Introduction

The human gut microbiota is populated with over 100 trillion cells of potentially 100–1000 microbial species [1, 2], which have been shown to be involved in maintenance of human health and pathogenesis of disease [3]. During the past several years, it has been widely demonstrated that the gut microbiota has coevolved with their mammalian host and has established a complex ecosystems [4]. Recent omic-based studies have further facilitated our understanding of dynamic interplay between multiple low molecular weight (LMW) substances produced by gut microbiota and gene expression regulation or posttranslation modification of host [5, 6]. Correspondingly, epigenetics is generally defined as the study of mechanisms that do not modify genomic DNA sequence yet regulate gene expression associated with physiologic and pathologic processes [7]. Furthermore, as epigenetic modifications are reversible depending on environmental stimuli [8], novel therapies that target aberrant epigenetic states and disease are gaining momentum. Here, we will discuss recent clues that implicate influences of gut microbiota on epigenetic modifications.

Epigenetic Modifications

The term epigenetic was first coined by Conrad Waddington for the underlying mechanisms that converted identical genetic information into diverse phenotypes [9]. Interesting epigenetic cases have been identified such as in reproductive status in honeybees [10] and turtle sex determination [11]. Currently, epigenetic modifications are considered to be heritable adaptive mechanisms that regulate gene expression patterns through mitosis and/or meiosis without change in primary DNA sequences [12]. The primary focus of this review is on the two main categories of epigenetic modifications: DNA methylation and histone modifications.

DNA Methylation

DNA methylation involves the transfer of methyl group from S-adenosyl methionine to 5 position of a cytosine pyrimidine ring (m5C), which mostly happens in the context of CpG dinucleotides [13]. The CpG dinucleotides are often clustered in regions called CpG islands [14], defined as sequence ranges of more than 200 bases where a value of observed to statistically expected CpG frequencies is greater than 0.6 and a G + C content is at least 50% [15]. Approximately 60% of gene promoters in human are associated with CpG islands [16] and are generally unmethylated [17]. However, in addition to CpG islands, CpG dinucleotides are predominately methylated [18].

In mammals, DNA methylation is catalyzed by three members of the DNA methyltransferase (DNMT) family: DNMT1, DNMT3a, and DNMT3b [15]. Typically, regulation of DNA methylation on gene expression will have exactly the reverse effect in different gene regions: methylation of gene promoters is in general related to transcriptional silencing [19], while methylation of gene body is often coupled with transcriptional activation [20].

Moreover, DNA methylation does not take place entirely at CpG islands. The term CpG island shores has recently been coined as an additional ingredient of methylation phenomena, referring to regions that flank CpG islands (~2 kb) with lower CpG frequencies [21]. Methylation of CpG island shores is closely related to transcriptional inactivation [15] and is certified to exhibit more specificity among normal tissue types, between cancerous and normal cells [22], thus could be potential candidates for tumor markers. However, to our best knowledge, limited studies focus on methylation of CpG island shores which are influenced by bacteria or by gut microbiota.

Histone Modifications

Histones are proteins that package and order eukaryotic genomes into chromatin, which play an important role in epigenetic regulation with varying covalent modifications [23]. Histones mainly contain five families: H1, H2A, H2B, H3, H4, and H5. Histones H2A, H2B, H3, and H4 are known as core histones to form the nucleosome core representing two H2A–H2B dimers and a H3–H4 tetramer [24]. Histones H1 and H5 are known as the linker histones, histone H1 seals off the nucleosome at the entry and exit locations of the DNA [25], and histone H5 is now recognized as an isoform of histone H1 [26]. All histones are subject to a variety of posttranscriptional modifications, of which the most studied are acetylation and methylation.

