1 Introduction

Melanoma is one of the most common tumours worldwide, exhibiting significant diversity and plasticity [1, 2]. Advanced melanoma typically leads to poor prognosis, and recent exciting progress has been made in the research of advanced melanoma and therapy-resistant melanoma. In addition to traditional chemotherapy, the US Food and Drug Administration (FDA) has approved a range of novel anti-melanoma drugs, including targeted therapy against point mutations and immune checkpoint therapy [3]. Simultaneously, a series of combination treatment strategies have entered phase III clinical trials [4, 5].

The study of the gene expression regulatory mechanisms in melanoma is of great significance for understanding the development of the disease. Many papers describe gene variations that affect cell signaling, transcription factors (TFs), epigenetic modifications, chromatin regulatory factors, and chromosomal structures, which may form a complex network [6]. The transcription program operating in normal or tumour cells determines the characteristics of the cells, established through sequential gene expression processes involving major TFs and more signaling molecules [7]. In addition to directly affecting the characteristics of melanoma itself, such as proliferation and metastasis, abnormalities in transcription regulation may indirectly impact the metabolism and immune microenvironment of tumour cells, resulting in poor response to treatment and prognosis [8].

This review provides a summary of the identified changes in dysregulated transcription programs in melanoma. We particularly emphasize the key components mediating transcriptional changes in proliferative or metastatic melanoma [1] which may contribute to controlling advanced melanoma [9]. Delving into the complexity of the melanoma genome and exploring dysregulated transcription programs as key drivers of tumour development aims to provide a comprehensive overview to advance melanoma research and the development of targeted therapeutic strategies.

2 Transcriptional programs in normal melanocytes

2.1 The origin of melanocytes

Melanocytes in the human body originate during neurogenesis in the embryonic stage, typically around 6–8 weeks, known as melanoblasts [6]. These melanoblasts develop from neural crest cells and migrate to various locations, including the eyes, inner ears, heart, adipose tissue, mucosa, brain, and skin [10]. Once within the epidermis, melanoblasts proliferate and differentiate, leading to colonization of the entire embryonic skin. During the 15th-24th weeks of human embryonic development, hair follicles begin to form. During this period, a portion of mature melanocytes differentiate and migrate to the bulb of the hair follicles, while others differentiate into melanocyte stem cells (MSCs), a population of embryonic progenitors with self-renewal capability, which also migrate and settle in the hair follicles [11, 12]. In the subsequent developmental process, melanocyte stem cells provide mature melanocytes to the hair bulb to produce pigment for the newly formed hair follicles, as well as supplying mature melanocytes to the interfollicular epidermis, serving as a reservoir for mature melanocytes in adult skin [13]. Initially existing mature melanocytes in the dermis undergo asymmetric division during the embryonic period to supplement the mature melanocytes in the epidermis and gradually disappear after childhood [14]. However, mature melanocytes persist in the dermis, resulting in blue or gray patches on the skin, known as nevi or Mongolian spots [14]. Therefore, melanin in adult skin is secreted by epidermal melanocytes, and two populations of melanocytes exist in hair follicles: mature melanocytes capable of producing melanin and melanocyte stem cells that serve as a reservoir for mature melanocytes [6].

2.2 Roles of MITF in melanocyte development, survival and proliferation

Microphthalmia-associated transcription factor (MITF) is a transcription factor that regulates the transcriptional expression of downstream genes by binding directly to DNA or in conjunction with cofactors. MITF is a basic helix-loop-helix-zip (bHLHZip) transcription factor encoded on chromosomes 3p12.3–14.115. In the process of melanocyte development, MITF plays a crucial role. MITF may be the first transcription factor to appear in specifying melanocytes during development [15]. In humans, melanoblasts first emerge as epidermal cells expressing MITF in embryos at 6–8 weeks of gestation. During melanoma development in mice, MITF is detected at approximately embryonic days 9–10, along with DCT, marking the establishment of melanocyte specificity [6, 16]. Loss of MITF leads to the complete absence of neurocristic pigment cells, resulting not only in a lack of pigmentation but also in abnormal development of the ears and eyes, as observed in patients with type 2A Waardenburg syndrome (WS) and Tietz syndrome [17].

