Curr Gastroenterol Rep (2015) 17: 15 DOI 10.1007/s11894-015-0439-z
INFLAMMATORY BOWEL DISEASE (S HANAUER, SECTION EDITOR)
IBD and the Gut Microbiota—from Bench to Personalized Medicine Emanuelle Bellaguarda & Eugene B. Chang
Published online: 12 March 2015 # Springer Science+Business Media New York 2015
Abstract Inflammatory bowel diseases (IBD) are chronic relapsing inflammatory disorders involving the gastrointestinal (GI) tract, which arise from the confluence of genetic, immunological, microbial, and environmental factors. Clinical, genetic, and experimental data support the role of gut microbiota in contributing to the etiopathogenesis of these diseases. In IBD, the development of gut dysbiosis and imbalances in host–microbe relationships contribute to the extent, severity, and chronicity of intestinal inflammation. With continued advances in knowledge, technology, bioinformatics tools, and capabilities to define disease subsets, we will be able to lower risk and improve clinical outcomes in IBD through individualized interventions that restore host–microbial balance. This article provides a critical review of the field, based on the latest clinical and experimental information.
Keywords IBD . Gut microbiome . Dysbiosis . Probiotics . Prebiotics . Antibiotics . Host–microbe interactions . Mucosal immunology . Intestinal inflammation . Crohn’s disease . Ulcerative colitis . Precision medicine . Western diet . Fecal microbiota transplantation
This article is part of the Topical Collection on Inflammatory Bowel Disease E. B. Chang (*) Section of Gastroenterology, Hepatology, and Nutrition, The University of Chicago, 5841 S. Maryland Avenue, MC 6084, Chicago, IL 60637, USA e-mail: [email protected] E. Bellaguarda Division of Gastroenterology and Hepatology, Northwestern University, 676 North Saint Clair suite 1400, Chicago 60611, IL, USA
Introduction Inflammatory bowel diseases (IBD) comprise a group of chronic relapsing inflammatory disorders involving the gastrointestinal (GI) tract, which arise from the confluence of genetic, immunological, microbial, and environmental factors. IBD have traditionally been categorized into two main clinical phenotypes: ulcerative colitis (UC) and Crohn’s disease (CD). UC is characterized by continuous, superficial inflammation involving the rectum and extending up to the cecum, while CD is characterized by deep ulcerations distributed in a skipped pattern anywhere from the mouth to the anus. CD is a complex entity, and, because of its penetrating nature, these patients can present with fistulas, bowel strictures, and abscesses. Moreover, this phenotype is more associated with extra-intestinal manifestations such as arthritis, uveitis, and perianal disease, whereas UC can be associated with primary sclerosing cholangitis and ankylosing spondylitis. The incidence and prevalence of IBD have been rising significantly in America, Europe, and Asia in the past few decades, and the burden to health care costs has significantly increased [1–4]. New therapies have come out in the recent years offering a better quality of life to many patients. However, the causes and more importantly ways to prevent these debilitating diseases have not yet been found. There are several theories to the pathogenesis of IBD. It is well established that there is genetic predisposition to both CD and UC, although, in 77 % of CD patients and in more than 80 % of UC patients, no genetic associations can be found [5]. This fact should not be viewed as an argument against a genetic basis for IBD but merely underscores the limitations of genome-wide association studies (GWAS) in being able to identify less common genetic variants. On the other hand, it is clear that genetic risk is not sufficient by itself in causing IBD in most cases, and there is very likely a role for other factors in the development of these diseases [6]. In fact, many
15 Page 2 of 13
Bwestern^ disorders such as IBD are likely related to dramatic changes observed in the collective human microbiota attributed to increased rates of C-section, sanitation of living spaces, dietary changes, and early and indiscriminant usage of antibiotics [7]. Therefore, the study of the gut microbiome is key in understanding worldwide trends in IBD and also other emerging metabolic and complex immune disorders of the modern age [8–10]. The association between intestinal microbiota and IBD dates back to the 1920s when Rettger et al. studied the effect of Bacillus acidophilus in the intestinal microbiota [11, 12]. In 1940, Kirsner evaluated the possible correlation between streptococci and UC and the effect of oral sulfonamide in fecal bacteria [13]. In the late 1990s, the association between fecal microbiota and CD was evident when recurrent inflammation was observed after the fecal stream was reestablished in postoperative CD patients [14, 15]. These and many other studies have provided a strong descriptive basis for IBD, but clearly a better understanding of the genetic, molecular, cellular, and microbial basis for IBD was needed. In recent years, advances in bioinformatics and in the development of culture-independent methods to define the composition and function of the intestinal microbiota have led to tremendous insights into the structure (membership) and function of complex microbial communities. As a consequence, a great deal of information about host–microbial interactions and their role in IBD pathogenesis and treatment has been gained. However, many questions still remain. Is the dysbiosis associated with active IBD a cause or consequence, and does it contribute to the natural history and clinical outcomes? Are the changes in gut microbiota stable or continuously changing during the course of this disease? How do standard medical therapy, diet, and other environmental factors affect the microbial function and contribute to mucosal healing? How can we leverage this knowledge to lower risk and improve clinical outcomes in IBD? Answers to these questions will significantly advance the field and improve our understanding on how to effectively treat IBD. In this article, we aim to review recent advances and developments in the field.
