DNA Methylation and Non-Coding RNAs in Metabolic Disorders: Epigenetic Role of Nutrients, Dietary Patterns, and Weight Loss Interventions for Precision Nutrition

The human cell epigenome is a highly dynamic and flexible system capable of responding to the environment factors such as nutrition, stress, toxicity, exercise, and drugs [1, 2]. Dysregulation of the epigenetics process and abnormal epigenetics are associated with various human metabolic disorders, including cancer [3, 4], obesity [5], type 2 diabetes [6], neurological disorders [7], as well as predictors of disease outcomes [8]. Nutrition is one of the environmental factors that can induce epigenetic changes both through direct exposure and transgenerational inheritance [9], which continuously alter gene expression and shape the phenotype. With the constant and overwhelming increase in the prevalence of obesity and metabolic disorders, precision nutrition, whose main goal is the development of personalized nutritional recommendations for disease prevention and treatment, represents a promising approach. It seeks to develop more comprehensive and dynamic recommendations by considering not only nutrition itself but also all mutable parameters to which the individual is subjected, such as genetics, epigenetics, and eating behavior, among others [10].

Currently, it is understood that diets can alter the epigenetic pattern at multiple levels: by regulating DNA methylation, by promoting histone modifications, and/or by changing non-coding RNA (ncRNA) expression [11, 12]. ncRNAs can be classified based on their length into short (<200 nucleotides) and long (>200 nucleotides) [13]. Advances in molecular biology have shown that gene expression is largely regulated not only by proteins but also by ncRNAs [14]. Among the ncRNAs, microRNA (miRNAs) and lncRNAs are particularly important in the regulation of gene expression. miRNAs are small RNA molecules ranging from 17 to 25 nucleotides. These molecules recognize target messenger RNA through sequence complementarity and regulate its protein translation [15]. In contrast, lncRNAs are involved in numerous gene regulatory activities, such as transcription, splicing, protein degradation, and chromatin modifications [16, 17], thereby modifying chromatin states and influencing gene expression. They are classified as a class of non-coding RNAs that consist of sequences longer than 200 nucleotides [18]. Although the biological functions of circRNAs are not yet fully elucidated, they have been shown to act as potent miRNA sponges via competition with miRNA/mRNA binding. CircRNAs are covalently closed single-stranded RNA rings generated from a process known as back-splicing or head-to-tail circle splicing, which involves the joining of a splice donor to an upstream splice acceptor of precursor mRNA (messenger RNA) [19, 20].

These molecules are components of complex regulatory mechanisms and pathways, interacting with gene expression regulation and consequently with epigenetics. In this context, they can modulate the epigenetic landscape in various ways: by altering the availability of substrates required for proper reactions or by directly suppressing DNA methylation or histone-catalyzing enzymes [21]. Additionally, miRNAs can regulate DNA methylation and histone modifications, while promoter methylation or histone acetylation can also modulate miRNA expression as part of a complex network involving feed-forward and feedback loops. Methyl- and folate-deficient diets can result in altered miRNA expression [11] and some studies suggest that weight loss interventions can modulate the expression of several miRNAs as well [22]. Modifications in DNA methylation of genes related to certain diseases could contribute to changes in gene expression and disrupt the function of critical metabolic pathways [23].

Epigenetic modifications are highly dynamic in response to environmental factors such as diet and exercise. Although most studies are correlational, several intervention studies have been conducted to dissect the impact of lifestyle modifications on the human epigenome. In this context, different dietary patterns, nutrients, and food components have been associated with epigenetic processes that may contribute to susceptibility to metabolic diseases, including obesity [24, 25]. Similarly, physical activity and exercise can alter epigenetic signatures [26]. This narrative review aims to compile scientific evidence demonstrating the potential epigenetic properties of nutrients, dietary patterns, and weight loss interventions in understanding the physiopathology of metabolic disorders and implementing precision nutrition.