Histone acetylation usually takes place on multiple lysine residues located within their N-terminal termini, primarily of histone 3 at lysine 9 (H3K9), lysine 14 (H3K14), lysine 18 (H3K18), lysine 23 (H3K23), histone 4 at lysine 5 (H4K5), lysine8 (H4K8), lysine 12 (H4K12), and lysine16 (H4K16) [27–30]. Histone acetylation and deacetylation is mediated by opposing practices of histone acetyl transferases (HATs) and histone deacetylases (HDACs), respectively [31]. HATs are believed to relax the chromatin structure and allow access for transcription factors, thereby promoting gene expression [32]. Conversely, HDACs promote tighter DNA–histone interaction and repress transcriptional activity [32]. Eighteen currently known HDACs are categorized into four classes depending on their homology with HDACs in yeast and subcellular location. Class I (HDACs 1–3 and 8), II (HDACs 4–7, 9, and 10), and IV (HDACs 11) require zinc for their enzyme activity, while the class III HDACs (sirtuins 1–7) are nicotine adenine dinucleotide (NAD+) dependent [33].

Histone methylation is generally involved in mono-, di-, or tri-methylation of histone 3 at lysine 4 (H3K4), lysine 9 (H3K9), lysine 27 (H3K27), lysine 36 (H3K36), lysine 79 (H3K79), lysine 20 (H3K20), of histone 3 at arginine 2 (H3R2), arginine 8 (H3R8), arginine 17 (H3R17), arginine 26 (H3R26), arginine 128 (H3R128), arginine 129 (H3R129), arginine 131 (H3R131), arginine 134 (H3R134), and methylated arginine 3 in the histone H4 (H4R3), methylated arginine 3 in the histone H2A (H2AR3) [34–46]. The consequences of histone methylation for gene expression are complex, as methylation may have opposite consequences depending on localization and number of methyl groups.

Influences of Gut Microbiota on Epigenetic Modification

An intricate and dynamic relationship exists between the gut microbiota, the immune system and epigenetic modifications [47]. The immune system develops symbiotic relationship with gut microbiota that shapes the diversity and abundance of microbiota [48]. In turn, gut microbiota and induced epigenetic modification is essential for the development and maturation of intestinal mucosal immune system of its hosts [49–53]. As epigenetic modifications of intestinal mucosal immune system mediate cross talk between gut microbiota and the mammalian host, we focus our attention on immune development. Evidence of gut-derived effector molecules effecting host epigenetics are summarized in Table 1.

Table 1 Evidence of gut-derived effector molecules effecting host epigenetics
Full size table

Influences of Gut Microbiota on DNA Methylation

Many essential micronutrients involved in the process of DNA methylation are associated with gut microbiota. S-adenosylmethionine (SAM) is the primary methyl donor in DNA methylation, which is produced by the biological process called one-carbon metabolism [64]. Folate, choline, methionine, and vitamins B2, B6, and B12 are heavily involved in one-carbon metabolism [65]. It is worth noting that B vitamins also act as cofactors in DNA methylation [66]. Long-term folic acid supplementation before and during pregnancy was associated with higher leptin (LEP) and retinoid X receptor alpha (RXRA) gene cord blood methylation, respectively [67]. Before pregnancy, higher intakes of betaine and methionine were associated with higher cord blood methylation levels of DNMT1 CpG4 and LEP CpG4, respectively. In the second trimester of pregnancy, high methyl group donor intakes (betaine, choline, and folate) were negatively associated with gene specific cord blood methylation (betaine with LEP CpG2; choline with DNMT1 CpG4; and folate with LEP CpG2 and DNMT1 CpG4). In the last trimester of pregnancy, a high intake of choline and folate was associated with higher methylation levels of DNMT1 CpG2 and lower methylation levels of RXRA CpG2, respectively [67]. The authors did not discuss the conflicting results; we hypothesized it may be related to the dynamic process of pregnancy. Hosts must obtain these essential micronutrients exogenously from diet and gut bacteria due to their lack of biosynthesis [54, 56]. The deteriorating gut microbiome and its metabolites will inevitably interfere biochemical process of diet fermentation and thus influence the normal provision of methyl donor for DNA methylation. In addition to global DNA methylation, other bacterial metabolites that relate to DNA methylation are summarized in Table 1.

