C. difficile is a Gram-positive spore-forming anaerobic bacterium, which is ubiquitous in nature and also colonizes the human intestinal tract (Ryan and Ray, 2004). C. difficile causes diarrhea and colitis, often in individuals who have been treated with antibiotics for other medical conditions.

The implications of C. difficile infection in the therapies for IBD relapses were suggested in 1980 (Lamont and Trnka, 1980). Later studies suggest that C. difficile may contribute to the relapse or trigger the development of IBD in some patients. Patients with IBD are at a higher risk of developing C. difficile infection. Studies of hospitalized patients from United States found that C. difficile was more common in patients with UC (3.9%) as compared to non-IBD patients (1.2%), and an adjusted odds ratio for C. difficile infection in patients with IBD, CD and UC were 2.9, 2.1, and 4.0, respectively (Meyer et al., 2004; Rodemann et al., 2007). Further studies from United States showed that enteric pathogens were detected in up to 20% of relapsed IBD patients with C. difficile being the most common pathogen (up to 90%) of the detected pathogens (Binion, 2012; Pant et al., 2013). IBD patients with C. difficile infection have worse outcomes and experience higher rates of recurrence (Issa et al., 2007; Jodorkovsky et al., 2010; Navaneethan et al., 2012). While these findings support a role of C. difficile infection in participating the pathogenesis in relapsed IBD, 10% of C. difficile infection occurred at the time of IBD diagnosis. This and the finding that IBD patients in remission had a significantly higher presence of toxigenic C. difficile in their intestinal tract as compared to healthy controls (8.2 vs. 1%) suggest that C. difficile infection may be a risk factor for the development of IBD (Clayton et al., 2009). A more recent study from the Netherlands employing both retrospective and prospective approaches found a low prevalence of C. difficile that was not associated with IBD. Their finding indicated that C. difficile is not a common trigger for exacerbations of IBD in clinical practice in the Netherlands (Masclee et al., 2013).

The major virulence factors of C. difficile are the two toxins encoded by tcdA and tcdB genes (McDonald et al., 2005; Belyi and Aktories, 2010; Aktories et al., 2017). These toxins inactivate host Rho and Ras family GTPase by glucosylation leading to the disruption of epithelial cytoskeleton and tight junctions, which may contribute to the increased epithelial permeability and luminal fluid accumulation associated with C. difficile infection (Kim et al., 2005). TcdA and TcdB toxins also cause cell death via caspase-dependent and caspase-independent apoptosis, further damaging the intestinal epithelial barrier (Hippenstiel et al., 2002; Qa’Dan et al., 2002; Matarrese et al., 2007; Nottrott et al., 2007). Both toxins also induce the production of proinflammatory cytokines in intestinal epithelial and immune cells, including IL-8, TNF-α, IL-1, and IL-6 (Savidge et al., 2003). By using multiomic analysis, it was found that the most discriminating metabolite between IBD and IBD infected with C. difficile was isocaproyltaurine, a covalent conjugate of a C. difficile fermentation product isocaproate and taurine from damaged tissue (Bushman et al., 2020).

In addition to C. difficile, other enteric pathogens were also detected in patients with flares of IBD. A recent study from United States found that compared to patients without IBD, patients with CD had a higher prevalence of norovirus (p = 0.05) and Campylobacter (p = 0.043). Compared to patients without IBD, patients with UC had a lower prevalence of norovirus (p = 0.049) and a higher prevalence of Campylobacter (p = 0.013), Plesiomonas (p = 0.049) and Escherichia coli (p < 0.001). However, these enteric infections did not impact IBD outcomes (Axelrad et al., 2018). A recent study from UK also reported that patients with IBD in primary care were more likely to present with common infections including gastrointestinal infections of C. difficile, Salmonella, Shigella and Campylobacter (Irving et al., 2021). Campylobacter species isolated in diagnostic laboratories are mainly Campylobacter jejuni.

