The Role of Enterobacteriaceae in Gut Microbiota Dysbiosis in Inflammatory Bowel Diseases

1Inflammatory bowel diseases (IBDs) are a heterogeneous group of chronic, relapsing–remitting, gastrointestinal (GI) inflammatory diseases with various degrees of damage, promoting the development of local and extra-intestinal complications. IBDs include ulcerative colitis (UC) disease and Crohn’s disease (CD); the first is restricted to the colon, while the second can affect different parts of the digestive tract.

2Despite the etiology of IBDs remaining unknown, it is reported that there is an involvement of the host’s genetics, gut microbiota, and immune system. Genetically susceptible individuals seem to be more prone to developing IBDs as a consequence of their atypical immune response to the autologous gut microbiota, following exposure to different environmental factors and stimuli. Indeed, recent studies have revealed that pathologic alterations in the GI microbiota activate a mucosal immune response, thus developing chronic intestinal inflammation. These alterations in the GI microbiota trigger the so-called dysbiosis. Thus, the disruption of the intestinal eubiotic status can be considered a cause rather than simply a consequence of the chronic GI inflammation in IBDs.

3Many studies have reported an increase in the proportion of potentially harmful Proteobacteria, especially of the Enterobacteriaceae family, in IBD patients. Therefore, it would appear that the host’s inflammatory response could be the trigger of gut microbiota imbalance, most likely caused by Enterobacteriaceae blooming, which leads to the persistence of IBD’s inflammatory state (Figure 1).

4The main aim of this review is to describe and clarify the overgrowth of Enterobacteriaceae in IBDs, especially focusing on their role during the inflammation process in relation to metabolism changes observed in the gut microbiota of IBD patients. The secondary aim is to provide an update on certain adherent-invasive Escherichia coli (AIEC) strains frequently isolated from IBD patients, and their potential role in both CD and UC disease, discussing the difficulties in the characterization at species level of Enterobacteriaceae by targeted metagenomic approaches.

5Within the GI microbiota, the most studied component is the bacterial microbiota. Humans have more than 200 different bacterial species, with highly variable abundance in the gut. Thirty percent of this population is shared by different subjects, generating the core microbiota.

6The bacterial microbiota plays important functions in the host, such as (i) the modulation of the immune system, (ii) the secretion of enzymes involved in the digestion of substrates not fully accessible to the host, and (iii) competition with pathogenic microorganisms for the same ecological niche. In this context, the functionality of the GI tract is ensured and maintained also by the gut microbiota, which plays a positive role in the preservation of the intestinal barrier (Figure 1A).

7In particular, eubiotic microbiota is necessary to maintain the host’s immune homeostasis by the induction and function of T cells. Alterations in the gut microbiota’s composition can lead to the imbalance of T-cell subpopulations, such as Th1, Th2, Th17, and Treg cells, causing gut inflammation.

8Numerous studies support the involvement of the GI microbiota in the pathogenesis of IBDs. Indeed, characteristic dysbiosis is often observed in subjects with CD and UC; often, the induction or the maintenance of remission in IBDs is frequently related to the use of antibiotics and probiotics. Several studies have demonstrated changes in microbial richness and evenness, and the relative abundance of specific bacterial taxa in IBD patients with respect to healthy subjects. In particular, at the phylum level, the proportion of both Firmicutes and Bacteroidetes appears decreased, while the proportion of Proteobacteria and Actinobacteria is increased.

9As shown by Geveres and co-workers, an increase in Veillonellaceae, Pasteurellaceae, Enterobacteriaceae, and Fusobacteriaceae, and a decrease in Bacteroidales, Erysipelotrichales, and Clostridiales, were correlated with IBDs’ clinical severity. Similarly, a study on pediatric patients with CD concluded that the presence and severity of inflammation depends on microbiota dysbiosis; nonetheless, this study found an independent correlation between dysbiosis and other factors such as diet and/or the usage of antibiotics.

