The process of migrasome production can be visualized dynamically via time-lapse imaging techniques (6,16). Subsequent studies have confirmed the widespread presence of migrasomes across a variety of species, tissues, organs and cell types, including rat eyes, lungs, intestines, zebrafish embryos, chick chorioallantoic membranes and human coronary artery endothelial cells (6,11,17–20) (Fig. 2). As a novel type of organelle, migrasomes are characterized as membrane-bound vesicles with an ellipsoidal shape that harbor numerous smaller vesicles. Their composition primarily comprises proteins, lipids and nucleic acids (6,21). Notably, migrasomes exhibit a distinctive protein profile, which includes membrane proteins, contractile proteins, cytoskeletal proteins, chaperones, vesicular trafficking proteins and cell adhesion proteins, >50% of which are membrane-related (6,21,22). These proteins are engaged predominantly in biological processes, including cell migration, cell matrix adhesion, lipid degradation, protein glycosylation and glycoprotein metabolism (21).

In comparison with the cell membrane and overall lipid composition of the cell, migrasomes are notably enriched in sphingolipids, such as sphingomyelin (SM), ceramide, monosialodihexosylgangliosides and glycosphingolipids, including monoglycosylceramide, diglycosylceramide and triglycosylceramide. In a study by Liang et al (23), it was demonstrated that ceramide and SM are essential for the formation and maintenance of migrasomes. Furthermore, filipin III staining and quantitative analysis revealed that migrasomes are enriched in cholesterol, an important component for the physical properties and structural integrity of the migrasome membrane; notably, TSPAN4, TSPAN7 and cholesterol assemble into TSPAN-enriched microdomain (TEMs), the enrichment of which stiffens the plasma membrane, facilitating migrasome initiation (9). Ongoing lipidomics analysis of migrasomes anticipates the discovery of additional lipid components, thereby verifying the growing profile of known biological characteristics and functions of migrasomes.

Migrasomes are rich in nucleic acids. A study by Zhu et al (24) employed SYTO^™ 14 fluorescence staining to demonstrate the presence of RNA within migrasomes. Sequencing analyses revealed that migrasomes predominantly contain long-chain mRNAs associated with cell metabolism, intracellular membrane transport, cell adhesion, vesicle fusion and the assembly of subcellular membrane structures. These mRNAs can be translated into proteins within recipient cells, participating in the biological responses of recipient cells and regulating their life processes. For example, the phosphatase and tensin homolog (PTEN) mRNA delivered from migrasomes to recipient cells is translated into the PTEN protein, which inhibits the proliferation of the recipient cells (24). Nonetheless, the mechanisms of migrasome RNA sorting and transport have yet to be elucidated, as does the presence of DNA within migrasomes, necessitating further investigation.

TSPAN4/7, and integrins α1, α3, α5 and β1, which are expressed on the migrasome membrane, serve as important structural markers of migrasomes, with TSPAN4 being the most distinguishing marker, which exhibits the clearest expression when examined by confocal microscopy (6). Nonetheless, the detection of migrasomes using fluorescently labeled marker proteins presents limitations, including complex procedures, extended experimental duration and difficulty (25). Consequently, Chen et al (25) investigated fluorescently-labelled wheat germ agglutinin (WGA), which is a more rapid, straightforward and less invasive marker than TSPAN4. However, WGA may non-specifically bind to structures containing sialic acid and N-acetyl-D-glucosamine, and its fluorescence intensity can be influenced by various factors; for example, it may also enter the cell and bind to intracellular components, leading to enhanced nonspecific signals. In addition, different cell lines or tissue types naturally exhibit variations in the glycosylation levels of their cell membrane surfaces, which can also interfere with the results, potentially limiting its application in quantitative migrasome analysis (25). Additional specific protein markers include N-deacetylase/N-sulfotransferase 1, phosphatidylinositol glycan anchor biosynthesis class K, carboxypeptidase Q and epidermal growth factor domain-specific O-linked N-acetylglucosamine transferase; however, due to marked variations in protein content across different cell types, these markers are not uniformly present in migrasomes derived from the same cell source (21,22). Furthermore, migrasome-associated mRNAs can be labelled with the nucleic acid stain SYTO 14, making them secondary markers for migrasomes (24,26).

