Thiostrepton suppresses intrahepatic cholangiocarcinoma progression via FOXM1-mediated tumor-associated macrophages reprogramming

Highlights

• •Thiostrepton (TST) effectively inhibited the proliferation and metastasis of intrahepatic cholangiocarcinoma (ICC) cells by targeting FOXM1.
• •TST reprogrammed tumor-associated macrophages (TAMs) towards the M1 phenotype, promoting anti-tumor immune responses in ICC.
• •Utilizing murine ICC cells innovatively for constructing a mouse tumor model, TST significantly reduced tumor burden in mouse models, highlighting its potential as a novel therapeutic strategy for ICC patients.

Abstract

Intrahepatic cholangiocarcinoma (ICC) is an aggressive cancer with an extremely poor prognosis, highlighting the urgent need for new treatment options. Recent studies increasingly suggest that the Forkhead box M1 (FOXM1) transcription factor may serve as a candidate target for cancer immunotherapy. However, its role and the underlying molecular mechanisms in ICC remain not fully understood. Here, we identify thiostrepton (TST) as a potent FOXM1 inhibitor, capable of exerting “dual anti-tumor” effects in ICC. On one hand, TST effectively suppresses tumor cell proliferation and metastasis. On the other hand, TST treatment improves the tumor immune microenvironment by reprogramming tumor-associated macrophages (TAMs), thereby enhancing anti-tumor immune responses. Mechanistically, TST directly alleviates ICC progression by arresting the cell cycle, promoting apoptosis, and inhibiting the epithelial-mesenchymal transition (EMT) process. Furthermore, TST-treated tumor cells secrete cytokines that drive TAMs repolarization toward the tumor-suppressive M1 phenotype. Overall, our results indicate that FOXM1 can serve as a novel target for ICC immunotherapy. By targeting FOXM1, TST exerts “dual anti-tumor” effects and has the potential to become a promising immunotherapy agent for ICC patients.

Introduction

Immune checkpoint blockade (ICB) has achieved considerable success in tumor treatment. However, a large proportion of patients with "cold tumors," including cholangiocarcinoma, exhibit minimal to no responsiveness to these treatments [1,2]. The immune-suppressive tumor microenvironment enriched with tumor-associated macrophages (TAMs) represents a major barrier to effective immunotherapy [3]. Intrahepatic cholangiocarcinoma (ICC), a highly heterogeneous and immunotherapy-resistant malignancy, is characterized by a tumor microenvironment (TME) with significant infiltration of TAMs [1,4]. TAMs are crucial in establishing and shaping the immunosuppressive microenvironment. They are extensively involved in both innate and adaptive immunity, inhibiting the anti-tumor functions of immune cells and diminishing the body's antigen-presenting capacity, thereby playing an indispensable role in tumor immune evasion [5]. There are two functionally contrasting subtypes of TAMs: classically activated M1 macrophages and alternatively activated M2 macrophages. Notably, TAMs, mainly referring to M2 phenotype, secrete inhibitory cytokines such as TGF-β and IL-10 to affect immune cells, creating a favorable immunosuppressive TME for tumor progression and immunotherapy resistance [6]. In contrast, M1-like TAMs exert anti-tumor activities, including mediating cytotoxicity and antibody-dependent cell-mediated cytotoxicity (ADCC) to eliminate tumor cells [7]. Interestingly, macrophages exhibit high plasticity, which means that reprogramming M2-like TAMs can potentially increase anti-tumor activity [[8], [9], [10]]. Therefore, repolarizing TAMs toward a pro-inflammatory, anti-tumorigenic M1-like state has the potential to reverse the immunosuppressive TME and convert "cold tumors" into "hot tumors".

