The effect of developmental iron deficiency on gene expression, TET proteins, and DNA hydroxymethylation in the rodent brain A DISSERTATION SUBMITTED TO THE FACULTY OF THE UNIVERSITY OF MINNESOTA BY Amanda Kathryn Barks IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Michael K. Georgieff, MD, Advisor June 2020 © Amanda Kathryn Barks, 2020 i Acknowledgements Thank you to my advisor, Dr. Michael Georgieff, for modeling for me how to be a successful physician scientist. Your calm, positive, and enthusiastic mentorship has helped to keep me (relatively) calm, positive, and enthusiastic through the challenges of graduate school. Thank you for teaching me and guiding me through my PhD, while also allowing me the academic and intellectual freedom to grow as an independent investigator. I am incredibly grateful for all of the mentorship and guidance that you have given me ever since I first set foot at University of Minnesota during my interview. Thank you to my co-mentor, Dr. Phu Tran, for guiding me through the lab, and for always making time for my many day-to-day questions. You have been a model of creativity and enthusiasm for science and I have greatly enjoyed our many wide-ranging conversations about science over the last five years. I would also like to thank my thesis committee members Dr. Yasushi Nakagawa, Dr. Lorene Lanier, and Dr. Timothy Hallstrom for their thoughtful guidance throughout my PhD training. Thank you to the members of the Gisslen, Rao, Paulsen, Ingolfsland, Satrom, and Bastian Neonatology labs. I feel very lucky to have had the opportunity to work with and learn from all of you in such a collaborative and supportive research environment. Special thanks to Montana Beeson, Garima Singh, and Dee Kulhanek for being not only wonderful coworkers, but wonderful friends. Thank you to the Graduate Program in Neuroscience and the Medical Scientist Training Program of the University of Minnesota for providing me with an amazing training environment. Special thanks to the MSTP entering class of 2013 and the GPN entering class of 2015 for your friendship and support. Finally, thank you to my family for supporting me over the last seven years and throughout my life. To my parents, Karen and John Barks, thank you for providing me with the best education, the best opportunities, and the best support always. To my sister, Rebecca, thank you for always teaching me about the things that you love, it always makes my life better. ii Abstract Fetal-neonatal iron deficiency (ID) has a lasting negative impact on neurodevelopment, resulting in significant cognitive, socio-emotional, and learning and memory deficits in adulthood, as well as increased risk for neuropsychiatric disease. Given that ID is the most common micronutrient deficiency worldwide, and that pregnant women and young children are disproportionately affected, it presents a significant public health concern. Preclinical models have demonstrated that the developing central nervous system (CNS) is particularly affected by ID, and that the deleterious neurodevelopmental effects and neuropsychiatric risks that follow are associated with dysregulation of CNS gene expression. Dysregulated genes map to signaling pathways and networks critical for neurodevelopment and neuronal function, suggesting that these critical functions are compromised by ID. If developmental ID is corrected by iron repletion within a critical period, correction of neurodevelopmental deficits is possible. However, if iron repletion occurs outside of the critical period, the phenotypic and gene expression changes persist into adulthood despite correction of the deficiency. While changes in gene expression can be understood as the proximate cause of the ID neurocognitive phenotype, it is still unclear what the ultimate cause is. As such, there is a gap in our understanding of how developmental ID establishes and maintains gene expression changes in the CNS. A potential mechanism by which iron could enact these changes is through Ten-Eleven Translocation (TET) enzymes, a family of iron-dependent hydroxylases that generate the epigenetic modification 5-hydroxymethylcytosine (5hmC), or DNA hydroxymethylation. Epigenetic modifications such as DNA hydroxymethylation have the ability to stably influence gene expression throughout the lifespan, and are known to iii be labile to environmental influences. Of particular relevance, 5hmC is more abundant in the brain than any other tissue, and it increases in enrichment as neurodevelopment progresses, particularly in genes critical for neuronal development and function. The central hypothesis of my thesis research is that dysregulation of TET enzymatic activity and 5hmC by fetal-neonatal ID drives gene expression changes in brain that contribute to the long-term neurocognitive phenotype of developmental ID. To test this hypothesis, the following aims were proposed: 1) Determine the effect of fetal-neonatal ID on TET activity and 5hmC in two regions of the developing rat brain, the hippocampus and the cerebellum, and 2) Determine whether treatment of developmental ID with dietary iron repletion can reverse the changes to this epigenetic system. Completion of these aims contributes to the long-term goal of understanding the cellular and molecular underpinnings of CNS dysfunction and increased neuropsychiatric disease risk following developmental ID. Because the standard therapy of iron repletion incompletely rescues the neurodevelopmental phenotype of ID, there is a need for better therapeutic options. By better understanding the underlying mechanisms of ID-related hippocampal dysfunction, it may be possible to identify new therapeutic targets for more effective treatment of iron deficiency. iv Table of Contents Acknowledgements i Abstract ii Table of Contents iv List of Tables v List of Figures vi List of Abbreviations vii Chapter 1: Introduction 1 Chapter 2: Methods 19 2.1. Methods for Chapter 3 19 2.2. Methods Common to Chapters 4 and 5 23 Chapter 3: Early Life Neuronal-Specific Iron Deficiency Alters the Adult Mouse Hippocampal Transcriptome 26 Chapter 4: Fetal-Neonatal ID Increases TET activity and DNA hydroxymethylation in the developing rat hippocampus 43 Chapter 5: Fetal-Neonatal ID Reduces TET activity and DNA hydroxymethylation in the developing rat cerebellum 56 Chapter 6: Discussion and Conclusion 67 References 79 v List of Tables Table 3.1. Top dysregulated diseases and functions in the adult chronically iron deficient mouse hippocampus 31 Table 3.2. Top dysregulated diseases and functions in the adult formerly iron deficient mouse hippocampus 34 Table 3.3. RT-qPCR validation of CREB1-regulated target genes in chronically and formerly iron deficient mouse hippocampus 38 vi List of Figures Figure 1.1. Iron-dependent TETs influence the epigenetic modifications DNA methylation and DNA hydroxymethylation. 16 Figure 3.1. Gene dysregulation in the chronically- and formerly iron deficient adult hippocampus. 30 Figure 3.2. Critical neuronal signaling pathways are altered by ID and rescued by its reversal at P21. 32 Figure 3.3. Genes permanently dysregulated by ID are involved in neurocognitive function and dysfunction. 35 Figure 3.4. CREB1 signaling is predicted to be altered by ID. 37 Figure 4.1. Characterization of iron status in P15 rats. 47 Figure 4.2. Hippocampal TET activity increases with decreasing iron status. 49 Figure 4.3. Hippocampal Tet expression increases with decreasing tissue iron status. 50 Figure 4.4. Global %5hmC is increased in the IDA hippocampus. 52 Figure 5.1. Characterization of cerebellar iron status at P15. 59 Figure 5.2. Cerebellar TET activity is reduced by developmental IDA. 61 Figure 5.3. Cerebellar %5hmC is modestly reduced by developmental IDA. 62 Figure 6.1. Genes enriched for 5hmC are differentially expressed in the CID hippocampus. 69 Figure 6.2. Model summarizing the effect of developmental ID on TET activity and DNA hydroxymethylation. 71 vii List of Abbreviations BDNF: Brain-derived neurotrophic factor CID: Chronically iron deficient CNS: Central nervous system CREB: cAMP response element-binding protein DN: Dominant negative DN-TFR1: Dominant negative transferrin receptor DOHaD: Developmental origins of health and disease Dox: Doxycycline FID: Formerly iron deficient ID: Iron deficiency IDA: Iron deficient anemic IS: Iron sufficient TIDA: Treated iron deficient anemic WT: Wildtype 1 Chapter 1: Introduction Text contains excerpts from: Barks A, Hall AM, Tran PV, Georgieff MK. Iron as a model nutrient for understanding the nutritional origins of neuropsychiatric disease. Pediatric Research. 2019;85:176-182. PMCID: PMC6353667. Early Life Nutrition and Risk of Neuropsychiatric Disorders Neurodevelopment is a complex process that involves not only establishment and expansion of neuronal and glial cell populations, but maturation of complex cell morphologies and establishment of appropriate connectivity among cells and among brain regions. The successful construction of these neural circuits during development is critical for the ability of the individual to perform complex behaviors. Increasing evidence supports the hypothesis that disruption of these processes during early development can result in neurocognitive and neuropsychiatric dysfunctions that appear later in life and can last across the lifespan (1). Although much of neurodevelopment proceeds according to a relatively hard-wired, time-locked, experience-independent genetic program, environmental experiences and exposures exert significant influences on neurodevelopment through the complex interplay between genes and the environment. Nutrition is one of the major environmental factors that influences early neurodevelopment. Although almost all nutrients are needed for normal brain development, a subset of nutrients plays a particularly significant role because they support the high rate of brain metabolism during late fetal and early postnatal life (2). These include macronutrients such as protein, long chain polyunsaturated fatty acids (LC- PUFAs), and glucose, as well as select micronutrients such as iron, zinc, and vitamins, which are involved in a variety of critical neurodevelopmental processes across brain 2 regions (Reviewed in detail in 3–5). Moreover, while it has been well established that fetal nutrition plays a role acutely in fetal development and pregnancy outcomes (6), it is increasingly understood that fetal nutrition also exerts long-term effects on offspring health and disease risk – including brain health and disease – into adulthood and throughout the lifespan. This growing body of both clinical and preclinical literature supports the hypothesis that fetal and early postnatal macro- and micro-nutritional status is linked to the risk of psychopathology later in life (7–9). Cohort studies of historical periods of famine, such as the Dutch Hunger Winter and the Chinese Famine, have provided foundational evidence for the role of fetal and early postnatal undernutrition in the etiology of neuropsychiatric disease later in life. These studies demonstrate that inadequate nutrition during fetal development is associated with an increased risk of schizophrenia and affective disorders, among individuals prenatally exposed to famine (10–12). Studies of preclinical models have helped to elucidate the specific roles and mechanisms of individual macro- and micronutrient deficiencies that are particularly associated with the development of neuropsychiatric pathologies. This broad framework constitutes the conceptual basis of my thesis research on the neurodevelopmental effects fetal-neonatal iron deficiency. The following sections will review the epidemiological, clinical, and preclinical studies that underlie our understanding of this framework, and lay the foundation for my thesis research. 3 The Developmental Origins of Health and Disease (DOHaD) Hypothesis The Developmental Origins of Health and Disease, or DOHaD, hypothesis posits that the early-life environment, particularly the fetal and early postnatal environment, influences later-life health outcomes and disease risks. The mechanism by which this effect occurs is the subject of considerable research effort. Originally the DOHaD hypothesis was applied to risk for adult cardiovascular disease (13,14). However, it is now understood that a wide range of early life exposures and conditions, including malnutrition, can affect lifetime health and disease risk across multiple organ systems. One of the classic examples with respect to this nutritional deprivation hypothesis comes from cohort studies of individuals who were exposed to gestational malnutrition during the Dutch Famine of 1944-1945. During this 5-month period, daily maternal rations fell to below 1,000 calories. As a result, pregnant women and their fetuses were exposed to famine during early, mid-, or late gestation (15). These cohort studies demonstrated that gestational exposure to famine was associated with a variety of adverse health outcomes in adulthood, including obstructive airway disease (16), impaired glucose tolerance (17), dyslipidemia (18), coronary heart disease (19), and schizophrenia (10,11). Interestingly, famine during different periods of gestation was associated with distinct health risks, suggesting that discrete critical periods exist during the fetal period for different organ systems and health outcomes. These observations also suggested that disruption of nutritional status during critical periods results in permanent compromise of organ systems and related health outcomes. 4 Nutrition and Critical Periods of Neurodevelopment A critical period in organ system development is characterized as a window of time in which there is a high degree of plasticity and thus increased malleability in response to environmental stimuli and exposures, including nutrition (20). Consequently, during a critical period of development, presence or absence of normal stimuli and exposures can have an enhanced and irreversible effect on how an organ system develops. With respect to nutrition, critical periods demarcate time windows in which a tissue is most sensitive to the presence or absence of a specific nutrient. Generally, a critical period coincides with the time when a given nutrient is in highest demand - often during a period of rapid tissue growth and development. Because tissue growth and development are highly metabolic processes, and sufficient nutrient supply is critical for energy and metabolism, tissues are most likely to suffer detrimental effects from insufficient supply during this window. Each tissue and cell type develops on a different time scale and has different metabolic and nutritional requirements for appropriate development (20,21), and so an exposure at a given time point may affect one tissue or function but not another. As such, there is no single critical period during development; rather, there are a series of nutrient-, tissue-, and cell type-specific critical periods throughout the broad time-span of organ system development (20). Relative to other primates, humans have larger and more highly metabolic brains than would be predicted based on average body size. The human brain has a higher energy demand per unit weight than any other tissue. In the adult, the brain accounts for 20 to 25% of basal energy supply despite comprising only 2% of total body mass (22). The developing fetal and early postnatal brain has an even higher metabolic demand than 5 the adult brain, accounting for as much as 60% of the total metabolic rate (22). As such, the developing brain is exquisitely sensitive to nutrient supply. The brain does not develop as a uniform tissue. Rather, each region develops on a unique timeline, with unique timing of nutrient demands, resulting in a series of brain region-specific critical periods. Thus, the effects of insufficient nutrient supply on the developing brain vary depending on the nutrient, the duration of the deficiency, and the neurodevelopmental timing of the deficiency relative to the various regional critical periods (23). Failure to construct a brain region during its critical period can lead to residual structural and connectivity defects, persistent neurochemical and electrophysiological abnormalities, or permanent dysregulation of gene expression. Such disruption by a nutritional deficiency ultimately manifests as functional deficits, including increased risk for neuropsychiatric disease (24). Iron deficiency (ID) serves as a particularly useful paradigm for understanding the effects of nutritional deficiency on neurodevelopment and neuropsychiatric disease risk, due to its extensive clinical and preclinical literature (25). ID has been implicated in multiple studies as a risk factor for developmentally-based psychopathologies in humans (26–28) and iron’s biology of action has been well-studied. The following sections will review in depth the clinical and epidemiological evidence linking ID to neuropsychiatric dysfunction, and review what is known about the structural, cellular, and molecular mechanisms that underlie its lasting effects. 6 Clinical Studies Linking Early-Life Iron Deficiency and Risk of Psychiatric Diseases ID is the most common micronutrient deficiency worldwide, affecting an estimated 2 billion people and 40-50% of pregnant women and preschool-aged children (29,30). While ID is most prevalent in developing countries, it is notable among nutrient deficiencies for its prevalence in industrialized countries, including the United States, where 18% of pregnant women and 14% of 1-2 year-olds are affected (31,32). Though ID is ubiquitous across age groups, ID in pregnant women and young children is of particular concern due to the well-established relationship between developmental ID and later-life neurologic dysfunction. Longitudinal cohort studies of formerly ID (FID) individuals who experienced iron deficiency during infancy have provided detailed information about the specific cognitive functions that are impaired by developmental iron deficiency. At 5 years of age, FID children exhibited slower perceptual speed and poorer understanding of quantitative concepts (33) and impaired language abilities (34). In early adolescence, FID individuals were characterized as having slower perceptual speed, poorer spatial memory, and developmental delays in a selective attention task (33). These impairments may contribute to the lower scores of FID adolescents on tests of reading and arithmetic as well as the increased likelihood to repeat a grade or receive referrals for tutoring (35). As young adults, FID individuals performed worse on measures of recognition memory and strategy shifting (36). Finally, outcome measures of employment and education at 25 years of age suggested that the cognitive deficits experienced throughout infancy, childhood, and adolescence had a significant impact on the lives of FID adults (37). 