Antibodies against GSDME (#AF4016), β-actin (#AF7018), and Caspase-3 (#AF7022) were purchased from Affinity Biosciences (Changzhou, China). Antibodies against CD45 (#756971), CD3 (#564008), CD4 (#557307), CD8 (#569870), and FoxP3 (#563101) were obtained from BD Biosciences (New Jersey, US). GSDME (#13075) and ubiquitin (#10201) antibodies from Proteintech Group (Wuhan, China) were used for immunoprecipitation (IP). An anti-PD-1 antibody for blocking PD-1/PD-L1 in mice was purchased from BioXCell (#BE0273, West Lebanon, US). Curcumin (CUR, >98% purity) and decitabine (DAC, >98% purity) were obtained from Energy-Chemica (Shanghai, China). DSPE-mPEG2000 was obtained from AVT Pharmaceutical (Shanghai, China). Detection kits for biochemical markers were purchased from Jiangsu Sinnowa Medical Technology (Zhenjiang, China).

Patient specimens and data were provided by Shanghai OUTDO Biotech Co., Ltd., National Engineering Research Center for Biochip, Shanghai, China. The collection and use of human tissue samples were approved by the Ethics Committee of the National Engineering Research Center, Shanghai (Approval No. SHYJS-CP-230902).

The CT26 and HT29 cell lines were maintained in DMEM supplemented with 10% FBS and 2 mM L-glutamine at 37 °C under 5% CO₂. Cell viability was assessed via a Cell Counting Kit-8 (APExBIO Technology, Texas, US), and extracellular LDH release was measured via an ELISA kit (Shanghai Enzyme-linked Biotechnology, Shanghai, China). The cell morphology was evaluated via microscopic examination.

The samples were lysed in RIPA buffer containing protease and phosphatase inhibitors (Servicebio, Wuhan, China). The lysates were subsequently centrifuged at 13,000 rpm for 15 min at 4 °C, after which the protein concentrations were determined via a BCA assay (Yeasen, Shanghai, China). Proteins were separated by SDS–PAGE, transferred to PVDF membranes, blocked with 5% skim milk, and incubated overnight at 4 °C with primary antibodies. After incubation with HRP-conjugated secondary antibodies, protein bands were visualized via enhanced chemiluminescence (ECL) (Servicebio, Wuhan, China).

The cells were lysed in IP buffer and centrifuged at 14,000 × g for 15 min at 4 °C, with 10% of the supernatant retained as input. The remaining lysate was incubated overnight at 4 °C with an anti-GSDME antibody and protein A/G magnetic beads (HY-K0202, MCE, New Jersey, US). The immunocomplexes were washed three times with PBS, boiled in sample buffer for 5 min, and analysed by immunoblotting with an anti-ubiquitin antibody.

After being washed with PBS, the cells were lysed in proteasome lysis buffer. The lysates were clarified by centrifugation at 4 °C, and the protein concentrations were quantified via a BCA assay. Equal aliquots (40 µg) of protein were dispensed into black 96-well plates and incubated with 50 µM Suc-LLVY-AMC (MCE, New Jersey, United States) at 37 °C for 60 min. Fluorescence (Ex 380 nm/Em 460 nm) was measured via a microplate reader, and proteasome activity was determined by subtracting the blank control signals.

The Gsdme-knockout CT26 cell line was generated as previously described (Zhang et al., 2020). Gsdme-targeting gRNAs were subsequently cloned and inserted into the lentiCRISPR-v2-puro vector. The plasmids were cotransfected with pSPAX2 and pCMV-VSV-G into HEK293T cells. Viral supernatants were harvested 48 h posttransfection and used to transduce CT26 cells. After 48 h, the cells were selected with puromycin (8 μg/mL).

DSPE-mPEG2000 and CUR were mixed at a 10:1 mass ratio and dissolved in anhydrous ethanol in a round-bottom flask. The solution was subjected to rotary evaporation, followed by vacuum drying for 24 h. Phosphate-buffered saline (PBS) was then added, and the mixture was sonicated for 20 min to prepare the CUR formulation.

