1. Long Non-Coding RNAs (LncRNAs) and Immune Evasion in Melanoma
Melanoma progression depends not only on the accumulation of genetic alteration and the acquisition of new phenotypic states but also on the interaction of melanoma cells with the immune system
[1]. The melanoma niche contains many immune-suppressive cells: regulatory T-cells, myeloid-derived suppressor cells, and tumor-associated macrophages
[2]. During melanoma progression, a number of soluble factors released by either tumor or ‘immune-suppressive cells inhibit the capacity of antigen-presenting cells, dendritic cells, or macrophages to process and correctly expose melanoma-associated new antigene, leading to impaired T-cell activation and immune evasion
[3][4][5]. Recently, pan-cancer genome-wide studies identified LncRNAs associated gene-expression signatures of the pro-tumor and anti-tumor TME, associated with response to immunotherapy
[6][7][8]. Notably, immune-related lncRNA signatures were shown to be predictive of prognosis, survival, and immunotherapeutic efficacy in breast cancer, lung adenocarcinoma, hepatocellular carcinoma, bladder cancer
[9][10][11][12][13], as well in melanoma
[7][8][13][14][15][16][17][18]. Four distinct LncRNAs immune-related signatures predictive of patient prognosis and survival were identified using the cutaneous skin melanoma (SKCM) TCGA dataset, including LncRNAs related to antigen processing and presentation, cytokines, Toll-like and cytokine receptor signaling pathway, and NK cell-mediated immunity
[13][15][17][18].
In recent years, advances in understanding the cellular and molecular mechanisms of tumor immune-escape led to the development and approval of many immunotherapies. Immune checkpoint inhibitors (ICI) targeting programmed death-1 (PD-1), its ligand PD-L1, and cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) have revolutionized the treatment of metastatic melanoma, leading to long-term remissions or cure. However, about 50–60% of patients still show primary resistance to ICI therapy or develop secondary resistance during treatment, and the underlying molecular mechanisms are not fully understood. Among the most characterized cell-autonomous resistance mechanisms is the downregulation of b2-microglobulin and human leukocyte antigens (HLA) expression, leading to tumor evasion from T-cell specific cytotoxicity. Primary ICI-resistance in melanoma is frequently associated with a low mutational burden
[19][20][21][22].
Yu et al. identified a 4-lncRNA signature (AC002116-2, AP000251-1, TMEM147-AS1, and NKILA) associated with a response to immunotherapy (anti-PD-1, CTLA4 and a cytokine tumor vaccine in 71 melanoma patients from the SKCM TCGA dataset)
[16]. In another study, Zhou and colleagues using two independent training and validation melanoma-datasets identified a 15-lncRNA signature (AC010904.2, LINC01126, AC012360.1, AC024933.1, AL442128.2, AC022211.4, AC022211.2, AC127496.5, NARF-AS1, AP000919.3, AP005329.2, AC023983.1, AC023983.2, AC139100.1, and AC012615.4) that predicts response to anti-PD-1 monotherapy
[14].
To date, the only lncRNA known to play a direct role in the anti-melanoma immune response is LIMIT, a lncRNA that induces MHC-I expression and tumor immunogenicity. The LIMIT gene is conserved in mice and humans, and its expression correlates with lymphocyte infiltration and expression of interferon response genes in the SKCM TCGA dataset. LIMIT expression is induced by IFNγ treatment in melanoma cells and
cis-activates the guanylate binding protein (GBP) gene cluster, leading to activation of the MHC-I antigen presentation machinery through the heat shock factor 1 (HSF1). Li and colleagues showed that increased LIMIT expression in melanoma patients and mouse transplanted melanomas correlates with MHC-I response, antitumor immunity, and enhanced efficiency of anti-PD-1 therapy. This is the first discovery of an immunogenic lncRNA in cancer, whose therapeutic targeting can represent a useful approach to increase the efficacy of cancer immunotherapy
[23].
