Fermentation of dietary fibers and indigestible carbohydrates by gut microbiota leads to the production of short-chain fatty acids (SCFAs). The main SCFAs are butyrate, acetate, and propionate. An important target cell is the regulatory T cell (Treg), a cell characterized by expression of CD25 and the transcription factor forkhead box P3 (FOXP3). Tregs maintain and restore the balance between Th1, Th2, Th17, and Tfh cells [12]. It has been extensively demonstrated that the induction of Tregs is mediated by butyrate, produced by gut microbiota. Butyrate enhances histone H3 acetylation in the promoter region of the (FOXP3) locus [13]. FMT leads to an increase in CD25⁺FOXP3⁺ Treg and concomitantly reduces the number and activity of Th17 cells and related cytokines IL-6 and IL-17 [14].

The induction of Tregs and repression of Th17 cells is the preferred type of immunomodulation in the case of inflammatory diseases as well as allergic diseases [15]. In the case of tumor immunology, activation of Tregs would be undesired [16]. Tumor tissue consists of the tumor cells but also an array of non-tumor cells, including invading cells of the innate and acquired immune system. Collectively, this is defined as the tumor microenvironment (TME). Within the TME, activation of Tregs can best be avoided. The TME contains different subsets of Tregs, including thymic-derived Tregs (tTregs), peripheral induced Tregs (pTregs), tissue-resident Tregs (tr-Tregs), and follicular Tregs (Tfr), but they all have immunosuppressive effects, therefore promoting tumor immune evasion. Indeed, it was found that patients with multiple myeloma or metastatic prostate carcinoma treated with anti-CTLA-4 responded significantly poorer to therapy than patients with high serum butyrate levels and higher numbers of blood Tregs [17].

Butyrate, apart from inducing Tregs, also stimulates the activation of cytotoxic T cells (CTLs), leading to increased production and release of γ-interferon and granzyme, thus contributing to tumor eradication (Figure 1A) [18, 19]. Acetate is also active in this respect [20]. In addition to their effect on Tregs and CTLs, SCFAs impact natural killer (NK) cells and macrophages within the TME [21]. On the other hand, butyrate and propionate promote an anti-inflammatory environment, which may, under certain conditions, favor tumor progression [22]. These different and sometimes opposing effects of SCFAs underscore the complexity of immune modulation in the TME by the gut microbiota.

Two different bacterial metabolites of tryptophan (Trp) have profound immune-modulating effects relevant for tumor eradication. The first one is indole-3-aldehyde (I3A), generated by Lactobacillus reuteri (L. reuteri) (Figure 1B), which activates CTL, leading to increased γ-interferon production. In preclinical melanoma models, L. reuteri produced I3A improves the efficacy of ICI therapy [24]. Suggestive of the clinical relevance is the observation that the overall survival of melanoma patients on ICI therapy is better in patients with high serum levels of I3A [24].

Other gut bacterial Trp metabolites like indole-3-propionic acid (IPA), produced in tandem by Lactobacillus johnsonii (L. johnsonii) and Clostridium sporogenes (C. sporogenes) (Figure 1B), have demonstrated the ability to alter the epigenetic states of immune cells through mechanisms such as H3K27 acetylation [21, 25, 26]. In this context, the effect of IPA on so-called precursor exhausted T cells (Tpex) is relevant. Within TME, CD8⁺ CTLs are crucial mediators of antitumor immunity. However, after prolonged and persistent T cell receptor stimulation, they may differentiate into a dysfunctional state, termed exhausted T cells (Tex). However, Tpex cells can be rejuvenated to effector CTLs [27]. IPA can activate the transcription factor T cell factor 1 (TCF-1) by H3K27 acetylation at the super enhancer region of the Tcf7 gene. TCF-1 induces Tpex activation and thus potentially enhances the effectiveness of ICI (Figure 1B). In an animal model, IPA is unable to enhance the efficacy of ICI therapy in Tcf7^-/- mice [26].

