As the most well-studied epigenetic modification in cancer, DNA methylation is the process of 5-methylcytosine (5-mC) formation, which is dynamically regulated by DNA methyltransferases (DNMTs) and ten-eleven translocation (TET) family members that are responsible for the transfer of methyl groups to cytosine residues and DNA demethylation via the oxidative catalysis of 5-mC to form 5-hmC (58). The downregulation of 5-hmC induced by the reduction of isocitrate dehydrogenase 2 (IDH2) and TET family members is a valuable epigenetic marker with relatively high diagnostic and prognostic significance in melanoma (59). Quite a few genes implicated in cell differentiation, epithelial–mesenchymal transformation, PI3K/mTOR signaling, and cell metabolism are differentially methylated in the promoter regions between melanomas and benign nevus (60, 61). Genome-wide DNA methylation analysis also unveils the epigenetic heterogeneity of melanoma metastases in specific organs (62). Moreover, several studies have demonstrated the engagement of DNA methylation in the regulation of specific oncogene or tumor suppressors in melanoma. For example, Field et al. showed that canonical tumor suppressor BAP1 is hypermethylated in uveal melanoma. In addition, integrative analysis reveals that SOX9 is methylated and lowly expressed in the highly proliferative group, and SOX9 overexpression results in decreased proliferation but increased invasion of melanoma cells (15). Therefore, DNA methylation plays a versatile role in mediating various aspects of melanoma cell biology.

In addition to tumor cell biology, the dysregulation of DNA methylation plays a critical role in regulating the immunologic characteristic of melanoma cells from various dimensions, including antigen presentation, inflammatory response pathways, and immune checkpoint molecules.

The effective recognition and targeting by tumor-infiltrating lymphocytes highly rely on the presence and levels of HLA class I antigen expression on the surface of tumor cells. The loss of the HLA class I antigen could help tumor cells to escape this killing effect, which to some extent results from the hypermethylation of the related genes (63, 64). This regulatory relationship is also supported by the finding that the treatment with DNA methylation inhibitor 5′-aza-2′-deoxycytidine (DAC) greatly contributes to the restoration of cell surface expression of HLA class I antigens and potentiates tumor cell recognition by MAGE-specific cytotoxic T lymphocytes (CTLs) (65). It is tempting to speculate that the hypermethylation-induced lack of HLA class I expression is the cause of the impaired response to vaccination, which provides a new route of escape from immune recognition. Similarly, Radosevich et al. also provided evidence that class II MHC molecules driven by the gene encoding the class II transactivator (CIITA) could also be significantly restored after the treatment with DAC in uveal melanoma cells, especially when stimulated by IFN-γ (66). Therefore, DNA methylation is a critical paradigm for the regulation of HLA antigens in melanoma cells.

The disordered DNA methylation in melanoma cells can also regulate inflammatory response pathways to modulate the immunologic characteristics. Chiappinelli et al. unveiled that the application of DNMT inhibitor prominently induces the expression of endogenous retrovirus (ERV) genes and the cytosolic sensing of double-stranded RNA (dsRNA) that cause a type I interferon response. Bioinformatics analysis displays that high viral defense signature expression in tumors is significantly associated with the durable clinical response of immunotherapy. Notably, DNMT inhibitor treatment sensitizes tumors to anti-CTLA-4 therapy in a preclinical melanoma model, which is attributed to the re-activation of the interferon response pathway (67, 68). Additionally, a recent study carried out by Falahat et al. with the employment of genome-wide DNA methylation profiling has noted that the promoter hypermethylation of cGAS and STING genes leads to their coordinated transcriptional silencing and contributes to the widespread impairment of the STING signaling function in clinically relevant human melanoma. Pharmacologic inhibition of DNA methylation could reinstate functional STING signaling, which is effective to restore STING signaling to improve the antigenicity of tumor cells through the upregulation of MHC class I molecules, facilitating the recognition and killing of melanoma cells by cytotoxic T cells within TME (18). Thus, epigenetic reprogramming of tumor cell-intrinsic STING function by modulating DNA methylation might be a promising approach to increase the efficacy of T cell-based immunotherapy. Moreover, it has been revealed that the deletion of TET2 in murine melanoma significantly reduces chemokine expression and TILs, thus enabling tumors to evade antitumor immunity and resist anti-PD-L1 immunotherapy. Conversely, to potentiate TET activity by systematic injection of its co-factor ascorbate or vitamin C could increase chemokine and TILs, leading to enhanced antitumor immunity and anti-PD-L1 efficacy and extending the lifespan of tumor-bearing mice. This investigation also suggests TET activity as a biomarker for predicting the efficacy and patient response to anti-PD-1/PD-L1 therapy (69).