Forkhead box M1 (FOXM1) is a key transcription factor in the FOX family. It significantly promotes tumor progression by regulating the cell cycle, enhancing invasion and metastasis, and interacting with multiple pro-oncogenic signaling pathways, including Wnt/β-catenin, TGF-β/SMAD3, and PI3K/AKT pathways [[11], [12], [13], [14]]. The upregulation of FOXM1 in tumor tissues, particularly in ICC, has been demonstrated to be substantially correlated with a poor clinical prognosis [[15], [16], [17], [18], [19], [20], [21], [22]]. Additionally, elevated levels of FOXM1 are commonly associated with chemotherapy resistance in cancer cells [23]. However, it has been demonstrated that the combination of FOXM1 inhibitors and 5-fluorouracil has a synergistic anti-tumor effect in ICC cells [24]. Furthermore, increasing studies have indicated that high FOXM1 expression contributes to an immunosuppressive microenvironment in tumor cells [25,26]. Targeting FOXM1 is anticipated to enhance the efficacy of lung cancer immunotherapy by reducing PD-L1 expression [27]. Nevertheless, since the impact of targeting FOXM1 in remodeling the immune microenvironment of ICC remains unclear, the anti-tumor efficacy of FOXM1 inhibition warrants further investigation. Thiostrepton (TST), a specific inhibitor of FOXM1, has shown considerable anti-tumor potential in previous preclinical studies [[28], [29], [30]]. Moreover, it is currently being evaluated in a phase II clinical trial to assess the safety and efficacy in treating patients with malignant pleural effusion (ClinicalTrials.gov identifier: NCT05278975).

In this study, we found that TST has a direct cytotoxic impact on ICC cells by specifically targeting FOXM1. Additionally, TST was able to reverse the immunosuppressive tumor microenvironment. TAMs were reprogrammed toward the M1 phenotype following TST treatment, resulting in an augmentation of immune responses against the tumor. Additionally, we further evaluated the safety of TST treatment, suggesting its potential and value for clinical application. Therefore, TST holds promise as a novel therapeutic agent for ICC treatment.

Materials and methods

Cell culture and transfection

The human ICC cell lines (CCLP-1 and HuCCT-1), the human monocytic leukemia cell line (THP-1) and the human embryonic kidney cell line (HEK-293T) were purchased from Shanghai Institutes of Biological Sciences, Chinese Academy of Sciences (Shanghai, China). The murine ICC cell line (MIC) was a gift from Fudan university (Shanghai, China). ICC, THP-1 and MIC cells were routinely cultured in RPMI-1640 medium [Biological Industries (BI), Israel] supplemented with 10% fetal bovine serum (BI, Israel), 100 μg/mL streptomycin (Sigma-Aldrich, Germany) and 100 U/mL penicillin (Sigma-Aldrich, Germany) at 37 °C in a humidified incubator (Thermo Scientific, USA) with 5% CO₂. HEK-293T cells were cultured under similar conditions, but in Dulbecco's modified Eagle's medium (BI, Israel) with the same supplements. Differentiation of monocytes (THP-1) toward naive macrophages (M0) was induced by stimulating with 100 nM phorbol 12-myristate 13-acetate (PMA) (MCE, USA) for 3 days, followed by an additional 24-hour incubation in fresh RPMI medium without PMA [31]. Thiostrepton (TST), Robert Costa Memorial drug-1 (RCM-1), and Forkhead domain inhibitor-6 (FDI-6) were purchased from MCE (USA); Siomycin A (SioA) was from Cayman chemical (USA).

Lentiviral plasmids encoding short hairpin RNA (shRNA) sequences targeting FOXM1, along with negative control RNAs, were generated by RepoBio (Hangzhou, China). Additionally, lentiviruses expressing FOXM1 and corresponding empty vectors were also prepared. These plasmids were co-packaged with pMD2G and psPAX2 plasmids and subsequently transfected into HEK-293T cells. Forty-eight hours post-transfection, lentiviral supernatants were collected and used to transfect the target cells. The transfected cells were selected using 3 μg/mL puromycin for a duration of one week. The efficacy of knockdown or overexpression was validated by comparing with negative control using qRT-PCR and Western blotting (Supplementary Fig. S1). Detailed sequences of the shRNAs are provided in Supplementary Table S1.

Human specimen

Human ICC tumor tissues, along with adjacent non-tumorous tissues, were obtained from the First Affiliated Hospital, Zhejiang University, School of Medicine. This project was approved by the Research Ethics Committee of the First Affiliated Hospital of Zhejiang University, School of Medicine (IIT20240765B). All patient samples were collected with informed consent, and the research procedures adhered strictly to the guidelines of the Declaration of Helsinki.

Mouse studies

All animal experiments were conducted in compliance with procedures approved by the Animal Care and Use Committee of the First Affiliated Hospital of Zhejiang University, School of Medicine. Male BALB/c nude and C57BL/6 mice, approximately 4 weeks old, were purchased from GemPharmatech (Jiangsu, China) and housed within a Specific Pathogen-Free (SPF) grade barrier system.

