What is epigenetics?
As mentioned above, the word “epigenetics” means “above genetics” and generally refers to mechanisms that control gene expression without changing the DNA sequence. Epigenetic changes encompass molecular modifications to DNA and chromatin.12,13 The most extensively investigated modifications are DNA methylation and changes to chromatin packaging of DNA by post-translational histone modifications.12,13 Other epigenetic mechanisms include regulation by noncoding RNAs, such as microRNAs, and mechanisms that control the higher-level organization of chromatin within the nucleus, which also affects gene expression.31 Two of the most comprehensively studied epigenetically regulated phenomena in mammals are X-chromosome inactivation32 and genomic imprinting, a genetic mechanism that controls gene expression in a parent of origin–specific manner.16,33 Epigenetic regulation is furthermore involved in tissue-specific gene expression and silencing.34
Disorders involving dysfunction of imprinted genes are likely to have an epigenetic component. Individuals have two copies, or alleles, of most genes. One is inherited from the mother and one from the father. When both copies of a gene are active, the system exhibits redundancy and hence is less susceptible to dysfunction, but with imprinted genes, one copy is turned off epigenetically. These imprinted genes are susceptibility loci for disease because a single genetic event can alter their normal function.35 Epigenetic events can also cause or contribute to diseases if imprinted genes are not completely turned off.
Epigenetic modifications alter gene expression but not gene sequence. Therefore, characterizing the expression profiles of epigenetically controlled genes should reveal epigenetic biomarkers for disease, exposure, intervention, and efficacy. In turn, this should enable early diagnosis of individuals with a propensity for adult-onset disease and could lead to novel therapeutic approaches that could prevent and treat diseases before symptoms develop. This is in contrast with current human disease management, which mainly treats diseases after onset. In particular, modern nutrition focuses on disease prevention and health maintenance, and epigenetics may provide the means to understand and achieve these goals. Ultimately, comprehensive knowledge of the human epigenome is required, because the epigenome is not only tissue and stage-of-life dependent, it also varies markedly between individuals and species.
Epigenetics provides a biological format of long-term effects of environmental experiences, metabolic programming, and adaptive development of an organism.31 Epigenetic changes occur most commonly during gestation, neonatal development, puberty, and old age. During early embryogenesis, the mammalian genome is “cleansed” of most epigenetic marks, which are progressively re-established during embryonic development.36 The epigenome is therefore most vulnerable to environmentally induced alterations during embryogenesis, when DNA synthesis is rapid and the DNA methylation pattern and chromatin structure required for normal development is established. Once established, these epigenetic alterations are passed on to the daughter cells during somatic cell division. Epigenetic marks that are not completely erased during gametogenesis or are not well re-established during embryonic development can affect health not only in the present, but also in future generations.35 In addition, epigenetic marks may be transmitted across generations, either directly, by persisting through meiosis, or indirectly, through replication in the next generation of the conditions in which the epigenetic change occurred.36 Compelling evidence supports both the developmental origins of health and disease and the underlying epigenetic mechanisms.36
There is little doubt that environmentally (e.g., nutritionally) induced changes may significantly affect the ontology of individual organisms. Epigenetic research involving monozygotic twins has provided evidence for this hypothesis.37 Such studies typically provide correlations between phenotypic differences and differences in DNA-methylation patterns. Little is known about genes that are particularly prone to undergo environmentally influenced epigenetic changes and about the role of such changes in the generation of specific phenotypes. Consequently, analysis of environmentally or nutritionally induced epigenetic changes should primarily focus on the few “metastable” epialleles that have been characterized to date,31 on imprinted genes, and on genes known to be associated with a particular phenotype, such as obesity. Regarding the latter gene type, experiments should concentrate on the genes carrying CpG islands, because these are frequently methylated.
