A comprehensive literature search was conducted to gather relevant studies for this review. The search was performed across major scientific databases including PubMed and Web of Science.

The search included combinations of keywords such as “cancer stem cells”, “melanoma”, “tumor microenvironment”, “biomarkers”, “signaling pathway”, “sphingolipids”, “S1P”, “exosomes”, “immunotherapy”, “Targeted therapy”, “drug resistance”, “immune modulation”, “immune cells” and “metabolic reprogramming”. Boolean operators (AND, OR) were used to refine the search, ensuring thorough coverage of all relevant topics.

Studies included in this review were selected based on relevance to the impact of the TME on MSCs, interactions between immune cells and MSCs, and molecular mechanisms within the TME that influence CSC survival and function. Only peer-reviewed articles written in English were considered. Studies were excluded if they focused on non-melanoma cancers or lacked sufficient data on the interaction between the TME and MSCs.

The review primarily includes studies published within the past 20 years (2004-2024), ensuring a focus on recent advancements while also capturing significant earlier research that has contributed to the understanding of TME-CSC interactions.

CD20, a non-glycosylated phosphoprotein, plays a regulatory role in the proliferation and differentiation of B lymphocytes. Given its ubiquitous expression on the surface of B lymphocytes and its tendency to bind readily to antibodies without shedding, monoclonal antibodies targeting CD20 have emerged as the standard treatment for non-Hodgkin’s lymphoma. In 2005, a study found that the tumor stem cell population was enriched in the CD20(+) fraction of melanoma cells. Fang et al. (11) identified the presence of multipotent spheroid cells within melanoma cell lines. Their study has revealed that the stem cell population within melanoma cells is notably enriched in the CD20(+) fraction, thus designating it as a MSCs population. Notably, the CD20(+) fraction demonstrated a tendency to form larger spheres in comparison to the CD20(-) fraction, along with a heightened potential for mesenchymal differentiation (11). It is crucial to highlight that CD20(+) cells, which play a pivotal role in tumor stemness, constitute a relatively small fraction, representing approximately 2% of the total melanoma cell population. Schlaak et al. treated patients with metastatic melanoma by injecting the therapeutic anti-CD20 antibody rituximab at the lesion site and showed that the use of rituximab yielded excellent efficacy, accompanied by a reduction in the melanoma serum marker S-100 to levels within the physiological range (29).

CD133, or Prominin-1, is considered to be the most important surface marker for identifying mesenchymal stem cells (30–32). Indeed, in various tumor types, such as glioma, colon cancer, pancreatic cancer, and liver cancer, CD133(+) cells consistently display characteristics resembling those of stem cells (32), underscoring the potential significance of CD133 as a marker for tumor stem cells. In the context of melanoma, only the CD133(+) subpopulation, not the CD133(-) counterpart of melanoma cells, were able to reform a Mart-1 (a characteristic melocytic marker) positive tumour in NOD-SCID mice (33). CD133 expression levels demonstrate an elevation in both primary and metastatic melanomas when contrasted with their normal pigmented nevi counterparts (34). Additionally, research findings from other studies have also indicated a correlation between increased CD133 expression levels and enhanced tumorigenicity and metastatic propensity in melanoma (30). However, doubts still abound, researchers sought to assess whether markers such as CD10, CD133, nestin and CD20 could assess the prognosis of advanced melanoma, the results showed that they were unable to detect a significant correlation between nestin or CD133 expression in melanoma and patient survival or clinical outcome (35), which implies that CD133 may not exert a significant influence as a prognostic factor.

