Diabetes and Epigenetics

Abstract

As we attempt to understand and treat diseases, the field of epigenetics is receiving increased attention. For example, epigenetic changes may contribute to the etiology of diabetes. Herein, we review the histology of the pancreas, sugar metabolism and insulin signaling, the different types of diabetes, and the potential role of epigenetic changes, such as DNA methylation, in diabetes etiology. These epigenetic changes occur at differentially-methylated sites or regions and have been previously linked to metabolic diseases such as obesity. In particular, changes in DNA methylation in cells of the pancreatic islets of Langerhans may be linked to type 2 diabetes (T2D), which in turn is related to peripheral insulin resistance that may increase the severity of the disease. The hypothesis is that changes in the epigenome may provide an underlying molecular mechanism for the cause and deleterious metabolic health outcomes associated with severe obesity or T2D. Conversely, reversing such epigenetic changes may help improve metabolic health after therapeutic interventions.

1. Introduction

Diabetes mellitus (DM) is a chronic metabolic disease in which either the pancreas produces very little to no insulin, termed type 1 diabetes (T1D), or insufficient insulin in the context of systemic insulin resistance, termed type 2 diabetes (T2D) [1]. Both of these conditions result in high levels of glucose in the bloodstream.

DM is associated with significantly elevated diabetic nephropathy, neuropathy, and retinopathy, which are microvascular complications, and cardiovascular conditions such as hypertension, atherosclerosis, and stroke, which are considered macrovascular diseases. DM is associated with genetic as well as environmental factors, with the cost of treatment and debilitating complications increasing dramatically due to an epidemic of DM worldwide.

However, the above statement is something of an oversimplification, because besides T1D and T2D, there are even more variants, and we will now look at all of these in more detail.

2. Type 1 diabetes

T1D represents only around 10% of the DM cases worldwide but is increasingly occurring earlier in life. It results from autoimmune destruction of the beta cells (β-cells) of the endocrine pancreas. As with cancer, obesity, and autoimmune diseases, T1D results from the interaction of genetic and non-genetic factors [2]. In autoimmune diseases, due to an immunological malfunction and lack of tolerance of self-antigens, the immune system destroys the body’s own tissues. More than 80 different diseases are considered autoimmune and affect approximately 100 million people worldwide [3]. T1D is one such example that results in the slow degeneration and destruction of the pancreas. However, approximately 10% of the affected patients are classified as subtype 1B, and the pathogenesis in these cases is considered idiopathic since there is no evidence of autoimmunity [4]. T1D can be correlated to ethnicity, gender, genetics, and environmental influences. For instance, it occurs more in children and those under the age of 20 and affects both male and female children equally. However, studies have found that males are disproportionally affected in areas with a high prevalence of T1D, whereas females are disproportionally affected in areas with a low prevalence of the disease. It is highest in non-Hispanic White people and lowest in Navajo groups. Moreover, T1D is common in families with a history of the disease. Epidemiological studies have found an association between T1D and environmental factors, and dietary and nutritional habits [5].

Essential mediators leading to β-cell destruction in T1D include the following pro-inflammatory cytokines: interleukin-1β (IL-1β); interferon-γ (IFN-γ); and tumor necrosis factor-α (TNF-α). These cytokines induce the overexpression of iNOS in β-cells, leading to an overproduction of NO that causes cytotoxicity. This suggests an important role for NO in the pathogenesis of DM [6].

3. Type 2 diabetes

In contrast to T1D, T2D is a defect in insulin secretion, insulin action, or both that leads to the development of a multifactorial and heterogeneous group of disorders. Changes in diet and physical activity levels have led to an increased worldwide prevalence of T2D over the past several decades. There is also strong evidence supporting a genetic component of T2D susceptibility, and several genes underlying monogenic forms of DM have already been identified. However, T2D likely results from the contribution of many genes interacting with different environmental factors to produce wide variations in the clinical course [7].

