Intestinal Barrier Dysfunction in Fatty Liver Disease: Roles of Microbiota, Mucosal Immune System, and Bile Acids

Abstract

Nonalcoholic fatty liver disease (NAFLD) describes a spectrum of progressive liver diseases ranging from simple steatosis to steatohepatitis and fibrosis. Globally, NAFLD is the leading cause of morbidity and mortality associated with chronic liver disease, and NAFLD patients are at a higher risk of developing cirrhosis and hepatocellular carcinoma. While there is a consensus that inflammation plays a key role in promoting NAFLD progression, the underlying mechanisms are not well understood. Recent clinical and experimental evidence suggest that increased hepatic translocation of gut microbial antigens, secondary to diet-induced impairment of the intestinal barrier may be important in driving hepatic inflammation in NAFLD. Here, we briefly review various endogenous and exogenous factors influencing the intestinal barrier and present recent advances in our understanding of cellular and molecular mechanisms underlying intestinal barrier dysfunction in NAFLD.

Graphical Abstract

Nonalcoholic fatty liver disease (NAFLD), commonly associated with metabolic syndrome (MetS) and type 2 diabetes, is a leading cause of chronic liver disease worldwide.^1–4 NAFLD is defined as a spectrum of progressive liver diseases ranging from bland steatosis or nonalcoholic fatty liver to its more advanced forms of nonalcoholic steatohepatitis (NASH) and fibrosis.^1,3 NASH patients are at a substantially higher risk of progressing to cirrhosis and ultimately hepatocellular carcinoma.^1,3,5,6 In the United States, the incidence of NAFLD in the general population is estimated at 30 to 40% and may be as great as 75 to 90% in obese individuals.^2,7 Modeling studies predict a substantial rise in the prevalence of NAFLD, NASH, and associated end-stage liver diseases over the next decade highlighting the importance of finding therapeutic interventions for NAFLD.⁸

The etiology of NAFLD is not clearly defined, but there is a consensus that inflammation plays key role in the progression of the disease.^5,9–14 Accumulating clinical and experimental evidence give credence to the idea that gut-derived antigens are the primary drivers of hepatic inflammation that promote NAFLD progression, distilling down to the importance of maintaining a healthy gut microbiota and functional intestinal barrier. Here, we review critical host and environmental factors that regulate the intestinal barrier in health and dysfunction in disease, and summarize recent developments in our understanding of cellular and molecular mechanisms underlying intestinal barrier dysfunction in NAFLD.

The Intestinal Barrier

The intestinal barrier, also referred to as the intestinal mucosal barrier, is a dynamic, multilayered, highly organized, and compartmentalized structure providing physical, biochemical, and immunological protection against luminal microorganisms, digestive enzymes, and dietary components (►Fig. 1).^15–18 The intestinal epithelium is formed by a monolayer of columnar epithelial cells interconnected by intercellular junctional complexes creating a physical barrier crucial for maintaining the architecture and selective permeability of the intestinal barrier.^15,17 Intestinal epithelial cells (IECs) are a functionally diverse group that include enterocytes responsible for the absorption of nutrients and water, goblet cells that produce mucus, enteroendocrine cells that secrete hormones, and Paneth cells that secrete antimicrobial factors. Mucosal epithelium also contains chemosensory tuft cells that are instrumental in defending against parasites, and microfold cells (M cells) that help maintain mucosal immune tolerance. M cells also provide mucosal immune surveillance by continuously sampling luminal contents and transporting antigens, including intact microorganisms, to antigen-presenting cells such as dendritic cells (DCs) and B cells in the submucosa for processing and leading to the initiation of immune response. The majority of IECs are common to both the small intestine and colon except for Paneth cells and M cells which are restricted to the small intestine. The small intestine contains both crypt and villi as well as a small number of goblet cells, whereas the large intestine possesses only crypts and a significantly larger population of goblet cells.^15,17,19

Fig. 1.

The intestinal barrier. The intestinal mucosal barrier is a dynamic, multilayered, highly organized, and compartmentalized structure providing physical, biochemical, and immunological protection against luminal contents. The intestinal epithelium is formed by a monolayer of columnar epithelial cells interconnected by intercellular junctional complexes creating a physical barrier crucial for maintaining the architecture and selective permeability of the intestinal barrier. Intestinal epithelial cells (IECs) are a functionally diverse group that include mucus-producing goblet cells, hormone-secreting enteroendocrine cells, and antimicrobial factor-secreting Paneth cells. The epithelial barrier is reinforced on the extraepithelial side by a mucus layer which forms the first line of defense against luminal microorganisms. The antimicrobial factors secreted by Paneth cells and enterocytes act as a biochemical barrier to further reinforce the mucus barrier in the extraepithelial side. The immunological barrier is maintained by a highly specialized immune network which fortifies the epithelial barrier on the basal side. The mucosal immune system regulates barrier function by expediently detecting and eliminating pathogens penetrating the epithelium while maintaining tolerance for commensal microbiota. Collectively, the diverse functions of the IECs along with the immune cells form a dynamic barrier that not only provides physical separation of the luminal content but also integrates and transmits signals from commensal bacteria to mucosal innate and adaptive immune cells to maintain intestinal mucosal homeostasis. AMP, antimicrobial peptides; IEL, intraepithelial lymphocytes; Tregs, regulatory T cells.