CCLP-1 cells (5 × 10⁶) were resuspended in 100 μL PBS and subcutaneously injected into the left flank of BALB/c nude mice to construct subcutaneous xenograft tumor models. On day 14 post-implantation, the mice were randomly assigned to two groups (n = 5) and received intraperitoneal injections of either TST (17 mg/kg, every other day) or DMSO as a control until sacrifice. Tumor volume and body weight were monitored every other day. Tumor volume was calculated using the formula: 1/2*(length × width²).

MIC cells (1 × 10⁷) resuspended in 100 μL PBS were subcutaneously injected into C57BL/6 mice to construct subcutaneous ICC models. On day 21 post-implantation, the mice were randomly assigned to two groups (n = 5) and received intraperitoneal injections of either TST or DMSO. After the final treatment, the mice were sacrificed, and the tumors were collected for histological and flow cytometric analyses.

Hydrodynamic tail vein injection was utilized to construct orthotopic ICC models [32]. Plasmids (25 μg pT3-myr-AKT-HA, 25 μg pT3-EF1a-NICD1, and 25 μg pCMV(CAT)T7-SB100) were diluted in 2 ml of saline and rapidly injected into the tail vein of C57BL/6 mice within 7 ss. On day 21 post-injection, the mice were randomly assigned to two groups (n = 5) and received intraperitoneal injections of either TST or DMSO. After the final treatment, the mice were sacrificed, and their livers were collected for further analysis.

Statistical analysis

Statistical analyses were conducted using GraphPad Prism 8.0.1 and R 4.4.0 software, with data presented as means ± standard deviation (SD). Significance was assessed using unpaired Student's t tests for comparisons between two groups, while one-way analysis of variance (ANOVA) was employed when there were more than two groups. The log-rank tests were used for Kaplan–Meier survival analysis. In this study, p < 0.05 was defined as statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001).

Results

TST effectively inhibits FOXM1 expression in ICC cells

In recent years, extensive research has demonstrated that FOXM1 significantly contributes to the initiation and progression of various cancers. We initially analyzed the publicly available gene expression datasets (TCGA, GSE26566, and GSE107943) and observed that FOXM1 mRNA expression was significantly upregulated in various tumors, including cholangiocarcinoma (Fig. S2A and B). Immunohistochemical (IHC) analysis of ICC clinical specimens revealed higher FOXM1 protein expression in tumor tissues compared to adjacent non-tumor tissues (Fig. S2C). Moreover, further analysis of the GEO database indicated that patients with elevated FOXM1 expression had shorter overall survival (OS) and disease-free survival (DFS), as well as increased tumor purity (Fig. S2D and E). These findings suggest a strong association between high FOXM1 expression and poor prognosis in ICC patients. Consequently, FOXM1 may be a potential therapeutic target for ICC treatment.

Here, we selected four candidate FOXM1 inhibitors for further investigation: thiostrepton (TST) , siomycin A (SioA) [33], Robert Costa Memorial drug-1 (RCM-1) [34], and Forkhead domain inhibitor-6 (FDI-6) [35] (Fig. 1A). Initially, we evaluated the effects of these compounds on FOXM1 transcription and protein levels in vitro. It was shown that, at a concentration of 5 μM, TST treatment significantly reduced mRNA and protein expression of FOXM1 in ICC cells, while the other three inhibitors showed no impact (Fig. 1, Fig. 1). Further studies revealed that these inhibitory effects of TST were concentration- and time-dependent (Fig. 1, Fig. 1). Additionally, we found that both SioA and TST could suppress the proliferation of tumor cells, with TST exhibiting a more pronounced effect (Fig. 1, Fig. 1). Moreover, the drug cytotoxicity experiment provided confirmation of the strong effectiveness of TST, as evidenced by the lowest half-maximal inhibitory concentration (IC₅₀) recorded at 5.130 μM (Fig. 1H). In summary, TST proved to be the most effective FOXM1 inhibitor in ICC cells.

Fig. 1.

TST effectively inhibits the expression of FOXM1 in ICC cells.

