Jenny Tung, Ph.D. - US grants

Project Summary The social environment has a clear and profound impact on human health and well being. Chronic social stress and reduced access to social support are strongly linked to major diseases of aging; as a result, social adversity is highly predictive of life expectancy itself. Recent evidence suggests that, while some of this relationship is explained by correlated factors such as smoking, obesity, and health care access, social stressors also have a direct impact on physiological function. Indeed, work in animal models has clearly demonstrated that the experience of social subordination alone can alter the function of the immune system, in part by altering gene regulation in immune cells. The goal of the proposed research is to address a key outstanding question that arises from these findings: when, and for whom, are chronic social stress effects on immune function most important? To do so, it will take advantage of dominance rank in female rhesus macaques as a model for chronic social stressor exposure in humans. Rhesus macaque females are excellent models for human social stress because they naturally organize into dominance rank hierarchies in which low ranking individuals experience increased rates of harassment, reduced social affiliation, and physiological markers of rank-related stress. Importantly, dominance rank assignments, and thus an individual's exposure to social stressors, can be manipulated in this species by manipulating group membership. Such manipulations yield a powerful experimental model for investigating the consequences of socially induced stress?an approach that is directly translatable to humans, but that is practically and ethically impossible in humans themselves. The proposed study will take advantage of this model to investigate how differential exposure to dominance rank-induced social stress causally influences gene expression in the immune system. Specifically, it will use an in vitro approach to efficiently screen for condition-specific social stress effects on gene expression levels across 30 physiologically relevant environmental conditions (e.g., pathogen exposure, steroid hormone signaling). It will complement the in vitro screen with an in vivo test of the gene regulatory and antibody response to influenza vaccination, a medical procedure in which variable responses are of particular concern as individuals age. Finally, it will test whether age, social behavior, and genotype can be used to predict interindividual variation in the strength of social stressor effects on immune regulation, and hence which individuals are most vulnerable. Together, the proposed analyses will provide much-needed insight into the factors that explain when and why individuals differ in their response to the same social stressors, as well as the potential consequences of these differences for medical treatment. The project's results will therefore have direct translational application to both identifying the most susceptible members of our aging population and suggesting tailored strategies for intervention.

Intellectual Merit:

A strong relationship between chronic social stress and adverse health outcomes, including higher mortality rates, is well established. However, the mechanisms through which this relationship arises, as well as predictors of variation in individual response to social stressors, remain poorly understood. Research on this topic is complicated by the need to integrate tools, analytical methods, and perspectives that bridge multiple disciplines, including animal behavior, psychology, sociology, and genetics and genomics. The research in this project draws on tools and theories from these fields to investigate two potential mechanisms that mediate the response to social stress: behavioral strategies - particularly the establishment of close social bonds; and genetic variation, which may alter individual susceptibility to these effects. To do so, it utilizes a powerful animal model for social stress, dominance rank in captive female rhesus macaques. Specifically, it tests how social bonds, genetic variation, and the combination of the two modify biological responses to social stress, as measured by genome-wide gene expression. This project therefore expands on previous demonstrations that dominance rank-induced stress has potent effects on gene regulation. The significance of this work is three-fold. First, it reveals how individual variation in behavioral strategies and genotype alleviate or exacerbate the detrimental effects of chronic stress. Second, it identifies what physiological processes are altered by behavioral strategies and genetic variation. Finally, it presents a valuable opportunity to train the PI (Fellow) in new areas, including psychology, genetics, and genomics, that will be important to his development as an independent interdisciplinary researcher.

Broader impacts:

The broader impacts of this proposal are three-fold: 1) It promotes teaching, training and learning not only through mentorship of the Fellow, but also by including him in the training of undergraduate students at Duke. Specifically, the Fellow trains two undergraduates in data collection, analysis, synthesis and presentation, with the ultimate goal of having the undergraduates present their findings as part of one of Duke's undergraduate research poster symposiums. 2) It promotes public interest and understanding of scientific research through presentations by the PI on animal behavior, evolution, and social and behavioral studies to local K-12 schools in Durham and Chapel Hill, and to a public audience at the Nature Research Center (part of the North Carolina Museum of Natural Sciences). 3) Finally, it contributes to society as a whole by improving the understanding of how adverse social environments, many of which are associated with low socioeconomic status, promote disparities in health and well being. Such findings are critical both to identifying the most susceptible members of the population, and to establishing effective interventions to alleviate the adverse effects of social stress.