Regarding this process, there is a decrease in β-cell mass in T2D with the primary implicated mechanism being the apoptosis of the cells. This type of dynamic cell death is increased in all diabetic individuals; β-cell mass depends on many factors, including cell size, cell renewal rate from proliferation of pre-existing cells or neogenesis (differentiation from other precursor cells), and speed of apoptosis. Also, β-cell failure during the progression to T2D can be caused by either chronic exposure of the β-cell to glucose, which is called “glucotoxicity,” or exposure to fatty acids, which is known as “lipotoxicity” [8].

4. Latent autoimmune diabetes in adults and maturity-onset diabetes of the young

Besides T1D and T2D, some forms of the disease do not fit neatly into those groups, namely latent autoimmune diabetes in adults (LADA) and maturity-onset diabetes of the young (MODY). LADA shares some type 1 and type 2 symptoms and treatments, is diagnosed during adulthood, and sets in gradually, like T2D [9]. On the other hand, MODY is caused by genetic changes that affect how well the body makes insulin [10].

5. Type 3 diabetes

There may be a connection between T2D and Alzheimer’s disease (AD). Indeed, it has been proposed that AD is actually a form of DM termed type 3 (T3D). Globally, the epidemic of T2D and the possibility of it contributing to the risk of AD have become a paramount health concern.

The hypothesis is that T3D corresponds to chronic insulin resistance plus an insulin-deficient state that is mostly confined to the brain [11, 12]. A deficit in glucose utilization is observed, which ultimately leads to cognitive dysfunction. Over time T3D steadily destroys cerebral functions due to insulin imbalance. Thus, the central nervous system develops insulin resistance, which leads to AD [13].

6. Gestational diabetes

Finally, in terms of DM classifications, the American Diabetes Association defines gestational DM (GDM) as DM seen during pregnancy. GDM occurs in approximately 5% of pregnancies, but rates can increase due to obesity. Pregnancies with a diagnosis of GDM present a risk to both mother and child. Women who have a record of GDM will typically develop T2D after pregnancy. Their children have a higher incidence of becoming obese and developing T2D early in life [14].

So, as we can see, ultimately, DM and its resultant health conditions are numerous and include many degenerative diseases such as the above-described six types of DM.

But possibly, this myriad of insidious conditions and diseases could be preventable if we focus research on the pancreatic β-cells (vital to the regulation of glucose levels in the bloodstream). But first, what exactly are β-cells? At this point, we will look at the histological composition of the pancreas in more detail.

7. The cells of the pancreas

Histologically, the adult pancreas consists of endocrine and exocrine cells, but these cells can change their state of differentiation in response to various stimuli (e.g., injury or stress). The exocrine portion of the pancreas produces and releases enzymes that digest proteins and lipids. In contrast, the endocrine portion produces hormones such as insulin and glucagon, which control blood glucose levels. During the cephalic phase of digestion, even before food enters the mouth, digestive enzymes and insulin are secreted to regulate and coordinate metabolic processes.

Over millions of years of evolution, large portions of what we today call the pancreas evolved originally from just exocrine tissue. As a result of this evolution, endocrine cells form encapsulated boundaries called islets of Langerhans, that separate the endocrine and exocrine acini within the pancreas [15].

Moreover, many cell types are present within these evolved islets: the alpha (α), beta (β), and delta (δ) cells produce the vital hormones glucagon, insulin, and somatostatin, respectively. A fourth cell type, known as the pancreatic polypeptide (PP) cell has the significant function of inhibiting glucagon release. Yet other cell types, ghrelin-positive cells, are mainly found in the gut and in the islet to inhibit insulin and somatostatin secretion and regulate the secretion of glucagon, PP, and somatostatin (Figure 1) [16]. Evidence in the literature, and discussed in detail by Da Silva Xavier [16], suggests transdifferentiation of α-cells via stimulation by gamma-aminobutyric acid and δ-cells into insulin-containing β (like)-cells. It is unknown whether the replenishment of β-cells from the transdifferentiation of α-cells is able to replace hub (stem) β-cells which influence the function of other β-cells (Figure 1) [16]. In any case, the pancreas does appear to have some reserve capacity for the regeneration of β-cells but this may be overwhelmed in DM.