7 Compared with non-ID individuals, a larger proportion of FID adults did not complete secondary school and were not actively pursuing further training or education (37). Likewise, ID infants and FID children and adults exhibit long-lasting socioemotional and affective deficits. At ages 4 to 5, FID children exhibited reduced positive affect during an interactive task with their mother (38) or a stranger (39) and were found to be more passive and unengaged than iron sufficient children (39,40). FID children also displayed poorer self-control on a delayed-gratification task (39). As early adolescents, FID individuals received higher ratings from parents or teachers on measures of anxiety and depression, social problems, delinquent behavior, and aggression (35). Finally, FID adults self-reported more feelings of detachment and negative emotions and had lower ratings for emotional health (37). Separate cohort studies have used measures of maternal iron intake and maternal iron deficiency during pregnancy to assess the effects of iron deficiency during fetal development on long-term neurodevelopmental outcomes of offspring. These studies have found a significant association between maternal iron deficiency and increased risk of schizophrenia spectrum disorders in adulthood (26,28), as well as an association between low maternal iron intake and increased risk of autism spectrum disorders (27). These findings were recently corroborated by a large-scale cohort study of >500,000 individuals where it was found that prevalence of autism spectrum disorders, attention- deficit/hyperactivity disorder (ADHD), and intellectual disability was increased in children of mothers diagnosed with anemia during the first 30 weeks of pregnancy (41). Notably, prevalence of these disorders was not increased when mother’s diagnosis of anemia occurred after 30 weeks of pregnancy. This study suggests that 8 neurodevelopmental events during the first two trimesters are perhaps most critical for development of these neuropsychiatric disorders, and that these neurodevelopmental processes are highly sensitive to or dependent on iron. Collectively, these human cohort studies point to a specific set of neurocognitive deficits and neuropsychiatric risks that arise from fetal and early postnatal iron deficiency, implicating a critical role for iron in the normal neurodevelopment. Biological Bases for the Long-term Effects of Early-life Iron Deficiency Understanding the potential mechanisms by which early-life ID influences neurodevelopment and subsequent risk for neuropsychiatric disease is critical for developing effective prevention and treatment strategies. The two central hypotheses for these effects are residual structural defects and dysregulation of genes involved in neuronal function. In the residual structural defects hypothesis, an early-life nutritional exposure, such as ID, during a critical period of development results in aberrant structural development ranging from gross structural abnormalities to fine ultrastructural changes (20,42,43). In the gene dysregulation hypothesis, the early-life nutritional environment aberrantly programs gene expression, which in turn stably alters the subsequent phenotype (44–48). Though these can be posed as two separate hypotheses, in fact, it is likely a combination of both, and their influence on each other, that drives aberrant neurodevelopment and lifetime risk for neuropsychiatric dysfunction. 9 Residual Structural Defects Rodent and porcine models of fetal-neonatal ID have demonstrated gross morphological changes to both grey and white matter in the brains of developmentally iron deficient individuals. In a porcine model, developmental ID during the first 30 days of life (approximately equivalent to the first 4 months of human postnatal development) results in a decreased total brain volume, with specific decreases in several brain regions including cortex, cerebellum, and hippocampus (49). This is accompanied by decreased total white matter volume and integrity (50). Rodent models of early life ID corroborate these effects on both hippocampal volume (51,52) and white matter volume/integrity (53). Importantly, these changes are partially reversible: iron restoration starting at 30 days of life is sufficient to normalize gross brain volume after only 30 days of iron repletion in the porcine model. However, disorganization of brain structure persists despite normalization of total brain volume, with remaining deficits in hippocampal volume and white matter organization/integrity (49). Together, these results indicate that while iron repletion may correct gross morphological measures such as total brain volume, ultrastructural changes to grey and white matter may persist beyond the period of iron deficiency and into adulthood. Rodent models, as well as in vitro primary culture models of ID have been critical in determining the ultrastructural effects of developmental ID, particularly on neurons. The neuronal ultrastructural effects of developmental ID were first characterized using a rat model of fetal-neonatal ID in which animals experience iron deficiency throughout gestation and early postnatal life. At postnatal day (P) 15 – the peak of hippocampal dendritogenesis – ID rats had truncated apical dendrites in hippocampal area CA1 (42). 10 Additionally, branches in the dendritic arbor of hippocampal neurons occurred more proximally to the soma and dendritic spine heads were smaller (54). These hippocampal neuronal findings have been corroborated using mouse models of non-anemic ID (43,55) and primary neuronal culture models of ID (56). In the rat model of ID, these hippocampal neuronal ultrastructural deficits persisted into adulthood despite early-life iron repletion (42,54). Further refinement of iron repletion experiments determined that there was a critical period for iron in hippocampal neurodevelopment. If iron was restored within the critical period, then adult hippocampal neuronal structure was normalized; however, if iron repletion occurred outside of the critical period, then hippocampal neuronal structure remained permanently altered (43). Increasing evidence shows that cortical neurons are also affected by developmental ID. Ultrastructural changes to both apical and basal dendritic arbors have been characterized in cortical neurons in rodent models of ID, primarily manifesting as decreased branching of dendrites proximal to the cell body (57). Correspondingly, in a study of infants of mothers with low iron intake during pregnancy, cortical grey matter fractional anisotropy (FA) was increased in infants with low maternal iron intake (58). In cortical gray matter, FA decreases as neurodevelopment progresses and dendritic arborization increases (59,60). Thus, the increased FA seen in infants of iron deficient mothers supports the preclinical model findings of compromised cortical neuron structure. Neuronal morphologic development is a highly metabolically demanding process that is dependent on adequate availability of energy substrates for development of complex cell morphology (61,62). Production of the metabolic substrate ATP by 11 mitochondria is iron-dependent, as several key enzymes in the electron transport chain are mechanistically dependent on iron for their enzymatic activity (63). Increasing evidence suggests that mitochondrial function of developing neurons, as indexed by measures such as basal respiration, ATP production, and glycolytic capacity, is compromised by developmental ID (56). Additionally, motility, size, and density of mitochondria are decreased in terminal dendrites of developing iron deficient neurons (64). These findings implicate mitochondrial dysfunction as a mechanism contributing to the above-described neuronal ultrastructural defects associated with developmental ID. Rodent models have also been critical in determining the ultrastructural effects of developmental ID on white matter. Oligodendrocytes, the myelinating cells of the central nervous system, stain more densely for intracellular iron than other cell types in the central nervous system (65), and the peak of iron uptake in the developing brain corresponds to the peak period of myelination (53,66), suggesting that adequate iron supply is important for their development and function. In both in vitro and in vivo models, oligodendrocyte precursor cell (OPC) differentiation to oligodendrocytes is impaired when iron uptake is impaired (67,68). The diameter of myelinated axons is significantly decreased in developmentally ID rats, even in adulthood, suggesting that iron is critical for establishment of myelination (57). This has been noted to correspond functionally with decreased conduction velocity (69). In order to understand the relationship between developmental ID and complex psychopathologies, it is important to understand not only its effects on individual brain regions and cell types, but also its effects on the developmentally driven interconnection of multiple brain regions and cell types. Because each brain area develops on a different 12 time scale, critical periods differ between and even within brain areas (21,23), and an exposure at a given time point in development may affect one brain region or physiologic function but not another. This non-uniform effect on the developing brain can disrupt the connectivity among brain regions that work together in complex circuits that are critical for higher-order behaviors in adulthood. For example, proper functional control of the ventral tegmental area (VTA) loop depends on the balance between tonic hippocampal input and intermittent frontal lobe input. Disruption of this balance is postulated to result in schizophrenia (70), a psychopathology that presents later in life but that has distinct developmental origins. Within the loop, the hippocampus develops earlier and more rapidly than the frontal cortex during late fetal/early neonatal life, and thus is far more vulnerable to the adverse effects of fetal deficiencies of nutrients such as iron that support rapid growth. This selective vulnerability of the hippocampus (relative to frontal cortex) to developmental iron deficiency has the potential to disrupt VTA circuit structure and function by unbalancing the inputs. Further contributing to the unbalance of VTA loop connectivity, the dopaminergic system, a key neurotransmitter system in VTA loop, is disrupted by ID due to the iron-dependence of enzymes in the dopamine synthesis pathway (71). Pre- clinical models of fetal/neonatal iron deficiency support this hypothesis (72,73). Emerging evidence in human cohorts also supports the hypothesis that functional connectivity of brain regions is disrupted by ID. A functional magnetic resonance imaging (fMRI) study of developmentally ID individuals in adulthood found significant changes in connectivity of the Default Mode Network (DMN), a major network of interconnected regions in the brain (74). Connectivity of the posterior cingulate cortex 13 within the DMN was particularly affected by ID (74). Similar disruptions to the DMN are seen in a variety of neurocognitive and neuropsychiatric disorders (75). Gene Dysregulation Widespread hippocampal gene dysregulation has been shown in both rodent and porcine models of developmental ID, both acutely (76,77) and in adulthood after complete resolution of ID (47,48). The dysregulated genes map to functions critical for neurodevelopment as well as neuropsychiatric disorders including mood disorders, pervasive developmental disorders, autism, and schizophrenia (47,48,76,77). These widespread and stable long-term alterations to gene expression implicate an underlying transcriptional regulatory mechanism that functions on a genome-wide scale, such as epigenetic modification. Epigenetics is the study of how supragenomic modifications to DNA and chromatin can alter gene expression and resultant phenotype, independent of alterations to the genome itself. Two major classes of epigenetic modification are histone modifications and cytosine modifications (including DNA methylation and DNA hydroxymethylation). Each of these classes of epigenetic modification is associated with distinct effects on transcriptional regulation. Additionally, all have been demonstrated to be environmentally labile, meaning that their prevalence and distribution in the genome can be altered, sometimes permanently, in response to environmental conditions and exposures such as stress (78,79), toxicants (80–82), and nutrition (44–46,83–85). Thus, epigenetic modification presents a plausible mechanism by which the environment, 14 including nutritional status, can exert an effect on central nervous system gene expression and subsequent neurodevelopmental and neuropsychiatric outcomes. Histone modifications are covalent modifications to the amino acid residues of histone tails, and include histone acetylation, methylation, phosphorylation, and ubiquitination. The histone code is complex and has not been studied in depth in the context of developmental ID. However, one family of histone modifications, lysine methylation, has been of particular interest due to its mechanistic dependence on iron. Removal of methyl groups from lysine residues is catalyzed by the JmjC ARID-domain containing histone demethylase (JARID) family proteins, which have an absolute requirement for iron for their enzymatic activity (86,87). Accordingly, in the adult, formerly iron deficient rat hippocampus, enrichment of histone methylation is significantly altered at the dysregulated Bdnf-IV promoter (84). H3K27me3 and H3K4me3, histone modifications associated with silenced genes, were found to be enriched at the Bdnf-IV promoter, while H3K4me3, a modification associated with active gene expression, was significantly depleted (84). Interestingly, properly timed dietary supplementation with the methyl donor choline mitigates the effects of ID on histone methylation in the hippocampus of preclinical models (84). This finding underscores the potential for dietary epigenetic “workarounds” in humans in the future. DNA methylation is an epigenetic modification generated by covalent addition of a methyl group to cytosine nucleotides, primarily at CpG dinucleotides. It is generally associated with gene silencing, particularly when present at promoter regions, and closed- state chromatin (88). On a genome-wide scale, the hippocampal DNA methylation pattern is significantly altered by developmental ID in a porcine model, with 853 15 differentially methylated CpG sites identified in the developmentally iron deficient hippocampus compared to controls (77). However, relatively few of these differentially methylated CpG sites map to differentially expressed genes (77). Similarly, in a rat model, 229 differentially methylated loci were identified in the developmentally iron deficient hippocampus, though again few of these regions mapped to known differentially expressed genes (89). Thus, while these studies clearly demonstrate altered patterns of DNA methylation in the developmentally ID brain, they leave open the question of the functional significance of these changes. DNA methylation has also been assessed at the gene-specific level at the brain derived neurotrophic factor (Bdnf) promoter. Bdnf, a gene critical for neuronal development, synaptic formation and plasticity, learning, and memory is downregulated in the hippocampus both acutely during the period of developmental ID (90), and in adulthood after complete resolution of ID (91). In adult, formerly iron deficient rats, total Bdnf-IV promoter methylation rate is significantly decreased, with 2 of 7 specific CpG sites significantly hypomethylated (84). Given that promoter hypomethylation is canonically associated with transcriptional activation (92), the functional significance of these changes at the downregulated Bdnf-IV promoter is again uncertain. Despite the evidence that developmental ID alters DNA methylation, there is no clear biological mechanism through which iron directly influences the establishment of DNA methylation. In contrast, active DNA demethylation relies on the iron-dependent family of TET methylcystosine dioxygenases which require iron for their enzymatic activity (Fig 1.1). TETs serially convert 5-methylcytosine (5mC) to 5- hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxycytosine (5caC) 16 in an iron-dependent manner (93,94). To complete the active demethylation process, 5fC and 5caC act as substrates for thymine DNA glycosylase (TDG) and the base excision repair pathway, which ultimately removes the modified cytosine base and replaces it with an unmodified cytosine (95). Thus, active DNA demethylation by iron-dependent TETs provides one plausible mechanism by which ID might influence DNA methylation status. Figure 1.1. Iron-dependent TETs influence the epigenetic modifications DNA methylation and DNA hydroxymethylation. Covalently modified cytosine bases are a major class of epigenetic modification. 5-methylcytosine (5mC), also known as DNA methylation, is generated from unmodified cytosine by the family of DNA methyltransferase (DNMT) enzymes. 5mC can be maintained long-term as a stable epigenetic modification, or it can be hydroxylated by TET methylcytosine dioxygenases to 5-hydroxymethylcytosine (5hmC), or DNA hydroxymethylation. 5hmC serves as a stable epigenetic modification in its own right. However, it also serves a dual purpose as an intermediate in the active DNA demethylation pathway, which returns modified cytosines to the unmodified state by serial modification of the base. TET enzymes require iron (Fe2+) as a cofactor for their enzymatic activity, which confers iron dependence to both the processes of active DNA demethylation and establishment of the stable epigenetic modification DNA hydroxymethylation. TDG: Thymine DNA glycosylase; BER: Base excision repair DNMT1 DNMT3A DNMT3B TET1 TET2 TET3 Fe2+ TET1 TET2 TET3 Fe2+ TET1 TET2 TET3 Fe2+ Stable Epigenetic Modification TDG/BER Active DNA Demethylation TDG/BER Stable Epigenetic Modification N H N NH2 O HO 5-Hydroxymethylcytosine (5hmC) 5-hydro eth lcytosine (5hmC) N H N NH2 O H3C 5-Methylcytosine (5mC) 5-methyl tosine (5 ) N H N NH2 O 5-Formylcytosine (5fC) O 5-for lcytosine (5fC) N H N NH2 O 5-Carboxycytosine (5caC) O HO 5-carboxycytosine (5caC) N H N NH2 O Cytosine (C) Cytosine ( ) 17 In addition to being a transient intermediate of DNA demethylation, 5hmC can also function itself as a stable epigenetic modification known as DNA hydroxymethylation (Fig 1.1) (96,97). In contrast to DNA methylation, DNA hydroxymethylation is associated with gene activation, particularly when present in the gene body (98). 5hmC is more abundant in the central nervous system than any other somatic tissue and is dynamically regulated throughout neurodevelopment, increasing in abundance in the brain as neurodevelopment progresses (99–102). This increase in 5hmC is particularly pronounced in genes critical for neurodevelopmental processes (100,102– 105). Together, this evidence suggests that 5hmC is critical in establishing proper gene regulation during neurodevelopment, and that it does so in an iron-dependent manner. While the iron-dependence of TET proteins has been clearly established (93), the effect of a clinically relevant degree of ID on TET protein activity and 5hmC in the central nervous system has never been studied. This gap in the literature has formed the basis of much of my thesis work. Aims of Dissertation My dissertation aimed to determine the global effect of iron deficiency on hippocampal gene expression, and to determine the role of a novel epigenetic system, DNA hydroxymethylation, in establishment and maintenance of these gene expression changes. Specifically, the aims were as follows. 1) In a genetic mouse model of hippocampal neuronal-specific ID, I aimed to determine the specific effects of chronic non-anemic ID on hippocampal gene expression. 18 2) Using the same mouse model, I aimed to determine whether or not these changes in gene expression were reversible with reversal of ID. 3) In a rat model of dietary ID anemia (IDA), I aimed to determine the effect of fetal neonatal ID on activity and expression of the iron-dependent TET methylcytosine dioxygenase enzymes, and subsequent global DNA hydroxymethylation in two brain regions. 4) Using the same rat model, I aimed to determine whether changes to this epigenetic system were reversible with postnatal dietary iron repletion. This work contributes to the overall understanding of the mechanisms by which fetal- neonatal ID drives its characteristic neurodevelopmental deficits and increased neuropsychiatric disease risks, with the goal of ultimately uncovering novel targets for intervention and treatment. 19 Chapter 2: Methods Text contains excerpts from: Barks A, Fretham SJB, Georgieff MK, Tran PV. Early life neuronal-specific iron deficiency alters the adult mouse hippocampal transcriptome. Journal of Nutrition. 2018;148:1521-1528. PMCID: PMC6258792. 