DESCRIPTION (provided by applicant): Advancing our understanding of social and behavioral effects in aging will increasingly require the integration of complex and often disparate data sets To measure such effects, investigators must manage demographic data on mortality and fertility, behavioral or survey data on social relationships, and, increasingly, biomarker data tha capture genetic and physiological variation. Currently, extant databases are designed to facilitate analysis of complex demographic and sociobehavioral data or complex genetic and genomic data: databases that meet the challenge of integrating all of these data types do not yet exist. Consequently, the research community is not reaping the full benefits of these data sets for understanding aging. Further, we lack models for how to house these data types together in a cohesive and accessible fashion. We propose to develop such a model by building on an existing database on wild primates that houses individual-based multidimensional, longitudinal phenotypic data that have already proved valuable for studies of social and behavioral effects on aging. We also have a growing set of complementary genetic and genomic data on the same individuals, including candidate gene and whole genome resequencing, gene expression, and epigenetic data sets that promise to capture physiological changes across the life course. In the proposed work, we will provide centralized, integrated archival storage for these multidimensional data sets, create a seamless integration of genetic and phenotypic information at the individual level, and provide a much needed, well documented model of such an integration. We will also work with the National Archive of Computerized Data on Aging (NACDA) to build mechanisms for sharing these data with other researchers in the field. Specifically, we propose to (1) build database modules to house our expanding multi-dimensional genetic and genomic data sets, (2) link these new genetics and genomics modules to our existing database (BABASE) and to each other, to integrate our genetic and phenotypic information, and (3) create a public portal to BABASE that will allow open access to all components of the genetics/genomics modules of the database, as well as open access to key aging- related components of the phenotypic data; this portal will be accessible through NACDA. Our new genetic modules will draw on database module designs pioneered by Chado, a branch of the Generic Model Organism Database project (GMOD). Together, these efforts will provide important archival storage for these valuable data sets, increase the efficiency of data analysis, and promote new, synergistic research directions, including collaborations with outside investigators that will allow us to gain deeper insights into aging in natural mammal populations. In addition, because all the code underlying BABASE and its new extensions will be open source, the proposed work will produce models for how other population studies focused on aging can achieve similar goals.

The objective of this study is to develop novel methods for generating and analyzing genome-scale data from biological samples that have been collected in a noninvasive manner. Noninvasive samples are often the only type of biological sample available for natural populations, especially in endangered or threatened species, but they yield tiny quantities of low-quality DNA. Noninvasive genetic analysis techniques have changed little in the past twenty years, meaning that the genomic revolution has not yet arrived for species for which only noninvasive samples are available. This study will address this problem by producing new lab protocols for increasing and purifying the DNA yield from noninvasive samples to a level appropriate for genomic analysis, and by developing new software to analyze the resulting genomic data. It will validate these approaches by comparing genomic data from noninvasive samples to data from high-quality DNA samples, using data for a known pedigree from the well-studied wild baboons of the Amboseli basin in East Africa.

This project will result in a major leap forward in tools for studying the genomics of many different species and sample types. Publicly available protocols and software will be released, which can be used to investigate evolutionary and population history in species for which such studies were previously impossible. It will also provide societal benefits in the form of tools for assessing the conservation status of vulnerable species. Finally, both graduate and undergraduate students will be trained as collaborators on the project.

DESCRIPTION (provided by applicant): Social adversity, both in early life and adulthood, can have major and long-lasting impacts on human health. Low social status and social isolation in early life have been linked to changes in the immune response and elevated rates of cardiovascular disease later in life, even when status differences are erased. At the same time, social adversity in adulthood - unconnected to any early life experience - has well-documented effects on mortality risk. Thus, traits that are influenced by social conditions early in life reman, at least to some degree, responsive to later social environments. However, the mechanisms that mediate these dual properties-long-term stability coupled with the potential for plasticity an change-remain poorly understood. In particular, we do not understand the molecular mechanisms that translate social experiences across the life course into physiological changes that affect health and disease risk. The goal of the proposed work is to leverage a powerful animal model for social adversity in humans-dominance rank in nonhuman primates-to investigate the relative contributions of early life and adult social status to variation in DNA methylation levels. DNA methylation is an epigenetic mechanism that may serve as an important link between social environment, physiology, and health. However, while this relationship has been investigated in detail for a handful of loci, we do not yet understand its importance genome-wide, or the degree to which changes in DNA methylation in response to the social environment depend on the timing of exposure. Intensively studied primate populations can serve as important models for these questions because known individuals are directly observed from conception to death, and social environmental effects are not confounded by other predictors of health, such as access to health care, diet, or smoking. This application takes advantage of one such population, the well-studied Amboseli baboon population of Kenya, to assess whether and how the relationship between social adversity and DNA methylation changes over the life course. Specifically, we propose to investigate the contribution of social status, in both early life and in adulthood, to patterns of genome-wide DNA methylation in blood. We will investigate targets of DNA methylation that are associated with early life social status, adult social status, or both; if both, we will further test whether these effects act independently. Because variation in DNA methylation levels does not always affect variation in other traits, we will also complement these data with data on gene expression levels from the same individuals, obtained during the same blood draw. The gene expression data will reveal the degree to which DNA methylation patterns that are sensitive to social status also influence downstream gene expression levels. Together, our results will help establish not only whether social status influences DNA methylation, but also the role of different stages of the life course in this relationship and the likelihood that epigenetic mechanisms explain broader relationships between social adversity and health.

