The role of non-coding RNAs in the regulation of cell death pathways in melanoma

Abstract

In recent years, due to increased exposure to ultraviolet rays, the incidence of melanoma has increased significantly. Despite the progress in early diagnosis and treatment methods, managing metastatic melanoma remains challenging. Resistance to programmed cell death (PCD) is a hallmark of various cancers, including melanoma. This resistance can affect cell growth, survival, metastasis, and resistance to therapies. In recent years, non-coding RNAs (ncRNAs) have been acknowledged as master regulators of various biological processes, including PCD. This regulatory network acts as a double-edged sword in melanoma progression, either promoting tumor survival by preventing PCD or acting as an anti-tumor agent by inducing it. Such diverse functions make ncRNAs invaluable as diagnostic and prognostic biomarkers and potential therapeutic targets. This review explores the multifaceted roles of ncRNAs in regulating different forms of PCD, comprising apoptosis, autophagy, necroptosis, pyroptosis, ferroptosis, cuproptosis, and anoikis, highlighting their potential applications in melanoma research and treatment.

Introduction

Melanoma, arising from melanocytes, is the third most common type of skin cancer after basal and squamous cell carcinoma. Its incidence is annually growing so that in 2022, 331,722 new cases and 58,667 deaths were reported for melanoma, representing 1.66% and 0.6% of all diagnosed cancer cases and cancer-related mortality, respectively [1]. Even though the incidence of melanoma is lower than other types of skin cancers, it is responsible for approximately half of skin cancer mortality (according to the Global Cancer Observatory [GLOBOCAN] 2018, 2020, and 2022 data) (Fig. 1) [1–3]. While a 5-year survival rate of 94% has been reported for all stages of cutaneous melanoma, this rate for metastatic melanoma is significantly lower [4]. In the United States, the 5-year survival rate for patients diagnosed with stage IV cutaneous melanoma in 2018 was estimated by the SEER database to be 29.8%. This is while the prognosis of non-cutaneous forms of melanoma is even worse [5].

Although thanks to the recent advances in diagnostic and treatment methods, the mortality of melanoma has decreased in recent years, the treatment of late-diagnosed melanoma still remains challenging [6]. Primary melanoma can be effectively managed by excision of the local tumor with a safety margin, while metastatic melanoma requires systemic therapies such as immunotherapy, targeted therapy, and chemotherapy in addition to local tumor resection [7].

As the dysregulation of PCD is one of the key characteristics of aggressive and metastatic cancers, induction of cell death in cancer cells is one of the primary objectives of anti-tumor therapy [8]. Before the 1970s, when apoptosis was identified, cell death was believed to occur passively and without regulation [9]. Today, it has been revealed that it can be classified into regulated and non-regulated forms based on the signal transduction dependency. While intracellular signaling pathways precisely regulate PCD, accidental events like cell injury account for non-programmed types of cell death. PCD can be subclassified into apoptotic and non-apoptotic cell death (autophagy, pyroptosis, ferroptosis, apoptosis, necroptosis, anoikis, and cuproptosis), which are different in morphological properties and molecular pathways [10].

Recent studies acknowledged non-coding RNAs (ncRNAs) as the master regulators of various biological processes, including PCD. It has been demonstrated that the Tumor microenvironment (TME) significantly influences the expression pattern of ncRNAs, affecting programmed cell death, tumor progression, immune evasion, and therapeutic resistance [11–14]. In melanoma, the cellular components of TME, including Immune cells (tumor-associated macrophages, regulatory T-cells, and myeloid-derived suppressor cells), cancer-associated fibroblasts (CAFs), endothelial cells, and non-cellular components, including hypoxia, extracellular matrix (ECM), cytokines (TGF-β, IL-6, TNF-α) can alter the expression pattern of various ncRNAs and interfere with PCD pathways including apoptosis, autophagy, pyroptosis, ferroptosis, necroptosis, anoikis, and cuproptosis [15–18]. Changes in the ncRNA expression pattern can affect the sensitivity of malignant cells to PCD. Thus, ncRNAs can have a considerable diagnostic/prognostic value in various tumors, including melanoma [19, 20]. Furthermore, by targeting ncRNAs associated with regulated cell death, novel therapeutic strategies can be developed to induce cell death in melanocytes, offering promising avenues for more effective treatments [20, 21].

This review critically examines the role of ncRNAs in modulating various forms of PCD in melanoma, highlighting their potential as key regulators in tumor progression and therapeutic responses. Understanding the underlying mechanisms of ncRNA-mediated regulation of PCD in melanoma can help to develop novel diagnostic and prognostic biomarkers and valuable therapeutic targets.

Fig. 1.

Comparison of the incidence and mortality rate of melanoma, non-melanoma skin cancers, and all types of cancer according to the GLOBOCAN 2018, 2020, and 2022 data [1–3]

Biological properties of regulatory NcRNAs

While more than 80% of the human genome undergoes transcription into RNA, only 2% of it encodes protein. These RNA molecules, which lack the ability to code for proteins, are known as ncRNAs. ncRNAs can originate from protein-coding sequences, enhancer regions, or even from transposons [22]. With the progress of RNA sequencing technologies and bioinformatic tools, more aspects of ncRNA biogenesis and function have been clarified in recent years. NcRNAs can be divided into housekeeping and regulatory ncRNAs based on their role in biological processes. Housekeeping ncRNAs, the most abundant and constitutively expressed type, are involved in mRNA translation. In contrast, regulatory ncRNAs play a role in gene expression regulation at the transcriptional, post-transcriptional, and epigenetic levels. Based on their length, regulatory ncRNAs are subdivided into long (lncRNAs) and small (sncRNAs), both of which have different subtypes that differ in structure, function, or location (Table 1). The biological properties of various regulatory ncRNAs will be discussed in the following sections.

Table 1.

Different types of ncRNAs and their properties and biological roles

Housekeeping RNAs	Type	Length (nucleotides)	Biological role
	rRNA	> 1500	Key component of ribosomes, crucial for mRNA translation and protein synthesis.
	tRNA	76–90	Facilitating mRNA translation by transporting amino acids to the ribosome.
	snRNA	100–300	Involved in pre-mRNA splicing
	snoRNA	60–200	Chemical modification of other RNAs (2′-O-methylation and pseudouridylation)
	Telomerase RNA	451	Template for the synthesis of telomeric sequences, maintaining telomere length
	Ribonuclease RNA	Varies	RNA processing, including the maturation and degradation
lncRNAs	lincRNA	1 kb	Regulation of gene expression level via affecting chromatin modifications and epigenetic changes
	NAT	> 200 (some of them are shorter than 200 neucleotide)	Inhibiting the function of their sense RNAs by forming double-stranded RNA structures.
	circRNA	100- >200	transcription regulation, cell cycle control, miRNA sponging, and RNA stability enhancement
	eRNA	50-2000	Regulation of gene expression level via enhancer activation and chromatin remodeling
sncRNAs	miRNA		Post-transcriptional regulation of gene expression by binding to complementary sequences in target mRNAs
	siRNA		Mediates RNA interference by guiding the degradation of complementary mRNA sequences
	piRNA		Protects the genome from transposon activity and regulates gene expression in germline cells.
	scaRNA		RNA splicing and modification od snRNAs

LncRNAs

Circular RNAs (circRNAs)

The discovery of circRNAs dates back to 1976 when a 3′-5′ covalently closed form of RNA was isolated from murine respirovirus (Sendai virus) and viroids separately by Kulakowski [23] and Sanger et al. [24]. Three years later, in 1979, the existence of circular RNAs in eukaryotic cells (HeLa cells) was first demonstrated by Hsu MT and Coca-Prados using electron microscopy [25]. In the later years, researchers discovered the first instances of circRNAs in humans, which were transcribed from the gene known as DCC (deleted in colon cancer) [26]. From the 1990s through the early 2000s, the existence of several other circRNAs in humans and other eukaryotes was documented [27], however, their role in biological processes wasn’t fully grasped, and they were considered byproducts of mis-splicing [28].

Thanks to the recent advancements in RNA sequencing techniques and bioinformatic tools, the biological importance of circRNAs has become more realized, and they become a spotlight in genomic research. In contrast to mRNAs and other lncRNA molecules produced through canonical pre-mRNA splicing, circRNAs are formed via a non-canonical splicing process known as “back splicing” (Fig. 2). In this process, a 5′ splice site acts as a donor site and is covalently joined to an upstream 3′ splice site (acceptor site) with a phosphodiester bond, which results in a circular form of RNA without 5′ caps and 3′ polyadenylated tail [29]. Owing to the relatively lower efficiency of back splicing, circRNAs are produced at a lower quantity than mRNAs; nonetheless, their high stability can render them more abundant than unstable mRNAs in specific cells [29]. Although most of the circRNAs are more than 200 nucleotides in length, some of the exon-derived circRNAs (ecircRNAs) and intronic circRNAs (ciRNAs) molecules can be found as small as 100–200 nucleotides [30]. Nonetheless, circRNAs are generally classified as lncRNAs.

Fig. 2.

CircRNA biogenesis and functions. The top panel shows the formation of circRNAs through back-splicing of a typical mRNA. The lower panel illustrates the different functions of circRNAs.RBP RNA-binding protein; eIF4G2 eukaryotic translation initiation factor 4 gamma 2, METTL3 methyltransferase like 3, Pol II Polymerase II

Exonic circRNAs (EcircRNAs), exon-intron circRNAs (EIciRNA), and circular intronic RNAs (ciRNAs) are the three main classes of circRNAs which are produced through exon skipping (lariat-driven circularization), intron pairing-driven circularization, and RNA binding protein (RBP)-driven circularization mechanisms, respectively [31]. Some alternative back-splicing mechanisms account for generating other less common types of circRNAs, such as intergenic circRNAs, read-through circRNAs (rt-circRNA), and fusion-circRNAs (f-circRNAs), commonly linked to pathological conditions [29]. Additionally, circRNAs display tissue-specific expression patterns, which can be affected by various conditions such as stress and different developmental stages [32].

Compared to their linear counterparts, which have a half-life of less than 10 h, the closed-loop configuration of circRNAs confers them resistance to exonucleases and increases their half-life to more than two days [33]. CircRNAs can be transcribed from protein-coding genes or non-coding loci. Moreover, a gene can produce more than one type of circRNA by different splicing patterns [34]. The biological role of circRNAs has not yet been fully deciphered; however, it has been revealed that they can participate in transcription regulation, cell cycle regulation, miRNA inhibition (miRNA sponging), and enhancing the stability of other RNA molecules (Fig. 2) [29, 32]. They can also interact with RNA-binding proteins (RBPs) and affect their localization and function [29]. Although circRNAs are non-coding, it has been shown that by adding internal ribosome entry sites (IRES) and open reading frame (ORF) to their structure, they can be translated both in vivo and in vitro. These synthetic circular RNAs can be used in gene therapy as an alternative to conventional mRNAs and provide a less expensive approach than mRNA-based approaches [35–37]. Bioinformatic analysis suggests that a limited number of endogenous circRNAs are protein-coding and can be translated [38].