2.1 Methods for Chapter 3 Animals Two transgenic mouse lines were used to generate litters as described in detail previously (43). Briefly, the first transgenic line carried a dominant-negative transferrin receptor (DN-TFR1) in the transgene TRE2-eGFP-DNTFR1 on a B6;CBA background. Mice from this line were mated with a second transgenic line B6;CBA-Tg(Camk2a- tTA)1Mmay/J (Jackson Laboratory). Resultant pups positive for both transgenes expressed DN-TFR1 at levels sufficient to disrupt hippocampal neuronal iron uptake, resulting in iron deficient CA1 hippocampal neurons with no effect on systemic tissue iron levels, as reported previously (43). These mice are referred to as DN. Resultant pups positive for only one or neither transgene were considered wildtype (WT) controls, because they do not express DN-TFR1 and have normal hippocampal neuronal iron uptake and storage, as reported previously (43). Upon weaning at P21, mice of both genotypes (DN and WT) were either maintained on standard non-purified diet (2018 Teklad Global 18% Protein Rodent Diet, Envigo) or switched to a doxycycline-containing diet (0.625g doxycycline/kg, TD.01306, Envigo). The doxycycline diet was nutritionally identical to the standard rodent diet, and both diets were nutritionally iron sufficient (200mg/kg). This resulted in two 20 experimental groups: chronically iron deficient (CID, DNnoDox), and formerly iron deficient (FID, DN+P21Dox). The WT doxycycline-untreated and treated groups were used as iron sufficient (IS) controls (WTnoDox and WT+P21Dox, respectively). Mice were housed in a 12-hour light, 12-hour dark cycle, with ad libitum access to food and water. All mice were maintained according to the Animal Use Policies and Guidelines outlined by the University of Minnesota Institutional Animal Care and Use Committee (IACUC). All protocols were approved by IACUC and complied with the Guide for the Care and Use of Laboratory Animals (106). Tissue Collection Adult male mice (2-9 months) were euthanized by intraperitoneal injection of Beuthanasia (10mg/kg) followed by rapid decapitation. The hippocampus was isolated and immediately flash frozen in liquid nitrogen. Tissues were stored at -80ºC until used. Preparation of RNA for Sequencing Total RNA was isolated from one hippocampal lobe per mouse using the RNAqueous RNA Isolation kit (Ambion). Four animals (n=4 biological replicates) were used per experimental group (ntotal=16). Sample size was determined based on previously described power calculations to optimize detection of differentially expressed genes (107). Isolated RNA was sent to the University of Minnesota Genomics Center for library preparation and next-generation sequencing. RNA was first quantified using the RiboGreen RNA Assay kit (Invitrogen), and assessed for quality using capillary electrophoresis (Agilent BioAnalyzer 2100, Agilent). Barcoded libraries were constructed 21 for each sample using the TruSeq RNA v2 kit (Illumina). Libraries were selected for fragments of ~200bp. Sequencing was performed using Illumina HiSeq 2500 to generate 50bp paired-end reads. Sequencing depth was >10 million reads per sample for all samples. RNA-Sequencing Data Analysis Methods Sequenced reads were subjected to quality control using FastQC in Galaxy. All samples had an average quality score >36. Following quality control, reads were aligned to the mouse reference genome (version mm10) using the TopHat package (v2.0.9) in Galaxy (108). The overall read alignment rate was >97% for all samples. Each experimental group (DNnoDox and DN+P21Dox) was compared to its age-matched control group (WTnoDox and WT+P21Dox, respectively) using CuffDiff (v2.2.1) in Galaxy (109). Differentially expressed genes were considered those with abs[log2(fold change)]>0.2 and FDR (q-value)<0.05 to account for multiple comparisons. Fisher’s exact test was performed to determine significance of the proportion of up- and down-regulated genes in experimental groups compared to their controls, with α set at p<0.05. Pathway Analysis Genes identified as differentially expressed were analyzed in Ingenuity Pathway Analysis (IPA, Qiagen) to predict canonical pathways, upstream regulators, cellular functions, and diseases altered or affected relative to controls as described previously (47). Briefly, Core Analyses were performed in IPA using the following settings: stringent filter, all data sources, direct relationships only, experimentally observed 22 confidence, mammalian species, and cells only, with cutoffs of abs[log2(fold change)]>0.2 and FDR (q-value)<0.05 to account for multiple comparisons. In IPA, statistical significance of genes mapping onto canonical pathways, molecular networks, cellular functions, and upstream regulators was determined by Fisher’s Exact test for each pathway, where a significant p-value (p<0.05, -log(p-value)>1.3) indicates that the mapping of the differentially expressed genes onto a given pathway, regulator, function, or disease was not due to chance alone. RT-qPCR RT-qPCR was performed on 5-6 animals (biological replicates) per group (ntotal=23). Sample size was chosen based on our previous reports (47,84). RNA samples used for RT-qPCR experiments were a mix of samples used in the RNA-Seq experiment and samples from different, age-matched mice. RNA was generated as described above. cDNA synthesis was performed using 1.0µg total RNA and the High Capacity RNA-to- cDNA kit (Applied Biosystems). qPCR was performed in singleplex using TaqMan Universal PCR Master Mix (Applied Biosystems) and TaqMan gene expression assays (Thermo Fisher Scientific) on a MX3000P instrument (Stratagene). TaqMan probe IDs are listed in Table 3.3. TATA-binding protein (Tbp) was used as an endogenous control (Mm01277042_m1), as expression of this gene remains unaffected by ID. Samples were run in duplicate, normalized to Tbp, and averaged to generate fold change relative to the respective control group using a standard ΔΔCt calculation. Results were analyzed by t- test, and α was set at p<0.05. 23 2.2 Methods Common to Chapters 4 and 5 Animals Timed pregnant gestational day (G)2-3 Sprague-Dawley rats (SD-400, Charles River) were fed either iron deficient (ID; 4 mg/kg Fe, TD 80396, Envigo) or iron sufficient control diet (IS; 200 mg/kg Fe, TD 01583, Envigo) upon arrival. At postnatal day (P)7, a subset of ID-fed litters were switched to IS diet, generating a treated ID anemic (TIDA) experimental group. The remaining subset of ID litters remained on ID diet until P15. This group was designated as IDA. All animals were housed in 12 hour light-dark cycle with ad libitum access to food and water, and were maintained according to the Animal Use Policies and Guidelines of the University of Minnesota Institutional Animal Care and Use Committee (IACUC). All protocols were approved by the University of Minnesota IACUC and complied with the Guide for the Care and Use of Laboratory Animals (106). Tissue Collection Rats were euthanized at P15 by rapid decapitation. Hippocampus and cerebellum were microdissected and flash frozen in liquid nitrogen. All tissues were stored at -80ºC until use. Nuclear Protein Isolation Nucleic acid-free nuclear proteins were isolated from one-half cerebellum or one hippocampal lobe per rat using the EpiQuik Nuclear Extraction Kit II (Epigentek). Nuclear protein concentration was quantified by a standard Bradford protein assay. 24 Nucleic Acid Isolation Total genomic DNA was isolated from approximately one-quarter flash frozen cerebellum or one-half flash frozen hippocampal lobe per rat using a standard phenol- chloroform-isoamyl alcohol protocol with RNAseA treatment. RNA was isolated from approximately one-quarter cerebellum or one-half hippocampal lobe per rat using the RNaqueous RNA Isolation Kit (Ambion). cDNA Synthesis and qPCR cDNA was generated from 1.0ug of RNA per sample, using the High Capacity RNA-to- cDNA kit (Applied Biosystems). qRT-PCR was performed on 5-8 biological replicates (n=5-8) per tissue, per experimental group, for both sexes. Sample size was chosen based on our previous reports. qPCR was performed in singleplex using Luna Universal Probe qPCR Master Mix (New England BioLabs), and TaqMan gene expression assays (Tfrc: Rn01474701_m1; Tet1: Rn01428192_m1; Tet2: Rn01522037_m1; Tet3: Rn01184937_g1; ThermoFisher Scientific) on a QuantStudio 3 Real-Time PCR System (ThermoFisher Scientific). All samples were run in duplicate. TATA-binding protein (Tbp: Rn01455646_m1) was used as an endogenous control. Fold change was calculated using a standard ΔΔCt calculation. TET Activity TET enzymatic activity was assessed from 12-20ug whole nuclear protein extract using the Epigenase 5mC-Hydroxylase TET Activity/Inhibition Assay kit (Epigentek). TET 25 activity was normalized to sample input protein amount (ug), and was quantified in 9-10 hippocampal biological replicates (4-5/sex) and 16-20 cerebellar biological replicates (8- 11/sex) per experimental group. All samples were run in duplicate, and results were calculated from the average of the replicates. Outliers were determined by ROUT method. 5hmC Quantification DNA hydroxymethylation was quantified as %5hmC using the Quest 5-hmC DNA ELISA Kit (Zymo Research). 100ng of genomic DNA per sample was used as input material. 5hmC was quantified in 7-8 hippocampal biological replicates (3-4/sex) and 18- 20 cerebellar biological replicates (8-10/sex) per experimental group. Statistical Analysis For all experiments, results were first analyzed by two-way ANOVA with Tukey post- hoc testing, where the two factors were sex and experimental group. Significance level was set at p<0.05. Sex was not found to be significant by two-way ANOVA for any endpoint. Subsequently males and females were combined, and results were analyzed by one-way ANOVA with Tukey post-hoc test, and α set at p<0.05. Correlation and simple linear regression analyses were performed with α set at p<0.05. For all linear regression analyses, iron status (systemic or tissue-level) served as the independent variable, and all other outcome variables (TET activity, Tet expression, %5hmC) served as the dependent variable. All statistical analyses were performed using Prism 8 (GraphPad Software). 26 Chapter 3: Early Life Neuronal-Specific Iron Deficiency Alters the Adult Mouse Hippocampal Transcriptome Barks A, Fretham SJB, Georgieff MK, Tran PV. Early life neuronal-specific iron deficiency alters the adult mouse hippocampal transcriptome. Journal of Nutrition. 2018;148:1521-1528. PMCID: PMC6258792. Introduction Micronutrient deficiencies are estimated to affect 2 billion people worldwide, with pregnant women and children under 5 years of age at greatest risk (110). Iron deficiency (ID) is the most common of the micronutrient deficiencies, with an estimated global prevalence of 25% (110,111). Among those affected, the burden falls disproportionately on pregnant women, infants, and preschool-aged children, with 40- 50% of individuals affected (111,112). A large body of data now supports that ID in these most vulnerable populations is detrimental to neurodevelopment (40,113). Human cohort studies have found an association between ID in infancy and slower perceptual speed, impaired language abilities, and difficulty with quantitative concepts – all deficits that may contribute to the finding that in adolescence, formerly iron deficient individuals have lower scores on reading and arithmetic tests, and increased likelihood to repeat a grade (35). Studies of this same cohort at age 25 show an association between early life ID and decreased likelihood of completing secondary school and pursuing higher education, and increased reports of negative emotions (37). Low maternal iron intake during pregnancy is associated with increased risk for schizophrenia (26) and autism (27). Together, these findings suggest that ID during fetal and early postnatal life has lasting negative effects on neurocognitive function and adult mental health. Furthermore, ID is chronic in many 27 populations around the world, and those children who experience chronic ID suffer more significant neurodevelopmental consequences than those who experience less severe ID (114). Preclinical rodent models have demonstrated that ID during the fetal and early postnatal periods negatively impacts the developing neurocognitive system, particularly the hippocampus. In a rat model of dietary iron deficiency anemia (IDA), chronic IDA beginning during gestation results in learning and memory behavioral deficits in adulthood (115). Importantly, adult formerly IDA rats also show behavioral deficits in hippocampal-dependent learning and memory despite iron repletion starting at postnatal day (P) 7 (72,73,116,117). Adult P65 formerly IDA rats have altered apical dendrite structure in hippocampal area CA1 (42,54), altered synaptic function (118), and altered expression of genes critical for neuronal morphogenesis, plasticity, and energy metabolism (54,76,91). On a genome-wide scale, formerly IDA rats continue to have widespread hippocampal gene expression alterations in adulthood despite iron treatment starting at the rodent developmental equivalent of term human birth (47). These altered genes map onto pathways and functions related to schizophrenia, autism, and mood disorders (47). A major limitation of the rat model of dietary IDA in deciphering the specific role of iron in long-term neurodevelopment is that animals are anemic (119). Because systemic anemia compromises tissue oxygen delivery, anemia is a confounding factor to understanding the effect of ID itself on the developing central nervous system. We have previously developed a conditional dominant negative transferrin receptor (DN-TFR1) genetic mouse model of hippocampal neuronal-specific ID in order to separate the effects 28 of neuronal ID from those of anemia (43). In this model, a point mutation in Tfr1 results in a dominant negative (DN), non-functional TFR1 protein, which impairs binding and cellular uptake of transferrin, the major iron carrier. By expressing the DN-TFR1 under a CaMKII-driven Tet transactivator (tTA), ID is restricted to hippocampal neurons and rapidly reversible with doxycycline treatment (43). Left chronically iron deficient, these mice have similar hippocampal structural and functional deficits as the IDA rat model in adulthood, indicating that ID, independent of anemia, does compromise the developing hippocampus (43). When hippocampal ID is reversed at P21, hippocampal dendrite morphology and learning and memory performance are grossly intact in adulthood, suggesting that P21 may be within the critical period for iron in neurodevelopment (43). However, it is unknown whether deficits to neuronal circuitry and gene expression remain. Currently, the underlying mechanisms of iron-dependent structural and functional changes to the developing nervous system remain incompletely characterized. Understanding hippocampal gene dysregulation following ID is an important step towards elucidating the mechanisms of adult hippocampal dysfunction. Unlike studies in the IDA rat model (47), whole-transcriptome analysis of the hippocampus has not been performed in a genetic mouse model of ID. The objective of the present study was to assess, at the transcriptomic level, two translationally-relevant questions: whether chronic ID without anemia alters gene expression in the adult hippocampus, and whether iron treatment during development completely rescues adult hippocampal gene expression. While the first question is relevant to human populations globally where chronic ID is 29 endemic, the second question relates to the public health policy of screening for childhood iron deficiency by measuring hematocrit. Results Chronic Non-Anemic Neuronal Iron Deficiency Alters Expression of Genes Involved in Development and Plasticity in the Adult Hippocampus To determine the effect of non-anemic hippocampal ID on the neuronal transcriptome, gene expression in the CID hippocampus was compared to age-matched IS controls. A total of 346 genes were differentially expressed in the CID hippocampus, representing approximately 1.5% of the sequenced transcriptome. Of these genes, a significant proportion (58%, p=0.047) were downregulated (Fig 3.1). The dysregulated genes mapped onto cellular functions that are critical for neurodevelopment, including shape change of neurites and axons, and branching of neurites and axons (Table 3.1). These neuronal functions were all predicted to have decreased activation (Z-score<-2.0) in the CID hippocampus, based on the observed changes in gene expression. Additionally, dysregulated genes mapped onto canonical signaling pathways that are involved in neurodevelopment and synaptic plasticity (Fig 3.2A), including axonal guidance, CXCR4, CDK5, Ephrin receptor, GABA receptor, Rac, and Neurotrophin/TRK signaling pathways. The majority of these pathways were predicted to have reduced activity (Z-score<-1.5) in the ID hippocampus. 30 Figure 3.1. Gene dysregulation in the chronically- and formerly iron deficient adult hippocampus. Chronic ID starting in late embryonic development results in dysregulation of 346 genes in the adult hippocampus. The majority of these genes are downregulated (p=0.047). When chronic ID is reversed at P21, 148 genes are dysregulated in the adult hippocampus. The majority are upregulated (p<0.0001). Of these 148 genes, 58 are genes that are also dysregulated in the chronically iron deficient hippocampus. The remaining 90 dysregulated genes are uniquely dysregulated in the formerly iron deficient hippocampus. CID: Chronically iron deficient; FID: Formerly iron deficient; IS: Iron sufficient. Total Genes 346 148 Upregulated (%) 146 (42%) 111 (75%) Downregulated (%) 200 (58%) 37 (25%) Fisher’s Exact Test P=0.047 P<0.0001 Upregulated Downregulated CID vs. IS FID vs. IS 288 58 90 31 D isease or function annotation A ctivation Z-score p-value G enes G enes, n C ell survival -3.47 1.16 x 10 -3 Bcl6, Bdnf, C cnd1, C ebpb, Cebpd, C xcl12, Egr1, Egr3, Fos, Foxo1, G fra1, H 2afx, Igf2, Il33, Jun, Lcn2, N pnt, N r4a1, N rn1, Ntf3, Prkcd, Ptgs2, Sm ad3, Spp1, Traf3, Vgf 26 A ccum ulation of cells -2.28 3.21 x 10 -3 C 4a/C 4b, C cnd1, C dkn1c, Cx3cr1, D usp1, Egr1, Enpp2, Il33, K rt18, Lcn2, Ltc4s, N tf3, Ptgds, Tnfrsf25 14 Shape change of axons -2.35 8.98 x 10 -6 A2m , Bdnf, C xcl12, Egr3, Fos, G rasp, K alrn, M id1, Tbr1 9 Shape change of neurites -2.28 1.31 x 10 -8 A2m , Bdnf, C hn1, Cxcl12, D ock10, Egr3, Fos, G it1, G rasp, K alrn, Lcn2, M ag, M id1, Neurod6, N rn1, Prss12, Ptgs2, Rapgef4, Ryr1, Scn4b, Sulf1, Tbr1, Vgf 23 B ranching of axons -2.16 1.76 x 10 -5 A2m , Bdnf, C xcl12, Egr3, Fos, G rasp, K alrn, M id1 8 B ranching of neurites -2.13 3.25 x 10 -8 A2m , Bdnf, C hn1, Cxcl12, D ock10, Egr3, Fos, G it1, G rasp, K alrn, Lcn2, M ag, M id1, Neurod6, N rn1, Prss12, Ptgs2, Rapgef4, Ryr1, Scn4b, Sulf1, Vgf 22 Table 3.1. Top dysregulated diseases and functions in the adult chronically iron deficient m ouse hippocam pus, annotated by Ingenuity Pathw ay A nalysis (IPA ) 32 Figure 3.2. Critical neuronal signaling pathways are altered by ID and rescued by its reversal at P21. In IPA, differentially expressed genes were mapped to canonical signaling pathways. (A) In the adult CID hippocampus, dysregulated genes mapped significantly onto canonical signaling pathways that function in neurodevelopment and neuroplasticity (p<0.05, -log(p-value)>1.3, dotted line). Based on the genes that were dysregulated and the directionality of their dysregulation, it was predicted that the majority of these pathways would have a net decrease in activity, indicated by a negative Z-score (Z<-1.5). (B) In the adult FID hippocampus, dysregulated genes no longer mapped significantly to the majority of these same pathways. Dysregulated genes still mapped significantly to Axonal guidance signaling and CXCR4 signaling pathways, however, there was no predicted effect on the net function of the pathways. CID: Chronically iron deficient; FID: Formerly iron deficient; IS: Iron sufficient. Negative Z-score Positive Z-score Z-score = 0 No Activity Pattern Available A B 33 Adult Gene Expression is only Partially Rescued by Iron Repletion at P21 To determine the effect of ID reversal on hippocampal gene expression in adulthood, iron status was normalized starting at P21, and hippocampal gene expression was assessed in adult FID animals compared to age-matched IS controls. A total of 148 genes were differentially expressed in the adult FID hippocampus despite iron repletion starting at P21, indicating a partial resolution of gene dysregulation compared to the CID hippocampus. Of these genes, a significant proportion (75%, p<0.0001) were upregulated (Fig 3.1). 90 genes were uniquely dysregulated in the FID hippocampus, and 58 were dysregulated in both the CID and FID hippocampus (Fig 3.1). Using IPA, the same set of neurodevelopmental and plasticity canonical signaling pathways were re-examined in the formerly iron deficient hippocampus. Compared to the CID hippocampus where these pathways were predicted to be significantly downregulated, the majority of these pathways were not affected in the FID hippocampus (Fig 3.2B). Axonal Guidance Signaling and CXCR4 signaling pathways had a significant number of genes differentially expressed, however, there was no predicted change to overall pathway activity. IPA was used to determine the potential functions and effects of the 148 genes that were dysregulated in the adult FID hippocampus. Genes mapped primarily to cellular functions involved in cell death (Table 3.2). Apoptosis, necrosis, and cell death were all predicted to have reduced activation (Z-score<-2.0), whereas cellular homeostasis, activation of cells, and cell movement were predicted to have increased activation (Z- score>2.0). 