10In particular, patients suffering from IBDs are associated with a decline in the abundance of protective anaerobic commensal bacteria, such as Faecalibacterium prausnitzii, Clostridium spp., and Bacteroidetes fragilis. F. prausnitzii has been shown to have anti-inflammatory properties, including the ability to modulate the host’s mucosal immune response by the production of short-chain fatty acids (SCFAs). SCFAs, such as acetate, propionate, and butyrate, are a primary energy source for colonic epithelial cells. Thus, the underrepresentation of this bacterium in IBDs results in a decrease in the beneficial effects of SCFAs, including the inhibition of pro-inflammatory cytokine expression, the production of mucin and antimicrobial peptides, and tight junction protein downregulation (Figure 1B).

11Moreover, the gut microbiota in IBDs is often characterized by an expansion of translocating facultative aerobic bacteria—possible pathobionts—particularly belonging to the family of Enterobacteriaceae, such as Escherichia and Shigella. Pathobionts are defined as commensal microorganisms that might become pathogens in the setting of a specific environmental stimulus in genetically susceptible subjects.

12Another bacterium abundantly found in patients suffering from UC and CD is Fusobacterium nucleatum. This bacterium showed the ability to adhere to and invade the intestinal barrier, inducing aberrant inflammation and exacerbating colitis, both in human and animal models. Moreover, it has been recently proposed that the detection of F. nucleatum in the fecal metagenome could be used as an early biomarker of gut dysbiosis in IBDs and colorectal cancer.

13Although fewer studies have examined the role of fungi and viruses specifically in the propagation of inflammation in IBDs, both of them are ubiquitous components of the intestinal microbiota. As an example, patients with IBDs show differences in the presence of certain fungi compared with non-IBD controls, and different studies have revealed that this fungal dysbiosis is associated with an increased ratio between Basidiomycota and Ascomycota at the phylum level, indicating the emerging need to characterize the gut microbiota in IBDs.

14Among viruses, bacteriophages are the predominant elements of the so-called virome in healthy humans. In this context, changes in bacteriophage composition in IBD patients have been described. However, whether bacteriophages could display a direct role in IBDs still remains unclear and has yet to be determined.

15These bacteria belonging to the Enterobacteriaceae family are localized near the intestinal mucosa due to their relatively higher tolerance of oxygen dispersed by the epithelium.

16These bacteria are the first colonizers in the aerobic GI tracts of newborns. In fact, they are facultative anaerobes, which deplete oxygen to create a reduced environment suitable for following the colonization of strict anaerobes. Following this, bacterial oligosaccharide fermenters, such as Bifidobacterium, bloom in the GI tract because of breast milk intake. Subsequently, the introduction of a solid food diet rich in polysaccharides produces the expansion of anaerobes and polysaccharide fermenters such as Bacteroides, Clostridium, and Ruminococcus, and a simultaneous decrease in Bifidobacterium and Enterobacteriaceae.

17In IBD mouse models, microbiota dysbiosis in response to inflammation has been described, with a large relative increase in Enterobacteriaceae. These studies support the notion that Enterobacteriaceae seem to have a growth advantage over other members of the gut microbiota commensals in the inflamed GI mucosa.

18Moreover, an increment in Enterobacteriaceae was also associated with chronic UC inflammation compared to an acute status, showing a positive correlation of this bacterial family with a severe disease stage.

19Furthermore, in a CD pediatric cohort, Enterobacteriaceae enrichment has been associated with an aggressive disease course with a higher risk of treatment failure.

20In IBD patients, the functions of the gut microbiota have also been found to be perturbed. As an example, bacterial amino acid biosynthesis and carbohydrate metabolism are diminished, whereas nutrient uptake is enhanced. In these patients, bacterial genes such as redox tolerance, secretion systems, and adherence and/or invasion, are overrepresented and the pathways linked to the production of bacterial SCFAs are repressed.

21In a mouse model of UC, the blooming of Enterobacteriaceae has been linked to tetrathionate respiration, a metabolic pathway promoting intestinal colonization by Salmonella enterica subsp. Typhimurium, and with benzoate degradation, a pathway linked to Enterobacteriaceae growth and virulence pathways. To date, the link between functional and compositional microbiota changes and IBD pathogenesis remains unclear. Moreover, it remains to be discovered whether the differences in the microbiota originating prior to the initiation of the disease process are a mere consequence of a hyperactive immune response.

22The mechanisms by which Enterobacteriaceae bloom in IBDs, particularly during inflammation, are not completely known. In order to explain this observation, the “oxygen hypothesis” has been formulated.