Various diseases, including cancers, give rise to the generation of specific circRNAs (intergenic circRNAs, rt-circRNA, and f-circRNAs), which play an essential role in disease development. For example, in various types of cancer cells, circRNAs can induce drug resistance, resistance to PCD, and tumor proliferation and progression [39]. In recent years, circRNAs have gained attention as potential diagnostic and prognostic biomarkers in cancer. Furthermore, the modulation of circRNA expression—either through upregulation or downregulation—has gained attention as a potential therapeutic solution, offering potential avenues for targeted treatment strategies [32].

LincRNA

Long intergenic noncoding RNAs (lincRNAs) are autonomously transcribed RNA molecules constituting more than 50% of lncRNAs [40]. The human genome has been revealed to encode more than 15,000 lincRNAs [41]. These molecules are called “intergenic” because they are transcribed from the non-coding region of the genome and do not physically overlap with protein-coding genes [42]. The initial evidence for lincRNAs came from tiling array investigations conducted across genomic sequences, which showed widespread transcription from areas without recognized coding genes [43–47]. Early evidence for the existence of active transcription units at the putative loci of these transcripts was provided by analyzing the fingerprints of chromatin states in several types of murine cells [48].

Like mRNA transcripts, lincRNAs are transcribed by RNA polymerase II; however, their transcription is approximately one-tenth that of mRNAs [49]. Compared to mRNAs, which have an average length of 2.9 kb and contain an average of 10.9 exons, lincRNAs are relatively shorter (1 kb) and contain fewer exons (2.9 exons on average) [49, 50]. It has been revealed that their expression pattern is particular for cells, tissues, developmental stages, and pathological conditions [40].

In recent years, with growing the knowledge about lincRNAs, their role in homeostasis and biological processes has been more clarified; however, fewer than 1% of lincRNAs have been identified in terms of their function [40]. LincRNAs can interact with chromatin and affect the expression of their own genes or other genes by epigenetic regulation and imposing either stable or repressive chromatin states [40]. They can alter the transcription level through the recruitment of chromatin remodeling complexes and the formation of R-loop at the promoter of target genes [42]. They can act as a decoy to sequester mRNAs, miRNAs, and proteins and suppress their function. These molecules also serve as scaffolds, facilitating the assembly of multiprotein complexes that are crucial for various cellular processes [51, 52]. In the cytoplasm, they can affect the stability, translation, and degradation of target mRNAs, adding another layer of control over gene expression [53]. Moreover, by functioning as miRNA sponges, lncRNAs can counteract the repressive effects of miRNAs on their target mRNAs, providing a mechanism for fine-tuning gene regulation [54].

Dysregulated expression of lincRNAs is the hallmark of various types of inherited or acquired diseases such as cancers. According to their disease-specific or disease stage-dependent expression pattern, they have the potential to be utilized as diagnostic or prognostic biomarkers as well as promising therapeutic targets [42].

Natural antisense transcripts (NATs)

NATs were discovered for the first time in bacteria in 1981 as molecules that are involved in the control of plasmid number [55]. During 1986–2002, the presence of NATs was also described in eukaryotes, including humans [56]. While historically NATs were classified as lncRNAs, today it has been revealed that NATs can also be shorter than 200 kb, and even some of them can be translated [57]. NATs are RNA transcripts synthesized from opposite strands of either coding or non-coding genes. They are complementary to sense RNAs and can interact with them wholly or partially and inhibit their function [58]. Like mRNAs, NATs undergo capping, poly-adenylation, and splicing [59]. NATs can act in cis (interact with RNAs from the same genomic loci) or trans (interact with RNAs from different genomic loci) [60]. Depending on their orientation towards the sense RNA, NATS are subclassified into three subsets comprising head-to-head (5′-regions overlap), tail-to-tail (3′-regions overlap), and embedded (One transcript is fully incorporated into the other) [57]. Through stimulating or inhibiting the translation of their corresponding sense RNAs, NATs play a critical role in gene expression regulation [61]. Dysregulation of NATs is one of the hallmarks of various types of cancers, making them a potential candidate for therapeutic purposes [57].

Enhancer RNAs (eRNAs)

eRNAs are RNA molecules with 50-2000 nucleotides in length that originate from the enhancer regions. In 2010, two groups of researchers separately identified that enhancers can be transcribed into RNA molecules called eRNAs [62, 63]. While the production of eRNAs is the hallmark of active enhancers, their role in enhancer activity has not yet been understood and remains controversial. While some researchers state that they are noise products and lack any significant biological role, others believe they are involved in enhancer activation [64]. Other suggestive roles of eRNAs include regulating gene expression levels by orchestrating chromatin remodeling and interaction with transcriptional regulators. The expression of eRNAs is particular to cell types and tissues. Dysregulated eRNA expression has been implicated in various cancer and plays a crucial role in activating tumor-promoting genes and contributing to the genomic instability observed in malignant cells [65].

MicroRNAs

Small non-coding RNAs (sncRNAs) are a subgroup of regulatory ncRNAs with a length of < 300 nt [66], which can modify gene expression by different mechanisms, including RNA interference, RNA modification, or spliceosomal involvement [67]. sncRNAs are regulators of gene expression and cancer signaling pathways; therefore, they are associated with different malignancies [68, 69]. The following section reviews the biological properties of the microRNAs (miRNAs), as the central subgroup of regulatory sncRNAs,. Due to the lack of studies about the role of other subgroups of sncRNAs in melanoma, this study has neglected to review them.

miRNAs are the prominent family of sncRNAs with a length of 19–25 nt, which are highly conserved molecules and exist in all eukaryotic cells [66]. Lee et al. reported the first miRNA, lin-4, in Caenorhabditis elegans (C. elegans), and this event marked a turning point in the history of molecular biology. The miRNA lin-4 was identified as a regulator of development events in C. elegans by decreasing the amount of the LIN-14 protein [70]. Seven years later, another miRNA, let-7, was identified in C. elegans as a regulator of the heterochronic gene lin-41 [71]. Currently, based on the latest release (v22) of miRBase (https://www.mirbase.org), the number of annotated hairpin precursors and mature sequences in humans has reached 1917 and 2654, respectively [72]. Discovering new miRNAs is continuing, and recognizing their role in biological processes is a hot topic in biological research.

The location of miRNAs varies in the genome; about half are intragenic, and the other half are intergenic. The miRNAs with intergenic origin have their promoter, and their regulation is independent of a host gene [73, 74]. Most miRNAs are processed by the canonical miRNA biogenesis pathway, which employs a microprocessor complex that exports 5/RanGTP complex and Dicer. miRNA biogenesis begins with transcribing miRNA coding transcripts into pri-miRNA by RNA polymerase II [75]. Then, the microprocessor complex, which comprises DGCR8 and Drosha (a ribonuclease III enzyme), processes pri-miRNA into pre-miRNA [76]. The processing of some types of pre-miRNAs, such as Mitrons (pre-miRNA-like hairpins) and 7-methylguanine capped (m⁷G)-pre-miRNA, in the nucleus, is independent of microprocessor complex function [77]. Next, pre-miRNA is exported into the cytoplasm by exportin 5/RanGTP complex. An RNase III endonuclease called Dicer cuts the terminal loop of pre-miRNA in the cytoplasm to create a mature miRNA duplex that is about 22 nt long [76, 78, 79]. Depending on factors such as thermodynamic asymmetry of the duplex and stability of base pairing at the 5’ end, two strands of mature miRNA duplex are separated [80]. Finally, either the 5p or 3p strands of the mature miRNA are loaded into the Argonaute (AGO) family of proteins (Ago2 in humans) and form a miRNA-induced silencing complex (miRISC) [81]. The miRNA strand of miRISC binds to the 3’ terminal untranslated region (UTR) of the target mRNA and, depending on the level of base complementarity, either breaks down the target (perfect match) or blocks its translation (partial match) [82].

At the post-transcriptional level, miRNAs inhibit gene expression by complementing with the target mRNA’s 3’ or 5’ UTR. The regulatory function of miRNAs is not limited to a specific gene and one single miRNA can interact with several genes. miRNAs can be considered one of the main regulators of gene expression in humans because about one-third of gene expression in the human genome may be controlled by miRNAs [83, 84]. According to recent studies, some miRNAs, called up-miRs (such as miR-221-5p and miR-212-5p) can upregulate the expression of the encoded protein of target mRNA [85]. Due to tissue and cell-specific expression of most miRNAs, they have the potential to be therapeutic, prognostic, and diagnostic biomarkers in different diseases such as cancer [86–88].

Numerous clinical trials are underway to assess miRNAs as cancer biomarkers for prognostication, diagnosis, and therapy response prediction purposes. The level of diagnostic and prognostic miRNA biomarkers has been tested in biofluids or tissue samples of cancer patients. For example, in a clinical trial (NCT03591367) miR-155 was used for diagnosis of non-muscle-invasive bladder cancer [89]. In another clinical trial (NCT05477667), let-7a and miR-124 were tested to diagnose non-Hodgkin’s and acute lymphoma. Moreover, some clinical trials have assessed miRNAs as predictors of cancer therapy efficacy. For instance, in a clinical trial (NCT04662996) researchers examined the potential of miRNAs to serve as indicators of chemotherapy effectiveness in patients with metastatic castration-resistant prostate cancer. Also, miRNAs have been used as therapeutics for cancer [89]. The phase 1 clinical trial using MRX34, a liposome-based miR-34 mimic, was the first miRNA treatment clinical trial conducted on humans [90].

The involvement of NcRNAs in distinct types of PCD within melanoma cells

PCD encompasses various forms of cell death mechanisms that differ in molecular regulation, morphology, cellular consequences, and biological outcomes and serve distinct biological purposes (Table 2) [91]. The following sections discuss these various forms of PCD in melanoma and the regulatory function of ncRNAs in these processes.

Table 2.