Circular RNAs are an emerging class of RNAs involved in the regulation of immune response
[24]. A recently described circRNA, circ_0020710, deriving from CD151 mRNA encoding an oncogenic transmembrane protein, promotes melanoma progression and contributes to melanoma immune evasion by sponging miR-370-3p and upregulating CXCL12 levels. It was shown that high levels of circ_0020710 correlate with cytotoxic T-lymphocyte exhaustion in melanoma patients and that combined treatment of the CXCL12 inhibitor AMD3100 and anti PD-(L)1 enhance its therapeutic efficacy in mouse xenograft melanoma models
[25]. Recent findings support novel functions of cirRNA in natural anti-cancer immunity. The innate immune system in mammals can distinguish between native vs. exogenous circRNA. N6-methyladenosine (m6A) methylation is critical for discrimination, which is the most frequent and abundant transcriptional modification in eukaryotic RNAs. While m6A-modified circRNAs inhibit immune response and are recognized as self cirRNAs, foreign m6A-unmodified circRNAs are potent activators of the anti-tumor immune response, opening the possibility of using circRNAs in cancer immunotherapy
[26]. Still, the role of m6A modification in anti-tumor immune response remains controversial. It was recently described that circNDUFB2 triggers an antitumor immune response in non-small lung cancer (NSCLC) mouse model independently of m6A modification
[27].
Increasing evidence suggests that small ORFs in the LncRNAs are translated into small peptides, representing a source of cancer-associated antigens (neoantigens) in different cancer types
[28][29]. Increased neoantigen load in cancer cells is reported to increase the efficiency of immune-checkpoint inhibitor (ICI) therapy
[30][31]. Qi et al. performed comprehensive proteogenomic profiling of HLA class 1-presented immunopeptides in melanomas with a high tumor mutation burden. The researchers identified 44 lncRNA-derived peptides presented by HLA class I. These data suggest that deregulation of lncRNA expression in melanoma can impact neoantigen load and predict the response of melanoma patients to immunotherapy.
2. LncRNAs and Drug Resistance in Melanoma
In recent years, the development of targeted small-molecule inhibitors has revolutionized melanoma standard-of-care and improved patient overall survival. The most frequently targeted cellular pathway in melanoma is the MAPK/ERK pathway, activated by the RAS and BRAF mutations in most melanoma patients. Both are gain-of-function mutations, driving MAPK pathway activation via MEK1/2 and ERK1/2
[32][33][34]. The discovery of the oncogenic BRAF mutation and the activation of the MAPK pathway signaling led to the development of successful combination therapies, targeting BRAF with either vemurafenib
[35][36] or dabrafenib
[37] and MEK kinase with trametinib
[38]. Although clinical trials with MEK inhibitors have shown less impressive responses than BRAF inhibitors in monotherapy, combination therapies have significantly improved initial results showing unprecedented clinical responses
[39].
However, intrinsic (5%) and acquired (50%) resistance to targeted therapy is still the major challenge for melanoma patients. Two major mechanisms underlie the onset of therapy resistance. The first is the genetic drug resistance which occurs via a selection of mutations inhibiting the response to selective, targeting inhibitors mainly through the acquisition of mutations in the drug-binding domain of targeted kinases and in the pathway downstream effectors
[40]. The second mechanism relies on the intrinsic phenotypic plasticity and intratumor heterogeneity of melanoma cells which trigger adaptive responses to therapy and acquisition of drug resistance phenotypes
[41]. The best-characterized mechanism is the phenotype switch from MITF
high/AXL
low drug-sensitive to MITF
low/AXL
high drug-resistant melanoma cells induced by inhibitors of the MAPK pathway
[41][42][43]. Unfortunately, most patients develop secondary resistance and eventually relapse despite the initial responses. Emerging evidence suggests that the same LncRNAs involved in the phenotype switch in melanoma are also involved in the acquired resistance to MAPK pathway inhibitors.