(A) Chemical structures of Thiostrepton, Siomycin A, RCM-1, and FDI-6. (B) The protein expression of FOXM1 in CCLP-1 cells treated with FOXM1 inhibitors at concentrations of 2.5, 5, 10, or 20 μM for 24 h. GAPDH was used as an internal control. (C) The mRNA expression of FOXM1 in CCLP-1 cells, treated with FOXM1 inhibitors at a concentration of 5 μM for 24 h. (D and E) Validation of FOXM1 expression after TST treatment for 24 or 48 h in the indicated cell lines using qRT‑PCR and Western blot. (F and G) EdU assay was applied to compare the cell proliferation ability in CCLP-1 cells, treated with FOXM1 inhibitors at a concentration of 5 μM for 24 h. Scale bar, 200 μm. (H) The half-maximal inhibitory concentrations (IC₅₀) of TST, SioA, RCM-1, and FDI-6 were determined after a 24-hour treatment in CCLP-1 cells. *P < 0.05, **p < 0.01, ***P < 0.001; ns, not significant. The data are expressed as the mean±SD of three independent experiments. TST, Thiostrepton; SioA, Siomycin A.

TST suppresses ICC cell proliferation in vitro

To investigate the effect of TST on tumor cell proliferation, various functional assays were conducted. As indicated by CCK-8 and colony formation assays, TST significantly inhibited the growth and colony formation of CCLP-1 and HuCCT-1 cells (Fig. 2A–C). EdU assays demonstrated a reduced DNA replication rate in ICC cells after TST treatment (Fig. 2, Fig. 2). Moreover, these inhibitory effect of TST on tumor proliferation became more significant with increasing concentration and prolonged treatment duration.

Fig. 2.

TST suppresses the proliferation of ICC cells in vitro.

(A) Effect of TST treatment on the proliferation of CCLP-1 and HuCCT-1 cells by CCK-8 assay. (B and C) CCLP-1 and HuCCT-1 cells were treated with different doses of TST and cell clonogenesis was measured by using a colony formation assay. Relative colony numbers are quantified in diagrams. (D and E) Representative images of EdU assays with corresponding quantified bar charts. Scale bar, 200 μm. *P < 0.05, **p < 0.01, ***p < 0.001; ns, not significant. The data are expressed as the mean±SD of three independent experiments.

We further explored how TST affects tumor cell proliferation using flow cytometry analysis. Consistent with our previous observations, TST treatment resulted in a concentration-dependent decrease in the S phase population of the cell cycle compared to the DMSO control group (Fig. 3A). Additionally, G1-S and G2-M phase arrest were observed in two different cell lines. Correspondingly, this treatment also led to a reduction in the levels of cell cycle-related proteins, including c-Myc, Cyclin B1, Cyclin E1, and Cyclin D1 (Fig. 3B). These results indicate that TST treatment alters cell cycle progression in ICC cells and plays a significant role in the G1-S and G2-M phase transitions. Moreover, there was a significant increase in the apoptotic cell population in ICC cells following TST treatment, accompanied by enhanced expression of apoptosis-related proteins such as the cleaved forms of PARP and caspase-9 (Fig. 3, Fig. 3). Collectively, these findings suggest that TST inhibits growth of ICC cells by arresting cell cycle progression and inducing apoptosis.

Fig. 3.

TST arrests the cell cycle and induces apoptosis in ICC cells.

(A) Flow cytometry was conducted to evaluate the cell cycle distribution in ICC cells treated with TST for 24 h. The percentages of cells in each phase of the cell cycle are quantified in diagrams. (B) Cell cycle-related protein levels in CCLP-1 and HuCCT-1 cells were assessed using Western blot. (C) Flow cytometry was conducted to evaluate the apoptosis cell population in ICC cells treated with TST for 24 h. The cells were stained with Annexin V and propidium iodide. The percentages of apoptotic cells are quantified in diagrams. (D) Apoptosis-related protein levels in CCLP-1 and HuCCT-1 cells were assessed using Western blot. **p < 0.01, ***p < 0.001; ns, not significant. The data are expressed as the mean±SD of three independent experiments.

TST inhibits the migration, invasion, and EMT process of ICC cells

FOXM1 is essential for the metastasis of tumor cells [36,37]. Wound healing and Transwell assays demonstrated that FOXM1 inhibition by TST suppressed the migration and invasion capacity of ICC cells (Fig. 4A and B). Besides, TST displayed a more potent capacity to impede the invasion and migration of tumor cells as the concentration increases. Epithelial-mesenchymal transition (EMT) contributes to the enhanced migration and metastasis of tumor cells. We subsequently assessed whether TST could interfere with the EMT process. The results showed that TST treatment upregulated E-cadherin expression while the expressions of N-cadherin, Snail, and SNAI2 were reduced in a time- and concentration-dependent manner (Fig. 4C). Taken together, our results reveal that TST efficiently limits the metastatic abilities of ICC cells.

Fig. 4.