Figure 1.

Schematic diagram showing the interaction of islet cells. Evidence points to the transdifferentiation (light blue arrow) of α-cells (red) via stimulation by gamma aminobutyric acid (GABA), and δ-cells (pink) into insulin-producing β-cells (green). It is unknown whether the replacement of β-cells by α-cells is able to take the place of hub β-cells (light blue) which influence the function of other β-cells (yellow arrows). Somatostatin released from δ-cells can inhibit the release of glucagon, insulin, and pancreatic polypeptide from α-, β-, and PP cells (purple), respectively. Pancreatic polypeptide released from PP cells can inhibit the release of glucagon. Ghrelin released from ghrelin-positive islet cells (orange) can inhibit insulin and somatostatin secretion.

In fact, the most abundant cells in the islets of Langerhans are the β-cells, comprising 55% of the cell number, and it has been suggested that they interact with other endocrine cells to influence the secretion of hormones [8]. Moreover, because the β-cells produce insulin they are the most critical pancreatic cell type involved in the etiology DM. It is therefore timely at this point to look at insulin production in more detail.

8. Insulin synthesis and secretion

Insulin production begins in the β-cells with the secretion of pre-proinsulin that is converted into proinsulin. Proinsulin is then transformed into insulin and C-peptide, which are stored in the form of secretory granules until they are triggered for release throughout the body during food ingestion. Insulin is mainly produced in response to glucose. This has been validated in vitro: it was found that when human islets or stem cell-derived β-cells were stimulated with glucose, they secreted insulin [17]. Other hormones, such as melatonin, estrogen, leptin, growth hormone, glucagon-like peptide 1, etc. can modulate the level of insulin secretion.

The primary signal that stimulates insulin exocytosis from granules is a process triggered by glucose (for instance, increased intake of dietary sugar) followed by a rise in intracellular calcium (Ca²⁺) [8, 18]. Calcium influx relies on many factors such as glucose transport, metabolic enzymes, and functioning potassium ion channels. Moreover, an elegant study showed that the growth and survival effects of glucose on β-cells require activation of proteins in the insulin signaling pathway via an autocrine mechanism (Figure 2) [19]. This, the fact that β-cells both secrete and respond to insulin via autoregulation may make they especially vulnerable to epigenetic changes induced by glucose (Figure 2), In the model proposed by Assmannet al. [19], we further posit that the identified targets may be exceptionally sensitive to epigenetic dysregulation in DM (in addition to transcriptional and/or translational dysregulation) (Figure 2). If these targets are epigenetically misregulated they may become difficult to normalize.

Figure 2.

Diagram of a link between glucose and insulin signaling in β-cells and indicating epigenetic effects (blue arrows). (A) Potential direct effects of glucose and/or its metabolites on proteins in the insulin/IGF-1 signaling pathway. (B) Potential indirect effects of glucose and direct effects of insulin following exocytosis of insulin. Akt, v-akt murine thymoma viral oncogene homolog; FoxO-1, forkhead box O1; GRB2, growth factor receptor-bound protein 2; PIP3, phosphatidylinositol (3,4,5)-trisphosphate; mTOR, mammalian target of rapamycin; 4EBP1, translation initiation factor 4e binding protein 1. We postulate that β-cells are especially sensitive to epigenetic perturbations (blue arrows) because unlike cells that do not produce insulin, β-cells also have an autocrine quality, in that they both produce insulin, and receive the insulin signal by receptor binding. This autoregulatory loop may be especially vulnerable to epigenetic dysregulation.