Male BALB/c mice (6–8 weeks old) were purchased from Jiangsu Huachuang Xinnuo Pharmaceutical Technology Co., Ltd. All the animal experiments were approved by the Institutional Animal Care and Use Committee and conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

The mice received intravenous injections of CUR or DAC at various doses for a total of 15 administrations over 30 days. After treatment, blood and tissues (heart, liver, spleen, lungs, kidneys, thymus, and femurs) were collected. Organ coefficients were calculated, and hematological, renal, and hepatic parameters were assessed. Tissues were fixed in 4% paraformaldehyde, embedded, sectioned, and stained with hematoxylin and eosin (H&E).

For the orthotopic tumor model, the mice were fasted for 12 h prior to anaesthesia. A midline abdominal incision was made to expose the colon, and CT26 cells (1 × 10⁶ cells per mouse) were injected into the colonic mucosa. The peritoneum, abdominal muscles, and skin were sequentially sutured, and the incision was disinfected with 75% ethanol. For the subcutaneous model, CT26 cells (1 × 10⁶ cells per mouse) were injected subcutaneously into the dorsal region after disinfection with 75% ethanol.

Tumor tissues and cancer cell lines were lysed via sonication in 8 M urea buffer containing protease inhibitors and EDTA, followed by centrifugation to isolate the lysates. Protein concentrations were determined via a BCA assay, after which the proteins were digested into peptides, desalted, and dried prior to DIA-based LC–MS/MS analysis. Differentially expressed proteins were subjected to Gene Ontology enrichment and pathway analysis, and expression patterns were visualized as heatmaps. All the experiments were performed by OE Biotech Co., Ltd. (Shanghai, China).

Single-cell suspensions were prepared from excised tumors by mechanical dissociation and enzymatic digestion, followed by filtration, red blood cell lysis, and Fc receptor blocking. The cells were stained with fluorophore-conjugated antibodies against surface and intracellular markers, with compensation set using single-stained beads. The data were acquired on a Beckman Coulter CytoFLEX cytometer and analysed via FlowJo software.

The data were analysed and visualized via GraphPad Prism 9.0. Statistical significance among treatment groups was determined via one-way ANOVA with Tukey’s post hoc test or two-tailed Student’s t-test, as appropriate. The data are presented as the means ± standard errors of the means (SEMs). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

In MSS CT26 tumors, GSDME promotes CD8⁺ T-cell infiltration and strengthens antitumor immune responses (Luo et al., 2024). To determine whether a similar feature exists in human MSS CRC, GSDME expression was assessed via immunohistochemistry (IHC) in tumor specimens from 94 MSS CRC patients (Figure 1A). Patients were classified into high (score >1.7, n = 49) and low (score ≤1.7, n = 45) GSDME expression groups on the basis of their IHC scores (Figures 1B,C). Survival analysis revealed markedly improved overall survival in the high-expression group, with 81.6% of patients alive at 90 months post-surgery, compared with 40.0% in the low-expression group (Figure 1D), underscoring the tumor-suppressive function of GSDME in MSS CRC. Notably, high GSDME-expressing MSS CRC tumors presented increased CD8⁺ T-cell infiltration and elevated PD-L1 expression (Figure 1E), both of which are predictive of a favourable response to PD-1/PD-L1 blockade. These findings suggest that GSDME may contribute to shaping a tumor immune landscape in human MSS CRC that favours responsiveness to PD-1/PD-L1 blockade.

Mouse (CT26) and human (HT29) CRC cell lines, both of which exhibit an MSS phenotype, were treated with CUR at 1, 5, or 10 μM for 72 h, covering concentrations around the IC₅₀ values for both cell lines (Figure 2A). Western blot analysis revealed the upregulation of GSDME and its pore-forming N-terminal fragment (GSDME-N) in both cell lines upon CUR treatment, with the maximal effect observed at 10 μM CUR (Figures 2B–D). Moreover, CUR treatment also activated caspase-3, which specifically cleaves GSDME to release GSDME-N (Wang et al., 2017). These results indicate that CUR upregulates GSDME and activates the caspase-3/GSDME axis to amplify pyroptosis in these two CRC cell lines.

