Etoposide activates CD8+ T cell anti-tumor immunity in osteosarcoma through MHC I upregulation via tumor-secreted IL-33 mediated signaling Authors: Xin He, Hanjun Li, Haoyu Wang, Dongqing Zuo, Haoru Dong, Haoran Mu, Binghui Yang, Yining Tao, Xiyu Yang, Bowen Zhao, Tao Zhang, Yafei Jiang, Zhuoying Wang, Hongsheng Wang, Liu Yang, Yingqi Hua, Zhengdong Cai, Chongren Wang, Mengxiong Sun, Jun Gui, Wei Sun Published: 2025 ABSTRACT AbstractBackgroundOsteosarcoma patients with high propensity for metastasis and recurrence generally encounter a poor prognosis. Despite the extensive exploration of immunotherapy, particularly the anti-programmed cell death protein 1 (anti-PD-1) antibody, in clinical trials, the efficacy remains unsatisfactory. A more profound comprehension of the resistance mechanisms and the development of innovative therapeutic strategies is imperative.MethodsA screening was performed for drugs capable of upregulating major histocompatibility class I (MHC I) expression among clinically common drugs. The effects of the drug on both T cells and tumor cells, as well as its combination efficacy with anti-PD-1 antibody, were studied in vitro and in vivo osteosarcoma models. The molecular mechanisms underlying these biological processes were explored via RNA sequencing analysis.ResultsEtoposide was shown to upregulate the MHC I expression in osteosarcoma cells, thereby enhancing the cytotoxicity of CD8+ T cells. Interleukin-33 (IL-33) played a dominant role in etoposide-activated anti-tumor immune response. Etoposide promoted the secretion of IL-33 and augmented the expression of IL-33 binding suppression of tumorigenicity 2 (ST2) receptor, which activated the nuclear factor kappa-B signaling pathway and resulted in MHC I upregulation. Furthermore, etoposide was demonstrated to improve the therapeutic efficacy of anti-PD-1 antibody.ConclusionsThis study revealed the molecular mechanism underlying etoposide-activated CD8+ T cell anti-tumor immunity. The combination of Etoposide and anti-PD-1 antibody has the potential to benefit patients with advanced osteosarcoma.Keywords: Immunotherapy, Bone Cancer, Human leukocyte antigen - HLA, Major histocompatibility complex - MHC, T cell WHAT IS ALREADY KNOWN ON THIS TOPIC Patients with advanced osteosarcoma often suffer from tumor metastasis or recurrence. Repeated surgeries and chemotherapy resistance leave them without more effective therapeutic options. Several lines of evidence have been suggested to indicate the inefficacy of immunotherapy in clinical trials. Further investigation is necessary to explore the novel strategies for sensitizing immunotherapy. Patients with advanced osteosarcoma often suffer from tumor metastasis or recurrence. Repeated surgeries and chemotherapy resistance leave them without more effective therapeutic options. Several lines of evidence have been suggested to indicate the inefficacy of immunotherapy in clinical trials. Further investigation is necessary to explore the novel strategies for sensitizing immunotherapy. WHAT THIS STUDY ADDS Through a drug screening, we identified that etoposide was able to upregulate major histocompatibility class I (MHC I) expression and activate CD8+ T cell anti-tumor immunity in osteosarcoma. Etoposide stimulated tumor-secreted interleukin-33 (IL-33)-mediated NF-κB signaling pathway, thereby promoting MHC I upregulation. Moreover, etoposide improved the therapeutic efficacy of anti-programmed cell death protein 1 antibody in multiple osteosarcoma models. Through a drug screening, we identified that etoposide was able to upregulate major histocompatibility class I (MHC I) expression and activate CD8+ T cell anti-tumor immunity in osteosarcoma. Etoposide stimulated tumor-secreted interleukin-33 (IL-33)-mediated NF-κB signaling pathway, thereby promoting MHC I upregulation. Moreover, etoposide improved the therapeutic efficacy of anti-programmed cell death protein 1 antibody in multiple osteosarcoma models. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY This study demonstrates the potential of etoposide in sensitizing osteosarcoma immunotherapy, providing a novel therapeutic strategy to benefit osteosarcoma patients. The molecular mechanism underlying etoposide-activated CD8+ T cell anti-tumor immunity, as well as the role of IL-33 in the tumor immune microenvironment of osteosarcoma, has been elucidated. This study demonstrates the potential of etoposide in sensitizing osteosarcoma immunotherapy, providing a novel therapeutic strategy to benefit osteosarcoma patients. The molecular mechanism underlying etoposide-activated CD8+ T cell anti-tumor immunity, as well as the role of IL-33 in the tumor immune microenvironment of osteosarcoma, has been elucidated. Introduction Osteosarcoma, recognized as a highly malignant tumor, is the most prevalent primary bone tumor among children and young adults.1 Despite the 5-year survival rate for osteosarcoma being approximately 60%,2 osteosarcoma exhibits a high propensity for local invasion and early metastasis. It has been determined that approximately 10–15% of patients who are newly diagnosed present with tumor metastasis, with lung being the most common metastatic site.3 Tumor recurrence in osteosarcoma is reported to occur at a rate of approximately 30–50%.4 The presence of tumor metastasis and recurrence in patients is indicative of a poor prognosis, with a 5-year survival rate that is less than 20%.2 The primary treatment for osteosarcoma is surgical, followed by adjuvant chemotherapy administered before and after surgery. The most frequently used chemotherapy regimen is the MAP, which consists of high-dose methotrexate, doxorubicin, and cisplatin.3 However, conventional first-line treatments have proven ineffective for patients with advanced osteosarcoma. To benefit these patients, a variety of novel therapeutic strategies, including immunotherapy, have been proposed in recent years. The success of immunotherapy in the treatment of solid tumors has prompted researchers to investigate its potential application in osteosarcoma. However, due to the low immunogenicity and immunosuppressive microenvironment, immunotherapy, especially anti-programmed cell death protein 1 (PD-1) antibody, fails to achieve the predefined treatment goal in osteosarcoma.5 The osteosarcoma tumor microenvironment (TME) is a complex network of cells that are intimately associated with the occurrence and development of the tumor. Immune cells in the TME play a crucial role in the anti-tumor response.6 The osteosarcoma TME is comprised of various immune cells, including T cells, B cells, natural killer cells (NK cells), tumor-associated macrophages, myeloid-derived suppressor cells (MDSCs), and dendritic cells. Among these, T cell-mediated anti-tumor immune response is considered to be one of the most critical elements.7 In the context of a tumor, tumor antigen peptides are loaded and presented on the cell surface by major histocompatibility class I (MHC I). The MHC I complexes are then scanned by T cell receptors, which allow CD8+ T cells to recognize the tumor antigen. Following antigen recognition, CD8+ T cells would be activated, inducing the immune killing function.8 However, tumor cells have been demonstrated to evade immune surveillance and attack through the downregulation of MHC I. The mechanisms underlying this evasion primarily involve genetic defects in the antigen presentation pathway, transcriptional silencing, and post-transcriptional silencing of MHC I regulatory factors. Alterations in the MHC I-mediated tumor antigen presentation pathway have been shown to be associated with acquired immunotherapy resistance.9 Osteosarcoma, which originates from mesenchymal stem cells, is characterized by its high tumor antigen heterogeneity and low tumor immunogenicity.10 Moreover, it has been observed to exhibit a loss