Key features of different types of programmed cell death

	Key molecules	Caspase dependency	Morphology	Cell membrane	DNA fragmentation	Inflammation
Apoptosis	Death receptors and their ligands, Bax, Bak, apoptosis inducing factors, caspases-8, -3, -9	Caspase dependent	Cell shrinkage; membrane blebbing	Unchanged	DNA degraded into 180–200 bp fragments	No
Autophagy	ULK1, PI3KIII, Autophagy-related genes, LC3	Caspase independent	Formation of intracellular vacuoles, Enlargement of organelles, membrane blebbing	Unchanged	No	No
Necroptosis	Death receptors, Toll-like receptors, T-cell receptors, RIPKs, MLKL	Caspase independent	Cell swelling and deformation	Membrane rupture	Random degradation of DNA	Yes
Ferroptosis	System Xc−, Glutathione peroxidase 4 (GPX4), Lipid reactive oxygen species (ROS)	Caspase independent	Mitochondria shrinkage, decreased cristae	Membrane rupture	No	Yes (in some cases)
Pyroptosis	NOD-like receptors (NLRs), AIM2-like receptors (ALRs), caspase-1, casp	Caspase dependent	Cell swelling and deformation	Membrane rupture	Random degradation od DNA	Yes
Cuproptosis	Copper (Cu)-induced stress, Ferredoxin 1 (FDX1), Glutaminase 2 (GLS2), Cyclin-dependent kinase inhibitor 2 A (CDKN2A)	Caspase-independent	Mitochondrial swelling, organelle disruption	Membrane rupture	No	Yes
Anoikis	Death receptors and their ligands, Bax, Bak, AIF, caspase-8, caspase-3, caspase-9	Caspase dependent	Cell shrinkage; membrane blebbing,	unchanged	DNA degraded into 180–200 bp fragments	No

Autophagy

Autophagy acts as a recycling system of the cells, in which cargos, such as damaged organelles and protein aggregates, degrade by a companion of lysosomes [92]. Therefore, autophagy can serve as a cellular homeostatic process, preventing healthy cells from transforming into malignant cells and cancer progression [93]. According to this, it is reasonable that tumor-inducing proteins inhibit autophagy, and tumor-suppressive proteins promote autophagy responses [94]. Surprisingly, there is evidence that autophagy may enhance tumor progression, metastasis, and therapy resistance. The tumorigenesis function of autophagy is carried out by maintaining the survival of cancerous cells under metabolic stress [95]. Our limited knowledge of the molecular regulation and application of autophagy in various contexts is reflected in the contradictory findings about the paradoxical role of autophagy in tumor biology [94].

There is an intricate cross-talk between TME and autophagy; various immune mediators in TME regulate autophagy, and meanwhile, autophagy regulates immune signaling pathways in TME. Numerous factors in TME can trigger autophagy. Hypoxia in TME is one of the stimulators of autophagy. In glioblastoma TME, hypoxia-induced IL-6 acts as an initiator of autophagy via p-STAT3-MIR155-3p-CREBRF-CREB3-ATG5 pathway [96]. Other inflammation mediators can trigger autophagy in TME, for example, IFN-γ-IDO1-kynureine pathways induce autophagy and enhance macrophage phagocytosis, which decreases cervical cancer progression [97]. It has been shown that inflammation-induced autophagy can limit the inflammation in TME by negative feedback [98]. The glucose level in TME is another factor that affects autophagy in tumor cells. The high glucose level in TME triggers autophagy via hyperglycemia-induced sterol regulatory element-binding protein 1 (SREBP1) and increases tumor growth [99]. The interaction between autophagy and TME has made it possible to reprogram TME by altering autophagy. For example, inhibition of autophagy by Chloroquine can shift macrophages towards the M1 phenotype, which has higher phagocytosis potency and eliminates tumor cells more efficiently compared with the M2 phenotype [100]. In hypoxic TME, secretion of some exosomes from tumor cells increases IL-6 and miR-155-5p in TME and induces M2 macrophage polarization via enhancing autophagy [101]. All this evidence suggests that various inflammatory molecules and cells in TME can influence autophagy.

Studies consistently reveal that autophagy is vital for melanoma development [102, 103]. During melanomagenesis, the amount of autophagy-associated genes and proteins is changed. These changes can be utilized as predictors of patients’ diagnosis, prognosis, survival, and drug susceptibility. Studies have shown that the expression of pro-autophagic proteins ATG5, BECN1, LC3A, and LC3B in early phases of melanoma growth is less than in the benign nevi phase [104]. ATG5 is a crucial factor for the expansion of autophagosome membrane. Its reduction has been shown to be associated with decreased autophagic activity and lower progression-free survival (PFS) rate in melanoma patients. Furthermore, it has been shown that in an in vitro model of melanomagenesis, reduction in ATG5 can promote tumor cell proliferation by inhibiting the oncogene-induced senescence mechanism [105]. As further evidence, in a murine model with an allele of the BRAF^V600E mutation, deletion of Atg7 significantly accelerated the onset of melanoma. Accelerated development of melanoma in the absence of Atg7 was dependent on the presence of PTEN [106]. In contradiction with this finding, Xie et al. revealed that autophagy is essential for melanomagenesis and deficiency of Atg7 prevented melanoma development and prolonged survival in a mouse model with BRAF^V600E mutation and allelic Pten loss [107]. The exact mechanism by which Atg deficiency causes increased melanomagenesis has not yet been precisely determined.

The following sections discuss the function of different ncRNAs in regulating autophagy in melanoma. Figure 3 shows the interaction of different ncRNAs with autophagy mediators in melanoma.

Nuclear enriched abundant transcript 1 (NEAT1) is a lncRNA that is over-expressed in melanoma tissues compared with paracancerous tissues. Targeting of NEAT1 has been proposed as a potential therapy for melanoma [108]. A study showed that the inhibition of NEAT1 by gambogenic acid triggered autophagy and ferroptosis in melanoma cells through AMPK/mTOR signal axis induction [109]. Also, it has been revealed that NEAT1 boosts the response of cancerous cells to chemotherapy by regulating autophagy [110]. Conversely, it has been shown that NEAT1 knock-down inhibits authophaghy and increase the sensitivity of colorectal cancer cells to 5-FU [111]. In non-small cell lung cancer (NSCLC) NEAT1 triggers authophagy through miR-128-3p axis [112]. Therefore, NEAT1’s regulation of authophagy exhibits complex, context-dependent effects across different cancer types, impacting therapeutic responses.

Fig. 3.

Regulation of autophagy by ncRNAs in melanoma. This figure shows the role of different ncRNAs in controlling autophagy, which is a crucial process in melanoma cells to evade the immune system or cause resistance to cancer therapeutics. Identifying these ncRNAs and their interaction with autophagy machinery may enhance the sensitivity of melanoma cells to anti-cancer treatments and consequently improve the clinical outcome of melanoma patients

It has been revealed that the amount of miR-18a-5p is elevated in clinical melanoma tissues and cell lines [113]. EPHA7 has been introduced as a target gene for miR-18a-5p in melanomagenesis. Changes in the expression of the EPH7A gene is also associated with other cancers such as gastric, prostate, and colorectal cancers [114–116]. In melanoma cells, the miR-18a-5p/EPHA7 axis modifies the expression of LC3-I/II and p62 which are two autophagy-related genes [113]. LC3II is an autophagosome-specific marker in mammals that is induced by autophagy-associated genes such as ATG3 and ATG7 [117]. On the other hand, p62 is a degradation factor for autophagy, and it has been shown that autophagy suppresses tumorigenesis through selective elimination of p62 [118, 119]. Guo et al. revealed that miR-18a-5p knockdown induces activation of LC3II, while it suppresses expression of p62 through overexpression of EPHA7 [113]. Liang et al. reported that miR-18a-5p acts as an oncogene by enhancing autophagosome formation and autophagy flux [120]. More research is required to fully understand the pathogenic role of autophagy change by miR-18a-5p in melanoma development.

miR-27a is another miRNA whose overexpression in melanoma tissues and cell lines has been reported [121, 122]. Also, in other types of cancers, miR-27a has been proven as an oncogene [123, 124]. In breast cancer cells, miR-27a reduces chemoresistance by impairing autophagy and degrading reactive oxygen species [125]. According to investigations of Tang et al., suppressing miR-27a triggers autophagy and apoptosis in melanoma cells by up-regulating SYK expression and activating the mTOR signaling pathway. As a prognostic factor, the expression of miR-27a has a positive correlation with tumor stage and lymph node metastasis [121].

Dysregulation of the miRNA profile can impact tumor invasion and metastasis [126]. Multiple miRNAs, including miR-9, miR-145, and miR-182, have been confirmed to regulate the invasion and metastasis of tumor cells in melanoma [127–129]. Therefore, some miRNAs show promise as biomarkers in the diagnosis and prognosis of melanoma. Guo et al. reported that serum and tumor tissue levels of miR-23a were down-regulated in melanoma patients, and this down-regulation was associated with poor clinical outcomes. They found out that miR-23a regulates the invasion and metastasis of melanoma cells through modifications in the autophagy mechanism. miR-23a targets ATG12 and reduces the invasion of melanoma cells through the AMPK-RhoA pathway. In breast cancer cells, it has been shown that miR-23a inhibits autophagy by targeting ATG3 [130]. Also, ectopic expression of miR-23a can inhibit melanoma metastasis in vivo [131]. Recent studies have revealed that highly metastatic melanoma cells can transfer their metastasis and invasion potency to low metastatic melanoma cells through the delivery of exosomal miRNAs, such as miR-1268a and miR4535, which are inhibitors of the autophagy pathway [132, 133].

Another important aspect of the implication of autophagy on melanoma is its role in resistance to treatments. In biopsies of melanoma patients with BRAF^V600E mutation who received BRAF inhibition (BRAFi) or a combination of BRAF and MEK inhibition, the level of autophagy increased compared to the baseline. This increase in autophagy was associated with a shorter duration of PFS. Some studies showed that the increase of autophagy upon treatment with BRAFi occurs through triggering ER stress [134]. However, other researchers mentioned that activation of TFEB is the reason for the autophagy increase in BRAFi-treated melanoma [135]. According to these new insights, targeted therapy for melanoma in combination with autophagy inhibitors may be beneficial for melanoma patients.

Drug-induced autophagy is one way that tumor cells employ to generate drug resistance and increase tumor cell survival [136, 137]. Treatment of BRAF^V600E mutated melanoma cell lines with BRAFi, induces cytoprotective autophagy, and inhibition of autophagy augments the efficiency of BRAFi in triggering cell death [134]. Many studies have revealed the role of miR216b in drug resistance in many cancers, including melanoma [138–140]. Vemurafenib, a BRAF inhibitor, suppresses miR-216b expression in melanoma. miR216b simultaneously suppresses the expression of multiple autophagy-related genes, namely Beclin-1, UVRAG, and ATG5. Therefore, this micro-RNA can disrupt vesicle nucleation and elongation during autophagy [140]. Luo et al. identified that ectopic expression of miR-216b can reduce Vemurafenib-induced autophagy. In conclusion, therapeutic strategies that use Vemurafenib in combination with increased expression of miR216b can suppress melanoma cell growth more effectively through inhibition of drug-induced autophagy [140]. Another study by Kim et al. identified that the expression of miR-1246 was higher in BRAF inhibitor-resistant cells than in BRAF inhibitor-sensitive cells. Also, miR-1246 can induce resistance to the BRAF inhibitor (PLX4720) in A375P/Mdr cells. PLX4720 causes a significant decrease in autophagy in miR-1246 transfected cells. Inhibition of autophagy by miR-1246 may play a role in resistance to BRAF inhibitors. Through inhibition of autophagy, miR-1246 increases the population of cells in the G2/M phase and facilitates their escape from cell death through mechanisms such as necrosis and apoptosis in response to PLX4720 [141]. Studies have demonstrated that miR-153-3p inhibits snail family transcriptional repressor-1 (SNAI1) in melanoma, thereby suppressing cell invasion and proliferation [142]. Also, miR-153-3p inhibits autophagy by suppressing ATG5 expression. A long intergenic non-coding RNA, LINC00641, triggers autophagy by regulating miR-153-3p. Down-regulation of miR-153-3p has been shown in melanoma tissues and cells [142].

