Unlocking caspase-8 enables Pt(IV) to trigger robust caspase-3/GSDME-mediated pyroptosis via drug co-delivery in chemoresistant bladder cancer

Abstract

Overcoming profoundly suppressed programmed cell death and an immunosuppressive tumor microenvironment (TME) remains a formidable challenge in gemcitabine-cisplatin-resistant muscle-invasive bladder cancer (GP-R MIBC). To address this, we developed a glutathione (GSH)-responsive nanosystem for co-delivering the cathepsin H (CTSH) inhibitor prodrug E64d and a cisplatin prodrug Pt(IV), implementing a spatiotemporally coordinated “unlock-and-trigger” strategy. Mechanistically, the activated E64 irreversibly inhibits CTSH, thereby unlocking the TNFα–caspase-8 axis to reinstate GSDMD-mediated pyroptosis. Concurrently, caspase-8 dually activates caspase-3 via both direct cleavage and the Bid-mediated mitochondrial pathway. This primed state ensures that the DNA damage induced by the reduced Pt(IV) robustly triggers caspase-3/GSDME-mediated pyroptosis. This synergistic dual-pyroptosis pathway potently stimulates immunogenic cell death (ICD) and reverses the immunosuppressive TME by promoting dendritic cell maturation and CD8⁺T lymphocyte infiltration. Consequently, when combined with an αPD-L1 antibody, this strategy effectively eradicated GP-R bladder tumors and suppressed recurrence in vivo, demonstrating compelling translational potential. In summary, our work presents a translatable nanotherapeutic strategy that resensitizes GP-R MIBC to chemo-immunotherapy by leveraging a synergistic “unlock-and-trigger” mechanism to reactivate pyroptosis.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04449-4.

Introduction

Bladder cancer is a significant global health burden, with around 549,000 new cases and 200,000 deaths annually [1]. Muscle-invasive bladder cancer (MIBC) accounts for 25% of cases and represents the most aggressive subtype due to its high invasiveness and mortality rate [2]. The first-line gemcitabine-cisplatin regimen (GP) improves overall survival by 6–8% in non-metastatic MIBC and yields > 5-year survival in 15% of metastatic patients [3, 4].

Cisplatin kills tumor cells by causing DNA damage. This damage activates caspase-3, which then cleaves GSDME to trigger pyroptosis—a form of inflammatory cell death [5–7]. This process initiates immunogenic cell death (ICD), thereby stimulating an anti-tumor immune response [8, 9]. However, 50% of patients develop adaptive chemoresistance within 9 months with concomitant failure of PD-L1 antibody therapy [10–12]. This phenomenon is the primary cause of mortality. This resistance arises not only from classical mechanisms like cisplatin inactivation by glutathione (GSH) but, more importantly, from fundamental defects in the programmed cell death (PCD) pathways within cancer cells [13–16]. To enhance cisplatin’s stability and DNA-targeting efficiency, we developed a liposoluble Pt(IV) prodrug. The therapeutic efficacy of Pt(IV) is still limited in resistant tumors due to persistent blockades in PCD pathways, demonstrating that simply increasing initial cytotoxic insult is inadequate to overcome resistance.

Recent studies have identified Cathepsin H (CTSH) as a key mediator of resistance to gemcitabine-cisplatin therapy [17]. CTSH operates through a non-canonical mechanism: it suppresses the TNFα–caspase-8 signaling axis. This inhibition simultaneously blocks two critical downstream pathways: (1) caspase-8/GSDMD-mediated pyroptosis, and (2) both the direct (caspase-8/−3) and indirect (caspase-8/Bid/−3) routes to caspase-3 activation. This multi-layered defense system not only drives chemoresistance but also, by preventing pyroptosis and ICD, fosters an immunosuppressive tumor microenvironment (TME) that further impairs immunotherapy outcomes [18–27]. Therefore, targeting CTSH is a promising strategy for reversing resistance. The cysteine protease inhibitor E64 can irreversibly inhibit CTSH without affecting the caspase family, making it a theoretically ideal candidate. However, E64 has poor membrane permeability due to its hydrophilic carboxyl group, and its effective in vitro concentration (IC₅₀ = 2.64 mM) is difficult to achieve in vivo [17], hindering clinical translation. Esterification of E64 to form the prodrug E64d improves cellular uptake. Yet, this prodrug lacks tumor specificity in circulation, and potential “off-target” toxicity presents new safety concerns—effectively transforming the pharmacological challenge into a drug delivery problem.

Consequently, the successful therapy of GP-resistant (GP-R) bladder cancer hinges on the temporal and spatial co-delivery of E64d and the Pt (IV) prodrug to tumor cells. Any disconnect in timing or location would compromise efficacy: unless E64d first lifts the CTSH blockade, the death signal from Pt (IV) cannot be fully amplified; unless both agents co-localize, a synergistic pyroptotic response cannot be achieved within the tumor population.

To address this, we developed a nano-platform that precisely executes an “unlock-and-trigger” strategy. Using a synthesized GSH-responsive amphiphilic polymer (PHMM), we constructed a nanoparticle system for the co-delivery of E64d and Pt(IV) (NP@Pt(IV)-E64d, or NP2) (Scheme 1 a). This system ensures specific tumor accumulation and synchronous intracellular release of both prodrugs. Inside the cell, E64d is rapidly hydrolyzed to active E64, which unlocks the TNFα–caspase-8 axis to reinstate GSDMD-mediated pyroptosis. Simultaneously, caspase-8 dually activates caspase-3 (via direct cleavage and the Bid-mediated mitochondrial pathway). This primes the cells so that the DNA damage induced by reduced Pt(IV) triggers robust caspase-3/GSDME-mediated pyroptosis. This spatiotemporally coordinated action not only resensitizes tumors to cisplatin but also powerfully enhances pyroptosis and ICD by concurrently activating both the caspase-8/GSDMD and caspase-3/GSDME pyroptosis axes (Scheme 1 b).

Scheme 1.

Schematic illustration of the spatiotemporally coordinated “unlock-and-trigger” strategy of NP@Pt(IV)-E64d (NP2) for overcoming chemoresistance and immunosuppression in gemcitabine-cisplatin resistant (GP-R) bladder cancer. (A) Preparation of NP2. (B) NP2-induced pyroptosis and immune activation are initiated by the CTSH inhibitor E64d, which de-represses caspase-8/GSDMD-mediated pyroptosis and primes caspase-3 activation. This priming, synergistically with Pt(IV)-induced DNA damage, robustly triggers caspase-3/GSDME-mediated pyroptosis. The resultant release of DAMPs and cytokines promotes DC maturation and CTL infiltration, transforming the TME from “cold” to “hot” and potentiating response to immune checkpoint blockade

Ultimately, this strategy reverses the immunosuppressive TME by promoting dendritic cell maturation and recruiting cytotoxic T cells. When combined with αPD-L1 antibody therapy, it achieves potent, immune-driven eradication of GP-R bladder cancer in recurrence models in vivo, validating its compelling translational potential. In conclusion, our study presents a promising and translatable nanoparticle-based strategy to overcome this form of clinical resistance.