M. avium is a mycobacterial species that is commonly present in the environment, which consists of four subspecies, including M. avium subspecies avium, M. avium subspecies paratuberculosis (MAP), M. avium subspecies hominissuis and M. avium subspecies silvaticum (Uchiya et al., 2017). MAP is present in about 70% of dairy herds without causing problems (Nielsen et al., 2000). However, it can cause Johne’s disease (chronic intestinal inflammation in distal small intestine) in cattle (Singh et al., 2014). Due to the similarities in histopathology between Johne’s disease and CD, it has been proposed that MAP may play a role in CD. In 1984, an unclassified mycobacterial species was isolated from gastrointestinal tissues of two patients with CD (Chiodini et al., 1984). A 7-day-old goat inoculated orally with 50 mg of the isolated organism developed a granulomatous lesion at the distal small intestine. The authors of this study concluded that their findings raise the possibility that a Mycobacterium plays an etiologic role in at least some cases of CD (Chiodini et al., 1984). Following this, multiple studies examined the association between MAP and CD, using bacterial cultivation or a PCR method targeting the insertion element (IS 900) in the genome of MAP. These studies generated controversial results and the details of these studies were previously reviewed by Naser et al. (2014).

Several clinical studies using anti-mycobacterial antibiotics in the treatment of CD reported beneficial effects. These studies used a combination of multiple antibiotics including rifabutin, clofazimine and clarithromycin, plus either ciprofloxacin, metronidazole, or ethambutol (Agrawal et al., 2020a,b). However, these studies did not identify whether patients included in the treatment were positive for MAP. Furthermore, the multiple antibiotics used in these studies would have had effects on other non-mycobacterial bacterial species. Although these studies did not prove MAP causes CD, they provided evidence supporting that bacterial species are involved in the pathogenesis of CD. Other studies also reported that different antibiotics such as nitroimidazoles and clofazimine also appear to be effective in the treatment of CD, further supporting that bacterial species are involved in the pathogenesis of CD (Feller et al., 2010).

The pathogenic mechanisms of MAP in Johne’s disease were studied. MAP is taken in by microfold (M) cells following recognition of the bacterium by the fibronectin receptor (Secott et al., 2004). MAP bacteria that travel across the intestinal epithelium are phagocytosed by macrophages via complement receptors. Infected macrophages form granulomas and become a latent infection or active disease (Birkness et al., 2007; Rice et al., 2019).

B. fragilis is a Gram-negative anaerobe, a commensal bacterium of the intestinal tract. However, a particular strain, namely enterotoxigenic B. fragilis strain, which produces an enterotoxin encoded by the bft gene, was reported to be associated with IBD disease activation or flare-up in year 2000 by a study from United States (Prindiville et al., 2000). The enterotoxin encoded by the bft gene is a secreted metalloprotease (Moncrief et al., 1995; Franco et al., 1997). A study in 2017 from Iran reported that bft gene was detected in 51.4% of intestinal biopsies from patients with UC and 1.6% non-IBD samples (Zamani et al., 2017). The B. fragilis enterotoxin cleaved the adherens junction protein E-cadherin and induced IL-8 production in intestinal epithelial cells (Wu et al., 1998, 2004; Kim et al., 2001). The enterotoxigenic B. fragilis strain induced persistent colitis with inflammatory cell infiltration, crypt abscesses and epithelial ulceration in wild type specific-pathogen free C57BL/6 mice and lethal colitis in germ-free C57BL/6 mice (Rhee et al., 2009). In multiple intestinal neoplasia mice, enterotoxigenic B. fragilis induced colitis via T17 response pathway and also induced colonic tumors (Wu et al., 2009). A recent study from the Netherlands reported a different result. They detected a higher prevalence of B. fragilis in patients with active CD as compared to patients in remission, however, bft-positive B. fragilis was not higher and they also found that bft-positive B. fragilis increased epithelial resistance rather than damaging the barrier (Becker et al., 2021).