The epithelial barrier is reinforced on the extraepithelial side by a mucus layer which forms the first line of defense against luminal microorganisms. Mucins, highly glycosylated proteins secreted by goblet cells, are the foundation of the mucus layer and form a net-like structure that occludes large particles, including bacteria, from directly contacting the epithelium. The organization of the mucus layer varies along the digestive tract according to function and luminal environment. The colonic mucosal layer is thicker since it harbors the great majority of gut microbiota and is comprised of two distinct layers. The inner layer is stratified, anchored to the epithelium, and is largely devoid of bacteria. The proteolytic degradation of the polymerized mucins in the inner layer gives rise to the looser and more permeable outer layer. The outer layer of the colon harbors microbiota that metabolize complex polysaccharides and dietary fibers to produce beneficial metabolites that serve as an energy source for the intestinal epithelium. Conversely, the mucus layer in the small intestine has only one loose, unattached, easily removable layer that limits bacterial growth by trapping and removing bacteria.^20,21 The antimicrobial factors secreted by Paneth cells and enterocytes act as a biochemical barrier to further reinforce the mucus barrier in the extraepithelial side. In the small intestine, the mucus contains high concentrations of antimicrobial peptides and proteins that neutralize and trap bacteria.^15,17 The immunological barrier is maintained by a highly specialized immune network which fortifies the epithelial barrier on the basal side. The mucosal immune system regulates barrier function by expediently detecting and eliminating pathogens penetrating the epithelium while maintaining tolerance for commensal microbiota. In response to pathogens and metabolic stress, the epithelial cells also secrete proinflammatory cytokines and reactive oxygen species to stimulate immune responses against invading organisms.^15,17,22 Collectively, the diverse functions of the IECs along with the immune cells form a dynamic barrier that not only provides physical separation of the luminal content but also integrates and transmits signals from commensal bacteria to mucosal innate and adaptive immune cells to maintain intestinal mucosal homeostasis.

Epithelial Junctional Complexes and Intestinal Barrier Permeability

The selective transport of solutes across the intestinal epithelium occurs through two major interdependent transport routes, transcellular, and paracellular. Transcellular transport of small solutes is primarily regulated by selective membrane-bound transporters, and transport of large macromolecules, including intact proteins, antigens, and bacteria is regulated by receptor-mediated endocytosis.²³ The M cells in the Peyer’s patches in the distal small intestine play a major role in the transcellular transport of luminal antigens and microorganisms. Recently, goblet cells have been shown to be also involved in the transcellular sampling of luminal antigens by forming goblet cell-associated antigen passages.¹⁹

Paracellular transport is the passive movement of water and solutes through the space between epithelial cells and is regulated by intercellular junctional complexes that include tight junction (TJ), adherens junction (AJ), and desmosomes localized at the apical-lateral membrane junction and along the lateral membrane (►Fig. 2). The TJ and AJ reside at the apical end of the lateral membrane and are often referred to as the apical junctional complex. AJs are primarily associated with the initiation and maintenance of cell-cell adhesion and are essential for TJ assembly and stabilization. The single-pass transmembrane proteins epithelial (E) cadherin and nectin are the major components of the AJs and play cooperative roles in their formation. The desmosomes are composed of transmembrane cadherins desmogleins and desmocollins and provide mechanical strength to the epithelium.^24–27 The TJs are responsible for sealing the paracellular space and their primary function is to restrict paracellular diffusion of solutes based on size and charge. The TJ complex, as revealed by freeze-fracture electron microcopy and immunoelectron microscopy, consists of a meshwork of fibrils formed by rows of transmembrane proteins. These are linked to an electron-dense cytoplasmic junctional plaque formed by a network of scaffolding and adaptor proteins, signaling components, and actin-binding cytoskeletal linkers. The transmembrane proteins claudins, occludin, junctional adhesion molecules (JAMs), tricellulin, MARVEL domain-containing protein 3, and coxsackievirus and adenovirus receptor are the major components of the transmembrane strands that represent the semipermeable paracellular diffusion barrier. A complex network of cytosolic proteins that includes the zonula occludens (ZO) and cingulin are the major components of the junctional plaque. Several key signaling proteins and their regulators are also associated with the TJ complex and play a role in regulating TJ function. Apart from their role in the regulation of paracellular transport, intercellular junctional complexes are crucial in maintaining the integrity of the epithelium by regulating proliferation, differentiation, migration, and epithelial cell polarity. A detailed review of the structure and functional properties of intercellular junctional proteins can be found elsewhere.^24,26–31

Fig. 2.

The intestinal epithelial intercellular junctional complexes. The intestinal epithelial cells are interconnected by three major intercellular junctional complexes, tight junction (TJ), adherens junction (AJ), and desmosomes. The TJ and AJ reside at the apical end of the lateral membrane and are referred to as apical junctional complexes (AJC). The TJs are responsible for sealing the paracellular space and their primary function is to restrict paracellular diffusion of solutes based on size and charge. The TJ complex consists of a meshwork of fibrils formed by rows of transmembrane proteins including occludin, claudins, junctional adhesion molecular A (JAM-A), and coxsackievirus and adenovirus receptors (CAR). These are linked to an electron-dense cytoplasmic junctional plaque formed by a network of scaffolding and adaptor proteins, signaling components, and actin-binding cytoskeletal linkers. AJs are primarily associated with the initiation and maintenance of cell-cell adhesion and are essential for TJ assembly and stabilization. The single-pass transmembrane proteins epithelial (E) cadherin and nectin are the major components of the AJs and play cooperative roles in their formation. The desmosomes are composed of transmembrane cadherins desmogleins and desmocollins and provide mechanical strength to the epithelium. Apart from their role in the regulation of paracellular transport, intercellular junctional complexes are crucial in maintaining the integrity of the epithelium by regulating proliferation, differentiation, migration, and epithelial cell polarity. ZO, zonula occludens.

