Intestinal Microbiome and the Liver

Abbreviations

ALD

alcoholic liver disease

GALT

gut-associated lymphoid tissue

IBD

inflammatory bowel disease

LEFSE

linear discriminant analysis effect size

MHE

minimal hepatic encephalopathy

OTU

operational taxonomic unit

The liver is the largest and one of the most intricate organs in the body, and is integral in regulating normal physiologic and biochemical processes. Liver diseases are among the prominent causes of chronic disease states and result in a significant global health burden. Well-established causes of liver diseases include viral infection, alcohol, obesity, insulin resistance, and autoimmune, metabolic, and genetic disorders. Irrespective of the underlying cause the pathophysiologic principles of liver injury usually involve inflammation, necrosis, apoptosis, oxidative stress, cholestasis, and fibrosis, which together can impact the natural history and progression of liver disease. On the other hand, dynamic repair and regenerative mechanisms are also active. The balance between these injury and repair spectrums influences the pathway to either recovery or further progression to cirrhosis. Emerging evidence clearly implicates the intestinal microbiome as an important link that can influence the fate of liver disease. To better understand the role of the intestinal microbiome as an important factor in liver disease, it is prudent to comprehend some of the normal but complex host-microbiome interactions that maintain the delicate balance between health and disease.

Historical Perspective and Technology Development

All bilaterian animals have an alimentary canal that transfers food to a digestive tract and have evolved in a symbiotic relationship with the microbial flora. These animals include the protostomes (arthropods, mollusks, and annelids) and the deuterostomes (echinoderms and chordates). Thus not only are all animals linked evolutionarily, they are linked ecologically through a shared symbiotic relationship with the microbial world. Our knowledge of the microbial world has its roots in antiquity. For example, humankind has been actively manipulating the microbial flora for food production for millennia. The documented history of fermentation of beer, wine, cheese, and yogurt goes back at least 10,000 years. The field of microbiology developed throughout the 18th and 19th centuries with the advent of microscopy and microbial culturing techniques that facilitated the discovery of how microbes were involved in health and disease.

The origins of the Human Microbiome Project go back some 100 years ago. For example, Ilya Mechnikov and Paul Ehrlich were awarded the Nobel Prize in Physiology or Medicine in 1908 for their work on immunity, but they both also made significant contributions to the scientific understanding of the human microbiome. Specifically, Mechnikov hypothesized that dysbiosis of the gut microflora was involved in mental health, aging, and disease. He even contributed to the concept of the use of a probiotic for its health benefit as he promoted the use of kefir, a fermented milk product, to prevent aging. Ehrlich, on the other hand, isolated a facultative anaerobe from infant stool that bears his name, Escherichia coli . It was originally thought that E. coli was one of the most abundant species in the human gut, whereas it is one that grows on aerobic plates. It is now known that the most abundant gut species are anaerobes.

Our main tools for characterizing the gut flora throughout most of the 20th century were aerobic and anaerobic culture–based techniques first introduced by Nobel laureate Robert Koch. The main limitation to culture methods is that less than 1% of environmental bacterial species grow on artificial media and thus cannot be identified by this method. Basically, we do not know what their nutrient requirements are, and there are many obligate interactions between the species of an ecosystem that cannot be replicated in culture.

The development of DNA-based analytic methods in the early 1970s revolutionized microbial ecology and molecular systematics. Carl Woese was one of the prominent pioneers of molecular systematics, and his identification of the three domains on life was seminal to our understanding of the microbial world. These techniques either involved the cloning and sequencing of PCR amplicons of the 16S ribosomal RNA gene from a community to produce an abundance species profile or resorted to a community fingerprinting technique such as denaturing gradient gel electrophoresis or length heterogeneity PCR fingerprinting, where one obtains an indication of the operational taxonomic units (OTUs) in the sample. The advantage of the latter is that one can profile an entire community in one run on a fluorescent sequencer; however, one does not necessarily identify individual species in the OTUs.

