EGCG Prevents High Fat Diet-Induced Changes in Gut Microbiota, Decreases of DNA Strand Breaks, and Changes in Expression and DNA Methylation of Dnmt1 and MLH1 in C57BL/6J Male Mice

Obesity as a multifactorial disorder involves low-grade inflammation, increased reactive oxygen species incidence, gut microbiota aberrations, and epigenetic consequences. Thus, prevention and therapies with epigenetic active antioxidants, (−)-Epigallocatechin-3-gallate (EGCG), are of increasing interest. DNA damage, DNA methylation and gene expression of DNA methyltransferase 1, interleukin 6, and MutL homologue 1 were analyzed in C57BL/6J male mice fed a high-fat diet (HFD) or a control diet (CD) with and without EGCG supplementation. Gut microbiota was analyzed with quantitative real-time polymerase chain reaction. An induction of DNA damage was observed, as a consequence of HFD-feeding, whereas EGCG supplementation decreased DNA damage. HFD-feeding induced a higher inflammatory status. Supplementation reversed these effects, resulting in tissue specific gene expression and methylation patterns of DNA methyltransferase 1 and MutL homologue 1. HFD feeding caused a significant lower bacterial abundance. The Firmicutes/Bacteroidetes ratio is significantly lower in HFD + EGCG but higher in CD + EGCG compared to control groups. The results demonstrate the impact of EGCG on the one hand on gut microbiota which together with dietary components affects host health. On the other hand effects may derive from antioxidative activities as well as epigenetic modifications observed on CpG methylation but also likely to include other epigenetic elements.


Introduction
Metabolic syndrome, a multifactorial disorder, results from a long-term imbalance of diet and physical activity, genetic predisposition, and an imbalanced gut microbiota influencing several metabolic pathways including epigenetic regulation. In 2015, the prevalence of metabolic syndrome was determined at 1.9 billion adults being overweight (BMI ≥ 25 kg/m 2 (body mass index)) and more than 600 billion obese (BMI ≥ 30 kg/m 2 ) [1]. Thus, this high incidence of obesity and associated diseases like type 2 diabetes (hyperglycemia on basis of an insulin resistance) are a challenge and financial burden for national health care systems.
Increased adipose tissue is an important energy storage but also a key endocrine organ with active metabolism relevant in energy homeostasis, lipid and glucose metabolism, fibrinolysis, coagulation, blood pressure, and inflammation (like interleukin-6 (IL-6), tumor necrosis factor (TNF-)) [2,3], but also in increased reactive oxygen species (ROS) and free fatty acids (FFAs) production, and increased oxidative 2 Oxidative Medicine and Cellular Longevity stress [4,5]. Oxidative stress in turn is associated with both genome-wide hypomethylation and promoter hypermethylation of the DNA [6], resulting in transcriptional silencing of key antioxidant enzymes as well as tumor suppressor genes [7,8]. Diets rich in bioactive anti-inflammatory compounds, such as polyphenols, have been recommended to reduce oxidative stress and to decrease inflammation [9].
According to Fang et al. (2003) MutL homologue 1 (MLH1) and O6-methylguanine-deoxyribonucleic acid methyltransferase (MGMT), both part of the DNA mismatch repair (MMR) system, are hypomethylated in the CpG rich promoter region [25] accompanied by a higher expression of mRNA [25]. In contrast, ROS-induced oxidative stress contributes to hypermethylation of normally unmethylated promoter regions, resulting in transcriptional silencing of key antioxidant enzymes as well as tumor suppressor genes [7,8]. EGCG also alters histone acetyl transferase (HAT) activity, resulting in aberrant chromatin structures, as well as miRNA expression in hepatocellular carcinoma cells [14].
However, above small-intestinal absorption, tea catechins could reach the large intestine and be processed by gut microbiota into gallic acid and epigallocatechin (EGC) [26,27]. Alterations in gut microbiota composition and differences in gut microbial metabolite profile due to different dietary feeding offer insights that may be relevant for several chronic conditions, including obesity. The obesity related gut microbiota is composed by a less bacterial diversity and altered abundance, gene-representation, and metabolic pathways [28]. These differences include a higher abundance of Firmicutes, whereas Bacteroidetes are reduced resulting in a higher Firmicutes/Bacteroidetes ratio in obese individuals.
An increased abundance of Lactobacilli is mentioned as a growth promoter and is associated with weight gain and inflammatory processes during obesity [29]. Genes encoding for carbohydrate metabolism enzymes are increased in the gut microbiome of obese mice, provoking a greater capacity to extract energy from the diet and to generate short-chain fatty acids (SCFAs) [30,31]. SCFAs are essential for the microbial community and play a role in regulation of energy balance, inflammatory processes, health, and obesity [32].
Meanwhile interventions with EGCG resulted in a reduction of Clostridium spp. abundance, increased Bacteroides, but also influenced Bifidobacterium and Prevotella, to a lesser extent resulting in lower levels of acetic and butyric acids and little influence on propionic acid levels in caecum. Thus, effects on weight reduction and weight maintenance due to dietary intervention with EGCG could be responsible for the regulation of energy metabolism in the body [26].
In the present study, we investigated the effects of a physiologically applicable dose of EGCG on gut microbiota, DNA damage, DNA methylation, and gene expression of inflammatory mediators: IL-6; Dnmt1, and DNA repair: MLH1 in liver and colon due to metabolic syndrome induced by a high fat diet (HFD) and due to control diet (CD) with a focus on associations between alteration of DNA repair processes, immune functions, and genomic instability. We investigated the colon as direct nutrient contact is given, but also hepatic tissue, the main organ in glucose and insulin metabolism. In addition, gut microbiota composition and diversity were analyzed on the basis of stool samples. Effects of dietary EGCG in alleviating conditions associated with obesity and metabolic syndromes are reported.

