The discovery of an “Achilles heel” in a type of gut bacteria that causes intestinal inflammation in patients with Crohn’s disease may lead to more targeted therapies for the difficult to treat disease, according to Weill Cornell Medicine and NewYork-Presbyterian investigators.
In a study published Feb. 3 in Cell Host and Microbe, the investigators showed that patients with Crohn’s disease have an overabundance of a type of gut bacteria called adherent-invasive Escherichia coli (AIEC), which promotes inflammation in the intestine.
Their experiments revealed that a metabolite produced by the bacteria interacts with immune system cells in the lining of the intestine, triggering inflammation. Interfering with this process, by either reducing the bacteria’s food supply or eliminating a key enzyme in the process relieved gut inflammation in a mouse model of Crohn’s disease.
“The study reveals a therapeutically targetable weak point in the bacteria,” said senior author Dr. Randy Longman, associate professor of medicine in the Division of Gastroenterology and Hepatology and the Director of the Jill Roberts Center for Inflammatory Bowel Disease at Weill Cornell Medicine and NewYork-Presbyterian/Weill Cornell Medical Center.
To find this “Achilles heel,” Dr. Longman and his colleagues, including Drs. Ellen Scherl and Chun-Jun Guo at Weill Cornell Medicine and collaborators Dr. Gretchen Diehl at Memorial Sloan Kettering and Dr. Kenneth Simpson at Cornell’s Ithaca campus, targeted a process the AIEC bacteria uses to convert a byproduct of sugar fermentation in the gut to grow.
Specifically, the AIEC uses 1,2-propanediol, a byproduct of the breakdown of a type of sugar called fucose that is found in the lining of the intestines. When the AIEC converts 1,2-propanediol, it produces propionate, which the study showed interacts with a type of immune system cell called mononuclear phagocytes that are also found in the lining of the gut. This sets off a cascade of inflammation.
Next, the investigators genetically engineered AIEC bacteria to lack a key enzyme in this process called propanediol dehydratase. Without propanediol dehydratase, the bacteria do not set off a cascade of inflammation in a mouse model of Crohn’s disease. Reducing the available supply of fucose in the animal’s gut also reduced inflammation.
“Changing one metabolic pathway in one type of bacteria can have a big impact on intestinal inflammation,” said the study’s co-lead author Dr. Monica Viladomiu, a post-doctoral associate in medicine in the Division of Gastroenterology and Hepatology and in the Jill Roberts Institute for Research in Inflammatory Bowel Disease at Weill Cornell Medicine. Maeva Metz, a Weill Cornell Medicine Graduate School of Medical Sciences doctoral candidate in Dr. Longman’s laboratory, is also co-lead author.
The discovery could lead to better treatments for Crohn’s disease, a type of inflammatory bowel disease that affects more than 4 million people worldwide. Currently, patients with Crohn’s disease are often treated with antibiotics, which can kill both beneficial and harmful bacteria causing unwanted side effects. But treatments that precisely target the inflammatory cascade discovered by Dr. Longman and colleagues might help reduce inflammation while preserving beneficial bacteria.
“If we can develop small molecule drugs that inhibit propanediol dehydratase or use dietary modifications to reduce the availability of fucose, we may be able to reduce intestinal inflammation in patients with Crohn’s disease with fewer side effects,” said Dr. Longman, who is also a member of the Jill Roberts Institute for Research in Inflammatory Bowel Disease.
One of the next steps for the team will be testing potential treatments. They also plan to study the potential role of an enzyme called fucosyltransferase 2 in protecting the gut against this inflammatory cascade.
Dr. Longman explained that many patients with Crohn’s disease have mutations in the gene that encodes this enzyme, rendering it nonfunctional.
“From a clinical perspective, that’s interesting because it may help us stratify people for whom one intervention or another maybe more useful,” Dr. Longman said.
Crohn’s disease (CD) is a worldwide chronic inflammatory bowel disease (IBD) whose incidence is increasing across Europe [1,2,3]. Although the precise aetiology of CD is unknown, it is well accepted that this disease is the consequence of immune-mediated injury to the gut mucosa inflicted by an overactive immune response towards environmental factors in a genetically predisposed host [4].
Sustained inflammation and chronic wound healing response often lead to intestinal fibrosis, a condition defined by an excessive accumulation of extracellular matrix (ECM) proteins produced by activated myofibroblasts, which are alpha smooth muscle actin (α-SMA)-expressing cells.
These cells not only derive from resident mesenchymal cells (fibroblasts and smooth muscle cells) but can also originate from epithelial and endothelial cells via epithelial/endothelial transition, as well as from stellate cells, pericytes and bone marrow stem cells [5]. Activation of myofibroblasts occurs in response to different stimuli, including growth factors such as transforming growth factor beta (TGF-β) and platelet-derived growth factor (PDGF), pro-inflammatory cytokines such as interleukin 1 (IL-1), IL-13 and IL-17, as well as CC chemokines like CCL2, CCL3 and CCL4, and lipid mediators released by immune and non-immune cells [6].