CD271, also termed the low-affinity nerve growth factor receptor or p75NTR, is a characteristic marker of mesenchymal stem cells (36, 37). Melanocytes are formed during the differentiation of multipotential neural crest stem cells (NCSCs) under precise regulatory mechanisms. CD271 exhibits analogous expression patterns in melanocytes, melanoma cells, and NCSCs (38). Furthermore, it plays a pivotal role in governing the preservation of cellular stemness and migratory characteristics via an intricate network of interconnected genes. The tumorigenic assay performed by Bolok et al. showed that CD271 was identified as a CSC marker that can identify and prospectively isolate MSCs (37). However, CD271(+) melanoma cells lack expression of typical melanoma cell surface markers such as TYR, MART, and MAGE, leading to the speculation that CD271+ melanoma cells may be in a state of incomplete differentiation. In addition, CD271 regulates phenotypic transition, a process that results in rapid and reversible conversion of the proliferative state to an invasive state or of the non-stem cell state to a stem cell state, by mechanisms that are not entirely understood (38). Cheli et al. considered that CD271 was not a perfect marker for MSCs, their study showed that not all CD271+ cells were tumorigenic, only the slow-growing CD271+ cell subpopulation was highly tumorigenic (39).

ABCB5, a member of the ABC family of transporter proteins, is highly expressed in CD133(+) melanoma cells. It is considered to be one of the markers of MSCs and is closely associated with chemotherapy resistance in MSCs (40, 41). In 2008, Schatton et al. confirmed that ABCB5 can be used as a molecular marker for MSCs, ABCB5(+) melanoma cells had a higher tumorigenic potential than the ABCB5(-) somatic cell population in subsequent mouse xenograft trials, which re-established clinical tumour heterogeneity. In vivo genetic genealogy tracing demonstrated the specific ability of the ABCB5(+) subpopulation to self-renew and differentiate compared to ABCB(-), as ABCB(+) cancer cells produced both ABCB5(+) and ABCB5- progeny, whereas the ABCB5(-) tumor population produced only ABCB5(-) cells (42). A study by Wang et al. confirmed that ABCB5 is a key factor in promoting melanoma metastasis. ABCB5(+) malignant MSCs showed a higher metastatic potential compared to ABCB5(-) melanoma subpopulation (43).

Aldehyde dehydrogenase (ALDH) is a polymorphic enzyme family responsible for the oxidation of aldehydes to carboxylic acids. ALDH has also emerged as a promising marker for CSCs, playing a role in resistance to various chemotherapeutic agents and immune responses across a range of human solid tumors (32, 44, 45). The human ALDH superfamily comprises 19 isoforms, each with distinct biological functions beyond their enzymatic role in detoxifying bioaldehydes and xenobiotics. Among these, several isoforms are closely associated with the characterization of CSCs and the acquisition of malignant properties and drug resistance. For instance, the ALDH1 family—particularly ALDH1A1, ALDH1A2, and ALDH1A3—has been identified as a critical marker for normal stem cells and CSCs. ALDH1 plays a key role in retinoic acid (RA) signaling, which is essential for maintaining the “stemness” properties of CSCs (46). The high ALDH activity detected in the Aldefluor assay is often attributed to ALDH1A1, which contributes to the aldehyde-induced fluorescence staining (47, 48). However, recent studies have shown that ALDH1A3 also significantly contributes to Aldefluor positivity, indicating a broader role for ALDH1 isoforms in CSC characterization (49). ALDH2, another isoform, has been strongly correlated with alcohol-related tumor formation, while ALDH3A1 is implicated in drug resistance through the oxidation of chemotherapeutic agents like oxazaphosphorines, including cyclophosphamide (CP) (47). In melanoma, ALDH(+) cells have been found to be more tumorigenic compared to ALDH(-) cells. Knockdown of ALDH expression using siRNA or shRNA leads to cell cycle arrest, apoptosis, reduced cell viability in vitro, and reduced tumorigenesis in vivo (50). Nonetheless, it is noteworthy that both ALDH(+) and ALDH(-) melanoma cell subpopulations displayed comparable tumorigenic capabilities in both in vivo and in vitro assays. Additionally, both subgroups exhibited similar responsiveness to anti-melanoma drugs, including dacarbazine and lexatumumab (51). Tumors from ALDH(-) cells largely maintained the parental ALDH(-) phenotype in vivo after 2-3 passages. This reaffirms that the observed tumorigenicity is indeed an inherent trait of ALDH(-) cells and is not attributed to inadequate segregation of the two distinct subpopulations. In contrast, tumors originating from ALDH⁺ cells display a mixed population—predominantly ALDH(+) cells, with 20-40% of cells lacking ALDH activity (51). While ALDH(+) melanoma cells show a higher capacity to generate phenotypic heterogeneity, the functional implications of this trait remain unclear. The inability of the ALDH phenotype to distinguish between cells responsible for tumor initiation and therapy resistance suggests that it may not necessarily mark the more aggressive subpopulations within malignant melanoma. These findings imply that the ability to reestablish tumor heterogeneity is not inherently tied to a more aggressive phenotype. Therefore, further research is needed to determine whether ALDH can serve as a reliable marker for MSCs.

