Accumulating evidence has demonstrated a strong link between glucose homeostasis and the growth of breast cancer cells. Cao et al. have demonstrated that miR-122, an EVs cargo derived from breast cancer cells, inhibits glycolysis and ATP-dependent insulin exocytosis and disrupts the critical function of islets to maintain normal blood glucose levels. This effect further suppresses insulin secretion, augments endogenous glucose production, impairs glucose tolerance, and disrupts fasting hyperglycemia. As a result, all these effects on glucose homeostasis could promote the rapid growth of breast cancer (5, 14, 15). In addition, breast cancer cells have been found to deliberately deviate trajectories during migration with the help of EVs and hyaluronic acid. This explains the role of EVs and hyaluronic acid in guiding the migration and metastasis of tumor cells (16). A study by Eckhard et al. confirmed the role of exosomal Rab27a in breast cancer development and metastasis, which indicated that targeting Rab27a reduced the release of exosomes, thus retarding breast cancer growth (17). Wanessa et al. have demonstrated that adhesion receptors, previously thought to be expressed only at the cellular level, are additionally present on SEVs membranes and thus participate in processes such as adhesion and uptake. Specifically, αvβ3 integrins, adhesion receptors present on SEVs, secreted by triple-negative breast cancer cells (MDA -MB-231) have been reported to accelerate invasion and organ-specific metastasis of breast cancer cells. Moreover, the downregulation of αvβ3 integrin expression further reduces the expression of the pro-oncoprotein CD63 present on the SEVs membrane, thereby inhibiting breast cancer progression (18). A in vitro study by Bertolini et al. highlighted that SEVs released from breast cancer cells activated NF-κB pathway in breast epithelium cells under hypoxic conditions, then promoted the release of inflammatory cytokines and mitochondria-mediated cell migration, which enabled the disruption of the mammary glandular vesicle structure. Simultaneously, this hypoxic microenvironment in patients with breast cancer commonly contributed to local carcinogenic and oncogenic inflammatory changes (19). Another study has found that Caveolin-1 is upregulated in metastatic breast cancer cell-secreted EVs and fosters breast cancer invasion and metastasis through the adhesion proteins (20). Besides, Wu et al. found that the bone marrow mesenchymal stem cell (BMSC)-derived EVs could be internalized by breast cancer cells and significantly improve the proliferation ability of these breast cancer cells (21). Adipocytes are the most abundant stromal cell component in breast cancer tissues and support the survival of metastatic breast cancer cells at distant sites (22). Reportedly, exosomes from adipose tissue-derived MSCs are internalized by breast cancer cells (MCF7), which activate the Hippo pathway to facilitate the proliferation and migration of breast cancer cells (23). Genomic instability (GI) is a major driver of tumorigenesis (24, 25). Specifically, genomic changes tend to affect cellular phenotypes and are evolutionary hallmarks of most cancers (26). As illustrated in the work of Siqi et al., three miRNAs (miR-421, miR-128-1, and miR128-2) were identified from 18 GI-associated miRNAs contained in EVs from serum of breast cancer patients, which were defined as GI-derived miRNA signature (miGISig). By analyzing its correlation with patient clinicopathological data from publicly available breast cancer -associated datasets (such as TCGA) and in vitro functional experiments, miGISig was found to be closely associated with poor prognosis of patients, indicating its high diagnostic potential for breast cancer (27). These results suggest that EVs derived from breast cancer cells can stimulate the growth and migration of tumor cells, which in turn facilitates the malignant progression of tumors.

