Individuality and epigenetics in obesity

Authors


Prof. J Alfredo Martínez, Department of Food Sciences, Physiology and Toxicology University of Navarra C/Irunlarrea s/n 31008 Pamplona Spain Phone: 0034 948 425600 Fax: 0034 948 425649. E-mail: jalfmtz@unav.es

Summary

Excessive weight gain arises from the interactions among environmental factors, genetic predisposition and the individual behavior. However, it is becoming evident that interindividual differences in obesity susceptibility depend also on epigenetic factors. Epigenetics studies the heritable changes in gene expression that do not involve changes to the underlying DNA sequence. These processes include DNA methylation, covalent histone modifications, chromatin folding and, more recently described, the regulatory action of miRNAs and polycomb group complexes. In this review, we focus on experimental evidences concerning dietary factors influencing obesity development by epigenetic mechanisms, reporting treatment doses and durations. Moreover, we present a bioinformatic analysis of promoter regions for the search of future epigenetic biomarkers of obesity, including methylation pattern analyses of several obesity-related genes (epiobesigenes), such as FGF2, PTEN, CDKN1A and ESR1, implicated in adipogenesis, SOCS1/SOCS3, in inflammation, and COX7A1 LPL, CAV1, and IGFBP3, in intermediate metabolism and insulin signalling. The identification of those individuals that at an early age could present changes in the methylation profiles of specific genes could help to predict their susceptibility to later develop obesity, which may allow to prevent and follow-up its progress, as well as to research and develop newer therapeutic approaches.

Introduction

The aetiology of obesity is multifactorial, involving complex interactions among the genetic background, hormones and different unhealthy environmental factors such as sedentarism or inadequate dietary habits (1). Up to now, obesity has been thoroughly studied from a thermoenergetic/lifestyle point of view, taking into consideration the balance between energy intake and expenditure. Lately, the genetic approach considered obesity in some circumstances as an inherited disease caused by gene mutations and polymorphisms, beginning with the thrifty gene hypothesis and keeping on with the study of candidate genes (2).

Among the different mechanisms that could lead to inter-individual differences in obesity, the epigenetic regulation of the gene expression has emerged in the last years as a potentially important contributor. Thus, epigenetics has been defined as the study of heritable changes in gene expression that occur in the absence of a change in the DNA sequence itself (3). These changes include DNA methylation, covalent modifications to histones, packaging of DNA around nucleosomes, chromatin folding and chromatin attachment to the nuclear matrix (4). More recently, the involvement of epigenetic mechanisms such as polycomb group (PcG) complexes and microRNAs (miRNAs) has been reported in mammals (5,6). The epigenetic state of the DNA and the associated phenotype can sometimes be inherited in what is called the transgenerational epigenetic inheritance (7). This process is explained because epigenetic modifications are not completely erased during gametogenesis and early embryogenesis, resulting in some memory of the epigenetic state persisting and being transferred to the next generation (8,9).

An adverse environment during in utero or lactation periods has been involved in the future development of obesity, suggesting that the mother's nutrition or perinatal lifestyle choices could alter the developmental programming of the foetus and pups (10). However, the role of nutrition during adulthood in the modification of the epigenetic pattern and the possible transmission through the gametes is a matter of debate (11). Thus, changes in DNA methylation patterns could be a result of the interplay of various dietary and environmental factors and also could be a source of inter-individual differences with respect to the susceptibility to develop obesity or other metabolic diseases.

Epigenetic mechanisms: when and where DNA methylation pattern is changing

Chromatin-remodelling mechanisms include DNA methylation, histone-tail acetylation, poly-ADP-ribosylation and ATP-dependent chromatin-remodelling processes (4). Methylation of CpG islands, defined as genomic regions that contain a high frequency of Cytosine-Guanine (CG) dinucleotides, results in the conversion of the cytosine to 5-methylcytosine and, in promoter regions, is often associated with gene silencing (12). There are three major enzymes involved in establishing and maintaining DNA methylation patterns: DNA methyltransferases (DMNT) 3A and 3B are de novo methyltransferases; DNMT1 ensures that methylation patterns are copied faithfully throughout each cell division (13). These enzymes cooperate with each other to establish, maintain the cellular DNA methylation patterns and also interact with histone deacetylases, histone methyltransferases and methyl-cytosine-binding proteins in a complex regulatory network. Regarding histone modifications, acetylation is associated with active gene transcription, whereas methylation of histone H3 lysine 9 (H3K9) is an indicator of condensed and inactive chromatin (12).

