cis ‐Regulatory Driver Mutations in Cancer Genomes

Abstract

The increasing availability of large‐scale whole cancer genome sequencing data sets has enabled research efforts to focus on cancer driver mutations within the noncoding genome. cis ‐Regulatory elements are sites responsible for the control of gene expression. Mutations at such loci may modify transcription factor binding sites, disrupt enhancer–promoter interactions or affect epigenetic marks. cis ‐Regulatory driver mutations have been found in some cancer types to‐date, with initial discoveries of point mutations in the TERT promoter, small insertions creating an enhancer regulating TAL1 and genomic rearrangements leading to simultaneous enhancer dysregulation of EVI1 and GATA2 . However, recent discoveries have also revealed mechanisms responsible for the accumulation of recurrent, and even functional, somatic single‐nucleotide mutations in regulatory regions, wholly apart from selective pressure. Close scrutiny of candidate cis ‐regulatory mutations is necessary in order to effectively identify mutations that have undergone positive selection and to accurately distinguish driver from passenger mutations. Key Concepts Improvements in DNA sequencing technology, and the commencement of large‐scale cancer genome sequencing projects, have made discoveries in the noncoding genome possible. cis ‐Regulatory elements, such as promoters and enhancers, are responsible for the control of gene expression. Transcription factors bind to cis ‐regulatory elements, influencing the rate of transcription of DNA to mRNA. Mutations in cis ‐regulatory elements can drive cancer development by altering transcription factor binding sites, enhancer–promoter interactions or the epigenetic landscape of the cell. Alterations to cis ‐regulatory elements can lead to gene dysregulation, potentially impacting the expression of oncogenes or tumour suppressor genes. Recurrent somatic driver mutations affecting cis ‐regulatory elements have already been identified in cancer genomes, including point mutations, insertions and deletions (indels) and structural rearrangements. cis ‐Regulatory elements may appear recurrently mutated in cancer due to selective pressures or due to other mechanisms driving the formation of mutation hot spots. Transcription factor binding can inhibit access to nucleotide excision repair (NER) enzymes, leading to increased mutation rates at transcription factor binding sites in some cancers. DNA modifications and nucleotide composition can both influence the propensity for certain sites to become mutated. Close scrutiny of candidate cis ‐regulatory mutations is necessary in order to effectively identify sites under selective pressure and to accurately distinguish driver from passenger mutations.

In search of non-coding driver mutations by deep sequencing of regulatory elements in colorectal cancer

“…Assigning function to a non-coding variant can be imprecise due to the variety of ways in which a variant can impact upon gene regulation (1), which can be difficult to capture via a single measure. Hence, in addition to our analyses of functional enrichment in genomic regions via OncodriveFML (28), we also considered base pair recurrence of somatic variants in our cohort.…”

“…For example, there are a plethora of ways in which a non-coding mutation may impact genome function. For example, a mutation may alter a transcription factor binding site, affect the partitioning of the genome into topologically-associating domains, or cause epigenetic changes by altering the binding of pioneer factors, nucleosome positioning, chromatin organisation or CpG methylation (1). In this study, we have proposed a list of single nucleotide variants and genomic windows containing recurrent indels, which may be functional mutations in the non-coding genome.…”

“…Such genes have been classified as cancer drivers because they harbour frequent high-impact somatic coding mutations in cancer genomes, with these mutations conferring a selective advantage to cells in certain tissues-types and resulting in oncogenesis. Identifying cancer driver mutations outside of protein-coding elements however, has proven to be a complex task as it can be difficult to assign function to some non-coding mutations (1). Despite a number of large-scale studies aimed at prioritising either recurrent or functional mutations (2-4), relatively few somatic driver mutations have yet been discovered in the noncoding genome.…”

Machine learning for multi-omics data integration in cancer

“…Outside of coding regions, genomic information can be used to understand elements of gene regulation and dysregulation in cancer ( Andersson and Sandelin, 2020 ). Somatic, germline, and epigenetic variation affecting these regions can have profound effects on gene expression in cancer ( Poulos and Wong, 2017 ). Although genomic technology is relatively more mature than other omics, discovering causal relationships in addition to associations remains as one of the biggest challenges ( McGuire et al., 2020 ).…”