Environmental influences can alter gene expression and change phenotype, in part by modifying the epigenome. If these environmentally induced epigenetic modifications occur at crucial stages of life, they can potentially change behavior, disease susceptibility, and survival. Today's mechanistic evidence of environmental epigenetic imprinting or programming, by nutritional or other means, mainly derives from animal models, but human epidemiological studies also point increasingly toward associations between environmental effects – especially prenatal and early postnatal – and long-term epigenetic modifications manifesting in health and disease phenotypes.38,39 Results of animal and human studies support the hypothesis that there is a fetal basis or developmental origin of adult-onset disease; evolution of developmental plasticity, which enables an organism to adapt to environmental signals during early life, can also increase the risk of chronic diseases when there is a mismatch between the early perceived environment and the one encountered in adulthood. Developmental plasticity manifests when environmental exposure produces a broad range of adult phenotypes from a single genotype by epigenetically altering gene expression.40,41
Early nutrition, late-life consequences, and epigenetics
The fetal environment can influence an individual's likelihood of developing chronic disorders during adulthood. Early evidence of this appeared in research showing higher rates of cardiovascular disease in historical cohorts that had experienced high infant mortality.36 Further studies revealed an inverse relationship between birth weight and the development of hypertension, cardiovascular morbidity, insulin resistance, type 2 diabetes mellitus, hyperlipidemia, and obesity.36 It was therefore hypothesized that fetal metabolic programming under nutritionally adverse circumstances (i.e., those under which fetal growth is restricted in order to ensure brain development) may result in greater risk of chronic disorders later in life,42 although other data (e.g., from survivors of the Dutch “hunger winter” of 1944/1945) indicate that individuals exposed to adverse conditions in utero may subsequently exhibit adverse effects even if they do not have a low birth weight.38 This observation is consistent with the complex relationship between birth weight and risk of cardiovascular disease43 and with the observation that metabolic dysfunction correlates more strongly with neonatal adiposity and maternal nutrition than with birth weight.39,44
Other studies have focused on the role of excess nutrition during pregnancy and rapid weight gain in infants,40 the risk of which is greater after impaired fetal growth. Also, children who are exposed to hyperglycemia in utero or whose mothers are obese are at greater risk of developing metabolic disorders, especially type 2 diabetes mellitus.41 Most studies performed so far, however, have focused on the pathways through which undernutrition during development translates into subsequent metabolic disease, despite the fact that fetal nutrition depends not on maternal food intake but on maternal metabolism, cardiovascular function, and, particularly, placental function.36
Developing organisms seem to be particularly susceptible to epigenetic changes. The periconceptional period is particularly important, as shown by the sensitivity to suboptimal nutrition that occurs during this developmental stage45,46 in which widespread reprogramming of the epigenome occurs.47 Nutritional constraints later in pregnancy38,48 and/or postnatal overnutrition that leads to rapid growth,40,48 as well as maternal–fetal overnutrition,49 can also cause metabolic dysfunction later in life. For each of these scenarios, relevant epigenetic changes have been reported.50–52
The importance of the periconceptional period, during which suboptimal nutrition can have long-lasting effects without causing any change in birth weight,38,46 underscores the importance of healthy nutrition during the prepregnancy period. During this time, nutrition may frequently be unbalanced, even in women in high-income countries.53 The related finding that the balance of micronutrients involved in one-carbon metabolism during pregnancy affects the subsequent metabolic health of the offspring54 has implications for the design of nutritional supplementation programs. Research in animal models has demonstrated that nutritional interventions (such as folate supplementation) or endocrinological interventions (such as neonatal leptin administration) can prevent55 or reverse52,56 the adverse effects of impaired early-life nutrition and the associated epigenetic changes.
