The FadA protein, a signature virulence factor of F. nucleatum, provides a direct link between a bacterial protein and a core cancer pathway. FadA binds to E-cadherin on colon epithelial cells (15), triggering the internalization of the E-cadherin/β-catenin complex. This releases β-catenin from the membrane, allowing it to translocate to the nucleus and activate the Wnt signaling pathway. This leads to the upregulation of oncogenes such as MYC and cyclin D1 (CCND1), fueling the uncontrolled proliferation characteristic of colorectal cancer (CRC) (16-19).

F. nucleatum also employs Fap2, a protein with dual pro-cancer functions. Firstly, Fap2 acts as an immunomodulator by binding to the TIGIT receptor on natural killer (NK) cells and cytotoxic T cells, delivering an inhibitory signal that shields the tumor from immune destruction (20). Secondly, Fap2 functions as a lectin, binding to Gal-GalNAc sugar moieties overexpressed on cancer cells. This interaction allows F. nucleatum to ‘hitchhike’ on circulating tumor cells, facilitating their adhesion to endothelial cells at distant sites and promoting metastasis (21,22).

H. pylori employ a complementary pair of adhesins. BabA binds to Lewis b (Le^b) antigens on healthy gastric cells, establishing the chronic infection required for delivering the CagA oncoprotein (23,24). As infection induces inflammation, the gastric environment changes, and H. pylori switches to SabA. SabA binds to sialyl-Lewis x (sLe^x) antigens, which are upregulated on inflamed tissue, creating a feedback loop that perpetuates chronic inflammation (25).

Streptococcus gallolyticus utilizes the PilG adhesin to bind collagen types I and IV, which are exposed in the disorganized TME but hidden in healthy tissue. This allows the bacterium to preferentially colonize colorectal tumors (26,27). Similarly, fimbrial adhesins such as FimA [Porphyromonas gingivalis (P. gingivalis) and FimH [adherent-invasive Escherichia coli (E. coli)] bind to host integrins and CEACAM6, respectively. These interactions activate Toll-like receptors (TLRs) or stabilize tumor cell adhesion, driving chronic inflammation and invasion (28-30).

Collectively, these adhesins demonstrate how microbes overcome the first hurdle of carcinogenesis: Physical persistence. However, they do more than simply hold on. By targeting molecules such as E-cadherin (FadA), TIGIT (Fap2) and CEACAM6 (FimH), these adhesins directly engage the ‘Proliferative Signaling’ and ‘Avoiding Immune Destruction’ Hallmarks. Within the ‘Evade-Endure-Colonize’ framework, adhesins serve as primary tools for the ‘Colonize’ phase, allowing microbes to establish a foothold in the tumor niche and physically bridge cancer cells to metastatic sites.

Colibactin, produced by pks+ E. coli, is a potent alkylating agent that creates DNA adducts, leading to double-strand breaks. It leaves a specific ‘mutational signature’ in human CRC genomes, serving as a molecular fingerprint of bacterial activity (32). Similarly, the CDT, found in H. pylori (gastric cancer) and E. coli (CRC), functions as a DNase. It translocates to the nucleus and cleaves chromosomal DNA, triggering cell cycle arrest and genomic instability (33,34).

H. pylori injects the CagA oncoprotein directly into host cells, where it disrupts cell polarity and promotes epithelial-mesenchymal transition (EMT) (35). Concurrently, the secreted VacA toxin disrupts epithelial barrier integrity and suppresses local T-cell function (36,37), while Tipα binds to STAT3, driving inflammation and proliferation (38).

BFT is a metalloprotease that cleaves E-cadherin. This disrupts the intestinal barrier and activates Wnt signaling, while also recruiting T helper 17 (Th17) cells to establish a pro-tumorigenic inflammatory environment (39,40). In non-gastrointestinal cancers, Chlamydia trachomatis secretes CPAF, a protease that degrades pro-apoptotic proteins and cell cycle regulators, promoting survival in cervical cells (41,42). Additionally, CNF1 from E. coli activates Rho GTPases, driving cytoskeletal rearrangement and motility (43,44).