Some nutrients, such as folate, vitamin B12, vitamin B6, vitamin B2, choline, and betaine, derived from diet (both natural and fortified foods) or supplements [9] act as methyl donors or co-enzymes for one-carbon metabolism. This metabolism regulates methyl transfer for DNA methylation and histone methylation reactions [27, 28]. Additionally, some nutrients and bioactive compounds can directly affect enzymes that catalyze DNA methylation and histone modifications, such as DNA methyltransferases (DNMTs), histone acetylases, and histone deacetylases [11].

DNA methylation reactions are part of one-carbon metabolism, a biochemical pathway that provides methyl groups for nucleotide synthesis and methylation reactions [29]. This cycle involves two methylation pathways: one folate-dependent and the other folate-independent, in which methionine is converted into S-adenosylmethionine (SAM), the donor of methyl groups. DNMTs transfer methyl groups from SAM to the carbon-5 position of cytosine bases, generating 5-methylcytosine [30, 31]. After this process, SAM is converted to S-adenosylhomocysteine [32]. Methyl groups derived from folate and choline are then used for the remethylation of homocysteine to produce methionine [32, 33]. This cycle relies on many enzymes and dietary micronutrients as cofactors, including the availability of folate, choline, and betaine through the diet [30]. Perturbations in one-carbon metabolism due to low intake of methyl-donor nutrients, malabsorption of these nutrients via disease or cellular conditions, and common polymorphisms in genes encoding important enzymes may alter the normal function and impact downstream biological processes, including DNA methylation [34, 35]. Deficiencies in certain nutrients (B vitamins, methionine, and/or choline) can alter SAM and S-adenosylhomocysteine levels, consequently affecting the DNA methylation process [11, 27]. Conversely, methyl-donor supplementation may modulate DNA methylation on a genome-wide level [36, 37].

In addition to the role in DNA methylation reactions, methyl-donors’ nutrients may also contribute to modify patterns of histone methylation through the provision of SAM and by the regulation of HMTs activity [38]. Beyond the bioavailability of substrates necessary for proper DNA methylation, nutritional factors can modulate the enzymatic activity of DNMTs and other enzymes involved in the methionine and folate cycles, such as B vitamins (B12), methionine, and choline [38], as well as C vitamin [39], folic acid, curcumin, resveratrol, and epigallocatechin-3-gallate (EGCG) [40‒42]. Recent studies have revealed that curcumin, a bioactive compound present in turmeric, can modulate the activities of DNMTs, thereby affecting DNA methylation [43]. Similarly, resveratrol, found in grapes and other plants, has demonstrated effects on DNMTs [44], by changing DNMT transcriptional pattern and, also, preventing the decrement of DNMTs due to an improvement in oxidative stress and inflammation [45, 46]. EGCG, a component of green tea, is also capable of influencing the activities of these enzymes [47]. In the one-carbon system, DNMTs require SAM as a cofactor for their full activation, and then, the intracellular concentration of SAM modulates DNMT activity [38]. Vitamin B6 controls the activity of serine hydroxymethyl-transferase which regulates the conversion of folic acid into 5,10-methylenetetrahydrofolate. Additionally, vitamin B12 acts as a cofactor for 5-methyltetrahydrofolate-homocysteine methyltransferase, which catalyzes the conversion of homocysteine into methionine [1].

Some studies have shown that a diet with low concentrations of folate is associated with leukocyte DNA hypomethylation [48, 49]. However, the evidence concerning the effects of folate supplementation on the DNA methylation profile remains controversial [50‒59]. Some authors have reported an increase in DNA methylation levels after supplementation [50, 58], while no changes are also founded [51‒53, 57, 60] (shown in Table 1). A recent meta-analysis on the effects of methyl-donor nutrient supplementation on DNA methylation levels indicated that folic acid supplementation did not significantly alter DNA methylation levels compared to a placebo [61].

Increasing evidence suggests that the expression of ncRNAs, as well as their downstream targets can be modulated by the intake of specific nutrients and natural compounds [66‒70, 73‒75] and, consequently, influence processes involved in the onset and progression of metabolic disease states (shown in Table 1). Vitamins have recently emerged as major epigenetic regulators, with evidence indicating that several vitamins regulate the expression of various ncRNAs. For example, vitamin A has been shown to regulate the expression of numerous miRNAs during embryonic development in both normal and neoplastic cells [76].