Results

CTSH-mediated pyroptosis suppression and immunosuppression underlie resistance to gemcitabine-cisplatin therapy

Muscle-invasive bladder cancer (MIBC) exhibits a propensity for early micrometastasis (Fig. 1a). Radical cystectomy (RC) reduces pelvic recurrence rates, preexisting micrometastases can drive distant metastatic progression. While adjuvant gemcitabine-cisplatin (GP) chemotherapy—the first-line systemic therapy for MIBC—effectively eliminates residual or micrometastatic lesions, acquired resistance in persistent micrometastases leads to therapeutic failure (Fig. 1b). Single-cell sequencing (scRNA-seq) analysis of the tumor microenvironment in patients receiving chemotherapy and immunotherapy revealed not only Treg cell enrichment but also abundant functionally exhausted CD8⁺ T cells (Tex) (Fig. 1c).

Fig. 1.

From CTSH-mediated pyroptosis suppression and immunosuppression in the clinic to validation of its overexpression in gemcitabine-cisplatin resistant (GP-R) models. (a) Schematic of chemotherapy-induced micrometastasis progression. (b) Representative axial CT image from a muscle-invasive bladder cancer patient (post-radical cystectomy and first-line GP therapy) demonstrating heterogenous enhancement in the pelvic and inguinal lymph node areas (red circle), suggestive of disease recurrence. (c) scRNA-seq UMAPs: Untreated (n = 2) vs. αPD-L1 + Chemo (n = 5). (d) CTSH expression association with TME scores (StromalScore, ImmuneScore, EstimateScore). (e-f) Establishment of the GP-R model. e) In vivo. f) In vitro. (g) IHC analysis of CTSH expression in tumor tissues after different chemotherapy cycles (scale bar: 50 μm). (h) Western blot (WB) analysis of CTSH expression in MB49-R cells (Vinculin as loading control)

Recent studies identified that cathepsin H (CTSH), highly expressed in gemcitabine–cisplatin-resistant (GP-R) bladder cancer, suppresses the TNFα–caspase-8 axis, thereby inhibiting pyroptosis. Analysis of 394 bladder cancer samples from TCGA showed that CTSH expression negatively correlates with tumor microenvironment (TME) score (Fig. 1 d) but positively correlates with Treg cell abundance (Figure S1a), suggesting its potential role in promoting an immunosuppressive microenvironment. This finding offers new insights for reversing chemoresistance and immunosuppression in GP-R bladder cancer.

To leverage this discovery, we established a GP-R mouse bladder tumor model (GP-R C57BL/6) (Fig. 1e) and derived resistant human (T24-R) and mouse (MB49-R) bladder cancer cell lines (Fig. f). In vivo models showed that GP-R C57BL/6 mice exhibited unrestrained tumor growth by day 21, demonstrating the emergence of drug resistance (Figure S1b). In vitro, resistant cells had significantly higher IC₅₀ values for gemcitabine and cisplatin compared to parental cells (Table S1), which was further confirmed by Annexin V-FITC/PI (Figures S1c, e) and Calcein-AM/PI staining (Figures S1d, f). Immunohistochemistry (IHC) and Western blot (WB) analyses consistently demonstrated upregulation of CTSH protein in both in vivo (Fig. g) and in vitro (Figs. h; S1g) resistant models.

Inhibition of CTSH potentiates Pt(IV)-induced pyroptosis

In GP-R bladder cancer, CTSH overexpression inhibits cisplatin-induced caspase-3/GSDME-mediated pyroptosis by suppressing the TNFα–caspase 8 axis and subsequent activation of caspase-3 by caspase-8. To address this, we employed the irreversible CTSH inhibitor E64. To overcome E64’s poor membrane permeability due to high hydrophilicity, we used its prodrug E64d—a carboxyl-esterified derivative with high lipophilicity that hydrolyzes intracellularly to release active E64, effectively inhibiting CTSH. Meanwhile, cisplatin was modified into a Pt(IV) configuration to enhance stability and reduce in vivo inactivation, thereby more effectively activating the caspase-3/GSDME pathway (Fig. 2a).

Fig. 2.

CTSH expression in GP-R tumors and the synergistic antitumor effect of its inhibition combined with Pt(IV) in triggering robust pyroptosis. (a) Schematic of cell-permeable E64d to potentiate Pt(IV)-induced pyroptosis via caspase-8/3 cascade activation. (b) MTT assay of cell viability in MB49-R cells treated for 72 h with E64d, Cispt, Pt(IV), or Pt(IV) + E64d. (c) Representative CLSM images of Calcein-AM (green, live) and PI (red, dead) dual staining in MB49 subtype cells (scale bar: 100 μm). (d) Representative Annexin V-FITC/PI staining flow cytometry plots of MB49-R cells after different treatments (bolded figures are percentages of the late apoptosis). (e) Representative CLSM images of MB49-R cells showing hallmark features of pyroptosis (indicated by white arrows) following 5 h treatment with the Pt(IV) and E64d combination (scale bar: 10–25 μm)

To investigate the synergistic anticancer effect of E64d and Pt(IV), we conducted systematic evaluations in CTSH-high MB49-R cells. MTT assays showed that low concentrations of E64d (< 10 µM) had no significant effect on proliferation, whereas Pt(IV) exhibited stronger cytotoxicity than cisplatin. When combined at a 1:1 ratio, Pt(IV) and E64d produced marked synergistic anti-proliferative effects at concentrations below 10 µM, significantly outperforming monotherapies (Fig. 2b). This synergy was consistently supported by Calcein-AM viability assays (Figs. 2c; S2a) and Annexin V-FITC/PI staining (Figs. 2 d; S2b), with statistically significant differences. Confocal laser scanning microscopy (CLSM) further revealed that, compared to Pt(IV) alone, the Pt(IV) + E64d combination induced earlier and more prominent pyroptotic morphology—including pyroptotic vesicle, embrane blebbing, and rupture (Fig. 2e)—morphologically confirming robust activation of the pyroptosis pathway and synergistic antitumor action. These findings substantiate that CTSH inhibition potentiates the anticancer activity of Pt(IV) in GP-R bladder cancer cells. This provides a mechanistic rationale for maximizing pyroptosis using relatively non-toxic CTSH inhibitors.

Preparation and characterization of NP@Pt(IV)-E64d nanoparticles

Although E64d and Pt(IV) prodrugs exhibit strong cell membrane penetration due to their lipophilicity, their lack of tumor targeting in systemic circulation can lead to off-target toxicity. Moreover, their synergistic effect critically depends on spatiotemporal co-delivery within tumor cells.

To address these challenges, we designed a glutathione (GSH)-responsive polymer, PHHM, and used nanoprecipitation to self-assemble Pt(IV) and E64d into GSH-responsive nanoparticles termed NP@Pt(IV)-E64d (Fig. 3a). As a control, NP@Pt(IV) was prepared similarly using only Pt(IV) and PHHM. Inductively coupled plasma mass spectrometry (ICP-MS) quantified platinum concentrations in NP@Pt(IV) and NP@Pt(IV)-E64d as 151,065 µg/L (775 µM) and 202,971 µg/L (1,040 µM), respectively. High-performance liquid chromatography (HPLC) determined the drug loading capacity of E64d in NP@Pt(IV)-E64d. The molar ratio of E64d to Pt(IV) in NP@Pt(IV)-E64d was approximately 1:1.

Fig. 3.