SOX10 is in the high-mobility-group (HMG)-box family of transcription factors and plays an important role in the development of melanocytes and other neural crest cells (52). The researchers conducted subcutaneous injections of melanoma cells with normal SOX10 expression and melanoma cells with SOX10 silenced by shRNA in immunodeficient NOD/SCID or nude mice. After an eight-week period, the melanoma cells expressing SOX10 consistently led to the development of substantial in vivo tumors (11 out of 14). In contrast, none of the subcutaneously injected melanoma cells expressing SOX10 shRNA (0 out of 16) resulted in tumor formation, even after an additional six weeks of observation.This strongly indicate that the silencing of SOX10 is highly efficacious in preventing the formation of tumors in melanoma cells in an in vivo setting (53). SOX10 also plays a significant role in mediating resistance to BRAF inhibition in melanoma. The downregulation of SOX10 triggers the activation of TGF-β signaling, which results in the upregulation of epidermal growth factor receptor (EGFR) and platelet-derived growth factor receptor β (PDGFRB), ultimately leading to the development of a slow-cycling phenotype. This, in turn, contributes to resistance against BRAF and MEK inhibition by bestowing survival signals that are independent of the MAPK pathway (54). Willis et al. conducted a study to investigate the staining sensitivity and specificity of SOX10 as a marker for metastatic melanoma. Their research focused on sentinel lymph nodes that had been previously diagnosed as positive or negative. The results revealed that SOX10 immunostaining successfully identified metastatic melanoma in all examined cases, achieving a 100% detection rate. Additionally, a statistically significant increase in staining intensity for SOX10 was observed when compared to S100 protein, HMB-45, and Melan-A (with respective p-values of 0.000, 0.000, and 0.003) (20). The preceding research indicates that SOX10 has the potential to serve as a marker for MSCs.

The Wnt signaling pathway is a crucial regulatory pathway that governs various cellular functions across different biological contexts, including embryonic development, tissue homeostasis, stem cell maintenance and cell proliferation and differentiation (56). The most prominent Wnt signaling pathways include canonical pathways known as the Wnt/β-catenin pathway and non-canonical pathways such as the Wnt/PCP (planar cell polarity) pathway, the Wnt/Ca2+pathway, and the WNT/RTK (receptor tyrosine kinase) pathway (57). Both canonical and non-canonical WNT signaling cascades play fundamental roles in shaping the development and evolution of CSCs (58). Active Wnt/β-catenin signaling was reported in ∼30% of melanoma tumors, indicating a potentially specific role for this signaling pathway in this aggressive type of cancer (59). Prior research has demonstrated that curtailing the heightened activation of β-catenin, a key gene in the Wnt signaling pathway, or its inactivation by inhibiting the APC gene can reduce the accumulation of β-catenin in cells, thus inhibiting the ability of tumor stem cells to self-renew and proliferate and differentiate, and restoring the sensitivity of CSCs to radiotherapy (60–62). The expression of the Frizzled 3 receptor (FZD3) is integral to the processes of proliferation and differentiation of melanocyte derivatives in an in vitro setting. It was established that FZD3 is essential for melanoma oncogenesis and establishes a positive feedback mechanism for activating the MAPK signaling network by initiating the non-canonical WNT pathway. Downregulation of FZD3 diminished the growth, colony-forming potential and invasive ability of melanoma cells, besides, the inhibition of FZD3 expression suppressed melanoma initiation and growth in vivo. Moreover, the clinical association of FZD3 expression with melanoma progression and reduced patient survival has been observed (63).