Many primates, including humans, live in complex social environments, and the capacity to deal with such environments can vary substantially across individuals. While some individuals form stable, positive social bonds, others do not; similarly, individuals vary in their achieved social status. Research in both humans and non-human primates suggests that variation in these social experiences can have profound effects on physiology, health, and survival. This project uses behavioral and biological data from wild baboons to test the hypothesis that social adversity influences health-related traits by altering the way that genes are expressed. Identifying these connections is crucial for understanding the evolution of primate sociality, and for addressing the well-documented health consequences of social adversity in humans. Other broader impacts of the project include fostering of research collaborations, undergraduate training, and dissemination of novel methods for genomic studies.

This two-year project will study 88 adult male baboons from the Amboseli, Kenya baboon population, to investigate whether low social status and/or social isolation lead to changes in DNA methylation, and whether these changes predict how individuals respond to an immune challenge. To do so, investigators will combine behavioral data with measurements of genome-wide DNA methylation levels and immune-related gene expression patterns. The resulting data will shed new light on whether, and to what degree, social experiences influence immunological traits that likely contribute to health and survival in wild primates.

Summary Stressful experiences in infancy and childhood can disrupt the process of normal development, producing life-long impacts on human health. Such experiences are powerful predictors of later life disease and mortality risk, and exposure to multiple early life stressors can have even more potent effects. These observations suggest that adverse early experiences become biologically embedded in human physiology. However, the molecular mechanisms that mediate the embedding process are not well understood, challenging our ability to predict susceptible individuals and develop effective intervention strategies. The goal of the proposed work is to leverage an emerging model for genomics in natural animal populations to investigate the effects of early life stress on genome-wide DNA methylation levels. DNA methylation is an epigenetic mechanism that is strongly influenced by early life conditions, can remain stable over time, and can influence downstream traits through its effects on gene regulation. However, we still know little about its importance in mediating the effects of early life stressors. Specifically, we do not understand the genes and pathways most affected by early life stress, the degree to which these effects persist over time, or the environmental, behavioral, or genetic factors that mediate inter-individual differences in susceptibility. Part of the challenge in answering these questions lies in the difficulty of collecting environmental data and biological samples for the same individuals and families over time. Animal models provide an opportunity to overcome this challenge. To do so, this proposal takes advantage of an intensively studied primate population, the wild baboons of the Amboseli ecosystem of Kenya, that have been the subjects of longitudinal study for up to 8 contiguous generations. In Amboseli, early life stressors have profound effects on fertility and survival, even in the absence of health risk behaviors like smoking, alcohol consumption, or poor diet. We propose to investigate the epigenetic consequences of these early life stressors on a genome-wide scale. Specifically, we will test the unique and cumulative effects of major early life stressors on DNA methylation levels in blood, investigate the relationship between early adversity- associated differential methylation and gene regulation, and investigate the physiological and behavioral pathways that connect early adversity to the epigenome later in life. We will also test the degree to which the signature of early life effects persists over time, using longitudinally collected repeated samples from individuals and families. Finally, we will investigate whether genotype, behavioral patterns, or environmental conditions affect rates of change, and use Mendelian randomization analyses to dissect the causal pathways that link the early environment to epigenetic patterns. Together, our results will provide an unusually comprehensive window into the relationship between early adversity and the epigenome. They will thus shed new light into the role of epigenetics in mediating the long-term effects of early adversity during development.

