Systematic investigation of disease-associated, regulatory variation; illuminating their prediction, molecular consequences and mechanisms contributing to risk of Parkinson disease (PD)

Summary – This proposal takes crucial next steps towards illuminating the mechanistic impact of regulatory variation underlying common disease risk and progression, focusing on Parkinson's disease (PD) as a model. The majority of disease-associated variants identified by genome-wide association studies (GWAS) lie in

noncoding DNA, likely influencing transcription of their cognate genes. Thus, improving our understanding of how regulatory variation can impact gene expression, and the downstream cellular mechanisms through which they modulate disease susceptibility, is crucial. The acquisition of biologically relevant genomic data, across the

cellular contexts in which the variants may exert their effects, is imperative for the prioritization and functional assay of variants within associated loci, as well as the determination of their mechanistic impact. Towards this end, we have already made significant strides in studying the chromatin and transcriptional

landscapes of gestational/early postnatal dopaminergic (DA) neurons and improved our understanding of how regulatory variation confers risk for PD. We have previously generated catalogs of open chromatin regions (OCRs) and similarly profiled gene expression of midbrain and forebrain DA neurons. We have developed and

implemented computational classifiers to identify key transcription factors (TFs) that actively influence gene expression and have identified PD-associated functional variation falling within novel enhancers. Although they establish a powerful precedent, these studies query only a snapshot early in normal DA neuron biology.

Here, we propose to link regulatory variants, the cell state(s) in which they act, the genes they influence, and the mechanisms through which they impact PD risk. We aim to define chromatin and transcriptional signatures derived from PD-vulnerable DA neurons over time and in response to PD-relevant insult of α-synuclein preformed

fibrils (Aim 1). Further, we will develop novel tools to computationally “learn” the sequence-basis of the cell type/state dependent OCRs, via machine learning, undertaking massively parallel reporter assays (MPRA) to test thousands of OCRs, and predicted disease risk variation therein, using DA neurons derived from human

induced pluripotent stem cells (hiPSC-DA) from unaffected individuals (Aim 2). In Aim 3, we will test the functional consequences of disrupting predicted key TFs and enhancers on a range of PD-relevant cellular phenotypes using hiPSC-DA neurons. We will similarly evaluate the molecular and cellular effects of risk and non-risk

variation therein, using hiPSC-DA harboring established PD mutations, to provide a greater opportunity of observing functional effects. Our proposal will advance our understanding of regulatory encryption and how noncoding, functional variation perturbs molecular mechanisms in common disease risk and progression,

particularly for PD. Additionally, our findings will inform mechanisms underpinning other PD-related disorders including diffuse Lewy Body Disease and Lewy Body Dementia and the increased risk of cognitive decline/dementia in Gaucher’s.