MITF is the most important transcription factor in melanoma development and the maintenance of melanocyte function. During human embryonic development, mature melanocytes actively proliferate in the epidermis and, under the influence of various transcription factors, including MITF, acquire the enzymes necessary for melanin production, eventually differentiating into mature melanocytes [6]. Under UV irradiation, MITF responds to upregulated α-melanocyte-stimulating hormone (α-MSH) and key melanocortin 1 receptor ligand (MC1R) signaling pathway to induce the expression of pigment-producing enzymes, including tyrosinase (TYR), tyrosinase-related protein 1 (TYRP1), and dopachrome tautomerase (DCT), participating in the regulation of melanin production [18, 19]. Another pathway regulating melanin production, the stem cell factor (SCF)-KIT pathway, also converges on MITF to activate genes involved in melanin production [17, 20].

MITF promotes differentiation-related functions in melanocytes, regulating genes associated with pigmentation such as TYR, TYRP1, DCT, MLANA, SILV, and SLC24A5, as well as cell adhesion-related genes including cancer embryonic antigen-related cell adhesion molecule 1 (CAECAM), and genes involved in melanosome transport such as RAB27a and MYOSIN5a (MYO5a) [15, 21]. Moreover, MITF plays a crucial role in the survival of melanocytes. It regulates survival through BCL-2 (an anti-apoptotic member of the BCL2 family), and decreases in MITF below a threshold level result in reduced BCL2, potentially inducing apoptosis in mature melanocytes [22]. The role of MITF in melanocyte proliferation is multifaceted. An optimal level of MITF promotes melanocyte proliferation, whereas low levels of MITF are associated with dedifferentiation, increased invasiveness, and elevated levels of p27 (CDKN1B), a cyclin-dependent kinase inhibitor protein, leading to reduced proliferation. Conversely, cells with high MITF activity induce cell cycle arrest through increased expression of p21 (CDKN1A) and p16 INK4a (CDKN2A), which negatively impacts cell proliferation [15, 23].

2.3 Roles of other transcription factors in melanocytes

In the differentiation, melanogenesis, survival, and proliferation of normal melanocytes, various transcription factors other than MITF play important roles. Sry Related HMG box 10 (SOX10) is a high-mobility group (HMG) transcription factor involved in melanocyte migration [24]. Concurrently, SOX10 direct activation of MITF transcription.

Additionally, signal transducer and activator of transcription 3 (STAT3) may directly contribute to melanocyte proliferation and may play a crucial role in α-MSH-induced melanin production [25, 26]. cAMP response element binding protein (CREB), as a direct effector of protein kinase A (PKA), plays an important role in the MC1R-induced melanin production pathway [27]. Recent studies have shown that the CNC family transcription factor Nrf3 is expressed in the basal layer of the epidermis, where melanocytes are located, and it coordinates melanin production through phagocytosis and autophagy [28].

UV radiation cause DNA damage and affect the specificity of TFs binding to DNA. Intermittent high-intensity UV radiation during childhood has been identified as a significant modifiable risk factor for melanoma in the white race [29]. But for the colored skin, the association between UV exposure and melanoma was weaker [30]. UV exposure reduces the DNA binding specificity of transcription factors (TFs) and leads to competition between TFs and DNA repair proteins, and as a result, this interference with DNA repair by TFs promotes melanoma development through UV-induced mutations [31].