Pyroptosis

Pyroptosis is an inflammatory form of necrosis initiated by microbial infections and some endogenous factors. Pyroptosis was initially discovered in 1986 when it was observed that anthrax toxin triggered rapid and widespread cell death in mouse peritoneal macrophages [143]. Although similar results were seen in follow-up studies, this form of cell death was initially misclassified as apoptosis [144, 145]. In 2000, cell death triggered by Salmonella was classified as a caspase 1-dependent necrosis, distinguishing it from apoptosis [146]. By 2001, this pro-inflammatory f orm of cell death was officially termed “pyroptosis,” derived from the Greek words “pyro,” meaning “fever,” and “ptosis,” meaning “falling.” [147]. While pyroptosis was initially recognized in macrophages that are triggered by infectious pathogens, later research revealed that it can occur in various other cell types as well [148]. Emerging evidence indicates that pyroptosis can also be induced in cancer cells as a response to chemotherapy [149]. It has been found that in addition to caspase 1, other proteins, including caspases 3, 7, 8, and gasdermin proteins (GSDM) are also involved in pyroptosis [150, 151]. In 2018, the “Nomenclature Committee on Cell Death” characterized pyroptosis as a form of PCD driven by the activation of GSDM family proteins, which create pores in the plasma membrane, typically as a consequence of inflammatory caspase activation [152].

Gasdermin D (GSDMD) and E (GSDME) proteins are two primary proteins responsible for executing pyroptosis, which is encoded by the GSDMD DFNA5 genes, respectively. GSDMA, GSDMB, GSDMC, and DFNB59 are other proteins that belong to the gasdermin family, and all account for membrane perforation during apoptosis [153, 154]. Pyroptosis can be triggered via two primary pathways: (i) The classical pathway involves the activation of the inflammasome, which subsequently activates caspase-1. Caspase-1 then cleaves and activates GSDMD. The N-terminal fragment of the cleaved GSDMD binds to the plasma membrane and mediates the formation of large, ring-shaped pores on it, leading to the cell rupture. Activation of the classical pathway also promotes the secretion of pro-inflammatory cytokines such as IL-1β and IL-18 [155]. (ii) In non-classical pathway, cleavage of GSDMD is mediated by activated caspase-4, -5, -11 [151]. Recent discoveries highlight that GSDME can shift apoptosis to pyroptosis in cancers during chemotherapy. Caspase-3 cleaves GSDME, releasing its N-terminal fragment, which forms pores in the plasma membrane, similar to GSDMD. These pores trigger inflammatory cell death, enhancing immune recognition of cancer cells and transforming the tumor microenvironment, making GSDME a key player in cancer therapy [156].

In contrast to most types of cancers, melanoma cells express GSDME at a higher level, which creates an opportunity to treat melanoma by inducing pyroptosis [157]. Nonetheless, the results of various studies indicate the dual role of pyroptosis in melanoma. For example, it has been revealed that pyroptosis can suppress melanoma growth, and the expression level of pyroptosis-related genes can be used as a prognostic marker for melanoma [158, 159]. Nonetheless, NLRP1, a pyroptosis-inducing protein, has been shown to promote tumor progression [160]. This is attributed to inflammatory factors that are released during pyroptosis and can promote tumor growth by creating an inflammatory microenvironment [161]. Atherosclerosis, neuroinflammation, and autoimmune diseases are other possible consequences of pyroptosis [159]. Melanoma poses several immunogenic mutations that result in the accumulation of immune cells in the tumor microenvironment. It has been shown that pyroptosis can lead to the remodeling of the melanoma microenvironment and the initiation of anti-tumor responses in the tumor microenvironment. For example, pyroptosis can lead to tumor infiltration of immune cells, activation of T and NK cells in the tumor microenvironment, M1 polarization of tumor-associated macrophages, and increased sensitivity to chemotherapy [159].

In 2019, Sun.L et al. conducted a bioinformatic study and identified 246 differentially expressed lncRNAs in patients with primary and metastatic melanoma, among which 184 lncRNAs were upregulated and others were downregulated. Notably, their findings linked the majority of these lncRNAs to the overall survival of melanoma patients, highlighting their potential as prognostic biomarkers [162]. Building on this foundation, In 202, Wu L et al., conducted a bioinformatics analysis to determine pyroptosis-related lncRNAs that their expression correlates with the overall survival of patients with skin cutaneous melanoma. Through rigorous bioinformatics analysis, they identified 22 hub pyroptosis-related lncRNAs with strong predictive potential, outperforming conventional risk signatures. These molecules comprise AC004847.1, USP30-AS1, AC082651.3, AL033384.1, AC138207.5, AC245041.1, U62317.1, AL512274.1, AC018755.4, MIR200CHG, LINC02362, LINC00861, AL683807.1, AC010503.4, AL512363.1, LINC02437, LINC01527, AL049555.1, AC245041.2, AL365361.1, AC015819.1, and MIR205HG. Among these, the expression levels of AC004847.1, AC015819.1, AC018755.4, AC138207.5, AL365361.1, AL683807.1, LINC00861, LINC02362, U62317.1, and USP30-AS1 were higher in melanoma samples than higher samples, while others showed the reverse pattern. Importantly, their lncRNA signature demonstrated robust predictive value for parameters critical to melanoma prognosis, including immune status, tumor infiltration of immune cells, tumor stemness, expression of m6A-related (ZC3H13, YTHDF1, FTO, YTHDC2, and WTAP) genes, and sensitivity to chemotherapy. However, the retrospective nature of this study and the lack of experimental validation limits its immediate translational impact [163]. Another bioinformatic study was conducted in 2021 by Xie J et al. in which 9 pyroptosis-related lncRNAs with prognostic value through correlations with m6A-methylation-related genes were identified. This lncRNA signature was consisted of AL121603.2, AC107464.2, AC245128.3, AC092171.5, AC242842.1, IRF2-DT, HLA-DQB1-AS1, AC004585.1, LINC00582. Nonetheless, in vitro or in vivo experimental validation was lacking in this study, which leaves a critical gap in functional understanding [164].

In 2023 Zeng B et al., applied a “dry-lab discovery/wet-lab validation” fusion approach to determine the possible miRNAs that have an important role in melanoma. First, by conducting a bioinformatic analysis they determined 71 upregulated exosomal-miRNAs in polymetastatic cell line (POL) and 35,603 target genes for these miRNAs. Their study revealed that four of the 71 upregulated miRNAs (miR-211-5p, miR-3184-3p, miR-383-5p, and miR-342-5p) are key regulators of prognosis, with miR-211-5p and miR-3184-3p strongly associated with poorer outcomes. Subsequent experiments revealed that overexpression of these miRNAs significantly enhances melanoma cell invasion and migration. Mechanistically, miR-211-5p targets GNA15, forming a regulatory axis that suppresses anti-tumor immunity by promoting glycolysis and inhibiting pyroptosis in the metastatic melanoma cells. Their Wet lab analysis verified that overexpression of miR-211-5p can inhibit both classical and non-classical pyroptosis pathways by downregulating caspase-1 and caspase-4, respectively [165]. These findings provide a deeper insight into the mechanism of modulating immune evasion and metastasis of melanoma cells by miRNAs, which can help in designing innovative therapeutic strategies.

The above-mentioned items underscore the critical role of ncRNAs in regulating pyroptosis and shaping the tumor-immune axis of melanoma. Although experimental evidence regarding the ncRNA-mediated regulation of pyroptosis in melanoma is currently lacking, there are various studies demonstrating the involvement of ncRNAs in pyroptosis of other types of cancer, comprehensively reviewed by Wang et al. [166]. These findings suggest potential similar mechanisms in melanoma. Nonetheless, it should be noted that bioinformatics-driven discoveries have laid a strong foundation, the translation of these insights into clinical applications requires robust experimental validation and mechanistic studies. Future research should focus on integrating these ncRNA signatures with advanced therapeutic strategies, such as targeted delivery systems or immune checkpoint inhibitors, to harness their full potential in combating melanoma. By bridging the gap between computational predictions and biological validation, we can pave the way for novel, personalized treatment approaches that enhance patient survival and improve outcomes.

Ferroptosis

Ferroptosis is an iron-dependent form of PCD that leads to lipid peroxidation of lipid bailer and cell rupture. The word “ferroptosis” was first used in 2012 [167]; however, cell death resembling ferroptosis has been identified much earlier. For the first time, Eagle et al., in the 1950s revealed that deprivation of the amino acid cyst(e)ine can induce cell death which can be prevented by transsulfuration, the endogenous pathway for the cysteine synthesis [168, 169]. Additionally, in 2001, a ferroptosis-like cell death induced by oxidative stress was discovered, called oxytosis [170]. In 2003, it was evidenced that a novel drug called erastin can induce a distinct form of cell death in RAS-expressing malignant cells that is different from apoptosis and necrosis [171]. Subsequent studies showed similar results and have also identified that this type of cell death can be inhibited using iron chelators [172, 173].

In recent years, the mechanistic knowledge of ferroptosis has progressed rapidly and it has been revealed that activation of ferroptosis can overcome the resistance of various cancers to chemotherapy [174]. Ferroptosis morphologically and biochemically differs from apoptosis and necrosis. During ferroptosis, mitochondria undergo atrophy, cristae are decreased, and plasma membrane density is increased, while despite the oxidative damage in the DNA, the nucleus size and chromatin condensation remain relatively unchanged [10]. The iron accumulation and lipid peroxidation of polyunsaturated fatty acid (PUFA)-containing phospholipids (PUFA-PLs) are the biochemical hallmarks of ferroptosis. In normal cells, Glutathione peroxidase 4 (GPX4) reduces lipid peroxides by consuming Glutathione (GSH). Thus, a decline in GPX4 activity and intracellular GSH depletion are the primary causes of ferroptosis initiation. When the conditions required to trigger ferroptosis surpass the buffering capacity of cellular antioxidant systems, the generated lipid peroxides are converted to harmful free radicals by Redox-active irons through the Fenton reaction [175]. Two distinct mechanisms account for the reduced activity of GPX4 or reduced level of cellular GSH: (i) direct binding of ferroptosis-inducing agents to GPX4 and prevent its function or promotes its degradation, (ii) inhibition of system Xc-, a cystine/glutamate antiporter critical for cellular cystine uptake and subsequent GSH synthesis. The latter underscores the importance of cystine in maintaining cellular antioxidant defenses [174].