It was not until the beginning of this century that classic molecular techniques were applied to investigate the human microbiome. For instance, molecular methods were first used to define the gut microbiome in inflammatory bowel disease (IBD) by Zoetendal et al. in 2002. They used denaturing gradient gel electrophoresis to distinguish the mucosal biofilm microbiome from the luminal microbiome. Swidsinski et al. used fluorescence in situ hybridization with 16S ribosomal RNA probes to demonstrate changes in the mucosal biofilm in IBD. Kleessen et al. used fluorescence in situ hybridization to demonstrate that the mucosal biofilm in IBD and identified Bacteroides and Firmicutes as the dominant taxa in the human gut. Seksik et al. used quantitative dot blot hybridization to identify enterobacteria in Crohn disease patients and noted a number of unknown species that were of unknown phylogenetic origin. Komanduri et al. used length heterogeneity PCR and cloning and sequencing to fully characterize the mucosal and luminal microbiome and compare the microbial species in ulcerative colitis, Crohn disease, and pouchitis. They demonstrated that there is an invasion of the mucosal biofilm by luminal species in the disease state. Their inference was that the mucosal biofilm develops in the neonate, is stable, and is recognized as self by the immune system. Thus the invasion of this protective biofilm by luminal species evokes an inflammatory response leading to the disease state. This early work led to the first microbiome patent for IBD. Ott et al. used single-stranded conformational polymorphism fingerprinting, a technique that is very similar to DDGE, to demonstrate a reduction in the abundance of normal anaerobic species such as Bacteroides and Lactobacillus in the colonic mucosal biofilm.

The advent of next-generation sequencing technology in the early years of this century has revolutionized microbial ecology and human microbiome analysis. For example, the Roche 454 technology essentially clones 16S ribosomal RNA PCR amplicons onto beads using emulsion PCR (PCR inside an oil droplet), producing 500,000 sequence reads in one run, dramatically increasing the throughput over classic cloning and sequencing technology by 5000-fold. One critical patented innovation was the development of the sample bar coding strategy that allows the mixing of multiple samples each with a unique barcode, running the mixture on the next-generation instrumentation, and then sorting the sequences into samples bins. Recent technologies such as Thermo Fisher Scientific's Ion Torrent technology and Illumina's polytomy technology have increased the sequencing throughput by several orders of magnitude. At this time, sequencing throughput is no longer a bottleneck in human microbiome research, and research in the field is impacted now by limitations in sample collection, database management, and bioinformatics analysis.

There are several distinct approaches to analyzing the human microbiome. The first is to identify the raw 16S ribosomal RNA gene sequences by comparing them against bacterial databases and then build relative abundance tables. A classic Linnaean system is used to identify bacterial species even though many would argue that this is somewhat artificial. For example, E. coli belongs to the kingdom Bacteria , phylum Proteobacteria , class Gammaproteobacteria , order Enterobacteriales , family Enterobacteriaceae , and genus Escherichia . Thus a phylum is a higher level in the taxonomic hierarchy with a large collection of related organisms grouped together. Thus, in healthy adults, the two most dominant phyla in the large intestine are Firmicutes (Gram-positive clostridia) and Bacteroidetes (mainly Gram-negative bacteria such as Bacteroides fragilis ). Computational limitations restrict this approach to algorithms that can compare millions of sequences against the bacterial databases. In this context we routinely use the Bayesian analysis tool provided by the Ribosomal Database Project to quickly annotate the raw reads and build relative abundance tables. The limitation of this tool is that it identifies taxa only down to the genus level, although this is usually sufficient taxonomic resolution to analyze most projects.

A second class of tools performs phylogenetic analysis that compares the taxa diversity between samples from various clinical classes (i.e., control and disease state). The two most popular tools are Quantitative Insights Into Microbial Ecology (QIIME) and mothur. Both tools cluster the raw reads, pick representative sequences for each cluster to define OTUs, construct relative abundance tables, construct a phylogenetic tree from the OTUs, and then perform various nonparametric and multivariate analyses of the data. One can make functional inferences from the taxa abundance tables by linking them to whole-genome annotations of taxa that have been completely sequenced. One such tool is Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt), which builds a relative abundance table of functional pathways on the basis of the Kyoto Encyclopedia of Genes and Genomes database. Another very useful tool that facilitates nonparametric analysis between sample classes of both phylogenetic and functional data is linear discriminant analysis effect size (LEFSE), which performs a Kruskal-Wallis nonparametric t test followed by linear discriminant analysis to identify the taxa that have significantly changed between classes. Another useful tool is Metasats, which performs Fisher's exact test to determine which differential taxa and is useful when there are fewer numbers of sample for each class.

The third class of tools take a metagenomics approach where shotgun sequencing of all of the DNA in a sample is performed, reads are identified from genome databases, and taxa abundance tables are reconstructed from the identified reads. However, questions have been raised about this method that suggest limitations in its ability to reconstruct accurate relative abundance tables because of the ligation bias inherent in the process. Thus the 16S ribosomal RNA amplification approach using fusion adapters is preferred as the barcodes and adaptors are introduced through the PCR process and not through biased ligation methods.

