Mechanisms and treatment modalities related to premature ovarian insufficiency in mitochondria: literature review.

Oxidative stress (OS) refers to a pathophysiological state characterized by an imbalance between the production of reactive oxygen species [ 12 ] and the capacity of the antioxidant defense system, under both physiological and pathological conditions. OS-induced damage denotes irreversible harm to cellular components resulting from this imbalance. Mitochondria are the primary source of ROS, which are short-lived, highly reactive oxygen-containing molecules. Consequently, mitochondria also serve as major targets of ROS-mediated damage during OS. Under physiological conditions, ROS act as signaling molecules that support ovarian cell maturation and follicular rupture, and facilitate the restoration of diploidy within the follicular microenvironment, thereby helping to terminate oocyte meiosis [ 13 ]. In contrast, pathological overproduction of ROS accelerates mitochondrial dysfunction and promotes the aging and functional decline of ovarian cells [ 14 ]. Miao C et al. [ 15 ] demonstrated that in POI patients, the reduction in oocyte and granulosa cell numbers is associated with increased ROS generation and decreased mitochondrial membrane potential, highlighting the significant impact of mitochondrial impairment on ovarian function. Furthermore, mitochondrial dysfunction can, in turn, exacerbate ROS production, creating a self-perpetuating vicious cycle that worsens both conditions [ 16 ]. OS interacts with multiple signaling pathways implicated in POI, including the PI3K/Akt/mTOR, MAPK, Keap1-Nrf2-ARE, and TGF-β/Smads pathways. The PI3K/Akt pathway, for instance, promotes cell cycle progression and inhibits apoptosis by activating downstream kinases, thereby regulating cell proliferation and differentiation [ 17 ]. Zhang W et al. [ 18 , 19 ] found that oral theophylline derivatives and yams improve the number and quality of ovulatory oocytes by activating the phosphatidylinositol 3-kinase (PI3K)/Akt pathway in oocytes to activate primordial follicles, ultimately enhancing fertility in naturally senescent female mice. Liang Y et al. [ 20 ] showed that extracts from Cuscuta species (CS) reduced ROS levels in KGN cells and partially activated the PI3K/Akt pathway to exert antioxidant effects in a POI model. The MAPK pathway in mammals includes several subfamilies: JNK 1/2/3, ERK1/2, p38 MAPK (p38α/β/γ/δ), ERK7/8, ERK3/4, and ERK5/BMK1 [ 21 ]. This pathway regulates diverse physiological processes such as inflammation, stress responses, cell growth, differentiation, and apoptosis through phosphorylation of transcription factors and other substrates [ 22 ]. Sun et al. [ 23 ] found that human umbilical cord mesenchymal stem cells (HUC-MSCs) enhanced the expression of γ-glutamylcysteine synthetase (γ-GCS) and glutathione (GSH) production by modulating the ERK-Nrf2-HO-1 pathway, thereby reducing ROS levels and improving ovarian antioxidant capacity. The Keap1-Nrf2-ARE pathway represents a central defense mechanism against exogenous and endogenous OS, and serves as a critical link in suppressing OS-related inflammatory responses, which are closely associated with aging and other inflammatory conditions [ 24 ]. Liang Y et al. [ 20 ] further confirmed that CS extracts also partially activated the Keap1-Nrf2/HO-1 signaling pathway to reduce ROS levels in KGN cells. Additionally, Nazdikbin Yamchi N et al. [ 25 ] f reported that transplantation of amniotic fluid-derived exosomes (AF-Exos) restored fertility in a POI rat model, potentially via the TGF-β/Smads signaling pathway. As a pivotal downstream mediator of oxidative stress-related signaling pathways, ROS play a central role in the pathogenesis of POI. The disruption of redox homeostasis leads to aberrant ROS levels, which induces mitochondrial dysfunction and contributes to the progression of POI. Consequently, therapeutic strategies that target key components within these signaling pathways to modulate ROS represent a promising approach for POI treatment. The pathways discussed above are summarized in Fig.  2 . Nevertheless, the therapeutic potential of targeting oxidative stress-mediated pathways in POI remains largely unexplored, warranting further in-depth investigation to translate these mechanistic insights into clinical applications. Fig. 2 Signaling pathways associated with mitochondrial oxidative stress in premature ovarian insufficiency (POI). The diagram illustrates key pathways and potential therapeutic agents. Abbreviations : CS  Cuscuta, HUC-MSCs  Human umbilical cord mesenchymal stem cells, AF-Exos  Amniotic fluid exosomes. Legend: Black arrows indicate activation or stimulation; red blunted lines denote inhibition; solid arrows represent direct interactions; dashed arrows represent indirect or potential interactions Signaling pathways associated with mitochondrial oxidative stress in premature ovarian insufficiency (POI). The diagram illustrates key pathways and potential therapeutic agents. Abbreviations : CS  Cuscuta, HUC-MSCs  Human umbilical cord mesenchymal stem cells, AF-Exos  Amniotic fluid exosomes. Legend: Black arrows indicate activation or stimulation; red blunted lines denote inhibition; solid arrows represent direct interactions; dashed arrows represent indirect or potential interactions Mitochondrial dynamics refers to the balanced processes of mitochondrial fission and fusion, which are regulated by specific proteins embedded in the mitochondrial membrane, including the fusion proteins mitofusin 1 (MFN1) and mitofusin 2 (MFN2), and the fission-related proteins dynamin-related protein 1 (DRP1) and mitochondrial fission 1 protein (FIS1) [ 26 , 27 ]. Mitochondrial fusion, mediated by MFN1 and MFN2, promotes the integration of the inner and outer mitochondrial membranes, facilitating the formation of an interconnected mitochondrial network essential for maintaining organelle quality and function. On the other hand, DRP1 acts as the core regulator of mitochondrial fission, enabling rapid adaptation of mitochondrial morphology in response to cellular demands. FIS1 supports the fission process by assisting in the recruitment of DRP1 to mitochondrial membranes. Alterations in the expression or activity of these dynamics-related proteins can lead to morphological and functional