The concept of dynamic balance regulation deserves particular attention because it distinguishes MSCs from nearly every other class of therapeutic agent in clinical use. Conventional small molecules and biologics are designed around fixed pharmacological targets; an anti-inflammatory drug lowers inflammatory mediators whether or not inflammation is still useful to the host, and a growth factor stimulates proliferation regardless of the surrounding context. MSCs, by contrast, appear to read the local chemical and mechanical cues of their surroundings and adjust their secretome accordingly. In a pro-inflammatory, oxidative environment, they tend to release factors that suppress T-cell proliferation, polarize macrophages toward reparative phenotypes, and induce regulatory T-cell activity. In a quiescent or regenerative setting, the same cells can promote angiogenesis, matrix deposition, and progenitor cell expansion. This context sensitivity, sometimes described in the literature as a yin-yang mechanism, means that a single cell population can theoretically both calm a destructive immune storm and then support reconstruction once the danger has passed, a sequence that no single conventional drug can currently execute.

The observation that transplanted MSCs rapidly undergo apoptosis in the host environment, and that this programmed cell death is itself therapeutically active, is among the most consequential findings highlighted in the review. Apoptotic extracellular vesicles, or apoVs, are membrane-bound particles released by cells in the final stages of programmed death. Far from being cellular debris, these vesicles carry a curated cargo of proteins, lipids, nucleic acids, and organelle fragments that reflect the donor cell’s regulatory state at the moment of death. Studies across multiple tissue systems have shown that apoVs can reproduce many of the immunomodulatory and pro-regenerative effects previously attributed to living MSCs, including suppression of pathological inflammation, promotion of macrophage efferocytosis, and stimulation of endogenous stem cell activity in tissues such as bone, dental pulp, and skin. The recognition that the dying cell is, in effect, packaging its regulatory wisdom into vesicles reframes apoptosis not as a failure of therapy but as an essential phase of it.

This insight carries practical implications for how future therapies may be designed and manufactured. Living cell products face formidable obstacles: donor-to-donor variability, the risk of uncontrolled proliferation or ectopic engraftment, sensitivity to cryopreservation and transport, senescence during ex vivo expansion, and the logistical burden of maintaining viable cells until the moment of administration. Cell-free vesicle preparations, by contrast, can in principle be produced under controlled bioreactor conditions, characterized biochemically, sterilized, stored at defined temperatures, and dosed in a manner analogous to conventional biologics. They also sidestep some immune-compatibility concerns, since vesicles present a less complex antigenic surface than intact cells. The review’s emphasis on apoVs therefore points toward a manufacturing paradigm in which the therapeutic product is not the cell itself but the standardized output of a carefully orchestrated cell culture process, harvested at the apoptotic stage when the vesicle payload is richest.

Quality control and standardization will determine whether this promise survives contact with regulatory reality. The field of extracellular vesicle research has historically been hampered by heterogeneous isolation methods, including ultracentrifugation, size-exclusion chromatography, polymer-based precipitation, and ultrafiltration, each of which yields preparations with different purity, particle size distributions, and residual contaminant profiles. The MISEV guidelines, most recently updated in the 2023 edition, were developed by the International Society for Extracellular Vesicles to address exactly this problem, prescribing minimal experimental requirements for characterizing vesicle preparations, reporting particle and protein ratios, and avoiding overclaiming about vesicle-specific mechanisms when preparations may contain non-vesicular contaminants. For a therapeutic modality seeking regulatory approval, adherence to such frameworks is not optional; it is the difference between a reproducible medicine and an irreproducible biological extract. The review’s insistence on stringent Chemistry, Manufacturing, and Controls standards reflects an awareness that early inconsistencies in the MSC literature, where poorly characterized products produced conflicting trial results, damaged the field’s credibility for years.