Single-Cell
Gel Electrophoresis (SCGE) Assay. DNA migration in an electric field was analyzed in hepatocytes and  colonocytes from mice in SCGE assays (comet assay) [33,34]. This approach has the advantage that measurement can be conducted with variety of organs [35] and is increasingly used in genetic toxicology Collins, (2015) [36]. A typical picture from a "comet" as marker for DNA-instability is shown in (Figure 1). Nuclei from livers and cells from colons were collected according to the method developed by   [37]. The procedure for the measurement of DNA migration in liver cells and colonocytes was employed by us in a number of earlier studies [38][39][40][41][42]. 1.0 g liver tissue was homogenized by use of a Potter Elvehjem-type (B. Braun, Melsungen, Germany) at 400 rpm in 4.0 mL chilled homogenization buffer (pH 7.5). Subsequently, the homogenates were centrifuged (800g, 10 min, 4 ∘ C). Colon cells were isolated by scratching mucosa from the colons and were kept on ice in 2.0 mL homogenization buffer. The nuclei were suspended in LMPA (0.5%, Gibco, Paisley, UK) and transferred to slides which were precoated with NMPA (1.0%, Gibco, Paisley, UK).
With nuclei from each organ, three slides were prepared per experimental point and 50 cells were evaluated from each slide. Slides were examined under a fluorescence microscope (Nikon EFD-3, Japan) using 25-fold magnification. DNA migration was determined with a computer aided comet assay image analysis system (Comet Assay IV, Perceptive Instruments, UK).

Gene Expression Analysis.
Liver and colon samples were stored at −80 ∘ C until RNA and DNA isolation using the All-Prep DNA/RNA/miRNA Universal Kit (Qiagen, Germany) according to the manufacturer's protocol. Concentration, respectively, purity was controlled with a Picodrop100 (Picodrop, UK). Complementary DNA (cDNA) was synthesized by reverse transcription using RT 2 First Strand Kit (Qiagen, Germany). cDNA was analyzed with real-time polymerase chain reaction (PCR) using qPCR Primer Assays (Qiagen, Germany) and RT 2 SYBR Green Master Mix (Qiagen, Germany) according to manufacturer's protocol. PCR conditions were as follows: initial step of 95 ∘ C for 10 min, followed by 40 cycles of 95 ∘ C for 15 s and 60 ∘ C for 1 min, ending with melting curve analysis (gradient melting of the products was performed at 0.5 ∘ C/10 s from 65 ∘ C to 95 ∘ C). Each sample was analyzed in duplicate with normalization to the housekeeping gene glyceraldehyde-3-phosphate-dehydrogenase (GAPDH).