Furthermore, the pancreatic endoplasmic reticulum kinase (ERK) plays a central role in regulating translational events. It regulates insulin translation through phosphorylation of eukaryotic initiation factor 2 alpha (eIF2a). ERK mutation is linked with permanent neonatal DM in humans [8]. One such example is Wolcott-Rallison syndrome (WRS), a rare autosomal recessive disease characterized by neonatal/early-onset non-autoimmune insulin-requiring DM associated with skeletal dysplasia and growth retardation [20]. Because glucose is critical in activating insulin signaling, let us look at glucose in more detail.

9. Glucose metabolism

Glucose activates other cell signals, such as cyclic AMP (cAMP), cyclic GMP (cGMP), inositol 1,4,5-trisphosphate (IP3), and diacylglycerol (DAG). When cAMP is produced, it then activates protein kinase A (PKA). cAMP may be the most crucial molecule that leads to insulin secretion and phosphorylation of proteins involved in insulin exocytosis via PKA. Incretin hormones also augment glucose-stimulated insulin secretion by stimulating the cAMP signaling pathway [8].

Incretin hormones are gut peptides secreted by enteroendocrine cells after feeding [21]; their function is to control the amount of insulin. In the pancreas, two kinds of incretins, glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1), share the same behavior, but outside the pancreas, they differ. They are both rapidly deactivated by an enzyme called dipeptidyl peptidase 4 (DPP4). A decrease in incretin secretion or an increase in incretin clearance is not a pathogenic factor in DM. However, in T2D, GIP no longer modulates glucose-dependent insulin secretion, even at supraphysiological (pharmacological) plasma levels. GIP incompetence is detrimental to β-cell function, especially after eating. On the other hand, GLP-1 is still insulinotropic in T2D, which has led to the production of compounds that activate the GLP-1 receptor intending to improve insulin secretion [22].

Furthermore, glucose metabolism is critical in insulin biosynthesis because it triggers insulin gene transcription and mRNA translation. The triggering of insulin gene transcription and mRNA translation is necessary for regulating insulin biosynthesis via modification of proinsulin mRNA expression and maintaining insulin mRNA stability [8]. mRNA has a vital role in regulating and controlling gene expression and this stability is affected by how RNA-binding proteins and structural elements interact with each other.

Regulation of mRNA stability is accomplished through various reactions to developmental stimuli (e.g., nutrient levels, cytokines, hormones, and temperature shifts or to different environmental stimuli such as stresses like hypoxia, hypocalcemia, viral infection, and tissue injury). However, deregulated mRNA stability can cause mRNA accumulation contributing to some forms of neoplasia, thalassemia, and AD [23]. The results from in vitro studies revealed that insulin mRNA stability decreases under lower glucose concentrations and increases under high glucose conditions [8]. In the absence of glucose, insulin mRNA levels in β-cells decrease sharply, which is reversed by elevating intracellular cAMP levels.

10. Transcription factors

Another stratification of regulation besides glucose signaling and mRNA expression is at the level of transcription factors. These play a central role in regulating gene expression by binding to specific consensus sequences, or cis-elements, within promoter regions [24]. Transcription factors are proteins that can be targets of modifications when they respond to cellular stimuli. This will affect their stability, activity, intracellular distribution, and interaction with other proteins [25]. One of these stimuli is insulin resistance, which affects many organs, mainly the liver, pancreas, adipose tissue, and muscle [26, 27].

Illustrating the power of transcription factors, pancreatic acinar cells can be reprogrammed to produce, process, and secrete insulin when forced to express the transcription factors Pdx-1, MafA, and Ngn3 [28].

Cellular differentiation processes can also be negatively impacted by gene expression [8]. For example, the deletion of pancreatic and duodenal homeobox 1 (PDX-1) from postnatal islets results in phenotypic loss of the β-cells. The Pdx-1 protein is a transcription factor responsible for the development of α and β-cells [8]. A second proposed explanation for the loss of β-cell phenotype is that there is an increase in α cells in the pancreatic islets due to a lack of Pdx-1 transcriptional processes, which convert α cells to β-cells [8]. This process would produce an overall imbalance between α and β-cells; hence the β-cells would ultimately be fewer, but it is unclear whether the loss of the β-cell phenotype was due to lack of Pdx-1 or lack of α to β-cell conversion [29].

