The emergence of Next Generation Sequencing (NGS) along with computational biology has broadened the scope in which diverse cellular processes can be interrogated. While there has been considerable progress in understanding the impact of genetic and epigenetic mechanisms in tumorigenesis using whole genomic, epigenomic and transcriptional analysis by NGS, there has been little consideration of the importance of interplay between these processes. We performed a comparative analysis of array and NGS technologies to identify differentially methylated CpG sites in colorectal cancer cell lines. NGS had more specificity in addition to profiling more CpG sites relative to Illumina 450K arrays.
Base-level resolution of sequencing data can identify any strand specific methylation bias. Our analysis shows that methylation frequency between the sense and antisense strand are highly correlated (average R2 ∼ 0.81), and coefficient of variance (CV) between the strands is generally low (about half of observed sites have <10% CV). However, a small percentage of bases had strand specific biases. Using a minimum of 100% CV and difference in methylation frequency greater than 50% as filtering criteria, we found 1210, 569, 638, and 1484 CpG sites have strand specific biases with 7 overlapping bases among the samples tested. Further investigation will identify whether these bases are random or reside within a particular region, where biases occur, and what genes are potentially affected.
We also used NGS and publically available gene expression datasets in colorectal cancer cell lines-HCT116 and HCT116 DKO (cell line with genetic knockouts of both DNA methyltransferases DNMT1 and DMNT3b) to identify roles of differential methylation in regulating gene expression. A majority of genes were down-regulated between HCT116 and HCT116 DKO cell lines including those involved in chromatin, nucleic acid, and nucleotide binding and cell cycle regulation. Interestingly, many differentially expressed genes are also involved in immune response. We then used bisulfite treated genomic data to evaluate genetic regulation of gene expression. For this, we converted bisulfite treated data into genomic space using custom in-house developed bio-informatics tools that were first tested using DNA isolated from NA12878 cell line. Our analysis showed that 65% of the known variants detected in NA12878 cell line by the genome in a bottle consortium can be identified by bisulfite sequencing of promoter associated CpG islands. One limitation of this analysis is the inability to identify C>T genomic variants. This data is being analyzed to evaluate effects of genetic mutations in promoter binding sites on gene expression in colorectal cancers. Comparative analysis of genetic and epigenetic regulation of gene expression will allow better understanding of gene regulatory networks in colorectal cancer.
Citation Format: Claire Olson, Fang Yin Lo, Kerry Deutsch, Sharon Austin, Kellie Howard, Amanda Leonti, Lindsey Maassel, Christopher Subia, Tuuli Saloranta, Nicole Christopherson, Kathryn Shiji, Shradha Patil, Steven Anderson, Anup Madan. Synergistic effects of promoter associated DNA methylation and genetic alterations to better understand oncogenic gene expression profiles. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4436.