Overview

Experience-sensitive epigenetic mechanisms, developmental plasticity, and the biological embedding of chronic disease risk Vincent T. Cunliffe∗ A wide range of developmental, nutritional, environmental, and social factors affect the biological activities of epigenetic mechanisms. These factors change spatiotemporal patterns of gene expression in a variety of different ways and bring significant impacts to bear on development, physiology, and disease risk throughout the life course. Abundant evidence demonstrates that behavioral stressors and adverse nutritional conditions are particularly potent inducers of epigenetic changes and enhancers of chronic disease risks. Recent insights from both human clinical studies and research with model organisms further indicate that such experience-dependent changes to the epigenome can be transmitted through the germline across multiple generations, with important consequences for the heritability of both adaptive and maladaptive phenotypes. Epigenetics research thus offers many possibilities for developing informative biomarkers of acquired chronic disease risk and determining the effectiveness of preventive and therapeutic interventions. Moreover, the experience-sensitive nature of these disease risks raises important questions about societal and individual responsibilities for the prevention of ill-health and the promotion of well-being during development, across the life course and between generations. Better understanding of how epigenetic mechanisms regulate developmental plasticity and mediate the biological embedding of chronic disease risks is therefore likely to shed important new light on the nature of the pathophysiological mechanisms linking social and health inequalities, and will help to inform public policy initiatives in this area. © 2015 Wiley Periodicals, Inc.

How to cite this article:

WIREs Syst Biol Med 2015, 7:53–71. doi: 10.1002/wsbm.1291

INTRODUCTION

A

growing body of evidence indicates that experiences acquired during development, childhood, or adulthood induce changes in gene expression, which impart cumulative, long-lasting effects on health, well-being, and vulnerability to disease. These experiences involve exposure to social, behavioral, ∗ Correspondence

to: [email protected]

Bateson Centre, Department of Biomedical Science, University of Sheffield, Sheffield, UK Conflict of interest: The author has declared no conflicts of interest for this article.

Volume 7, March/April 2015

environmental, and nutritional signals, and are communicated within the body by sensory, endocrine, and developmental regulatory mechanisms. Some of the resulting gene expression changes can be transmitted across multiple generations of offspring, and they are accompanied by heritable changes in biology, behavior, and disease risk. Experience-sensitive changes to gene expression are generated by the epigenetic machinery, which introduces covalent modifications to chromatin structure that affect gene transcription, without changing the coding properties of DNA sequences. Epigenetic changes also involve noncovalent structural remodeling of chromatin and the altered expression of noncoding RNAs that modulate

© 2015 Wiley Periodicals, Inc.

53

wires.wiley.com/sysbio

Overview

Histone modifications

DNA-binding transcription factors

DNA methylation

Gene expression

Histone variants

Chromatin remodeling

Noncoding RNA

FIGURE 1 | Diverse molecular mechanisms mediate epigenetic regulation of gene expression at transcriptional and post-transcriptional levels. A wide variety of covalent histone modifications and DNA methylation marks signify distinct transcriptional states of associated genes, and chromatin remodeling factors regulate accessibility of histone-modifying enzymes and transcription factors to their targets. Long noncoding RNAs interact with chromatin components and short noncoding RNAs modulate stability of cognate mRNAs.

the stability and translatability of specific messenger RNAs (Figure 1). The main aims of this review are to assess progress in understanding the functional characteristics of experience-regulated epigenetic mechanisms and their pathogenetic roles in chronic diseases. The wide range of extrinsic factors that modulate chronic disease risk via their effects on the epigenetic machinery will be discussed, with particular emphasis on the impacts of behavioral stressors and poor nutrition. A number of recent studies, showing that experience-sensitive phenotypic and epigenetic changes exhibit transgenerational inheritance, will also be reviewed. The broader implications of this new knowledge, for medicine and society, will also be considered. Epigenetics research provides a new perspective from which to understand the impact of context-sensitive developmental mechanisms on health and well-being, and will provide significant impetus for policy initiatives to improve public health. Moreover, epigenetic biomarkers capable of detecting and predicting disease risks could also be very useful in determining the effectiveness of medical interventions, and would help considerably to reduce the global burden of chronic noncommunicable diseases.

EPIGENETIC REGULATION OF GENE EXPRESSION Through their interactions with genomic DNA sequences, transcription factors implement the logical rules of multicellular development, establishing the body plan, generating distinct cell lineages, facilitating cell interactions, and enabling the emergence of complex physiological functions. The accuracy and reliability of the developmental decisions determined 54

by these rules is influenced by an ‘epigenetic’ code of patterned covalent modifications, which is introduced into both the DNA and the histone components of chromatin. The epigenetic code is a second layer of information that is created and updated during development, providing cells with a physical trace of their developmental history, as well as a coded representation of their differentiation status and a prescription of their future developmental potential. Moreover, the cell machinery responsible for generating epigenetic modifications mediates the impacts of intercellular, endocrine, and environmental signals on the mechanisms that regulate gene transcription. Thus, while it has long been recognized that epigenetic mechanisms impart robustness and stability to developmental decisions, it is becoming increasingly evident that the epigenetic machinery also generates functional plasticity in response to changing circumstances, which may be either phenotypically adaptive or maladaptive. Within chromatin, DNA is modified mostly by methylation of the cytosines within CpG dinucleotide pairs that are enriched at gene promoters. DNA methyltransferases produce 5-methylcytosine, which is recognized by methyl-CpG-binding proteins MeCP2 and MBDs1–4,1 triggering a cascade of protein interactions that lead to stable transcriptional silencing of the methylated locus. Chemical derivatives of 5mC such as 5-formylcytosine, 5-carboxylcytosine, and 5-hydroxymethylcytosine have recently been detected, which may be demethylation intermediates involved in transcriptional reactivation.2 Enzymes involved in this demethylation are specifically recruited to regions of methylated DNA by the adaptor protein Gadd45a.2 Gadd45a interacts both with the TET enzymes, which produce 5-formylcytosine and 5-carboxylcytosine, and AID/apobec, which converts 5-hydroxymethylcytosine into thymine, producing a base mismatch that is then repaired by DNA glycosylases.3,4 There is therefore a remarkable close functional interplay between DNA methylation, demethylation, and repair mechanisms, potentially linking aberrations of transcription silencing and DNA repair processes to increased risks of disease-causing mutations in somatic tissues. The core histones of chromatin are major substrates for enzymes that catalyze the addition or removal of acetyl and methyl groups to basic residues within their N-terminal domains, and other enzymes that regulate phosphorylation, glycosylation, sumoylation, monoubiquitination, and poly-ADP-ribosylation of these proteins.5 Much progress has been made in elucidating the transcriptional and developmental roles of histone methylation and acetylation, but the biological

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

functions of the other modifications remain much less well understood. Many of the proteins responsible for generating and recognizing histone methylation and acetylation marks are encoded by members of the Polycomb (PcG) and Trithorax (TrxG) groups of genes, which encode repressors and activators of target gene transcription, respectively.6–8 Some of the better understood PcG proteins encode adaptors, co-factors, and catalytic subunits of protein assemblies that generate, recognize, or remove the histone lysine methylation marks that are associated with transcriptional silencing. Other, less well-characterized PcG proteins mediate histone polyubiquitination and O-linked glycosylation. By contrast, some TrxG proteins are involved in lysine methylation of histones that is associated with transcriptional activation of target genes, whereas others bind specifically to acetylated histones to promote gene expression. Different histone methyltransferases modify distinct lysine or arginine residues in distinct core histones, and they can catalyze either mono-, di-, or tri-methylation of these substrates, which makes the repertoire of histone modifications remarkably complex. Histone methyltransferase and acetyltransferase activities are counteracted by histone demethylases and deacetylases, two additional large groups of enzymes that also exhibit variable degrees of amino acid residue specificity. Moreover, these four broad classes of enzyme are frequently found in physical association with DNA methyltransferases and the functionally antagonistic components of the DNA repair machinery, which remove methylcytosine from CpG dinucleotides. Thus, the chromatin-modifying machinery introduces a complex set of functionally interconnected structural modifications into chromatin domains that specify the circumstances under which gene transcription may occur. Regulation of noncovalent interactions between core histones, DNA-binding transcription factors, and DNA also has major impacts on gene expression, because DNA that is packaged into highly condensed chromatin is difficult to transcribe. By contrast, DNA sequences in decondensed chromatin are more readily accessed by transcription factors and RNA polymerases. In their roles as the basic building blocks of chromatin around which DNA can be wound, core histones regulate the accessibility of genes to the transcription machinery, and their interactions with DNA are regulated by ATP-dependent chromatin remodeling factors and histone chaperones. Together with covalent chromatin modifications and atypical variant core histones such as H2A.Z and H3.3, chromatin remodeling factors and histone chaperones can radically transform chromatin domains between transcriptionally inactive Volume 7, March/April 2015