34 D isease or function annotation A ctivation Z-score p-value G enes G enes, n A poptosis -3.82 3.87 x 10 -5 Ace, Ascl1, Btg2, C am k2d, Casp9, Cckbr, Cd200, Cd44, Cnp, Ctgf, C yr61, D cn, D usp1, Egr1, Fm od, Fos, Igf2, Junb, Kitlg, K l, K lf5, M ef2c, N pas4, N r4a1, N tsr1, Pcp4, Prkar2b, Prkca, Ptgds, Sept4, Sfrp1, Slc4a2, Spp1, Trh, Ttr, Vstm 2l 32 N ecrosis -3.23 8.18 x 10 -5 Ascl1, Btg2, C am k2d, C asp9, C d200, Cd44, C np, C ol6a1, Ctgf, D cn, D usp1, Egr1, Fos, G jb2, Igf2, Junb, K itlg, K l, K lf5, Lbp, M ef2c, N pas4, N r4a1, Pcp4, Prkar2b, Ptgds, Sfrp1, Spp1, Trh, Ttr, Vstm 2l 31 C ellular hom eostasis 2.95 4.98 x 10 -4 Atp2b4, C asp9, C ldn1, C ldn2, D cn, Egr1, F5, H tr2c, Igf2, Junb, K itlg, M ef2c, N r4a1, Prkca, Ryr1, Sfrp1, Slc31a1, Slc4a2, Spns2, Spp1 20 C ell death -2.88 7.03 x 10 -6 Ace, Antxr1, Ascl1, Atp2b4, Btg2, C am k2d, C asp9, C ckbr, Cd200, C d44, C np, C ol6a1, Ctgf, C yr61, D cn, D usp1, Egr1, Fm od, Fos, G jb2, Igf2, Junb, Kitlg, Kl, K lf5, Klk8, Lbp, M ag, M ef2c, N pas4, N r4a1, N tsr1, Pcp4, Prkar2b, Prkca, Ptgds, Ryr1, Sept4, Sfrp1, Slc31a1, Slc4a2, Spp1, Trh, Ttr, Vstm 2l 45 A ctivation of cells 2.54 2.61 x 10 -4 Ace, C d200, C d44, C hrna4, C tgf, D usp1, F5, G jc2, H tr2c, Igf2, K itlg, Lbp, N r4a1, Prkca, Rab34, Rab3b, Slc4a2, Spp1 18 C ell m ovem ent 2.47 9.14 x 10 -6 Ascl1, Atp2b4, Btg2, C d200, C d44, Ctgf, C yr61, D cn, Efna5, Enpp2, Fnbp1l, Fos, G jb2, Igf2, K itlg, K lf5, Lam c2, Lbp, M ef2c, M inos1nbl1/N bl1, N r2f2, N r4a1, Plxnd1, Prkca, Ptgds, Sept4, Sfrp1, Slc4a2, Spns2, Spp1, Sulf1, W ls 32 Table 3.2. Top dysregulated diseases and functions in the adult form erly iron deficient m ouse hippocam pus, annotated by Ingenuity Pathw ay A nalysis (IPA ) 35 A Subset of Hippocampal Genes Are Permanently Dysregulated by Developmental ID Of the 148 genes that were dysregulated in the FID hippocampus, a subset of 58 are genes were also dysregulated in the CID hippocampus, suggesting that expression of these genes is highly sensitive to early life ID and insensitive to its reversal (Fig 3.1). Although these genes were dysregulated in both the CID and FID hippocampus, the dysregulation did not always have the same directionality; 67% of genes were dysregulated in the same direction in both conditions while the remaining 33% of genes were dysregulated in opposite directions (Fig 3.3A). Utilizing IPA, these 58 genes mapped onto functions related to neurocognitive function and diseases, including cognition, learning, schizophrenia, and Alzheimer’s disease (Fig 3.3B). Figure 3.3. Genes permanently dysregulated by ID are involved in neurocognitive function and dysfunction. (A) Of the 58 genes dysregulated in both the CID and FID hippocampus, 9 (15%) are downregulated in both conditions, and 30 (52%) are upregulated in both conditions. The remaining 19 genes (33%) are dysregulated with opposite directionality in the two conditions. Upward arrow indicates gene upregulation; downward arrow indicates gene downregulation. (B) When mapped to diseases and functions in IPA, these genes map to the functions of cognition and learning, as well as schizophrenia and Alzheimer’s disease. CID: Chronically iron deficient; FID: Formerly iron deficient; IS: Iron sufficient. CI D vs . I S FI D vs . I S A B down down down up up down up up 52% 24% 9% 15% 36 CREB1 is Differentially Activated in the Chronically and Formerly ID Hippocampus IPA was also used to predict which upstream regulators of gene expression might account for the gene dysregulation associated with ID. cAMP response element-binding protein (CREB) signaling was chosen for further analysis because of its critical role in late-stage long term potentiation (LTP), and thus, long term memory. As a transcription factor, CREB serves as the upstream transcriptional regulator of LTP-associated gene expression changes that underlie memory (120). Additionally, CREB is activated by BDNF signaling (121), which is known to be reduced by ID (91). Based on the observed gene expression changes, CREB1 activity was predicted to be altered in both the CID and FID hippocampus. In the CID hippocampus (Fig 3.4A), 13 genes regulated by CREB1 signaling were downregulated, predicting a net decrease in CREB1 activity (Z-score=- 3.4, p<0.001). In contrast, CREB1 activity was predicted to be increased in the FID hippocampus (Z-score=2.6, p<0.001), where 7 genes known to be regulated by CREB1 signaling were upregulated (Fig 3.4B). Of these CREB1-regulated genes, 6 were dysregulated in both conditions, but in opposite directions. RT-qPCR was used to validate this subset of 6 CREB1-regulated genes (Table 3.3). 37 Figure 3.4. CREB1 signaling is predicted to be altered by ID. (A) 13 genes involved in CREB1 signaling were found to be significantly downregulated in the CID hippocampus. Based on these gene expression changes, CREB1 activity is predicted by IPA to be significantly decreased (Z-score=-3.42, p<0.001). (B) In the FID hippocampus, 7 CREB1-regulated genes were found to be significantly upregulated, which predicts a significant increase in CREB1 activity (Z-score=2.61, p<0.001). Genes outlined in bold are dysregulated in both chronic- and former ID. A B 38 Table 3.3. RT-qPCR validation of CREB1-regulated target genes in chronically and formerly iron deficient mouse hippocampus Fold Change (Relative to IS Control) CID FID Gene Probe ID RNA-Seq§ RT-qPCR# RNA-Seq§ RT-qPCR# Arc Mm01204954_g1 0.51 0.65* 1.49 1.48* Btg2 Mm00476162_m1 0.65 0.75* 1.48 1.2** Egr1 Mm00656724_m1 0.75 0.79 1.64 1.29* Fos Mm00487425_m1 0.51 0.43** 3.12 2.73 Junb Mm04243546_s1 0.65 1.02 1.58 1.23 Nr4a1 Mm01300401_m1 0.59 0.67** 1.42 1.05 § All target genes had p<0.05 by RNA-Seq analysis, after correcting for multiple comparisons * p<0.05, ** p<0.01 compared to the respective IS control group (WTnoDox or WT+P21Dox) # n=5-6/group CID: Chronically iron deficient; FID: Formerly iron deficient; IS: Iron sufficient 39 Discussion Using a unique non-anemic mouse model of hippocampal neuronal-specific ID, we demonstrate that chronic hippocampal ID, without anemia, results in altered expression of approximately 1.5% of the adult hippocampal transcriptome. These altered genes map to canonical signaling pathways and functions that are critical for normal neurodevelopment and plasticity, predicting a net decrease in these functions in the iron deficient hippocampus. These observed gene expression changes are consistent with findings from our primary hippocampal neuronal culture model of pure neuronal ID, which demonstrates decreased branching and complexity of developing dendritic arbors (56) as well as our previous findings using this mouse model, which demonstrate compromised apical dendrite structure in hippocampal area CA1 and poorer spatial learning and memory in permanently iron deficient mice (43). The vast majority of previous studies of ID have utilized a rat model of dietary ID in which animals become anemic. While the formerly IDA rat hippocampus exhibits significant, transcriptome-wide changes in gene expression in adulthood (47), it has been unclear how many of these gene expression changes are due to ID itself rather than anemia. The present results indicate that while a smaller percentage of the hippocampal transcriptome is dysregulated by non-anemic ID than by IDA (1.5% and 5.7%, respectively) (47), ID alone is sufficient to disrupt hippocampal gene expression. Translationally, this is important for two reasons. First, non-anemic ID is most prevalent in populations that are at risk for suffering its neurodevelopmental effects, including preschool-aged children (122), pregnant women (123), and certain infant populations such as premature infants, infants of diabetic mothers, and intrauterine growth restricted 40 infants who are at risk for non-anemic ID from birth due to low iron endowment (124). Second, iron is prioritized to red blood cells over other tissues, including the central nervous system (CNS) (125). Thus, chronic ID may be present in the CNS long before it affects prioritized tissues such as red blood cells and is detected peripherally as anemia (126). Our results indicate that individuals experiencing this type of non-anemic ID, particularly during neurodevelopment, are potentially at risk for transcriptome-level changes to the CNS that may affect CNS function in adulthood. One question that has remained unanswered is the exact timing of the critical period for iron in hippocampal development. Defining this period is critical for understanding when treatment with iron will completely restore hippocampal function and thus result in no long-term deficits. In previous studies, normalizing iron status within the putative critical period for iron in neurodevelopment (P10-P30) was sufficient to prevent gross hippocampal morphological and functional abnormalities in adulthood (43,127). However, normalization of this set of parameters does not preclude the possibility that more subtle deficits remain in the formerly iron deficient hippocampus. The present transcriptomic analysis indicates that reversal of chronic hippocampal ID at P21 does not fully normalize the hippocampal transcriptome in adulthood. While many critical neuronal signaling pathways are no longer predicted to be altered in the FID hippocampus, those genes that do remain dysregulated map to functions such as learning, cognition, and schizophrenia, and predict altered function of key neuronal actors such as CREB. Importantly, these predicted diseases and functions are consistent with the human cohort literature, which shows that low maternal iron intake during pregnancy is 41 associated with increased risk of schizophrenia (26), and that early-life iron deficiency is associated with increased risk of learning and cognition deficits (35). Interestingly, we found that 33% of the genes dysregulated in both the CID and FID hippocampus were dysregulated in opposite directions. A possible explanation for this phenomenon is that chronically iron deficient tissue reaches a new homeostatic set point by P21, and that re-introduction of iron disrupts this homeostasis by iron- overloading a tissue that has adapted, via compensatory changes in gene expression, to low iron levels. In this case, re-introduction of iron could result in a rebound dysregulation of genes in the opposite direction of that in the iron deficient condition. Additionally, we found that 90 genes were dysregulated uniquely in the FID hippocampus. Together, these data raise the possibility that iron repletion of a previously iron deficient tissue may have effects of its own. Further research is needed to understand the functional implications of these gene expression changes, and the ideal timing for iron therapy in order to avoid any potential lasting or deleterious effects of iron repletion. Finally, the mechanisms by which developmental iron status drives the reported long-term gene expression changes are incompletely understood. Epigenetic modification of DNA and chromatin is one mechanism by which environmental factors in early life, such as nutrition, can program gene expression into adulthood (128). Several proteins involved in establishment of epigenetic modifications require iron for their enzymatic activity, including the JmjC ARID-domain containing histone demethylase (JARID) proteins and TET methylcytosine dioxygenases (TETs). JARIDs require iron for their enzymatic removal of methyl groups from lysine residues of histone tails (86,87). We have previously reported that the formerly IDA rat hippocampus undergoes differential 42 histone methylation at the BDNF-IV promoter compared to IS controls, concomitant with decreased BDNF expression, likely due to decreased JARID enzymatic activity during iron deficiency (84). TETs also have an absolute requirement for their enzymatic conversion of 5-methylcytosine (the epigenetic modification commonly known as DNA methylation) to 5-hydroxymethylcytosine (93), however, they have never been studied in the context of developmental ID. Because epigenetic modifications regulate gene expression throughout the genome, alteration of iron-dependent epigenetic modifications presents a plausible mechanism by which broad, genome-wide changes in gene expression could be indirectly driven by iron deficiency. Further investigation of these iron-dependent epigenetic modifications will advance our understanding of the underlying mechanisms by which non-anemic ID enacts its lasting effects. 43 Chapter 4: Fetal-Neonatal ID Increases TET activity and DNA hydroxymethylation in the developing rat hippocampus Introduction There is an increasing understanding that the early-life environment, including the nutritional environment, can have a lasting impact on health. Iron deficiency (ID) is estimated to affect 40-50% of pregnant women and preschool-aged children worldwide (30,111). It is now well-established that ID during the fetal and early childhood periods has a significant effect on neurodevelopment, resulting in cognitive, socio-emotional, and learning and memory deficits that last into adolescence and early adulthood (36,37,127). It also carries long-term health risks including increased risk for neuropsychiatric disorders including autism spectrum disorders, ADHD, and schizophrenia (26,27,41). By studying preclinical models of fetal-neonatal ID, it has been established that these neurophenotypical changes are underlain by significant dysregulation of gene expression in the central nervous system (CNS) as a whole (129). Gene expression is also widely dysregulated specifically in the hippocampus, a brain region which has been studied extensively in the context of developmental ID (76,77). Not only are changes in gene expression present acutely during ID, but they persist into adulthood despite treatment of ID, propagating the potential for neuronal dysfunction into adulthood (47,48). These ID-induced changes in CNS gene expression are likely the proximate cause of the abnormal neurocognitive and neuropsychiatric phenotype of developmental ID. However, the ultimate mechanism by which early-life ID establishes and maintains altered gene expression across the lifespan remains unknown. Thus, there is a need to 44 understand the mechanisms by which early life ID drives gene expression changes in the developing brain, leading to permanently altered gene expression in the adult brain. Epigenetic modification is one mechanism by which environmental influences such as ID can enact permanent changes in gene expression (130,131). Our group and others have previously demonstrated that two major epigenetic mechanisms, histone methylation and DNA methylation, are both altered in the developing hippocampus by fetal-neonatal ID (77,84,89). However, given the complexity of the epigenome, developing a full understanding of how ID reprograms gene expression requires a more complete understanding of how ID effects all components of the epigenetic code. A potentially important iron-dependent epigenetic mechanism that remains unexplored in developmental ID is the modified cytosine base 5-hydroxymethylcytosine (5hmC), or DNA hydroxymethylation. 5hmC is closely related to 5-methylcytosine (5mC), or DNA methylation, and both are stable epigenetic modifications that play a role in gene transcriptional regulation. 5hmC is generated by TET methylcytosine dioxygenases, which enzymatically convert 5mC to 5hmC (93). TET proteins belong to the large class of iron- and α-ketoglutarate (αKG)-dependent oxygenases, and as such, they have an absolute requirement for iron for their enzymatic activity (25). If the iron binding site is mutated, or if iron is depleted, TET proteins do not efficiently convert 5mC to 5hmC, resulting in significantly decreased global levels of 5hmC (93). While the iron-dependence of TET proteins has been clearly established, the effect of a clinically relevant degree of iron deficiency on TET protein activity in the CNS has never been assessed. 45 The mammalian TET family of proteins contains three members: TET1, TET2, and TET3. All three are expressed at high levels in the CNS, particularly during neurodevelopment (103). In parallel, 5hmC, the enzymatic product of TET proteins, is more enriched in the CNS than other tissues, where it accounts for approximately 17.2% of all modified cytosines, and more in select neuronal populations (99–101). TET expression and enrichment of 5hmC increases in the CNS during neurodevelopment, from low levels in neural progenitor cells and immature neurons to the highly enriched state specific to mature neurons (102–104). This enrichment occurs particularly at genes that are critical for neurodevelopment and neuronal function (100,102–105). As neurons mature from neural progenitor cells to mature neurons, 5hmC increases in the intragenic regions of genes critical for neuronal differentiation, migration, and axon guidance (103). Additionally, in the mouse hippocampus, 5hmC levels increase significantly (2.57-fold) between P7 and adulthood, predominantly within the exons of genes that become activated between these two time points (102). Globally, 5hmC is enriched in the gene bodies of genes that are highly transcribed (100). Finally, genes in the CNS that are differentially hydroxymethylated compared to other tissue types map to synapse-related functions and neuronal development (102,105). Taken together, this evidence demonstrates that TET proteins are highly enzymatically active in the CNS, particularly during neurodevelopment when they actively establish neuronal 5hmC – a stable epigenetic modification that plays a critical role in the establishment and regulation of CNS-specific gene expression. The iron-dependence of this epigenetic system, together with its apparent importance during neurodevelopment, suggests that it may be vulnerable to disruption by 46 ID, and that this disruption could have significant consequences for neurodevelopment. Thus, the aim of the present study is to determine the effect of fetal-neonatal iron deficiency anemia (IDA) on TET enzymatic activity and establishment of its enzymatic product, DNA hydroxymethylation, in the developing rat hippocampus. The second major aim is to determine whether changes to this epigenetic system are reversible with early treatment of ID. Our results suggest that altered hippocampal TET activity and DNA hydroxymethylation may contribute to the hippocampal gene dysregulation and subsequent hippocampal dysfunction associated with fetal-neonatal ID. Results Assessment of Systemic and Tissue-Level Iron Status at P15 Hematocrit, or the volume percentage of the blood composed of red blood cells, was quantified from peripheral blood at P15 to assess systemic iron status of pups (Fig 4.1A). Relative to IS controls, IDA pups had significantly reduced hematocrit (28.1±3.6% vs. 10.2±3.9%, p<0.0001) indicating that they were anemic, a common end- stage of severe or prolonged ID. Treatment with iron-replete diet starting at P7 resulted in partial recovery of hematocrit in the TIDA group relative to their untreated IDA counterparts (20.9±4.8% vs. 10.2±3.9%, p<0.0001). However, relative to IS controls, TIDA pups continued to have significantly lower hematocrit (p<0.0001), indicating that recovery of their systemic iron status was still ongoing at P15. To assess tissue-level iron status, mRNA expression of Transferrin Receptor-1 (Tfrc) was quantified (Fig 4.1B). Transferrin Receptor-1 (TfR1) is the main cellular surface receptor involved in iron import into cells, and its expression is upregulated in 47 conditions of iron deficiency (133). IDA pups had significantly increased hippocampal Tfrc expression relative to IS controls (p<0.0001), indicating a tissue-level iron deficiency in the hippocampus. TIDA pups continued to have significantly increased hippocampal Tfrc expression relative to IS controls (p=0.0191), but with significant recovery relative to untreated IDA counterparts (p<0.0001). Hematocrit and hippocampal Tfrc expression were analyzed by correlation analysis to confirm the relationship between systemic iron status and hippocampal tissue iron status (Fig 4.1C). As expected, there was a significant inverse correlation between hematocrit and Tfrc expression (R2=0.8061; p<0.0001). Figure 4.1. Characterization of iron status in P15 rats. (A) Reduction in systemic iron status, indexed by hematocrit, at P15 following fetal-neonatal dietary iron deficiency. The greatest reduction in hematocrit was seen in the untreated IDA group. Dietary iron repletion resulted in partial recovery of hematocrit. (B) Reduction in tissue iron status of the P15 hippocampus, as indexed by Tfrc mRNA expression. Treatment with dietary iron repletion resulted in significant but incomplete recovery of tissue iron status. (C) Hippocampal tissue iron status is significantly correlated with systemic iron status. Error bars indicate SD; Dashed lines indicate 95% CI; * indicates p<0.05; **** indicates p<0.0001; IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. 