The regulatory pathway itself deserves closer examination. In Europe, MSC-based products have been approved under the advanced therapy medicinal product framework, which demands full characterization of the cell source, manufacturing process, and clinical benefit-risk profile. Japan has taken a somewhat different route, with conditional and time-limited approvals that allow earlier market entry contingent on post-marketing evidence generation. The United States Food and Drug Administration, by contrast, has generally required that MSC products navigate the investigational new drug and biologics license application pathway, treating them as biological products subject to the full weight of clinical trial evidence requirements. This transatlantic divergence in regulatory philosophy has practical consequences: products available to patients in one jurisdiction may remain unavailable in another, and developers must design manufacturing and clinical strategies that can satisfy the strictest of the standards they intend to target. For MSC-derived vesicles, which are even newer as a product category, the regulatory template is still being written, and early engagement with agencies will likely shape the standards that later entrants must meet.

The integration of artificial intelligence into this therapeutic landscape, mentioned in the source material as a direction for optimized treatments, aligns with broader trends in biomedical engineering. Machine learning approaches are increasingly applied to predict donor cell potency from omics data, to identify culture conditions that maximize the yield and consistency of vesicle production, and to model the dose-response behavior of complex biological therapeutics that defy classical pharmacokinetic description. Because MSC regulatory effects are inherently multi-factorial and context-dependent, they are well suited to computational approaches that can integrate hundreds of variables simultaneously. An AI-optimized pipeline could, in principle, screen candidate donor lines for secretory profiles associated with specific disease indications, tune bioreactor parameters to steer vesicle cargo toward desired compositions, and even predict which patient subpopulations are most likely to respond based on the inflammatory state of their own microenvironment. Such stratification would address a persistent weakness of MSC trials to date, namely the enrollment of heterogeneous patient populations in which average treatment effects may mask meaningful responses in identifiable subgroups.

The clinical indications where this paradigm may first mature are those in which chronic inflammation and dysregulated repair are central features. Periodontal disease, craniofacial defects, and other oral-craniofacial conditions, areas where the research group behind the review has deep expertise, offer tractable models because the tissues are accessible, the inflammatory microenvironment is well characterized, and endogenous progenitor populations are known to persist in adult structures. Beyond the craniofacial domain, autoimmune disorders, graft-versus-host disease, inflammatory bowel disease, and various fibrotic conditions of the liver, lung, and kidney represent settings in which the immunomodulatory face of MSC activity is most relevant, while bone defects, wound healing, and ischemic tissues call upon the pro-regenerative face. The dynamic balance concept suggests that the same therapeutic platform could eventually serve across this spectrum, with the host microenvironment, rather than the product label, determining which regulatory program is activated.

Several scientific questions remain open as the field advances. The precise molecular identity of the factors that allow MSCs to sense their microenvironment in real time is only partially mapped; candidate mechanisms include toll-like receptor activation, cytokine receptor signaling, and mechanical sensing through integrins and the cytoskeleton. The dose-response relationship for vesicle therapies is similarly undefined, since vesicles do not behave like classical drugs and their persistence, biodistribution, and cellular uptake in vivo are still being characterized. Questions of source selection also persist: MSCs derived from bone marrow, adipose tissue, dental pulp, umbilical cord, and other tissues differ in secretory profile, immunogenicity, and expansion behavior, and no consensus has emerged on which source is optimal for which indication. Finally, the long-term safety of repeated vesicle administration, including any potential for immunogenicity against vesicle-associated antigens or unintended modulation of immune surveillance, will require dedicated study before chronic use can be contemplated.

What the review ultimately conveys is a maturing discipline that has moved past its initial reductionist assumptions and arrived at a more accurate, and more useful, understanding of how cell-based therapeutics actually work. The shift from viewing MSCs as replacement parts to viewing them as environmental engineers, and from transplanting living cells to deploying their standardized vesicular messengers, mirrors a broader evolution in medicine toward therapies that modulate systems rather than substitute for them. If the manufacturing, characterization, and regulatory frameworks described in the review can be established in practice, the result may be a class of treatments that harness the body’s own regenerative plasticity with a precision and safety profile that living cell transplants never quite achieved. The next decade of clinical translation will reveal whether this conceptual clarity, hard-won over decades of sometimes contradictory evidence, can be converted into reliable therapies for the complex diseases that have thus far resisted conventional approaches.

