit

Adipose factors derived from adipose tissue can cross the blood-brain barrier and regulate the function of the hypothalamus through receptor binding, thereby regulating appetite, energy metabolism, and neuroendocrine activities. Leptin, a key adipokine secreted by WAT, regulates appetite and energy homeostasis [69]. Leptin acts as a “sympathetic stimulant” that activates sympathetic nerve activity regulated by the hypothalamus. Animal studies have shown that leptin directly increases sympathetic innervation of adipose tissue, liver, and ovaries by injecting into the ventricles [70, 71]. In PCOS, elevated levels of leptin in the blood cross the blood-brain barrier and bind to leptin receptors in the ARC of the hypothalamus, upregulating POMC expression while inhibiting the activity of NPY neurons, thereby suppressing appetite and enhancing sympathetic excitation [72,73,74] [Supplemental Table]. In animal models of PCOS, central leptin resistance has been observed, manifested by the inability to inhibit food intake or weight gain after leptin injection, which is associated with chronic inflammation of the hypothalamus. Therefore, as a result, ARC is unable to respond to elevated leptin levels, loses its appetite suppression effect, leads to weight gain, and drives obesity [75].

Adiponectin, a hormone secreted by fat cells, crosses the blood-brain barrier to inhibit microglia activation, thereby reducing saturated fatty acid-induced hypothalamic inflammation [76,77,78] [Supplemental Table]. The increase in sympathetic nerve activity was significantly associated with a decrease in high molecular weight lipocalin (HMW-adiponectin) levels. In cross-sectional studies, sympathetic nerve activity was negatively correlated with lipocalin. This suggests that elevated sympathetic nerve activity may directly inhibit lipocalin secretion or synthesis [79]. Adiponectin also inhibits NPY expression while stimulating POMC production, reducing food intake. In PCOS, hypoadiponectinemia disrupts these regulatory pathways, promoting hyperphagia, obesity, and IR. High-fat diets exacerbate this dysfunction by amplifying hypothalamic inflammation [80]. Synergistic effect: Leptin and adiponectin, through their ratio (L: A), are predictive factors of metabolic risk and insulin resistance in PCOS [5].

BAT secretes batokines such as FGF21 that activate hypothalamic β-klotho receptors, suppressing NPY neurons, enhancing sympathetic output, and mitigating neuroinflammation [81]. Cold exposure reduces nerve growth factor (NGF) synthesis in BAT, while reserpine-induced NE depletion upregulates NGF production [82]. This reveals a short-loop feedback mechanism: sympathetic NE release inhibits BAT-derived NGF, preventing excessive sympathetic activation. Notably, NGF modulates ovarian and sympathetic functions, suggesting its dual role in PCOS pathogenesis may offer therapeutic potential.

Effects of adipose browning on ovarian function

Adipose browning, particularly the activation of BAT or conversion of WAT to beige adipocytes, plays a critical role in regulating ovarian function through metabolic and endocrine pathways. Studies have shown that xenografted rat BAT is functional and does not cause harmful tissue rejection in aging mice. It improves ovarian function by increasing follicular reserve and oocyte quality, significantly enhancing fertility in aging mice. Additionally, it also improves the metabolism and overall health of aging mice, and some of the improved genes in transplanted mice can be passed on to offspring [83]. Mechanistically, BAT transplantation restores metabolic homeostasis, including systemic energy expenditure and adipokine profiles like adiponectin, which are essential for ovarian folliculogenesis. Notably, improved gene expression patterns in transplanted mice, such as those related to mitochondrial function and anti-inflammation, are heritable to offspring, suggesting transgenerational effects of BAT-mediated metabolic reprogramming [84]. It has also been shown that perinatal exposure to pesticides or their metabolites can impair thermogenesis by interfering with the connectivity patterns of sympathetic circuits that regulate BAT [85]. These findings highlight BAT’s therapeutic potential in age- or obesity-related ovarian dysfunction, though human translational studies are warranted to validate safety and efficacy.

Estrogenic regulation in BAT

The regulation of estrogen in BAT is important, and its pathological significance in PCOS cannot be ignored. Estrogen mainly acts directly on brown adipocytes through the nuclear receptor ERα, promoting mitochondrial biogenesis and norepinephrine-induced lipolysis to activate their thermogenesis [86]. In addition, estrogen signaling in the central nervous system (such as the hypothalamus) indirectly enhances BAT activity by activating the sympathetic nervous system [87]. Martinez et al. found that the central role of estradiol increases sympathetic nervous system activation of BAT thermogenesis through the VMH [88]. Estrogen also has a protective effect on BAT inflammation by inhibiting oxidative stress and immune cell infiltration [89]. In addition, some studies have suggested that BAT oxidative metabolism and glucose uptake in postmenopausal women (low estrogen levels) are significantly lower than those in premenopausal women; BAT oxidative metabolism decreases to postmenopausal levels when ovarian function is suppressed in premenopausal women, resulting in a decrease in circulating estrogen [90]. Therefore, it is inferred from the above results that estradiol loss may inhibit BAT mitochondrial activity [90]. Animal models have shown that activation of ovarian sympathetic nerves (SON) significantly reduces ovarian blood flow and estradiol (estradiol) secretion [91]. This change leads to abnormal follicular development and decreased estrogen synthesis. In summary, the effect of estrogen on BAT is positive, and treatment of sympathetic and estrogen may have therapeutic potential for PCOS [Supplemental Figure].

