The microbial saboteur: Role of bacteria in cancer chemoresistance

Chemoresistance – the acquired or intrinsic ability of a tumor to evade cytotoxic therapy – has long been attributed to tumor-cell-intrinsic factors: Genomic mutations, epigenetic plasticity, and pro-survival signaling. Yet a mounting body of evidence compels a fundamental reassessment: Bacteria residing within tumor tissues are active and mechanistically sophisticated contributors to treatment failure, operating through direct drug catabolism, epigenetic reprogramming of cell-death pathways, and immunological subversion of the tumor microenvironment. The existence of a resident intratumoral microbiome, distinct from the gut flora and detectable across pancreatic, colorectal, breast, lung, and ovarian cancers, is now firmly established by microbial sequencing, fluorescence in situ hybridization, and culture-based approaches. The most biochemically direct mechanism is enzymatic inactivation of chemotherapy agents by bacterial enzymes expressed within the tumor itself. The landmark demonstration by Geller et al., in Science, showed that Gammaproteobacteria – detectable in 76% of human pancreatic ductal adenocarcinoma (PDAC) specimens – express a long isoform of the bacterial cytidine deaminase enzyme that converts gemcitabine into its inactive metabolite 2′,2′-difluorodeoxyuridine, rendering the cornerstone of PDAC chemotherapy ineffective before it ever reaches the tumor cell. This resistance was abrogated by co-treatment with ciprofloxacin in murine models, directly validating the therapeutic potential of antibiotic co-administration – an approach now being evaluated clinically. Analogous enzymatic biotransformation of fludarabine, cladribine, and the prodrug CB1954 by intratumoral Escherichia coli strains establishes that bacterial drug inactivation is neither species-specific nor confined to a single chemotherapy class. Perhaps no single organism better encapsulates the bacterial threat to chemotherapy than Fusobacterium nucleatum (Fn), an oral anaerobe that ectopically colonizes colorectal, esophageal, breast, and pancreatic tumors and whose elevated intratumoral burden independently predicts poor chemotherapy response and reduced survival. Its resistance mechanisms span multiple programmed cell-death pathways: Fn activates autophagic flux in colorectal cancer (CRC) cells via its surface adhesin FadA – engaging E-cadherin to trigger Wnt/β-catenin signaling – and through lipopolysaccharide-mediated TLR4/MyD88/NF-κB activation that upregulates miR-21, shielding tumor cells from 5-fluorouracil- and oxaliplatin-induced apoptosis. Beyond autophagy, Fn inhibits pyroptosis via the Hippo pathway and suppresses ferroptosis, functioning as a broad-spectrum antagonist of chemotherapy-triggered tumor cell death. A striking cell-nonautonomous mechanism involves Fn-infected cancer cells releasing exosomes laden with specific circular RNAs that transfer resistance to neighboring, uninfected tumor cells by alleviating endoplasmic reticulum stress-dependent apoptosis – demonstrating that bacterial chemoresistance can propagate across the tumor without requiring direct bacterial contact. The broader intratumoral bacterial community compounds these individual species-level effects through several additional mechanisms. Polymicrobial biofilms construct physical barriers that impede drug penetration, while simultaneously generating hypoxic, acidic niches intrinsically refractory to conventional cytotoxics.10 Bacteria actively remodel tumor immune tone: Fn recruits myeloid-derived suppressor cells (MDSCs) and tumor-associated macrophages while suppressing NK cell and cytotoxic T-lymphocyte activity, blunting the immune effectors that facilitate clearance of chemotherapy-damaged cells and underpin immunotherapy efficacy. Lactobacillus iners in cervical cancer confers chemo-radioresistance by reprogramming tumor metabolism toward elevated lactate production, while Peptostreptococcus anaerobius in CRC abrogates anti-PD-1 immunotherapy response by activating intratumoral immunosuppressive MDSCs. From the gut, microbiome-derived metabolites – including indoles that impair innate immune activation within the tumor milieu – extend bacterial influence on drug efficacy far beyond the immediate tumor microenvironment. The translational implications are substantial and increasingly actionable. Antibiotic-chemotherapy co-administration, precision intratumoral antibacterial nanomedicine, and fecal microbiota transplantation from chemotherapyresponder donors each represent therapeutic strategies under active investigation, while intratumoral microbiome profiling is emerging as an independent predictor of treatment response that may refine precision oncology protocols. Yet a necessary caution applies: The intratumoral microbiome is not uniformly deleterious, with certain bacterial populations sensitizing tumors to immunotherapy, and indiscriminate antibiotic strategies risk disrupting both systemic microbiome homeostasis and the very immune competence that underpins therapeutic response. Bacteria, long absent from the canonical models of chemoresistance, are microbial saboteurs operating at the heart of treatment failure. Decoding and strategically disrupting their contribution– through enzymatic drug metabolism, cell-death pathway subversion, and immune remodeling – is no longer a frontier curiosity but a clinical imperative that may prove decisive in overcoming one of oncology’s most enduring challenges.