The formation of migrasomes is contingent upon cellular migration, a finding initially presented by Ma et al (6). A study revealed that, among 563 compounds capable of reducing migrasome numbers in cultured cells, 507 also decreased RF formation, reinforcing the association of migrasome production with cell migration (27). Subsequent research conducted by Fan et al (28) revealed a lower migrasome count in turning cells compared with those migrating in a straight line, highlighting the importance of migration continuity and speed in migrasome formation. Directional changes during migration lead to fewer RFs and migrasomes, with the removal of vimentin from cells having been shown to impair migration and reduce migrasome numbers (29,30). In conclusion, these findings suggested that the formation of migrasomes is markedly regulated by cell migration behavior.

SM, synthesized from ceramide by SMS, is a key component of the plasma membrane, and is involved in signaling and membrane transport (23). A previous study has shown that SM and ceramide are enriched on migrasomes and are present at the sites of migrasome formation; furthermore, ceramide is unevenly distributed on different migrasomes, and as migrasome biogenesis proceeds, SM levels continuously increase, indicating that ceramide can be converted into SM on migrasomes (23). Hydrolyzing SM on migrasomes or knocking out ceramide synthase 5 to reduce SM synthesis severely impairs the formation of migrasomes, demonstrating the importance of SM for migrasome formation (23).

Mammalian cells contain two principal types of SMS: i) SMS1, which is localized in the Golgi apparatus; and ii) SMS2, which is located in both the Golgi apparatus and the plasma membrane (31). SMS2 synthesizes SM from ceramide in the plasma membrane (23,31). Consequently, it is plausible that SMS2 modulates migrasome formation by facilitating SM production. Experimental evidence, including the knockout of SMS2 and treatment with SMS2 inhibitors that hinder SM synthesis, has demonstrated that migrasome formation and growth are impeded in SM-depleted conditions, whereas the reintroduction of SM restores migrasome production (23). Furthermore, impaired SM synthesis reduces cholesterol recruitment, thereby affecting TEM assembly, since TSPAN4, TSPAN7 and cholesterol assemble into TEMs, the enrichment of which stiffens the plasma membrane, which is an important condition for migrasome formation (9).

Research has revealed that SMS2 localizes to foci on the basal membrane at the leading edge of a cell, which predestines migrasome formation sites (23). These foci mature into migrasomes during cell migration (23). Intracellular SMS2 foci initially adhere to the section of the membrane of the migrating cell that is interacting with a substrate to generate motility, resembling focal adhesions (FAs), which are the sites where cells are linked to the ECM in which integrins are highly enriched. However, previous studies have failed to detect FA markers within migrasomes, and active forms or markers of FA components such as integrin α5, integrin β1 or the FA kinase paxillin do not colocalize with intracellular SMS2 foci, underscoring that SMS2 focus formation is independent of FAs (23,32). To elucidate the role of SMS2 foci in migrasome formation, researchers identified an SMS2 mutant, S217A, that cannot form foci. The formation of migrasomes in cells expressing this mutation has been shown to be markedly diminished (23). Notably, the inability to form SMS2 foci prevents migrasome formation, even when exogenous SM is added (23). These findings suggest that SMS2 foci not only regulate migrasome formation by synthesizing SM, but may also be involved in other important intracellular signaling processes that are required for migrasome biosynthesis and function. Future investigations should explore the precise mechanisms of SMS2 focus assembly, the selection of assembly sites and the adhesive mechanisms of SMS2 foci to fully elucidate the contribution of SMS2 foci to migrasome formation and function.

TSPANs constitute a family of small hydrophobic proteins characterized by four transmembrane domains, comprising a total of 33 distinct members in mammalian cells. These proteins facilitate the organization of functional higher-order protein complexes on the cell membrane through interactions with adhesion molecules, enzymes and signaling proteins, thereby forming the structures known as TEMs (33,34). Initial research identified TSPAN4 as a marker of migrasomes; however, subsequent investigations have suggested that the roles of TSPAN family members may extend beyond this initial characterization. By establishing a stable normal rat kidney (NRK) cell line that expresses various levels of TSPAN4 and green fluorescent protein (GFP), a study by Huang et al (9) demonstrated that the overexpression of 14 types of TSPAN family members enhances migrasome formation in a dose-dependent manner, with TSPAN4 exhibiting the most pronounced effect. Conversely, knockout of the TSPAN4 gene was shown to markedly reduce the number of migrasomes, underscoring the importance of TSPAN4 for migrasome formation. During the migrasome formation process, the signal from TSPAN4-GFP during the growth phase of migrasomes rapidly re-emerges on their surface, whereas no such recovery occurs when migrasomes tend to mature and stabilize (9). Furthermore, high-speed imaging revealed that TSPAN4-GFP forms rapidly-moving-discrete spots on RFs and migrasomes that assemble on the surface of migrasomes during their growth phase. Taken together, these findings suggest that TSPAN4 is recruited to migrasomes specifically during the growth phase (9).