This RAPID award takes advantage of an ongoing experiment in the Kalahari Desert of South Africa on the Damaraland mole rat, a highly social mammal that cooperatively raises its young. The goal is to test hypotheses about how breeding drives bone growth and remodeling in females. Two bones that do and do not elongate in reproductive female mole rats (lumbar vertebrae and femur, respectively), and the same bones in nonreproductive females, will be compared for bone growth and bone-derived stem cell activity, mineralization potential, gene expression, and responsiveness to hormonal signals. The project will provide new information about how hormonal signals and hormonally-mediated gene expression changes result in bone growth that is related to fitness in this unusually cooperative mammal. Results will be broadly applicable to understanding the control of skeletal remodeling and growth in mammals. The project involves training of a postdoctoral fellow and volunteers at the Kalahari Research Centre in South Africa, and public education at the North Carolina Museum of Natural Sciences.


This project will provide novel insight into the physiological and genomic targets of social evolution in the Damaraland mole rat. In females of this eusocial species, secondary skeletal growth (lumbar vertebrae elongation and widening of the pelvic girdle) is a function of reproductive activity and results in increased litter sizes. The project will investigate gene regulatory differences in bone-forming mesenchymal stem cells that explain this breeding female-specific growth phenotype. It will test the hypothesis that estrogen and insulin-like growth factor signaling differentiate reproductive females and non-breeders, and bones that experience secondary growth from those that do not. Differences in stem cell gene expression and chromatin accessibility at estrogen-receptor transcription-factor binding sites will be quantified. By elucidating the physiological mechanisms underlying permanent, secondary skeletal growth in mole rat females, the results will lead to a better understanding of how skeletal growth in adult mammals, including humans, is controlled.

Project Summary The social environment has a clear and profound impact on human health and well being. Chronic social stress and reduced access to social support are strongly linked to major diseases of aging; as a result, social adversity is highly predictive of life expectancy itself. Recent evidence suggests that, while some of this relationship is explained by correlated factors such as smoking, obesity, and health care access, social stressors also have a direct impact on physiological function. Indeed, work in animal models has clearly demonstrated that the experience of social subordination alone can alter the function of the immune system, in part by altering gene regulation in immune cells. The goal of the proposed research is to address a key outstanding question that arises from these findings: when, and for whom, are chronic social stress effects on immune function most important? To do so, it will take advantage of dominance rank in female rhesus macaques as a model for chronic social stressor exposure in humans. Rhesus macaque females are excellent models for human social stress because they naturally organize into dominance rank hierarchies in which low ranking individuals experience increased rates of harassment, reduced social affiliation, and physiological markers of rank-related stress. Importantly, dominance rank assignments, and thus an individual's exposure to social stressors, can be manipulated in this species by manipulating group membership. Such manipulations yield a powerful experimental model for investigating the consequences of socially induced stress?an approach that is directly translatable to humans, but that is practically and ethically impossible in humans themselves. The proposed study will take advantage of this model to investigate how differential exposure to dominance rank-induced social stress causally influences gene expression in the immune system. Specifically, it will use an in vitro approach to efficiently screen for condition-specific social stress effects on gene expression levels across 30 physiologically relevant environmental conditions (e.g., pathogen exposure, steroid hormone signaling). It will complement the in vitro screen with an in vivo test of the gene regulatory and antibody response to influenza vaccination, a medical procedure in which variable responses are of particular concern as individuals age. Finally, it will test whether age, social behavior, and genotype can be used to predict interindividual variation in the strength of social stressor effects on immune regulation, and hence which individuals are most vulnerable. Together, the proposed analyses will provide much-needed insight into the factors that explain when and why individuals differ in their response to the same social stressors, as well as the potential consequences of these differences for medical treatment. The project's results will therefore have direct translational application to both identifying the most susceptible members of our aging population and suggesting tailored strategies for intervention.