The paracellular transport system is further divided into two routes based on size and charge, the pore pathway and the leak pathway (►Fig. 3). The pore pathway is a high-capacity, charge- and size-selective system which allows the passage of small molecules and specific ions. The pore pathway is primarily regulated by claudins, a large family of integral membrane proteins forming either anion-selective (claudins10a and 17) or cation-selective (claudins 2, 10b, and 15) paracellular channels. Claudins are further grouped into tight claudins (1, 3, 4, 5, and 18) and leaky claudins (2, 7, 10, and 15) based on their effect on epithelial permeability. Presence of tight claudins is associated with increased barrier tightness whereas the leaky claudins are associated with increased permeability. Expression of specific claudins varies along the villus-crypt axis according to the mucosal permeability found along the intestine in homeostasis, but can be modified by inflammatory cytokines such as tumor necrosis factor-α (TNFα), interferon-γ (IFNγ) and interleukin (IL)-1-β. The TJ protein JAM-A also plays a critical role in regulating paracellular permeability and genetic deletion of JAM-A in mice results in increased intestinal permeability.^{18,23,29–33}

Fig. 3.

The paracellular transport pathways. Paracellular transport is the passive movement of water and solutes through the space between epithelial cells and is regulated by intercellular junctional complexes localized at the apical-lateral membrane junction and along the lateral membrane. The paracellular transport is facilitated by two major pathways: the pore pathway and the leak pathway. The pore pathway is a high-capacity, charge-and size-selective permeability pathway that permits the passage of specific ions and small molecules with diameters generally less than 8 Å. The leak pathway is a low-capacity, charge- and size-independent permeability mechanism that allows the passage of macromolecules as large as 100 Å, including proteins and bacterial lipopolysaccharides. In addition to the pore and leak pathways, events featuring severe epithelial damage ensue an unrestricted pathway, wherein bulk luminal contents including live bacteria can freely diffuse across the intestinal epithelial barrier in an unregulated manner.

The leak pathway is governed by occludin, ZO-1, and myosin light chain kinase (MLCK) and is a low-capacity-charge and size-independent transport mechanism permeable to larger macromolecules including proteins and bacterial lipopolysac-charides (LPS). Despite advances in our understanding of the molecular mechanisms that enable ion permeability, macromolecular diffusion across the TJ is not well understood. Paracellular transport is altered in various pathological situations including inflammatory bowel disease (IBD) where modifications in TJ function and composition allow higher molecular weight molecules to diffuse across the epithelial layer. In addition to the leak and pore pathways, in the event of severe epithelial damage as observed in necrotizing enterocolitis, luminal content including live bacteria can cross the intestinal barrier in an unregulated manner resulting in severe mucosal inflammation (►Fig. 3).^{18,31,34–38}

Intestinal Barrier Dysfunction in NAFLD

Evidence that gut microbial antigens have a role in metabolic diseases comes from clinical and preclinical studies finding an association between systemic increase in LPS termed “metabolic endotoxemia” and low-grade inflammation in obesity and type 2 diabetes.^39–44 Studies by Pendyala et al found that even short-term exposure to a diet rich in fat, fructose, and cholesterol (western diet [WD]) increases endotoxemia by 71% in healthy subjects.⁴⁴ A study comparing the effect of glucose and fructose on endotoxemia in pediatric NAFLD patients found that consumption of beverages containing fructose but not glucose results in an immediate increase in serum LPS which remained high for the duration of the 4-week study.⁴⁵ Prolonged fructose consumption has also been found to increase endotoxemia in nonhuman primates and is associated with increased incidence of NAFLD and type 2 diabetes.⁴⁶ Cani et al found that mice infused with low-dose LPS for 4 weeks to induce metabolic endotoxemia develop obesity, steatosis, and hepatic insulin resistance, a phenotype similar to mice fed a WD for 4 weeks.⁴⁰ These studies along with evidence that loss of Toll-like receptor (TLR) function or depletion of gut microbiota mitigates metabolic abnormalities associated with nutrient excess support the hypothesis that an impaired intestinal barrier allows gut microbial antigens to translocate thereby contributing to metabolic diseases.^42,47–49

The earliest evidence of increased gut permeability in NAFLD patients was provided by Miele et al who found that biopsy-proven NAFLD patients have increased intestinal permeability and decreased expression of TJ protein ZO-1 in the intestine compared with healthy subjects.⁵⁰ Increased intestinal permeability and serum endotoxin levels have also been reported in children with NAFLD.^51–53 Several studies reported an association between higher serum endotoxin levels and hepatic TLR4 expression in NAFLD/NASH patients suggesting a role for gut-derived endotoxins in promoting hepatic inflammation through TLR4 activation.^54,55 A recent study found that, compared with healthy controls, NAFLD patients exhibit a lower level of TLR4 codon 299 heterozygous gene mutation (Asp299Gly), which is associated with blunting the negative impacts of endotoxins.⁵⁶ Furthermore, a meta-analysis of clinical studies reporting intestinal permeability in NAFLD patients found that 40% of NAFLD and 50% of NASH patients have increased intestinal permeability.⁵⁷ Consistent with these epidemiological studies that report an association between increase intestinal permeability and hepatic injury, our laboratory group recently demonstrated that mice with a defect in intestinal epithelial barrier, JAM-A knockout (KO) mice (JAM-A encoded by F11r), develop severe hepatic inflammation and fibrosis compared with controls when fed a WD for 8 weeks. In the WD-fed F11r^−/− mice, liver injury was associated with increased mucosal inflammation, disruption of TJ integrity, and increased intestinal permeability to bacterial endotoxins. Investigation of the gut microbiota in F11r^−/− mice fed WD revealed an increase in Proteobacteria and Firmicutes but decrease in Bacteroidetes, which resembles the composition of gut microbiota in NASH patients with advanced fibrosis.⁵⁸ Antibiotic treatment to deplete gut microbiota not only alleviated steatohepatitis in the F11r^−/− mice but also improved metabolic derangements associated with NAFLD in both the KO and wild-type (WT) mice, underscoring the importance of gut microbial antigen translocation in promoting proinflammatory immune responses in the liver. Our studies investigating JAM-A expression in the colonic mucosa of a small cohort of NAFLD patients revealed decreased JAM-A protein expression in NAFLD patients compared with controls suggesting a role of JAM-A in promoting intestinal barrier dysfunction in NAFLD. This study provides experimental evidence for the concept that gut microbial antigens are a significant driver of hepatic inflammation and fibrosis as well as metabolic comorbidities associated with NAFLD (►Fig. 4). From a therapeutic perspective, these findings reinforce the hypothesis that restoring the integrity of intestinal epithelial barrier could prove an effective therapeutic strategy for NAFLD.⁵⁹

Fig. 4.