3 Transcriptional dysregulation and the impacted factors in melanoma

3.1 Key transcription factors

Transcription factor dysregulation, epigenetic modification, and metabolic reprogramming all contribute to melanoma development (Fig. 1). First and foremost, abnormal expression of key transcription factors plays a significant role in the development of melanocytes into melanoma cells [32]. MITF, isolated approximately 30 years ago, has been extensively studied by numerous scientists and found to serve as a key coordinator of various aspects of melanocyte and melanoma biology [15]. In normal melanocytes, MITF regulates melanocyte development, melanin synthesis, and melanocyte survival [6]. In melanoma cells, MITF plays a crucial role in determining the phenotype of melanoma cells, with high MITF levels associated with a more proliferative state and low MITF levels linked to increased invasiveness, but both are capable of tumour formation [21, 33].

Fig. 1
figure 1

Dysregulation of transcription factors, epigenetic modification, and metabolic reprogramming in melanoma development. A During the process of melanocyte development into melanoma cells, transcription factors such as MITF, SOX10, and SOX9 become dysregulated. B Epigenetic modifications of genetic material such as DNA methylation, histone acetylation, chromatin remodeling, and non-coding RNAs. C Metabolic reprogramming of carbohydrates, lipids, and proteins in a relatively hypoxic and nutrient-limiting environment

Full size image

Early studies on the systematic analysis of the cancer genome revealed that MITF is a target of amplification in melanoma cells [34]. MITF amplification has been identified in approximately 20% of melanoma populations, and studies on melanoma cell lines have shown that superenhancer, BRAF mutations and p16 inactivation can lead to MITF amplification [35]. MITF amplification is more prevalent in metastatic disease and is associated with decreased overall patient survival; however, it is not related to the chemotherapy response in melanoma patients [34, 36]. Recently, single-cell sequencing data has revealed the upregulation of MITF in acral melanoma, promoting lymph node metastasis through the activation of the fatty acid oxidation (FAO) pathway [37]. In addition to gene amplification, mutations in the MITF gene are also associated with melanoma development, where MITF is like an oncogene in melanoma [32]. The E318K mutation in MITF can lead to both familial and sporadic melanoma [38], while the MITF-E87R mutation promotes tumour progression by enhancing invasion through increased expression of S100A4 and reduced adhesion [39]. MITF is also regulated by factors other than genetic factors. High expression of MITF in melanoma is associated with Super enhancers (SE) [35]. SEs are clusters of enhancers bound by a high density of transcription factors and co-factors, often associated with genes that control and define cell identity [40, 41]. SOX10, a known transcriptional regulator of MITF, has been identified as an SE marker in some melanoma cell lines with high MITF expression [42, 43]. Several other transcription factors, such as PAX3, CREB, ZEB2, and LEF1, also promote MITF mRNA transcription in melanoma cells [44, 45]. Moreover, MITF can cooperate with lymphoid-enhancing factor-1 (LEF-1) as a non-DNA-binding co-activator to enhance its self-expression in melanoma [46]. Conversely, various transcription factors, including ATF2, c-MYC, HOXA1, GLI2, BRN2, and NF-κB can inhibit MITF mRNA expression [47]. In addition to various splice isoforms, the MITF protein undergoes extensive post-translational modification. Serine and threonine phosphorylation affect protein stability, ubiquitination possibly regulates target specificity and protein stability, acetylation may decrease MITF DNA-binding affinity and residence time, and tyrosine phosphorylation enhances MITF’s ability to regulate a range of target gene transcription [15, 48]. Therefore, MITF expression and regulation are complex processes.

SOX10, a lineage-specific transcription factor crucial for MITF expression, plays a significant role in melanoma cell growth [42]. It mediates phenotypic switching in melanoma, facilitating transitions between states with distinct proliferative, invasive, and drug-resistant characteristics. High expression of SOX10 positively regulates melanoma cell proliferation, contributing to tumour growth. Conversely, the loss of SOX10 results in a shift toward an invasive phenotype, marked by increased expression of mesenchymal genes [49]. This adaptability allows melanoma cells to thrive in diverse microenvironments. Moreover, SOX10 is linked to the immune phenotype in melanoma, with high SOX10 expression associated with lower PD-L1 levels - a sign of negative regulation of PD-L1 expression [50]. Research indicates that SOX10 serves as a highly specific marker for lymph node melanoma metastasis [51].