Project Summary The incidence, prevalence, and burden of disease are unequally distributed within and across human populations. This heterogeneity is due in part to differences in exposure to social adversity, which is in turn patterned by variation in socioeconomic status, access to social support, and early life disadvantage. Indeed, experimental studies in animal models indicate that social adversity per se, even in the absence of differences in health care access or health risk behaviors, can increase disease susceptibility and shorten lifespan. They have also shown that social disadvantage both increases the expression of inflammation-related genes and alters the genome-wide immune response to bacterial and viral antigens. The goal of the proposed research is to investigate the translatability of these findings to human populations by studying the effects of social disadvantage on immune gene regulation in the context of health disparities. Specifically, the proposed study will characterize the relationship between socioeconomic status, past trauma, and peripheral blood mononuclear cell (PBMC) gene expression in samples collected by the Detroit Neighborhood Health Study (DNHS), a population-representative study of urban Detroit. The DNHS sample is ideal for this work because it is complemented by extensive information on individual and neighborhood-level socioeconomic disadvantage and, unusually for such studies, cryopreserved PBMCs for a representative subsample of study participants. Such samples are precisely the type used in studies of chronic social stress and immune gene regulation in nonhuman primates, thus maximizing comparability against findings from animal models. Notably, previous studies in rhesus macaques have shown that the effects of low social status on gene regulation are exaggerated after immune challenge. Such observations suggest that social disadvantage is particularly important in shaping the response to pathogens. However, while the effects of genotype, age, and sex on the genome-wide gene expression response to immune stimulation are well studied, little is known about the role of chronic social stress in humans. The proposed study will address this gap by investigating how social disadvantage patterns immune gene expression in cryopreserved PBMCs from the DNHS sample, both at baseline and following exposure to the bacterial endotoxin lipopolysaccharide. It will also investigate the relative contribution of social adversity and genetic ancestry in shaping the immune response. By comparing these data to data generated using a similar approach in nonhuman primate models, this approach will highlight the degree to which the causal effects of social adversity in animal models are mirrored in humans. It will therefore address key questions about the genomic mechanisms through which social disadvantage translates into health outcomes, with direct application to identifying the sources of health disparities during aging.

CORE C: EXTERNAL NETWORK Research on the social, economic, and biodemographic dimensions of aging depends on successful interinstitutional exchange and interdisciplinary collaboration. To foster these relationships, Duke?s Center for Population Health and Aging (CPHA) includes an External Network Core (ENC) that connects faculty, staff, and trainees at Duke with the national and international aging research community. The ENC will take advantage of the strengths of biosocial research on aging in the local North Carolina Research Triangle to support collaborative pilot proposals and an interinstitutional working group on the Social and Biological Determinants of Health, which features a unique, cross-disciplinary focus on integrating research in human populations and animal models. Through these activities, as well as support for conference travel and short-term collaborative visits, the ENC will identify promising faculty and trainees with an interest in aging research and provide crucial early stage support to grow the aging research community. Additionally, to showcase cutting-edge, emerging work in behavioral and social research on aging, the ENC will support themed workshops and symposia on topics of interest to the CPHA community. These events, in combination with the established, CPHA-supported annual meeting, ?Demography Daze,? will feature important research advances, catalyze new opportunities for collaboration, and motivate new research directions within and outside Duke. The ENC thus takes advantage of the exceptional community of biosocial researchers at Duke, while simultaneously building connections with faculty and trainees on a regional, national, and global scale. It therefore advances the overarching aims of the P30 proposal as a whole, with particular relevance to integrating social science and life science approaches to understand aging across the life course.

Recent findings suggest that intermixing with now-extinct relatives like the Neanderthals may have helped humans adapt to new environments during their rapid expansion around the globe, and may still contribute to human differences in disease susceptibility today. However, the extent to which benefits to hybridization are also found in other primates is not well understood. This doctoral dissertation project will use a model for natural hybridization in primates to investigate signatures of beneficial transfer of genetic variants between species. It will also ask whether hybridization affects hair color and maturation timing in baboons, traits that may be important in mate competition, fertility, or survival. The research will provide novel insight into hybridization?s role in primate evolution and shed light on how hybridization may influence primate diversity in the future. It will also contribute to STEM education at the elementary, undergraduate, and graduate levels through the design of new educational activities, mentored research opportunities for undergraduate students, and support for the co-PI?s PhD training. Finally, this project will facilitate public engagement with science through free-to-the-public presentations that will also be accessible online.