Similar to the Wnt signaling pathway, the Notch signaling pathway also assumes a significant role in the regulation of the maintenance and self-renewal of CSCs (64). The Notch pathway is composed of four different transmembrane receptors, Notch1 to Notch4, and their membrane-bound ligands, Jagged1, Jagged2, and Delta1, Delta3, and Delta4 (65). Previous studies have shown that Notch1 is highly expressed in 50-60% of melanomas and 65% of melanoma cell lines, while it is very low or undetectable in normal melanocytes and pigmented nevi (66). In a separate study, researchers have conducted an analysis of Affymetrix expression profiles employing a dataset derived from 44 samples obtained from patients diagnosed with metastatic melanoma. The findings from this analysis indicated a potential linkage between the activation of the Notch signaling pathway and the prognosis of individuals afflicted with metastatic melanoma (67, 68). Huynh et al. used RO4929097(a novel γ-secretase inhibitor) to target Notch signaling, the results revealed that it affected the oncogenic and stem cell-like properties of primary melanoma cells. Furthermore, through the utilization of in vivo tumorigenic assays, their research unequivocally substantiated the capability of RO4929097 to effectively curtail the growth of both primary and metastatic melanoma cells (67). This observation further underscores the potential of targeting the Notch signaling pathway as a promising therapeutic strategy for the eradication of melanoma.

The Hh signaling pathway is a well-conserved cascade essential for embryonic development, tissue homeostasis, and the maintenance of stem cells (69, 70). The Hh signaling pathway exhibits distinct functions in different types of cancer (71). Throughout tumor development, Hh signaling plays three major roles: driving tumor initiation, promoting tumor growth, and regulating residual cancer cells following therapy (55). Additionally, emerging evidence suggests that the Hh pathway is involved in promoting the self-renewal of CSCs (72). Evidence indicates that the Hh pathway is crucial for the oncogenic characteristics of melanoma, SHH-GLI signaling regulates the proliferation and survival of human melanoma (73). In 2012, Roberta et al. demonstrated for the first time that blocking of the Hh-GLI pathway decreased self-renewal and tumorigenicity of ALDH high MSCs (61), suggesting that the Hh-GLI1 signaling pathway significantly influences the self-renewal and tumorigenesis of MSCs. Additionally, the investigators observed that targeting the Hh transcription factors GLI1 and GLI2 restored the sensitivity of human melanoma cells resistant to vemurafenib and even inhibited acquired chemotherapy resistance in melanoma patients (74).

Increasing evidence indicates that the TME critically regulates the maintenance of CSCs, thereby controlling cancer progression and metastasis (75). The TME is a markedly heterogeneous and intricate milieu, comprising a complex network of cells and extracellular macromolecules. This encompasses stromal cells, immune cells, epithelial cells(ECs), and supporting cells within the extracellular matrix (ECM) (76). Moreover, various microenvironmental factors, such as the perivascular niche and hypoxia, have also been substantiated (77). The constituents within the TME play a pivotal role in facilitating the growth, sustenance, and differentiation of CSCs (78). Furthermore, they are involved in conferring therapeutic resistance by shielding tumor cells from damage caused by treatment interventions (79, 80). It has also been shown that CSCs also contribute to the re-establishment of the TME by transdifferentiating into a spectrum of normal stroma-like cells, such as vascular ECs, vascular pericytes or fibroblasts (81, 82). Understanding the intricate interactions within the TME, including those between CSCs and various cellular components, is crucial for unraveling the mechanisms that underlie tumor progression and therapeutic resistance.

MSCs interact with immune cells in the TME to evade immune detection and promote tumor growth.

