Over the past two decades, research has demonstrated a potential role of the gut microbiome in the development and progression of colorectal cancer (CRC) (1). Among the microbial taxa implicated in CRC, the genus Fusobacterium has emerged as a key player (2). Fusobacterium is a Gram-negative, obligate anaerobic bacterium commonly located in the oral cavity. For over a decade, it has been studied for its role as an onco-microbe. Already in 2012, Kostic et al. reported increased levels of F. nucleatum in CRC tumors (3). Ever since, there have been accumulating studies linking Fusobacterium to the development and progression of CRC (4).

F. nucleatum was initially classified as a single species comprising several subspecies: F. nucleatum ssp. nucleatum, F. nucleatum ssp. animalis, F. nucleatum ssp. vincentii, and F. nucleatum ssp. polymorphum (5). However, advancements in genomic studies have shown significant divergence across these subspecies, leading to the elevation of F. animalis and F. vincentii to species designations (5, 6). F. animalis has emerged as the predominant species detected in CRC tumors (7–9). In contrast, Fusobacterium species associated with other diseases, such as periodontitis, brain abscesses, and adverse pregnancy outcomes, are primarily F. nucleatum (2). This distinction indicates that F. animalis is specifically adapted to the CRC tumor environment. Other species, such as F. nucleatum, F. necrophorum, and F. periodonticum, are also reliably found in lesser proportions among CRC patients (7–9). As a result, a virulent genus-level Fusobacterium CRC complex, likely including multiple Fusobacterium species, has been proposed, redirecting focus from rigid taxonomic classification (10, 11).

Fusobacterium species have several putative virulence factors, including adhesion proteins (FadA), the arginine-inhibitable adhesin RadD, and the galactose-inhibitable adhesion protein 2 (Fap2), which may contribute to CRC progression at different stages (12–14). In recent years, there has been increasing focus on the Fap2 protein and its role in colorectal carcinogenesis, and F. nucleatum Fap2 has been identified as a key factor in the invasion of cultured cancer cells and the subsequent secretion of pro-inflammatory cytokines (15). Fap2 is a type 5 secretion system (T5SS) autotransporter that mediates Fusobacterium to interact with its environment, particularly host cells. Fap2 has a hook-like structure that binds to the lectin N-acetylgalactosamine (Gal-GalNac) that is highly expressed on the surface of cancer cells and to the T cell immunoreceptor with Ig and ITIM domains (TIGIT) present on the T cells and natural killer cells (13, 16). Additionally, Fap2, acting as an autotransporter, facilitates the export of FadA to the bacterial surface, where it contributes to adhesion and invasion of host cells (17). Yeoh et al. identified 754 putative fap2 homologs in 288 Fusobacterium genomes and demonstrated that fap2 homologs are widely dispersed in Fusobacterium genomes, although with significant sequence diversity (11). It is not yet known whether the diversity in the fap2 sequence results in different Fap2 binding abilities to cancer cells or if Fap2 is more frequently present in F. animalis compared to other Fusobacterium species. Some studies have proposed that clinical strains of Fusobacterium naturally lacking fap2 show reduced binding to Gal-GalNac-expressing CRC cells (13). However, expression studies of the fap2 gene are limited, and it remains unclear whether fap2 expression is increased in the tumor microenvironment.