Gut bacteria have other direct or indirect effects (via metabolites) on the tumor cells and/or the TME, rendering tumors more susceptible to ICI therapy. Polyphenol-derived microbial metabolites, including hydroxyphenyl-γ-valerolactones (PVLs), can suppress the Wnt/β-catenin signaling pathway, a pathway commonly dysregulated in colorectal cancer (CRC) [22]. By altering cancer-associated fibroblasts (CAFs) and reducing tumor proliferation, polyphenol metabolites show promise as therapeutic agents that modulate the metabolic profile of the TME [28]. Gut bacteria can alter the metabolism of chemotherapeutic drugs such as irinotecan, affecting their efficacy and toxicity [29–31]. Organoid studies have revealed that microbiome manipulation induces phenotypic changes in tumor cells. These include altered expression of drug efflux pumps and anti-apoptotic pathways, which may lead to chemoresistance or chemosensitivity [32, 33]. Such findings emphasize the need for microbiome-targeted strategies to improve treatment outcomes. While L. johnsonii and similar microbes enhance immune therapies, identifying additional beneficial strains remains crucial [26] along with a better understanding of the mechanisms through which microbial metabolites enhance CTL activation, so that these attributes can be harnessed to improve immunotherapy responses across various cancer types.

Detailed investigations into specific microbial interactions with immune checkpoint molecules such as PD-1/PD-L1 could uncover novel ways to augment ICI therapies. Routy et al. [7] found that the effectiveness of ICI therapy, such as those targeting the PD-1/PD-L1 pathway, is influenced by the gut microbiome composition. For instance, the abundance of Clostridioides was found to be higher in melanoma patients who responded to ICIs, aligning with similar observations in the gut microbiome of responders. Conversely, Gardnerella vaginalis was more abundant in NRs [34]. Furthermore, the presence of certain bacteria, such as Saccharopolyspora, Pseudoxanthomonas, and Streptomyces, in pancreatic adenocarcinoma tissues has been linked to enhanced recruitment and activation of CD8⁺ T cells, a crucial component of the adaptive immune response against tumors [25, 35, 36]. This further supports the idea that specific bacterial taxa within the TME might modulate the efficacy of ICIs.

Originally, it was hypothesized that screening of the gut microbiota of patients who have responded favorably to ICI therapy could lead to the identification of universal FMT donors. Developing personalized FMT protocols tailored to individual microbiome profiles could optimize inflammation reduction, immune activation, and drug efficacy [37]. FMT, utilizing feces from a healthy donor, has demonstrated high success rates in treating recurrent Clostridioides difficile infection, with cure rates reaching up to 90%. This success stems from Clostridial infections being more responsive to microbial restoration via FMT. However, the efficacy of FMT in other (chronic) diseases such as inflammatory bowel diseases (IBD; ulcerative colitis and Crohn’s disease), irritable bowel syndrome, and metabolic disease has been less consistent, with clinical remission rates ranging from 24% to 50%. This variability has been attributed to IBD’s complex nature as a microbiome-driven immunological disorder influenced by host genetics and environmental factors [38, 39].

Dietary interventions could be an additional venue to optimize the effects of FMT. Dietary interventions aimed at promoting the production of beneficial microbial metabolites, such as polyphenols that modulate Wnt/β-catenin signaling, hold significant potential for CRC and other cancers [40]. Long-term dietary changes or supplementation with fibers and polyphenols could reshape gut microbiota composition, supporting anti-cancer metabolite production. Clinical trials assessing these interventions in combination with standard therapies are a critical next step. The first studies on the potential improvement of FMT success rate by dietary intervention showed mixed results. In patients with metabolic syndrome, the Mediterranean diet, neither for the donor nor the recipient, had no effect on the outcome of the FMT [41]. On the other hand, in ulcerative colitis patients, an anti-inflammatory diet did improve the outcome of FMT intervention on induction and maintenance of remission of the disease [42]. In an animal model of colitis-associated cancer (CAC), a high-fat diet was associated with reduced butyrate and accelerated CAC progression [43]. A number of studies on the effects of dietary intervention, either as a single intervention or combined with FMT, on the outcome of ICI therapy for cancer are ongoing [44].

Bacteria can influence tumors not only through local interactions but also through systemic effects that are mediated by microbial metabolites. As discussed above, SCFAs and other bacterial metabolites that are produced in the gut can circulate throughout the body and modulate immune responses in distal tumors [53]. Additionally, the gut microbiome plays an important role in regulating systemic immune activation, which could impact the efficacy of ICIs in cancers outside of the gastrointestinal tract [54, 55].

