Epigenome-metabolome-microbiome axis in health and IBD

Abstract

Environmental triggers in the context of genetic susceptibility drive phenotypes of complex immune disorders such as Inflammatory bowel disease (IBD). One such trigger of IBD is perturbations in enteric commensal bacteria, fungi or viruses that shape both immune and neuronal state. The epigenome acts as an interface between microbiota and context-specific gene expression and is thus emerging as a third key contributor to IBD. Here we review evidence that the host epigenome plays a significant role in orchestrating the bidirectional crosstalk between mammals and their commensal microorganisms. We discuss disruption of chromatin regulatory regions and epigenetic enzyme mutants as a causative factor in IBD patients and mouse models of intestinal inflammation and consider the possible translation of this knowledge. Furthermore, we present emerging insights into the intricate connection between the microbiome and epigenetic enzyme activity via host or bacterial metabolites and how these interactions fine-tune the microorganism-host relationship.

Epigenetics as a contributor to immunological disease, including IBD

DNA is organized in each cell nucleus by interacting with histones to form chromatin. Chromatin is regulated by modifications on DNA or histones and protein complexes that remodel its architecture [1]. At the interface of the environment and gene expression, chromatin alterations enable cell type-specific and kinetic regulation of gene expression from identical DNA sequences and are essential for proper cell development and function. More than 100 unique chromatin modifications have been identified, including common ones like methylation, acetylation, phosphorylation, and less studied modifications like proprionylation, butyrylation, crotonylation, and succinylation [1,2]. These modifications are regulated by over 400 chromatin modifying enzymes. Epigenetic ‘writers’ catalyze post-translational modifications on histones or DNA (e.g. DNA and histone methyltransferases, histone acetyltransferases), while ‘erasers’ remove these dynamic modifications (e.g. demethylases, histone deacetylases (HDACs), TET proteins). Finally, epigenetic modifications are interpreted by ‘readers’, that dock to defined modified histones via distinct protein domains (e.g. bromodomain, PHD, YEATS) and recruit appropriate transcriptional machinery to facilitate or prevent transcription [1].

Genome wide association studies (GWAS) have identified over 200 genetic susceptibility loci for IBD [3] comprising Crohn’s disease (CD) and Ulcerative Colitis (UC). Many of these mutations result in loss-of-function variants of immune components that are normally responsible for protective responses to enteric microorganisms, emphasizing the importance of the host-microorganism relationship for intestinal homeostasis [3]. However, the majority of susceptibility gene mutations contribute individually to IBD with low odds ratio [3], have low concordance rates in monozygotic twins [3] and are in non-coding regions [4]. Moreover, the rapid rise in the incidence of complex immune disorders such as IBD indicates that additional factors to genetics must contribute to disease. Recent studies in monozygotic twins showed that by 65 years of age almost 70% of the immune epigenome is environmentally derived [5], thus demonstrating the profound impact of non-heritable environmental influences on immune output and disease susceptibility. Furthermore, dysregulation of epigenetic enzymes and chromatin regulatory processes is an established event in human disease, but current knowledge is limited mostly to cancer. This has made proteins that regulate the epigenome some of the most promising and intently pursued targets in drug discovery today [6,7]; yet, a critical gap exists in our understanding of altered epigenetic enzymes or chromatin that result in immune disorders, including IBD. Chromatin alterations have been consistently observed in IBD patients and recent GWAS studies have reveals mutants of certain epigenetic enzymes significantly associated with IBD susceptibility. Furthermore, intestinal microorganisms and their metabolites are increasingly being recognized for their ability to alter activity of host chromatin modifying enzymes in order to immunomodulate and affect disease susceptibility. A greater mechanistic understanding of epigenetics in IBD will illuminate important, potentially reversible, mechanisms that link genetic predisposition and intestinal microorganisms in the pathogenesis of disease.

Epigenetic alterations as a trigger and target in intestinal inflammation?