abnormalities in mitochondria, contributing to the pathogenesis of POI. Zhang M et al. [ 28 , 29 ] reported that deficiencies in MFN1 and MFN2 impair the development of follicular cells and oocyte maturation, ultimately resulting in reduced fertility. Similarly, the expression and phosphorylation of DRP1 are significantly suppressed in aged ovarian cells, and DRP1 deficiency has been associated with abnormal mitochondrial morphology in oocytes, indirectly compromising ovarian function [ 30 ]. Recently, Ding SM et al. [ 31 ] demonstrated that melatonin (MT) treatment in a chronic unpredictable mild stress (CUMS) mouse model restored the expression of FIS1 and MFN1, reflecting the reestablishment of mitochondrial homeostasis and leading to improved ovarian function. These dynamics-related proteins collectively contribute to mitochondrial quality control by maintaining structural integrity, regulating membrane potential, and facilitating the removal of damaged mitochondria through mitophagy. In aging ovarian cells, mitochondrial dynamics are markedly altered, characterized by diminished fusion and fission activity. Such disruption indicates a loss of mitochondrial homeostasis, which impairs overall organelle quality and is closely associated with the functional decline of ovarian cells in POI. Mitophagy is a selective autophagic process that maintains cellular homeostasis by removing damaged mitochondria, thereby improving overall mitochondrial quality. The PINK1/Parkin pathway plays a central role in regulating this process [ 32 ]. Under conditions of mitochondrial damage, the loss of mitochondrial membrane potential (ΔΨm) prevents PINK1 from entering the inner mitochondrial membrane. Instead, PINK1 accumulates on the cytoplasmic side of the outer mitochondrial membrane, where it undergoes autophosphorylation and recruits Parkin, leading to its activation via ubiquitination [ 33 ]. Parkin-mediated ubiquitination then facilitates the binding of adapter proteins to LC3, initiating the formation of autophagosomes around damaged mitochondria [ 34 ]. Mitophagy has been shown to play stage-specific roles during ovarian follicular development. For instance, anti-Müllerian hormone (AMH), which acts as an inhibitor of primordial follicle activation, suppresses the mitophagy regulator Foxo3a and modulates the PINK1/Parkin pathway [ 28 ]. The PINK1/Parkin pathway mediates the selective clearance of damaged mitochondria in mammalian cells under physiological conditions [ 35 ]. In atretic follicles, follicle-stimulating hormone (FSH) can also influence mitophagy by inhibiting the translocation of Parkin to mitochondria [ 36 ]. In the context of premature ovarian insufficiency (POI), which is closely associated with aberrant ovarian cell development, elevated mitophagy levels have been observed in POI mouse models. These models exhibit increased numbers of autophagosomes and autolysosomes in granulosa cells, along with upregulated mRNA and protein expression of mitophagy-related regulators such as PINK1, Parkin, and LC3 [ 15 ]. Collectively, these findings suggest that mitophagy may be downregulated in aged oocytes, and impairments in autophagic flux can lead to the accumulation of damaged mitochondria and a decline in mitochondrial quality. This functional deterioration ultimately contributes to ovarian aging. However, most current studies on mitophagy in ovarian aging focus on signaling pathways, and further mechanistic investigations are needed to fully elucidate its role in POI pathogenesis. Mitochondrial biogenesis marks the formation of new mitochondria, a process initiated by the coordinated synthesis of nuclear and mitochondrial DNA-encoded proteins, along with their proper localization and folding into functional complexes. These steps enable the biosynthesis of mitochondrial membranes and respiratory chain subunits, which constitute the structural and functional foundation for mitochondrial physiological activities [ 37 ]. In ovarian cells, this process is critically regulated by peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α), which activates downstream signaling molecules such as Nrf1, Nrf2, TFAM, and TFB1/2 M. These factors collectively drive mtDNA replication and transcription, facilitate the translation of mitochondrial proteins, and promote the assembly of new mitochondrial structures [ 38 ]. Current studies indicate that PGC-1α expression is downregulated in cyclophosphamide-induced mouse models of ovarian injury [ 39 ]. Several interventions have been shown to counteract this decline. For example, in animal studies, acupuncture needling from “Zhibian” (BL54) to “Shuidao” (ST28) was found to activate the SIRT1/PGC-1α/Nrf2 signaling pathway, thereby ameliorating POI [ 40 ]. Similarly, combined treatment with mesenchymal stem cell-derived mitochondria (MSC-Mito) and pyrroloquinoline quinone (PQQ) enhanced mitochondrial biogenesis via upregulation of SIRT1 and PGC-1α [ 41 ]. Additionally, coenzyme Q10 has been reported to improve ovarian function by increasing SIRT1 expression, which in turn upregulates PGC-1α activity [ 42 ]. In summary, mitochondrial biogenesis represents a vital self-renewal mechanism that regulates both the quality and quantity of mitochondria. The downregulation of key regulators such as PGC-1α leads to a reduction in mitochondrial mass and functional capacity, ultimately contributing to the decline in ovarian function observed in POI. Mitochondrial DNA (mtDNA) copy number and adequate ATP production have been established as critical for oocyte development and maturation [ 43 ]. During oogenesis, mtDNA copy number increases substantially—from approximately 2,000 copies in primordial germ cells to between 150,000 and 250,000 copies in mature oocytes [ 44 , 45 ]. This quantitative expansion is functionally significant, as mtDNA copy number in both oocytes and granulosa cells is positively correlated with ovarian cell quality [ 46 ]. However, the relationship between mtDNA copy number and specific reproductive disorders remains complex. For instance, X Wan et al. [ 47 ] reported no causal link between mtDNA copy number and several female infertility conditions, including premature ovarian insufficiency (POI), polycystic ovary syndrome (PCOS), and endometriosis. In contrast, May-Panloup P et al. [ 48 ] observed a significant