As a result, the formation of fibrotic scars in the intestinal wall results in a narrowing of the intestinal lumen and generates strictures and fistulae, or stenosis, in approximately 50% of CD patients. Despite the availability of treatments that target inflammation, no effective anti-fibrotic therapies exist, with surgical intervention being the only curative option although inflammation and stenosis may reoccur [5, 7, 8].
There is therefore an urgent need for a better understanding of the pathophysiology and identification of potential therapeutic targets in intestinal fibrosis.
A growing body of evidence suggests that an imbalance in the gut microbiota, or dysbiosis, is highly associated with CD pathogenesis, where it modulates the inflammatory status [9]. The bacterial microbiome has also been linked to intestinal fibrosis; however, the direct correlation between specific microbial species and fibrogenesis is still uncertain.
CD patients are characterized by having fewer bacterial phyla, Firmicutes and Bacteroidetes, with anti-inflammatory properties, and more Actinobacteria and Proteobacteria, with proinflammatory roles [9, 10]. Among the Gram-negative Proteobacteria, pathogenic adherent-invasive Escherichia coli (AIEC) has been preferentially observed in ileal CD [11, 12].
The prototype AIEC strain, LF82, colonizes the intestinal mucosa and induces proinflammatory cytokines in a number of acute dextran sulfate sodium (DSS)-induced colitis mouse models [13,14,15]. Adhesion to intestinal epithelial cells (IECs) has been shown to be mediated by the interaction of LF82 type 1 pili with the abnormally expressed human carcinoembryonic antigen-related cell adhesion molecule 6 (CEACAM6) [14], the binding of flagella to TLR5 and IPAF flagellin receptors [13], or via chitin-binding domains, encoded by bacterial chitinase ChiA, that interact with human chitinase CHI3L1 expressed on IECs under inflammatory conditions [15].
In addition to bacteria, fungal microbiota dysbiosis is also considered a possible cause of CD. A number of studies have shown a decrease in levels of non-pathogenic Saccharomyces cerevisiae yeasts [16, 17] and an increase in abundance of Candida species in CD patients [18].
Candida albicans is the most prevalent opportunistic fungal pathogen in the intestine. Our group has previously demonstrated that colonic inflammation induced by DSS promotes C. albicans colonization in mice. In turn, C. albicans, via the β-galactoside-binding lectin receptor galectin-3, augments inflammation, as revealed by increased expression of TLR-2 and TNF-α, and triggers antibody generation directed against C. albicans antigens and also anti-S. cerevisiae antibodies (ASCA), which are serological markers of CD [19, 20]. Together, these findings indicate that microorganisms and their products can be profibrogenic in the gut.
In the current study, we investigated the effects of the AIEC strain LF82 and C. albicans on chronic intestinal inflammation and fibrosis progression.
Discussion
Intestinal fibrosis is the most common complication in patients with CD. It is the final outcome of the gut mucosal reaction to chronic inflammation and repair, which results in excessive deposition of extracellular matrix, leading eventually to intestinal dysfunction.
Despite substantial efforts to identify the triggers for intestinal inflammation, the mechanisms underlying fibrosis remain poorly characterized and delay the development of effective anti-fibrotic therapies. Mounting evidence indicates that microorganisms can correlate directly with intestinal inflammation and could predict fibrosis. Here, we identified the AIEC strain LF82 as a new candidate organism affecting intestinal fibrogenesis, while the fungus C. albicans had no profibrogenic effect.
Most previous studies have shown that alterations in the bacterial microbiome are implicated in intestinal fibrosis. Antibiotic treatment in rats with chronic colitis significantly prevented TGFβ-1, collagen production and stricture formation [22], and intramural injection of faecal material or extracts from anaerobic bacteria into the intestinal wall induced chronic colitis with fibrosis and elevated levels of TGFβ-1 in colonic tissue [22, 23].
Additionally, a recent study showed that Tl1a-overexpressing mice had reduced colonic collagen deposition under pathogen-free conditions and interestingly proved causality by demonstrating that specific bacteria or bacterial consortia including groups of mucolytic bacteria such as Mucispirillum schaedleri, Ruminococcus, Anaeroplasma and members of the Streptococcus and Lactobacillus genera are directly correlated with the degree of fibrosis and fibroblast phenotype [24]. Severe and persistent intestinal fibrosis also occurred in mice infected chronically with Salmonella enterica [25].
AIEC is an E. coli pathotype that is present in high numbers in the inflamed gut of CD patients and its role in chronic colitis-associated fibrosis was first demonstrated in a chronic infection model of fibrosis using the AIEC strain NRG857c. NRG857c colonization persists for months in WT mice and is accompanied by chronic transmural inflammation and fibrosis [26].
This model was used in an attempt to avoid the use of LF82, which does not colonize conventional mice beyond 7 days and requires the expression of human CEACAM6 receptors to develop intestinal inflammation after acute DSS exposure [14]. In the present study, we observed the effects of LF82 on intestinal inflammation and fibrosis in WT mice in response to chronic DSS exposure by repeated LF82 challenges every 7 days. We demonstrated that LF82 worsened fibrosis, revealed by increased collagen deposition in the colon subepithelium and serosal areas and enhanced expression of the main fibrotic genes Col1a1, Col3a1, Fn1 and Vim.
