The global prevalence of type 2 diabetes mellitus (T2DM) is increasing rapidly (1–3). Yet, there is no marketed therapeutic drug that indefinitely cures or prevents the occurrence and progression of diabetes (4). Glucose-lowering agents alone have limited efficacy in preventing T2DM progression (5). Thus, with rapidly increasing rate of diabetes, a deeper understanding of its underlying molecular mechanisms is urgently needed to develop better therapies.

Several lines of evidence suggest that chronic activation of pro-inflammatory pathways in insulin target tissues, such as the adipose tissue, liver, muscle, and pancreatic islets may contribute to obesity, insulin resistance (IR), and T2DM (6). These evidence contributed to the term immunometabolism, which highlights the potential interplay between immune functions and metabolic defects (7). Meanwhile, persistent chronic inflammation can lead to scarring, decreased function, and organ failure, eventually leading to a rise in diabetes-related deaths (8). The vital role of inflammation in the occurrence and progression of diabetes has sparked great interest in exploring immune imbalances as therapeutic targets (4, 9).

The first trigger(s) of immune imbalance accompanying metabolic dysfunction have not been revealed, but the involvement of epigenetic modifications in the upstream regulation of inflammation has been recognized (10). Environmental factors such as altered nutritional status can induce epigenetic changes (11). Subsequently, pre-existing epigenetic marks in genes activate or repress gene transcription in response to environmental stimuli. Thus, epigenetic changes serve as key bridges in the complicated interaction between the environment and genetics to actuate the inflammatory reaction accompanying metabolic disorders.

In this review, we explain the relationship between epigenetics, inflammation and T2DM, and discuss the potential mechanisms by which epigenetic factors contribute to diabetes by regulating inflammation remodeling. We also highlight how the rapidly increasing knowledge base of epigenetics can open a door of opportunities to improve the clinical management of diabetes.

In the past, metabolism and immunity were often considered as two separate phenomena, wherein the main function of metabolism was to maintain the transformation of the body’s substances and energy while that of immunity was to protect against foreign invaders and remove hazardous substances produced by the body itself. Recently, with a more comprehensive understanding of metabolism and immunity, the cross-talk and interaction between the two in the body’s physiological state and during disease development have been recognized (12, 13).

First, immune system function is based on the proliferation and differentiation of immune cells as well as on various types of cytokines. Immune cells need metabolic support to provide energy and substrates, such as glucose, amino acids, phospholipids and fatty acids; thus, the immune system controls metabolism to ensure its energy supply (14, 15). Cytokines in immunometabolism not only play a role in immune regulation but also act as agents of energy metabolism. Some metabolites, such as glucose, can also activate a response to pathogens as a signal medium, besides supplying energy. Overall, immunity plays a vital role in metabolic homeostasis, and an imbalance in immunity could lead to metabolic diseases including T2DM.

T2DM occurrence is consistent with an alternative immune cell profile, including changes in the mononuclear macrophage (Mϕ) system, B lymphocytes, T lymphocytes, natural killer (NK) cells, and innate lymphoid cells (ILCs). Overall, the alteration of the immune cell profile showed a pro-inflammatory proliferation, differentiation and phenotypes, including the increase in M1-like Mϕs, T helper (Th)1 cells, Th17 cells, CD8+ cells, antibody-producing B-2 cells (16–18), and downregulation of M2-like Mϕs, Th2 cells, regulatory T cells (Treg), IgM-producing B-1 cells, and ILCs subsets (such as ILC2s and ILC3s) (19–21). This immunocytic cross-talk causes pathogenic inflammation via the release of cocktail pro-inflammatory cytokines (such as tumor necrosis factor (TNF)-α, TNF-β, interleukin (IL)-1β, IL-2, IL-6, IL-17, interferon (IFN)-α, IFN-γ) and IgG in circulation and glycometabolic tissues such as adipose tissue, liver, muscle, and pancreas, which further disturbs metabolism, ultimately resulting in pancreatic β-cell dysfunction, glucose intolerance, and IR (4).