F. varium is a Gram-negative bacterium. It was reported in 2002 from Japan that F. varium in colonic mucosa of patients with UC was significantly higher than that in CD and healthy controls detected using immunohistochemical method (Ohkusa et al., 2002). This study also detected a significantly higher antibodies to extracts of F. varium antigens (Ohkusa et al., 2002). In a following study, the same research group examined the cytotoxicity of supernatants of 20 bacterial species isolated from intestinal tissues of patients with UC. Only the supernatant of F. varium, which was found to contain high concentration of butyric acid, killed Vero cells (Ohkusa et al., 2003). Enemas containing F. varium culture supernatants given to mice induced colonic ulceration, crypt abscesses, inflammatory cell infiltration and apoptotic changes (Ohkusa et al., 2003). In a different study, Tahara et al. (2015) detected Fusobacterium in inflamed intestinal tissues of 152 Japanese UC patients using quantitative real-time PCR and found that Fusobacterium species were commonly present in patients with UC and was associated with mild clinical phenotypes of UC. However, the PCR primers that Tahara et al. (2015) used were not specific for F. varium.

A recent study in 2016 from Korea isolated Fusobacterium species from colonic biopsies of 51 (15.6%) patients with IBD (26 CD and 25 UC). Fusobacterium bacteria were isolated from eight of the 51 patients with IBD, however, only two were F. avium (Lee et al., 2016).

E. coli is a Gram-negative facultative bacterium, which is a commensal bacterium of the human gut. However, some strains of E. coli are human enteric pathogens. The adherent-invasive Escherichia coli (AIEC) was reported to be associated with ileal CD in the adult population (Darfeuille-Michaud et al., 2004; Martin et al., 2004; Sasaki et al., 2007). AIEC strains were defined based on their ability to adhere to and invade intestinal epithelial cell lines assessed (Darfeuille-Michaud et al., 2004). Various cell lines such as Caco-2, intestine-407 and HT-29 cells were used in assessing AIEC strains. AIEC strains do not have specific virulence factors. Comparative genomic analysis of CD associated AIEC strains and non-AIEC strains did not identify any molecular marker that can differentiate AIEC strains from non-AIEC strains (O’Brien et al., 2017). In the past 20 years, extensive studies have been conducted to examine the pathogenic mechanisms of AIEC strains.

AIEC strains adhere to and invade intestinal epithelial cells. AIEC adhesion to intestinal epithelial cells was via interactions of the common type 1 pili adhesion FimH and carcinoembryonic antigen cell adhesion molecule 6 (CEACAM6) (Boudeau et al., 2001; Barnich et al., 2007; Carvalho et al., 2009; Dreux et al., 2013). CEACAM6 is a glycoprotein that is abnormally expressed in intestinal epithelial cells of patients with ileal CD (Barnich et al., 2007). AIEC strains also used long polar fimbriae to interact with M cells on the surface of Peyer’s patches, which facilitated the translocation of AIEC across monolayers of M cells (Chassaing et al., 2011).

AIEC strains resist the defense from macrophages. Macrophages play a critical role in clearance of bacterial infection. In macrophages, while AIEC bacteria targeted early by the autophagic machinery were degraded, those that escaped early uptake by autophagy were able to survive and replicate inside macrophages (Lapaquette et al., 2012). Impaired ATG16L1, IRGM, or NOD2, which are molecules affecting autophagy or bacterial sensing, favored AIEC persistence in macrophages and induced increased production of tumor necrosis factor (TNF)-α and IL-6 (Lapaquette et al., 2012). AIEC induced granulomas in vitro (Lapaquette et al., 2012). AIEC strains promote inflammation. AIEC strains induced production of cytokines such as IL-8, TNF-α, and IL-6 in both epithelial cells and macrophages (Subramanian et al., 2008; Lapaquette et al., 2012). Survival and replication of AIEC in macrophages did not cause macrophage death, but increased production of TNF-α and IL-6 (Lapaquette et al., 2012). AIEC strain LF82 induced intestinal inflammation in CEABAC10 mice, transgenic mice that express human CEACAMS (Carvalho et al., 2009). In conventional mice treated with streptomycin, AIEC strains also induced chronic intestinal inflammation and intestinal fibrosis (Small et al., 2013). CD8⁺ T cells showed a protective role in AIEC infection in streptomycin-treated mice, and depletion of CD8⁺ T cells increased cecal AIEC load, intestinal pathology, and fibrosis in C56BL/6 mice (Small et al., 2013).

