The role of gut microbiota in intestinal and liver diseases

Abstract

The world-wide incidence of many immune-mediated and metabolic diseases, including those of the intestines and liver, is steadily increasing. Gut microbiota plays a central role in the pathogenesis of these diseases as it mediates environmental changes to the intestinal immune system. Various environmental factors including diet, food additives and medication also trigger the compositional and functional alterations of microbiota, that is, dysbiosis, and this dysbiosis is closely associated with many chronic inflammatory diseases. However, the causal relationship remains unclear for the majority of these diseases. In this review, we discuss essential epidemiological data, known pathogenetic factors including those of genetic and environmental nature, while mainly focusing on the role of gut microbiota in the development of selected intestinal and liver diseases. Using specific examples, we also briefly describe some of the most widely-used animal models including gnotobiotic models and their contribution to the research of pathogenetic mechanisms of the host–microbiota relationship.

Introduction

For many decades, the incidence of numerous chronic immune-mediated and metabolic diseases, including intestinal and liver diseases, has been increasing. These diseases, initially affecting mainly affluent Western countries, are rapidly spreading to developing countries with dietary and lifestyle changes likely being key risk factors promoting their incidence. Modern processed food and microbiome alterations are often incompatible with the age-old human genome which requires thousands of years at least to adjust to dietary changes. This fact has been well documented by ancient human genome-sequencing projects focusing on evolutionary adaptations to a high-starch diet and dairy products. ¹ Our microbiome has evolved along with us for millions of years, shaping our immune system and development, ² so it is not surprising that these changes may have far-reaching consequences to our health.

Indeed, compositional and functional changes of gut microbiota, known as dysbiosis, are associated with a whole range of civilisation diseases. These three major features typically characterise dysbiosis: (a) the decrease in microbiota diversity which is associated with many chronic inflammatory diseases; (b) the reduction or complete loss of beneficial microbes; and (c) the increased number of potentially pathogenic microbes (pathobionts). In a healthy gut ecosystem, pathobionts represent a relatively low percentage of gut microbiota. However, the outgrowth of pathobionts, such as Enterobacteriaceae, is frequently observed in many infectious and immune-mediated diseases.^3,4

Dysbiosis could be caused by host-derived factors such as genetic background, compromised general health (infections, inflammation) and lifestyle (lack of physical exercise, stress) or, even more importantly, by environmental factors such as diet (high in sugars, low in fibre), xenobiotics (antibiotics and other medication, food additives, chlorine in water), or hygienic surroundings. Interestingly, in addition to diet, caging factors including ventilation and bedding type are the most significant gut microbiota-altering environmental factors in laboratory mice. ⁵

Dysbiotic microbiota can influence the host’s immune system and mucosal integrity through a variety of mechanisms. These include the modulation of inflammasome signalling through microbial metabolites, the modulation of Toll-like receptor (TLR) and NOD-like receptor (NLR) signalling, the degradation of secretory IgA (sIgA), shifting in the balance between regulatory and pro-inflammatory T cell subsets, direct mucolytic activity and others. ⁶ Whether dysbiosis is the direct cause of a disease or a consequence of disease-related changes in the host’s physiology remains unclear. However, there are several examples, including the type 1 diabetes (T1D), coeliac disease and Parkinson’s disease, where alterations in gut microbiota precede the onset of illness.^7–9Also, microbiota transplantation from patients or diseased mice to germ-free or antibiotic-treated mice shows that it is possible to transfer a disease phenotype. This suggests a causative role of gut microbiota in disease pathogenesis.^10,11Significantly, a disease severity rate of more than 70% in relation to gut microbiota composition has been shown by correlation studies for some diseases, including the type 2 diabetes and atopic dermatitis. ¹²

In this review, we discuss essential epidemiological data, known pathogenetic factors including those of genetic and environmental nature; focusing mainly on the role of gut microbiota in the development of selected gastrointestinal and liver diseases. Using specific examples, we also briefly describe some of the most widely-used animal models including gnotobiotic mice and their contribution to the research of pathogenetic mechanisms of the host-microbiota relationship.

