Erin C. Dunn - US grants

DESCRIPTION (provided by applicant): Depression is a major public health problem among youth, currently estimated to impact between 1% to 9% of U.S. children and adolescents and twice as many females as males. Researchers have explored gene- environment interactions (GxE) in order to more deeply understand the complex etiology of depression. Despite advances, knowledge remains limited on how genetic and environmental factors combine to contribute to depression. Moreover, the field's conceptualization of "environment" is sorely narrow;very little empirical work has tried to understand the salience of both proximal and distal environments and how both types of environments interact with genetic risks. These gaps in knowledge are problematic;without a more inclusive understanding of the role of "environment," GxE research may ignore environmental risks and pathogens important in the lives of youth. Guided by a "cells to society" approach, this study will use data (Waves I - IV) from the National Longitudinal Study of Adolescent Health (Add Health), a nationally- representative study of youth ages 12-19 at enrollment, to investigate the following aims: (1) examine whether three genes found to be linked to depression (SLC6A4, DATI, DRD4) modify the association between proximal family environments (characterized by parental abuse and maltreatment of the child and low perceived closeness between parent and child) and depression in adolescents;(2) explore, after controlling for individual risks and family environments, the independent contribution of distal environments (schools and neighborhoods) on youth depression and disentangle the unique effect of these environments;(3) investigate whether SLC6A4, DAT1, and DRD4 modify the association between distal environments (schools/neighborhoods) and depression in adolescents. These aims will be accomplished using Add Health's unique genetically-informative, multi-level, longitudinal design. These aims are aligned with NIMH's Strategic Plan: Strategy 1.2 ("Identify the genetic and environmental factors associated with mental disorders") and Strategy 2.3 ("Develop tools to better define/identify risk and protective factors for mental illness across the lifespan") and the Division of Developmental Translational Research (DDTR), whose high priority is to "test integrative models incorporating biological, behavioral, and experiential factors in the development of psychopathology, and utilize longitudinal research to track trajectories of risk and protection based on the combined and interactive influences among these factors." Results of this innovative, multi- level study will advance interdisciplinary knowledge on how to measure environment and when and where to intervene in the lives of youth to reduce their risk depression and promote their resilience.

DESCRIPTION (provided by applicant): This Mentored Research Scientist Career Development Award (K01) will provide the candidate with the necessary skills and knowledge to develop an independent program of research that uses epidemiological methods to identify genetic and environmental determinants of psychiatric disorders. Although psychiatric symptoms and disorders, such as depression and anxiety, have a complex etiology and emerge through the effect of genes, experience, and their interaction (e.g., gene-environment interaction; GxE), efforts to identify GxE interactions have had mixed success. The overall aim of the current proposal is to test the hypothesis that GxE effects are strongest during sensitive periods in development, or windows of time in the lifespan when the developing human brain is particularly vulnerable or sensitive to experience, including exposure to social adversity (e.g., child maltreatment, social deprivation). This main hypothesis will be tested in three aims. Aim 1 will examine the effect of timing of exposure to adversity on emotion recognition skills and subsequent development of internalizing symptoms. Aim 2 will examine the effect of genetic variation in sensitive period relevant gene pathways on emotion recognition skills and internalizing symptoms. Sensitive period relevant gene pathways will be defined by two sets of genes. The first will be sets of genes identified in the NIMH-funded Brain Cloud resource, a database of temporal patterns of gene expression in the human prefrontal cortex. Genes selected in this gene set will be highly and differentially expressed in the early years of life. Te second gene set will be the human orthologues of genes shown in animal studies to regulate the timing of sensitive periods (i.e., gad2, otx2, rtnf, lynx1, and bdnf). SNPs will be mapped to a gene using ProxyGeneLD. All SNPs in the pathway will be modeled simultaneously; thus no SNP-level tests will be conducted. Aim 3 will investigate gene-by adversity interactions (GxE), focusing on whether genetic variation in sensitive period relevant gene pathways modifies the association between timing of adversity and both emotion recognition and internalizing symptoms. Findings generated from the proposed research can help identify periods in the lifespan when interventions could be most effective in preventing internalizing symptoms, mechanisms by which adversity increases risk for psychopathology, and possible targets to treat internalizing symptoms and disorders. The training component of the proposed award, centered in the Psychiatric and Neurodevelopmental Genetics Unit at the Massachusetts General Hospital, is designed to provide the candidate with the skills and knowledge necessary to reach her career goals and complete the K01 research aims. The candidate, Dr. Erin C. Dunn, has a background in social and psychiatric epidemiology, but no training in the use of bioinformatics tools, large-scale genomic data, and mechanisms linking exposure to adversity to subsequent risk for internalizing symptoms. Her long-term career goal is to become an independent, translational epidemiologist, with the skills and knowledge to translate findings from basic science research and improve population-level health. To accomplish this goal, Dr. Dunn will be trained in two novel areas: (1) genomic and bioinformatics tools for epidemiology; and (2) developmental neuroscience of psychopathology. Within these training areas, Dr. Dunn will develop the skills to use bioinformatics resources, conduct pathway and gene set analyses, and identify patterns of gene expression. These newly-acquired skills will be integrated with conceptual and methodological strategies to identify sensitive periods in development and the pathways through which exposure to adversity increases risk for psychopathology. Training activities include coursework, workshops, conference attendance, as well as supervised research projects, individual training, and ongoing supervision and consultation. Training will be overseen by two internationally recognized mentors (Dr. Jordan Smoller and Dr. Charles Nelson) and a panel of expert consultants. This interdisciplinary mentorship and consultant team is comprised of experts in molecular and statistical genetics, gene expression, bioinformatics, developmental neuroscience, GxE, and psychiatric disorders. The training aims will be applied in the research component of the award. The research component consists of two studies designed by Drs. Dunn, Smoller, and Nelson to test the hypotheses linked to the aims noted above. The studies use data from the two largest, individually- genotyped general population datasets in the world, in which internalizing symptoms, emotion regulation skills, and exposure to adversity have been deeply phenotyped from early childhood through adolescence. These studies are: the Avon Longitudinal Study of Parents and Children (n=13,988) and the Generation R study (n=9,745). Together, these training and research projects will constitute the basis for an R01 proposal that Dr. Dunn will prepare in the third year of the award period. This R01 will employ the training and research completed during the award period and will seek to replicate and extend findings that support a GxE informed model to identify sensitive periods. Depending on the K01 results, such a model could include specific gene sets, measures of emotion recognition skills and other social competencies, timing of adversity, and measures of depression, anxiety, and internalizing symptoms. It could also include other factors that mitigate or reverse the effects of exposure to adversity during the sensitive period (e.g., fetal drug exposure).