Angiogenesis, the formation of new blood vessels from pre-existing blood vessels, is an important cause of tumor growth and blood-borne metastasis. This is a dynamically complex process involving multiple mechanisms, regulated by a variety of cancer related molecules (28). In healthy tissues, angiogenesis is tightly regulated by a precise balance between stimulating and inhibiting signals. When this balance is disturbed, abnormal vascular growth occurs and is a major cause of breast cancer invasion and metastasis (29). EVs are important mediators of angiogenesis through constant communication with the environment through a variety of para-secretory factors and cell-to-cell and cell-to-matrix interactions (30). Moreover, EVs released by primary tumors also promote angiogenesis and cancer progression. Aslan et al. have found that docosahexaenoic acid downregulates the expression of pro-angiogenic genes, such as HIF1-α, TGF-β, SOX2, Snail1, Snail2, and VEGFR, in breast cancer cell-derived exosomes. Furthermore, the docosahexaenoic acid inhibits tumor angiogenesis by upregulating tumor-suppressing miRNAs (miR-101, miR-199, and miR-342) and downregulating oncomiRs (miR-382 and miR-21) in exosomes (31). In addition, MSC-derived EVs repress the expression of vascular endothelial growth factor (VEGF) by miR-100/mTOR/HIF1α signal axis, ultimately retarding the formation of capillary-like tubules in endothelial cells (32). Meanwhile, MSC-derived exosomes deliver miR-16 to repress the expression of its target VEGF, thus curtailing angiogenesis in breast cancer (33). Sayantan et al. have demonstrated that, in primary breast cancer cell-derived SEVs, annexin II induces angiogenesis both in vitro and in vivo by upregulating the expression of the tissue fibrinogen activator tPA, which in turn promotes lung and brain metastases of breast cancer (34). Moreover, breast cancer cell-derived exosomes stimulate endothelial cell angiogenesis in the TME via the circHIPK3/miR-124-3p/MTDH axis (35). These data suggest that breast cancer cell-derived EVs can stimulate angiogenesis in the TME, which in turn promote breast cancer development and progression. Although many studies have demonstrated the effect of EVs on angiogenesis, based on the presented findings and clinical experiences, we can easily imply that another severe adverse effect associated with EVs administration may be an increased risk of thrombosis. Therefore, further preclinical models are needed to better evaluate the risk and benefit and more attention should also be paid to this point.

Cancer metastasis can be promoted by forming a supportive TME at secondary organ sites, called PMN, wherein the EVs secreted by tumor cells can modulate TME changes (36). Busatto et al. have suggested that EVs secreted from brain metastatic cells of breast cancer induce low-density lipoprotein aggregation, which accelerates uptake of these metastatic breast cancer cells by monocytes, and promotes the formation of a PMN in the brain (37). Reportedly, EVs secreted by the triple-negative breast cancer cell lines MDA-MB-231 and SUM159 have been found to induce lung fibroblasts to express the extracellular matrix (ECM) proteins fibronectin, tenascin-c, and periostin, thereby contributing to the formation of PMN in the brain (38). Moreover, LC3-loaded EVs released by breast cancer cells could induce lung fibroblasts to produce CCL2 by the HSP60-TLR2-MyD88-NF-κB pathway. Then, CCL2 recruits monocytes to the lung and subsequently form a pulmonary PMN characterized by monocytes, high infiltration of macrophages, immunosuppression, and enhanced vascular permeability, ultimately accelerating the formation of breast cancer lung metastasis (39). Lin28B has been considered a potent inducer of breast cancer metastasis, and neutrophil N2 conversion is a key step in the suppression of T-cell function. Specifically, Qi et al. have found that Lin28B is involved in neutrophil recruitment and N2 conversion through let-7s-carrying exosomes to alter the immune status in the TME, thereby establishing an immunosuppressive PMN and ultimately inducing lung metastasis of breast cancer (40). According to the findings of Wu et al., miR-19a and integrin-binding sialoprotein (IBSP) expression in breast cancer cells fail to alter the in vitro growth or migration ability of breast cancer cells. However, following encapsulation and release by exosomes, IBSP can attract osteoclasts and form an osteoclast-rich microenvironment to assist the delivery of exosomal miR-19a to osteoclasts, thus further inducing osteoclast formation. This subsequently augment PMN formation and bone metastasis in estrogen receptor (ER)-positive breast cancer (41). As previously reported, exosomes secreted by the mildly metastatic MDA-MB-231 cell line and the highly metastatic SCP28 breast cancer cell line could release miR-21 and subsequently stimulate PMN formation by upregulating programmed cell death (PD) 4 protein levels, whereby strengthening osteoclast differentiation and activation, ultimately leading to bone metastasis of breast cancer (42). From above findings, we could imply that the formation of PMN caused by tumor-derived EVs could significantly assist in organ colonization and metastasis of breast cancer, such as brain, lung and bone. As a result, patients died from tumor metastasis. In the future, we should focus on developing the novel methods to block the formation of PMN caused by tumor-derived EVs, and effectively prevent the happening of organ-specific tumor metastasis.