Methylation Analysis.
Genomic DNA was bisulfite converted with EpiTect5 Fast Bisulfite Conversion kit (Qiagen, Germany) and amplified by PCR using the PyroMark PCR Kit (Qiagen, Germany) according to manufacturer's instructions with primers for Dnmt1 and MLH1 designed by PyroMark Assay Design SW 2.0 Software ( Table 2).
The PCR was carried out in a total reaction volume of 25.0 L, containing 12.5 L PyroMark 2x PCR master mix, 10 pmol (Dnmt1) or 7 pmol (MLH1) of each primer, 2.5 L CoralLoad Concentrate 10x (Qiagen, Germany), and 10.0 ng (Dnmt1) or 15.0 ng (MLH1) bisulfite converted DNA. Thermocycling started with initial denaturation at 95 ∘ C for 15 min, followed by 45 cycles at 94 ∘ C for 30 s, 55.5 ∘ C for 45 s, 72 ∘ C for 45 s, and a final extension at 72 ∘ C for 10 min. PCR product quality was investigated with agarose gel electrophoresis. CpG methylation analysis was performed in a PyroMark Q24 MDx (Qiagen, Germany).

Gut Microbiota Analysis.
Stool was collected before intervention and continuously after 1 month until the end of the study period and stored at −20 ∘ C until microbial DNA extraction by using the QIAamp5 Fast DNA Mini Kit (Qiagen, Germany) following the manufacturer's protocol including two steps of 45 sec beadbeating at 4000 rpm with a 60 sec break in between to increase the DNA yield. DNA concentration and purification were measured using Pico100 (Picodrop Ltd., Cambridge, UK).
The abundance of gut microbial subgroups was determined by 16s rDNA using quantitative real-time PCR with SYBR green or TaqMan-Probe Master Mix with specific primer pairs (Tables 3 and 4) in a Rotor Gene 3000 (Corbett Life Science, Australia). The PCR reaction mixtures and serial DNA dilution of typical strains were prepared according to Remely et al. (2013) [44].

Statistical Analyses.
For statistical analyses of SCGE assay results GraphPad Prism 5.02 (GraphPad Software, USA) was used. The means and standard deviation (SD) of % DNA     in the comet tails of the nuclei from the different treatment groups were calculated. Comparisons of groups were done with Student's t-test based on the means of three slides/animal. All statistical analyses of gene expression and methylation analysis were performed using IBM SPSS Advanced Statistics 20.0 (SPSS, USA). All data are shown means ± SD. ΔCT values were calculated by normalization to GAPDH (ΔCT = CT-Target − CT-GAPDH). The ΔΔCT value shows the difference between the two groups (ΔΔCT = ΔCT + EGCG − ΔCT-Control). Relative changes in gene expression between the intervention and control group are determined by the 2 −ΔΔCT equation (fold change = 2 −ΔΔCT ). Kolmogorov-Smirnov-Test was used to test the normalization of the data. The Mann-Whitney -Test was used to examine significant relationships. For all comparisons values ≤ 0.05 were considered as statistically significant.

SCGE Experiments with Nuclei from Colon and Liver
Cells. HFD induced significant DNA damage in liver and colon compared with CD. In liver of animals fed with HFD and supplemented with EGCG in drinking water the extent of DNA migration was significantly decreased by 31.5% ( ≤ 0.05) after supplementation (Figure 3(a)). No effect was detected in colon of HFD group after EGCG supplementation Figure 3(b). Supplementation with EGCG in CD group in colon of animals caused slight DNA damage, while no effect was detected between CD and CD + EGCG in liver.

Relative Gene Expression and CpG Methylation of the Promoter Region of MLH1 in Liver and Colon Cells.
In liver cells the relative gene expression of MLH1 decreased significantly in HFD fed mice compared to the CD (49%, ≤ 0.001). EGCG supplementation showed significant reduction in CD + EGCG compared to CD (56%; ≤ 0.001) and in HFD + EGCG compared to HFD (44%; ≤ 0.01). Moreover the protective effect by EGCG was significantly higher in HFD + EGCG than in CD + EGCG (38%; ≤ 0.05) (Figure 4(b)). The relative gene expression of MLH1 in colon did not result in any significant difference between the interventions (Figure 4(a)).

Relative Gene Expression of IL-6 in Colon.
A significant lower expression of IL-6 has been measured between CD and HFD animals ( ≤ 0.01) and also in CD and HFD + EGCG mice ( ≤ 0.01) in the colon (Figure 7). In liver the gene expression of IL-6 was under detection limit in the study group.

