Epacadostat Overcomes Cetuximab Resistance in Colorectal Cancer by Targeting IDO-Mediated Tryptophan Metabolism

3.1 Cetuximab Increases IDO1/IDO2 Expression in CRC Patients

To investigate the mechanism of cetuximab resistance in CRC, RNA sequencing from four liver metastatic biopsies of CRC from two patients with pre-treated and acquired resistance to cetuximab was analyzed (GSE84267). GO pathway analysis of differentially expressed genes (DEGs) revealed significant enrichment in the tryptophan metabolism pathway (Figure 1A). Notably, key enzymes in the tryptophan metabolism pathway, including IDO1 and IDO2, were significantly upregulated among the top 40 most dysregulated genes, shedding light on the potential alterations in tryptophan metabolism associated with cetuximab resistance (Figure 1B).

To validate our findings, we collected serum from CRC patients treated with or without cetuximab. Following cetuximab treatment, the levels of the kynurenine/tryptophan (K/T) ratio in patient serum, serving as a measure of IDO activity, showed a significant increase compared to the patients with only neoadjuvant chemotherapy (Figure 1C). Detailed ELISA kit information could be found in Table S1. Furthermore, CRC patients treated with cetuximab exhibited increased IDO1 and IDO2 in tumor tissues, as determined by qPCR (Figure 1D) and Western blot (Figure 1E). Detailed sequences of qPCR primers and antibodies for Western Blot analysis could be found in Table S2 and S3. IHC analysis also revealed a significant elevation in the expression levels of IDO1 and IDO2 in tumor tissues after treatment (Figure 1F). Detailed information on antibodies for IHC analysis could be found in Table S4. These observed phenomena in tumor tissues were consistent with the changes in IDO1 and IDO2 expression levels in patient serum, indicating that cetuximab treatment could promote IDO expression in CRC patients.

3.2 Cetuximab Increases IDO1/IDO2 Expression in RAS/RAF Mutant CRC Cells and Mouse Models

To investigate whether cetuximab induced IDO1 and IDO2 expression in vitro and in vivo, we employed genetically mutated CRC cell lines, HCT116 (KRAS mutant) and HT29 (BRAF mutant), as our cellular and mouse models. These cells were treated with DMSO, 10 μg/mL cetuximab, followed by the assessment of cell viability using CTG. Notably, cetuximab treatment did not significantly alter cell viability after 96 h (Figure S1A). Furthermore, flow cytometry (FC) results demonstrated that cetuximab did not considerably induce apoptosis in HCT116 and HT29 cells (Figure S1B). Likewise, clonogenic assays and transwell assays revealed that cetuximab had no significant effect on the proliferation and migration ability of HCT116 and HT29 cells (Figure S1C). These results indicated that cetuximab monotherapy may not produce cytotoxic effects on HCT116 and HT29 RAS/RAF mutant CRC cells by inhibiting cell proliferation, viability, or migration, or inducing apoptosis, in contrast to the effects seen in RAS/RAF wild-type CRC cells like Caco-2 (Figure S2A–E).

However, HCT116 and HT29 cells treated with cetuximab significantly upregulated the IDO1 and IDO2 mRNA expression (Figure 2A). Consistent with these findings, ELISA analysis revealed an increase in the K/T ratio in the cell culture supernatants following cetuximab treatment, suggesting enhanced cellular IDO activity (Figure 2B). In addition, the protein expression levels of both IDO1 and IDO2 were markedly elevated in the cetuximab-treated group (Figure 2C). These results indicated that cetuximab upregulated IDO1 and IDO2 expression in vitro.