Other recently described mechanisms, as PcG complexes and miRNAs, could help regulate gene expression by epigenetic mechanisms. Thus, PcG complexes are a family of proteins that can remodel chromatin (e.g. histone H3K27 methylation or H2AK119 ubiquitylation) maintaining epigenetically repressed states by preventing transcription factors to bind to promoter sequences in DNA (5). On the other hand, miRNAs are non-protein-coding RNA genes that mature into 22-nucleotide-long sequence-specific gene regulators with a role in controlling gene expression (14). These miRNAs are trancriptionally regulated by DNA methylation and histone modifications, and they may affect epigenetic mechanisms by targeting key enzymes involved in controlling chromatin structure and histone modifications (12).

Epigenetic marks, including CpG methylation, have been usually considered stable in somatic cells, although it has been observed that some environmental factors can cause variation or reversibility in the DNA methylation patterns in postnatal life. Thus, the pattern changes with ageing in a tissue-specific fashion (15), and an age-related increase in methylation is negatively associated with the expression of genes such as hepatic glucokinase (16), suggesting that DNA methylation could be involved in increasing age-dependent susceptibility to hepatic insulin resistance and diabetes. Other factors that are able to alter DNA methylation patterns such as inflammation (17), oxidative stress (18) and hypoxia (19) are exacerbated in adipose tissue of obese subjects (20). In this regard, the relationship between obesity and the epigenetic regulation of gene expression has been recently reported by our group (21), showing that the successful response to a hypocaloric diet could be related to a lower methylation of the tumour necrosis factor-alpha promoter in PBMC cells.

In this context, epigenome, referred to the overall epigenetic state of a cell, is particularly susceptible to be dys-regulated during gametogenesis and between fertilization and blastocyst formation (22), although subtle alterations in DNA methylation pattern are maintained during gestation, lactation and throughout the life course (Fig. 1). Alterations of DNA methylation patterns occur during both cell lineage specification and ageing, suggesting that a well-functioning epigenetic homeostatic mechanism must be in place to prevent and repair epigenetic abnormalities (23). Alterations in this regulatory circuit would lead to an accumulation of epigenetic lesions beyond repair in cells, which is an essential step towards the cancerous state or nutrition-related diseases such as obesity.

Figure 1.

Periods of life in which DNA methylation processes take place. Erasure of methylation imprints is almost exclusively of two stages: primordial germ cells and blastocyst development. Maternal care, ageing and different environmental factors (such as dietary components, some toxins and drugs, inflammation and perhaps physical activity) regulate the methylation processes in the different periods of life.

Dietary factors and epigenetic regulation

In the last years, different examples of dynamical changes in DNA methylation patterns because of the restriction or supplementation with different nutrients have been reported (Table 1). Some of these findings have been observed during the perinatal period. Thus, maternal caloric restriction has been related to increased methylation of some genes such as Hras (24), whereas the reduction of rat uterine blood flow altered the methylation status of p53 in the kidney of the offspring (25). Protein restriction during pregnancy altered epigenetic regulation of some genes in the newborns (26) such as the glucocorticoid receptor, which appeared hypomethylated (26,27), or peroxisome proliferator activated receptor alpha (PPAR-alpha) (28). The restriction of methyl donors such as vitamins of the B complex (cobalamin and folate) or amino acids (methionine) led to DNA methylation in the preovulatory oocyte and the pre-implantation embryo (29), while mice fed on a folate-deficient diet during the post-weaning period significantly increased genomic DNA methylation in rat liver, which persisted into adulthood (30).