Interaction between ovary and extraovarian adipose tissue

Removal of POAT can result in increased expression of Fshr gene, decreased expression of steroidogenic enzymes gene, altered levels of growth factors and adipokines in ovary, decreased levels of adiponectin, lipocalin- 2 and vascular endothelial growth factor, decreased lipid accumulation and increased abundance of enzymes involved in lipogenesis in ovary, induced caspase activation and follicle apoptosis in ovary [92]. It also altered the activation of signaling molecules involved in maintaining intracellular homeostasis, suggesting that a large number of ovarian adipokines and growth factors originated from POAT. In addition, in the mouse model, bilateral POAT removal surgery was performed at 6 weeks of age, and after 2 weeks, abnormal follicular development, decreased serum sex hormone levels, and abnormal expression of genes related to follicle development were observed in the operating mice, accompanied by abnormal lipid metabolism, which was manifested by a decrease in fat mass, an increase in energy expenditure, and an up-regulation of the expression of genes and proteins involved in lipolysis [93]. Collectively, these findings establish POAT as a key source of paracrine factors and lipid regulators for ovarian function. Disruptions in this interaction may contribute to reproductive disorders like PCOS, emphasizing the need to investigate neuroendocrine mechanisms underlying adipose-ovarian crosstalk for targeted therapeutic development.

Initiation and amplification of hypothalamic inflammation

Hypothalamic inflammation contributes significantly to metabolic dysregulation in PCOS. Insulin receptors expressed on hypothalamic astrocytes mediate metabolic control by modulating neuronal activity in the PVN and ARC, suppressing appetite and enhancing sympathetic output [94,95,96]. NEFA and TNF-α released from adipose tissue traverse the blood-brain barrier, activating the NF-κB pathway in hypothalamic astrocytes and polarizing microglia. These events ultimately amplify GnRH pulsatility. In PCOS animal models, dual mechanisms drive hypothalamic dysfunction [97]:on the one hand, activated microglia near GnRH neurons exhibit enhanced phagocytosis of GABAergic terminals, reducing inhibitory neurotransmission and increasing GnRH pulse frequency; on the other hand, elevated circulating anti-Müllerian hormone (AMH) induces cytoskeletal remodeling in hypothalamic tanycytes, triggering rapid axonal retraction. This disrupts the physical barrier between GnRH nerve terminals and fenestrated capillaries, facilitating excessive GnRH release and subsequent ovarian hyperandrogenism [98, 99]. Paradoxically, heightened GnRH pulsatility further exacerbates hypothalamic inflammation by polarizing microglia toward the proinflammatory M1 phenotype, which releases IL-6, TNF-α, and nitric oxide. Concurrently, upregulated suppressors of cytokine signaling 3 (SOCS3) and protein tyrosine phosphatase 1B (PTP1B) impair leptin and insulin signaling pathways, fostering central leptin/insulin resistance and perpetuating a self-reinforcing cycle of metabolic and reproductive dysfunction [13, 100] [Supplemental Table]. This leptin resistance disrupts hypothalamic appetite regulation and energy expenditure, exacerbating metabolic dysfunction in PCOS [101, 102].

Sympathetic regulation of adipose tissue immune cells

Inflammation caused by fat cell hypertrophy reduces insulin receptor density, ultimately impairing glucose uptake [103] [Supplemental Table]. Hyperandrogenism and inflammation jointly polarize macrophages toward the proinflammatory M1 phenotype in PCOS-WAT. M1 macrophages secrete MCP-1 to recruit additional immune cells and release TNF-α, creating a self-amplifying cycle of IR and localized inflammation [104, 105]. Notably, cold exposure induces a class of cold-induced neuroimmune cells (CINCs) in subcutaneous WAT. CINCs respond to sympathetic NE signals by releasing BDNF [106]. Murine studies reveal that myeloid-specific BDNF deletion reduces sympathetic innervation of WAT, suggesting NE-activated macrophages regulate BDNF-mediated neuro-adipose crosstalk [107]. BDNF is a large molecular protein with a high molecular weight that makes it difficult to penetrate the blood-brain barrier (BBB) [108]. However, based on animal models, inflammatory factors can damage the blood-brain barrier under chronic inflammatory conditions [109]. Therefore, it is hypothesized that BDNF has the potential to cross the BBB, thereby increasing hypothalamic-sympathetic output, which may be a potential pathway, and that anti-inflammatory therapy for PCOS to improve the integrity of the BBB may help maintain homeostasis in the central nervous system. However, the existing studies are mostly based on hypothetical conditions and lack animal and human models, and more targeted confirmatory studies could be conducted in the future.

