We propose to characterize associations among the fecal microbiome, the fecal glycome, and measures of household environmental exposures in infants who do and do not subsequently develop autism spectrum disorder (ASD) from the MARBLES cohort. One of the most common co-morbidities in autism are gastrointestinal problems, and the presence of frequent symptoms of diarrhea or constipation is associated with more severe symptoms. However, virtually all research of GI dysfunction in ASD to date has been conducted after the ASD diagnosis has been made, thus not allowing for examination of temporal relationships between GI dysbiosis and the onset of ASD. Moreover, few underlying biologic mechanisms have been identified. Increasingly, the prominent but insufficiently characterized, role of the microbiota in human health has been recognized. Environmental influences on individual gut microbiota profiles are also coming under scrutiny, but there has been very little work on the impact of chemical exposures on the microbiome. Taking advantage of data and samples available from a large, prospective pregnancy study of high-risk infant siblings of children with autism, this project seeks to investigate the development in early postnatal life of the individual profiles of the gut microbiome, the environmental chemical influences on these, and their relationship to GI symptoms and to the subsequent development of autism and its early signs. Our overarching hypothesis is that environmental exposures common in developing countries influence the developing intestinal microbiota and intestinal permeability in the first year of life and that the resultant dysbiosis and gut leakiness increase risk for development of ASD. With an established interdisciplinary team at the cutting edge of the microbiome and glycome measurement, we will use recently developed effective techniques to quantify fecal milk glycans and milk glycan monomers that are clear drivers for intestinal health or dysbiosis in the developing infant gut microbiome. We will apply an innovative mechanistic framework that incorporates a number of known or suspected factors in GI dysfunction in ASD, including a compromised intestinal barrier, and links exposure to environmental toxins, GI outcomes, and ASD. Establishing associations between the maternal and child environment, the developing infant gut microbiome, and onset of ASD symptomology and diagnosis would set the stage for mechanistic studies examining ways to shift the infant microbiota away from onset of dysbiosis during the first year of life a critical developmental period?with potential implications for neurodevelopmental outcomes.

Placental tissue is normally discarded at birth, but is essentially a molecular time capsule for gene by environmental interactions and dysregulated molecular and cellular pathways that can be revealed at the level of the epigenome. Identifying epigenetic biomarkers at birth that reflect in utero exposures or predict adverse neurodevelopmental outcomes is an important goal that has been limited by prior technologies or lack of relevant tissue availability. Our team of currently collaborating interdisciplinary scientists within the Children?s Center for Environmental Health plans to use existing placental samples from a prospective high-risk cohort study (MARBLES) to identify epigenetic biomarkers at birth for in utero exposure to polychlorinated biphenyls (PCB) and neurodevelopmental outcomes by age three. Using unbiased whole genome bisulfite sequencing (WGBS), we have previously demonstrated that placental tissues retain the distinctive DNA methylation patterns of the preimplantation embryo and so can capture the molecular state in very early development, a feature that is conserved across mammalian species, including mouse. The new hypothesis to be tested in this proposal is that perinatal exposures to PCB adversely impact neurodevelopment and leave a lasting molecular signature over genes relevant to neurodevelopment that can be detected in placenta. The proposed PCB Epigenomic Brain & Behavior Lasting Effects Study (PEBBLES) will combine the analysis of human placental samples from the high-risk MARBLES cohort with the analysis of placenta and brain tissues and sorted cell types derived from a mouse model of perinatal exposure to the same mixture of PCB congeners detected in MARBLES mothers. This study will leverage existing neurological and behavioral analyses and samples to examine the relationship between PCB-induced perturbations of DNA methylation marks with adverse neurotoxic outcomes. Epigenomic analyses of placenta and brain as well as sorted cellular subtypes from each of these tissues will include WGBS for methylome, RNA-seq for transcriptome, and ATAC-seq for chromatin accessibility. Bioinformatic and statistical analyses will integrate the genomic data sets with behavioral and molecular outcome measures and determine whether similar epigenetic marks are observed in placenta that could be used to predict long-lasting adverse brain and behavioral outcomes in humans.