Increasing in vitro and in vivo evidence suggests that bioactive food compounds, such as resveratrol, can directly target several ncRNAs in hepatic cells [77], kidney tissues [78], and preadipocytes [79]. Resveratrol, a component of plant-based foods especially red wine, has garnered significant attentions because of its established anti-inflammatory and antioxidant effects [80]. It appears to modulate several miRNAs involved in inflammatory pathways, including the pro-inflammatory miR-155, the anti-inflammatory miR-663, and the oncogenic miR-21 [81]. Similarly, proanthocyanidins have been shown to suppress miR-33a/b and miR-122. MiR-122 is strongly associated with the risk of developing metabolic syndrome and type 2 diabetes [81, 82], whereas the downregulation of hepatic miR-33 improves metabolic homeostasis and liver function [83]. In addition, resveratrol, and epigallocatechin gallate modulated miR-33 and miR-122 concentrations by direct binding, and the effect on miRNA expression was structure-dependent [80]. The specific binding of polyphenols to miRNAs represents a new post-transcriptional mechanism by which polyphenols can modulate metabolism. Over 100 miRNAs have been modulated by polyphenols, with this interaction justified by the molecular structure of the polyphenol determining the nature of its association with certain miRNAs [84] Furthermore, EGCG treatment has been found to modulate the expression of 61 miRNAs in human hepatocellular carcinoma HepG2 cells [85]. Curcumin can modulate histone deacetylases and acetyltransferases, DNMT I, and miRNA expression [86].

Furthermore, supplementation with extra-virgin olive oil and its phenolic compounds could induce different modifications in DNA methylation and miRNAs expression, which can be related to its cardiovascular effects [87]. For example, after a Mediterranean diet supplemented with extra-virgin olive oil, some authors observed that the association between miR410 and stroke incidence could be modulated by extra-virgin olive oil consumption [88]. Other study evaluating the effects of 1-year adherence to different diets evidenced 413 differentially expressed circulating exosomal lncRNAs (118 upregulated and 295 downregulated) after a Mediterranean-based diet enriched in extra-virgin olive oil [89].

Human diseases have been a significant concern since the origin of humanity. Understanding the causes, treatments, and cures for various diseases is critical. With the advent and advancement of technology, studying disease mechanisms, progression, and therapies at the molecular level has become more rational and reliable compared to macroscopic approaches. Epigenetic mechanisms have now emerged as indispensable factors in the diagnosis, development, and therapy of several diseases, including metabolic disorders [90, 91].

In recent years, numerous academic publications have highlighted the contribution of DNA methylation levels to metabolic disorders. Many studies have demonstrated associations between DNA methylation and BMI, obesity, abdominal adiposity [92, 93], and components of metabolic syndrome [94, 95], as well as, with results from weight loss interventions [96, 97]. More specifically, results from an epigenome-wide association study (EWAS) with 8,165 participants and data from 6 independent cohorts, 2 case-control, and 8 retrospective studies, evidenced possible association among DNA methylation levels in blood cells and adipose tissue and body adiposity [23]. In fact, changes in specific CpGs methylation levels in organs and tissues may be related to the pathogenesis of metabolic disorders. In patients with obesity, DNA methylation profile in leukocytes is altered and associated with metabolic disturbances [98]. Additionally, patients with type 2 diabetes showed 77 differentially methylated regions, most of them hypomethylated, when compared with the control groups [6].

Furthermore, the circulating miRNAs have emerged as potential biomarkers of diseases because they are protected from RNAs in the extracellular environment, especially when associated with microvesicles, exosomes, or protein/lipoprotein complexes, thus participating in cell-to-cell communication. In addition, miRNAs are highly stable at room temperature and during repeated freeze-thaw cycles [99]. The levels of circulating miRNAs generally reflect a systemic response to an external stimulus; therefore, the expression levels of these molecules can be used as non-invasive diagnostic, prognostic, and therapeutic biomarkers of human diseases [100, 101].

