Can DNA Methylation Predict the Lipid Response to Dietary Intervention? There's a Fat Chance

High circulating triglycerides (TG), carried by triglyceride-rich-lipoproteins (TRL), are associated with insulin resistance, obesity, metabolic syndrome, and type 2 diabetes and are considered a major risk factor for cardiovascular disease. Although strategies that lower TG and TRL to improve cardiometabolic outcomes are of major interest, there are still many unknowns that impede therapeutic development and thwart therapeutic success. For example, what increases the susceptibility of certain individuals to unfavorable lipid profiles? And why are some study participants more responsive to diet and lifestyle interventions? While considerable efforts have been made to identify genetic variants that may account for metabolic changes across individuals, variants with functional significance are rare, and those that have been identified do not fully explain the high variability of lipid profiles. Hence, attention has been refocused towards the epigenetic regulation—highly dynamic and heritable changes in gene function without changes in DNA sequence—of genes involved in lipid and lipoprotein metabolism. While many studies have linked altered epigenetic processing of key metabolic genes to distinct metabolic phenotypes [reviewed in (1)], whether specific epigenetic marks can predict the success of lifestyle interventions remains unknown. In the present study, Li et al show that study participants with higher levels of DNA methylation (DNAm) in the first intron of carnitine palmitoyltransferase 1 A (CPT1A)—the rate limiting enzyme in mitochondrial fatty acid uptake and oxidation—show greater improvements in TG and TRL when adhering to a low-fat weight-loss diet (2). Therefore, this study reports an important observation for the first time and directly links altered epigenetic processing of a key lipid metabolism gene to favorable lipid profiles following diet modification.

DNAm of cytosine residues within CpG islands in promoter regions and other cis regulatory DNA elements is a highly dynamic epigenetic mark associated with chromatin remodeling and gene transcription. In the present study, Li et al were motivated by recent findings from epigenome-wide associations studies linking increased levels of CPT1A DNAm to reduced CPT1A expression and increased circulating TG and TRL (3). In a separate study, dietary fat appears to modify CPT1A DNAm (4), leading the authors to hypothesize that adjusting dietary fat intake could alter the relationship between CPT1A DNAm and circulating lipid profiles. This study leveraged the POUNDS lost trial: a 2-year randomized clinical trial comparing the ability of 4 diets with varying macronutrient content to promote weight loss. In brief, the study involved 811 overweight adults that were assigned to either (1) 20% fat, 15% protein, 65% carbohydrate; (2) 20% fat, 25% protein, 55% carbohydrate; (3) 40% fat, 15% protein, 45% carbohydrate; or (4) 40% fat, 25% protein, 35% carbohydrate (5). DNAm was assessed in 538 study participants by high-resolution methyl-capture sequencing and presented as percent of methylation over a given CpG containing region. Importantly, standard (eg, TG, total cholesterol, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol) and specialized (eg, very low density lipoprotein with or without apolipoprotein C-III) lipids were measured. Notably, the authors found that individuals with a higher baseline level of regional DNAm at CPT1A had lower circulating TG, consistent with previous studies (3). In addition, the authors report for the first time that higher CPT1A DNAm was associated with even greater reductions in total TG and TRL, including very low density lipoprotein particles without apolipoprotein C-III. Although the data presented do not probe the molecular mechanisms leading to these observations, the findings show for the first time that baseline methylation levels of distinct regions of a lipid processing gene can predict changes in the metabolic parameters following dietary intervention. Since CPT1A is primarily involved in fatty acid catabolism, Li et al's findings suggest that reduced dietary lipids alter the transcriptional regulation of CPT1A to increase fatty acid oxidation, leading to reduced circulating lipids within their associated lipoprotein class.

DNAm is one of the best-characterized, yet controversial, epigenetic marks. Although the prevailing dogma links DNAm to reduced transcription factor (TF) binding and transcriptional repression, with the advent of more sensitive techniques it has become clear that this model needs to be revisited. Recent models suggest that (1) TFs bind to methylated CpGs to activate transcription, (2) DNAm is a consequence of TF binding, and (3) in some cases DNAm does not influence TF binding [reviewed in (6)]. Therefore, the outcomes of DNAm are largely dependent on the specific TFs and baseline transcriptional state within a cell. In the present study, the authors show that higher levels of DNAm within intron 1 of CPT1A are associated with lower circulating lipids, which suggests that CPT1A expression, and hence lipid catabolism, is increased. While this directionality is plausible, given our evolving understanding of the dynamic role of DNAm in transcriptional regulation, our interpretation of the study is limited due to the lack of CPT1A expression data. In addition, the dynamic methylation within intron 1 of the CPT1A gene suggests that this region is actively modified in response to nutrient, and specifically dietary lipid, status, but why remains unclear. Since intron 1 contains putative lipogenic TF binding sites (eg, SREBP1c), future studies aimed at determining the effect of DNAm on the binding of specific TFs within the CPT1A gene would be particularly informative. Comparison of chromatin structure and accessibility (eg, ATAC seq) within the CPT1A gene in response to dietary changes would be even more powerful and could highlight novel lipid-sensitive regulatory elements to dramatically transform our understanding of the nutrient–epigenome interface.

Overall, Li et al present intriguing findings that hold promise for the development of precision dietary intervention approaches. While further studies are needed to determine mechanism, evidently CPT1A is dynamically regulated in response to metabolic status, and its methylation could be used to assess which dietary interventions might have the most favorable effects on circulating lipids to mitigate the association with cardiometabolic disease in susceptible individuals.

Disclosures

The authors have no conflicts of interest or disclosures to declare.