Factors Determining the Assemblage of Gut Microbiota and IBD Risk The human microbiome project, funded by the National Institutes of Health in 2007, has catalogued, to date, 60 millions predicted genes, 600 microbial reference genomes, and 700 metagenomes [16]. An individual’s collective microbiome is believed to have as many as 10 to 100 trillions of cells, 10 times the number of eukaryotic cells in the human body. Most of our microbiota reside in the GI tract, which contains 10 out of 50 known bacterial phyla. Firmicutes and Bacteroidetes together account for more than 90 % of the intestinal
Curr Gastroenterol Rep (2015) 17: 15
microbiota. The remaining 10 % includes phyla such as Actinobacteria, Proteobacteria, Fusobacteria, Verrucomicrobia, and Cyanobacteria [17]. The biodiversity of gut microbes is believed to be in the range of 500 to 2, 000 different species [18]. In addition to the commensal gut bacteria, Archaea, fungi, and viruses are present, the latter primarily comprised of bacteriophages that are as numerous and far more diverse than their bacterial counterparts [19–21]. GWAS have identified over 160 single nucleotide polymorphisms (SNPs) associated with IBD. Many of these genes are involved in pathways that modulate how the host responds to microbial stimuli. Nucleotide-binding oligomerization domain-containing protein 2 (NOD2), also known as caspase recruitment domain-containing protein 15 (CARD15), was the first gene identified, supporting the link between gut microbiome and IBD etiopathogenesis. This gene is located on chromosome 16 and is expressed in monocytes/Paneth cells in the terminal ileum, encoding an intracellular receptor for the bacterial peptidoglycan muramyl dipeptide (MDP) [22, 23]. Thus, NOD2 recognizes components in the bacterial cell membrane and, together with other genes associated with IBD susceptibility (e.g., ATG16L1, CARD 9, IL23R), determines how patients will respond to microbial stimuli. The responses can then lead to changes in epithelial, mesenchymal, and immune functions that are involved in the maintenance of intestinal homeostasis. For example, NOD2-deficient mice exhibit increased colonization in the terminal ileum by Bacteroides, Firmicutes, and Bacillus and decreased ability to prevent proliferation of pathogenic bacteria. Moreover, germ-free mice have a decreased expression of NOD2 in terminal ileum enterocytes that can be reversed upon conventionalization [24]. In CD patients carrying NOD2 mutations, the emergence of certain pathobionts can potentially trigger a decrease in the immunomodulatory cytokine, IL-10 [22]. CD patients carrying NOD2 risk alleles (Leu1007fs, R702W, or G908R) can develop a dysbiosis characterized by a decrease in the Clostridium groups XIVa and IV and an increase of Actinobacteria and Proteobacteria [25]. NOD2 is also associated with more severe disease outcomes including higher rates of stricturing, fistulizing disease, and recurrent surgeries [26]. Furthermore, mutations of NOD2 and ATG16L1 are associated with increased bacterial translocation to the bloodstream in patients with active CD. NOD2-variant genotypes are associated with decreased phagocytic and bactericidal activity in the peripheral neutrophils as well as decreased anti-TNF trough levels [27•]. ATG16L1 is a gene that has an important role in autophagy and a variant of it, T300A, is associated with increased endoplasmic reticulum stress in Paneth cells as well as increased IL1-beta production in response to MDP in the terminal ileum of CD patients [28]. IBD patients with the T300A variant appear to have greater representation in the ileum of adherent-invasive Escherichia coli, which may be associated
Curr Gastroenterol Rep (2015) 17: 15
with increased risk for severe disease, development of fistulas, and surgery. However, it is also clear that the ATG16L1 risk variant is quite common in the general Caucasian population, and most of these individuals never develop IBD. Thus, other factors such as those encountered in the environment, from alterations in gut microbiota, epigenetic modifications, and other disease modifiers may play a role in the disease pathogenesis [29]. For example, mice hypomorphic for the ATG16L1 mutation and infected with murine norovirus develop mucosal inflammation resembling CD but exhibit far less pathology when infected with non-replicative murine norovirus [30]. These observations underscore the multifactorial nature of IBD and the complex interplay between host genetic, environmental, and gut microbial factors that can contribute to the pathogenesis of these diseases (Table 1).
Western Diet-Induced Changes in the Gut Microbiome and Triggering of Mucosal Inflammation in Genetically Susceptible Hosts The rising incidence and prevalence of IBD in Asian countries are believed to be related to rapid changes in our environment, societal norms, and increased industrialization. Among the many factors that may underlie this trend is the consumption of western type diets high in fat, processed sugars, protein, and overall calories [31]. Diet can significantly affect the gut microbial composition of individuals, which, in turn, can modulate both intestinal and immune homeostasis. The gut microbiota among primates are significantly different, which may be related to differences in diet (herbivore, carnivore, and omnivore) [32]. In humans, a high-fat, low-fiber diet (western diet) is associated with a predominance of Bacteroidetes and Actinobacteria, while Firmicutes and Proteobacteria are associated with a low-fat/high-fiber diet—high in carbohydrates and simple sugars (agrarian societies) [33]. Similar differences were seen in European children consuming western diets in contrast to children from Burkina Faso Africa who were found Table 1
Page 3 of 13 15
to have high concentrations of fecal short-chain fatty acids (SCFAs) and decreased abundance of Enterobacteriaceae (e.g., Shigella and Escherichia) in their gut microbiota [34•]. In mice, diets high in saturated (milk-derived) fat cause large shifts in gut microbiota along with the bloom of a rare sulfite-reducing pathobiont, Bilophila wadsworthia. In IL-10deficient mice that are genetically prone to the development of spontaneous colitis, this shift is associated with a proinflammatory response and increased incidence of colitis [35••]. Others have shown that a high-fat diet promotes colonization by adherent-invasive E. coli (AIEC) in the gut that is associated with decreases in mucus layer thickness, increased intestinal permeability, induction of NOD2 and TLR5 gene transcription, increased TNF-α secretion, and more severe inflammation in IL-10-deficient mice [36]. In contrast, CD patients fed with a Mediterranean diet exhibited a decreasing trend in CRP levels, decreased abundance of Proteobacteria and Bacillaceae, and increases in Bacteroidetes and Clostridium clusters. A significant change in the expression of over 3,000 genes before and after the intervention diet was observed in the host lymphocytes, but the clinical significance of this finding is still unknown [37]. Although several studies evaluating the effect of diet modification in IBD have been performed, there have been no large-scale randomized trials that have provided insights into the potential therapeutic actions of diet and dietary supplementation (Table 2).