While diverse in mechanism, these toxins converge functionally to enable the ‘Genomic Instability’ and ‘Tumor-Promoting Inflammation’ Hallmarks. Genotoxins such as Colibactin and CDT directly mutagenize the host genome, providing the genetic variation required for tumor evolution (the Endure phase). Meanwhile, modulating toxins such as CagA and BFT dismantle cell-cell junctions and induce EMT. This plasticity is essential for cancer cells to detach from the primary tumor, initiating the Evade phase of metastasis.

Gingipains, cysteine proteases from P. gingivalis, degrade collagen and fibronectin. In oral squamous cell carcinoma, this activity breaks down the basement membrane, paving the way for invasion (28,46). Similarly, hyaluronidases secreted by Staphylococcus and Clostridium species cleave hyaluronic acid. This ‘liquefies’ the ECM in diverse cancer settings - from skin and breast cancer to urogenital tract malignancies, reducing physical resistance to cancer cell migration and facilitating angiogenesis (47,48).

Collagenases from bacteria such as Clostridium histolyticum degrade the dense collagen scaffold of the ECM. In the TME, this activity assists cancer cells in breaching the tumor capsule and entering the vasculature (49). Interestingly, this mechanism is being explored therapeutically to ‘soften’ desmoplastic tumors (such as pancreatic cancer) to improve drug delivery (50).

These enzymes function as the tumor's ‘demolition crew’. By degrading the basement membrane and ECM, they directly enable the ‘activating invasion and metastasis’ Hallmark. In the metastatic cascade, these factors are critical for the transition from the ‘Evade’ phase (local invasion) to the ‘Endure’ phase (intravasation into blood vessels). Without this enzymatic assistance, tumor cells would remain physically confined regardless of their genetic mutations.

The E6 and E7 proteins of high-risk HPV tear down the p53 and Retinoblastoma (Rb) tumor suppressors, respectively. This removes cell cycle checkpoints and prevents apoptosis, driving the uncontrolled proliferation observed in cervical and head-and-neck cancers (53,54). The HTLV-1 Tax protein functions as a transcriptional activator, driving the expression of IL-2 and its receptor to create a malignant autocrine loop in T-cell leukemias (55).

EBV's latent membrane protein 1 (LMP1) mimics a constitutively active CD40 receptor, driving survival signaling via NF-κB and MAPK pathways (56), while EBV Nuclear Antigen 1 ensures viral persistence and immune evasion (57). In liver cancer, HBV X protein and HCV Core protein act as promiscuous regulators, interacting with Wnt/β-catenin and generating reactive oxygen species to promote both proliferation and genomic instability (58-61).

Unlike bacteria that manipulate cells from the exterior, oncoviruses bypass the ‘Evade’ phase and jump directly to hijacking the cell's central command. By dismantling tumor suppressors (p53 and Rb) and mimicking growth signals (vGPCR and LMP1), these viral proteins directly enable the Hallmarks of ‘Enabling Replicative Immortality’ and ‘Sustaining Proliferative Signaling’. This allows the infected cell to bypass natural checkpoints, ensuring the survival and expansion required for the ‘Endure’ phase of malignancy.

LPS from Gram-negative bacteria activates TLR4, driving NF-κB-mediated inflammation. Previous evidence highlights the role of LPS in determining organotropism; circulating LPS can ‘prime’ the lungs for metastasis by upregulating inflammatory adhesion molecules, creating a receptive ‘pre-metastatic niche’ for breast cancer cells (63,64).

Fungal β-glucans: The microbiome is not limited to bacteria; the fungal ‘mycobiome’ is also a key resident of tumors. β-glucans, major structural components of the fungal cell wall, are potent immunomodulators recognized by host receptors such as Dectin-1(65). In pancreatic cancer, fungi such as Malassezia migrate to the pancreas, where their cell wall β-glucans activate the complement cascade. This activation promotes inflammation and has been shown to accelerate tumor progression (66).