Numerous genome wide methylation profiling studies in humans have shown aberrant DNA methylation in sera and intestinal tissues of young and adult IBD patients [8^•,9,10]. In addition, histone methylation profiling of intestinal epithelial cells (IECs) from a CD cohort revealed differences in H3K4me3 signatures [11], with enriched H3K4me3 patterns in genes associated with immune processes to bacterial factors, cytokine signaling and reduced H3K4me3 signatures associated with intestinal epithelial ion absorption and lipid and protein metabolism [11]. Moreover, some studies revealed the highly differentially methylated or histone modified-genes were in fact for IBD risk loci, connecting possible epigenetic signatures with predisposed risk loci [9]. However, it is often challenging in human studies to cleanly separate correlation from causation.

In mice, demonstration that perturbations of individual epigenetic enzymes profoundly influence both immune and intestinal homeostasis in vivo has provided some evidence of causation. Loss of DNA ‘writers’ (DNMT1), histone ‘writers’ (SETDB1, EZH2, ASH1L), DNA ‘erasers’ (MBD2, TET), histone ‘erasers’ (HDACs 2, 3, 7), histone ‘readers’ (UHRF1, TRIM28, SP140) or chromatin remodelers (SMARCA4, PBRM1) disrupted cell identity and prevented appropriate integration of microbiota cues and maintenance of intestinal homeostasis (Table 1). Importantly, deletion of some epigenetic enzymes protected in models of IBD and thus may offer some therapeutic avenues for amelioration of intestinal inflammation. For instance, deletion of histone ‘writers’ (G9a, KAT8), histone ‘erasers’ (JMJD3, KDM2B, HDAC 6, 9, 10) or chromatin remodeler SMARCD1 protected in various models of IBD by altering the T regulatory (Treg):T helper (Th)17 ratio, suppressing macrophages or promoting protective IEC functions (Table 1).

Table 1.

Epigenetic regulators in mouse IBD models

Gene/Protein	Epigenetic enzyme function	Mouse disease phenotype	Cell(s) Implicated
		Exacerbated intestinal inflammation
DNMT1	Writer of DNA methylation	Spontaneous early epithelial and crypt damage with IEC specific deletion; however, recover with Dnmt3b compensation. Dnmt1-Dnmt3b double IEC KO don’t recover [12]	IEC
SETDB1	Writer of H3K9me3	Spontaneous and lethal terminal ileitis and colitis with IEC specific deletion [13••]	IEC
EZH2	Writer of H3K27me3	↑ DSS and TNBS colitis with IEC specific deletion [14]	IEC
ASH1L or KMT2H	Writer and Reader of H3K4me, H3K36me and H3K9me1	↑ TNBS colitis and adoptive transfer colitis model with deletion [15]	Treg
MBD2	Eraser and Reader of CpG island Hemi-methylated DNA	↑ DSS colitis with deletion [16]	IEC DC T cells
TET2	Eraser of DNA to produce 5-hydroxymethylcytosine	↑ DSS colitis with deletion [17]	Macrophages DCs
HDAC2	Eraser	↑ DSS colitis with IEC specific deletion [18] ↑ Adoptive T cell transfer colitis and TNBS colitis model with deletion in CD4 T cells [19]	IEC Th17
HDAC3	Eraser	↑ Adoptive T cell transfer colitis with Treg specific deletion [20] ↑ DSS colitis with IEC specific deletion [21] ↑ Citrobacter rodentium infection with IEC specific deletion [22•]	Treg IEC
HDAC7	Eraser	Mice with phosphorylation deficient human HDAC7 (HDAC7-ΔP) developed spontaneous autoimmunity [23]	iNKT
SIRT6	Eraser of H3K9Ac and H3K56Ac	↑ DSS colitis with IEC-specific deletion [24]	IEC
UHRF1	Reader of H3K9me2/ 3 Hemi-methylated DNA	Spontaneous development of colitis with CD4 T cell specific deletion [25] ↑ DSS colitis with myeloid specific deletion or mutation in the DNA methylation site (Uhrf1YP187/188AA) [26]	Treg Macrophages
TRIM28	Reader H3K9me3	Autoimmunity with global T cell deletion [27,28] ↓ Adoptive transfer T cell colitis with CD4+ T cell deletion [29]	Treg
SP140	Reader	↑ DSS colitis with hematopoietic knockdown [30••]	Macrophages
SMARCA4 or BRG1	Chromatin remodeler, Reader	Spontaneous colitis with IEC specific deletion and ↑ DSS colitis [31] Spontaneous colitis with ILC3 specific deletion of Rag1−/− background [32]	IEC ILC3
PBRM1/Baf180	Chromatin remodeler, Reader	↑ Citrobacter rodentium infection with IEC deletion [33]	IEC
		Protected from intestinal inflammation
G9a	Writer of H3K9me1/2	↓ Adoptive T cell transfer colitis with G9a deletion in CD4+ T cells [34]	Th17 Treg
KAT8/Mof	Writer of H416ac	↓ DSS colitis with conditional deletion [35]	Th17
JMJD3	Eraser of H3K27me3	↓ Adoptive T cell transfer colitis with CD4+ T cell specific deletion [36]	Th17
KDM2B/JHDM1B	Eraser of H3K4me3 and H3K36me3	↓ DSS colitis with deletion [37]	Macrophages DCs
HDAC6	Eraser	↓ Adoptive T cell transfer colitis with Treg specific deletion [38]	Treg
HDAC9	Eraser	↓ DSS colitis with complete deletion [39] ↓ Adoptive T cell transfer colitis with Treg specific deletion [40]	Treg
HDAC10	Eraser	↓ Adoptive T cell transfer colitis with Treg specific deletion [41]	Treg
SMARCD1 or BAF60A	Chromatin remodeler	↓ DSS colitis with IEC specific deletion [42•]	IEC