IBD and Gut Dysbiosis: Cause or Effect? Many studies have now documented large changes in the gut microbiota (dysbiosis) in patients with active IBD. Decreased diversity and richness of intestinal microbiota have been reported among UC patients [38]. A recent study investigating twin pairs discordant for IBD revealed a reduction in microbial diversity in the healthy sibling mirroring the changes seen in the UC-affected twin, suggesting that alterations in the intestinal microbiota precede the development of disease [39].
Genetics, microbiome, and IBD Table 2
Diet, microbiome, and IBD
NOD2 mutation Animal models NOD2 knockout mice— increased terminal ileum colonization by Bacteroides, Firmicutes, and Bacillus [24]
CD patients Lower levels of IL-10 [22] Dysbiosis with lower counts of Clostridium groups XIVa and IV and an increase of Actinobacteria and Proteobacteria [25] Worse phenotypes (stricturing and fistulizing) and outcomes (higher rates of surgery) [26] Increased bacterial translocation into the bloodstream and decreased anti-TNF trough levels [27•].
High-fat, low-fiber (western diet) diet • Dysbiosis in humans as well as in mice [33, 34•] • Increased counts of sulfite-reducing bacteria (Bilophila wadsworthia) and adherent-invasive E. coli (AIEC) [35••, 36] Low-fat, high-fiber diet • High fecal concentration of SCFAs [34•] • Decreased abundance of Enterobacteriaceae [34•] Mediterranean diet • Lower CRP levels [37] • Decreased abundance of Proteobacteria and Bacillaceae [37] • Increased abundance of Bacteroidetes and Clostridium [37]
15 Page 4 of 13
In UC, the gut dysbiosis is often characterized by a reduction of Firmicutes [38, 40] and Bacteroidetes and a concomitant increase of Proteobacteria and Actinobacteria in both tissue and fecal samples [38, 39]. Individuals who are steroid responsive have a more diverse microbiota when compared to non-responders (Shannon index 338±62 vs 142±49; P= 0.013) [38]. A multicenter cohort pediatric study enrolling treatment-naive and newly diagnosed patients showed an increased abundance of Enterobacteriaceae, Pasteurellacaea, Veillonellaceae, and Fusobacteriaceae and a decreased abundance of Erysipelotrichales, Bacteroidales, and Clostridiales in ileal and rectal biopsies [41••]. Machiels et al. demonstrated a significant reduction of Faecalibacterium prausnitzii and Roseburia hominis in active UC patients vs control subjects. Moreover, they observed a significant inverse correlation between disease activity and the abundance of R. hominis and F. prausnitzii even in quiescent UC [42]. Wang et al. assessed the representation of Bifidobacterium and Lactobacillus in mucosal biopsies and fecal samples of 21 healthy controls, 15 CD, and 29 UC patients. F. prausnitzii was again significantly decreased in active CD as well as UC patients when compared to control. On the other hand, the abundance of E. coli was increased in both groups. Moreover, the increased proportion of E. coli in active UC was greater than that in active CD patients [43]. A dense bacteriophage community is associated with the gut mucosa, reaching 1010/mm3 of tissue. Interestingly, significantly more bacteriophages were detected in the intestinal mucosa of CD patients, especially in non-ulcerated mucosa, when compared to healthy subjects, raising the hypothesis that virus-like particles (VLP) may play an important role in the development and functional impact of intestinal dysbiosis [44]. Along these lines, a more recent study showed decreased diversity of VLP in fecal samples of CD patients when compared to controls [45]. The clinical significance of these findings remains unknown, due to lack of well-established genome inventories for identification and functional characterization of bacteriophages. While we have made a significant progress in understanding the role of bacteria in IBD, the role of fungi and Archaea in gut homeostasis and IBD etiopathogenesis is less well understood. Li et al. evaluated mucosal and luminal fungal microbiota of 19 active CD patients and 7 healthy controls, showing increased biodiversity and expansion of Candida spp., Gibberella moniliformis, Alternaria brassicicola, and Cryptococcus neoformans in the colonic mucosa. Candida albicans was only seen in the inflamed mucosa. In the analysis of fecal samples, there was an increased abundance of C. albicans, Aspergillus clavatus, and C. neoformans, the last two only seen in active inflammation. Interestingly, the diversity of the gut fungal community correlated with mucosal inflammation and elevated levels of TNF-α, and IFN-γ, and also with low levels of IL-10 and more severe disease activity in CD
Curr Gastroenterol Rep (2015) 17: 15
patients. The authors suggested that fungal microbiota in feces might be useful as a disease activity marker and predictor of response to medications in the future [46]. While much work has been done in describing and establishing associations between the gut dysbiosis of IBD and clinical phenotypes and outcomes, this information has not been useful in establishing causality. The interpretation of these data is also difficult because most studies have not factored in the many confounding factors that can affect microbial assemblage (e.g., medications, diet, surgery). In addition, most studies have been cross-sectional, i.e., where samples are taken at one point in time and often at different stages of disease. In moving forward, investigators have to consider the importance of prospective study design, uniformity in sample acquisition, and the necessity to defining more homogeneous patients’ subsets in order to better understand the role of gut microbiota in IBD (Table 3).