Studies link Fusobacterium to oncogenic mutations, epigenetic changes, and microsatellite instability (MSI-H) in CRC (12, 18, 19). However, it remains unclear whether Fusobacterium drives these changes or simply thrives in the evolving tumor microenvironment. Additionally, Fusobacterium has been demonstrated to enhance a pro-inflammatory tumor microenvironment (TME) through its interactions with immune cells and the induction of inflammatory cytokine secretion. This response may facilitate tumor growth, invasion, and immune evasion (18, 20, 21). An increasing number of studies have shown that the binding and invasion of cancer cells by Fusobacterium species stimulate the secretion of pro-inflammatory cytokines such as tumor necrosis factor (TNF), interleukin-6 (IL6), interleukin-1 beta (IL1β), and C-X-C motif chemokine ligand 8 (CXCL8 or IL8). Casasanta et al. reported an Fap2-dependent invasion of F. nucleatum in an HCT116 cell line, resulting in the secretion of CXCL8 and C-X-C motif chemokine ligand 1 (CXCL1), thereby enhancing the metastatic potential of the cancer cells (15). Previous analyses of the present cohort revealed significantly elevated immune gene expression (CXCL1, IL1B, IL6, CXCL8) in most CRC tumors (22). Cell culture studies conducted by Udayasuran et al. showed that CXCL8 and CXCL1 gene expression is induced under hypoxic conditions independently of Fusobacterium infection, with hypoxia promoting Fusobacterium invasion and exacerbating the hypoxia-induced effects (21). However, the in vivo correlation between increased CXCL8/CXCL1 expression and Fusobacterium presence remains insufficiently explored. Moreover, the majority of in vitro studies have utilized single strains of Fusobacterium, predominantly F. nucleatum, thereby leaving a gap in our understanding of the impact of different Fusobacterium species, in particular F. animalis.

In addition to its role in promoting metastasis, CXCL8 has been shown to induce an immunosuppressive TME by promoting M2 macrophages and inhibiting CD8+ T cell infiltration (23). Tumor-associated macrophages (TAMs) can polarize into M1 (pro-inflammatory) or M2 (anti-inflammatory) phenotypes, influencing tumor progression, metastasis, and prognosis (24). In both colorectal and oral cancer, Fusobacterium has been shown to impair TAMs function, which plays a significant role in tumor progression (25). Studies conducted in vitro and in situ demonstrated that F. nucleatum drives M2 macrophage polarization via the IL6/Signal Transducer and Activator of Transcription 3 (STAT3)/c-Myc signaling pathway and the Toll-like receptor 4 (TLR4)/Nuclear factor kappa B (NF-κB)/S100 calcium-binding protein A9 (S100A9) cascade (26, 27). M2 macrophages are characterized by the production of anti-inflammatory and pro-tumorigenic factors such as interleukin-10 (IL10), transforming growth factor beta (TGFβ), and arginase-1 (ARG1). These factors suppress anti-tumor immune responses, degrade the extracellular matrix, and promote cancer cell migration and invasion, creating a supportive TME (27). Additionally, the M2 polarization helps tumors evade immune surveillance and drives CRC metastasis (24). However, it is important to emphasize that in vitro studies may not fully elucidate the complex interactions among various cell types within the colorectal cancer tumor environment.

We have previously characterized the immune and bacterial profiles in colorectal tumors from a small cohort of Norwegian patients. Our results indicated a profound inflammatory response in the CRC tumors and defined a distinctive immunological signature (22). Furthermore, our microbiome analysis demonstrated a high prevalence of Fusobacterium in these tumor biopsies and identified the Fusobacteria to the species level (8, 28, 29). The integration of these data may provide valuable insights into the association of Fusobacterium and immunological gene expression in CRC. The aim of this study was to examine the link between immune gene expression in CRC tumors, Fusobacterium species, their virulence factor Fap2, and M2 macrophage polarization.

This study aimed to investigate Fusobacterium-associated molecular and immunological changes in colorectal tumor biopsies from Norwegian patients, building on our earlier findings of a significantly elevated prevalence of Fusobacterium within this cohort (8, 28). In this study we examined the relationship between Fusobacterium, fap2, and immune gene expression in CRC tumors. Additionally, we investigated the association of Fusobacterium with oncogenic mutations, microsatellite instability, and macrophage polarization.

Our study demonstrated a significant association between Fusobacterium and microsatellite instability-high (MSI-H) tumors. Except for one, all samples harbouring BRAF V600E mutations are in addition MSI-H status ( Table 2 ). These findings are consistent with the previous research linking Fusobacterium with microsatellite instability (18, 20, 34). This association may have significant implications for CRC prognosis, as MSI-high tumors are associated with specific therapeutic responses, particularly to immune checkpoint inhibitors (35). Patients with BRAF-mutated CRC have a poor prognosis because their tumors are aggressive and resistant to current therapies (36). MSI-H BRAF-mutant tumors react better to immune checkpoint inhibitors such as anti-PD-1 therapy, which improves survival in this group (37). However, emerging evidence suggests that Fusobacterium infection in MSI-H cancers may impair immunotherapy. F. nucleatum has been shown to suppress T cell infiltration and produce an immunosuppressive tumor microenvironment, perhaps aiding immune evasion (38). Although causality remains unproven, our findings support the hypothesis that Fusobacterium influences CRC progression via oncogenic pathways (12, 13, 15, 18, 27).