Many evidences have shown a close relation between gut microbiota and DNA methylation, both in basic researches and clinical findings. The influence of gut microbiome on the epigenetic regulation of host genes has been demonstrated with Toll-like receptor 2 (TLR2)-knockout mice [68, 69], in which the DNA methylation and gene expression in colonic mucosa of wild-type and Tlr2−/− C57BL/6 mice were interrogated. Average DNA methylation in the promoter regions of two genes which related to immune processes, Anpep and Ifit2, increased in Tlr2−/− colonic mucosa. This indicates that alterations in mucosal microbial composition induced by TLR2 deficiency contribute to transcriptomic and epigenomic modifications. Furthermore, repression of the TLR4 gene in intestinal epithelial cells (IECs) through DNA methylation can play a role in maintaining intestinal homeostasis and regulating the mucosal immune system in the gut [70]. Dapito et al. [71] in an elegant series of experiments have demonstrated that TLR4 activation by LPS generated from the intestinal microbiota contributes to hepatocellular carcinoma promotion. Interestingly, Kumar et al. [72] revealed that DNA methylation profiles in blood of pregnant women correlated with gut microbiota patterns. Eight well-matched pregnant women were classified into two groups depending on their dominant gut microbiota, i.e., Bacteroidetes, Proteobacteria, and Firmicutes. Next-generation sequencing of DNA methylomes indicated a clear correlation between predominant phyla and epigenetic patterns. Promoter DNA methylation status of genes in the HighFirm group was functionally associated with genes specifically involved in cardiovascular diseases, lipid metabolism, obesity, and the inflammatory response. This is one of the first studies that underline the association of the gut microbiota with DNA methylation profiles.

Influences of Pathogenic Bacteria on DNA Methylation

Examples of DNA methylation effects of pathogenic bacteria, such as Helicobacter pylori and Klebsiella spp, on the host have also been characterized recently. Chronic inflammation induced by Helicobacter pylori infection is responsible for DNA methylation in gastric mucosae of genes which are closely associated with gastric cancer risk [73]. H. pylori infection induces the inactivation of miR-210 gene expression in the gastric epithelium with chronic inflammation compared with uninfected gastric epithelium, with surprisingly higher rate of CpG methylation within the miR-210 locus in healthy individuals with H. pylori than healthy people without H. pylori [74]. Significantly higher levels of aberrant methylation (5.4- to 303-fold) were present in eight regions of CpG islands in the gastric mucosae of healthy volunteers with H. pylori infection, which strongly indicated that H. pylori infection potently induces aberrant methylation in multiple CpG islands [75]. Similarly, Hp-positive healthy volunteers had a higher level of DNA methylation of three miRNA genes (miR-124a-1, miR-124a-2, and miR-124a-3) in gastric mucosae than Hp-negative individuals, indicating that H. pylori infection can induce DNA methylation of miRNA genes, in addition to protein-coding genes. Infection of immature intestinal epithelial cells with Lactobacillus acidophilus/Bifidobacterium infantis and Klebsiella spp. results in over 200 regions of differential DNA modification [76].

Influences of Gut Microbiota on Histone Modifications

The gut microbiota plays a role in uptaking and secreting minerals that act as cofactors for enzymes participating in epigenetic regulation, such as cobalt, iodine, selenium, and abovementioned zinc. Furthermore, a wide range of enzymes such as the methyltransferases, acetyltransferases, deacetylases, Bir A ligase, phosphotransferases, kinases, and synthetases are produced by the gut microbiota. Various key energy metabolites including SAM, acetyl-CoA, NAD+, α-KG, and ATP serve as primary cofactors for many enzymes that regulate epigenetic modification [57]. Epicatechin and epigallocatechin-3-gallate (EGCG) can function as histone acetyltransferase (HAT) inhibitor [58, 59], which could be degraded and metabolized by the intestinal microbiota [77]. Similarly, short-chain fatty acids (SCFAs), isothiocyanate, and allyl compounds derived from human gut microflora have been shown to inhibit histone deacetylase (HDAC) activity [78, 79].