Melanoma cells can shape their microenvironment through both direct cellular interactions and the release of soluble factors, which are essential for their growth and metastatic potential (112). These soluble factors include cytokines, chemokines, growth and angiogenic factors and other signaling molecules (113). For instance, factors released by macrophages can stimulate nearby melanoma cells, leading to an increased production of melanoma inhibitory activity (MIA) in vitro. This elevated MIA, in turn, enhances the invasive capacity of melanoma cells by modulating their attachment to the ECM (114). Additionally, TGF-β, IL-6, and IL-8 released by CAFs and TAM can promote CSCs self-renewal and tumor progression by activating pathways like STAT3, NF-κB, and Wnt (80, 81). VEGF and other angiogenic factors secreted by TME cells enhance neovascularization, creating a vascular niche that further supports CSCs survival and expansion (115). Conversely, MSCs can contribute to an inflammatory microenvironment within the TME by secreting cytokines and chemokines, such as IL-6, IL-8, and TNF-α. These inflammatory factors recruit immune cells and hematopoietic stem cells into the TME, fostering an environment conducive to tumor growth and metastasis (112). Moreover, these soluble factors not only promote CSCs growth but also contribute to the establishment of an immunosuppressive microenvironment by recruiting Tregs and MDSCs, thereby inhibiting effective immune surveillance (83).

Neurons and nerve fibers are critical components of the TME, playing significant roles in immunomodulation and the regulation of various signaling pathways in tumor cells and other cellular elements within the TME (135). These neural elements are key facilitators of pathways that drive tumor growth and metastasis (136). The interactions between neurons, nerve fibers, and the TME can be broadly categorized into two main types: (1)Perineural Invasion (PNI), which describes the process where tumor cells invade and grow along nerves. PNI is a common feature in many solid tumors and is associated with a poor prognosis; and (2) Neural-Tumor Interactions, which refer to the biochemical communication among immune cells, malignant cells, and nerve fibers within the TME (135, 137). Moreover, nerve fibers within the TME can secrete a variety of growth factors, neurotrophins, matrix metalloproteinases, neuropeptides, and neurotransmitters, which activate membrane receptors on tumor cells and support tumor progression (138). Recent studies have highlighted the role of the sympathetic nervous system (SNS) in melanoma progression through its action on β-adrenergic receptors (β-ARs) expressed on tumor cells (135). Activation of these receptors triggers the release of pro-tumorigenic cytokines and metalloproteinases in melanoma cell lines, which further facilitates melanoma growth and metastasis (139).

Endothelial cells (ECs) play a critical role not only in vascular functions such as alloimmunity, immune cell recruitment, immune tolerance, and vascular inflammation, but also in modulating immune responses across various tissues and organs (140). In malignant tumors, angiogenesis is primarily driven by multiple pro-angiogenic factors, including growth factors, cytokines, and ECM proteins, which act on vascular ECs (141). In melanoma, increased ALDH1A1 expression and activity upregulates the release of pro-angiogenic factors, and these factors modulate the angiogenic profile of ECs by rearranging the Notch pathway (142). Compared to non-tumor ECs, those within the tumor environment exhibit enhanced proliferation, migration, and tube formation in response to these stimuli. Furthermore, they show increased resistance to therapy under the influence of VEGFA signaling (101, 143). Additionally, ECs derived from CSCs have been identified in various solid tumors, such as gliomas, renal cancers, and breast cancers (81, 144, 145). CSCs can differentiate not only from epithelial cells but also from ECs (145). This further supports the hypothesis that the CSC population has stem cell characteristics associated with tumor growth and vascularization.