Despite significant advances in novel anti-cancer therapies for advanced breast cancer in recent years, many patients with breast cancer remain incurable and metastasis remains the leading cause of death in patients with advanced breast cancer (43, 44). Tumor metastasis is a multistep process involving the spread of cancer cells from the primary tumor to the blood or lymphatic system, then survival in circulation, subsequent migration to target organs, and eventually distant implantation and proliferation (45, 46). These effects require tight regulation of the cellular mechanisms by which tumor cells detach from the primary tumor and transfer to the metastatic site, in which EVs also play a crucial role (47, 48). The work of Shechter et al. has exhibited that highly metastatic or chemo-resistant breast cancer cells release more EVs than breast cancer cells with low metastatic potential. Additionally, these EVs reduce cell adhesion by substantially expressing CD44, which subsequently disrupt actin filament structure and ultimately induces tumor spread, thus contributing to breast cancer metastasis (49).

Bone is the most likely site of metastasis for all molecular subtypes of breast cancer. Patients with bone metastases often have other serious complications, such as severe bone pain, fractures, severe hypercalcemia, and nerve compression syndrome, This seriously affects the life expectancy and the quality of life (50). Under normal conditions, bones undergo a dynamic balance of bone resorption and bone formation mediated by osteoclasts and osteoblasts, respectively. However, bone metastases usually show an imbalance in this process (51), and osteolytic bone metastases is the most common type (52). EVs also promote this process. EVs can introduce miR-940 into the bone to promote osteogenic differentiation of mesenchymal stem cells by targeting Rho GTase-activating protein 1 (ARHGAP1) and FAM134A. As a result, promoting osteolytic bone metastasis (53). MDA-MB-231 triple-negative breast cancer cells also secrete EVs loaded with miR-218 to block osteoblast differentiation, tip the balance toward osteolysis, and form metastatic bone niches to facilitate osteolytic bone metastasis (54). Meanwhile, ER⁺ breast cancer cells can also produce EVs loaded with miR-19a, and promote osteolytic bone metastasis by inhibiting PTEN expression and inducing NF-κB and AKT pathways (41). So, in the future, targeting EVs to inhibit the formation of osteolytic bone metastasis may be a promising research direction in the field of cancer metastasis.

Lung is another important target organ of frequent breast cancer metastasis. Pulmonary capillaries consist of endothelial cells surrounded by basal membranes and adjacent alveolar cells. Tumor cells need to adhere to this endothelial membrane and penetrate into the lung parenchyma to establish a metastatic tumor (55, 56). Tumor-derived EVs secreted by highly metastatic breast cancer cells could load miR-122 to increase nutrient availability in lung metastatic cancer cells by down-regulating the glycolytic enzyme pyruvate kinase in lung fibroblasts, as a result, facilitating the formation and development of breast cancer lung metastasis (57). EVs also mediate miRNA transfer from cancer cells to other cells for tumor microenvironmental remodeling and promote the formation of pre-metastatic niches in lung. miR-138-5p and miR-183-5p loaded by EVs and secreted by cancer cells can regulate the activity of tumor-associated macrophages (TAM) by targeting KDM6B and PPP2CA respectively, and to promote the formation of breast cancer lung metastasis (58, 59). Besides miRNA, other non-coding RNA loaded in EVs also contribute to the formation of breast cancer lung metastasis. Abnormal expression of lncRNA in EVs promoted the formation of lung metastasis microenvironment (60, 61). Moreover, Exosome circPSMA1 could up-regulate Akt1 expression and affect the expression of downstream genes such as β-catenin and cyclin D1. circPSMA1/miR637/Akt1/β-catenin (cyclin D1) signal axis promotes the lung metastasis of triple-negative breast cancer (62). We believe that blocking the formation of lung-specific breast cancer metastasis could also affectively improve the survival outcomes of patients with breast cancer.