The lncRNA TSLNC8 is downregulated in melanoma patient samples with acquired resistance to vemurafenib and vemurafenib-resistant melanoma cell lines. Mechanistically, TSLNC8 binds to and regulates the subcellular localization of the catalytic subunit of protein phosphatase 1α (PP1α). Knockdown of TSLNC8 results in the cytoplasmic accumulation of PP1α and re-activation of the MAPK signaling
[44].
The lncRNA SAMMSON, which controls mitochondrial metabolism and functions as an oncogene in melanoma, is upregulated following inhibition of the ERK signaling in mutant BRAF melanoma cells, and its depletion sensitizes melanoma cells to BRAF inhibitors
[45]. Conversely, ectopic expression of SAMMSON confers resistance of melanoma cells to vemurafenib
[46].
The long intergenic lncRNA
MIRAT is upregulated following prolonged MAPK inhibition in
NRAS mutant melanomas and modulates MAPK signaling by binding to the MEK scaffold protein IQGAP1
[47]. The lncRNA POU3F3 promotes resistance of melanoma cells to the alkylating agent dacarbazine, upregulating the DNA-methyltransferase MGMT levels by sponging miR-650
[48]. Finally, the LncRNAs XIST, H19, MEG3, and LINC01291 were shown to confer resistance of melanoma cells to platinum compounds by regulating miR-21/PI3KR1 miR-18/IGF1, miR-499-5p/CYLD, and miR-625-5p/IGF-1R axis, respectively
[49][50][51][52].
The lncRNA TINCR plays a novel role in drug resistance in melanoma patients. TINCR knockdown leads to ATF4 activation, downregulation of MITF, and acquisition of an invasive phenotype, accompanied by increased resistance to vemurafenib and trametinib. Importantly, overexpression of TINCR in melanoma patient-derived xenografts partially reduced melanoma cell invasiveness and sensitized cells to the MEK1/2 inhibitor trametinib
[53].
3. LncRNAs as Biomarkers in Melanoma
LncRNAs can be secreted by cells and detected in different body fluids, such as blood, plasma/serum, or urine
[54]. They originate from apoptotic and necrotic cells or are actively secreted by living cells through extracellular vesicles. Secreted vesicles protect LncRNAs from degradation by RNAses and are therefore good candidates as stable markers for diagnosis or prognostic stratification
[55].
Several circulating LncRNAs have been described in melanoma patients. lncRNA HOTAIR was found in the plasma of patients with advanced melanoma, while HOTAIR expression in melanoma tumors strongly correlates with tumor stage
[56]. The lncRNA LINC01638 is significantly upregulated in the plasma of melanoma patients and predicts local recurrence
[57]. Similarly, SPRY4-IT1 expression is increased in the plasma of melanoma patients compared to healthy individuals, and its expression highly correlates with tumor site and stage
[58]. The plasmacytoma variant translocation 1 (PVT1) lncRNA was also detected at elevated levels in the serum of melanoma patients, and its expression correlates with tumor stage and is a marker of postoperative disease dynamics
[59] Kolenda and colleagues identified a 17-LncRNAs signature in the plasma of melanoma patients that distinguishes healthy individuals and melanoma patients. Three of these LncRNAs-IGF2AS, MEG3, ZEB2-AS1—were identified as independent prognostic factors in
BRAF-mutant advanced melanoma patient sera treated with vemurafenib
[60].
4. LncRNAs as Drug Targets
Given their function in cancer progression, LncRNAs represent attractive targets for developing new drugs. Different RNA-based approaches were employed to create lncRNA-targeting drugs, such as posttranscriptional downregulation using antisense oligonucleotides (ASOs), small interfering RNAs (siRNAs), small hairpin (shRNAs), therapeutic circular RNAs, and CRISPR-Cas9 gene editing approaches
[61][62][63]. The last generation of ASOs includes phosphorothioate oligonucleotides and locked nucleic acids (LNA) modifications, chimeric RNA-DNA-RNA ASOS (GAPmers) bearing LNA, and S-constrained ethyl modifications that have demonstrated improved potency and in vivo stability
[45][64].