By developing an in vitro system to simulate migrasome formation, researchers have demonstrated that TSPAN4 and cholesterol are sufficient to reconstitute migrasome-like structures, further validating their role in this process (9). To elucidate how the assembly of TEMs promotes migrasome formation, a theoretical model was proposed, identifying three key energetic drivers: i) The bending energy of the migrasome and RF membranes; ii) the membrane tension energy; and iii) the boundary energy at the migrasome-RF interface (9). A subsequent study using a biomimetic system divided migrasome growth into two phases: i) A local swelling phase, driven by membrane tension and potentially other cellular factors (such as the cytoskeleton, specific lipids, ion fluxes, mechanosensitive signaling and adhesion complexes); and ii) a TSPAN4 migration-enrichment phase mediated by TSPAN family proteins (35). Notably, the biomimetic system lacks cytoskeletal components, which may regulate migrasome formation in vivo; therefore, discrepancies between artificial and cellular systems may exist (6,17,21,22). TSPAN4, TSPAN7 and cholesterol assemble into TEMs, whose enrichment stiffens the plasma membrane, facilitating migrasome initiation. Changes in TSPAN4 curvature will affect the swelling of the plasma membrane, the reduced intrinsic curvature of TSPAN4 directs TEMs to low-curvature membrane swellings, stabilizing these protrusions and promoting migrasome maturation (9,10,18,35,36). In summary, TEMs composed of TSPAN4 and cholesterol are important for migrasome biogenesis.

PIP2, a plasma membrane lipid synthesized by phosphatidylinositol 4-phosphate 5-kinase type-1α kinases, regulates key subcellular processes, including the regulation of ion channels and transporters, clathrin-dependent and -independent endocytosis, exocytosis regulation, actin polymerization, efficient FA turnover and the regulation of cell-cell contacts) (39,40). PIP2 localizes to migrasomes, as confirmed by phospholipase Cγ-PH domain-GFP fusion protein probes and antibody staining (41). Kinetic experiments have revealed that PIP2 recruitment precedes TSPAN4 and integrin α5 recruitment, both of which are important for migrasome biogenesis. Phosphatidylinositol-4-phosphate 5-kinase type-1α (PIP5K1A) inhibition disrupts migrasome formation, implicating PIP2 synthesis in this process (41).

PIP2 likely functions by recruiting binding partners, such as Rab35, a GTPase involved in organelle biogenesis, endosomal trafficking and actin regulation (42). Rab35 is recruited to migrasome formation sites in a PIP2-dependent manner, and its depletion disrupts RF elongation and migrasome assembly (41). Rab35 also interacts with integrin α5 via the GFFKR motif, recruiting integrins α5 to migrasome sites, although the direct binding mechanism remains unconfirmed (41,43). Thus, the PIP2-Rab35 axis orchestrates migrasome formation by coupling lipid signaling with integrin trafficking. Future studies should address PIP5K1A recruitment dynamics, Rab35-integrin α5 interactions and the clinical relevance of the PIP2-Rab35 signaling axis in cancer and migration-associated diseases.

Through a chemical genetic screen designed to identify compounds and protein targets that disrupt migrasome formation, a study by Lu et al (27) identified SAR407899, an inhibitor of ROCK1 and ROCK2, which stably suppresses migrasome biogenesis. ROCK1 and ROCK2 are serine/threonine kinases that act downstream of the small GTPase Ras homolog family member A (RhoA), regulating diverse cellular processes, including actin cytoskeleton organization, cell adhesion and migration, via the ROCK-Rho signaling pathway (44). Although ROCK2 knockdown does not affect migrasome formation, ROCK1 depletion markedly reduces the migrasome abundance per cell (27). Notably, ROCK1 knockdown also impairs cell migration. To distinguish the effects of RFs and migration from migrasome formation, researchers have quantified the number of migrasomes per 100 µm RFs (27). The findings of this experiment revealed that ROCK1-depleted cells generate notably weaker traction forces than healthy cells, supporting the premise that migrasome formation depends on ROCK1-mediated traction and other ECM components (such as fibronectin) adhesion (27).

