A Multi-omics approach to Environment and Depression in Parkinsons disease (MOOD-PD)

项目摘要 我们建议使用代谢和多组学标记来阐明慢性，长期农药暴露如何影响 生活在农村环境中的老年人和帕金森病患者中抑郁和焦虑的发生率 (PD)和抑郁症抑郁症和焦虑症是老年人最常见的精神障碍类型。 在长期护理环境中，抑郁症患病率高达25%，这对共病患者的医疗保健使用有很大贡献。 疾病1 -5。PD患者中的重度抑郁发作和临床焦虑发生率要高得多， PD诊断1，2.提高我们对抑郁症和焦虑症中可改变的风险因素和机制的理解 是一个迫切的公共卫生问题。在过去的二十年里，我们的团队已经产生了一个独特的数据， 资源，现在为我们提供了机会，调查共同的和重要的贡献， 环境暴露（农药）抑郁和焦虑，并探索疾病的过程， 多维生物网络在先前的资助下，我们收集了终身抑郁/焦虑诊断， 治疗史以及大规模人群中抑郁和焦虑症状的当前（和随访）状态- 在加州中央谷的居民中进行了一项基于病例对照的PD研究，包括500多名PD患者， 近十年来，在这些人中，多达38%的PD患者和27%的老年人没有 PD在其生命中的任何时间报告了抑郁症或焦虑症的诊断。对于所有研究参与者， 基因组（Illumina Global Screening Array，660 K标记）数据可用，我们还生成了约800个标记 表观基因组（Illumina Infinium 450 K DNA平台;全基因组DNA甲基化）和非靶向代谢组学数据（在 300名参与者的两个时间点）。我们开发了一个纵向的地理信息系统（GIS）为基础的评估 对于农药暴露，将国家规定的关于所有农业农药的类型、日期和地点的信息联系起来， 自1974年以来，加州记录的土地使用地图和研究参与者的住所和工作场所的应用程序。 此外，我们还收集了有关职业、家庭和园艺使用杀虫剂的广泛信息。生物 过程，包括对慢性毒物暴露或疾病过程的生物（代谢、表观遗传）反应， 动态的，但也依赖于遗传易感性。在这里，我们建议联合收割机结合一个强大的系统生物学分析， 询问遗传学中锚定的表观遗传和代谢组学数据以识别特定农药特征的方法 暴露（有机磷、拟除虫菊酯、烟碱类）和与PD抑郁相关的外周疾病过程。 我们的数据使我们能够有效地进行调查多维网络的高风险试点研究 使用监督机器学习方法来识别不同生物系统中的慢性反应模式 这些分子层可能与暴露和/或疾病过程有关。将我们的多组学数据组合成 多维网络将解决我们目前关于分子机制的知识差距， 抑郁症和焦虑症的PD，一个主要的神经退行性疾病的日益重要的老龄化社会。

Project Abstract We propose to use metabolic and multi-omic markers to elucidate how chronic, long-term pesticide exposures affect the occurrence of depression and anxiety in elderly living in a rural environment and those who develop Parkinson’s disease (PD) and depression. Depression and anxiety disorders are the most common types of mental disorders in older adults. Depression prevalence is as high as 25% in long-term care settings and strongly contributes to health care use for comorbid illnesses1-5. Major depressive episodes and clinical anxiety occur at a much higher rate among PD patients even well before PD diagnosis1,2. Improving our understanding of modifiable risk factors and mechanisms involved in depression and anxiety amongst the elderly is an urgent public health matter. Over the past two decades, our team has generated a unique data resource that now provides us with the opportunity to investigate the contributions of common and important environmental exposures (pesticides) to depression and anxiety and to explore disease processes through multidimensional biologic networks. With prior funding, we have collected lifetime depression/anxiety diagnosis and treatment histories as well as current (and follow-up) status of depressive and anxiety symptoms in a large population- based case control study of PD among residents of the California Central Valley, including more than 500 PD patients we closely followed for over a decade. Among these individuals, as many as 38% of PD patients and 27% of elderly without PD reported a diagnosis of depression or anxiety disorder at any time during their life. For all study participants we have genome (Illumina Global Screening Array, 660K markers) data available, and for ~800 additionally we generated epigenome (Illumina Infinium 450K DNA platform; genome-wide DNA methylation) and untargeted metabolomic data (at two time points for 300 participants). We developed a longitudinal geographic information system (GIS) based assessment for pesticide exposures that links state-mandated information on type, date, and location of all agricultural pesticide applications in California recorded since 1974 to land use maps and study participants’ residences and work places. Additionally, we collected extensive information on occupational, home and gardening use of pesticides. Biologic processes, including biologic (metabolic, epigenetic) responses to chronic toxicant exposures or disease processes are dynamic but also dependent on genetic susceptibilities. Here, we propose to combine a powerful systems biology analytic approach to interrogate epigenetic and metabolomic data anchored in genetics to identify signatures for specific pesticide exposures (organophosphates, pyrethroids, neonicotinoids) and peripheral disease processes related to depression in PD. Our data uniquely position us to efficiently conduct a high-risk pilot study that investigates multidimensional networks using supervised machine learning methods to identify chronic response patterns in biologic systems across different molecular layers that can be linked to exposure and/or disease processes. Assembling our multi-omic data into multidimensional networks will address gaps in our current knowledge concerning molecular mechanisms contributing to depression and anxiety disorders in PD, a major neurodegenerative disorder of growing importance in aging societies.