Final Report Summary - EPINORC (Epigenetic Disruption of Non-Coding RNAs in Human Cancer)
Most of the currently developed cancer research has been protein-coding gene-centered and it has paid insufficient attention to alterations in other type of DNA sequences. In this regard, many efforts have been devoted to messenger RNA, but non-coding RNAs are also an essential player in tumorigenesis. The main goal of this project was to investigate the epigenetic alterations that noncoding genes suffer in cancer. In this regard, in addition to the well recognized epigenetic defects in miRNA genes, be it at the level of DNA methylation (Lujambio et al, Cancer Res 2007, Lujambio et al, PNAS 2008) or histone modification (Iliou et al., Epigenetics 2012), our recent studies have also shown that epigenetic silencing by CpG island hypermethylation of other classes of non-coding RNAs, such as transcribed ultraconserved regions (T-UCRs), small nucleolar RNAs (snoRNAs) or PIWI-interacting RNAs (piRNAs) also occur in human neoplasia. Following our original observation of the transcriptional inactivation of T-UCR in malignancies (Lujambio et al, Oncogene 2010), we have been working in deciphering the activity of these poorly studied transcripts, and we have recently identified that they can play an important role in the regulation of microRNA biogenesis (Liz et al, Molecular Cell 2014). Related to snoRNAs, we have performed a candidate gene approach to identify those that are targeted by cancer-specific gain of DNA methylation and identified three candidate snoRNAs commonly inactivated in transformed cultured cells and primary tumors (Ferreira et al., RNA Biology 2012). piRNAs represent another great enigma in the ncRNA community. Because a number of studies have clearly suggested a role for these small RNAs in male germ cell development, we have wondered about their putative disruption in human pathologies that target the male reproductive system. Herein we have found that male infertility disorders of non-genetic cause (Krausz et al, PloS ONE 2012; Heyn et al., PLoS ONE 2012) and testicular tumors (Ferreira et al., Epigenetics 2013) share the common defect of the DNA methylation-associated silencing of the PIWI proteins in association with downstream downregulation of piRNAs and hypomethylation of LINE1 target sequences. We have also addressed the epigenetic changes occurring at long noncoding genes; in this case, the emerging picture shows a great level of complexity, as expected. In addition to lncRNAs undergoing CpG island methylation-linked loss that we are starting to characterize, histone modifications also exert an important role. For example we have recently shown that the histone methyltransferase EZH2 binds to an antisense lncRNA that it is transcribed within the genomic locus of another histone methyltransferase, SMYD3 (Guil et al., Nat Struct Mol Biol 2012). This newly identified lncRNA behaves as a tumor suppressor. lncRNAs also became aberrantly expressed upon disruption of proteins that bind methylated cytosine, such as MeCP2 (Petazzi et al., RNA Biology 2013). Another antisense lncRNA recently characterized in the laboratory can act as an activator of transcription of its host gene (the epithelial-mesenchymal marker vimentin) through formation of a hybrid DNA:RNA structure called R loop (Boque-Sastre et al, PNAS 2015). All these findings, together with work from many other groups, have uncovered a mesmerizing network of RNAs disrupted in cancer whose intricacies we are just beginning to understand (Guil and Esteller, Nat Struct Mol Biol 2012; Guil and Esteller, TiBS 2015).