Consistent with the in vitro findings, cetuximab treatment in HCT116 and HT29 subcutaneous xenograft immune-deficient mice did not significantly inhibit tumor growth compared to the control group, indicating limited anti-tumor efficacy of cetuximab monotherapy (Figure S2F). Similarly, H&E staining showed no significant morphological changes in tumor tissues from cetuximab-treated mice (Figure S2G). Notably, IHC staining revealed a marked increase in IDO1 and IDO2 protein levels in cetuximab-treated tumors (Figure 2D). This upregulation was further confirmed by qPCR and western blot analyses, which demonstrated elevated mRNA and protein expression of IDO1 and IDO2, respectively (Figure 2E,F). Additionally, the K/T ratio in cetuximab-treated mice was significantly increased, consistent with observations in patient serum samples following cetuximab therapy (Figure 2G). Collectively, these in vitro and in vivo results demonstrate that cetuximab upregulates IDO1 and IDO2 expression in RAS/RAF mutant CRC cells and mouse models.

3.3 IDO Inhibitor Overcomes Cetuximab Resistance In Vitro

By inhibiting IDO1, epacadostat reduces the consumption of tryptophan and the accumulation of kynurenine, thereby enhancing the immune response to tumors and potentially slowing tumor growth or even inducing tumor regression. It has been investigated as a potential treatment for various types of cancer, particularly in combination with other immunotherapies or chemotherapy agents [18]. Accordingly, we hypothesized that cetuximab upregulates IDO1 and IDO2, thereby promoting CRC progression and immune evasion.

To evaluate the effects of IDO1 inhibition, we first examined the migratory capacity of CRC cells. Under the combined treatment of cetuximab and epacadostat, the migratory ability of CRC cells was significantly reduced compared to the epacadostat group (Figure 3A). Next, we assessed the clonogenic ability of CRC cells and found that the combination of cetuximab and epacadostat significantly inhibited colony formation compared to epacadostat treatment alone (Figure 3B). We then evaluated cell viability using CTG assays. Following 96 h of treatment, cetuximab monotherapy had no significant impact on the viability of CRC cells, while epacadostat alone reduced cell viability compared to the control group. Notably, the combination of cetuximab and epacadostat resulted in a significantly greater reduction in cell viability compared to epacadostat alone (Figure 3C). To further investigate the mechanisms underlying the observed effects, we analyzed apoptosis-related markers. The combination treatment markedly increased the Bax/Bcl-2 ratio and the levels of cleaved-caspase3/caspase3, indicating enhanced apoptosis compared to epacadostat alone (Figure 3D). Consistent with these results, FC analysis demonstrated that epacadostat sensitized HCT116 and HT29 cells to cetuximab, with a higher apoptotic cell rate observed in the combination group compared to epacadostat monotherapy (Figure 3E). In summary, these findings demonstrate that the combination of cetuximab and epacadostat exerts a significant anti-tumor effect by enhancing cytotoxic and antiproliferative activities in vitro.

To further validate the role of IDO1 in CRC progression, we conducted IDO1 knockdown experiments using small interfering RNA (siRNA; si-IDO1). Consistent with the effects observed with epacadostat, si-IDO1 significantly attenuated the migratory capacity of HCT116 and HT29 cells in transwell assays compared to the control group (Figure S3A). Moreover, the combination of cetuximab and si-IDO1 further reduced cell migration compared to si-IDO1 alone (Figure S3A). CTG assay results demonstrated a significant reduction in cell viability upon IDO1 depletion, and the combination showed a more pronounced decrease in cell viability than si-IDO1 alone (Figure S3B). Additionally, FC analysis revealed increased apoptosis in the cetuximab-treated group with IDO1 knockdown, suggesting that cetuximab exerts stronger anti-tumor effects in the absence of IDO1 (Figure S3C). These findings collectively demonstrate that IDO1 plays a pivotal role in promoting CRC cell survival, migration, and resistance to apoptosis.

3.4 IDO Inhibitor Overcomes Cetuximab Resistance In Vivo

Based on the previous findings, we next investigated whether inhibition of IDO sensitized CRC to cetuximab in vivo. To achieve this, we established a subcutaneous tumor model using the mouse cell line MC38. Once the tumors were palpable (100–150 mm³), mice were divided into four treatment groups: control (DMSO), cetuximab (50 mg/kg, once weekly), epacadostat (300 mg/kg, once daily), or a combination of cetuximab and epacadostat (Figure 4A).