Gut Microbial Functions That Modulate Intestinal Inflammation The gut microbiota can significantly impact the production of micronutrients such as SCFAs, synthesis of vitamins, renewal of gut epithelial cells, development of immune responses and tolerance, and the fitness and virulence of pathogenic organisms [47, 48]. It is well known that germ-free mice have incomplete development of T and B cell functions, reflected by decreased numbers of circulating CD4+ T cells and reduced antibody production [49]. Segmented filamentous bacteria (SFB) induce T helper 17 (TH17) cell differentiation. Th17 cells induce secretion of IL-17, IL-22, and IL-21, which are necessary to mount an immune response against fungi and bacterial infections [50••]. In addition, certain strains of Clostridia and Bacteroidetes have been shown to increase the abundance of CD4+ Foxp3+ regulatory cells, which promotes tolerance to commensal bacteria and stimulates epithelial Table 3
Dysbiosis and IBD
Dysbiosis and IBD • Alterations in the intestinal microbiota may precede the development of IBD [39] • Dysbiosis may lead to metabolic pathways disruption and increased inflammation [47, 48] • Bacteriophage may play a role in intestinal dysbiosis and inflammation [44, 45] • Increased abundance of A. clavatus, and C. neoformans is seen in active inflammation [46] • F. prausnitzii has anti-inflammatory properties, and its abundance is decreased in UC and CD patients [42, 43, 52••, 53••] • Decreased production of SCFA is seen in CD and UC patients [54, 56] • Increased production of intestinal H2S in seen in UC patients due to increased abundance of Desulfovibrio and Bilophila wadsworthia [57–59]
Curr Gastroenterol Rep (2015) 17: 15
repair through NF-kB-dependent signaling. Perturbations in the gut microbiota can lead to an exacerbated inflammatory response contributing to chronic inflammation. A recent study demonstrated that dysbiosis in the oral cavity is associated with inflammatory responses in IBD patients, possibly reflecting a generalized disturbance in gut microbiota. Bacteroidetes was significantly increased with a concurrent reduction in Proteobacteria in salivary samples from IBD patients. Moreover, elevated levels of IgA, TNF-α, and IL-1β and a decreased level of lysozyme (an antimicrobial protein) were found [51]. Sokol et al. showed the protective effect of F. prausnitzii in a trinitrobenzene sulfonic acid (TNBS)-induced acute colitis appeared to be mediated by inhibition of NF-κB activation and IL-8 production [52••]. The antiinflammatory properties of F. prausnitzii were also observed in a dinitrobenzene sulfonic acid (DNBS) chronic colitis model. Live F. prausnitzii or its culture supernatant (SN) given for 7 to 10 days to mice with moderate and severe DNBS-induced chronic colitis resulted in a significant improvement of chronic colitis histological score, myeloperoxidase (MPO) levels, and pro-inflammatory colonic cytokines (IL-6, TNF-α, IFN-γ). Moreover, mice treated with F. prausnitzii or SN were protected against recurrent colitis when they were rechallenged with DNBS 3 days prior to sacrifice [53••]. Attempts to identify and isolate specific pathogens triggering IBD have not been successful, raising the possibility that disturbances in community-wide gut microbial dynamics and function are involved. In this regard, metagenomic and metaproteomic studies have shown that up to 12 % of total metabolic pathways of gut microbiota are perturbed in active IBD patients when compared to controls, while only 2 % of genera changes in stool and intestinal biopsies specimens of these subjects [54]. Moreover, decreased expression of microbial genes involved in butanoate and propanoate metabolism was observed in ileal CD [54] that correlated with decreased production of butyrate, acetate, and other SCFAs. There was also increased degradation of mucin, which was attributed to reduced abundance of Firmicutes in active CD and UC patients [42, 55]. Several studies have also shown that SCFAs affect the host immune balance by regulating the number and function of colonic T regulatory (Treg) cells. Observed decreases in SCFA production would therefore lead to enhancement of the pro-inflammatory state [56]. The intestinal dysbiosis, and the loss of beneficial microbial products, can facilitate the proliferation of diseasepromoting bacteria that produce pro-inflammatory metabolites. UC patients appear to produce more intestinal hydrogen sulfide (H2S) due to a combination of increased abundance of sulfite-reducing bacteria such as Desulfovibrio and low concentrations of thiosulfate sulfur transferase (TST), an enzyme responsible for H2S detoxication [57–59]. In IBD, increased expression of genes metabolizing the sulfur-containing amino acid cysteine has been observed, along with increased sulfate
Page 5 of 13 15
transport. H2S can impair DNA repair and inhibit SCFA oxidation and consequently its beneficial properties [60, 61]. Moreover, the metagenome is adapted to manage oxidative stress, which favors the survival of these types of microbes in hostile inflammatory environments. Enterobacteriaceae such as E. coli can utilize nitrate-derived products generated in the inflamed gut and outcompete commensal bacteria that require fermentation substrates [62]. These organisms can therefore perpetuate and proliferate in the setting of active inflammation, suppressing the commensal flora and perpetuating the inflammatory state (Table 3). In summary, despite a wealth of data that demonstrate a strong relationship between intestinal dysbiosis and IBD, mechanistic insights are still incomplete.