or downregulation of MHC I.11 12 These characteristics contribute to the immune evasion of osteosarcoma, thereby representing a significant factor in the immunotherapy resistance. Consequently, the upregulation of MHC I expression has emerged as a breakthrough in osteosarcoma immunotherapy. For better adaptation to clinical usage, a drug screening was performed on a series of clinically common drugs for osteosarcoma treatment. We identified that etoposide significantly upregulated the MHC I expression in osteosarcoma cells, resulting in enhanced CD8+ T cell anti-tumor immunity. Transcriptional analysis of T cells and non-immune cells sorted from osteosarcoma revealed that etoposide-stimulated interleukin-33 (IL-33) secretion from tumor cells, which activated nuclear factor kappa-B (NF-κB) signaling pathway, thereby promoting MHC I expression. Furthermore, IL-33 itself has been shown to activate CD8+ T cells to power the anti-tumor immune response. Etoposide combined with anti-PD-1 antibody notably blunts osteosarcoma progression, proposing a valuable strategy for osteosarcoma treatment. Methods Study cohort and human samples RNA sequencing data of Therapeutically Applicable Research to Generate Effective Treatments of Osteosarcoma (TARGET-OS) were downloaded from the Genomic Data Commons (https://portal.gdc.cancer.gov/). Our study adhered to the Declaration of Helsinki and was approved by the Institutional Research Ethics Committee of Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine. Peripheral blood samples were harvested from an osteosarcoma patient, from whom informed consent was obtained. The therapeutic regimens were etoposide at 500 mg/m2 and anti-PD-1 antibody tislelizumab at 200 mg Q28D. Peripheral blood mononuclear cells (PBMCs) were isolated from peripheral blood samples by Ficoll-Paque PLUS density gradient media (Cytiva, cat#17144002) and stored at −80℃ for fluorescence-activated cell sorting (FACS) analysis. Cell lines The human osteosarcoma cell lines HOS (ATCC, cat#CRL-1543), MNNG-HOS (ATCC, cat#CRL-1547), 143B (ATCC, cat#CRL-8303), SJSA-1 (ATCC, cat#CRL-2098), U-2 OS (ATCC, cat#HTB-96), MG-63 (ATCC, cat#CRL-1427), and the mouse osteosarcoma cell lines K7 (Dr. Kleinerman’s lab of MD Anderson Cancer Center), K7M2 (ATCC, cat#CRL-2837) were cultured in high glucose Dulbecco’s Modified Eagle Medium (Gibco, cat#11965092). All cell lines were cultured in the media supplemented with 10% fetal bovine serum (WISENT, cat#085–450) and 100 U/mL Penicillin Streptomycin (Gibco, cat#15140122) at 37℃ under 5% CO2. Cell line authentication was performed on all of the cell lines using short tandem repeat DNA profiling. All cell lines were preserved at the Shanghai Bone Tumor Institute and tested for mycoplasma-free regularly. Mice and in vivo treatments 6-week-old female BALB/c mice (BALB/cAnNCtrl), obtained from Laboratory Animal Center of Shanghai General Hospital, were used in this study. Mice were housed under specific pathogen-free conditions and in vivo treatments were performed by following the protocols in accordance with the guidelines of the Laboratory Animal Center of Shanghai General Hospital. The Clinical Center Laboratory Animal Welfare and Ethics Committee of Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine approved used protocols in this study (No. 2024AW014). For syngeneic transplanted osteosarcoma models, 5×106 K7M2 in 50 µL phosphate-buffered saline (PBS) (Biotend, cat#RE100001) were intramedullary injected into the marrow space of the proximal tibial of BALB/c mice. Tumor progression was monitored by vernier caliper measuring the tumor volume, calculated by the following equation: volume =0.2618 × (width × length × (width + length)). Tumor burden was monitored by following the tumor volume, which was permitted by the Clinical Center Laboratory Animal Welfare and Ethics Committee of Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine. For etoposide treatment, 8 mg/kg Etoposide (Selleck, cat#S1225) was administered intraperitoneal injection (i.p). every 2 days starting from the day tumors were palpable. For CD8 depletion, 100 µg/mouse of anti-mouse CD8α-InVivo (Selleck, cat#A2102) or Rat IgG2b isotype control-InVivo (Selleck, cat#A2116) were administered i.p every 4 days starting 1 day prior to tumor implantation. For anti-PD-1 treatment, 10 mg/kg of InVivoMab anti-mouse PD-1 (CD279) (BioXCell, cat#BE0146) was administered i.p every 3 days starting from the next day of etoposide chemotherapy. For anti-IL-33 treatment, 2 µg/mouse of Mouse IL-33 Antibody (R&D, cat#AF3626) was administered intravenously every 3 days starting 1 day prior to etoposide chemotherapy. For IL-33 treatment, 1 µg/mouse of Mouse IL-33 Recombinant Protein (PeproTech, cat#210–33) was administered i.t every 2 days starting from the next day of etoposide chemotherapy. Cell viability assay and drugs screening Cells were plated onto 96-well plates (Corning, cat#3798) at a density of 5000 cells per well, then treated with or without drugs on different concentrations in complete medium for 24 hours. Drugs include etoposide (Selleck, cat#S1225), gemcitabine (Selleck, cat#S1714), methotrexate (Selleck, cat#S1210), topotecan (Selleck, cat#S9321), anlotinib (Selleck, cat#S8726), pazopanib (Selleck, cat#S3012), apatinib (Selleck, cat#S5248), regorafenib (Selleck, cat#S1178), sorafenib (Selleck, cat#S7397), rapamycin (Selleck, cat#S1039), everolimus (Selleck, cat#S1120) and cabozantinib (Selleck, cat#S1119). Cell viability was measured by Cell Counting Kit-8 (NCM; cat#C6005) and detected by SpectraMax M3 Microplate Reader (Molecular Devices). Inhibitory concentration (IC50) values were determined through curve fitting. For drug screening, cells that were around 70–80% confluent were treated with drugs at a dose of IC50 for 24 hours. Then cells were harvested for FACS analysis. CD8+ T cell cytotoxicity assay Naïve CD8+ T cells from splenocytes of BALB/c mice were isolated by EasySep Mouse Naïve CD8+ T Cell Isolation Kit (STEMCELL, cat#19858) and were activated 72 hours by complete RPMI-1640 Medium (Gibco, cat#11875093) containing 5 µg/mL Ultra-LEAF Purified anti-mouse CD28 Antibody (BioLegend, cat#102116) and 1 µL/10 ml mouse IL-2 Recombinant Protein (PeproTech, cat#212–12) in 96-well plates which were pretreated with 10 µg/mL Ultra-LEAF Purified anti-mouse CD3ε Antibody (BioLegend, cat#100340). For IL-33 treatment, 200 ng/mL Mouse IL-33 Recombinant Protein (PeproTech, cat#210–33) was treated with T cell activation. K7M2 was pretreated with etoposide for 24 hours. For MHC I blockade, InVivoMAb anti-mouse MHC Class I (H-2) (BioXCell, cat#BE0077) was added 2 hours before etoposide treatment. For IL-33 blockade, 200 ng/mL Mouse IL-33 Antibody (R&D, cat#AF3626) was added 2 hours before etoposide treatment. Activated CD8+ T cells were co-cultured with K7M2 at the 1:1 ratio for 24 hours. K7M2 was harvested and detected by Annexin V-FITC/7-AAD Apoptosis Detection Kit (YEASEN, cat#40 311ES20). FACS analysis For analysis of cell surface markers, cells were harvested by Cell Dissociation Buffer Enzyme-Free (Gibco, cat#13151014) and incubated with Purified anti-mouse CD16/32 Antibody (BioLegend, cat#101302) on ice for 15 min to block non-specific receptors binding. Then cells were stained with phycoerythrin (PE) anti-mouse H-2Kd Antibody (BioLegend, cat#116607) and PE Mouse IgG2a, κ Isotype Ctrl Antibody (BioLegend, cat#400211), PE anti-mouse IL-33Rα Antibody (BioLegend, cat#145303) and PE Rat IgG2a, κ Isotype Ctrl Antibody (BioLegend, cat#400507) on ice for 30 min. After washing with PBS, cells were acquired by C6 Plus Flow Cytometer (BD) and data were analyzed with FlowJo V.10.8.1 (BD). For human cells, Human TruStain FcX (Fc Receptor Blocking Solution) (BioLegend, cat#422301), PE anti-human human leukocyte antigen (HLA)-A,B,C Antibody (BioLegend, cat#311405) and PE Mouse