and active states.9 As might be expected for proteins with such important roles in regulating gene expression, the enzymes that regulate histone acetylation and methylation, together with many of their auxiliary regulatory and specificity-determining cofactors, are extensively implicated in tumorigenesis.10 One surprising discovery is that RNA transcripts perform direct physical roles in epigenetic regulation. For example, X-chromosome inactivation involves the decoration of the inactivated X chromosome with long noncoding RNA (lncRNA) transcripts from the X-chromosome-derived Xist locus. Repeat sequence elements within the Xist RNA stably interact with PcG proteins such as EZH2 and SUZ12 and recruit Polycomb PRC2 and PRC1 complexes to the X-chromosome, leading to the establishment of widespread methylation of core histones and DNA that underpins stable X-inactivation.11 Other examples of regulatory mechanisms involving lncRNAs include PcG-mediated repression in the Beckwith–Wiedemann syndrome-imprinted region of human chromosome 1112 and PcG-mediated repression at the mouse Bdnf locus.13 Short noncoding RNAs (sncRNAs) have also been identified as epigenetic regulators of gene expression, but unlike lncRNAs, they function by targeting transcripts with which they exhibit partial complementarity for degradation by specialized protein complexes.14 One such class of sncRNAs are piRNAs, which form complexes with Piwi proteins to target the transcripts of transposable elements for degradation. Another class of sncRNAs are microRNAs (miRNAs), which target transcripts of protein-coding genes, particularly transcription factors, for degradation, making the expression of such genes dependent on continued activity of the regulatory signals eliciting their transcription.14 The functional complexity, diversity, and reversibility of epigenetic machinery create remarkable sophisticated and dynamic networks of flexibly linked information-processing devices that regulate gene expression (Figure 1). This machinery integrates developmental signals with information that is acquired from endocrine, sensory, and cognitive cues, to allow cells, tissues, and whole organisms to detect and respond to extrinsic stimuli throughout the life course. The broad range of reversible post-translational modifications that influence gene transcription suggests that a correspondingly wide variety of developmental, endocrine, and neural activity-dependent signals regulate the types and distributions of epigenetic modifications that exist in chromatin (Figure 2). An important challenge therefore has been to understand how cells are able to signal to chromatin in ways that alter gene transcription.

© 2015 Wiley Periodicals, Inc.

55

wires.wiley.com/sysbio

Overview

Developmental signals

Nutritional signals

Environmental signals

Social signals

Epigenetic machinery

FIGURE 2 | The epigenetic machinery is a molecular interface for interpreting the developmental, nutritional, environmental, and social signals underlying adaptive and maladaptive phenotypic plasticity.

Adaptive responses

Recent studies have begun to clarify the key features of these mechanisms, offering new insights into how extrinsic experiences impact on physiology, behavior, and risks to health and well-being.

SIGNALING TO EPIGENETIC MACHINERY Intercellular signaling mechanisms make up the communication networks that drive multicellular development and integrate body functions across the life course, producing different cell types, coordinating cell movements, forming tissues, and regulating physiological activity within and between organs. A wide variety of intercellular signals, including neurotransmitters, hormones, secreted polypeptides, and transmembrane-spanning proteins, selectively control the transcriptional activation and repression of specific genes by regulating the recruitment of activated DNA-binding transcription factors to DNA, altering the patterns of covalent histone modifications in chromatin, or replacing typical core histones with atypical variant histones. Intercellular signaling proteins such as epidermal growth factor (EGF), fibroblast growth factor (FGF), and brain-derived neurotrophic factor (BDNF) signal to the nucleus through receptor- and membrane-associated, kinase-linked protein cascades, which couple events at the plasma membrane to gene transcription (Figure 3). Signaling by proteins such as FGFs and BDNF converges on the intracellular kinase ERK, which phosphorylates nuclear kinases such as MSK1 and RSK1, as well as DNA-binding transcription factors such as CREB, ATF1, and ELK1, enabling them to bind to their cognate binding sites in the regulatory elements of direct target genes, known 56

Robustness to change

Maladaptive responses

as ‘immediate-early genes’, such as cfos and egr1.15 The binding of these transcription factors to target DNA sites is accompanied by recruitment of the SWI/SNF Brg1/Brm chromatin remodeling machinery, replacement of core histones with histone variants, along with S10 phosphorylation, K14 acetylation, and K4 methylation of core histone H3. Histone variants such as H3.3 are enriched at the promoters of immediate-early genes such as cfos, when they are transcriptionally active or poised for transcriptional activation, suggesting that their incorporation into chromatin is also part of the process that initiates and maintains inducible gene transcription.16 Further recent studies indicate that variant histone exchange involves ADP-ribosylation by the poly(ADP-ribose) polymerase-1 (PARP-1),15 which is also implicated in DNA methylation, chromatin architecture, and DNA repair.17 Steroid hormones regulate gene transcription in many different cell types, including neurones, via interactions with nuclear hormone receptors, which then bind to target sites in regulatory DNA elements to elicit changes in chromatin structure and promoter activation. Receptor binding to cognate regulatory elements can either activate or repress gene transcription, depending on the physiological context and functional properties of the target gene. Detailed insights into the relationship between chromatin structure, epigenetic modifications, and inducible gene transcription have been obtained from studies of steroid hormone receptors,18–20 one of the best understood examples of which is the glucocorticoid receptor (GR), whose in vivo activity is regulated by the stress hormones cortisol and corticosterone. Glucocorticoid-induced binding of GR to its regulatory DNA target sites in transcriptionally inducible genes is closely accompanied by recruitment of the chromatin

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

EGF, FGF, BDNF

P

P

ERK P ELK1

ERK

P

MSK1 RSK1

ELK1 MSK1 RSK1

P

+1 c-fos

P SWI/SNF Brg1

ELK1

P300/CBP

c-fos H3.3

Ac

P

Ac

Ac

Ac

P

Ac

FIGURE 3 | Developmentally regulated intercellular signaling to chromatin. Induction of immediate early gene transcription by intercellular

signaling proteins, such as epidermal growth factor, fibroblast growth factor, and brain-derived neurotrophic factor.15 Receptor activation by ligand leads to ERK phosphorylation. Phospho-ERK then phosphorylates nuclear kinases MSK1/RSK1 and transcription factor ELK1. MSK1/RSK activation leads to histone H3 S10 phosphorylation and recruitment of chromatin remodeler Brg1. Variant histone H3.3 is enriched at loci poised for transcription. ELK1 phosphorylation allosterically activates p300, leading to increased local histone acetylation and onset of transcription.

remodeling SWI/SNF complex together with increased local acetylation and activation-associated methylation of core histones. Variant histones, such as H3.3 and H2A.Z, are also incorporated into activated GR-associated nucleosomes. In addition to its role as a transcription activator, GR is also involved in ligand-induced transcriptional repression of specific target genes (Figure 4), where it interacts with the SWI/SNF factor Brg1 and the histone deacetylase HDAC2, and attenuates both histone acetylation and RNA polymerase II function.21 In neurones, binding of the excitatory neurotransmitter glutamate to its receptors at synapses elicits cation influx across postsynaptic membranes, causing phosphorylation of the CREB transcription factor and its recruitment, along with the histone acetyltransferase CBP and the MSK1/RSK1 kinases, to promoter regions of immediate-early genes (Figure 5). The binding of these proteins to promoter sequences is accompanied by S10 phosphorylation, K14 acetylation, and K4 methylation of histone H3 molecules in the vicinity of the regulatory elements of these genes along with the binding of the chromatin remodeling protein SWI/SNF factor Brg1. These events lead to transcriptional activation of target genes, in a manner Volume 7, March/April 2015

that is highly reminiscent of how FGFs and BDNF promote transcription of their targets.22 Taken together, the above examples show how distinct extrinsic signal types deploy molecular mechanisms with common characteristics to regulate transcription of their specific target genes. In each case, transformation of an extracellular signal into a transcriptional decision is accompanied by a range of biochemical changes that involve histone-modifying enzymes, DNA methyltransferases, chromatin remodeling factors, as well as ligand-binding or phosphorylatable DNA-binding transcription factors.

EPIGENETIC REGULATION OF BIOLOGICAL RESPONSES TO STRESS The brain directs behavior in response to a combination of extrinsic sensory stimuli, physiological signals received from within the body, and cognitive activity in the brain itself. The resulting behaviors help to maintain homeostasis and produce adaptive responses to changing environmental and social conditions. An important mediator of environmental and social stress in humans is the steroid hormone cortisol, which is

© 2015 Wiley Periodicals, Inc.

57

wires.wiley.com/sysbio

Overview

GR Cortisol GR HDAC2

NGF1B

Ac

P

SWI/SNF Brg1

Ac

Ac

HDAC2

pomc

Ac

NGF1B

GR

Pitx1

P

Ac

P

SWI/SNF Pitx1 Brg1

Ac

+1 pomc

P

FIGURE 4 | Endocrine signaling to chromatin via the glucocorticoid receptor (GR). Repression of pomc transcription in hypothalamic-pituitary cells by stress hormone-activated GR. Cortisol-bound GR binds to target sites in the pomc promoter accompanied by the histone deacetylase HDAC2. Reduced local histone acetylation is accompanied by attenuation of pomc transcription while liganded GR and HDAC2 remain bound to the promoter.21

Glutamate

Glutamate receptor Ca2+ influx ERK P

ELK1

ERK

CaMKII PKA

P

MSK1 RSK1

CREB

ELK1 MSK1 RSK1

P

P P

P

CREB

+1 c-fos

SWI/SNF Brg1

CREB

ELK1

P300/CBP

c-fos

H3.3 Ac

P

Ac

Ac

Ac

P

Ac

FIGURE 5 | Neural activity-dependent signaling to chromatin in neurones. Induction of immediate early gene transcription in neurones by excitatory neurotransmission at synapses. Activation of glutamate receptor by glutamate causes an increase in intracellular calcium, activating CaMKII, protein kinase A, and ERK, which then phosphorylate and activate DNA-binding transcription factors such as CREB/ELK1 and nuclear kinases such as MSK1/RSK1. MSK1/RSK activation leads to histone H3 S10 phosphorylation and recruitment of chromatin remodeler Brg1. Variant histone H3.3 is enriched at loci poised for transcription. CREB and ELK1 phosphorylation allosterically activates CBP and p300, leading to increased local histone acetylation and onset of transcription.

