Article is available in full to IFST members and subscribers.

Register on the FST Journal website for free

Click the button to register to FST Journal online for free and gain access to the latest news


If you are an IFST member, please login through the Members Area of the IFST website. 













Nutrition, epigenetics and health

Nutrition, epigenetics and health

John C. Mathers of Newcastle University explains how gene expression is regulated in the body and discusses how diet may induce heritable changes to the genome.

Epigenetics – the basics

Every nucleated cell in the human body contains approximately three billion base pairs (bp) of DNA and the DNA in each of these cells is exactly the same. The particular sequence of bases in the DNA for any individual was determined at conception and, except for monozygotic (identical) twins, each of us is genetically unique. From that fertilised egg onwards, at each cell division, our unique DNA sequence was copied faithfully to the next generation of cells. This means that each of our cells has the potential (the DNA blueprint) to make any of our proteins and to carry out any of our cellular functions. All cells express some genes in common i.e. the so-called ‘house-keeping’ genes that encode the proteins that are needed to maintain the basic functions of all cells e.g. transporting in oxygen and nutrients, generating ATP to fuel the work of the cell and exporting waste products. However, a liver cell is very different from a bone cell and since the work done by each cell type differs, there are characteristically different patterns of genes expressed in different cell types. In addition, during embryonic and fetal development, some genes are required for developmental functions that are not needed later in post-natal life. This means that some parts of the genetic code are ‘redundant’ in adult cells and there need to be mechanisms to control (or regulate) how the information in DNA is used to make the complement of proteins required by each particular cell in each particular circumstance. That regulation is provided, in part, by the cell’s epigenetic machinery. Epigenetics describes heritable changes to the genome without changes to the DNA sequence per se[1].

In cancers and many other diseases, abnormal patterns of gene expression are associated with methylation of the promoter regions of specific genes leading to loss of function of the corresponding gene.

Epigenetic marks and molecules

The genetic code written in DNA is relatively simple. A unique linear sequence of adenine (A), guanine (G), thymine (T) and cytosine (C) bases provides the genetic blueprint for each individual. In contrast, regulation of the genome is necessarily complex to ensure fidelity of that unique DNA sequence and to ensure that the information in DNA is used in ways that are appropriate for the specific cell or tissue and any particular circumstance. For example, just after a meal, when we are absorbing nutrients that are required for immediate use, we need to switch on genes that are involved in nutrient storage e.g. for glycogen synthesis in liver and muscle. In addition, as noted above, specific cells, e.g. immune cells, neurones and muscle cells, carry out unique functions which mean that characteristically different patterns of genes are ’switched on’ (expressed) in each cell whilst others are ‘switched off’ (silenced). The epigenetic machinery that contributes to this regulation includes a multi-layered system of chemical ’marks’ and molecules together with a consortium of proteins described as ‘readers, writers and erasers’ that uses this epigenetic information[2].

i) Marks on DNA

Some cytosine (C) residues within DNA are modified by the addition of a methyl (CH3) group to the 5’ position to generate 5-methylcytosine (5mC). This occurs normally when the C is followed by a guanine (G) in a so-called CpG dinucleotide. The pattern of methylation marks on DNA is dynamic with methyl groups being added by a family of DNA methyl transferase enzymes (DNMT, an epigenetic ‘writer’) and being removed by a series of reactions involving a group of enzymes known as ten-eleven translocation methylcytosine dioxygenases (TET, an epigenetic ‘eraser’). During the latter process, 5-hydroxymethylcytosine (5hmC) is produced as an intermediate (1). Recently, it has been discovered that in some body tissues, particularly the brain and embryonic stem cells, there is a considerable density of 5hmC as well as 5mC. Although the role of 5hmC is not yet well-understood, the role of 5mC in silencing genes has been established for many years. For example, about 40% of the human genome consists of repetitive elements, such as retrotransposons, which make up the large part of heterochromatin. Retrotransposons have been acquired after exposure to viruses over tens of millions of years and are potentially harmful. However, these areas of the genome are heavily methylated and ‘silenced’ so that the potentially damaging repetitive elements cannot be expressed. In addition, in cancers and many other diseases, abnormal patterns of gene expression are associated with methylation of the promoter regions of specific genes leading to loss of function of the corresponding gene[2].