While at first, it was believed that ferroptosis exclusively occurs in cancer cells with RAS mutations, subsequent findings determined that the occurrence of ferroptosis is independent of RAS mutation. Many cancers exhibit altered thiol metabolism, resulting in increased iron accumulation and heightened ferroptosis susceptibility. For instance, in melanoma, the differentiation status of cancer cells inversely correlates with their vulnerability to ferroptosis. During tumor progression, less differentiated melanomas demonstrate greater sensitivity to ferroptosis [176]. Ferroptosis can influence the immune landscape of the TME by shaping the composition and function of immune cells within the TME. Oxidized lipid mediators and damage-associated molecular patterns (DAMPs) that are released from the ferroptotic melanoma cells serve as “find me” signals for immune cells such as macrophages, leading to the attraction of phagocytes to the tumor site [177]. Ferroptosis is also involved in the response to T-cell-activating immunotherapy approaches in patients with melanoma. In this regard, activated T cell-derived interferon-gamma (IFNγ) suppresses the expression of SLC3A2 and SLC7A11, subunits of system Xc-, which leads to the occurrence of ferroptosis. It has been shown that downregulation of SLC3A2 is a good prognosis for the treatment of melanoma with nivolumab (an anti-PD1 monoclonal antibody) [178, 179]. The proto-oncogene BRAF, frequently mutated in melanoma, decreases reliance on oxidative phosphorylation, preventing ROS formation and ferroptosis. However, BRAF inhibitors can reverse this effect and sensitize melanoma cells to ferroptosis inducers, offering therapeutic potential [180]. More differentiated melanoma cells tend to be less sensitive to ferroptosis [180]. This highlights the need to investigate and characterize the processes involved in regulating ferroptosis in melanoma, especially in its more aggressive forms, to uncover new ways to predict how melanoma will behave. Moreover, this can help to develop more effective treatments that target these processes and overcome resistance in difficult-to-treat differentiated melanoma cells.

Emerging research highlights the involvement of ncRNAs in ferroptosis regulation. For example, Wu et al. have revealed that AGAP2 Antisense RNA 1 (AGAP2-AS1) is upregulated in cutaneous melanoma and correlates with poor prognosis [181]. It has also been revealed that in addition to melanoma, AGAP2-AS1 is involved in tumorigenesis, tumor progression, invasion of various non-melanoma malignancies [182]. AGAP2-AS1 participates in tumor progression by induction of tumor resistance in melanocytes; thus, downregulation of AGAP2-AS1 is a potential therapeutic strategy to enhance ferroptosis in melanocytes and inhibit tumor progression [183]. Similarly, NEAT1 lncRNA inhibits ferroptosis and autophagy in melanoma, promoting tumor growth and reducing treatment sensitivity. NEAT1 overexpression suppresses SLC7A11 expression, decreasing cystine uptake and GPX4 activity [109]. The regulatory role of NEAT1 in ferroptosis and autophagy has also been demonstrated in non-melanoma cancer [110, 111, 184]. Thus, targeting NEAT1 may have therapeutic benefits for melanoma patients. Hanniford et al. identified a circRNA, cerebellar degeneration-related 1 antisense (CDR1as), as a marker of melanoma differentiation status and response to treatment. High CDR1as expression sensitizes melanoma cells to GPX4 inhibitors-mediated ferroptosis, while its silencing increases melanoma invasion and metastasis [185]. In a recently published study, Zhang et al. identified a circRNA known as circPIAS1 that hinders immune checkpoint blockade therapy-induced ferroptosis in melanoma, inducing treatment resistance and cancer progression. They revealed that circPIAS1 encodes an oncogenic variant of PIAS1 protein that is distinct from canonical PIAS1 protein. This protein reduces the phosphorylation of STAT1 while increasing its SUMOylation, leading to the activation of SLC7A11/GPX4 signaling pathway and restriction of immunotherapy-induced ferroptosis. Targeting circPIAS1 or its regulatory pathway could be an efficient approach to enhance the efficacy of immune checkpoint blockade therapy via restoration of treatment-induced ferroptosis [186].

Bioinformatic studies have further expanded our understanding of ferroptosis-related lncRNA signatures in melanoma prognosis. For example, Sun et al. identified a prognostic model based on 18 ferroptosis-related lncRNA signatures in skin cutaneous melanoma. First, they identified the top three mutated genes involved in ferroptosis—TP53, ACSL5, and TF—and showed that their higher expression correlates with increased immune cell infiltration into the tumor. After that, the authors constructed a model incorporating 18 lncRNAs, which includes USP30-AS1, LINC01871, AC026369.3, AL606807.1, AC021078.1, AC093297.2, AC004865.2, AC010245.2, AC018645.3, AC011511.5, AL021368.2, AC024909.1, KANSL1L-AS1, PPP1R26-AS1, AC100778.3, AC069222.1, AL592211.1, and MALINC1. This model holds promise for improving prognosis prediction and guiding immunotherapy strategies for patients with skin cutaneous melanoma [187]. Similarly, Ma et al. proposed a ferroptosis-related lncRNA signature comprising five molecules for uveal melanoma. Experimental analysis confirmed the role of three out of these lncRNAs (LINC00963, PPP1R14B.AS1, and ZNF667.AS1) in promoting the invasion of C918 cells. Furthermore, the study demonstrated that downregulating these lncRNAs effectively inhibits cellular invasion [188]. Other studies have also identified additional lncRNA-based models, including those by Xu et al. [189], Guo et al. [190], Rao et al. [191], and Ping et al. [192], which incorporate various ferroptosis-related lncRNAs to predict the prognosis of melanoma patients and their response to immunotherapy. Figure 4 shows some of these ferroptosis-related lncRNAs, identified in these studies as influencing melanoma prognosis.

Fig. 4.

cRNAs act as a double-edged sword in melanoma. Some ncRNAs act as oncogene or drug-resistance inducers, while others act as tumor suppressors. This figure shows the discussed ncRNAs in this review that have either oncogenic or tumor suppressor roles in melanoma

MiRNAs also play pivotal roles in ferroptosis regulation. For example, Luo et al. have revealed that miR-137 acts as a negative regulator of ferroptosis in melanocytes by inhibiting glutaminolysis via targeting SLC1A5, a glutamine transporter. They showed that the knockdown of miR-137 increases the erastin-induced ferroptosis in vitro and in vivo, highlighting the potential of miR-137 as a therapeutic target in melanoma [193]. MiR-9 is another negative regulator of ferroptosis in melanocytes. It directly engages with the 3’-UTR of glutamic-oxaloacetic transaminase (GOT1) mRNA and prevents its translation, inhibiting erastin- and RSL3-mediated ferroptosis. Thus, its inhibition can be a new avenue for therapeutic intervention [194]. The inhibitory effect of miR-9 on pyroptosis has also been observed in other diseases. For example, Jeyabal et al. have shown that miR-9 inhibits hyperglycemia-induced pyroptosis in human ventricular cardiomyocytes by directly targeting and downregulating ELAVL1 [195].

In conclusion, the items mentioned above underscore the crucial role of ferroptosis in the biology and treatment of melanoma. Further investigation of the underlying mechanisms of ferroptosis and its molecular regulatory network, including the role of lncRNAs and miRNAs, can lead to the identification of novel prognostic and predictive biomarkers as well as therapeutic targets.

Apoptosis

Apoptosis is a natural mechanism for eliminating senescent, damaged, mutant, and infected cells to preserve tissue homeostasis. Two main pathways can induce apoptosis: the mitochondrial and receptor-mediated pathways. Intracellular triggers (e.g. stress, DNA damage) and extracellular signals, such as FasL and TRAIL secreted by other cells can initiate the receptor-mediated pathways [196, 197]. Apoptosis plays a protective role against tumor formation and any malfunction in the regulating systems of apoptosis or deficiency of apoptosis mediators may lead to carcinogenesis. Although apoptosis is a non-immunogenic PCD, cellular and molecular changes during apoptosis can affect components of TME. For instance, the release of purine nucleotides from apoptotic cells can attract macrophages and dendritic cells (DCs) to TME [198]. Moreover, the release of some molecules such as sphingosine-1-phosphate (S1P) from apoptotic cells can remodel the stroma component [199]. Contrariwise, the TME can regulate apoptosis in tumor cells. For example, the presence of cytokines such as IFNs, CD137, and IL-24 in TME can enhance apoptosis in tumor cells. Also, the expression of TRAIL by neutrophils, monocytes, and macrophages in TME can trigger TRAIL-mediated apoptosis in cancer cells [200–202].

Tumor cells evade apoptosis by several strategies. Understanding these strategies is beneficial for developing therapeutics based on triggering apoptosis in tumor cells [196, 203]. In the following parts, the function of various ncRNAs in the regulation of apoptosis is discussed, and potential biomarkers for treatment and prognosis of melanoma are suggested.

miR-150 has been found as a tumor inducer in a variety of tumors [204, 205]. In malignant melanoma, miR-150 is up-regulated and is associated with lower long-term survival in metastatic melanoma [206, 207]. Interestingly, higher serum levels of miR-150 are associated with a lower risk of melanoma recurrence, highlighting its complex role [208]. The PCD protein-4 (PDCD4) is an apoptotic protein [209], that suppresses tumorigenesis and its down-regulation is associated with poor prognosis and higher metastasis in various cancers [209], including melanoma [210]. PDCD4 triggers apoptosis by preventing protein translation through binding to eIF4A, which is an essential factor for the initiation of protein translation [211]. Wan et al. revealed that PDCD4 is one of the target genes of miR-150 in melanoma. They reported that in melanoma tissues the level of miR-150 is increased, while the expression of PDCD4 is decreased significantly. Also, the inhibition of miR-150 induced cell apoptosis and triggered the expression of caspase-8, an apoptotic protein, in the human melanoma A357 cell line. Similarly, Zhang et al. revealed that miR-150 triggers proliferation and metastasis of cervical cancer cells via targeting PDCD4 [212]. Therefore, miR-150 is a promising biomarker for the treatment and prognosis of melanoma [213].

