Recent Advances in Genetically Engineered Mouse Models of Melanoma: Insights into Tumor Initiation and Oncogenic Signaling Pathways Authors: Hyeonji Choi, Jung Hyun Lee, Hye Ji Cha, Yoonho Shin, Sekyu Choi Published: 2026 ABSTRACT Genetically engineered mouse (GEM) models are essential for elucidating the cellular and molecular mechanisms underlying various human diseases. Because mouse skin closely resembles the structure and lineage organization of human skin, melanoma GEM models enable the investigation of melanoma initiation in melanocytes and melanocyte stem cells, along with its progression from the epidermis to the dermis. Although early melanoma development is mainly regulated by major oncogenic signaling pathways, including MAPK and PI3K, recent studies have uncovered additional regulatory pathways, including Cdk4-mediated cell-cycle control, Jak–Stat-mediated transcriptional regulation, Hippo–Yap/Taz signaling, the p62–Igf2bp1 axis, and epigenetic regulators such as Tet2, which contribute to melanoma heterogeneity and phenotypic plasticity. This review summarizes the recent advances in melanoma GEM models, highlights the oncogenic signaling pathways active in these systems, and discusses the molecular insights derived from these models regarding melanoma initiation, progression, and intratumoral diversity. These findings highlight the utility of GEM models for elucidating melanoma biology and identifying potential therapeutic targets. Introduction Cutaneous melanoma is one of the most aggressive skin cancers, and its incidence and mortality are increasing worldwide (1-3). Although other skin cancers primarily arise from epithelial cells, melanoma originates from melanocytes and melanocyte stem cells (MeSCs), which are highly migratory and multipotent neural crest–derived cells (4-7). Compared with skin epithelial cells such as keratinocytes, melanocytes readily reactivate their neural crest–like migratory and plasticity programs (6, 8). These intrinsic characteristics may contribute to their high metastatic potential upon malignant transformation, making melanoma the deadliest form of skin cancer (4). Several key aspects must be clarified to better understand the unique cellular properties of melanomas. First, it is essential to identify the intracellular signaling pathways that are altered during tumorigenesis, including the genetic mutations that enable melanocytes and MeSCs to progress into melanoma and the distinct features they acquire as a result (9). Notably, many patients with melanoma have mutations in the mitogen-activated protein kinase (MAPK) pathway; approximately 50% and 20% of melanoma cases carry BRAF and NRAS mutations, respectively (10). These mutations alter fundamental cellular processes, including those of proliferation, survival, and differentiation, as well as the cellular plasticity of melanoma by modulating transcription factors (TFs) (11). Second, elucidating the pathogenesis of melanoma is critical as it would provide insights into how these oncogenically mutated cells interact with environmental factors (12). Melanomas, driven by oncogenic mutations initially grow in the epidermis, gradually invade the dermis (7), and readily metastasize to other organs. Particularly, a few oncogenically mutated melanocytes and MeSCs start growing laterally in the epidermis, a process termed the “radial growth phase,” followed by vertical expansion. After the “vertical growth phase,” melanoma cells invade the dermis and systemically disseminate (13). Melanoma cells encounter diverse cell types during these stages, including immune cells, epithelial cells, and cancer-associated fibroblasts. These interactions collectively drive phenotypic switching and eventually contribute to the high intratumoral heterogeneity of melanoma (13, 14). To understand the complex propagation mechanisms of melanoma in terms of its intrinsic cellular properties and pathogenesis, various experimental models have been used, including human patient-derived cells, genetically engineered mouse (GEM) models, animal model–derived cells, xenograft models, and tumor spheroids (15). Although many of these systems can recapitulate the properties of melanoma cells, they do not fully capture the structural complexity of the skin or reproduce the interactions with surrounding cell types. In contrast, GEM models closely reproduce the structural organization of human skin, including that of the epidermis and dermis. Therefore, GEM models allow a more comprehensive understanding of human melanoma pathogenesis (16). Hence, in this review, we aimed to introduce recently developed GEM models and summarize the major signaling pathways associated with potential oncogenic events. Additionally, we discuss secondary signaling pathways that contribute to melanoma growth, invasion, and heterogeneity following the first oncogenic hit. Finally, we outline the emerging research directions and applications of GEM