reduction in mean mtDNA copy number in the oocytes of women with POI. Although these findings appear divergent, they highlight the need for further investigation into the role of mtDNA copy number in POI pathogenesis. It remains unclear whether mtDNA copy number can distinguish POI from other ovarian disorders. Should such a distinction be validated, it could enhance the precision of diagnostic and therapeutic strategies for patients. In addition to intracellular mtDNA, extracellular cell-free mtDNA (cf-mtDNA) has gained attention as a potential biomarker. Cf-mtDNA is released into the plasma following mitochondrial damage and is significantly upregulated in certain pathological states. Recent evidence suggests an association between elevated cf-mtDNA levels and POI [ 49 ]. Currently, cf-mtDNA is recognized as a reliable indicator in various inflammation-related diseases. While a link with POI has been established, it remains uncertain whether cf-mtDNA can serve as a novel diagnostic marker for POI, and further validation is required to determine its clinical utility. MtDNA in oocytes undergoes significant replication during maturation, yet it lacks the robust protective and repair mechanisms available to nuclear DNA. This inherent vulnerability renders mtDNA more prone to various forms of damage, including point mutations, oxidative base modifications, and single-strand breaks [ 50 ]. Clinically, elevated levels of both serum mtDNA and reactive oxygen species [ 12 ] have been observed in patients with POI [ 51 ]. The increased ROS production, driven by multiple pathological factors, further exacerbates the susceptibility of mtDNA to mutagenesis [ 52 , 53 ]. Several genetic studies underscore the impact of mtDNA integrity on ovarian function. For instance, H Zhao et [ 54 ]. demonstrated that oocyte-specific deletion of the Fbxw7 gene results in follicular depletion and POI in mice. Moreover, mutations in genes such as FANCA, TWNK, and TP63 have been recently implicated in follicular loss and ovarian damage, ultimately manifesting as POI [ 55 – 58 ]. Heteroplasmy—the coexistence of mutant and wild-type mtDNA within cells—is common in oocytes. An organelle bottleneck during maternal transmission of mtDNA reduces the heteroplasmy load in offspring, which may explain the relatively lower incidence of inherited POI [ 59 ]. Nevertheless, it remains unclear whether fully developed ovarian cells, which remain quiescent for decades after birth, undergo mtDNA alterations due to internal or external environmental factors. In summary, mtDNA mutations are closely associated with the pathogenesis of POI. As the genetic foundation for mitochondrial function, alterations in mtDNA inevitably disrupt oxidative phosphorylation, calcium homeostasis, and redox balance, thereby contributing to POI. Key questions remain unresolved: whether existing mtDNA mutations can be corrected through targeted interventions, and whether protective strategies can be implemented to prevent their occurrence. These avenues represent promising directions for future research. The mitochondrial membrane potential (MMP) is an electrochemical gradient generated by the asymmetric distribution of protons across the inner mitochondrial membrane during energy production. This potential is essential for driving oxidative phosphorylation and ATP synthesis, facilitating the conversion of ADP to ATP. The stability of the MMP is therefore critical for maintaining normal cellular physiological functions. During apoptosis, knockdown of lactate dehydrogenase (LDH) induces a decrease in MMP and an increase in reactive oxygen species [ 12 ] and Bax levels. This cascade promotes the release of cytochrome c (cyt-c), ultimately triggering apoptosis [ 60 ]. Consequently, MMP dissipation is a hallmark of mitochondrial dysfunction in apoptotic and senescent cells, and assessment of MMP is indispensable for evaluating mitochondrial function in ovarian cells from individuals with POI. The plasticity model represents the prevailing framework for understanding the structural and functional organization of the mitochondrial respiratory chain [ 61 ]. According to this model, electrons are transferred sequentially through distinct respiratory complexes, generating an electrochemical gradient across the inner mitochondrial membrane that drives ATP synthesis [ 62 ]. This process, known as oxidative phosphorylation, depends on the coordinated activity of the electron transport chain (ETC)—comprising complex I (NADH: ubiquinone oxidoreductase), complex II (succinate dehydrogenase), complex III (cytochrome c reductase), and complex IV (cytochrome c oxidase)—together with ATP synthase (complex V), which collectively produce ATP [ 63 ]. Coenzyme Q (CoQ), particularly in its ubiquitous form (Coenzyme Q10)CoQ10, serves as a lipid-soluble electron carrier central to the Q cycle in mitochondrial electron transport [ 64 ], In aging humans and murine models, CoQ expression is notably diminished. A study by Ben-Meir et [ 65 ]. demonstrated that CoQ10 supplementation can restore cumulus cell function in aged female mice. Further supporting this, X Nie et [ 66 ]. reported that CoQ10 enhances mitochondrial respiratory chain activity and ATP production, suggesting its potential relevance in improving oocyte quantity and quality, optimizing follicular fluid oxidative metabolism, and promoting ovarian functional recovery in aging ovaries. Impairment of the respiratory chain disrupts cellular energy supply, leading to mitochondrial dysfunction and contributing to the pathogenesis of premature ovarian insufficiency (POI).

This study conducted a systematic search of the literature available up to 2025 in databases including PubMed, Web of Science, and CNKI, covering recent advances in genetics, biochemistry, and clinical research. The inclusion criteria comprised in vitro and in vivo experiments, clinical trials, and review articles related to the mitochondrial mechanisms underlying primary ovarian insufficiency (POI). A total of 93 articles were ultimately selected for comprehensive analysis. The findings were organized into six key aspects: mitochondrial dysfunction, alterations in mitochondrial dynamics, mitochondrial biogenesis, mitochondrial DNA changes, mitochondrial membrane potential (MMP), and mitochondrial electron transport chain function.