Cellular markers of MSCs are closely linked to their tumorigenic potential, making them a promising target for therapeutic intervention. Targeting specific stemness markers on MSCs has shown potential in suppressing melanoma progression. For instance, monoclonal antibodies directed against ABCB5 have been demonstrated to induce antibody-dependent cell-mediated cytotoxicity (ADCC) in ABCB5-positive MSCs. Systemic administration of these antibodies has resulted in significant tumor-suppressive effects (157, 158). Similarly, monoclonal antibodies targeting distinct epitopes of the CD133 protein have exhibited cytotoxic effects on melanoma cells, effectively inhibiting tumor growth (159). Suppression of CD133 expression was further associated with reduced metastatic potential and decreased levels of MMP2 and MMP9 (159, 160). CD44, another stemness marker, is expressed across various CSC populations, including melanoma. Shen et al. reported that lipid nanoparticles coated with hyaluronic acid (HA)—a natural ligand for CD44—and loaded with paclitaxel analogs (PTX-loaded HA-SLNs) induced significant apoptosis in CD44-positive B16F10 melanoma cells, both in vitro and in vivo ( 28, 161). As mentioned earlier, CD20 is a well-known cell surface marker of MSCs. Schlaak et al. treated a patient with chemotherapy-refractory metastatic melanoma with the anti-CD20 monoclonal antibody rituximab in combination with low-dose dacarbazine, which showed that the patient achieved complete remission of all but one metastasis that was partially stabilized, that serum malignant black markers S-100 were reduced to normal levels, and that no treatment-related toxicity was observed (29). Moreover, Morita and colleagues developed a humanized anti-CD271 monoclonal antibody (hCD271 mAb) that specifically targeted CD271-positive MSCs. Their in vivo studies revealed that this antibody significantly reduced melanoma growth in xenograft models by depleting the CD271-positive cell subpopulation (162). These findings suggest that targeting MSC-specific markers could provide a means to impair tumor growth and reduce recurrence rates in patients.

The maintenance of biological characteristics and self-renewal of MSCs is closely related to multiple signaling pathways (60). Key pathways such as Hh, Notch, and Wnt/β-catenin, which are essential for the formation of melanocyte lineages, have been shown to be strongly associated with the malignant behavior of melanoma cells (163). Inhibition of the Hedgehog pathway, particularly targeting the activator smoothened (SMO) using small interfering RNA (siRNA) and the small molecule inhibitor NVP-LDE-225, has been effective in curbing melanoma growth in vitro. Additionally, NVP-LDE-225 not only induced apoptosis in vitro but also inhibited tumor growth in xenograft models (164). SMO inhibitors like Vismodegib and Sonidegib have been approved for clinical use; however, their effectiveness has been limited by the development of drug resistance and notable side effects (165). Recent research by Pietrobono et al. led to the development of novel SMO inhibitors based on acylguanidine or acylthiourea scaffolds, which specifically target SMO in melanoma cells, thereby reducing Gli1 expression, inducing DNA damage and apoptosis, and inhibiting the self-renewal of MSCs (166). In the presence of Hh pathway repression, gene expression data also revealed compensatory upregulation of two other developmental pathways, Notch and Wnt (164). The Notch signaling pathway has also been explored as a therapeutic target. Chanh et al. demonstrated that RO4929097, a gamma-secretase inhibitor, reduced the oncogenicity and stem-like properties of melanoma cells in vitro, suggesting that Notch inhibition could be a promising strategy for targeting MSCs (67). Moreover, Notch inhibitors have shown potential as radiosensitizers, as their combination with radiation reversed the radiation-resistant phenotype of melanoma in vitro and reduced cell migration (167). Additionally, the FDA-approved antipsychotic drug Pimozide has shown efficacy in inhibiting the Wnt/β-catenin signaling pathway, displaying specific cytotoxicity against cancer cells, including melanoma (168). Previous studies have demonstrated that DAPT (a Notch inhibitor), Cyclopamine (an Hh signaling inhibitor), and XAV939 (a Wnt signaling inhibitor) (169), which target CSCs, may represent potentially effective strategies for treating patients with melanoma (22). In addition, Demcizumab (anti-Notch ligand, DLL4 antibody), OMP-52M51 (anti-Notch1 antibody), and OMP-18R5 (anti-Wnt receptor, FZD monoclonal antibody) are expected to be better therapeutic agents in the future against melanoma (21, 22).