One of the main sources of circulating miRNAs associated with metabolic disorders, especially obesity, is the adipose tissue [102‒104]. In patients with lipodystrophy, the circulating levels of miRNAs are significantly decreased compared to healthy individuals [104]. In the same way, studies have suggested that there is a relationship between the circulating miRNAs expression and degrees of obesity [105]. In this context, patients with body mass index (BMI) ≥40 kg/m² showed higher expression levels of miR-222, miR-142-3p and miR-140-5p and decreased levels of miR-221, miR-15a, miR-520c-3p, miR-130b, and miR-423-5p compared to subjects with BMI <40 kg/m². Furthermore, the combination of elevated levels of miR-423-5p, miR-520c-3p, and miR-15a in the plasma of men were able to predict degree obesity with 93.5% certainty. In addition, the expression of miR-520c-3p was negatively associated with BMI, fat mass, waist circumference, blood glucose levels, HbA1c glycated hemoglobin levels, and blood lymphocyte count [105].

Although most studies have focused on the expression of miRNAs, increasing evidence has also associated lncRNAs with metabolic diseases. lncRNAs are a class of non-coding RNAs comprising sequences longer than 200 nucleotides, with estimates suggesting there are between 16,000 and 100,000 lncRNAs in the human genome [18]. They can interact with DNA, RNA, and proteins, and possess a variety of functions, including chromatin regulation, nuclear organization, assembly and function of nuclear bodies, stability or translation of miRNAs, and regulation of signal transduction pathways [106]. Using the approach of meta-analysis of public datasets, several brown adipose tissue (BAT) and white adipose tissue (WAT)-specific lncRNAs have been identified [107, 108], and suggested as having potential role in the pathophysiology of metabolic diseases. The most prominent lncRNAs is lnc-Lep, which is transcribed from an enhancer region upstream of leptin [109]. Functional studies indicate that lnc-Lep is essential for adipogenesis and required for the maintenance of adipose leptin expression. Importantly, large-scale genetic studies of humans reveal a significant association of single-nucleotide polymorphisms in the region of human lnc-Lep with lower plasma leptin levels and obesity [110]. Interestingly, lncRNAs with specific expression on BAT are called lncBATEs. Among these groups of lncRNAs, the lncBATE1 and lncBATE10 are in an intergenic locus targeted by CCAAT/enhancer binding protein a C/EBPa, C/EBPb, and PPARc [107, 108]. Knockdown of lncBATE1 or lncBATE10 results in significant downregulation of BAT marker genes including CIDEA, C/EBPb, PGC-1a, PRDM16, PPARa, and UCP1, with reduced mitochondrial content and oxygen consumption, indicating impaired BAT thermogenesis [107, 108]. Moreover, these two lncBATEs are also required for browning of iWAT [107, 108].

Emerging evidence suggests that circRNAs may play an important role in regulating adipose function and energy metabolism. Differentially expressed circRNAs in the adipose tissue from subjects with obesity and lean individuals have been reported using circRNA microarrays. Among these, circSAMD4A can act as a miRNA sponge by interacting with miR-138-5p. Knockdown of circSAMD4A inhibits adipocyte differentiation [111]. Several other circRNAs such as ciRS-133 (circRNA sponge for miR-133) and circNrxn2 (miR-103 sponge) may promote WAT browning [112]. However, not all circRNAs regulate adipose function through miRNA sponges. Deep sequencing of visceral and subcutaneous fat identifies thousands of adipose circRNAs, many of which are dynamically regulated during adipogenesis and obesity. Among the regulated circRNAs, circArhgap5-2 is required for adipogenesis not through sponging miRNAs [113].

Epigenetic processes may contribute to an increase in the susceptibility for metabolic disorders. Despite the role of specific nutrients on epigenome, some evidence demonstrated that dietary patterns and lifestyle interventions may promote changes in DNA methylation levels [38] (Table 2). In this sense, modifications of the DNA methylation in genes related to some diseases could contribute to changes in gene expression and disrupt the function of important metabolic pathways [23]. However, epigenetic modifications induced by diet and behaviors may occur when dietary interventions take place over a long period of time (ranging from months to years in humans) and, when there is a transition from a previous diet to a new type of diet [38]. For example, the introduction of a calorie-restricted diet can alter the levels of substrates and cofactors involved in DNA methylation reactions (e.g., SAM, methionine, and cysteine) [114], by modulating the activity of enzymes involved in the methionine and folate cycles, such as DNMT1 [115] and DNMT3a [116]. Some disease-related physiological factors have been associated with the dysregulation of epigenetic machinery. For example, physiological factors related to obesity and low-grade chronic inflammation can induce changes in the epigenetic pattern [23]. Importantly, a modulation of these processes through diet and lifestyle interventions may prevent diseases and help in health maintenance [117].