models. Melanocyte Stem Cell Dynamics in the Skin MeSCs are undifferentiated cells of neural crest origin located in the bulge and hair germ regions of the hair follicle (HF) (17-19). These compartments form distinct niches; the bulge primarily sustains cell quiescence, whereas the hair germ serves as a more dynamic compartment where cells are poised for activation (19). MeSC activity is regulated by the hair growth cycle (20): they are quiescent during the telogen phase and activate at the onset of the anagen phase, leading to the production of proliferating progenitor cells. These progenitors migrate to the hair bulb, where they develop into mature melanin-producing melanocytes that provide pigmentation to the newly growing hair shaft (21). In the catagen phase, mature melanocytes in the hair bulb undergo apoptosis, thereby terminating melanin production (22). In contrast, MeSCs survive catagen by homing back to the hair germ, where they dedifferentiate and re-enter a quiescent state (20). MeSCs can migrate upward to the interfollicular epidermis (IFE) in response to ultraviolet (UV) radiation, where they enhance epidermal pigmentation as a protective mechanism by increasing melanin deposition (23). This plasticity underscores the capacity of MeSCs to shift between the hair follicular and epidermal niches in response to physiological demands. Comparative Localization of Melanocytes and Melanocyte Stem Cells in Human and Mouse Skin MeSCs exhibit important positional differences between humans and mice. In human skin, the IFE contains both melanocyte progenitors and differentiated melanocytes (24, 25). Within the HF, MeSCs are located in the bulge region, whereas differentiated melanocytes are present in the bulb (20). Conversely, melanocyte-lineage cells in mouse dorsal skin are located predominantly within the HF and not in the IFE (20, 25). Hence, mouse tail skin has been used as an alternative model for studying interfollicular melanocyte-lineage cells that more closely resemble those in human skin (26, 27). This tissue exhibits a more clearly stratified epidermal architecture, including the basal layer, spinous layer, stratum granulosum, and stratum corneum and contains melanocytes as well as melanocyte progenitor–like cells within the IFE (26, 27). Melanocyte Stem Cells and Melanocytes as Cells of Origin in Melanoma The origin of melanoma has remained controversial. Although melanoma arises from the melanocyte lineage, it is still unclear which cellular state is most susceptible to malignant transformation. Early studies suggested that MeSCs, which possess high developmental plasticity, may serve as a melanoma-initiating population (28, 29). In mice, MeSCs are largely confined to the HF, and this anatomical feature has enabled lineage-tracing approaches to examine their contribution to melanoma initiation (20, 28). Quiescent MeSCs remain resistant to tumorigenesis even in the presence of oncogenic alterations such as the BrafV600E mutation and phosphatase and tensin homolog (Pten) loss (28). However, they acquire migratory potential upon activation by external stimuli such as UV exposure or depilation, relocate to the IFE, and subsequently invade the dermis during their hyperproliferative expansion, ultimately giving rise to melanoma (28). These findings suggest that the cellular state of MeSCs, in addition to oncogenic mutations, is a crucial determinant of melanoma initiation. However, interpretation of this study is largely restricted to follicular MeSCs. In contrast, studies using mouse tail skin, have provided evidence that interfollicular melanocytes may also be the origin of melanoma (27). The mouse tail epidermis contains mature melanocytes within the scales and amelanotic dormant melanocytes, which are stem-like cells, within the interscale regions (27). Among these populations, mature melanocytes appear more permissive to clonal expansion and tumor formation, whereas MeSC niches in the interscale epidermis appear to suppress tumorigenic proliferation despite the presence of oncogenic mutations (27). Thus, mature melanocytes may also function as melanoma-initiating cells, undergoing malignant transformation through transcriptional reprogramming and dedifferentiation (27). Collectively, MeSCs and differentiated melanocytes are plausible cellular origins of melanoma, and both may exist within the human epidermis (27, 28). Their tumorigenic potential is determined not only by intrinsic oncogenic events but also by external stimuli and the surrounding niche. Accordingly, mouse models that preserve skin architecture and recapitulate diverse niche and microenvironmental interactions are better suited for investigating the complex mechanisms of melanoma initiation. Experimental Models of Melanoma Various experimental models have been utilized in cancer research, each with distinct advantages. Studies using mouse or human cell lines provide a straightforward and powerful platform