Microbiota and inflammatory bowel disease

Inflammatory bowel disease (IBD) is a group of multifactorial disorders characterised by chronic relapsing inflammation of the intestinal wall and extra-intestinal organs. ⁶ IBD consists of two major types, Crohn’s disease (CD) and ulcerative colitis (UC). CD can affect any part of the gastrointestinal tract from mouth to anus, yet most commonly occurs in the terminal ileum. UC usually affects only the rectum and colon causing continuous lesions of the mucosa and superficial submucosa, whereas CD is characterised by transmural ‘skip lesions’ of the intestinal wall. The most commonly affected extraintestinal sites are the joints, skin and eyes. ¹³ The typical symptoms of active IBD are diarrhoea, fever and fatigue, abdominal pain, blood in the stool, poor appetite and weight loss. The incidence of IBD is stabilising in North America and Europe, but its prevalence continues to rise and exceeds 0.3%. However, since 1990, the IBD incidence has been rising in newly industrialised countries with westernised societies in Africa, Asia and South America. ¹⁴

The pathogenesis of IBD is not yet fully known, but it is widely accepted that chronic inflammation in IBD is driven by dysregulated immune response to either gut microbiome components including bacteria, fungi, viruses, or protozoa or dietary antigens in a genetically susceptible host. Genome-wide association studies (GWAS) have revealed 240 single nucleotide polymorphisms (SNPs) that are associated with IBD. ¹⁵ Many of the genes, such as Nod2, TLR4, ATG16L1, IRGM, IL-10R, or IL-23Rplay an essential role in microbial sensing, autophagy, or immune response and may affect gut microbiota composition. ¹⁶ However, genetic factors, having remained relatively stable for hundreds of years, cannot explain the increase in incidence. So, changing lifestyle and environmental factors including a lack of physical activity, stress, dietary changes, chlorine in water, xenobiotics such as food additives and medication are most likely the main culprits.^6,17,18

Not surprisingly, IBD is associated with compositional and functional alterations of gut microbiota. The most consistent finding in patients with IBD is reduced microbiota diversity represented mainly by a decrease in the relative abundance of Firmicutesphylum and an increase in Proteobacteriaphylum.^4,19Changes in the abundance of Bacteroidetesphylum are less consistent with some studies reporting a decrease,^4,19but others reporting an increase in relative abundance.^20–22At the species level, there is a decrease in short-chain fatty acids (SCFA) producing bacteria, such as Faecalibacterium prausnitziiand the Clostridiumclusters IV, XIVa and XVIII,^23,24an increase in sulphate-reducing bacteria (SRB) such as Desulfovibrio,^25,26an increase in mucolytic bacteria for instance Ruminococcus gnavusand Ruminococcus torques, ²⁷ and in addition, a disbalance of inflammatory and anti-inflammatory species.^4,28Interestingly, some studies report the increased abundance of human enteropathogens such as pathogenic Escherichia coli, Clostridium difficile, Fusobacterium nucleatumand/or Campylobacterspecies.^20,29,30However, the causal relationship, for a single bacterium, has not yet been proved.

Due to high interindividual variations of faecal microbiome profiles of both CD and UC patients, it is difficult to discriminate between these diseases on the basis of microbiota composition. ⁴ However, a recent study by Pascal et al. ³¹ shows that the microbiome of CD patients is less diverse and more unstable than the microbiome of UC patients and that CD patients have a higher relative abundance of Fusobacteriumand Escherichiaand a lower abundance of Faecalibacterium, Anaerostipes, or Colinsellacompared to UC patients.

On a functional level, these compositional alterations have various effects. SCFA including butyrate and propionate are necessary for the differentiation and proliferation of regulatory T cells (Tregs), the production of mucin and antimicrobial peptides, and are a major source of nutrition for intestinal epithelial cells.^32,33The hydrogen sulphide produced by SRB is toxic to intestinal epithelial cells and triggers inflammation.^25,34Mucolytic bacteria degrade mucins in the mucus layer and utilise it as a source of energy. ²⁷ The increase in pro-inflammatory bacteria such as segmented filamentous bacteria (SFB) stimulates the differentiation of T helper 17 (Th17) cells. ³⁵ Conversely, Tregs are induced by selected strains of Clostridia/Clostridiumspp. ³⁶ or by the polysaccharide A of Bacteroides fragilis. ³⁷

