BeskrivelseThe epigenetic modification is essential for controlling normal development and homeostasis and allows organism to integrate and react to environments. The epigenetic changes could be influenced by many factors including age, sex as well the lifestyle. Since both genetic and environmental factors play important roles in complex disease development, in epigenetic studies, it will considerably increase the power by controlling the genetic effect. For this reason, the twin designs have been widely used in epigenetic studies. In theory, using monozygotic (MZ) twins that share genetic makeups in epigenetic studies provides distinctive advantage over traditional case-control design. However, no power assessment has been properly performed on the discordant twin design although it is popular in use. To address this concern, we conducted a computer simulation study to investigate and compare the statistical power of discordant twin design and ordinary case-control design. We used a liability threshold model and simulated both genetic and environmental contributions to disease. In this way, the heritability, which is represented by the genetic contribution to disease, can be put under control. Simulation results suggest that the discordant twin design has higher power estimates when disease is with moderate to high heritability (>30%), and only in cases when the influence of genetic contribution is nominal, the discordant twin design has lower power. The results indicate that the discordant twin design is indeed a powerful tool for epigenetic association studies. After power assessment using computer simulation, we applied the discordant twin design to real twin data on epigenome-wide association analysis of body mass index (BMI). BMI is a simple measurement based on weight and height, and studies have shown that BMI is associated with many traits relevant to metabolism and future heart disease risk, such as cholesterol and blood pressure. It is also often used as a screening tool for obesity which is a common condition that significantly declines quality of life and is closely related to diabetes as well as many cardiovascular diseases. There are many factors that influence body weight including genetic, behavioral and hormonal. Meanwhile, the heritability estimates for BMI are generally higher than 47%, which makes BMI a good candidate for epigenetic association analysis using the discordant twin design. In an effort to look for epigenetic changes associated with BMI in the Chinese population, epigenetic data was collected on 30 pairs of MZ twins with methylation profiles from whole blood samples analyzed using the reduced representation bisulfite sequencing (RRBS) technique. Epigenome-wide association study on this data found 11 differentially methylated regions (DMRs) and several biological pathways associated with BMI. Some identified pathways are related to extracellular matrix (ECM), neuronal system and signaling. Within the DMRs, expression of DPYSL3 has been associated with nonalcoholic fatty liver disease and CWF19L1 has been shown to influence liver fat deposition. By regulating gene expression through transcriptional and epigenetic regulation as well as alternative splicing in the nucleus, the long non-coding RNAs (lncRNAs) are important regulators of the epigenetic status of the human genome. LncRNA expression data on 220 MZ twin pairs from Danish Twin Register was also collected for association with BMI. There are 6 significant pseudogenes related to BMI variation after FDR correction. Causal inference suggests that all 6 pseudogenes are in response to BMI changes. By using GREAT software, we have found that most of the identified GO biological processes are related to kidney development, regulation of lipid biosynthetic process, regulation of circadian rhythm, regulation of Notch signaling and other processes that are related to BMI or obesity. Overall, both assessment and application of the discordant twin design have shown the values of using twins in epigenetic studies and the potential for extension to be applied to other complicated diseases or traits.
|Periode||15. okt. 2016 → 14. okt. 2019|