ii) Marks on histones

Within the cell nucleus, DNA is wrapped around bundles of eight globular proteins called histones in a form of ‘smart packaging’, which regulates access to DNA for key cellular processes. Specific amino acid residues in the N-terminal tails of histones are ‘decorated’ by the addition of multiple small chemical groups including methyl, acetyl, phosphate and ubiquitin. These post-translational modifications appear to form a complex information system known as the ’histone code’[3]. The histone code hypothesis suggests that the specific patterns of these histone modifications recruit other proteins (epigenetic ‘readers’) which alter chromatin structure or enable gene transcription. Knowledge of the biological meaning of the various histone modifications is still rather sketchy although the role of a few of these marks has been elucidated. For example, chromosomal condensation that occurs in mitosis and meiosis is associated with phosphorylation of serine residues 10 and 28 on histone H3. Similarly, methylation of lysine residues at positions 4, 36 and 79 on histone H3 is associated with active transcription, whereas methylation marks on lysines at positions 9 and 27 on histone H3 and lysine 20 on histone H4 are associated with gene silencing. Importantly, it seems that DNA methylation and histone ‘decoration’ work together to regulate DNA[2].

iii) Non-coding RNAs

Less than 20 years ago, a new large family of RNA species was discovered that is now called non-coding RNAs (ncRNA) because these RNAs carry out their biological functions directly as RNA and do not need to be translated into proteins. From an epigenetic perspective, the best characterised ncRNA are the microRNA (miRNA) which are typically ≈22bp long. The human genome encodes >1000 miRNA which regulate about 60% of all of our genes. Binding of particular miRNA in a sequence-specific manner to the 3’ untranslated region (UTR), to the coding sequence or to the 5’ UTR of the target messenger RNA (mRNA) inhibits translation or causes mRNA degradation. Again, there is coordination between types of epigenetic regulator and transcription of many miRNA is regulated by DNA methylation[2].

iv) The epitranscriptome

In the last few years, there has been an explosion of information about a constellation of post-transcriptional modifications, e.g. addition of methyl groups, on RNA species now known as the epitranscriptome[3]. This is an emerging field and much remains to be discovered but it appears that the chemically modified nucleosides in RNA play an important role in RNA homeostasis by regulating the amounts and activities of particular RNA species[4]. So far, little is known about whether, or how, environmental factors, such as diet, influence the epitranscriptomic landscape[1] but a potential link with obesity is discussed below.

Impact of dietary factors on the epigenome

The honey bee provides a striking example of the importance of diet in influencing the epigenome and subsequent effects on phenotype. Genetically identical female larvae are fed initially with ‘royal jelly’ but then those larvae destined to become worker bees are switched to a combination of pollen and nectar. In contrast, the tiny minority of bee larvae destined to become new queens continue to be fed royal jelly. This dietary difference results in two very different phenotypes – worker bees that do not reproduce and that have a short lifespan (a few weeks in summer) and the much larger queen bees, which can lay up to 2000 eggs per day and may live for three-four years (Figure 1). Several components in royal jelly are epigenetically active and it is becoming apparent that effects on a particular DNA methyl transferase (DNMT3) that is involved in ’writing’ new methylation marks on DNA (de novo DNA methylation) may be central to the different reproductive potential of queen bees compared with worker bees that is induced by royal jelly[2].

Figure 1 The contrasting phenotypes of queen and worker honey bees results from differences in nutrition during larval growth and development altering epigenetic trajectories