0 10 20 30 40 0 1 2 3 4 P15 Hippocampus Tfrc vs. Hct Hematocrit Tf rc E xp re ss io n Goodness of Fit: R2=0.8061 Is slope significantly non-zero? p<0.0001 CB IS IDA TIDA 0 1 2 3 4 Tfrc Expression P15 Hippocampus R el at iv e Ex pr es si on **** **** * A IS IDA TIDA 0 10 20 30 40 P15 H em at oc rit **** **** **** P<0.0001 R2=0.8061 48 These results indicate that IDA pups were iron deficient at both the systemic and hippocampal tissue levels relative to IS controls. Iron status of TIDA pups at both the systemic and tissue levels was partially recovered by P15, giving them an intermediate iron status and demonstrating that they are responsive to iron therapy. Finally, systemic iron status, as quantified by hematocrit, was significantly correlated with and reflective of tissue iron status. TET Activity and Expression are Increased in the P15 IDA Hippocampus The effect of developmental IDA on enzymatic activity of TET was assessed in the P15 hippocampus (Fig 4.2A). Relative to IS controls, the IDA hippocampus had significantly increased TET activity (p<0.0001). The TIDA hippocampus continued to have significantly increased TET activity relative to IS controls despite dietary iron repletion (p=0.012). To determine whether TET activity had a continuous dose-dependent relationship with iron status, regression analysis of TET activity was performed against hematocrit and hippocampal Tfrc expression (Fig 4.2B). TET activity had a significant inverse relationship with hematocrit (p=0.0236, R2=0.2536), and a positive correlation with hippocampal Tfrc expression (p=0.1243, R2=0.2693), both indicating that TET activity increased with decreasing iron status. 49 Figure 4.2. Hippocampal TET activity increases with decreasing iron status. (A) TET activity, quantified from whole nuclear protein extract, is increased in the IDA and TIDA hippocampus. Dietary iron repletion does not normalize TET activity in the TIDA hippocampus (B) Regression analysis of TET activity, as a factor of systemic iron status (hematocrit) or hippocampal iron status (Tfrc expression). TET activity has a significant inverse relationship with hematocrit, indicating that decreasing systemic iron status is associated with increasing TET activity, across all three experimental groups. Error bars indicate SD; Dashed lines indicate 95% CI; * indicates p<0.05; **** indicates p<0.0001; IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. Expression of Tet1-3 mRNA was quantified to determine whether the increase in TET activity reflected an increase in enzyme functionalization or increased expression of Tets (Fig 4.3A). Tet1 and Tet2 expression were unchanged in IDA and TIDA hippocampus compared to IS controls. In contrast, expression of Tet3 was significantly upregulated in IDA (p<0.0001) and TIDA (p=0.0462) hippocampus. Regression analysis demonstrated that expression of both Tet2 and Tet3 was significantly correlated with hippocampal iron status, as indexed by Tfrc expression, in a dose-dependent manner (Fig 4.3B; Tet2: p=0.001, R2=0.2767; Tet3: p<0.0001, R2=0.4238). IS IDA TIDA 0.0 0.5 1.0 1.5 2.0 2.5 P15 Hippocampus A ct iv ity (R el at iv e) **** * 0 10 20 30 40 0.0 0.5 1.0 1.5 2.0 2.5 Hematocrit A ct iv ity (R el at iv e) Transform of HPC: Hct vs. TET Activity p=0.0236 R2=0.2536 0 1 2 3 4 0.0 0.5 1.0 1.5 2.0 2.5 Tfrc (Relative Expression) A ct iv ity (R el at iv e) Transform of HPC: TET activity vs. Tfrc p=0.1243 R2=0.2693 A B 50 Figure 4.3. Hippocampal Tet expression increases with decreasing tissue iron status. (A) Expression of Tet3 mRNA is increased in the IDA hippocampus. Dietary iron repletion significantly reduces Tet3 expression relative to the untreated IDA group, but not to control levels. Tet1 and Tet2 mRNA expression are not significantly affected by IDA or its treatment. (B) Regression analysis of Tet mRNA expression as a function of hippocampal tissue iron status. Both Tet2 and Tet3 expression have a significant positive correlation with Tfrc mRNA expression, indicating that Tet mRNA expression increases with decreasing tissue iron status. Error bars indicate SD; Dashed lines indicate 95% CI; * indicates p<0.05; **** indicates p<0.0001 IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. A IS IDA TIDA 0.8 1.0 1.2 1.4 R el at iv e Ex pr es si on Tet1 ns 0 1 2 3 4 5 0.8 1.0 1.2 1.4 HPC: Tet1 vs. Tfrc Tfrc (Relative Expression) Te t1 (R el at iv e Ex pr es si on ) p=0.1317 R2=0.0656 IS IDA TIDA 0.8 1.0 1.2 1.4 R el at iv e Ex pr es si on Tet2 ns 0 1 2 3 4 5 0.8 1.0 1.2 1.4 Tfrc (Relative Expression) Te t2 (R el at iv e Ex pr es si on ) P15 Hippocampus Tet2 vs. Tfrc p=0.001 R2=0.2767 IS IDA TIDA 0.8 1.0 1.2 1.4 R el at iv e Ex pr es si on Tet3 **** * * 0 1 2 3 4 5 0.8 1.0 1.2 1.4 P15 Hippocampus Tet3 vs. Tfrc Tfrc (Relative Expression) Te t3 (R el at iv e Ex pr es si on ) p<0.0001 R2=0.4238 B 51 Together, these results demonstrate that developmental IDA results in increased TET activity, and that dietary iron repletion is not sufficient to completely normalize TET activity within the developmental timeframe studied. The Tet expression analyses suggest that increased expression of Tet2 and Tet3 may be driving the increased TET activity, rather than an increase in enzyme functionalization. DNA hydroxymethylation is Increased in the P15 IDA Hippocampus To determine the functional effect of increased TET activity, the enzymatic product of TET, 5-hydroxymethylcytosine (5hmC), was quantified from genomic DNA. 5hmC was quantified as a percentage of total nucleotides (global %5hmC). In parallel with TET activity, global %5hmC was elevated in the IDA hippocampus relative to IS controls (Fig 4.4A; p=0.0105). In contrast, while TET activity remained elevated in the TIDA hippocampus, %5hmC returned to IS control levels with dietary iron repletion. Regression analysis demonstrated that hippocampal %5hmC was significantly correlated with systemic iron status (p=0.0236, R2=0.2308), and marginally correlated with tissue iron status (p=0.0672, R2=0.4124), where increasing %5hmC was correlated with decreasing iron status (Fig 4.4B). These results suggest that developmental IDA-induced changes in TET activity result in functional changes in establishment of the enzymatic product, 5hmC. Additionally, altered global %5hmC is rapidly reversible with dietary iron repletion within this developmental timeframe, and is reversible without complete recovery of TET activity. 52 Figure 4.4. Global %5hmC is increased in the IDA hippocampus. (A) Global %5hmC is increased in the P15 IDA hippocampus. Dietary iron repletion normalizes %5hmC. (B) Regression analysis of %5hmC as a function of systemic (hematocrit) or hippocampal (Tfrc expression) iron status. There is a significant inverse relationship between %5hmC and hematocrit. Error bars indicate SD; Dashed lines indicate 95% CI; * indicates p<0.05; ** indicates p<0.01 IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. Discussion In the present study, we sought to determine whether activity of the iron- dependent TET enzymes, and establishment of their enzymatic product DNA hydroxymethylation, is altered in the developing rat hippocampus by fetal-neonatal ID. While it has been established that TETs have an absolute requirement for iron for their enzymatic activity, the effect of a clinically relevant level of ID on TETs has never been assessed. Previous studies performed in cell culture have shown that when the iron binding site of TETs was mutated or when iron was depleted from culture media, TET activity and establishment of 5hmC was almost completely eliminated (93). In contrast, the present study finds that in the developing rat hippocampus, physiological IDA, which typically results in a 40% reduction of brain tissue iron content (134) results in increased TET activity and increased 5hmC. IS IDA TIDA 0.0 0.1 0.2 0.3 0.4 0.5 % 5h m C P15 Hippocampus *** 0 10 20 30 40 0.0 0.1 0.2 0.3 0.4 0.5 Hematocrit % 5h m C P15 Hippocampus %5hmC vs. Hct Goodness of Fit: R2=0.2308 Is slope significantly non-zero? p=0.0236 p=0.0236 R2=0.2308 0 1 2 3 4 0.0 0.1 0.2 0.3 0.4 0.5 Tfrc (Relative Expression) % 5h m C P15 Hippocampus %5hmC vs. Tfrc Goodness of Fit: R2=0.4124 Is slope significantly non-zero? p=0.0672 (ns) p=0.0672 R2=0.4124 A B 53 Broadly, the main mechanisms by which TET activity can be increased are through increased functionalization of TET enzymes or increased TET expression. The present mRNA expression data supports the increased TET expression hypothesis, finding that expression of both Tet2 and Tet3 increase with decreasing hippocampal iron status. This upregulation appears to compensate, or even overcompensate, for any decrease in TET enzymatic activity that ID might drive, resulting in a net increase in TET enzymatic activity. While the exact mechanism by which ID upregulates Tet expression is unclear, possible mechanisms include chromatin remodeling and differential expression/recruitment of transcriptional machinery by ID. Both of these process have previously been demonstrated to occur at the Bdnf promoter in developmental ID, driving differential gene expression (84). Further study of chromatin state and transcriptional machinery recruitment to the Tet2 and Tet3 promoters has the potential to provide insight into the mechanism by which ID increases Tet expression in the developing hippocampus. A limitation of these data is that Tet expression was quantified at the level of mRNA. It will be important in future experiments to quantify protein expression of TETs to determine whether increased mRNA expression translates to an increase in protein expression of TETs. ID and anemia have previously been shown to downregulate signaling of the mTOR pathway, which is critically involved in signaling of protein synthesis in response to cellular energy and metabolic status (135,136). Thus, it is possible that there may be a disconnect between mRNA expression and protein expression in the IDA hippocampus. 54 The second major aim of the study was to determine whether the observed changes in TET activity and 5hmC were reversible with treatment of ID, in the form of dietary iron repletion starting at P7. This is intended to model a clinical scenario in which maternal-fetal iron deficiency during pregnancy is detected at birth and treated, as a P7 rat is the rough neurodevelopmental equivalent of a term human infant (137). The present study finds that TET activity is not fully recovered by P15 despite dietary iron repletion. Concomitant Tet expression data suggests that this is likely due to the continued upregulation of Tet in the TIDA hippocampus at P15. RNA sequencing of the TIDA hippocampus in adulthood (P65) suggests that Tet upregulation does not persist long-term (47). Thus, it is possible that with continued dietary iron repletion and full recovery of systemic- and tissue-level iron status, TET activity may also recover. Future studies using additional timepoints aged beyond P15 can determine whether TET activity recovers fully with full recovery of systemic and tissue-level ID, as well as the timeline of recovery. Unlike TET activity, global %5hmC recovers to IS control levels with dietary iron repletion by P15. However, normalization of global %5hmC does not preclude the possibility that changes in 5hmC might persist at the level of individual genetic loci. This is the case for DNA methylation, where our lab has previously shown that there is no increase or decrease in global DNA methylation (%5mC) in the P15 TIDA rat hippocampus (89). In contrast, at the locus-specific level, 229 individual loci are differentially methylated, mapping to 108 genes (89). This same phenomenon may be true for DNA hydroxymethylation, and future studies of locus-specific 5hmC can reveal whether individual loci remain differentially hydroxymethylated despite treatment of 55 developmental ID. Similarly, in the IDA hippocampus where %5hmC is globally increased, future locus-specific studies have the potential to lend insight into what role altered 5hmC might play in the gene dysregulation that occurs during developmental ID. Given the long-term cognitive and neuropsychiatric consequences of developmental ID, it is of critical importance to understand the underlying mechanisms by which ID programs the developing brain for later dysfunction. Understanding these mechanisms will ultimately allow for more effective treatment, or even prevention, of these long-term neurodevelopmental consequences. The present results demonstrate that a novel epigenetic system, DNA hydroxymethylation, is disrupted in the developing rat hippocampus by fetal-neonatal ID, through disruption of the iron-dependent TET enzymes. DNA hydroxymethylation has been implicated in the dynamic gene regulation required throughout neurodevelopment, as well as critical processes such as dendritic arborization (100,104). Thus, dysregulation of the DNA hydroxymethylation system by developmental ID may contribute to the neurodevelopmental disruption and subsequent long-term cognitive and neuropsychiatric phenotype of developmental ID. 56 Chapter 5: Fetal-Neonatal ID Reduces TET activity and DNA hydroxymethylation in the developing rat cerebellum Introduction Development of the cerebellum is extremely protracted relative to most other brain regions, stretching from early gestation well into postnatal development. For example, development of cerebellar granule cells, the most numerous neuronal cell type in the CNS, begins with birth of granule cell progenitors by embryonic day (E)13 in rodents, and continues with migration and robust expansion of granule cells spanning from E15 to postnatal day (P)15 (138). Finally, maturation of morphology and connectivity continues through the third to fourth week of life (139). Cerebellar neurodevelopment is similarly protracted in human development, and has a significant postnatal component stretching through the first year of life (140). Given that developing cells and organ systems are most vulnerable to disruption by environmental insults, the cerebellum, with its extended neurodevelopmental trajectory, is highly sensitive to environmental insults (141–144). It has long been noted that developmental iron deficiency (ID) is associated with impaired motor development and function. Like the cognitive and socioemotional effects associated with developmental ID, the motor functional effects also last well beyond the initial period of ID and into childhood. ID diagnosed in early development between 1 and 2 years of age is associated with persistent motor deficits at 5 and 10 years of age despite treatment and resolution of ID (33,35). More recently, fetal-neonatal iron status has also been associated with poorer locomotor and gross motor function at 9 months of age 57 (145), and poorer fine motor function at 5 years of age (34). These findings suggest that ID during fetal and early postnatal development impairs the neurodevelopment of motor systems in which the cerebellum plays a central role. The cerebellum is also understood to play a role in cognitive and affective processes, many of which are compromised by developmental ID (146). Cerebellar dysfunction has also been implicated in the pathogenesis of neuropsychiatric diseases such as autism spectrum disorders and ADHD – disorders for which fetal ID is a risk factor (41,147). Despite the fact that the cerebellum plays a central role in motor and cognitive function and dysfunction, it is a brain region that remains relatively understudied in the context of developmental ID. In a porcine model of ID beginning at birth, relative cerebellar volume is decreased at postnatal day (PND) 61 – the rough neurodevelopmental equivalent of 1 year-of-age in humans – despite dietary iron repletion beginning at PND 32 (49). Cerebellar fractional anisotropy, a measure of tissue organization, is also significantly decreased at both PND 32 and PND 61, indicating the presence of ultrastructural disruption to the cerebellum in addition to gross volumetric disruption (49). However, in rodent models, very little is known about how ID affects the cerebellum compared to better-studied brain regions such as the hippocampus. The epigenetic modification 5-hydroxymethylcytosine (5hmC), or DNA hydroxymethylation, is highly enriched in the brain, and is established by the iron- dependent family of TET enzymes (93,99–101). Additionally, this iron-dependent epigenetic system has been shown to play a critical functional role in cerebellar neurodevelopment. Relative to NeuN+ mature neurons, immature neurons and neuronal precursors in the cerebellum are relatively depleted for 5hmC, and in mice, global 58 %5hmC increases 4-fold between P7 and P42, and 5-fold by 1 year of age (102). In cerebellar granule cells, expression of TETs is developmentally regulated, with expression of Tet1 peaking at P12, and expression of Tet3 peaking between P18-21 (104). Knockdown of Tet1 and Tet3 at P7 in developing cerebellar granule cells impairs not only the developmental accumulation of 5hmC, but also dendritic arborization of granule cells and expression of critical neurodevelopmental genes (104). Together, these studies suggest that this iron-dependent epigenetic system plays a critical functional role in cerebellar neurodevelopment. In the previous chapter, it was shown that hippocampal TET enzymes and DNA hydroxymethylation are significantly perturbed by fetal-neonatal ID. Given that this epigenetic system is critical for cerebellar neurodevelopment, and given the vulnerability of the cerebellum to developmental exposures such as ID, the aim of the present study was to determine the effect of fetal-neonatal iron deficiency anemia (IDA) on TET enzymatic activity and DNA hydroxymethylation in the developing cerebellum. The second major aim was to determine whether treatment of ID with dietary iron repletion can reverse the changes to this epigenetic system. Our results suggest that activity of the TET/DNA hydroxymethylation epigenetic system is significantly reduced by developmental ID, through reduced functionalization of TET enzymes, and that these changes are only partially reversible within this developmental timeframe. These results suggest that further investigation into the effects of ID on the cerebellum are warranted in order to define the genes that are affected by disrupting this epigenetic system, and how they might contribute to the neurodevelopmental phenotype of fetal-neonatal ID. 59 Results Assessment of Cerebellar Iron Status at P15 To determine tissue-level iron status of the cerebellum at P15, mRNA expression of Tfrc was quantified (Fig 5.1). IDA pups had significantly increased cerebellar Tfrc expression relative to IS controls (p<0.0001). TIDA pups also had significantly increased Tfrc expression relative to controls despite dietary iron repletion starting at P7 (p=0.0004). However, TIDA pups did demonstrate significant recovery of cerebellar iron status relative to untreated IDA counterparts (p<0.0001). Together with the previously shown hematocrit data (Fig 4.1A), these data demonstrate that IDA pups were iron deficient at the systemic level as well as the cerebellar tissue level at P15. Dietary iron repletion starting at P7 resulted in significant but incomplete recovery of iron status at both the systemic and cerebellar tissue levels. IS IDA TIDA 0 1 2 3 4 Tfrc Expression P15 Cerebellum R el at iv e Ex pr es si on **** **** *** Figure 5.1. Characterization of cerebellar iron status at P15. Reduction in tissue iron status of the P15 cerebellum, as indexed by a 3-fold increase in Tfrc mRNA expression. Treatment with dietary iron repletion starting at P7 resulted in significant but incomplete recovery of tissue iron status. Error bars indicate SD; *** indicates p<0.001; **** indicates p<0.0001; IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. 