Intestinal barrier dysfunction in nonalcoholic fatty liver disease (NAFLD). The intestinal epithelial cells (lECs) along with the immune cells form a dynamic barrier that not only provides physical separation of the luminal content but also integrates and transmits signals from commensal bacteria to mucosal innate and adaptive immune cells to maintain intestinal mucosal homeostasis (left). Commensal microbiota maintains mucosal homeostasis through various mechanisms; their ability to promote expansion of regulatory T cells (Tregs) at the expense of Th17 cells is crucial to this role. Tregs through production of anti-inflammatory cytokines such as interleukin (IL)-10 not only maintain immune tolerance in the mucosa but also promote antimicrobial defense by inducing mucus production by goblet cells, antimicrobial peptide (AMP) production by Paneth cells and enterocytes, and immunoglobulin A (IgA) production by plasma cells. IL10 also suppress unwanted activation of mucosal immune cells, including pathogenic Th17 cells, effector T cells, and innate immune cells. Microbial metabolites such as short-chain fatty acids (SCFAs) also participate in maintaining mucosal homeostasis by promoting Treg differentiation and function and by enhancing enterocyte function. In NAFLD, gut dysbiosis characterized by a higher abundance of Gram-negative bacteria belonging to the phylum Proteobacteria and a reduction in bacteria from the Bacteroidetes and Firmicutes phyla disrupts the healthy crosstalk between the microbiota and mucosal immune cells resulting in the loss of mucosal immune tolerance (right). The resulting increase in resident immune cell activation and recruitment of proinflammatory effector T cells to the mucosa plays a central role in fueling chronic inflammation by increasing proinflammatory cytokines such as tumor necrosis factor-α (TNFα) and interferon-γ (IFNγ) in the mucosa. These proinflammatory cytokines disrupt intestinal barrier resulting in increased translocation of microbial products which further fuels mucosal inflammation and deteriorates intestinal barrier function. Microbial dysbiosis in NAFLD also increase colonic bile acids (BAs), contributing to the loss of barrier function. The resulting increase in the translocation of gut microbial products to the liver fuels hepatic inflammation and fibrosis resulting in the progression of NAFLD. Apart from the innate immune cells including recruited monocytes in the liver, our studies show that diet-induced, microbiota-dependent recruitment of α₄β₇+ CD4 T cells to the liver also critically drive hepatic inflammation and fibrosis in NAFLD. Unpublished data from our group indicates a role of α₄β₇/MAdCAM-1 axis in the regulation of monocyte recruitment to the liver in NAFLD.

Regulation of Intestinal Barrier by the Mucosal Immune System

The immune system in the gastrointestinal (GI) tract performs a delicate balancing act of maintaining homeostasis—mediating tolerance against food antigens, commensals, and self-antigens—while poised to mount a brisk and effective immune response against harmful microbes. Critical to this balance is an intact epithelial barrier fortified by a highly specialized immune network collectively referred to as the gut-associated lymphoid tissue (GALT).^22,60,61 The GALT is a component of mucosa-associated lymphoid tissue, an intricate and relatively autonomous immune network comprised of both innate and adaptive immune cells in mucosal tissues including the respiratory tract, digestive tract, and urogenital tract. Important lymphoid elements associated with the gut include the ileal Peyer’s patches, mesenteric lymph nodes, lamina propria mononuclear cells, and intraepithelial lymphocytes. Based on anatomical and functional properties as it relates to priming the immune response, the GALT can be differentiated into inductive and effector sites. Within the GALT, Peyer’s patches and the mesenteric lymph nodes serve as inductive sites specialized in fostering antigen presentation by resident DCs for T cell priming/activation and generation of effector T cells essential for adaptive immune responses. Interaction of T cells with resident DCs imprints gut-specific homing in effector cells by inducing expression of adhesion molecules such as α₄β₇ and the chemokine receptor CCR9, in a retinoic-acid-dependent manner. These effector T cells egress the GALT to reenter the circulation via the efferent lymphatics and subsequently home back to the intestinal lamina propria, a key effector site within the GI tract.^61–63

In addition to the acquisition of homing receptors, T cell priming in inductive lymphoid sites by DCs dictates CD4 T helper differentiation. Upon priming, commensal-specific T cells in the gut mucosa differentiate into either T helper 17 (Th17) or regulatory T cells (Treg)—subsets critical for intestinal mucosal homeostasis. The acquisition of CD4 effector cell differentiation is specified by key lineage-determining cytokines, IL-6, IL-12, IL-23, and the chemokine transforming growth factor (TGF)-β. The cytokine IL-6 together with TGFβ and IL-23 drive Th17 differentiation. When there are high concentrations of TGFβ, regulatory T cell differentiation is induced at the expense of Th17 cells via activation of the transcription factor forkhead box P3 (Foxp3).^64,65 IL-22 production by Th17 cells stimulates epithelial cells to enhance antimicrobial defenses and enhances epithelial barrier integrity, while regulatory T cells are critical for the maintenance of an anti-inflammatory milieu in the intestine.^66,67 Consequently, the balance of Th17/Tregs/Th1 cells is vital to GI health.

