Background: Aberrant methylation of DNA acts epigenetically to skew the gene transcription rate up or down. In this study, we have developed a comprehensive computational framework for the stage-specific analysis of methylation patterns in colorectal cancer.
Methods: Combining public-domain methylation and clinical data from TCGA, the methylation β-matrix was converted into M-value matrix, annotated with sample stages and analysed for stage-specific genes using multiple approaches involving stage-differentiated linear modelling of methylation patterns or their correlation with the phenotype and expression patterns. Differentially methylated genes (DMGs) were identified using a contrast against control samples (adjusted p-value <0.001 and |log fold-change of M-value| >2). These results were further filtered using a series of all possible pairwise stage contrasts (p-value <0.05) to obtain stage-specific DMGs. These were then subjected to a consensus analysis, followed by Kaplan-Meier survival analysis to explore the relationship between methylation and prognosis for the consensus stage-salient biomarkers.
Results: We found significant genome-wide changes in methylation patterns in cancer samples relative to controls agnostic of stage. Our stage-differentiated analysis yielded the following stage-salient genes: one stage-I gene (FBN1), one stage II specific gene (FOXG1), one stage III specific gene (HCN1) and four stage IV specific genes (NELL1, ZNF135, FAM123A, LAMA1), which could be used for stage-specific diagnosis. All the biomarkers were hypermethylated, indicating down-regulation and signifying a CpG island Methylator Phenotype (CIMP) manifestation. A prognostic signature consisting of FBN1 and FOXG1was significantly associated with patient survival (p-value < 0.01) and could be used as a biomarker panel for early-stage CRC prognosis.
Conclusion: Our workflow for stage-differentiated consensus analysis has yielded stage-salient diagnostic biomarkers as well as an early-stage prognostic biomarker panel. In addition, our studies have affirmed a novel CIMP-like signature in colorectal cancer, urging clinical validation.