The MARBLES (Markers of Autism Risk in Babies: Learning Early Signs) Study was launched in 2006 by the UC Davis MIND Institute and Center for Children's Environmental Health as the first epidemiologic cohort of younger siblings of children with autism spectrum disorders (ASD) to begin follow-up during (and before) the prenatal period when ASD and other neurodevelopmental outcomes are likely to originate.1-7 In contrast to population-based cohorts, which require very large sample sizes given a relatively low prevalence of ASD (currently 1 in 68)8 and typically are not able to conduct gold standard diagnosis of ASD, the enriched risk design takes advantage of participants at exceptionally high risk for developing ASD and other neurodevelopmental outcomes, achieving tremendous efficiencies. Previous cohorts of high-risk younger siblings recruited postnatally have not addressed non-inherited and potentially modifiable etiologic factors.9 Early enrollment provides an opportunity to examine a broad array of environmental exposures and their mechanisms, while simultaneously allowing thorough search for early biologic markers.10 Given increasing prevalence of ASD,8,11 it is more critical now than ever to invest in studies identifying emerging environmental factors responsible for increasing risk of these neurodevelopmental disorders and the mechanisms underlying their etiology, which are currently not well-understood. This project addresses both gaps by maintaining and enhancing the resource infrastructure of the MARBLES enriched-risk cohort. Retention of this cohort with deep evaluation of risk factors, mechanistic markers, and outcomes will be critical to evaluation and early identification of newly emerging etiologic factors for ASD in a susceptible population, serving as a canary in a coal-mine for identifying exposures that influence neurodevelopment. Our cohort's data and specimens will permit analyses of new questions on exposures in relation to well-defined clinical neurodevelopmental outcomes, and our rich characterization of mechanistic biomarkers could identify pathways and signatures of susceptibility involved. We propose to continue to enroll and follow-up participants of one of the only enriched-risk ASD U.S. cohorts with prospectively collected pregnancy data and biosamples that is currently in no cost extension (NCE). In addition, we propose to further develop and maintain our repository infrastructure for the expansive data and sample repositories in order to expand collaborative sharing and facilitate investigation of newly-emerging environmental exposures and molecular mechanistic markers in relation to risk and presentation of ASD and other adverse neurodevelopmental conditions. Finally, we plan to validate and reliability test environmental exposure questions and measures for use in future studies. Completion of these aims will assist collaborative sharing of our vast collection of data and biosamples to facilitate investigation of newly-emerging environmental exposures and molecular mechanistic markers in relation to risk and presentation of ASD and other adverse neurodevelopmental conditions.

This study is designed to test a novel paradigm that the human genome dynamically interacts with the environment and that epigenetic mechanisms are at the interface of genome- environment interactions. The human genome is marked by structural variations including large copy number variations and differences in repetitive sequences. Environmental pollutants such as polychlorinated biphenyls (PCBs) are associated with DNA methylation changes. Chromosome 15q11-q13 duplication syndrome (Dup15q), is one of the most common copy number variations (CNV) observed in neurodevelopmental disorders. This proposal is based on the findings that human brain samples with Dup15q syndrome showed significantly higher levels of the persistent organic pollutant PCB 95 than controls or idiopathic autism cases, and that the interaction of both Dup15q and PCB 95 cumulatively impacted the methylation of over 1000 genes, enriched for autism candidate genes with synaptic functions. Three aims are proposed to test the overarching hypothesis that the cumulative impact of multiple risk factors converge on common gene pathways and are detectable as epigenetic signatures in early life tissues. The first aim seeks to determine the mechanism by which UBE3A overexpression interacts with PCB-95 exposure on histone H2A.Z biding sites genome-wide using next generation sequencing, siRNA knockdown, and CRISPR/Cas9 technologies. The second aim investigates the interaction between a human-relevant mixture of PCB congeners and UBE3A overexpression on the methylome and transcriptome in dental pulp stem cells derived from baby teeth from Dup15q and control children. The third aim seeks to identify potential methylation biomarkers of POP exposures and CNV in placenta samples from a human prospective autism study and a rhesus macaque model of maternal obesity. The results of these studies are expected to formally test the hypothesis that DNA methylation levels reduced by cumulative impacts of environmental pollutants and genetic changes may result in alterations in chromatin, leading to transcriptional changes and adverse health outcomes. In addition, epigenetic alterations of autism candidate genes involved in synaptogenesis are expected to be uncovered by this approach and be established as perinatal epigenetic biomarkers of cumulative risk assessments, as well as guiding future behavioral and pharmacological treatments. Finally, the results of these studies are expected to be broadly relevant to understanding the relationship between the genome, environmental exposures, and the epigenome in human health and disease.

