The role of diet and exercise in the transgenerational epigenetic landscape of T2DM

Journal name:
Nature Reviews Endocrinology
Volume:
12,
Pages:
441–451
Year published:
DOI:
doi:10.1038/nrendo.2016.87
Published online

Abstract

Epigenetic changes are caused by biochemical regulators of gene expression that can be transferred across generations or through cell division. Epigenetic modifications can arise from a variety of environmental exposures including undernutrition, obesity, physical activity, stress and toxins. Transient epigenetic changes across the entire genome can influence metabolic outcomes and might or might not be heritable. These modifications direct and maintain the cell-type specific gene expression state. Transient epigenetic changes can be driven by DNA methylation and histone modification in response to environmental stressors. A detailed understanding of the epigenetic signatures of insulin resistance and the adaptive response to exercise might identify new therapeutic targets that can be further developed to improve insulin sensitivity and prevent obesity. This Review focuses on the current understanding of mechanisms by which lifestyle factors affect the epigenetic landscape in type 2 diabetes mellitus and obesity. Evidence from the past few years about the potential mechanisms by which diet and exercise affect the epigenome over several generations is discussed.

At a glance

Figures

  1. The main forms of epigenetic modifications.
    Figure 1: The main forms of epigenetic modifications.

    Several types of epigenetic modifications have been identified. (1) Modification of nucleosides in DNA such as by methylation and hydroxymethylation. (2) Post-translational modification of histone proteins by methylation, acetylation, ubiquitylation, SUMOylation, citrullination and ADP-ribosylation. (3) Changes in small noncoding RNA expression. Nucleoside and histone modifications regulate gene transcription by modulating the conformation of the chromatin and the access of DNA binding factors. Small noncoding RNAs (such as microRNAs) regulate gene expression by prompting mRNA degradation or modulating protein translation.

  2. Putative effects of exercise and obesity on the predisposition to metabolic diseases.
    Figure 2: Putative effects of exercise and obesity on the predisposition to metabolic diseases.

    a | In response to exercise or diet, the epigenome of gametes (depicted here for spermatozoa) is remodelled. b | After fertilization, these epigenetic changes will affect the developmental programming of the embryo. In utero, the developing embryo is susceptible to other environmental influences from the mother that can potentiate or suppress signals from the gametes. c | Exercise might influence the spermatozoan epigenome and enable transmission of epigenetic information to affect the development of the central nervous system and brain processes such as feeding behaviour. This process could lead to positive metabolic outcomes. Alternatively, according to the thrifty phenotype hypothesis, if exercise triggers an epigenetic response that is similar to caloric restriction of the gamete, the offspring might have a predisposition for metabolic disease when faced with a state of food abundance due to an increased fat storage capacity. Obese fathers could transmit epigenetic markers on genes that regulate brain development and appetite control, thereby predisposing their offspring to obesity. d | Epigenetic inheritance might be propagated to successive generations by so-called transgenerational epigenetic inheritance.

  3. Potential effect of environmentally induced epigenetic changes on gene expression.
    Figure 3: Potential effect of environmentally induced epigenetic changes on gene expression.

    In this hypothetical model, gene A is hypomethylated, while gene B is hypermethylated in response to environmental cues. Gene expression changes might not occur until a secondary environmental stress, or specific physiological state is altered, thereby enabling specific transcription factors to bind to the hypomethylated gene A, but not to hypermethylated gene B, which results in transcription of gene A. This model also explains the often observed discrepancy between DNA methylation and transcriptional changes. In this model, the same epigenetic modification in multiple tissues might only be functionally relevant in a given tissue if a specific transcription factor is also activated.

Key points

  • Epigenetic processes have been implicated in the pathogenesis of type 2 diabetes mellitus
  • Diet and exercise might affect the epigenome over several generations
  • Epigenetic changes can be driven by DNA methylation and histone modification in response to environmental stressors
  • Regulation of gene expression by DNA methylation and histone modification occurs by a mechanism that impairs the access of transcriptional machinery to the promoters
  • Studying the epigenetic signatures of insulin resistance and the adaptive response to exercise might provide insight into gene–environment networks that control glucose and energy homeostasis.

Introduction

Type 2 diabetes mellitus (T2DM) is a life threatening metabolic disease that is reaching epidemic proportions1. Defects in multiple organ systems that control glucose homeostasis, including the brain, pancreas and peripheral tissues (liver, adipose tissue and skeletal muscle), lead to impaired insulin action and secretion and ultimately the clinical manifestation of T2DM2, 3. Although the molecular basis for this pathology is incompletely understood, genetic and environmental factors, probably in a synergistic manner, contribute to the risk of developing T2DM4. The genetic basis of T2DM has been largely deciphered through genome-wide association studies of diverse clinical cohorts5, 6, 7. Through these efforts, over 100 loci have been identified, which collectively explain ~10% of the variation in the predisposition to T2DM8. The modest effect of these genetic variants on the risk of developing T2DM has heightened interest in identifying the 'missing heritability' of this disease9. Environmentally driven epigenetic modifications of the genome provide a potential molecular basis for the missing heritability in T2DM. Epigenetic processes have been implicated in the predisposition to and progression of T2DM in response to a variety of lifestyle factors or environmental exposures, including undernutrition, obesity, physical inactivity, stress and toxins10, 11, 12, 13. However, direct evidence that epigenetic factors drive the inheritance of T2DM in humans is lacking.

Etymologically, epigenetics means 'aside' genetics, and thus refers to almost all biochemical processes that influence gene expression without changing the genetic code itself14. This broad definition has been refined to include alterations of the DNA molecule (such as DNA methylation) and chromatin structure (by histone positioning and post-translational modification)15. These changes might or might not be heritable, but they affect how cells turn genes on or off. Regulation of gene expression by DNA methylation and histone modifications occurs by a mechanism that impairs the access of transcription machinery to the promoters (Fig. 1). Methylation of specific residues in the genome has the potential to modulate the binding capacity of cis-regulatory elements, whereas modifications on histone proteins regulate the binding of chromatin modifiers, which relax or compact DNA. The expression patterns of small noncoding RNAs that have the potential to be inherited through successive generations constitute an additional class of epigenetic modifications. Epigenetic inheritance refers to epigenetic changes that are inherited by the next generation, whereas transgenerational epigenetic inheritance refers to epigenetic modifications that are heritable over several generations and affect evolution16.

Figure 1: The main forms of epigenetic modifications.
The main forms of epigenetic modifications.

Several types of epigenetic modifications have been identified. (1) Modification of nucleosides in DNA such as by methylation and hydroxymethylation. (2) Post-translational modification of histone proteins by methylation, acetylation, ubiquitylation, SUMOylation, citrullination and ADP-ribosylation. (3) Changes in small noncoding RNA expression. Nucleoside and histone modifications regulate gene transcription by modulating the conformation of the chromatin and the access of DNA binding factors. Small noncoding RNAs (such as microRNAs) regulate gene expression by prompting mRNA degradation or modulating protein translation.

Here, we review how epigenetic control can be exerted through several mechanisms, including DNA methylation, histone modification and microRNA (miRNA)-mediated processes (Fig. 1). We consider the evidence regarding mechanisms by which transgenerational epigenetic inheritance and transient epigenetic modifications influence metabolic health. We also describe the potential mechanisms by which epigenetic changes in T2DM and obesity might influence insulin sensitivity and pancreatic β-cell function. Finally, we consider the emerging biology related to the mechanisms by which modifiable lifestyle factors, including diet and exercise, might alter the plasticity of the epigenome and potentially influence the pathogenesis of T2DM.

Epigenetics in metabolic dysfunction

The thrifty phenotype hypothesis. Poor nutrition at several critical windows of development both in utero and during childhood is associated with the development of metabolic and cardiovascular diseases later in life17, 18. This phenomenon has been referred to as the thrifty phenotype hypothesis19, 20. For the aetiology of T2DM, the thrifty phenotype hypothesis postulates that poor nutrition of the fetus and infant during early development programs gene expression and the metabolic profile of the offspring to anticipate a life of starvation19, 20. This phenotype might become maladaptive when the offspring has abundant access to food and could lead to the development of T2DM. For example, in rodents, early growth retardation in utero due to specific maternal diets is associated with detrimental changes to the structure and function of the β cells of the islets of Langerhans, as well as impaired insulin signalling and the development of T2DM later in life18, 21, 22, 23. Calorie restriction and a low-protein diet during fetal development in rodents might trigger epigenetic changes that could affect the metabolic health of the offspring19, 20, 24. Strikingly, even in humans, nutrient availability during famine exposure in utero and during early childhood and adolescence can affect cardiovascular and metabolic health not only in the current generation, but over several generations19, 20, 25, 26, 27, 28, 29, 30.