Even more starkly, mutation in both PDX-1 and PTF1A results in pancreatic agenesis [8]. PTF1A is a gene that is a component of transcription factor 1 complex (PTF1) in the pancreas and encodes a protein that functions in embryonic pancreatic development. It is crucial and determines if cells in the pancreatic buds go on either towards pancreatic organogenesis or return to duodenal fates [18, 30].

Although we see that transcription factors are powerful effectors of cellular behavior, there are deeper layers than this, namely epigenetic, which we will begin to cover in more detail.

11. Etiology of type 2 diabetes mellitus

We have covered the different types of diabetes, the cells that comprise the pancreas, glucose signaling, transcription, and some of the genes that govern pancreatic cellular behavior and differentiation, so we will now look more closely at the etiology of T2D, ultimately moving into the epigenetic layer.

The etiology of T2D is complex and multifactorial since it is affected by genetic predisposition [31] and behavioral influences, such as diet and physical activity [32]. As previously stated, T2D is often characterized by β-cell dysfunction, insulin resistance, and hyperinsulinemia [33]. These factors and symptoms depend on the disease phase and how insulin affects and regulates the bloodstream’s high level of glucose [34]. Essentially, genetic, epigenetic, and non-genetic factors influence the pathogenesis of T2D [35].

Firstly, as far as genetics goes, genome-wide association studies have identified associations between single-nucleotide polymorphisms (SNPs) and disease in large case-control cohorts and family-based studies. However, although over a hundred genetic variants have been identified that are associated with T2D risk, they can explain only a modest portion of T2D heritability [36].

There also are non-genetic risk factors for T2D, such as age, physical inactivity, and energy-rich diets that result in obesity [35]. However, it is not necessary to be obese to have this type of DM. Most patients who have DM are not obese, have an incommensurable reduction of insulin secretion, and are less insulin resistant than obese individuals. It was discovered that T2D could also exist in the absence of an obese phenotype by studying non-obese rodent models [34].

There is also evidence that DM in adulthood can be caused by intrauterine or fetal malnutrition. This type of malnutrition is vital to comprehending adult DM because the genetic abnormalities and imbalances in the mother’s uterus can affect the probability of her child developing DM even as an adult. Another study has shown that low birth weight leads to T2D development or insulin resistance. Moreover, factors such as the mother having DM, low birth weight of the child, and fetus malnutrition, work in a complex manner with a variety of epigenetic regulators (guided by α and β-cell-type-specific transcription factors such as Pdx1, mentioned earlier) and result in abnormal β-cell maturation and differentiation causing adult DM [37].

Thus, other explanations for T2D heritability have been proposed, including alterations in epigenetic patterns [35]. We likely need a more holistic understanding of epigenetics to obtain a complete picture of the etiology of DM, especially environmental-epigenetic interactions.

But what exactly does “epigenetic” mean? It is here that we can delve more into the molecular aspects of DNA and chromatin and how they relate to gene expression and disease etiology.

12. Role of epigenetics in diabetes mellitus

Epigenetics addresses the relationship between genes, environmental exposure, and disease development. Additionally, epigenetics concerns heritable gene expression changes without changes in the DNA sequence itself, affecting how cells “read” genes. Many factors affect epigenetic modifications, such as age, lifestyle, family history, and disease status. Today, three major epigenetic systems are recognized: DNA methylation, histone modifications (the most well-characterized being acetylation), and non-coding RNA (ncRNA)-associated gene silencing (Figure 3).

Figure 3.

