The origin of MSCs can be traced back to the mid-to-late 20th century (30). After decades of exploration, researchers have now developed a relatively systematic understanding of these cells (Fig. 1). As early as 1960-1975, researchers first identified a type of non-hematopoietic stem cell in the bone marrow with clonogenic capacity and adherence-dependent growth (31). Friedenstein et al (32) were the first to systematically isolate and characterize these cells, which is considered the starting point of research on MSCs. This discovery demonstrated for the first time that within the complex hematopoietic microenvironment of the bone marrow, there exists a unique group of non-hematopoietic precursor cells, laying the theoretical foundation for subsequent studies on MSCs.

From 1975 to 1980, research further elaborated on the biological characteristics of MSCs, such as the cloning formation characteristics and radiation resistance (33). Further studies gradually revealed the multi-lineage differentiation potential of these cells. In 1988, Grigoriadis et al (34) found that a subset of mesenchymal progenitor cells could differentiate into four distinct cell types, namely, muscle, adipose tissue, cartilage and bone, in vitro under the induction of glucocorticoids. This discovery suggested that MSCs possessed multi-directional differentiation capabilities, thus opening up the possibility of application in regenerative medicine. In 1995, research further confirmed the presence of MSCs in periodontal tissues (35). This finding indicated that the source of MSCs expanded from bone marrow to other tissues, providing an important basis for the subsequent isolation of MSCs from various tissues, such as fat and the umbilical cord, and also revealed the extensive distribution of MSCs in the body.

From 2000 to 2015, the mechanisms of MSCs were extensively explored and researchers gradually recognized that MSCs possessed multiple functions in disease contexts, including immunomodulation, inflammation suppression, promotion of repair (for example, repairing bones, cartilage, tendons and cardiac muscle) and specific homing to tumor sites (36,37). In 2006, the International Society for Cellular Therapy proposed the minimal defining criteria for MSCs: i) Adherence to plastic; ii) specific expression of CD105, CD73 and CD90; iii) lack of expression of surface markers CD45, CD34, CD14, CD11b, CD79α, CD19 and human leukocyte antigen-DR isotype; and iv) multi-lineage differentiation potential. This criteria established a foundation in standardizing research on MSCs (38). The proposal of this standardized definition marks a notable milestone in the development of this field. It enables the results of different researchers to be similar and establishes standards for the basic research of MSCs.

Between 2015 and 2020, extracellular vesicles (EVs) secreted by MSCs, particularly exosomes, were identified as key carriers mediating their effects and gradually became a research hotspot in regenerative medicine as an alternative to cell therapy (39). From 2020 to the present, clinical trials associated with MSCs have been widely conducted. Certain studies have demonstrated that MSCs exhibit notable therapeutic efficacy (40,41), while others have reported potential negative effects; for instance, there are complex interactions between MSCs and tumor cells, which can create a microenvironment conducive to tumor cell proliferation, angiogenesis, migration, invasion and metastasis, thereby promoting tumor progression (42). Furthermore, in another study, it was found that MSCs co-cultured with colorectal cancer cells exhibited enhanced invasiveness and proliferation ability due to changes in their tumor protein p53/transforming growth factor β1 (p53/TGF-β1) levels (43). The core contribution of this stage is that it has pushed the research on MSCs to enter a novel phase from ‘verification of effectiveness’ to ‘precision of treatment’ (44). These results have prompted researchers to reflect, suggesting that this bidirectional nature of effects may be associated with factors such as cell source, route of administration and dosage (45,46). Of note, MSCs can be isolated from various biological tissues. In addition to common sources such as bone marrow, adipose tissue, umbilical cord, amniotic fluid, placenta and menstrual blood (47–51), several studies have also reported that MSCs can be successfully extracted and isolated from teeth (52), tonsils (53), as well as visceral tissues including the brain, spleen, kidney and liver (54–56).

EVs are heterogeneous collections of particles enclosed by a lipid bilayer membrane. EVs are typically classified into three main types based on their biogenesis pathways and size (57). Among them, exosomes are small nanovesicles (diameter, 30–150 nm) derived from the endosomal system (58), which distinguishes them from the larger microvesicles (diameter, 100–1,000 nm) and the distinctively originating apoptotic bodies (diameter, 500–5,000 nm) (59,60) (Fig. 2A). The biogenesis of exosomes is a complex process (Fig. 2B). It begins with endocytosis at the cell membrane surface, forming early endosomes. Early endosomes further mature, receive transport vesicles from organelles such as the Golgi apparatus and undergo processes such as acidification to transform into late endosomes. Within late endosomes, the membrane invaginates inwardly a second time, forming multiple intraluminal vesicles (ILVs) through budding (61–63). Late endosomes containing numerous ILVs are termed multivesicular bodies (MVBs). MVBs undergo either of two pathways: Most MVBs fuse with lysosomes, leading to the degradation of the ILVs, while a small portion of MVBs fuse with the plasma membrane, releasing the ILVs into the extracellular environment. These released ILVs are termed exosomes (61,64,65). Similar to MSCs, the sources of exosomes are diverse (Fig. 3). Besides being present in MSCs and various immune cells (66–68), exosomes are also widely found in tumor cells (69,70). Furthermore, exosomes exist in various bodily fluids, including blood, urine, breast milk and semen (71,72). Exosomes have also been detected in certain foods and plants (73–77).

