Pre-diagnosis blood DNA methylation profiling of twin pairs discordant for breast cancer points to the importance of environmental risk

Abstract

AbstractBackgroundBreast cancer risk assessment typically relies on consideration of mammography, family history, reproductive history, in addition to assessment of major mutations. However, quantifying the impact of environmental factors, such as lifestyle, can be challenging. DNA methylation (DNAm) presents a promising opportunity for additional information, as it captures effects from both genetics and environment. Previous studies have identified associations and predicted the risk of breast cancer using DNAm in blood, however, these studies did not distinguish between genetic and environmental influences.ResultsPre-diagnosis blood samples were obtained from 32 monozygotic (MZ)and 76 dizygotic (DZ) female twin pairs discordant for breast cancer. DNAm was profiled using Illumina 450k and EPIC platform. Samples were collected at the mean age of 56 years (standard deviation (SD) =9.90), while the mean age at diagnosis was 66.76 years (SD=6.64). The mean age for censoring controls was 75.20 years (SD=9.45). To identify individual DNAm sites (Cytosine-phosphate-Guanine, CpG) and differentially methylated regions (DMRs) associated with breast cancer risk, survival analysis using paired Cox proportional hazard modeling was performed. Due to the paired modeling, shared genetic and environmental effects, age at entry, and age at sampling were controlled for.We identified 212 CpGs (p<6.4*10^-8) and 15 DMRs associated with breast cancer risk across all pairs, with three DMRs overlapping with the individual CpGs. All but one of the 212 CpGs had lower DNAm in cases, suggesting a prevailing trend of hypomethylation of blood DNA prior to breast cancer diagnosis. Altogether 197/212 significant CpGs were also differentially methylated within the 32 MZ twin pairs discordant for breast cancer, suggesting that DNAm at these CpGs is likely independent of genetic effects. Prior research suggests that estrogen regulates at least five of the top CpGs identified. Hence, methylation at these sites may reflect individual differences in estrogen exposure.ConclusionIn conclusion, most of the identified CpGs associated with future breast cancer diagnosis appear to be independent of genetic effects. This suggests that DNAm could potentially serve as a valuable biomarker for environmental risk factors for breast cancer and may offer potential benefits as a complementary tool to current risk assessment procedures.