Epigenetic regulation of gene expression. Epigenetic alterations such as DNA methylation and/or histone modifications alter the accessibility of genes to the transcriptional machinery by inducing either a relaxed/open or condensed/closed chromatin state. Non-coding RNAs such as miRNAs also regulate the cell phenotype by repressing or enhancing the expression of gene transcripts. Conversely, these non-coding RNAs can themselves be epigenetically regulated.

Epigenetic alterations such as DNA methylation and/or histone modifications alter the accessibility of genes to the transcriptional machinery by inducing either a relaxed/open or condensed/closed chromatin state. In general DNA methylation, principally of cytosines in gene promoters, condenses DNA and leads to gene silencing, whereas acetylation of histones opens up chromatin and is associated with gene activation (Figure 3). Non-coding RNAs such as miRNAs also regulate the cell phenotype by repressing or enhancing the expression of gene transcripts (Figure 3). Conversely, these non-coding RNAs can themselves be epigenetically regulated. Epigenetic changes often occur during an organism’s lifetime and are sometimes transmitted to the next generation [38].

Several studies suggest that epigenetics plays a vital role in the pathology of DM, especially T2D. Common T2D is likely to result from many genes interacting with different environmental factors (Figure 4) to produce a wide variation in the disease’s clinical course [7], and as previously described for other multifactorial diseases such as hypertension [39]. In the model proposed by Arif et al. [39], epigenetic and genetic factors regulate phenotypes. Specifically, in addition to heritable Mendelian genetics, polygenic phenotypes, such as DM, are significantly affected by gene-environment interactions triggering epigenetic modifications (Figure 4). Indeed, previous studies have shown that epigenetic mechanisms can predispose individuals to the diabetic phenotype. Also, the altered homeostasis in T2D, such as prolonged hyperglycemia, dyslipidemia, and increased oxidative stress, could result from, and cause, epigenetic changes associated with the disease [40].

Figure 4.

Influences on the expression of phenotypes. Development of polygenic conditions, such as diabetes, depend on complex and interacting genetic and environmental pathways.

As previously stated, the main insulin-producing cells in the pancreas are the β-cells, and epigenetic modifications play a critical role in establishing and maintaining their identity and function in physiological conditions [41]. Stable β-cell function is vital to the regulation of glucose levels in the bloodstream. In the case of diabetes, epigenetic dysregulation may result in the reduction of the expression of genes essential for β-cell function, the ectopic expression of genes that are not supposed to be expressed in β-cells, and loss of genetic imprinting, leading to loss of β-cell identity [40]. Consequently, this may lead to β-cell dysfunction and impaired insulin secretion, impairing the function of the pancreas, and in turn, causing widespread sequalae and finally disease in the whole organism, Thus, a causal chain is established whereby the environment causes disease in the following sequence: environment ⟶ chromatin ⟶ genes ⟶ cells ⟶ organs ⟶ organism [42]. The model proposed by Liu et al. goes a long way in establishing this causal chain (Figure 5) [42].

Figure 5.

A stratified view of gene-environment interactions during development and disease. Environmental effects are incorporated by epigenetic processes including chromatin remodeling to either inhibit or enhance gene expression. These effects are then manifested hierarchically in the sequence of cells to organs (i.e. pancreas) to organism. Disease etiology (for example diabetes) occurs in this hierarchical sequence.

It’s important to realize that many risk factors may lead to epigenetic dysregulation by causing this initial “disruption” to chromatin, such as hyperglycemia, physical inactivity, parental obesity, mitochondrial dysfunction, aging, and an abnormal intrauterine environment. Those factors can affect the epigenome at different time points throughout the lifetime of an individual. Moreover, the epigenome can change due to environmental factors, such as diet and exercise, because of the epigenome’s plasticity. As a result, the epigenome is a good target for epigenetic drugs that may be used to induce insulin secretion and treat DM [40].

Ultimately, epigenetics is vital to research DM and possible future treatments. It could be the solution to early detection and treatment, via timely detection and modification of relevant genes and reversal (normalization) of signaling pathways. But how do we identify these genes and pathways?