Mϕs, an important component of innate immunity, plays a crucial role in metabolic inflammation in T2DM. In humans, Mϕs can usually be divided into tissue-resident Mϕs, such as resident intestinal Mϕs, adipose tissue macrophages (ATMs) and liver macrophages (Kupffer cells, KCs), and recruited Mϕs, which are derived from blood monocytes through the binding of monocyte chemoattractant protein-1 (MCP-1,also named CCL-2) and C-C chemokine receptor 2 (CCR2) (22). Primarily, gut microbial composition and metabolic function could be disturbed by a high-fat diet (HFD), and the subsequent upregulation of lipopolysaccharide (LPS) could directly stimulate resident intestinal Mϕs to transfer into M1-like Mϕs, and promote the secretion of pro-inflammatory cytokines such as IL-1β, IL-6, TNF-α, and chemokines CCL-2. Meanwhile, intestinal epithelial cells can also produce CCL2 and cooperatively recruit blood monocytes into gut lumen. The monocytes gradually differentiate towards into resident mature M1-like Mϕs with loss of Ly6C/CCR2 and gain of CD64/MHII expression, and also show high levels of IL-1β, IL-6, or TNF-α and hyper-responsiveness to inflammatory stimuli (23). Through a similar mechanism, because of increased fatty acids, a hypoxic environment and local stimulation of TNF-α, IL-1β, IL-6, and other pro-inflammatory cytokines, resident tissue Mϕs transform to M1-like Mϕs in the adipose, liver, muscle tissue, and pancreatic islets. Furthermore, it contributes to blood monocyte recruitment into these tissues and converts them into M1-like Mϕs to promote an inflammatory environment. It inhibits the insulin pathway, causes IR, and impairs insulin production (24, 25). In the intestine, a HFD induces a pro-inflammatory shift in T cells, characterized by an increase in INF-γ⁺ Th1 and CD8⁺T cells, and a subsequent decrease in IL-10⁺ Tregs and IL17⁺ Th17 cells (26). In addition, there is a decrease in IgA⁺ antibody-secreting B cells and ILC3s in HFD-fed mice, and the reduced levels of colonic secretory IgA+ and IL-22 derived from ILC3s may be associated with IR (26, 27). In adipose tissue, CD8+ T cells are more abundant and can promote monocyte recruitment and differentiation into M1-like ATMs (16). Pro-inflammatory Th1 cells are also increased in obese adipose tissue leading to pro-inflammatory cell infiltration by the production of IFN-γ (28). ILC2s and eosinophils regulate adipose immune homeostasis by inhibiting M1 polarization of Mϕs based on the secretion of IL-13 (derived from ILC2s) and IL-4 (derived from eosinophils) and alleviate adipose inflammation (29). The aggregation of free fatty acids released by the liver and adipose tissue can simultaneously lead to an increase in neutrophils and CD8+ T cells, thereby inducing liver inflammation (30). Neutrophil elastase can be released and taken up by hepatocytes (31), promoting the intracellular degradation of insulin receptor substrate 2 (IRS-2), thereby enhancing IR in hepatocytes.Overall, in T2DM, the cross-talk between metabolism and immunity has been recognized by researchers, and the immune mechanism of T2DM has been partially revealed. However, in the context of T2DM, the internal homeostasis of immunity and its interaction mechanism with glycolipid metabolism need to be further explored.

The Diabetes Control and Complications Trial (DCCT) (32) showed that intensive glycemic control could better delay theprogression of microvascular complications, compared to conventional therapy in patients with T1DM. During the following observationalEpidemiology of Diabetes Intervention and Complications (EDIC) study (33), patients who had been in the conventional therapy group were also switched to receive intensive glycemic control, and both groups successfully achieved similar mean hemoglobin A1c (HbA1c) levels of approximately 8%. Nevertheless, the risk of macrovascular and microvascular complications in patients with early intensive glycemic control in DCCT was significantly higher than that in patients without early intensive treatment (34–37). This has been explained by a phenomenon called “metabolic memory”, which suggests that a ‘memory’ of previous glucose exposures in target cells causes its deleterious effects to persist long after glycaemic control has been established. Growing evidence suggests that epigenetic alterations in target cells are an important cause of high-glucose ‘memory’. Thus, epigenetics, as a link between metabolic memory and the occurrence and development of diabetes, has attracted considerable attention.

Epigenetics refers to the heritable modifications in gene expression without changes in the DNA sequence that regulates cell differentiation, cell-specific gene expression, parental imprinting, X chromosome inactivation, and genomic stability and structure. Epigenome encompasses genome-wide DNA methylation, histone modifications, and the expression of small non-coding RNAs, primarily microRNAs (miRNAs), and also chromatin accessibility (10). As a mechanisms linking environmental factors to altered gene activity, epigenetic processes could be dynamically altered under the influence of short-term and/or long-term environmental exposures, such as drugs, diet, sedentary lifestyle, obesity, elevated blood glucose levels, and aging. Researchers have found hundreds of epigenetic alterations in relation to inflammation, obesity, and T2DM in human tissues that are relevant to metabolism. Of these, epigenetic alterations directly involved in IR and impaired insulin secretion have been reviewed elsewhere (38–40) and will not be covered in the present review. In this review, we focus on epigenetic alterations associated with immune inflammation in the context of T2DM ( Figure 1 ).