Epidemiological studies. Following families over several generations in epidemiological studies has been extremely valuable in providing evidence that exposure to certain environmental factors influences the risk of mortality associated with cardiovascular disease and diabetes mellitus. The effect of prenatal undernutrition on subsequent metabolic health has been investigated in a study of the Dutch famine in 1944 (Ref. 25). The timing of the famine, particularly if it occurred during late gestation, had a profound effect on glucose tolerance of the offspring25. People who experienced the famine in late gestation were more likely to have poor glucose tolerance than those who experienced the famine in early gestation. People experiencing the famine in utero during any stage of gestation were found to have elevated glucose levels and impaired glucose tolerance in adulthood26. Further evidence for a link between food restriction and T2DM has been found in studies of people with in utero famine exposure during the Chinese famine (1959–1961)27, the Ukrainian famine (1932–1933)28 and in conjunction with three major famines in Austria (1918–1919, 1938 and 1946–1947)29. The link is probably due to an effect of food restriction on key organs and tissues during critical periods of development.

Variations in the access to food during childhood might also influence the risk of developing T2DM. Studies of the Överkalix parish in northern Sweden reveal that access to food during a child's slow growth period, before their prepubertal peak in growth velocity, seems to affect their descendants' risk of developing cardiovascular disease or T2DM30. Although mortality attributed to cardiovascular disease in the proband was low if food was not readily available during the slow growth period of the parent, mortality associated with T2DM increased if the grandparent was exposed to a surfeit of food during his or her slow growth period30. The fact that food restriction during the prepubertal period in children affects metabolic health at a point in time well beyond any critical window of fetal development supports the notion that environmental stressors might affect the epigenome at multiple points throughout development19, 20. Moreover, these findings suggest that parental or ancestral early life exposures, such as diet or stress, have a transgenerational effect on health and development in subsequent generations. Curiously, these transmissions were sex-specific; female offspring were affected by their grandmother's, but not their grandfather's, access to food, with the opposite being found in male offspring. Similar intergenerational, sex-specific responses have been also observed in rodents31, 32, 33 and could explain, at least in part, the epidemiological evidence for higher transmission of T2DM by the mother and higher transmission of type 1 diabetes mellitus by the father34. These sex-specific effects probably share mechanisms involved in X-chromosome inactivation, whereby epigenetic processes suppress the expression of one allele on a sex chromosome35. Collectively, these epidemiological studies suggest that nutritional cues affect the genome through transgenerational epigenetic inheritance.

Animal studies. Although definitive evidence supporting a role for food restriction or food excess and stress hormones on epigenetic inheritance in humans is lacking, animal studies are supportive. For example, male offspring of female rats exposed prenatally to dexamethasone, a model of fetal exposure to glucocorticoid stress hormones, showed reduced birth weight, glucose intolerance and elevated hepatic phosphoenolpyruvate carboxykinase (PEPCK) activity36, 37. Strikingly, this phenotype persisted through several generations and was transmitted by either maternal36 or paternal37 lines. Moreover, rats exposed to a low-calorie diet for over 50 generations showed higher susceptibility to diabetes mellitus and perturbed epigenetic markers in the promoter of the insulin gene than control rats on chow diet38. This phenotype was only partially reversed by a normal diet over the next two generations. Given that the animals used were not genetically identical (outbred), the observed persistent transgenerational effect could be caused by genetic selection and not stable epigenetic reprogramming.

A high-fat diet also has a transgenerational influence on insulin sensitivity. Animal studies reveal that either maternal or paternal exposure to a high-fat diet leads to insulin resistance in subsequent generations33, 39. Both maternal and paternal exposure to a high-fat diet seems to have a profound effect on pancreatic β-cell function in rodent offspring throughout subsequent generations33, 40. A maternal high-fat diet induces insulin resistance and deterioration of pancreatic β-cell function, with marked sex-specific differences in adult offspring that are accompanied by adipose tissue inflammation and liver steatosis in male, but not female, mice40. A paternal high-fat diet increases body weight and adiposity and impairs glucose tolerance and insulin sensitivity in adult female, but not male, offspring, with associated changes in the expression of pancreatic islet genes and hypomethylation of the anti-inflammatory gene, Il-13 receptor subunit alpha-2 (Il13ra2)33. Similarly, paternal prediabetes leads to glucose intolerance and insulin resistance in offspring, with altered gene expression patterns in pancreatic islets and downregulation of several genes involved in glucose metabolism and insulin signalling pathways41, as well as increasing the risk of obesity42. Collectively, these studies in rodents reveal that the metabolic state of either the mother or the father can have a deleterious effect on the pancreatic islets and peripheral tissues that control glucose and energy homeostasis in their offspring. Given that both maternal and paternal diets influence pancreatic biology and metabolic health33, 39, factors beyond fetal exposure to the in utero environment seem to have a role.

The effect of excessive caloric intake on the metabolic health of the offspring seems to be evolutionarily conserved across several species. Indeed, a high-sugar maternal diet alters body composition of Drosophila melanogaster larval offspring for at least two generations and confers an obese-like phenotype under suboptimal (high-calorie) feeding conditions in adult offspring43. In addition to calorie excess, low-protein diets also lead to metabolic changes in the offspring. For example, hepatic expression of genes involved in lipid and cholesterol biosynthesis is increased and levels of cholesterol and cholesterol esters are decreased in offspring of male mice fed a low-protein diet44. Epigenomic profiling of livers from the offspring of male mice fed a low-protein diet show modifications in cytosine methylation of peroxisome proliferator-activated receptor α (PPARα), a key nuclear receptor protein that regulates expression of genes involved in lipid metabolism44. Undernutrition during prenatal life seems to alter the male germline methylation at discrete loci, with differentially methylated regions associated with altered gene expression during the embryonic life of the offspring45. Thus, nutritional perturbations in utero, even when the nutritional status is normalized later in life, seem to compromise male germline development and epigenetic reprogramming. These nutritional perturbations seem to permanently alter DNA methylation in the germline of the adult offspring. Collectively, these results support the mounting epidemiological evidence that the parental diet can affect the metabolic and cardiovascular health of the offspring and support a model of environmental reprogramming that influences the heritable epigenome.

In addition to dietary factors, toxins can influence the epigenome and lead to metabolic disease in offspring. Exposure of female rats to jet fuel increased the incidence of multiple pathologies in the F1 generation, including kidney abnormalities in both female and male offspring, prostate and pubertal abnormalities in male offspring and primordial follicle loss and polycystic ovarian disease in female offspring13. The pathologies were also seen in the F3 generation, as demonstrated by the increased incidence of primordial follicle loss and polycystic ovarian disease in female offspring, and obesity in both female and male offspring. Ancestral exposure of female rats to the insecticide dichlorodiphenyltrichloroethane (DDT) also leads to kidney disease, prostate disease, ovary disease and tumour development in F1 adults and obesity in F3 adult male and female rats46. Thus, multiple environmental factors, including hormones, diet, toxins and other stressors, seem to affect the epigenome over several generations. Clearly, studies tracking the effect of maternal and paternal nutrition on subsequent generations are more challenging to perform in humans than in rodents, given the evidence that multiple environmental stressors are likely to influence the epigenome. Extensive biobanks and clinical records of diet exposure along with readily accessible biomarkers will be needed to decipher the effect of different diets on metabolic outcomes over several generations in humans.

Carriers of the epigenetic information

Genetic and epigenetic information is passed on to the next generation by the gametes. After fertilization, epigenetic information carried by the spermatozoa and the oocyte is integrated in the developing embryo by cellular events referred to as epigenetic reprogramming. During epigenetic reprogramming, both maternal and paternal DNA is subjected to gradual demethylation before implantation, followed by de novo methylation after implantation47. The molecular mechanism by which the parental diet affects the metabolic and cardiovascular health of the offspring is elusive, but a role for the gametes in conveying information and cues about the environment to the next generation must be considered.