A new approach to lncRNA targeting involves the steric inhibition of lncRNA interactions with nucleic acids and proteins or targeting a small molecule of lncRNA elements responsible for secondary and tertiary RNA structure formation. These small molecules can potentially destabilize lncRNA molecules or inhibit their biological functions by inhibiting their molecular interactions. The small molecule NP-C86 has been shown to disrupt the interaction of the tumor suppressor GAS5 lncRNA with UPF1, a protein involved in nonsense-mediated decay, leading to upregulation of GAS5 RNA
[65]. Two other studies identified small molecules targeting the 3′ triple-helix element of MALAT1 lncRNA, leading to a significant reduction of its expression
[66][67]
Natural Antisense RNAs have been used to
cis-regulate neighboring genes, as shown for the CDKN2B and CDKN1A loci by the lncRNA ANRIL
[68] and LincRNA-p21
[69], respectively. The long intergenic noncoding RNA p21 has also been described as an inducer of the transcriptional activity of wild-type p53
[70]. Currently, several chemically modified single-stranded ASOS, named antagoNATs, are under preclinical and clinical development
[71][72].
In addition, recent advances in RNA-based therapeutics approaches and applications open novel possibilities for using LncRNAs as drug themselves
[62][63]. The delivery system is key to properly exploiting the envisioned drug role for LncRNAs. To be satisfactory, LncRNAs should be conveyed with high stability, specificity, cell permeability, and low immunogenicity. The most frequently used carriers are viral vectors, both lentiviral and adenoviral vectors, which can rapidly infect dividing and non-dividing cells and have very long-expression time. However, the safety of viral vectors in systemic drug administration has been a matter of discussion, and these vectors have been recently replaced by non-viral vectors, such as liposomes, lipid nanoparticles (NPs), and exosomes
[62].
Liposomes are an ideal type of nanoparticles for drug and DNA/RNA delivery, as they show reduced toxicity and immunogenicity, and encapsulation improves drug stability. Several liposome-encapsulated LncRNAs are used in preclinical experiments in vitro and in vivo
[73][74][75].
Lipid NPs are similar to liposomes, but they are more suitable for encapsulating LncRNAs. They proved to inhibit tumor growth in preclinical animal trials, with no serious side effects
[76]. Recently, ASO-gold-TAT NPs were used to target the nuclear lncRNA
MALAT1 in a mouse model of xenotransplanted lung cancer, showing the efficient suppression of metastasis dissemination and increased overall survival
[77].
Exosomes are membrane-bound, endogenous vesicles containing lipids, proteins, DNA, mRNAs, miRNAs, and LncRNAs that are naturally secreted by cells
[78]. Exosomes are very promising delivery vectors and can be easily engineered to convey LncRNAs. They show lower immunogenicity and higher stability in vivo but an average packaging efficiency compared to liposomes. For this reason, exosomes can now be integrated with liposomes and NPs to improve specificity and deliverability. Liposomes, lipid NPs, and exosomes can be functionalized and delivery optimized by surface modification through genetic engineering or chemical modifications. Vesicles can be coupled with nucleic acid aptamers, antibodies, peptides, protein ligands, polymers, or small molecules
[62].
Even if several LncRNAs proved to be promising drug targets or drug themselves in both in vitro and in vivo models, up to now, none of the lncRNA or small molecule targeting LncRNAs have entered clinical trials. Moreover, the initial clinical evaluation of RNA-based drugs in cancer is still controversial: four drugs entered phase II or III clinical trials, but seven were withdrawn because of the off-target effects and lack of efficiency
[62]. The delivery, specificity, and immunogenicity of RNA-based drugs remain the major challenges in developing non-coding RNA therapeutics.