The amplification of pyroptosis was further confirmed by assessing lactate dehydrogenase (LDH) release and cell morphology. CUR induced a concentration-dependent increase in LDH release in both cell lines (Figures 2E,G). Consistently, CUR-treated cells displayed a swollen pyroptotic morphology, with the proportion of pyroptotic cells increasing with increasing CUR concentration (Figures 2F,H). Collectively, these findings suggest that CUR enhances pyroptosis in CT26 and HT29 cells.

To investigate the mechanism underlying the CUR-mediated upregulation of GSDME, quantitative proteomic analysis was performed on CUR-treated CT26 and HT29 cells. Principal component analysis (PCA) revealed distinct proteomic profiles between treated and untreated cells (Figures 3A,D). Differentially expressed proteins (DEPs) were identified in each cell line using a threshold of P < 0.05 and a fold change (FC) ≥ 1.5.

GO analysis of DEPs in CT26 and HT29 cells revealed significant enrichment of intracellular processes, particularly those associated with the ubiquitin–proteasome system (UPS), which regulates intracellular protein levels (Figures 3B,E). Hierarchical clustering (Figure 3C) revealed that, following CUR treatment, Usp20 (ubiquitin-specific protease 20), a deubiquitinating enzyme (Li et al., 2022), was upregulated, whereas several ubiquitination-related enzymes, including Map3k1 (mitogen-activated protein kinase kinase kinase 1) (Suddason and Gallagher, 2015), Cdc34 (cell division cycle 34) (Jie et al., 2025), Rnf135 (RING finger protein 135) (Hu et al., 2025), and Rnf141 (RING finger protein 141) (Zhang et al., 2021), were downregulated in CT26 cells. In CUR-treated HT29 cells, SASH1 (SAM and SH3 domain-containing protein 1), UBB (ubiquitin B), CDC26 (cell division cycle 26), and UBE2L6 (ubiquitin-conjugating enzyme E2 L6) were downregulated (Figure 3F); these proteins exhibit ubiquitin ligase activity either directly or indirectly (Orfali et al., 2020; Masuda et al., 2015; Dauphinee et al., 2013; Oh et al., 2013).

Previous studies have shown that CUR broadly affects UPS activity, with specific molecular targets varying across cell types (Torghabe et al., 2025). Although proteomic data revealed distinct expression patterns of ubiquitination-related proteins in CT26 and HT29 cells following CUR treatment, these observations suggest that CUR may influence GSDME levels in both cell lines through modulation of UPS activity. Consistent with this interpretation, immunoprecipitation (IP) and proteasome activity assays revealed reduced GSDME ubiquitination and decreased proteasome activity upon CUR treatment in both cell lines (Figures 3G–J). Taken together, these findings support a working model in which CUR promotes intracellular accumulation of GSDME by attenuating its ubiquitination and subsequent proteasome-mediated turnover.

CT26 tumor-bearing mice were used to evaluate CUR-enhanced, GSDME-dependent pyroptosis in vivo. Decitabine (DAC), a DNA methyltransferase inhibitor known to upregulate GSDME and induce pyroptosis in various tumor types (Gong et al., 2023; Chen et al., 2025), was included as a control. The mice received daily intravenous injections of CUR or DAC at 1, 5, or 10 mg/kg for five consecutive days. To facilitate CUR administration, a hydrophilic CUR-loaded nanoformulation was prepared via DSPE-mPEG2000. This nanoformulation markedly enhanced CUR dispersion in phosphate-buffered saline and resulted in a well-defined spherical morphology, uniform particle size, and sustained CUR release (Supplementary Figures S1–S4).