Maternal overnutrition may contribute to trans-generational amplification of obesity in humans.57 In the Agouti mouse model, dietary supplementation with substrates and cofactors of DNA methylation abrogate such female-line transmission of obesity.58 Also, early treatment with a histone deacetylase inhibitor, a drug class for clinical oncological trials, can reverse the phenotypic and epigenetic consequences of intrauterine growth retardation for pancreatic development and adult-onset diabetes mellitus.59 Offspring of rats given a low-protein diet during pregnancy have a different hepatic gene expression, gene promoter methylation, and histone acetylation of metabolically relevant receptors: the glucocorticoid receptor and the peroxisome proliferator-activated receptor α. Later in life, the offspring develop abnormalities of metabolic control and cardiovascular function. Concurrently supplementing the diet of the pregnant dam with folate, which promotes methyl group provision, prevents these effects.55 Studies in pregnant rodents subject to a variety of dietary challenges or to exposure to glucocorticoids show a relatively consistent outcome for the offspring, i.e., abnormalities of insulin secretion and sensitivity, sarcopenia, appetite disturbance, obesity, and endothelial dysfunction.60 These features are particularly apparent when the animals are placed on a high-fat diet after weaning.48
Primary candidates for genes holding epigenetic memories of early-life experiences are those that are directly associated with energy acquisition, storage, and use. Leptin is thought to be one of the best gene candidates involved in the development of obesity, because it encodes a hormone that regulates energy intake and expenditure.61 Epigenetic variants of leptin could possibly explain low plasma concentrations of this fat hormone. The promoter region of leptin is methylated in somatic tissues of humans and mice and displays epigenetic variation.62–64 It is speculated that leptin is sensitive to environmental cues and can acquire a thrifty epigenotype. Other promising candidates are genes listed in the human obesity gene map,65 imprinted genes, and genes close to or disrupted by transposable elements.
Genes, nutrition, and the brain
A great deal of research has focused on the respective binary research interfaces of genomics, cognitive science, and nutrition research, but little is known about the intersection of all three disciplines. For example, how does nutrition affect the brain at the whole genome level? And how does an individual's genetic make-up determine their nutritional response with consequences for cognitive performance and for predisposition toward cognitive decline and disease?
A substantial body of knowledge has already been gathered regarding the interface between nutrition and genes; the fields of nutrigenomics (e.g., how the body responds to diet at the genomic scale)66 and nutrigenetics67 (e.g., how individuals are predisposed to respond to food in a certain way and not in another) are rapidly expanding this knowledge.68 The nutrigenetic success story of G-protein polymorphisms was discussed in the preceding section.
Competence is also improving in the interface between nutrition and the brain; that is, how diet affects neurodevelopment,69 neurodecline,70 and cognitive performance.71 However, such studies are typically epidemiological, observational, or associative.
Furthermore, knowledge is accumulating regarding the interface of the brain and the genes (e.g., how individuals are predisposed to the development of cognitive decline72,73 and disease74 and to respond to nutritional intervention aimed at improving neurodevelopment and cognition as well as preventing neurodecline75). In recent pioneering studies, initial understanding of how epigenetic and genetic mechanisms may influence cognitive function76 and neuronal disease has been gained.77 Kalueff78 reviewed the neurobiology of memory and anxiety with the ultimate goal of making the leap from genes to behavior; he summarized experimental data from neurogenetics, neurochemistry, behavioral pharmacology, and clinical findings reflecting mechanisms of the interplay between memory and anxiety, including brain neurochemistry and circuitry, pharmacology, neuroplasticity, genetics, and interactions between genes and environment.78 Burmeister et al.79 recently updated the science community on progress and controversy in psychiatric genetics; although several psychiatric diseases such as bipolar disorder, schizophrenia, and autism are highly heritable, identifying their genetic basis and replicating the findings are challenging. Nevertheless, the incorporation of intermediate traits (endophenotypes) and environmental factors into genetic analyses and the identification of rare inherited variants and novel structural mutations have advanced the field. The same is expected from larger sample sizes and meta-analyses of such studies and from paying more attention to unique families and rare variants.80
