Several studies on F. nucleatum have suggested the involvement of this bacteria in triggering epithelial-mesenchymal transition (EMT) in colon tissues. F. nucleatum is known to adhere and invade epithelial cells by employing several virulence factors, such as Fusobacterium adhesin A (Fad A), Fusobacterium autotransporter protein 2 and Fusobacterium outer membrane protein A responsible for promoting pro-oncogenic signaling and CRC development (34-36). In order to provide further insight into the metabolite alteration leading to EMT induced by these virulence factors in colorectal carcinogenesis, Ma et al carried out a detailed analysis on the interactions of these virulence factors with E-cadherin/β-catenin resulting in an altered molecular mechanism upon the infection of epithelial cells with F. nucleatum (37). Upon infection, F. nucleatum was found to increase the phosphorylation of p65, a crucial component of NF-κB signaling and to enhance the production of interleukin (IL)-6, IL-1β and matrix metalloproteinase (MMP)-13. This activation of NF-κB signaling indicated that F. nucleatum induced inflammation in colon tissue, which subsequently led to tumorigenesis (37). From previous studies conducted on cell adhesion, it is a well-known fact that the loss of E-cadherin and β-catenin-mediated cell adhesion is a crucial factor for inflammation and the induction of tumorigenesis (38,39). Surprisingly, F. nucleatum did not alter the expression levels of E-cadherin and β-catenin in epithelial cells and was found to interact only with E-cadherin to induce inflammatory response and promote malignant phenotype of CRC (37). Another study by Abed et al revealed that Fap2, a Gal-GalNAc binding protein, is expressed on the surface of F. nucleatum (36). Fap 2 functions as Gal-GalNAc lectin in CRC cells, where it binds to Gal-GalNAc overexpressed in CRC cells and causes the expansion of CRC by inhibiting antitumor immune activity by targeting T cell immunoglobulin and ITIM domain (TIGIT) (35,36). This indicates that virulence factors, such as FadA along with surface proteins of F. nucleatum may be a potential target for the development of an immunotherapy against CRC.

The role of F. nucleatum in experimental colitis was recently confirmed and it was found that F. nucleatum along with dextran sodium sulfate (DSS) synergistically promoted EMT and the aggressiveness of CRC (40). F. nucleatum was found to play a pro-tumorigenic role by activating the EGFR, AKT and ERK signaling pathway in azoxymethane (AOM)/DSS-induced CRC. The F. nucleatum cell surface protein, FadA, induces DNA damage by upregulating chk2 in CRC progression. Treatment with F. nucleatum resulted in DNA damage in adenomatous polyposis coli (APC) (−/+) mice (41). The number, size and tumor burden were reduced following treatment with FadA^−/− F. nucleatum as compared to WT, demonstrating a direct role of the cell surface protein of F. nucleatum in tumor progression (41). Recently, Okita et al described a heavy to moderate load of F. nucleatum (Fn) DNA associated with high microsatellite instability (MSI-H) and L/E [L: MSI-L/E: Elevated level of microsatellite alterations at selected tetra-nucleotide repeats (EMAST)] in two CRC cohorts; they further presented evidence that Fn activated factors that promote γ-H2AX, a marker for DNA damage (42). In another study, murine and human enteroid-derived monolayers (EDMs) co-cultured with Fn exhibited a downregulation of a key DNA repair protein, NEIL2 (DNA glycosylase) (43). NEIL^−/− EDMs exhibited increased DNA damage and elevated cytokine levels. Furthermore, NEIL2 downregulation is mainly observed in microsatellite stable (MSS) CRC, as compared to MSI CRC, indicating that Fn accumulation induces NEIL2 downregulation, resulting in DNA damage and thus leading to CRC progression. A recent study on microRNA (miRNA/miR) profiles and proteomic analysis from Fn-infected and non-infected cells found that the generation of miR1246/92b-3p/27a-3p-rich and CXCL16/RhoA/IL-8 exosomes promoted pro-metastasis (44). Bone-marrow derived mesenchymal stem cells (BMSCs) infected with Fn have also exhibited increased tumor growth and an enhanced susceptibility to tumors an APC (−/+) mouse transplantation model. Fn-infected BMSCs also displayed accelerated cancerinitiating potential and invasiveness in a nude murine model by activating Wnt-β-catenin-transforming growth-interacting factor (TGIF) signaling pathways (45). The co-culture of human CRC cell lines (SW480 and HCT116) with Fn was also shown to result in an increased expression of mesenchymal markers, such as Vimentin, Snail and Zeb1. Furthermore, Fn infection increased the capability of intrusion, migration and tumor-sphere formation along with a high expression of CD44, indicating that Fn plays a critical role in EMT and the cancer stem cell phenotype (46). These finding confirmed that the Fn play important role in the progression of CRC.