To investigate the potential benefits of hybridization in primates, the project will integrate genomic and observational data from wild baboons. It will test the hypothesis that hybridization facilitates the transfer of a small number of advantageous genetic variants between yellow baboons and anubis baboons. To do so, genomic data will be generated for multiple populations spanning a baboon hybrid zone. These data will be subjected to three complementary methods to identify genetic variants affected by adaptive introgression. Cases of potential adaptive introgression will be investigated for overlap with genes that are also differentially expressed between yellow and anubis baboons. The project also will test the hypothesis that genetic ancestry predicts hair color and maturation timing in a well-studied baboon population at the center of the hybrid zone. It will assess whether ancestry effects on these traits differ in number and magnitude across the genome, based on evidence that hair and skin color traits tend to have a simpler genetic basis than life history traits like maturation timing. Finally, the project will test whether regions in the genome that contribute to ancestry-related differences in hair color or maturation timing are candidates for adaptive introgression.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Project Summary As the population of the United States ages, the health burden imposed by diseases of aging is expected to increase concomitantly. Social factors, including low socioeconomic status, social isolation, and low social support, are among the best predictors of susceptibility to diseases of aging, as well as lifespan itself. Nevertheless, key questions about the causal relationship and the biological mechanisms that link social experiences to health and aging remain unanswered. Animal models are a powerful tool to address these questions. Like humans, other social mammals exhibit strong associations between social adversity, health, and mortality. Unlike humans, though, they experience less complex environments, have shorter generation times, and can be subjected to experimental manipulation in controlled environments. The goal of this proposal is to build a Research Network on Animal Models to Understand Social Dimensions of Aging. By supporting interdisciplinary communication and pilot research from both human and nonhuman animal researchers, we aim to maximize the impact of animal model research on understanding the social determinants of health and aging. A Research Network is essential because current research in this area is distributed across many different disciplines, there is no standard set of conferences or publication venues where researchers with related interests overlap, and there are high barriers to entry for animal model work. The proposed network will overcome these challenges by supporting scientific meetings and workshops that build contacts across disciplines and among researchers at all career levels. It will also recruit new, diverse investigators into the field by providing opportunities for pilot project support, travel fellowships, and mentorship by experienced senior investigators. As part of these activities, the network will provide hands-on training for animal model research in rodents and nonhuman primates and for comparative research that leverages human and animal data sets. To build visibility, it will support targeted symposia at meetings where animal model work on the social dimensions of aging has not traditionally been represented. Finally, it will identify priority areas for animal model research and promote data and protocol dissemination. These activities will help set a research agenda that extends beyond the network?s immediate activities. Thus, the proposed network will transform a weakly connected community into a self-sustaining field of researchers equipped to conduct impactful research on the social dimensions of aging. Areas of particular interest include models to test the causal effects of social interactions on health; methods that can be flexibly deployed in both nonhuman animals and human populations; and research that investigates the benefits of behavioral interventions for alleviating the costs of social adversity. Advances in these areas will have direct translational application to human health and well-being during aging.

Project Summary Social support and social integration are some of the most robust predictors of morbidity and mortality identified to date. This relationship arises from increased susceptibility to several of the top causes of death in the United States, including major diseases of aging such as heart disease and cancer. Recent studies suggest that a signature of social relationships is also detectable in data on gene regulation, highlighting a potential pathway through which social ties get ?under the skin? to influence health. However, despite abundant evidence that the health effects of social relationships begin early in life, no studies have related the full life course trajectory of social relationships to data on gene regulation, or used these data to investigate why some individuals appear more susceptible to social isolation than others. The goal of this study is to address these gaps by linking fine-grained, longitudinal data on social relationships to unbiased surveys of the molecular signature of social experience. To do so, we will leverage an established model for social relationships and health in natural animal populations, the baboons of the Amboseli ecosystem of Kenya. This population has been the subject of longitudinal study for up to 9 generations, revealing that social isolation predicts shortened lifespan in a manner highly analogous to humans. We propose to link annual measures of social relationships, from birth through adulthood, with gene expression and chromatin accessibility data collected both at baseline and following ex vivo challenge with bacterial and viral mimics. This strategy will allow us to investigate the types of social relationships that matter most, the timing of their effects on gene regulation, and their relevance for immune function, a primary contributor to variation in health during aging. Using these data, we will address three aims. First, we will characterize the gene regulatory signature of variation in social relationships across the life course. We will investigate the relative roles of early life, cumulative experience, and social relationships close to the time of biological sample collection, as well as the relative importance of social relationship quantity versus quality. Second, we will assess whether individuals vary in their sensitivity to social environments based on genotype, by identifying gene-social relationship interactions that affect gene expression. Finally, we will test the consequences of social relationship-associated gene regulation for immune defense and lifespan. In doing so, this work will shed important light on whether gene regulatory signatures of social relationships are likely to be mechanistically implicated in the link between social relationships and health, or instead serve as passive biomarkers. Together, our results will provide the most comprehensive window into the functional genomic signature of social relationships available to date. By revealing when, how, and for whom social relationships matter most, they will therefore address three questions of outstanding importance to understanding the role of the social environment in human health.