Subject of Research: Beyond cell transplants: The next frontier in mesenchymal stem/stromal cell therapy and regenerative medicine

Article Title: Beyond cell transplants: The next frontier in mesenchymal stem/stromal cell therapy and regenerative medicine

Article References: Beyond cell transplants: The next frontier in mesenchymal stem/stromal cell therapy and regenerative medicine. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: Mesenchymal stem cells, Regenerative medicine, Extracellular vesicles, Cell-free therapy, Immunomodulation, Tissue repair, Stem cell therapy, Science Bulletin, Apoptotic bodies, Clinical translation, Beyond, cell

The case for treating large extracellular vesicles as their own category rests on a fundamental difference in origin. Small extracellular vesicles arise from two well-characterised sources: the endosomal system, which produces exosomes when multivesicular bodies fuse with the plasma membrane, and the plasma membrane itself, which buds off small ectosomes. Most large extracellular vesicles, by contrast, stem directly from the plasma membrane, although the review emphasises that the various subtypes follow strikingly different biogenesis routes. This distinction matters because it shapes what the vesicles carry, how they are released, and what they do once they arrive at a recipient cell. A vesicle born from a dying cell’s collapsing membrane carries a very different cargo than one pinched off from the rear of a migrating cell or extruded from a neuron under proteotoxic stress.

The classical cast of large vesicles includes apoptotic bodies, microvesicles and large oncosomes, each recognised for years and each associated with distinct biology. Apoptotic bodies form as dying cells disassemble themselves, a process driven by caspase-mediated activation of ROCK1, which triggers the actomyosin contraction underlying membrane blebbing. These fragments are far more than debris: they broadcast find-me and eat-me signals that orchestrate the orderly clearance of dead cells by phagocytes, and recent work shows they can stimulate stem cell proliferation during tissue maintenance and couple bone resorption to bone formation. Microvesicles, shed directly from the plasma membrane of viable cells, depend on the ARF6–myosin light chain–actomyosin axis for their scission, and tumour cells exploit this pathway to spread oncogenic receptors such as EGFRvIII and drug-resistance proteins like P-glycoprotein to neighbouring cells. Large oncosomes, the giant vesicles shed by cancer cells, have been linked to loss of the formin DIAPH3 and to the reprogramming of fibroblasts in the tumour microenvironment through MYC-dependent mechanisms.

The most exciting chapters of the review concern the newer subtypes, several of which were unknown when the extracellular vesicle field consolidated its nomenclature under the MISEV guidelines. Migrasomes, discovered in 2015 by Li Yu’s group, form on retraction fibres left behind by migrating cells and serve as organelles for controlled release of cytoplasmic contents during cell migration. Their formation depends on the assembly of micron-scale tetraspanin-enriched macrodomains, driven by sphingomyelin synthase 2, phosphatidylinositol 4,5-bisphosphate signalling through Rab35, and palmitoylation of the tetraspanin Tspan4. Migrasomes have since been implicated in zebrafish organ morphogenesis, embryonic angiogenesis, mitochondrial quality control through a process termed mitocytosis, and even coagulation, with neutrophil-derived migrasomes identified as essential components of the clotting system.

Equally remarkable are the vesicles that cells use to jettison damaged components. Exophers, first described in Caenorhabditis elegans neurons in 2017, are enormous vesicles through which stressed neurons extrude toxic protein aggregates and dysfunctional mitochondria, providing a proteostasis escape route with obvious implications for neurodegenerative disease. Mitophers, reported in 2023, represent a form of mitochondria-specific ectocytosis through which sperm cells regulate their mitochondrial complement and, consequently, their fertility, establishing a paradigm for cargo-selective packaging of entire organelles into large vesicles. Blebbisomes, characterised in 2025, are large, organelle-rich vesicles with startling cell-like properties, carrying broad organellar cargo and expanding the functional repertoire attributed to vesicle-mediated communication. Large ageing neutrophil-derived vesicles, or LAND-Vs, add an anti-inflammatory dimension: they control complement activation and facilitate the resolution of inflammation, revealing that large vesicles can dampen immune responses as well as provoke them.