for precisely analyzing intracellular signaling pathways (30-32). In addition to established cell lines (33), primary cultures derived from patients with melanoma or mouse models can offer insights into cellular characteristics that more closely reflect the in vivo context (34). Beyond conventional 2D culture systems, recent advances in 3D culture techniques have led to the development of human-derived organoid models (35), which can better recapitulate tumor cell heterogeneity and help preserve tumor-specific characteristics under 3D culture conditions. In addition, the ex vivo tissue slice culture has been widely used as a valuable approach for maintaining tissue architecture outside the body while enabling the investigation of interactions among diverse cell types (36). Moreover, conventional in vivo melanoma models include transplantation models using human- or mouse-derived cells and GEM models, which are established through activation of oncogenes and inactivation of tumor suppressor genes (Table 1) (37-42). However, as the scope of each model differs, it should be selected according to the research question. Although in vitro models are highly useful for directly assessing cell-intrinsic changes, they do not fully recapitulate skin structure or interactions with surrounding cell types. Furthermore, given the marked heterogeneity of melanoma, in vitro systems often represent only a subset of tumor cell states and are not well suited to reflect tissue-level changes across diverse cellular populations. Even though primary cultures and mouse- and human-derived organoids preserve cellular heterogeneity better than conventional single cell lines, they do not fully recapitulate interactions with surrounding cell types and may therefore overlook phenotypic changes driven by such interactions (Table 1) (43). Ex vivo models are typically based on the culture of tissue specimens derived from human patients (36, 44). These models are useful for experimentally validating the relevance of findings in human tissues because they partially retain skin architecture (44). However, these models are difficult to maintain over long periods, and once tissues are removed from the body, the extracellular matrix (ECM) gradually changes and inflammatory or injury responses may be induced (44). As a result, the cultured tissue may progressively diverge from its original in vivo state (44). Because genetic and environmental manipulation are technically challenging in ex vivo systems, these models are not ideally suited for establishing causal relationships (Table 1) (45). Traditional in vivo models include allograft and xenograft systems (43). In both approaches, cells are inoculated into mice after ex vivo manipulation, which allows genetic modification and partially reproduces the in vivo environment. However, because tumors are typically induced through subcutaneous injections, they usually develop within the dermis. Thus, although they are useful for studying interactions with immune cells, the surrounding ECM, and dermal cells, they are less suitable for understanding native skin structure and the pathogenesis of melanoma within its physiological context (Table 1). In this regard, GEM models provide unique advantages, as they allow genetic manipulation while enabling the investigation of melanoma development in a more physiologically relevant setting, including the progression from the epidermis to the dermis. Accordingly, GEM models offer a powerful platform to define the cellular origin of melanoma, elucidate how transformed cells interact with neighboring cell types during tumorigenesis, and determine how these interactions contribute to cancer initiation and progression (Table 1). Melanoma Genetically Engineered Mouse Model Numerous studies have used GEM models to investigate melanoma. Mouse skin is structurally similar to human skin, comprising both epidermis and dermis, as well as melanocytes and stem cells residing within HFs (18, 46, 47). Therefore, melanoma GEM models provide an effective system for studying the initiation of melanoma from melanocyte lineage cells and tracking the spatial migration of melanoma cells, which structurally progress from the epidermis into the dermis. Moreover, because the entire process can be observed in vivo, it is possible to monitor changes in melanoma cell characteristics through their dynamic interactions with other cells including immune cells, stromal components, and fibroblasts. Melanoma GEMs are typically generated by introducing patient-derived oncogenic mutations, either individually or in combination, into melanocyte-lineage cells (Table 2) (37, 48). Technological advancements in mouse genetics have enabled the precise expression of oncogenic drivers, specifically in melanocyte-lineage cells (38). Recently, melanoma GEM models have been reported to use not only major patient-derived signaling pathways, but also additional regulatory genes