The association analyses between Fusobacterium and immune gene expression identified 25 genes present in the nCounter Immunology V2 panel that were associated with high levels of Fusobacterium ( Table 2 ). Of the 25 genes, CXCL8 (IL8) stood out with the highest fold change and an exceptionally high count in the Fusobacterium high group, suggesting that the presence of Fusobacterium in tumor tissue may stimulate CXCL8 production. Subsequent immunohistochemical analyses demonstrated that CXCL8 was mainly produced by the tumor cells. This is in line with in vitro studies performed by Casasanta et al., who showed that a human CRC tumor cell line produces CXCL8 when incubated with a laboratory strain of F. nucleatum (15). Others have shown in tumor cell lines and animal models that CRC tumor cells express one of the receptors for CXCL8, C-X-C motif chemokine receptor 2 (CXCR2), and that stimulation with CXCL8 induces tumor cell growth and metastasis (39, 40). In human CRC studies, an association between CXCL8 levels and prognosis has been suggested (40–44). Taken together, our study points toward the possibility that Fusobacterium may promote tumor growth by inducing CXCL8 production in human CRC cancer cells. Several other chemokines were associated with Fusobacterium, including the CXCL chemokines C-X-C motif chemokine ligand 10 (CXCL10) and C-X-C motif chemokine ligand 11 (CXCL11) and the (C-C motif) ligand (CCL) chemokines CCL3 and CCL4. CXCL10 and CXCL11 may be primarily produced by myeloid cells in the tumor microenvironment (18, 27, 45–47), and several reports suggest that CXCL10 and CXCL11 have anti-tumor effects in CRC, possibly by recruiting cytotoxic lymphocytes to the tumor (48–51). Interestingly, we observed that Fusobacterium load was also associated with MSI-H status and elevated levels of IFNγ ( Table 2 , Figure 5 ), both associated with a cytotoxic anti-tumor response. On the other hand, CXCL10 and CXCL11 have been described to enhance cancer development, possibly by promoting angiogenesis and metastasis (51, 52). Like CXCL10 and CXCL11, CCL3 and CCL4 have been attributed both tumor-promoting and anti-tumor effects (53). While this study suggests an association between Fusobacterium and these chemokines, the causality between Fusobacterium and these chemokines remains to be investigated.

The expression of SPP1 and IL6 was associated with Fusobacterium load, both displaying a high fold change in the DEG analysis. Both molecules are associated with colorectal cancer progression, where SPP1 might promote cancer cell migration and invasion (54) and where IL6 is a pro-inflammatory cytokine that may promote cancer cell proliferation and survival (55). The survival effects of IL6 were shown to be mainly mediated by the activation of the transcription factor STAT3 in a mouse model of colitis-associated cancer (56). Interestingly, we identified high expression of the STAT3 inhibitor, suppressor of cytokine signaling 3 (SOCS3), in CRC samples with Fusobacterium, suggesting that STAT3 activation is inhibited in these samples ( Table 3 ). However, further investigations are needed to dissect a possible connection between Fusobacterium load and IL6-mediated cancer progression.

Most of the genes that were elevated in the Fusobacterium high group in the present study also showed increased expression in tumor samples compared to non-neoplastic tissue in our previous study, regardless of Fusobacterium levels (22). This indicates that Fusobacterium may not trigger a unique immune response beyond what is already initiated in the tumors, but it could enhance the existing response. This is in line with Udayasuryan et al., who recently showed in an in vitro study that CXCL8 and CXCL1 are induced under hypoxic conditions independently of Fusobacterium infection, with hypoxia promoting Fusobacterium invasion and exacerbating the hypoxia-induced effects (21).