IgG2a, κ Isotype Ctrl Antibody (BioLegend, cat#400213) were used instead. For IL-33 blockade, 200 ng/mL Mouse IL-33 Antibody (R&D, cat#AF3626) was added 2 hours before etoposide treatment. For NF-κB blockade, 10 µM BAY 11–7082 (MCE, cat#HY-13453) was added 2 hours before etoposide treatment. For analysis of tumor-infiltrating T cells, tumors were harvested and dissociated mechanically and enzymatically in RPMI-1640 medium containing 1 mg/mL Collagenase IV (Gibco, cat#17104019) and 100 µg/mL DNase I (Invitrogen, cat#18047019) with continuous agitation at 37℃ for 1 hour. The digestion mixture was passed through 70 µm cell strainers (Beyotime, cat#FSTR072), lysed with Red Blood Cell Lysis Buffer (Beyotime, cat#C3702), and then incubated with Purified anti-mouse CD16/32 Antibody (BioLegend, cat#101302) on ice for 15 min to block non-specific receptors binding. Cells were stained with FITC anti-mouse CD45 Antibody (BioLegend, cat#103107), PerCP/Cyanine 5.5 anti-mouse CD3 Antibody (BioLegend, cat#100217), APC anti-mouse CD4 Antibody (BioLegend, cat#100411), PE anti-mouse CD8b Antibody (BioLegend, cat#126607) on ice for 30 min. After washing with PBS, cells were acquired by LSRFortessa Flow Cytometer (BD) and data were analyzed with FlowJo. For analysis of T cells cytotoxic molecules, cells were preincubated in complete medium containing Cell Activation Cocktail (with Brefeldin A) (BioLegend, cat#423303) for 6 hours. Cell surface markers were stained with FITC anti-mouse CD45 Antibody (BioLegend, cat#103107), APC anti-mouse CD3 Antibody (BioLegend, cat#100235), PE anti-mouse CD8b Antibody (BioLegend, cat#126607) on ice for 30 min. Then cells were washed with PBS and fixed with Intracellular Fixation & Permeabilization Buffer (eBioscience, cat#88-8823-88) for 60 min. Cell intracellular markers were stained with Pacific Blue anti-mouse Perforin Antibody (BioLegend, cat#154311), Alexa Fluor 700 anti-mouse interferon (IFN)-γ Antibody (BioLegend, cat#505823), PerCP anti-human/mouse Granzyme B Recombinant Antibody (BioLegend, cat#396415), Brilliant Violet 605 anti-mouse tumor necrosis factor (TNF)-α Antibody (BioLegend, cat#506329) in permeabilization buffer for 60 min. After washing with permeabilization buffer, cells were acquired by LSRFortessa Flow Cytometer (BD) and data were analyzed with FlowJo. For human samples, FITC anti-human CD3 Antibody (BioLegend, cat#317305), APC anti-human CD4 Antibody (BioLegend, cat#300514), PE anti-human CD8 Antibody (BioLegend, cat#344705) and PerCP anti-human/mouse Granzyme B Recombinant Antibody (BioLegend, cat#396415) were used instead. RNA sequencing For cells RNA collection, K7M2 tumor-bearing BALB/c mice with vehicle or etoposide treatment were executed. The single-cell suspension of tumor tissues was prepared as described above. After staining with FITC anti-mouse CD45 Antibody (BioLegend, cat#103107), APC anti-mouse/human CD11b Antibody (BioLegend, cat#101211) and PE anti-mouse CD3 Antibody (BioLegend, cat#100205), the CD45+CD11b-CD3+ T cells and CD45-CD11b-CD3- non-immune cells were sorted out by FACSAria III Cell Sorter (BD). RNA was extracted from sorted cells by Trizol (Beyotime, cat#R0016). For RNA sequencing, Sinotech Genomics performed RNA sequencing. Total RNA was tested for integrity by Agilent Bioanalyzer 2100 (Agilent Technologies). Concentration and purity were assessed by Qubit 3.0 Fluorometer (Life Technologies) and Nanodrop One spectrophotometer (Thermo Fisher Scientific). Sequencing was performed using NovaSeq 6000 (Illumina) according to the manufacturer’s protocol. The raw data were first processed with FastQC to filter out adapters and low-quality sequences. Pair-end reads were aligned to mouse mm10 genome using STAR (V.2.7.6a). Reads with good mapping quality (>30) that aligned to genomic exons were counted using featureCounts (mm10 Ensembl 93) to generate a table with counts for each gene. Differential gene expression was identified with logarithm base two-fold changes of expression (log2FC) ≥2 and false discovery rate (FDR) <0.05. The gene functional enrichment was performed by gene set enrichment analysis (GSEA) (http://www.gsea-msigdb.org/). The enrichment magnitude and statistical significance were quantified by Normalized Enrichment Score and FDR, respectively. Reverse transcription-PCR Total RNA was extracted by Trizol (Beyotime, cat#R0016) according to the manufacturer’s instructions and reversed transcribed to cDNA by HiScript III All-in-one reverse transcription (RT) SuperMix Perfect for qPCR (Vazyme, cat#R333-01). RT-PCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, cat#Q711-03). The expression of each gene was calculated based on the cycle threshold which was set within the linear range of DNA amplification. The relative expression was calculated by the cycle threshold method, with normalization of raw data to a housekeeping gene. The sequences of the primers used were as follows (table 1). Table 1Primer sequences for RT-PCR.GeneFrontReverseIL-33CTACTGCATGAGACTCCGTTCTGAGAATCCCGTGGATAGGCAGAGST2TGTATTTGACAGTTACGGAGGGCACTTCAGACGATCTCTTGAGACACXCL9GGAGTTCGAGGAACCCTAGTGGGGATTTGTAGTGGATCGTGCCXCL10CCAAGTGCTGCCGTCATTTTCGGCTCGCAGGGATGATTTCAAGAPDHGTCTCCTCTGACTTCAACAGCGACCACCCTGTTGCTGTAGCCAACXCL, C-X-C motif chemokine ligand; GAPDH, glyceraldehyde -3-phosphate dehydrogenase; IL-33, interleukin-33; ST2, suppression of tumorigenicity 2. Primer sequences for RT-PCR. CXCL, C-X-C motif chemokine ligand; GAPDH, glyceraldehyde -3-phosphate dehydrogenase; IL-33, interleukin-33; ST2, suppression of tumorigenicity 2. Western blot Protein was extracted by RIPA Lysis Buffer (Beyotime, cat#P0013B) added with Protease and phosphatase inhibitor cocktail (Beyotime, cat#P1050). Protein concentration was quantified by BCA Protein Assay Kit (Beyotime, cat#P0012). Lysate with SDS-PAGE Sample Loading Buffer (Beyotime, cat#P0015) was boiled for 10 min and loaded on SDS-PAGE gels (Epizyme, cat#PG112). Protein was electrophoresed and transferred to 0.45 µm polyvinylidene fluoride membranes (Millipore, cat#IPFL00010). After being blocked with 5% non-fat milk tris buffered saline with tween-20 for 1 hour, membranes were incubated with the primary and secondary antibodies. After that, blots were developed by ECL reagent with ChemiDoc Touch Imaging System (BIO-RAD). The image was subjected to gray value analysis using ImageJ. Antibodies we used were as follows: IL-33 Polyclonal antibody (Proteintech, cat#12372–1-AP), Anti-beta Actin antibody (Abcam, cat#ab8226), NF-κB p65 Rabbit mAb (CST, cat#8242), Phospho-NF-κB p65 Rabbit mAb (CST, cat#3033), HRP-conjugated Goat Anti-Rabbit IgG(H+L) (Proteintech, cat#SA00001-2). Chromatin immunoprecipitation sequencing analysis For chromatin iImmunoprecipitation sequencing (ChIP-seq) library preparation, live K7M2 cells treated with etoposide or vehicle were processed in accordance with standard protocols by Sinotech Genomics, where sequencing was performed. To visualize the sequencing data, normalized signal tracks in BigWig format were generated and loaded into the Integrative Genomics Viewer (IGV, Broad Institute) along with the mouse reference genome (mm10). Enrichment profiles across specific genomic regions—including the H2-K1 and H2-D1 loci—were compared between experimental groups. Signal intensity, representing read coverage and binding enrichment, was displayed to illustrate differences in chromatin association among samples. All IGV snapshots were exported using consistent scaling and resolution settings to ensure comparability across conditions. Transient degradation small interfering RNA delivery The small interfering RNA (siRNA) for ST2, which was obtained from QEgene, was diluted in Opti-MEM (Gibco, cat#31985070). Lipofectamine 3000 Transfection Reagent (Invitrogen, cat#L3000075) was diluted in Opti-MEM. Cells that were around 30–50% confluent were incubated with the