Mouse models of IBD are often associated with a typical microbiotic signature. For example, the increased numbers of Bacteroides distasonis, Clostridium ramosum, Akkermansia muciniphila, or Enterobacteriaceaecorrelate to disease activity in the dextran sulphate sodium (DSS) model of colitis. ³⁸ The trinitrobenzene sulfonic acid (TNBS) colitis, which is a T cell-mediated, IL-12 driven intestinal inflammation, is characterised by increased numbers of Enterobacteriaceaeand Bacteroides. ³⁹ The colitis in IL-10^−/−mice is associated with increased numbers of Enterobacteriaceaeand adherent-invasive E. coli(AIEC). ⁴⁰ Nod2-deficient mice, a model for Crohn’s disease, have a diminished capacity to produce antimicrobial peptides by their Paneth cells, resulting in altered gut microbiota. ⁴¹ Paneth cell dysfunction has been associated with Crohn’s disease. ⁴² Nod2^-/-mice are unable to regulate the H. hepaticusload in the terminal ileum and develop a granulomatous ileitis, enlarged Peyer’s patches and mesenteric lymph nodes. ⁴³ The TRUC (T-bet^-/-Rag^-/-) mice, which lack both adaptive immunity and T-bet-dependent components of innate immunity (ILC1 and NKp46⁺ILC3), develop colitis when colonised by Klebsiella pneumoniaeand Proteus mirabilis, but only in the presence of commensal microbiota. ⁴⁴ It is notable that colitogenic microbiota can be transplanted to wild-type mice and will induce colitis. ⁴⁵ To induce colitis in the adoptive T-cell transfer model, specific bacteria including Helicobacter muridarum, Helicobacter hepaticusor SFB and the presence of specific pathogen-free (SPF) microbiota are required.^46–48Interestingly, some commensal strains or their components, such as Parabacteroides distasonisor Lactobacillus casei, were shown to reduce intestinal inflammation in the DSS-induced colitis model.^49,50

To sum up, experimental animal models of IBD, both chemically induced and genetic, provide strong evidence for the role of gut microbiota in the pathophysiology of colitis. Since, in most models, germ-free animals do not develop the disease and the antibiotic treatment of colitic mice typically reduces intestinal inflammation.^51–54Frequently, the disease phenotype could be transplanted to healthy mice along with the microbiota.

Microbiota and colorectal carcinoma

Colorectal carcinoma (CRC) is the third most commonly diagnosed cancer and the fourth leading cause of cancer death in the world. In some developed countries, incidence and mortality rates have been gradually decreasing, possibly due to preventative screening followed by the removal of colonic polyps. However, in many low- and middle-income countries, CRC incidence and mortality rates are increasing rapidly and are expected to rise globally to as many as 2.2 million new cases and 1.1 million deaths annually by 2030. ⁵⁵

The etiological factors contributing to CRC initiation, progression and dissemination include genetic and environmental factors. But it is increasingly evident that gut microbiota and its interactions with the host’s immune system play a pivotal role in CRC pathogenesis.^56,57Interestingly, gut microbiota is also a key factor influencing the efficacy and toxicity of anticancer therapy. ⁵⁸

CRC patients have, as do IBD patients, compositional alterations of gut microbiota characterised by a reduced diversity of both mucosa-associated and faecal microbiota,^59,60a decreased relative abundance of Firmicutesand an increased relative abundance of Proteobacteria.^61–63Interestingly, CRC and CD patients share similar microbiota alterations characterised by increased abundance of F. nucleatumand E. coli. ³¹

Moreover, IBD patients are more susceptible to later CRC development which suggests a causal relationship between microbiota, inflammation and CRC.