Enterotoxigenic Bacteroides fragilis (ETBF) is an enterotoxin producing bacterium found in an abundance in fecal samples obtained from patients with CRC is known to cause virulence with the aid of a metalloprotease holotoxin, BFT, also known as fragylisin (47). BFT is associated with acute diarrhea, IBD and CRC. This BFT binds to the colonic epithelial cell (CEC) surface protein receptor, initiating E-cadherin cleavage (47). E-cadherin in its intracellular domain is bound to β-catenin, which is a very crucial component of the Wnt signaling pathway. The degradation of E-cadherin releases β-catenin and facilitates EMT, and increases cell permeability by decreasing cell adhesion (48). The degradation of E-cadherin also facilitates the production of chemokines, such as IL-6, IL-8 and IL-1β, which activate NF-κB and MAPK signaling in CEC, generating inflammatory response in cells followed by carcinogenesis (49). The binding of BFT to CEC receptor also leads to the activation of key cellular pathways, such as signal transducer and activator of transcription (STAT)3 along with the formation of reactive oxygen species and nitrogen species which are known to cause DNA damage in epithelial cells and this genomic instability caused due to damaged DNA enhances expression of several oncogenes such as c-Myc by multi-folds, leading to the development of adenomas and carcinomas in the colon and rectum (50). Zamani et al investigated the mucosal colonization of ETBF to find the potential association of ETBF in benign and malignant lesions. B. fragilis was abundantly associated with patient samples as compared to healthy control (51). ETBF was increasingly associated with serrated lesions and with low-grade dysplasia adenoma. Liu et al found that ETBF increased the stemness of CRC by upregulating Nanog and Sox2 expression. They found that mechanistically, ETBF significantly elevated JMJD2B by activating TLR4 pathway (52). In a colitis CRC model, with ETBF colonization in wild-type BALB/c mice, the administration of AOM/DSS increased the rapid development of large number of polyps mainly in the colon and rectum (53). Roberti et al studied the contribution of gut microbes to the elicitation of the follicular helper T (T_FH) response (54). Ileal microbiota (Bacteroides fragilis and Erysipelotrichaceae) was involved in the protective immune response against colon cancer. The ileal microbiome plays an important role in immune surveillance and in the prognosis of proximal colon cancer and chemotherapy-mediated ileal crypt apoptosis (54). Bacteroides fragilis-associated long non-coding RNA (lncRNA1: BFAL1) mediates ETBF cancer progression in CRC; BFAL1 expression is increased in CRC compared with adjacent normal tissues. ETBF promotes CRC tumor growth through BFAL1 by activating RHEB (Ras homolog)/mTOR signaling. BFAL1 and ETBF are highly expressed in tumor tissues and predict poor outcomes in CRC (55).