Dysregulation of mucosal immune balance with a resulting increase in proinflammatory cytokines, most notably IFNγ and TNFα, has deleterious consequences for the intestinal barrier. Increased frequencies of IFNγ-producing proinflammatory Th1 CD4 subsets are reported in inflammatory GI diseases.^68–70 The cytokine IL-12 specifies Th1 polarization and Th1 effectors are induced when the balance of cytokine lineage is shifted in favor of IL-12 in response to inflammation or injury. Increased intestinal permeability following T cell activation is mediated by TNFα-dependent inhibition of the epithelial Na⁺/K⁺-ATPase enzyme as well as MLCK-dependent TJ disruption.^71,72 IFNγ disrupts the intestinal barrier through mechanisms associated with a decrease or delocalization of TJ proteins ZO-1, JAM-A, occludin, and claudins as well as nuclear factor kappa B (NF-κB)-dependent disruption of actin cytoskeleton.^73–76 An excess of Th1 CD4 T cells in the absence of restraining anti-inflammatory cytokines, such as IL-10 or TGFβ, underlies the pathology of several chronic inflammatory GI diseases.^77–79 This pathophysiology is orchestrated in conjunction with several key players of the innate immune system such as neutrophils, monocytes, and macrophages, as described elsewhere.^80–84

In addition to T cells and innate immune cells, innate counterparts of T cells referred to as innate lymphoid cells (ILCs) are increasingly recognized as important components in the pathogenesis of chronic inflammatory GI diseases.^85–87 ILCs are abundant at the mucosal barriers where they are primed to rapidly respond to pathogens and tissue damage by secreting cytokines. These include classic Th1 and Th17 cytokines such as IFNγ, IL-12, IL-17, IL-18, IL-22, and IL-23.^88,89 Studies in humans and animal models suggest that ILCs cooperate with T cells to maintain barrier integrity. As example, depletion of ILCs in a mouse model of nonobese diabetic-recombination activating gene-1 (NOD-Rag1 null) IL-2 receptor common gamma chain double-deficient mice resulted in systemic inflammation triggered by increased translocation of commensal bacteria. The inflammatory phenotype in these mice was rescued by IL-22 suggesting that ILC-derived IL-22 contributes to barrier function.⁹⁰ On the other hand, accumulation of ILC1s in human IBD patients and ILC3s in inflamed colons of mice infected with Helicobacter hepaticus and other models of intestinal inflammation have been reported.⁹¹ Therefore, the emerging role of ILCs suggests they are beneficial in the maintenance of mucosal homeostasis, but untoward activation of ILCs may aggravate mucosal injury and cause loss of barrier function.^{85,87,89,91,92}

The highly interconnected and specialized immune repertoire of the GALT with a panoply of cytokines and cellular contributors actively maintains intestinal homeostasis. Mucosal immune cell subsets maintain the integrity of the intestinal barrier including TJs, mucous production, and mucosal remodeling and repair. Disruption of this balance can potentiate immunopathology; these findings underscore the need to understand whether the relative balance of mucosal immune cell subsets is skewed in NAFLD and if restoring mucosal immune homeostasis represents an effective strategy to prevent disease progression.

Dysregulation of Mucosal Immune Homeostasis in NAFLD

Aberrant activation of mucosal immune cells is the most common mechanism underlying mucosal inflammation in inflammatory GI diseases as well as metabolic diseases.^93,94 While little is known about the role of the mucosal immune system and its influence on the integrity of the intestinal barrier, increased immune cell infiltration and proinflammatory cytokine levels indicative of mucosal inflammation are reported in both humans and animal models of NAFLD and NASH.⁵⁹ Suggesting a role of mucosal immune cells in the loss of barrier function in NAFLD, studies from our group show that mucosal inflammation in mice fed a WD was associated with increased infiltration of proinflammatory effector CD4 T cells in the intestine.⁹⁵ Further, data from WD-fed mice revealed an increase in systemic CD4 T cells expressing the heterodimeric integrin receptor α₄β₇, and increased recruitment of these α₄β₇+ CD4 T cells in the colonic mucosa. This recruitment correlated with microbiota-dependent expression of MAdCAM-1 in the colonic tissue. Monoclonal antibody (mAb)-mediated blockade of α₄β₇+ CD4 T cell recruitment to the intestine reduced mucosal inflammation, improved the intestinal epithelial barrier, and reduced hepatic inflammation and fibrosis. Interestingly, α₄β₇ mAb treatment also reduced hepatosteatosis and improved MetS indices in the WD-fed mice suggesting a role of α₄β₇/MAdCAM-1 axis in modulating hepatic metabolism. Our studies also revealed the involvement of α₄β₇/MAdCAM-1 axis in the recruitment of α₄β₇⁺ CD4 T cells to the NASH liver.⁹⁵ While the exact sequelae, the tissue compartment, that is, the liver versus the intestine where α₄β₇ is induced, are unclear, this new evidence points to the critical role of the mucosal immune system in the intestinal barrier dysfunction in NAFLD (►Fig. 4). Additional research is required to pinpoint the role of specific immune subsets in modulating intestinal barrier function in NAFLD and other metabolic diseases.