Brain metastasis occur in about 10%–30% of patients with breast cancer (63). Since most chemotherapy drugs can’t cross the blood-brain barrier (BBB), brain metastasis is refractory to conventional chemotherapy and exhibit extremely poor survival outcomes (64, 65). Therefore, preventing breast cancer cells from infiltrating the BBB is an important step in preventing and treating the brain metastases of breast cancer. Specifically, cytoskeleton-associated proteins, which are involved in endothelial cell adhesion and migration at the BBB, are important targets in preventing breast cancer metastases into the brain (66). The tight association between the increased expression of the cytoskeleton-associated protein tubulin tyrosine ligase like 4 (TTLL4) and brain metastasis of breast cancer was proposed by Arnold et al. They reported that TTLL4-dependent β-tubulin glutamylation increased the secretion of EVs synthesized by multivesicular bodies. These EVs increased the ability of breast cancer cells to adhere to BBB endothelial cells as well as the permeability of these endothelial cells, ultimately promoting breast cancer brain metastasis (67). Furthermore, EVs originated from vascular endothelial cells were found to induce the adhesion of triple-negative breast cancer cells to endothelial cells, and this process could be enhanced by cirGal-3. Further mechanistic exploration revealed that the enhanced glycolysis of endothelial cells induced by cirGal-3 could increase ICAM-1 expression; and EVs-dependent delivery of cirGal-3 to triple-negative breast cancer cells (MDA-MB-231) increased adhesion between the two cell types, ultimately leading to brain metastases of breast cancer (68). Moreover, Sirkisoon et al. have revealed that miR-1290-containing EVs derived from breast cancer activate astrocytes in brain TME via the FOXA2-CNTF axis. This subsequently augmented intracranial colonization and growth of breast cancer cells, leading to the progression of brain metastasis (69).

However, how do breast cancer cells acquire the ability to release specific EVs cargo to maintain the plasticity of breast cancer and the rapid spread of tumor metastasis? Further research should be performed to answer this complex question and more strategies should be proposed to block the process.

Cancer-associated fibroblasts (CAFs) are a group of activated fibroblasts in the TME with significant heterogeneity and plasticity. Particularly, CAFs are involved in tumor progression-related regulatory processes, such as the regulation of tumorigenesis, metastasis, and therapeutic resistance (76). Fibroblasts can engulf EVs secreted by breast cancer cells, in turn, fibroblasts are activated to become CAFs and stimulate the carcinogenic phenotype of breast cancer (77). Moreover, EVs released from CAFs can be engulfed by breast cancer cells to subsequently drive their migration, invasion, proliferation, epithelial-mesenchymal transition (EMT), chemoresistance, and ECM remodeling (78). Prior evidence by Baroni et al. suggests that triple-negative breast cancer cells stimulate the conversion of fibroblasts to CAFs through EVs cargo miR-9. Moreover, these newly transformed CAFs can secrete miR-9 to augment breast cancer invasion and metastasis by increasing E-cadherin expression (79). Specifically, Vu et al. found that the transfer of miR-125b-containing EVs from breast cancer cells to normal fibroblasts significantly increased the expression of the CAFs activation markers Acta2, MMP2, and MMP3, thereby promoting the conversion of CAFs phenotype (80). Concurring results by Yang et al. demonstrated that EVs released from breast cancer cells (MDA-MB-231 and MCF-7), which transferred miR-146a to normal fibroblasts, then activated the Wnt/β-catenin pathway and subsequently induced the malignant phenotypes of CAFs (81). Nevertheless, Tao et al. highlighted that EVs released from CAFs can impede the growth and metastasis of breast cancer cells by inhibiting EMT and proliferation via upregulating miR-1-3p expression (82). We can easily imply from these reported results that both breast cancer cells and CAFs can secrete EVs to regulate breast cancer progression, but the specific oncogenic or tumor-suppressive effect depends on the different effector proteins. However, more depth mechanisms underlying the effect of EVs on the behaviors of CAFs require further investigation.