E. coli is a Gram-negative commensal bacterium of the human microbiota and represents the most common cultivable and aero-anaerobic bacteria. A number of studies have demonstrated a clear link with CRC (56-60). According to the acquisition of factors of virulence, there are four E. coli phylogenic groups (A, B1, B2 and D). Groups A and B1 are generally not pathogenic; however, groups B2 and D are involved in intestinal and extra-intestinal pathogenesis. Crohn's disease, a chronic IBD known to be a risk factor for CRC, is caused by some strains of phylogroup B2 that are associated with it (60,61). Swidsinski et al (58) and Martin et al (59) demonstrated that mucosa-associated E. coli resided in higher numbers in patients with CRC than in the controls, supporting the central role of these bacteria in CRC development. Pathogenic E. coli strains synthesize various virulence factors (62). These factors consist of several toxins known as cyclomodulins, such as cytolethal distending toxins (CDT), cytotoxic necrotizing factor (CNF), cycle inhibiting factor and colibactin. Cyclomodulins are genotoxic and known to alter cell cycle progression, proliferation, cell differentiation and apoptosis (63-68). Cuevas-Ramos et al stated that a specific type of bacterial strain belonging to the B2 phylogenetic group possessed a unique gene island termed 'pks', which translates into the genotoxic compound, colibactin, possessing genotoxic properties causing DNA double-strand breaks and chromosomal instability (CSI) in human cells (68). As regards cyclomodulin-producing E. coli, a previous study revealed an increased prevalence of cyclomodulin-producing B2 E. coli in colon tumor biopsies, suggesting a possible role of such pathogenic E. coli in colon carcinogenesis (30).

This genotoxic compound generates double-strand breaks in DNA causing damage to several portions of DNA, rendering it unstable and this instability of the genome paves the way for the upregulation of expression of oncogenes, such as c-Myc, which results in the formation of adenomas and the development of CRC (56,67).

Recent studies have found that patients with an enrichment of Peptostreptococcus stomatis (P. stomatis) and Peptostreptococcus anaerobius (P. anaerobius) have a higher risk of CRC development (69-71). Patient stool and tissue are enriched with P. anaerobius in CRC. Tsoi et al indicated the pro-tumorigenic role of P. anaerobius modulating TLR2 and 4 leading to reactive oxygen species (ROS) accumulation, resulting in cholesterol biosynthesis and proliferation (70). Purcell et al found the enrichment of P. stomatis in consensus molecular subtype 1 (CMS1) of CRC tumor tissue (71). P. stomatis supports bacteria colonization by promoting acidic and hypoxia around tumor microenvironment (71). Long et al found that P. anaerobius promoted CRC in APC (−/+) mice and altered tumor immunity by the PCWBR2-integrin α₂/β₁ -PI3K-AKT-NFkB signaling pathways, identifying the PCWBR2-integrin α₂/β₁ axis as a potential therapeutic target in CRC (72).

S. gallolyticus has long been associated with CRC. Kumar et al reported significant role of few selected strains of Streptococcus gallolyticus subsp. gallolyticus (SGG) in CRC and upon a detailed analysis of its metabolite profile and mechanisms of pathogenesis, S. gallolyticus was found to enhance the proliferation of CRC cells by targeting Wnt/β-catenin signaling and upregulating β-catenin in cells, thereby elevating the expression of target oncogenes c-Myc and cyclin-D, resulting in the adenocarcinoma of CRC tissue (73). Kwong et al found that CRC diagnosis was associated with S. gallolyticus by perturbing barrier function (69). Another study found that tumorigenic conditions promoted the S. gallolyticus colonization of gut by activation of SGG-specific bacteriocin and replacing commensal enterococci (74). Some strains of S. gallolyticus have been found to induce cell proliferation designated as proliferation-promoting Sg: PP-Sg and others were not found to stimulate cell proliferation and therefore classified as non-proliferation-promoting: NP-Sg. PP-Sg found to higher colonization in mice as compared to NP-Sg, owing to the better interaction of PP-Sg with host epithelial cells. Furthermore, PP-Sg promoted CRC development in an AOM mouse model (73) (Table I).