58

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

(a)

Experience-sensitive epigenetic mechanisms

Stressful stimuli

(b)

Chronic stress

Hypothalamus

Arginine Corticotropinvasopressin releasing hormone (AVP) (CRH)

High levels of circulating cortisol for prolonged periods

Prefrontal cortex

Pituitary gland Pomc

Amygdala

Hippocampus

Adrenocorticotrophic hormone (ACTH)

Blunted HPA axis activity, fearfulness, aggression, anxiety, PTSD, memory impairment

Adrenal gland

Glucocorticoids (cortisol, corticosterone)

FIGURE 6 | (a) The hypothalamo-pituitary-adrenal (HPA) axis. Cortisol production stimulates a glucocorticoid receptor (GR)-mediated negative feedback loop that limits further cortisol production. (b) Chronic persistent activation of the HPA axis causes fearfulness, aggression, and anxiety, which over prolonged periods can cause post-traumatic stress disorder, memory loss, and eventually the negative feedback elicited by elevated cortisol leads to stably attenuated responsiveness of the HPA axis to stressful stimuli.

produced by the adrenal glands and secreted into the circulation to provide integrated, body-wide control of behavior, metabolism, and immune responses.23 However, while short periods of exposure to adrenal hormones can trigger adaptive responses that restore homeostatic balance, prolonged production of stress hormones such as cortisol can overload the body, leading to psychiatric disorders such as anxiety and depression, and also predispose to cardiometabolic diseases.24,25 Production of cortisol is regulated by neurones in the hypothalamus, which process signals from other regions of the brain that receive sensory and metabolic input (Figure 6(a)). The hypothalamus secretes corticotrophin-releasing hormone (CRH) and arginine vasopressin (AVP), which act on the adjacent pituitary gland, triggering transcription of the pro-opiomelanocortin (POMC) gene, one of the products of which, adrenocorticotrophic hormone (ACTH), is secreted into the bloodstream.26 On reaching the adrenal glands, ACTH induces synthesis and release of cortisol, which acts systemically, mostly through two cortisol-activated transcription factors, GR and mineralocorticoid receptor (MR). In the brain, the best understood impacts of cortisol are found in the hippocampus, amygdala, prefrontal cortex, and hypothalamus, where it modulates neurotransmitter release, expression of neurotransmitter receptors and ion channels, dendritic arborization, and remodeling of synapses. In the hippocampus, while brief exposure to cortisol Volume 7, March/April 2015

can be memory-enhancing, chronic exposure actually impairs memory.23 In the amygdala, chronic exposure to cortisol enhances fear, aggression, and anxiety, which are frequently influenced by additional changes in the prefrontal cortex (Figure 6(b)). Cortisol promotes homeostatic feedback repression of CRH and POMC transcription in the hypothalamus and pituitary gland, which is mediated directly by cortisol-bound GR. This feedback thus makes ACTH secretion and cortisol synthesis strictly dependent on continued production of CRH and AVP in response to persistence of the extrinsic stressor (Figure 6(a)). Acutely stressful experiences can leave behind strong memories that intensify responses when similar stimuli are encountered again. The molecular basis of this form of memory is poorly understood, but epigenetic mechanisms play important roles. Classic studies in rodent animal models and humans have demonstrated a close relationship between elevated stress in early life and the appearance of behavioral disorders in later life.27 The pups of rats that provide high levels of maternal care for their offspring developed into adults with low levels of stress reactivity, anxiety, and fearfulness. In contrast, rat pups with experience of lower levels of maternal care developed into adults with elevated levels of stress reactivity, anxiety, and fearfulness. Moreover, the hippocampi of these poorly cared-for pups exhibited lower levels of GR expression, reduced levels of GR promoter-associated histone acetylation together with

© 2015 Wiley Periodicals, Inc.

59

wires.wiley.com/sysbio

Overview

(a)

Persistently high HPA axis activity, high Crh, Avp expression, high serum cortisol

Elevated HPA axis activity

Early-life stress: reduced maternal care

Methylation and silencing of Nr3C1, Bdnf

Anxiety, elevated stress reactivity, fearfulness in adulthood, care-giving deficits

Altered HPA axis activity?

(b)

?

?

Early-life stress: childhood adversity and abuse

Increased expression of FKBP5, genome-wide DNA methylation, silencing of NR3C1

Severe adult psychiatric disorders, depression, violence, suicide, and PTSD

FIGURE 7 | Parallels between the effects of early-life stress on (a) hypothalamo-pituitary-adrenal (HPA) axis function, epigenetic modifications, gene transcription, and adult behavior in rodents, and (b) its impacts on the human epigenome and psychiatric disorders in adulthood, potentially via HPA axis dysregulation.

increased DNA methylation of the GR promoter, which were accompanied by correspondingly higher levels of hypothalamo-pituitary-adrenal (HPA) axis activity and serum cortisol28,29 (Figure 7(a)). All of these phenotypes, along with behavioral responses to stress, were reversed by infusion of a histone deacetylase inhibitor,28 implying a causal role for epigenetic changes in regulation of HPA axis activity and raising the possibility of therapeutic intervention by targeting epigenetic machinery. Another form of early-life stress is caused by the separation of infant rats from their mothers for extended periods each day during the first 2 weeks of life. This separation stress reduced methylation of DNA sequences within the AVP promoter and increased expression of both AVP in the hypothalamus and the AVP target gene POMC in the pituitary gland30 (Figure 7(a)). The consequences of early-life stress for neural activity-regulated gene expression and care-giving behaviors are further illustrated by experiments in which rat pups, reared by experimentally stressed mothers displaying abusive maternal behaviors, exhibited reduced transcription and increased CpG methylation of the synaptic plasticity gene Bdnf , along with significant deficits in their later ability as adults to care for their own offspring.31 As might be expected for such a powerful biological mechanism, the epigenetic effects of stress are not 60

solely confined to early-life stages. Chronic social stress in adult mice, which has considerable parallels to clinical depression-related stress in human patients, caused persistent demethylation and transcription of the Crh gene, and social avoidance behavior in these adults.32 Interestingly, experimental inhibition of stress-induced Crh expression in this model also attenuated stress-induced social avoidance, demonstrating a causal link between the stress-induced increase in Crh gene transcription and the reduction in social behavior. Links between epigenetic modifications, gene transcription, adverse early-life experiences, and later-onset psychiatric disorders have been discovered in human patients, and they are remarkably similar to those observed in rodents. The first such human study reported reduced levels of mRNA from the GR gene NR3C1 that were correlated with increased NR3C1 CpG methylation, in postmortem hippocampal tissue from adult suicide victims with a documented history of childhood abuse33 (Figure 7(b)). Consistent changes in NR3C1 transcription and DNA methylation were not observed in suicide victims without a history of childhood maltreatment, suggesting that stressful early-life experiences were responsible for epigenetic changes in the brain, which could have dysregulated affective behavior in later life. Recent

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

follow-up studies have compared the hippocampal transcriptomes and epigenomes of suicide victims with a history of childhood abuse and adult rats subjected to early-life stress. This work has uncovered the existence of evolutionarily conserved, widespread transcriptomic, and epigenomic responses to early-life stress that are associated with severe adult psychiatric disorder.34–36 The observation of similar links between a history of adversity in childhood and increased DNA methylation at the NR3C1 locus in the white blood cells of healthy adults37 further underscores the pervasiveness and persistence of changes to gene expression mechanisms that are wrought in the human body by early-life adversity. Additional insights into how stressful experiences in later life erode resilience to further stress have emerged from recent studies of post-traumatic stress disorder (PTSD). PTSD is characterized by a breakdown of psychological resilience to a traumatic stressor, leading to recurrent, frightening memories of the traumatic experience, which can be accompanied by elevated stress sensitivity, chronic anxiety, increased startle responses, physical incapacity, and/or cognitive dysfunction.38 The neural structures that are thought to be affected in PTSD lie mostly within the forebrain, and include the hypothalamus, hippocampus, and amygdala. Patients with PTSD exhibit reduced ACTH and serum cortisol concentrations, indicating that normal responsiveness of the HPA axis is impaired. These observations contrast starkly with the situation in patients with depression, where serum cortisol levels are often increased.39 Precisely how HPA axis responsiveness to stressors becomes blunted in patients with PTSD, thereby compromising their ability to cope, is poorly understood. However, a history of early-life adversity is associated with an increased risk of PTSD as an adult, suggesting potential roles for developmental and epigenetic mechanisms in its etiology.39,40 A recent study compared epigenomic and transcriptomic profiles in the peripheral blood leukocytes of adult PTSD patients with either a documented history or no history of abuse in childhood.41 The results showed that patients with a history of childhood abuse exhibited distinct and well-correlated patterns of transcriptomic and DNA methylation changes in comparison to those exhibited by PTSD patients with no such history, which were much more variable.41 These insights suggest that epigenetic changes elicited by early-life stress might canalize gene expression in a particular direction that predisposes to PTSD in adulthood. The hypothesis that epigenetic changes may dysregulate HPA axis activity is supported further by studies showing that adults with PTSD and a history of childhood trauma exhibited increased Volume 7, March/April 2015

stress-dependent transcription of the gene encoding the GR inhibitor FKBP5, and allele-specific, childhood trauma-dependent decreases in DNA methylation at glucocorticoid-response elements within the FKBP5 promoter.42 Childhood traumatic abuse is associated with increased risks of violence, delinquency, hazardous behavior, and a range of other chronic diseases besides PTSD43 (Figure 7(b)), and a recent study further demonstrated strong associations between a history of abuse during childhood and the presence of specific patterns of genomic DNA methylation in white blood cells of adults.36 Thus, adverse psychological stress during childhood is accompanied by epigenetic changes that may lead to a variety of severe health risks in adulthood. Further studies in animal models and human cohorts will help to determine the roles of epigenetic mechanisms in mediating the effects of stressful experiences on disease risk.