Short-chain fatty acids (SCFAs) generated through dietary carbohydrates fermentation by the gut microbiota, many of which in the class Clostridia of Firmicutes phyla, are thought to be mainly responsible for histone modifications. Several promising HDAC inhibitors (HDACi) are presently being evaluated in various stages of clinical trials for the therapy of a wide range of cancer, as well as inflammatory and degenerative disorders [80, 81]. Intriguingly, SCFAs, predominantly acetate, propionate, butyrate, caproate, and valerate have recently attracted considerable attention for their HDAC inhibitory activity [60, 62, 63, 82]. Major SCFAs-producing bacteria and products with the corresponding substrates are listed in Table 2. Schwiertz et al. [84] found elevated SCFA concentrations in overweight and obese subjects with altered gut microbiota compared to healthy controls; they speculate these changes in SCFA may play a role in the development or maintenance of obesity. The affinity of SCFAs to various G protein-coupled receptors (free fatty acid receptors) may be involved in inflammatory conditions, as well as interactions between the gut microbiota and epigenetic regulation [62, 85].

Table 2 A list of key SCFAs-producing bacteria
Full size table

SCFAs, primarily acetate, propionate, and butyrate have recently been revealed for their beneficial effects of histone deacetylase inhibitory activity. Vinicius Andrade-Oliveira et al. [86] evidenced acetate treatment inhibited the activity of HDACs and increased global methylation status in kidney tissue undergoing ischemia and reperfusion injury. Nagendra Singh et al. reported that butyrate and propionate blocked the generation of dendritic cells from bone marrow stem cells; this effect is associated with decreased expression of the transcription factors PU.1 and RelB and is dependent on the ability of these two bacterial metabolites to inhibit histone deacetylases [87]. Butyrate is one of the most potent HDACi which inhibits most HDACs except class III and HDAC 6 and 10 of class II with approximately 80% inhibition of HDAC1/2 and a Ki value of 58 μM [88, 89]. A large number of histone hyperacetylation effects exerted by butyrate have been suggested, which can induce cell differentiation, apoptosis, and inhibit proliferation of tumor cells. In human colon tumor cell lines, butyrate treatment resulted in overexpression of p21WAF1 via hyperacetylation of the gene-associated histones, and arrest of G1 cell cycle [63]. Butyrate blocked human leukemic lymphoblasts in the G2/M phase and caused apoptotic cell death [90], and induced H4 acetylation and myeloid maturation in acute myeloid leukemia (AML) [91] and growth inhibition in human cervical carcinoma and prostate cancer cells [92, 93]. In Bcr/Abl + human leukemia cells, a model of apoptosis induction by butyrate was proposed to involve a coordinated obstruction of the cytoprotective Raf/MEK/ERK pathway in conjunction with the reactive oxygen species-dependent activation of JNK [94]. Also, butyrate-induced histone H3 acetylation and significantly up-regulated expression of Fas, P21, and P27 to induce apoptosis of lymphoma tumor cells [95] and treatment of butyrate led to an increase in histone H3 lysine 9 acetylation (H3K9Ac) levels at the promoter regions of Nos2, Il6, and Il12b [96].

A novel contributory mechanism to the anticarcinogenic effect of butyrate is the down-regulation of the key apoptotic and angiogenesis regulator neuropilin-1 (NRP-1) transcription, which has been shown to promote tumor cell migration and survival in colon cancer in response to vascular endothelial growth factor (VEGF) binding [97]. Butyrate also plays a promising role in inflammatory response by suppressing the nuclear factor κB (NF-κB) activation via inhibition of HDAC [98]. NF-κB regulates the production of pro-inflammatory cytokines involved in early immune inflammatory responses, including IL-1b, IL-2, IL-6, IL-8, IL-12, tumor necrosis factor-alpha (TNF-α), inducible nitric oxide synthase (iNOS), cyclooxygenase-2 (COX-2), intercellular adhesion molecule-1 (ICAM-1), vascular cellular adhesion molecule-1 (VCAM-1), T cell receptor-α (TCR-α), and MHC class II molecules [99–106].

