Epigenetic regulation in MASLD Exploring the role of DNA methylation

In this thesis we focused on the role of DNA methylation in Metabolic Dysfunction- associated Steatotic Liver Disease (MASLD) and the relationship with the intestinal microbiome and its produced metabolites. We started by reviewing epigenetic mechanisms in metabolic diseases, followed by a review of the relevant aspects of the main cause of death in MASLD: atherosclerotic cardiovascular disease, asCVD. We analyzed DNA methylation after donor Fecal Microbiota Transplant (FMT) in patients with metabolic syndrome and in patients with MASLD, systematically reviewed hepatic DNA methylation in MASLD and Alcoholic Liver Disease (ALD) and concluded by comparing DNA methylation at different disease stages of MASLD.
In Chapter 2, we reviewed epigenetic mechanisms in various diseases such as obesity, insulin resistance/metabolic syndrome, type 2 diabetes and MASLD. We reviewed the current literature on the use of surrogate tissues, such as blood, to study DNA methylation and gain insights into the processes occurring in adipose tissue. Concordance was found in the methylation of genes related to basic cellular functions and discordance, with high variation, in genes related to specific tissue functions. Subcutaneous adipose tissue was found to be more hypomethylated than cells in blood. One of the studies on white adipose tissue (WAT) and brown adipose tissue found a histone repressor marker related to the differentiation WAT-BAT. When specific methylation was compared between omental and subcutaneous adipose tissue, it was noted that the PPARG promoter region is hypermethylated in omental tissue, accompanied by a decrease in gene expression compared to subcutaneous adipose tissue. We review the role of metabolites derived from the intestinal microbiota, such as butyrate - a short-chain fatty acid involved in epigenetic processes - which has been shown to inhibit histone deacetylase 1 (HDAC1), which in turn is related to a decrease in the activity of inflammatory pathways. Furthermore, on the interaction of microbiome and epigenetic alteration, we mention probiotic supplementation with Lactobacillus rhamnosus GG and Bifidobacterium lactis Bb12, as well as DNA methylation of the promoter region of genes associated with obesity, FTO and MC4R. Next, we review the relationship between genetics and epigenetics in MASLD, such as the hypermethylation of the PNPLA3 regulatory region accompanied by lower mRNA levels in patients with MASLD in advanced stages. We highlight the cross-communication between epigenetic mechanisms in type 2 diabetes, where there is an association between specific histone modifications and hypomethylated regions in the genome.
When reading epigenetic studies, it is necessary to be aware of certain points, such as evaluating the genomic context in which methylation occurs, for example, the importance of DNA methylation in the promoter region or in repeated DNA sequences, where the effect can be to up-regulate or down-regulate, respectively, gene expression. Also important is to appreciate used techniques estimating degree of methylation and correlate these findings with the gene transcriptome and analyzing the possible implications of using surrogate tissue and cross-communication between epigenetic mechanisms.
In Chapter 3, we reviewed the correlation between MASLD and the leading cause of death: asCVD. We noted that there is a greater risk of asCVD in in advanced stages of MASLD compared to simple steatosis. We analyzed markers of atherosclerosis in patients with MASLD, such as the coronary artery calcium (CAC) score, where we observed a faster progression in patients with MASLD than in those without it. As for carotid intima-media thickness (cIMT) ultrasound, several studies have shown a higher incidence of pathological cIMT in patients with MASLD compared to those without it. One of the studies reviewed suggests the need to use a specific cardiovascular risk score for MASLD. We studied metabolic factors in MASLD that may influence atherosclerotic cardiovascular disease, such as mixed hyperlipidemia and hypercoagulable state due to increased hepatic secretion of procoagulant factors such as fibrinogen and plasminogen activator inhibitor 1. We draw attention to factors that can influence both phenotypes, such as insulin resistance/metabolic syndrome, hypertension, obstructive sleep apnea, intestinal microbiota dysbiosis and chronic low-grade inflammation. We summarized the drugs being studied for the treatment of MASLD: GLP1-RA, TZD, PPAR agonist, THR-β, FXR agonist, FGF19 analogue and C-C chemokine receptor type 2 inhibitor. We conclude by emphasizing the importance of raising awareness about the co-occurrence of these two diseases.