The effects of a grandmother's suboptimal diet on the reproductive potential of granddaughters, in the absence of any further dietary manipulations in the daughters, was studied in a low-protein diet model in rats48. When the grandmother received a low-protein diet, granddaughters had decreased ovarian reserve and increased intra-abdominal fat mass. These changes were coincident with accelerated accumulation of oxidative stress and mitochondrial DNA copy number instability in the ovaries. Moreover, telomere length of ovarian cells declined more rapidly in the granddaughters of female rats on a low-protein diet compared with those on a control diet, which is indicative of accelerated ageing in the reproductive tract due to the grandmother's diet. Although the female reproductive tract seems to be vulnerable to transgenerational programming via the maternal line, one cannot exclude the possibility that this phenotype is caused by altered development of the reproductive tract in utero of the mother due to her mother's low-protein diet. Moreover, the direct effect of different dietary manipulations and environmental stressors on the oocyte remains largely unknown.

Preclinical studies in rodent models provide evidence that the diet of the father (that is, high in fat33 or low in protein44) or the presence of paternal prediabetes41 has a profound effect on the metabolic health of the offspring. Investigation into the epigenome of the father's sperm reveals that prediabetes alters the overall DNA methylome, with a large portion of differentially methylated genes overlapping altered gene expression patterns in the pancreatic islets in their offspring41. A high-fat diet alters the chromatin structure by modulating histone H3 occupancy at genes involved in the regulation of embryogenesis, as well as differential H3K4me1-enrichment at transcription regulatory genes, concomitant with altered gene expression in the livers of male offspring49. Despite these findings, the effect of diet and environmental stressors on the epigenetic landscape of spermatozoa, particularly in humans, is elusive.

Exercise and epigenetic inheritance

Regular exercise training has widespread beneficial effects, including increased oxidative capacity, improved cardiovascular function, enhanced whole-body glucose homeostasis and fatty acid oxidation50, 51. Pregnant women are often encouraged to participate in regular exercise programs given the overall beneficial outcomes for both the mother and the child52. Indeed, moderate exercise during pregnancy reduces the risk of the offspring developing obesity in childhood and preadolescence, even after adjusting for various maternal and offspring characteristics53. Nevertheless, the molecular mechanisms by which regular exercise by the mother either before or during pregnancy affects the metabolic health of the child or even grandchild are elusive.

One study54 focused on the effects of diet and exercise on peroxisome proliferator-activated receptor γ co-activator-1α (PGC-1α), a transcriptional co-activator that has a key role in mitochondrial biogenesis and oxidative metabolism55. Inbred female mice (C57BL/6) were exposed to chow, high-fat diet or high-fat diet with voluntary wheel exercise for the 6 weeks before conception and throughout pregnancy54. Maternal high-fat diet increased methylation of the Ppargc1a promoter and decreased mRNA expression in skeletal muscle from neonatal and 12-month-old offspring, and led to age-associated metabolic dysfunction. Maternal exercise exposure prevented maternal high-fat diet-induced Ppargc1a hypermethylation and enhanced the expression of Ppargc1a and its target genes, coincident with mitigation of age-associated metabolic dysfunction in the offspring. Although this study was rather limited and focused on only one metabolic gene, the evidence suggests that maternal exercise prevented maternal high-fat diet-induced epigenetic and metabolic dysregulation in the offspring. Future studies on the effects of exercise on promoter methylation and gene expression at the genome-wide level are warranted. Given that regular exercise training has multiple effects on whole-body physiology51, tissues beyond skeletal muscle should also be surveyed.

Evidence from numerous animal studies suggests that regular exercise training by either the mother or the father confers a health benefit on the first generation56, 57, 58, 59. Nevertheless, little is known about the mechanism by which exercise effects various metabolic phenotypes over several generations. One study compared and contrasted the effects of maternal and paternal exercise training in rodents on anatomical and metabolic characteristics, including gene expression in skeletal muscle, to that of their sedentary offspring over two generations60. Exercise training resulted in sex-dependent metabolic improvements, as demonstrated by reduced body weight and fat mass in male offspring and increased skeletal muscle mass and improved glucose tolerance in female offspring. Overall, maternal and paternal exercise training altered the expression of several mRNAs in gastrocnemius muscle through two generations of sedentary offspring. Sedentary F1 female offspring with exercise ancestry were lighter, with lower muscle and omental fat mass, than F1 offspring with sedentary ancestry. F1 male offspring with exercise ancestry had lower skeletal muscle mass than F1 male offspring with sedentary ancestry, but in contrast to female offspring with exercise ancestry, omental fat mass was unaltered. The mechanism for the sex-dependent effect of exercise ancestry on fat mass between male and female offspring is unknown, as a molecular interrogation of omental gene expression profiles or systemic factors controlling lipid metabolism was not performed.

Glucose tolerance was impaired in F2 female offspring with exercise ancestry60, compared with F2 offspring with sedentary ancestry, but this finding is somewhat counterintuitive given the health-promoting benefits of exercise in the parents. Differences in gene expression patterns in skeletal muscle were observed between multiple generations of sedentary male or female mice with either exercise or sedentary ancestry, which suggests in utero effects are involved rather than transgenerational effects. Moreover, the expression pattern of several genes was inverted between F1 male offspring and F2 female offspring, with the expression of metabolic genes such as Alas1, Hk2, Ppard and Ppargc1a found to be reduced in F1 male offspring with exercise ancestry, whereas the expression of these genes was increased in F2 female offspring with exercise ancestry. Overall, exercise ancestry was associated with increased mRNA expression of the aforementioned genes in female offspring and reduced mRNA expression of these genes in male offspring. Although this study highlights the fact that metabolic phenotypes and gene expression profiles in offspring are partly influenced by whether the mother or father undergoes an exercise program, it falls short of delivering mechanistic insight into the way in which this epigenetic information is conveyed.

Studies of oocytes or sperm are required to investigate the mechanism by which epigenetic information is communicated to the next generation. Few studies have examined the epigenetic landscape of the gametes of animals under different environmental stress. A study of inbred (C57B/6J) mice revealed that sperm of exercise-trained fathers provides a potential mechanism for the intergenerational inheritance of metabolic traits58. In this study, offspring of fathers that were exposed to long-term exercise training (wheel running) were more metabolically efficient than offspring from sedentary fathers and consequently more susceptible to metabolic dysfunction when fed a high-fat diet58. Even though exercise training increased body weight and fat mass in the fathers, their offspring had reduced energy efficiency and increased risk of obesity and insulin resistance on a high-fat diet, and these mice also had aberrant expression of several genes in skeletal muscle that are involved in metabolic processes58. Interestingly, analysis of parental sperm from the exercise trained mice revealed some of the genes that were differentially methylated in sperm (H19, Ptpn1, Ogt and Oga) were also differentially expressed in skeletal muscle from the offspring. A relative increase in H19 methylation was also observed in the H19 imprinting control region incorporating CCCTC-binding factor (CTCF)-4 binding site in sperm from the exercise-trained mice, compared with sedentary controls.

Finally, a preliminary analysis of a subset of sperm-borne miRNAs revealed that levels of miR-483-3p, miR-431 and miR-21 were increased, and levels of miR-221 were decreased after exercise training in the sperm of fathers58. The observation that chronic exercise training by the father confers a thrifty phenotype in the offspring, programming them to become metabolically efficient, prone to develop obesity on a high-fat diet and at greater risk of developing insulin resistance might seem counterintuitive, given the health promoting effects of exercise training for the father. In principle, this thrifty phenotype affects the offspring and might increase the risk of T2DM in much the same way as early life exposure (both in utero and during adolescence) to high-fat or low protein diets and/or stressors19, 20, 25 (Fig. 2). Although exercise in lean male rodents before conception induces a thrifty phenotype in the next generation, a study in which dietary-induced (high-fat fed) obese male mice were exposed to an 8 week swim training program before conception provides evidence to suggest that exercise prevents the deleterious effect of obesity on the metabolic features of the next generation61. Exercise restored insulin sensitivity in obese founder mice, concomitant with a degree of normalization of X-linked sperm miRNAs61. Moreover, ancestry exposure to short-term exercise training normalized adiposity and restored insulin sensitivity in female offspring.

