Methylation profiling by array for Parkinson's disease using brain tissue

Parkinson’s disease is second most prevalent neurodegenerative disorder. Parkinson’s disease is hypothesized to be caused by a multifaceted interplay between genetic and environmental factors. We profiled the DNA methylation using the frontal lobe from people who died from Parkinson’s disease and compared them with age-, and sex-matched controls. We hypothesized that the complex regulation of DNA methylation and metabolome leads to deficits that cause irregular motor function among Parkinson’s disease subjects and the metabolomic disparities may reflect the effect of these complex interactions. Herein, we aim to integrate quantitative metabolomics and DNA methylation of brain tissue from primary motor cortex (Brodmann area 4) from Parkinson’s disease sufferers to better understand the biochemistry associated with the onset of the disease. Parkinson’s disease (PD) is second most prevalent neurodegenerative disorder following Alzheimer’s disease. Parkinson’s disease is hypothesized to be caused by a multifaceted interplay between genetic and environmental factors. Herein, and for the first time, we describe the integration of metabolomics and epigenetics (genome-wide DNA methylation; epimetabolomics) to profile the frontal lobe from people who died from PD and compared them with age-, and sex-matched controls. We identified 48 metabolites to be at significantly different concentrations (FDR q<0.05), 4,313 differentially methylated sites [5'-C-phosphate-G-3' (CpGs)] (FDR q<0.05) and increased DNA methylation age in the primary motor cortex of people who died from PD. We identified Primary bile acid biosynthesis as the major biochemical pathway to be perturbed in the frontal lobe of PD sufferers, and the metabolite taurine (p-value = 5.91E-06) as being positively correlated with CpG cg14286187 (SLC25A27; CYP39A1) (FDR q = 0.002), highlighting previously unreported biochemical changes associated with PD pathogenesis. In this novel multi-omics study, we identify regulatory mechanisms which we believe warrant future translational investigation and central biomarkers of PD which require further validation in more accessible biomatrices. A total of 78 brain tissue samples were acquired from the NIH NeuroBioBank (40 brain tissue samples were from PD cases and 38 were from cognitively healthy subjects). The frozen brain tissues were obtained from the primary motor cortex (Brodmann area 4) and processed using Infinium MethylationEPIC Array following the manufacturer’s instructions (Illumina Infinium HD Methylation Assay Experienced User Card, Automated Protocol 15019521 v01). Raw signal intensity data were preprocessed from raw *.idat files using minfi Bioconductor package. The samples were normalized using Noob normalization. Methylation beta values were extracted for the remaining probes. Sample position effect on the EPIC array was adjusted using empiricalBayesLM function from WGCNA package. The effect of sample position was modeled as polynomial of second degree. Proportion of neuronal cells in each sample was then estimated using flow sorted PFC samples and estimateCellCount function from minfi package.