POI severely impacts the physiological health and reproductive potential of women, with current treatments offering limited efficacy in restoring ovarian function. Mitochondrial dysfunction has emerged as a pivotal underlying mechanism, converging the damaging effects of diverse etiologies. This review has detailed how deficits in oxidative stress, dynamics, biogenesis, mtDNA integrity, membrane potential, and electron transport collectively drive follicular depletion. Consequently, mitochondria-targeted therapies represent a promising frontier for POI intervention. Agents like CoQ10, along with strategies such as mitochondrial transplantation and biogenesis activators, have demonstrated potential in preclinical studies to rescue oocyte quality and granulosa cell function by restoring energy metabolism and reducing oxidative damage. However, the clinical translation of these approaches faces significant hurdles, including a reliance on animal models with inherent biological differences, a lack of standardized protocols and validated biomarkers, and unresolved challenges in targeted delivery, immune compatibility, and the ethics of genetic manipulation. Therefore, future progress hinges on validating mitochondrial biomarkers in multi-center cohorts, developing targeted delivery systems to enhance efficacy and safety, and conducting rigorous clinical trials. By systematically addressing these translational gaps, mitochondrial medicine holds the potential to evolve into an effective strategy for preserving fertility and improving outcomes for patients with POI.

A global decline in female fertility is evidenced by increasing rates of infertility and miscarriage, with POI being a significant etiology [ 92 ]. Compelling evidence from in vivo and in vitro studies has established that mitochondrial function is a critical determinant of oocyte quality. The ovaries of individuals with POI consistently present a spectrum of mitochondrial impairments, including compromised energy metabolism, disrupted dynamics, reduced biogenesis, alterations in mtDNA copy number and mutation load, diminished membrane potential, and decreased ETC activity. These mitochondrial defects are collectively recognized not as an independent initiating factor, but as a convergent hub downstream of various etiologies, such as genetic, autoimmune, chemotherapeutic, and environmental insults. Although anti-Müllerian hormone (AMH) is the most commonly used biomarker, it exhibits insufficient sensitivity, as approximately 6% of patients with POI present with normal AMH levels. Meanwhile, in animal models, mitochondrial parameters—such as mtDNA copy number, cell-free mtDNA (cf-mtDNA), and mitochondrial membrane potential (MMP)—show strong correlations with ovarian function. However, these findings lack validation in large human cohorts, and substantial species differences—such as the several dozen-fold higher mtDNA copy number in human oocytes compared to mice and distinct antioxidant systems—hinder the direct extrapolation of animal data to humans. Currently, animal experiments have explored the changes in mitochondrial function during ovarian aging, and most of them have detected changes in mitochondria-related indicators, but there is a lack of more reliable and rigorous clinical experiments to verify the specific role of mitochondria-related mechanisms. This is mainly due to the fact that it takes less time to construct animal models of natural aging or chemotherapy-induced POI and is easier to observe than human POI. More notably, the diagnosis of POI requires good patient compliance, which is difficult to guarantee. Therefore, this begs the question: are there more efficient diagnostic markers that can accurately detect premature ovarian insufficiency, better preserve the patient’s fertility, and save the worsening of POI in a timelier manner? Current evidence for mitochondrial-targeted therapies in POI is stratified. CoQ10 demonstrates the highest level of evidence, supported by findings from animal studies and preliminary clinical trials that confirm its role in improving ETC function, reducing oxidative stress, and slowing follicular depletion [ 93 ]. Other nutraceuticals, including resveratrol, melatonin, and NAD⁺ enhancers, currently lack human data, with their potential still under investigation at the preclinical level. Beyond supplementation, more radical approaches like mitochondrial transplantation, mtDNA gene editing, and stem cell-based therapies have shown promise in murine models for improving oocyte competence. However, their path to the clinic is blocked by major technical and ethical barriers, such as immune rejection, delivery inefficiency, off-target effects, and ethical debates.Major methodological bottlenecks currently impede advancement in this field. These include single-center studies with small, heterogeneous cohorts, the logistical difficulties of procuring human ovarian tissue, and long-term follow-up studies plagued by poor patient compliance. The situation is further exacerbated by a lack of standardized sampling and detection protocols, which collectively result in poor reproducibility and low comparability of findings across studies. To advance the diagnosis of POI, a dual-pronged strategy is proposed. First, multi-center cohorts should be established to validate an integrated diagnostic algorithm that combines non-invasive mitochondrial biomarkers—such as circulating cf-mtDNA, mtDNA copy number, and mutation profiles—with AMH for improved diagnosis and fertility assessment. This clinical approach should be complemented by a deep mechanistic investigation, which involves creating a high-resolution imaging repository of ovarian energy metabolism (via ultrasound, MRI, and PET-CT) and correlating it with synchronous single-cell multi-omics analyses (transcriptomics, proteomics, metabolomics). The convergence of these data will ultimately enable the construction of a comprehensive “Mitochondria-Ovary Aging Digital Atlas,” providing an unprecedented systems-level view of ovarian aging. In addition to existing therapeutic drugs and methods, multi-omics analysis is booming, which provides more advanced technical support for the discovery of new therapeutic targets for the treatment of POI. Perhaps the application of new technologies can lead to the discovery of drug targets similar to CoQ10 drugs at different levels, bringing new hope for the treatment of mitochondrial dysfunction and POI. Mitochondrial transplantation is also one of the emerging therapeutic directions, but there are problems that are difficult to solve: exogenous mitochondria may trigger immune rejection or heterogeneous fusion (e.g., mitochondrial DNA heterogeneity) and inefficient delivery: only a small amount of stem cells are homing to the ovaries after intravenous injection, and local delivery systems (e.g., ovarian arterial perfusion) need to be