Several bacteria including F. nucleatum, genotoxic E. coli, enterotoxigenic B. fragilis, or Helicobacterspp. are associated with CRC development, yet it remains unclear whether these bacteria are CRC drivers or passengers. ⁶⁴ Driver bacteria induce DNA damage via genotoxins produced either directly by bacteria or indirectly by inflamed tissue. Conversely, ‘passenger’ bacteria take advantage of the unique tumour microenvironment and outcompete other microbes. For example, possible CRC ‘drivers’ are specific strains of E. coliproducing genotoxins such as colibactin and cytolethal distending toxin (CDT). Recent study shows that CRC patients with micro-satellite instability have markedly increased colibactin-producing E. coli.^65,66Colibactin triggers CRC development through the induction of DNA damage and cellular senescence.^40,67Also, mono-colonisation with E. coliNC101, a mouse commensal with genotoxic capabilities, promotes invasive carcinoma in azoxymethane (AOM)-treated IL10^-/-mice. ⁴⁰ Significantly, some probiotic strains such as E. coliNissle 1917 harbour genotoxicity islands which are also necessary for their probiotic functions. ⁶⁸ A possible ‘passenger’ strain, frequently isolated from colorectal adenoma or carcinoma patients, is F. nucleatum: ⁶⁹ an invasive anaerobe that has previously been linked to periodontitis and appendicitis and, in CRC patients, is associated with shorter survival times. ⁷⁰ Kostic et al. ⁶⁹ showed that F. nucleatumpromotes intestinal tumorigenesis in the ApcMin/+ mouse model.

The role of depleted bacterial strains, such as Lactobacillior Bifidobacteria, in CRC pathogenesis is less well studied. However, it is possible that the depletion of commensals reduces colonisation resistance and enables the outgrowth of genotoxic strains. Also, the decreased production of SCFA by commensals may play a key role in tumorigenesis as SCFA have been shown to have anti-inflammatory, antitumorigenic and antimicrobial effects. ⁷¹

To conclude, gut microbiota and their metabolites are key players in CRC pathogenesis. Future preventative, diagnostic and therapeutic strategies must be based on knowledge of the microbiome and its interactions with the tumour’s environment. The manipulation of the gut microbiome by dietary changes, prebiotics, probiotics, specific antibiotics, or faecal microbiota transplant (FMT) will become an indispensable tool in CRC management.

Microbiota and coeliac disease

Coeliac disease is a lifelong, immunologically mediated disease induced in genetically predisposed individuals by gluten in food. This disease is the most common form of immune-mediated food intolerance, affecting about 1% of the European population. It is characterised by mucosal atrophy and increased cellularity of the small intestine which can lead to malabsorption. The autoimmune nature of this disease is documented by the presence of autoimmune mechanisms directed against several autoantigens, including the most important diagnostic autoantigen, that is, tissue transglutaminase.

Compared to healthy subjects, the presence of changes in intestinal microbiota composition (intestinal dysbiosis) was described both in patients with active coeliac disease and patients treated by following a gluten-free diet.^72–74Most studies have consistently reported low levels of Lactobacilliand Bifidobacteria, and increased levels of Proteobacteriain both children and adults with active coeliac disease.^75,76Interestingly, the increased abundance of Proteobacteriacorrelated with disease activity. Other studies have reported also increases in the abundance of Bacteroidesand E. coli. ⁷² A recent study in infants with familial risk of coeliac disease demonstrates that gut microbiota changes, such as loss of diversity, a decreased abundance of Bifidobacterium longum, or increase in Enterococcusspp., precede the onset of disease. ⁹

In contrast to IBD where numerous mouse and rat models have been described and often used, good experimental models of coeliac disease had been lacking for a long time. ⁷⁷ To analyse the complex mechanisms involved in the pathogenesis of coeliac disease, animal models have recently been developed, with some of them being used to determine the role played by gut microbiota in disease development.^78,79Gut microbiota in rat and mouse models of coeliac disease can be manipulated by gnotobiotic technology, changing the diet, antibiotic treatment and by the administration of pathogenic or beneficial (probiotic) microorganisms. The gnotobiotic model of coeliac disease in germ-free rats in which gluten enteropathy (flattening and increased cellularity of the mucosa) was induced by the administration of gliadin soon after birth was developed in our lab. The role of intraepithelial lymphocytes in this model was analysed and has been established. ⁸⁰ How the composition of gut microbiota affects intestinal barrier function was tested using intestinal loops of germ-free rats in the presence of various intestinal bacteria, gliadin and interferon gamma. Translocation through the gut mucosa of only small amounts of gliadin was found in the presence of Bifidobacterium bifidum. In contrast, in the presence of Shigella,increased amounts of translocating gliadin and the impairment of mucosal tight junctions were detected. ⁸¹