TST inhibits the migration, invasion and EMT process of ICC cells.

(A) Migration capacity of ICC cells treated with TST was determined by wound healing assays in CCLP-1 and HuCCT-1 cells. Scale bar, 200 μm. Percentages of wound healed areas are quantified in diagrams. (B) Invasion and migration capabilities of CCLP-1 and HuCCT-1 were detected by Transwell assays. Scale bar, 200 μm. Relative count of invasion and migration cells is quantified in diagrams. (C) The protein levels of EMT markers in CCLP-1 and HuCCT-1 cells were detected by Western blot. *P < 0.05, **p < 0.01, ***p < 0.001; ns, not significant. The data are expressed as the mean±SD of three independent experiments.

TST mitigates ICC tumor burden in vivo

To further evaluate the therapeutic potential of TST in vivo, subcutaneous xenograft tumor models were established in nude mice. We randomly assigned mice to receive either vehicle control or TST (17 mg/kg) every two days via intraperitoneal injection (Fig. 5A). Consistent with the results in vitro, it is encouraging to observe that the TST-treated group showed a reduction of roughly 45% in tumor weight compared to the control group, with a corresponding slowing of tumor growth (Fig. 5, Fig. 5). No discernible difference in body weight was observed between the groups (Fig. 5D).

Fig. 5.

TST inhibits the growth of subcutaneous ICC models.

(A) Timeline of in vivo experiments performed on BALB/c nude mice. CCLP-1 cells (5 × 10⁶) were injected subcutaneously and left for 2 weeks to form tumors. On day 14, mice were injected with TST at a dose of 17 mg/kg at indicated time points represented by black arrows. (B) Representative images of tumor size in nude mice injected subcutaneously with CCLP-1 cells and the statistical analysis of tumor weigh in different groups. (C and D) Statistical analyses of the tumor growth curves and body weight change in different groups. (E) Timeline of in vivo experiments performed on C57BL/6 mice. MIC cells (1.0 × 10⁷) were injected subcutaneously and left for 3 weeks to form tumors. On day 21, mice were injected with TST at a dose of 17 mg/kg at indicated time points represented by black arrows. (F and G) Representative images of tumor size in C57BL/6 mice injected subcutaneously with MIC cells and the statistical analysis of tumor weigh in different groups. (H) H&E staining and immunohistochemical analysis of CK19, Ki67, PCNA, and FOXM1 in tumor tissues. Scale bar, 100 μm. *P < 0.05, ***p < 0.001; ns, not significant. i.p., intraperitoneal injection.

What's more, in order to investigate possible systemic impact, we initially carried out a trial on a subcutaneous murine tumor model in C57BL/6 mice (Fig. 5E). It was noticed that tumor growth was severely delayed in mice administered with TST whereas in those treated with the DMSO, the tumor grew rapidly (Fig. 5, Fig. 5). This was also confirmed by IHC, which showed that proliferation-related markers in cholangiocarcinoma decreased following the reduction of FOXM1 expression (Fig. 5H).

Following this, we next analyzed the impact of TST in a mouse orthotopic tumor model (Fig. 6A). As shown in Fig. 6, Fig. 6, TST treatment considerably attenuated the growth of tumor. Besides, we observed abundant E-cadherin expression in specimens with low FOXM1 expression following TST administration. Conversely, the percentage of N-cadherin as well as Vimentin substantially decreased in response to TST (Fig. 6D). This may imply that the EMT process in the tumor is hindered after TST treatment. Furthermore, the biological safety of TST was evaluated in mice. Serum biomarker levels showed no significant changes between the two groups (Fig. 6E). The H&E staining results also indicated no noticeable damage or inflammatory lesions in major organs, including the heart, lungs, spleen, kidneys, and brain (Fig. 6F). Therefore, TST appears to be a promising candidate for ICC treatment due to its efficacy and favorable safety profile.

Fig. 6.

TST reduces tumor burden in orthotopic ICC models.