The Papassotiropoulos group has shed more light on the genetic basis of cognitive performance and decline; as a methodological recommendation, they suggest identifying the genetic basis for complex disorders using pooling-based genome-wide SNP association studies as a low-cost alternative to individually genotyping thousands of samples on high-density SNP microarrays.81 They also reviewed the application of genetics, transcriptomics, and proteomics to Alzheimer's disease (AD),81 discussing, for example, rare mutations of APP and presenilin 1/2 and their importance for amyloid metabolism81 and apolipoprotein E (APOE) as an important susceptibility gene for late-onset AD.82 The same group, together with others, has also published a number of research articles on associations between genes and performance, with a focus on memory; except for a genome survey of human short-term memory,83 memory performance was concretely associated with genetic variability in a gene locus encoding the brain protein KIBRA84; APOE, the major genetic AD risk factor85 (with the APOE ε4 allele associated with better episodic memory than the APOE ε2 and ε3 alleles in young, healthy persons); a gene cluster encoding proteins of a signalling cascade encompassing the N-methyl-D-aspartic acid and metabotropic glutamate receptors, adenylyl cyclase, Ca 2+/calmodulin-dependent protein kinase II, protein kinase (pK) A, and pKC86; and the calmodulin-binding transcription activator 1 gene87 (which influences episodic memory). In an extension of the genetic determination of memory, de Quervain et al.88 assessed emotional memory and identified an association with a deletion variant of the α2b-adrenoceptor, which is related to differential amygdala activation during encoding of emotional memories.89 Payton et al.90 assessed the influence of serotonin transporter gene polymorphisms on cognitive decline and cognitive abilities in older adults without dementia, focusing on two functional SNPs within the SLC6A4 and VNTR2 genes. They found subjects homozygous for the VNTR2 12 allele to decline more rapidly on all cognitive tests, with significance reached for tests of fluid intelligence, semantic memory, and general cognitive ability. VNTR2 is only the second gene reported to regulate the rate of cognitive decline in individuals without dementia, and it may become a target for the treatment of cognitive impairment in older adults.90
Although these pioneering studies point toward genetic susceptibility loci, replication of such findings in genome-wide association studies (GWAS) has been challenging, as with most complex disorders and phenotypes; Need et al.91 performed a GWAS of 11 cognitive phenotypes from the Cambridge Neuropsychological Test Automated Battery. They showed these measures to be heritable by comparing the correlation in 100 monozygotic and 100 dizygotic twin pairs; the full battery was tested in approximately 750 subjects, and a further 500 individuals were investigated for spatial and verbal recognition memory to search for smaller genetic effects. Unfortunately, the authors were not able to establish any significant genome-wide associations with SNPs or copy number variants, nor could they replicate any polymorphism previously associated with cognition. The group also investigated – as a separate set – SNPs in genomic loci that had been shown to harbor rare variants associated with neuropsychiatric disorders. Only NRXN1 showed evidence of significant association with cognition. Need et al.91 concluded that common genetic variation has little influence on cognition in healthy subjects and that cognitive measures do not represent a more-tractable genetic trait than clinical endpoints.
In summary, we are learning how gene variants may influence cognitive performance on the one hand and the outcome of nutritional interventions on the other hand. In contrast, the intersection of genomics, cognitive science, and nutrition research, that is, how nutrition affects the brain genome-wide and how (epi)genetics may determine an individual's nutritional response with consequences for cognitive performance and decline, remains largely uncharted terrain. This is especially true for human studies, and one reason is the difficult sampling; genome-wide analyses of mechanisms of neuronal decline in the brain are limited to animal studies or to postmortem brain biopsies of affected subjects. In the search for early biomarkers, cerebrospinal fluid (the liquid surrounding the brain) may offer an alternative,92 and cerebrospinal fluid studies have identified candidate protein biomarkers for brain diseases.93 GWAS have revealed (candidate) genetic risk loci for diseases including AD and Parkinson's disease.74 To our knowledge (epi)genetic predisposition to cognition-promoting nutritional interventions has not been investigated.