F. animalis has emerged as the most prevalent Fusobacterium species in colorectal tumors (8, 9). However, our research, along with that of others, has identified multiple Fusobacterium species in tumor biopsies, albeit with lower prevalence ( Table 2 ). In the present study, we observed four samples in the Fusobacterium high group that contained other Fusobacterium species. As illustrated in Figure 2 , both F. animalis and other Fusobacterium species contributed to the correlation between Fusobacterium relative abundance and immune gene expression. This supports that several Fusobacterium species may contribute to CRC, yet the predominance of F. animalis in tumors suggests an underlying cause that warrants further investigation. Schmidt et al. showed that F. animalis, followed by F. nucleatum, are two of the oral species with the highest oral-fecal transmission rates, particularly in colorectal cancer patients (57). The increased transmission of F. animalis could possibly be related to its colonization of tumors, however; it does not explain its predominance relative to F. nucleatum.

The importance of Fap2 is therefore an area of interest. A better overview of Fap2 in the tumor environment may be useful for the understanding of the role of different Fusobacterium strains associated with CRC. Sivertsen et al. showed that not all F. animalis genomes contain the prototypic fap2 gene, and others have suggested that Fusobacterium strains without Fap2 may not bind to tumor cells (13, 58). We therefore tested all biopsy samples for the presence of fap2 and identified fap2 DNA in the majority of biopsy samples from cancer patients that contained F. animalis, F. nucleatum, or F. vincentii. Its presence in these samples may indicate that Fusobacterium strains that colonize tumor lesions express Fap2. Since the fap2 PCR assays did not detect fap2 from other Fusobacterium species, the presence of fap2 in these samples could not be established. We also measured fap2 RNA and used the ΔCt method to evaluate differences in fap2 gene expression between the different colonic sites. We found no significant differences between tumor samples and adjacent non-neoplastic tissues. Our findings therefore align with studies showing that fap2 is constitutively expressed. Cochrane et al. observed no changes in fap2 expression during the initial phase of infection in an in vitro cell model (59). Similarly, Ponath et al. created global RNA maps for five clinically relevant Fusobacterium species and found no variation in fap2 expression across different in vitro conditions, concluding that fap2 expression was growth independent (60). Thus, while fap2 appears essential for infection, it does not seem to function as an inducible virulence factor regulated by quorum sensing or other environmental signals. The constitutive expression may explain its ability to colonize various body regions, including extensive portions of the colon, as shown in the present study. Although fap2 gene expression did not correlate with activated immune genes, the total abundance of fap2 and nusG RNA in each sample correlated with the expression of CXCL8, IL6, SPP1, and IDO1, suggesting that the load of active Fusobacterium cells is related to, and possibly influences, immune gene expression. However, the constitutive nature of fap2 expression argues against its regulation by the inflammatory microenvironment. Taken together the bacterium’s association with pro-inflammatory gene expression in colorectal cancer and supports the notion that Fusobacterium has a role in the stimulation of the observed immune response (26, 27).

Additionally, we investigated Fusobacterium-associated polarization of TAMs in colorectal cancer tumors. TAMs play an important role in tumor progression and metastasis. Macrophages are highly versatile immune cells that exhibit significant heterogeneity and plasticity, adapting to various signals within the TME (61). In CRC, macrophages can adopt different phenotypes, primarily classified as M1 and M2 macrophages, each playing distinct roles in tumor progression. M1 macrophages produce pro-inflammatory cytokines such as INF-γ and TNF-α, which can inhibit tumor growth and promote tumor immunity. On the other hand, M2 macrophages secrete anti-inflammatory cytokines such as IL-10 and TGF-β that promote tissue repair, angiogenesis, and tumor progression (24). The impact of Fusobacterium infection on M2 polarization of macrophages has been investigated in vitro, as well as in Apc ^Min/⁺ mice and in CRC tumor biopsies (26, 27). Chen et al. used immunofluorescent labeled antibodies against M1 and M2 macrophages (anti-CD86 and anti-CD206, respectively) and showed significantly higher amounts of M2 macrophages in F. nucleatum-positive tumor biopsies (27). However, our results are contradictory, showing that the increased amounts of M2 macrophages are not associated with the abundance of Fusobacterium in the tumor biopsies ( Figures 5G1–G3 ). We further examined the tumor samples for expression of known M1 and M2 macrophage markers using quantitative RT-PCR assays, and the results showed no significant differences between Fusobacterium-high and Fusobacterium-low tumor samples ( Figures 5A–F ). Using cultured macrophages, Chen et al. showed that infection with F. nucleatum promoted M2 polarization of macrophages, and they proposed that the polarization is mediated by the IL6/STAT3/C-MYC signaling pathway via overexpression of TLR4, STAT3, and c-myc genes (27). Their results also indicated a significant increase in expression of IL6 and STAT3 in tumor biopsies. Our findings showed a significant increase in IL6 gene expression, which correlated with the Fusobacterium load in the tumor biopsies ( Table 3 ). In contrast, there was no notable increase in STAT3 gene expression associated with Fusobacterium. Interestingly, we identified high expression of SOCS3 in Fusobacterium-high samples, suggesting that JAK/STAT signaling is inhibited in this group. Furthermore, we examined expression of the c-myc and cyclin D1 genes in the tumor tissues. The expression of these genes in CRC tumors is mediated by the IL6/STAT3 signaling pathway (62). We examined the expression of these two genes as indicators for activation of IL6/STAT3 signaling, and our results showed no correlation between Fusobacterium abundance and higher expression of c-myc and cyclin D1 genes ( Figure 6 ). Putting it together, our findings did not support Fusobacterium-associated M2 macrophage polarization in vivo, nor did they indicate that Fusobacterium promotes IL6/STAT3-mediated c-myc gene expression in tumor biopsies.

