In a new publication Gadd et al. apply blood-derived epigenetic signatures of lifestyle traits to matched blood and brain samples, uncovering variability in how well blood translates across brain regions and a relationship between smoking and the prefrontal cortex. These preliminary results contribute to understanding how lifestyle-related DNA methylation affects the brain.
These factors have been extensively linked with blood-based genome-wide DNA methylation, but it is unclear if the signatures from blood translate to the target tissue of interest-the brain. To investigate this, the authors apply blood-derived epigenetic predictors of four lifestyle traits to genome-wide DNA methylation from five post-mortem brain regions and the last blood sample prior to death in 14 individuals in the Lothian Birth Cohort 1936 (LBC1936).
This cohorte is a longitudinal study of healthy ageing in individuals who reside in Scotland. Participants completed a childhood intelligence test at age 11 years in 1947 and were then recruited for this cohort at the mean age of 70 years. They have been followed up approximately every 3–4 years (currently at the sixth wave), collecting a series of cognitive, physical, clinical and social data, along with blood donations that have been used for genetic, epigenetic, and proteomic measurement. Approximately 15% of individuals in the LBC1936 have consented to post-mortem tissue collection.
The LBC1936 group were born in 1936 and there are socioeconomic and cultural trends in their lifestyles that must be taken into consideration; many may have worked in factories or in shipbuilding yards, exposing them to high levels of respiratory pollutants and poorer socioeconomic status in this era was related to behaviours such as smoking. Though Scotland has a historic and sustained high prevalence of unhealthy lifestyle behaviours such as smoking and alcohol consumption, with a large proportion of the current population either overweight or obese, cohort profiling has found both the Generation Scotland and Lothian Birth Cohorts to be healthier and of higher socioeconomic status than the general population.
The majority of the 14 individuals within the brain bank subset either had been or were still smokers (86%) at the time of death, which was a higher proportion than in the reference group (46%). Most of the 14 individuals (86%) had high HDL cholesterol (> 1 mmol/l), drank alcohol weekly (92%) and had mean BMI of 25.5 kg/m2. Five of the individuals did not smoke.
Using these matched samples, the authors found that correlations between blood and brain DNA methylation scores for smoking, high-density lipoprotein cholesterol, alcohol and body mass index were highly variable across brain regions.
Smoking scores in the dorsolateral prefrontal cortex had the strongest correlations with smoking scores in blood and smoking behaviour. This was also the brain region which exhibited the largest correlations for DNA methylation at AhRR/site cg05575921 - which is the single strongest correlate of smoking in blood-in relation to blood and smoking behaviour. The aryl hydrocarbon receptor (AhRR) gene may act as a tumor suppressor but it is known that Dioxins and dioxin-like compounds are teratogens that exert their effects through the AhRR. This suggested a particular vulnerability to smoking-related differential methylation in this region.
The authors' work contributes to understanding how lifestyle factors affect the brain and suggest that lifestyle-related DNA methylation is likely to be both brain region dependent and in many cases poorly proxied for by blood. Though these pilot data provide a rarely-available opportunity for the comparison of methylation patterns across multiple brain regions and the blood, due to the limited sample size available authors' results must be considered as preliminary and should therefore be used as a basis for further investigation.
All R code used in this study is available with open access at the following Gitlab repository: https://gitlab.com/dannigadd/blood-brain-lifestyle-traits.