The administration of CUR or DAC at 10 mg/kg significantly upregulated and activated GSDME, respectively, in CT26 tumors (Figures 4A–F). Compared with the controls, CUR induced ∼1.5- and ∼2.6-fold increases in total GSDME and GSDME-N, respectively, whereas DAC induced ∼1.9- and ∼1.7-fold increases, comparable to those observed with CUR. To assess safety, healthy mice received intravenous CUR or DAC at 1, 5, or 10 mg/kg every other day for 30 days, with saline-treated mice serving as controls. CUR was well tolerated, as indicated by normal weight gain and the absence of mortality (Figures 5A,B). In contrast, DAC exhibited dose-dependent toxicity: treatment with 5 or 10 mg/kg resulted in significant weight loss and substantial mortality (Figures 5A,B). High-dose DAC also caused pronounced thymic atrophy, as evidenced by a reduced thymus index and loss of the medullary region enriched in immature T cells (Figures 5C,D), along with dose-dependent decreases in femoral bone marrow density and peripheral blood counts of WBCs, monocytes, granulocytes, and lymphocytes (Figures 6A,B). These results indicate that CUR effectively upregulates and activates GSDME to increase pyroptosis in CT26 tumors, with an efficacy comparable to that of DAC but minimal toxicity.

To assess the impact of CUR-enhanced pyroptosis on the TME in CT26 tumors, immune cell subsets were characterized by staining with fluorescently labelled antibodies against lineage-specific surface and intracellular markers (Figure 7A).

Compared with the control, CUR treatment significantly increased the proportion of CD45⁺ leukocytes among total tumor cells in both the subcutaneous and orthotopic CT26 tumor models, indicating enhanced immune cell recruitment to the tumor site (Figures 7B,C). Further analysis of T-cell subsets revealed a substantial increase in CD3⁺ T-cell infiltration following CUR treatment. Notably, the proportions of antitumor effector CD8⁺ and CD4⁺ T cells were significantly elevated, whereas the fraction of immunosuppressive regulatory T cells (Foxp3⁺CD4⁺ Tregs) within the CD4⁺ population was markedly reduced (Figures 7B,C). These findings suggest that CUR-induced pyroptosis promotes the expansion of antitumor effector cells while suppressing immunosuppressive subsets.

The therapeutic efficacy of CUR in combination with an anti-PD-1 monoclonal antibody was evaluated in both subcutaneous and orthotopic CT26 tumor models. CUR and anti-PD-1 were administered every 2 days for 2 weeks at 10 mg/kg each. In the subcutaneous tumor model, both agents were delivered via intravenous injection, whereas in the orthotopic model, CUR was administered orally, and anti-PD-1 was administered intraperitoneally.

In the subcutaneous model, CUR monotherapy resulted in negligible antitumor activity, and PD-1 blockade alone was similarly ineffective because of the immunosuppressive TME (Figures 8A,B). In contrast, combination therapy with CUR and anti-PD-1 markedly suppressed tumor growth compared with either monotherapy, as reflected by reduced tumor volumes and increased tumor inhibition rates (Figures 8A,B). This combination also significantly prolonged overall survival (Figure 8C) while maintaining stable body weight throughout the treatment period (Figure 8D).

In the orthotopic model, control mice developed substantial ascites, with 50% exhibiting severe accumulation (Figure 8E). CUR monotherapy partially alleviated ascites, whereas anti-PD-1 therapy alone and combination therapy achieved greater reductions. The mean tumor weights were lowest in the combination group, intermediate in the anti-PD-1 and CUR groups, and highest in the control group (Figure 8F). The tumor inhibition rates were 28.2% (CUR), 42.2% (anti-PD-1), and 73.3% (combination) (Figure 8G), demonstrating that CUR potentiates PD-1 blockade. Moreover, the combination treatment significantly prolonged overall survival compared with either monotherapy (Figure 8H).

To confirm the essential role of GSDME in the synergistic effect of CUR and PD-1 blockade, subcutaneous tumors were established via the use of GSDME-knockout CT26 cells (Figure 9A). Loss of GSDME abrogated CUR-induced pyroptosis in CT26 cells, as indicated by diminished LDH release and an apoptotic morphology instead of pyroptotic features (Figures 9B,C). In GSDME-deficient tumors, CUR failed to reprogram immune cell subsets or synergize with PD-1 blockade (Figures 9D–G). These results demonstrate that the cooperative antitumor efficacy of CUR and PD-1 blockade in CT26 tumors is critically dependent on GSDME-mediated tumor cell pyroptosis.