60 TET Activity is Reduced by Iron Deficiency in the P15 Cerebellum To determine the effect of IDA on activity of TET enzymes, TET activity was quantified from P15 cerebellar nuclear proteins (Fig 5.2A). IDA cerebellum had significantly reduced TET activity relative to IS controls (p=0.0016). Dietary iron repletion starting at P7 resulted in normalization of TET activity relative to untreated IDA counterparts (p<0.0001) that was indistinguishable from IS control levels. Regression analysis was performed to determine whether there was a continuous dose- dependent relationship between cerebellar tissue iron status, as indexed by Tfrc expression, and TET activity. This analysis demonstrated a significant inverse relationship between Tfrc expression and TET activity (Fig 5.2B; p=0.0003, R2=0.3028). Expression of Tet1-3 mRNA was quantified to determine whether reduced TET activity reflected a reduction in expression or a reduction in enzymatic functionalization. No significant change in expression of Tet1, Tet2, or Tet3 was found in IDA or TIDA cerebellum compared to IS controls (Fig 5.2C). Together, these data demonstrate that developmental IDA results in reduced TET activity in the P15 cerebellum. Dietary iron repletion restores TET activity within this developmental timeframe, prior even to complete normalization of systemic and tissue level iron status. Unchanged Tet expression suggests that the reduced TET activity reflects a true iron-dependent reduction in TET enzymatic activity. 61 Figure 5.2. Cerebellar TET activity is reduced by developmental IDA. (A) TET activity, quantified from whole nuclear protein extract, is reduced in the IDA cerebellum, but not in the TIDA cerebellum. (B) Regression analysis of TET activity, as a factor cerebellar iron status (indexed by Tfrc expression). TET activity has a significant inverse relationship with Tfrc expression. (C) Cerebellar Tet mRNA expression is unchanged by IDA or its treatment. Error bars indicate SD; Dashed lines indicate 95% CI; ** indicates p<0.01; **** indicates p<0.0001 IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. Cerebellar DNA hydroxymethylation is Reduced by Iron Deficiency In order to determine the functional effect of reduced TET activity, the enzymatic product of TET, 5hmC, was quantified from P15 cerebellar genomic DNA. Global %5hmC was significantly reduced in the TIDA group relative to IS controls, with IS IDA TIDA 0.0 0.5 1.0 1.5 Tet1 R el at iv e Ex pr es si on ns IS IDA TIDA 0.0 0.5 1.0 1.5 Tet2 R el at iv e Ex pr es si on ns IS IDA TIDA 0.0 0.5 1.0 1.5 Tet3 R el at iv e Ex pr es si on ns 0 1 2 3 4 5 0.0 0.5 1.0 1.5 2.0 Transform of CRB: TET activity vs. Tfrc Tfrc (Relative Expression) A ct iv ity (R el at iv e) p=0.0003 R2=0.3028 IS IDA TIDA 0.0 0.5 1.0 1.5 2.0 P15 Cerebellum A ct iv ity (R el at iv e) ** **** A C B 62 %5hmC reduced, but not statistically significant, in the IDA group relative to IS controls (Fig 5.3A). Regression analysis demonstrated a modest but statistically significant relationship between systemic (p=0.0356, R2=0.2971) and tissue-level (p=0.0561, R2=0.1004) iron status and %5hmC, with poorer iron status being correlated with decreasing %5hmC (Fig 5.3B). These results suggest a functional effect of reduced TET enzymatic activity during iron deficiency. Unlike TET activity however, the cerebellar %5hmC reduction does not appear to be as rapidly reversible with dietary iron repletion within this developmental timeframe. Figure 5.3. Cerebellar %5hmC is modestly reduced by developmental IDA. (A) Global %5hmC is significantly reduced in the P15 TIDA cerebellum. The IDA cerebellum did not reach statistical significance. (B) Regression analysis of %5hmC as a function of systemic (hematocrit) or cerebellar (Tfrc expression) iron status. There is a significant positive relationship between %5hmC and hematocrit, and a marginally significant inverse relationship between %5hmC and Tfrc mRNA expression. Both demonstrate a modest relationship between decreasing iron status and reduction in %5hmC. Error bars indicate SD; Dashed lines indicate 95% CI; * indicates p<0.05 IS: Iron sufficient; IDA: Iron deficient anemic; TIDA: Treated iron deficient anemic. 0 1 2 3 4 5 0.0 0.1 0.2 0.3 Tfrc (Relative Expression) % 5h m C P15 Cerebellum %5hmC vs. Tfrc Goodness of Fit: R2=0.1004 Is slope significantly non-zero? p=0.0561 0 10 20 30 40 0.0 0.1 0.2 0.3 0.4 Hematocrit % 5h m C P15 Cerebellum %5hmC vs. Hct Goodness of Fit: R2=0.2971 Is slope significantly non-zero? p=0.0356 A B p=0.0561 R2=0.1004 p=0.0356 R2=0.2971 IS IDA TIDA 0.0 0.1 0.2 0.3 P15 Cerebellum Non-linear Curve % 5h m C ANOVA overall p=0.0257 IS vs. IDA: p=0.1139 IS vs. TIDA: p=0.0249 IDA vs. TIDA: p=0.8146 * 63 Discussion In the present study, we sought to determine whether activity of iron-dependent TET enzymes, and establishment of their enzymatic product DNA hydroxymethylation, is altered in the developing rat cerebellum by fetal-neonatal ID. In the previous chapter, it was shown that physiologic levels of ID were sufficient to increase activity of TETs and subsequent establishment of DNA hydroxymethylation by upregulating Tet expression in hippocampus. By studying this same epigenetic system in the cerebellum, we sought to determine whether it was similarly affected across brain regions at this stage of neurodevelopment. The first major finding of these studies is that, unlike the hippocampus, cerebellar TET enzymatic activity is significantly reduced by developmental IDA. Because Tet expression remained unchanged, the reduction most likely reflects a reduced enzyme functionalization due to loss of the critical cofactor iron from the TET active site. This is consistent with previous reports showing that chelation of iron completely ablates TET enzymatic activity; however, iron chelation models a system where there is a complete loss if iron from cells (93). The present results are novel because they demonstrate that a physiologic level of iron deficiency, where the loss of iron is not absolute, is sufficient to compromise TET enzymatic function. This likely reflects a scenario in which intracellular iron is not absent, but it is insufficient to outcompete other divalent metal cations for binding to the TET active site, resulting in a mismetallated and catalytically inactive enzyme (148,149). An additional finding was %5hmC was only modestly reduced in the IDA cerebellum despite the significant reduction of TET enzymatic activity. This suggests that at baseline, TET activity might be present in excess, and that a 64 more severe loss of TET enzymatic function might be required to more severely affect establishment of 5hmC. The second major aim of the study was to determine whether the observed changes in TET activity and DNA hydroxymethylation were reversible with treatment of IDA by dietary iron repletion. The present study finds that TET enzymatic activity recovers fully with dietary iron repletion starting at P7. There are two potential mechanisms by which this recovery could occur. First, existing TETs could become metallated with iron as iron status begins to recover and iron becomes available. Second, protein turnover between P7 and P15 could generate newly synthesized TETs that are appropriately metallated with iron, overcoming the TET activity deficit associated with mismetallated TETs. The present experiments cannot distinguish which of these mechanisms are responsible for the restoration of TET activity, however, future experiments in protein expression and metalation dynamics have the potential to elucidate the mechanism. Notably, TET enzymatic activity recovered in the TIDA group despite the fact that both systemic- and tissue-level iron status are not completely recovered to baseline by P15. This suggests that there may be a threshold of ID above which cells continue to be able to appropriately metallate TETs with iron despite suboptimal iron levels, but beyond which there is no longer sufficient intracellular iron for TETs to be appropriately metallated. Future studies of TET activity at different levels of recovery of iron status between P7 and P15 might indicate where this threshold lies. In contrast to TET activity, global %5hmC does not recover to baseline levels with dietary iron repletion by P15. Global DNA hydroxymethylation in the rodent CNS 65 increases as much as 5-fold between birth and adulthood (102). However, understanding when the bulk of this accumulation occurs, and whether it accumulates steadily or during specific developmental windows is critical to understanding how ID affects 5hmC both acutely and long-term. Two separate studies have found that relatively little accumulation of 5hmC occurs in the rodent brain between birth and P12-14, and that the bulk of 5hmC accumulation occurs after P14 (104,150). This is consistent with data showing that expression of TETs peaks between P12-18 (104). Thus, one possible explanation for the present finding is that recovery of 5hmC lags behind recovery of TETs because the TET/5hmC epigenetic system is not particularly active between P7 and P15 when iron treatment is administered, making recovery slow. In such a scenario, recovery of global 5hmC might be possible later in postnatal development as the TET/5hmC system naturally becomes more active. Alternatively, the failure of %5hmC to recover with dietary iron repletion might indicate that a critical period for establishment of DNA hydroxymethylation was missed. While this might be true at the level of individual genetic loci, the fact that the bulk of 5hmC accumulation occurs after P15 indicates that this possibility is less likely. Quantifying 5hmC in the TIDA cerebellum at later developmental timepoints could help to determine whether %5hmC does recover with time, and whether focal deficits in 5hmC remain. The cerebellum remains an understudied brain region in the field of developmental ID, though emerging data suggests that, like better studied regions such as the hippocampus, the cerebellum is also affected by ID (49). The data presented in this chapter demonstrate that the cerebellum experiences significant reduction in activity of the TET/DNA hydroxymethylation epigenetic system during developmental ID, though 66 the exact consequences of this are unclear. Transcriptomic analysis of the developmentally IDA cerebellum combined with locus-specific analyses of DNA hydroxymethylation has the potential to lend insight into how perturbation of this system by ID affects gene expression in the cerebellum. Additionally, given the role that TETs and DNA hydroxymethylation play in dendritic arborization of cerebellar neurons (104), structural analysis of developmentally ID cerebellar neurons may provide insight into how ID affects the cerebellum at the ultrastructural level. Developing a better understanding of how the cerebellum is affected by ID, and what mechanisms drive these changes, will ultimately lead to a better understanding of how cerebellar dysfunction contributes to the neurodevelopmental phenotype of fetal-neonatal ID. 67 Chapter 6: Conclusion and Future Directions It has long been established that fetal-neonatal iron deficiency has a lasting negative impact on neurodevelopment. These lasting impacts of ID range from subtle cognitive and socioemotional deficits to increased risk for frank neuropsychiatric dysfunction such as autism, ADHD, intellectual disability, and schizophrenia (25,41). Decades of studies in preclinical models have gradually specified the gross structural, ultrastructural, and molecular underpinnings of these lasting deficits and increased disease risks (127). The data presented in this dissertation builds on this large body of work by identifying the specific role of ID in neuronal gene dysregulation, as well as characterizing the effect of developmental IDA on a novel iron-dependent epigenetic system in the developing brain. In Chapter 3, we found that chronic ID, independent of anemia, is sufficient to cause widespread dysregulation of hippocampal gene expression, with dysregulated genes mapping to critical neurodevelopmental signaling pathways and functions. Additionally, early restoration of iron status does not fully restore adult hippocampal gene expression, raising the question of how developmental ID is able to reprogram gene expression beyond the initial period of ID. In Chapters 4 and 5, the novel iron-dependent epigenetic modification 5-hydroxymethylcytosine (5hmC), or DNA hydroxymethylation, was studied in the developmentally IDA brain as a potential mechanism by which these ID-induced gene expression changes might be established and maintained. Here we found that DNA hydroxymethylation is significantly perturbed in the developmentally ID hippocampus and cerebellum, through perturbation of the iron-dependent TET enzymes 68 that generate DNA hydroxymethylation. In considering the significance and implications of these experimental results as a whole, several additional questions and directions for future studies become apparent. Determining the relationship between DNA hydroxymethylation and gene dysregulation Dysregulation of the epigenome by environmental exposures is of significant concern because of the role that the epigenome plays in regulation of gene expression. If the epigenome and its associated machinery are disrupted, then downstream gene expression may also be disrupted. In the context of neurodevelopment, where precisely timed regulation of gene expression is critical for proper structural and functional development of the brain, dysregulation of gene expression through perturbation of the epigenome could have significant consequences for brain structure and function. Linking epigenetic changes to the gene expression changes they enact is an important step towards understanding the mechanisms by which environmental exposures, such as developmental ID, drive functional phenotypic changes and increased disease risks. In Chapter 3, it was shown that hippocampal neuronal-specific ID results in widespread dysregulation of hippocampal gene expression, and that dysregulated genes map to canonical signaling pathways and functions that are critical for neurodevelopment and neuronal function. In Chapter 4, it was shown that hippocampal establishment of the iron-dependent epigenetic modification 5hmC is significantly increased by developmental IDA, through increased expression and activity of hippocampal TET enzymes. Demonstrating a tentative link between these two studies, an analysis of the 346 genes dysregulated in the CID mouse hippocampus (Fig 3.1) found that 29% of these 69 dysregulated genes are known to be enriched for 5hmC in neurons (Fig 6.1) (103). This analysis suggests that genes normally enriched for 5hmC might be particularly vulnerable to dysregulation by ID, given the sensitivity of 5hmC to ID. However, further studies are required in order to solidify the link between this epigenetic modification and gene dysregulation by developmental ID. Figure 6.1. Genes enriched for 5hmC are differentially expressed in the CID hippocampus. 346 genes were dysregulated in the CID mouse hippocampus compared to iron sufficient controls (see Fig. 3.1). Of these genes, 100 have been found in a previous study (103) to be enriched for 5hmC. Pathway analysis was performed on these 100 genes, and found that they mapped to several pathways critical for neuronal development, function, and synaptic plasticity. CID: Chronically iron deficient. Quantification of 5hmC at the regional or base-pair level using unbiased genome- wide techniques has the potential to make the link between 5hmC and gene expression when combined with transcriptomic data. Examples of such techniques include hydroxymethylated DNA immunoprecipitation followed by sequencing (hMeDIP-seq), which provides region-level resolution of 5hmC, and oxidative bisulfite sequencing, which provides base pair-level resolution of 5hmC. These techniques, combined with 100 (29%) 246 (71%) Enriched for 5hmC Not enriched for 5hmC Pathways: • Actin cytoskeleton signaling • Axonal guidance signaling • CXCR4 signaling • Eprin/Ephrin Receptor signaling • ErbB signaling • Erk/MAPK signaling 70 transcriptomics, can identify specific changes to 5hmC that can be correlated to changes in expression of corresponding genes. In the rat model of dietary IDA presented in Chapters 4 and 5, performing such experiments on the P15 IDA and TIDA hippocampus and cerebellum could identify genes whose differential expression may be attributable to differential 5hmC. Analysis of the IDA tissues would identify genes affected by active, ongoing ID, whereas analysis of TIDA tissues would identify the group of genes that become permanently dysregulated despite iron repletion by reprogramming of 5hmC. Identifying these changes would be an important step towards understanding how reprogramming of the epigenome contributes to the neurodevelopmental phenotype of ID. Response of the TET/5hmC system to IDA is tissue-specific Chapters 4 and 5 demonstrated that the TET/DNA hydroxymethylation epigenetic system is significantly altered at P15 by developmental IDA in both the hippocampus and cerebellum. However, the directionality of change was different in the two brain regions, as summarized in Figure 6. In the IDA hippocampus, TET activity was increased, likely due to increased expression of Tet2 and Tet3, resulting in increased global %5hmC (Fig 6.2A). In contract, cerebellar TET activity was reduced by IDA, resulting in a modest reduction in global %5hmC (Fig 6.2B). Because cerebellar Tet expression remained unchanged, this reduction in activity most likely reflected reduced enzyme functionalization due to loss of the iron cofactor from the TET active site. 71 Figure 6.2. Model summarizing the effect of developmental ID on TET activity and DNA hydroxymethylation. (A) In the P15 IDA hippocampus, TET activity is significantly increased, likely driven by increased Tet expression. This ultimately leads to an increase in the TET enzymatic product, 5hmC. (B) In the P15 IDA cerebellum, TET activity is significantly reduced, independent of a change in Tet expression, likely reflecting decreased enzymatic function due to loss of the iron cofactor from the TET active site. This leads to a reduction in the TET enzymatic product, 5hmC. That TETs remain metallated with iron and functional in one tissue, and insufficiently metallated with iron to maintain enzymatic activity in another raises the question of whether the IDA hippocampus and cerebellum experience the same degree of iron deficiency at P15. Comparison of Tfrc upregulation demonstrates that the IDA cerebellum mounts a slightly greater relative upregulation of Tfrc (3.10-fold increase) than the IDA hippocampus (2.85-fold increase) compared to IS controls. This greater upregulation of Tfrc suggests that the IDA cerebellum is more iron deficient than the IDA hippocampus. One possible explanation for this difference is that the hippocampus might have greater intracellular iron storage capacity compared to the cerebellum. Storage iron, in the form of ferritin, could buffer intracellular iron levels even under conditions of systemic A BHippocampus Cerebellum 72 iron deficiency (148). This would allow the hippocampus to withstand systemic iron deficiency without experiencing significant intracellular effects such as impaired protein metallation. Quantification of tissue iron content in the P15 IDA hippocampus and cerebellum could confirm whether there is indeed a difference in degree of tissue iron deficiency that might underlie this difference in response. Another possible explanation is that the P15 cerebellum might have greater iron demand than the hippocampus, making it more vulnerable to iron depletion under conditions of low systemic iron. Directly comparing baseline, non-ID expression of cellular iron uptake machinery in the two brain regions could confirm whether the cerebellum has a greater iron demand at P15. Likely at the root of both of these possibilities is that the hippocampus and cerebellum have different neurodevelopmental trajectories. In the rodent hippocampus, the bulk of neurons become post-mitotic prenatally, between E13 and E20 (137,151). The exception is dentate granule cells, which