(A) Timeline of in vivo experiments performed on C57BL/6 mice. The mixed plasmids were injected through the tail vein using hydrodynamic pressure and left for 3 weeks to form tumors. On day 21, mice were injected with TST at a dose of 17 mg/kg at indicated time points represented by black arrows. (B) Representative images of orthotopic ICC in C57BL/6 mice, alongside statistical analyses of liver weights and liver-to-body weight ratios. (C) Immunohistochemical analysis of Ki67, PCNA, and FOXM1 in tumor tissues. Scale bar, 100 μm. (D) Immunohistochemical analysis of EMT markers in tumor tissues. Scale bar, 100 μm. (E) Serum biochemical data of the mice in the DMSO- and TST-treated groups (n = 3). (F) Representative images of H&E-stained tissue specimens (heart, kidney, spleen, brain, and lung). Scale bar, 100 μm. **p < 0.01; ns, not significant. i.p., intraperitoneal injection. Alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), total bilirubin (TBIL), creatine kinase-MB (CK-MB), creatinine (Crea), blood urea nitrogen (BUN).

RNA-seq analysis reveals that FOXM1 potentially mediates the crosstalk between tumor cells and TAMs

TST effectively suppresses tumor progression by targeting FOXM1. To explore the potential mechanisms through which FOXM1 regulates ICC progression, RNA-seq was performed on HuCCT-1 cells transfected with shFOXM1 and shNC vectors. Notably, FOXM1 knockdown significantly reduced the expression of several key oncogenic factors, including FGFR2, highlighting its critical role as a regulator of multiple tumorigenic pathways in ICC progression (Fig. 7A and B). Surprisingly, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed a strong association between FOXM1 expression and immune-related pathways, particularly the cytokine‒cytokine receptor interaction pathway (Figs. 7B and S4). Gene Ontology (GO) enrichment analysis and Gene Set Enrichment Analysis (GSEA) further supported the involvement of FOXM1 in regulating cytokine activity and chemokine-mediated signaling pathways (Fig. 7, Fig. 7 and S5). To explore the molecular mechanisms underlying these pathways, protein–protein interaction (PPI) analysis identified eight hub genes in the cytokine–cytokine receptor interaction pathway that were closely associated with FOXM1 (Fig. 7E). Subsequent qRT-PCR and ELISA analyses validated that FOXM1 inhibition significantly reduced CCL2 expression while enhancing GM-CSF levels in ICC cells (Fig. 7, Fig. 7). Conversely, FOXM1 overexpression led to higher CCL2 levels but suppressed GM-CSF expression. CCL2 is a potent chemokine that recruits monocytes and macrophages, while GM-CSF promotes their differentiation and activation. Consistent with this, immune infiltration analysis revealed that the expression levels of CCL2 and GM-CSF were significantly correlated with macrophage infiltration in cholangiocarcinoma (Fig. 7H). GSEA further indicated that downregulation of FOXM1 may influence tumor-macrophage interactions in the TME, potentially promoting a shift toward an M1-like macrophage phenotype (Fig. 7I). These findings suggest that FOXM1 may play a crucial role in the crosstalk between tumor cells and TAMs by regulating cytokines.

Fig. 7.

RNA-seq analysis reveals that FOXM1 potentially mediates the crosstalk between tumor cells and TAMs.

(A) Volcano plot of DEGs. Red dots on the volcano plot represent upregulated genes. Blue dots represent downregulated genes in the shFOXM1 group compared with shNC group, with a fold change>1.5 or < 1/1.5 and p-value≤0.05. (B and C) KEGG enrichment and GO enrichment analyses of DEGs. (D) GSEA showed the close association between FOXM1 expression and the activity of chemokine- and cytokine-related signaling pathways in ICC. (E) PPI analysis identifying eight hub genes in the cytokine‒cytokine receptor interaction pathway closely related to FOXM1. (F) qRT-PCR analysis detecting the effect of FOXM1 on the expression of the identified hub genes in HuCCT-1 cells. (G) ELISA analysis showing the effect of FOXM1 on the secretion levels of CCL2 and GM-CSF in the supernatants of HuCCT-1 cells. (H) The correlations of CCL2 and CSF2 expression with immune cell infiltration in the TCGA-CHOL cohort by ssGSEA. (I) GSEA revealed enrichment of macrophage-related pathways, including those associated with M1 macrophages, in relation to FOXM1 expression in ICC. *P < 0.05, **p < 0.01, ***p < 0.001; ns, not significant. The data are expressed as the mean±SD of three independent experiments.