developed. The mtDNA editing tools (e.g., mito-TALENs) are still experimental, and germ cell editing is ethically controversial at the genetic level. Improving the quality of oocytes by improving mitochondrial quality is likely to be a new way to manage and enhance reproductive outcomes in patients with POI. For further exploration of treatment strategies, it is necessary to integrate clinical data more tightly, which is a key step in translating animal experimental results into clinical trials. The combination of multi-center data modality, multi-omics technology, multi-modal imaging and mitochondrial index detection has an obvious effect on improving the conversion rate between the two, which is an efficient means to break the barriers between time, space and ethnicity to promote POI research. In addition to integrating existing research, it is also necessary to conduct new research, and the following is what I personally believe is a feasible direction. First, nanocarrier technology is used for precise targeted delivery: mitochondria-targeted nanoparticles (e.g., TPP-modified liposomes) are designed to encapsulate antioxidants (e.g., melatonin) or siRNAs (e.g., targeting DRP1) to improve ovarian-specific accumulation. Alternatively, exosome engineering can be carried out using follicular fluid exosomes loaded with PGC-1α mRNA and delivered to granulosa cells through membrane fusion. Secondly, the combination of therapies is optimized, such as the sharing of antioxidant and energy supplementation: the combination of CoQ10 (to improve ETC function) and α-lipoic acid (to scavenge ROS) synergistically improves oocyte quality; Stem cell binding epiregulation: MSCs in combination with histone deacetylase inhibitors (e.g., NAD + precursor NMN) enhance mitochondrial biosynthesis through the SIRT1/PGC-1α pathway. Alternatively, mitochondrial replacement and gene repair: autologous mitochondrial transplantation, in which healthy mitochondria are extracted from the patient’s adipose stem cells and microinjected into autologous oocytes to avoid immune rejection (reported in preclinical trials); Base editing technology: mito-CRISPR is used to correct pathogenic mtDNA point mutations (e.g. m.11778G > A), and progress has been made in the Leber hereditary optic neuropathy model. Finally, the new strategy of integrated traditional Chinese and Western medicine and the mechanism analysis of traditional Chinese medicine compounds: for example, “He’s Yulin Fang” can alleviate oxidative stress through the Nrf2/HO-1 pathway, and in the future, it can be combined with network pharmacology to screen key active ingredients (such as baicalin); Metabolic regulation of acupuncture: electroacupuncture “Guanyuan” and “Sanyinjiao” acupoints may up-regulate mitochondrial function through AMPK/PGC-1α, and the changes of ovarian glucose metabolism should be further verified by PET-CT. Of course, the above treatment methods are theoretically reasonable, but their clinical application needs to be studied. While the studies included in this review provide valuable insights into the role of mitochondria in POI and potential therapeutic strategies, it is crucial to acknowledge the potential biases inherent in this body of evidence. First, publication bias is likely prevalent, as studies reporting positive outcomes (i.e., effective interventions) are more likely to be published and accessible than those with negative or null results. This may lead to an over-optimistic perception of the efficacy of certain therapies, such as CoQ10 and TCM formulations. Second, significant model limitations exist in preclinical research, particularly concerning the use of animal models like chemotherapy-induced POI in mice. These models may not fully recapitulate the complex etiology and pathophysiology of human POI, thereby limiting the translational relevance of the findings. Additionally, many mechanistic studies, for instance those investigating specific signaling pathways, are conducted in highly controlled in vitro settings, which may introduce experimental bias and fail to mirror the intricate in vivo microenvironment. Finally, research on complex interventions like TCM formulations (e.g., HSYLF, BSNXD) and stem cell therapies often faces challenges related to inadequate standardization of components and insufficient mechanistic elucidation. This lack of standardization hinders the reproducibility of results and complicates cross-study comparisons. Moving forward, mitigating these biases will require pre-registered study protocols, rigorously designed randomized controlled trials (especially in the clinical translation phase), and the systematic reporting of negative outcomes. Such measures are essential for an objective and accurate assessment of the true potential of mitochondria-targeted therapies for POI.

A variety of antioxidant compounds have been shown to improve ovarian function in POI by counteracting mitochondrial oxidative stress. These include proanthocyanidins [ 67 ], He Shi Yu Lin Formula (HSYLF) [ 68 ], deer blood hydrolysate [ 69 ], Bu-Shen-Ning-Xin Decoction (BSNXD) [ 70 ], Icariin [ 71 ], and CoQ10 [ 72 ]. It is noteworthy that these agents have so far been validated only in animal models. They primarily function by alleviating redox imbalance, thereby ameliorating POI manifestations in mice. Feasibility and Efficacy: While these compounds demonstrate promising efficacy in animal models by alleviating oxidative stress, their clinical translation is severely hampered by a near-total absence of human data. The critical question of whether these preclinical benefits can be replicated in humans remains unanswered. Particular challenges exist for complex herbal formulations (e.g., HSYLF, BSNXD), where ill-defined active ingredients and unclear molecular targets preclude standardization, quality control, and precise dosing. Deer blood hydrolysate, as a biological product, carries potential risks of immunogenicity and viral contamination. Among these, CoQ10 possesses the most substantial evidence base, including preliminary clinical observations; however, its low oral bioavailability and the need for confirmatory large-scale trials are significant limitations. Side Effects and Risks: Because systematic toxicology and side-effect profiles are conspicuously lacking, the multi-component nature of herbal formulas introduces a risk of unforeseen drug-herb interactions and long-term toxicity [ 12 ]. Mitochondrial dynamics are regulated by key proteins, and their imbalance is implicated in POI. Melatonin (MT) has been shown to restore the diminished expression of fission protein FIS1 and fusion proteins OPA1 and MFN1, thereby reestablishing mitochondrial homeostasis [ 31 ]. Similarly, the herbal formulation Jian-Pi-Yi-Shen (JPYS) upregulates the levels of Drp1 and Fis1, further promoting balanced fission [ 73 ]. In addition, human umbilical mesenchymal