Apart from these bacterial species mentioned above, there are several other microbes predominantly found in the fecal and mucosal samples of patients with CRC, and play a key role in pathogenesis and carcinogenesis. Microbiota analysis of healthy individuals indicated that several species of LAB namely, Lactobacillus acidophilus, L. casei, L. rhamnosus exert anti-inflammatory effects in maintaining gut homeostasis (75,76). L. rahmnosus was found to reduce the levels of β-catenin and NF-κB p65 proteins, and to induce the expression of tumor suppressor gene p53 and anti-apoptotic factor BAX in colon epithelial cell, thus preventing CRC (76). Bifidobacterium, another LAB, exerts a negative effect on CRC proliferation due to its property to reduce β-glucuronidase activity in the gut which enhances the chemotherapeutic efficacy of CPT-11, hence providing a beneficiary role in the treatment of CRC (77). The detailed role of these commensal LABs in the prevention of CRC has yet to be identified and can be explored in the near future in order to gain better insight into potential treatment strategies for CRC. The abundance of Parvimonas micra has been found in the stool of patients with CRC. P. micra was found to inhibit the NOD2 signaling, giving rise to an inflammatory and pro-tumorigenic microenvironment (78). Xu et al found that P. micra abundance was elevated in patients with CRC and was low in healthy individuals and patients with colorectal adenoma (79). The overabundance of Porphyromonas gingivalis has been found in patients with CRC (80). Yang et al reported an association between Prevotella intermedia with a higher risk of CRC development (81). Another study identified P. intermedia in a multinational cohort of fecal samples of CRC (82). Gemella morbillorum has been found to regulate IL-12 production and thereby, immunoregulation in CRC. Other species of Gemella regulate the protective function of the adaptive immune response at the mucosal surface by cleaving IgA1 (83,84). Still, however, no direct role of this bacteria has been reported in CRC development. A schematic diagram of the mechanisms of action of major species of bacteria involved in the development of CRC in humans is presented in Fig. 1.

Allen and Sears highlighted the impact of the dysbiosis of the gut microbiome on genome and epigenomics of the colon epithelial cells which increased cell proliferation and tumor formation, growth and metastasis (86). DNA methylation patterns and histone marks on several promoters and enhancers were found to be dysregulated, leading to the downregulation of tumor suppressor genes (TSGs) and the upregulation of oncogenes. Several of the miRNAs (oncomiRs and anti-oncomiRs) and long non-coding RNAs have also been found to be dysregulated and associated with CRC (86). B. fragilis has been found to induce the hypermethylation of several TSGs, such as Hoxa5, Polg, Runx1, Runx3, CD37, Stx11, Tceb2, Lgr6, Cdx1 and Fut4, causing carcinoma development in the colon and rectum (86). Xia et al identified F. nucleatum and Hungatella hathewayi to be highly associated with the upregulation of DNA methyltransferases which cause the hypermethylation of promoter of TSG CDX2 and MLH1 (87). In a previous study, the APC gene and DNA mismatch repair (MMR) system were the prime genetic factors identified, which if altered or silenced, led to the development of CRC (88). In an extensive clinical analysis performed by Gagnière et al the MMR pathway was shown to be most affected by entero-pathogenic E. coli. They reported that pks⁺ E. coli strains downregulated the expression of MLH1 genes and inhibited the formation of MLH1 MMR protein in T-84 cells, which led to MSI and tumorigenesis (89). In addition, methylation in APC and INK4a TSG also promoted tumor development in the case of colitis-associated CRC (90-92). Streptococcus species have also been identified to be associated with APC gene hypermethylation (87).

Apart from methylation, dysregulation in the expression of miRNAs serves as a major epigenetic marker responsible for CRC. F. nucleatum is often found to increase RASA1 expression by inhibiting miR-21 expression, thus leading to chronic inflammation in the intestine, initiating carcinogenesis (93). F. nucleatum has also been found to downregulate the expression of miR-4802 and miR-18a, which lead to resistance against chemotherapeutic drugs administered to patients suffering from CRC (94). Several other oncomiRs and anti-oncomiRs were analyzed from fecal samples and gut mucosa, which revealed differential miRNA profiles in the presence of different gut microbiome, and this variation in miRNA profiles can also serve as a fingerprint for the detection and diagnosis of CRC (95,96).