The fourth approach is the use of correlation network analysis and correlation difference network analysis. In the correlation network approach, data for the microbiome, immunome, and metabolome and quantitative clinical data are combined into feature tables for each clinical class (i.e., control or disease state) and Spearman correlations are calculated between all features. A network is plotted in Cytoscape to visualize the statistically significant correlations, and these are used to develop a hypothesis about the interactions between the features from the gut-brain-liver-immunome-microbiome axis that we call the metabiome . Subsequently, a correlation difference network can be created, where the Spearman correlations that are significantly different between the classes are plotted in Cytoscape. This allows the inference of interaction that changes between the clinical classes (i.e., what interactions changed between the control state and the disease state).

All biologic ecosystems are dynamic nonlinear systems, and the interaction and abundances of the components are constantly oscillating. This is especially true of the gut microbiome and explains the large variation seen in the relative abundances of taxa even in the normal population. To complicate matters further, different taxa in the microbiome may be functionally equivalent in different individuals, further confounding the analysis of the observed nonparametric sparse datasets that are based on simple phylogenetic taxa information. Thus the field needs to be moving to functional metrics of the system. Additionally, a cross-sectional clinical study of the microbiome ecosystem is subsampling time points from these innate ecosystem oscillations. However, we hypothesize that the correlation analysis approaches capture the signal in these ecologic oscillations, and these correlation methods have been used successfully to investigate the metabiome in the oral microbiome, the oral mycobiome, the vaginal microbiome, and the gut microbiome in various disease states such as HIV infection, bacterial vaginosis, cirrhosis, and hepatic encephalopathy.

The final approach that has become popular in the microbiome analysis field is the use of machine learning as a prediction tool to develop diagnostics. There are several open-source GUI-driven packages available, such as Orange and WEKA, to perform the modeling, and they have been used in a number of models of disease such as IBD and colorectal cancer.

In summary, the microbiome field has exploded in the last decade since the start of the Human Microbiome Project in 2005, with a logarithmic increase in the number of citations of microbiome each year ( Fig. 3-1 ). It has become apparent that the human microbiome is implicated in social behavior, reproduction, growth, cognition, metabolism, the immune system, and disease. Thus the human microbiome is an integral component of the human ecosystem and is a major driver of homeostasis. One could even say that the human host exists merely to propagate the “selfish microbiome.” In the following sections we will review the perturbations of the microbiome in liver disease.

Fig. 3-1, Citations per year in PubMed for microbiome .

General Principles and the Conceptual Construct: the Gut-Liver Axis in Homeostasis and Implications for Liver Diseases

The host-microbe interactions significantly contribute in shaping the landscape of normal health and disease. The liver and the intestine are integrally linked in regulating the host-microbe interactions. Together the liver and the intestine form a close nexus better known as the gut-liver axis . The human intestinal tract is a reservoir of trillions of diverse microorganisms together called the microbiota , and their collective genomes are referred to as the intestinal microbiome . The human intestinal microbiome has coevolved with the host immune system. The intestinal microbiome influences the development of intestinal immune responses, and these immune responses also reciprocally regulate the structure and composition of the intestinal microbiome, demonstrating a symbiotic relationship facilitated by quorum-sensing small molecules.

Both the liver and the intestine are constantly exposed to myriad products derived from food, medications, toxins, and resident microbes. All of these have dynamic variations and can cause an exogenously induced shift in the composition of the luminal microbiome and its associated functions. Such a change can trigger an imbalance promoting dysbiosis and can tip the balance from health to disease. Several unique characteristics of the liver and the intestine ensure that the diverse microbiota is anatomically contained and remains in an appropriate balance between tolerance and vigilance to guard against pathobionts and opportunistic pathogens. Although numerous factors participate in establishing and maintaining this highly regulated homeostasis, the structural and functional attributes of the liver and the intestine such as intestinal barrier and immune tolerance are crucial. The subsequent sections highlight the conceptual details about the elements orchestrating the homeostatic milieu.