Relative Gene Expression and CpG Methylation of the Promoter Region of Dnmt1 in Liver and Colon Cells.
Relative gene expression of Dnmt1 in liver cells was lower in HFD compared to CD (61%; ≤ 0.01). Supplementation with EGCG resulted in a significantly reduced gene expression compared to, respectively, CD (75%; ≤ 0.01) and HFD (51%; ≤ 0.01) (Figure 8(b)).
In colon the relative gene expression of Dnmt1 decreased significantly in HFD compared to CD (69%, ≤ 0.01) although the decrease of HFD compared to CD is compensated by EGCG in HFD with a more than three times higher gene expression of Dnmt1 compared to HFD ( ≤ 0.01, Figure 8(a)).
In colon cells CpG2 and CpG4 showed significant changes in the methylation status of Dnmt1. A significant higher methylation status in CpG2 was observed due to EGCG in the HFD group (19%; ≤ 0.05). In CD + EGCG supplementation resulted in a significant hypermethylation at CpG4 (28%; ≤ 0.05). No significant changes were observed at methylation status of Dnmt1 at CpG1 and CpG3 (Figure 9(a)).
The Firmicutes/Bacteroidetes ratio is significantly higher in both HFD groups compared to CD groups ( < 0.0001). EGCG treatment induced a significantly lower ratio in HFD + EGCG compared to HFD ( < 0.0001) but a significantly higher ratio in CD + EGCG compared to CD ( < 0.0001; Figure 10(b)). Lactobacilli decreased with EGCG intervention. Both clostridial clusters (Clostridium cluster IV, Clostridium cluster XIVa) were significantly lower in HFD groups compared to CD groups ( < 0.0001). Clostridium cluster IV significantly increased in HFD + EGCG compared to HFD ( = 0.005). Clostridium cluster XIVa increased in CD + EGCG compared to CD ( = 0.189) and was significantly more abundant in comparison to HFD + EGCG ( < 0.0001). In turn, F. prausnitzii was less abundant in HFD groups compared to CD groups ( < 0.0001) and was more abundant in the CD groups compared to all other groups (CD + EGCG : HFD + EGCG : < 0.0001; CD : CD + EGCG: = 0.001; Figure 10(c)).
In HFD mice, butyryl CoA: acetate CoA-transferase gene was significantly lower in comparison to CD mice ( < 0.0001). Intervention with EGCG in the CD group resulted in a significant decrease ( < 0.0001) whereas with HFD no significant effect was shown. Results of the butyrate kinase gene showed similar results whereas HFD + EGCG group showed a significant increase (CD : HFD < 0.0001; CD : CD + EGCG = 0,001; HFD : HFD + EGCG = 0.005).
Bacteroidetes were significantly lower abundant in HFD compared to CD ( < 0.0001) and significantly increased with EGCG intervention in HFD mice ( = 0,001). An increase was shown in CD mice due to EGCG treatment.
Akkermansia showed a lower abundance in HFD fed mice compared to CD ( = 0.092). EGCG treatment resulted in a lower abundance in CD + EGCG ( = 0,001) but no significant change in abundance of HFD + EGCG was observable ( = 0.574).