HLA transgenic mice, that is, mice expressing human DQ8, one of two DQ alleles associated with coeliac disease were used to study the effect of microbiota. ⁸² Germ-free non-obese diabetic DQ8 mice developed severe gluten-induced enteropathy compared to colonised mice. These results suggest that intestinal microbiota could reduce the pro-inflammatory effects of gluten after ingestion.^83,84The decisive role of gut microbiota composition in gluten-degradation patterns was demonstrated by colonisation of germ-free C57BL/6 mice with bacteria isolated from the small intestines of coeliac disease patients or healthy controls. This significant finding suggests that the presence of dysbiosis in coeliac patients could modulate autoimmune risks in genetically susceptible individuals. ⁸⁵ An interesting and important study performed on mice was published very recently demonstrating that reovirus, a common and otherwise harmless virus present in the gut during the postnatal period, can trigger a damaging immune response to gluten and contribute to the development of coeliac disease. ⁸⁶

Despite the fact that animal models of human diseases have their limitations, they help us to understand the complex mechanisms occurring during host–microbe interactions and enable us to analyse the effects of microbiota on immune reactivity to food components (especially gluten). ⁸⁴

Microbiota and non-alcoholic fatty liver disease

Compositional and functional alterations of gut microbiota are also associated with the development and progression of a number of liver diseases including non-alcoholic fatty liver disease (NAFLD), alcoholic liver disease (ALD), cirrhosis, or hepatocellular carcinoma. ⁸⁷ NAFLD, the hepatic manifestation of insulin resistance and metabolic syndrome, is characterised by hepatic triglyceride accumulation without significant alcohol consumption. ⁸⁸ The global prevalence of NAFLD has been rapidly increasing and is currently estimated at 24%. ⁸⁹ An inflammatory form of NAFLD, non-alcoholic steatohepatitis (NASH), can potentially progress to advanced liver disease, cirrhosis and hepatocellular carcinoma. ⁹⁰ The factors contributing to the progression of NAFLD towards NASH remain incompletely understood.

Gut microbiota is an important factor involved in NAFLD pathogenesis. There is a direct anatomical link, the portal vein, between the gut and liver. Thus, gut microbiota and its metabolites could easily influence liver physiology. When the gut barrier, made of a single layer of epithelial cells sealed with tight-junction proteins, is compromised then the impact of gut microbiota and food-derived molecules on the liver will be even more pronounced. Cani et al. showed that dietary fat and glucose could lead to gut barrier injury and increased gut permeability to bacteria, its components and metabolites. ⁹¹ Therefore, pro-inflammatory bacteria-derived molecules, such as lipopolysaccharide (LPS), can induce liver injury by the activation of Kupffer cells via TLR4 signalling. ⁹² Several human ⁹³ and animal^94–96studies suggest that a high intake of carbohydrates, particularly fructose, is closely linked to the development of NAFLD. NASH patients, as well as obese patients, have dysbiosis characterised, at the phylum level, by an increase in Bacteroidetesand Proteobacteriaand a decrease in Firmicutes. ⁹⁷ Also, fructose-induced NAFLD is associated with alterations in gut microbiota composition as well as increased gut permeability resulting in endotoxin translocation, activation of Kupffer cells via TLR4 and M1 macrophages which leads to chronic hepatic inflammation and injury. ⁹⁶ Interestingly, probiotics in animal studies were shown to correct dysbiosis by inhibiting proliferation of harmful bacteria and to improve gut barrier function by upregulating the expression of tight junction proteins. ⁹⁸