complex of siRNA and transfection reagent for 24 hours. The silencing efficiency of the siRNA was validated by RT-PCR and FACS analysis. For etoposide and IL-33 treatment, when transfection finished, cells’ transfection medium was replaced by complete medium with vehicle, 10 µM etoposide (Selleck, cat#S1225) or 200 ng/mL Mouse IL-33 Recombinant Protein (PeproTech, cat#210–33). After being treated for 24 hours, cells were harvested for western blot analysis. Quantification and statistical analysis Statistical analysis was carried out using Prism V.9 (GraphPad). Two-tailed Student’s t-test was applied to determine the significance of differences between two groups of independent samples. A one-way analysis of variance test was applied to determine the significance of differences between three or more groups of independent samples. Spearman’s correlation analysis was performed to determine the correlation between two groups of variables. The Kaplan-Meier survival curves compared by two-sided log-rank test were performed. Schematic diagrams were created using BioRender (BioRender). Results Etoposide upregulates MHC I expression in osteosarcoma cells To evaluate the association between MHC I expression and clinical outcomes in osteosarcoma, we analyzed transcriptomic and genomic data from the Shanghai General Hospital-Osteosarcoma (SGH-OS) cohort (n=91).13 Patients with low expression of HLA (n=23) demonstrated a reduced survival compared with those with high HLA expression (n=68) (figure 1A). Analysis of the TARGET-OS database further revealed a positive correlation between immune score and expression of HLA-A, HLA-B, and HLA-C (figure 1B). Furthermore, T cell infiltration levels were also positively associated with HLA-A/B/C expression (figure 1C). Together, these data suggest that low MHC I expression is associated with immunosuppression phenotype and unfavorable prognosis in osteosarcoma patients. Etoposide upregulates MHC I expression in osteosarcoma cells. (A) Analysis of the survival probability of osteosarcoma patients classified by HLA expression in SGH-OS cohort. (B) Analysis of the correlation between immune score and HLA-A, HLA-B, HLA-C expression in TARGET-OS. (C) Analysis of the correlation between T cells and HLA-A, HLA-B, HLA-C expression in TARGET-OS. (D) Schematic illustration of drug screening. Detecting the toxicity of clinically commonly used drugs and calculating IC50 in K7M2 and 143B osteosarcoma cell lines. MHC I expression was detected by FACS after cells were treated for 24 hours at the dose of IC50. The top one drug was selected. (E) Cell viability of K7M2 and 143B treated with etoposide by Cell Counting Kit-8 assay. Data are presented as mean±SD (n=6). (F) Relative MHC I expression in K7M2 and 143B treated with drugs. Data are presented as mean±SD (n>7). (G) Representative FACS analysis and quantification of MHC I expression in K7M2 and 143B treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=3/4). (H) Tumor growth curve and weight of K7M2 tumor-bearing mice treated with vehicle or etoposide (i.p, 8 mg/kg, every 2 days). Data are presented as mean±SD (n=4 mice per group). (I) Representative FACS analysis and quantification of MHC I expression in mice treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=4 mice per group). *, p<0.05; ***, p<0.001; ****, p<0.0001. FACS, fluorescence-activated cell sorting; HLA, human leukocyte antigen; IC50, inhibitory concentration; i.p, intraperitoneal injection; MFI, mean fluorescence intensity; MHC I, major histocompatibility class I; SGH-OS, Shanghai General Hospital-Osteosarcoma; TARGET-OS, Therapeutically Applicable Research to Generate Effective Treatments of Osteosarcoma. To identify representative cellular models for studying low HLA-expressing osteosarcoma, we evaluated MHC I levels in commonly used human and mouse osteosarcoma cell lines. Among these, the 143B human and K7M2 mouse cell lines exhibited relatively low basal MHC I expression (online supplemental figure S1A), and both are associated with highly malignant phenotype.14 We next screened 12 drugs commonly used in osteosarcoma treatment for their ability to upregulate MHC I expression (figure 1D). The toxicity was assessed in 143B and K7M2 cells to determine the half-maximal IC50 for each compound (figure 1E and online supplemental figure S1B). Subsequently, cells were treated at the IC50 dose for 24 hours, and MHC I expression was measured. The results indicated that etoposide emerged as the most effective agent in upregulation of MHC I expression in both cell lines (figure 1F,G). To evaluate the anti-tumor effect of etoposide in vivo, we treated K7M2 tumor-bearing mice (online supplemental figure S2A) and observed significant suppression of tumor growth (figure 1H, online supplemental figure S2B). Consistent with the in vitro findings, etoposide also enhanced MHC I expression in tumor tissues (figure 1I). Etoposide enhances the cytotoxicity of CD8+ T cells To assess the effect of etoposide on the tumor immune microenvironment of osteosarcoma, we analyzed tumor-infiltrating lymphocytes from K7M2 tumor-bearing mice. Etoposide treatment significantly increased the proportions of total T cells (CD3+) as well as CD4+ and CD8+ T cell subsets within tumors (figure 2A). Moreover, CD8+ T cells from treated mice produced higher levels of the cytotoxic mediators perforin and TNF-α (figure 2B). In contrast, other immune populations such as monocytes and neutrophils showed no significant changes in infiltration (online supplemental figure S2C). These data suggest that etoposide enhances anti-tumor immunity primarily by promoting CD8+ T cell infiltration and function. Etoposide enhances the cytotoxicity of CD8+ T cells. (A) Representative FACS analysis and quantification of the percentage of tumor-infiltrating CD3+, CD4,+ and CD8+ T cells in mice treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=4 mice per group). (B) Representative FACS analysis and quantification of the percentage of perforin and TNF-α in tumor-infiltrating CD8+ T cells in mice treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=4 mice per group). (C) Schematic illustration of analysis of T cell cytotoxicity to K7M2. (D) Representative FACS analysis and quantification of apoptosis percentage of K7M2 pretreated with vehicle, etoposide, or etoposide added with anti-MHC I antibody and co-cultured with activated T cells. Quantitative data are presented as mean±SD (n=4). (E) Tumor growth curve and weight of K7M2 tumor-bearing mice treated with vehicle, etoposide, or etoposide added with anti-CD8α antibody (i.p, 100 µg/mouse, every 4 days). Data are presented as mean±SD (n=5 mice per group). ns, p>0.05; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. 