Review maps extracellular vesicle opportunities and limits in PMOS

A 2026 narrative review concludes that extracellular vesicles, which carry proteins, lipids and nucleic acids released by cells, may influence ovarian signaling and systemic metabolism in polyendocrine metabolic ovarian syndrome, formerly called polycystic ovary syndrome. Studies summarized in the review report effects on granulosa-cell survival, steroidogenesis, mitochondrial quality control, insulin resistance and oocyte maturation in cell cultures and mouse or rat models. The evidence is promising but overwhelmingly preclinical: the authors say incomplete vesicle characterization, inconsistent dosing and models, limited biodistribution data and unresolved long-term reproductive safety prevent extrapolation to patient care. No clinically validated diagnostic test or approved extracellular-vesicle treatment for PMOS is identified by the review.

The terminology matters because extracellular vesicles are not a single product. They are lipid-bilayer particles formed through distinct processes, including endosomal release, plasma-membrane budding and fragmentation of apoptotic cells. The review follows MISEV2023 guidance by using the general term extracellular vesicle, or EV, unless the origin of a preparation is established. Small EVs, commonly defined operationally as particles below approximately 200 nanometres, are also called sEVs, but their small size does not prove that they arose inside endosomes. Likewise, morphology, electron microscopy and detection of proteins such as CD9, CD63 and CD81 alone cannot establish exosome biogenesis. This distinction is important for interpreting which biological effects can reasonably be attributed to EVs rather than to co-isolated proteins, RNA, lipids or other material.

Clinical observations point toward both diagnostic and disease-mechanistic possibilities, while falling short of establishing clinical utility. One study reported increased annexin V-positive platelet microparticles in women with PCOS. Another compared follicular-fluid particle fractions and granulosa-cell small-RNA profiles in women with and without PCOS, finding that body-mass index and the sampled particle fraction influenced the apparent RNA signatures. The six upregulated and sixteen downregulated small RNAs reported in that comparison came from an overall follicular-fluid analysis rather than from an exclusively small-EV population. Follicular fluid is usually collected during oocyte retrieval, making it an invasive specimen rather than a non-invasive screening tool. Circulating EVs are more compatible with accessible sampling, but disease-specific sensitivity, specificity and added clinical value have not yet been established.

Functional experiments suggest that follicular EVs can carry pathological as well as potentially protective signals. Elevated EV-associated LIPE-AS1 was linked to altered steroid metabolism and impaired oocyte maturation through a LIPE-AS1/miR-4306/LHCGR regulatory axis, while adipocyte-derived EV-associated miR-26b was reported to promote cumulus-cell apoptosis by targeting JAG1. In an IVF/ICSI cohort, higher miR-143-5p in follicular-fluid EVs suppressed RASAL2 and promoted granulosa-cell proliferation; inhibiting the miRNA in a letrozole-induced rat model improved ovarian and hormonal abnormalities. Because increased proliferation can coexist with arrested follicular development, the finding does not make miR-143-5p a therapeutic supplement. Instead, it supports its candidacy as a biomarker or inhibitory target, subject to independent validation, prespecified thresholds and assessment beyond standard clinical variables.

Several experimental interventions suggest that stromal-cell-derived EVs can alter cellular pathways associated with ovarian dysfunction. Umbilical-cord stromal-cell EVs reduced METTL3 expression, PINK1 messenger-RNA m6A modification and PINK1 transcript stability in DHEA-treated KGN cells and mice, suppressing excessive PINK1/Parkin-associated mitophagy and apoptosis. Overexpression of METTL3 partly reversed the effect. Adipose stromal-cell EV-associated miR-323-3p acted through PDCD4 to promote cumulus-cell proliferation and reduce apoptosis. Another study involving follicular-fluid EV-associated miR-18b-5p proposed PTEN suppression and PI3K/Akt/mTOR activation in KGN cells and letrozole-induced rats; treatment was associated with lower FSH, LH and testosterone and higher estradiol. These studies connect vesicles to defined mechanisms, but they do not show that unselected patient follicular EVs are uniformly protective.

Metabolic effects provide a second route of potential benefit. In PCOS rats, adipose stromal-cell EVs transferred miR-21-5p to the liver, reduced BTG2 and activated IRS1/AKT signalling; metabolic improvement accompanied changes in ovarian phenotype and fertility. This supports an extraovarian mechanism rather than selective ovarian delivery. Brown-adipose-tissue EVs administered to letrozole-induced mice were reported to improve insulin resistance, estrous cyclicity and fertility, with lower LH and testosterone, reduced ovarian STAT3 and increased GPX4. Menstrual-blood stromal-cell EVs improved mitochondrial biogenesis, oxidative-stress measures and estradiol-related endpoints in granulosa cells isolated from treated rats, but that experiment did not involve administration to intact animals or demonstrate fertility rescue. Differences in source, cargo, dose, route and model mean that similar endpoints across studies cannot establish equivalent efficacy.