Figure 2: Putative effects of exercise and obesity on the predisposition to metabolic diseases.
Putative effects of exercise and obesity on the predisposition to metabolic diseases.

a | In response to exercise or diet, the epigenome of gametes (depicted here for spermatozoa) is remodelled. b | After fertilization, these epigenetic changes will affect the developmental programming of the embryo. In utero, the developing embryo is susceptible to other environmental influences from the mother that can potentiate or suppress signals from the gametes. c | Exercise might influence the spermatozoan epigenome and enable transmission of epigenetic information to affect the development of the central nervous system and brain processes such as feeding behaviour. This process could lead to positive metabolic outcomes. Alternatively, according to the thrifty phenotype hypothesis, if exercise triggers an epigenetic response that is similar to caloric restriction of the gamete, the offspring might have a predisposition for metabolic disease when faced with a state of food abundance due to an increased fat storage capacity. Obese fathers could transmit epigenetic markers on genes that regulate brain development and appetite control, thereby predisposing their offspring to obesity. d | Epigenetic inheritance might be propagated to successive generations by so-called transgenerational epigenetic inheritance.

From a metabolic standpoint, the results from these studies support the notion that exercise might either be beneficial or deleterious to the next generation, depending on the metabolic phenotype or level of adiposity of the father. Even though exercise seemed to alter the sperm epigenome, definitive evidence that any of the identified exercise-response miRs confer a deleterious metabolic outcome in the offspring is lacking. Future studies will require additional proof-of-principle, whereby the differentially expressed sperm miRs identified in various paternal stress models (high-fat diet, low-protein diet, environmental toxins or exercise training) are microinjected into zygotes to assay for effects on reprogramming gene expression in the offspring and/or recapitulating the metabolic phenotype. Studies to assess the effect of the interplay between genes and the environment on the epigenome, as well as the mechanisms by which these factors modify the transcriptional potential of a cell or organ to adapt to changes in diet or exercise to improve overall metabolic health are required62.

Exercising for the next generation

Epigenetic information is transferred to the next generation by the gametes and consequently an effect of diet or weight loss on spermatozoa or oocytes might be more critical for passing on this information to the next generation than any epigenetic change that occurs in peripheral tissues. Indeed, evidence that diet alters the spermatozoa epigenome in adult men is emerging. Comprehensive profiling of the epigenome of spermatozoa from lean and obese men reveal striking differences in small noncoding RNA expression and DNA methylation patterns, with similar histone positioning63. Weight loss induced by bariatric surgery was associated with a dramatic remodelling of sperm DNA methylation, notably at genetic locations implicated in the central control of appetite, including melanocortin-4 receptor (MC4R), brain-derived neurotrophic factor (BDNF), neuropeptide Y (NPY), cannabinoid receptor type 1 (CR1) and cocaine and amphetamine regulated transcript (CART), as well as genes related to obesity and metabolism including fat mass and obesity associated (FTO), carbohydrate sulfotransferase 8 (CHST8) and SH2 binding domain-containing protein 1 (SH2B1)63. Just as the epigenome in human skeletal muscle and adipose tissue exhibits plasticity, weight loss induced by bariatric surgery also seems to remodel spermatozoa DNA methylation, notably at genetic locations implicated in the central control of appetite63. The implications of these epigenomic signatures for the metabolic health of children of obese men are unknown. In adult men, 3 months of exercise training alters global and genome-wide sperm DNA methylation, with exercise-induced changes in DNA methylation in genes related to numerous diseases such as schizophrenia and Parkinson disease64. Whether these exercise-induced adaptations in sperm affect metabolic health in humans is unknown, but experimental studies in mice suggest such exercise-induced changes in the spermatozoa epigenome have a deleterious effect on offspring, leading to reduced energy expenditure and increased risk of obesity58. Clearly, additional clinical studies to ascertain the effects of the paternal and maternal diet and exercise status on the metabolic health of the children are needed before any firm conclusions can be made.

Epigenetic impacts on metabolism

Given epidemiological, as well as experimental, evidence that crude changes in environmental exposure such as a high-fat diet or a low-protein diet, calorie restriction or stress hormones modify the epigenome, it is likely that multiple factors, some obvious and some discrete, might affect the epigenome over a person's lifetime, which could influence their metabolic health (Fig. 3). The cumulative effects of distinct environmental stressors might be difficult to determine in a real world setting. Findings published in 2016 highlighted that epigenetic events can directly regulate key gene expression patterns that have a detrimental effect on metabolic health, with the discovery that Trim28 orchestrates an epigenetic switch to promote obesity by suppressing a cluster of genes with high expression levels in the hypothalamus and adipose tissue65. However, if reliable tools to track an individual's lifetime exposure to dietary insults are developed and utilized, coupled with biomarkers of distinct epigenetic signatures, the field might one day approach meaningful dissection of the individual variations in epigenetics and how these differences could affect treatment of metabolic disease. Before one can move to 'personalized epigenetics', the landscape of the epigenomic signature(s) of key cells or organs that control glucose homeostasis in healthy people versus those with insulin resistance or T2DM, as well as the effect of diet and exercise on the epigenome of tissues, must be determined.

Figure 3: Potential effect of environmentally induced epigenetic changes on gene expression.
Potential effect of environmentally induced epigenetic changes on gene expression.

In this hypothetical model, gene A is hypomethylated, while gene B is hypermethylated in response to environmental cues. Gene expression changes might not occur until a secondary environmental stress, or specific physiological state is altered, thereby enabling specific transcription factors to bind to the hypomethylated gene A, but not to hypermethylated gene B, which results in transcription of gene A. This model also explains the often observed discrepancy between DNA methylation and transcriptional changes. In this model, the same epigenetic modification in multiple tissues might only be functionally relevant in a given tissue if a specific transcription factor is also activated.

Epigenomic modifications in T2DM

Increasing evidence suggests that dynamic modifications in the epigenome can occur within an individual's lifetime and these changes might affect metabolic health. Although monozygotic twin pairs are considered genetically identical, phenotypical discordance exists, particularly in terms of the incidence of T2DM and cardiovascular disease66. Clinical and molecular studies of twin pairs are particularly informative in terms of teasing out the potential contribution of environmental factors to the manifestation of a variety of diseases.

One of the earliest studies advancing this notion comes from a survey of tissue-specific epigenomes in 3-year-old and 50-year-old monozygotic twin pairs67. Global and locus-specific differences in DNA methylation and histone acetylation were determined in lymphocytes from this cohort of twins67. Interestingly, age-dependent epigenomic differences were found. The older twin pairs were epigenetically more distinct than the younger twin pairs. The epigenomic differences and mRNA profiles were also greater between the older twin pairs that had different lifestyles and had spent less of their lifetime together, or had different health histories as determined by a questionnaire. The ageing-associated differences in the epigenomic profile were not altered between replicate samples from a subgroup of the older twin pairs taken over a short-term period (2–12 weeks), but seemed to be associated with the long-term exposure to environmental cues67. A secondary screen performed in epithelial mouth cells and intra-abdominal fat and skeletal muscle biopsy samples revealed marked epigenomic differences in the various tissues from the older monozygotic twin pairs that mirrored the profiles derived from lymphocytes. The fact that the DNA methylation and histone acetylation patterns diverge with ageing in multiple tissues from the monozygotic twin pairs indicates that age-related epigenetic drift might contribute to the divergence in the onset of ageing-related diseases68. Whether germ-line cells are more protected from epigenomic drift than somatic cells remains to be determined. Similar epigenomic changes have been observed in whole-blood or leukocytes from monozygotic twin pairs discordant for obesity69, insulin resistance70 or T2DM71.