Interventions to Reshape the Gut Microbiota and Restore Host–Microbe Balance Probiotics in IBD The efficacy of probiotics for the treatment of IBD is controversial and difficult to assess due to differences in preparations (types of microbes and quality control), administration, patient groups, and endpoints. In general, probiotics in CD patients have not shown any additional benefit in inducing clinical remission or preventing clinical relapse [63•]. To date, the best evidence of probiotics in IBD is for the treatment and management of pouchitis [64, 65]. More recently, Persborn et al. enrolled 16 UC patients with severe pouchitis and 13 UC patients with a healthy pouch in a prospective clinical study evaluating the effect of probiotic supplementation (Ecologic 825, Winclove, Amsterdam, the Netherlands). The study showed a significant increase of E. coli K12 and horseradish peroxidase (HRP) permeability during active inflammation (3.7 (3.4–8.5) vs 1.7 (1.0–2.4) in controls (P
INFLAMMATORY BOWEL DISEASE (S HANAUER, SECTION EDITOR)
IBD and the Gut Microbiota—from Bench to Personalized Medicine Emanuelle Bellaguarda & Eugene B. Chang
Published online: 12 March 2015 # Springer Science+Business Media New York 2015
Abstract Inflammatory bowel diseases (IBD) are chronic relapsing inflammatory disorders involving the gastrointestinal (GI) tract, which arise from the confluence of genetic, immunological, microbial, and environmental factors. Clinical, genetic, and experimental data support the role of gut microbiota in contributing to the etiopathogenesis of these diseases. In IBD, the development of gut dysbiosis and imbalances in host–microbe relationships contribute to the extent, severity, and chronicity of intestinal inflammation. With continued advances in knowledge, technology, bioinformatics tools, and capabilities to define disease subsets, we will be able to lower risk and improve clinical outcomes in IBD through individualized interventions that restore host–microbial balance. This article provides a critical review of the field, based on the latest clinical and experimental information.
Keywords IBD . Gut microbiome . Dysbiosis . Probiotics . Prebiotics . Antibiotics . Host–microbe interactions . Mucosal immunology . Intestinal inflammation . Crohn’s disease . Ulcerative colitis . Precision medicine . Western diet . Fecal microbiota transplantation
This article is part of the Topical Collection on Inflammatory Bowel Disease E. B. Chang (*) Section of Gastroenterology, Hepatology, and Nutrition, The University of Chicago, 5841 S. Maryland Avenue, MC 6084, Chicago, IL 60637, USA e-mail: [email protected] E. Bellaguarda Division of Gastroenterology and Hepatology, Northwestern University, 676 North Saint Clair suite 1400, Chicago 60611, IL, USA
Introduction Inflammatory bowel diseases (IBD) comprise a group of chronic relapsing inflammatory disorders involving the gastrointestinal (GI) tract, which arise from the confluence of genetic, immunological, microbial, and environmental factors. IBD have traditionally been categorized into two main clinical phenotypes: ulcerative colitis (UC) and Crohn’s disease (CD). UC is characterized by continuous, superficial inflammation involving the rectum and extending up to the cecum, while CD is characterized by deep ulcerations distributed in a skipped pattern anywhere from the mouth to the anus. CD is a complex entity, and, because of its penetrating nature, these patients can present with fistulas, bowel strictures, and abscesses. Moreover, this phenotype is more associated with extra-intestinal manifestations such as arthritis, uveitis, and perianal disease, whereas UC can be associated with primary sclerosing cholangitis and ankylosing spondylitis. The incidence and prevalence of IBD have been rising significantly in America, Europe, and Asia in the past few decades, and the burden to health care costs has significantly increased [1–4]. New therapies have come out in the recent years offering a better quality of life to many patients. However, the causes and more importantly ways to prevent these debilitating diseases have not yet been found. There are several theories to the pathogenesis of IBD. It is well established that there is genetic predisposition to both CD and UC, although, in 77 % of CD patients and in more than 80 % of UC patients, no genetic associations can be found [5]. This fact should not be viewed as an argument against a genetic basis for IBD but merely underscores the limitations of genome-wide association studies (GWAS) in being able to identify less common genetic variants. On the other hand, it is clear that genetic risk is not sufficient by itself in causing IBD in most cases, and there is very likely a role for other factors in the development of these diseases [6]. In fact, many
15 Page 2 of 13
Bwestern^ disorders such as IBD are likely related to dramatic changes observed in the collective human microbiota attributed to increased rates of C-section, sanitation of living spaces, dietary changes, and early and indiscriminant usage of antibiotics [7]. Therefore, the study of the gut microbiome is key in understanding worldwide trends in IBD and also other emerging metabolic and complex immune disorders of the modern age [8–10]. The association between intestinal microbiota and IBD dates back to the 1920s when Rettger et al. studied the effect of Bacillus acidophilus in the intestinal microbiota [11, 12]. In 1940, Kirsner evaluated the possible correlation between streptococci and UC and the effect of oral sulfonamide in fecal bacteria [13]. In the late 1990s, the association between fecal microbiota and CD was evident when recurrent inflammation was observed after the fecal stream was reestablished in postoperative CD patients [14, 15]. These and many other studies have provided a strong descriptive basis for IBD, but clearly a better understanding of the genetic, molecular, cellular, and microbial basis for IBD was needed. In recent years, advances in bioinformatics and in the development of culture-independent methods to define the composition and function of the intestinal microbiota have led to tremendous insights into the structure (membership) and function of complex microbial communities. As a consequence, a great deal of information about host–microbial interactions and their role in IBD pathogenesis and treatment has been gained. However, many questions still remain. Is the dysbiosis associated with active IBD a cause or consequence, and does it contribute to the natural history and clinical outcomes? Are the changes in gut microbiota stable or continuously changing during the course of this disease? How do standard medical therapy, diet, and other environmental factors affect the microbial function and contribute to mucosal healing? How can we leverage this knowledge to lower risk and improve clinical outcomes in IBD? Answers to these questions will significantly advance the field and improve our understanding on how to effectively treat IBD. In this article, we aim to review recent advances and developments in the field.