involved in various pathways (Table 2). It has been revealed that these additional genes are important for phenotype switching toward more malignant states. In addition to genetic alterations, these models are often combined with carcinogenic stress to better mimic melanoma initiation and progression in vivo. For example, UVB irradiation or chemical carcinogen treatment such as 7,12-dimethylbenzanthracene can be applied to GEM to induce additional DNA damage and tumor-promoting stress (Table 2) (49-51). These approaches enable researchers to examine how oncogenic mutations cooperate with environmental or chemical drivers to accelerate melanoma development, alter tumor latency, and influence malignant progression. Signaling Pathways Involved in Melanoma Development Primary driver mutations that are important for melanoma initiation predominantly arise within major signaling pathways in human patients, such as the MAPK and phosphoinositide 3-kinase (PI3K) pathways (Fig. 1). However, in mouse models, a single genetic alteration, such as Braf activation, is often insufficient to induce aggressive melanomas (52). Therefore, GEM models typically incorporate combinations of mutations to better recapitulate human melanoma development and progression (Table 2) (48). In addition to alterations in major signaling pathways, secondary or cooperative mutations play crucial roles in pathogenesis (Fig. 2, 3). Importantly, recent studies have identified various signaling pathways that contribute to the heterogeneous characteristics of melanoma, such as resistance to immunotherapy, invasiveness, migratory capacity, and proliferative status (Fig. 2, 3). In the following section, we describe the signaling pathways that have been investigated using these melanoma GEM models and discuss how these models have advanced our understanding of melanoma. Major Signaling Pathways MAPK pathway The MAPK pathway, which in melanoma includes NRAS, BRAF, MAP2K1/MAP2K2, and KIT mutations, contains a cascade of RAS, RAF, MEK, and ERK. In melanocyte-lineage cells, Nras and Braf activate downstream Erk (53-55). Activated Erk promotes the cell cycle by increasing cyclin D1 and inhibiting p27Kip1 expression (56). However, it can suppress apoptosis by inhibiting the expression of pro-apoptotic factors or inducing their degradation (57). Moreover, Erk modulates microphthalmia- associated transcription factor (Mitf) expression, altering the differentiation status and overall transcriptional program of melanocytes (58, 59). Furthermore, the MAPK pathway is involved in phenotypic switching in melanoma, contributing to cellular plasticity and altered invasion (Fig. 1) (37, 38, 58, 60-62). PI3K pathway PIP2 is phosphorylated in the plasma membrane by activated PI3K to form PIP3 (63, 64). PIP3 subsequently promotes activation of downstream effectors, including Akt and mechanistic target of rapamycin (mTOR) (65). The activated components transmit various signals and regulate cell growth, survival, and metabolic processes (66). Loss-of-function mutations in PTEN, which are frequently observed in patients with melanoma, result in overactivation of the PI3K pathway (67). PTEN inhibits PIP3 synthesis, and its loss leads to increased PIP3 levels, thereby stimulating downstream signaling. In addition to PTEN alterations, various point mutations in the catalytic subunit of PI3K and loss-of-function mutations in PIK3R1, which encodes a regulatory subunit of PI3K, have also been reported in melanoma cells (Fig. 1) (68). Akt-mTOR pathway The Akt–mTOR axis is a representative downstream signaling of the PI3K pathway (69). Activating mutations or overexpression in Akt are rarely detected in melanoma cells (70). The mutations inhibit apoptosis-inducing proteins such as BAD or FOXO1/3 (71, 72), and WNT signaling is activated through inhibition of GSK3β (72). These mutations also promote cell growth and survival (71). Furthermore, Akt activates the mTOR signaling pathway. Mechanistic target of rapamycin complex 1 (mTORC1) enhances cell growth through regulation of protein synthesis (73) and induces angiogenesis via Hif1a (74, 75). Mechanistic target of rapamycin complex 2 (mTORC2)/rapamycin-insensitive companion of mTOR (RICTOR) signaling regulates cytoskeletal organization, migration, and invasion in melanoma cells, and increased mTORC2 activity is generally associated with metastatic potential (76). However, in some therapeutic contexts, reduced mTORC2/RICTOR activity has also been linked to drug resistance (77). p16Ink4a and p19Arf Both p16Ink4a and p19Arf are encoded by the Cdkn2a locus and are crucial in the cell cycle. Nuclear p16Ink4a inhibits the formation of the cyclin D–Cdk4 complex, thereby suppressing retinoblastoma protein (Rb) phosphorylation and blocking G1–S phase progression (78-80). p19Arf inhibits Mdm2, thereby stabilizing p53 and inducing cell-cycle arrest and apoptosis (81). The expression of these proteins is regulated by various