Helicobacters are Gram-negative bacterial species. H. pylori is the well-known human gastric pathogen. Some Helicobacter species were also isolated from intestinal tract of humans and animals, which are referred to as enteric Helicobacter species. By using a genus PCR, a study from Germany in 2004 detected a higher presence of Helicobacter species in colonic and ileal biopsies of patients with CD (12%), UC (17%) than controls (4%), however the difference did not reach statistical significance (Bohr et al., 2004). Sequencing of the genus PCR products showed that the detected Helicobacter DNAs were non-pylori Helicobacter species (Bohr et al., 2004). By using a Helicobacteraceae family specific PCR, a significantly higher presence of members of Helicobacteraceae were found in intestinal biopsies and fecal samples from children with CD (Zhang et al., 2006; Man et al., 2008; Kaakoush et al., 2010). Further identification of the PCR amplified DNA revealed that most of the detected members of Helicobacteraceae were enteric Helicobacter species (non-pylori Helicobacter species that colonize the intestinal tract). In intestinal biopsies and luminal washings collected from adult patients with IBD and controls, a study from UK detected non-pylori Helicobactor by PCR in 2.9% patients with IBD and 8.1% controls (Basset et al., 2004). Studies of detection of enteric Helicobacter species in IBD were mainly conducted by PCR. The lack of isolated enteric Helicobacter species from patients with IBD has prevented the study of pathogenic mechanisms of these bacterial species.

C. concisus is an oral bacterium and some strains can also colonize the intestinal tract (Zhang, 2015; Liu et al., 2018b). The association between C. concisus and pediatric CD was first reported using PCR detection of C. concisus 16S rRNA gene in intestinal biopsies and bacterial isolation (Zhang et al., 2009). The association between pediatric CD and C. concisus was then demonstrated using fecal samples (Man et al., 2010b). Later studies also reported the association between C. concisus and IBD, including both CD and UC, in the adult population (Mahendran et al., 2011; Mukhopadhya et al., 2011; Kirk et al., 2016). Genome analysis showed that C. concisus contains two genomospecies, and C. concisus strains isolated from intestinal tract of patients with IBD were predominately genomospecies 2 (Chung et al., 2016; Wang et al., 2017; Cornelius et al., 2021). Analysis of genomes of C. concisus strains isolated from patients with IBD and healthy controls also identified molecular markers for strains that were associated with severe CD and UC, these markers were from genomospecies 2 C. concisus strains (Liu et al., 2018a; Liu F. et al., 2020). Several studies also examined the pathogenic mechanisms that may contribute to the pathogenesis of IBD.

C. concisus damages the intestinal epithelial barrier. The bacterium induced apoptosis in human intestinal epithelial HT-29 cells following 48 h incubation and the expression of barrier-forming tight junction protein claudin-5 reduced to 66% (Nielsen et al., 2011). Tight junction protein redistribution caused by C. concisus was also observed (Man et al., 2010a; Nielsen et al., 2011). C. concisus strain adhered to and some strains were invasive to intestinal epithelial cells (Man et al., 2010a; Ismail et al., 2012). Some C. concisus strains have acquired prophages that contain the zot gene (Liu et al., 2016). The C. concisus Zot protein caused cell death in human intestinal epithelial Caco-2 cells (Mahendran et al., 2016).

C. concisus enhances the responses of intestinal epithelial cells and macrophages to commensal bacteria. Preincubation of THP-1 macrophage-like cells and HT-29 cells with C. concisus Zot protein significantly enhanced the production of TNF-α by these cells in response to E. coli (Mahendran et al., 2016). Incubation of HT-29 cells with C. concisus upregulated the expression of TLR4 and MD-2, which primarily recognize LPS (Ismail et al., 2013). LPS is a core component of the outer membrane of Gram-negative bacteria. A low level of expression of TLRs is a mechanism of maintain the tolerance to gut bacteria (Santaolalla and Abreu, 2012). Increased expression of TLR4 and MD-2 on intestinal epithelial cells may promote these cells to respond to commensal gut microbes that they usually do not respond under homeostasis, which causes intestinal inflammation. C. concisus has a spiral to curved shape, a morphology that better equips bacterial species with abilities to travel in mucus and get close to epithelium.

