Cross-Linking Ligation and Sequencing of Hybrids (qCLASH) Reveals an Unpredicted miRNA Targetome in Melanoma Cells

Abstract

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.

Sequencing of Argonaute-bound miRNA/mRNA hybrids reveals regulation of the unfolded protein response by microRNA-320a

“…Bioinformatic predictions based on seed binding of miRNA with mRNA 3′ UTR have expanded our understanding of miRNAs' functions. However, recent studies using CLASH have identified significant miRNA binding events that do not require perfect seed base-pairing as well as miRNA binding to regions other than the 3′ UTR (33)(34)(35)(36)80). This study identified seven high-confidence non-canonical miR-320a targets including eleven different target sites in both the 3′ UTR and the CDS.…”

Sequencing of Argonaute-bound microRNA/mRNA hybrids reveals regulation of the unfolded protein response by microRNA-320a

“…Bioinformatic predictions based on seed binding of miRNA with mRNA 3′ UTR have expanded our understanding of miRNAs’ functions. However, recent studies using CLASH have identified significant miRNA binding events that do not require perfect seed base-pairing as well as miRNA binding to regions other than the 3′ UTR [ 34 – 37 , 82 ]. This study identified seven high-confidence non-canonical miR-320a targets including eleven different target sites in both the 3′ UTR and the CDS ( S7B Fig ).…”

A noncanonical microRNA derived from the snaR-A noncoding RNA targets a metastasis inhibitor

“…As opposed to traditional CLIP methods in which sequenced miRNAs and targets must then be analytically assigned, the hybrid reads generated by CLASH allow unambiguous pairing between miRNAs and their in vivo targets. While these methods have been effectively used for miRNA target identification in several studies ( Helwak et al 2013 ; Helwak and Tollervey 2014 ; Gay et al 2018 ; Sethuraman et al 2018 ; Bullard et al 2019 ; Ungerleider et al 2020 ; Kozar et al 2021 ; CJ Fields and M Xie, in prep. ), we were interested in determining whether CLASH data could be used for identification of previously unknown miRNA sequences, and specifically for identification of new miRNAs arising via noncanonical biogenesis pathways.…”