7-AAD, 7-Aminoactinomycin D; FACS, fluorescence-activated cell sorting; i.p, intraperitoneal injection; MHC I, major histocompatibility class I; TNF, tumor necrosis factor. The activation of naïve CD8+ T cells depends on the recognition of tumor antigens presented by MHC I.15 To investigate whether etoposide enhances CD8+ T cell cytotoxicity via MHC I upregulation, we performed a co-culture assay with T cells and K7M2 cells that were pre-treated with or without etoposide (figure 2C). Intriguingly, etoposide-pre-treated tumor cells were more effectively killed by T cells. However, this enhanced cytotoxicity was diminished when MHC I was blocked on the tumor cells (figure 2D). These findings demonstrate that etoposide-induced sensitization of tumor cells to T cell attack is MHC I-dependent. To further establish the functional contribution of CD8+ T cells in etoposide-mediated anti-tumor activity, we depleted these cells using anti-CD8α antibody prior to tumor implantation (online supplemental figure S2D). CD8+ T cell depletion significantly abrogated the anti-tumor effect of etoposide (figure 2E and online supplemental figure S2E,F), confirming that CD8+ T cells are essential for treatment efficacy. Collectively, these results indicate that etoposide potentiates CD8+ T cell-mediated anti-tumor immunity in osteosarcoma through a mechanism that depends on MHC I upregulation. Etoposide-stimulated IL-33 secretion promotes CD8+ T cell activity To characterize the transcriptional changes within the osteosarcoma TME, we isolated T cells and non-immune cells from K7M2 tumor-bearing mice by flow cytometry and performed RNA sequencing on these populations (online supplemental figure S3A). In tumor-infiltrating T cells, etoposide treatment led to a marked upregulation of genes encoding cytotoxic molecules (figure 3A). In non-immune cells, GSEA revealed heightened cytokine activity and elevated transcriptional activity (figure 3B,C). Correspondingly, the expression of Cxcl9 and Cxcl10, cytokines involved in T cell recruitment, was increased (figure 3C and online supplemental figure S3B). These results imply a pivotal role for cytokine signaling in the etoposide-driven immune response. Etoposide-stimulated IL-33 secretion promotes CD8+ T cell activity. (A) Volcano plot of differentially expressed genes for tumor-infiltrating T cells sorted from tumor-bearing mice after vehicle or etoposide treatment. (B) GSEA analysis of enrichment of cytokine activity in the non-immune cells sorted from K7M2 tumor-bearing mice treated with vehicle or etoposide. (C) Heatmap of cytokines in the non-immune cells sorted from K7M2 tumor-bearing mice treated with vehicle or etoposide. (D) RT-PCR analysis of IL-33 in K7M2 treated with vehicle or etoposide. Data are presented as mean±SD (n=4). (E) Western blot and quantification of IL-33 in K7M2 treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=3). (F) Representative FACS analysis and quantification of MHC I expression in K7M2 treated with vehicle, etoposide, or etoposide added with anti-IL-33 antibody (200 ng/mL). Quantitative data are presented as mean±SD (n=4). (G) Representative FACS analysis and quantification of apoptosis percentage of K7M2 pretreated with vehicle, etoposide, or etoposide added with anti-IL-33 antibody and then co-cultured with activated T cells. Quantitative data are presented as mean±SD (n=4). (H) Tumor growth curve and weight of K7M2 tumor-bearing mice treated with vehicle, etoposide, or etoposide added with anti-IL-33 antibody (intravenous, 2 µg/mouse, every 3 days). Data are presented as mean±SD (n=4 mice per group). (I) Representative FACS analysis and quantification of the percentage of granzyme B and IFN-γ in activated CD8+ T cells treated with IL-33. Quantitative data are presented as mean±SD (n=5). (J) Representative FACS analysis and quantification of apoptosis percentage of K7M2 co-cultured with activated T cells which were pretreated with vehicle or IL-33 (200 ng/mL). Quantitative data are presented as mean±SD (n=4). (K) Representative FACS analysis and quantification of the percentage of tumor-infiltrating CD8+ T cells in mice treated with vehicle, etoposide, IL-33 (i.t, 1 µg/mouse, every 2 days), etoposide added with IL-33 or etoposide added with anti-IL-33 antibody. Quantitative data are presented as mean±SD (n=4 mice per group). (L) Representative FACS analysis and quantification of the percentage of granzyme B and TNF-α in tumor-infiltrating CD8+ T cells in mice treated with vehicle, etoposide, IL-33, etoposide added with IL-33, or etoposide added with anti-IL-33 antibody. Quantitative data are presented as mean±SD (n=4 mice per group). ns, p>0.05; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. 7-AAD, 7-Aminoactinomycin D; FACS, fluorescence-activated cell sorting; GSEA, gene set enrichment analysis; i.t, intratumoral injection; IFN, interferon; IL-33, interleukin-33; MFI, mean fluorescence intensity; MHC I, major histocompatibility class I; RT-PCR, reverse transcription PCR; TNF, tumor necrosis factor. Transcriptional profiling identified IL-33 as one of the most strongly upregulated cytokines (figure 3C). This increase was confirmed at the mRNA and protein levels by RT-PCR and western blot, respectively (figure 3D,E). To evaluate the functional contribution of IL-33 to etoposide-induced MHC I upregulation, we treated K7M2 cells with etoposide in the presence of a neutralizing anti-IL-33 antibody. IL-33 blockade significantly suppressed the enhancement of MHC I expression by etoposide (figure 3F). We further examined whether IL-33 is required for the enhancement of CD8+ T cell cytotoxicity by etoposide. Using the co-culture killing assay, neutralization of IL-33 markedly impaired the ability of etoposide to promote T cell-mediated tumor killing (figure 3G). Consistently, the anti-tumor effect of etoposide in vivo was substantially reversed on IL-33 blockade (figure 3H and online supplemental figure S3C,D). Previous studies have established that IL-33 directly promotes T cell immune responses.16 To define its specific impact on CD8+ T cells, we treated them with escalating concentrations of recombinant IL-33. We found that IL-33 induced a dose-dependent increase in the production of granzyme B and IFN-γ (figure 3I). Accordingly, IL-33-stimulated CD8+ T cells exhibited significantly enhanced tumor-killing capacity (figure 3J). We further explored the potential role of IL-33 in the tumor immune microenvironment of osteosarcoma. Since IL-33 mediates its biological effects primarily through its receptor ST2,17 we first assessed the expression of ST2 on the surface of different subtypes of tumor-infiltrating immune cells. For CD8+ T cells, we found that CD8+ T cells exhibited low surface expression of ST2 under baseline conditions; however, following etoposide treatment, ST2 expression was markedly upregulated (online supplemental figure S3E). This finding implies that IL-33 signaling may act on CD8+ T cells in osteosarcoma TME after etoposide treatment. We subsequently evaluated the infiltration and functional status of CD8+ T cells in the TME. The results demonstrated that both etoposide and IL-33 monotherapy increased the percentage of tumor-infiltrating CD8+ T cells to a similar extent, while their combination led to an additive enhancement. Notably, the pro-infiltrative effect of etoposide was abolished by anti-IL-33 antibody blockade (figure 3K). With respect to CD8+ T cell cytotoxicity, similar trends were observed: both etoposide and IL-33 individually enhanced the expression of cytotoxic molecules, including granzyme B and TNF-α, and their combination showed a synergistic effect. Importantly, the enhancement of CD8+ T cell function by etoposide was again reversed on IL-33 neutralization (figure 3L). Beyond T cells, we also examined the potential role of IL-33 in myeloid cells. We observed detectable surface expression of the ST2 receptor on macrophages, which was significantly upregulated on etoposide treatment as well (online supplemental figure S3F). The per cent of tumor-associated macrophages was increased in all the treated groups (online supplemental figure S3G). Moreover, the expression of immune-active markers, including CD80 and CD86, was elevated on treatment with etoposide or IL-33 individually or in combination, and these effects were abolished by IL-33 blockade. Conversely, IL-33 alone or in combination with etoposide led to reduced expression of CD206, which is a marker associated with M2-like macrophages. Notably, in the presence of etoposide and anti-IL-33 antibody, CD206 expression was markedly increased (online supplemental figure S3H). Collectively, these results suggest that etoposide-IL-33 signaling promotes macrophage reprogramming toward an M1-like phenotype. Analysis of MDSCs showed that the ST2 receptor was expressed on both monocytes and neutrophils, and its expression level was not altered by etoposide treatment (online supplemental figure S3I). Flow