Factors Determining the Assemblage of Gut Microbiota and IBD Risk The human microbiome project, funded by the National Institutes of Health in 2007, has catalogued, to date, 60 millions predicted genes, 600 microbial reference genomes, and 700 metagenomes [16]. An individual’s collective microbiome is believed to have as many as 10 to 100 trillions of cells, 10 times the number of eukaryotic cells in the human body. Most of our microbiota reside in the GI tract, which contains 10 out of 50 known bacterial phyla. Firmicutes and Bacteroidetes together account for more than 90 % of the intestinal
Curr Gastroenterol Rep (2015) 17: 15
microbiota. The remaining 10 % includes phyla such as Actinobacteria, Proteobacteria, Fusobacteria, Verrucomicrobia, and Cyanobacteria [17]. The biodiversity of gut microbes is believed to be in the range of 500 to 2, 000 different species [18]. In addition to the commensal gut bacteria, Archaea, fungi, and viruses are present, the latter primarily comprised of bacteriophages that are as numerous and far more diverse than their bacterial counterparts [19–21]. GWAS have identified over 160 single nucleotide polymorphisms (SNPs) associated with IBD. Many of these genes are involved in pathways that modulate how the host responds to microbial stimuli. Nucleotide-binding oligomerization domain-containing protein 2 (NOD2), also known as caspase recruitment domain-containing protein 15 (CARD15), was the first gene identified, supporting the link between gut microbiome and IBD etiopathogenesis. This gene is located on chromosome 16 and is expressed in monocytes/Paneth cells in the terminal ileum, encoding an intracellular receptor for the bacterial peptidoglycan muramyl dipeptide (MDP) [22, 23]. Thus, NOD2 recognizes components in the bacterial cell membrane and, together with other genes associated with IBD susceptibility (e.g., ATG16L1, CARD 9, IL23R), determines how patients will respond to microbial stimuli. The responses can then lead to changes in epithelial, mesenchymal, and immune functions that are involved in the maintenance of intestinal homeostasis. For example, NOD2-deficient mice exhibit increased colonization in the terminal ileum by Bacteroides, Firmicutes, and Bacillus and decreased ability to prevent proliferation of pathogenic bacteria. Moreover, germ-free mice have a decreased expression of NOD2 in terminal ileum enterocytes that can be reversed upon conventionalization [24]. In CD patients carrying NOD2 mutations, the emergence of certain pathobionts can potentially trigger a decrease in the immunomodulatory cytokine, IL-10 [22]. CD patients carrying NOD2 risk alleles (Leu1007fs, R702W, or G908R) can develop a dysbiosis characterized by a decrease in the Clostridium groups XIVa and IV and an increase of Actinobacteria and Proteobacteria [25]. NOD2 is also associated with more severe disease outcomes including higher rates of stricturing, fistulizing disease, and recurrent surgeries [26]. Furthermore, mutations of NOD2 and ATG16L1 are associated with increased bacterial translocation to the bloodstream in patients with active CD. NOD2-variant genotypes are associated with decreased phagocytic and bactericidal activity in the peripheral neutrophils as well as decreased anti-TNF trough levels [27•]. ATG16L1 is a gene that has an important role in autophagy and a variant of it, T300A, is associated with increased endoplasmic reticulum stress in Paneth cells as well as increased IL1-beta production in response to MDP in the terminal ileum of CD patients [28]. IBD patients with the T300A variant appear to have greater representation in the ileum of adherent-invasive Escherichia coli, which may be associated
Curr Gastroenterol Rep (2015) 17: 15
with increased risk for severe disease, development of fistulas, and surgery. However, it is also clear that the ATG16L1 risk variant is quite common in the general Caucasian population, and most of these individuals never develop IBD. Thus, other factors such as those encountered in the environment, from alterations in gut microbiota, epigenetic modifications, and other disease modifiers may play a role in the disease pathogenesis [29]. For example, mice hypomorphic for the ATG16L1 mutation and infected with murine norovirus develop mucosal inflammation resembling CD but exhibit far less pathology when infected with non-replicative murine norovirus [30]. These observations underscore the multifactorial nature of IBD and the complex interplay between host genetic, environmental, and gut microbial factors that can contribute to the pathogenesis of these diseases (Table 1).
Western Diet-Induced Changes in the Gut Microbiome and Triggering of Mucosal Inflammation in Genetically Susceptible Hosts The rising incidence and prevalence of IBD in Asian countries are believed to be related to rapid changes in our environment, societal norms, and increased industrialization. Among the many factors that may underlie this trend is the consumption of western type diets high in fat, processed sugars, protein, and overall calories [31]. Diet can significantly affect the gut microbial composition of individuals, which, in turn, can modulate both intestinal and immune homeostasis. The gut microbiota among primates are significantly different, which may be related to differences in diet (herbivore, carnivore, and omnivore) [32]. In humans, a high-fat, low-fiber diet (western diet) is associated with a predominance of Bacteroidetes and Actinobacteria, while Firmicutes and Proteobacteria are associated with a low-fat/high-fiber diet—high in carbohydrates and simple sugars (agrarian societies) [33]. Similar differences were seen in European children consuming western diets in contrast to children from Burkina Faso Africa who were found Table 1
Page 3 of 13 15
to have high concentrations of fecal short-chain fatty acids (SCFAs) and decreased abundance of Enterobacteriaceae (e.g., Shigella and Escherichia) in their gut microbiota [34•]. In mice, diets high in saturated (milk-derived) fat cause large shifts in gut microbiota along with the bloom of a rare sulfite-reducing pathobiont, Bilophila wadsworthia. In IL-10deficient mice that are genetically prone to the development of spontaneous colitis, this shift is associated with a proinflammatory response and increased incidence of colitis [35••]. Others have shown that a high-fat diet promotes colonization by adherent-invasive E. coli (AIEC) in the gut that is associated with decreases in mucus layer thickness, increased intestinal permeability, induction of NOD2 and TLR5 gene transcription, increased TNF-α secretion, and more severe inflammation in IL-10-deficient mice [36]. In contrast, CD patients fed with a Mediterranean diet exhibited a decreasing trend in CRP levels, decreased abundance of Proteobacteria and Bacillaceae, and increases in Bacteroidetes and Clostridium clusters. A significant change in the expression of over 3,000 genes before and after the intervention diet was observed in the host lymphocytes, but the clinical significance of this finding is still unknown [37]. Although several studies evaluating the effect of diet modification in IBD have been performed, there have been no large-scale randomized trials that have provided insights into the potential therapeutic actions of diet and dietary supplementation (Table 2).