23In physiological conditions, enterocytes reduce oxygen levels in the gut lumen by beta-oxidation processes, creating an anaerobic environment. In IBD patients in which intestinal inflammation occurs, beta-oxidation is reduced, and the oxygen level is increased. This event leads to an increase in facultative aerobes such as Proteobacteria/Enterobacteriaceae and, as a consequence, intestinal dysbiosis.

24Moreover, in inflammatory conditions, an increment in nitrate production by the host has been demonstrated. The expression of the gene encoding nitric oxide synthase is inhibited by the SCFA butyrate, via the activation of the peroxisome proliferator-activated receptor (PPAR). Butyrate represents the major energy source for enterocytes and it is involved in the maintenance of colonic mucosal health. Intestinal inflammation leads to a reduction in healthy butyrate-producing microbiota, which leads to an increase in nitrate production and then the blooming of Enterobacteriaceae. Consequently, the increased levels of both oxygen and nitrate observed in IBD patients could lead to Enterobacteriaceae overgrowth.

25Enterobacteriaceae contain molecular components that directly enhance the inflammatory response. Microbe-associated molecular patterns (MAMPs) are molecules located on the bacterial surface. These molecules interact with the receptors on immune cells to trigger inflammation. One of the most potent MAMPs is the endotoxin lipopolysaccharide (LPS), which interacts with Toll-Like Receptor (TLR) 4 on immune cells. The lipid-A of LPS can be acylated with five or six acyl chains. Hexa-acylated LPS shows immunostimulant activity that is 100-fold more active than penta-acylated LPS. Enterobacteriaceae, and, in general, Proteobacteria, possess a more immunostimulatory version of LPS. Moreover, Enterobacteriaceae present unmethylated immunostimulatory motifs that can induce an immune response through an interaction with TLR-9. The blooming of Enterobacteriaceae, which contain these motifs, leads to a more selective pressure and shift toward an Enterobacteriaceae-dominated community.

26Some scientific evidence leads to the hypothesis that dysregulation in innate immune systems induces Proteobacteria overgrowth, which promotes intestinal inflammation. Regarding this, it has been reported that mice lacking TLR-5 developed transmissible spontaneous colitis associated with an abnormal expansion of Enterobacteriaceae. Moreover, spontaneous colitis has also been reported in interleukin (IL)-10-deficient mice, due to their intolerance of intestinal microbiota. Indeed, IL-10 is the main anti-inflammatory cytokine required for immune tolerance to naïve microbiota, and in this mouse model, the onset and progression of gut inflammation has been correlated with an increase in E. coli.

27Several studies also reported the role of Enterobacteriaceae in inducing the secretion of IL-8, tumor necrosis factor (TNF)-α, and IL-1β, and in the disruption of intestinal mucosa tight junctions, followed by an increment in intestinal permeability and inflammation.

28Furthermore, a positive correlation between the nucleotide-binding oligomerization domain (NOD)-2 risk allele and Enterobacteriaceae was described in intestinal biopsies from an adult IBD patient cohort. Since NOD-2 is an intracellular bacterial-sensing receptor that drives inflammatory signaling, this evidence strengthens the hypothesis of the direct involvement of Enterobacteriaceae in inflammation.

29In IBDs, an impairment in the microbial bile acid (BA) metabolism has been demonstrated, with high levels of fecal-conjugated BAs and low levels of secondary BA production. These impairments in deconjugation and transformation abilities were associated with microbiomes.

30In physiological conditions, cholesterol and BAs regulate multiple metabolic processes in the host. Hepatocytes synthesize primary BAs from cholesterol conjugated to glycine and taurine. These conjugated BAs are stored in the gall bladder and are involved in lipid absorption and fat emulsification. Gut microbiota carry out numerous BA biotransformation reactions, including the hydrolysis of conjugated BAs and glycine or taurine by bile salt hydrolase (BSH), and BAs’ 7α/7β-dehydroxylation by a multistep biochemical pathway found only in anaerobic bacteria.