Beyond the effect on tumor intrinsic pro-inflammatory pathways, a series of recent investigations demonstrate the involvement of DNA methylation in regulating the expressions of multiple immune checkpoint molecules like PD-L1, PD-L2, CTLA-4, LAG3, TIM-3, and Galectin-9. For example, the integrated genomic analysis identified that global DNA methylation regulates PD-L1 expression in melanoma. The inhibition of global methylation by decitabine increases PD-L1 expression in melanoma cells (70). Furthermore, PD-L2 expression is also under the control of DNA methylation. The methylation status of PD-L2 gene can be used to predict the progression-free survival of melanoma patients who have received anti-PD-1 antibody treatment (71, 72). Similar to PD-L2, the methylation status of CTLA-4 is also inversely correlated with its mRNA level in melanoma tissue. The evaluation of CTLA-4 methylation provides paramount information for selecting patients more likely to respond to immunotherapy, especially to anti-CLTA-4 antibody treatment (73, 74). Moreover, several studies also illustrate the effect of DNA methylation on the mRNA level of LAG3, TIM-3, and Galectin-9 (75, 76).

In aggregate, the immunophenotypes of melanoma cells are largely influenced by DNA methylation, which is thus involved in regulating antitumor immunity and the response to immunotherapy ( Table 1 ).

The alteration of DNA methylation also displays a prominent effect on the features and functions of immune cells within TME. DNA methylation profile analysis of melanoma tissues has identified a subgroup of patients with methylation patterns resembling stromal cells and leukocytes, an overexpressed immune signature, and improved survival rates (77). Furthermore, the global methylation profile is regarded as an independent indicator of tumor cell immune evasion and response to immune checkpoint inhibition in melanoma (89, 90). The analysis of methylation profile then more specifically focuses on tumor-infiltrating immune cells with the method to extract methylation of immune cell type-specific genes from genome-wide methylation arrays. Intriguingly, multiple immune methylation clusters are related to patients’ survival. Moreover, low-dimensional projection based on immune cell type-specific methylation reveals the grouping of the solid tumors in line with melanoma immune methylation clusters rather than tumor types (91). These results highlight the close linkage between immune methylation signature and immune cell function.

In terms of the mechanistic study, heat shock protein gp96 engages conventional and plasmacytoid DCs (pDCs) through CD91, with the involvement of global DNMT-dependent epigenetic modifications, to change the protein expression within these antigen-presenting cells (78). As a result, pDCs upregulate neuropilin-1 to enable the long-term interactions of pDCs with Treg cells, thereby enhancing the suppression of Th1 antitumor immunity. Moreover, a recent report pointed out that pharmacological agent leading to DNA hypomethylation in CD8⁺ T cells vastly regulates activation-related transcriptional networks, which helps to potentiate effector cytokine production and antitumor activity. Therefore, in vivo employment of DNA hypomethylating agent promotes the infiltration of CD8⁺ T cells and suppresses tumor growth via CD8⁺ T cell-dependent cytotoxicity in melanoma (79). Taken together, the antitumor capacity of immune cells within TME is tightly regulated by DNA methylation in melanoma, which underpins developing therapeutic strategy of combining DNA methylation-regulatory agent and immune checkpoint blockades ( Table 1 ).