Similarly, SOX9, another member of the SOX family characterized by the conserved high-mobility group (HMG) DNA-binding domain, acts similarly in adult skin as SOX10 does during embryogenesis [17]. SOX9 also contributes to NEDD9 expression along with high SOX10, thereby enhancing metastatic properties. Elevated SOX9 expression restores the invasiveness of SOX10 knockdown cells in melanoma, and heightened SOX9 levels promote metastasis within heterogeneous melanoma populations [52].

Paired box 3 (PAX3) is a member of the paired box domain family of transcription factors [53], and LEF-1 is a transcription factor involved in the Wnt signaling pathway [54]. They all contribute to the growth and development of normal melanocytes through their interaction with MITF, tyrosinase-related protein-2 (TRP-2), tyrosinase (Tyr), and cyclin-dependent kinase 2 (Cdk2), but also participate in the malignant melanoma development [55]. Importantly, canonical Wnt signaling has been greatly implicated in the phenotype switching in tumor progression, and LEF1 is a crucial a transcription factor that promotes the canonical Wnt/β-catenin signaling pathway in melanoma [56].

3.2 Epigenetic modifications

Beyond gene mutations and TFs, epigenetics emerges as a pivotal player in the early diagnosis, disease monitoring, and prognosis evaluation of melanoma, offering new insights for targeted therapy [57]. DNA methylation, among the earliest and extensively studied epigenetic modifications, leads to gene silencing in melanoma, particularly in the highly methylated promoters of relevant tumour suppressor genes. This phenomenon may contribute to the initiation and progression of tumour cells [58, 59].

Histone acetylation typically correlates with activation of gene expression. Heightened activity of histone deacetylases (HDAC) can downregulate multiple tumour suppressor genes, fostering melanoma cell progression and inducing resistance to alkylating drugs [58]. Notably, acetylation of lysine 27 on histone H3 (H3K27ac) has recently emerged as a crucial mechanism regulating MITF expression, thereby enhancing the metastatic potential of melanoma cells. Additionally, histone methylation, occurring on lysine and/or arginine residues, may directly influence gene transcription, thus playing a direct role in the transformation of melanocytes into melanoma cells [60, 61].

Chromatin remodelling entails the reorganization of chromatin from a condensed state to a transcriptionally accessible state and relies on the activity of polycomb group (PcG) proteins. This process influences the development of melanoma cells by affecting the expression of relevant genes [58].

Non-coding RNAs are mainly divided into two categories: long non-coding RNAs (lncRNA) and small non-coding RNAs (sncRNA). Among sncRNAs, microRNAs (miRNA) (19–22 bps) primarily disrupt post-transcriptional gene expression by interfering with the 3’UTR region of mRNA [62]. SAMMSON, a lncRNA associated with melanoma, is considered one of the key lncRNAs linked with melanoma occurrence, with expression observed in over 90% of melanomas. SAMMSON regulates tumour metabolism by activating the mitochondrial p32 protein, thereby maintaining oxidative phosphorylation and mitochondrial homeostasis [63].

3.3 Metabolic reprogramming

Metabolic reprogramming stands out as a defining characteristic of cancer. Within tumours, enhanced aerobic glycolysis is a common phenomenon, and in melanoma, the prevalent BRAF mutation provides an additional boost to the glycolytic capacity of melanoma cells, resulting in tumour growth and progression [64]. Tumour-specific nitrite production can inhibit the tricarboxylic acid cycle in tumours, restoring the tumour immune microenvironment and sensitizing tumour cells to immune therapy [65]. Moreover, oxidative phosphorylation in mitochondria, regulated by MITF-coexpressed lncRNA SAMMSON, promotes melanoma growth by directly regulating the mitochondrial master regulator p32 [62]. However, high level of oxidative phosphorylation in mitochondria disturb melanoma occurrence.