The abnormal proliferation of tumor cells is a central aspect of tumor progression. A study based on in vitro cell models found that MSCs derived from CC cells (CC-MSCs) upregulated the expression level of CD73 in tumor cells, thereby accelerating the progression of CC (78). Another study using both in vitro and in vivo models further revealed that the expression level of the transcription factor NANOG in CC cell-derived MSCs can promote the proliferation of CC SiHa cells, thereby driving CC growth (81). Notably, further research has observed that the effect of bone marrow-derived MSCs (B-MSCs) on the proliferation of CC cells exhibits a dose-dependent pattern: Low doses can promote proliferation, whereas high doses inhibit proliferation; this phenomenon may be associated with the modulation of the mitogen-activated protein kinase/phosphatidylinositol 3-kinase (PI3K) signaling pathways (46). This finding suggested that the dosage of MSCs used may be one of the key factors influencing whether MSCs exert pro-tumor or antitumor effects.

Due to the high demand for oxygen and nutrients by tumor cells, tumor angiogenesis is key to sustaining the early stages of tumor development and its progression (82). Previously, a study based on in vitro and in vivo models reported that adipose tissue-derived MSCs (AD-MSCs) may promote CC angiogenesis by activating the nuclear factor κ-light-chain-enhancer of activated B cells signaling pathway, thereby accelerating tumor growth and metastasis (79). Similarly, umbilical cord-derived MSCs (UC-MSCs) may enhance in vivo angiogenesis in CC HeLa cells by secreting vascular endothelial growth factor (VEGF), thereby promoting the progression of CC (83).

MSCs can promote cancer progression by modulating the immune response within the TME. For instance, CC-MSCs exhibit a stronger ability to induce M2 macrophage polarization compared with MSCs derived from normal cervical tissue, thereby fostering an immunosuppressive microenvironment (84,85). Furthermore, it has been reported that CC-MSCs can promote the secretion of TGF-β1 and the expression level of programmed death-ligand 1 through an adenosine-dependent pathway, subsequently inhibiting the antitumor activity of CD8⁺ T lymphocytes and contributing to cancer promotion (80). Another study also revealed that, compared with MSCs from normal cervical cells, MSCs derived from CC cells highly express CD39 and CD73 on their cell membrane. These molecules serve a notable role in extracellular adenosine generation and immune evasion (86). However, the aforementioned mechanisms are primarily based on in vitro experiments and their in vivo effects require further validation.

MSCs and their derived exosomes can inhibit the growth of CC through multiple signaling pathways. For example, in vitro cell models have reported that UC-MSC-Exos can induce apoptosis in CC HeLa cells and suppress the expression levels of epithelial-mesenchymal transition (EMT)-related proteins, thereby exerting antitumor effects (87). Additionally, microRNA (miR)-370-3p carried by such exosomes can inhibit the proliferation and migration of CC cells by targeting 24-dehydrocholesterol reductase in in vitro cell models (88). It has also been reported that UC-MSC-Exos can promote squamous differentiation of CC CaSki cells by activating the Notch signaling pathway, thereby inhibiting their growth and metastasis (89). In in vivo animal models, AD-MSCs carrying the herpes simplex virus thymidine kinase lentiviral vector can induce cancer cell apoptosis, thereby delaying CC progression (90). Furthermore, miR-144-3p in B-MSC-Exos can inhibit the proliferation and invasion of CC cells and promote their apoptosis by targeting centrosomal protein 55 (91). Menstrual blood-derived MSCs can inhibit the proliferation of CC HeLa cells through the TGF-β1-mediated c-Jun N-terminal kinase/p21 signaling pathway, thereby exerting antitumor effects (92).

Due to the immunosuppressive properties of MSCs, MSCs are often perceived more as tumor-promoting rather than antitumor agents (93–95). However, research has also suggested that MSCs may exert antitumor effects by modulating inflammatory responses. For example, a previous study by Yi et al (96) demonstrated that UC-MSC-Exos can markedly reduce the expression levels of pro-inflammatory factors such as tumor necrosis factor-α (TNF-α), IL-1β and IL-6, while increasing the expression of the anti-inflammatory factor IL-10, thereby alleviating cervical inflammation and potentially reducing the risk of carcinogenesis by inhibiting the EMT process. This finding indicated that the anti-inflammatory regulatory functions of MSCs and their derived exosomes may be associated with their antitumor effects.

