The emergence of developmental behavioral epigenomics

Thousands of epigenetic studies have been conducted in the past 50 years. However, the integration of epigenetic mechanisms to the study of human behavior, or ‘human behavioral epigenetics’, has gained traction in the last two decades and provides an unprecedented opportunity to identify molecular processes underlying child behavior and development. Not surprisingly, much of this work has focused on the early years of life, as this is a time of rapid brain growth and dynamic epigenetic changes. Children’s brains grow and change more from conception to age 2 years than at any other time; this ‘First Thousand Days’ framework has been endorsed by the American Academy of Pediatrics and highlights the importance of optimizing early experience to support children’s long-term health and development [1].

Epigenetics has provided us with new perspectives as we learn more about the extent to which early experiences and exposures can be biologically embedded and have long-term consequences for behavior and development. Biological embedding of risk factors for poor developmental outcomes occurs during sensitive periods when the brain is more vulnerable to environmental signals, such as during the first 2 years of life. Adversity is transmitted into long-term developmental risk through responses in biological pathways, including epigenetic responses beginning in the prenatal period. Early-life stress can become toxic stress when it is chronic, strong, cumulative and frequent and results in prolonged activation of the hypothalamic–pituitary–adrenal axis stress response system, causing wear and tear on adaptive regulatory pathways and increasing risk for poor outcome [2].

Epigenetic differences, both in stress response genes and throughout the genome, have been related to prenatal factors including diet, psychosocial conditions, substance use, environmental contaminants, adversity, stress and physical health [3]. Similarly, epigenetic alterations have also been related to postnatal environmental, psychosocial and physical stressors [3]. These signals from the environment trigger molecular changes in numerous tissues and cells throughout the body, presumably including brain cells, potentially leading to or correlating with altered behavior. Epigenetic alterations participate in the malleability of the developing central nervous system, referred to as neural plasticity, enabling changes in neuronal organization and programming in response to environmental stimuli. Additionally, epigenetic features are more dynamic and responsive to environmental stimuli during the early developmental periods and thus provide a potential mechanism through which early life events can have long-term impacts, particularly in the context of neurobehavioral development.

This also makes preterm birth (gestational age <36 weeks) a particularly useful model for the study of the infant at risk and for examining epigenetic differences that are related to development, as these infants experience a critical developmental life stage ex utero, while their term-born counterparts experience that same life stage in utero. This contrast is even more pronounced for infants born very preterm (<30 weeks), who experience much, or even all, of what would be the third trimester of gestation in an external environment. These infants have to cope not only with the effects of the extrauterine environment on an immature nervous system, but also with medical conditions, including brain injury, that are more prevalent in this population. Preterm birth is also of substantial public health importance as it affects one in ten births in the USA, causes suffering and hardship for the infants and their families and is associated with annual costs of over $26 billion in the USA [4].

While survival rates among infants born prematurely have increased with advancements in perinatal care, less progress has been made in reducing the rates of developmental impairments in this population. Approximately one-third of these infants will have a clinically significant developmental impairment by 2 years of age [5]. Importantly, even though these infants are at increased risk of impairment, many do not experience developmental deficits as they age. Recent studies have suggested that the epigenome can provide insights into the impact of preterm birth, and the unique postnatal environment that these infants experience, on long-term health.

Preterm birth itself has been associated with DNA methylation differences in the placenta, cord tissues, cord blood, peripheral blood, buccal epithelium and saliva, demonstrating that the changes to the epigenetic landscape are systemic [6]. While some candidate gene studies have suggested that the changes in DNA methylation tend to resolve in early childhood or adolescence [7,8], others have identified preterm birth-associated changes in young adults that were replicated in an independent sample of older adults [9], suggesting that some epigenetic changes are persistent across the life course. Clarifying the impact of preterm birth on the epigenome, whether these epigenetic differences are persistent or dynamic with aging and development, and how they correlate with the adverse health and developmental sequelae of preterm birth, is of immense interest.