EPIGENETIC EMBEDDING OF DISEASE RISK DURING DEVELOPMENT During mammalian fetal gestation, the maternal environment provides a powerful combination of hormonal and nutritional signals, whose effects on fetal development and physiology persist well beyond childhood and into adult life. Barker and Osmond first postulated that maternal experience, particularly nutrition, exerts significant influence over the susceptibility of the developing fetus to chronic diseases during adulthood.44 It was noted that offspring with a low birthweight had increased risks of coronary heart disease, hypertension, stroke, and diabetes.44,45 Moreover, individuals who experienced a period of maternal undernourishment during the first trimester of fetal gestation had increased risk of hypertension and stroke as adults, whereas exposure to a period of maternal undernourishment during the second trimester of fetal gestation was associated with an increased risk of diabetes and heart disease in adulthood. Furthermore, poor maternal nutrition in the third trimester was linked to clotting disorders in adulthood.46 Such developmentally regulated, differential disease susceptibility is a clear indication that the plasticity of fetal development facilitates the emergence of distinct, experience-dependent phenotypes that can be maladaptive in later life.47 Thus, it has been suggested that in times of poor nutrition, the fetus responds to maternal signals of adverse nutritional conditions by making ‘thrifty’ endocrine and metabolic adaptations which enable energy sources to be conserved, in anticipation of potentially adverse postnatal conditions. However, if nutrition then becomes plentiful in childhood or adulthood,

© 2015 Wiley Periodicals, Inc.

61

wires.wiley.com/sysbio

Overview

it has been further proposed that the persistence of the thrifty adaptations made during fetal gestation elicits behaviors and/or altered metabolism that lead to excess consumption of energy-dense food, much of which is then converted into fat for storage in adipose tissue, increasing the risk of disorders such as cardiovascular disease, stroke, obesity, and diabetes. The thrifty phenotype hypothesis is supported by epidemiological research on communities exposed to extreme hardship. Over a period of 7 months during the winter of 1944–1945, severe food rationing was imposed on the people of the western Netherlands by the occupying German army. During this food blockade, in which adults experienced conditions of severe malnutrition, the health records of babies conceived or born during this period continued to be taken meticulously. Many years later, it became apparent that people who had been exposed to the food blockade as fetuses had increased likelihoods of developing diabetes, mood disorders, renal dysfunction, and obesity as adults.48–50 Moreover, fetal exposure to famine was associated with altered patterns of DNA methylation near genes likely to be involved in fetal growth and development.49,50 For example, analysis of DNA methylation patterns in blood samples of individuals collected at age 59 showed that exposure to famine or at around the time of conception was associated with much lower levels of DNA methylation at the IGF2 growth regulatory locus than was observed in DNA samples from their unexposed, same sex siblings.49 By contrast, individuals exposed to famine in late gestation showed no differences in the pattern of DNA methylation at the IGF2 locus in comparison to their unexposed siblings, implying that the response to famine was developmentally regulated.49 Reports of prenatal famines in a variety of geographical locations, including Scandinavia, Russia, the Gambia, and China, provide broader support for the link between prenatal famine and adult onset diseases such as diabetes, and some of these reports also suggest a role for epigenetic mechanisms in mediating this link.51 A recent study of rural communities in the Gambia found that seasonal variation in the level of methyl-donor nutrient intake by mothers at conception had dramatic effects on the pattern of DNA methylation in somatic tissues of the resulting offspring,52 although it remains unclear whether these altered patterns of DNA methylation are associated with altered disease risk in offspring. Another recent study of multigenerational, isolated preindustrial communities in Finland identified a robust relationship between the occurrence of limited food availability during human fetal gestation and lower life expectancy of offspring after famine in 62

later life, suggesting that limited nutrition during early development erodes physiological resilience to later adversity.53 Experiments in rodents have confirmed that an environmental mismatch between nutritional conditions during fetal development and in adult life predisposes to disorders such as diabetes and hypertension. Thus, severe, prolonged, global undernourishment of pregnant female rats caused low birthweight in offspring, who subsequently developed high blood pressure, endocrine abnormalities, reduced movement, hyperphagia, and obesity54 (Figure 8(a)). A protein-restricted diet administered to pregnant rats similarly caused cardiometabolic disorders in their offspring, which included hypertension, obesity, and a preference for high-fat foods.55–57 The low-protein diet-induced hypertension was associated with increased expression and decreased DNA methylation of the angiotensin 1b receptor Agtr1b in the adrenal gland, all of which were suppressed by inhibition of glucocorticoid synthesis.58 DNA methylation has been implicated as a mechanism through which nutritional deprivation signals transmit information from mother to fetus, thereby influencing gene transcription and physiology in offspring.59,60 In the offspring of parents maintained on a restricted diet, DNA methylation and transcription of genes encoding the GR and PPAR𝛼 transcription factor in rodent offspring were particularly sensitive to the effects of parental dietary restriction,61–63 which fits well with other studies showing that HPA axis activation and excessive glucocorticoid-mediated stress responses underpin the low birthweight and later cardiometabolic disorders found in the offspring of undernourished mothers.25,64 Neuroendocrine consequences of reduced diet during pregnancy have also been observed in the sheep fetal nervous system. When pregnant ewes were maintained on a restricted diet over a period extending from 2 months before the end of the first month of pregnancy, a specific reduction in DNA methylation of both the NR3C1 promoter and the POMC promoter was found in hypothalamic tissue obtained from late gestation fetuses.65 Moreover, in dissected hypothalamic tissues, both NR3C1 and POMC promoter DNA sequences were specifically associated with hyperacetylated, H3K4 hypermethylated, and H3K27 hypomethylated histones in hypothalamus tissue from maternally underfed fetuses, consistent with these genes being located within chromatin that becomes transcriptionally more permissive in adverse circumstances. Taken together, human epidemiology and experimental studies in model organisms indicate that developmental plasticity allows the mammalian

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

(a)

Experience-sensitive epigenetic mechanisms

Hypomethylation and transcription of Agtr1b Nr3c1, Pomc, Ppara

? Hypertension, obesity, endocrine dysfunction, hyperphagia in offspring

Maternal undernourishment during pregnancy Low birthweight of offspring

(b)

?

(c) Hypomethylation and increased transcription of II13ra2, II1rI1

Methylation and silencing of Ppara in offspring liver Paternal high-fat diet

Paternal undernourishment

Pancreatic β cell dysfunction in offspring

Decreased serum glucose levels in offspring

FIGURE 8 | Maternal and paternal undernourishment induce cardiometabolic disorders in offspring, which are accompanied by epigenetic and transcriptional changes.

fetus to respond adaptively to environmental and nutritional cues, but some physiological changes that are acquired during fetal development can be maladaptive and cause chronic metabolic disease, particularly if the external environment changes radically in later life (Figure 8(a)). There is, however, a growing body of evidence which suggests that in addition to the physiological changes acquired by somatic tissues, environmentally induced modifications occur in the germline of exposed individuals that confer disease risks to their offspring and to their grandchildren. Epidemiological research on inhabitants of the remote, agricultural region of Overcalix in northern Sweden showed that exposure of paternal grandfathers to a good food supply in their slow-growth years, just before puberty, was associated with a significantly increased risk of early mortality for their grandsons but not for their granddaughters.66–68 By contrast, a period of poor food supply (due to poor harvests) for paternal grandfathers, in their slow-growth years just before puberty, was associated with reduced risk of early mortality to their grandsons. Moreover, exposure of paternal grandmothers to periods of good food supply in the slow-growth years, just before puberty was associated with an increased risk of early mortality for their granddaughters, but not for their grandsons. Neither pattern of exposure–phenotype relationship was observed for maternal grandparents and their offspring, indicating that the altered risk Volume 7, March/April 2015

of early mortality due to grandparental nutritional experience around puberty required transmission through the father’s germline. Such paternally transmitted transgenerational effects of altered diet have also been observed in rodents. For example, male mice fed a poor diet low in protein sired offspring that exhibited altered transcription of genes involved in lipid metabolism, including the metabolic regulatory gene, Ppara.69 The reduced transcription of Ppara was accompanied by increased DNA methylation in a known Ppara enhancer element in offspring, indicating that paternal dietary restriction induced heritable epigenetic changes in sperm that directly affected metabolic gene expression in their progeny.69 A similar conclusion was drawn from a separate rodent study which showed that preconceptional fasting of fathers also led to reduced serum glucose concentrations in their offspring.70 Another report further demonstrated that reduced nutrition of fetal male mice in utero caused them to develop obesity, glucose intolerance, and altered liver metabolism through reprogramming of gene expression and modification of the liver epigenome.63 The progeny of these males exhibited glucose intolerance as well as changes to genomic DNA methylation and metabolic gene expression, demonstrating a robust, heritable association between the environmentally triggered epigenomic and physiological changes.63 Other reports indicate that paternal high-fat diets alter the

© 2015 Wiley Periodicals, Inc.