Disruption of lipid metabolism emerges as another distinctive metabolic feature of melanoma. Reports indicate overweight/obese (OW/OB) patients with metastatic melanoma show unexpected improvements in prognosis after treatment with immune checkpoint inhibitors (ICIs) and BRAF-targeted therapy [66]. The higher expression of PD-1 on T cells in obese individuals enhancing sensitivity to anti-PD1 may explain this phenomenon. Peroxisomes, crucial players in lipid metabolism, are implicated, and mouse experiments suggest that targeting peroxisome biogenesis can heighten melanoma sensitivity to MAPK pathway inhibition, thereby overcoming MAPKi resistance [67].

Imbalances in amino acid metabolism, particularly involving serine, glutamine, and branched-chain amino acids (BCAA), are intricately linked to melanoma’s pathogenesis. 3-phosphoglycerate dehydrogenase (PHGDH) serves as the rate-limiting enzyme in de novo serine synthesis. Melanoma exhibits sensitivity to serine synthesis restrictions under nutrient-limiting conditions. Dietary supplementation of serine or increased PHGDH expression may promote melanoma progression. Glutamine playsv a vital role in normal cellular metabolism. Regional deficiency leads to excessive histone methylation, inducing cell dedifferentiation and enhancing melanoma cells’ resistance to targeted therapy. Conversely, dietary glutamine supplementation can inhibit melanoma growth in vivo through epigenetic suppression. Furthermore, branched-chain amino acids (BCAA), including leucine (Leu), isoleucine (Ile), and valine (Val), promote tumour cell proliferation by supporting acetyl-CoA production in the tricarboxylic acid cycle [64].

4 The melanoma plasticity and the relationship with treatment resistance

In normal developmental processes, cells undergo specific differentiation programs to generate cells and tissues with distinct functions. However, when exposed to particular environmental stimuli, cells with differentiation potential undergo non-heritable phenotypic changes, continuously adapting to the environment for survival [68]. The transition between epithelial and mesenchymal phenotypes, known as epithelial-mesenchymal transition (EMT), enables tumors to switch between proliferative and invasive phenotypes (Fig. 2A). This phenomenon is especially prevalent in tumour cells, enabling them to promptly respond to microenvironmental stress and treatment interventions, leading to metastasis and drug resistance in clinical practice [69]. Cutaneous malignant melanoma bears a high burden of gene mutations, besides genetic changes, melanoma lineage plasticity also plays an important role in the process of metastasis and drug resistance, which is affected by key factors like MITF, ZEB family, SNAI1, and twist-related protein 1 (TWIST1) & TWIST2, as well as epigenetic modifications such as DNA methylation and histone acetylation (Fig. 2). The multilayered regulation of phenotypic plasticity, including epigenetic and transcriptional control, is essential in determining and maintaining melanoma cell states [68, 70,71,72]. A comprehensive understanding of the dynamic nature of phenotypic plasticity is crucial for addressing two major unresolved clinical challenges – cancer metastasis and treatment resistance.

Fig. 2
figure 2

Mechanisms of lineage plasticity in melanoma. A Melanocytes exhibit two distinct phenotypes: those undergoing EMT transition to an invasive phenotype, and those undergoing MET transition to a proliferative phenotype. B Transcription factors involved in EMT, such as SNAI and ZEB, along with epigenetic modifications of genetic material, influence the phenotypic transition of melanoma cells. C Melanocytes undergoing EMT can form vascular structures and promote the growth of melanoma masses