Table 1.  Some examples of dynamical changes in DNA methylation pattern and histone modifications because different nutritional conditions
Nutritional conditionModelDietary/nutritional conditionDose/manipulationTreatment durationMagnitude of epigenetic and/or metabolic effectReference
Maternal alterationsRatReduction of uterine blood flow during pregnancyLigation of uterine arteries3 dLess than 43% in p53 methylation (–179 bp)(25)
RatDietary protein restriction during pregnancy90 protein vs. 180 g kg–1PregnancyAbout <10% in PPAR-alpha and glucocorticoid receptor methylation(26)
Methyl donorsSheepRestriction of cobalamin, folate and methionineSulphur < 50% Cobalt < 95%Pregnancy80% hypomethylated, 20% hypermethylated(29)
RatFolate deficiency in post-weaning period0 mg folic acid kg–1 vs. 2 or 8 mg kg–1Post-weaning periodMore than 34–48% in DNA methylation(30)
RatFolate supplementation in elderly period18 µmol folate kg–18/20 weeksMore than 30% in DNA methylation(38)
MouseDietary methyl supplementation with folic acid, vitamin B12, choline and betaine5 g kg–1 choline, 5 g kg–1 betaine, 5 g kg–1 folic acid, 0.5 mg kg–1 vitamin B12PregnancyLower body weight(34)
MicromineralsHumanMicronutrient (folate, vitamins A and B1) intakeConsume less methyl donorsLong durationIncrease of p16(INK4a), p14(ARF) methylation(43)
RatSelenium deficiency0.2 mg selenium kg–1Weanling ratsLess than 20% in DNA methylation(44)
Caco-2 cellsArsenite deficiency0, 1 or 2 µmol arsenite L–17 dLess than 20% in DNA methylation(45)
Vegetarian dietHumanVegetarian vs. omnivore dietNot showed or determined (unapplicable)YearsLess than 30–40% in MnSOD methylation(46)
Fatty acidsHCT116/HT-29 cellsButyrate supplementation4 mmol L−124 hLess than 70% in RARbeta2 methylation(48)
Other compoundsKYSE 510/150 cellsSoy genistein supplementation5, 10, or 20 µmol L−16 dReversal of gene hypermethylation(50)
OSCC cellsTea polyphenols (epigallocatechin-3-gallate) supplementation20 or 50 µmol L−16 dInhibition of DNA methyltransferase(51)
Rat colonocytesDiallyl disulfide supplementation200 µmol L−13–6 hIncrease of histone acetylation(52)
MouseSulforaphane oral administration10 µmol L−16 hLess than 50–65% in histone deacetylase activity(57)
HoneybeeRoyal jelly intakeFed with royal jellyNot showed or determined (unapplicable)Less than 10–15% in dynactin p62 methylation(58)
Toxins/drugsRatBisphenol A during pregnancy and lactation50 mg kg−1 dietPerinatalLess than 15% in CabpIAP methylation and body weight gain(32)

More specifically related to obesity, in utero or neonatal exposure to a DNA-hypomethylating compound such as bisphenol A has been associated with higher body weight (31), although its effects were prevented by supplementing the diet with different methyl donors such as folic acid or genistein (32). Genistein itself induced hypermethylation of the Agouti gene, decreasing the expression of this gene and protecting offspring from obesity (33). In fact, transgenerational amplification of body weight has been prevented by a promethylating dietary supplement (34). In this sense, maternal supraphysiological methyl group (folate, cobalamin, choline and betaine) supply (35), and a low-protein diet in rodents throughout pregnancy (36) modified DNA methylation of some key metabolic genes (agouti, glucocorticoid receptor and peroxisomal proliferator-activated receptor-alpha). All these data strongly suggest that epigenetic mechanisms may be boosted or impaired by dietary factors in the mother and could be involved in obesity susceptibility in the offspring. Moreover, significant variations in choline dietary requirements can be explained by single nucleotide polymorphisms (SNPs) in genes involved in choline and folate metabolism (37). Some of these SNPs increase the risk of choline deficiency, which could influence DNA methylation status.

In the adult state, some examples of diet-induced epigenetic changes have also been reported. In this sense, folic acid (38,39) has been linked to DNA methylation in a dose-dependent manner, as reviewed by Kim et al. (40). Other dietary factors involved in DNA methylation within the gastrointestinal tract and colorectal cancer development are alcohol (39), vitamin B6(41,42), vitamin A and some minerals (43). Selenium deficiency also seems to be an important modifier of methyl metabolism in some tissues (44), while arsenite deprivation hypomethylates Caco-2 cells (45).