63

wires.wiley.com/sysbio

Overview

patterns of chromatin modifications in sperm, leading to pancreatic 𝛽-cell dysfunction in offspring.71,72 Taken together, the evidence that paternal diet confers metabolic disease risks to offspring via epigenetic modification of the germline is compelling (Figure 8). The studies of human communities exposed to temporary but harsh nutritional conditions have revealed remarkable connections between transient, adverse early-life experiences and the later emergence of chronic disease states in adulthood.51 Moreover, the epigenomic analyses that have been performed to date, in both humans and experimental animals, indicate that stable experience-dependent epigenetic changes are likely involved in altering disease risk, both in nutritionally deprived individuals and in their progeny. While the human studies are relatively rare, they reveal that when communities experience major transitions from inadequate to abundant food supplies, a burden of chronic disease is likely to be acquired. Today, this can be most clearly observed where the resource-intensive forces of globalization are transforming poor, underdeveloped societies into emerging market economies with rapidly expanding consumer populations and rising affluence. As citizens of these developing countries make the transition from undernutrition to overnutrition, it is clear that the rates of cardiometabolic disorders are rising astonishingly quickly.73,74 In order to tackle the increasing prevalence of these diseases, a key role can be envisioned for epigenetic biomarker analysis in human cohort studies both to elucidate the pathogenetic mechanisms and monitor the effectiveness of specific interventions.

TRANSGENERATIONAL INHERITANCE OF ACQUIRED DISEASE RISKS The Overcalix studies demonstrated that grandparental exposure to famine before puberty conferred sex-specific, decreased risks of early mortality to grandchildren.66–68 Although the mechanisms underlying the male-specific transmissibility of this altered mortality risk remain poorly understood, the patterns of inheritance of these risks strongly imply that an epigenetic memory of famine was somehow established within germline chromatin of the exposed individuals and transmitted only through their sons’ germline and only to offspring that were of the same sex as each famine-exposed grandparent. The fact that both grandmothers and grandfathers transmitted similarly increased risks of mortality through their sons to their grandchildren eliminated the possibility that the increase in mortality was a specific effect of famine on maternal physiology that was experienced 64

by progeny during fetal gestation. Studies of the grandchildren of women who experienced the Dutch famine of 1944–1945 revealed that the children of their male offspring were more obese than the children of their female offspring,75 further suggesting that epigenetic reprogramming during male germline development is particularly sensitive to nutritional factors. As reviewed above, many experimental studies in rodents provide compelling evidence that unbalanced nutrition in adulthood induces epigenetic changes in germline chromatin which are transmitted to offspring via gametes, affecting gene transcription and increasing transgenerational susceptibilities to adult metabolic disease.63,69,71,72,76 In seeking to explore the possibility that additional environmental or behavioral exposures beyond nutrition may determine metabolic disease risk, recent studies of the ALSPAC cohort have identified a strong and specific link between prepubertal onset of paternal tobacco smoking and increased obesity in sons.77 Indeed, a compelling body of evidence demonstrates that cigarette smoking is a powerful modulator of DNA methylation at specific genomic loci, in both adult smokers and children whose mothers smoked during pregnancy. Several groups have identified robust smoking-related changes in white blood cell DNA methylation patterns close to genes such as F2RL3 and AHRR.78–82 Some of these DNA methylation changes persist in white blood cells for many years after smoking cessation,80 implying that these changes are likely acquired as stable epigenetic marks by the hematopoietic stem cells which give rise to white blood cells. The transgenerational transmission of a smoking-related increased risk of obesity77 may similarly involve persistence of smoking-induced DNA methylation patterns or other epigenetic changes in the germline. Given the powerful links between smoking and other chronic diseases such as cancer and cardiovascular disease, it is possible that persistent smoking-induced epigenetic modifications to chromatin structure also play key roles in tumorigenesis and atherogenesis. Other environmental factors that confer transgenerationally inherited disease susceptibilities include exposure to persistent organic pollutants or endocrine disrupting agents, which are accompanied by altered patterns of epigenetic modifications in the chromatin of both germline and somatic tissues. Exposure of pregnant female rats to the synthetic fungicide and antiandrogen vinclozolin, during the period of fetal gonadal sex determination when extensive chromatin remodeling occurs in germ cells, induced high levels of spermatogenic defects in the testes of male offspring with little corresponding effect on the fertility

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

of female progeny.83 Moreover, these defects persisted for four generations of male offspring, and they correlated with the appearance of altered patterns of spermatogenic gene expression, as well as altered DNA methylation in the sperm chromatin from affected males, whereas no such effects were seen in female offspring. Multigenerational transmissible defects of spermatogenesis, as well as increased incidence of adult-onset chronic disorders such as obesity, have been shown to result from transient exposure of pregnant female rats to other environmental toxicants during fetal gonadal sex determination. Substances that have been demonstrated to elicit such epigenetic effects include dioxin, the pesticides permethrin, diethyltoluamide (DEET), and dichlorophenyltrichloroethane (DDT), plastics-derived endocrine disruptors such as bisphenol A (BPA), and hydrocarbon fuel.84–88 The resulting developmental abnormalities were accompanied by altered patterns of sperm DNA methylation in the vicinity of a subset of genes, some of which have previously been implicated in obesity. During pregnancy, the maternal HPA axis is activated by exposure to increased levels of circulating glucocorticoid hormone, which affects both fetal development and maternal physiology. Studies in humans and mice indicate that an increased level of maternal glucocorticoid induces CRH production by the placenta, which activates both maternal and fetal HPA axes, and is associated with low birthweight, cardiometabolic disorders, and impaired cognitive function in offspring.25 Multigenerational inheritance of prenatal glucocorticoid-induced low birthweight and glucose intolerance has been observed in the offspring of rats exposed to high levels of maternal glucocorticoids during fetal gestation,89–91 suggesting a likely corticosteroidal effect on the fetal germline. Furthermore, studies of the impact of reduced maternal care on the behavior of newborn offspring and their descendants strongly suggest that early-life stresses such as chronic maternal separation stress induce depressive behaviors in offspring on reaching adulthood, which are accompanied by stable changes in expression and epigenetic modifications of genes involved in regulating HPA axis activity.24,92,93 The altered behaviors documented in these studies, as well as the epigenetic and transcriptional changes that accompany them, were transmitted from the F1 to the F2 and F3 generations, through both male and female germlines. Furthermore, the transgenerational inheritance of both the depressive behaviors and the accompanying epigenetic modifications was independent of rearing conditions, because these Volume 7, March/April 2015

changes persisted whether offspring were raised by biological or surrogate mothers.92–94 Additional studies of the effects of vinclozolin exposure further indicate that transgenerational behavioral effects of this chemical are transmitted through the male germline. When compared to offspring of controls, third-generation adult offspring of vinclozolin-treated adult males exhibited altered behavioral reactivity to a chronic restraint stress experienced during adolescence, which was accompanied by altered gene expression in the brain.95 Interestingly, exposure of adult male mice to prolonged, chronic stress caused robust increases in the expression of a specific group of miRNAs in their germ cells, which were accompanied by substantial reductions in HPA axis responsivity in offspring and altered hypothalamic gene expression.96 This discovery suggested a novel epigenetic mechanism for transgenerational impacts of behavioral stressors via modification of sperm miRNA content. The stress-responsive miRNA hypothesis is supported by another study which demonstrated that early-life stress induced transgenerationally inherited metabolic and behavioral abnormalities, along with altered expression of sperm miRNAs, in male progeny.97 Remarkably, microinjection of RNAs extracted from sperm of traumatized adult males into fertilized eggs was sufficient to induce the same metabolic and behavioral changes in the resulting progeny that were transmitted through the germline. Sperm miRNAs have also been implicated in the transgenerational inheritance of metabolic disturbances caused by high-fat-diet-induced obesity in male mice,76 further supporting the emerging view that like chromatin modifications, sperm-derived sncRNAs are heritable developmental signals elicited by environmental cues and transmitted to progeny at fertilization, allowing them to modulate gene expression in potentially any cell type (Figure 9). A particularly compelling example of a transgenerationally inherited behavioral change is illustrated by experiments using a classical behavioral conditioning paradigm to train mice to fear the odorant acetophenone, which is specifically recognized by the olfactory odorant receptor Olfr151.98 When the ability of offspring to exhibit fear-related startle responses to acetophenone was measured, both male and female offspring of fear-conditioned male mice exhibited elevated startle responses to acetophenone but not to control odorants, which persisted in the F2 generation.98 The mechanism underlying the germline transmission of the behavior produced by paternal experience remains incompletely understood, but expression of an Olfr151-lacZ transgene was increased in the olfactory system of F1 and

© 2015 Wiley Periodicals, Inc.

65

wires.wiley.com/sysbio

Overview

(a)

Increased expression of specific miRNAs in sperm

? ? Altered transcription of hypothalamic genes

Exposure of adult males to chronic stress

(b)

Altered sperm miRNA profile

?

Metabolic and behavioral abnormalities in offspring

Neonatal males early-life stress

(c)

Reduced HPA axis activity in offspring

Altered sperm ncRNA profile

?

Metabolic abnormalities in offspring

Adult males on high-fat diet

FIGURE 9 | Transgenerational inheritance of dietary- and stress-induced metabolic and behavioral phenotypes in mice is accompanied by altered expression of microRNAs in sperm, suggesting a potential regulatory role for these molecules in epigenetic programming.