While these approaches hold the potential for disrupting the tumorigenic properties of MSCs, they are not without challenges. The Hh, Notch, and Wnt pathways are integral to normal melanocyte development and other physiological functions, which increases the risk of toxicity when employing inhibitors against these pathways. Therefore, the clinical development of these targeted therapies will require a careful approach, emphasizing precise drug delivery systems and guided administration to minimize off-target effects and improve therapeutic selectivity. Future research should focus on optimizing these delivery strategies and validating their safety and efficacy in clinical settings to ensure that the benefits outweigh potential risk.

The diverse biological characteristics of MSCs are closely influenced by their surrounding microenvironment, which plays a crucial role in maintaining their stem-like properties and contributing to therapy resistance (170). Studies have shown that modulating this microenvironment can significantly impact the behavior of MSCs. For example, Kim et al. have demonstrated that SDF-1/CXCR4 played an important role in the lymphatic metastatic microenvironment of chemoresistant melanoma cells. Targeting this axis could potentially inhibit the lymphatic metastasis of CD133(+) chemoresistant melanoma cells, offering a promising therapeutic approach (171). Moreover, research indicates that TAMs can interact with CSCs, including MSCs, to promote therapeutic resistance. Mitchem et al. showed that targeting TAM could effectively overcome CSC-mediated therapeutic resistance (172).

In addition to TAMs, other immune cells, such as Treg, are also capable of promoting stemness and treatment resistance in tumor stem cells and can serve as potential targets for the immune microenvironment modulation of CSCs (86). Targeting the hypoxic microenvironment has also shown promise in treating MSCs, with both HIF-1α and HIF-2α emerging as valuable targets (173). Kim et al. found that the HIF-1α inhibitor IDF-11774 inhibited the growth and metastasis of B16F10 melanoma by downregulating HIF-1α expression (174).

Advances in CAR-T cell therapy have introduced new strategies for targeting MSCs. CAR-T cells engineered to recognize CSC surface markers have shown promising results (175). CAR-T cells targeting TYRP1 demonstrated enhanced anti-tumor activity in vitro and in vivo in both skin and rare melanoma subtypes, with minimal toxicity observed in preclinical models, and Phase I clinical trial is currently in preparation based on these promising outcomes (176).

Recent studies have also indicated that CSC-based vaccines could inhibit the metastasis of primary tumors by inducing humoral and cellular immune responses that lead to the lysis of these targeted CSCs (177). Yin et al. developed a novel Melanoma stem cells (MSCs)-based vaccine that induces CD8+ T cells to specifically target MSCs. The vaccine was found to promote the maturation of DCs, activate CD8+ T cells, inhibit the expression of CTLA-4, PD-1 and Tim-3, and increase the expression of IFN-γ and GzmB in CD8+T cells. The specific targeted killing effect of the vaccine inhibited melanoma growth and metastasis (178). DCs, as the most powerful antigen- presenting cells, have also been used in experimental models and various clinical trials to induce effective antitumor immune responses (25). For example, a novel vaccine targeting CD133(+) CD44(+) MSCs in melanoma, expressing a 6kDa early secretory antigenic target (ESAT-6) and secreting interleukin (IL)-21, significantly suppressed melanoma growth and metastasis in mice, extending survival (179).

ICIs and monoclonal antibodies have reactivated cancer immunotherapy (180, 181). However, while some patients showed partial or even complete responses to immunotherapy, a substantial number did not respond favorably (181). Researchers believed that cancer vaccines may produce tumor-specific T cells, which may be enhanced by ICIs that counteract immunosuppressive mechanisms (182). To investigate this hypothesis, Zheng et al. explored this strategy by combining a MSC-lysate pulsed dendritic cell (MSC-DC) vaccine with PD-L1 and CTLA-4 inhibitors in a mouse model. This dual immune blockade led to significantly greater tumor regression compared to the vaccine alone, suggesting that this combination could enhance immune-mediated anti-tumor responses (182).