Full size image

4.1 Epithelial-mesenchymal transition in melanoma regulated by EMT-TFs

The epithelial–mesenchymal transition (EMT) is characterized by the loss of epithelial phenotype and the acquisition of mesenchymal properties. Specifically, it involves the loss of cell–cell adhesion and apical-basal polarity functions, while cells acquire enhanced migratory and invasive mesenchymal characteristics to varying degrees [8, 73], melanoma cells undergoing EMT transformation lose e-cadherin, a crucial cell-adhesion glycoprotein [72]. The reverse process is described as mesenchymal- epithelial transition (MET). EMT is a common occurrence in various cancer cells, recent studies reveal that melanoma cells undergo EMT or MET, transitioning between proliferative and invasive phenotypes [68, 72, 74]. MITF is recognized as a key transcription factor in EMT [75]. Additionally, EMT is regulated by several other transcription factors (EMT-TFs) [76], including zinc finger E-box binding homeobox 1 and 2 (ZEB1 and ZEB2) [75, 77], snail family transcriptional repressor 1 and 2 (SNAI1 and SNAI2) [78], and TWIST1 [79].

ZEB2 is normally expressed in melanocytes, but its expression gradually decreases during the transition to melanoma, while ZEB1 expression increases [79]. In human melanoma, the upregulation of ZEB2 promotes cell growth and proliferation by activating the expression of MITF, and reduces invasion ability [80]. However, recent analysis using single-cell RNA sequencing has provided evidence that ZEB1 negatively regulates melanocyte proliferation programs and upregulates invasive/stem-like programs [79]. Early studies have demonstrated that the loss of Snai1 has a profound effect on the formation of the mesoderm during embryonic development [81, 82]. Besides, it has been observed that the expression of membrane transport proteins can be induced by members of the SNAI family, contributing to the process of EMT [83]. Studies in non-small cell lung cancer have demonstrated that excessive expression of TWIST1 mediates resistance to MET-dependent TKIs by suppressing p27 expression [84]. The acetylation status of TWIST1 plays a regulatory role in EMT, where diacetylated TWIST1-acK73/76 interacts with BRD8 (a component of the TIP60-Com complex) and histone H4-acK5/8, recruiting the TIP60-Com complex to activate mesenchymal target genes and MYC, thereby promoting target gene expression and cancer metastasis [78].

4.2 Epigenetic regulation in melanoma through interaction with EMT-TFs

Although transcription factors can directly regulate target genes, their binding to genes is ultimately temporary. In this context, epigenetic regulation, mainly involving three levels: DNA methylation, histone modification, and RNA interference, stabilizes cellular plasticity [85,86,87]. EMT-TFs interact synergistically with epigenetic enzymes to regulate downstream effectors of EMT.

DNA methylation is catalyzed by a group of proteins called DNA methyltransferases (DNMTs), and high methylation in gene promoter regions can lead to transcriptional repression [87]. DNA hypermethylation has been associated with the invasive phenotype of malignant melanoma [88]. DNMT interacts with EMT-TFs such as SNAI and ZEBs [89], recruited to the promoter region of the E-cadherin gene, leading to downregulation of E-cadherin expression [90], which may induce melanoma cells to undergo EMT. In other tumours, the expression of EMT-TFs has been shown to be regulated by DNA methylation [76, 91].

Histone acetylation modifications also play a role in melanoma plasticity. Numerous histone-modifying enzymes form complexes with EMT-TFs, affecting chromatin condensation and site accessibility during gene regulation. Classical complexes include the NuRD Complex, PRC1 Complex, PRC2 Complex, and SWI/SNF Chromatin-Remodeling Complex [92]. During the process of EMT triggered by the transforming growth factor-beta (TGF-β) signaling, Histone Deacetylase 1 (HDAC1) mediates global histone deacetylation and the acquisition of specific histone H3 lysine 27 acetylation (H3K27ac) marks on enhancers, thereby promoting the occurrence of this process [89].

Long non-coding RNAs (lncRNAs) have emerged as potential regulators of melanoma lineage plasticity. Additionally, the lncRNA SAMMSON acts as a target of the lineage-specific transcription factor SOX10, and targeting SAMMSON has shown promise in enhancing the sensitivity to MAPK-targeted therapy [92].