Microbiota and ALD

ALD, which is caused by chronic excess alcohol consumption, is a major cause of liver-related morbidity and mortality. Like NAFLD, early stage ALD, represented by asymptomatic and reversible steatosis, may progress to steatohepatitis, fibrosis, cirrhosis and its complications (such as portal hypertension, ascites, variceal bleeding or hepatic encephalopathy), or even hepatocellular carcinoma. The ALD pathogenetic mechanisms are not entirely understood, but recent research indicates that gut microbiota including its fungal component may play an essential role. ALD is associated with gut dysbiosis and increased permeability resulting in increased translocation of microbial and fungal products from the gut, via the portal vein, into the liver. ALD patients, as well as NAFLD patients, often demonstrate small intestinal bacterial overgrowth (SIBO).^99,100ALD-associated dysbiosis is characterised by an increase in Enterobacteriaceaeand a decrease in Bacteroidetes, Lactobacillusand A. muciniphila.^101–105Interestingly, the susceptibility to alcohol-induced liver injury is transmissible from alcoholic hepatitis (AH) patients to both germ-free and conventional mice by FMT and, conversely, the gut microbiota transferred from alcoholic patients without AH shows a protective effect. ¹⁰⁶ Also, in a different study, alcohol sensitivity was shown to be dependent on gut microbiota, because FMT from alcohol-resistant to alcohol-sensitive mice prevented steatosis, liver inflammation and restored gut homeostasis. ¹⁰⁷ Recent research shows that supplementation of depleted commensals might be beneficial in ALD treatment. ¹⁰⁵ For example, A. muciniphila,a Gram-negative intestinal commensal, is heavily depleted in both ALD patients and ethanol-fed mice and its abundance indirectly correlates with disease severity. The A. muciniphilasupplementation promotes intestinal barrier integrity and ameliorates liver injury in ethanol-fed mice. ¹⁰⁵

Interestingly, recent work by Yang et al. ¹⁰⁸ shows that intestinal fungi also contribute to ALD development. In this study, ALD patients demonstrated decreased fungal diversity and richness and significant Candidaovergrowth accompanied by an increased systemic exposure and immune response to fungi which correlated to mortality. Also, findings from the ethanol-treated mouse model supported the role of fungi in ALD as the treatment with an antifungal drug, amphotericin B, protected mice from intestinal fungal overgrowth, reduced beta-glucan translocation and ameliorated ethanol-induced liver disease. ¹⁰⁸ So, it seems that chronic exposure to ethanol resulting in fungal overgrowth and translocation of fungal products into the liver may lead to liver inflammation and cirrhosis. To sum up, the manipulation of both microbial and fungal components of the gut microbiome might be a promising strategy in the treatment of ALD.

Microbiota and cirrhosis

Cirrhosis, a final stage of many chronic liver diseases including NAFLD, ALD, cholestatic liver diseases and viral hepatitis, is characterised by replacement of healthy liver tissue with non-functional fibrous tissue. Cirrhosis, as well as NAFLD and ALD, is associated with gut microbiota alterations, including dysbiosis and SIBO. Qin et al. showed that the intestinal tract of cirrhosis patients is massively colonised by microorganisms usually associated with the oral cavity, such as Veillonelaand Streptococcus. ¹⁰⁹ The overrepresentation of oral microbiota in the small intestines of cirrhosis patients was also confirmed by Chen et al. ¹¹⁰ Like ALD patients, the cirrhosis patients demonstrate a significant fungal dysbiosis. Also, a bacterial–fungal ratio, specifically the Bacteroidetes/Ascomycotaratio, was shown to reliably predict the length of hospitalisation of these patients. ¹¹¹

The treatment of cirrhosis and its complications include reduction of the gut microbiota load with non-absorbable antibiotics. ¹¹² In animal models, therapeutic strategies based on the blocking of microbiota sensing receptors, such as TLRs or NLRP3 proved also to be effective in decreasing liver inflammation and fibrogenesis.^113,114

To conclude, gut microbiota is a critical player in liver disease pathogenesis and it should be given the highest attention. However, further research is needed to fully expand our knowledge of the interactions among diet, microbiota and host in the prevention and treatment of liver diseases.

Conclusions

In this review, we have summarised available data from both human and animal studies on the role of gut microbiota in the pathogenesis of selected intestinal and hepatic diseases. Human gut microbiota is being altered primarily by lifestyle and environmental factors including, among others, lack of physical activity, stress, dietary changes, xenobiotics and medication. The altered, that is, dysbiotic, microbiota is associated with multiple chronic inflammatory diseases. Whether dysbiosis is a direct cause or a consequence of the disease remains unclear, but animal models, especially gnotobiotic models, prove to be a tremendously useful tool when addressing these questions. Undoubtably, gut microbiota is a central player in the pathogenesis of chronic inflammatory diseases. However, further well-designed studies are needed to increase our knowledge of host-microbiome physiology to enable us to create new diagnostic, preventative and therapeutic strategies.