Genetic and epigenetic plasticity allows tumors to evade single-targeted treatments. Here we direct Bcl2-specific short interfering RNA (siRNA) with 5'-triphosphate ends (3p-siRNA) against melanoma. Recognition of 5'-triphosphate by the cytosolic antiviral helicase retinoic acid-induced protein I (Rig-I, encoded by Ddx58) activated innate immune cells such as dendritic cells and directly induced expression of interferons (IFNs) and apoptosis in tumor cells. These Rig-I-mediated activities synergized with siRNA-mediated Bcl2 silencing to provoke massive apoptosis of tumor cells in lung metastases in vivo. The therapeutic activity required natural killer cells and IFN, as well as silencing of Bcl2, as evidenced by rescue with a mutated Bcl2 target, by site-specific cleavage of Bcl2 messenger RNA in lung metastases and downregulation of Bcl-2 protein in tumor cells in vivo. Together, 3p-siRNA represents a single molecule-based approach in which Rig-I activation on both the immune- and tumor cell level corrects immune ignorance and in which gene silencing corrects key molecular events that govern tumor cell survival.

Recent studies revealed trajectories of mutational events in early melanomagenesis, but the accompanying changes in gene expression are far less understood. Therefore, we performed a comprehensive RNA-seq analysis of laser-microdissected melanocytic nevi (n = 23) and primary melanoma samples (n = 57) and characterized the molecular mechanisms of early melanoma development. Using self-organizing maps, unsupervised clustering, and analysis of pseudotime (PT) dynamics to identify evolutionary trajectories, we describe here two transcriptomic types of melanocytic nevi (N1 and N2) and primary melanomas (M1 and M2). N1/M1 lesions are characterized by pigmentation-type and MITF gene signatures, and a high prevalence of NRAS mutations in M1 melanomas. N2/M2 lesions are characterized by inflammatory-type and AXL gene signatures with an equal distribution of wild-type and mutated BRAF and low prevalence of NRAS mutations in M2 melanomas. Interestingly, N1 nevi and M1 melanomas and N2 nevi and M2 melanomas, respectively, cluster together, but there is no clustering in a stage-dependent manner. Transcriptional signatures of M1 melanomas harbor signatures of BRAF/MEK inhibitor resistance and M2 melanomas harbor signatures of anti-PD-1 antibody treatment resistance. Pseudotime dynamics of nevus and melanoma samples are suggestive for a switch-like immune-escape mechanism in melanoma development with downregulation of immune genes paralleled by an increasing expression of a cell cycle signature in late-stage melanomas. Taken together, the transcriptome analysis identifies gene signatures and mechanisms underlying development of melanoma in early and late stages with relevance for diagnostics and therapy.

BackgroundLymphocyte activating 3 (LAG3, LAG-3, CD223) is a promising target for immune checkpoint inhibition in clear cell renal cell carcinoma (KIRC). The aim of this study was to investigate the epigenetic regulation ofLAG3in KIRC by methylation.MethodsWe correlated quantitativeLAG3methylation levels with transcriptional activity, immune cell infiltration, and overall survival in a cohort of n=533 patients with KIRC and n=160 normal adjacent tissue (NAT) samples obtained from The Cancer Genome Atlas (TCGA). Furthermore, we analyzedLAG3methylation in peripheral blood mononuclear cells (PBMCs) and KIRC cell lines. We validated correlations between LAG3 expression, immune cell infiltrates, survival, and methylation in an independent KIRC cohort (University Hospital Bonn (UHB) cohort, n=118) by means of immunohistochemistry and quantitative methylation-specific PCR.ResultsWe found differential methylation profiles among PBMCs, NAT, KIRC cell lines, and KIRC tumor tissue. Methylation strongly correlated with LAG3 mRNA expression in KIRCs (TCGA cohort) and KIRC cell lines. In the UHB cohort, methylation correlated with LAG3-positive immune cells and tumor-intrinsic LAG3 protein expression. Furthermore,LAG3methylation strongly correlated with signatures of distinct immune cell infiltrates, an interferon-y signature (TCGA cohort), and immunohistochemically quantified CD45+, CD8+, and CD4+immune cell infiltrates (UHB cohort). LAG3 mRNA expression (TCGA cohort), methylation (both cohorts), and tumor cell-intrinsic protein expression (UHB cohort) was significantly associated with overall survival.ConclusionOur data suggest an epigenetic regulation of LAG3 expression in tumor and immune cells via DNA methylation. LAG3 expression and methylation is associated with a subset of KIRCs showing a distinct clinical course and immunogenicity. Our study provides rationale for further testingLAG3DNA methylation as a predictive biomarker for response to LAG3 immune checkpoint inhibitors.