PD-L1 is best known as an immune checkpoint molecule that facilitates cancer cell migration (45). Emerging evidence has indicated that PD-L1 has an intrinsic capability to facilitate sustained cellular migration, a process important for migrasome biogenesis. High-resolution time-lapse microscopy has demonstrated that PD-L1 accumulates at the trailing edge of migrating cancer cells, where it forms distinct structures that move directionally during retraction (46). Given that RFs form at the rear of the cell during migration, a subsequent study demonstrated that PD-L1 localizes not only in RFs, but also in migrasomes. Notably, PD-L1 enhances migrasome formation independently of cell migration (46). PD-L1 closely associates with integrin β4, co-localizing at the rear of the cell, and later in RFs and migrasomes. A further investigation revealed that PD-L1 recruits integrin β4 to the trailing edge; this recruitment stimulates contractility via the dynamics of the cell, a mechanism by which PD-L1 maintains rear polarity and reduces membrane tension (46). Additionally, PD-L1 activates RhoA by coupling integrin β4 to the cytoskeleton, further promoting rear contraction (actin cytoskeleton-mediated contractility) (46). Taken together, these findings underscore the dual role of PD-L1 in facilitating cell migration and migrasome formation. However, the precise mechanisms of PD-L1 in migrasome biogenesis, and its broader functional implications, warrant further exploration.

Migrasomes constitute a notable secretory pathway in migrating cells, encapsulating diverse cytoplasmic contents (6,51). Migrasomes form and mature on RFs of migrating cells before they detach, rupture and release their contents (6) (Fig. 3A). Migrasomes can also be internalized by neighboring cells, or adhere to cell surfaces or the ECM without being engulfed. For example, interactions with the ECM enable migrasomes to attach to specific cell surfaces or tissues, facilitating intercellular communication (32). These vesicles transport intracellular biomolecules, including proteins, mRNAs and microRNAs. Recent studies have indicated that secretory proteins, such as signaling molecules, are actively transported from migrating cells into migrasomes via kinesin-mediated carriers, akin to targeted neurotransmitter release in neuronal systems (51). These molecules can be transferred to adjacent or distant cells, altering recipient cell-gene expression and function. For example, Zhu et al (24) added purified migrasomes derived from L929 cells to U87-MG, MDA-MB-468 and PC3 cells that do not express the PTEN protein, and demonstrated that migrasomes mediate the transfer of PTEN mRNA and protein, inhibiting proliferation in recipient cells.

Migrasomes are enriched with cytokines, including chemokines and morphogens, which enables them to act as signaling molecule-carriers, and thereby influence cell behavior. In zebrafish embryos, gastrula-derived migrasomes contain high levels of C-X-C motif chemokine ligand (CXCL)12a (18). These migrasomes, produced during mesoderm and endoderm cell gastrulation, regulate organ development. Mutations that reduce migrasome numbers lead to severe developmental defects, which can be rescued by migrasome supplementation, underscoring their role in organogenesis (18). Further studies have revealed that CXCL12 in migrasomes interacts with C-X-C chemokine receptor type 4 (CXCR4) on dorsal precursor cells (DFCs), inducing chemotaxis and ensuring typical organ morphogenesis (35). Migrasomes thus spatially and temporally distribute signaling molecules, releasing CXCL12 upon rupture to modulate DFC behavior. Similarly, adipose-derived stem cells produce CXCL12-enriched migrasomes that promote adipose tissue regeneration via CXCR4/RhoA signaling (50).

A study by Zhang et al (11) identified migrasome-producing monocytes in the chorioallantoic membrane of chicken embryos, where these vesicles are rich in proangiogenic factors, including TGF-β3, VEGFA and CXCL12. In this context, migrasomes induce endothelial tube formation; their depletion via monocyte clearance or TSPAN4 knockout inhibits angiogenesis, whereas supplementation restores monocyte recruitment and capillary formation (11). Monocytes leverage migrasomes to deliver angiogenic factors, thereby creating a microenvironment conducive to blood vessel growth. Notably, migrasome-derived CXCL12 recruits additional monocytes, forming a positive feedback loop (11,52).