Chromatin flexibly switches from a dense heterochromatin state with poor transcriptional activity to a relaxed euchromatin state with high transcriptional activity. Histone modification usually occurs at the N-terminal “tail” of histones, displaying prominent effects on shaping chromatin structure and regulating gene expression (92). Typical histone modification paradigms include acetylation, methylation, phosphorylation, and ubiquitination. Specifically, histone acetylation refers to the addition of acetyl-to-lysine residues, which is reversely regulated by histone acetyl transferases (HATs) and histone deacetylation enzymes (HDACs) (93). Furthermore, histone methylation occurs at lysine or arginine residues at specific sites of histone H3 and H4, dynamically modulated by lysine-specific demethylase (LSD) and a variety of methyl-transferases (94). A recent systematic epigenomic profiling analysis found frequent loss of histone acetylations and H3K4me2/3 on regulatory regions proximal to cancer-regulatory genes implicated in melanoma-driving pathways, suggesting the extensive participation of histone acetylation and methylation in melanoma (12, 13). The expression levels of histone acetylation-modulatory enzymes are frequently dysregulated in melanoma (95–97). Our group has previously reported that histone deacetylase SIRT6 plays a bimodal role in melanoma via regulating IGF1R-Akt axis and the related autophagy (98). In addition, histone acetyltransferase P300 directly binds to the promoter region of melanocytic-lineage factor MITF to enhance mitochondrial oxidative phosphorylation and thereby melanoma growth (99). Additionally, the role of histone methylation in melanoma biology has also been well illustrated. H3K9 demethylases LSD1 and JMJD2C jointly contribute to melanoma progression by facilitating the transcription and expression of E2F target genes (100). Intriguingly, H3K9 methylation regulated by SETDB1, LSD1, and JMJD2C appears to play quite opposite roles in the occurrence and development of melanoma (12, 13). While SETDB1-induced increase of H3K9 methylation exerts a prominent carcinogenic effect in established tumors (101, 102), the restoration of H3K9 methylation by targeting LSD1 and JMJD2C abolishes the malignant transformation from melanocytes to melanoma driven by oncogenic mutations (103). Therefore, histone methylation plays a bimodal role in melanoma carcinogenesis and development, and targeting histone methylation should conform to the stage status in melanoma therapy.

Apart from tumor cell biology, the dysregulation of histone modification also exerts prominent influences on the immunologic characteristic of melanoma cells, including antigen presentation, immune checkpoint molecules, and inflammatory pathways.

The initial investigation on the role of histone modification in tumor cell immunogenicity was centered on the expression of genes related to MHC class I or class II antigen presentation. Khan et al. firstly reported that histone deacetylase inhibitors could induce the expression of molecules like TAP1, TAP2, LMP2, LMP7, and Tapasin, which are involved in antigen processing and presentation via the MHC class I pathway in melanoma cells (80). The facilitative role of HDAC inhibitor trichostatin A on the expression of antigen processing machinery is forwardly confirmed by another study (81). Of note, the in vivo antitumor effect of TSA is found to be entirely dependent on the alteration of immunogenicity of tumor cells, as the tumor-suppressive effect does not occur in immune-deficient mice (81). However, the pharmacological toxicity experiment reveals that the IC₅₀ of pan-HDAC inhibitor panobinostat is much lower in peripheral blood mononuclear cells (PBMCs) than that in melanoma cells, indicating that HDAC inhibitor is cytotoxic in PBMCs at concentrations much lower than those required for melanoma antitumor activity (104). Therefore, a more detailed analysis of HDAC family molecules in melanoma tumor immunology was further carried out. Woan et al. reported that HDAC6 mediates the expression of tumor-associated antigens and MHC class I expression. Targeting HDAC6 helps to achieve a pronounced delay of in vivo melanoma tumor growth that partly depends on an intact immunity (105).