Genetic variation within epigenetic regulators and regulatory regions associate with IBD

In humans, variants of epigenetic regulators or regulatory regions are emerging as direct contributors to IBD susceptibility. Of the 90% of IBD-associated genetic variants that occur in non-coding regions, many are in chromatin regulatory regions, such as immune cell enhancers marked by H3K27ac, and often near sites for master regulators for immune cell identity, differentiation, and stimulus-specific gene expression [4]. Indeed, an IBD risk loci in a distal enhancer of CD4+ Tregs was necessary for limiting colitis [43^••]. Furthermore, genetic variants within loci that code for three epigenetic regulators: SP140, DNMT3A, and DNMT3B have been found to associate with increased susceptibility to IBD using GWAS (Table 2). Fine mapping GWAS and studies in non-European targeted cohorts have identified newer risk loci associated with epigenetic regulators, such as Baz1a and HDAC11. In addition, profiling of colonic biopsy samples have revealed epigenetic regulators, EZH2, SIRT6, ASH1L, SETDB1 and BRG1, which were identified to have colitis phenotypes in mice (Table 1) had reduced expression levels in human IBD [13^••,14,24,31].

Table 2.

Variants of epigenetic regulators that associate with susceptibility of IBD

Gene	IBD-associated SNP	Epigenetic enzyme subclass	Protein domain/Function	Validated or predicted outcome of SNP	Validated in mice
SP140	rs6716753 [45–47] rs28445040 [45,46] rs7423615 [46,48]	Reader	SAND, PHD Bromodomain	Validated Loss-of-function	YES
DNMT3A	rs13428812 [48] rs201014116 [49]	Writer	DNA methyltransferase	Unknown	NO
DNMT3B	rs4911259 [45,47] rs6058869 [50] rs6087990 [51,52] rs1474738 [52]	Writer	DNA methyltransferase	Unknown	NO
BAZ1A	rs712303 [49]	Reader, Chromatin Remodeler	Bromodomain	Unknown	NO
BRD2	rs1049526 [53]	Reader	Bromodomain	Predicted loss-of-function	NO
HDAC7	rs11168249[45]	Eraser	Histone deacetylase	Unknown	NO
HDAC11	rs2655211[54]	Eraser	Histone deacetylase	Unknown	NO