4.3 EET and EMT contribute to MAPKi resistance in melanoma therapy

During treatment, melanoma cells undergo phenotypic transitions that contribute to the development of drug resistance [1, 93]. Zhou et al. observed EMT-like phenotypic changes in melanoma cells during MAPKi treatment [74]. Geoffrey Richard et al. confirmed a correlation between high levels of ZEB1 expression and intrinsic resistance to MAPKi in BRAFV600 mutant cell lines and tumours. ZEB1 overexpression is sufficient to drive resistance to MAPKi by promoting a reversible transition to a MITFlow/p75high stem cell-like and tumorigenic phenotype [94]. Therefore, a better understanding of the role of EMT-TFs in melanoma cell plasticity will aid in designing novel combination therapies targeting specific EMT-TFs and the MAPK pathway.

Meanwhile, during tumour growth, epithelial–endothelial transition (EET) occurs (Fig. 2C), where stem cell-like cancer cells differentiate into new blood vessels to supply blood to tumour cells [8]. This proces, also known as angiogenic mimicry (VM), refers to a functional microcirculation pattern associated with endothelium-dependent and mosaic vessels in most malignant tumours. The use of inhibitors of VEGF receptor tyrosine kinase activity may facilitate the occurrence of this process [95]. A recent study reveals that endothelial transformation of metastatic melanoma cells can survive as resting cells and may regain metastatic properties through endothelial interstitial transformation (EndMT) for potential spread [96]. However, the role of EMT-TFs in the process of melanoma cell dormancy and reactivation remains unclear, leaving significant research gaps.

4.4 Impacts of melanoma plasticity on anti-tumour immunotherapy

In addition to regulating transcription factors, EMT also promotes immune suppression in melanoma by upregulating PD-L1 expression, inducing immunosuppressive Treg cells, inhibiting CD8+ cytotoxic cells and NK cells, and promoting antigen presentation, thus affecting tumour immune microenvironment [97,98,99]. These alterations in the tumour immune microenvironment facilitate the development of an invasive phenotype in melanoma cells, increasing their propensity for metastasis. During treatment, melanoma cells undergo phenotypic transitions leading to the development of drug resistance [1, 93]. Zhou et al. observed EMT-like phenotypic changes in melanoma cells during MAPKi treatment [74] This phenotypic plasticity also affects immunotherapy, as evidenced by the innate anti-PD1 resistance signature (IPRES), which shares similarities to MAPKi resistance features, including the expression of AXL, ROR2, and WNT5A [93]. Additionally, highly invasive melanoma cells also exhibit higher intrinsic drug resistance. Therefore, targeting characteristic molecules during the EMT process is emerging as a promising therapeutic approach to address melanoma metastasis and chemotherapy resistance.

5 Clinical implications by targeting dysregulated transcription

Lineage plasticity and dysregulated transcriptional regulation play pivotal roles in melanoma progression, influencing treatment responses and disease outcomes. In this context, targeting specific molecular elements associated with dysregulated transcription holds significant clinical implications. The following outlines potential clinical applications.

5.1 Targeting dysregulated transcriptional factors in melanoma

Key TFs also play pivotal roles in transcriptional dysregulation, which makes the approach to target TFs in melanoma seem to be feasible. For instance, the small molecular inhibitor, ML329, was tested effective to inhibit cell survival in two MITF-dependent melanoma cell lines, SK-MEL-5 and MALME-3 M, rather than in the MITF-independent A375 cell line [100]. However, the translational application by targeting the MITF has not been well progressed, since the role of MITF in mediating melanoma biological properties and responses to targeted therapies have been ambiguous. Moreover, the paired-like homeodomain transcription factor 1 (PITX1) shows certain treatment potential by regulating SOX gene family mRNA transcription, resulting in melanoma suppression with impaired cell proliferation and increased apoptosis [101], although further investigations are still in need.