Discussion
We showed that HFD induces significant DNA damage in liver and colon compared with CD. These results were also reflected by a lower gene expression of Dnmt1 in liver and colon of HFD fed mice. In turn, DNA methylation status was higher in this group. MLH1 methylation status was also higher compared to CD, but gene expression lower due to HFD feeding in both organs. In particular CpG1 showed a decreased methylation status in contrast to CpG4 with an increased methylation in liver. In colon CpG2, CpG4, CpG5, and CpG6 of MLH1 promoter region were affected due to different feeding. IL-6 gene expression was significantly higher in HFD compared to CD.
The gut microbial composition differs between CD and HFD: lower total bacterial abundance due to HFD, lower microbial diversity, higher Firmicutes/Bacteroidetes ratio, lower abundance of F. prausnitzii, and Akkermansia, and reduced incidence of butyryl CoA: acetate CoA-transferase gene and butyrate kinase gene.
On the basis of microbial analysis we are able to support the results of our previous publications on human fecal analysis [44]. Mice fed a HFD differ in microbial subpopulations especially the Firmicutes/Bacteroidetes ratio which is already handled as a marker in obesity epidemic. Although other researches also show converse or not diverging results [45][46][47][48][49]. However, additionally we can support results of gut microbial metabolites/cell wall components influencing the host via epigenetic mechanisms too. An increased Firmicutes/Bacteroidetes ratio occurs or even causes lowgrade inflammation, increased IL-6 gene expression, but also increased DNA damage and by increased MLH1 methylation status and reduced expression of the gene. However, the aspect if gut microbial dysbiosis first induces immunological disequilibrium or genomic instability remains. Pateras et al. (2015) extensively investigated the "cross-talk" between genomic instabilities and alteration of immune functions in regard to health implications in humans [50]. Ray and Kidane (2016) indicated a microbial dysbiosis as conducive for the release of bacterial metabolites triggering chronic inflammation followed by DNA damage [51]. Karakasilioti et al. (2013) support another theory that persistent DNA damage triggers chronic inflammation in adipose tissue [52]. Inhibition of cyclooxygenase 2, implicated in the generation of prooxidant eicosanoids, resulted in a decrease of 8-isoprostane, plasma thiobarbituric acid reactive substance, and an increased GSH/GSSG (glutathione/glutathione-disulfide) ratio in rats [53]. Thus, microbial dysbiosis may induce four different mechanisms of DNA damage: (1) mitogen released by a dysbiotic microbiota possesses the capacity to enter the cell and enhance reactive oxygen and nitrogen species (RONs) through enzyme expression, (2) the gut microbial dysbiosis provokes chronic inflammation through pattern recognition receptors to initiate DNA damage and cellular transformation, (3) the dysbiotic microbiota directly generates RONs to induce DNA damage and base excision repair (BER) intermediates, and (4) the release of bacterial genotoxins and metabolites causes chronic inflammation, which in turn promotes DNA damage [51].
Interventions with EGCG as an epigenetic active antioxidant may provide valuable impact in the therapy of metabolic syndrome. Bose et al. (2008) showed an effect of EGCG on HFD fed mice: the percentage of body fat and the visceral fat weight were reduced significantly ( ≤ 0.05) due to the supplementation of EGCG (3.2 g/kg diet) for 16 weeks when compared to control mice [11]. However, the bioavailability is low; intragastric administration of EGCG (75 mg/kg) resulted in 20.6 ng/g in the small intestine and 3.6 ng/g in colon.
Oral administration of EGCG (equivalent with two or three cups of green tea) induced plasma levels of 0.2-0.3 M. It is regulated by the active efflux through the multidrug resistance-associate protein 2 on the apical surface of the intestine and the liver. The uptake predominantly takes place from enterocytes into the intestinal lumen and from the liver to the bile with excretion in the feces and little to none in the urine. EGCG and other tea catechins undergo extensive biotransformation: methylated by catechol-O-methyltransferase Oxidative Medicine and Cellular Longevity 13 (COMT), glucuronidated by UDP-glucuronosyltransferases, and sulfated by sulfotransferases [54]. However, depending on polymerization degree, EGCG also influences gut microbiota composition: higher polymerization results in a higher bioavailability in the gut as the absorption in the small intestine is negligible in contrast to low degree of polymerization [55]. Methylated catechins, ring fission products (like valerolactone), and phenolic products further degraded to phenylacetic and phenylpropionic acids are indicative of gut microbial transformations. However, the biological activities of these metabolites are lower than the activities of their parent compounds [54]. On the other hand EGCG has valuable direct impact as antioxidant, but also as epigenetic active substance influencing histone modification and/or DNA methylation patterns [27].