Intestinal Barrier Function and Gut Microbiota

The human GI tract harbors a complex and dynamic microbial community that live in a symbiotic relationship with the host. The large intestine is home to almost 100 trillion microorganisms including bacteria, archaea, fungi, and viruses collectively referred to as the gut microbiota. Gut bacteria are primarily comprised of bacteria belonging to the phyla Bacteroidetes, Firmicutes, Actinobacteria, Proteo-bacteria, Fusobacteria, and Verrucomicrobia, with Bacteroidetes and Firmicutes accounting for over 90% of the community. Gut microbiota, at various anatomical sites within the GI tract, is conserved at the higher levels of classification, but there is a high degree of variability at the species and strain level across healthy individuals due to various host and environmental factors, including diet, early microbial exposure, and host genetics. Antibiotics and pharmaceuticals have a profound effect on host microbial composition, as well. Despite the taxonomic diversity, biochemical functions encoded by microbial metagenome are more stable, and there appears to be a baseline stability in the biochemical functions of gut microbiome in healthy individuals.^96,97 The established biochemical functions of the microbiome critical for host physiology include synthesis of enzymes, hormones, vitamins, and neurotransmitters, absorption of lipids, bile acids (BAs), and amino acids, and metabolism of dietary fibers to produce bioactive short-chain fatty acids (SCFAs).^98–100 In addition to the SCFAs constituting an energy source for colonocytes, enhancing epithelial barrier integrity, and elaborating local and systemic immunomodulatory functions, gut microbiota produce numerous primary and secondary metabolites that contribute to a myriad of biochemical functions critical for metabolic, neurological, and endocrine homeostasis.^101–103 Emerging evidence reveal that gut microbiota also play critical roles in the induction, education, and function of the host immune system.^104,105 Consequently, disruption of the symbiotic host-microbial interaction due to changes in the composition and the function of gut microbiota profoundly impact host physiology and have been linked to numerous chronic health conditions including obesity, MetS, cardiovascular disease, NAFLD, IBD, and cancer.^106–113 A more detailed examination of the gut microbiota is beyond the scope of this review and the interested reader is directed to several excellent recent reviews on the topic.^{96,97,114–116}

The crosstalk between the luminal microbiota and the host is essential for the development and maintenance of the intestinal epithelial barrier. The IECs and immune cells are equipped with pattern recognition receptors such as TLRs, NOD-like receptors (NLR), and G-protein coupled receptors that recognize specific microbiota-associated molecular patterns (MAMPs) and microbial metabolites. Mechanistic studies demonstrate that mice deficient in TLR or NLR signaling develop more severe mucosal injury when treated with colitogenic agents due to defects in proliferation and survival of IECs.^117–119 Similarly, illustrating the protective role of microbiota in restoring mucosal homeostasis following injury, disruption of microbiota with antibiotics exacerbates mucosal injury in mice treated with colitogenic agents.¹²⁰ Antimicrobial peptide secretion by Paneth cells and mucin secretion by goblet cells are severely impaired in mice deficient in TLR and NLR signaling as well as in mice treated with antibiotics and enhance susceptibility to opportunistic infection by enteric pathogens.¹²¹ The importance of microbial-host crosstalk is best illustrated in germ-free mice that exhibit significant abnormalities in intestinal morphology and architecture, including shorter villi and crypts and atrophied Peyer’s patches.^122,123 Proliferation and renewal of IECs are significantly blunted in germ-free mice who also exhibit a thinner mucosal layer ascribed to defects in mucus secretion making them highly susceptible to intestinal injury and infection.^121,122,124

Studies in germ-free mice also illustrate the importance of gut microbiota in the development and maintenance of local and systemic innate and adaptive immune functions. Germ-free mice have underdeveloped innate and adaptive immune systems, exhibit severely stunted GALT formation, and are susceptible to microbial infections.¹²⁵ Compared with conventionally raised mice, germ-free mice have fewer Treg cells, Th17 cells, CD4 + CD8α+ double-positive intraepithelial lymphocytes, and immunoglobulin A (IgA)-producing plasma cells signifying the role of microbiota in modulating specific lineage differentiation in immune cells.^126–130 Several microbial taxa that play a direct role in promoting differentiations of both protective and tolerogenic immune cell lineages have been identified. Examples include Bifidobacterium adolescents and segmented filamentous bacteria that promote Th17 cell differentiation.^127,131,132 Surface polysaccharide A of Bacteroides fragilis promotes expansion of Foxp3+ Tregs through TLR-dependent mechanisms.¹³³ Microbial fermentation byproducts, SCFAs produced by Clostridium species belonging to clusters IV, XIVa, and XVIII, as well as Akkermansia muciniphila and Faecalibacterium prausnitzii induce Treg differentiation as well as suppress inflammatory effector T cells through specific SCFA receptors expressed by immune cells.^132,134,135 These findings suggest that the intestinal epithelium and underlying immune cells must contend with baseline exposure to MAMPs and microbial metabolites to sustain homeostasis and repair. Taken together, accumulating evidence indicates that microbial sensing by the host influences multiple aspects of epithelial barrier function, including epithelial cell renewal, wound healing, TJ repair, as well as the secretion of mucin, antimicrobial peptides, and IgA to limit invasion of pathogens.^21,136–139 Thus, disruption of microbial-host crosstalk brought about by pathogenic microorganisms, environmental factors, or defects in microbial recognition and processing by IECs or immune cells precipitates the onset of inflammation ultimately leading to loss of barrier function.