Intestinal Barrier

The intestinal tract is the largest mucosal barrier that interfaces with the intestinal microbiome. On the surface it appears as a physical barrier only, however, it actively participates in key processes involved with host physiology. The intestinal barrier comprises (1) the mucus layer, (2) the single-layer epithelium with six cell types—interstitial epithelial cells (enterocytes or colonocytes), microfold cells (M cells), goblet cells, Paneth cells (present only in the small intestine), enteroendocrine cells, and stem cells—and (3) the lamina propria ( Fig. 3-2 ). Each of the cells is responsive to and conditioned by the intestinal microbiota. Also, the intestinal microbial colonization supports the initiation, development, and maturation of the intestinal mucosa and its gut-associated lymphoid tissue (GALT), including Peyer patches, isolated lymphoid follicles, and mesenteric lymph nodes, the elements central to the innate and adaptive immune responses. Together these multiple layers have evolved to orchestrate highly specialized barrier defenses incorporating antimicrobial, physiologic, and immune functions that maintain intestinal homeostasis and microbial ecology. Their roles are further highlighted below:

1.
The mucus layer: The mucus is formed of mucin glycoproteins secreted by the goblet cells. In the small intestine the mucus is discontinuous and is less well defined, presumably to support active absorptive requirements. In contrast, the colon has thick, double-layered mucus with a dense inner layer devoid of bacteria and a loose outer layer in direct contact with the microbiota. The mucus structure thus has a physiologic role in shaping the mucosa-associated microbiota by providing glycans as a nutrient source in the outer layer, whereas the dense inner layer limits the direct contact of these bacteria with epithelial cells.
2.
The epithelium and its specialized cells: The goblet cells , aside from their role in mucus production detailed above, produce trefoil factors and resistin-like molecule β, which aid in maintaining barrier integrity by stabilizing mucin polymers and reducing the susceptibility to inflammation. Paneth cells are exclusively found in the small intestine and are strategically located close to epithelial stem cells in the crypt. Paneth cells secrete antimicrobial peptides such as defensins and C-type lectins, which cross-link with the mucus layer and are critical in containing the microbiota. Whereas defensins target all members of the microbiota, the C-type lectin regenerating islet-derived IIIγ (RegIIIγ) is effective in preventing epithelial contact by Gram-positive bacteria. Importantly, the loss of Paneth cells results in increased invasion of the epithelial barrier by pathogenic and symbiotic microbes. The M cells are another type of specialized intestinal epithelial cells. They are found in the epithelium of Peyer patches, which transport organisms and particles from the intestinal lumen to intraepithelial immune cells. The dendritic cells (DCs) in the Peyer patches access microbes either through the M cells or directly from the lumen through transepithelial extension.
3.
The lamina propria: The lamina propria is just below the intestinal epithelia and harbors a rich repertoire of intestinal immune cells. The lamina propria also has Peyer patches that serve as lymphoid follicles, where naïve immune cells differentiate into a variety of mature immune cell subsets. The DCs in the lamina propria coordinate stimulation of B cells in Peyer patches to produce IgA, which acts as a link between the innate and adaptive arms of the immune system. Whereas the innate response involves nonspecific IgA that binds to microbial surface glycans and causes bacterial agglutination, the microbe-specific IgA drives the main adaptive immune response in controlling the microbiota. The DCs can also induce T-cell differentiation or T cell–dependent B-cell maturation into the germinal centers. Naive T cells (T _H 0 cells) can differentiate into effector T _H 1, T _H 2, or T _H 17 cells or into regulatory forkhead box protein 3–positive regulatory T cells or type 1 regulatory T cells.
4.
Pattern recognition receptors (PPRs): Toll-like receptors (TLRs) and nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs), collectively referred to as pattern recognition receptors (PRRs), recognize a variety of microbial components and, hence, are important in establishing an interface between the host and the microbiota. The intestinal epithelial cells express PRRs that sense the microbe-associated molecular patterns expressed by commensal and pathogenic microbes. The recognition of microbe-associated molecular patterns by neonatal intestinal epithelial cells is necessary for stimulation of the development of isolated lymphoid follicles and the formation of lymphoid structures that are capable of supporting maturation of B cells and secretion of IgA. In a pioneering study, mice deficient in TLR signaling or greatly depleted of commensal bacteria exhibited increased susceptibility to experimentally induced intestinal inflammation, implicating the role of commensal microorganism–dependent signals in the regulation of intestinal homeostasis and response to injury. This and other studies defined the beneficial roles of intestinal epithelial cell–intrinsic TLR signaling through cytoprotective heat-shock proteins, epidermal growth factor receptor ligands, and trefoil factor 3, as well as enhanced integrity of apical tight junction complexes. Moreover, inflammasomes formed by caspase 1 and NLRs in constitutive knockout mouse models showed a complex response with inflammation and epithelial repair, suggesting dual roles of inflammasomes. Together this evidence illustrates that microbial recognition promotes intestinal epithelial cell health and function in maintaining homeostasis.
5.
Tight junctions: Tight junctions are transmembrane protein complexes between the epithelial cells and are inherently involved in the maintenance of the barrier function. They create a selectively permeable barrier that allows paracellular nutrient transport and impedes passage of the microbiota and its products. The claudins, occludin, and junctional adhesion molecules are the three main transmembrane proteins, of which claudin is considered to be the main structural component of the tight junctions. In addition, tight junctions also contain a complex system of adaptor molecules and scaffold proteins that mediate cross-links between the transmembrane proteins and the actin cytoskeletons within epithelial cells. Together the tight junctions offer an indispensable second layer of barrier protection.