TST reprograms TAMs toward the M1 phenotype by inhibiting FOXM1-mediated pathways

To clarify how TST affects macrophage infiltration and functional states by inhibiting FOXM1, we performed flow cytometric analyses on mouse orthotopic and subcutaneous ICC tissues. The results illustrated that TST treatment remarkably elevated the proportion of tumor-infiltrating M1 macrophages within the TME (Fig. 8A and B). Intriguingly, we noticed that tumors in TST-treated mice showed a notable rise in the overall proportion of CD3⁺ T cells among CD45⁺ immune cells, while there was no significant difference in the proportions of CD4⁺ or CD8⁺ T cells (Fig. 8, Fig. 8 and S7). Immunofluorescence analysis further confirmed that TST treatment increased the infiltration of M1 macrophages and decreased the infiltration of M2 macrophages (Fig. 8E). Furthermore, TST administration significantly reduced serum CCL2 levels in tumor-bearing mice. Concurrently, it markedly increased the secretion of the pro-inflammatory cytokine IL-12p70, potentially stimulating immune response against the tumor (Fig. 8F).

Fig. 8.

TST reprograms TAMs toward the M1 phenotype.

Flow cytometric assays were conducted on mouse orthotopic and subcutaneous ICC tissues. Representative flow cytometry and statistical analysis showed the proportion of M1 macrophages in orthotopic ICC models (A) and subcutaneous ICC models (B). Representative flow cytometry and statistical analysis showed the proportion of CD3⁺ T cells in orthotopic ICC models (C) and subcutaneous ICC models (D). (E) Immunofluorescence staining was performed to demonstrate the proportion of M1 macrophages (CD86⁺+ F4/80⁺) and M2 macrophages (CD206⁺+ F4/80⁺) in orthotopic ICC tissues. Scale bar, 100 μm. (F) Differences in cytokine levels in the serum of mice between the TST treatment group and the DMSO control group (n = 3). *P < 0.05, **p < 0.01.

To further validate the effect of TST on the reprogramming of TAMs in vitro, THP-1 cells were co-cultured with conditioned medium (CM) derived from TST-treated HuCCT-1 cells. Compared to the CM from untreated tumor cells, TST significantly upregulated the expression of M1-associated genes (iNOS, CD86, and CD68) while downregulating the expression of M2-associated genes (IL-10 and CCL22) (Fig. 9A). Notably, a comparison between Groups 1 and 2 revealed that in the absence of CM, TST alone had no significant effect on the polarization of THP-1 cells toward M1. In contrast, TST-treated CM markedly counteracted the influence of CM on promoting the polarization of THP-1 cells toward M2 phenotype. These findings suggest that TST influences macrophage polarization indirectly by affecting tumor cell.

Fig. 9.

TST induces macrophages into pro-inflammatory state by inhibiting FOXM1-mediated pathways.

(A) THP-1 cells were cultured in normal medium with or without TST (2.5 μM or 5μM) for 24 h (group 1), or in conditioned medium (CM) from HuCCT-1 tumor cells pre-treated with or without TST (2.5 μM or 5μM) for 24 h (group 2). The transcript levels of the indicated genes were quantified by qRT‑PCR. (B and C) THP-1 cells were co-cultured with transfected HuCCT-1 cells with or without TST (5 μM) for 24 h. The transcript levels of the indicated genes were quantified by qRT‑PCR. (D and E) The secretion levels of TNF-α and IL-1β from THP‑1 cells co‑cultured with HuCCT-1 cells were detected using ELISA. (F and G) The mRNA and protein expression levels of phagocytosis checkpoints in transfected HuCCT-1 cells were detected by qRT-PCR and Western blot, respectively. *P < 0.05, **p < 0.01, ***p < 0.001; ns, not significant. The data are expressed as the mean±SD of three independent experiments.

Afterwards, to determine whether the TST-induced reprogramming of TAMs is mediated by FOXM1, we co-cultured THP-1 cells with transfected HuCCT-1 cells. It was shown that when THP-1 cells were co-cultured with FOXM1-knockdown tumor cells, there was an upregulation of M1-associated genes (CD86, CD68) and a downregulation of M2-associated genes (IL-10, CCL22) (Fig. 9B). Additionally, the secretion of pro-inflammatory cytokines TNF-α and IL-1β in the cell supernatant was significantly increased (Fig. 9D). In the contrary, the overexpression of FOXM1 led to contrasting outcomes, favoring M2 polarization (Fig. 9, Fig. 9). However, the addition of TST to the FOXM1-overexpression group reversed these unfavorable effects, promoting the repolarization of TAMs toward a tumor-suppressive phenotype. Furthermore, we observed that CD24, a novel "don't-eat-me" signal, is positively regulated by FOXM1. Silencing FOXM1 significantly decreased CD24 mRNA and protein levels, whereas FOXM1 overexpression upregulated CD24 expression (Fig. 9, Fig. 9). This suggests that elevated FOXM1 expression enables tumor cells to evade macrophage-mediated phagocytosis by upregulating CD24. Collectively, these findings suggest that TST reprograms TAMs toward the M1 phenotype, enhances immune surveillance, and reshapes the immunosuppressive microenvironment by inhibiting FOXM1-mediated pathways.