The gut microbiota is known to contribute immensely towards the maintenance of the immune system. Fluctuations in the dynamic equilibrium of this microbiota composition leads to defects in the immune system, resulting in inflammation and in tumor initiation. Genotoxins from B. fragilis and pks⁺ E. coli are known to induce inflammation in colon epithelial cells and this inflammatory response by the host induces genetic and epigenetic alterations, which contribute to CRC development (97). Mutations in the tumor suppressor p53 gene are commonly associated with cancer initiation, as well as progression, and they have been found to prolong the effects of NF-κB signaling, which generates the inflammatory response in cells (98). Inflammation triggers oxidative stress, increasing DNA damage and causing mutations in genes, such as APC, directing the Wnt signaling pathway and KRAS, which initiates adenoma formation and is followed by the loss of chromosome 18q and mutations in TP53, resulting in CSI and in the formation of carcinomas (99,100). Inflammation in cells often regulates the production of chemokines and cytokine-driven signaling pathways, such as the NF-κB, PI3K, Akt and ERK pathways. These signaling pathways are responsible for the initiation of tumorigenesis by either upregulating Wnt signaling, which promotes cell proliferation or by inhibiting apoptosis (101). F. nucleatum has also been found to enhance pro-inflammatory markers and the infiltration of CD11b⁺ myeloid immune cells and few macrophages, activating Th17 cells and TGFβ signaling, which promotes tumor initiation and angiogenesis (102).

This interaction of F. nucleatum with the host immune system suggests its prime role in CRC. F. nucleatum induces cancer development by altering the host immune response in the tumor microenvironment. F. nucleatum expresses various cell surface proteins (FadA, Fap2 and RadD) which can activate inflammatory factors and favor an environment for tumor growth by recruiting inflammatory cells. F. nucleatum can promote the immune suppression of the intestinal mucosa by inhibiting the function of T cells, natural killer (NK) cells and macrophages, resulting in CRC progression (103). Cancer-targeting NK cells are also inhibited by F. nucleatum, due to the binding of Fap2 protein on T-cells (35). It is the combination of CSI and MSI mechanisms that govern the development of adenoma-carcinoma in colorectal tissues. CSI is under the influence of immune factors, which is in a synergy with the MSI pathway characterized by changes in the epigenetics of colon epithelial cells (100); the three commonly found pro-carcinogenic microbial strains, F. nucleatum, B. fragilis and E. coli, have been observed to exert an equal influence on both the pathways driving towards CRC. By contrast, several protective microbial species known to produce butyrate and propionate exert an inhibitory effect on CRC by producing SCFAs and activating cancer-preventing phytochemicals (100). A schematic illustration of the crosstalk between several factors which play a key role in the development of CRC is presented in Fig. 2.

The composition of microbes in the host are a continuous influence on the environment and in order to protect themselves, microbes belonging to specific species tend to form mucosal biofilms. These microbial biofilms exert a protective function for the microbiota from immune factors present in the host (104). The study by Dejea et al demonstrated that biofilms were characteristically found in almost all the patients with right-sided CRC (105). The microbial composition of these biofilms was then further studied and it was predominantly found that microbes belonged to Bacteroides, Fusobacteria, Clostridia, Bifidobacterium and E. coli (104,105). These biofilms in the colon have been found to exert pro-carcinogenic effects by disrupting E-cadherin, and enhancing IL-6, Ki63 and pSTAT3 expression in the colon epithelium (105), more specifically on the right side. The presence of these biofilms often results in a poor prognosis (100). Thus, it can be conveniently concluded that biofilms can serve as a signature biomarker for CRC.