In addition to the effect on antigen presentation, HDAC inhibitor also prominently regulates immune checkpoint expression (e.g., PD-L1) in the tumor. Mechanistically, HDAC inhibitor induces rapid upregulation of histone acetylation of the PD-L1 gene to elicit enhanced and durable gene expression, which thereby helps to increase the efficacy of anti-PD-1 antibody (20). Intriguingly, histone methylation is another important paradigm implicated in tumor cell immunologic characteristics. The ablation of LSD1 increases tumor immunogenicity by upregulating ERV transcripts and suppressing the RNA-induced silencing complex, which is responsible for the diminished resistance to anti-PD-1 immunotherapy in melanoma (82). According to this notion, targeting LSD1 is of great translational potential in optimizing the efficacy of immunotherapy. Moreover, the expression of histone methylase EZH2 is highly associated with suppressed IFN-γ signature in the tumor. EZH2 inhibitors may be most effectively targeted to immunologically cold melanoma to achieve both the direct cytotoxicity and increased immune responses in the context of checkpoint inhibitor immunotherapy (83).

Taken together, dysregulated histone acetylation and methylation in melanoma regulate tumor cell immunogenicity through manipulating various approaches including antigen presentation, immune checkpoint molecules, and inflammatory pathways.

Accumulating evidence has revealed that histone modification is also central to regulating the function of different immune cells in TME, including T cells, NK cells, and macrophages.

Histone acetylation and methylation were previously proved to impact the antitumor activity of different subtypes of T cells. Specifically, the activation of histone deacetylase SIRT1 might impede the differentiation of Th17 cells via reducing STAT3 acetylation and relevant secretion of pro-angiogenic factors, thus restraining melanoma growth (84). Furthermore, while pan-HDAC inhibitor displays toxicity on immune cells, specific inhibition of HDAC6 by ACY-1215 results in the downregulation of the Th2 transcription factor GATA3, upregulation of the Th1 transcription factor T-Bet, accumulation of central memory phenotype T cells, reduced exhaustion-associated phenotypes, and enhanced killing in mixed lymphocyte reactions. Therefore, the observation that HDAC6-selective inhibitor augments T-cell immune properties in melanoma patients provides a rationale for translational investigation of their potential clinical efficacy (85). Additionally, Knox et al. proved that HDAC6 inhibitor facilitates the transition of the TME from “cold” to “hot” and profoundly synergizes with immune checkpoint blockade therapies in controlling tumor growth. Of note, the combined effect is associated with enhanced infiltration of immune cells, increased central and effector T-cell memory, and significantly reduced pro-tumorigenic M2 macrophages within TME (106). Similarly, histone methylation also regulates the function of infiltrating T cells. It is reported that nuclear LSD1 phosphorylated at serine 111 (nLSD1p) is enriched in PD-1⁺CD8⁺ T cells from resistant melanoma patients. Targeting the LSD1p nuclear axis effectively confers the infiltration of IFN-γ/TNF-α-expressing CD8⁺ T cells into the tumor, which is further augmented by combined immunotherapy. In addition, nLSD1p is regulated by the key T-cell exhaustion transcription factor EOMES in dysfunctional CD8⁺ T cells, and nLSD1p regulates EOMES nuclear dynamics via demethylation/acetylation switching of critical EOMES residues. Accordingly, there is a positive feedback loop consisting of LSD1 and EOMES in impairing the antitumor capacity of CD8⁺ T cells (87).

The dysregulation of histone modification also affects the function of alternative types of immune cells in melanoma. Systemic overexpression of SIRT2 promotes melanoma progression in immune-competent mice by suppressing the infiltration and function of NK cells in TME. Pharmacological inhibition of SIRT2 potentiates NK cell-dependent antitumor immunity and the growth of allograft melanoma (86), which might be of translational implication in increasing the efficacy of immunotherapy. Moreover, EZH2-mediated histone H3 methylation of HIF1α in macrophages leads to the silence of HIF1α expression and reprograms the immune suppressive microenvironment to the facilitative one, which alleviates the tumor burden and prolongs the overall survival of mice implanted with tumor (88).

Taken together, histone modification simultaneously affects tumor cell immunogenicity and the function of immune cells within TME, so as to mediate tumor immunology mainly through these two approaches ( Table 1 ).