Preparation and characterization of NP@Pt(IV) and NP@Pt(IV)-E64d nanoparticles. (a) Synthesis strategy of NP@Pt(IV)-E64d for minimizing off-target effects and ensuring co-localization. (b-f) Physicochemical characterization of NP@Pt(IV) and NP@Pt(IV)-E64d nanoparticles: (b) Representative TEM image (scale bar: 200 nm). (c) Hydrodynamic size distribution by DLS. (d) Size stability over 13 days. (e) Polydispersity index (PDI). (f) Zeta potential. (g-i) The GSH-triggered process of NP@Pt(IV)-E64d: (g) Pt drug release kinetics in 10 mM GSH. (h) Nanoparticle size change after 24 h in GSH. (i) Structural degradation visualized by TEM (scale bar: 100 nm). (j-m) Cellular uptake and intracellular platinum content of NP@Pt(IV)-E64d. (j) CLSM images of MB49-R cells after 7 h incubation with Cy5.5-labeled nanoparticles (red). Nuclei (blue, DAPI) and cytoskeleton (green, Actin). Scale bar: 10 μm. (k) Flow cytometry analysis of cellular uptake over time. (l) Quantification of MFI from (k) (n = 3). (m) Intracellular platinum content measured by AAS after treatment with 5 µM platinum-equivalent doses (n = 3). (n-q) In vivo biodistribution and tumor accumulation of NP@Pt(IV)-E64d. (n) In vivo fluorescence imaging of GP-R tumor-bearing mice after intravenous injection of Cy7.5-labeled nanoparticles. (o) Time-dependent quantification of MFI in tumors. (p) Ex vivo fluorescence imaging of major organs and tumors at 48 h post-injection. (q) Corresponding quantitative MFI of dissected tissues.

We first characterized the structure and physicochemical properties of NP@Pt(IV)-E64d. The disulfide bonds in the polymer backbone are cleaved in the high-GSH tumor microenvironment, enabling controlled drug release (Figure S3), while Pt(IV) is reduced by GSH to active cisplatin (Figure S4). Transmission electron microscopy (TEM) revealed that both NP@Pt(IV) and NP@Pt(IV)-E64d are spherical with diameters of ~150 nm and ~114 nm, respectively (Figs. 3b; S5a). Dynamic light scattering (DLS) analysis (Fig. 3c) showed hydrodynamic sizes of ~147 nm and ~124 nm, suitable for tumor-targeted delivery. Furthermore, both nanoparticles maintained size stability in PBS and serum-containing medium at 37 °C for 13 days (Figs. 3 d; S5b), and their low polydispersity index (PDI) values (< 0.1) indicated excellent particle uniformity (Figs. 3e; S5b). Moreover, zeta potentials were approximately −19.1 ± 0.9 mV and −18.5 ± 0.7 mV, respectively (Fig. 3f). GSH-responsive drug release was confirmed by comparing platinum release with and without GSH. In the presence of 10 mM GSH, over 60% of platinum was released within 24 h, versus only ~11% without GSH (Figs. 3 g; S5c). After GSH treatment, the particle size distribution shifted from monomodal to bimodal (Fig. 3 h), and TEM showed structural disintegration (Fig. 3i), collectively verifying GSH-triggered disassembly.

Subsequently, we evaluated the anticancer activity of NP@Pt(IV)-E64d by studying its uptake in MB49-R cells. CLSM imaging of Cy5.5-labeled nanoparticles (NP@Cy5.5) revealed time-dependent cellular uptake, with clear cytoplasmic fluorescence at 4 h that increased over 7 h (Figs. 3j; S5d-e). Flow cytometry further showed that the mean fluorescence intensity at 7 h was 10.7- and 2.4-fold higher than at 1 h and 4 h, respectively (Figs. 3k-l). Meanwhile, using atomic absorption spectroscopy (AAS) to quantify intracellular platinum accumulation, NP@Pt(IV)-E64d treatment (5 µM Pt) for 1 h resulted in significantly higher platinum content (69.8 ± 10.1 ng/10⁶ cells) compared to Pt(IV), NP@Pt(IV), or cisplatin. Accumulation further increased at 4 h and 7 h (Fig. 3 m), confirming efficient tumor cell uptake.

Finally, we assessed the in vivo biodistribution of NP@Pt(IV)-E64d in a GP-R C57BL/6 model. Following intravenous injection of Cy7.5-labeled nanoparticles (NP@Cy7.5), In Vivo Imaging System (IVIS) showed detectable tumor signals at 8 h, peaking at 24 h and persisting for up to 48 h (Figs. 3n), indicating good tumor targeting and long circulation. Ex vivo organ imaging at 48 h revealed the strongest fluorescence in the kidneys and liver (3.98 × 10⁸ p s⁻¹ cm⁻² sr⁻¹ and 3.65 × 10⁸ p s⁻¹ cm⁻² sr⁻¹, respectively), followed by tumors (2.80 × 10⁸ p s⁻¹ cm⁻² sr⁻¹), with tumor signals

significantly higher than those in the heart, spleen, and lungs (Figs. 3p).

NP@Pt(IV)-E64 enhances caspase-3/GSDME-mediated pyroptosis via caspase-8 activation

To evaluate the anticancer activity of different nanoparticles in vitro, we compared the effects of free Pt(IV), NP@Pt(IV), and NP@Pt(IV)-E64d in GP-R mouse MB49-R and human T24-R bladder cancer cells. MTT assays showed that NP@Pt(IV)-E64d exhibited the strongest cytotoxicity in both cell lines (Figs. 4a-b), with an IC₅₀ of 2.9 ± 0.4 µM in MB49-R cells, significantly lower than that of free Pt(IV) (6.3 ± 0.3 µM) (Table 1). To quantitatively assess the ability of the NP@Pt(IV)-E64d to overcome drug resistance, we calculated the resistance reversal fold (IC50, free Pt(IV)/IC50, NP@Pt(IV)-E64d). This analysis yielded a reversal fold of 2.2. These results suggest that NP@Pt(IV)-E64d not only facilitates more efficient drug release but also effectively circumvents the established resistance mechanisms in bladder cancer cells.

Fig. 4.

Evaluation of the anticancer effects of NP@Pt(IV)-E64d nanoparticles through caspase-8-mediated caspase-3 activation. (a-b) Cell viability in MB49-R (a) and T24-R cells treated with Pt(IV), NP@Pt(IV), and NP@Pt(IV)-E64d for 72 h. (c) CLSM images of cleaved caspase-8, GSDMD-N, cleaved caspase-3, GSDME-N expression (red) in MB49-R cells after different treatments (n = 3). Nuclei (blue, DAPI) and cytoskeleton (green, actin). Scale bar: 20 μm. (d) Schematic illustration of mechanism: CTSH inhibition by E64d activates caspase-8, which initiates pyroptosis via GSDMD and amplifies the Pt(IV)-triggered pyroptosis through caspase-3/GSDME. (e-f) Quantification of cleaved caspase-3 expression based on MFI in CLSM images (e) and on western blot band intensity adjusted to vinculin (f) (n = 3). (g) Western blot analysis of the full-length and cleaved forms of caspase-8, GSDMD, caspase-3, and GSDME expression levels

Table 1.