Recent findings indicate that bacteria may also systemically influence metastatic potential by altering tumor cell metabolism. Bacillus thermoamylovorans was found to promote metastatic disease through the activation of enhanced glycolytic and nucleotide synthesis pathways, both of which have been associated with increased tumor aggressiveness [56]. In contrast, a related species, B. subterraneus, did not induce these changes. This highlights how functional bacterial genes, rather than bacterial presence alone, may drive disease progression. These findings suggest that systemic bacterial metabolites could selectively fuel metastatic dissemination and shape tumor evolution beyond the primary site.

One of the key mechanisms through which bacteria within tumors exert local effects on the TME is immune modulation. Bacterial ligands can alter the behavior of tumor-associated macrophages (TAMs) to become more pro-tumorigenic, supporting the growth of tumors and suppressing immune responses [34, 57]. This is achieved through the activation of cell signaling pathways like NF-κB by the bacteria, which then activate the transcription of genes that drive inflammation and aid in tumor development [46]. However, not all intratumor bacteria promote cancer progression (Table 1). The oral administration of S. mitis isolated from gastric cancer tumors has been found to inhibit tumor growth, reducing tumor volume and extending survival in tumor-bearing mice [58]. This effect was mediated through the suppression of M2 macrophage polarization and infiltration, which reduced pro-tumor immune signaling.

Beyond the modulation of the immune system, certain bacteria actively modify important components of the TME to promote anti-tumor immune responses. Salmonella species have been demonstrated to exert varying effects on immune cells, tumor cells, and stromal components that promote tumor regression [66]. Salmonella can directly kill tumor cells by apoptosis, autophagy, pyroptosis, and ferroptosis. Through immunomodulation, Salmonella aids in recruiting and activating CTLs and NK cells and in reducing immunosuppressive Tregs and myeloid-derived suppressor cells. Additionally, the interaction of Salmonella with tumor stroma can lead to the remodeling of the ECM. This potentially impedes tumor invasion and metastasis. These findings suggest that besides playing a role in the modification of the immune landscape, Salmonella is also able to create a hostile environment that promotes tumor progression through the disruption of key structural and metabolic components of the TME.

Apart from Salmonella, other bacterial species also affect the structure and metabolism of the TME, facilitating tumor progression. Prevotella bacteria can change the composition of the ECM in such a way that it better facilitates tumor invasion and metastasis [52]. Other bacteria can also secrete angiogenic factors that alter the behavior of endothelial cells, which leads to the formation of new blood vessels that help sustain the growth of tumors [53, 57, 60]. These bacterial species include A. muciniphila, B. fragilis, E. faecalis, F. nucleatum, and Lactobacillus. rhamnosus GG (LGG) [59, 60, 63–65]. The angiogenic effects of A. muciniphilia, B. fragilis, and LGG are of interest because these species are either widely used probiotic bacteria (LGG) or otherwise associated with beneficial effects.

Certain bacteria have also been shown to directly interact with and stimulate oncogenic pathways. Certain strains of Escherichia coli, carrying the polyketide synthase (pks) operon, produce colibactin. This metabolite causes DNA damage and mutagenesis, thus contributing to the progression of CRC [55, 67]. In contrast, a direct reduction in tumor cell proliferation has been observed in mice treated with S. mitis, as indicated by a lowered expression of proliferation markers such as PCNA and Ki-67 [58]. Changes in the microbial composition within mouse tumors were observed after oral administration of this bacterium. This suggests that S. mitis can translocate from the GI tract to the site of the tumor, influence the structure of the TME, and impact tumor progression.

A recent study was able to identify distinct microbiomes and metabolomes in different stages of colorectal tissue transformation. This included bacterial families like Lachnospiraceae being depleted in adenomatous polyp environments and F. nucleatum subsp. animalis being more abundant in cancerous tissues [45]. These bacteria have been associated with the accumulation of beta-hydroxybutyric acid and fumarate within the TMEs, which may also contribute to the modification of the TME and colorectal tumor progression.