31An impairment in the gut microbiota’s metabolism significantly harms BAs’ metabolism and, consequently, the host’s glucose and cholesterol homeostasis. Moreover, a decrease in BAs seems to favor the blooming of pathogenic and pro-inflammatory members of Enterobacteriaceae. These changes in the gut microbiota’s ability of BA modification may be a significant factor in the onset or progression of IBDs. Furthermore, modified BAs display altered binding profiles for BA receptors (i.e., Farnesoid X receptor (FXR)), inferring that the dysbiosis of the gut microbiota’s characteristic of IBDs may alter the capacity for BA modification in this community. Cirrhotic patients have demonstrated a positive correlation between BA dysmetabolism and Enterobacteriaceae overgrowth. Moreover, in obese patients the administration of vancomycin increased the abundance in Proteobacteria and, concomitantly, decreased fecal secondary BAs, with a simultaneous postprandial increase in primary BAs in plasma. Finally, Heinken and co-workers demonstrated through a computational model that the dysbiotic IBDs’ microbiota were enriched in pathogenic Escherichia spp., which highly contributed to the BAs’ deconjugation transformation.

32Future studies should be undertaken to validate and strengthen the evidence that Enterobacteriaceae play a significant role in BA dysmetabolism in IBDs, and to explore if BA modification driven by dysbiosis might represent a marker of IBDs’ onset or progression.

33One of the well-studied members of the Enterobacteriaceae family is E. coli, which is frequently able to colonize the human intestine. This bacterium has been proposed as a possible cause of the beginning of disease in IBD patients; indeed, several studies have found an increased number of virulent E. coli strains isolated from IBD patients with respect to healthy controls.

34In the genome of E. coli strains, there are three nitrate reductases and three nitric oxide reductases. Hence, non-fermentable nitrates can be used by E. coli as a substrate for nitrate respiration and converted into fermentable nitrates. As a consequence, the expansion of E. coli/Enterobacteriaceae in the lumen of the inflamed gut observed in IBD subjects could reflect the high levels of non-fermentable nitrate generated in the gut by the host’s inflammatory response. Consequently, in human and animal models, the presence of E. coli strains in IBDs has been reported (Table 1). However, whether E. coli strains may specifically benefit from nitrate respiration during IBDs has yet to be studied.

35Most of the E. coli strains identified in IBD samples are AIEC. AIEC members are (i) able to adhere to and invade epithelial cells utilizing a process dependent on actin microfilaments and microtubule recruitment; (ii) able to survive within the macrophage phagolysosome; (iii) able to induce the production of tumor necrosis factor (TNF) from infected macrophages; and, finally, lacking any defined invasive element. The adaptive evolution of the genome of AIEC strains in a susceptible host and their ability to promote inflammation makes these microorganisms potential pathobionts.

36Several studies have reported that different E. coli strains persist within epithelial cells and macrophages and selectively colonize the ileum of CD patients. In particular, AIEC strains adhere to epithelial cells by the binding between the type 1 pili, expressed by bacteria, and the carcinoembryonic antigen-related cell adhesion molecule (CEACAM) 6, expressed on the apical surface of ileal epithelial cells. Epithelial expression of CEACAM 6 is stimulated by interferon (IFN)-γ and TNF, which are overexpressed by macrophages harboring AIEC strains, generating a positive feedback loop.

37E. coli strains from ileal biopsy from CD patients express several virulence genes mediating epithelial adherence and iron acquisition. Moreover, AIEC strains possess a wide arsenal of genes, such as fliC, OmpC, and others that regulate type 1 pilus expression and mediate epithelial cell adherence/invasion.

38Further observations suggest that UC-associated AIEC could be distinct from those observed in CD, with greater invasive properties or displaying different types of gene expression. Although several studies confirmed the presence of AIEC strains in a higher proportion with respect to CD patients, others reported a lower prevalence compared with that of CD. In the systematic review of Nadalian and co-workers, an increase in AIEC strains in UC patients emerged with respect to non-IBD controls, as well as the possible involvement of these bacterial strains in the pathogenesis of this disease.

39Some studies of animal models suggest that the accumulation of AIEC and other E. coli strains in the gut could be a consequence of inflammation. Moreover, several studies revealed that AIEC infection could induce changes in the gut microbiota. Nowadays, it still remains unclear whether AIEC strains cause intestinal inflammation leading to IBDs or whether their blooming could be a consequence of inflammation, which leads to the aggravation of these diseases.