As the most extensively studied epigenetic mechanism, DNA methylation regulates gene expression and maintains chromosomal stability. DNA methylation mainly refers to the covalent bonding of a methyl group, namely, 5-methylcytosine (5mC) at the CG context or the so-called CpG sites on cytosine, under the control of specific enzymes known as DNA methyltransferases (DNMTs) including DNMT1, DNMT3A, and DNMT3B (41). Methylation of DNA at gene promoter regions is typically associated with gene suppression through different mechanisms, including the recruitment of transcriptional repressors and interference with transcription factor binding. DNA methylation can be reversed by ten-eleven translocation proteins that convert 5mC to 5-hydroxymethylcytosine. A growing body of research suggests a link between DNA methylation and diabetes (42). Identified by epigenome-wide association studies, differentially methylated CpG sites annotated to several candidate genes for T2D, such as ABCG1, SREBF1, TXNIP, and LGALS3BP, have been found in the bloodc (43). Increased methylation in the ABCG1 and SREBF1 genes, and the subsequent downregulation of their expression, are thought to be associated with T2DM. Both ABCG1 and SREBF1 are transcriptional activators that are required for lipid homeostasis. ABCG1 mediates cholesterol efflux to mature HDL (44). Studies have shown that ABCG1 deficiency in myeloid cells promotes activation of the NLRP3-inflammasome and atherogenesis (45). In addition, decreased methylation of the TXNIP gene is also robustly associated with prevalent T2DM. TXNIP is a key regulator of oxidative stress and has been linked to inflammation; its upregulation can also facilitate the activation of the NLRP3-inflammasome and the release of inflammatory mediators (46). LGALS3BP encodes a secretory glycoprotein regulated by the NF-κB pathway. Previous studies have observed an upregulation of LGALS3BP in TNFα-treated adipocytes, promoting macrophage recruitment, suggesting a potential pro-inflammatory action of LGALS3BP (47, 48). In addition, in a study that focused on the methylation of genes involved in inflammation in the peripheral blood of obese and lean individuals, 28 significantly hypomethylated proinflammatory genes were found in obese individuals. Fifteen of these genes, including CXCL6, TLR5, IL6ST, EGR1, IL15RA, and histone deacetylase (HDAC) 4, showed significantly high mRNA levels. The degree of methylation was negatively correlated with fasting plasma insulin, serum IL6, C-reactive protein, and arteriolar reactive oxygen species (49). Since epigenetic patterns are cell-specific, other insulin target tissues such as adipose tissues (AT) are also of concern. Combining the transcriptomes and methylomes of subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT) samples from obese T2DM individuals and using tissue-specific regulatory networks, Jing et al. identified epigenetically dysregulated gene modules in adipose tissues. Two SAT modules were predicted to be involved in regulating adipocyte differentiation and promoting obesity-related inflammatory responses to impair insulin signaling (50).