cytometry analysis revealed that the proportions of tumor-infiltrating monocytes and neutrophils remained equivalent across all treatment conditions: either etoposide or IL-33 was administered individually or in combination, or when etoposide was co-administered with anti-IL-33 antibody (online supplemental figure S3J). These data indicate that etoposide-IL-33 axis may not function on MDSCs. Taken together, these findings suggest that etoposide-triggered IL-33 release reshapes tumor immune microenvironment of osteosarcoma, which not only directly potentiates CD8+ T cell activity, but also promotes M1-like macrophage polarization, thereby fostering anti-tumor immunity. Etoposide-promoted MHC I expression through IL-33-activated NF-κB signaling pathway To explore the signaling mechanisms by which IL-33 contributes to MHC I upregulation, we interrogated the RNA sequencing data. GSEA analysis revealed that a significant enrichment of NF-κB signaling pathway in etoposide treatment group when compared with vehicle control (figure 4A). We then performed a more detailed differential expression analysis focusing on key NF-κB-related genes. The results showed that most of these genes, including H2-K1 and H2-D1, were significantly upregulated on etoposide treatment (online supplemental figure S4A). Consistently, flow cytometric analysis demonstrated that treatment with NF-κB inhibitor abolished etoposide-induced upregulation of MHC I (figure 4B). To determine whether NF-κB signaling directly regulates MHC I genes in osteosarcoma cells, we conducted ChIP-seq analysis of the p65 subunit. The results revealed enhanced binding of NF-κB p65 to the promoter regions of H2-K1 and H2-D1 following etoposide exposure (figure 4C). Furthermore, analysis of HLA-related regulatory pathways in the SGH-OS cohort indicated strong enrichment of NF-κB signaling activity (online supplemental figure S4B). Quantitative evaluation of NF-κB signaling pathway score demonstrated a significant positive correlation with the expression of HLA-A, HLA-B, and HLA-C (figure 4D). Etoposide-promoted MHC I expression through IL-33-activated NF-κB signaling pathway. (A) GSEA analysis of enrichment of NF-κB signal pathway in the non-immune cells sorted from K7M2 tumor-bearing mice treated with vehicle or etoposide. (B) Representative FACS analysis and quantification of MHC I expression in K7M2 treated with vehicle, etoposide, or etoposide added with NF-κB inhibitor (10 µM). Quantitative data are presented as mean±SD (n=3). (C) ChIP-seq tracks of NF-κB p65 binding at the H2-K1 and H2-D1 genomic regions in K7M2 treated with vehicle or etoposide. (D) Analysis of the correlation between pathway score of NF-κB signaling and HLA-A, HLA-B, HLA-C expression in SGH-OS cohort. (E) Western blot and quantification of p-p65/p65 in K7M2 treated with vehicle, etoposide, or etoposide added with anti-IL-33 antibody. Quantitative data are presented as mean±SD (n=3). (F) Representative FACS analysis and quantification of ST2 expression in K7M2 treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=5). (G) Western blot and quantification of p-p65/p65 in WT and ST2-knockdown K7M2 tumor cells treated with vehicle or etoposide. Quantitative data are presented as mean±SD (n=3). (H) Western blot and quantification of p-p65/p65 in WT and ST2-knockdown K7M2 tumor cells treated with vehicle or IL-33. Quantitative data are presented as mean±SD (n=3). **, p<0.01; ***, p<0.001; ****, p<0.0001. ChIP-seq, chromatin iImmunoprecipitation sequencing; FACS, fluorescence-activated cell sorting; GOBP, gene ontology biological process; GSEA, gene set enrichment analysis; HLA, human leukocyte antigen; IL-33, interleukin-33; MFI, mean fluorescence intensity; MHC I, major histocompatibility class I; NF-κB, nuclear factor kappa-B; ST2, suppression of tumorigenicity 2; si-ST2, small interfering RNA of suppression of tumorigenicity 2; WT, wild-type. We next investigated whether etoposide activated NF-κB signaling in osteosarcoma cells via IL-33. Western blot analysis confirmed that etoposide strongly induced phosphorylation of p65, indicating NF-κB activation. Importantly, this effect was substantially attenuated on IL-33 neutralization (figure 4E). Of note, we observed that ST2 was expressed on tumor cells, which was significantly upregulated following etoposide treatment (figure 4F). We then knocked down ST2 using siRNA (online supplemental figure S4C,D). In contrast to wild-type (WT) cells, etoposide failed to activate NF-κB signaling in ST2-knockdown cells (figure 4G). Similarly, recombinant IL-33 activated NF-κB signaling in WT cells, but not in ST2-knockdown cells (figure 4H). In summary, these findings indicate that etoposide upregulates MHC I expression through IL-33-ST2-NF-κB signaling axis. Etoposide improves the therapeutic efficacy of anti-PD-1 therapy in osteosarcoma Current clinical evidence indicates that the majority of immune checkpoint inhibitors have shown limited efficacy in osteosarcoma patients.4 In this study, we demonstrated that etoposide upregulates MHC I expression and activates CD8+ T cell-mediated anti-tumor immunity in osteosarcoma. Based on these findings, we speculated that etoposide has the potential to improve the therapeutic response to immune checkpoint inhibitors, such as anti-PD-1 antibody. An in vivo test was performed to evaluate the effect of anti-PD-1 antibody, both alone and in combination with etoposide, on osteosarcoma growth (online supplemental figure S5A). While anti-PD-1 monotherapy showed minimal tumor suppression, the combination with etoposide significantly inhibited tumor progression (figure 5A and online supplemental figure S5B). Consistent with the proposed mechanism, MHC I expression was markedly elevated in both the etoposide alone and combination groups (figure 5B). We further characterized the tumor immune microenvironment following treatment. The proportion of tumor-infiltrating CD8+ T cells was significantly increased across all treatment groups (figure 5C). Moreover, CD8+ T cells in the combination treatment group exhibited enhanced cytotoxic function, as evidenced by elevated production of perforin and IFN-γ (figure 5D). Together, these data suggest that etoposide sensitizes osteosarcoma to anti-PD-1 therapy. Etoposide improves the therapeutic efficacy of anti-PD-1 antibody in osteosarcoma. (A) Tumor growth curve and weight of K7M2 tumor-bearing mice treated with vehicle, etoposide, anti-PD-1 antibody (i.p, 10 mg/kg, every 3 days) or etoposide combined with anti-PD-1 antibody. Data are presented as mean±SD (n=4 mice per group). (B) Representative FACS analysis and quantification of MHC I expression in mice treated with vehicle, etoposide, anti-PD-1 antibody, or etoposide combined with anti-PD-1 antibody. Quantitative data are presented as mean±SD (n=4 mice per group). (C) Representative FACS analysis and quantification of the percentage of tumor-infiltrating CD8+ T cells in mice treated with vehicle, etoposide, anti-PD-1 antibody, or etoposide combined with anti-PD-1 antibody. Quantitative data are presented as mean±SD (n=4 mice per group). (D) Representative FACS analysis and quantification of the percentage of perforin and IFN-γ in tumor-infiltrating CD8+ T cells in mice treated with vehicle, etoposide, anti-PD-1 antibody, or etoposide combined with anti-PD-1 antibody. Quantitative data are presented as mean±SD (n=4 mice per group). (E) Representative FACS analysis and quantification of the percentage of tumor-infiltrating effector memory T cells from mice with recurrent tumors, treated with vehicle, etoposide, anti-PD-1 antibody, or etoposide combined with anti-PD-1 antibody. Quantitative data are presented as mean±SD (n=3 mice per group). (F) Representative FACS analysis and quantification of the percentage of tumor-infiltrating exhausted T cells from mice with recurrent