Of the GWAS identified IBD-associated epigenetic enzyme variants, SP140 is the only one that has been functionally validated thus far. SP140 is an immune-restricted PHD and bromodomain containing epigenetic ‘reader’ [44] and CD-associated SP140 SNPs resulted in altered splicing of SP140 mRNA and ultimately diminished SP140 protein. SP140 was identified to be essential for macrophage transcriptional programs that dictate cell identity by preferentially occupying silenced, lineage-inappropriate genes bearing H3K27me3 and ensuring their repression [30^••]. Consequently, loss of SP140 in mouse or human macrophages and cells carrying CD-associated SP140 SNPs displayed hypo-responsiveness to Toll-like Receptor (TLR) stimulation suggesting a critical role for this epigenetic reader in immune response to microbiota. Importantly, a loss of this epigenetic reader is causative of intestinal inflammation since hematopoietic-specific knockdown of Sp140 in mice resulted in exacerbated DSS–induced colitis [30^••]. Thus, an inability to interpret the epigenetic landscape in immune cells due to loss of SP140 contributes to intestinal inflammation.

Global effects of microbiota on host chromatin

Globally, enteric microbiota and their metabolites profoundly influence host chromatin, both at the level of DNA methylation as well as histone post translational modifications (PTMs). Employment of germ-free (GF) raised mice or antibiotic (ABX) treated mice demonstrated that absence of microbiota greatly disrupts host chromatin in the intestine [55,56^••,57]. Microbiota maintains DNA methylation of genes regulating bacterial sensing, immunity and homeostasis. In a recent study, DNA methylation at regulatory elements was reduced in intestinal epithelial cells (IECs) of conventional (CV) mice compared with GF mice and associated with inflammatory genes and bacterial sensing pathways, such as MyD88 and this demethylation was mediated by TETs [56^••]. Differential DNA methylation occurred early in postnatal development in mice pointing to a critical developmental window for imprinting by the microbiota, in a Dnmt1-dependent manner, which correlated with transcription of genes responsible for mucus barrier formation and intestinal maturation [58], as well as innate immune and phagocytosis genes [59].

Similarly, global alterations in histone modifications (H3K4me3, H3K27me3) were observed in GF and ABX treated mice. In particular, microbiota-dependent histone modifications were enriched for innate and adaptive immune related and metabolic pathways [11,55,60]. In addition, histone crotonylation, a modification that associates with active transcription, was abundantly observed in the colon and was reduced with ABX treatment [61^•]. Furthermore, chromatin accessibility, particularly in the enhancer landscape of various intestinal cells, is altered by enteric microbiota that is in concordance with the direction of gene expression changes [56^••,60,62]. Conventionalization of GF mice and re-establishing commensal bacteria rescued DNA methylation, histone modification and chromatin accessibility changes [55,56^••,63]. Hence, the microbiota is a major influencer of host epigenetic state.

Epigenetics as a sensor of metabolic state

As has been thoroughly reviewed previously [64], epigenetics and metabolism are intimately linked. Almost all chromatin modifying enzymes rely on available metabolites for their catalytic activity and thus will underpin the intricate co-evolutionary relationship between microorganisms and epigenetics for adaptive cellular transcription. DNA methyltransferases (DNMTs) require the metabolite S-adenosyl methionine (SAM) and histone acetyltransferases require acetyl coenzyme coA (acetyl-CoA) [64]. Other metabolites such as α-ketoglutarate, S-adenosyl homocysteine (SAH), β-oxybutyrate similarly serve as substrates, donors, cofactors, or competitive inhibitors for various epigenetic enzymes [64] (Figure 1). Recent discoveries of novel epigenetic marks also exemplify this tight coupling, including histone succinylation, crotonylation and lactylation which rely on substrates succinyl-coA, crotonyl-coA and lactate levels, respectively [65–67]. These metabolites are generated in diverse cellular metabolic pathways, including tricarboxylic acid (TCA) and methionine cycle. However, the microbiome is also a major source of metabolites in mammals, including lipids, vitamins, organic acids, indoles, oxazoles and secondary bile acids (SBAs) [68], and is also emerging as prominent regulator of host epigenetic enzyme activity. In particular, microbiota exclusively metabolize complex carbohydrates derived from dietary fibers via fermentative reactions to produce small organic acids in the colonic lumen, the bulk of which are short chain fatty acids (SCFAs): acetate, propionate and butyrate – all of which are HDAC inhibitors [21,69,70^••]. Notably, supplementation of SCFAs in GF mice could recapitulate a large portion of global chromatin states and gene expression patterns observed with complete gut colonization [55]. Insights gleaned from colorectal cancer mouse models also show how other microbial derived metabolites, such as gallic acid, influence chromatin by promoting TCF4-chromatin association and subsequent H3K4me3 at WNT promoters [71^•]. Thus, both cellular and microbial metabolites are major drivers of epigenetic activity for malleable transcriptional and cellular output.