In Chapter 4, we compare the effect of two types of FMT: allogenic (donor) and autologous (own), on the intestinal microbiota composition and function as well as changes in , produced plasma metabolites and DNA methylation of peripheral blood mononuclear cells in treatment naïve individuals with metabolic syndrome.T he analyses were conducted at the beginning of the study and after 6 weeks. We studied insulin sensitivity and applied machine learning methodology to identify the bacteria, metabolites and DNA methylation that differentiate the two groups studied. We found that there was a greater change in the intestinal microbiota in the allogenic group than in the autologous group, with the interesting observation that the introduction of Prevotella ASVs is correlated with the methylation of AFAP1 - a gene involved in mitochondrial function - and is also correlated with insulin sensitivity.
In Chapter 5, we evaluated the effect of autologous versus vegan allogenic FMT on hepatic DNA methylation, gutmicrobiome composition and function, as well as plasma metabolome of patients with MASLD. We noticed a significant increase in E. siraeum and B. wexlerae after allogenic transplantation, both species previously studied in the context of MASLD, (childhood) obesity and type 2 diabetes, respectively. In the same group, we observed an increase in the liver metabolites phenylacetylglutamine (PAG) and phenylacetylcarnitine (PAC). This increase was related to the presence of the bacteria Gemmiger formicillis and Firmicutes bacterium_CAG-170. PAG has previously been associated with MASLD, and PAC is involved in the transport of fatty acids in mitochondria. FMT correlated with altered hepatic methylation of TARS, RPTOR and ZFP57. Multiple CpGs were differentially methylated in ZFP57, an unusual finding in previous studies that appears to be a response to signals derived from the gut microbiota. We conclude that manipulation of the microbiome is related to altered metabolites and hepatic DNA methylation in individuals with MASLD.
In Chapter 6, we conducted a systematic review of liver methylation in humans with MASLD and ALD, as well as an analysis of metabolic pathways. We included 15 articles, and it was noteworthy that all of them presented a moderate to high risk of bias. Among the pathways common to both phenotypes, related to genes with differentiated hepatic methylation, we highlight oncogenic pathways, PPAR signaling pathways and senescence. In MASLD, phospholipase D, the insulin signaling pathway and, in ALD, the MAPK and apelin signaling pathways appear individually. Considering that several genes related to oncogenic pathways were differentially methylated, we suggest further studies to evaluate the possibility of using liver DNA methylation as a risk marker for cancer development.
In Chapter 7, we studied hepatic DNA methylation in liver biopsies from 59 obese patients divided into 3 groups according to liver histopathology: normal, metabolic dysfunction-associated steatosis of the liver (MASL) and metabolic dysfunction -associated steatohepatitis (MASH). In this study we used two different and complementary methods to assess DNA methylation: DMR analysis and machine learning and correlated our findings with the transcriptome. We found a DMR in Pyruvate carboxylase, with hypermethylation of 4 CpGs accompanied by decreased gene expression in the MASH group compared to the other two groups. This finding contributes to the notion that there is a moment in the progression from steatosis to fibrosis when the mitochondria can no longer adapt to the increased influx of acetyl-CoA, and the citric acid cycle can no longer maintain its rhythm. The decrease in PC activity possibly reflects a reduction in its anaplerotic capacity to supply TCA with oxaloacetate in the MASL to MASH transition. In this study, we also found hypomethylation of EMP1 in MASH compared to MASL, with corresponding changes in the transcriptome. EMP1 appears to be related to fibrosis and hepatic stellate cells, as well as tumor progression in several types of cancer. Its ablation in mice prevented the occurrence of colorectal cancer metastases.
Our conclusion is that the study of DNA methylation in MASLD is still at an early stage, and major steps are needed in the appropriate design of studies, standardization of methods and analysis of results to obtain valid and clinically meaningful results. Despite these challenges, DNA methylation is proving to be an interesting tool as 1. an indicator of new pathological pathways not yet fully studied in MASLD, for example mitochondrial dysfunction; 2. as a possible risk marker for the development of cancer; and 3. as an additional tool for individualizing treatment when altered methylation of specific pathways are identified in each patient.