F2 progeny, and a specific CpG dinucleotide in the Olfr151 locus was hypomethylated in the sperm of both F0 fear-conditioned mice and their F1 male progeny. Taken together, these results suggest that an epigenetic change at the Olfr151 locus in sperm may be responsible for both the transgenerationally stable, increased expression of Olfr151 in the nervous system and the accompanying increase in behavioral sensitivity to acetophenone.

CONCLUSION AND SOCIETAL IMPLICATIONS This review draws together recent evidence clarifying the roles of epigenetic mechanisms in facilitating experience-dependent changes to health and well-being, not only during development but also across the life course and between generations. Epigenetic changes to chromatin structure and function underlie maladaptive responses to adverse behavioral and physiological stressors, such as poor parental care and experimentally conditioned fear memories. Epigenetic processes are also implicated in the biological embedding of cardiometabolic disease risks by poor nutrition and the adverse impacts of endocrine-disrupting environmental pollutants on 66

health. One of the clearest and perhaps most provocative insights emerging from this research is that chemical, nutritional, and behavioral exposures confer increased risks for some disorders that are heritable, through the introduction of stable epigenetic changes and altered gene expression in multiple generations of progeny, both in human populations and in experimental animals. These discoveries have done much to reignite the nature-nurture debate, by revealing the existence of mechanisms whereby some acquired traits, particularly chronic diseases and behavioral disorders, are determined by early-life events but only expressed much later. Most intriguingly, some of the known traits are transmitted to subsequent generations independently of the need for mutations or re-exposure to the initial phenotype-altering stimulus. Better understanding of how social, environmental, and nutritional factors modify the developmental processes underlying disease risk will be invaluable for improving public health.99,100 Epigenetic biomarkers that can help to identify pathological causes and effects will also be of considerable utility in stratifying populations, monitoring at-risk individuals, evaluating the effectiveness of measures to reduce or eliminate contextual risks, and facilitating the development of new therapies for overt disease.

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

As the number and variety of links between epigenetics and experience-sensitive phenotypes has grown, the wider societal implications of these findings have begun to receive greater attention, particularly in relation to understanding how social experience affects health and well-being, and the pathophysiological mechanisms which link social and health inequalities.101 Two recent studies describe significant correlations between socioeconomic position and DNA methylation in adult white blood cells,102,103 one of which further indicated links between socioeconomic status, the level of global DNA methylation, and the incidence of cardiovascular disease.103 While these reports offer only partial glimpses of potential mechanisms through which human socioeconomic status could become biologically embedded, they identify social status as a powerful source of psychological and material signals that could engender chronic disease through epigenetic mechanisms. As discussed in this review, behavioral stressors and poor nutrition are good examples of two distinct, but intimately related categories of extrinsic signals, and more research will now be required to explore their mechanisms of action. Ongoing epigenetic analysis of large, phenotypically detailed human cohorts will generate and evaluate specific hypotheses about the relationships between experience-derived signals, epigenetic changes, and altered behavioral and chronic disease risk. To further investigate causal interactions, these hypotheses will require rigorous testing in experimentally tractable, phenotypically relevant model organisms. The possibility that epigenetically programmed diseases are induced by avoidable environmental factors raises important ethical questions about collective and individual responsibilities to minimize known health risks, develop human capabilities, and apply principles of distributive justice for current and future generations.100,104–106 While the health impacts of adverse maternal experiences during fetal gestation provide a basis for targeting public health advice to pregnant mothers, there is compelling evidence that the male germline is also vulnerable to environmental impacts which confer

substantial health risks on offspring. The clear implication of these findings is that effective mitigation of environmental health risks is unlikely to be achieved by sex- or life-stage-specific behavior change, but will require action that recognizes the much greater breadth of these risks across the life course. The developmental biologist Conrad Waddington originally coined the term epigenetics to describe ‘the science concerned with the causal analysis of development’.107 Waddington sought to understand the dynamic interactions between genes, cells, and tissues during embryogenesis, finding inspiration for his thinking about these interactions in early computational research on the behavior of complex systems.108 Recognizing that the embryo is a dynamic biological system in which feed-forward and feedback signals modulate the stability and responsiveness of phenotypes to internal and external signaling processes, Waddington demonstrated that heritable adaptations to environmental factors could be acquired through a process he termed ‘genetic assimilation’.109 Waddington was also among the first to understand that the adverse effects of human activity on the natural world and within society would produce environmental, societal, and behavioral changes that could compromise human health and well-being over the long term,110 and we now have a much clearer view of the epigenetic nature of some of the interactions underlying this phenotypic plasticity, important examples of which are described above. Viewed in this light, it is apparent that Waddington’s wide-ranging contributions to developmental biology, epigenetics, and environmental science have paved the way for the emergence of a consilient, systems-level understanding of the causal interconnections linking human behavior to rising health inequalities and adverse environmental change. Practical applications of this rapidly growing body of knowledge are now needed to preserve and improve human health in the 21st century, and epigenetics research will provide an important scientific evidence base supporting the pursuit of these aims.

REFERENCES 1. Clouaire T, Stancheva I. Methyl-CpG binding proteins: specialized transcriptional repressors or structural components of chromatin? Cell Mol Life Sci 2008, 65:1509–1522.

3. Rai K, Huggins IJ, James SR, Karpf AR, Jones DA, Cairns BR. DNA demethylation in zebrafish involves the coupling of a deaminase, a glycosylase, and Gadd45. Cell 2008, 135:1201–1212.

2. Niehrs C, Schäfer A. Active DNA demethylation by Gadd45 and DNA repair. Trends Cell Biol 2012, 22:220–227.

4. Zhu J-K. Active DNA demethylation mediated by DNA glycosylases. Annu Rev Genet 2009, 43: 143–166.

Volume 7, March/April 2015

© 2015 Wiley Periodicals, Inc.

67

wires.wiley.com/sysbio

Overview

5. Lee J-S, Smith E, Shilatifard A. The language of histone crosstalk. Cell 2010, 142:682–685. 6. Simon JA, Kingston RE. Occupying chromatin: polycomb mechanisms for getting to genomic targets, stopping transcriptional traffic, and staying put. Mol Cell 2013, 49:808–824. 7. Schuettengruber B, Martinez A-M, Iovino N, Cavalli G. Trithorax group proteins: switching genes on and keeping them active. Nat Rev Mol Cell Biol 2011, 12:799–814.

22. Maze I, Noh K-M, Allis CD. Histone regulation in the CNS: basic principles of epigenetic plasticity. Neuropsychopharmacology 2013, 38:3–22. 23. McEwen BS. Physiology and neurobiology of stress and adaptation: central role of the brain. Physiol Rev 2007, 87:873–904.

8. Noordermeer D, Duboule D. Chromatin architectures and Hox gene colinearity. Curr Top Dev Biol 2013, 104:113–148.

24. Franklin TB, Saab BJ, Mansuy IM. Neural mechanisms of stress resilience and vulnerability. Neuron 2012, 75:747–761.

9. Millar CB. Organizing the genome with H2A histone variants. Biochem J 2013, 449:567–579.

25. Reynolds RM. Glucocorticoid excess and the developmental origins of disease: two decades of testing the hypothesis—2012 Curt Richter Award Winner. Psychoneuroendocrinology 2013, 38:1–11.

10. Chen T, Dent SYR. Chromatin modifiers and remodellers: regulators of cellular differentiation. Nat Rev Genet 2014, 15:93–106. 11. Gendrel A-V, Heard E. Noncoding RNAs and epigenetic mechanisms during X-chromosome inactivation. Annu Rev Cell Dev Biol 2014, 30:561–580. 12. Kanduri C. Kcnq1ot1: a chromatin regulatory RNA. Semin Cell Dev Biol 2011, 22:343–350. 13. Modarresi F, Faghihi MA, Lopez-Toledano MA, Fatemi RP, Magistri M, van der Brug MP, Wahlestedt C. Inhibition of natural antisense transcripts in vivo results in gene-specific transcriptional upregulation. Nat Biotechnol 2012, 30:453–459. 14. Stuwe E, Tóth KF, Aravin AA. Small but sturdy: small RNAs in cellular memory and epigenetics. Genes Dev 2014, 28:423–431. 15. O’Donnell A, Odrowaz Z, Sharrocks AD. Immediateearly gene activation by the MAPK pathways: what do and don’t we know? Biochem Soc Trans 2012, 40:58–66. 16. Michod D, Bartesaghi S, Khelifi A, Bellodi C, Berliocchi L, Nicotera P, Salomoni P. Calcium-dependent dephosphorylation of the histone chaperone DAXX regulates H3.3 loading and transcription upon neuronal activation. Neuron 2012, 74:122–135. 17. Gibson BA, Kraus WL. New insights into the molecular and cellular functions of poly(ADP-ribose) and PARPs. Nat Rev Mol Cell Biol 2012, 13:411–424. 18. Wiench M, Miranda TB, Hager GL. Control of nuclear receptor function by local chromatin structure. FEBS J 2011, 278:2211–2230. 19. Voss TC, Hager GL. Dynamic regulation of transcriptional states by chromatin and transcription factors. Nat Rev Genet 2014, 15:69–81. 20. Reid G, Gallais R, Métivier R. Marking time: the dynamic role of chromatin and covalent modification in transcription. Int J Biochem Cell Biol 2009, 41:155–163. 21. Bilodeau S, Vallette-Kasic S, Gauthier Y, FigarellaBranger D, Brue T, Berthelet F, Lacroix A, Batista

68

D, Stratakis C, Hanson J, et al. Role of Brg1 and HDAC2 in GR trans-repression of the pituitary POMC gene and misexpression in Cushing disease. Genes Dev 2006, 20:2871–2886.

