Abstract

Simple Summary

   Aberrant epigenetic modifications in oncogenic pathways can lead to the
   onset of different cancers. This study aims to explore the role of
   differential DNA methylation in the regulation of oncogenic signaling
   pathways by integrating data from multiple sources including methylome,
   transcriptome and clinical presentation to uncover the effect of
   methylation changes acting on the four most common cancers. We utilized
   a differential methylation-parsing pipeline, which extracted
   differentially methylated biomarkers based on feature selection.
   Extracted biomarkers were integrated with the matching RNA-Seq and
   clinical data to determine if these differentially methylated CpGs
   could serve as potential diagnostic candidates for the four most common
   cancers. Our results suggested differential methylation of the genes
   within the NRF2-PI3K pathway may lead to the presentation of various
   cancer and serve as potential epigenetic biomodifiers.

Abstract

   Disruption of signaling pathways that plays a role in the normal
   development and cellular homeostasis may lead to the dysregulation of
   cellular signaling and bring about the onset of different diseases,
   including cancer. In addition to genetic aberrations, DNA methylation
   also acts as an epigenetic modifier to drive the onset and progression
   of cancer by mediating the reversible transcription of related genes.
   Although the role of DNA methylation as an alternative driver of
   carcinogenesis has been well-established, the global effects of DNA
   methylation on oncogenic signaling pathways and the presentation of
   cancer is only emerging. In this article, we introduced a differential
   methylation parsing pipeline (MethylMine) which mined for epigenetic
   biomarkers based on feature selection. This pipeline was used to mine
   for biomarkers, which presented a substantial difference in methylation
   between the tumor and the matching normal tissue samples. Combined with
   the Data Integration Analysis for Biomarker discovery (DIABLO)
   framework for machine learning and multi-omic analysis, we revisited
   the TCGA DNA methylation and RNA-Seq datasets for breast, colorectal,
   lung, and prostate cancer, and identified differentially methylated
   genes within the NRF2-KEAP1/PI3K oncogenic pathway, which regulates the
   expression of cytoprotective genes, that serve as potential therapeutic
   targets to treat different cancers.

   Keywords: DNA methylation, machine learning, cancer biomarkers, gene
   expression, p53, NRF2, Wnt, Hippo, signaling pathways

1. Introduction

   Over the past few decades, the field of cancer research has
   continuously evolved [[30]1]. Different techniques have been applied in
   the detection and treatment of cancer, including screening for cancers
   during early stages of disease, in order to identify different cancers
   before the onset of symptoms [[31]2,[32]3,[33]4]. Furthermore, new
   strategies to predict the outcome of cancer treatments has also emerged
   [[34]5,[35]6,[36]7]. With the advent of new technologies in the field
   of oncology, large amounts of cancer data have collected (including
   genomic, transcriptomic and epigenomic data). However, the accurate
   prediction of disease outcome and differentiation between different
   cancers remains one of the most interesting and challenging tasks for
   researchers. Machine-learning (ML) methods have become a popular tool
   in recent years to assist medical researchers to discover and identify
   patterns and relationships between different cancers, in order to
   create more effective biomarker panels which will be able to predict
   potential biomarkers that are present in different cancers
   [[37]8,[38]9,[39]10,[40]11,[41]12]. ML defines the ability of machines
   to learn and predict future events and outcomes based on large
   datasets, and has been utilized in health care to improve the
   interpretation of medical data especially in the development of novel
   computational tools for stratification, grading, and prognostication of
   patients with the goal of improving patient care [[42]8,[43]13,[44]14].

   Genomic data has been associated with rich clinical annotations, and
   DNA sequencing has now been incorporated into standard clinical
   practice [[45]15]. An emerging trend in application of machine learning
   in oncology is the integration and analysis of mixed data (e.g.,
   genomic, transcriptomic, and epigenomics)—multi-omics to uncover
   patterns that are reflected across the different data. These studies
   have painted a portrait of the mutation landscape of multiple cancers
   including breast, colorectal, lung and prostate
   [[46]16,[47]17,[48]18,[49]19]. However, some cancers may have
   relatively few genetic mutations, with their biology largely driven by
   other types of variations such as aberrant epigenetic modifications.

   Epigenetics is the study of heritable phenotypic changes that do not
   involve alteration in the genomic sequence, and includes mechanisms
   such as histone modification, altered micro (mi) RNA, or long
   non-coding (lnc) RNA [[50]20]. Mediated by DNA methyltransferases, DNA
   methylation is a major epigenetic mechanism that occurs when a methyl
   group is transferred to the C5 position of the carbon to form
   5-methylcytosine (5mC). This modification plays an essential role in
   various biological processes including the regulation of gene
   expression, genomic imprinting, cell differentiation and normal
   development [[51]21,[52]22,[53]23]. Aberrant DNA methylation has been
   associated with various diseases, including cancers
   [[54]24,[55]25,[56]26]. DNA hypermethylation acts as a regulator of
   gene expression, and methylation-mediated silencing of tumor suppressor
   genes or regulatory regions within the genome can lead to dysregulation
   of cell growth or altered response to cancer therapies
   [[57]27,[58]28,[59]29,[60]30]. Despite the varied and complex nature of
   modifications in the epigenetic landscape, many cancers display a high
   degree of similarity both across different tissues and within the
   tissues of origin [[61]31,[62]32]. Therefore, abnormal DNA methylation
   has been viewed as an attractive avenue for development of cancer
   diagnostics and therapeutics.

   Although thousands of studies have highlighted the value of using
   changes in DNA methylation as candidate biomarkers in the detection and
   treatment of various cancers, only a handful of these targets have been
   approved for clinical use (as summarized in [63]Table 1).
   Traditionally, these studies have focused on the association of
   differential methylation and its effects on individual genes in the
   phenotypic presentation of different carcinogenesis. However, few
   studies have investigated the effects of epigenetic changes (especially
   DNA methylation) on individual and complex signaling pathways across
   cancers at different anatomical sites. Cell signaling (signal
   transduction) plays a vital role within biological systems by relaying
   extracellular signals in order to regulate intracellular gene
   expression. The signal transduction process is typically initiated when
   a ligand binds to a membrane-bound receptor, triggering a cascade of
   intercellular signaling activities through activation of multiple
   kinases. This ultimately has an effect on the downstream gene
   expression activation of different signaling pathways and can lead to
   various physiological or cellular responses (e.g., cell proliferation,
   differentiation, and metabolism), and any disruptions within these
   intra and/or extracellular communication chains can lead to the
   development of different diseases, including cancer
   [[64]33,[65]34,[66]35]. As DNA methylation is a dynamic and reversible
   process, changes in methylation could act as regulators of certain
   oncogenic pathways, leading to development of diseases, and provide an
   attractive target for development of biomarkers and therapies.

Table 1.

   Summary of DNA methylation biomarkers in five common cancers that is
   approved for clinical use.
   Cancer Methylated Gene Description References