are generated through the first postnatal week (137,151). In contrast, cerebellar neurogenesis stretches from E12 through P15 (147,152). This is largely due to the extended period of neurogenesis of cerebellar granule cells (147). Thus, at P15 when tissues were collected and analyzed for the described experiments, the hippocampus and the cerebellum are at significantly different stages of neurodevelopment. The more mature hippocampus is post-mitotic and undergoing morphologic and synaptic development, whereas the cerebellum is completing neurogenesis and beginning morphologic development. In the context of this different developmental timing hypothesis, the cerebellum may have a greater iron demand than the hippocampus at P15 because it is undergoing neurodevelopmental processes that have greater energy and iron substrate requirements 73 such as terminal differentiation. Another possibility is that the hippocampus might have greater accumulated intracellular iron stores than the cerebellum due to the fact that the majority of hippocampal neurons have been post-mitotic since before birth. In contrast, many cerebellar granule cells are only recently post-mitotic, giving them little time to build intracellular iron stores to buffer their iron status in the face of systemic ID. Ongoing studies of the TET/DNA hydroxymethylation system in a more neurodevelopmentally immature hippocampus aim to determine whether the difference in TET response to ID between the two brain regions can be attributed to degree of neurodevelopmental maturity. The role of TET enzymes in DNA demethylation Disruption of TET activity by developmental ID also has implications for global DNA methylation, given the central role of TETs in active DNA demethylation (Fig 1.1). Like DNA hydroxymethylation, DNA methylation is dynamic and changes across neurodevelopment (100). Thousands of genetic loci become hypomethylated in the adult brain compared to the fetal brain, and this developmentally-timed loss of methylation is TET2-dependent (100). Hypomethylation of these loci is accompanied by gain of histone modifications associated with active chromatin and increased DNAse sensitivity, indicating that these changes likely have functional significance for gene expression (100). Thus, in order to understand the full impact of ID-induced TET disruption on gene expression, it will be important to study global DNA methylation at a locus-specific level in addition to DNA hydroxymethylation. Genome-wide bisulfite sequencing has previously identified 229 differentially methylated loci in the P15 TIDA hippocampus, 74 that map to 108 genes (89). Similar experiments in the untreated IDA hippocampus and cerebellum at P15 could lend further insight into how disruption of DNA methylation, via disruption of TET activity, might contribute to gene dysregulation by developmental ID. Iron deficiency and the epigenome The data presented here adds to a body of work demonstrating that developmental ID significantly reprograms the epigenome of the developing brain by reprogramming DNA methylation (77,89) and histone methylation (84). Here we present evidence that an additional epigenetic modification, DNA hydroxymethylation, is also sensitive to developmental ID. In considering how these epigenetic changes affect gene expression, it is important to also consider how the epigenome functions as a whole. Each epigenetic modification has a canonical effect on gene expression, and altering that epigenetic modification might be expected to affect the way that a gene is expressed in accordance with the canonical function of that modification. However, in an environmental exposure paradigm such as developmental ID, multiple co-occurring epigenetic modifications are altered. Therefore, in considering the effect of an exposure on gene expression, one has to consider not only the effects of individual modifications on gene expression, but also how changes to multiple co-occurring modifications throughout a gene and its regulatory elements might cumulatively interact with each other the shape gene expression. A broader goal of this research is to understand the contributions of each component of the epigenome and how the various components of the epigenome interact to shape the altered gene expression seen both acutely during ID and chronically after recovery. 75 Epigenetics and the DOHaD hypothesis Finally, it is important to consider the implications of these findings in the context of the Developmental Origins of Health and Disease (DOHaD) hypothesis. Fetal-neonatal ID perfectly models the DOHaD hypothesis, which posits that the fetal and early postnatal environment shapes later life health and disease risk (153). In the case of fetal- neonatal ID, developmental exposure to ID is associated with lifetime neuropsychiatric disease risk, including increased risk for disorders such as intellectual disability, autism spectrum disorders, ADHD, and schizophrenia (26,27,41). There are two main hypotheses for how developmental exposure to ID drives this dysfunction and enacts the DOHaD hypothesis. The first hypothesis is that ID reprograms gene expression in a way that permanently disrupts function. The second hypothesis is that ID during neurodevelopment causes permanent structural, ultrastructural, and connectivity defects in the brain, and that these structural defects drive dysfunction. What role does disruption of the epigenome play in the DOHaD paradigm, and the two hypotheses underlying it? First, disruption of the epigenome is clearly implicated in the gene dysregulation hypothesis, as the epigenome regulates gene expression. As demonstrated here and in the literature (77,84,89), several epigenetic modifications with a clear mechanistic iron-dependence become dysregulated under conditions of developmental iron deficiency. Relevant to the DOHaD hypothesis, changes to these epigenetic modifications persist despite treatment of ID (77,89) and even into adulthood (84). This suggests that ID may induce a permanent reprogramming of the epigenome, and thus, permanent reprogramming of gene expression and function. 76 Disruption of the epigenome by ID also plays a significant role in the structural defects hypothesis. Structural development of the central nervous system requires precisely timed gene expression during specific critical periods of development. Developmentally-timed changes to the epigenome at least in part orchestrate these dynamic changes in gene expression, and it is well-characterized that the epigenome is highly dynamic across neurodevelopment (100,102,103). If the iron-dependent machinery that orchestrates these epigenetic changes, such as histone demethylases (84) or TETs, is compromised by developmental ID, then the dynamics of gene expression will be compromised as well. Without the ability to express the required genes at the required times, structural development of the CNS will be compromised under conditions of ID. Considering these two hypotheses in the context of developmental ID, it becomes apparent that the gene dysregulation and structural defects hypotheses are not mutually exclusive, but rather, it is their combined and interactive effect that enacts the DOHaD hypothesis. As an example, it has previously been shown by our lab that developmental ID results in compromised dendritic morphology of hippocampal neurons (42,43,54–56). Underlying these morphological changes are changes in expression of critical genes involved in neuronal morphological development (48,56,90,91). For at least one of these genes, the neurotrophic factor Bdnf, it is known that the dysregulated gene expression is driven by underlying iron-dependent modulation of multiple epigenetic modifications (84). In this conceptualization, underlying modulation of the epigenome by developmental ID ties the gene dysregulation and structural defects hypotheses together to ultimately enact the DOHaD hypothesis. 77 Because of the central role that the epigenome plays in the DOHaD hypothesis, use of epigenetic modulators in treatment of ID is enticing. For example, supplemental choline – a donor into the one-carbon metabolism pathway – has shown some efficacy in treatment of ID. In preclinical models, supplemental choline during late gestation partially rescues gene expression in the ID hippocampus (47), and postnatal choline supplementation rescues recognition memory deficits in adult formerly iron deficient rats (154). However, there are several caveats to the use of epigenetic modulators in treatment of ID. First, given that multiple epigenetic modifications are altered by ID, the therapeutic efficacy of modulating any single modification is unclear. Second, the efficacy of epigenetic modulators has to be considered in the context of critical periods of neurodevelopment. For example, once dendritic arborization is complete, altering gene expression by treatment with an epigenetic modulator cannot reverse the structural damage of ID on dendritic arbors – the developmental process is complete and the critical period is closed. For this reason, appropriately timed use of epigenetic modulators in relation to neurodevelopmental critical periods might prove to be most efficacious. Clinical Implications Due to the high worldwide prevalence of general and nutrient-specific malnutrition, particularly among pregnant women and young children, understanding the effects of nutrient deficiencies on long-term neurocognitive outcomes and risk for neuropsychiatric disorders is critical because sound nutritional policy and practice can potentially mitigate risk. Prevention of nutrient perturbations during fetal and early postnatal life is an accomplishable goal with long-term societal implications. Women of 78 child-bearing age should be in optimal nutritional health entering into pregnancy, particularly with respect to critical macronutrients such as protein, long chain polyunsaturated fatty acids (LC-PUFAs), and glucose, as well as select micronutrients such as iron, zinc, iodine, and vitamins. Prevention strategies to optimize fetal brain nutritional status throughout pregnancy include not only maintaining maternal nutritional sufficiency but also reducing conditions in nutritionally sufficient women that nevertheless alter fetal nutritional status. These conditions include maternal hypertension, the most common cause of intrauterine growth restriction (IUGR) in developed countries (155), gestational diabetes mellitus (156), maternal smoking (157), obesity/excessive gestational weight gain (158,159), and maternal stress (160). Postnatally, provision of human milk and maintenance of the status of critical nutrients are key prevention strategies. While promising preclinical data suggests that dietary epigenetic modulators such as choline may have a beneficial therapeutic effect on neurodevelopment in the face of nutritional insufficiency (47,154), it is too soon to propose these methyl diets as routine therapeutic interventions until more is known about the optimal timing, dose and duration of these agents (161). 79 References 1. Cuthbert BN, Insel TR. Toward the future of psychiatric diagnosis: The seven pillars of RDoC. BMC Med. 2013 Dec 14;11(1):126. 2. Georgieff MK, Brunette KE, Tran P V. Early life nutrition and neural plasticity. Dev Psychopathol. 2015 May 1;27(2):411–23. 3. González HF, Visentin S. Micronutrients and neurodevelopment: An update. Arch Argent Pediatr. 2016;114(6):570–5. 4. Cusick SE, Georgieff MK. The Role of Nutrition in Brain Development: The Golden Opportunity of the “First 1000 Days.” J Pediatr. 2016;175:16–21. 5. Krebs NF, Lozoff B, Georgieff MK. Neurodevelopment: The Impact of Nutrition and Inflammation During Infancy in Low-Resource Settings. Pediatrics. 2017 Apr 1;139(Supplement 1):S50–8. 6. Abu-Saad K, Fraser D. Maternal nutrition and birth outcomes. Epidemiol Rev. 2010 Apr 1;32(1):5–25. 7. Bale TL, Baram TZ, Brown AS, Goldstein JM, Insel TR, McCarthy MM, et al. Early life programming and neurodevelopmental disorders. Biol Psychiatry. 2010;68(4):314–9. 8. Davis J, Eyre H, Jacka FN, Dodd S, Dean O, McEwen S, et al. A review of vulnerability and risks for schizophrenia: Beyond the two hit hypothesis. Neurosci Biobehav Rev. 2016;65:185–94. 9. Brown AS, Susser ES. Prenatal nutritional deficiency and risk of adult schizophrenia. Schizophr Bull. 2008 Aug 20;34(6):1054–63. 10. Susser ES, Lin SP. Schizophrenia After Prenatal Exposure to the Dutch Hunger Winter of 1944-1945. Arch Gen Psychiatry. 1992 Dec 1;49(12):983–8. 11. Susser E, Neugebauer R, Hoek HW, Brown AS, Lin S, Labovitz D, et al. Schizophrenia after prenatal famine. Further evidence. Arch Gen Psychiatry. 1996 Jan 1;53(1):25–31. 12. St Clair D, Xu M, Wang P, Yu Y, Fang Y, Zhang F, et al. Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959-1961. J Am Med Assoc. 2005 Aug 3;294(5):557–62. 13. Barker DJP. Fetal nutrition and cardiovascular disease in later life. Br Med Bull. 1997 Jan;53(1):96–108. 14. Gluckman PD, Hanson MA. Living with the past: Evolution, development, and patterns of disease. Science (80- ). 2004 Sep 17;305(5691):1733–6. 15. Painter RC, Roseboom TJ, Bleker OP. Prenatal exposure to the Dutch famine and disease in later life: an overview. Reprod Toxicol. 2005 Jan;20(3):345–52. 16. Lopuhaa CE. Atopy, lung function, and obstructive airways disease after prenatal exposure to famine. Thorax. 2000 Jul 1;55(7):555–61. 17. Ravelli A, van der Meulen J, Michels R, Osmond C, Barker D, Hales C, et al. Glucose tolerance in adults after prenatal exposure to famine. Lancet. 1998 Jan;351(9097):173–7. 18. Lumey L, Stein A, Kahn H, Romijn J. Lipid profiles in middle-aged men and women after famine exposure during gestation: the Dutch Hunger Winter Families Study. Am J Clin Nutr. 2009 Jun;89(6):1737–43. 80 19. Roseboom, TJ; van der Meulen, JHP; Osmond, C; Barker, DJP; Ravelli, ACJ; Schroeder-Tanka, JM; van Montfrans, GA; Michels, RPJ; Bleker O. Coronary heart disease after prenatal exposure to the Dutch famine, 1944-45. Heart. 2000 Dec 1;84(6):595–8. 20. Hensch TK. Critical Period Regulation. Annu Rev Neurosci. 2004 Jul 21;27(1):549–79. 21. Clancy B, Darlington RB, Finlay BL. The course of human events: Predicting the timing of primate neural development. Dev Sci. 2000 Mar 1;3(1):57–66. 22. Holliday MA. Body composition and energy needs during growth. In: Falkner F, Tanner JM, editors. Human Growth: A Comprehensive Treatise. Boston, MA: Plenium Press; 1986. p. 101–17. 23. Kretchmer N, Beard JL, Carlson S. The role of nutrition in the development of normal cognition. Am J Clin Nutr. 1996 Jun 1;63(6):997S-1001S. 24. Georgieff MK, Tran P V, Carlson ES. Atypical fetal development: Fetal alcohol syndrome, nutritional deprivation, teratogens, and risk for neurodevelopmental disorders and psychopathology. Dev Psychopathol. 2018;30(3):1063–86. 25. Lozoff B, Beard J, Connor J, Barbara F, Georgieff M, Schallert T. Long-lasting neural and behavioral effects of iron deficiency in infancy. Nutr Rev. 2006 May;64(5 Pt 2):S34-43; discussion S72-91. 26. Insel BJ, Schaefer CA, McKeague IW, Susser ES, Brown AS. Maternal iron deficiency and the risk of schizophrenia in offspring. Arch Gen Psychiatry. 2008 Oct 6;65(10):1136–44. 27. Schmidt RJ, Tancredi DJ, Krakowiak P, Hansen RL, Ozonoff S. Maternal intake of supplemental iron and risk of autism spectrum disorder. Am J Epidemiol. 2014 Nov 1;180(9):890–900. 28. Sørensen HJ, Nielsen PR, Pedersen CB, Mortensen PB. Association between prepartum maternal iron deficiency and offspring risk of schizophrenia: Population-based cohort study with linkage of danish national registers. Schizophr Bull. 2011 Sep 1;37(5):982–7. 29. McLean E, Cogswell M, Egli I, Wojdyla D, de Benoist B, KOLLER O, et al. Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993–2005. Public Health Nutr. 2009 Apr 23;12(04):444. 30. Yip R. Iron deficiency: contemporary scientific issues and international programmatic approaches. J Nutr. 1994 Aug;124(8 Suppl):1479S-1490S. 31. Mei Z, Cogswell ME, Looker AC, Pfeiffer CM, Cusick SE, Lacher DA, et al. Assessment of iron status in US pregnant women from the National Health and Nutrition Examination Survey (NHANES), 1999-2006. Am J Clin Nutr. 2011 Jun 1;93(6):1312–20. 32. Cogswell ME, Looker AC, Pfeiffer CM, Cook JD, Lacher DA, Beard JL, et al. Assessment of iron deficiency in US preschool children and nonpregnant females of childbearing age: National Health and Nutrition Examination Survey 2003- 2006. Am J Clin Nutr. 2009 May 1;89(5):1334–42. 33. Lozoff B, Jimenez E, Wolf AW. Long-Term Developmental Outcome of Infants with Iron Deficiency. N Engl J Med. 1991 Sep 5;325(10):687–94. 34. Tamura T, Goldenberg RL, Hou J, Johnston KE, Cliver SP, Ramey SL, et al. Cord 81 serum ferritin concentrations and mental and psychomotor development of children at five years of age. J Pediatr. 2002;140(2):165–70. 35. Lozoff B, Jimenez E, Hagen J, Mollen E, Wolf AW. Poorer behavioral and developmental outcome more than 10 years after treatment for iron deficiency in infancy. Pediatrics. 2000 Apr;105(4):E51. 36. Lukowski AF, Koss M, Burden MJ, Jonides J, Nelson CA, Kaciroti N, et al. Iron deficiency in infancy and neurocognitive functioning at 19 years: evidence of long-term deficits in executive function and recognition memory. Nutr Neurosci. 2010 Apr 19;13(2):54–70. 37. Lozoff B, Smith JB, Kaciroti N, Clark KM, Guevara S, Jimenez E. Functional Significance of Early-Life Iron Deficiency: Outcomes at 25 Years. J Pediatr. 2013;163(5):1260–6. 38. Corapci F, Radan AE, Lozoff B. Iron deficiency in infancy and mother-child interaction at 5 years. J Dev Behav Pediatr. 2006 Oct;27(5):371–8. 39. Chang S, Wang L, Wang Y, Brouwer ID, Kok FJ, Lozoff B, et al. Iron-deficiency anemia in infancy and social emotional development in preschool-aged Chinese children. Pediatrics. 2011 Apr;127(4):e927-33. 40. Lozoff B. Iron deficiency and child development. Food Nutr Bull. 2007;28(4 Suppl):S560-71. 41. Wiegersma AM, Dalman C, Lee BK, Karlsson H, Gardner RM. Association of Prenatal Maternal Anemia With Neurodevelopmental Disorders. JAMA Psychiatry. 2019 Sep 18;1. 42. Jorgenson LA, Wobken JD, Georgieff MK. Perinatal iron deficiency alters apical dendritic growth in hippocampal CA1 pyramidal neurons. Dev Neurosci. 2003;25(6):412–20. 43. Fretham SJB, Carlson ES, Wobken J, Tran P V, Petryk A, Georgieff MK. Temporal manipulation of transferrin-receptor-1-dependent iron uptake identifies a sensitive period in mouse hippocampal neurodevelopment. Hippocampus. 2012 Aug;22(8):1691–702. 44. Tyagi E, Zhuang Y, Agrawal R, Ying Z, Gomez-Pinilla F. Interactive actions of Bdnf methylation and cell metabolism for building neural resilience under the influence of diet. Neurobiol Dis. 2015;73:307–18. 45. Zeisel S, Steven. Choline, Other Methyl-Donors and Epigenetics. Nutrients. 2017 Apr 29;9(5):445. 46. Ly A, Ishiguro L, Kim D, Im D, Kim SE, Sohn KJ, et al. Maternal folic acid supplementation modulates DNA methylation and gene expression in the rat offspring in a gestation period-dependent and organ-specific manner. J Nutr Biochem. 2016;33:103–10. 47. Tran P V, Kennedy BC, Pisansky MT, Won K-J, Gewirtz JC, Simmons RA, et al. Prenatal Choline Supplementation Diminishes Early-Life Iron Deficiency-Induced Reprogramming of Molecular Networks Associated with Behavioral Abnormalities in the Adult Rat Hippocampus. J Nutr. 2016 Mar;146(3):484–93. 48. Barks A, Fretham SJ, Georgieff MK, Tran P V. Early-life neuronal-specific iron deficiency alters the adult mouse hippocampal transcriptome. J Nutr. 2018 Oct 1;In Press(10):1521–8. 82 49. Mudd AT, Fil JE, Knight LC, Dilger RN. Dietary iron repletion following early- life dietary iron deficiency does not correct regional volumetric or diffusion tensor changes in the developing pig brain. Front Neurol. 2018 Jan 11;8(JAN):735. 50. Leyshon BJ, Radlowski EC, Mudd AT, Steelman AJ, Johnson RW. Postnatal Iron Deficiency Alters Brain Development in Piglets. J Nutr. 2016 Jul 1;146(7):1420–7. 51. Ranade SC, Rose A, Rao M, Gallego J, Gressens P, Mani S. Different types of nutritional deficiencies affect different domains of spatial memory function checked in a radial arm maze. Neuroscience. 2008;152(4):859–66. 52. Rao R, Tkac I, Schmidt AT, Georgieff MK. Fetal and neonatal iron deficiency causes volume loss and alters the neurochemical profile of the adult rat hippocampus. Nutr Neurosci. 2011 Mar;14(2):59–65. 53. Connor JR, Menzies SL. Relationship of iron to oligodendrocytes and myelination. Glia. 1996 Jun;17(2):83–93. 