IC₅₀ (µM) of anticancer drugs in GP-R cells

Drugs	MB49-R (n = 3, Mean ± SD)	T24-R (n = 3, Mean ± SD)
Pt(IV)	6.3 ± 0.3	4.4 ± 0.1
NP@Pt(IV)	5.2 ± 0.4	2.6 ± 0.4
NP@Pt(IV)-E64d	2.9 ± 0.4	1.2 ± 0.1

Based on previous CLSM observations that co-treatment with Pt(IV) + E64d can rapidly induce morphological features of pyroptosis in GP-R cells, we further investigated the underlying mechanism. CLSM images revealed an abundant cytosolic accumulation of cleaved caspase-8 and pronounced activation of downstream GSDMD-N after NP@Pt(IV)-E64d treatment—levels higher than those with free Pt(IV) + E64d (Fig. 4c). This suggests that nanoparticle co-delivery more rapidly unloks CTSH-mediated inhibition of caspase-8, initiating GSDMD-dependent pyroptosis. We also assessed key markers of the caspase-3/GSDME pathway and observed significantly elevated fluorescence signals of cleaved caspase-3 in the nucleus and GSDME-N in the cytoplasm/membrane in the NP@Pt(IV)-E64d group (Figs. 4c; S6a-d). Semi-quantitative analysis (Figs. 4e; S6e-h) and western blot (WB) (Fig. 4f) consistently confirmed upregulation of cleaved caspase-8, GSDMD-N, cleaved caspase-3, and GSDME-N in NP@Pt(IV)-E64d group compared to NP@Pt(IV) and Pt(IV)+E64d groups.

These results demonstrate that NP@Pt(IV)-E64d by co-delivering Pt(IV) and E64d spatiotemporally, rapidly activates the caspase-8/GSDMD pyroptosis pathway. The activated caspase-8 subsequently activates caspase-3, thereby amplifying the Pt(IV)-induced caspase-3/GSDME pathway (Fig. 4 d).

NP@Pt(IV)-E64d-triggered pyroptosis remodels the immune landscape

Pyroptosis is a form of inflammatory programmed cell death. To investigate whether NP@Pt(IV)-E64d promotes immune landscape remodeling by inducing pyroptosis (Fig. 5a), we measured the release of ATP and a panel of inflammatory cytokines, including IL-18, IL-1β, TNF-α, IL-6, and IFN-β, from drug-treated MB49-R cells. The results showed that, compared to the Pt(IV) and NP@Pt(IV) groups, NP@Pt(IV)-E64d treatment significantly enhanced the release of ATP (53.4 ± 1.2 nmol/L), IL-18 (45.9 ± 1.4 pg/mL), IL-1β (3.9 ± 0.1 pg/mL), TNF-α (4.15 ± 0.8 pg/mL), IL-6 (36.4 ± 0.46 pg/mL), and IFN-β (9.3 ± 0.1 pg/mL). Specifically, relative to the NP@Pt(IV) group, NP@Pt(IV)-E64d increased extracellular levels of ATP by 20%, IL-18 by 10%, IL-1β by 10%, TNF-α by 67%, IL-6 by 11%, and IFN-β by 7%, respectively (Figs. 5b-d; S7a-c). CLSM imaging further revealed that NP@Pt(IV)-E64d markedly enhanced calreticulin (CRT) membrane expression (Figs. 5e; S8a-b) and reduced nuclear HMGB1 fluorescence intensity (Figs. 5 F; S8c-d), indicating the activation of key damage-associated molecular patterns (DAMPs). These findings suggest that NP@Pt(IV)-E64d-induced pyroptosis is accompanied by substantial release of inflammatory factors and DAMPs (including ATP, CRT, and HMGB1), potentially initiating immunogenic cell death (ICD) and laying the groundwork for antitumor immune activation. Concurrently, NP@Pt(IV)-E64d treatment promoted PD-L1 upregulation on tumor cells (Figures S8e-f), suggesting a compensatory immune suppression mechanism. This finding rationalizes the combination with anti-PD-1/PD-L1 therapy to achieve enhanced therapeutic efficacy.

Fig. 5.

The release of cytokines and DAMPs from NP@Pt(IV)-E64d-triggered pyroptosis remodels the immune landscape. (a) Schematic illustration of immune landscape remodeling: NP@Pt(IV)-E64d-triggered pyroptosis promotes DC maturation, M1 macrophage polarization through the release of cytokines (IL-1β, IL-18) and DAMPs (ATP, CRT, HMGB1), leading to compensatory PD-L1 upregulation. (b-d) Quantification of ATP (b), IL-1β (c), and IL-18 (d) in the culture supernatant of MB49-R cells after different treatments (n=3). (e-f) Representative CLSM images of CRT (e) and HMGB1 (f) expression in MB49-R cells after different treatments. Scale bar: 20 μm. (g-h) Flow cytometry analysis of (g) bone marrow-derived dendritic cell (BMDC) maturation (CD11c⁺CD80⁺CD86⁺ population) and (h) bone marrow derived macrophage (BMDM) M1/M2 ratio (CD80⁺CD206⁺ population).

We further evaluated the effect of this process on immune cell function. Bone marrow-derived dendritic cells (DCs) or macrophages were co-cultured with MB49-R cells pretreated with different drugs. Flow cytometry analysis showed that the NP@Pt(IV)-E64d group exhibited significantly higher DC maturation (36.8% vs. 25.2%) (Fig. 5 g) and M1/M2 macrophage ratio (3.6 vs. 2.2) (Fig. 5 h) compared to the NP@Pt(IV) group, representing approximately 1.5-fold and 1.6-fold increases, respectively (Figures S8g-h). The corresponding values in the PBS group were 11.7% and 0.9. In summary, these results demonstrate that NP@Pt(IV)-E64d-triggered pyroptosis and associated immunogenic signals are sufficient to initiate a potent innate immune response in vitro—namely, DC maturation and macrophage M1 polarization—strongly supporting its potential to activate antitumor immunity in vivo. In response to the concurrent induction of PD-L1 upregulation, combining NP@Pt(IV)-E64d with immune checkpoint inhibitors (anti-PD-L1 antibodies) may represent a promising strategy to enhance antitumor immune effects.

NP@Pt(IV)-E64d exhibited potent antitumor efficacy and promoted immune activation in the GP-R C57BL/6 model

To evaluate the antitumor activity and immunomodulatory potential of NP@Pt(IV)-E64d in immunocompetent mouse models, we established a GP-R model using C57BL/6 mice. On day 0, mice were treated via intravenous or intraperitoneal injection with PBS, Cispt, Pt(IV), Pt(IV) + E64d, NP@Pt(IV), or NP@Pt(IV)-E64d. Treatment efficacy and alterations in the tumor immune microenvironment were assessed on day 14 (Fig. 6a).

Fig. 6.