How Does the Host Maintain Homeostasis With Continued Exposure to an Enormous Microbial Load?

The intestinal barrier actively interacts with the resident intestinal microbiome and generates a tolerant immune response to maintain homeostasis through multiple firewalls. These firewalls are broadly constructed along three main domains of compartmentalization, microbial hyporesponsiveness of resident intestinal macrophages, and immune suppression, and are illustrated in Fig. 3-3 .

Fig. 3-3, Multiple firewalls limit immune responses against resident intestinal bacteria.

Interactions of the Liver With the Intestinal Microbiome

The liver is a unique organ that modulates host physiology across the portosystemic interface through complex metabolic and immune regulatory functions. The normal liver can be viewed as two closely integrated compartments: the vascular and hepatobiliary systems. The distinct features of the vascular compartment of the liver include predominant venous blood supply, receiving approximately 70% of it from the intestine via the portal vein. Moreover, the liver exhibits a high degree of vascularization, associated with blood flow through highly permeable fenestrated endothelia, allowing the hepatic tissue to be in direct contact with the bloodstream. Consequently, the liver, much like the intestine, is continually exposed to intestinal-derived products. This systemic load to the liver is received by the hepatobiliary compartment, which comprises parenchymal cells (~80%)—that is, the hepatocytes and cholangiocytes—and nonparenchymal cells (~20%) such as hepatic stellate cells (HSCs), Kupffer cells, and liver sinusoidal endothelial cells (LSECs). The PRRs are expressed on many of these cells, which sense the constant influx of these microbial-derived products from the intestine and trigger immune responses. Additionally, both vascular and biliary systems communicate with the extrahepatic cells and tissues. The gut-liver axis interactions are further modulated by well-regulated bile acid signaling and coordinate a highly complex interorgan cross-talk.

The following section provides a brief overview of the interactive role of PRRs (TLRs and NLRs) and intestinal microbial products, with different hepatic cells. The ensuing innate and adaptive immune responses can either lead to tolerance or initiate and propagate inflammation, immune dysregulation, and fibrosis. Together, these pathologic hallmarks cut across the spectrum of most liver diseases of various causes. The TLRs/NLRs are expressed in the intestine and the liver. The intestinal regulation of PRRs and its implications in host physiology are beyond the scope of this chapter, and are reviewed elsewhere. Further, as discussed below, increasing evidence indicates an active role of TLRs/NLRs in liver diseases.

How Is the Intestinal Microbiome Related to Liver Disease?

The liver and the intestine combine to form the largest immune system in the body. Both are constantly exposed to foreign antigens in a variety of forms, which in the context of the current chapter relate to the microbiome. It is now well recognized that the intestinal microbiome contains 10-fold more cells than the total number of human cells and 100-fold more genes than the human genome. The gastrointestinal system harbors one of the highest concentrations of a microbiome on the planet. The symbiotic relationship (mutualistic, commensal, or parasitic) between the human host and the intestinal microbiome has coevolved and actively contributes to the development and maintenance of homeostasis. For instance, the distribution and composition of the gastrointestinal microbiome is guided by regional differences affecting the microbial niche and reflects part of this symbiotic relationship ( Fig. 3-4 ). In addition to their host-mediated effect on microbial assembly, composition, and activities, microbial niches are also under intense pressure from other surrounding bacteria. Bacteria use sophisticated intercommunication systems to help maintain their niches; consequently, this microbial network is essential to host homeostasis. These microbial relationships can be antagonistic or mutualistic, depending on the nature of the species, and are summarized in Fig. 3-5 . Our understanding regarding the functional significance of the intestinal microbiome is still in its infancy, although the relationship between the intestinal microbiome and liver diseases has been well established.

Fig. 3-4, Regional differences in the gastrointestinal tract affect the microbial niche.

Fig. 3-5, Bacteria use a complex communication network to thrive in an environment.