C. concisus induces intestinal inflammation. C. concisus induced the production of IL-8 in intestinal epithelial cells (Man et al., 2010a; Ismail et al., 2013), and the production of IL-8 and TNF-α in human macrophages (Man et al., 2010a; Mahendran et al., 2016).

C. concisus also has the potential to affect the adaptive immune system. C. concisus upregulated epithelial expression of the immune checkpoint protein PD-L1 in interferon (IFN)-γ sensitized HT-29 cells, suggesting that while C. concisus induces innate inflammatory responses, it has the potential to inhibit the function of T cells (Lee et al., 2021). However, the upregulated expression of PD-L1 was only examined at an mRNA level. Whether the upregulated PD-L1 has functional bioactivity at a protein level is not yet known.

An individual’s gastrointestinal environment has an impact on the enteric pathogenicity of C. concisus. The growth of C. concisus is highly dependent on the presence of H₂ (Lee et al., 2014). In the human intestinal tract, the level of H₂ gas generated by bacterial species is greatly affected by diet and gut bacterial composition, which may play a role in determining C. concisus intestinal colonization (Strocchi and Levitt, 1992).

F. nucleatum is a Gram-negative anaerobe that colonizes the human oral cavity and intestinal tract. A study from Canada in 2011 reported a significantly higher isolation of F. nucleatum from intestinal biopsies of patients with IBD as compared with healthy controls (Strauss et al., 2011). Recently, F. nucleatum was reported to be abundant in intestinal tissues of patients with UC and IBD, and the abundance was associated with disease severity (Liu H. et al., 2020; Su et al., 2020).

The pathogenic mechanisms of F. nucleatum relating to IBD include damaging intestinal barrier and promoting inflammation. F. nucleatum strains isolated from inflamed biopsies of patients with IBD were more invasive to Caco-2 cells (Strauss et al., 2011). In mice with dextran sodium sulfate (DSS)-induced colitis, F. nucleatum promoted intestinal inflammation (Liu H. et al., 2020; Su et al., 2020). F. nucleatum damaged intestinal epithelial barrier by inducing autophagic epithelial cell death, affecting the expression and distribution of zonula occludens-1 (ZO-1) and occludin and skewing proinflammatory M1 macrophage (Liu L. et al., 2019; Liu H. et al., 2020; Su et al., 2020). In mice with DSS-induced colitis, F. nucleatum induced Th1 and Th17 subset T cell differentiations, promoted the secretion of proinflammatory cytokines such as TNF-α, IFN-γ, IL-1β, IL-6, and IL-17 (Liu L. et al., 2019).

F. nucleatum is also associated with colorectal cancer (Lee et al., 2019). Using colorectal cancer cell lines (HCT116, DLD1, SW480, and HT-29) and xenograft mice, it was shown that F. nucleatum promotes colorectal cancer cell growth by modulating E-cadherin and promotes cancer cell proliferation via FadA adhesin, activating TLR4 signaling to nuclear factor-kB (Rubinstein et al., 2013; Yang et al., 2017).

In recent years, numerous studies have compared the gut microbiota in patients with IBD and controls. These studies were recently reviewed by Pittayanon et al. (2020) in a systematic review published in Gastroenterology. Pittayanon et al. (2020) reviewed 48 such studies and found the following: (1) The diversity of microbiota was either decreased or not different in patients with IBD vs. controls. (2) For patients with CD, three studies reported a decrease of Christensenellaceae and Coriobacteriaceae as compared to controls. Faecalibacterium prausnitzii was reported to decrease in 6 of 11 studies. Two studies each of Actinomyces, Veillonella and E. coli revealed an increase of these organisms. (3) For patients with UC, Eubacterium rectale and Akkermansia were decreased in three studies, E. coli was increased in four of nine studies. (4) Pittayanon et al. (2020) stated that they cannot make conclusions due to inconsistent results and methods among studies.