Antitumor efficacy and immune activation of NP@Pt(IV)-E64 in the GP-R C57BL/6 models. (a) Schematic of GP-R C57BL/6 establishment and treatment timeline. (b) Representative tumor images from each treatment group. Scale bars: 1 cm. (c) Tumor growth curves of GP-R-bearing mice in different treatment groups. (d) Final tumor weight measurements. (e) Body weight monitoring during treatment. (f) H&E staining, TUNEL assay of tumor tissues from different groups. Scale bars: 50 μm (H&E), 200 μm (TUNEL). (g) Flow cytometry analysis of CD4⁺/CD8⁺T cells in tumor. (h) Quantification of CD8⁺T cell ratios. (i) Quantification of mature (CD80⁺CD86⁺) DCs in tumor. (j) Quantification of mature (CD80⁺CD86⁺) DCs in lymph nodes. (k) Quantification of PD-1⁺T cells among CD8⁺T cells in tumor

In the evaluation of antitumor efficacy, NP@Pt(IV)-E64d exhibited the strongest inhibitory effect: it yielded the smallest endpoint tumor volume (Fig. 6b), and the tumor growth curve (Fig. 6c) indicated significantly superior suppression compared to the NP@Pt(IV) and Pt(IV) groups. Quantitative analysis further confirmed that the average tumor weight in the NP@Pt(IV)-E64d group was reduced by more than 90% compared to the PBS control group (Fig. 6 d). To assess the biosafety profile of this formulation, we monitored body weight changes and performed histopathological and biochemical analyses. Throughout the treatment period, mice treated with NP@Pt(IV)-E64d maintained stable body weight, with no statistically significant difference in terminal body weight compared to the PBS group (Fig. 6e), and performed significantly better than all free drug groups—indicating favorable systemic tolerability. Furthermore, H&E staining of major organs (heart, liver, spleen, lungs, and kidneys) revealed no obvious pathological abnormalities (Figure s10). Bochemical analysis showed no significant differences between the NP@Pt(IV)-E64d group and the PBS group in the levels of Alanine aminotransferase (ALT), creatinine (CREA), Creatine Kinase-MB (CK-MB), and Aspartate aminotransferase (AST) (Figures S9a-d), suggesting negligible liver, renal, or cardiac toxicity. Together, these results demonstrate the potent antitumor efficacy and excellent biosafety of NP@Pt(IV)-E64d. Histopathological analysis of tumor tissues revealed distinct changes across treatment groups. H&E staining showed looser tumor cell arrangement, increased nuclear pyknosis, and expanded necrotic areas in the NP@Pt(IV)-E64d group (Fig. 6f). TUNEL staining further confirmed significantly higher levels of cell death compared to the Pt(IV) and NP@Pt(IV) groups, corroborating the potent induction of tumor cell death by this formulation. Collectively, NP@Pt(IV)-E64d demonstrated both favorable biosafety and significant antitumor efficacy in the GP-R C57BL/6 model.

Regarding immunomodulation, flow cytometry results revealed that NP@Pt(IV)-E64d significantly enhanced the infiltration of CD8⁺T cells within the tumor (26.4 ± 8.9%), markedly higher than that in the Pt(IV) (13.0 ± 1.6%) and NP@Pt(IV) (14.1 ± 1.8%) groups (Figs. 6g), indicating effective activation of cytotoxic T cell responses. Concurrently, NP@Pt(IV)-E64d exhibited the highest proportion of mature dendritic cells in both tumors (48.2 ± 4.8%) (Figs. 6i; S11a) and lymph nodes (45.5 ± 8.0%) (Figs. 6j; S11b), suggesting enhanced antigen presentation capacity conducive to subsequent T cell activation. However, we also observed an elevated proportion of PD‑1⁺ CD8⁺ T cells (13.1 ± 1.3%) in the NP@Pt(IV)-E64d group, significantly higher than in the Pt(IV) (5.6 ± 0.6%) and NP@Pt(IV) (8.1 ± 0.4%) groups (Figs. 6k and k and 118c), indicating concomitant T cell exhaustion alongside immune activation. This result aligns with the PD‑L1 upregulation observed in vitro, suggesting that treatment-induced immune activation triggers compensatory immunosuppression. These findings provide a rationale for combining NP@Pt(IV)-E64d with immune checkpoint inhibitors to overcome immune suppression and enhance antitumor immunity.

NP@Pt(IV)-E64d synergized with αPD-L1 antibody to promote durable antitumor immunity in the GP-R recurrence model

Muscle-invasive bladder cancer (MIBC) frequently develops early pelvic lymph node metastases. These micrometastases often exhibit resistance to chemotherapy and are accompanied by an immunosuppressive tumor microenvironment, characterized by T cell exhaustion and upregulation of immune checkpoints such as PD-1/PD-L1. These factors limit the efficacy of monotherapies, underscoring the necessity for combination strategies to overcome immune evasion.

To evaluate the synergistic effect of NP@Pt(IV)-E64d with an αPDL-1 antibody, we established a GP-R recurrence model (Fig. 7a). The primary tumor was implanted in the right flank of mice on day −7. Mice received intravenous NP@Pt(IV)-E64d on days 1, 3, and 5, and intraperitoneal αPD-L1 on days 2, 4, and 6. To model distal recurrence, a secondary tumor was implanted in the left flank on day 7, and all treatments were discontinued. Analysis of tumor volume, weight, and growth curves on day 22 revealed that the combination of NP@Pt(IV)-E64d and αPDL-1 exhibited the strongest suppression of primary tumor growth (Figs. 7b). Crucially, this regimen also significantly inhibited the growth of the distal tumors established after treatment cessation (Figs. 7d), indicating the activation of a potent and sustained systemic antitumor immune response.

Fig. 7.

Combination therapy of NP@Pt(IV)-E64 and αPD-L1 induces durable antitumor immunity and prevents micrometastases in a GP-R recurrence model. (a) Schematic of the recurrence model and treatment regimen. (b) Representative images of primary tumors from each group. (c) Primary tumor growth curves. (d) Representative images of distant tumors. (e) Distant tumor growth curves. (f) Flow cytometry analysis of CD8⁺T cells (CD3⁺CD8⁺) in tumors. (g) Quantification of CD8⁺T cells populations. (h) Flow cytometry analysis of TCM (CD44⁺CD62L⁺) in spleen. (i) Quantification of TCM populations in spleen. (j) Quantification of TCM populations in lymph nodes. (k) Quantification of MDSC populations in tumor. (l) Quantification of LAG3⁺T cells populations in tumor. (m) Quantification of NK cells populations in tumor

To elucidate the underlying immunologic mechanisms, we analyzed the tumor immune microenvironment and peripheral lymphoid organs by flow cytometry. The combination of NP@Pt(IV)-E64d and αPDL-1 significantly enhanced the infiltration of CD8⁺ T cells within tumors (25.7 ± 5.5%) (Figs. 7f). More importantly, it substantially increased the frequency of central memory T (TCM) cells in the spleen (Figs. 7h) and lymph nodes (Figs. 7j; S12a) (11.8 ± 2.7% and 7.4 ± 1.4%, respectively), which were 5.1-fold and 1.5-fold higher, respectively, than in the αPD-L1 monotherapy group. This suggests that TCM cell-mediated long-lasting immunologic memory is key to controlling distal tumor outgrowth. Furthermore, the combination of NP@Pt(IV)-E64d and αPDL-1 effectively remodeled the immunosuppressive tumor microenvironment, as evidenced by a significant reduction in myeloid-derived suppressor cells (MDSCs) (Figs. 7k; S12b) and LAG3⁺T cells (Figs. 7 l; S12c), alongside a 3.8-fold increase in NK cell proportions compared to the PBS group (Figs. 7 m; Sd). These findings indicate that the strategy concurrently reverses immunosuppression and enhances innate immune cytotoxicity.

In summary, the combination of NP@Pt(IV)-E64d and αPD-L1 effectively controls both primary tumor progression and distal recurrence by synergistically activating durable, TCM-dependent immune memory, reversing immunosuppression, and potentiating innate immune function.

Discussion

The management of advanced bladder cancer, particularly the development of resistance to first-line gemcitabine-cisplatin regimens, remains an urgent and unmet clinical need. Confronting this challenge, we focused on CTSH—a key driver robustly linked to the clinical chemoresistant phenotype.