In addition to DNA methylation and histone modification, non-coding RNA and m⁶A RNA methylation are also critical types of epigenetic modification. The expression profiles of microRNAs, lncRNAs, and circRNAs in melanoma have been extensively studied and are found to be associated with a variety of tumor hallmarks (107–110). Studies of miRNAs in melanoma initially focused on their roles as potential biomarkers (111). For example, the level of either serum or tissue miR-16 is highly associated with the American Joint Committee on Cancer (AJCC) stage and the prognosis of melanoma patients (112). Thereafter, the function of miRNAs in angiogenesis, metastasis niche formation, and T-cell dysfunction within TME has also received extensive attention recently. For example, Pencheva et al. uncovered a convergent and cooperative miRNA network consisting of miR-1908, miR-199a-5p, and miR-199a-3p, which drive the metastasis and angiogenesis in melanoma (113). In addition, miR-30b/30d promotes the metastatic behavior of melanoma cells by directly targeting the GalNAc transferase GALNT7, leading to reduced immune cell activation and recruitment (114). Of note, the biological functions of some specific non-coding RNAs are highly correlated with lineage-specific factors in melanoma. In particular, the sequestration of miR-16 by the mRNA of pigmentation-associated enzyme TYRP1 promotes tumor growth (115, 116). Moreover, a recent annotated lncRNA SAMMSON is regulated by melanocytic-specific transcriptional factor SOX10, and the deficiency of SAMMSON is capable of restraining tumor progression and increasing the efficacy of MAPK-inhibition targeted therapy (109, 117, 118).

In addition to non-coding RNAs, m⁶A RNA methylation is another critical chemical modification found in mRNA and non-coding RNAs in eukaryotic cells (119). Jia et al. discovered that the downregulation of the m⁶A level is a characteristic of ocular melanoma and indicates a poor prognosis (120). However, the increased m⁶A RNA methylation of BACE2 induces its upregulation, which accelerates the tumorigenesis of ocular melanoma (121). These two reports indicate the paradoxical role of m⁶A methylation in ocular melanoma. In uveal melanoma, the expression of the main m⁶A regulatory enzyme METTL3 is found to be significantly increased. The knockdown of METTL3 prominently suppresses cell proliferation, colony formation, migration, and invasion via the downregulation of c-Met (122), indicating that METTL3 is a promising target for the treatment of uveal melanoma.

The roles of ncRNAs in melanoma biology have been well elucidated. Recently, some investigations have focused on clarifying the role of ncRNAs in melanoma immunology, as well as the underlying mechanism. Through the transcriptional analysis of the process responsible for the conversion of monocytes into MDSCs mediated by melanoma extracellular vesicles (EVs), a set of microRNAs (miR-146a, miR-155, miR-125b, miR-100, let-7e, miR-125a, miR-146b, and miR-99b) is found to be markedly associated with the function of MDSCs and the resistance to treatment with immune checkpoint inhibitors in melanoma patients (123). Forwardly, Mastroianni et al. reported that miR-146a controls the immune response in melanoma TME. MiR-146a level is significantly increased in melanoma tissue. Furthermore, T cells isolated from miR-146a-knockout mice exhibit higher expression of miR-146a-targeted gene STAT1 and STAT1-regulated cytokine IFN-γ. Interestingly, the combined inhibition of PD-1 and miR-146a is proved as a promising therapeutic approach to enhance the antitumor immune response elicited by checkpoint therapy (124). In terms of CD8⁺ T cell-dependent tumor immunology, the senescence and exhaustion of T cells are the major barriers toward successful cancer immunotherapy. MiR-155 is identified to enhance the antitumor function of CD8⁺ T cells by restraining T cell senescence and functional exhaustion through epigenetic silencing of drivers of terminal differentiation. Mechanistically, miR-155 enhances polycomb repressor complex 2 (PRC2) activity indirectly by promoting the expression of the PRC2-associated factor Phf19 through the downregulation of the Akt inhibitor, Ship1 (125). Martinez-Usatorre et al. then elucidated the mechanism responsible for the increased expression of miR-155 in tumor-infiltrating CD8⁺ T cells and demonstrated that high miR-155 expression in tumor-infiltrating CD8⁺ T cells may be a surrogate marker of the relative potency of in situ antigen-specific CD8⁺ T-cell response (126). Apart from the role of intrinsic miRNAs within CD8⁺ T cells, CD8⁺ T cells can also internalize exosomes from melanoma cells in TME, which induces the downregulation of T-cell response through decreased T-cell receptor (TCR) signaling and diminished cytokine and granzyme B secretion. Therefore, miRNAs in melanoma-derived exosomes also contribute to tumor immune evasion and could be a valuable therapeutic target (127).