Milagro et al. evaluated men before and after an 8-week energy-restricted diet (30% energy restriction) and observed 170 CpG sites differentially methylated after intervention. Indeed, the authors suggested that DNA methylation differences may be used as biomarkers for the effects of weight loss programs [118]. In agreement, other authors suggested that lifestyle modifications may modify global DNA methylation of overweight and sedentary female breast cancer survivors [121]. On the other hand, no significant difference in Long Interspersed Nuclear Elements (LINE-1) methylation levels was observed after 12 months of a reduced-calorie weight loss diet, and exercise program [119]. Moreover, exploratory previous results showed that hypocaloric dietary intervention and bariatric surgery may modify DNA methylation profile of patients with severe obesity [126, 127].

Taken together, these data suggest that weight loss resulting from different interventions, such as energy-restricted diets (combined or not with exercise) and bariatric surgery, may lead to both decreases and increases in DNA methylation at global and gene-specific levels. These changes are primarily observed in genes associated with metabolic pathways, including those involved in carcinogenesis, inflammation, and insulin metabolism. Additionally, some studies have identified no significant effects on DNA methylation.

As previously described, growing evidence have shown that circulating miRNAs can be regulated by nutrients or bioactive compounds, which could explain the changes in the expression pattern in some miRNAs in response to a specific diet [137]. In this context, results suggest that dietary patterns and weight loss interventions can modulate the expression of several miRNAs (shown in Table 2). Aiming to identify potential biomarkers for weight loss, differential miRNA expression in peripheral blood mononuclear cells (PBMCs) in high responders compared to low responders after an 8-week energy-restricted diet has been reported. Additionally, seven miRNAs (miR-34a, miR-208, miR-193a, miR-320, miR-433, miR-568, and miR-181a) may be potential biomarkers of beneficial effects of weight loss in women with obesity [22]. Similarly, weight loss induced by a high-protein diet can decrease miR-223 levels in men with obesity [131].

In this way, Milagro et al. [132], characterized the expression of miRNAs in PBMCs of women with obesity submitted to an energy-restricted diet for 8 weeks. After the intervention, the population was classified into responders (>5% loss of body mass) and non-responders (<5% loss of body mass). At baseline, miR-935 and miR-4772 were increased, while miR-223, miR-224, and miR-376b were decreased in the non-responder group compared with those who responded to diet-induced weight loss. Notably, miR-935, miR-4772, and miR-376b also showed a relevant association with magnitude of weight loss, making them valuable candidate biomarkers for weight loss and dietary response [132].

Another study evaluated miRNA alterations in white blood cells obtained from patients with metabolic syndrome who participated in the RESMENA nutritional trial. Subjects were classified as high responders and low responders when weight loss after 8 weeks was greater or less than 8%, respectively. Six miRNAs (miR-1237, miR-1976, miR-642, miR-636, miR-612, miR-193b) both hypomethylated and increased were identified in high responders compared to low responders [134]. The same authors also reported on the regulation of miR-548q and miR-1185-1 in high responders compared to low responders matched for sex and gender. In a functional assay, miR-548q and miR-1185-1 reduced glycogen synthase kinase-3B (GSK3B) gene expression [138]. In addition, the expression of miR-1185-1 was negatively correlated with serum levels of IL-6 [134].

In addition, the expression profile of miRNAs may depend on the type of diet used during the weight loss intervention [135]. In this sense, among miRNAs differentially expressed in individuals with obesity compared to controls, seven miRNAs (miR-130a-3p, miR-142-5p, miR-144-5p, miR-15a-5p, miR-22-3p, miR-221-3p, and miR-29c-3p) were significantly associated with the response to a low-fat diet, being able to further categorize responders from non-responders to this intervention [135].