In gemcitabine-cisplatin–resistant (GP-R) bladder cancer cells, elevated cathepsin CTSH expression confers resistance by suppressing the TNF–caspase-8 signaling pathway, thereby blocking two key upstream routes for caspase-3 activation. This blockade inhibits GSDME cleavage and subsequent pyroptosis, enabling cancer cells to survive cisplatin-induced damage.

To overcome this resistance, we developed glutathione-responsive nanoparticles designed for the synchronous co-delivery of a Pt(IV) prodrug and the CTSH inhibitor E64d. This nanoplatform leverages the elevated intracellular glutathione levels in tumor cells to trigger targeted burst release, enhancing therapeutic efficacy while reducing off-target effects. Importantly, rather than simply increasing drug dosage, our strategy directly dismantles drug resistance at its source: the co-delivered E64d irreversibly inhibits CTSH activity. This positions caspase-8 as a central regulatory hub, initiating a dual pyroptotic signaling cascade. Specifically, activated caspase-8 directly cleaves GSDMD to trigger pyroptosis, while simultaneously restoring upstream signals required for caspase-3 activation. The reactivated caspase-3 then cooperates with Pt(IV) to potently activate GSDME-mediated pyroptosis. As a result, our system not only reinstates inflammatory programmed cell death but also restores the intrinsic efficacy of chemotherapy, enabling the synergistic eradication of resistant tumors.

Pyroptosis induced by NP@Pt(IV)-E64d serves as the upstream trigger that fundamentally links the nanoplatform’s direct cytotoxic effect to its subsequent remodeling of the immune microenvironment [28]. Specifically, this form of immunogenic cell death drives the robust release of DAMPs—including extracellular ATP, surface-exposed CRT, and HMGB1—along with pro-inflammatory cytokines such as IL-1β, IL-18, IL-6, TNF-α, and IFN-β. These danger signals function as the mechanistic bridge between the two key observations: they are the direct products of pyroptosis that, in turn, orchestrate the observed immune activation. This is evidenced by subsequent dendritic cell maturation and a marked increase in tumor-infiltrating CD8⁺ T cells, demonstrating that the initial cytotoxic event has been translated into a remodeled, immunologically active tumor microenvironment. Importantly, this potent immune activation simultaneously triggers an adaptive resistance mechanism—the upregulation of PD-L1 on tumor cells, coupled with elevated PD-1 expression on activated CD8⁺ T cells. This feedback loop not only validates the biological relevance of our induced immune response but also provides the mechanistic rationale for combination with αPD-L1 antibody. By strategically blocking this inhibitory axis, our approach unleashes a durable T cell-mediated antitumor response that significantly suppresses tumor recurrence, thereby establishing a causal continuum from nanoparticle-induced pyroptosis through immune activation to effective combination immunotherapy.

We must point out that Gasdermin (GSDM) family proteins (e.g., GSDMD, GSDME) mediate inflammatory, lytic cell death. This process results in cellular swelling and membrane rupture, releasing damage-associated molecular patterns (DAMPs; e.g., HMGB1, ATP) and pro-inflammatory cytokines (e.g., IL-1β, IL-18), thereby effectively activating adaptive immunity and generating robust anti-tumor immune responses. In contrast, in GP-R cells treated with NP@Pt(IV)-E64d, the substantial exposure of calreticulin (CRT) on the cell membrane appears to originate primarily from Pt(IV)-induced DNA damage and endoplasmic reticulum (ER) stress. This mechanism involves the Bax/Bak-mediated increase in ER membrane permeability and calcium release, which promotes CRT translocation from the ER lumen to the cell surface. There, CRT acts as an ‘eat-me’ signal, triggering anti-tumor immunity independently of the Gasdermin protein family. Notably, CTSH protein is highly expressed in GP-R cells and inhibits caspase-8 activation. Our study shows that E64d in the co-delivery system can effectively overcome this suppression and activate caspase-8. Activated caspase-8 cleaves the ER membrane protein BAP31, generating a p20BAP31 fragment that specifically recruits and activates Bax at the ER membrane [29]. This exacerbates ER stress and calcium release more efficiently and specifically, establishing a positive feedback loop that amplifies the CRT exposure signal. Therefore, through follow-up experiments, it will be crucial to investigate the respective contributions of the initial damage induced by Pt(IV) and the specific amplification signal provided by E64d via the caspase-8/p20BAP31 axis. This investigation will clarify the unique mechanism of our system and guide the synergistic optimization of components in such combination therapies.

We acknowledge several limitations in the current study. First, the in vivo therapeutic efficacy was evaluated using a subcutaneous xenograft tumor model. While this model is valuable for initial proof-of-concept studies, the tumor microenvironment, including vascularization, immune cell infiltration, and stromal interactions, may not fully recapitulate the complexity of orthotopic or metastatic disease. Future studies employing more clinically relevant models will be essential to validate these findings. Second, the long-term biodistribution and potential chronic toxicity of the nanomaterial require further investigation. The enhanced tumor accumulation of NP@Pt(IV)-E64d observed in our study is primarily attributed to the enhanced permeability and retention (EPR) effect—the theoretical foundation of passive targeting for nanomedicines. This mechanism leverages the high vascular permeability and impaired lymphatic drainage of tumors to facilitate nanoparticle extravasation and retention within tumor tissues, thereby increasing local drug concentration and enhancing antitumor efficacy, which is consistent with our in vivo experimental results. During the treatment period, NP@ Pt(IV)-E64d also exhibited preliminary in vivo safety: no significant differences were observed in body weight, liver and kidney function, or myocardial enzyme levels in tumor-bearing mice compared to the PBS group, and H&E staining of major organs revealed no apparent tissue structural damage. However, the in vivo fate of nanomedicines often represents a “double-edged sword.” Although the EPR effect promotes tumor-targeted drug delivery, systemically administered nanoparticles inevitably interact with the mononuclear phagocyte system (MPS). As the primary MPS organs, the liver and spleen constitute formidable anatomical barriers that rapidly capture and clear nanoparticles from the bloodstream. This clearance not only limits the amount of drug reaching the tumor site but also raises two concerns requiring long-term attention: First, the accumulation of drug-loaded nanoparticles in MPS-rich organs such as the liver may lead to potential chronic off-target toxicity [30]. Second, the MPS not only mediates acute nanoparticle clearance but may also serve as a long-term reservoir. Accumulating evidence suggests that even materials generally considered biocompatible can induce late-onset inflammatory responses or even fibrotic changes following prolonged retention in MPS organs [31]. Therefore, although our short-term data support the favorable preliminary safety profile of NP@ Pt(IV)-E64d, its long-term in vivo fate—including biodegradation behavior, clearance kinetics, and potential chronic organ toxicity—warrants in-depth investigation in future studies. Such research is essential to bridge the gap between promising preclinical efficacy and safe clinical translation.

In conclusion, Our study demonstrates that CTSH-targeted NP@Pt(IV)-E64d nanoparticles induce pyroptosis and achieve profound and sustained tumor suppression in authentic GP-R models, thereby establishing a novel therapeutic avenue capable of reversing chemoresistance where standard chemotherapy fails.

