Recent research has identified critical roles for microRNAs in a large number of cellular processes, including tumorigenic transformation. While significant progress has been made towards understanding the mechanisms of gene regulation by microRNAs, much less is known about factors affecting the expression of these noncoding transcripts. Here, we demonstrate for the first time a functional link between hypoxia, a welldocumented tumor microenvironment factor, and microRNA expression. Microarray-based expression profiles revealed that a specific spectrum of microRNAs (including miR-23, -24, -26, -27, -103, -107, -181, -210, and -213) is induced in response to low oxygen, at least some via a hypoxia-inducible-factor-dependent mechanism. Select members of this group (miR-26, -107, and -210) decrease proapoptotic signaling in a hypoxic environment, suggesting an impact of these transcripts on tumor formation. Interestingly, the vast majority of hypoxiainduced microRNAs are also overexpressed in a variety of human tumors.

The functional roles of SNPs within the 8q24 gene desert in the cancer phenotype are not yet well understood. Here, we report that CCAT2, a novel long noncoding RNA transcript (lncRNA) encompassing the rs6983267 SNP, is highly overexpressed in microsatellite-stable colorectal cancer and promotes tumor growth, metastasis, and chromosomal instability. We demonstrate that MYC, miR-17-5p, and miR-20a are up-regulated by CCAT2 through TCF7L2-mediated transcriptional regulation. We further identify the physical interaction between CCAT2 and TCF7L2 resulting in an enhancement of WNT signaling activity. We show that CCAT2 is itself a WNT downstream target, which suggests the existence of a feedback loop. Finally, we demonstrate that the SNP status affects CCAT2 expression and the risk allele G produces more CCAT2 transcript. Our results support a new mechanism of MYC and WNT regulation by the novel lncRNA CCAT2 in colorectal cancer pathogenesis, and provide an alternative explanation of the SNP-conferred cancer risk.[Supplemental material is available for this article.]Notwithstanding the considerable advancements in our understanding of the molecular genetic basis of cancer, in the majority of cancer-associated genomic regions, the responsible protein-coding genes have not been identified yet. The discovery of short (19-22 nt), noncoding RNAs (ncRNAs)-called microRNAs (miRNAs) (Ambros 2001)-not only revealed a novel mechanism of gene regulation but also led to the identification of miRNAs directly involved in cancer development (Spizzo et al. 2009). It is therefore plausible that as-yet-unidentified members of the broader category of ncRNA mapping to cancer-associated genomic regions play ratelimiting roles in tumor initiation and/or progression (Rinn and Chang 2012). For instance, we previously reported that highly conserved genomic regions (ultraconserved regions, or UCRs) (Bejerano et al. 2004) are frequently transcribed as long (>200 bp) ncRNAs (lncRNAs) in both normal and tumor tissues (Calin et al. 2007). Furthermore, germline mutations, as well as single nucleotide polymorphisms (SNPs) in ultraconserved ncRNAs, were found to occur more frequently in patients with colon cancer and chronic leukemia than in the general population (Wojcik et al. 2010).The rs6983267 SNP, mapping to the 8q24.21 chromosomal region, has been consistently associated with an increased risk of colorectal cancer (CRC) (Haiman et al. 2007): The G allele was associated with greater predisposition to CRC than the T allele (odds ratios of 1.27 and 1.47 for heterozygotes and homozygotes, respectively; P = 1.27 3 10 À14 ) (Tomlinson et al. 2007). The increased cancer risk from this SNP variant was also observed in other cancer types, including prostate, ovarian, and inflammatory breast cancer (Ghoussaini et al. 2008;Bertucci et al. 2012). Despite the consistent association between rs6983267 and cancer risk, the underlying molecular and cellular mechanisms remain largely unknown. The genomic region spanning rs6983267 was found to contain DNA (Pom...

Motivation Deciphering the language of non-coding DNA is one of the fundamental problems in genome research. Gene regulatory code is highly complex due to the existence of polysemy and distant semantic relationship, which previous informatics methods often fail to capture especially in data-scarce scenarios. Results To address this challenge, we developed a novel pre-trained bidirectional encoder represen-tation, named DNABERT, to capture global and transferrable understanding of genomic DNA sequences based on up and downstream nucleotide contexts. We compared DNABERT to the most widely used programs for genome-wide regulatory elements prediction and demonstrate its ease of use, accuracy, and efficiency. We show that the single pre-trained transformers model can simultaneously achieve state-of-the-art performance on prediction of promoters, splice sites, and transcription factor binding sites, after easy fine-tuning using small task-specific labelled data. Further, DNABERT enables direct visualization of nucleotide-level importance and semantic relationship within input sequences for better interpretability and accurate identification of conserved sequence motifs and functional genetic variant candidates. Finally, we demonstrate that pre-trained DNABERT with human genome can even be readily applied to other organisms with exceptional performance. We anticipate that the pre-trained DNABERT model can be fined tuned to many other sequence analyses tasks. Availability The source code, pretrained and finetuned model for DNABERT are available at GitHub https://github.com/jerryji1993/DNABERT Supplementary information Supplementary data are available at Bioinformatics online.

SUMMARY Adenosine deaminases acting on RNA (ADARs) are involved in RNA editing that converts adenosine residues to inosine specifically in double-stranded RNAs. In this study, we investigated the interaction of the RNA editing mechanism with the RNA interference (RNAi) machinery and found that ADAR1 forms a complex with Dicer through direct protein-protein interaction. Most importantly, ADAR1 increases the maximum rate (Vmax) of pre-microRNA (miRNA) cleavage by Dicer and facilitates loading of miRNA onto RNA-induced silencing complexes, identifying a new role of ADAR1 in miRNA processing and RNAi mechanisms. ADAR1 differentiates its functions in RNA editing and RNAi by formation of either ADAR1/ADAR1 homodimer or Dicer/ADAR1 heterodimer complexes, respectively. As expected, expression of miRNAs is globally inhibited in ADAR1−/− mouse embryos, which in turn alters expression of their target genes and might contribute to their embryonic lethal phenotype.

Unrestrained E2F activity forces S phase entry and promotes apoptosis through p53-dependent and -independent mechanisms. Here, we show that deregulation of E2F by adenovirus E1A, loss of Rb or enforced E2F-1 expression results in the accumulation of caspase proenzymes through a direct transcriptional mechanism. Increased caspase levels seem to potentiate cell death in the presence of p53-generated signals that trigger caspase activation. Our results demonstrate that mitogenic oncogenes engage a tumour suppressor network that functions at multiple levels to efficiently induce cell death. The data also underscore how cell cycle progression can be coupled to the apoptotic machinery.