In insulin-sensitive organs such as skeletal muscle and adipose tissue, epigenetic modifications might be important for the pathogenesis of T2DM, as the changes alter the profile of genes that control glucose and lipid metabolism. Although quantitatively the acquired modifications in genome-wide promoter DNA methylation in skeletal muscle or subcutaneous adipose tissue from twin pairs discordant for T2DM are small, distinct promoter-specific changes in key metabolic genes have been identified72. Promoters of several genes are differentially methylated in subcutaneous adipose tissue (ADCY5, CAV1, CIDEC, CDKN2A, CDKN2B, DUSP9, HNF4A, IDE, IRS1, KCNQ1, MTNR1B, TSPAN8 and WFS1) and skeletal muscle (CDKN2A, DUSP9, HNF4A, HHEX, KCNQ1, KLF11, PPARGC1A and SLC30A8) from the twin pairs discordant for T2DM72. Somewhat similar studies of adipose tissue of monozygotic twin pairs discordant for T2DM73 or obesity74 reveal changes in DNA methylation and expression of genes involved in oxidative phosphorylation, carbohydrate, amino acid and lipid metabolism, inflammation and extracellular matrix remodelling. These epigenetic changes might result from chemical modifications to the DNA or alternatively, from discrete shifts in the cellular composition of the tissues during ageing.

Epigenetics and T2DM: cause or effect?

Over the past decade, evidence has emerged to suggest that the epigenetic landscape is altered in multiple organs in patients with T2DM, including pancreatic islets75, 76, 77, 78, skeletal muscle79, 80, 81, adipose tissue82 and liver83, 84. Perhaps unsurprisingly, given the nature of the underlining pathophysiology of the disease, unbiased screens have identified differential methylation and expression of subsets of genes associated with a wide range of metabolic processes and mitochondrial functions in peripheral tissues of patients with T2DM79, 81. Moreover, epigenetic modifications and signatures of disturbed lipid oxidative metabolism and mitochondrial function have been observed in adipose tissue85, 86, 87, 88, skeletal muscle81 and liver84, 89 from individuals with obesity but not T2DM. Nevertheless, longitudinal studies tracking measures of insulin sensitivity and T2DM pathogenesis with genomic modifications in blood, pancreatic islets or peripheral tissues that control whole-body insulin sensitivity are lacking. Given the evidence of age-associated changes in the epigenomes of monozygotic twins67, these tissue-specific modifications might be acquired throughout an individual's lifetime, and, by an as yet unknown mechanism, lead to the development of islet-cell dysfunction, insulin resistance and the manifestation of T2DM. Several lines of independent evidence now point to an emerging view that distinct diabetes-associated epigenetic signatures characterize major organs responsible for whole body glucose and lipid homeostasis. Whether these changes constitute tissue-specific signatures or more global signatures of the pathogenesis of T2DM is unknown.

On the basis of the findings that changes in the epigenome with T2DM and obesity can be tissue-specific, the notion of epigenomic plasticity has emerged. Cell-based approaches have been useful to test the hypothesis that exposure to systemic factors associated with T2DM (that is, hyperinsulinaemia, hyperglycaemia and elevated levels of lipids or cytokines) impart environmental cues on the epigenome and whether such environmentally induced epigenetic marks can be reversed76, 80, 90, 91, 92, 93. For example, short-term (48 h) exposure of cultured human skeletal muscle cells to elevated levels of lipids or the cytokine tumour necrosis factor (TNF), but not insulin or glucose, induces promoter methylation of PPARGC1A79. The deleterious effects of lipids or TNF on PPARGC1A promoter methylation are prevented by gene silencing of the de novo DNA methyltransferase DNMT3B. These dynamic changes in the epigenome are not limited to skeletal muscle. Palmitate treatment (48 h) also alters both global and specific changes in the DNA methylation and expression profiles of human islets, with modifications to specific genes implicated in T2DM, including TCF7L2, GLIS3, HNF1B and SLC30A8 (Ref. 91). Collectively, these studies provide evidence that epigenetic modifications induced by lipids or cytokines might increase the risk of T2DM. However, these acute insults to the epigenome are not limited to increased levels of lipids or cytokines. In primary human vascular cells, genes and pathways associated with endothelial dysfunction are induced by hyperglycaemia via modulation of acetylated H3K9/K14 levels and CpG methylation changes93. Thus, a variety of cell-specific responses to systemic factors might affect the epigenome and gene expression signature.

Diet and exercise affect plasticity

The finding from cell-based studies that systemic factors associated with T2DM exert a degree of plasticity on the epigenome is compelling79, 91, 93. But perhaps more physiologically relevant is the observation that lifestyle modifications that influence insulin sensitivity, such as diet or exercise, affect the epigenome. A short-term high-fat diet (50% extra calories distributed as 60% fat, 32.5% carbohydrate and 7.5% protein for 5 days) in healthy men leads to widespread changes in the skeletal muscle epigenome, with multiple genes involved in metabolic regulation affected, including AKT2, PDX1/IPF1, SLC30A8, CDKN2A, CDKN2B and PPARG94. When the participants returned to the control diet, the changes induced by the high-fat diet in the skeletal muscle methylation signature were partly and nonsignificantly reversed after 6–8 weeks, which indicates that the time course for the demethylation process might be attenuated. Weight loss following bariatric surgery is also associated with dynamic epigenetic changes in human skeletal muscle81 and adipose tissue82, 95. In skeletal muscle, methylation and gene signatures for metabolic pathways that control lipid metabolism and mitochondria function are improved81. In adipose tissue, weight loss surgery modifies the expression of genes linked to adipogenesis95, as well as genes involved in insulin-mediated glucose uptake82. The heterogeneity of the response of select tissues to overnutrition and weight loss must be considered in future studies given that genetics or environmental exposures either in utero or throughout an individual's lifetime affect the epigenome.

Exercise your epigenome

The effects of diet or weight loss on the epigenome might not be entirely unexpected given the epidemiological evidence that nutrient availability influences metabolic health over several generations25, 26, 27, 28, 29, 30. Exercise also seems to modify the epigenome, which might constitute a mechanism to remodel skeletal muscle to support improvements in glucose and lipid metabolism96. Early studies to assess the effects of exercise on epigenetic responses addressed the involvement of histone deacetylases (HDAC) in regulating expression of genes involved in metabolic pathways through the transcription factor myocyte enhancer factor 2 (MEF2)97, 98, 99. Acute exercise increases signal transduction via mitogen-activated protein kinase (MAPK), 5′-AMP-activated protein kinase (AMPK) and calcium-calmodulin-dependent protein kinase II (CaMKII) signalling cascades97, 99. These signalling events are associated with phosphorylation and dissociation of specific HDAC isoforms from the transcription factor MEF2, thereby removing the transcriptional repressive function to drive exercise-responsive GLUT4 expression98, 99.

Exercise also remodels DNA methylation in skeletal muscle96. Acute exercise in healthy men and women decreases whole genome methylation in skeletal muscle, as well as intensity-dependent promoter hypomethylation of PPARGC1A, PDK4 and PPARD concomitant with increased mRNA levels of each respective gene96. Exercise-mediated changes in the epigenome are also observed in skeletal muscle from people at risk of developing T2DM100, 101. Interestingly, the calcium-releasing agent caffeine mimics the effects of exercise on both gene expression and promoter hypomethylation96, as well as hyperacetylation of HDAC102 in cultured cells, which implies that muscle contraction and calcium release are early triggers for these epigenetic changes with exercise. The effects of exercise on the epigenome are not limited to the working skeletal muscle. One study provides evidence that adipose tissue undergoes epigenetic remodelling in response to exercise training103. Thus, exercise might affect the epigenome of multiple organs to support the adaptive response to improve work capacity.

Exercise-induced changes in skeletal muscle miRNA profiles might also drive the adaptive response to acute or chronic training104, 105, 106, 107. Acute exercise alters the expression of a number of miRNAs that are enriched in skeletal muscle, which might modulate the abundance of proteins that regulate glucose and lipid metabolism by fine tuning the expression of target genes to ultimately improve insulin sensitivity. Beyond tissue-specific regulation, circulating miRNAs released from a variety of tissues including, but not limited to, skeletal muscle, adipose tissue and liver form the basis of a communication system by which gene expression signatures in one organ might be influenced by other exercise-responsive tissues108, 109, 110. The diversity in the different miRNAs released into the circulation in response to different modes and intensities of exercise or during the recovery period from exercise in trained versus untrained individuals might contribute to the diverse metabolic, cardiovascular and other whole-body adaptations that occur in conjunction with an acute exercise bout or from regular exercise training62.