TFs. p16Ink4a expression is regulated by Ets1/2 (82, 83) and Ap1 (82) activated by Erk in the MAPK pathway. p19Arf expression can be altered by E2F1 and E2F3 (84). These genes are important for cell death and proliferation as they directly regulate the cell cycle (78, 85). CDKN2A is one of the most frequently mutated tumor suppressor loci in patients with melanoma, and approximately 40% of familial melanoma cases harbor germline CDKN2A mutations (Fig. 1) (86). Additional Signaling Pathways Cdk4 Mutations at codon 24 of CDK4, predominantly CDK4R24C and CDK4R24H, are frequently observed in patients with melanoma (87). As one of the most frequently observed mutations in melanoma GEMs, the R24C/H mutation blocks the binding of Cdk4 to p16Ink4a, leading to sustained activation of Cdk4 (85). Activated Cdk4 binds to cyclin D, promoting Rb phosphorylation (78, 85). Ultimately, the cells continuously enter the S phase (Fig. 2) (78). Cdk5 Recently, Cdk5 has shown a crucial role in the metastatic spread of melanoma (88). The role of Cdk5 in melanoma progression was investigated by introducing a Cdk5fl/fl allele into the TyrCreER; BrafV600E; Ptenfl/fl (TBP) mouse model (89). In this mouse model, the loss of Cdk5 was found to significantly inhibit metastasis (89). Mechanistically, Cdk5 is involved in vimentin phosphorylation that interferes assembly of vimentin filaments (89). This ultimately affects cellular migration activity, and the process favors metastatic characteristics, enabling tumor cells to spread to distinct organs (Fig. 2) (89). Jak-Stat The Jak-Stat pathway regulates the expression of various genes crucial for cell survival, cell cycle, and metabolic control during homeostasis. Particularly, it has been revealed in melanoma that Jak1 regulates Stat3 activity, thereby controlling the activity of enhancers such as CEBPA/B/D and ultimately altering Mitf expression levels (90). This regulation of Mitf expression determines whether melanoma cells exhibit increased proliferation or acquire invasive and metastatic capacities (90). Beyond studies using in vivo models, Jak1/2 mutations have been shown to block the efficacy of anti-PD1 therapy via impaired interferon-gamma signaling, indicating that mutations in the Jak-Stat pathway can confer resistance to melanoma immunotherapy (Fig. 2) (91, 92). p62/Sqstm1 p62/SQSTM1 is highly expressed in metastatic melanoma lesions of patients harboring NRAS mutations (93, 94). Indeed, in mice, knocking out Sqstm1 in a Braf- and Pten-mutant background significantly reduced melanoma induction (94). The detailed mechanism involves p62 enhancing the mRNA stability of pro-metastatic factors (94). Particularly, p62 interacts with the RNA-binding Igf2bp1 to increase the stability of Fermt2, which is crucial for cell adhesion and migration (94). This also explains the process through which NRAS-mutated melanoma cells acquire migratory activity (Fig. 2) (94). Epigenetic modulation Reduced levels of 5-hydroxymethylcytosine (5hmC) are epigenetic hallmarks of melanoma (95, 96). Typically, 5hmC is generated from 5mC through TET family enzymes including TET2 (97). During this process, α-ketoglutarate is converted into succinate (98). Recently, it has been revealed that melanocyte-specific conditional knockout of Tet2 promotes melanoma incidence and progression, ultimately reducing survival rates (99). Tet2 knockout reduces hydroxymethylome plasticity in melanoma (99). This results in continued cell division and increased expression of phosphorylated Akt, Pcna, and cyclin A, all of which are associated with cell proliferation (99). Furthermore, the Tet2 knockout melanoma mouse model exhibited lower 5hmC levels. This was accompanied by increased expression of proteins that activate the cell cycle and decreased expression of senescence-related proteins such as p15Ink4b, p16Ink4a, and p19Arf (99). This imbalance may exacerbate cancer progression (Fig. 3) (99). Lkb1 pathway Analysis of The Cancer Genome Atlas–based human melanoma tumors demonstrated that approximately 50% of melanoma cases exhibited significantly reduced expression of serine/threonine protein kinase STK11 (LKB1) (100). Under normal conditions, Lkb1 responds to DNA damage by activating AMP-activated protein kinase, which restores cellular energy homeostasis and metabolic adaptation (101, 102). When Lkb1 was depleted in BrafV600E-mutated melanocytes and DNA damage was induced by UVB irradiation, melanoma progression accelerated, inducing multiplicity and dedifferentiation. This is presumed to result from activation of the mTORC1/2 signaling pathway (Fig. 3) (100). Hippo pathway In animal models, a single oncogenic Braf mutation is more likely to lead to oncogene-induced senescence (OIS) instead of direct progression to melanoma (48). However, recent reports indicate that OIS-independent activation of the Hippo pathway due to cellular overgrowth can prevent progression to melanoma (103). Upon activation