Maternal folic acid, the synthetic form of folate, is one of the first modifiable factors identified to date with the potential to reduce occurrence of autism spectrum disorders (ASD) by 40% if taken near conception. Folic acid appears to protect against ASD especially in mothers and children who are genetically susceptible to inefficient folate-dependent one-carbon and methylation metabolism, but this finding needs replication. In addition to being essential for neurodevelopment, folate is a primary methyl-donor for methylation reactions, including DNA methylation. The time near conception is an especially critical period for adequate methyl supply during cycles of active demethylation and re-methylation of the genome during embryogenesis. Our preliminary data show that folic acid supplementation is associated with even greater reductions in ASD risk, by 75%, in younger siblings of children with ASD. In addition, we found DNA methylation differences associated with no maternal use of folic acid supplements in birth tissues (placenta and cord blood) for genes with nuclear regulatory and brain development functions that could have implications for ASD. The goal of the proposed work is to leverage data and samples from mother-child pairs in two large prospective pregnancy cohorts of high-risk infant siblings to examine specific pathways for prevention of ASD through maternal dietary and supplemental folate intake. This work will build on previous studies by examining exposures collected prospectively, with more accuracy in terms of timing and dose. We will be first to examine whether other methyl-donor B-vitamins have associated effects. We will include information on dietary and supplemental folate and B-vitamin intake from validated instruments, and measurements from maternal first trimester serum, cord blood, and fetal placental tissue. We will examine folic acid interactions with genetic susceptibility factors. Finally, we will increase understanding of the underlying mechanisms by identifying DNA methylation changes associated with folate status, and investigate whether these changes overlap with methylation patterns observed in autistic brains, using innovative methods, multiple platforms, and replication across tissue type and study population, to address challenges associated with DNA methylation measurement. Through completion of the proposed project, we will gain a better understanding of how strongly folic acid is associated with reduced risk of ASD in high-risk families, when intake is most associated and at what levels, how this association differs based on genetic susceptibility, and whether maternal folic acid intake alters DNA methylation profiles in ways that could decrease ASD risk. These findings will have great clinical and public health implications, informing autism prevention trial and ultimately changes in recommendations and policy.

Because little is known about non-genetic causes of autism spectrum disorders (ASD), this proposal seeks a novel approach to advance our knowledge of the complex etiology of autism due to environmental chemical exposures from the use of common consumer products. Phthalates are selected as a chemical class of interest because they are ubiquitous in personal care products (e.g., cosmetics, fragrances, shampoos) and indoor residential environments (e.g., polyvinyl chloride (PVC) flooring and plastics, children's toys, vinyl tiles, shower curtains), have been shown to influence sex steroids critical n early brain development, and have neuro- developmental toxicity in infants and children. A recent epidemiologic study also found that having PVC flooring material (a source of airborne phthalates) in a parent's or child's room was associated with an increased risk of ASD. The overall goal of this project is to determine whether exposure to phthalates during pregnancy is associated with an increased risk of ASD. To estimate prenatal exposure to phthalates, this study will leverage existing resources from a unique longitudinal study initiated under the NIEHS-funded UC Davis Center for Children's Environmental Health known as "MARBLES" (Markers of Autism Risk in Babies - Learning Early Signs). MARBLES is a prospective investigation, enrolling pregnant women who already have a child with ASD and are therefore at high risk for delivering another child with autism, designed to identify causes and early markers of autism. From this study, we have available multiple urine samples collected from the mother throughout pregnancy, a key feature for improving the exposure estimates for each infant and mother due to the fact that phthalates are metabolized and excreted quickly, with elimination half-lives on the order of hours. Moreover, two hundred fifteen of the children from this pregnancy cohort will have been assessed for autism at the age of 36 months during the time of this R21, enriching the value of these prenatal biological samples. Close to one in five of the first 147 children to reach 36 months has been confirmed with a diagnosis of ASD. Other developmental diagnoses, including adaptive function and cognitive development, are also available for these children. Therefore, multiple gestational urine samples collected from mothers and confirmed diagnoses of autism in MARBLES are invaluable resources that will enable rigorous analyses of the association between prenatal exposure to phthalates and the risk for ASD or other developmental concerns. Up to four urine samples were collected each trimester for each woman. To reduce analytical costs while maintaining some information about variability, we will analyze the first sample collected each trimester as an individual sample and pool all remaining samples for that trimester. To assess variability of maternal exposures during pregnancy, longitudinal individual urine samples collected approximately one week apart from 9 mothers will be analyzed for phthalate metabolite levels. This study will markedly advance understanding of the role of a class of endocrine disrupting compounds common in consumer products on ASD.