The human GI tract is also colonized by a diverse population of deoxyribonucleic acid and ribonucleic acid viruses (the gut virome), including eukaryotic viruses, retroviruses, bacteriophages, and archaeal viruses that are capable of infecting host cells and enteric microorganisms.^140–142 The human gut virome is equivalent in number to gut bacterial cells in the lumen, but their abundance can reach as high as 20 times the bacterial cells on gut mucosal surfaces and within the mucus layer.¹⁴³ While, little is known about the role gut virome play in human health and disease, changes in gut bacterial composition have been shown to parallel changes in viral abundance and diversity in several GI and systemic disorders, including IBD, acquired immunodeficiency syndrome, obesity and type 2 diabetes, malnutrition, NAFLD, and alcoholic liver disease (ALD).^144–157 Current evidence suggest that gut virome, specifically gut bacteriophages, the most dominant member accounting for over 90% of the human gut virome, play a crucial role in shaping gut microbial communities through predation and horizontal gene transfer.^143,158 Implying a protective role, studies in animal models of colitis have shown that depletion of gut viruses with antiviral cocktail increases the susceptibility of mice to dextran sulfate sodium-induced colitis. This anti-inflammatory effect is due to the ability of enteric viruses to induce IFNβ secretion by plasmacytoid DCs and is dependent on TLR3 and TLR7 activation by the enteric viruses.¹⁵⁹ Furthermore, bacteriophages present on gut mucosal surfaces have been shown to provide ubiquitous nonhost derived antimicrobial immunity by regulating mucosa-associated gut bacterial population.^160,161 In contrast to the reported anti-inflammatory and protractive role of gut virome, members of the gut virome can also cause gut inflammation. This was apparent in studies demonstrating an expansion of phages that target beneficial microbe Faecalibacterium prausnitzii in IBD.^145,162 Furthermore, reports of bacteriophages from IBD patients that can elicit heightened T cell immune responses by stimulating phage-specific as well as nonspecific production of IFNγ, a proinflammatory cytokine implicated in the pathogenesis of IBD, demonstrate the dark side of gut virome in human diseases.¹⁶³ Due to a lack of adequate tools to identify, classify, and isolate gut viruses, this component of the gut microbiome has been significantly understudied and their role in modulating intestinal barrier is not clear. However, given the richness and the complexity of the human gut viruses, their ability to interact with other components of the gut microbiota and enterocytes, and the extensive host immune repertoire that can be influenced by gut viruses, development of tools to enable in-depth understanding of the role of gut virome in human health and diseases is urgently needed.

The human intestine is also home to a diverse population of fungi collectively termed the gut mycobiome. Unlike gut bacteria, a core gut mycobiome in humans is yet to be established, but evidence till date suggest that members of the fungal genus Candia are the true resident and the most abundant fungi present in healthy human intestine.¹⁶⁴ Members of Saccharomyces and Malassezia genera are also dominant in human stool samples, but due to high dietary and environment abundance of the members of Saccharomyces and Malassezia genera, it is not clear whether they are true residents of human gut.¹⁶⁴ Similar to gut bacteria, gut mycobiome also play a critical role in maintaining intestinal homeostasis and host immunity.^165–167 Dysregulation of human mycobiota primarily indicated by an expansion of Candida spp. has been linked to Crohn’s disease, ulcerative colitis, obesity, type 2 diabetes, NAFLD, ALD, and autoimmune liver diseases.^{157,166–176} A comprehensive review of the literature related to gut mycobiome is beyond the scope of this review and readers are directed to some excellent reviews on this topic.^{165–167,170,174,175,177}

Gut Dysbiosis in NAFLD

A growing body of evidence suggests that the crosstalk between gut and liver plays a pivotal role in the initiation as well as progression of NAFLD. Alterations in gut microbiota (dysbiosis) is commonly observed in NAFLD and metabolic comorbidities associated with NAFLD, such as obesity, diabetes, and cardiovascular diseases. To find a correlation between compositional changes in microbiota and disease severity, several studies have analyzed gut flora from adult and pediatric NAFLD and NASH patients using quantitative reverse transcription polymerase chain reaction, 16s rRNA sequencing, and whole genome metagenomic analyses. While the results were not consistent, they all pointed to a significant alteration in the composition of gut microbiota: NAFLD/NASH patients demonstrated a higher abundance of Gram-negative bacteria belonging to the phylum Proteobacteria and a reduction in bacteria from the Bacteroidetes and Firmicutes phyla.^{58,112,178–181} Studies comparing intestinal microbial composition in NAFLD patients without fibrosis to NAFLD patients with stage 2 fibrosis or higher revealed higher abundance of Bacteroides and Ruminococcus and lower abundance of Prevotella in patients with fibrosis.¹⁷⁹ Bastian et al looked at NAFLD patients with a fibrosis stage of 2 or higher and found that the abundance of Bacteroidetes was greater than in patients with stage 0 to 1.¹⁸² Demonstrating an association between gut dysbiosis and advanced stages of NAFLD, Loomba et al reported lower abundance of Firmicutes and higher abundance of Proteobacteria in the patients with advanced fibrosis. This study also observed a higher abundance of Escherichia coli and Bacteroides vulgatus in NAFLD patients with advanced fibrosis.⁵⁸ Apart from gut dysbiosis, small intestinal bacterial overgrowth is also frequently reported in NAFLD/NASH patients with an estimated prevalence of 50 to 78%.^55,183,184 These clinical studies along with preclinical studies in gnotobiotic animals as well as animals treated with antibiotics to deplete gut microbiota have corroborated the role of gut microbiota in the progression of NAFLD but the underlying mechanisms remain poorly defined (►Fig. 4).