Finally, not only micronutrients can induce epigenetics changes. A threefold increase in the expression of the human manganese superoxide dismutase (MnSOD) gene has been associated with a decreased CpG methylation comparing a vegetarian with an omnivore group (46). Other dietary factors such as fatty acids are likely to participate in epigenetic regulation by DNA methylation (47). For instance, butyrate induced demethylation of RARbeta2 in cancer cells (48). Tocopherols have been related to epigenetic modifications of histones (49). Other compounds present in different plant foods such as the isoflavone genistein (50), tea polyphenols (51) and garlic's diallyl disulfide (52) are also able to regulate DNA methyltransferases or histone acetylation in cultured cancer cells. In this respect, the interesting effect of genistein described by Fang et al. (50) indicated that this substance reactivates methylation-silenced genes, partially through a direct inhibition of DNA methyltransferases, suggesting a reversibility of the process. Therefore, animal studies have shown that genomic demethylation may protect against some tumours such as colorectal cancer (53) but at the same time may promote chromosomal instability and increase the risk of lymphoma and sarcoma (54). The potential reversibility of DNA methylation makes DNA and chromatin changes attractive targets for therapeutic intervention (55). Although nucleoside analogue inhibitors of DNMTs (e.g. procainamide, hydralazine or zebularine) have been widely used in in vitro cell–culture systems to reverse abnormal DNA hypermethylation and restore-silenced gene expression (56), new studies are necessary to investigate the role of dietary factors in the reversion of DNA methylation.

Another interesting example is sulforaphane, an isothiocyanate that inhibits histone deacetylase activity in human colon and prostate cancer lines, which induces an increase in global and local histone acetylation status (57). Finally, a recent finding by Kucharski et al. (58) evidenced that epigenetic information can be differentially altered by nutritional input in honeybees and that the flexibility of epigenetic modifications may provoke profound shifts in developmental fates, including reproductive and behavioural status. Advances in this area will open the door to study the effect of different nutrients and quantitative requirements in DNA methylation, histone modifications and other mechanisms involved in the epigenetic regulation of specific genes related to weight gain and the occurrence of different metabolic diseases.

Epiobesigenic genes

Continuous scientific advances are promoting research about the implication of epigenetic mechanisms in the onset, development and therapy of diseases such as cancer, type 2 diabetes and obesity (4). In this context, the search for gene promoters susceptible to epigenetic regulation and with a role in the development of obesity (epiobesigenic genes) could be of great interest. Some examples of human genes that have been described as regulated by epigenetic mechanisms in relation to a disease or physiological mechanism have been reported in Table 2(59–82). Most of them produce hypermethylation in the promoter (CASP8, IGFBP3, RARA) and are related to enhanced susceptibility to several kinds of tumours (64,76,82), although some of them are involved in different metabolic diseases such as hypertension (HSD11B2) (81), atherosclerosis (PPARG) (62), diabetes (PPARGC1) and endotoxin tolerance (71). Remarkably, some genes of this list may play a role in obesity development and related processes, especially in adipogenesis (59–67,69,70,81), inflammation (66,67,71–76) or insulin signalling (71,78–82).

Table 2.  Examples of several human genes recently described as regulated by epigenetic mechanisms and involved in obesity (epiobesigenes)
Role in obesityGene symbolNameEpigenetic evidenceReferences
AdipogenesisPPARGC1APeroxisome proliferator-activated receptor gamma-coactivator-1-alphaImportant in human islet insulin secretion(59)
AdipogenesisNR0B2Nuclear receptor small heterodimer partnerTumour suppressor methylation-related(60)
AdipogenesisFGF2Fibroblast growth factor-2Homocysteine disrupts endothelial cells through altered promoter DNA methylation(61)
AdipogenesisPPARGPeroxisome proliferator-activated receptor gammaChanges in DNA methylation during cellular ageing and atherosclerosis(62)
AdipogenesisPTENPhosphatase and tensin homologueEpigenetic role in colorectal cancer and gliomas(63)
AdipogenesisRARARetinoic acid receptor-alphaPromoter hypermethylationis associated with prostate and mammary cancer(64)
Adipogenesis and cell cycleCDKN1ACyclin-dependent kinase inhibitor 1A (p21, Cip1)Aberrant promoter methylations related to cancer(65)
Adipogenesis and inflammationLEPLeptinPost-zygotic development, adipocyte maturation and cellular ageing(62,66–68)
AdipogenesisESR1Oestrogen receptor-alphaPrognostic value of Oestrogen receptor hypermethylation(69)
AdipogenesisNR3C1Glucocorticoid receptorMethylation status is sensitive to prenatal maternal mood(70)
Inflammation and insulin resistanceTNFTumour necrosis factor (TNF superfamily, member 2)Epigenetic silencing during endotoxin tolerance and myeloid differentiation(71)
InflammationPLA2G4APlasma secretory phospholipase A(2) type IIAMalignant prostate cells(72)
InflammationSOD3Extracellular superoxide dismutaseDevelopment of foam cells(73)
InflammationSOCS1Suppressor of cytokine signalling 1Severity of liver fibrosis and hepatocarcionma(74)
InflammationSOCS3Suppressor of cytokine signalling 3Role in cellular growth and migration and melanomas(75)
Inflammation and apoptosisCASP8Caspase 8Hypermethylation in neuroblastomas and medulloblastomas(76)
Energy metabolismCOX7A1Cytochrome c oxidase subunit VIIa polypeptide 1 (muscle)Age influences DNA methylation(77)
Fat metabolismLPLLipoprotein lipaseChanges in DNA methylation during cellular ageing(62)
Fat metabolismFABP4Fatty acid-binding protein 4, adipocyteChanges in DNA methylation during cellular ageing(62)
Insulin signallingCAV1Caveolin-1Aberrant methylation is associated with hepatocellular carcinoma(78)
Insulin signallingPIK3CGPhosphoinositide-3-kinase, catalytic, gamma ppCpG hypermethylation is associated with progression of colorectal cancer(79)
Insulin resistanceSSTR2Somatostatin receptor 2Tissue-specific related(80)
Insulin resistance and adiposityHSD11B211 beta-hydroxysteroid dehydrogenase 2In vivo epigenetic repression and relation to hypertension(81)
Insulin resistanceIGFBP3Insulin-like growth factor-binding protein 3Hypermethylation is associated with non-small cell lung cancer(82)