26. De Kloet ER, Joels M, Holsboer F. Stress and the brain: from adaptation to disease. Nat Rev Neurosci 2005, 6:463–475. 27. Meaney MJ. Maternal care, gene expression, and the transmission of individual differences in stress reactivity across generations. Annu Rev Neurosci 2001, 24:1161–1192. 28. Weaver ICG, Meaney MJ, Szyf M. Maternal effects on the hippocampal transcriptome anxiety-mediated behaviors in the offspring are reversible in adulthood. Proc Natl Acad Sci 2006, 103:3480–3485.

care and that USA

29. Weaver ICG, Cervoni N, Champagne FA, D’Alessio AC, Sharma S, Seckl JR, Dymov S, Szyf M, Meaney MJ. Epigenetic programming by maternal behavior. Nat Neurosci 2004, 7:847–854. 30. Murgatroyd C, Patchev AV, Wu Y, Micale V, Bockmühl Y, Fischer D, Holsboer F, Wotjak CT, Almeida OF, Spengler D. Dynamic DNA methylation programs persistent adverse effects of early-life stress. Nat Neurosci 2009, 12:1559–1566. 31. Roth TL, Lubin FD, Funk AJ, Sweatt JD. Lasting epigenetic influence of early-life adversity on the BDNF gene. Biol Psychiatry 2009, 65:760–769. 32. Elliott E, Ezra-Nevo G, Regev L, Neufeld-Cohen A, Chen A. Resilience to social stress coincides with functional DNA methylation of the Crf gene in adult mice. Nat Neurosci 2010, 13:1351–1353. 33. McGowan PO, Sasaki A, D’Alessio AC, Dymov S, Labonté B, Szyf M, Turecki G, Meaney MJ. Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse. Nat Neurosci 2009, 12:342–348. 34. Labonte B, Yerko V, Gross J, Mechawar N, Meaney MJ, Szyf M, Turecki G. Differential glucocorticoid receptor exon 1(B), 1(C), and 1(H) expression and methylation in suicide completers with a history of childhood abuse. Biol Psychiatry 2012, 72:41–48.

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

35. Suderman M, McGowan PO, Sasaki A, Huang TC, Hallett MC, Meaney MJ, Turecki G, Szyf M. Conserved epigenetic sensitivity to early life experience in the rat and human hippocampus. Proc Natl Acad Sci USA 2012, 109(suppl):17266–17272.

49. Heijmans BT, Tobi EW, Stein AD, Putter H, Blauw GJ, Susser ES, Slagboom PE, Lumey LH. Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proc Natl Acad Sci USA 2008, 105: 17046–17049.

36. Suderman M, Borghol N, Pappas JJ, Pinto Pereira SM, Pembrey M, Hertzman C, Power C, Szyf M. Childhood abuse is associated with methylation of multiple loci in adult DNA. BMC Med Genomics 2014, 7:13.

50. Tobi EW, Lumey LH, Talens RP, Kremer D, Putter H, Stein AD, Slagboom PE, Heijmans BT. DNA methylation differences after exposure to prenatal famine are common and timing- and sex-specific. Hum Mol Genet 2009, 18:4046–4053.

37. Tyrka AR, Price LH, Marsit C, Walters OC, Carpenter LL. Childhood adversity and epigenetic modulation of the leukocyte glucocorticoid receptor: preliminary findings in healthy adults. PLoS One 2012, 7:e30148. 38. Yehuda R, LeDoux J. Response variation following trauma: a translational neuroscience approach to understanding PTSD. Neuron 2007, 56:19–32. 39. Raabe FJ, Spengler D. Epigenetic risk factors in PTSD and depression. Front Psychiatry 2013, 4:80. 40. Yehuda R, Daskalakis NP, Desarnaud F, Makotkine I, Lehrner AL, Koch E, Flory JD, Buxbaum JD, Meaney MJ, Bierer LM. Epigenetic biomarkers as predictors and correlates of symptom improvement following psychotherapy in combat veterans with PTSD. Front Psychiatry 2013, 4:118. 41. Mehta D, Klengel T, Conneely KN, Smith AK, Altmann A, Pace TW, Rex-Haffner M, Loeschner A, Gonik M, Mercer KB, et al. Childhood maltreatment is associated with distinct genomic and epigenetic profiles in posttraumatic stress disorder. Proc Natl Acad Sci USA 2013, 110:8302–8307. 42. Klengel T, Mehta D, Anacker C, Rex-Haffner M, Pruessner JC, Pariante CM, Pace TW, Mercer KB, Mayberg HS, Bradley B, et al. Allele-specific FKBP5 DNA demethylation mediates gene-childhood trauma interactions. Nat Neurosci 2013, 16:33–41. 43. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009, 373:68–81. 44. Barker D, Osmond C. Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet 1986, 327:1077–1081. 45. Barker DJP. The developmental origins of well-being. Philos Trans R Soc Lond B Biol Sci 2004, 359:1359–1366. 46. Barker DJ, Clark PM. Fetal undernutrition and disease in later life. Rev Reprod 1997, 2:105–112. 47. Bateson P, Barker D, Clutton-Brock T, Deb D, D’Udine B, Foley RA, Gluckman P, Godfrey K, Kirkwood T, Lahr MM. Developmental plasticity and human health. Nature 2004, 430:419–421. 48. Lumey LH, Stein AD, Kahn HS, van der Pal-de Bruin KM, Blauw GJ, Zybert PA, Susser ES. Cohort profile: the Dutch Hunger Winter families study. Int J Epidemiol 2007, 36:1196–1204.

Volume 7, March/April 2015

51. Lumey LH, Stein AD, Susser E. Prenatal famine and adult health. Annu Rev Public Health 2011, 32:237–262. 52. Dominguez-Salas P, Moore SE, Baker MS, Bergen AW, Cox SE, Dyer RA, Fulford AJ, Guan Y, Laritsky E, Silver MJ, et al. Maternal nutrition at conception modulates DNA methylation of human metastable epialleles. Nat Commun 2014, 5:3746. 53. Hayward AD, Rickard IJ, Lummaa V. Influence of early-life nutrition on mortality and reproductive success during a subsequent famine in a preindustrial population. Proc Natl Acad Sci USA 2013, 110:13886–13891. 54. Woodall SM, Johnston BM, Breier BH, Gluckman PD. Chronic maternal undernutrition in the rat leads to delayed postnatal growth and elevated blood pressure of offspring. Pediatr Res 1996, 40:438–443. 55. Bellinger L, Lilley C, Langley-Evans SC. Prenatal exposure to a maternal low-protein diet programmes a preference for high-fat foods in the young adult rat. Br J Nutr 2004, 92:513. 56. Langley-Evans SC. Nutritional programming of disease: unravelling the mechanism. J Anat 2009, 215:36–51. 57. Langley-Evans S, Gardner DS, Jackson AA. Maternal protein restriction influences the programming of the rat hypothalamic-pituitary-adrenal axis. J Nutr 1996, 126:1578–1585. 58. Bogdarina I, Haase A, Langley-Evans S, Clark AJL. Glucocorticoid effects on the programming of AT1b angiotensin receptor gene methylation and expression in the rat. PLoS One 2010, 5:e9237. 59. Burdge GC, Slater-jefferies J, Torrens C, Phillips ES, Hanson MA, Lillycrop KA. Dietary protein restriction of pregnant rats in the F 0 generation induces altered methylation of hepatic gene promoters in the adult male offspring in the F 1 and F 2 generations. Br J Nutr 2007, 97:435–439. 60. Lillycrop KA, Phillips ES, Jackson AA, Hanson MA, Burdge GC. Dietary protein restriction of pregnant rats induces and folic acid supplementation prevents epigenetic modification of hepatic gene expression in the offspring. J Nutr 2005, 135:1382–1386. 61. Burdge GC, Lillycrop KA. Nutrition, epigenetics, and developmental plasticity: implications for

© 2015 Wiley Periodicals, Inc.

69

wires.wiley.com/sysbio

Overview

understanding human disease. Annu Rev Nutr 2010, 30:315–339. 62. Lillycrop KA. Effect of maternal diet on the epigenome: implications for human metabolic disease. Proc Nutr Soc 2011, 70:64–72. 63. Martínez D, Pentinat, T, Ribó S, Daviaud C, Bloks VW, Cebrià J, Villalmanzo N, Kalko SG, Ramón-Krauel M, Díaz R, et al. In utero undernutrition in male mice programs liver lipid metabolism in the second-generation offspring involving altered Lxra DNA methylation. Cell Metab 2014, 19:941–951. 64. Drake AJ, McPherson RC, Godfrey KM, Cooper C, Lillycrop KA, Hanson MA, Meehan RR, Seckl JR, Reynolds RM. An unbalanced maternal diet in pregnancy associates with offspring epigenetic changes in genes controlling glucocorticoid action and foetal growth. Clin Endocrinol (Oxf) 2012, 77:808–815. 65. Begum G, Stevens A, Smith EB, Connor K, Challis JR, Bloomfield F, White A. Epigenetic changes in fetal hypothalamic energy regulating pathways are associated with maternal undernutrition and twinning. FASEB J 2012, 26:1694–1703. 66. Kaati G, Bygren LO, Edvinsson S. Cardiovascular and diabetes mortality determined by nutrition during parents’ and grandparents’ slow growth period. Eur J Hum Genet 2002, 10:682–688. 67. Pembrey ME, Bygren LO, Kaati G, Edvinsson S, Northstone K, Sjöström M, Golding J, ALSPAC Study Team. Sex-specific, male-line transgenerational responses in humans. Eur J Hum Genet 2006, 14:159–166. 68. Bygren LO. Intergenerational health responses to adverse and enriched environments. Annu Rev Public Health 2013, 34:49–60. 69. Carone BR, Fauquier L, Habib N, Shea JM, Hart CE, Li R, Bock C, Li C, Gu H, Zamore PD, et al. Paternally induced transgenerational environmental reprogramming of metabolic gene expression in mammals. Cell 2010, 143:1084–1096. 70. Anderson LM, Riffle L, Wilson R, Travlos GS, Lubomirski MS, Alvord WG. Preconceptional fasting of fathers alters serum glucose in offspring of mice. Nutrition 2006, 22:327–331. 71. Ng SF, Lin RC, Laybutt DR, Barres R, Owens JA, Morris MJ. Chronic high-fat diet in fathers programs 𝛽-cell dysfunction in female rat offspring. Nature 2010, 467:963–966. 72. Wei Y, Yang CR, Wei YP, Zhao ZA, Hou Y, Schatten H, Sun QY. Paternally induced transgenerational inheritance of susceptibility to diabetes in mammals. Proc Natl Acad Sci USA 2014, 111:1873–1878. 73. Wells JCK. Maternal capital and the metabolic ghetto: an evolutionary perspective on the transgenerational basis of health inequalities. Am J Hum Biol 2010, 22:1–17.

