DNA methylation linked to all-cause mortality in older people: An epigenome-wide association study

Lund, J. (Speaker)

Activity: Talks and presentationsConference presentations

Description

Abstract: The risk of death can be affected by demographic, epidemiological and clinical factors. At the molecular level, epigenetic mechanisms have been shown to be influential. Using genome-wide DNA methylation data from large-scale cohort studies of older Scottish individuals, we performed an epigenome-wide association study to identify DNA methylation sites related to all-cause mortality and verified these using independent data on Danish twins. By fitting the Cox proportional hazard models, we identified 2,552 CpG sites (linked to 2,036 genes) with a genome-wide significance of FDR < 0.05. Among them, 1,403 positively and 1,149 negatively correlated with mortality. A total of 57 CpGs and 330 genes were replicated in Danish twin cohorts. In addition, we performed gene set enrichment analyses for genes linked to significant CpGs and found 41 and 49 over-represented gene-sets for CpGs showing positive and negative correlation with mortality respectively, verified in Danish twins. These gene-sets are dominated by pathways related to cell-cell signaling, extracellular matrix, neurological functioning, and G protein-coupled receptors. In addition, we performed de-novo pathway enrichment analysis based on the verified genes and extracted the top 20 largest pathways found. Aggregating the top genes in the enriched pathways, a total of 118 unique genes were found, among these, 26 (22%) linked to neuronal, 21 (17.8%) linked to muscle skeleton, and 11 (9.3%) linked to cancerous disorders. Our results provide further evidence of the association between epigenetic factors and   the risk of death in the older populations.
Background and Aim:The risk of death can be affected by demographic, epidemiological and clinical factors. At the molecular level, epigenetic mechanisms have been shown to be influential. Therefore, we seek to investigate the evidence of DNA methylation effects with respect to mortality, using DNA methylations levels extracted from whole blood samples.Design and Methods:Using genome-wide DNA methylation data from large-scale cohort studies of older Scottish individuals, we performed an epigenome-wide association study to identify DNA methylation sites related to all-cause mortality and verified these using independent data on Danish twins. For this, we fit the Cox proportional hazard model to find likely influential causes of increased and decrease in mortality within the segment of older individuals.
Primary variables:Whole blood DNA methylation from Illumina 450K Methylation chips, gene-set enrichment, de-novo pathway enrichment, large-scale cohorts.
Preliminary results:By fitting the Cox proportional hazard models, we identified 2,552 CpG sites (linked to 2,036 genes) with a genome-wide significance of FDR < 0.05. Among them, 1,403 positively and 1,149 negatively correlated with mortality. A total of 57 CpGs and 330 genes were replicated in Danish twin cohorts. In addition, we performed gene set enrichment analyses for genes linked to significant CpGs and found 41 and 49 over-represented gene-sets for CpGs showing positive and negative correlation with mortality respectively, verified in Danish twins. In addition, we performed gene set enrichment analyses for genes linked to significant CpGs and found 41 and 49 over-represented gene-sets for CpGs showing positive and negative correlation with mortality respectively, verified in Danish twins. As well as, we performed de-novo pathway enrichment analysis based on the verified genesConclusions:We have performed an epigenetic association study on all-cause mortality based on whole blood samples of LBC cohorts, have confirmed our findings in a similar study on elderly Danish twins and sought further support from a study on age-related DNA methylation patterns. Likewise, gene-sets enriched by significant CpGs displayed high overlap with that from the replication cohort and from previously published studies. Finally, our de-novo pathway analysis revealed gene networks implicated in cancer, anatomical and neurological diseases. Our results provide further evidence of the association between epigenetic factors and the risk of death in the older populations.Presentation language:English
Period23. Apr 2018
Event typeConference
LocationFredericia, Denmark
Degree of RecognitionRegional