EGCG Protected DNA Damage Caused by HFD in the
Liver. EGCG supplementation caused a decrease of DNA migration in liver while no effects were found in the colon in the HFD group. In the CD + EGCG even increased DNA damage was found in the colon. The divergent effects in the two organs may be due to the effect that catechin is present in the colon in higher concentration as in the liver. It was shown earlier that high doses of EGCG cause formation of ROS, as a consequence DNA damage in human cells [56].
As mentioned above, comets reflect single and double strand breaks as well as apurinic sites; our findings are indicative for a protective effect of the green tea catechin toward these types of lesions. Our findings are in partial agreement with results published by Kager et al. (2010) who gave EGCG orally to normal weight mice (as in the present experiment). The authors showed no evidence for a protective effect in standard SCGE experiments in colons and livers. However, a clear protective effect was seen in hepatocytes but not in colonocytes in this experiment in regard to formation of oxidized DNA bases [41]. The molecular mechanisms which account for the DNA-protective properties of EGCG involve scavenging of radicals that are typical for obesity [57] which has been found in vivo in rodents and also in vitro [58,59] alternatively; indirect effects caused by activation of antioxidant enzyme via the transcription factors nuclear factor erythroid 2-related factor 2 (Nrf2) may play a role [60] Oršolić et al. (2013) published results which were obtained with diabetic mice; they found even an increase of DNA migration in the liver after injection of EGCG for 7 days [61]. Notably, the authors used a relatively high dose of the catechin in this experiment (i.e., 50 mg/kg body weight) and it is known from in vitro studies that high concentrations of EGCG and other phenolics cause DNA damage as a consequence of radical formation [62,63].
The patterns of gene expression of Dnmt1 and MLH1 which were observed in the liver are in partial agreement with results from the SCGE experiments; a decrease was observed in hepatocytes in obese animals compared to controls. However, EGCG did not compensate this effect but caused a further decline of the expression of both genes.
On the contrary a clear increase of Dnmt1 was seen with the catechin in the colons of HFD animals while obesity itself had no impact on the transcriptional activity of both genes in colonocytes. It is well documented that the MLH1 and Dnmt1 play a key role in DNA repair processes, in particular mismatch repair [64,65]. However, distinct differences which we found between the induction of DNA migration in SCGE experiments and decreased gene expression levels distract from the assumption of a direct relation between comet formation and repair processes which are controlled by MLH1 and Dnmt1.

EGCG Decreased Inflammatory IL-6 and MLH1 Gene Expression Reflected by Higher MLH1 Promoter Methylation.
In colon a significantly lower IL-6 gene expression was induced by EGCG but did not affect significantly MLH1 gene expression. In liver EGCG reduced MLH1 gene expression in both diet groups. The mean methylation was higher in CD + EGCG compared to CD whereas in HFD + EGCG a decrease was shown in comparison to HFD. However, methylation status varies CpG specific and additionally diet specific.
EGCG also reduced the expression levels of TNF-, IL-6, IL-18, and IL-1 mRNAs, the serum levels of TNF-, and the activation of Stat3 and JNK proteins in diethylnitrosamine-(DEN-) induced liver tumor genesis treated C57BL/KsJdb/db (db/db) obese mice [13], but also IL-6 synthesis in rat adjuvant-induced arthritis by administration of 100 mg/kg EGCG, intraperitoneally daily [66]. Ahmed et al. (2008) showed an increase in the synthesis of soluble gp130 protein, an endogenous inhibitor of IL-6 signaling and transsignaling [66]. Additionally EGCG induced a concentration and time dependent reversal hypermethylation of tumor suppressor genes such as p16, RAR, MGMT, and MLH1 genes in human esophageal cancer cells [12].

EGCG Increased Dnmt1 DNA Methylation Resulting in
Tissue Specific Variances in Gene Expression. EGCG supplementation resulted in a significantly reduced gene expression of Dnmt1 compared to, respectively, CD and HFD in the liver. In the colon EGCG compensates the decrease in gene expression due to HFD and results equalized to CD and with three times higher gene expression of Dnmt1 compared to HFD. The methylation status in the promoter region of Dnmt1 was higher in supplemented groups compared to both control groups (CD, HFD). A significant increase was shown in CpG1 and CpG3 in the liver. In colon CpG2 and CpG4 were affected, respectively, in HFD + EGCG or in CD + EGCG.
EGCG was already shown to be the most efficacious inhibitor of enzymatic DNA methylation in vitro in comparison to other tea polyphenols (catechin, epicatechin) and bioflavonoids (quercetin, fisetin, and myricetin). Inhibitory effects for SssI nmt-and Dnmt1-mediated DNA methylation were shown at a half maximal inhibitory concentration (IC50) of 0.21 and 0.47 M, respectively. Inhibitory mechanisms are mentioned on one hand via direct pathways and on the other hand indirect via Dnmt-mediated DNA methylation through increased formation of SAH, a potent inhibitor of S-adenosylmethionine-(SAM-) mediated reactions [67] and via altering the availability of methyl groups which are used to methylate catechol groups on polyphenols by catechol-O-methyltransferase [27]. Although  mentioned rather an important influence of the presence of a physiologically relevant concentration of Mg2+ (such as 2 mM) on inhibitory potency of EGCG compared to a dependence on its own methylation [67]. EGCG is also suggested to induce Foxp3 promoter demethylation inducing differentiation and expansion of Treg via Dnmt inhibitory activity and to reduce T cell proliferation and cytokine production [68]. Inhibition of Dnmts together with an inhibition of histone deacetylase is suggested to prevent the hypermethylation and the silencing of key genes [67]. Either gut microbial derived metabolites of EGCG, gallic acid (GA), and epigallocatechin (EGC) influence epigenetic gene expression via HAT inhibitors to a lesser extent [27].