Figure 1.

Intimate relationship between metabolism and epigenetics. Host cellular and bacteria-derived metabolites as essential co-factors for chromatin modifying enzyme function. Green arrows indicate promotion of enzyme activity, red arrows indicate inhibition of enzyme activity. This figure was created using Biorender.

Any alteration in microbiota and their metabolic products, a hallmark of IBD, therefore will profoundly alter host epigenetic enzyme activity. Certainly, a reduction of species diversity in microbial communities is consistently observed in IBD patients [8^•,72^••,73^••]. Importantly, this decrease in microbial diversity parallels a decrease in the metabolite pool with reduced bacteria strains being ones that produce beneficial metabolites such as butyrate and phytate [72^••,74^••]. In addition to SCFAs, pantothenate and nicotinate (vitamins B5 and B3) are also depleted, while polyunsaturated fatty acids such as adrenate and arachidonate are increased in IBD patients [72^••]. Other metabolites like bile acids and acylcarnitines are also differential in IBD patients [72^••,73^••]. Alterations in nicotinate affect the NAD+ pool and thus would be predicted to directly impact the activity of the Sirtuin subfamily of HDACs, while changes in acylcarnitines alters acetyl-coA pools and thus would be predicted to influence HAT activity [64]. However, the precise effects, both acutely and permanently, of all these altered metabolites in IBD on epigenetic enzymes as the critical intermediaries between host and microbiota in human IBD will be a fertile area of future investigation.

Crosstalk between the microbiome and epigenome in intestinal epithelial cells

Intestinal epithelial cells (IECs) act both as a physical barrier to maintain intestinal homeostasis and express innate immune receptors that recognize and respond to commensal microorganisms. It is well established that microbiota regulate intestinal epithelial cell (IEC) barrier integrity, especially at the suckling-to-weaning transition, where onset of solid food ingestion and metabolites is critical for establishing barrier integrity [75]. Lactate produced from commensal bacteria Bifidobacterium and Lactobacillus spp. led to expansion of intestinal stem cells and differentiation into Paneth and goblet cells [76]. Interesting, histone lactylation in macrophages was recently proposed to drive expression of genes involved in tissue homeostasis [67], suggesting its role in IECs may be similar. One of the seminal papers highlighting the critical platform IECs hold for epigenetic programming by microbiota came from Alenghat et al. that depleted histone ‘eraser’ HDAC3 specifically in IECs (HDAC3^ߡIEC) [21]. HDAC3^ΔIEC mice exhibited greater weight loss, disease severity, loss of crypt architecture, edema and inflammation following DSS-induced colitis. Moreover, a large portion of gene expression regulated by HDAC3 was microbiota-dependent [21]. Epithelial HDAC3 was also essential in integrating metabolic cues and regulated diet-induced obesity in mice [77]. Recent insights expand on how microbiota programmed circadian rhythmicity of H3K4me3 and HDAC3-dependent H3K9ac and H3K27ac, in small intestines and colons [78,79]. HDAC3 increased in IECs after conventionalization of GF mice and further could be sufficiently induced by monocolonization of GF mice with Bacteriodetes thetaiotamicron [78]. Moreover, binding of HDAC3 to its target genes was reduced in the absence of circadian clock repressor REV-ERBα [78], implicating the requirement for sensing circadian rhythmicity. While HDAC3 activity was inhibited with colonization of butyrate producing Faecalibacterium prausnitzii, HDAC3 activity in IECs actually increased in microbiome-replete mice compared to GF mice. It was subsequently revealed that microbial-metabolite inositol triphosphate (IP3) and phytate could activate HDAC activity [74^••] demonstrating divergent HDAC modulation by microbial derived metabolites. IECs from mice colonized with phytate producing Escherichia coli had decreased H3K9ac at loci regulated by HDAC3 [74^••] and administration of phytate attenuated DSS colitis in mice [74^••]. In terms of innate immune functions of IECs, microbiota-dependent transcriptional suppression of the C-type lectin receptor Clec2e in IECs correlated with decreased histone acetylation and recruitment of HDAC3 to reduce adherence of Citrobacter rodentium [80]. In another study, enteric infection with C. rodentium led to overall increased HDAC activity in IECs which was required for proper IFNγ production by interspersed CD8 lymphocytes to clear infection [22^•]. Thus, a key mechanism of how microbiota dictate intestinal epithelial cell function is via histone deacetylases, particularly HDAC3, that senses metabolism, circadian rhythmicity and microbial cues in order to coordinate responses and maintain gut homeostasis (Figure 2).