Discussion

Despite the increasing incidence and mortality rates, only modest advancements have been made in the treatment of advanced cholangiocarcinoma until recently. Over the past decade, the combination of gemcitabine and cisplatin as a first-line chemotherapy regimen has shown limited efficacy, improving patients' progression-free survival (PFS) by only 8.0 months [38]. However, recent clinical trials have increasingly recognized the therapeutic promise of immunotherapy in ICC [39]. Here, we propose that TST has the ability to inhibit FOXM1 and exert “dual anti-tumor” effects, indicating its potential as a highly promising therapeutic agent for ICC treatment.

Previous research on FOXM1 has primarily focused on its direct oncogenic effects in tumors, while recent studies have increasingly recognized its significant role in promoting immune evasion and influencing immune cell infiltration [12,40]. However, despite TAMs being among the most crucial immunosuppressive cells, the relationship between FOXM1 and TAMs remains unclear. TST has demonstrated strong tumor-killing effects by targeting FOXM1 and has showed potential in improving the tumor immune microenvironment [27,41,42]. In this study, we discovered that TST could reprogram TAMs toward the M1 phenotype through FOXM1-mediated pathways. Specifically, FOXM1 inhibition in ICC reduced tumor cell-derived CCL2 secretion and increased GM-CSF levels. CCL2 promotes monocyte and macrophage recruitment and M2-like TAMs polarization, whereas GM-CSF induces monocyte differentiation into pro-inflammatory M1 macrophages [43,44]. These changes in cytokine levels directed TAMs toward an M1-like phenotype, enhanced pro-inflammatory cytokine secretion, and stimulated T cell recruitment, thereby boosting anti-tumor immunity. Furthermore, the downregulation of CD24 caused by FOXM1 inhibition further enhanced macrophage-mediated phagocytosis. Notably, we innovatively employed murine ICC cells for the construction of a mouse tumor model, which more accurately simulates the tumor microenvironment of ICC. Interestingly, although the ratio of CD4⁺/CD8⁺ T cells did not substantially increase, their absolute cell counts did rise within TME. However, the functional status of these T cells, as well as their crosstalk with TAMs in modulating the immune microenvironment, remains to be clarified. Additionally, the efficacy of combining TST with ICB in treating ICC is still under investigation.

In conclusion, we demonstrated that TST exerts "dual anti-tumor" effects in ICC by targeting FOXM1 (Fig. 10). On one hand, TST effectively inhibited the proliferation and metastasis of ICC cells. On the other hand, it reduced tumor-derived CCL2 while increasing GM-CSF, leading to the reprogramming of TAMs toward the M1 phenotype, which in turn promoted anti-tumor activity. Moreover, TST showed a favorable safety profile, providing a basis for its possible clinical applications. Currently, TST is being evaluated in a phase II clinical trial for the treatment of malignant pleural effusion. Therefore, our findings suggest that TST has the potential to become a novel immunotherapy drug for ICC patients.

Fig. 10.

Thiostrepton (TST) exerts "dual anti-tumor" effects in ICC by targeting FOXM1. On one hand, TST effectively inhibits the expression of FOXM, thereby arresting the cell cycle, promoting apoptosis, and suppressing the metastasis of tumor cells. On the other hand, TST-treated tumor cells secrete cytokines that induce TAMs to reprogram toward the M1 phenotype, thereby exerting an anti-tumor immune effect.

Consent for publication

Not applicable.

CRediT authorship contribution statement

Yu Li: Writing – original draft, Validation, Methodology, Investigation, Data curation. Yifan Jiang: Writing – review & editing, Methodology, Investigation, Conceptualization. Rongliang Tong: Visualization, Supervision, Methodology. Bo Ding: Validation, Software, Methodology. Jiangzhen Ge: Resources, Methodology. Keyi Du: Validation, Supervision. Jingqi Sun: Supervision, Methodology. Zheng Tang: Resources, Methodology. Diyu Chen: Writing – review & editing, Visualization, Funding acquisition, Conceptualization. Jian Wu: Supervision, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The supporting data for the study is available in the article's supplementary material.