Conclusion

In this study, we engineered glutathione (GSH)-responsive nanoparticles for the co-delivery of the cathepsin H (CTSH) inhibitor prodrug E64d and a cisplatin prodrug Pt(IV) to implement a spatiotemporally coordinated “unlock-and-trigger” strategy for treating gemcitabine-cisplatin resistant (GP-R) bladder cancer. The nanoparticles overcome GP-R by synchronously unlocking CTSH inhibition and inducing DNA damage, thereby collaboratively triggering robust pyroptosis via both GSDMD and GSDME. This potent cell death ignites a profound inflammatory cascade, releasing DAMPs and cytokines that effectively convert the tumor microenvironment from an immunologically “cold” to a “hot” state. Consequently, the nanoparticle platform, particularly when combined with αPD-L1, fosters a favorable immune landscape—characterized by enhanced dendritic cell maturation, robust CD8⁺T cell infiltration, expansion of TCM cells in peripheral lymphoid organs, and a reduction in immunosuppressive populations—which collectively establishes durable immune memory and suppresses tumor recurrence. This “unlock-and-trigger” strategy, which re-sensitizes GP-R bladder cancer to chemo-immunotherapy, represents a promising and translatable nanoparticle-based strategy to overcome this form of clinical resistance.

Materials and methods

Synthesis and characterization of nanoparticles

Synthesis of PHHM

1, 2, 4, 5-Cyclohexanetetracarboxylic dianhydride (HPMDA, 246.6 mg, 1.1 mM) and 2-hydroxydecanoic acid (2-HD, 154.2 mg, 1 mM) were placed in a 50 mL round-bottom flask. N, N-Dimethylformamide (DMF, 10 mL) was added, and the mixture was stirred continuously for 48 h. Methoxy poly(ethylene glycol) (mPEG5k-OH, 1 g, 0.2 mmol) was then introduced, and the reaction proceeded for an additional 24 h. The final product, poly(2-HD-co-HPMDA)-mPEG (denoted as PHHM), was purified by dialysis and lyophilized under vacuum.

Synthesis of Pt(IV)

Step 1: Preparation of Pt(IV)-OH.

Cisplatin (CisPt, 1 g, 3.3 mmol) was added to a 100 mL round-bottom flask containing 30% hydrogen peroxide (H₂O₂, 36 mL). The flask was heated in an oil bath at 55 °C for 2 h, during which the solution turned yellow and turbid. The temperature was then raised to 100 °C, and the reaction continued for an additional 30 min until the solution gradually clarified. The flask was removed from the oil bath, allowed to react at room temperature for 30 min, and then stored at 4 °C overnight for crystallization. The resulting pale-yellow crystalline product, Pt(IV)-OH, was filtered, washed, and vacuum-dried overnight.

Step 2: Synthesis of Pt(IV).

Pt(IV)-OH (334 mg, 1 mmol) and dodecyl isocyanate (465 mg, 2.2 mmol) were added to a dry reaction flask. Anhydrous N, N-dimethylformamide (DMF, 10 mL) was introduced, and the mixture was heated at 55 °C in an oil bath for 12 h, yielding a transparent yellow solution. The DMF solvent was removed using a rotary evaporator. The product was precipitated by adding anhydrous methanol (1 mL) and ice-cold anhydrous diethyl ether (20 mL), and this purification step was repeated three times. The final product, Pt(IV), was obtained as a Light-gray powder after vacuum drying.

Synthesis of NP@Pt(IV)-E64d

PHHM (100 mg), E64d, and Pt(IV)-C12 (10 mg) were co-dissolved in 1 mL dimethyl sulfoxide (DMSO). The mixture was dropwise added to 10 mL distilled water under continuous stirring at room temperature for 10 min. DMSO was removed by dialysis (molecular weight cutoff: 3,500 Da). The supernatant was collected via centrifugation (3,000 rpm, 3 min) to obtain PHHM-E64d-Pt(IV) (denoted as NP@Pt(IV)-E64d). For the control nanoparticle (NP@Pt(IV)), the same procedure was repeated without E64d.

The morphology of NP@Pt(IV)-E64d was analyzed by transmission electron microscopy (TEM, Hitachi HT-7700, Japan), revealing a spherical structure with uniform size. Dynamic light scattering (DLS, Malvern Nano ZS90, UK) confirmed a hydrodynamic diameter of ~120 nm and a zeta potential of −18.6 mV. Inductively coupled plasma mass spectrometry (ICP-MS) quantified platinum concentrations in NP@Pt(IV) and NP@Pt(IV)-E64d as 151,065 µg/L (775 µM) and 202,971 µg/L (1,040 µM), respectively. High-performance liquid chromatography (HPLC) determined the drug loading capacity of E64d in NP@Pt(IV)-E64d. The molar ratio of Pt(IV) to E64d in NP@Pt(IV)-E64d was approximately 1:1.

Cellular uptake of nanoparticles

Cellular uptake was evaluated using confocal laser scanning microscopy (CLSM) and flow cytometry. MB49-R cells (8 × 10⁴ cells/well) were seeded into 24-well plates with coverslips and treated with NP@Cy5.5 (2 µM Pt) for 1, 4, or 7 h. After PBS washing, cells were fixed with 4% paraformaldehyde, stained with DAPI (nucleus, blue) and Alexa Fluor 488-phalloidin (cytoskeleton, green), and imaged using a CLSM (LSM-800, Zeiss, Germany) to visualize Cy5.5 (red) fluorescence. For flow cytometry, MB49-R cells (3 × 10⁵ cells/well) were treated similarly, harvested, and analyzed for Cy5.5 fluorescence intensity.

Cellular platinum uptake

MB49-R cells (5 × 10⁵ cells/well) were seeded in 6-well plates and treated with CisPt, Pt(IV), NP@Pt(IV), or NP@Pt(IV)-E64d (5 µM Pt) for 1, 4, or 7 h. Cells were washed with ice-cold PBS, lysed with concentrated nitric acid, and intracellular platinum content was quantified by atomic absorption spectroscopy (AAS).

Platinum release kinetics of NP@Pt(IV)-E64d

To evaluate the platinum (Pt) release profile of NP@Pt(IV)-E64d, phosphate-buffered saline (PBS, pH 7.4) and PBS containing 10 mM glutathione (GSH) were prepared as release media (50 mL each). NP@Pt(IV)-E64d dispersion (2 mL) was loaded into pre-swollen dialysis bags (molecular weight cutoff: 3.5 kDa) and immersed in 50 mL of the respective medium under constant agitation (37 °C). At predetermined time points, 1 mL of the external dialysate was collected and replaced with an equal volume of fresh medium to maintain sink conditions. The cumulative Pt release was quantified by atomic absorption spectroscopy (AAS) and calculated using the formula: Cumulative release (%) = (Total Pt in dialysate/Total Pt loaded in NP@Pt(IV)-E64d) × 100%.

Biodistribution in chemoresistant tumor-bearing mice

NP2@Cy7.5 was prepared by conjugating NP with Cy7.5 dye (mass ratio 1:5) via 2-h stirring (4 °C, dark), followed by purification using Zeba^™ spin desalting columns (Thermo Scientific). NP2@Cy7.5 (3 mg Pt/kg) was intravenously injected into chemoresistant mice. In vivo fluorescence signals (ex/em: 745/800 nm) were captured using an IVIS Spectrum imaging system (PerkinElmer) at designated time points. Ex vivo organ imaging (heart, liver, spleen, lung, kidney, tumor) was performed 48 h post-injection, and fluorescence intensities were quantified with Living Image software.