Beyond macrophages, multiple immune cells contribute to adipose inflammation in PCOS: Circulating mononuclear cells (MNCs) infiltrate dysfunctional WAT, differentiating into macrophages that secrete proinflammatory cytokines (IL-6, IL-1β, TNF-α), directly exacerbating IR and inflammation [110]. Ovarian MNC-derived macrophages may further stimulate CYP17 (a key androgen synthesis enzyme), promoting hyperandrogenism [111]. T cells, B cells, and neutrophils are implicated in adipose inflammation, though their mechanistic roles in PCOS require further investigation.

Conclusion

The pathogenesis of polycystic ovary syndrome (PCOS) has evolved from the traditional focus on hypothalamic-pituitary-ovarian axis (HPOA) dysregulation to a more comprehensive paradigm emphasizing the intricate interplay within the hypothalamic-sympathetic nerve-adipose tissue axis (HSF axis). Research demonstrates that hypothalamic nuclei—including the ARC, PVN, and DMH—integrate metabolic signals to dynamically regulate adipose tissue metabolism and function via sympathetic pathways. Under physiological conditions, sympathetic activation stimulates β-adrenergic receptors in adipose tissue, driving UCP1-mediated thermogenesis in BAT, browning of WAT, and lipolysis to maintain energy homeostasis. Concurrently, adipose-derived adipokines provide feedback to suppress hypothalamic appetite centers and enhance sympathetic output, forming a bidirectional regulatory network.

In PCOS, however, the HSF axis is profoundly dysregulated. Hyperandrogenemia suppresses BAT thermogenic gene expression while paradoxically activating sympathetic overdrive to WAT, resulting in BAT hypofunction, WAT lipolytic hyperactivity, and chronic inflammation. Concurrently, adipose metabolic dysfunction disrupts hypothalamic energy sensing and neuroendocrine regulation via blood-brain barrier impairment, exacerbating IR and androgen synthesis. NEFA and proinflammatory cytokines released from adipose tissue further aggravate hypothalamic dysfunction, while TNF-α from adipose-resident macrophages induces central IR and amplifies sympathetic imbalance in a chronic low-grade inflammatory milieu. These abnormalities coalesce into a self-perpetuating cycle of “hypothalamic inflammation to sympathetic hyperactivity to adipose metabolic dysfunction”, ultimately manifesting as PCOS hallmarks: obesity, IR, hyperandrogenemia, and ovulatory disorders [112].

Notably, neuroendocrine-metabolic interactions can exacerbate this vicious cycle. Hypothalamic dysregulation, an imbalance in NPY/POMC neurons, promotes overeating and reduced energy expenditure, leading to obesity. Hyperandrogenism has an inhibitory effect on BAT activity and beige adipogenesis, thus reducing metabolic efficiency. Moreover, there is a sympathetic-IR feedback loop where IR itself enhances sympathetic excitation, creating a feedforward mechanism that further exacerbates the situation. Collectively, these multilayered interactions underscore the centrality of HSF axis dysregulation in PCOS pathophysiology [113].

Hypothalamus-sympathetic nervous system-brown adipose tissue axis research: current status and challenges

In the research on the HSF axis, the existing literature still has significant limitations. Most functional mechanism studies heavily rely on rodent models, but their neural anatomical structures differ significantly from those of humans—for instance, mice lack fine anatomical features of the sympathetic nervous system. This discrepancy limits the extrapolation of research findings to humans and may lead to misjudgments about the neural mechanisms controlling human brown fat [114]. Secondly, the research content generally focuses on the sympathetic nerve regulatory pathways, but little is known about the mechanisms of sensory nerve innervation and the neural microenvironment. Reports in related fields on the interaction between adipose tissue nerves and immunity, and the co-release of neurotransmitters are almost non-existent [115]. Moreover, regarding the clinical application hypotheses of selective neural modulation techniques, due to the limited sample size and insufficient data from human subjects, their effectiveness still lacks direct experimental evidence.

Future research efforts should focus on clarifying the neural projection networks of key hypothalamic nuclei (ARC, PVN, DMH) and their regulatory roles in adipose-sympathetic interactions, unraveling the short-loop feedback mechanisms through which adipokines like leptin and NGF modulate sympathetic activity, and exploring depot-specific variations in adipose-hypothalamic signaling. Using cross-synaptic tracing techniques, to analyse the common neural input pathways of hypothalamic nuclei to brown fat, white fat, and beige fat, and to reveal the core mechanism of CNS ‘one nucleus, multiple targets’ regulation in energy metabolism allocation. These advancements will establish a foundation for developing targeted therapies that intervene in the HSF axis to concurrently address metabolic dysfunction and reproductive disorders in PCOS [116].