IBD and Gut Dysbiosis: Cause or Effect? Many studies have now documented large changes in the gut microbiota (dysbiosis) in patients with active IBD. Decreased diversity and richness of intestinal microbiota have been reported among UC patients [38]. A recent study investigating twin pairs discordant for IBD revealed a reduction in microbial diversity in the healthy sibling mirroring the changes seen in the UC-affected twin, suggesting that alterations in the intestinal microbiota precede the development of disease [39].
Genetics, microbiome, and IBD Table 2
Diet, microbiome, and IBD
NOD2 mutation Animal models NOD2 knockout mice— increased terminal ileum colonization by Bacteroides, Firmicutes, and Bacillus [24]
CD patients Lower levels of IL-10 [22] Dysbiosis with lower counts of Clostridium groups XIVa and IV and an increase of Actinobacteria and Proteobacteria [25] Worse phenotypes (stricturing and fistulizing) and outcomes (higher rates of surgery) [26] Increased bacterial translocation into the bloodstream and decreased anti-TNF trough levels [27•].
High-fat, low-fiber (western diet) diet • Dysbiosis in humans as well as in mice [33, 34•] • Increased counts of sulfite-reducing bacteria (Bilophila wadsworthia) and adherent-invasive E. coli (AIEC) [35••, 36] Low-fat, high-fiber diet • High fecal concentration of SCFAs [34•] • Decreased abundance of Enterobacteriaceae [34•] Mediterranean diet • Lower CRP levels [37] • Decreased abundance of Proteobacteria and Bacillaceae [37] • Increased abundance of Bacteroidetes and Clostridium [37]
15 Page 4 of 13
In UC, the gut dysbiosis is often characterized by a reduction of Firmicutes [38, 40] and Bacteroidetes and a concomitant increase of Proteobacteria and Actinobacteria in both tissue and fecal samples [38, 39]. Individuals who are steroid responsive have a more diverse microbiota when compared to non-responders (Shannon index 338±62 vs 142±49; P= 0.013) [38]. A multicenter cohort pediatric study enrolling treatment-naive and newly diagnosed patients showed an increased abundance of Enterobacteriaceae, Pasteurellacaea, Veillonellaceae, and Fusobacteriaceae and a decreased abundance of Erysipelotrichales, Bacteroidales, and Clostridiales in ileal and rectal biopsies [41••]. Machiels et al. demonstrated a significant reduction of Faecalibacterium prausnitzii and Roseburia hominis in active UC patients vs control subjects. Moreover, they observed a significant inverse correlation between disease activity and the abundance of R. hominis and F. prausnitzii even in quiescent UC [42]. Wang et al. assessed the representation of Bifidobacterium and Lactobacillus in mucosal biopsies and fecal samples of 21 healthy controls, 15 CD, and 29 UC patients. F. prausnitzii was again significantly decreased in active CD as well as UC patients when compared to control. On the other hand, the abundance of E. coli was increased in both groups. Moreover, the increased proportion of E. coli in active UC was greater than that in active CD patients [43]. A dense bacteriophage community is associated with the gut mucosa, reaching 1010/mm3 of tissue. Interestingly, significantly more bacteriophages were detected in the intestinal mucosa of CD patients, especially in non-ulcerated mucosa, when compared to healthy subjects, raising the hypothesis that virus-like particles (VLP) may play an important role in the development and functional impact of intestinal dysbiosis [44]. Along these lines, a more recent study showed decreased diversity of VLP in fecal samples of CD patients when compared to controls [45]. The clinical significance of these findings remains unknown, due to lack of well-established genome inventories for identification and functional characterization of bacteriophages. While we have made a significant progress in understanding the role of bacteria in IBD, the role of fungi and Archaea in gut homeostasis and IBD etiopathogenesis is less well understood. Li et al. evaluated mucosal and luminal fungal microbiota of 19 active CD patients and 7 healthy controls, showing increased biodiversity and expansion of Candida spp., Gibberella moniliformis, Alternaria brassicicola, and Cryptococcus neoformans in the colonic mucosa. Candida albicans was only seen in the inflamed mucosa. In the analysis of fecal samples, there was an increased abundance of C. albicans, Aspergillus clavatus, and C. neoformans, the last two only seen in active inflammation. Interestingly, the diversity of the gut fungal community correlated with mucosal inflammation and elevated levels of TNF-α, and IFN-γ, and also with low levels of IL-10 and more severe disease activity in CD
Curr Gastroenterol Rep (2015) 17: 15
patients. The authors suggested that fungal microbiota in feces might be useful as a disease activity marker and predictor of response to medications in the future [46]. While much work has been done in describing and establishing associations between the gut dysbiosis of IBD and clinical phenotypes and outcomes, this information has not been useful in establishing causality. The interpretation of these data is also difficult because most studies have not factored in the many confounding factors that can affect microbial assemblage (e.g., medications, diet, surgery). In addition, most studies have been cross-sectional, i.e., where samples are taken at one point in time and often at different stages of disease. In moving forward, investigators have to consider the importance of prospective study design, uniformity in sample acquisition, and the necessity to defining more homogeneous patients’ subsets in order to better understand the role of gut microbiota in IBD (Table 3).