tumors, treated with vehicle, etoposide, anti-PD-1 antibody, or etoposide combined with anti-PD-1 antibody. Quantitative data are presented as mean±SD (n=3 mice per group). (G) Representative images of H&E staining and count of pulmonary nodules in K7M2 tumor lung metastasis mice treated with vehicle, etoposide, anti-PD-1 antibody, or etoposide combined with anti-PD-1 antibody (scale bar=1 mm). Data are presented as mean±SD (n=3 mice per group). *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. anti-PD-1, anti-programmed cell death protein 1; FACS, fluorescence-activated cell sorting; i.p, intraperitoneal injection; IFN, interferon; MFI, mean fluorescence intensity; MHC I, major histocompatibility class I. To evaluate the long-term efficacy of etoposide combined with anti-PD-1 therapy, we established a murine postoperative recurrence model (online supplemental figure S5C). Tumor-bearing mice first received two cycles of individual or combined treatment with etoposide and anti-PD-1 antibody, after which the tumor-bearing limb was surgically resected. Following recovery, tumor cells were re-implanted into the contralateral limb to model local recurrence, mimicking the clinical scenario in osteosarcoma patients. The results demonstrated that both etoposide monotherapy and the combination treatment significantly suppressed osteosarcoma growth and also inhibited tumor recurrence (online supplemental figure S5D,E). Flow cytometric analysis of the recurrent tumors showed that either etoposide or anti-PD-1 monotherapy enhanced CD8+ T cell cytotoxicity, and this effect was further strengthened in the combination treatment group (online supplemental figure S5F). To assess the long-term immunomodulatory impact of these treatments, we also analyzed tumor-infiltrating effector memory T cells. Results indicated a significant increase in effector memory T cells infiltration in both monotherapy groups, with the highest levels observed in the combination group (figure 5E). Conversely, the proportion of exhausted T cells was markedly reduced in all treatment groups, with the most pronounced decrease in the combination setting (figure 5F). These findings suggest that the combination of etoposide and anti-PD-1 antibody promotes sustained T cell activity and improves long-term anti-tumor immunity in osteosarcoma. Given the high incidence of pulmonary metastasis in osteosarcoma,1 we further evaluated the effect of combination therapy on metastatic spread using a lung metastasis model, in which mice were injected with K7M2 cells via tail vein (online supplemental figure S5G). Both etoposide monotherapy and combination therapy significantly suppressed the formation of pulmonary metastatic nodules (figure 5G). Finally, we present the case of a patient with stage III osteosarcoma who presented with progressive pulmonary metastasis. The patient had a history of multiple surgical interventions, including amputation, and had developed resistance to conventional chemotherapy. Following two cycles of combination therapy with etoposide and anti-PD-1 antibody (figure 6A), chest CT imaging showed a reduction in the size of the pulmonary metastatic nodule (figure 6B). To assess immunologic changes, we isolated PBMCs and analyzed CD8+ T cell function. After combination treatment, the frequency of CD8+ T cells increased (figure 6C), and these cells exhibited enhanced production of granzyme B (figure 6D). In conclusion, these findings suggest that the combination of etoposide and anti-PD-1 antibody has the potential to benefit patients with advanced, treatment-refractory osteosarcoma. Combination of etoposide and anti-PD-1 antibody has the potential to benefit patients with advanced osteosarcoma. (A) Schematic illustration of osteosarcoma patient treated with etoposide and tislelizumab. (B) Chest CT of osteosarcoma patient before and after treated with etoposide and tislelizumab. (C) Representative FACS analysis and quantification of the percentage of CD8+ T cells in osteosarcoma patient’s PBMCs before and after treatment with etoposide and tislelizumab. Quantitative data are presented as mean±SD (n=3). (D) Representative FACS analysis and quantification of the percentage of granzyme B in CD8+ T cells of osteosarcoma patient’s PBMCs before and after treatment with etoposide and tislelizumab. Quantitative data are presented as mean±SD (n=3). anti-PD-1, anti-programmed cell death protein 1; FACS, fluorescence-activated cell sorting; PBMC, peripheral blood mononuclear cell. Discussion In the clinical management of advanced osteosarcoma, therapeutic resistance remains a major challenge. Although immunotherapy has been widely explored, its efficacy in osteosarcoma remains limited, as reported in previous studies.4 5 Our results are consistent with these observations: anti-PD-1 monotherapy, a commonly used immunotherapeutic approach, showed only modest anti-tumor activity. Numerous clinical trials have attempted to combine anti-PD-1 antibodies with other anti-tumor agents. However, in the absence of a clear mechanistic rationale, most of these efforts have failed to demonstrate synergistic benefit.18 In this study, we screened multiple clinically relevant agents and identified etoposide as a potent inducer of MHC I expression in tumor cells and an enhancer of CD8+ T cell cytotoxicity. We further uncovered that etoposide stimulates the secretion of IL-33 and upregulates its receptor ST2 in osteosarcoma cells. Mechanistically, IL-33 activates NF-κB signaling pathway, leading to MHC I upregulation and enhanced CD8+ T cell-mediated anti-tumor immunity. In addition, IL-33 directly activates CD8+ T cells, further amplifying the immune response (figure 7). These findings establish a scientific rationale for combining etoposide with anti-PD-1 therapy in osteosarcoma, highlighting the role of MHC I regulation and its underlying mechanisms. Moreover, by repurposing clinically available drugs, this strategy may circumvent the long development cycles and uncertainties associated with novel drug discovery, thereby accelerating the translation of research findings into clinical practice. A proposed model for the study etoposide activates CD8+ T cell anti-tumor immunity in osteosarcoma through MHC I upregulation via tumor-secreted IL-33 mediated signaling. IFN, interferon; IL-33, interleukin-33; MHC I, major histocompatibility class I; NF-κB, nuclear factor kappa-B; PD-1, programmed cell death protein 1; ST2, suppression of tumorigenicity 2; TCR, T cell receptor; TNF, tumor necrosis factor. Etoposide, a topoisomerase II inhibitor widely used in cancer therapy, has been part of osteosarcoma treatment regimens since the mid-1980s.19 It is frequently administered in combination with ifosfamide and remains a standard option for salvage chemotherapy in recurrent or metastatic osteosarcoma.20,22 Notably, in small cell lung cancer, the combination of platinum-etoposide chemotherapy with programmed cell death ligand 1 blockade has been shown to significantly improve patient survival.23 In the present study, we investigated the immunomodulatory effects of etoposide within the tumor immune microenvironment of osteosarcoma. Our findings demonstrate that etoposide upregulates MHC I expression and promotes anti-tumor immunity by enhancing CD8+ T cell cytotoxicity and increasing their infiltration into tumors. Beyond its established role as a second-line chemotherapeutic agent, our findings position etoposide as a promising candidate for combination therapy with anti-PD-1 antibodies, particularly in osteosarcoma patients with low HLA expression. Further clinical trials evaluating the efficacy of etoposide combined with immune checkpoint blockade are warranted to translate these findings into improved patient outcomes. Loss of MHC I-mediated antigen presentation represents a common mechanism of immune evasion in multiple