Elevated levels of miR-21 have been indicated in human melanoma cell lines and positively correlated with tumor invasiveness [214]. SPRY1 and PDCD4 are two targets of miR-21 [215]. SPRY1 plays a crucial role in the proliferation, migration, and apoptosis of cancerous cells by inhibiting MAPK/ERK signaling [216, 217]. Mao et al. found up-regulation of miR-21 and down-regulation of SPRY1 and PDCD4 genes in melanoma tissues compared to adjacent normal tissues. They suggested that miR-21 promotes proliferation and migration and prevents apoptosis in human melanoma A375 by inhibiting SPRY1 and PDCD4 via the ERK/NF-κB signaling pathway [218]. Another study by Satzger et al. demonstrated a significant increase of miR-21 in primary melanoma tissues and melanoma cell lines. Up-regulation of miR-21 was associated with higher proliferation and lower apoptosis in melanocytes [219].

In the early stages of melanomagenesis, the loss of miR-101-3p may facilitate tumor progression by influencing genomic integrity and apoptosis in melanoma cells. Malignant melanoma cells express miR-101-3p at lower levels than normal epidermal melanocytes, supporting the potential role of miR-101-3p in preventing melanomagenesis. It has been shown that ATRX, CASP3, LMNB1, and PARP are target genes of miR-103-p. miR-101-3p acts as a tumor suppressor in melanoma by causing genomic instability which leads to DNA damage and followingly apoptosis induction [220]. The tumor-suppressor role of miR-103-p has been shown in other cancers. For example, miR-103-p triggers apoptosis in oral cancer, induces autophagy in endometrial carcinoma cells, and prevents the progression of lung squamous cell carcinoma [221–223].

Dacarbazine is a chemotherapy drug for the treatment of melanoma; however, low efficiency of dacarbazine in melanoma patients has been reported [224, 225]. Treatment with dacarbazine leads to a rise in early apoptotic cells in melanoma. Overexpression of miR-204-5p reduces the proportion of dacarbazine-induced apoptosis in melanoma cells. A recent study by Lapkina et al. showed that miR-204-5p overexpression decreases BCL2 and SIRT1 expression in primary melanoma tumors and increase the efficiency of Dacarbazine [226]. Therefore, modulating the expression of miR-204-5p in combination with dacarbazine may increase the efficiency of dacarbazine in melanoma patients [227].

miR-424-5p has been revealed to have an oncogenic function in different tumors, such as thyroid [228], gastric [229], and colorectal [230] cancers. Han et al. [231] reported down-regulation of a lncRNA, namely TINCR, and, conversely, overexpression of miR-424-5p in cutaneous malignant melanoma (CMM) tissues. Overexpression of TINCR decreases proliferation and invasion, and increases apoptosis of CMM cell lines. TINCR is negatively correlated with miR-424-5p in CMM tissues [232]. LATS1, a tumor suppressor gene, has been introduced as a target for miR-424-5p [233]. TINCR can positively regulate the expression of LATS1 and, consequently, enhances apoptosis in CMM cells [231]. This evidence supports the tumor-suppressive role of TINCR in CMM.

Elevated expression of the lncRNA HOXA11-AS and ITGA9 and down-regulation of miR-152-3p have been reported in melanoma. miR-152-3p acts as an onco-supressor in melanoma and this is achieved partially by suppressing the expression of integrin alpha9 (ITGA9). ITGA9 can induce proliferation, metastasis, and EMT while reducing apoptosis. HOXA11-AS knock-down is reduced ITGA9 expression via up-regulation of miR-152-3p. Therefore, HOXA11-AS is a potential biomarker for the treatment of cutaneous melanoma, and its inhibition can suppress tumor progression by increasing apoptosis and reducing the proliferation and metastasis of melanoma cells [234]. Similar results have been reported by silencing HOXA11-AS in various tumors such as uveal melanoma [235], glioma [236], and breast cancer [237].

Melanoma stem cells (MSCs) play a crucial role in melanomagenesis, making them a promising target for melanoma treatment. A dysregulated gene expression profile is one of the hallmarks of MSCs [238]. lncRNAs and miRNAs regulate MSCs’ characteristics and stemness by controlling their gene expression [239]. Elevated levels of LINC00698 in MSCs have been reported in a recent study. They revealed that LINC00698 sponges miR-3132 to upregulate the TCF7 gene, which plays a critical role in melanomagenesis by MSCs. Moreover, LINC00698 triggers stemness and inhibits apoptosis in MSCs by binding to the hnRNPM protein and increasing its stability [240]. Therefore, LINC00698-miR-3132-TCF7/hnRNPM axis is a potential therapeutic target for treating melanoma. LHFPL3-AS1-long is another regulator of the tumorigenic potency of MSCs. By sponging miR-181a-5p, LHFPL3-AS1-long prevents the degradation of Bcl-2 mRNA in vivo and in vitro, causing inhibition of apoptosis in MSCs. The overexpression of LHFPL3-AS1 has been observed in melanoma patients compared with healthy donors suggesting that LHFPL3-AS1-miR-181a-5p-Bcl-2 axis has a potential to be considered as a target for melanoma treatment [241].

A recent study by Feichtenschlager et al. identified a novel melanoma-related lncRNA namely T-RECS (Transcript Regulating Cell Survival) which was up-regulated in NRAS- or BRAF-mutated melanoma cell lines and patients. T-RECS stabilizes the hnRNPA2/B1 protein (a regulator of MAPK signaling) which can justify coexpression of both of them in skin samples of melanoma patients. Targeting T-RECS by antisense oligonucleotides diminished the growth rate and increased apoptosis in melanoma cells while it had a low effect on normal cells [242].

Elevated expression of another lncRNA, GAS6-AS2, in melanoma tissues is correlated with poor prognosis and advanced stages. Also, increased expression of GAS6-AS2 inhibits apoptosis and promotes proliferation in melanoma cells. Similarly, GAS6-AS2 knockdown reduces proliferation and metastasis but triggeres apoptosis via targeting miR-493-5p in hepatocellular carcinoma [243]. It has been shown that GAS6-AS2 induces melanoma xenograft growth in vivo. GAS6-AS2 promotes the expression of the GAS6, an activator of AXL/AKT/ERK signaling. Therefore, GAS6-AS2 can serve as an oncogenic lncRNA in melanoma by triggering the GAS6/AXL/AKT/ERK signaling pathway. Knock-out of GAS6-AS2 can serve as a potential treatment for melanoma by the mechanism of suppressing the progression of cancer via promoting apoptosis [244].

The expression of GAS5, a lncRNA, is reduced in samples of malignant melanoma patients and its level is negatively correlated with larger tumor size, higher TNM stage, and increased risk of metastasis [245, 246]. GAS5 knock-down increases viability and proliferation of melanoma cells by preventing apoptosis via increasing the expression of Bcl-2 and also inducing cell cycle progression through up-regulation of Cyclin D1, CDK4, and p27 expression [246].

Lu et al. reported up-regulation of a circRNA, circ_0079593, in melanoma and also, showed that silencing this circRNA increased apoptosis and repressed proliferation, migration, invasion, and glucose metabolism in vitro. Moreover, they reported that circ_0079593 knock-down arrested tumor growth in mice xenograft melanoma model. Therefore, circ_0079593 can be considered an oncogene in melanoma [247]. miR-516b is down-regulated in melanoma, and circ_0079593 proved to be a sponge for it. GRM3 is the target of miR-516b, and its oncogenic function has been revealed in many cancers, including melanoma [247–249]. Circ_0079593 inhibits apoptosis in melanoma cells by triggering GRM3 through sponging miR-516b. In conclusion, circ_0079593 might serve as a biomarker for the treatment of melanoma [247]. In glioma it has been shown that suppressing circ_0079593 reduced proliferation by enhancing apoptosis in cancer cell [250].

Another circular RNA, circ-FOXM1, is increased in melanoma and its reduction provoked apoptosis and inhibited proliferation, invasion, and glycolysis in melanoma cells in vitro. Similarly, silencing circ-FOXM1 can induce apoptosis in osteosarcoma [251]. By sponging miR-143-3p, circ-FOXM1 induces overexpression of FLOT2, the target of miR-143-3p, and consequently reduces apoptosis in melanoma cells and facilitates melanomagenesis. Therefore, targeting circ-FOXM1 is a potential treatment for restoring apoptosis and hampering progression in melanoma cells [252]. It has been shown that overexpression of miR-143 hampers proliferation and promotes apoptosis in cervical cancer cell lines [253]. Nabipoorshrafi et al. have revealed that overexpression of miR-143 triggers apoptosis in melanoma cell lines. Therefore, miR-143 exhibits tumor suppressor activity in melanoma; however, the exact mechanism of miR-143-induced apoptosis in melanoma cells has not been understood yet [254].

Downregulation of miR-485-5p has been found in melanoma cells. Overexpression of miR-485-5p triggers apoptosis while reducing proliferation and metastasis in melanoma and breast cancer cells [255, 256]. Therefore, miR-485-5p has a tumor suppressor function in melanoma cells. Expression of PRRX1, a target for miR-485-5p, is increased in melanoma cells, and its silencing promotes apoptosis and diminishes viability and metastasis in melanoma cells. PRXX1 mediates these functions through the activation of TGFβ. Thus, miR-485-5p can be considered a biomarker for the detection and treatment of melanoma [256].

miR-26b is an important miRNA widely known due to its inhibitory role in cancer progression [257, 258]. Investigations have indicated down-regulation of miR-26b in human melanoma specimens [259, 260]. Li et al. illustrated that in malignant melanoma cells, miR-26b acts as a tumor suppressor by regulating MAPK signaling and apoptosis. Also, they revealed that treatment of melanoma cells can change the expression of Bcl-2 protein, whereas the expression of Bax, Bid, or Bcl-XL was unchanged. Therefore, there is a functional interaction between miR-26b and Bcl-2, but, understanding the exact mechanism needs more investigation [260].

Necroptosis

Necroptosis, a type of cell death that is not apoptotic and does not require caspase, was described by Degterev et al. in 2005 [261]. Receptor-interacting protein kinase 1 (RIPK1), RIPK3, and mixed lineage kinase domain-like protein (MLKL) make up the necrosome, which mediates necroptosis [262]. Extracellular stress triggers necroptosis through pathways such as TNF-α/TNFR [263], Fas ligand/FAS [264], interferon-gamma (IFN-γ)/IFNAR1 [265], double-stranded RNA/Toll-like receptor 3 (TLR3) [266], and double-stranded DNA/Z-DNA binding protein 1 (ZBP1) [267]. Necroptosis causes inflammation following a rupture in the plasma membrane and leakage of damage-associated molecular patterns (DAMPs) from the cells [268]. Therefore, necroptosis can trigger anti-tumor responses by activating multiple immune mediators such as DCs, macrophages, natural killer cells, and CD4+/CD8 + T lymphocytes [269, 270].