70

74. Wells JCK. The evolution of human adiposity and obesity: where did it all go wrong? Dis Model Mech 2012, 5:595–607. 75. Veenendaal MVE, Painter RC, de Rooij SR, Bossuyt PM, van der Post JA, Gluckman PD, Hanson MA, Roseboom TJ. Transgenerational effects of prenatal exposure to the 1944–45 Dutch famine. BJOG 2013, 120:548–553. 76. Fullston T, Ohlsson Teague EM, Palmer NO, DeBlasio MJ, Mitchell M, Corbett M, Print CG, Owens JA, Lane M. Paternal obesity initiates metabolic disturbances in two generations of mice with incomplete penetrance to the F2 generation and alters the transcriptional profile of testis and sperm microRNA content. FASEB J 2013, 27:4226–4243. 77. Northstone K, Golding J, Davey Smith G, Miller LL, Pembrey M. Prepubertal start of father’s smoking and increased body fat in his sons: further characterisation of paternal transgenerational responses. Eur J Hum Genet 2014, 22:1382–1386. 78. Joubert BR, Håberg SE, Nilsen RM, Wang X, Vollset SE, Murphy SK, Huang Z, Hoyo C, Midttun Ø, Cupul-Uicab LA, et al. 450K Epigenome-wide scan identifies differential DNA methylation in newborns related to maternal smoking during pregnancy. Environ Health Perspect 2012, 120:1425–1431. 79. Monick MM, Beach SR, Plume J, Sears R, Gerrard M, Brody GH, Philibert RA. Coordinated changes in AHRR methylation in lymphoblasts and pulmonary macrophages from smokers. Am J Med Genet B Neurospychiatr Genet 2013, 159B:141–151. 80. Shenker NS, Polidoro S, van Veldhoven K, Sacerdote C, Ricceri F, Birrell MA, Belvisi MG, Brown R, Vineis P, Flanagan JM. Epigenome-wide association study in the European Prospective Investigation into Cancer and Nutrition (EPIC-Turin) identifies novel genetic loci associated with smoking. Hum Mol Genet 2013, 22:843–851. 81. Zeilinger S, Kühnel B, Klopp N, Baurecht H, Kleinschmidt A, Gieger C, Weidinger S, Lattka E, Adamski J, Peters A, et al. Tobacco smoking leads to extensive genome-wide changes in DNA methylation. PLoS One 2013, 8:e63812. 82. Harlid S, Xu Z, Panduri V, Sandler DP, Taylor JA. CpG sites associated with cigarette smoking: analysis of epigenome-wide data from the sister study. Environ Health Perspect 2014, 122:673–678. 83. Anway MD, Cupp AS, Uzumcu M, Skinner MK. Epigenetic transgenerational actions of endocrine disruptors and male fertility. Science 2005, 308:1466–1469. 84. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner MK. Dioxin (TCDD) induces epigenetic transgenerational inheritance of adult onset disease and sperm epimutations. PLoS One 2012, 7:e46249. 85. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner MK. Pesticide and insect repellent mixture (permethrin and DEET) Induces epigenetic transgenerational

© 2015 Wiley Periodicals, Inc.

Volume 7, March/April 2015

WIREs Systems Biology and Medicine

Experience-sensitive epigenetic mechanisms

inheritance of disease and sperm epimutations. Reprod Toxicol 2012, 34:708–719. 86. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner MK. Plastics derived endocrine disruptors (BPA, DEHP and DBP) induce epigenetic transgenerational inheritance of obesity, reproductive disease and sperm epimutations. PLoS One 2013, 8:e55387. 87. Skinner MK, Manikkam M, Tracey R, GuerreroBosagna C, Haque M, Nilsson EE. Ancestral dichlorodiphenyltrichloroethane (DDT) exposure promotes epigenetic transgenerational inheritance of obesity. BMC Med 2013, 11:228. 88. Tracey R, Manikkam M, Guerrero-Bosagna C, Skinner MK. Hydrocarbons (jet fuel JP-8) induce epigenetic transgenerational inheritance of obesity, reproductive disease and sperm epimutations. Reprod Toxicol 2013, 36:104–116. 89. Drake AJ, Liu L. Intergenerational transmission of programmed effects: public health consequences. Trends Endocrinol Metab 2010, 21:206–213. 90. Drake AJ, Walker BR, Seckl JR. Intergenerational consequences of fetal programming by in utero exposure to glucocorticoids in rats. Am J Physiol Regul Integr Comp Physiol 2005, 288:R34–R38. 91. Matthews SG, Phillips DIW. Minireview: transgenerational inheritance of the stress response: a new frontier in stress research. Endocrinology 2010, 151:7–13. 92. Franklin TB, Russig H, Weiss IC, Gräff J, Linder N, Michalon A, Vizi S, Mansuy IM. Epigenetic transmission of the impact of early stress across generations. Biol Psychiatry 2010, 68:408–415. 93. Franklin TB, Linder N, Russig H, Thöny B, Mansuy IM. Influence of early stress on social abilities and serotonergic functions across generations in mice. PLoS One 2011, 6:e21842. 94. Weiss IC, Franklin TB, Vizi S, Mansuy IM. Inheritable effect of unpredictable maternal separation on behavioral responses in mice. Front Behav Neurosci 2011, 5:3. 95. Crews D, Gillette R, Scarpino SV, Manikkam M, Savenkova MI, Skinner MK. Epigenetic transgenerational inheritance of altered stress responses. Proc Natl Acad Sci USA 2012, 109:9143–9148. 96. Rodgers AB, Morgan CP, Bronson SL, Revello S, Bale TL. Paternal stress exposure alters sperm microRNA content and reprograms offspring HPA stress axis regulation. J Neurosci 2013, 33:9003–9012.

Volume 7, March/April 2015

97. Gapp K, Jawaid A, Sarkies P, Bohacek J, Pelczar P, Prados J, Farinelli L, Miska E, Mansuy IM. Implication of sperm RNAs in transgenerational inheritance of the effects of early trauma in mice. Nat Neurosci 2014, 17:667–669. 98. Dias BG, Ressler KJ. Parental olfactory experience influences behavior and neural structure in subsequent generations. Nat Neurosci 2014, 17:89–96. 99. Rozek LS, Dolinoy DC, Sartor MA, Omenn GS. Epigenetics: relevance and implications for public health. Annu Rev Public Health 2014, 35:105–122. 100. Vineis P, Stringhini S, Porta M. The environmental roots of non-communicable diseases (NCDs) and the epigenetic impacts of globalization. Environ Res 2014, 133:424–430. 101. Landecker H, Panofsky A. From social structure to gene regulation, and back: a critical introduction to environmental epigenetics for sociology. Annu Rev Sociol 2013, 39:333–357. 102. Borghol N, Suderman M, McArdle W, Racine A, Hallett M, Pembrey M, Hertzman C, Power C, Szyf M. Associations with early-life socio-economic position in adult DNA methylation. Int J Epidemiol 2012, 41:62–74. 103. McGuinness D, McGlynn LM, Johnson PC, MacIntyre A, Batty GD, Burns H, Cavanagh J, Deans KA, Ford I, McConnachie A, et al. Socio-economic status is associated with epigenetic differences in the pSoBid cohort. Int J Epidemiol 2012, 41:151–160. 104. Rothstein MA, Cai Y, Marchant GE. The ghost in our genes: legal and ethical implications of epigenetics. Health Matrix Clevel 2009, 19:1–62. 105. Khan F. Preserving human potential as freedom: a framework for regulating epigenetic harms. Health Matrix Clevel 2010, 20:259–323. 106. Loi M, Del Savio L, Stupka E. Social epigenetics and equality of opportunity. Public Health Ethics 2013, 6:142–153. 107. Waddington CH. The Epigenetics of Birds. Cambridge: Cambridge University Press; 1952. 108. Waddington CH. The Strategy of the Genes. London: Allen and Unwin; 1957. 109. Waddington CH. Genetic assimilation of an acquired character. Evolution 1953, 7:118–126. 110. Waddington C. The Man Made Future. London: Croom Helm; 1978.

© 2015 Wiley Periodicals, Inc.

71