Histones are highly conserved proteins in eukaryotic cells and nucleosomes and DNA are the basic units of chromatin structure. Histone modifications, including acetylation, methylation, and phosphorylation, which occur on the N-terminal tails of histones H3 and H4, are important for gene regulation. It can regulate gene expression by altering the affinity of histones to DNA, thereby enabling the transformation of chromatin between loose and dense states. Histone methylation and acetylation have been found to play important roles in the regulation of several key genes associated with diabetes. Histone acetylation states is catalyzed by histone acetyltransferases (HATs) and HDACs. Generally, histone lysine acetylation is associated with transcriptional activation, whereas acetylation removal is associated with transcriptional repression (51). The sirtuin (SIRT) family of deacetylases, specifically SIRT1, regulates several factors involved in metabolism, including adiponectin secretion, inflammatory responses, and levels of reactive oxygen species, which together contribute to the development of IR (52). Histone methylation, in contrast to acetylation, is more constant and long-standing and is associated with both active and silent genes depending on the specific location and degree of modification. Despite being relatively stable, histone methylation can be dynamically modified through the concerted actions of histone methyltransferases and histone demethylases (53). Together with the increased histone methyltransferase Set7 expression in peripheral blood mononuclear cells, the Set7-dependent monomethylation of lysine 4 of histone 3 on NF-kB p65 promoter was also found in patients with T2DM. This epigenetic changes were related to vascular dysfunction via upregulation of NF-kB subsequent transcription of oxidant-inflammatory genes, and increased plasma levels of intercellular cell adhesion molecule-1 and monocyte chemoattractant protein-1 (54). Animal experiments revealed that a series of genes, mainly those enriched in the MAPK signaling pathway, in adipose tissue from HFD-fed mice were activated by histone 3 lysine 9 methylation (H3K9me)2, H3K9me3, H3K4me1, and histone H3 K27 acetylation (H3K27ac). Of these, MAP3K5, MET, and VEGFA may be involved in inflammation-related energy metabolism via activation of the MAPK signaling cascades (55). miRNAs are short non-coding RNAs that can silence the expression of target genes by binding to the 3′-untranslated region of target mRNAs, leading to translational repression and/or mRNA degradation (56). They provide a rapid but reversible means of gene regulation, as a response to environmental stimuli at both the tissue and systemic levels without changing the DNA sequence itself (57).Over 2,500 mature human miRNAs have been identified, and they are thought to regulate up to 60% of human protein-coding genes (58). Some miRNAs are associated with chronic inflammation in T2DM; for example, miR-146a is downregulated in the serum of T2DM patients (59). Another miRNA associated with inflammation in T2DM is miR-147, which is overexpressed in the serum of diabetic and obese rats with periodontitis; miR-147 overexpression is believed to activate macrophages and increase the expression of pro-inflammatory markers such as TNF-α and IL-12 (60). In addition, chronic and transient hyperglycemia was also found to change the levels of miRNAs (miR-26a-5p, miR-26b-5p, let-7d-5p, let-7e-5p, miR-365a-3p, and miR-146a-5p) in adipocytes, which mostly converged to alter IL-6 transcription and can be instrumental in the development of inflammation and metabolic dysregulation of VAT (61).

A growing body of literature links epigenetic modifications to crucial pathways in the pathogenesis of diabetes. As described above, besides altered metabolic processes, DNA methylation, covalent modification of histones, and the expression of non-coding RNA, in particular miRNAs, are also involved in the occurrence and development of chronic inflammation in diabetes. The epigenetic regulation of immune cell recruitment, proliferation, differentiation, and functional phenotypes plays a key role in promoting IR and impairing β-cell insulin production ( Table 1 ).

With the global pandemic of T2DM and the serious harm caused by its complications, studies to unravel the pathogenesis of diabetes and the development of new hypoglycemic drugs have continued. With the help of high-throughput sequencing methods such as genomics, transcriptomics, proteomics, and metabolomics, rapid screening of the key targets of diabetes has been achieved at multiple levels, from genes to downstream functional proteins and metabolites, and has enabled a more comprehensive and profound understanding of diabetes (124). Although the incidence of T2DM has increased dramatically over the past few decades (125), this short period is unlikely to cause significant changes in the human genome, possibly because the expression of key diabetes-related genes is regulated by environmental factors. Epigenetic modifications can reveal the effects of environmental factors such as diet, physical activity, stress, and temperature on T2DM. Since the first epigenetic study reported the changes in DNA methylation in pancreatic islets and skeletal muscle in T2DM individuals, numerous studies have revealed the many epigenetic modifications of islet mass and insulin secretion function, including in PDX1, INS, ADCY5, CDKN1A, PDE7B, PPARGC1A, among others. Besides, there are a number of epigenetic modifications that occur in target glycometabolic organs such as adipose tissue (in ATP10A, IRS1, PPARG, JARID2, TCF7L2, etc.), liver (in ABCC3, GRB10, MOGAT1, PDGFA, etc.), skeletal muscle (in PPARGC1A, MAPK1, FADS2, etc.), and blood (in ABCG1, FAM123C, FHL2, KLF14, etc.). Meanwhile, many studies have revealed the relationship between diet, physical activity, aging, and T2DM with respect to epigenetics (38).