Mucosal and Liver Tolerance

To better understand the role of the intestinal microbiome in relation to liver diseases, it is important to recognize how the normal intestine and liver, despite constant exposure to the microbiota and microbial derivatives, are able to maintain a steady-state homeostasis. This is due to the phenomena of mucosal and liver tolerance. The tolerance is mainly influenced by (1) continual low-dose microbial antigen exposure leading to induction of regulatory T cells and (2) high-load exposure resulting in tolerance induction preferentially through the mechanisms of anergy and deletion.

The key players involved in the mechanisms for mucosal tolerance relate to (1) GALT through M cells and (2) DCs. On exposure to the intestinal microbiota and microbial products, GALT through M cells senses the message, which is passed on to DCs in the lamina propria. DCs are critical for inducing tolerance to antigens in the small intestine. Experimental evidence also suggests that GALT-associated DCs have a unique capacity to direct differentiation of regulatory T cells from forkhead box protein 3–negative T cells. Similarly, recent work suggests that interleukin (IL)-22–producing innate lymphoid cells help regulate the anatomic microbial containment, evident by peripheral dissemination of commensal bacteria and systemic inflammation following depletion of innate lymphoid cells, which is prevented by the administration of IL-22. The disseminating Alcaligenes species originated from the host lymphoid tissues. Also, Alcaligenes is associated with systemic inflammation after depletion of innate lymphoid cells in mice. Further, Alcaligenes -specific systemic immune responses are observed in the serum of pediatric Crohn disease patients and plasma of chronic hepatitis C patients. Collectively these data indicate that innate lymphoid cells regulate selective containment of lymphoid-resident bacteria to prevent systemic inflammation associated with chronic diseases.

In the context of the intestinal microbiome and liver tolerance, it is worth revisiting the three specialized resident hepatic antigen-presenting cells: (1) Kupffer cells, (2) LSECs, and (3) DCs. The Kupffer cells are preferentially located within the sinusoidal vascular space, predominantly in the periportal area, well suited to clear endotoxins from the passing blood and to phagocytose debris and microorganisms. Their slow migration along the liver sinusoids causes frequent perturbations and even temporary stasis of the sinusoidal blood flow, thereby facilitating close contact to passing lymphocytes and microbial products. The LSECs are the predominant nonparenchymal liver cell population (~50%) and have a characteristic fenestrated endothelium. LSECs express molecules that promote antigen uptake, including the mannose receptor and the scavenger receptor, and molecules that promote antigen presentation, including MHC class I and class II and the costimulatory molecules CD40, CD80, and CD86. The resident hepatic DCs are typically located around the central veins and portal tracts. Resting DCs can inhibit proliferation and cytokine production of activated, tissue-infiltrating lymphocytes through cytotoxic T lymphocyte–associated protein 4 and programmed death ligand 1. Although hepatocytes may also present antigens to liver-infiltrating T cells, they are more commonly regarded as targets for cellular immune responses, and their role in the induction of primary immune responses, although suggested, is still less clear.

How do the specialized resident hepatic antigen-presenting cells (Kupffer cells, LSECs, and DCs) induce endotoxin tolerance? As the liver is continually exposed to low levels of gut-derived endotoxins, some proposed mechanisms for promoting tolerance include release of inhibitory factors such as programmed death ligand 1, cytotoxic T lymphocyte–associated protein 4, and prostaglandin E ₂ by Kupffer cells and DCs. Also, the uninflamed liver provides a tolerogenic microenvironment by secreting immunosuppressive factors such as IL-10, transforming growth factor β (TGF-β), retinoic acid, and prostaglandin E ₂ . Further, the natural killer cells and natural killer T cells are either suppressed or programmed to exert antiinflammatory properties by secreting IL-4. Moreover, the interaction with adaptive lymphocytes favors induction of tolerogenic T-cell responses or deletion of autoreactive T cells.