of the Hippo signaling cascade, Lats1/2 phosphorylates the transcriptional coactivators Yap and Taz, a modification that promotes their cytoplasmic retention and subsequent proteolytic turnover (104-107). Senescent nevi and BrafV600E-expressing melanocytes show increased levels of phosphorylated Yap/Taz (103). Therefore, when Lats1/2 is conditionally knocked out in the TyrCreER; BrafV600E model, nevi develop into melanomas (103). This demonstrated that Yap activation is crucial in melanoma initiation (Fig. 3). Chemokine signaling C-X-C motif chemokine receptor 2 (Cxcr2) ligands recruit Cxcr2-expressing myeloid-derived suppressor cells (MDSCs) or tumor-associated neutrophils to form a pre-metastatic niche during cancer development. They can also influence tumor growth through direct autocrine signaling (108). In the context of melanoma, higher expression of ligands such as CXCL1, CXCL2, CXCL3, CXCL5, and CXCL8 was confirmed in melanoma lesions compared with those in nevi (109). Furthermore, high expression of the receptor CXCR2 is associated with poor prognosis. The conditional deletion of Cxcr2 in two melanoma mouse models, TyrCreER; BrafCA/+; Ptenlox4-5 and TyrCreER; NrasQ61R; Ink4a−/−, suppresses tumor progression (109). Moreover, the loss of Cxcr2 altered the tumor immune microenvironment by increasing the number of CD8+ T cells and decreasing the number of macrophages and MDSC-like cells (Fig. 3) (109). Discussion GEM models have been instrumental in elucidating the mechanisms underlying melanoma initiation and progression. Compared with cell line–based systems, GEM models better preserve tissue architecture, cellular heterogeneity, and microenvironmental context, thereby providing critical insights into how signaling networks are dynamically reorganized during melanoma development. These models have particularly enabled the investigation of how distinct melanocyte-lineage populations respond to oncogenic events and external stimuli in vivo. A defining strength of GEM models is their versatility in modeling melanoma across diverse anatomical contexts. Studies using murine dorsal, tail, ear, or acral skin have provided critical insights into region-specific melanocyte biology and have demonstrated that melanoma can arise from distinct cellular origins within different tissue niches. This capacity is of particular relevance to human melanoma biology, as human skin comprises various melanocyte-lineage populations, including differentiated melanocytes, bulge MeSCs, and epidermal melanocyte-lineage cells, each representing a potential cellular source of melanoma. As multiple melanocyte-lineage populations are present within the same tissue context, an unresolved question is whether these populations independently drive tumor initiation and progression or engage in competitive or cooperative interactions that shape disease trajectory. Such cellular diversity complicates a comprehensive understanding of melanoma, as tumors may arise from distinct melanocyte-lineage compartments and follow different trajectories depending on their cellular origin and surrounding microenvironment. GEM models, through lineage-specific genetic manipulation and in vivo fate-tracing, are uniquely positioned to dissect how each population contributes to melanoma initiation and progression within its native tissue context. Building on the strengths of GEM-based approaches, the integration of organoid technology represents a promising direction for further advancing melanoma research (110-112). Organoids generated under defined genetic conditions offer an opportunity to examine how specific mutations interact with tissue architecture and the local microenvironment (113). Current HF-bearing skin organoids contain pigment-producing melanocytes, and recent human pluripotent stem cell-derived appendage-bearing skin organoids have shown melanocytes not only in follicle but also in epidermal compartments, suggesting improved recapitulation of human skin organization (114-116). Such organoid platforms, when combined with in vivo findings from GEM studies, will enable more detailed investigation of the distinct pathological stages of melanoma. Nevertheless, several important challenges remain. Whether these organoid systems can stably maintain long-term melanocyte-lineage populations, particularly cells with MeSC-like properties, remains unclear. If future models succeed in preserving such populations, GEM-integrated organoid or CRISPR–editable human-derived organoid platforms could serve as powerful next-generation systems for investigating melanoma initiation from diverse cellular origins while simultaneously addressing genetic and microenvironmental determinants of disease. Importantly, GEM models will continue to serve as the foundational model upon which these emerging technologies are built, ensuring that the mechanistic insights gained from in vivo studies remain central to our understanding of melanoma biology.