Chronic inflammation is an important driver of tumorigenesis and progression (83, 84). Neutrophils are the most abundant leukocytes in the immune system and the first to migrate to sites of inflammation (85). These cancer-associated neutrophils exhibit both anti-tumor (N1) and pro-tumor (N2) phenotypes, with N1 neutrophils having a stronger cytotoxic effect on tumor cells and N2 neutrophils supporting tumor cell growth, invasion, and metastasis (86). As identified by Amorim et al., EVs released from breast cancer cells (MDA-MB-231) induced the N2-phenotype transformation of neutrophils. These MDA-MB-231-secreted EVs enhanced the production of neutrophil extracellular traps, interleukin (IL)-8, and VEGF from N2 neutrophils, as well as increased the expression of Arg-1, MMP9, and the N2 neutrophil marker CD184, which subsequently promoted breast cancer cell invasion (87). Reportedly, PD-L1 is overexpressed in breast cancer cells and is considered a key tumor immunosuppressor (88, 89). Specifically, PD-L1 maintains the immunosuppressive microenvironment by interacting with PD-1 on lymphocytes (90, 91). CAF-derived exosomes overexpress miR-92 via YAP1, and this increases PD-L1 expression to impair the function of tumor-infiltrated immune cells to drive breast cancer progression (88, 92). As previously documented by Xie et al, breast cancer cell-derived EVs transfer active TGF-β type II receptors (TβRII) to activate TGF-β signaling in target cells. The uptake of TβRII-containing EVs by breast cancer cells with low metastatic potential triggers EMT, which enhances the stemness and metastatic ability of breast cancer cells. Meanwhile, EVs-dependent delivery of TβRII to CD8⁺ T cells also induces the activation of SMAD3 and TCF1, further causes CD8⁺ T cell depletion and subsequent failure of tumor immunotherapy (93). Besides, Njock et al. identified that three miRNAs (miR-142-5p, miR-183-5p, and miR-222-3p) were released from endothelial cells in the TME and were transferred via EVs to macrophages. In a mouse model of breast cancer, EVs enriched with these miRNAs contributed to the differentiation of macrophages toward the immunosuppressive M2 phenotype, which augmented breast cancer growth (94).

Myeloid-derived suppressor cells (MDSCs) are closely associated with breast cancer progression (95, 96). Jiang et al. highlights that miR-9 and miR-181a shuttled by EVs from breast cancer cells could activate the JAK/STAT signaling pathway by targeting SOCS3 and PIAS3. This fosters the expansion of early MDSCs and compromises T-cell proliferation and promotes T-cell apoptosis, which subsequently accelerates the growth of breast cancer cells both in vivo and in vitro (97). Biswas et al. further indicated that exosomes secreted by MSCs can harbor high levels of TGF-β, which subsequently induced the malignant progression of breast cancer by facilitating the differentiation of immature MDSCs into M2 macrophages with stronger immunosuppressive activity. Exosomes derived from MSCs, but not from breast cancer cells, contain significant amounts of TGF-β, C1q, and semaphorins. Moreover, the induced PD-L1 overexpression in immature myelomonocyte precursors and CD206⁺ macrophages, and the fostered differentiation of MHC class II macrophages enhance Arg-1 activity and IL-10 secretion to promote myeloid tolerance activity. This ultimately leads to the formation of immunosuppressive M2 macrophages and results in the rapid growth of breast cancer cells (98). The interactions between EVs cargoes and immune cells significantly contribute to the rapid progression of breast cancer, and also facilitate to the formation of immunosuppressive tumor microenvironment.

Tumor cells undergo EMT before the occurrence of tumor metastasis. During EMT, tumor cells lose epithelial features such as cell adhesion and acquire mesenchymal features, which allow them to migrate and metastasis to occur (99–101). When tumor cells reach the metastatic site, they then undergo the reverse process called mesenchymal-epithelial transformation, which allows these cells to colonize their metastatic site (102–104). However, in the early stages of metastatic spread, disseminated tumor cells may remain dormant in the G0/G1 phase to accumulate the capacity for bone marrow metastasis; this process is also observed in metastatic breast cancer cells (105). Casson et al. revealed that bone marrow MSC (BMSC)-derived EVs increase the adhesion capacity of ER⁺ breast cancer cells (MCF7) to a dormant epithelial cell phenotype. This facilitates their colonization in the bone marrow in a circulating quiescent state, ultimately leading to bone metastases of breast cancer and resistance to chemotherapeutic agents (106). Further evidence has suggested that breast cancer cells augment BMSCs to release EVs containing miR-222 and miR-223. These EVs, in turn, stimulate the dormancy of disseminated breast cancer cells, causing resistance to chemotherapeutic agents. Based on this, Bliss et al. found that, in a mouse model, targeting dormant breast cancer cells with miR-222 and miR-223 antagonists (delivered by EVs secreted by BMSC) significantly improved the susceptibility of breast cancer cells to carboplatin and also extended the survival time of mice (107). Therefore, we believe that targeting dormant breast cancer cells by EVs may be an effective strategy to improve the response to routine therapy and even prevent breast cancer recurrence and metastasis.