Intestinal Barrier Function and Bile Acids

BAs are synthesized by hepatocytes from cholesterol, stored in the gallbladder, and released postprandially into the small intestine where they aid in the micellar solubilization and absorption of dietary lipids and fat soluble vitamins.^185,186 The majority of BAs secreted into the intestine are reabsorbed (~90–95%) and recirculated to the liver via enterohepatic circulation with most of the reabsorption occurring in the terminal ileum through the apical sodium-dependent BA transporter (ASBT) specific for the transport of conjugated BAs.¹⁸⁷ A small percentage of BAs are also reabsorbed in the colon facilitated by gut microbiota which converts the conjugated BAs into more hydrophobic molecules allowing passive diffusion.^188–190

The enterohepatic cycling of BAs is a highly efficient and regulated process that restricts BAs to the intestinal and hepatobiliary compartments.¹⁸⁷ Disruption of this process, primarily due to alterations in the expression of BA transporters and function or excessive biotransformation of BAs by gut microbiota which impedes transport via ASBT transporters, can affect enterohepatic cycling resulting in excess BAs in the colon.^{188,191–193} At high concentrations, most BAs are cytotoxic to IECs; however, the extent of BA-induced tissue injury depends on the duration of exposure and BA conjugation status. For instance, in vitro studies in human adenocarcinoma cells revealed that hydrophobic BAs chenodeoxycholic acid (CDCA) and deoxycholic acid (DCA) induce apoptosis,^194–196 whereas the hydrophilic BA ursodeoxycholic acid (UDCA) inhibits apoptosis.¹⁹⁷ Exposure of IECs to DCA and taurodeoxycholic acid induces NF-κB-dependent IL-8 secretion,^198,199 whereas UDCA and tauroursodeoxycholic acid inhibit TNFα and IL-1β secretion.^200,201 With respect to length of exposure, short-term exposure to high concentrations of BAs disrupts cell membranes causing oxidative/nitrosative stress and apoptosis,^195,202,203 whereas prolonged exposure results in genomic instability and resistance to apoptosis.^204,205 BA-induced tissue injury or protection is also mediated through their ability to induce or suppress proinflammatory cytokine production by IECs. Apart from their direct effect on IECs, recent studies suggest that the BA-receptor farnesoid X receptor (FXR) plays a role in modulating the effects of BAs in the intestine by demonstrating that activation of FXR induces apoptosis and inhibits proliferation of IECs.^203,206 However, the precise mechanisms underlying FXR-mediated regulation of intestinal barrier function is not clear as demonstrated by recent studies reporting higher intestinal permeability in whole body FXR KO mice but no changes in intestinal permeability in liver or intestine-specific FXR KO mice.²⁰⁷ Interestingly, ablation of FXR in the liver increased expression of FXR and FXR-targeted genes in the colon compared with intestinal specific FXR KO and WT mice.²⁰⁷ Collectively, these observations underscore the need for further studies to understand the underlying mechanisms of BA-mediated disruption of intestinal barrier in chronic GI and liver diseases.

Role of Bile Acids in Promoting Intestinal Barrier Dysfunction in NAFLD

Consistent with the reported increase in fecal BAs in NAFLD patients,²⁰⁸ our studies in a murine model of NAFLD demonstrate that a WD increases passage of primary unconjugated BAs including CDCA, cholic acid, lithocholic acid (LCA), α- and β-muricholic acid, and UDCA to the large intestine.²⁰⁹ Use of sevelamer hydrochloride to sequester intestinal BAs reduced mucosal inflammation and improved intestinal barrier function. Improvement in intestinal barrier was associated with reduced liver injury, reduced hepatosteatosis, and improved MetS indices. In vitro and in vivo permeability assays revealed that CDCA increases gut permeability suggesting that a diet-induced increase in the passage of CDCA to the large intestine may contribute to the loss of barrier function. Increased passage of primary unconjugated BAs to the large intestine correlated with an increase in Clostridium and Bacteroides, microbial taxa with bile salt hydrolase activity.²⁰⁹ Bacteria belonging to the taxa Clostridium are among a small number of intestinal anaerobes with 7-dehydratase activity, the enzyme required for production of DCA and LCA.^188,210,211 These findings along with previous reports that the BA pool in germ-free and antibiotic-treated mice consist of primary conjugated BAs, and fecal excretion of BAs is decreased in germ-free mice suggest the involvement of microbiota in increasing unconjugated BAs in NAFLD.^212,213 While further studies are needed to determine the molecular mechanisms involved, evidence to date suggests that dysbiotic microbiota contribute to the loss of barrier function by increasing the passage of unconjugated BAs to the large intestine which is not equipped to handle this influx (►Fig. 4).^193,203,214

Conclusion

The intestinal epithelial barrier, comprised of IECs, a mucosal layer, vascular endothelium, and the mucosal immune system, constitute a highly selective partition facilitating absorption of nutrients and essential molecules while limiting access to pathogens and harmful luminal contents to maintain homeostatic balance. Impairment in intestinal barrier function is implicated in the etiology and pathogenesis of chronic GI disorders, autoimmune diseases including type 1 diabetes, rheumatoid arthritis, multiple sclerosis, and metabolic diseases where barrier dysfunction contributes to disease pathology through systemic activation of immune cells. While previously underappreciated, loss of intestinal barrier function has emerged as a major factor profoundly impacting progression of NAFLD and may explain why fatty liver progresses to steatohepatitis and cirrhosis in a subset of individuals. This concept is further strengthened by preclinical studies in animal models of NAFLD and clinical studies in NAFLD patients where improvement in liver pathology is linked to improved intestinal permeability. A recent double-blinded, placebo-controlled, randomized control trial reported improvement in liver function in NAFLD patients treated with the laxative lubiprostone that improves intestinal barrier and reduces intestinal permeability.²¹⁵ While additional studies are needed to determine the efficacy of lubiprostone treatment in NAFLD patients with advanced disease, findings from the lubiprostone study demonstrate the therapeutic potential of directly targeting intestinal barrier in NAFLD. Therefore, efforts to comprehensively understand the complex interplay between microbiota, epithelial cells, and the mucosal immune system in modulating intestinal barrier function is urgently needed as this could aid in the genesis of new therapeutic strategies to prevent and treat not only NAFLD but multiple human infectious, inflammatory, and metabolic diseases.