stem cells (hUMSCs) contribute to POI recovery by modulating mitochondrial kinetic homeostasis [ 74 ], suppressing ferroptosis [ 75 ], and improving endometrial receptivity [ 76 ]. Feasibility and Efficacy: Although melatonin shows promise, its use in POI is experimental. Defining the optimal dosage, treatment timing, and potential impacts on the hypothalamic-pituitary-gonadal axis in humans requires rigorous clinical investigation. The herbal formula JPYS faces the same standardization and mechanistic elucidation challenges as other complex mixtures. Human umbilical cord mesenchymal stem cells (hUMSCs) present profound feasibility issues, including risks of tumorigenicity, immunogenicity, low cell survival post-transplantation, and high costs, rendering them currently impractical for widespread clinical use. Side Effects and Risks: Potential side effects of melatonin include drowsiness, headaches, and hormonal disturbances [ 77 ]. hUMSC therapy carries risks of teratoma formation, microembolism, and immune overactivation (cytokine release syndrome) [ 78 , 79 ]. The regulation of mitophagy presents a promising therapeutic avenue for POI. In a cyclophosphamide-induced POI model, the combination of bone marrow mesenchymal stem cells with moxibustion (BMSCs-MOX) was found to ameliorate the condition by modulating mitophagy [ 80 ]. Conversely, electroacupuncture (EA) has been shown to protect granulosa cells by promoting mitophagy through the Hippo-YAP/TAZ pathway [ 81 ]. I This suggests that the therapeutic benefits may arise from restoring mitophagic balance, either by enhancing deficient clearance or suppressing excessive degradation. In practice, mitophagy is often favorably modulated as part of the mechanism of action for many effective drugs. Feasibility and Efficacy: The combination of bone marrow mesenchymal stem cells with moxibustion (BMSCs-MOX) represents a highly complex and non-standardized intervention, making it impossible to attribute efficacy to a specific component or to replicate the protocol consistently. While electroacupuncture (EA) is intriguing, its operator-dependent nature and the difficulty in blinding present major methodological challenges for rigorous randomized controlled trials (RCTs). The proposed mechanism via the Hippo-YAP/TAZ pathway requires more direct molecular validation. Side Effects and Risks: The clinical application of BMSCs-MOX is associated with notable risks, including biological challenges such as immune rejection, undesirable inflammatory responses, and potential viral contamination, alongside technical and logistical hurdles like low cell survival rates and difficulties in transportation and storage [ 82 ]. EA is generally safe but may cause minor bleeding, bruising, or, rarely, vasovagal syncope [ 83 ]. Enhancing mitochondrial biogenesis has emerged as a key strategy for POI treatment. Both He’s Yangchao recipe [ 84 ] and the combination of pyrroloquinoline quinone (PQQ) with MSC-derived mitochondria (MSC-Mito) [ 41 ] have been demonstrated to stimulate this process. The mechanisms involve the upregulation of key regulators such as SIRT1 and PGC-1α, which drive the synthesis of new mitochondria. This is evidenced by increased biogenesis markers in granulosa cells, particularly following estrogen receptor β (ERβ) blockade. Feasibility and Efficacy: He’s Yangchao recipe is hampered by the inherent limitations of uncharacterized herbal mixtures. The strategy of combining PQQ with MSC-derived mitochondria (MSC-Mito) is scientifically novel but clinically unfeasible with current technology. The challenges of mitochondrial isolation, delivery, and the unknown stability of the purported synergistic effect relegate this approach to the realm of exploratory basic science. Side Effects and Risks: The combination of PQQ and MSC-Mito introduces unpredictable risks [ 85 ]. Mitochondrial transplantation can provoke immune responses and lead to heteroplasmy, with long-term safety profiles being entirely unknown [ 86 ]. Therapeutic strategies targeting mitochondrial DNA (mtDNA) focus on improving its quality, quantity, and overall integrity. While Nicotinamide Mononucleotide (NMN) does not significantly increase mtDNA copy number, it markedly enhances oocyte quality by promoting normal spindle formation and ensuring accurate chromosome alignment [ 87 ]. In contrast, He’s Yangchao recipe effectively boosts mtDNA copy number and reduces mitochondrial structural abnormalities [ 84 ]. Furthermore, the combination of MSC-Mito and PQQ acts synergistically to inhibit mtDNA damage [ 41 ]. Beyond pharmacological approaches, mitochondrial transplantation has emerged as a promising technique to directly improve both the quantity and quality of mtDNA [ 88 ]. Feasibility and Efficacy: The fact that NMN improves oocyte quality without increasing mtDNA copy number suggests an indirect mechanism of action, potentially through improving overall cellular metabolic health rather than direct mtDNA repair. Mitochondrial transplantation is the most technologically ambitious and least mature approach. Beyond efficacy concerns, critical issues of immune rejection, delivery efficiency, and stable functional integration of transplanted mitochondria remain unresolved. Furthermore, its application to germ cells raises profound ethical concerns and is subject to strict regulatory restrictions. Side Effects and Risks: Mitochondrial transplantation and gene editing carry risks of off-target effects and metabolic disorders due to heteroplasmy, alongside significant ethical controversies. CoQ10, a fat-soluble quinone and endogenous component of the mitochondrial respiratory chain, represents one of the most extensively studied therapeutic agents for enhancing mitochondrial electron transport. In addition to its central role in facilitating electron transfer within the electron transport chain (ETC), CoQ10 exerts potent antioxidant effects that protect lipid membranes from peroxidation, thereby preserving the structural integrity of cellular and organellar membranes and maintaining normal cell morphology [ 89 ]. Furthermore, CoQ10 has been shown to reduce ROS levels in oocytes, inhibit apoptotic pathways, and improve overall mitochondrial function, collectively contributing to the enhancement of ovarian reserve in preclinical models [ 90 ]. Feasibility and Efficacy: CoQ10 stands out as one of the most viable candidates due to its established safety profile as a dietary supplement and the existence of supporting clinical data. However, the strength and consistency of its therapeutic effect must be conclusively demonstrated through large-scale, multi-center, placebo-controlled clinical trials before it can be recommended as a standard treatment for POI. Side Effects and Risks: CoQ10 is generally well-tolerated, with mild and rare side effects, such as gastrointestinal discomfort, anorexia, and skin rashes [ 91 ]. An overview of the aforementioned therapeutic agents, including their mechanisms of action and supporting evidence, is provided in Table  1 . Although numerous mitochondrial-targeting drugs have demonstrated efficacy in ameliorating POI in animal studies, their translation into clinical practice remains limited, and robust statistical analyses in human populations are still lacking. Several potential therapies aimed at improving mitochondrial quality are currently under development or in early-stage clinical trials. However, no single treatment has proven universally effective in preventing ovarian aging. Given the multifactorial etiology and complex mechanisms underlying POI, it is unlikely that any monotherapy will fully prevent or treat the condition. Moreover, even for interventions that have entered clinical use, the intricate regulation of human reproduction necessitates thorough safety and efficacy evaluations to ensure minimal adverse effects before widespread application can be recommended. Table 1 Premature ovarian insufficiency is treated with different mitochondrial aspects Different aspects Drug name Reference Model Sample size Mechanism of action Outcomes Study limitations Improve mitochondrial dysfunction proanthocyanidins (PC) [ 67 ] mouse 20 SESTRIN2-NRF2 pathway PC prevents and ovarian function by activating a protective SESTRIN2-NRF2 pathway against oxidative stress. Diverse 3-NPA protocols and small samples curb generalizability. He Shi Yu Lin formula(HSYLF) [ 68 ] mouse 50 HSYLF treats POI by decreasing mitochondrial ROS, increasing membrane potential, and improving mitochondrial function. Mitochondrial markers only; no human data limit translation. deer blood hydrolysate [ 69 ] mouse 40 DBH can be further developed as a functional food for the prevention of POF and the retardation of the POF process. No long-term safety, human data, and Nrf2/HO-1-Bcl-2/Bax/caspase-3 mechanistic depth. Bu-Shen-Ning-Xin decoction [ 70 ] mouse 30 rno_circRNA_012284/rno_miR-760–3p/HBEGF axis. The BSNXD moiety ameliorates POI by attenuating OS in ovarian granulosa cells by rno_circRNA_012284/rno_miR-760–3p/HBEGF axis. Rat-only, no human tissue; long-term safety and BSNXD pharmacokinetics untested. Icariin [ 71 ] mouse 88 Nrf2 pathway ICA treats POF by modulating the Nrf2 pathway to ameliorate oxidative stress, ferroptosis, and apoptosis. Ovarian protection by icariin is confined to tumor-bearing mice; its effect on cisplatin sensitivity across tumor types is unknown. Coenzyme Q10(CoQ10) [ 72 ] mouse 32 CoQ10 protects POF through antioxidant and proliferation. Balance mitochondrial fusion and fission melatonin [ 31 ] mouse 40 eIF2α-AFT4 pathway MT mitigates POI by inhibiting the eIF2α-AFT4 pathway. Mechanistic insight remains superficial, with bioinformatics approaches yet to be deployed. human umbilical cord mesenchymal stem cells(hUMSCs) [ 74 ] mouse 48 hUMSCs treatment can restore the imbalance of mitochondrial dynamics and restart testosterone synthesis of TCs by suppressing GSK3β expression, ultimately alleviating POI damage. How hUMSCs phosphorylate GSK3β to repair mitochondrial dynamics is still speculative and needs deeper study. Jian-Pi-Yi-Shen(JPYS) [ 73 ] mouse 40 ASK1/JNK pathway JPYS decoction improves mitochondrial function and alleviates apoptosis through ASK1/JNK pathway. Multi-omics analyses are required to fully clarify JPYS’s therapeutic mechanism. Decrease mitophagy bone marrow mesenchymal stem cells combined with moxibustion (BMSCs-MOX) [ 80 ] mouse 48 Moxibustion enhanced the migration and homing of BMSCs following transplantation and improves their ability to repair ovarian damage. Rat data, canonical Pink1/Parkin axis only; primate/organoid validation and non-classical mitophagy await study. electroacupuncture (EA) [ 81 ] mouse 24 Hippo-YAP/TAZ pathway EA at the Guanyuan (CV4) acupoint protected the granulosa cell by inhibiting cell apoptosis and promoting mitophagy, which was mediated by the Hippo-YAP/TAZ pathway. Mouse-only evidence, single-pathway focus; human relevance and broader mitophagy routes remain to be tested. Restore mitochondrial biogenesis He’s Yangchao recipe [ 84 ] mouse 42 ERβ/PGC1α/TFAM pathway He’s Yangchao recipe can regulate mitochondrial biogenesis through ERβ/PGC1α/TFAM pathway to improve ovarian function in POI mice. Mouse-only POI model, no human validation; Yangchao recipe composition and chronic safety await study. Pyrroloquinoline quinone (PQQ) in combination with MSC-Mito [ 41 ] mouse 35 SIRT1/ATM/p53 pathway PQQ facilitates MSC-Mito proliferation and, in combination with MSC-Mito, ameliorates chemotherapy-induced POI through the SIRT1/ATM/p53 signaling pathway. Single-model, short-term data; long-term scope and broader POI causes await testing. Ameliorate mitochondrial DNA Nicotinamide mononucleotides (NMN) [ 87 ] mouse 62 The regenerative effects of nicotinamide mononucleotide can be a potential treatment option for elderly patients seeking to become mothers through autologous oocytes. Limited access to chemotherapy-exposed human oocytes constrained cross-species validation. He’s Yangchao recipe [ 84 ] mouse 42 ERβ/PGC1α/TFAM pathway He’s Yangchao recipe can regulate mitochondrial biogenesis through ERβ/PGC1α/TFAM pathway to improve ovarian function in POI mice. Mouse-only POI model, no human validation; Yangchao recipe composition and chronic safety await study. Pyrroloquinoline quinone (PQQ) in combination with MSC-Mito [ 41 ] mouse 35 SIRT1/ATM/p53 pathway PQQ facilitates MSC-Mito proliferation and, in combination with MSC-Mito, ameliorates chemotherapy-induced POI through the SIRT1/ATM/p53 signaling pathway. Single-model, short-term data; long-term scope and broader POI causes await testing. Strengthen mitochondrial electron transport Coenzyme Q10(CoQ10) [ 89 ] mouse 32 The protective effect of CoQ10 on the ovaries is exerted through antioxidant and proliferative properties. In vitro and xenograft models limit clinical translation of FSP1-targeted ferroptosis therapy. Premature ovarian insufficiency is treated with different mitochondrial aspects proanthocyanidins (PC) The primary limitations include: (1) the significant gap between rodent models and human pathophysiology, limiting the extrapolation of findings; (2) a pervasive lack of standardization and mechanistic clarity, particularly for herbal and cellular therapies; and (3) a notable absence of systematic safety data for nearly all interventions beyond CoQ10.Future research must pivot towards addressing these translational challenges. The highest priorities should be the validation of efficacy in human cohorts, the development of standardized and scalable manufacturing protocols, and the rigorous establishment of long-term safety profiles. Without focused efforts on these fronts, the promising field of mitochondrial medicine for POI will remain largely confined to preclinical exploration.