Figure 2.

Epigenetic mechanisms as the intermediary of microorganism-host communication. Schematic of enteric microorganism and metabolic cues that regulate intestinal epithelial cell and immune cell development and function via epigenomic mechanisms. Individual chromatin modifying enzymes demonstrated to be important in this cross-talk are indicated. This figure was created using Biorender.

Crosstalk between the microbiome and epigenome in intestinal-resident immune cells

Intestinal immune cells play an important role in sensing and discriminating between symbionts and pathobionts. Microbiota drive differentiation of unique immune cell populations, including colon macrophage or innate lymphoid cell populations and are indispensable for imprinting tissue specific functions on these cells [60,81] (Figure 2). For instance, macrophages differentiated in the presence of microbial-derived metabolite butyrate displayed enhanced antimicrobial activity against pathogenic bacteria [70^••]. This was less so for propionate and not noted with acetate, suggesting unique roles of the various SCFAs [70^••]. Other studies similarly observed increased macrophage antibacterial activity [82], reduced NF-κB activation and cytokine production with butyrate or propionate co-stimulation [83,84] (Figure 2). The converse was observed in intestinal macrophages from ABX treated mice [85]. Importantly, butyrate was unable to enhance antimicrobial activity in macrophages with HDAC3 silencing [70^••]. Lactobacillus reuteri, a common commensal bacteria, produced a novel form of folate (vitamin B9) that can transfer 2 carbons onto homocysteine resulting in ethionine [86]. Mass spectrometry of human cell line THP-1 histones revealed incorporation of ethionine instead of methionine into proteins, a reduction of histone methylation, and increased ethylation of histone lysine residues [86]. Furthermore, ethionine pre-treatment in THP-1 cells prevented NF-κB activation and TNF transcription following LPS stimulation [86]. Conversely, lactate derived from pathogenic Staphylococcus aureus inhibited HDAC11, unchecked activation of HDAC6, and increased histone 3 acetylation at the IL10 promoter, resulting in enhanced IL10 transcription in mouse host macrophages [87]. These findings present a complex picture of how metabolites from commensal or pathogenic bacteria immunomodulate host chromatin in diverse ways and the ability of epigenetic enzymes to calibrate host responses to diverse microbial signals. Recent findings further support the idea that microbiota epigenetically ‘train’ the state of immune cells. Intestinal bacterium Clostridium scindens provided microbial processed-bile salt, deoxycholate mediated-protection against later Entamoeba histolytica infection by altering H3K4me3 and H3K27me3 in bone marrow granulocyte progenitor cells, leading to expansion and differentiation of intestinal neutrophils [88]. Similarly, microbiota were critical for dendritic cell (DC) responses since DCs from GF and ABX mice could not adequately produce type I interferon (IFN) [89] (Figure 2). This was not due to a defect in intrinsic innate signaling but rather due to chromatin changes with lower levels of H3K4me3 on transcriptional start sites of many inflammatory primary response genes or IFN stimulatory genes and enriched H3K27me3 on metabolic processes that were critical in priming subsequent T cell responses [89,90^••]. Thus, commensal microbiota and their metabolites instruct intestinal innate immune cells via chromatin mechanisms to calibrate and fine-tune transcriptional responses.