It is worth noting that epigenetic information can also be inherited across several generations; this phenomenon is called intergenerational and transgenerational epigenetic inheritance (IEI and TEI). This indicates that environmental factors can influence not only individuals directly but also indirectly through their parents via IEI or TEI. Retrospective and prospective studies of human cohorts revealed that parental impact via overnutrition or undernutrition could deteriorate the metabolic health of their offspring (126). As shown in the Newborn Epigenetics Study (NEST) cohort, parental obesity can cause susceptibility to weight gain or obesity in offspring through altered small RNA and DNA methylation in human spermatozoa (127). In addition, DNA methylation of the offspring’s umbilical cord blood leukocytes has been identified (128). However, research on the relevant mechanisms of IEI or TEI is limited because of the requirement of long periods of human studies and ethical issues. The use of rodents as model systems enables overcoming some limitations of human studies. Increasing evidence has confirmed that paternal or maternal overnutrition, overweight, and T2DM could all increase offspring’s risk for obesity and diabetes. The core mechanisms by which IEI or TEI manifests its effects include the programming regulation of the offspring’s pancreas and adipose tissue by epigenetic inactivation of DNA methylation and microRNA in germ cells during the process of development and functional differentiation (126, 129). However, there are only a few studies on the mechanisms of IEI and TEI from an immunological perspective and should be studied to reveal the mechanisms of inherited metabolic disorders to explore prevention and treatment methods.

The inflammatory state caused by immune imbalance plays an important role in the development of T2DM. Therefore, epigenetic modifications of immune cells can indirectly regulate the inflammatory state of the body and affect insulin resistance and insulin secretion dysfunction.In our review, we focused on summarizing the changes in immune-related epigenetic modifications in the context of T2DM, especially the epigenetic regulation of macrophages, T and B lymphocyte recruitment, proliferation, differentiation, and functional phenotypes. In addition, we also identified some possible directions for future research, such as histone lactylation of macrophages in the background of T2DM and epigenetic modification of ILCs, which are widely involved in glucolipid metabolism and immune homeostasis. These findings provide novel insights for revealing the pathogenesis of T2DM from the perspective of epigenetic immune modification and can provide new directions to reveal the pathogenesis of T2DM with respect to epigenetics. However, several issues need to be addressed. For example, as it is difficult to obtain diabetic tissue samples, there are relatively few epigenetic studies based on the local tissues of patients, and most studies are carried out on the epigenetics of blood. Thus, larger cohorts must be conducted through worldwide collaborations, and owing to the tissue- and cell-specific nature of epigenetic modifications, more studies on human biopsies are desirable. Meanwhile, most of the studies at this stage are descriptive studies, and although they provide a macroscopic understanding of glycolipid metabolic functions and immune-related epigenetic changes in T2DM, the lack of detailed studies leaves us with uncertainty about its targeting, which further affects the application of research findings in clinical diagnosis and treatment.

The application of epigenetic findings facilitates the clinical diagnosis and treatment of T2DM. First, the prevention of T2DM, is crucial for the diagnosis and treatment of diabetes; therefore, the identification of epigenetic biomarkers indicative of T2DM susceptibility is of great significance. Notably, biomarkers need to be identified in easily accessible human samples, and blood samples are a good choice. Although blood epigenetic biomarkers have been used for tumor prediction, in T2DM, definitive epigenetic biomarkers still need to undergo extensive validation (130). Clinical translational research based on epigenetic reversibility has also been conducted, and some drugs that affect DNA methylation and histone modification, such as histone deacetylase inhibitors (valproic acid, sodium phenylbutyrate, vorinostat, and givinostat), histone acetyltransferase inhibitors (curcumin), protein arginine methyltransferase inhibitors (AMI-1), DNA methyltransferase inhibitors (hydralazine, procainamide, RG108, MG98), histone demethylating inhibitors (tranylcypromine), and sirtuin-activating compounds (resveratrol), have been developed (131). The efficacy and application of these drugs in alleviating metabolic diseases included obesity, diabetes, fatty liver, and metabolic syndrome have been preliminarily confirmed, and corresponding clinical studies, including those involving valproic acid, sodium phenylbutyrate, resveratrol, etc., are being carried out gradually. However, the low specificity and global action of these medicines may lead to side effects. Therefore, the development of highly specific epigenetic drugs targeting specific diseases and cell functions may be a more suitable direction for future research.

QD and ZG are co-responsible for the collection, collation, and writing of the original manuscript. LZ designed and revised the manuscript. KC, QZ, and SH are responsible for the concept development, revision, and review of the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 81704067) and Innovation Team and Talents Cultivation Program of National Administration of Traditional Chinese Medicine (Nos.ZYYCXTD-D-202001).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.