40The number of genera and species in the Enterobacteriaceae family has markedly increased. Currently, this family comprises >60 genera and 250 species. To better define the role of Enterobacteriaceae in gut inflammation or in intestinal microbiota dysbiosis, it is important to distinguish each family component at genus or species level. As reported in Table 1, the main studies in which Enterobacteriaceae species have been correlated to IBDs are based on bacterial identification using culture-based techniques. In fact, the low sequence variance in 16S rRNA partial gene sequencing, usually selected in 16S rRNA-based metagenomics, does not allow resolution at genus/species level.

41Thus, it is mandatory to select other molecular targets that offer a superior discriminatory power of Enterobacteriaceae at the level of genus/species, especially for gut microbiota studies with targeted metagenomics approaches. In the past, different genes, such as tuf, atpD, dnaJ, rpoB, infB, and gyrB, have been explored using molecular biology techniques.

42Phylogenetic analyses of the tuf and atpD gene sequences have demonstrated controversial results in the identification of some Enterobacteriaceae species from isolated microbiological culture.

43The dnaJ gene encodes the heat shock protein 40 and contains regions highly conserved in the Enterobacteriaceae family, suitable for broad-specificity primer design. Despite this target demonstrating high performance in the unambiguous identification of Enterobacteriaceae species, it remains suboptimal for Enterobacteriaceae discrimination from direct specimens. In fact, the amplicon generated by polymerase chain reaction (PCR) is relatively long (758 base pair (bp)) for metagenomic-based approaches, which require short-length fragments.

44The rpoB gene encodes the RNA polymerase subunit, and a partial region of this locus has been employed in describing nematode microbiota. This study revealed that this target could be considered a highly appropriate marker for assessing the taxonomic structure of mock communities; however, further studies confirming this are required.

45The infB gene is a universally distributed gene in all prokaryotes and encodes for translation initiation factor 2 (IF2). In Hedegaard and colleagues’ study, a partial sequence of infB was used to develop a method for identifying enterobacteria at species level. The authors concluded that infB can be used to both discriminate and classify strains of the same species.

46The gyrB gene, which encodes the subunit B protein of DNA gyrase, has been previously used for describing phylogenetic relationships between Enterobacteriaceae. A fragment of the gyrB gene (gyrBint) of 506 bp has been successfully employed to identify Enterobacteriaceae, proving its applicability in a clinical laboratory.

47Nowadays, shotgun metagenomics approaches could allow gut microbiota taxonomic profiling, including Enterobacteriaceae, at species level, from direct samples. However, sample contamination by high levels of the host’s DNA and the high cost of the analysis represent drawbacks of shotgun metagenomics with respect to targeted metagenomics. Moreover, comparative studies on gut microbiota composition using both approaches have highlighted the differential depiction of bacterial profiles, although there was a consistent description of core microbiota. Thus, both metagenomics approaches could be considered complementary in the genus-/species-level description of Enterobacteriaceae.

48There is a plethora of evidence on the contribution of microbiota dysbiosis to gut inflammation in IBDs. The loss of beneficial microbes that produce anti-inflammatory molecules favors Enterobacteriaceae as the environment for the expansion of pathogens and/or pathobionts. The proinflammatory nature of Enterobacteriaceae has been well established. These bacterial family members contain molecules such as MAMPs and LPS that directly enhance the inflammatory response of the host. Moreover, Enterobacteriaceae could also participate in intestinal dysmetabolism, altering the BA metabolism, formerly impaired in IBDs.

49Amongst Enterobacteriaceae, AIEC strains colonize the intestinal mucosa of IBD patients. Several studies in animal models correlated the intestinal overgrowth of these bacteria to the gut inflammation trigger. However, the contribution of specific Enterobacteriaceae species or shifts in the IBD gut microbiota remains under debate.

50In fact, is emerged the difficulty to discriminate by the 16S rRNA-based metagenomics the effective composition of Enterobacteriaceae in IBDs and to define the active role of each component in these diseases.

51The identification of Enterobacteriaceae components and the deduction of their function in intestinal dysbiosis in IBDs represents a stimulating challenge, requiring expertise and effort in order to improve the knowledge and the comprehension of the fine mechanism of IBDs’ pathogenesis and progression.