In parallel, the microbiota is critical for regulation of adaptive immune responses and maintenance of tolerance towards commensal microbes via the epigenome (Figure 2). Butyrate, and to a lesser extent propionate, programs generation of T regulatory cells (Tregs) in the periphery and colon by inhibiting HDAC activity at the Foxp3 locus [91–93]. Conversely, butyrate suppressed Th17 differentiation [94–96]. Butyrate, or supernatant of F. prausnitzii, administration in vivo alleviated colitis [94,97]. Other less abundant SCFAs, such as pentanoate, also affect Th17 proliferation and IL17A production [98^•]. Comparative assessment of SCFAs revealed pentanoate, butyrate and propionate had potent HDAC-inhibitory activity in T lymphocytes, while acetate and hexanoate had little activity [93,98^•]. Pentanoate impacted B regulatory cells (Bregs) but had no impact on Tregs [98^•]. Other metabolites such as secondary bile acids (SBA) also modulate Th17 and Treg differentiation with some evidence that this was via chromatin architecture [99,100^•]. 3-oxolithocholic acid and isoallolithocholic acid had reciprocal roles with 3-oxoLCA inhibiting Th17 cells differentiation while isoalloLCA increased Tregs [100^•]. Administration of SBAs, lithocholic acid, deoxycholic acid and ursodeoxycholic acid alleviated DSS-induced, TNBS-induced, T cell transfer colitis and Clostridiodes difficile infection models [73^••,101]. Furthermore, there is some evidence that they may similarly act as an HDAC inhibitor [102]. These studies collectively highlight the cell-specific, metabolite-specific programming by microbiota. However, most work has focused thus far on SCFAs and HDACs in this process, as HDAC activity is an easier output to assay. Understanding the underlying chromatin mechanisms governing immunomodulation by other microbial metabolites, at defined loci, will be an exciting avenue of future research.

Concluding remarks, perspectives and future directions

IBD is a multifactorial disease, that is more likely multiple subsets of diseases, making both accurate diagnosis and therapeutic management challenging. Substantial progress has been made in the genetic understanding of IBD subtypes and onset, yet genetic factors explain only a small portion of overall disease incidence. Environmental triggers such as perturbation of enteric microorganisms are also a major factor for disease stratification; however, pathways underlying how cues from microbiota are integrated during development of IBD remain poorly understood. Epigenetics links genetic predisposition with disturbances of microorganisms in the pathogenesis of IBD and may offer novel precision medicine diagnosis and therapeutic opportunities.

Indeed, the importance of epigenetics in mediating host-microorganism-metabolism relationships in health and IBD is rapidly emerging. Furthermore, alterations in the epigenome or genetic variants of chromatin enzymes is materializing as a causative factor in IBD. However, our grasp of the mechanisms, the relationship between lesser-studied microorganisms such as enteric fungi and viruses and the epigenome, the role of novel microorganism-derived metabolites in altering the host epigenome and the extent to which altered epigenetics in specific cell types directly contribute to disease is still in its infancy. One of the fundamental questions that still remains unanswered is whether rescue of epigenetic, microbiome or metabolic alterations with the ever-expanding array of epigenetic-targeting drugs is a feasible strategy in complex immune disorders such as IBD. Deciphering the critical epithelial, immune or neuronal pathways specifically regulated by certain chromatin-modifying enzymes in response to changing commensal microorganisms and their metabolites and how this contributes to distinct phenotypes of IBD will be essential. The rapid improvement of chromatin techniques in limited cell numbers that are becoming more accessible to immunologists will help achieve this goal. Finally, a more expansive and unbiased assessment of regulation of lesser-studied aspects of the epigenome by commensal microbiota or novel metabolites in individual cell types is warranted and may also reveal unexpected avenues for therapeutic intervention of the epigenome-metabolome-microbiome axis.