Mounting evidence suggests that necroptosis may act as a double-edged sword in the progression of cancer. The occurrence of necroptosis in tumor cells causes the release of DAMPs, cytokines, and chemokines, which attract various APCs, including DCs, macrophages, and monocytes to TME. APCs take up released antigens from necroptotic tumor cells and cross-prime naïve CD8 + T cells to activate immunity against tumor cells [271]. Therefore, induction of necroptosis using various agents such as nanoparticles can be a promising strategy for cancer treatment [272]. However, some other studies have investigated that triggering necroptosis can turn TME in favor of tumor growth by recruiting immunosuppressive cells. Seifert et al. showed that necrosomes can trigger oncogenesis in pancreatic cells in two ways. First, activated necrosome induces CXCL1 production, which attracts myeloid-derived suppressor cells (MDSCs) into TME. MDSCs are immunosuppressive cells that diminish anti-tumor immune responses and induce tumor progression. Second, activation of necrosome induces Mincle production which reduces DC activation and thereby inhibits T cell anti-tumor responses [273]. Moreover, cytokines released during necroptosis of cancer cells can induce proliferation of tumor cells, angiogenesis, and metastasis [274]. Based on this evidence, necroptosis has a dual role in cancer. Further exploration is required to investigate the complex cross-talk between necroptosis and the TME for developing new cancer therapies by modulating necroptosis.

There is a complex interaction between ncRNAs and necroptosis that profoundly influences cellular fate. Until now, multiple ncRNAs have known to regulate the function of RIPK1, RIPK3, and MLKL, which are the key components of necrosome [275]. The lncRNA PACER case is one example that clearly demonstrates how ncRNAs interact with the necroptosis pathway. The lncRNA PACER demonstrates a context-dependent dual role in necroptosis, either triggering or inhibiting the process by controlling the expression of RIPK1 and RIPK3. Bozgeyik et al. showed that PACER may function as a tumor promoter by inhibiting necroptosis cell death signaling pathways via control of RIPK3 [276]. miR-141-3p is another regulator of necroptosis that can directly interact with RIPK1 and suppress necroptosis [277]. miR-425-5p directly targets RIP1 mRNA and reduces its expression, which leads to inhibition of necroptosis. Some other miRNAs can trigger necroptosis, for example, miR-200a-5p induces necroptosis by interacting with RIPK3 in vitro and in vivo [278]. It has been shown that miR-7 induces necroptosis by targeting SLC25A37 and TIMM50 in the rhabdomyosarcoma tumor model and cancer cell lines [279]. Highlighting their potential in the predictive modeling of cancer, necroptosis-related lncRNAs are being explored to construct prognostic tools for cancer. Chen et al. developed a necroptosis-related lncRNA model to predict immunotherapy response in breast cancer. The most important necroptosis-related lncRNAs involved in building the predictive model were SH3BP5-AS1, AC012073.1, AC120114.1, LINC00377, AL133467.1, and AC036108.3 [280]. In conclusion, ncRNAs intricately control necroptosis in cancer, acting as both activators and inhibitors by targeting core pathway components like RIPK1 and RIPK3, thereby influencing tumor cell fate.

Evidence such as low expression of CYLD and RIPK3 in cell lines supports the idea that the role of necroptosis is limited in melanoma [281–283]. This could be a natural characteristic of melanoma cells to limit necroptosis and restrict immune responses. Also, low expression of RIPK1 has been reported in melanoma cells, which triggers apoptosis and ER-stress-induced autophagy [284]. Manipulating necroptosis machinery and activating this cell death type has been assumed as a potential therapeutic for targeting melanoma cells. Treating B16 melanoma cells with a pan-caspase inhibitor in combination with radiochemoimmunotherapy decreases tumor growth by triggering necroptosis and increasing infiltration of immune cells such as DCs and CD8 + T lymphocytes in TME [285]. Transferring p19^Arf and interferon-β genes to melanoma cells triggers necroptosis and activation of NK cells, T lymphocytes, and neutrophils [286]This evidence suggests that manipulating the necroptosis pathway may be a useful therapeutic approach for melanoma. Therefore, clarifying the mechanism of necroptosis in melanoma and identifying necroptosis-related ncRNAs can lead to finding new targets for the treatment of melanoma.

To our knowledge, until now, only two studies have been conducted to decipher the role of necroptosis-related ncRNAs in melanoma. In a bioinformatics study by Zhang et al., USP30-AS1, LINC01711, LINC00520, NRIR, BASP1-AS1, and LINC02178 were necroptosis-related hub lncRNAs. They revealed that USP30-AS1 is correlated with RIPK3 and MLKL, and they also showed that increased USP30-AS1 expression is associated with decreased risk scores in skin cutaneous melanoma (SKCM) [287]. As shown in multiple studies, overexpression of RIPK3 and MLKL in cancers is associated with a higher occurrence of necroptosis and activation of immune responses against tumor cells [288, 289]. Liu et al. identified seven necroptosis-related lncRNAs, including EBLN3P, AC093010.2, LINC01871, IRF2-DT, AL162457.2, AC242842.1, HLA-DQB1-AS1. Between these lncRNAs, AL162457.2 was overexpressed in the melanoma high-risk group. Therefore, they performed a series of functional cell assays to investigate the role of AL162457.2 in melanoma. Silencing AL162457.2 decreased proliferation, invasion, and migration of tumor cells in vitro [290]. Further research is needed to shed light on the role of necroptosis-related ncRNAs in melanoma.

Anoikis

Anoikis is an apoptotic form of cell death that occurs due to disruption of cell-cell or cell-extracellular matrix attachments. It serves both as a normal physiological process and a pathological defense mechanism that helps maintain tissue homeostasis by removing misplaced or detached cells. Importantly, anoikis can be considered a mechanism that prevents cancer cell metastasis by impairing their adherent-independent growth and invasion into distant areas [291]. Nevertheless, resistance to anoikis has become a hallmark of aggressive, metastatic cancers. Malignant tumors, including melanoma, often overcome anoikis through poorly understood mechanisms that allow cells to evade anoikis and spread through metastasis [292].

Anoikis resistance is one of the main reasons for the metastatic and aggressive nature of melanoma. While the underlying mechanism(s) of anoikis resistance in melanoma cells have not yet been fully understood, it has been shown that the TME plays a pivotal role in promoting anoikis resistance, facilitating melanoma progression and metastasis. Alterations in the ECM composition and stiffness (such as overexpression of integrins, fibronectin, laminin, and syndecans) activate pro-survival signaling pathways like PI3K/Akt and ERK, promoting anchorage-independent growth and anoikis resistance. The hypoxic condition within the TME stabilizes HIF-1α and induces glycolytic metabolism to counteract anoikis-induced metabolic stress. Epithelial-mesenchymal transition is another contributor to anoikis resistance, as during this process, cells acquire stem-like properties that support their anchorage-independent growth [293, 294]. Aberrations in B-RAF, upregulation and phosphorylation of STAT3, activation of the yes-associated protein (YAP), and expression of anti-apoptotic factors such as Bcl2 and FLIP have also been observed in anoikis-resistant melanoma cells [291, 295].

Recent studies have provided valuable insights into the molecular players contributing to anoikis resistance in melanoma. It has been revealed that the expression of miR-214 is upregulated in metastatic melanoma, where it facilitates cell movement and metastasis. MiR-214 contributes to melanoma invasion and metastasis by direct inhibition of TFAP2C and ITGA3 mRNAs along with other surface proteins involved in cell adhesion. It has been shown that overexpression of miR-214 or silencing TFAP2C leads to metastasis of melanoma, while downregulation of miR-214 or concurrent overexpression of miR-214 and a TFAP2C mutant lacking its 3′UTR reduces the invasive potential of melanoma. Anoikis assays in melanoma cell cultures also support the role of miR-214 in conferring anoikis resistance [296]. Additionally, the miR-200 family (including miR-200a, miR-200b, miR-200c, miR-141, and miR-429) and miR-205 are involved in suppression of anoikis resistance, epithelial-mesenchymal transition, and cell motility by targeting ZEB1 and SIP1 [297]. Downregulation of miR-200 family members is correlated with invasion and metastasis of various cancers [298–300].

Recently, Zhong et al. developed an anoikis-related lncRNA signature to predict the prognosis and immune status of patients with cutaneous melanoma. They identified six anoikis-related lncRNAs with prognostic value, among which four lncRNAs (AC083799.1, VIM − AS1, AC005261.1, and LINC01819) were associated with good prognosis and have protective roles, and two other lncRNAs (DLEU1 and AC090409.1) were associated with poor prognosis. This novel risk model not only aids in predicting patient outcomes but also offers a tool for assessing the immune landscape of melanoma, which can help to develop personalized therapeutic approaches for melanoma patients [301].

Identifying the regulatory network of anoikis, including the role of ncRNAs, can lead to the development of accurate prognostic biomarkers. Moreover, by targeting the key ncRNAs involved in anoikis resistance, the metastatic potential of melanoma can be reduced, leading to an improvement in treatment outcomes in melanoma patients.

Cuproptosis

In 2022, for the first time, Tsvetkov et al. discovered a novel type of PCD that was copper-dependent, called cuproptosis [302]. Excess amounts of intracellular copper can lead to aggregation of dihydrolipoamide S-acetyltransferase (DLAT), which causes proteotoxic stress and, finally, cell death [303]. More and more studies are employing bioinformatics analysis to focus on the critical connection between cuproptosis and the cancer process. These bioinformatics studies attempt to identify the association between cancer and cuproptosis with the help of cuproptosis-related genes known by Tsvetkov et al.’s study [304].

According to Tsvetko et al., increased mitochondrial reactivity oxidative phosphorylation (OXPHOS) is linked to cuproptosis [302]. Formerly, it was thought that the metabolism of tumor cells mainly relies on glycolysis; however, recent studies have highlighted the role of OXPHOS in ATP synthesis in many tumor types [305]. A significant increase in OXPHOS in melanoma has been reported by Aminzadeh-Gohari et al. [306]. Therefore, cuproptosis may play a role in melanoma; however, it is not clear whether its role is oncogenic or tumor-suppressive. Until now, few papers have been published to investigate cuproptosis-related ncRNA signatures in melanoma.