Inflammation in CD has been shown to be driven predominantly by Th1 and Th17 responses [27]. We observed high gene expression of Il17, as well as of Ifng, which was consistent with elevated mRNA expression of the Th1 transcription factor Tbet, suggesting a Th1 response in our model. Other experimental colitis models have shown that both Th1- and Th2-mediated pathways contribute to the pathogenesis of CD, where they participate in different stages of chronic colitis development and affect diverse components of the inflammatory response [28, 29].
This may explain the increase in the Th2 transcription factor Gata3 in LF82-challenged mice after DSS exposure. Mucosal Treg cells and activated macrophages are also increased in paediatric and adult CD patients [30]. In our system, the Treg transcription factor Foxp3 was expressed at increased levels. Furthermore, there was also an upregulation in gene expression of the monocyte/macrophage marker Adgre1 and the neutrophil marker Ly6g, as well as higher expression levels of the proinflammatory cytokines related to them such as Il1b, Il6 and Il12b. Our results are similar to those observed with chronic NRG857c inflammation involving Th1 and Th17 responses and a significant role for macrophages and Treg cells [26].
Since EMT has been identified as a key contributor to the pool of activated fibroblasts associated with fibrosis in a mouse model of chronic colitis and to fistula formation in CD patients [31, 32], we evaluated the in vitro effect of LF82 on EMT and myofibroblast activation in TGF-β1-stimulated human IEC Caco-2 cells. LF82 aggravated TGF-β1-stimulated myofibroblast activation of these cells by EMT, as revealed by highly increased gene expression of mesenchymal cell markers FN1 and VIM and downregulated expression of the IEC marker OCLN.
TGF-β signalling in epithelial cells appears to play an anti-inflammatory role [33]; however, we did not observe any effect of TGF-β on expression of the proinflammatory cytokines IL1B, IL6, IL12B and IL8 in our model. LF82 had strong proinflammatory properties and thus was able to overexpress all of these genes in IECs in the presence of TGF-β. Overall, these findings indicate that LF82 is associated with severe intestinal inflammation and fibrosis, and can affect fibroblast function directly or possibly via its products. Future studies should identify the mechanism for LF82-stimulated intestinal inflammation and fibrosis.
The importance of the fungal microbiome has received little attention in the context of intestinal fibrosis. Here, we evaluated whether C. albicans, the most prevalent fungal species in CD patients, is positively or negatively correlated with fibrosis severity. Despite the pro-inflammatory effects observed, we found that C. albicans did not affect fibrosis severity in DSS-treated mice.
This could be due to certain products of this fungus that may have opposing effects on intestinal inflammation compared to fibrosis and this requires further investigation. In vitro studies showed no impact of C. albicans on myofibroblast activation and proinflammatory properties of TGF-β1-stimulated IECs, thus confirming the lack of a profibrogenic effect of C. albicans and that its effect on inflammation depends on different cell types.
Epithelial integrity is compromised in DSS-induced colitis suggesting penetration of microbes and diffusion of associated antigens into the mucosa [34]. One could thus think that the observed effects in our experiments had nothing to do with LF82 itself but reflect a difference between yeasts and bacteria, or may be attributed to both LF82 and other microbes colonizing the gut including other E. coli strains.
Pathogenic and commensal E. coli were both increased in CD [35], but there was no direct evidence that they could directly affect intestinal fibrosis. However, it has been shown that TLR4, which could be activated by Gram-negative derived LPS, mediates chronic intestinal inflammation and fibrosis by regulating cytokine expression on intestinal macrophages and myofibroblasts and inducing epithelial–mesenchymal transition [36].
These findings suggest a role for E. coli in inducing colitis and fibrosis. Additionally, it has been reported that stimulation of intestinal myofibroblasts with LPS can upregulate TLRs (2, 3, 4, 6, 7) and their accessory molecules (MyD88, TIRAP), activate the MAPK pathway and increase IL-8 secretion, thus indicating that intestinal myofibroblasts participate in the immune response in the intestine after activation by bacterial products and may play a role in CD-associated fibrosis [37].
In our study, the effects of DSS on gut epithelial injury, resulting in increased intestinal permeability, changes in the microbiota and immunological and fibrotic alterations, as observed in the DSS-treated group, were in the same range as in the LF82-treated and C. albicans-treated DSS-induced groups.
On the other hand, LF82 enhanced fibrogenesis compared to the DSS-treated group, meaning that we had an additive or synergistic proinflammatory and profibrogenic role of LF82. In addition, in in vitro experiments, LF82 enhanced the effect of the fibrogenic cytokine TGFβ inducing EMT and myofibroblast activation.
reference link: https://gutpathogens.biomedcentral.com/articles/10.1186/s13099-021-00401-z
More information: Monica Viladomiu et al. Adherent-invasive E. coli metabolism of propanediol in Crohn’s disease regulates phagocytes to drive intestinal inflammation, Cell Host & Microbe (2021). DOI: 10.1016/j.chom.2021.01.002