Gut Microbial Functions That Modulate Intestinal Inflammation The gut microbiota can significantly impact the production of micronutrients such as SCFAs, synthesis of vitamins, renewal of gut epithelial cells, development of immune responses and tolerance, and the fitness and virulence of pathogenic organisms [47, 48]. It is well known that germ-free mice have incomplete development of T and B cell functions, reflected by decreased numbers of circulating CD4+ T cells and reduced antibody production [49]. Segmented filamentous bacteria (SFB) induce T helper 17 (TH17) cell differentiation. Th17 cells induce secretion of IL-17, IL-22, and IL-21, which are necessary to mount an immune response against fungi and bacterial infections [50••]. In addition, certain strains of Clostridia and Bacteroidetes have been shown to increase the abundance of CD4+ Foxp3+ regulatory cells, which promotes tolerance to commensal bacteria and stimulates epithelial Table 3
Dysbiosis and IBD
Dysbiosis and IBD • Alterations in the intestinal microbiota may precede the development of IBD [39] • Dysbiosis may lead to metabolic pathways disruption and increased inflammation [47, 48] • Bacteriophage may play a role in intestinal dysbiosis and inflammation [44, 45] • Increased abundance of A. clavatus, and C. neoformans is seen in active inflammation [46] • F. prausnitzii has anti-inflammatory properties, and its abundance is decreased in UC and CD patients [42, 43, 52••, 53••] • Decreased production of SCFA is seen in CD and UC patients [54, 56] • Increased production of intestinal H2S in seen in UC patients due to increased abundance of Desulfovibrio and Bilophila wadsworthia [57–59]
Curr Gastroenterol Rep (2015) 17: 15
repair through NF-kB-dependent signaling. Perturbations in the gut microbiota can lead to an exacerbated inflammatory response contributing to chronic inflammation. A recent study demonstrated that dysbiosis in the oral cavity is associated with inflammatory responses in IBD patients, possibly reflecting a generalized disturbance in gut microbiota. Bacteroidetes was significantly increased with a concurrent reduction in Proteobacteria in salivary samples from IBD patients. Moreover, elevated levels of IgA, TNF-α, and IL-1β and a decreased level of lysozyme (an antimicrobial protein) were found [51]. Sokol et al. showed the protective effect of F. prausnitzii in a trinitrobenzene sulfonic acid (TNBS)-induced acute colitis appeared to be mediated by inhibition of NF-κB activation and IL-8 production [52••]. The antiinflammatory properties of F. prausnitzii were also observed in a dinitrobenzene sulfonic acid (DNBS) chronic colitis model. Live F. prausnitzii or its culture supernatant (SN) given for 7 to 10 days to mice with moderate and severe DNBS-induced chronic colitis resulted in a significant improvement of chronic colitis histological score, myeloperoxidase (MPO) levels, and pro-inflammatory colonic cytokines (IL-6, TNF-α, IFN-γ). Moreover, mice treated with F. prausnitzii or SN were protected against recurrent colitis when they were rechallenged with DNBS 3 days prior to sacrifice [53••]. Attempts to identify and isolate specific pathogens triggering IBD have not been successful, raising the possibility that disturbances in community-wide gut microbial dynamics and function are involved. In this regard, metagenomic and metaproteomic studies have shown that up to 12 % of total metabolic pathways of gut microbiota are perturbed in active IBD patients when compared to controls, while only 2 % of genera changes in stool and intestinal biopsies specimens of these subjects [54]. Moreover, decreased expression of microbial genes involved in butanoate and propanoate metabolism was observed in ileal CD [54] that correlated with decreased production of butyrate, acetate, and other SCFAs. There was also increased degradation of mucin, which was attributed to reduced abundance of Firmicutes in active CD and UC patients [42, 55]. Several studies have also shown that SCFAs affect the host immune balance by regulating the number and function of colonic T regulatory (Treg) cells. Observed decreases in SCFA production would therefore lead to enhancement of the pro-inflammatory state [56]. The intestinal dysbiosis, and the loss of beneficial microbial products, can facilitate the proliferation of diseasepromoting bacteria that produce pro-inflammatory metabolites. UC patients appear to produce more intestinal hydrogen sulfide (H2S) due to a combination of increased abundance of sulfite-reducing bacteria such as Desulfovibrio and low concentrations of thiosulfate sulfur transferase (TST), an enzyme responsible for H2S detoxication [57–59]. In IBD, increased expression of genes metabolizing the sulfur-containing amino acid cysteine has been observed, along with increased sulfate
Page 5 of 13 15
transport. H2S can impair DNA repair and inhibit SCFA oxidation and consequently its beneficial properties [60, 61]. Moreover, the metagenome is adapted to manage oxidative stress, which favors the survival of these types of microbes in hostile inflammatory environments. Enterobacteriaceae such as E. coli can utilize nitrate-derived products generated in the inflamed gut and outcompete commensal bacteria that require fermentation substrates [62]. These organisms can therefore perpetuate and proliferate in the setting of active inflammation, suppressing the commensal flora and perpetuating the inflammatory state (Table 3). In summary, despite a wealth of data that demonstrate a strong relationship between intestinal dysbiosis and IBD, mechanistic insights are still incomplete.
Interventions to Reshape the Gut Microbiota and Restore Host–Microbe Balance Probiotics in IBD The efficacy of probiotics for the treatment of IBD is controversial and difficult to assess due to differences in preparations (types of microbes and quality control), administration, patient groups, and endpoints. In general, probiotics in CD patients have not shown any additional benefit in inducing clinical remission or preventing clinical relapse [63•]. To date, the best evidence of probiotics in IBD is for the treatment and management of pouchitis [64, 65]. More recently, Persborn et al. enrolled 16 UC patients with severe pouchitis and 13 UC patients with a healthy pouch in a prospective clinical study evaluating the effect of probiotic supplementation (Ecologic 825, Winclove, Amsterdam, the Netherlands). The study showed a significant increase of E. coli K12 and horseradish peroxidase (HRP) permeability during active inflammation (3.7 (3.4–8.5) vs 1.7 (1.0–2.4) in controls (P