NP@Pt(IV)-E64d -mediated inhibition of resistant cell viability

Cell viability was assessed via MTT assay. T24-R and MB49-R cells (3,000 cells/well) in 96-well plates were treated with CisPt, Pt(IV), NP@Pt(IV), or NP@Pt(IV)-E64d (0.05–40 µM Pt) for 72 h. MTT solution (5 mg/mL, 10 µL/well) was added, followed by 4-h incubation. Formazan crystals were dissolved with 10% SDS, and absorbance (570 nm, reference 650 nm) was measured using a SpectraMax microplate reader. Cell viability (%) was calculated as: (Experimental OD570–650/Control OD570–650) × 100.

Cell and animal models

Establishment of chemoresistant bladder cancer cell lines

T24 and MB49 (chemosensitive parental lines) were seeded into 100 mm dishes at a density of 2 × 10⁶ cells/dish. After 12 h of adhesion, cells were treated with 2 µM cisplatin and 0.2 µM gemcitabine for 48 h. Floating dead cells were discarded, and adherent viable cells were collected and cultured in drug-free medium until normal proliferation resumed. A portion of cells was cryopreserved at −80 °C, while the remaining cells underwent repeated drug exposure/recovery cycles for 60 days (60 × 24 h) to establish stable chemoresistant sublines (T24-R and MB49-R).

Chemoresistant bladder cancer animal model

Male C57BL/6 mice (4 weeks old) were acclimatized for 1 week, followed by subcutaneous injection of 100 µL Matrigel-saline suspension (1:1, v/v) containing 1.5 × 10⁶ MB49 cells. When tumors reached ~100 mm³, mice were randomized into PBS control (intraperitoneal PBS) and CisPt + Gem treatment groups (cisplatin [3 mg/kg] + gemcitabine [50 mg/kg], administered intraperitoneally once weekly for 5 weeks). Tumor length (L) and width (W) were measured weekly, and tumor volume was calculated as (L × W²)/2 to monitor chemoresistance development.

Bilateral chemoresistant bladder cancer (recurrence) model

On day −7, primary tumors (2 mm³ fragments) were implanted into the right flank of 5-week-old male C57BL/6 mice. On day 0 (tumor volume: 80–100 mm³), mice were randomized into four groups (n = 5): PBS, NP@Pt(IV)-E64d (3 mg Pt/kg, i.v., days 1/3/5), α-PD-L1 antibody (20 mg/kg, i.p., days 2/4/6), and NP@Pt(IV)-E64d + α-PD-L1. Distant metastatic tumors were established on the left flank (day 7) using fresh tumor fragments. Tumor volumes and body weights were monitored every 2 days. Mice were euthanized on day 22 for immune cell infiltration analysis (flow cytometry).

Detection of ATP (find-me signal, ICD biomarker)

MB49-R cells (4 × 10⁵ cells/well) were seeded in 12-well plates and treated with PBS, CisPt, Pt(IV), Pt(IV)+E64d, NP@Pt(IV), or NP@Pt(IV)-E64d (5 µM Pt/E64d) for 12 h. Supernatants were collected, centrifuged (1,000 ×g, 3 min), and extracellular ATP levels were quantified using an Enhanced ATP Assay Kit (Beyotime) according to the manufacturer’s protocol. Fluorescence intensity was measured with a SpectraMax microplate reader (Molecular Devices, USA).

Evaluation of maturation of bone marrow-derived dendritic cells (BMDCs) in vitro

Bone marrow cells isolated from femurs of 5–6-week-old male C57BL/6 mice were cultured in RPMI-1640 medium containing 10% FBS, 20 ng/mL GM-CSF, and 10 ng/mL IL-4 (Beyotime) for 5 days (37 °C, 5% CO₂). Pretreated MB49-R cells (PBS, CisPt, Pt(IV), Pt(IV)+E64d, NP@Pt(IV), or NP@Pt(IV)-E64d; 5 µM Pt) were co-cultured with BMDCs at a 10:1 ratio (tumor cells: BMDCs) for 24 h. BMDCs were stained with anti-CD11c-PE, anti-CD80-FITC, and anti-CD86-APC antibodies (30 min, 4 °C, dark) and analyzed by flow cytometry (CytoFLEX, Beckman Coulter). Maturation was quantified as CD80⁺CD86⁺/CD11c⁺ cell percentage.

Western blotting

MB49-R cells (1 × 10⁶ cells/well) were seeded in 6-well plates and treated with PBS, CisPt, Pt(IV), Pt(IV)+E64d, NP@Pt(IV), or NP@Pt(IV)-E64d (5 µM Pt/E64d) for 12 h. Cells were lysed in RIPA buffer containing protease/phosphatase inhibitors (30 min, ice), and supernatants were collected after centrifugation (12,000 ×g, 15 min). Protein concentrations were quantified using a BCA assay (Beyotime), and equal amounts (40 µg) were separated by SDS-PAGE (Bio-Rad Mini-PROTEAN system) and transferred to PVDF membranes (0.22 μm). Membranes were blocked, incubated with primary antibodies (e.g., anti-Caspase-3, 1:1000, Cell Signaling Technology) overnight at 4 °C, washed with TBST, and probed with HRP-conjugated secondary antibodies (1:1000, 1 h, room temperature). Bands were visualized using ECL substrate (Beyotime, P0018AS) and imaged with an Amersham Imager 600 (GE Healthcare).

Flow cytometry of tissue immune cells

Single-cell suspensions from tumors, lymph nodes, and spleens were stained with fluorochrome-conjugated antibodies: anti-CD3-PE, anti-CD4-APC, anti-CD8-FITC (T cells); anti-CD11c-PE, anti-CD80-FITC, anti-CD86-APC (DCs); anti-NK1.1-PE (NK cells); and anti-CD62L-APC/CD44-FITC (central memory T cells). Data were acquired using a CytoFLEX flow cytometer (Beckman Coulter) and analyzed with CytExpert software.

Statistical analysis

Quantitative data are expressed as mean ± standard deviation (SD). Differences between two groups were assessed using unpaired two-tailed Student’s t-tests (for normally distributed data) or Wilcoxon rank-sum tests (for non-normally distributed data). Correlations were analyzed using Spearman’s rank correlation. All analyses were performed in GraphPad Prism 9.5. Replicates: Values of n refer to the number of biologically independent experiments (defined as experiments performed on separate cell culture passages/mouse cohorts). Reporting: Exact n values for each experiment are provided in the corresponding figure legends. Results of statistical tests are reported in the figure legends.

Author contributions

J.W.: Conceptualization, Writing - original draft, Investigation, Data curation, Methodology. B.Z.: Writing - original draft, Visualization, Validation. H.Z.: Investigation, Methodology, Project administration. Y.T.: Validation, Visualization. H.H.: Validation, Visualization. H.X.: Resources. B.G.: Writing - review & editing, Conceptualization, Methodology, Funding acquisition, Supervision. All authors reviewed the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 82460158); Guangxi Key Research and Development Program (Grant No. Guike AB25069043); The Natural Science Foundation of Guangxi Zhuang Autonomous Region (Grant No. 2023GXNSFAA026169); Guilin Scientific Research and Technology Development Plan (Grant No. 20230135-1-1).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All animal experiments complied with institutional guidelines and were approved by the Animal Ethics Committee of of Guilin Medical University (Approval No.GLMC20241012). The CT image in Figure. 1b was provided by The Second Affiliated Hospital of Guilin Medical University with written informed consent from patient.

Competing interests

The authors declare no competing interests.