This project seeks to determine if prenatal air pollution exposures increase the risk of cognitive delays and autistic traits. Emerging evidence suggests that air pollutant exposure may increase risk of neurodevelopmental disorders and autism spectrum disorder (ASD). Our research demonstrates that exposure during gestation and early life to traffic related air pollution (TRP) and regional particulate matter (PM) were associated with risk of autism and neurodevelopmental delays. Other studies indicate neurodevelopmental effects based on prenatal biomarker measures of polycyclic aromatic hydrocarbon (PAH) exposure. However, we do not know if these associations persist over time, nor has a critical period of development been pinpointed. In this study we propose to examine the relationship between prenatal air pollution exposure and early longitudinal measures of cognitive ability and ASD in two prospective samples - Markers of Autism Risk in Babies, Learning Early Signs (MARBLES) and Early Autism Risk Longitudinal Investigation (EARLI). MARBLES and EARLI are both longitudinal studies of pregnant mothers, who have had one child with ASD, increasing the risk for abnormal development and ASD in the subsequent child. Approximately 40% of the infant children from MARBLES and EARLI will have a DD and 1/7 will be diagnosed with an ASD, which enriches our ability to study a broad range of neurodevelopmental outcomes. We will leverage the valuable phenotype and biospecimen resources of MARBLES and EARLI for the first prospective study of prenatal air pollution effects on autistic traits and the trajectory of cognitive development over the first thirty-six months of life. Specific aims are: (1) to assign air pollutant exposure using state of the art modeling techniques for multi- site studies estimating TRP and PM as well as measure novel biomarker measurements of nitro-PAH exposure to freeway-based diesel exhaust and gaseous nitro-PAH pollutants from vehicular and other sources; (2) examine the effect of these exposures on the trajectory of cognitive development using repeated administrations of the Mullen Scales of Early Learning (MSEL); and (3) evaluate the effect of these exposures on autistic traits and ASD diagnoses. Because air pollutant exposure is common and can be mitigated, the potential public health impact of this study is large. Additionally, identification of a risk factor in very young children provides the opportunity for early intervention where reduction of risk for disordered development is still possible.

Project 1, Epidemiology & Environment, will leverage existing resources from two large epidemiologic studies to address the potential contribution from several common household exposures to risk for an autism spectrum disorder (ASD), separately and in combination with certain genomic or epigenetic profiles. The two existing investigations are: the population-based case-control CHARGE (Childhood Autism Risk from Genetics and Environment) Study, and MARBLES (Markers of Autism Risk in Babies Learning Early Signs), a cohort study following pregnant women who previously delivered a child that developed autism to understand what influences the outcome of the younger sibling and to identify early markers of ASD. Chemical classes of interest for Project 1 are: polybrominated diphenyl ethers (PBDEs, used as flame retardants), their hydroxylated metabolites, and pyrethroid insecticides (used in sprays and foggers to control ants, cockroaches, flies and mosquitos, and on pets to control fleas). Chemical determinations made in Core C in various media (plasma, urine, breastmilk) will be supplemented by toxicokinetic modeling and exposures assessed by questionnaire and other data sources. These will be examined in relation to child's developmental status, i.e., ASD, developmental delay (DD), specific speech /language delay, and other trajectories, and for associations with markers of immune function generated in Project 3, and epigenetic markers from Project 2. Differential impact of the PBDEs and pyrethroids will be evaluated based on relevant genetic polymorphisms, CNVs, or measures of global DNA methylation, considering mechanistic pathways that link to these compounds. Finally, with the bio-informatics team at Pennsylvania State University, we will explore, in discovery-oriented mode, a wide array of exposures from biologic specimens, interviews, and medical records, along with the genetic and epigenetic data.

Epigenetic mechanisms act at the interface of genetic and environmental risk factors in autism. Project 2 is designed to investigate the epigenetic mark of DNA methylation, as environmental toxins have been demonstrated to reduce global DNA methylation levels while methyl-donor nutrients can be protective. This project will make use primarily of human cord blood samples from the MARBLES study in order to test the hypothesis that epigenetic patterns laid down in early life that regulate synapse maturation and immune responses will be impaired in autism through interactions between genetic and environmental factors. The first aim is designed to perform a genome-wide analysis of DNA methylation and copy number variation and to study the association of differences in genetics and epigenetics with environmental exposures (from Project 1 and Core C) and nutrients. The second aim will investigate methylation of a specific gene locus, FOXP3, as an epigenetic marker of regulatory T cells and will make use of immunology expertise and existing participant samples from both MARBLES and CHARGE from Project 3. The third aim will test a multifactorial mechanistic model of transcription-induced epigenetic memory of perinatal gene x environment interactions at two specific loci, F0XP3 and FMR1, through interactions with Projects 3 and 4. Together these studies will increase understanding of the epigenetic interface between genetic and environmental risk factors in autism, leading to improved diagnosis, prevention, and therapies.