Extensive research demonstrates that prematurity and early postnatal stress are associated with epigenetic changes in genes in the hypothalamic–pituitary–adrenal axis (primarily, but not limited to NR3C1) and in the serotonin transporter gene (SLC6A4) [10]. For very preterm infants, the early postnatal period is spent in the neonatal intensive care unit (NICU), where they experience a unique cadre of medical, environmental and psychosocial stressors [11], and thus epigenetic changes in these genes may represent important programming events related to stress response systems. Such candidate gene approaches have allowed for elegant investigations of alterations to biologically hypothesized pathways. However, complex traits such as behavior have multifactorial and polygenic etiologies that are beyond the scope of candidate gene methods. Our group and others have performed epigenome-wide association studies to demonstrate that the perinatal methylome captures information about prenatal and perinatal stress, as well as the health and behavior of infants that were born very prematurely, often occurring in genes and pathways that would not have previously been considered [12–14]. While numerous studies have examined differences in epigenetic patterns between preterm versus term-born infants, children or adults, there is a sparsity of published literature from prospective longitudinal studies of infants who were born prematurely. These studies are critical to deepening our understanding of the impacts of perinatal and postnatal factors that influence epigenetic and developmental trajectories, and whether these foster resiliency or vulnerability to developmental impairments.

Not only do epigenetic marks capture information about prenatal and postnatal programming, but epigenetic processes have also been shown to predict child outcomes across many functional domains. We previously reviewed the literature linking placental epigenetics to neurodevelopmental outcomes in healthy infants, finding that DNA methylation differences across the placental epigenome are associated with neonatal neurobehavior, cortisol reactivity and cry acoustics [15]. These studies suggest that placental epigenetics are one likely mechanism explaining the consistent and enduring relationships observed between early-life (i.e., prenatal) stress and child neurodevelopmental trajectories. Studies of neonatal peripheral tissues, including cord blood and buccal swabs, have also been performed in relation to diverse neonatal and childhood neurobehavioral outcomes. Variations in DNA methylation across many biologically relevant gene loci have been linked to child performance on standardized cognitive assessments, including measures of IQ and executive functioning [16], as well as neurological disorders [17] and symptoms and trajectories of psychopathology (e.g., attention-deficit/hyperactivity disorder, conduct problems) [18]. Links between epigenetics and more subtle variations in infant behavioral attributes, known collectively as temperament, are also beginning to be appreciated [19].

Understanding the relationship between epigenetics and early-life neurodevelopmental and behavioral outcomes is a critical venture. Not only can neurobehavioral variations be reliably measured as early as the neonatal period, as has been demonstrated with the NICU Network Neurobehavioral Scale [20], but these characteristics are related to epigenetic markers and can also serve as early warning signs for later developmental impairment or clinically relevant psychological disorders, many of which cannot be formally diagnosed until early childhood. The ability to identify epigenetic markers of infant neurobehavioral variations could not only help uncover the biological pathways underlying typical and atypical development, but could also be used as a tool for predicting which children are at highest risk for long-term problems and could therefore benefit most from early intervention. These endeavors could particularly benefit children born prematurely who are at highest risk for later developmental impairment, including elevated risk for psychopathology. The ability to identify, by NICU discharge, which individual infants are at greatest risk for adverse developmental outcome could lead to intervention programs to mitigate, if not prevent, these outcomes.

In light of all the progress made in the area of behavioral epigenetics, several areas remain ripe for future inquiry. Most research to date has investigated associations between DNA methylation at individual gene loci and environmental exposures or outcomes. While these studies have helped pinpoint individual genes and gene networks that may underlie developmental programming or behavioral development, it is clear that multiple genes work together to jointly influence the workings of biological and behavioral systems. Highly powered epigenome-wide association studies with large sample sizes, such as are afforded by the NIH nationwide Environmental Exposures on Child Health Outcomes program [21] or through multi-cohort consortia, and polyepigenetic approaches, are therefore needed. These efforts have begun in arenas such as the development and utilization of epigenetic clocks, predictors that make use of DNA methylation across multiple genes to reliably estimate an individual’s chronological or biological age [22], including in the special case of preterm birth [23]. In addition to examining multiple genes simultaneously, ongoing growth and advancement in omics technologies mean that future studies should incorporate multiple levels of analysis (e.g., transcriptomics, proteomics, metabolomics) to understand the functional relevance of alterations in children’s epigenetic profiles. Finally, rather than single-time-point or short-term longitudinal data collection, parallel data collection of epigenetic, exposure and phenotypic data via long-term longitudinal studies is needed to better understand the forces that influence biological and behavioral trajectories. The emergence of a developmental behavioral epigenetics paradigm will help us understand how and when the associations between environment, epigenetics and behavior unfold, important mediators and modifiers of these associations, and whether there are sensitive periods or inflection points during which interventions may have optimal benefit.