Dose and delivery also complicate the therapeutic picture. In an estradiol-valerate mouse model, adipose stromal-cell-derived exosome-like nanovesicles were associated with a lower LH/FSH ratio, higher estradiol and changes in FOXO3, Map1lc3b and SF-1 expression. The strongest histological response occurred at 75 micrograms per kilogram, whereas 100 micrograms per kilogram was associated with inflammation and partial tissue degeneration. The observation cautions against assuming that greater exposure is necessarily safer or more effective. Engineered small EVs have also been used to improve delivery rather than to demonstrate general ovarian homing. A CD63-binding peptide, CP05, was used to attach the PI3K agonist 740Y-P to umbilical-cord stromal-cell sEVs; intravenous treatment in DHEA-induced mice improved reproductive and glucose-insulin endpoints relative to model controls. The study found no difference in ovarian internalization of modified and unmodified sEVs at the assessed time, so the result supports enhanced payload delivery rather than a general increase in EV targeting.

Evidence concerning fertility remains mixed and strongly dependent on biological context. Follicular EVs from nonhyperandrogenic PCOS patients impaired maturation of mouse oocytes exposed to the preparations, with increased oxidative stress and mitochondrial and spindle abnormalities. The selected phenotype and cross-species assay limit generalization, but the experiment demonstrates that patient-derived follicular vesicles may transmit harmful signals. By contrast, Wharton’s-jelly stromal-cell conditioned medium and EVs both improved selected in-vitro maturation outcomes in mouse oocytes, although conditioned medium generally produced greater benefit and significant restoration of maturation-related gene expression was reported mainly in conditioned-medium groups. Adjacent studies involving chemotherapy-related ovarian injury, premature ovarian insufficiency or thin endometrium are not direct PMOS treatment evidence. Any future strategy for PMOS-associated infertility would need to assess oocyte competence, endometrial receptivity and implantation rather than infer fertility from cell viability alone.

The review’s descriptive appraisal of five studies with accessible main texts highlights weaknesses in reporting against MISEV2023 recommendations. Two described morphology and particle size without an EV-associated protein panel. Two reported EV-associated proteins but did not describe a negative-marker or contaminant panel, while the engineered sEV study included positive and cellular negative markers but could benefit from broader cytosolic-marker and co-isolate assessments. These are reporting limitations, not proof that vesicles were absent; supplementary or inaccessible information could alter individual judgments. Relevant controls include EV-depleted fractions, medium processed without producer cells, cargo loss-of-function and rescue experiments, and labeling controls for uptake studies. Particle numbers and total preparation protein are also different dosing measures and should not be treated as interchangeable.

Translation also requires direct measurement of where administered vesicles travel and how long they remain active. Experimental biodistribution indicates substantial hepatic and splenic accumulation, with distribution influenced by cell source, dose, route and targeting strategy. Studies should therefore compare ovarian exposure with liver, spleen and other organs, assess retention and clearance, and include free-label controls. Engineered products require separate measurements of loading efficiency, removal of unbound drug, batch consistency and carrier-versus-cargo effects. Repeated-dose safety should address inflammation, endocrine perturbation, ovarian reserve, ectopic proliferation and offspring outcomes. The review’s central conclusion is that extracellular vesicles connect ovarian cell communication with systemic metabolism and offer biomarker candidates, inhibitory targets and experimental delivery systems, but their effects are source- and context-specific. Independent clinical validation, standardized characterization, quantitative biodistribution and long-term reproductive safety remain the decisive steps between today’s animal models and patient care.

Subject of Research: Extracellular-vesicle mechanisms, biomarkers and experimental therapies in polyendocrine metabolic ovarian syndrome

Article Title: Research progress of extracellular vesicles in PMOS: molecular mechanisms, diagnostic value, and therapeutic strategies

Article References: Qi, L., Jiang, Y., Han, Y., Feng, C., Jia, L., & Zhang, Y. (2026). Research progress of extracellular vesicles in PMOS: molecular mechanisms, diagnostic value, and therapeutic strategies. Frontiers in Endocrinology, 17, Article 1976561. https://doi.org/10.3389/fendo.2026.1976561

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1976561

Keywords: extracellular vesicles, polyendocrine metabolic ovarian syndrome, PCOS, ovarian dysfunction, biomarkers, mesenchymal stromal cells, granulosa cells, oocyte maturation, mitochondrial signaling, insulin resistance, Research, progress