Beneath this diversity lies a set of shared physical and biochemical principles that the review dissects with technical precision. Membrane budding and scission require the coordination of actomyosin contractility, hydrostatic pressure and lipid remodelling. The transbilayer distribution of phospholipids emerges as a recurring regulatory node: flippases maintain lipid asymmetry, while scramblases such as TMEM16F collapse it in a calcium-dependent manner, exposing phosphatidylserine on the outer leaflet and promoting both vesicle release and the engulfment signals that mark apoptotic material for clearance. Disruption of flippase localisation in C. elegans increases extracellular vesicle budding, demonstrating that lipid asymmetry actively suppresses vesiculation in healthy membranes. Protein crowding at the membrane, tetraspanin microdomains, lipid rafts and caveolin-dependent dynamics each contribute to curvature generation at specific sites, while the ESCRT machinery contributes its membrane-scission capabilities to several of these pathways.

Cargo sorting into large vesicles follows equally mechanistic rules. The ARF6–Exportin-5 axis delivers pre-microRNA cargo into tumour microvesicles, Cdc42 acts as a regulatory node for microvesicle biogenesis, and RNA-binding proteins govern which transcripts are packaged. Zipcode-like sequence elements enrich specific mRNAs in microvesicles, and integrin pairing with extracellular matrix proteins determines where migrasomes form on retraction fibres. Delivery to recipient cells proceeds through clathrin-independent endocytosis and macropinocytosis, although the review is careful to note that the field still lacks definitive experimental support for functional cargo transfer in many contexts, a caution that reflects growing scrutiny of the extracellular vesicle cargo-transfer hypothesis.

The pathological implications are broad and, in several cases, already translating toward clinical applications. In cancer, large vesicles transfer oncogenic receptors, promote angiogenesis through CD147 and VEGF signalling, carry retrotransposon elements and amplified oncogene sequences, and mediate immune evasion through vesicle-associated PD-L1, which Epstein–Barr virus LMP1 enhances. In neurodegeneration, microglia convert aggregated amyloid-beta into neurotoxic forms through microvesicle shedding, while macrophage-derived migrasomes activate complement-dependent blood–brain barrier damage in cerebral amyloid angiopathy and promote paracrine senescence in ageing brain. Inflammation, thrombosis and kidney injury each feature large-vesicle mechanisms, with urinary podocyte-derived migrasomes emerging as early indicators of glomerular damage and plasma apoptotic bodies serving as non-invasive biomarkers of apoptosis in ischemic stroke.

Methodological rigour occupies a central place in the review’s argument. The authors catalogue the technologies now enabling multidimensional characterisation, from cryo-electron microscopy and super-resolution structured illumination microscopy to nanoparticle tracking analysis, tunable resistive pulse sensing, asymmetric flow field-flow fractionation and microfluidic affinity capture. Deep-learning platforms for automated detection of vesicle exocytosis and self-supervised image enhancement for long-term subcellular recording are extending what live imaging can reveal. Yet the review is equally candid about pitfalls: fluorescent lipophilic dyes can confound uptake studies, isolation methods such as ultracentrifugation and tangential flow filtration yield different preparations, and loss-of-function studies and mechanistic analyses must be appropriately integrated to ensure rigour and reproducibility. Engineered migrasomes that serve as thermally stable vaccination platforms hint at the therapeutic horizon, but the authors stress that biodistribution depends on cell source, administration route and targeting ligands, and that the field’s progress will depend on pairing new tools with the disciplined standards set out in the MISEV2023 framework. What emerges is a portrait of a maturing discipline: one in which the largest and most structurally complex members of the extracellular vesicle family are finally being studied with the systematic attention their biology demands.

Subject of Research: Subtypes, biogenesis and functions of large extracellular vesicles

Article Title: Towards a systematic understanding of the diverse subtypes, biogenesis and functions of large extracellular vesicles

Article References: Huang, Y., Xie, R., Ma, J., & Yu, L. (2026). Towards a systematic understanding of the diverse subtypes, biogenesis and functions of large extracellular vesicles. Nature Reviews Molecular Cell Biology. https://doi.org/10.1038/s41580-026-01021-8

Image Credits: AI Generated

DOI: 10.1038/s41580-026-01021-8

Keywords: large extracellular vesicles, migrasomes, apoptotic bodies, microvesicles, large oncosomes, exophers, blebbisomes, mitophers, plasma membrane budding, tetraspanins, extracellular vesicle biogenesis, cell-cell communication