Liver and Pattern Recognition Receptors

As alluded to earlier and extensively discussed in previous sections, the liver continually experiences low-grade exposure to the intestinal microbial products. These microbial components are sensed by the resident hepatic antigen-presenting cells, which also express the PRRs, notably TLRs. Briefly, the TLRs recognize pathogen-associated molecular patterns and participate in the innate and adaptive immune responses of the liver generally favoring tolerance. In addition to resident hepatic antigen-presenting cells (Kupffer cells, DCs, and LSECs), other liver cells such as hepatocytes and HSCs also express TLRs. Of the 13 identified mammalian TLRs, humans express TLR1 to TLR10. A general overview of the TLRs, their ligands, and their downstream signaling pathways is provided in Fig. 3-6 , which clearly illustrates how different microbial components interact with the TLRs and NLRs. Most TLRs are membrane associated but some (TLR3, TLR7, TLR8, and TLR9) are intracellular. The classic danger sensing receptor TLR4 recognizes endotoxin lipopolysaccharide (LPS) and binds with coreceptor CD14 and myeloid differentiation protein 2. TLR2 heterodimerizes with TLR1 or TLR6 and is essential for the recognition of a variety of pathogen-associated molecular patterns from Gram-positive bacteria, including bacterial lipoproteins, lipomannans, and lipoteichoic acids. Bacterial flagellin binds to TLR5. TLR9 is required for response to unmethylated CpG DNA. Also, TLR7 and TLR8 recognize small synthetic antiviral molecules, and their natural ligand is single-stranded RNA. TLR3 acts via a retroviral double-stranded RNA ligand. TLR3 is the only TLR that activates the toll–IL-1 receptor domain–containing adapter inducing interferon-β (TRIF)-dependent interferon signaling pathway. In contrast, all other TLRs use the myeloid differentiation primary response gene 88 (MyD88)-dependent pathway that activates nuclear factor κB (NF-κB). The bacterial peptidoglycans can interact with either cell surface or intracellular PRRs. The cell surface recognition of peptidoglycan is facilitated by membrane-bound CD14 and TLR2, whereas the intracellular recognition of peptidoglycan is mediated by two members of the NLR family, NOD1 and NOD2.

Fig. 3-6, Toll-like receptors and downstream signaling pathways.

Together the intestine and liver through specialized cells and PRRs mediate a tolerant response to the microbiota and intestinal microbial products. The innate and adaptive immune mechanisms initially favor a more immunoregulatory response to achieve homeostasis but can also result in host tissue damage.

Microbiota, Hepatic Inflammation, and Fibrosis

Inflammation and fibrosis are the two hallmarks of chronic liver diseases of any underlying cause (viral hepatitis, alcohol abuse, nonalcoholic steatohepatitis [NASH], metabolic disorder or cholestasis, etc.). The liver is the first site to encounter delivery of the intestinal microbiome and microbial products, and as such is enriched with immune cells. The innate immune response through activation of PRRs initially promotes inflammation to limit injury and infection. The cross-talk between these innate immune cells and the resident hepatic parenchymal and nonparenchymal cells can influence recovery, cell injury, or cell death. Cell injury and cell death can further perpetuate an inflammatory cascade through damage-associated molecular patterns, often referred to as sterile inflammation . Accordingly, these interactions can affect the development of acute and chronic liver diseases. The subsequent sections focus on the role of the intestinal microbiota as a driver of hepatic inflammation and fibrosis.

The Intestinal Microbiota as a Mediator of Hepatic Inflammation

The intestinal microbiota is a generous source that can trigger and maintain hepatic inflammation. As detailed earlier, the interactions of the microbial components and products with parenchymal and nonparenchymal hepatic cells that express PRRs can result in activation and secretion of various inflammatory mediators:

1.
The intestinal microbiota induces hepatic inflammation via TLRs: Several hepatic cells execute TLR-mediated response. The best described and probably the most important in the context of hepatic inflammation is the TLR4 pathway. LPS is the major structural component of the outer wall of all Gram-negative bacteria. TLR4 requires association with LPS-binding protein, CD14, and myeloid differentiation protein 2 to recognize LPS. On TLR4 ligation, the intracellular domain of TLR4 recruits toll–IL-1 receptor domain–containing adapter protein and MyD88 for MyD88-dependent signaling, and TRIF-related adaptor molecule bridges TRIF for MyD88-independent signaling. The Kupffer cells are crucial in this cascade and are among the first to encounter gut-derived toxins, including LPS. The Kupffer cell TLR4 and LPS binding activates NF-κB, mitogen-activated protein kinase, extracellular signal–regulated kinase 1, p38, c-Jun N-terminal kinase, and interferon regulatory factor 3, triggering production of proinflammatory cytokines and type I interferon. These proinflammatory stimuli contribute to enhanced hepatocyte damage and increased leukocyte infiltration, thereby amplifying liver injury. At the same time, cell death is the most important trigger for leukocyte infiltration and inflammation. Together the result is a vicious cycle of perpetuating inflammation, cell injury, and cell death.