Humans and other mammals do not exhibit the dramatic phenotypic plasticity that is evident in honey bees but there is growing evidence that a large number of components of human foods impact on epigenetic marks and molecules. For obvious reasons, most of the experimental work that provides evidence of causal relationships between food components and the epigenome has been carried out in mice or other model organisms. However, there is considerable empirical evidence that nutrients and other food-derived substances also influence the human epigenome. The molecular mechanisms underlying links between dietary factors and the epigenome are not yet well understood but include the role of dietary factors in i) providing substrates e.g. methyl and acetyl groups for marking DNA, RNA and histones and ii) affecting the activity of the epigenetic ’readers, writers and erasers’ i.e. the proteins that are responsible for creating and using epigenetic information[5]. Folate, betaine and choline that contribute methyl groups for one-carbon metabolism influence cellular availability of S-adenosylmethionine (SAM), which is the universal donor used in methylating DNA, RNA and proteins, such as histones. The short-chain fatty acid butyrate, which is produced in large amounts in the large bowel from fermentation of fibre, was one of the earliest nutrients shown to have epigenetic effects. At physiological concentrations, butyrate is a potent inhibitor of histone deacetylase (HDAC) enzymes (which are epigenetic ‘erasers’) and, consequently, has widespread effects on gene expression. This action of butyrate is believed to be one of the mechanisms through which higher fibre intake protects against colorectal cancer and other large bowel diseases. Butyrate is also present in milk and milk products, including parmesan cheese[3].

Many plant secondary compounds (phytochemicals) that we value in foods because of their colours, flavours and health-promoting effects influence epigenetic marks and molecules. These effects occur largely though influences on the activities of epigenetic ‘readers, writers and erasers’ and some examples are given in Table 1.

Although the biological information in each individual’s genetic make-up (their genotype) sets boundaries for their individual phenotype, genetics is not destiny. Each person’s phenotype, at any particular time, is plastic and is a consequence of the cumulative effects of interactions between their genotype and environmental factors, including diet, over the life-course. The types of evidence discussed above supports the idea that epigenetics has a causal role in mediating the effects of interactions between diet and genetics in determining phenotype. This central role for the epigenome (illustrated in Figure 2) allows each individual (and, indeed, individual cells within each person) to respond to their nutritional environment and to ensure that an appropriate constellation of genes is switched on (and switched off) to help to main homeostasis and functional integrity[3].

Figure 2 The epigenome is central to interactions between diet and the genome in determining phenotype
(Mathers, 2017)


Table 1 Examples of epigenetically active compounds from plant foods

Early life nutrition, epigenetics and health trajectories

Ground-breaking epidemiological studies by David Barker and his colleagues, supported by experimental studies in model organisms, have provided firm evidence for the idea that nutrition (and other environmental exposures) in early life cast a long shadow on health across the life-course. These findings gave rise to the Developmental Origins of Health and Disease (DOHaD) hypothesis, which posits that the early life environment may alter ‘programming’ of the phenotype with potential to influence the risk of chronic diseases from childhood to old age. Whilst several potential mechanisms to explain this lifelong ‘memory’ of early life events and exposures have been proposed, including abnormal development of organs and tissues, epigenetic mechanisms appear to be particularly important because they provide a well-recognised pathway for embedding durable evidence of such exposures, which persists across the multiple cell generations. Indeed, epigenetic marks and molecules may be especially sensitive to environmental exposures during very early development since this is a life-stage in which genome-wide patterns of DNA methylation undergo radical change[3].

Much DOHaD-related research to date has addressed maternal nutrition during pregnancy and, to a more limited extent, early post-natal nutrition but now the focus is shifting to earlier stages and to include investigation of the effects of pre-and peri-conceptual nutrition in both (potential) mothers and fathers[6]. In addition to the biological plausibility that nutritional status and exposure at this life-stage will influence epigenetic marks and molecules, targeting nutritional advice at potential (or aspiring) parents may be especially effective since it may increase the likelihood that children are conceived under more optimal nutritional (and epigenetic) conditions.


We are familiar with the ways in which excess body fat increases risk of multiple complex diseases through endothelial dysfunction, oxidative stress, insulin resistance, dyslipidaemia and systemic inflammation. In addition, there is now strong evidence that the epigenome is also disrupted in obese individuals. For example, an epigenome-wide association study has revealed multiple changes in DNA methylation associated with body mass index (BMI)[7]. Obesity is a risk factor for many age-related diseases and reduces lifespan and there is now evidence that obesity is associated with acceleration of the ‘epigenetic clock’, a panel of CpG sites at which methylation changes track with ageing[8]. Very recently, a comparison of DNA methylation in obese individuals and in those with bowel cancer has shown hypermethylated CpG islands (clusters of CpG sites which occur, often, in the promoter regions of genes), which may account for the epigenetic instability that drives cancer initiation in obesity[9].