Microbial extracellular vesicles (MEVs): An increasing area of research focuses on MEVs - nanosized lipid bilayers released by bacteria. These vesicles act as long-distance delivery vehicles for virulence factors. In lung cancer, MEVs have been shown to enter host cells and modulate signaling pathways that suppress immune surveillance and promote EMT (67). Recent findings indicate that MEVs can alter the lung microenvironment to favor tumor colonization, representing a novel mechanism of host-microbe communication (68).

Secondary bile acids: Gut bacteria metabolize primary bile acids into secondary forms such as deoxycholic acid. In the liver, high levels of hydrophobic bile acids can induce DNA damage and senescence in stellate cells, creating a pro-inflammatory environment that facilitates hepatocellular carcinoma and liver metastasis from CRC (69).

Short-chain fatty acids (SCFAs) and hydrogen sulfide (H₂S): While often protective, SCFAs such as butyrate can be co-opted by cancer cells as an energy source (Warburg effect) (70). Similarly, H₂S produced by F. nucleatum promotes angiogenesis and fuels tumor cell mitochondrial metabolism (71).

While proteins act as targeted weapons, these structural and metabolic factors function as the ‘fertilizer’ for the TME. By maintaining chronic inflammation (LPS and β-glucans) and providing alternative fuel sources (SCFAs and H₂S), they enable the ‘Tumor-Promoting Inflammation’ and ‘Deregulating Cellular Energetics’ hallmarks. Crucially, factors including circulating LPS and MEVs act as long-range signals that prepare distant organs for the ‘Colonize’ phase, establishing the pre-metastatic niche before tumor cells arrive.

As understanding deepens, the future of cancer diagnostics lies in assessing the functional threat posed by the microbiome rather than just its taxonomic composition. The presence of the pks genomic island (encoding Colibactin) or specific alleles of fadA or vacA could serve as powerful prognostic biomarkers (73,74). Detecting the DNA of these virulence factors in a tumor biopsy or ‘liquid biopsy’ could identify patients at high risk for metastasis. For example, quantifying F. nucleatum load via fadA detection in blood shows promise for screening and predicting recurrence in CRC. Furthermore, the composition of the gut microbiome is now a validated predictive biomarker for patient response to immune checkpoint inhibitors (75). The future will involve developing precisely defined microbial signatures to predict treatment responses and determine if microbiome modulation is required before therapy begins.

Targeting the microbial drivers of cancer represents a paradigm shift, offering therapies that complement and enhance traditional oncology. These strategies can be grouped into three main categories:

Disarming the pathogen (anti-virulence therapy): This precise approach aims to neutralize specific virulence factors without inducing broad-spectrum dysbiosis. Strategies include small molecule inhibitors designed to block the active sites of microbial enzymes, such as gingipains or bacterial collagenases, to prevent tissue invasion (76). Additionally, monoclonal antibodies could block critical adhesin-receptor interactions; for instance, blocking the Fap2 adhesin to prevent it from binding TIGIT on NK cells could restore antitumor immunity (77).

Precision microbiome editing: This strategy aims to selectively remove harmful ‘oncomicrobes’ or introduce beneficial ones. Bacteriophage therapy offers a highly specific method to eliminate bacteria such as F. nucleatum while leaving the beneficial microbiota unharmed (78). Furthermore, the development of preventative vaccines against oncogenic agents such as H. pylori or EBV remains a major goal for long-term cancer prevention (79).

Reshaping the ecosystem: This strategy aims to engineer the microbial community to support cancer therapy. This includes next-generation probiotics engineered to produce anti-inflammatory molecules or compete with pathogenic species (80). It also involves metabolic interventions, such as using prebiotics to favor butyrate-producing bacteria or drugs that inhibit the conversion of primary to secondary bile acids (81).

To translate these concepts into clinical reality, research must move from establishing correlations to proving causation. While animal models provide compelling evidence, a major hurdle remains in definitively proving that a specific microbial virulence factor is a driver, not just a ‘passenger’, in human cancer. This will necessitate sophisticated multi-omics analyses of large, longitudinal patient cohorts. Furthermore, recreating the complexity of the TME including hypoxia, immune cell infiltrate, and polymicrobial communities, through advanced co-culture systems and human tumor organoids will be essential for validating these new therapeutic targets.