Other TLRs and nonparenchymal cells are less prominently involved in contributing to hepatic inflammation. In an experimental model, CpG-containing DNA through TLR9-mediated signaling induced IL-1β production by Kupffer cells. More recently, in a murine model, overactivation of TLR5 signaling by high-dose flagellin caused acute inflammatory responses, neutrophil accumulation, and oxidative stress in the liver, which contributed to the progression and severity of flagellin-induced liver injury. Although LSECs express TLR4 and its activation can result in the production of NF-κB-induced tumor necrosis factor (TNF-α) and reactive oxygen species, the LPS-mediated LSEC damage is reduced in Kupffer cell–inactivated animals, suggesting a limited direct role of TLR4 in LSECs in vivo. This evidence clearly supports the role of the intestinal microbiota in hepatic inflammation mediated through TLR signaling.
2.
The intestinal microbiota induces hepatic inflammation via NOD-like receptors: The intracellularly located NLRs can interact with pathogens through either pathogen-associated molecular patterns or endogenous danger signals through damage-associated molecular patterns, and can trigger hepatic inflammation through a cytosolic multiprotein complex called the inflammasome . The inflammasome consists of an NLR (absent in melanoma 2 in response to cytoplasmic double-stranded DNA), the adapter molecule apoptosis-associated specklike protein containing a caspase-associated recruitment domain (ASC), and the effector procaspase 1. The inflammasome serves as a platform for activation of caspase 1 which in turn results in the maturation of proinflammatory cytokines, predominantly IL-1β. Activation of the inflammasome is a two-step process in which the priming step (injury, infection, or sterile inflammation) results in up-regulation of inflammasome expression and the second step triggers functional inflammasome activation by an inflammasome activator. A mechanistic overview of NLR family pyrin domain–containing 3 (NLRP3) activation in response to microbial infection is illustrated in Fig. 3-7 . In endotoxin-induced liver injury, LPS is a potent inducer of messenger RNA expression of all inflammasome components (NLRP3, ASC, caspase 1, and pannexin 1), pro-IL-1β, and pro-IL-18 via NF-κB activation. Together these observations indicate the role of intestinal microbiota and microbial product–related hepatic inflammation mediated through inflammasome activation in liver diseases.

The Intestinal Microbiota as a Mediator of Hepatic Fibrosis

Hepatic fibrosis is the most important determinant of liver disease progression and is a highly conserved response to liver injury irrespective of the cause. The mechanisms of development and progression of fibrosis are covered in detail in Chapter 5 . The most important participants in fibrogenic response are the HSCs. The cellular cross-talk in the perturbed hepatic microenvironment, which is generally proinflammatory, influences their activation and survival of activated HSCs. Pertinent to the current chapter, the role of the intestinal microbiota and microbial products as an important factor in fibrosis will be covered. As discussed earlier, HSCs provide a link between the gut and the liver through their high expression of TLRs, which can participate in HSC activation and fibrosis.

How can HSCs and TLR signaling contribute to hepatic fibrosis? The two main mechanisms proposed are:

1.
TLR4-induced chemokines : Following interaction between HSCs and TLR4, there is increased expression of chemokines (monocyte chemotactic protein 1, macrophage inflammatory protein 1α, macrophage inflammatory protein 1β, and RANTES) and adhesion molecules (intercellular adhesion molecule 1, vascular cell adhesion molecule 1, and E-selectin). Moreover, HSC-derived monocyte chemotactic protein 1 and RANTES act in an autocrine manner to activate HSCs. Genetic or pharmacologic inhibition of chemokines (RANTES, monocyte chemotactic protein 1) or chemokine receptors (CCR1, CCR2, CCR5) reduces liver fibrosis.
2.
C ross-talk between TLR4 and TGF-β signaling : TGF-β is a potent fibrogenic cytokine that activates HSCs and induces liver fibrosis. In quiescent HSCs, bone morphogenetic protein and activin membrane-bound inhibitor (BAMBI), an endogenous decoy receptor of TGF-β receptor, is highly expressed, and directly interacts with Smad7 to interfere with the association between type I and type II TGF-β receptors and Smad3, inhibiting TGF-β signaling. In contrast, HSCs are activated following TLR4 stimulation because of down-regulation of BAMBI in an MyD88- and NF-κB-dependent manner.

Moreover, TLR4, by stimulating LSECs directly or indirectly, can promote portal hypertension and angiogenesis through fibronectin production from HSCs. Clinically, the presence of inhibitory TLR4 single-nucleotide polymorphism is associated with decreased fibrosis in hepatitis C patients. Together the available evidence illustrates a link between HSCs, the intestinal microbiota, and microbial products through TLR signaling.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here