tumor types. Consequently, strategies aimed at restoring MHC I expression are considered promising for improving response to immunotherapy.9 One well-established approach involves the IFN signaling pathway, where chemotherapy-induced type I IFN has been shown to enhance MHC I expression and CD8+ T cell cytotoxicity.24 Activation of the NF-κB signaling pathway represents another potential route for elevating MHC I levels. For instance, a known suppressor of NF-κB signaling acts as a negative regulator of MHC I, and its activator mimetic has been shown to upregulate MHC I, sensitize tumor cells to T cell-mediated killing, and synergize with immune checkpoint inhibitors.25 Additionally, multiple oncogenic pathways, including mitogen-activated protein kinase and epidermal growth factor receptor signaling, have been reported to correlate negatively with MHC I expression, suggesting that their inhibition may help restore anti-tumor immunity.9 In this study, we identified a mechanism of MHC I upregulation in osteosarcoma: etoposide stimulates tumor-secreted IL-33, which in turn activates the NF-κB signaling pathway, leading to enhanced MHC I expression on osteosarcoma cells. IL-33, identified in 2005 as a member of the IL-1 cytokine family, signals through its receptor ST2.17 It is known to promote wound-healing responses in various tissues and has been implicated in inflammatory and allergic diseases.26 However, its role in tumor development remains controversial.27 Several studies have suggested a tumor-promoting function for IL-33 in cancers such as breast cancer28 and epithelial ovarian cancer,29 where it facilitates tumor growth and metastasis. IL-33 also acts as a key modulator of the TME. In colon cancer, it exerts dual effects by enhancing cancer stem-like properties in tumor cells while also influencing macrophage behavior.30 In glioblastoma, IL-33 expression correlates with increased tumor-associated macrophages, monocytes, and microglia, collectively fostering a pro-tumorigenic niche.31 Similarly, in squamous cell carcinoma, IL-33 released by cancer stem cells promotes the differentiation of macrophage precursors toward an immunosuppressive phenotype.32 Furthermore, in pancreatic cancer, intratumoral fungi have been shown to induce IL-33 secretion and type 2 immunity, thereby accelerating tumor progression.33 Conversely, a substantial body of evidence supports the anti-tumor functions of IL-33. In models of B16 melanoma and Lewis lung carcinoma, IL-33 activated CD8+ T cells and NK cells, thereby suppressing pulmonary metastasis.34 It has also been shown to enhance the efficacy of 5-fluorouracil chemotherapy in colorectal cancer16 and to exert immunoadjuvant effects in the immunotherapy of human papillomavirus-associated cancers.35 In pancreatic cancer, IL-33-activated group 2 innate lymphoid cells promoted the formation of tertiary lymphoid structures, and an engineered recombinant IL-33 protein demonstrated anti-tumor activity.36 Furthermore, in a model of gastric cancer peritoneal metastasis, IL-33 combined with tumor-associated macrophage reprogramming reshaped the tumor immune microenvironment and improved treatment outcomes.37 Together, these studies highlight the context-dependent nature of IL-33, which can function as a double-edged sword in cancer immunobiology. However, the role of IL-33 in osteosarcoma remains incompletely understood. While few studies have investigated its function in this malignancy, elevated IL-33 expression has been correlated with improved overall survival in one clinical analysis.38 In our study, we identified IL-33 as a key cytokine secreted by osteosarcoma cells on etoposide stimulation. We further dissected its dual roles in modulating both tumor cells and immune cells. Specifically, IL-33 promoted MHC I expression in tumor cells via the NF-κB signaling, thereby enhancing CD8+ T cell-mediated tumor recognition and killing. In parallel, IL-33 directly potentiated CD8+ T cell activation, amplifying the overall anti-tumor immune response. In addition, IL-33 promoted M1-like macrophage polarization, further contributing to an immune-supportive microenvironment. While the present work has focused primarily on the effects of IL-33 on CD8+ T cells, future studies are warranted to investigate its mechanism underlying macrophage polarization in etoposide-treated osteosarcoma. Collectively, our findings establish IL-33 as a critical mediator of etoposide-mediated remodeling of the tumor immune microenvironment of osteosarcoma and highlight its potential to improve the efficacy of anti-PD-1 therapy. IL-33 typically localizes to the nucleus under homeostatic conditions, where it functions as a nuclear factor. However, in response to cellular stress or alterations in the TME, it can be released into the extracellular space, where it acts as an alarmin.39 Cellular injury or necrosis within the TME represents a major trigger for IL-33 release. Furthermore, phagocytosis of tumor debris by medullary sinus macrophages has been shown to promote IL-33 secretion via toll-like receptor 9 activation.40 In our study, we found that etoposide chemotherapy induced IL-33 secretion in osteosarcoma cells, although the precise mechanism remains unclear. As a topoisomerase II inhibitor, etoposide primarily acts by inducing DNA damage. Further study is required to determine whether IL-33 release in this context results directly from etoposide-induced DNA damage or other forms of cellular stress. Our study has primarily examined the role of IL-33 within osteosarcoma cells, with a particular emphasis on its capacity to activate the NF-κB signaling pathway and upregulate MHC I expression. Beyond its effects on tumor cells, we also demonstrated that IL-33 can directly activate CD8+ T cells. This observation is consistent with previous reports as annexin A1 was shown to enhance IL-33-driven CD8+ T cell anti-tumor responses through the PI3K-AKT/p65 signaling axis.16 In addition, IL-33 was found to be essential for CD8+ T cell activation and therapeutic synergy when cisplatin was combined with immune checkpoint inhibitors.41 Moreover, engineering chimeric-antigen receptor T cells to secrete IL-33 has been shown to remodel the TME and suppress the growth of several solid tumors.42 In summary, within the tumor immune microenvironment of osteosarcoma, IL-33 exerts dual effects on both tumor cells and CD8+ T cells to enhance anti-tumor immunity. Supplementary material Acknowledgments We are grateful to Dr Sumiya Dalangood (State Key Laboratory of Systems Medicine for Cancer, Shanghai Jiao Tong University School of Medicine) and Dr Chenwei Yuan (Renji Hospital, Shanghai Jiao Tong University School of Medicine) for valuable suggestions. We also appreciate Dr Wen Qiao (State Key Laboratory of Systems Medicine for Cancer, Shanghai Jiao Tong University School of Medicine), Dr Cegui Hu (State Key Laboratory of Systems Medicine for Cancer, Shanghai Jiao Tong University School of Medicine), Dr Na Jing (State Key Laboratory of Systems Medicine for Cancer, Shanghai Jiao Tong University School of Medicine) and Dr Songlin Li (State Key Laboratory of Systems Medicine for Cancer, Shanghai Jiao Tong University School of Medicine) for helpful assistance. Footnotes Funding: This study was supported by the National Natural Science Foundation of China (82473469, 82373177, 82272997, 32422029, 82472915, 82172366). Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: This study involves human participants and was approved by the Institutional Research Ethics Committee of Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine (No. 2024-070). Participants gave informed consent to participate in the study before taking part. Data availability statement No data are available. References Associated Data Supplementary Materials Data Availability Statement No data are available.