After 28 days of treatment, the tumor size gradually increased in both the control and cetuximab treatment groups. However, the administration of epacadostat alone and the combination group both led to a suppression in tumor growth (Figure 4B), and the combination group exhibited a remarkable reduction in tumor volume and tumor weight (Figure 4C,D). Additionally, the K/T ratio in the serum of mice in the epacadostat and combination therapy groups decreased, indicating that epacadostat inhibits tryptophan metabolism (Figure 4E). Next, we conducted H&E staining on mouse tumor tissues to assess the pathological status, followed by the evaluation of tumor proliferation and apoptosis using Ki-67 and TUNEL staining, respectively. The results of H&E staining indicated a decrease in tumor cell count and proliferation of peritumoral fibrous tissue in the epacadostat group, while these post-treatment pathological features were even more pronounced in the combination group (Figure 4F). Then, from the results of the Ki-67 and TUNEL assay, we found that compared to the inapparent therapeutic effect of cetuximab monotherapy, epacadostat showed significant efficacy, whereas the combination therapy exhibited higher apoptosis and lower proliferation than the epacadostat monotherapy group (Figure 4F). Additionally, compared to the control group, IDO1 expression was significantly reduced in the epacadostat monotherapy group, and a further decrease was observed in the combined treatment group compared to the cetuximab monotherapy group (Figure 4F). Detailed information on antibodies and the TUNEL kit for IHC and IF analysis could be found in Table S4. These in vivo results further supported the in vitro evidence that IDO inhibition overcame cetuximab resistance in CRC.

3.5 Cetuximab-Induced IFN-γ Signaling Promotes the Expression of IDO

To elucidate the mechanism by which cetuximab promotes IDO1 and IDO2 expression, we performed GSEA analysis on RNA sequencing data obtained from liver metastatic biopsies of CRC patients who acquired resistance to cetuximab. The GSEA analysis indicated an upregulation of the IFN-γ signaling pathway in the cetuximab-resistant group (Figure 5A). Evidence suggested that cetuximab could activate immune cells, resulting in the release of various cytokines, including IFN-γ. This has been validated in models such as breast cancer [19]. Furthermore, elevated levels of IFN-γ in the TME are often associated with the release of damage-associated molecular patterns (DAMPs), which are generated in response to or damage induced by therapies such as cetuximab. DAMPs, in turn, can further amplify IFN-γ production, creating a positive feedback loop that promotes immune escape mechanisms [20]. Therefore, we aimed to investigate the impact of cetuximab on IFN-γ signaling and to elucidate the regulatory role of IFN-γ in modulating IDO expression in CRC. We examined the expression of IFN-α, IFN-β, and IFN-γ mRNA in the tumor tissues of the subjects. The results showed that the expression of IFN-γ mRNA was significantly increased in the cetuximab treatment group (Figure 5B), while the levels of IFN-α and IFN-β remained relatively unchanged (Figure S4A). Additionally, serum IFN-γ concentrations were significantly elevated in CRC patients treated with cetuximab (Figure 5C), and IHC staining of tumor tissues corroborated these findings (Figure 5D). To investigate whether cetuximab-induced IDO upregulation was associated with the MAPK pathway, a key resistance mechanism in cetuximab-resistant tumors [1, 3, 5, 6, 16], we treated CRC cells with the MEK inhibitor trametinib. Our results demonstrated that trametinib did not attenuate cetuximab-induced IDO expression, suggesting that the upregulation of IDO by cetuximab is independent of the MAPK pathway (Figure S4B,C). Notably, we observed increased expression of calreticulin (CRT), a well-established marker of DAMPs, in HCT116 and HT29 (Figure S4D). This suggests that cetuximab may also enhance IFN-γ levels in CRC through the release of DAMPs.