EGCG Changed Obese Gut Microbial Profile to Lean
Phenotype. HFD feeding caused a significant lower bacterial abundance in both HFD groups (HFD, HFD + EGCG) resulting in a lower microbial diversity compared to CD. The Firmicutes/Bacteroidetes ratio is significantly lower in HFD + EGCG but higher in CD + EGCG compared to corresponding control group. Main changes due to EGCG intervention are decreased Lactobacilli abundance and also a lower abundance of F. prausnitzii in HFD groups with highest abundance in CD. Butyryl CoA: acetate CoA-transferase gene significantly increased in CD + EGCG whereas in HFD + EGCG the butyrate kinase significantly increased. Akkermansia are reduced to be abundant with EGCG supplementation.
It is already known that HFD, Western lifestyle, impacts gut microbiota composition. Diet quality and quantity are important influencing factors on bacterial community composition and metabolic/immunological activity of the host gut microbiota. Microbiota-mediated genomic instability but also immunological disequilibrium may be reduced due to phytochemicals. However, they are generally poorly absorbed in the small intestine; thus the impact of the close contact with the gut microbiota affects health benefits attributed to natural compounds. Results from green tea polyphenols, EGCG, give evidence to have a positive influence on gut microbiota composition. Unno et al. (2014) show not only changes in body and stool composition due to EGCG treatment but also changes in gut microbiota composition. However, these changes were dependent on dosage: 0.3% EGCG supplementation induced Bifidobacterium, Lactobacillales, and Bacteroides but reduced the abundance of Clostridium clusters. A concentration of 0.6% EGCG supplementation increased the abundance of Lactobacillales and Bacteroides but nearly depleted the abundance of Clostridium clusters and Bifidobacterium [26]. However, Unno et al. (2014) used Wistar rats and fed a commercial chow; thus comparison of both projects in gut microbiota composition, especially Lactobacilli, may be impaired on induced metabolic syndrome due to HFD feeding. However, we showed also an increase in Bacteroidetes and a decrease in Clostridium clusters impairing butyrate metabolism. Unno et al. (2014) showed lower levels of acetic and butyric acid but little influence on propionic acid due to 0.6% EGCG supplementation [26]. We even showed differences in butyrate formation pathways. The butyrate kinase pathway is more related to a Western diet, reflected in HFD fed mice (main fat resource: lard). Meanwhile the butyryl CoA: acetate CoA-transferase gene is associated with vegetarian feeding and to a lesser content available in omnivores [69].
In addition to EGCG as an epigenetically active antioxidant we also supplemented C57BL/6J mice with vitamin E (4.5 mg/kg b.w.) as a positive control which showed similar results compared to EGCG but with profound effects which will be published elsewhere. DNA damage showed a significant reduction in both organs in the HFD group but not in the CD group (tail intensity in CD-animals + vitamin E: 6.03 ± 0.95 in liver and 6.00 ± 1.13 in colon). The relative gene expression of Dnmt1 in the colon increased with vitamin E supplementation in CD but significantly decreased in HFD. In all intervention groups a decreased gene expression was noted in the liver. The methylation status of Dnmt1 in colon was lower. Furthermore, a positive correlation of Dnmt1 mean methylation and DNA damage has been observed in liver whereas in CD a correlation has been found in the colon. Vitamin E supplementation affects specific CpG sites of MLH1 inducing a lower gene expression of MLH1 with HFD in liver and colon.

Conclusions
According to our results, EGCG might be suggested for the potential use for the prevention or in the therapy of obesity related and oxidative stress-induced health risks. One effect may derive from changes in GI microbiota and their anti-inflammatory effects by metabolites. Another effect may derive from antioxidative activities as well as epigenetic modifications observed on CpG methylation but also likely to include other elements of the epigenetic machinery. Interactions between antioxidative and epigenetic effects, for example, via ROS mediated breaks of Dnmt pathways, need to be explored.