In contrast to the pro-angiogenic effects of AD-MSC-Exos (79) and UC-MSC-Exos (83), a previous study based on an in vitro cell model indicated that amniotic fluid-derived MSC-Exos can inhibit the proliferation of CC HeLa cells by suppressing tumor angiogenesis, thereby exerting antitumor effects (97). This opposing role in tumor vascular regulation exhibited by MSCs and their exosomes derived from different tissues suggests that the cellular source may be a key factor in determining their functional direction.

MSC-Exos inherently carry various bioactive substances, such as lipids, proteins and nucleic acids, serving as crucial mediators in intercellular communication (98). Furthermore, following engineering modifications, MSC-Exos can also function as drug delivery systems to enhance therapeutic precision. For example, UC-MSC-Exos can serve as effective delivery vehicles for paclitaxel, augmenting its antitumor efficacy (79). Furthermore, B-MSC-Exos can regulate the methylation level of LIM zinc finger domain 2 in CC cells by delivering miR-331-3p, thereby inhibiting tumor progression (99).

In summary, MSCs and their exosomes exhibit a distinct dual role in CC (Fig. 4): MSCs and their exosomes can both promote tumor cell proliferation, angiogenesis and immune escape, as well as inhibit tumor cell proliferation, induce apoptosis and modulate the immune microenvironment. This bidirectional effect is likely closely associated with their tissue origin, dosage, administration route and the specific signaling molecules carried by the exosomes. These aspects will be discussed in detail in the following sections.

The differences in tissue origin may be an important factor influencing the direction of the role of MSCs and their derived exosomes in CC. A previous study by Song et al (102) demonstrated that MSCs from different sources exhibited notable differences in their chemotactic ability towards CC cells. This research conducted a series of in vitro experiments to compare the chemotactic effects of MSCs from adipose tissue, umbilical cord, amniotic membrane and chorion on CC cells. The results demonstrated that the MSCs from chorion had the strongest chemotactic ability. Underlying this phenomenon may involve the epigenetic signatures carried by MSCs from different tissue origins. A previous study demonstrated that MSCs retain the epigenetic characteristics of their tissue of origin during development, thereby influencing their gene expression profiles and secretome (103). For instance, B-MSCs may be more inclined to express factors associated with osteogenesis, whereas AD-MSCs are enriched in molecules associated with lipid metabolism and angiogenesis (104). This finding suggested that although MSCs from different tissue sources are similar in morphology, immunophenotype and basic activity, their biological functions may differ, which was consistent with the conclusion of Kern et al (100).

The different dosages used may also result in opposite effects of the same source of MSCs on CC cells. Long et al (46) conducted co-culture experiments of B-MSCs with CC cells HeLa in vitro. The results indicated that a low proportion of B-MSCs could promote the proliferation of CC cells, while a high proportion inhibited their proliferation. This dose-dependent effect may be associated with a threshold effect in signaling pathways. Specifically, at low doses, growth factors secreted by MSCs (such as VEGF and TGF-β) might primarily activate pro-survival signaling pathways (for example, PI3K/protein kinase B) in tumor cells, thereby promoting proliferation. However, when the MSCs dose increases to a certain level, the concentration of their secreted factors also rises, potentially exceeding a threshold and instead activating pro-apoptotic or anti-proliferative signaling pathways (involving molecules such as interferon-γ or TNF-α). Nevertheless, this hypothesis requires further investigation for confirmation. The aforementioned phenomenon suggests that when applying MSCs in CC therapy, the selection of dosage is critically important and may directly impact therapeutic outcomes.

Although exosomes are derived from MSCs and share high functional similarity with MSCs, they may exert different or even opposing effects during the development of CC. For instance, UC-MSCs can promote tumor progression by inducing tumor angiogenesis (83), whereas their derived exosomes demonstrate antitumor effects (87). This functional discrepancy is closely associated with the distinct mechanisms of action between the two. Specifically, MSC-Exos can modulate their secretion and uptake behavior through various means and interventions, thereby exhibiting functional characteristics that differ from those of their parent cells (105). This understanding suggests that in future research, the relationship between the two should be regarded as both interconnected and independent, warranting separate investigation rather than being simplistically treated as a single entity.

MSCs and their derived exosomes can be administered through various routes and the method of administration may influence their ultimate effects. A previous study has reported that in mouse models, local injection of B-MSCs exacerbate tumor growth, whereas no notable difference was observed between the treatment and control groups after intravenous injection (101). Although this study was based on an ascites cancer cell model, the results suggested that the route of administration may affect the function of MSCs. Based on this phenomenon, the present review hypothesizes that local injection may result in a high concentration of MSCs aggregating at the tumor site, yet the sphere of action may be limited and MSCs might be induced to exhibit pro-tumorigenic effects due to pro-tumoral signals within the local microenvironment. Furthermore, different routes of administration lead to distinct patterns of contact between MSCs and the immune system of the host, potentially triggering different rates of immune recognition and clearance (106). This could, in turn, affect their persistence and functional duration in vivo, which may also contribute to their differential effects. Therefore, using CC models in future research is necessary to further elucidate the differential impacts of various administration routes in CC.