54. Brunette KE, Tran P V, Wobken JD, Carlson ES, Georgieff MK. Gestational and neonatal iron deficiency alters apical dendrite structure of CA1 pyramidal neurons in adult rat hippocampus. Dev Neurosci. 2010 Aug;32(3):238–48. 55. Carlson ES, Tkac I, Magid R, O’Connor MB, Andrews NC, Schallert T, et al. Iron is essential for neuron development and memory function in mouse hippocampus. J Nutr. 2009 May;139(4):672–9. 56. Bastian TW, von Hohenberg WC, Mickelson DJ, Lanier LM, Georgieff MK. Iron Deficiency Impairs Developing Hippocampal Neuron Gene Expression, Energy Metabolism, and Dendrite Complexity. Dev Neurosci. 2016 Sep 27;38(4):264–76. 57. Greminger AR, Lee DL, Shrager P, Mayer-Proschel M. Gestational Iron Deficiency Differentially Alters the Structure and Function of White and Gray Matter Brain Regions of Developing Rats. J Nutr. 2014 Jul 1;144(7):1058–66. 58. Monk C, Georgieff MK, Xu D, Hao X, Bansal R, Gustafsson H, et al. Maternal prenatal iron status and tissue organization in the neonatal brain. Pediatr Res. 2016 Mar;79(3):482–8. 59. Ball G, Srinivasan L, Aljabar P, Counsell SJ, Durighel G, Hajnal J V, et al. Development of cortical microstructure in the preterm human brain. Proc Natl Acad Sci. 2013 Jun 4;110(23):9541–6. 60. Kroenke CD, Van Essen DC, Inder TE, Rees S, Bretthorst GL, Neil JJ. Microstructural Changes of the Baboon Cerebral Cortex during Gestational Development Reflected in Magnetic Resonance Imaging Diffusion Anisotropy. J Neurosci. 2007 Nov 14;27(46):12506–15. 61. Fukumitsu K, Fujishima K, Yoshimura A, Wu YK, Heuser J, Kengaku M. Synergistic Action of Dendritic Mitochondria and Creatine Kinase Maintains ATP Homeostasis and Actin Dynamics in Growing Neuronal Dendrites. J Neurosci. 2015;35(14):5707–23. 62. Oruganty-Das A, Ng T, Udagawa T, Goh ELK, Richter JD. Translational control of mitochondrial energy production mediates neuron morphogenesis. Cell Metab. 2012;16(6):789–800. 63. Dallman PR. Biochemical Basis for the Manifestations of Iron Deficiency. Annu Rev Nutr. 1986 Jul;6(1):13–40. 64. Bastian TW, Hohenberg WC von, Georgieff MK, Lanier LM. Chronic Energy 83 Depletion due to Iron Deficiency Impairs Dendritic Mitochondrial Motility during Hippocampal Neuron Development. J Neurosci. 2019;39(5):802. 65. Connor JR, Pavlick G, Karli D, Menzies SL, Palmer C. A histochemical study of iron-positive cells in the developing rat brain. J Comp Neurol. 1995 Apr 24;355(1):111–23. 66. Taylor EM, Morgan EH. Developmental changes in transferrin and iron uptake by the brain in the rat. Dev Brain Res. 1990;55(1):35–42. 67. Morath DJ, Mayer-Pröschel M. Iron modulates the differentiation of a distinct population of glial precursor cells into oligodendrocytes. Dev Biol. 2001;237(1):232–43. 68. Morath DJ, Mayer-Pröschel M. Iron deficiency during embryogenesis and consequences for oligodendrocyte generation in vivo. Dev Neurosci. 2002;24(2– 3):197–207. 69. Greminger AR, Mayer-Pr??schel M. Identifying the threshold of iron deficiency in the central nervous system of the rat by the auditory brainstem response. ASN Neuro. 2015;7(1):1–10. 70. Lisman JE, Coyle JT, Green RW, Javitt DC, Benes FM, Heckers S, et al. Circuit- based framework for understanding neurotransmitter and risk gene interactions in schizophrenia. Trends Neurosci. 2008;31(5):234–42. 71. Beard J, Erikson KM, Jones BC. Neonatal Iron Deficiency Results in Irreversible Changes in Dopamine Function in Rats. J Nutr. 2003 Apr 1;133(4):1174–9. 72. Pisansky MT, Wickham RJ, Su J, Fretham S, Yuan L-L, Sun M, et al. Iron deficiency with or without anemia impairs prepulse inhibition of the startle reflex. Hippocampus. 2013 Oct;23(10):952–62. 73. Schmidt AT, Waldow KJ, Grove WM, Salinas JA, Georgieff MK. Dissociating the long-term effects of fetal/neonatal iron deficiency on three types of learning in the rat. Behav Neurosci. 2007;121(3):475–82. 74. Algarin C, Karunakaran KD, Reyes S, Morales C, Lozoff B, Peirano P, et al. Differences on Brain Connectivity in Adulthood Are Present in Subjects with Iron Deficiency Anemia in Infancy. Front Aging Neurosci. 2017 Mar 7;9:54. 75. Buckner RL, Andrews-Hanna JR, Schacter DL. The brain’s default network: Anatomy, function, and relevance to disease. Ann N Y Acad Sci. 2008 Mar 1;1124(1):1–38. 76. Carlson ES, Stead JDH, Neal CR, Petryk A, Georgieff MK. Perinatal iron deficiency results in altered developmental expression of genes mediating energy metabolism and neuronal morphogenesis in hippocampus. Hippocampus. 2007 Jan;17(8):679–91. 77. Schachtschneider KM, Liu Y, Rund LA, Madsen O, Johnson RW, Groenen MAM, et al. Impact of neonatal iron deficiency on hippocampal DNA methylation and gene transcription in a porcine biomedical model of cognitive development. BMC Genomics. 2016 Dec 3;17(1):856. 78. Lubin FD, Roth TL, Sweatt JD. Epigenetic regulation of BDNF gene transcription in the consolidation of fear memory. J Neurosci. 2008 Dec 1;28(42):10576–86. 79. McGowan PO, Suderman M, Sasaki A, Huang TCT, Hallett M, Meaney MJ, et al. Broad epigenetic signature of maternal care in the brain of adult rats. Sirigu A, 84 editor. PLoS One. 2011 Feb 28;6(2):e14739. 80. Anderson OS, Nahar MS, Faulk C, Jones TR, Liao C, Kannan K, et al. Epigenetic responses following maternal dietary exposure to physiologically relevant levels of bisphenol A. Environ Mol Mutagen. 2012 Jun;53(5):334–42. 81. McBirney M, King SE, Pappalardo M, Houser E, Unkefer M, Nilsson E, et al. Atrazine induced epigenetic transgenerational inheritance of disease, lean phenotype and sperm epimutation pathology biomarkers. Óvilo C, editor. PLoS One. 2017 Sep 20;12(9):e0184306. 82. Perkins A, Lehmann C, Lawrence RC, Kelly SJ. Alcohol exposure during development: Impact on the epigenome. Int J Dev Neurosci. 2013 Oct;31(6):391– 7. 83. Ke X, Xing B, Yu B, Yu X, Majnik A, Cohen S, et al. IUGR disrupts the PPARγ- Setd8-H4K20me1and Wnt signaling pathways in the juvenile rat hippocampus. Int J Dev Neurosci. 2014 Nov;38:59–67. 84. Tran P V, Kennedy BC, Lien Y-C, Simmons RA, Georgieff MK. Fetal iron deficiency induces chromatin remodeling at the Bdnf locus in adult rat hippocampus. Am J Physiol Regul Integr Comp Physiol. 2015 Feb 15;308(4):R276-82. 85. Waterland RA, Jirtle RL. Transposable elements: targets for early nutritional effects on epigenetic gene regulation. Mol Cell Biol. 2003 Aug 1;23(15):5293– 300. 86. Chen Z, Zang J, Whetstine J, Hong X, Davrazou F, Kutateladze TG, et al. Structural insights into histone demethylation by JMJD2 family members. Cell. 2006 May 19;125(4):691–702. 87. Tsukada Y, Fang J, Erdjument-Bromage H, Warren ME, Borchers CH, Tempst P, et al. Histone demethylation by a family of JmjC domain-containing proteins. Nature. 2006 Feb 18;439(7078):811–6. 88. Smith ZD, Meissner A. DNA methylation: Roles in mammalian development. Nat Rev Genet. 2013 Mar 12;14(3):204–20. 89. Lien Y-C, Condon DE, Georgieff MK, Simmons RA, Tran P V, Lien Y-C, et al. Dysregulation of Neuronal Genes by Fetal-Neonatal Iron Deficiency Anemia Is Associated with Altered DNA Methylation in the Rat Hippocampus. Nutrients. 2019 May 27;11(5):1191. 90. Tran P V, Carlson ES, Fretham SJB, Georgieff MK. Early-life iron deficiency anemia alters neurotrophic factor expression and hippocampal neuron differentiation in male rats. J Nutr. 2008 Dec;138(12):2495–501. 91. Tran P V, Fretham SJB, Carlson ES, Georgieff MK. Long-term reduction of hippocampal brain-derived neurotrophic factor activity after fetal-neonatal iron deficiency in adult rats. Pediatr Res. 2009 May;65(5):493–8. 92. Moore LD, Le T, Fan G. DNA methylation and its basic function. Vol. 38, Neuropsychopharmacology. 2013. p. 23–38. 93. Tahiliani M, Koh KP, Shen Y, Pastor WA, Bandukwala H, Brudno Y, et al. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science. 2009 May 15;324(5929):930–5. 94. Ito S, Shen L, Dai Q, Wu SC, Collins LB, Swenberg JA, et al. Tet proteins can 85 convert 5-methylcytosine to 5-formylcytosine and 5-carboxylcytosine. Science. 2011 Sep 2;333(6047):1300–3. 95. Wu X, Zhang Y. TET-mediated active DNA demethylation: Mechanism, function and beyond. Nat Rev Genet. 2017;18(9):517–34. 96. Bachman M, Uribe-Lewis S, Yang X, Williams M, Murrell A, Balasubramanian S. 5-Hydroxymethylcytosine is a predominantly stable DNA modification. Nat Chem. 2014 Sep 21;6(12):1049–55. 97. Wen L, Tang F. Genomic distribution and possible functions of DNA hydroxymethylation in the brain. Genomics. 2014;104(5):341–6. 98. Shi D-Q, Ali I, Tang J, Yang W-C. New Insights into 5hmC DNA Modification: Generation, Distribution and Function. Front Genet. 2017 Jul 19;8:100. 99. Kriaucionis S, Heintz N. The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain. Science (80- ). 2009;324(5929):929–30. 100. Lister R, Mukamel EA, Nery JR, Urich M, Puddifoot CA, Johnson ND, et al. Global epigenomic reconfiguration during mammalian brain development. Science. 2013 Aug 9;341(6146):1237905. 101. Münzel M, Globisch D, Brückl T, Wagner M, Welzmiller V, Michalakis S, et al. Quantification of the Sixth DNA Base Hydroxymethylcytosine in the Brain. Angew Chemie Int Ed. 2010 Jun 25;49(31):5375–7. 102. Szulwach KE, Li X, Li Y, Song CX, Wu H, Dai Q, et al. 5-hmC-mediated epigenetic dynamics during postnatal neurodevelopment and aging. Nat Neurosci. 2011 Oct 30;14(12):1607–16. 103. Hahn MA, Qiu R, Wu X, Li AX, Zhang H, Wang J, et al. Dynamics of 5- Hydroxymethylcytosine and Chromatin Marks in Mammalian Neurogenesis. Cell Rep. 2013;3(2):291–300. 104. Zhu X, Girardo D, Govek E-E, John K, Mellén M, Tamayo P, et al. Role of Tet1/3 Genes and Chromatin Remodeling Genes in Cerebellar Circuit Formation. Neuron. 2016;89(1):100–12. 105. Khare T, Pai S, Koncevicius K, Pal M, Kriukiene E, Liutkeviciute Z, et al. 5-hmC in the brain is abundant in synaptic genes and shows differences at the exon-intron boundary. Nat Struct Mol Biol. 2012 Sep 9;19(10):1037–43. 106. National Research Council (U.S.). Committee for the Update of the Guide for the Care and Use of Laboratory Animals., Institute for Laboratory Animal Research (U.S.). Guide for the care and use of laboratory animals. National Academies Press; 2011. 220 p. 107. Ching T, Huang S, Garmire LX. Power analysis and sample size estimation for RNA-Seq differential expression. RNA. 2014 Nov;20(11):1684–96. 108. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013 Apr 25;14(4):R36. 109. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, van Baren MJ, et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol. 2010 May 2;28(5):511–5. 110. Bailey RL, West KP, Black RE. The epidemiology of global micronutrient 86 deficiencies. Ann Nutr Metab. 2015;66(Suppl. 2):22–33. 111. McLean E, Cogswell M, Egli I, Wojdyla D, de Benoist B. Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993-2005. Public Health Nutr. 2009 Apr 1;12(4):444–54. 112. Stevens GA, Finucane MM, De-Regil LM, Paciorek CJ, Flaxman SR, Branca F, et al. Global, regional, and national trends in haemoglobin concentration and prevalence of total and severe anaemia in children and pregnant and non-pregnant women for 1995-2011: A systematic analysis of population-representative data. Lancet Glob Heal. 2013;1(1):e16–25. 113. Doom JR, Georgieff MK. Striking while the iron is hot: Understanding the biological and neurodevelopmental effects of iron deficiency to optimize intervention in early childhood. Curr Pediatr Rep. 2014;2(4):291–8. 114. Lozoff B, Clark KM, Jing Y, Armony-Sivan R, Angelilli ML, Jacobson SW. Dose- Response Relationships between Iron Deficiency with or without Anemia and Infant Social-Emotional Behavior. J Pediatr. 2008 May;152(5):696-702.e3. 115. Unger EL, Hurst AR, Georgieff MK, Schallert T, Rao R, Connor JR, et al. Behavior and Monoamine Deficits in Prenatal and Perinatal Iron Deficiency Are Not Corrected by Early Postnatal Moderate-Iron or High-Iron Diets in Rats. J Nutr. 2012 Nov 1;142(11):2040–9. 116. Schmidt AT, Alvarez GC, Grove WM, Rao R, Georgieff MK. Early iron deficiency enhances stimulus-response learning of adult rats in the context of competing spatial information. Dev Cogn Neurosci. 2012 Jan;2(1):174–80. 117. Felt B, Beard J, Schallert T, Shao J, Aldridge J, Connor J, et al. Persistent neurochemical and behavioral abnormalities in adulthood despite early iron supplementation for perinatal iron deficiency anemia in rats. Behav Brain Res. 2006 Aug 10;171(2):261–70. 118. Jorgenson LA, Sun M, O’Connor M, Georgieff MK. Fetal iron deficiency disrupts the maturation of synaptic function and efficacy in area CA1 of the developing rat hippocampus. Hippocampus. 2005 Jan;15(8):1094–102. 119. Carlson ES, Magid R, Petryk A, Georgieff MK. Iron deficiency alters expression of genes implicated in Alzheimer disease pathogenesis. Brain Res. 2008;1237:75– 83. 120. Lonze BE, Ginty DD. Function and regulation of CREB family transcription factors in the nervous system. Vol. 35, Neuron. 2002. p. 605–23. 121. Cunha C, Brambilla R, Thomas KL. A simple role for BDNF in learning and memory? Front Mol Neurosci. 2010 Feb 9;3:1. 122. Asobayire FS, Adou P, Davidsson L, Cook JD, Hurrell RF. Prevalence of iron deficiency with and without concurrent anemia in population groups with high prevalences of malaria and other infections: a study in Côte d’Ivoire. Am J Clin Nutr. 2001 Dec 1;74(6):776–82. 123. Engmann C, Adanu R, Lu TS, Bose C, Lozoff B. Anemia and iron deficiency in pregnant Ghanaian women from urban areas. Int J Gynecol Obstet. 2008;101(1):62–6. 124. Siddappa AM, Rao R, Long JD, Widness JA, Georgieff MK. The assessment of newborn iron stores at birth: A review of the literature and standards for ferritin 87 concentrations. Neonatology. 2007;92(2):73–82. 125. Wallace DF. The Regulation of Iron Absorption and Homeostasis. Clin Biochem Rev. 2016;37(2):51–62. 126. Rao R, Ennis K, Lubach GR, Lock EF, Georgieff MK, Coe CL. Metabolomic analysis of CSF indicates brain metabolic impairment precedes hematological indices of anemia in the iron-deficient infant monkey. Nutr Neurosci. 2016 Aug 6;1–9. 127. Fretham SJB, Carlson ES, Georgieff MK. The role of iron in learning and memory. Adv Nutr. 2011 Mar;2(2):112–21. 128. Barrett JR. Programming the Future: Epigenetics in the Context of DOHaD. Environ Health Perspect. 2017 Mar 31;125(4):A72. 129. Clardy SL, Wang X, Zhao W, Liu W, Chase GA, Beard JL, et al. Acute and chronic effects of developmental iron deficiency on mRNA expression patterns in the brain. In: Oxidative Stress and Neuroprotection. Vienna: Springer Vienna; 2006. p. 173–96. 130. Feil R. Environmental and nutritional effects on the epigenetic regulation of genes. Mutat Res Mol Mech Mutagen. 2006 Aug;600(1–2):46–57. 131. Dolinoy DC, Jirtle RL. Environmental epigenomics in human health and disease. Environ Mol Mutagen. 2008 Jan;49(1):4–8. 132. Salminen A, Kauppinen A, Kaarniranta K. 2-Oxoglutarate-dependent dioxygenases are sensors of energy metabolism, oxygen availability, and iron homeostasis: potential role in the regulation of aging process. Cell Mol Life Sci. 2015 Oct 29;72(20):3897–914. 133. Ponka P, Lok CN. The transferrin receptor: role in health and disease. Int J Biochem Cell Biol. 1999 Oct;31(10):1111–37. 134. Petry CD, Eaton MA, Wobken JD, Mills MM, Johnson DE, Georgieff MK. Iron deficiency of liver, heart, and brain in newborn infants of diabetic mothers. J Pediatr. 1992 Jul;121(1):109–14. 135. Ndong M, Kazami M, Suzuki T, Uehara M, Katsumata S ichi, Inoue H, et al. Iron deficiency down-regulates the Akt/TSC1-TSC2/mammalian Target of Rapamycin signaling pathway in rats and in COS-1 cells. Nutr Res. 2009;29(9):640–7. 136. Wallin DJ, Zamora TG, Alexander M, Ennis KM, Tran P V, Georgieff MK. Neonatal mouse hippocampus: Phlebotomy-induced anemia diminishes and treatment with erythropoietin partially rescues mammalian target of rapamycin signaling. Pediatr Res. 2017 Sep;82(3):501–8. 137. Workman AD, Charvet CJ, Clancy B, Darlington RB, Finlay BL. Modeling transformations of neurodevelopmental sequences across mammalian species. J Neurosci. 2013;33(17):7368–83. 138. Carletti B, Rossi F. Neurogenesis in the cerebellum. Vol. 14, Neuroscientist. Sage PublicationsSage CA: Los Angeles, CA; 2008. p. 91–100. 139. Hatten ME, Heintz N. Mechanisms of Neural Patterning and Specification in the Development Cerebellum. Annu Rev Neurosci. 1995 Mar;18(1):385–408. 140. Ábrahám H, Tornóczky T, Kosztolányi G, Seress L. Cell formation in the cortical layers of the developing human cerebellum. Int J Dev Neurosci. 2001 Feb 20;19(1):53–62. 88 141. Koning I V., Dudink J, Groenenberg IAL, Willemsen SP, Reiss IKM, Steegers- Theunissen RPM. Prenatal cerebellar growth trajectories and the impact of periconceptional maternal and fetal factors. Hum Reprod. 2017 Jun 1;32(6):1230– 7. 142. Koning I V., Groenenberg IAL, Gotink AW, Willemsen SP, Gijtenbeek M, Dudink J, et al. Periconception maternal folate status and human embryonic cerebellum growth trajectories: The Rotterdam predict study. Pawluski J, editor. PLoS One. 2015 Oct 22;10(10):e0141089. 143. De Zeeuw P, Zwart F, Schrama R, Van Engeland H, Durston S. Prenatal exposure to cigarette smoke or alcohol and cerebellum volume in attention- deficit/hyperactivity disorder and typical development. Transl Psychiatry. 2012 Mar 6;2(3):e84–e84. 144. Yawno T, Sutherland AE, Pham Y, Castillo-Melendez M, Jenkin G, Miller SL. Fetal Growth Restriction Alters Cerebellar Development in Fetal and Neonatal Sheep. Front Physiol. 2019 May 22;10:560. 145. Santos DCC, Angulo-Barroso RM, Li M, Bian Y, Sturza J, Richards B, et al. Timing, duration, and severity of iron deficiency in early development and motor outcomes at 9 months. Eur J Clin Nutr. 2018 Mar 6;72(3):332–41. 146. Buckner RL. The cerebellum and cognitive function: 25 years of insight from anatomy and neuroimaging. Neuron. 2013 Oct 30;80(3):807–15. 147. Sathyanesan A, Zhou J, Scafidi J, Heck DH, Sillitoe R V., Gallo V. Emerging connections between cerebellar development, behaviour and complex brain disorders. Nat Rev Neurosci. 2019 May 28;20(5):298–313. 148. Foster AW, Osman D, Robinson NJ. Metal preferences and metallation. J Biol Chem. 2014 Oct 10;289(41):28095–103. 149. Barwinska-Sendra A, Waldron KJ. The Role of Intermetal Competition and Mis- Metalation in Metal Toxicity. In: Advances in Microbial Physiology. 2017. p. 315–79. 150. Wagner M, Steinbacher J, Kraus TFJ, Michalakis S, Hackner B, Pfaffeneder T, et al. Age-Dependent Levels of 5-Methyl-, 5-Hydroxymethyl-, and 5-Formylcytosine in Human and Mouse Brain Tissues. Angew Chemie - Int Ed. 2015 Oct 12;54(42):12511–4. 151. Khalaf-Nazzal R, Francis F. Hippocampal development - Old and new findings. Neuroscience. 2013;248:225–42. 152. Rahimi-Balaei M, Bergen H, Kong J, Marzban H. Neuronal Migration During Development of the Cerebellum. Front Cell Neurosci. 2018 Dec 17;12:484. 153. Heindel JJ, Vandenberg LN. Developmental origins of health and disease: A paradigm for understanding disease cause and prevention. Curr Opin Pediatr. 2015;27(2):248–53. 154. Kennedy BC, Tran P V, Kohli M, Maertens JJ, Gewirtz JC, Georgieff MK. Beneficial Effects of Postnatal Choline Supplementation on Long-Term Neurocognitive Deficit Resulting from Fetal-Neonatal Iron Deficiency. Behavioural Brain Research. 2017. 155. Xiong X, Mayes D, Demianczuk N, Olson DM, Davidge ST, Newburn-Cook C, et al. Impact of pregnancy-induced hypertension on fetal growth. Am J Obstet 89 Gynecol. 1999 Jan 1;180(1 I):207–13. 156. Chockalingam UM, Murphy E, Ophoven JC, Weisdorf SA, Georgieff MK. Cord transferrin and ferritin values in newborn infants at risk for prenatal uteroplacental insufficiency and chronic hypoxia. J Pediatr. 1987;111(2):283–6. 157. Bosley AR, Sibert JR, Newcombe RG. Effects of maternal smoking on fetal growth and nutrition. Arch Dis Child. 1981 Sep;56(9):727–9. 158. Sullivan EL, Nousen EK, Chamlou KA. Maternal high fat diet consumption during the perinatal period programs offspring behavior. Physiol Behav. 2014;123:236– 42. 159. Shen Y, Dong H, Lu X, Lian N, Xun G, Shi L, et al. Associations among maternal pre-pregnancy body mass index, gestational weight gain and risk of autism in the Han Chinese population. BMC Psychiatry. 2018 Dec 17;18(1):11. 160. Monk C, Georgieff MK, Osterholm EA. Maternal prenatal distress and poor nutrition - Mutually influencing risk factors affecting infant neurocognitive development. J Child Psychol Psychiatry. 2013;54(2):115–30. 161. Shorter KR, Felder MR, Vrana PB. Consequences of dietary methyl donor supplements: Is more always better? Prog Biophys Mol Biol. 2015;118(1–2):14– 20.