To further elucidate the relationship between IFN-γ and IDO in CRC, we investigated whether IFN-γ directly regulated IDO1 transcription and expression. Dual luciferase reporter assays conducted in HCT116 and HT29 cells revealed that cetuximab activated the transcription of IDO1. Additionally, treatment with the IFN-γ cytokine (10 ng/mL) also activated IDO1 transcription in both cell lines, confirming that cetuximab might promote IDO1 transcription activity via upregulating IFN-γ (Figure 5E). Subsequently, siRNA-mediated knockdown of IFN-γ resulted in the downregulation of IDO1 and IFN-γ expression in HCT116 and HT29 cells (Figures 5F,G and S4E). Next, we examined the type II IFN signaling and the IDO (IDO1 and IDO2) in the TCGA-CRC cohort. As anticipated, IFN-γ showed a strong positive correlation with the expression of IDO1 and IDO2 (Figure 5H). These data indicated that cetuximab induced type II IFN signaling activation, enhancing IDO expression in cetuximab-resistant CRC.

3.6 Epacadostat Overcomes Cetuximab-Resistant Tumor by Promoting the Polarization of Macrophage M1

Tryptophan and its metabolites influence tumor growth and immune evasion. Increased IDO activity could lead to local depletion of tryptophan and accumulation of immunosuppressive kynurenine metabolites, creating an immunosuppressive milieu that promotes tumor immune evasion and tolerance [21]. Therefore, immune cells were isolated by FC and stained with specific markers for further analysis (Figure S5A).

In peripheral blood, no significant differences were observed in the populations of CD4⁺ T lymphocytes (CD45⁺/CD3⁺/CD4⁺) across the treatment groups (Figure 6A). However, epacadostat treatment alone increased the levels of CD8⁺ T lymphocytes (CD45⁺/CD3⁺/CD8⁺), with a more pronounced elevation observed in the combination treatment group compared to epacadostat monotherapy (Figure 6A). Consistent with this, the levels of IFN-γ⁺ CD8⁺ T cells mirrored the changes in CD8⁺ T lymphocytes (Figure 6B). Furthermore, the combination treatment significantly increased the proportion of M1 macrophages while reducing the proportion of M2 macrophages in peripheral blood (Figure 6C). In tumor tissues, similar patterns were observed. The combination treatment significantly enhanced the infiltration of CD8⁺ T lymphocytes and IFN-γ⁺ CD8⁺ T cells compared to epacadostat alone (Figure 6D,E). Additionally, the combination therapy shifted the macrophage polarization toward an M1 phenotype, as evidenced by an increased M1/M2 ratio (Figure 6F). In the spleen, the immune cell profiles aligned with those in peripheral blood and tumor tissues. The combination treatment elevated the levels of CD8⁺ T lymphocytes and IFN-γ⁺ CD8⁺ T cells (Figure S5B,C) and promoted M1 macrophage polarization (Figure S5D). However, neutrophil (CD45⁺/Ly6G⁺/CD11b⁺) populations remained unchanged across all treatment groups in peripheral blood, spleen, and tumor tissues (Figure S5E–G).

CD8⁺ T lymphocytes are pivotal effector cells in tumor immunity, capable of directly killing tumor cells and secreting cytokines such as IFN-γ to promote anti-tumor immune responses and inhibit tumor growth [22]. These findings demonstrated that the increases in infiltration of CD8⁺ T lymphocytes and IFN-γ⁺ CD8⁺ T cells in the TME were involved in the process of combination therapy against CRC. Moreover, these results also suggested that the polarization of M1 macrophages was involved in the anti-tumor immune response, whereas the polarization of M2 macrophages exhibited a tumor-promoting effect. Detailed information on antibodies and related reagents for FC analysis could be found in Table S5. FC analysis collectively indicated a positive correlation between therapeutic response and the activation of CD8⁺ T lymphocytes and M1 macrophages in our models, suggesting that combination therapy enhanced apoptosis in cetuximab-resistant CRC by activating immune effector cells in the TME (Figure 7).