Multipotent mesenchymal stem cells (MSCs) are precursors to various cell types, with the ability to support tissue homeostasis, promote hematopoiesis and interact with cancer cells. Previous research on MSCs has focused on MSC-derived EVs (MSC-EVs), indicating that MSC-EVs have functions similar to those of MSC (53–58). A recent investigation demonstrated that MSCs generate migrasomes that attract leukemia KG-1a cells and CD34⁺ hematopoietic stem cells via the CXCR4-CXCL12 axis (59). These migrasomes, enriched in TSPANs, including CD166 and TSPAN4, and endosomal markers, such as Rab7 and CD63, are influenced by ECM components such as fibronectin and laminins. In addition to influencing cell migration and thereby affecting the formation of retraction fibers, ECM components can also impact the process of TEM formation (59). Migrasomes produced by MSCs attract leukemia KG-1a cells and CD34⁺ hematopoietic stem cells through the CXCR4-CXCL12 axis, thereby facilitating communication between the MSCs and these cells, thus highlighting the role of migrasomes in MSC-hematopoietic cell interactions, offering insights into their functions in health and disease.

Neutrophil-derived migrasomes adsorb coagulation factors from the plasma; migrasomes actively bind and enrich coagulation factors on their surface by virtue of their unique membrane composition, particularly cholesterol esters, and then localize to injury sites and modulate coagulation by activating platelets (60). These migrasomes also regulate immune responses by delivering cytokines and chemokines to injury sites, enhancing immune cell recruitment during inflammation (60). Conversely, migrasomes have been shown to suppress immunity by transporting immunosuppressive molecules, thereby maintaining immune tolerance (46).

Mitochondria, the cellular ‘powerhouses’, maintain homeostasis through energy production and quality control mechanisms, such as mitophagy (61,62). A previous study demonstrated that migrasomes mediate mitocytosis, a process in which damaged mitochondria are selectively expelled to sustain cellular health (12) (Fig. 3C). Under stress conditions, such as carbonyl cyanide m-chlorophenyl hydrazone treatment, mitochondria undergo fragmentation via kinesin-5B (KIF5B)-mediated transport and myosin head domain-containing protein 1 (MYO19)/density-regulated protein 1-dependent fission, accumulating at the cell periphery for subsequent migrasome encapsulation. Knocking out TSPAN9 or taking other measures to block the formation of migrasomes has been shown to cause migrating cells to lose their mitocytosis capability, leading to the accumulation of damaged mitochondria within the cells and affecting cell viability (12,63,64).

Although both mitocytosis and mitophagy contribute to cellular homeostasis, their mechanisms differ markedly. Mitocytosis, which is mediated by migrasomes, selectively removes mildly damaged mitochondria from migrating cells. In this process, damaged mitochondria selectively bind to intracellular dynein to facilitate their transport out of the cell, they are then transported towards the cell periphery by outwards motor proteins, such as KIF5B. MYO19 further facilitates this process by anchoring mitochondria to cortical actin, thereby promoting their incorporation into migrasomes (12,63,64). By contrast, mitophagy primarily degrades damaged mitochondria via the autophagy pathway. Upon mitochondrial damage, PTEN-induced kinase 1 accumulates on the outer mitochondrial membrane, where it recruits parkin to ubiquitinate outer membrane proteins. This ubiquitination marks the mitochondria for autophagosomal engulfment, and subsequent lysosomal degradation (61,62).

Mitocytosis complements mitophagy by incrementally clearing mildly damaged mitochondria. Migrasomes may also transfer mitochondrial components, including mitochondrial DNA and mitochondrial microRNAs, to recipient cells, influencing their function.

Migrasomes facilitate viral dissemination, facilitating the evasion of antiviral therapies (Fig. 3B). Vaccinia virus induces migrasomes containing viral particles, enabling their spread despite tecovirimat treatment (13,65). Similarly, herpes simplex virus-2 exploits migrasomes as ‘Trojan horses’ for intercellular transmission (66). Taken together, these findings illuminate novel viral spread mechanisms and suggest potential targets for antiviral strategies.