In skeletal muscle, contraction induces phosphorylation of histone deacetylases, allowing relaxing of chromatin at regulatory regions of genes that respond to exercise111, 112. Methylation of DNA is also acutely changed, notably in the proximity of genes that control glucose and lipid metabolism96. Such epigenetic response to exercise, both after acute or chronic training, suggests that epigenetic factors have a role in the regulation of gene expression after exercise96, 99. Whether individual aerobic capacities or the adaptability to specific exercise training regimes will influence the distinct epigenetic response to exercise is unknown. Differences in the molecular fingerprint for the response to exercise training at the level of skeletal muscle could explain some of the intraindividual differences in the magnitude of exercise-induced gene expression. For example, in a previous study conducted in young men and women, we showed that the effect of acute exercise on DNA hypomethylation was not associated with the relative responsiveness to exercise or baseline aerobic capacity96. This observation was made in a fairly homogeneous group of individuals with similar levels of fitness. More investigations in various age groups and in individuals representing more extreme percentiles of aerobic capacity should, therefore, be conducted to determine if epigenetic mechanisms are involved in the plethora of responses to exercise in humans. Similarly, the effect of exercise on the epigenome of gametes could vary across individuals. If epigenetic factors limit the adaptive response to exercise training in some individuals, this might also affect the efficacy of exercise programs that aim to ameliorate insulin resistance in patients with T2DM or obesity in one individual or in the next generation.

Conclusion

Currently, no available treatments can effectively cure T2DM. However, lifestyle factors have a critical role in the prevention and treatment of T2DM by reducing the burden on the patient. Increasingly, it is becoming clear that our own environmental exposure, for good or for bad, also has an effect on the metabolic health of our children and grandchildren. The coming decades should bring some clarity to the molecular nature of this epigenetic inheritance. Although different diets and exercise training programs affect whole-body glucose metabolism and energy homeostasis, the epigenetic fingerprint in different stages of T2DM and the response to a variety of dietary and exercise regimes is unknown. By performing detailed epigenetic studies in both animal models and clinical cohorts and subsequently validating markers in cell-based systems, gene–environment networks underlying the pathogenesis of insulin resistance in T2DM and the adaptive response to different diet and exercise regimes might be revealed. Such information on the influence of environmental factors in metabolic health might also affect health care to reduce the risk of T2DM on several levels including prenatal care, early childhood and adolescence and throughout adulthood.