In addition, butyrate as well exerts the anti-inflammatory effect by inhibiting interferon γ (IFN γ) signaling and up-regulating the expression of peroxisome proliferator-activated receptor γ (PPAR γ) [107, 108]. Evidence indicates a role for butyrate in strengthening innate immunity by enhancing the expression of the LL-37 gene, the only cathelicidin-derived peptide expressed in humans, with increased histone acetylation of the cathelicidin promoter and mitogen-activated protein (MAP) kinase signaling [63, 109–111]. What’s more, butyrate induces histone H3 lysine 27 (H3K27) acetylation of the FoxP3 locus and thus promotes differentiation of regulatory T cells in the colon [82].

Histone modifications of commensal bacteria or probiotics have also been well characterized. High concentrations of butyrate produced by Porphyromonas gingivalis could reactivate latent HIV-1 integrated in the host genome as proviral DNA copies via HDAC inhibition [69, 112]. Bacteroides vulgatus, a commensal of the gut microbiota, can induce acetylation of histone H3 through an inflammatory signaling cascade [113]. Also, B. vulgatus induces inhibition of NF-κB transcriptional activity as well as IL-6 mRNA accumulation mediated by TGF-β1, which in turn induces histone H3 deacetylation via histone deacetylase recruitment at IL-6 promoter [113]. This may contribute to the maintenance of intestinal homeostasis by inhibiting commensal bacteria to activate inflammatory responses in the intestine. Furthermore, Bifidobacterium breve and Lactobacillus rhamnosus GG, representatives of commensal probiotics, diminished the LPS-induced expression of IL-17, IL-23, and CD40 with epigenetic processes involving the inhibition of histone acetylation and the optimal enhancement of DNA methylation [114].

Influences of Pathogenic Bacteria on Histone Modifications

Specific effectors produced by pathogenic bacteria can also induce specific histone modifications. Listeria monocytogenes has been shown to increase IL-8 gene expression by inducing acetylation (lysine 8) of histone H4 and phosphorylation/acetylation (serine 10/lysine 14) of histone H3 globally and at the il8 promoter in human umbilical vein endothelial cells, as well as recruitment of the histone acetylase CREB-binding protein through a nucleotide-binding oligomerization domain 1 (Nod1) proteins-dependent activation of p38 MAPK signaling and NF-κB [69, 115, 116]. Recently, listeriolysin O (LLO)-induced K+ efflux was unveiled as an important signal leading to histone modification [117]. The increased expression of p21WAF1 induced by Helicobacter pylori is associated with the release of HDAC-1 from the p21WAF1 promoter and hyperacetylation of histone H4 [118].

Conclusion

Recent studies and ‘omic’-based technologies suggest that diverse metabolites and signaling molecules produced by gut microbiota may regulate host gene expression through epigenetic modification, explaining a significant part of host–microbiota interactions (Fig. 1). However, we should not stop here; clarifying the underlying molecular mechanisms of these biological processes are still in their infancy and warrant further investigation. Specifically, it needs to be deciphered how DNA methylation or histone modification occurs at specific gene loci. What’s more, novel epigenetic modifications and enzymes which may participate in epigenetic regulation need to be discovered. Continued basic and applied studies are also needed to focus our efforts toward utilizing epigenetic-targeting drugs on microbiota and relevant microbiota-induced cancer and metabolic syndrome, an area that promises exciting preventive and therapeutic applications.

Fig. 1
figure 1

Interaction between gut microbiota and epigenetic modification. The gut microbiota could regulate epigenetic modification with varying energy metabolites and signaling molecules such as minerals, SCFAs, enzymes, and so on. Conversely, epigenetic modification of immune system can also plays a role in intestinal homeostasis. Dash lines indicate elements that remain to be elucidated or elaborated

Full size image