In a bioinformatics study, Zhou et al. reported the increased expression of some cuproptosis-related lncRNAs, including LINC01150, EBLN3P, MIR100HG, WAC-AS1, LINC00339, and USP30-AS1, and decreased expression of AC009495.1 in low-risk melanoma in comparison with high-risk group [307]. Interestingly, some of these up-regulated lncRNAs in low-risk groups, such as MIR100HG, LINC00339, and WAC-AS1, have previously been considered as oncogenes in other cancer types [308–310]. These paradoxical observations may be justified by the tissue and cell-dependent function of lncRNAs. There is evidence that the same lncRNA can act as both tumor oncogene and suppressor depending on the cell and tissue environment [311, 312]. Another bioinformatics study by Yang et al. identified five protective lncRNAs, including VIM-AS1, AC012443.2, MALINC1, AL354696.2, and HSD11B1-AS1 for melanoma. By constructing a risk-predictive model, they revealed that these five lncRNAs act as protective factors to decrease the risk of skin cutaneous melanoma [313]. More investigations are needed to understand the exact role of cuproptosis-related lncRNAs in the prognosis, diagnosis, and treatment of melanoma.

Clinical translation of NcRNAs in melanoma: diagnostic, prognostic, and therapeutic perspectives

In the previous sections, the role of ncRNAs as pivotal regulators of PCD in melanoma was comprehensively explored and it has discussed how disregulan of these ncRNAs can affect tumor progression and response to treatment. These findings highlight the potential of ncRNAs as diagnostic/prognostic biomarkers and therapeutic targets/agents.

Diagnostic and prognostic applications

Owing to the cell- and tissue-specific expression pattern, stability in the body fluids, and association with disease progression and therapeutic resistance, ncRNAs hold great potential as diagnostic and prognostic indicators in melanoma [314]. Although the diagnostic and prognostic role of ncRNAs in melanoma has not yet been widely explored in clinical settings, the experiences from other types of cancer are encouraging. For example, PCA3 (prostate cancer antigen 3) (marketed as Progensa) is an approved urinary biomarker for the diagnosis of prostate cancer [315]. Numerous clinical trials are also underway to assess miRNAs as cancer biomarkers for prognostication, diagnosis, and therapy response prediction purposes. For example, in a clinical trial (NCT03591367) miR-155 was used for diagnosis of non-muscle-invasive bladder cancer [89]. In another clinical trial (NCT05477667), let-7a and miR-124 were tested to diagnose non-Hodgkin’s and acute lymphoma. Moreover, some clinical trials have assessed miRNAs as predictors of cancer therapy efficacy. For instance, in a clinical trial (NCT04662996), researchers examined the potential of miRNAs to serve as indicators of chemotherapy effectiveness in patients with metastatic castration-resistant prostate cancer. These studies support the translational potential of ncRNAs in melanoma as diagnostic/prognostic biomarkers and predictors of response to therapy. Nonetheless, large-scale and multicenter clinical studies are required to evaluate the sensitivity, specificity, and clinical applicability of these biomarkers. Integrating ncRNA profiling into the currently used diagnostic/prognostic approaches can significantly improve early detection, disease progression monitoring, patient stratification, and therapy monitoring. Nonetheless, clinical translation of ncRNAs as reliable biomarkers faces several technical challenges. The RNA sequencing results can significantly vary based on the techniques used, which can cause reproducibility issues [316]. Due to the lack of advanced targeted enrichment techniques, the detection of ncRNAs that are expressed at low levels can be problematic [317]. Moreover, there is no universal consensus on ncRNA expression cutoffs to make a clinical decision [318]. On the other hand, RNA sequencing results can be considerably affected by preanalytical variables such as sample collection, storage, and RNA isolation [319]. It also should be noted that generally, a single ncRNA lacks sufficient diagnostic/prognostic power. In this regard, combining multiple ncRNAs into diagnostic/prognostic signatures can compensate variability of individual ncRNAs and improve the accuracy.

ncRNA-based therapeutic interventions

Therapeutic modulation of the ncRNA network is a promising strategy for the treatment of melanoma, which can be performed either by restoring tumor-suppressor ncRNAs or inhibiting oncogenic ncRNAs. NcRNA therapeutics can be combined with other therapeutic approaches to enhance their efficacy and overcome therapeutic resistance. Here we explore various strategies for downregulation or overexpression of ncRNAs and discuss their advantages, disadvantages, and clinical applicability.

Downregulation strategies

Downregulation of oncogenic ncRNAs can restore the susceptibility of cancer cells to PCD and facilitate tumor clearance. Antisense oligonucleotides (ASOs), RNA interference (RNAi), and CRISPR/Cas9-mediated gene editing are common approaches to the downregulation of ncRNAs. ASOs are single-strand sequences that bind to complementary RNA molecules, leading to the degradation of target RNAs by recruitment of RNase H. Several FDA-approved ASO products indicate their safety and efficacy [320]. Nonetheless, their usage in clinical settings faces several challenges, including delivery challenges, in vivo stability, immunogenicity, off-target RNA binding, difficulty of targeting nuclear ncRNAs, and steric hindrance against ncRNAs with secondary/tertiary structures (322). There are various potential strategies to overcome these hurdles. For example, 2′-O-Methoxyethyl (2′-MOE) or Locked Nucleic Acids (LNAs) modifications can be utilized to enhance the stability and binding affinity of ASOs [322]. In recent years, lipid nanoparticles (LNPs), pH-sensitive polymers, and natural cell-derived exosomes have emerged as targeted delivery vehicles capable of delivering ncRNAs into tumor sites and protecting them from endosomal degradation [323, 324]. RNA interference (RNAi) using siRNAs or shRNAs is another reliable approach to downregulation of oncogenic ncRNAs. LNP-encapsulated siRNAs (e.g., Patisiran) have shown remarkable efficacy in non-cancer diseases, raising hopes for oncology applications [325]. Slow release formulations, such as poly(lactic-co-glycolic acid) (PLGA) microspheres, can reduce dosing frequency by sustaining therapeutic levels of ASOs/siRNAs, leading to the reduction of toxicity risk [326]. The CRISPR system is another option that can be used to disrupt oncogenic ncRNAs either at the DNA or RNA levels. The traditional CRISPR/Cas9 system enables permanent disruption of promoter regions or entire loci of oncogenic ncRNAs; nonetheless, this strategy is associated with the risk of genotoxicity [327]. In recent years, some CRISPR-based technologies have emerged, enabling the downregulation of ncRNAs without altering genomic sequences. For example, CRISPR interference (CRISPRi) and CRISPR/Cas13 platforms allow epigenetic silencing and direct RNA editing, respectively; nonetheless, they have not yet been widely adopted for clinical applications [328].

Upregulation strategies

Restoring the expression or function of ncRNAs in melanoma can reactivate PCD pathways, reverse drug resistance, and enhance tumor regression. While most therapeutic strategies focus on the downregulation of oncogenic ncRNAs, the upregulation of tumor-suppressive ncRNAs can be equally important. There are various strategies enabling the transient or stable restoration of ncRNAs. For example, synthetic miRNA agonists can transiently restore the tumor suppressive function by mimicking endogenous miRNAs [329]. In 2020, MRX34, a liposomal miR-34a mimic, was evaluated in a phase I clinical trial for patients with solid tumors, including melanoma. Despite clinical efficacy, this trial was ended due to severe immune-related adverse events [330]. Gene therapy using viral vectors, particularly lentiviral and adeno-associated viral (AAV) vectors, enables stable expression of ncRNAs and overcomes the need for repeated dosing of synthetic oligonucleotides. While the lentiviral platform is widely used in ex vivo gene therapy, it can also be utilized for in vivo gene delivery. Nonetheless, the risk of insertional mutagenesis and labor-intensive manufacturing processes are the main roadblocks to its clinical use [331]. In contrast, the AAV platform is widely adopted for in vivo gene delivery, as evidenced by various FDA-approved AAV-mediated in vivo gene therapies. However, due to the limited cargo capacity of AAV vectors (~ 4.7 kb), the delivery of long lncRNAs is problematic. Neutralization of AAV vectors in some individuals due to the presence of pre-existing antibodies is another hurdle for their clinical applications [332]. CRISPR activation (CRISPRa) system and epigenetic modulators are other emerging technologies for inducing the transcription of ncRNA loci [333, 334]; however, their applications are still in preclinical studies and have not yet been widely adopted for clinical settings.

Conclusions and future perspectives

Dysregulation of PCD is one of the underlying mechanisms involved in the progression, metastasis, and drug resistance of melanoma. Thus, focusing on PCD and its regulatory processes can lead to a better understanding of melanoma and the design of new therapeutic approaches. NcRNAs, which represent a significant portion of the human genome’s transcriptional output, serve as master regulators of a wide array of biological functions, including the intricate control of PCD. In this review, the role of ncRNAs in the regulation of different types of PCD, including apoptosis, autophagy, necroptosis, ferroptosis, pyroptosis, anoikis, and cuproptosis in melanoma, has been comprehensively discussed (Fig. 4). Given the target-specific function and tissue-, disease-, and disease developmental stage-specific expression profile, ncRNAs hold considerable potential as both diagnostic and prognostic biomarkers as well as therapeutic targets. The construction of ncRNA-based prognostic models can significantly enhance the ability to predict patient risk, immune status, and responsiveness to treatment, enabling more personalized therapeutic approaches. Nonetheless, the intricate relationship between ncRNAs and PCD, specifically the non-apoptotic form, has not yet been comprehensively understood and requires further in-depth investigations and validations. Moreover, variability in sequencing techniques is another hurdle for the use of ncRNAs as diagnostic/prognostic biomarkers, underscoring the urgent need for developing standardized approaches for ncRNA detection and quantification.

On the other hand, despite outstanding progress, the use of ncRNA-based therapeutics in clinical settings faces several hurdles, including stability, off-target delivery, and therapeutic resistance. Structural modifications and nanocarrier-based encapsulation can protect therapeutic ncRNAs from degradation and facilitate their targeted delivery. In this regard, emerging in vivo delivery platforms such as LNPs, polymeric nanoparticles, and exosomes hold great promise. Therapeutic resistance is another concern regarding the use of ncRNA-targeted therapeutic interventions, since the cancer cells can reverse the efficacy of these interventions via activation of parallel compensatory pathways. Integrating ncRNA modulating approaches into other therapies like chemotherapy or immunotherapy can help to overcome therapeutic resistance. It should be noted that ncRNAs have a context-dependent function. This means that the same ncRNA may play two opposite roles in two different cellular contexts. Integrating multi-omics data can help to provide a more holistic understanding of the ncRNA regulatory network in melanoma.

Future research, encompassing both in vitro and in vivo studies, alongside rigorous clinical validation, is essential to advance our knowledge of the ncRNA-driven regulation of PCD in melanoma. Such efforts have the potential to catalyze significant breakthroughs in melanoma diagnosis, prognosis prediction, and treatment, ultimately leading to more targeted and effective therapeutic interventions.

Author contributions

YM: Conceptualization, Methodology, Investigation, Project administration, Writing – original draft.MS, BN, FJ: Investigation, Validation, Writing – original draft.MJ, NG: Investigation, Writing – original draft.EK, EA: Conceptualization, Project administration, Supervision, Writing – review & editing.

Funding

Not applicable.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.