Another study proposed an S100A9-mediated mechanism for M2 polarization of macrophages (26). S100A9 is a pro-inflammatory mediator that has been shown to regulate myeloid-derived suppressor cells (MDSCs), which are a major component of the immunosuppressive TME (63). Our results showed that the expression of the S100A9, as well as S100A8, mRNAs were induced in Fusobacterium-high tumor samples ( Supplementary Table 3 ). However, we did not find any correlation between expression of the S100A9 gene and infiltration of M2 macrophages in the tumor biopsies in our cohort.

In summary, our results showed no indication for Fusobacterium-mediated M2 polarization of macrophages in the CRC tumor samples but, in contrast, a pro-inflammatory TME driven by inflammatory cytokines such as CXCL8, IL-6, and TNF-α. This discrepancy may reflect differences in sample type (in vitro vs. in vivo), tumor stage, or local microenvironmental signals. It should be noted that our study cohort mainly consisted of tumor biopsies taken under colonoscopy examination and represents tumors in their early stages, mostly Dukes stages A and B, while other studies showing M2 polarization of macrophages was conducted on tumor biopsies in later Dukes stages ( Table 1 ) (26). Our results revealed a predominantly pro-inflammatory tumor microenvironment, driven by cytokines such as CXCL8, IL-6, and TNF-α, which create a favorable niche for tumor growth. Studies have shown that in later stages, the recruitment and polarization of M2 macrophages lead to the secretion of anti-inflammatory cytokines, including IL-10 and TGF-β (24). This results in an immunosuppressive tumor microenvironment that supports angiogenesis and facilitates tumor invasion and metastasis.

A strength of this study is the use of actual patient samples, which provides a more physiologically relevant perspective on tumor-microbiome interactions compared to in vitro models. However, the use of patient biopsies introduces variability that may obscure certain associations observed in controlled experimental settings. Additionally, our cohort size is limited, and transcriptional analyses alone may not capture post-translational modifications or protein-level interactions that influence immune responses. Future research should focus on expanding sample sizes to validate these findings and incorporate spatial transcriptomics to map Fusobacterium-immune interactions within the tumor microenvironment. Future functional studies using advanced patient-derived techniques to localize gene expression in tissue sections may help clarify Fusobacterium’s role in immune modulation.

Our findings align with previous research on how Fusobacterium influences the tumor microenvironment. According to our in vivo investigations, increased expression of several immune genes, such as CXCL8, IL6, SPP1, and IDO1, was associated with a high Fusobacterium load of several Fusobacterium species but dominated by F. animalis. The fap2 gene was detected in most F. animalis- and F. nucleatum-positive samples and was actively transcribed, suggesting constitutive expression. Notably, tumors with high Fusobacterium abundance exhibited no significant change in M2 macrophage polarization.