Project Summary Childhood adversity (e.g., abuse, poverty) is a potent risk factor for depression, increasing lifetime risk of this common and burdensome disorder by at least two-fold. While the association between adversity and depression risk is well documented, the biological mechanisms explaining this relationship are poorly understood. In this proposal, we will address this gap by testing the central hypothesis that vulnerability to adolescent- and young adult-onset depression arises, in part, via the effects of adversity-induced epigenetic changes during an early sensitive period that occurs in the first five years of life. Sensitive periods are life stages when the brain is highly plastic and experience (e.g., adversity) can impart enduring effects. This hypothesis will be prospectively tested across three aims ? in a discovery and replication approach ? using data from two large birth cohorts: (1) the Avon Longitudinal Study of Parents and Children and (2) Generation R. In Aim 1, we will investigate the extent to which the developmental timing of exposure to adversity predicts blood DNA methylation (DNAm). We will use an innovative two- stage structured lifecourse statistical modeling approach to investigate the role of repeated exposure to seven distinct types of adversities during early life (up to age 7) on DNAm in middle childhood (age 7). For each type of adversity, we will investigate the following theoretical models to determine which one or more are best supported by the data: (1) a sensitive period model, in which the effect of presence or absence of exposure to adversity on DNAm depends on the time period of the exposure; (2) an accumulation model, in which the effect of exposure to adversity on DNAm increases with the number of occasions exposed, regardless of timing; and (3) a recency model, in which the effect of exposure to adversity on DNAm is stronger for more proximal events. In Aim 2, we will use regression and causal inference-based mediation approaches and Mendelian randomization to determine the degree to which age 7 DNAm changes predict adolescent-onset depression and mediate the effect of adversity on adolescent depression. In Aim 3, we will determine the short- vs. longer-term effects of DNAm on risk for young-adult depression by examining: (a) the persistence of DNAm profiles from age 7 to age 17; and (b) the relative contribution of early vs. adolescent adversity on age 17 DNAm and risk for young-adult onset depression. Throughout, we will control for genetic factors shown to explain variability in DNAm and depression. This research will identify molecular biomarkers of exposure to adversity and risk for depression and determine the age stages when adversity is most likely to affect this biomarker. These findings will inform our understanding of the high-risk/high-reward stages of development when adversity is most harmful and when public health investments could be most efficacious in preventing depression.