Additionally, several recent reports have revealed the close linkage between m⁶A RNA methylation and antitumor immunity in melanoma ( Figure 1 ). Yang et al. reported that under metabolic starvation stress, fat mass and obesity-associated protein (FTO) is prominently induced under the control of autophagy and the NF-κB pathway in melanoma. The genetic knockdown of FTO induces the increase of m⁶A methylation in the critical pro-tumorigenic melanoma cell-intrinsic genes including PD-1, CXCR4, and SOX10, which results in an increased RNA decay through the m⁶A reader YTHDF2. Therefore, the deficiency of FTO can help to sensitize melanoma cells to IFN-γ stimulation and increase the efficacy of anti-PD-1 immunotherapy in mice bearing melanoma xenograft tumor, which is associated with the activation of adaptive immunity (128). This study takes the lead in unveiling the crosstalk between m⁶A RNA methylation and tumor cell immunogenicity in melanoma. Then, the role of another m⁶A demethylase ALKBH5 is elucidated in melanoma immunology. In particular, ALKBH5 modulates Mct4/Slc16a3 expression and lactate content in TME, so that the composition of tumor-infiltrating Tregs and MDSCs is altered. Of note, the ALKBH5 gene mutation and expression status of ALKBH5 in melanoma tissues are highly correlated with the response to immunotherapy in patients, indicating that ALKBH5 is not only a therapeutic target but also a reliable biomarker in predicting the efficacy of immunotherapy (129). Moreover, the depletion of methyltransferases METTL3 and METTL14 in tumor cells is found to prominently induce the activation of the IFN-γ-STAT1-IRF1 signaling axis through the stabilization of the STAT1 and IRF1 mRNA via YTHDF2. This regulatory effect leads to the elevated secretion of IFN-γ, CXCL9, and CXCL10, as well as the enhanced infiltration of cytotoxic CD8⁺ T cells within TME, ultimately resulting in potentiated response to anti-PD-1 antibody treatment. Notably, there is a negative correlation between METTL3/METTL14 and STAT1 in tumor tissues of 59 patients. In this case, METTL3 and METTL14 are also potential therapeutic targets in anticancer immunotherapy (130). Furthermore, the role of METTL3 in melanoma immunology is also attributed to its effect on tumor-infiltrating macrophages. METTL3-depleted macrophages were observed to reshape the TME via the enhancement of M1- and M2-like TAM and Treg infiltration in a transgenic mouse model with genetic ablation of METTL3 in myeloid cells. This effect is caused by the impairment of YTHDF1-mediated translation of SPRED2, which induces subsequent activation of NF-κB and STAT3 through the ERK pathway, thus leading to precipitated tumor growth and metastasis. Intriguingly, the therapeutic efficacy of PD-1 checkpoint blockade is robustly attenuated in METTL3-deficient mice, indicating METTL3 as a potential therapeutic target for tumor immunotherapy, including in melanoma (131). Combined, these reports highlight that m⁶A RNA methylation is an important paradigm of epigenetic modification involved in the regulation of tumor immunology in melanoma, from the perspective of either tumor cell immunogenicity or the functions of tumor-infiltrating cells like CD8⁺ T lymphocytes and macrophages.