Cross-Linking Ligation And Sequencing Of Hybrids (Qclash) Reveals An Unpredicted Mirna Targetome In Melanoma Cells

Simple SummarymiRNAs are omnipresent short non-coding RNA molecules, which post-transcriptionally fine-tune the expression of most protein-coding genes in health and disease. In cancer, miRNAs have been described to directly regulate the amounts of targeted tumor suppressors and oncogenes. Although their canonical mechanism of action is well understood, the correct prediction of miRNA target genes is still a challenge. We describe here the unbiased investigation of the miRNA targetome in cancer (melanoma) cells using a technique, which allows to physically link the miRNA to its target gene. The herein identified miRNA-target interactions reveal further layers of complexity of post-transcriptional gene regulation in cancer cells and shed new light on miRNA-target gene interactions.MicroRNAs are key post-transcriptional gene regulators often displaying aberrant expression patterns in cancer. As microRNAs are promising disease-associated biomarkers and modulators of responsiveness to anti-cancer therapies, a solid understanding of their targetome is crucial. Despite enormous research efforts, the success rates of available tools to reliably predict microRNAs (miRNA)-target interactions remains limited. To investigate the disease-associated miRNA targetome, we have applied modified cross-linking ligation and sequencing of hybrids (qCLASH) to BRAF-mutant melanoma cells. The resulting RNA-RNA hybrid molecules provide a comprehensive and unbiased snapshot of direct miRNA-target interactions. The regulatory effects on selected miRNA target genes in predicted vs. non-predicted binding regions was validated by miRNA mimic experiments. Most miRNA-target interactions deviate from the central dogma of miRNA targeting up to 60% interactions occur via non-canonical seed pairing with a strong contribution of the 3 ' miRNA sequence, and over 50% display a clear bias towards the coding sequence of mRNAs. miRNAs targeting the coding sequence can directly reduce gene expression (miR-34a/CD68), while the majority of non-canonical miRNA interactions appear to have roles beyond target gene suppression (miR-100/AXL). Additionally, non-mRNA targets of miRNAs (lncRNAs) whose interactions mainly occur via non-canonical binding were identified in melanoma. This first application of CLASH sequencing to cancer cells identified over 8 K distinct miRNA-target interactions in melanoma cells. Our data highlight the importance non-canonical interactions, revealing further layers of complexity of post-transcriptional gene regulation in melanoma, thus expanding the pool of miRNA-target interactions, which have so far been omitted in the cancer field.