References

  1. World Health Organization. Global report on diabetes. http://www.who.int/diabetes/global-report/en/ (2006).
  2. Wilcox, G. Insulin and insulin resistance. Clin. Biochem. Rev. 26, 1939 (2005).
  3. Moller, D. E. & Kaufman, K. D. Metabolic syndrome: a clinical and molecular perspective. Annu. Rev. Med. 56, 4562 (2005).
  4. Franks, P. W., Pearson, E. & Florez, J. C. Gene–environment and gene–treatment interactions in type 2 diabetes: progress, pitfalls, and prospects. Diabetes Care 36, 14131421 (2013).
  5. Billings, L. K. & Florez, J. C. The genetics of type 2 diabetes: what have we learned from GWAS? Ann. NY Acad. Sci. 1212, 5977 (2010).
  6. Bonnefond, A. & Froguel, P. Rare and common genetic events in type 2 diabetes: what should biologists know? Cell Metab. 21, 357368 (2015).
  7. Prasad, R. B. & Groop, L. Genetics of type 2 diabetes-pitfalls and possibilities. Genes (Basel) 6, 87123 (2015).
  8. McCarthy, M. I. Genomic medicine at the heart of diabetes management. Diabetologia 58, 17251729 (2015).
  9. Groop, L. & Pociot, F. Genetics of diabetes — are we missing the genes or the disease? Mol. Cell. Endocrinol. 382, 726739 (2014).
  10. Kirchner, H., Osler, M. E., Krook, A. & Zierath, J. R. Epigenetic flexibility in metabolic regulation: disease cause and prevention? Trends Cell Biol. 23, 203209 (2013).
  11. Ling, C. & Groop, L. Epigenetics: a molecular link between environmental factors and type 2 diabetes. Diabetes 58, 27182725 (2009).
  12. Ong, T. P. & Ozanne, S. E. Developmental programming of type 2 diabetes: early nutrition and epigenetic mechanisms. Curr. Opin. Clin. Nutr. Metab. Care 18, 354360 (2015).
  13. Tracey, R., Manikkam, M., Guerrero-Bosagna, C. & Skinner, M. K. Hydrocarbons (jet fuel JP-8) induce epigenetic transgenerational inheritance of obesity, reproductive disease and sperm epimutations. Reprod. Toxicol. 36, 104116 (2013).
  14. Bird, A. Perceptions of epigenetics. Nature 447, 396398 (2007).
  15. Costa, F. F. Epigenomics in cancer management. Cancer Manag. Res. 2, 255265 (2010).
  16. Youngson, N. A. & Whitelaw, E. Transgenerational epigenetic effects. Annu. Rev. Genomics Hum. Genet. 9, 233257 (2008).
  17. Pembrey, M. E. et al. Sex-specific, male-line transgenerational responses in humans. Eur. J. Hum. Genet. 14, 159166 (2006).
  18. Fernandez-Twinn, D. S. et al. Maternal protein restriction leads to hyperinsulinemia and reduced insulin-signaling protein expression in 21-mo-old female rat offspring. Am. J. Physiol. Regul. Integr. Comp. Physiol. 288, R368R373 (2005).
  19. Hales, C. N. & Barker, D. J. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia 35, 595601 (1992).
  20. Hales, C. N. & Barker, D. J. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. 1992. Int. J. Epidemiol. 42, 12151222 (2013).
  21. Ozanne, S. E., Sandovici, I. & Constancia, M. Maternal diet, aging and diabetes meet at a chromatin loop. Aging (Albany NY) 3, 548554 (2011).
  22. Petry, C. J., Dorling, M. W., Pawlak, D. B., Ozanne, S. E. & Hales, C. N. Diabetes in old male offspring of rat dams fed a reduced protein diet. Int. J. Exp. Diabetes Res. 2, 139143 (2001).
  23. Sandovici, I. et al. Maternal diet and aging alter the epigenetic control of a promoter–enhancer interaction at the Hnf4a gene in rat pancreatic islets. Proc. Natl Acad. Sci. USA 108, 54495454 (2011).
  24. Dahri, S., Snoeck, A., Reusens-Billen, B., Remacle, C. & Hoet, J. J. Islet function in offspring of mothers on low-protein diet during gestation. Diabetes 40 (Suppl. 2), 115120 (1991).
  25. Ravelli, A. C. et al. Glucose tolerance in adults after prenatal exposure to famine. Lancet 351, 173177 (1998).
  26. Lumey, L. H. et al. Cohort profile: the Dutch Hunger Winter families study. Int. J. Epidemiol. 36, 11961204 (2007).
  27. Li, Y. et al. Exposure to the Chinese famine in early life and the risk of hyperglycemia and type 2 diabetes in adulthood. Diabetes 59, 24002406 (2010).
  28. Lumey, L. H., Khalangot, M. D. & Vaiserman, A. M. Association between type 2 diabetes and prenatal exposure to the Ukraine famine of 1932–33: a retrospective cohort study. Lancet Diabetes Endocrinol. 3, 787794 (2015).
  29. Thurner, S. et al. Quantification of excess risk for diabetes for those born in times of hunger, in an entire population of a nation, across a century. Proc. Natl Acad. Sci. USA 110, 47034707 (2013).
  30. Kaati, G., Bygren, L. O. & Edvinsson, S. Cardiovascular and diabetes mortality determined by nutrition during parents' and grandparents' slow growth period. Eur. J. Hum. Genet. 10, 682688 (2002).
  31. Faulk, C., Barks, A., Liu, K., Goodrich, J. M. & Dolinoy, D. C. Early-life lead exposure results in dose- and sex-specific effects on weight and epigenetic gene regulation in weanling mice. Epigenomics 5, 487500 (2013).
  32. de Castro Barbosa, T. et al. High-fat diet reprograms the epigenome of rat spermatozoa and transgenerationally affects metabolism of the offspring. Mol. Metab. 5, 184197 (2016).
  33. Ng, S. F. et al. Chronic high-fat diet in fathers programs β-cell dysfunction in female rat offspring. Nature 467, 963966 (2010).
  34. Warram, J. H., Krolewski, A. S., Gottlieb, M. S. & Kahn, C. R. Differences in risk of insulin-dependent diabetes in offspring of diabetic mothers and diabetic fathers. N. Engl. J. Med. 311, 149152 (1984).
  35. Lee, J. T. & Bartolomei, M. S. X-inactivation, imprinting, and long noncoding RNAs in health and disease. Cell 152, 13081323 (2013).
  36. Weaver, I. C. et al. Epigenetic programming by maternal behavior. Nat. Neurosci. 7, 847854 (2004).
  37. Drake, A. J. et al. Reduced adipose glucocorticoid reactivation and increased hepatic glucocorticoid clearance as an early adaptation to high-fat feeding in Wistar rats. Endocrinology 146, 913919 (2005).
  38. Hardikar, A. A. et al. Multigenerational undernutrition increases susceptibility to obesity and diabetes that is not reversed after dietary recuperation. Cell Metab. 22, 312319 (2015).
  39. Dunn, G. A. & Bale, T. L. Maternal high-fat diet promotes body length increases and insulin insensitivity in second-generation mice. Endocrinology 150, 49995009 (2009).
  40. Yokomizo, H. et al. Maternal high-fat diet induces insulin resistance and deterioration of pancreatic β-cell function in adult offspring with sex differences in mice. Am. J. Physiol. Endocrinol. Metab. 306, E1163E1175 (2014).
  41. Wei, Y. et al. Paternally induced transgenerational inheritance of susceptibility to diabetes in mammals. Proc. Natl Acad. Sci. USA 111, 18731878 (2014).
  42. Fullston, T. et al. 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. 27, 42264243 (2013).
  43. Buescher, J. L. et al. Evidence for transgenerational metabolic programming in Drosophila. Dis. Model. Mech. 6, 11231132 (2013).
  44. Carone, B. R. et al. Paternally induced transgenerational environmental reprogramming of metabolic gene expression in mammals. Cell 143, 10841096 (2010).
  45. Radford, E. J. et al. In utero undernourishment perturbs the adult sperm methylome and intergenerational metabolism. Science 345, 1255903 (2014).
  46. Skinner, M. K. et al. Ancestral dichlorodiphenyltrichloroethane (DDT) exposure promotes epigenetic transgenerational inheritance of obesity. BMC Med. 11, 228 (2013).
  47. Sanford, J. P., Clark, H. J., Chapman, V. M. & Rossant, J. Differences in DNA methylation during oogenesis and spermatogenesis and their persistence during early embryogenesis in the mouse. Genes Dev. 1, 10391046 (1987).
  48. Aiken, C. E., Tarry-Adkins, J. L. & Ozanne, S. E. Transgenerational developmental programming of ovarian reserve. Sci. Rep. 5, 16175 (2015).
  49. Terashima, M. et al. Effect of high fat diet on paternal sperm histone distribution and male offspring liver gene expression. Epigenetics 10, 861871 (2015).
  50. Egan, B. & Zierath, J. R. Exercise metabolism and the molecular regulation of skeletal muscle adaptation. Cell Metab. 17, 162184 (2013).
  51. Hawley, J. A., Hargreaves, M., Joyner, M. J. & Zierath, J. R. Integrative biology of exercise. Cell 159, 738749 (2014).
  52. [No authors listed.] Impact of physical activity during pregnancy and postpartum on chronic disease risk. Med. Sci. Sports Exerc. 38, 9891006 (2006).
  53. Mourtakos, S. P. et al. Maternal lifestyle characteristics during pregnancy, and the risk of obesity in the offspring: a study of 5,125 children. BMC Pregnancy Childbirth 15, 66 (2015).
  54. Laker, R. C. et al. Exercise prevents maternal high-fat diet-induced hypermethylation of the Pgc-1α gene and age-dependent metabolic dysfunction in the offspring. Diabetes 63, 16051611 (2014).
  55. Lin, J., Handschin, C. & Spiegelman, B. M. Metabolic control through the PGC-1 family of transcription coactivators. Cell Metab. 1, 361370 (2005).
  56. Stanford, K. I. et al. Exercise before and during pregnancy prevents the deleterious effects of maternal high-fat feeding on metabolic health of male offspring. Diabetes 64, 427433 (2015).
  57. Sheldon, R. D. et al. Gestational exercise protects adult male offspring from high-fat diet-induced hepatic steatosis. J. Hepatol. 64, 171178 (2016).
  58. Murashov, A. K. et al. Paternal long-term exercise programs offspring for low energy expenditure and increased risk for obesity in mice. FASEB J. 30, 775784 (2016).
  59. Carter, L. G., Qi, N. R., De Cabo, R. & Pearson, K. J. Maternal exercise improves insulin sensitivity in mature rat offspring. Med. Sci. Sports Exerc. 45, 832840 (2013).
  60. Guth, L. M. et al. Sex-specific effects of exercise ancestry on metabolic, morphological and gene expression phenotypes in multiple generations of mouse offspring. Exp. Physiol. 98, 14691484 (2013).
  61. McPherson, N. O., Owens, J. A., Fullston, T. & Lane, M. Preconception diet or exercise intervention in obese fathers normalizes sperm microRNA profile and metabolic syndrome in female offspring. Am. J. Physiol. Endocrinol. Metab. 308, E805E821 (2015).
  62. Zierath, J. R. & Wallberg-Henriksson, H. Looking ahead perspective: where will the future of exercise biology take us? Cell Metab. 22, 2530 (2015).
  63. Donkin, I. et al. Obesity and bariatric surgery drive epigenetic variation of spermatozoa in humans. Cell Metab. 23, 369378 (2016).
  64. Denham, J., O'Brien, B. J., Harvey, J. T. & Charchar, F. J. Genome-wide sperm DNA methylation changes after 3 months of exercise training in humans. Epigenomics 7, 115 (2015).
  65. Dalgaard, K. et al. Trim28 haploinsufficiency triggers bi-stable epigenetic obesity. Cell 164, 353364 (2016).
  66. Vaag, A. & Poulsen, P. Twins in metabolic and diabetes research: what do they tell us? Curr. Opin. Clin. Nutr. Metab. Care 10, 591596 (2007).
  67. Fraga, M. F. et al. Epigenetic differences arise during the lifetime of monozygotic twins. Proc. Natl Acad. Sci. USA 102, 1060410609 (2005).
  68. Martin, G. M. Epigenetic drift in aging identical twins. Proc. Natl Acad. Sci. USA 102, 1041310414 (2005).
  69. Ollikainen, M. et al. Genome-wide blood DNA methylation alterations at regulatory elements and heterochromatic regions in monozygotic twins discordant for obesity and liver fat. Clin. Epigenetics 7, 39 (2015).
  70. Zhao, J., Goldberg, J., Bremner, J. D. & Vaccarino, V. Global DNA methylation is associated with insulin resistance: a monozygotic twin study. Diabetes 61, 542546 (2012).
  71. Yuan, W. et al. An integrated epigenomic analysis for type 2 diabetes susceptibility loci in monozygotic twins. Nat. Commun. 5, 5719 (2014).
  72. Ribel-Madsen, R. et al. Genome-wide analysis of DNA methylation differences in muscle and fat from monozygotic twins discordant for type 2 diabetes. PLoS ONE 7, e51302 (2012).
  73. Nilsson, E. et al. Altered DNA methylation and differential expression of genes influencing metabolism and inflammation in adipose tissue from subjects with type 2 diabetes. Diabetes 63, 29622976 (2014).
  74. Pietilainen, K. H. et al. DNA methylation and gene expression patterns in adipose tissue differ significantly within young adult monozygotic BMI-discordant twin pairs. Int. J. Obes. (Lond.) 40, 654661 (2016).
  75. Dayeh, T. et al. Genome-wide DNA methylation analysis of human pancreatic islets from type 2 diabetic and non-diabetic donors identifies candidate genes that influence insulin secretion. PLoS Genet. 10, e1004160 (2014).
  76. Volkmar, M. et al. DNA methylation profiling identifies epigenetic dysregulation in pancreatic islets from type 2 diabetic patients. EMBO J. 31, 14051426 (2012).
  77. Yang, B. T. et al. Increased DNA methylation and decreased expression of PDX-1 in pancreatic islets from patients with type 2 diabetes. Mol. Endocrinol. 26, 12031212 (2012).
  78. Stitzel, M. L. et al. Global epigenomic analysis of primary human pancreatic islets provides insights into type 2 diabetes susceptibility loci. Cell Metab. 12, 443455 (2010).
  79. Barres, R. et al. Non-CpG methylation of the PGC-1α promoter through DNMT3B controls mitochondrial density. Cell Metab. 10, 189198 (2009).
  80. Kulkarni, S. S. et al. Mitochondrial regulators of fatty acid metabolism reflect metabolic dysfunction in type 2 diabetes mellitus. Metabolism 61, 175185 (2012).
  81. Barres, R. et al. Weight loss after gastric bypass surgery in human obesity remodels promoter methylation. Cell Rep. 3, 10201027 (2013).
  82. Multhaup, M. L. et al. Mouse-human experimental epigenetic analysis unmasks dietary targets and genetic liability for diabetic phenotypes. Cell Metab. 21, 138149 (2015).
  83. Nilsson, E. et al. Epigenetic alterations in human liver from subjects with type 2 diabetes in parallel with reduced folate levels. J. Clin. Endocrinol. Metab. 100, E1491E1501 (2015).
  84. Kirchner, H. et al. Altered DNA methylation of glycolytic and lipogenic genes in liver from obese and type 2 diabetic patients. Mol. Metab. 5, 171183 (2016).
  85. Xu, X. et al. A genome-wide methylation study on obesity: differential variability and differential methylation. Epigenetics 8, 522533 (2013).
  86. Agha, G. et al. Adiposity is associated with DNA methylation profile in adipose tissue. Int. J. Epidemiol. 44, 12771287 (2015).
  87. Guenard, F. et al. Differential methylation in visceral adipose tissue of obese men discordant for metabolic disturbances. Physiol. Genomics 46, 216222 (2014).
  88. Keller, M. et al. Global DNA methylation levels in human adipose tissue are related to fat distribution and glucose homeostasis. Diabetologia 57, 23742383 (2014).
  89. Horvath, S. et al. Obesity accelerates epigenetic aging of human liver. Proc. Natl Acad. Sci. USA 111, 1553815543 (2014).
  90. El-Osta, A. et al. Transient high glucose causes persistent epigenetic changes and altered gene expression during subsequent normoglycemia. J. Exp. Med. 205, 24092417 (2008).
  91. Hall, E. et al. Effects of palmitate on genome-wide mRNA expression and DNA methylation patterns in human pancreatic islets. BMC Med. 12, 103 (2014).
  92. Ishikawa, K. et al. Long-term pancreatic beta cell exposure to high levels of glucose but not palmitate induces DNA methylation within the insulin gene promoter and represses transcriptional activity. PLoS ONE 10, e0115350 (2015).
  93. Pirola, L. et al. Genome-wide analysis distinguishes hyperglycemia regulated epigenetic signatures of primary vascular cells. Genome Res. 21, 16011615 (2011).
  94. Jacobsen, S. C. et al. Effects of short-term high-fat overfeeding on genome-wide DNA methylation in the skeletal muscle of healthy young men. Diabetologia 55, 33413349 (2012).
  95. Dahlman, I. et al. The fat cell epigenetic signature in post-obese women is characterized by global hypomethylation and differential DNA methylation of adipogenesis genes. Int. J. Obes. (Lond.) 39, 910919 (2015).
  96. Barres, R. et al. Acute exercise remodels promoter methylation in human skeletal muscle. Cell Metab. 15, 405411 (2012).
  97. Yu, M. et al. Metabolic and mitogenic signal transduction in human skeletal muscle after intense cycling exercise. J. Physiol. 546, 327335 (2003).
  98. McGee, S. L. & Hargreaves, M. Exercise and myocyte enhancer factor 2 regulation in human skeletal muscle. Diabetes 53, 12081214 (2004).
  99. McGee, S. L., Fairlie, E., Garnham, A. P. & Hargreaves, M. Exercise-induced histone modifications in human skeletal muscle. J. Physiol. 587, 59515958 (2009).
  100. Nitert, M. D. et al. Impact of an exercise intervention on DNA methylation in skeletal muscle from first-degree relatives of patients with type 2 diabetes. Diabetes 61, 33223332 (2012).
  101. Rowlands, D. S. et al. Multi-omic integrated networks connect DNA methylation and miRNA with skeletal muscle plasticity to chronic exercise in type 2 diabetic obesity. Physiol. Genomics 46, 747765 (2014).
  102. Mukwevho, E. et al. Caffeine induces hyperacetylation of histones at the MEF2 site on the Glut4 promoter and increases MEF2A binding to the site via a CaMK-dependent mechanism. Am. J. Physiol. Endocrinol. Metab. 294, E582E588 (2008).
  103. Ronn, T. et al. A six months exercise intervention influences the genome-wide DNA methylation pattern in human adipose tissue. PLoS Genet. 9, e1003572 (2013).
  104. Aoi, W. et al. The microRNA miR-696 regulates PGC-1α in mouse skeletal muscle in response to physical activity. Am. J. Physiol. Endocrinol. Metab. 298, E799E806 (2010).
  105. Nielsen, S. et al. Muscle specific microRNAs are regulated by endurance exercise in human skeletal muscle. J. Physiol. 588, 40294037 (2010).
  106. Russell, A. P. et al. Regulation of miRNAs in human skeletal muscle following acute endurance exercise and short-term endurance training. J. Physiol. 591, 46374653 (2013).
  107. Tonevitsky, A. G. et al. Dynamically regulated miRNA–mRNA networks revealed by exercise. BMC Physiol. 13, 9 (2013).
  108. de Gonzalo-Calvo, D. et al. Circulating inflammatory miRNA signature in response to different doses of aerobic exercise. J. Appl. Physiol. (1985) 119, 124134 (2015).
  109. Parrizas, M. et al. Circulating miR-192 and miR-193b are markers of prediabetes and are modulated by an exercise intervention. J. Clin. Endocrinol. Metab. 100, E407E415 (2015).
  110. Wardle, S. L. et al. Plasma microRNA levels differ between endurance and strength athletes. PLoS ONE 10, e0122107 (2015).
  111. McKinsey, T. A., Zhang, C. L., Lu, J. & Olson, E. N. Signal-dependent nuclear export of a histone deacetylase regulates muscle differentiation. Nature 408, 106111 (2000).
  112. Smith, J. A., Kohn, T. A., Chetty, A. K. & Ojuka, E. O. CaMK activation during exercise is required for histone hyperacetylation and MEF2A binding at the MEF2 site on the Glut4 gene. Am. J. Physiol. Endocrinol. Metab. 295, E698E704 (2008).