5.2 Targeting epigenetic regulation-related pathways

Addressing epigenetic regulation-related pathways is another crucial strategy. However, the epigenome-based insights into melanoma regulatory transcription, as well as the implications with targeted therapies are generally limited. Some promising examples with profound clinical benefit include inhibitors targeting EZH2 and DOT1L, LSD1, BET, which not only reverse abnormal differentiation states in melanoma, but also increase the sensitivity to treatment [102]. Besides, the expression of non-coding RNA, SAMMON, which possesses as an informative biomarker of melanoma biology and BRAF inhibitor resistance could be targeted by the above mentioned PITX1 mRNA in a SOX10-dependent manner [101]. Besides, cinobufagin that suppressed Wnt/β-catenin target genes such as Axin-2, cyclin D1, and c-Myc in melanoma via LEF1 inhibition also turns out to be a potential anti-melanoma drug [55].

5.3 Targeting metabolic reprogramming in transcriptional dysregulation

Metabolic reprogramming is a distinctive feature in melanoma, and targeting this pathway holds promise for intervening in melanoma metabolic reprogramming [103]. The adipocyte enhancer-binding protein 1 (AEBP1), which is a transcriptional repressor influencing genes relating to the metabolic state of melanoma cells, was proved as a potential biomarker for cancer prognosis and a therapeutic target [104, 105]. Another effective targets modulating melanoma redox is the nuclear factor erythroid 2-related factor 2 (NRF2) in response to the oxidative stress burden during melanoma development [106].

5.4 Targeting plasticity-related metastasis and drug resistance

Regulating metastasis and drug resistance involves targeting specific transcription factors. Transcription factors like RUNX2, SOX9, and TWIST2, associated with metastasis and phenotype switching, could be targeted to prevent melanoma metastasis [107]. For instance, RUNX2 mediates resistance to BRAF inhibitor Vemurafenib, suggesting the potential benefits of combination therapy to enhance effectiveness [108, 109]. Besides, the metabolic reprogramming-related AEBP1 gene greatly contributes to the metastatic abilities of malignant melanoma, which was reported to upregulate EMT-related genes [105].

Additionally, targeting c-JUN, HIF-1, STAT3, and NF-κB, which regulate PD-L1 expression, may reduce the risk of treatment resistance [110]. c-JUN, involved in inflammation-induced dedifferentiation and the acquisition of pro-inflammatory characteristics, stands out as a potential key regulator of treatment resistance [111, 112]. Besides, STAT1, a regulator of PD-L1 expression, emerges as a promising target for combination immunotherapy [113]. Therefore, a comprehensive understanding and targeted intervention in dysregulated transcriptional control elements present promising avenues for developing precise and effective therapeutic strategies, ultimately improving treatment responses in melanoma patients (Table 1).

Table 1 Targeting melanoma plasticity and its related malignant features
Full size table

6 Future challenges

Melanoma plasticity is a process wherein cells alter their phenotype, presenting different molecular and/or histological characteristics, playing a crucial role in cancer progression and treatment resistance [93]. While potential genetic changes within tumours can enhance lineage plasticity, it is primarily a dynamically controlled process by transcriptional and epigenetic dysregulation [1]. This review extensively explores the categories of genetic mutations, transcription, and epigenetic regulatory factors influencing lineage amplification and plasticity in melanoma. Additionally, it delves into their interactions with other malignant features such as energy metabolism, metastasis, and drug resistance. Strategies for detecting and treating highly plastic tumours are also discussed.

The future of this field holds significant promise through the advancement of high-throughput single-cell and spatial analysis techniques. These developments, coupled with tools for tracing individual cells, are anticipated to drive significant progress. Similarly, accurately recapitulating various manifestations of plasticity in tumour models, sampled at relevant time points, will be crucial for advancing our understanding. Effectively deciphering and utilizing the cell atlases generated by these comprehensive large-scale approaches will be a major challenge, and high-throughput genetic perturbation strategies are expected to assist in this respect. Overall, future research is poised to unveil the molecular mechanisms driving melanoma plasticity and leverage this knowledge to formulate more precise and effective therapeutic strategies, ultimately enhancing treatment responses and prognostic outcomes for melanoma patients.