In addition to dietary interventions, weight loss achieved with exercise can also influence the expression of miRNAs [139]. In a subgroup of obese patients, a 3-month individualized intervention that included aerobic and resistance training sessions decreased plasma levels of miR-146a-5p [140]. Likewise, a 16-week therapeutic exercise intervention modified the circulating levels of miR-192 and miR-193b, both specifically expressed in the pre-diabetic state relative to healthy people and patients with type 2 diabetes [141]. In addition, miR-935 expression was higher in low responders compared to high responders at a 16-week intervention, aimed at generating a 500-kcal (∼250-kcal diet-induced energy deficit and ∼250 kcal induced by daily exercise) [136].

Taken together, these data suggest that dietary restriction and exercise are associated with precisely increased or decreased levels of specific circulating miRNAs associated with obesity. However, it is unclear whether these changes in miRNA levels are a cause or a consequence of improvements in metabolism.

Metabolic disorders comprise dysfunction in multiple specific cell types and tissues with a wide range of associated etiological factors. The studies aiming to investigate epigenetic aspects related to metabolic disorders are in constant development and are very heterogeneous. In the current literature, there is a broad diversity of analyzed tissue samples, study designs, populations, experimental protocols, and other methodological aspects related to epigenetic mechanisms including targeted and/or genome-wide approaches and data analysis datasets. This heterogeneity poses significant challenges to the generalizability and applicability of research findings. Diverse methodologies can result in inconsistent results, making it difficult to draw definitive conclusions. Moreover, the lack of standardized protocols across studies can further exacerbate these inconsistencies, highlighting the need for more rigorous and harmonized approaches in research design to ensure more reliable and comparable outcomes. The study of the role of histones in metabolic disorders is still in its early stages, indicating a significant gap in our current understanding and therapeutic strategies. Therefore, it is imperative that future research focuses on elucidating these mechanisms to harness their therapeutic potential.

However, despite these challenges and barriers faced in the field of metabolic disorders epigenetics, the potential benefits of this research are immense and still largely untapped. Clinical applications of epigenetics in metabolic disorders are rapidly advancing, offering promising avenues for diagnosis, treatment, and prevention. By understanding the epigenetic modifications that contribute to these disorders, clinicians can develop targeted therapies that precisely address the underlying molecular mechanisms. Epigenetic biomarkers can aid in early diagnosis and prognosis, allowing for more personalized and effective treatment strategies. Additionally, epigenetic interventions hold the potential for reversing adverse metabolic states, ultimately improving patient outcomes and quality of life.

In conclusion, the rapid development of epigenomic technology and the expanding body of epigenomic data offer unique opportunities to elucidate the interplay between genetic, environmental, and epigenetic mechanisms in metabolic disorders. The reciprocal relationship between nutrients, dietary patterns, lifestyle, and the epigenome paves the way for nutritional interventions focused on DNA methylation and miRNA. Certain DNA methylation marks and miRNA expressions can help predict the risk of metabolic disorders and serve as biomarkers to forecast the success of therapies in the context of precision medicine and nutrition. This integration of epigenetics into clinical practice promises more personalized and effective strategies for preventing and managing metabolic disorders.

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this manuscript.

This study was supported by São Paulo Research Foundation (FAPESP) (Grant No. #2020/01893-2, #2020/15126-3, #2022/09301-2) and from “FACINGLCOVID-CM project, Funding REACT EU Program (Comunidad de Madrid and the European Regional Development Fund [ERDF], European Union)”. Also, Metacategorización personalizada de procesos inflamatorios asociados a síndrome metabólico, enfermedades autoinmunes y virales para medicina de precisión (METAINFLAMACIÓN, Ref: Y2020/BIO-6600) is gratefully acknowledged.

Carolina Ferreira Nicoletti and Tais Silveira Assman conceived the idea for the manuscript. Carolina Ferreira Nicoletti, Tais Silveira Assman, and Leticia Lobato Souza performed the literature review and wrote the manuscript. José Alfredo Martinez edited and critically reviewed the manuscript. All authors approved the final version of the manuscript.