The presence of tumor-associated bacteria significantly influences the efficacy of ICIs. In cases of esophageal squamous cell carcinoma (ESCC), enrichment of Streptococcus was associated with an increase in CD8⁺ T cell infiltration and a more favorable response to anti-PD-1 treatment [68]. Similarly, supplementation with L. johnsonii has been correlated with the upregulation of progenitor exhausted CD8⁺ T cells and enhanced effectiveness of anti-PD-1 immunotherapy in melanoma, breast cancer, and CRC [26]. In melanoma and lung cancer, a more diverse gut microbiome has been linked to better responses to ICI therapy [54]. However, some bacteria have also been found to reduce the efficacy of ICIs. F. nucleatum, a bacterial species commonly found in CRC and NSCLC, is associated with reduced survival and poor response to ICI therapy. This is potentially due to its role in promoting a more immunosuppressive TME through its inhibition of CTL activity [52]. Irrespective of ICI, Gammaproteobacteria and also F. nucleatum are able to breakdown gemcitabine, a common chemotherapy for pancreatic duct adenocarcinoma [29]. In those cases, antibiotic treatment prior to chemotherapy can improve the outcome and survival, but only from a median of 7 months to 14 months [69].

The growing insights into the roles of bacteria in the TME have paved the way for the development of more innovative combination therapies. Combining the targeting of the microbiota with ICIs could potentially optimize immune responses. Bacteria identified as immunosuppressive, such as Fusobacterium, could be eliminated with the use of antibiotics, while the abundance of bacteria identified as immunogenic, such as Streptococcus and L. johnsonii, could be increased by FMT from selected donors, by probiotics, or genetically engineered bacteria [34, 70]. It has been suggested that bacteria invading tumors at distal from the intestines are likely transported through systemic circulation via blood or lymph, or by immune cell transport. However, a therapeutic approach targeting these routes raises concerns about bacteria gaining access to other (healthy) tissues, which could lead to unintended side effects. A potential solution to this latter effect is engineering targeted delivery mechanisms that are capable of minimizing systemic side effects while optimizing the therapeutic effects of the tumor-associated bacteria.

Certain bacteria colonize hypoxic regions of the TME. Clostridium novyi (C. novyi)-NT targets hypoxic tumor cores, releasing toxins that selectively kill tumor cells. This forms the basis of bacteriolytic therapy, a promising experimental approach [29, 71]. Combining bacteriolytic therapies, such as those using C. novyi-NT, with chemotherapy, radiotherapy, or ICI therapy presents a promising avenue for treatment. Through genetic modification, other bacterial species could be created to preferentially target and colonize tumor tissues based on hypoxia.

Beyond influencing immune responses through microbiome composition, certain bacterial species may enhance the efficacy of ICIs by actively modifying the TME. Salmonella-based cancer therapies have been explored for their potential to synergize with ICIs [66]. By inducing pro-inflammatory signaling within the tumor, Salmonella enhances the infiltration and activation of cytotoxic immune cells, thereby increasing tumor sensitivity to checkpoint blockade. These effects suggest that Salmonella could serve as an adjuvant to ICI therapy, improving patient response rates while potentially overcoming resistance mechanisms. The ability of Fusobacetrium nucelatum to access human tumors has been demonstrated. This characteristic is also explored to be harnessed to use in cancer-targeting therapies [72]. One challenge is enhancing the specificity of the therapy to ensure that engineered bacteria target tumor tissue without affecting healthy organs.

The bacterial and metabolite composition of the CRC TME presents an opportunity to refine immunotherapy approaches. While a healthy gut microbiome and metabolome can help prevent malignancies, emerging research suggests that bacterial and metabolite signatures in precancerous adenomatous polyps may serve as valuable diagnostic markers for early CRC detection [45]. Additionally, leveraging TME composition to predict cancer progression and therapeutic response could help in the development of novel treatments and improve patient prognosis. Tumor-associated bacteria could be used as predictive biomarkers for therapy response. For example, a favorable response to chemoimmunotherapy could be predicted from the presence of Streptococcus in ESCC and L. johnsonii in melanoma, breast cancer, and CRC, while a resistance to ICIs could be predicted from high levels of Fusobacterium in colorectal and lung cancers [26, 52, 68]. The incorporation of microbiota profiling through tumor biopsy into the pre-treatment evaluation process for cancer patients could allow for personalized cancer treatment and improve the identification of patients more likely to benefit from ICIs. Depending on the tumor-associated bacterial species present, additional measures could also be taken to inhibit or enhance the patient’s microbiota.