In humans, the most common epitranscriptomic mark is N6- methyladenosine (m6A) in which a methyl group is attached to the nitrogen (N) at position 6 in the adenosine base within mRNA. This provides a potential link with nutrition because the fat mass and obesity associated protein (FTO) is a demethylase (an epigenetic ‘eraser’) that converts m6A back to adenosine. The FTO gene is polymorphic and people carrying the unusual variant of the FTO gene (those having an A rather than a T at rs9939609) are heavier and more likely to become obese. The mechanism linking FTO genotype with excess weight is poorly understood but may relate to perceived responses to food since the FTO genotype regulates expression of ghrelin, the hunger hormone[10]. This provides a potential link between m6A and behavioural responses to food that predispose to increased energy intake and obesity[10].

Future perspectives

Research on understanding the effects of diet, and its interactions with genotype, on the epigenome and the consequences for health and well-being is moving fast. Advances in the field are being accelerated by better tools for investigating epigenetic marks and molecules. For example, commercial epigenetic arrays make it possible to interrogate the methylation status of >800,000 individual CpG sites simultaneously using DNA extracted from small amounts of human blood or tissue. In addition, next generation sequencing provides in-depth analysis of patterns of miRNA expression and of DNA methylation whilst ultra-high performance liquid chromatography (UHPLC) coupled with tandem mass spectrometry (MS/MS) can provide quantitative profiles of RNA modifications (the epitranscriptome). However, the complexity of histone ‘decorations’ is more challenging so that an integrated analysis of all components of the epigenetic machinery is unlikely to be routine for some time. Nevertheless, this novel research promises to deliver new insights into links between diet and health and is an attractive area in which ambitious young researchers can build their careers.

John C. Mathers

Human Nutrition Research Centre, Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, NE2 4HH

email john.mathers@


Acknowledgement My research is supported by the Medical Research Council through the Centre for Ageing and Vitality (award number MR/ L016354/1).


1. Malcomson FC & Mathers JC (2017) Nutr. Bull. 42, 254-265. “Nutrition, epigenetics and health through life”.

2. Ideraabdullah FY & Zeisel SH (2018) Physiol. Rev. 98, 667-695. “Dietary modulation of the epigeneome”.

3. Mathers JC (2017) “Diet and epigenetics.” In: Human Nutrition (C. Geissler, HJ Powers, editors), pp.579-592. Oxford University Press, Oxford, UK.

4. Esteller M & Pandolfi PP (2017) Cancer Discovery 7, 359-368. “”The epitranscriptome of noncoding RNAs in cancer”.

5. Kadayifci FZ, Zheng S & Pan Y-X (2018) Int. J. Mol. Sci. 19, 4055. “Molecular mechanisms underlying the link between diet and DNA methylation”.

6. Fleming TP, Watkins AJ et al. (2018) Lancet 391, 1842-1852. “Origins of lifetime health around the time of conception: causes and consequences”.

7. Wahl S, Drong A et al. (2017) Nature 541, 81-86. “Epigenome-wide association study of body mass index, and the adverse outcomes of adiposity”.

8. Quach A, Levine ME et al. (2017) Ageing (Albany NY) 9, 419-446. “Epigenetic clock analysis of diet, exercise, education, and lifestyle factors”.

9. Dong L, Ma L et al. (2019) Sci. Reports 9, 5100. “Genome-wide analysis reveals DNA methylation alterations in obesity associated with high risk of colorectal cancer”.

10. Karra E, O’Daly OG et al. (2013) J. Clin. Invest. 123, 3539-3551. “A link between FTO, ghrelin, and impaired brain food-cue responsivity”.





New FS and T website!

We encourage you to visit the new home for our Food Science and Technology quarterly magazine. This now sits as part of Wiley's Online Library. There, you can access all our past issues from 2017 up to our current issue. Access the new website here: 

View the latest digital issue of FS&T or browse the archive


Click here

Become a member of the Institute of Food Science and Technology