Download references

Acknowledgements

The authors are supported by grants to R.B. and J.R.Z. from the Novo Nordisk Foundation, and to J.R.Z. from the Swedish Research Council, The European Research Council and the Strategic Program in Diabetes Research at Karolinska Institutet.

Author information

Affiliations

  1. The Novo Nordisk Foundation Center for Basic Metabolic Research, Section of Integrative Physiology, Faculty of Health and Medical Sciences, University of Copenhagen, Blegdamsvej 3B, Copenhagen 2200, Denmark.

    • Romain Barrès &
    • Juleen R. Zierath
  2. Department of Molecular Medicine and Department of Physiology and Pharmacology, Section of Integrative Physiology, Karolinska Institutet, von Eulers väg 4a, SE 171 77 Stockholm, Sweden.

    • Juleen R. Zierath

Contributions

Both authors researched data for the article, contributed to discussion of the content, wrote the article and reviewed and/or edited the article before submission.

Competing interests statement

The authors declare no competing interests.

Corresponding author

Correspondence to:

Author details

  • Romain Barrès

    Romain Barrès is Associate Professor at the Novo Nordisk Center for Basic Metabolic Research at the University of Copenhagen, Denmark. His research focuses on mechanisms by which environmental factors induce epigenetic modifications, thus predisposing or protecting from diabetes mellitus.

  • Juleen R. Zierath

    Juleen R. Zierath is a Professor at the Karolinska Institutet, Sweden, and the Novo Nordisk Center for Basic Metabolic Research at the University of Copenhagen, Denmark. Her research focuses on the pathogenesis of type 2 diabetes mellitus and the adaptive response of skeletal muscle to exercise.

Additional data