In general, hypermethylation of normally unmethylated CpG islands correlates with transcriptional repression, and the presence of CpG islands in promoter regions suggests that the gene might be regulated at least in part through CpG methylation (81). Mapping of methylation patterns in CpG islands has become an important tool for understanding both normal and pathologic gene expression events (83). A deep analysis of the abundance of CpG islands in these potential epiobesigenic genes (Fig. 2) revealed up to 10 promoters, specially susceptible to epigenetic regulation according to Methprimer (83) accepting their default criteria. Thus, genes involved in adipogenesis such as fibroblast growth factor-2, phosphatase and tensin homologue, cyclin-dependent kinase inhibitor 1A and oestrogen receptor-alpha have enough CpG islands in their promoter to be potential epigenetic targets in relation to obesity and adipose tissue control (20,84–86). Suppressors of cytokine signalling 1 and 3 participate in leptin regulation and in a negative feedback loop to attenuate cytokine signalling (87,88), which are affected in the obese state (89). Genes participating in energy homeostasis such as cytochrome c oxidase subunit VIIa polypeptide 1 (COX7A), lipoprotein lipase and insulin signalling-related genes (such as caveolin-1 and insulin-like growth factor-binding protein 3) could also be potential targets for epigenetic regulation of obesity (1,90–92). Undoubtedly, other promoters (Table 2) that do not contain CpG-rich sequences (not present in Fig. 2) could also be implicated in epiobesigenic mechanisms regulating body weight (71). Glucocorticoid receptor and tumour necrosis factor-alpha are two examples of genes without remarkable CpG islands but under epigenetic control and with an important role in obesity development (27,71,93).

Figure 2.

CpG plots, including CpG islands (shadowed), of several human epiobesigenic genes. Islands are characterized by CpG dinucleotide content of at least 60% that would be statistically expected (∼4–6%), whereas the rest of the genome has much lower CpG frequency (∼1%). Each graph indicates the relative abundance percentage of CpGs (y-axis) in the gene promoter sequence (x-axis). At the bottom of each gene, as small lines, the putative CpG sites and their positions (as base pairs respect to the transcription start or 0 bp) are shown. The grey shadowed areas indicate the CpG islands susceptible to epigenetic regulation.

Conclusion

This review has been focused on the importance of epigenetic regulation in the aetiology and development of obesity and reports some epiobesigenic genes potentially involved in these processes. Two main objectives are currently pursued in the field of dietary factors influencing epigenetics: first, the identification (at an early age) of individuals that could present specific methylation profiles in specific genes suggesting a susceptibility to different metabolic diseases, including excess body weight gain and type 2 diabetes, which would permit the prevention and monitoring of their evolution; and second, the use of dietary supplementation as a means of counteracting adverse epigenomic profiles in an individualized manner, similar to the administration of inhibitors of DNMTs and histone deacetylase inhibitors in cancer therapy.

Conflict of Interest Statement

No conflict of interest was declared.

Ancillary