Although autism spectrum disorder (ASD) primarily affects brain function, our data have also identified widespread changes in the immune system of children with autism (AU), both at the systemic and cellular levels. Characterization of the relationship between the immune and neuronal systems and their synergy with respect to environmental exposure is key to understanding the mechanisms through which toxicants can alter neurodevelopment. Ca2+ dependent signaling, for example through the mTOR pathway, provides a denominator that is common to both the neural and immune systems. Our data converges on specific neuro and immune-modulatory effects, implicating mTOR pathways, following ex vivo exposure of cells to congeners of PBDE (polybrominated diphenyl ethers). Project 3 hypothesizes that the maternal gestational body burden of non-coplanar environmental toxicants such as PBDE will correlate to immune dysregulation. Furthermore, we hypothesize that children with AU will exhibit increased sensitivity to PBDE exposure. We propose that increased PBDE body burden will lead to changes in the profiles of cytokines/chemokines production and that ex vivo toxicant exposure of peripheral blood mononuclear cells collected during gestation from mothers who give birth to an ASD child will be exaggerated. We further propose that children with ASD will have increased sensitivity to ex vivo exposure to PBDE leading to differentially altered immune cell function that will correlate with the altered expression of specific mTOR pathway related genes. In particular, we hypothesize that alterations in mTOR signaling will affect regulation of immune responses including changes in DNA methylation of F0XP3 expressed in regulatory T cells (Project 2). In conjunction with Project 4, we propose that there is a direct relationship between cytokine/chemokine profiles and changes in neuronal development. We have formulated these hypotheses on the basis of our published work and preliminary data demonstrating that PMBC from children with ASD respond differentially to ex vivo PBDE exposure. We will: 1) Examine the maternal gestational environment by leveraging samples taken during each trimester from mothers enrolled in the MARBLES (Markers of Autism Risk in Babies - Learning Early Signs) Study; 2) Examine the maternal gestational environment by using samples taken during each trimester from mothers enrolled in the MARBLES Study; and 3) Determine F0XP3 and global methylation on DNA from existing samples of PBDE-exposed peripheral blood mononuclear cells (CHARGE) to determine if methylation differences are reflective of differential cell function including cytokine/chemokine production.

Despite the fact that little is known about non-genetic causes of autism spectrum disorders (ASD), currently, few rigorous investigations of environmental factors in the etiology of this condition are underway. This proposal extends the epidemiology project initiated under the NIEHS-funded UC Davis Center for Children's Environmental Health (CCEH) known as "MARBLES" (Markers of Autism Risk in Babies - Learning Early Signs). MARBLES has enrolled close to 200 pregnant women who already have a child with ASD and therefore are at high risk for delivering an infant who will also develop ASD. In this R01, we will recruit an additional 250 pregnant mothers and follow their pregnancy and their child. Through longitudinal collection of 1) extensive behavioral, medical, and exposure data, 2) biologic specimens from the mother and child, and 3) detailed psychometric assessments from birth to three years of age, MARBLES has created an infrastructure for addressing etiologic questions and searching for early pathophysiologic markers. The exposures of interest are two classes of compounds common in household products, and having known neuro- or neurodevelopmental toxicity: pyrethroid pesticides and polybrominated diphenyl ethers (flame retardants). First, we will assess associations of self-reported exposure, measurements of internal dose, and toxicologically-derived estimates of biologically effective dose, on the one hand, with risk for ASD or other impairments in neurobehavioral development on the other. Secondly, we will examine whether the exposure or dose estimates are associated with markers of aberrant immune responses or mitochondrial dysfunction and whether these markers predict clinically confirmed child developmental status at three years of age. Thus, this project begins with the macro-level associations typical of black-box epidemiology, refines the estimates of dose, and then explores more deeply into the mechanisms by which environmental exposures might alter neurodevelopment and lead to clinical outcomes. This integrated approach is made possible by an interdisciplinary team that brings together molecular and basic research, epidemiologic population-based approaches, and clinical sciences. If our hypotheses are supported, this will be the first evidence of higher autism risk based on prospective measurements of compounds in commonly used household products, and one of the first to identify modifiable risk factors. The results will thereby open the door to prevention at the individual behavioral and societal level, e.g., potentially through education, product labeling, and/or regulatory action. Results may also lead to targeted interventions to alter children's developmental trajectory. Overall, we expect this study to stimulate a paradigm shift in the field, towards a more multifactorial and mechanistically driven research agenda, with significantly more attention to environmental exposures and the development of inter-disciplinary approaches.