Premature ovarian insufficiency (POI) is a clinical condition defined as the cessation of ovarian function before the age of 40. It is characterized by amenorrhea, elevated serum follicle-stimulating hormone (FSH) levels (>40 IU/L), and estrogen deficiency. This hormonal disruption results in atrophy of the sexual organs, persistent amenorrhea, and increased gonadotropin levels. Clinically, patients often experience varying degrees of perimenopausal symptoms, marking the stage of declining ovarian function due to hormonal shifts [ 1 ]. Over recent decades, the increasing incidence of premature ovarian insufficiency has placed affected women at a higher risk for a spectrum of complications, encompassing premature mortality, cardiovascular and cerebrovascular diseases, mental health conditions, sexual dysfunction, and infertility [ 2 ], which affects at least 3.5% of women of reproductive age globally [ 3 ]. Unfortunately, POI has a complex and multifactorial etiology, and current treatments cannot fully restore ovarian function [ 4 ]. As a result, the quality of life and mental health of affected patients continue to be adversely affected. POI is strongly associated with advanced age and is closely linked to mitochondrial dysfunction, although its precise pathogenesis remains incompletely understood. Mitochondria are highly dynamic, double-membrane-bound organelles with a diameter of 0.5–1.0 μm and contain their own genetic material. Their function is regulated by both the mitochondrial and nuclear genomes. Human mitochondrial DNA (mtDNA) is a circular molecule comprising 16,569 base pairs. It encodes 13 proteins, all of which are subunits of the electron transport chain (ETC). These mtDNA-encoded proteins associate with nuclear-encoded proteins to form functional complexes essential for oxidative phosphorylation (OXPHOS). As the primary sites of cellular oxidative metabolism, mitochondria are responsible for the final oxidation of energy substrates and the release of metabolic energy. They mediate key oxidative pathways—the tricarboxylic acid (TCA) cycle and OXPHOS—through which cells accomplish the second and third stages of aerobic respiration. Notably, oocytes contain a greater number of mitochondria and a higher mtDNA copy number compared to somatic cells. This reflects the elevated demand for energy and biosynthetic precursors to support normal oocyte maturation and early embryonic development [ 5 ]. Mitochondrial replication occurs actively by the blastocyst stage [ 6 ], after which it becomes quiescent until implantation, when replication resumes to support further development [ 5 ]. Compared to oocytes from younger women, those from older women exhibit a decline in both the quality and quantity of mitochondria [ 7 ]. Mitochondria in the aging ovary undergo significant morphological alterations, including irregular swelling, cristae fragmentation, and the formation of vacuoles. Concurrently, senescent ovarian cells demonstrate reduced mtDNA content compared to their younger counterparts [ 8 ]. Notably, strategies aimed at improving mitochondrial function have shown therapeutic potential for POI. These include attenuating mitochondrial oxidative stress [ 9 ], enhancing ovarian dynamics [ 10 ], and increasing the mitochondrial membrane potential (MMP) [ 11 ]. Although the precise relationship between mitochondrial dysfunction and POI pathogenesis remains to be fully elucidated, Fig. 1 summarizes several proposed mechanistic links. This article, therefore, focuses on elucidating the connection between mitochondrial integrity and POI and discusses related therapeutic strategies that target mitochondrial function. Fig. 1 Mechanisms linking mitochondrial dysfunction to premature ovarian insufficiency (POI). (Created with BioRender.com.) The diagram illustrates six major aspects of mitochondrial pathology in POI: (1) Mitochondrial dysfunction, primarily manifested as a disruption of redox homeostasis and increased oxidative stress. (2) Dysregulation of Mitochondrial Dynamics, referring to an imbalance between mitochondrial fission and fusion, as well as impaired mitophagy. (3) Dysregulation of Mitochondrial Biogenesis, which involves the synthesis of mitochondrial DNA and proteins necessary for new organelle formation. (4) Mitochondrial DNA Aberrations, including quantitative changes (copy number depletion) and qualitative changes (mutations). (5) Destabilization of Mitochondrial Membrane Potential, where a decrease in MMP is a hallmark of functional decline. (6) Mitochondrial Electron Transport, as ATP production depends on efficient electron flow, which provides the proton motive force for oxidative phosphorylation Mechanisms linking mitochondrial dysfunction to premature ovarian insufficiency (POI). (Created with BioRender.com.) The diagram illustrates six major aspects of mitochondrial pathology in POI: (1) Mitochondrial dysfunction, primarily manifested as a disruption of redox homeostasis and increased oxidative stress. (2) Dysregulation of Mitochondrial Dynamics, referring to an imbalance between mitochondrial fission and fusion, as well as impaired mitophagy. (3) Dysregulation of Mitochondrial Biogenesis, which involves the synthesis of mitochondrial DNA and proteins necessary for new organelle formation. (4) Mitochondrial DNA Aberrations, including quantitative changes (copy number depletion) and qualitative changes (mutations). (5) Destabilization of Mitochondrial Membrane Potential, where a decrease in MMP is a hallmark of functional decline. (6) Mitochondrial Electron Transport, as ATP production depends on efficient electron flow, which provides the proton motive force for oxidative phosphorylation