13. DNA methylation, and the pathogenesis and potential treatment of type 2 diabetes mellitus

As mentioned, DNA methylation and histone modifications (typically acetylation but there are others such as phosphorylation, ribosylation, ubiquitylation, sumoylation, and citrullination [43]), are the main mechanisms in which epigenetics affects cell phenotype and biological processes (Figure 3).

Of the two, DNA methylation has been the most well-studied by microarray. During methylation of DNA, 5-methylcytosine is created by DNA methyltransferases modifying cytosines. Most of this occurs in CpG islands in the promoter regions in multiple protein-coding genes. Methylation of cytosines at the promoter regions is associated with the repression of transcription. Repressors that bind to methylated CpG islands then initiate a cascade that results in the second primary mechanism of epigenetic regulation: namely histone modifications and the recruitment of histone deacetylases (or transferases) [27]. Repression of multiple genes could then lead to the DM phenotype. But we need to identify what these genes are.

Bansal and Pinney [44] reviewed studies where both DNA methylation and gene expression changes were reported. DNA methylation status had a strong inverse correlation with gene expression, suggesting that this may be a potential future therapeutic target. They highlighted the emerging use of genome-wide DNA methylation profiles as biomarkers to predict patients at risk of developing diabetes or specific complications of diabetes.

Indeed, developing predictive models that incorporate both genetic information and DNA methylation changes may be effective diagnostic approaches for all types of diabetes and could lead to additional innovative therapies.

For example, one study used the genome-wide Infinium 450K array and identified 1,649 CpG sites, and 853 genes that include TCF7L2, FTO, KCNQ1, IRS1, CDKN1A, and PDE7B. Significant changes in DNA methylation were found in donors that have T2D compared to controls. Also, increased DNA methylation at the promoter of CDKN1A and PDE7B was associated with decreased transcriptional activity in clonal in vitro, as well as impaired glucose-stimulated insulin secretion [45].

Another genome-wide study of DNA methylation using the Infinium27K array found 276 differentially-methylated CpG sites, of which 96% were hypomethylated in islets of diabetic compared to non-diabetic donors [46]. Changes in differential DNA methylation were correlated with expression changes of 34 genes assessed by microarray [46].

We are also conducting our own genome-wide methylation studies using a human in vitro model of diabetes based on induced pluripotent stem cell-derived β-cells.

Interestingly, bariatric surgery appears to be capable of partially reversing the obesity-related and diabetic epigenome [47]. The identification of potential epigenetic biomarkers predictive of the success of bariatric surgery may open new doors to personalized therapy for severe obesity and diabetes, which is cause for great optimism [47].

14. Conclusions

DNA methylation changes at differentially methylated sites or regions have been linked to metabolic diseases such as obesity and T2D. Thus, changes in the epigenome may provide an underlying molecular mechanism for the deleterious metabolic health outcomes associated with these conditions. Conversely, coordinated reversal of these changes may improve metabolic health after therapeutic intervention, and this provides optimism for the future.

Acknowledgments

Special thanks to Saudi Culture Mission (SACM) in the USA and King Saud Bin Abdulaziz University for Health Sciences in Riyadh, KSA, for their immense help and financial support to R.A.A. This publication resulted in part from research support to A.B.C. from the National Institute of Health (NIH) R25 Resource Grant (1 R25 AG047843-01) and to T.H. from the National Science Foundation [NSF IOS-1355034], Howard University College of Medicine, and the District of Columbia Center for AIDS Research, an NIH funded program [P30AI117970], which is supported by the following NIH Co-Funding and Participating Institutes and Centers: NIAID, NCI, NICHD, NHLBI, NIDA, NIMH, NIA, NIDDK, NIMHD, NIDCR, NINR, FIC, and OAR. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conflict of interest

The authors declare no conflict of interest.