A bladder microenvironment-adaptive ceria nanozyme hydrogel enables urea-responsive intravesical chemo-immunotherapy

Abstract

Bladder cancer intravesical therapy remains limited by rapid urinary washout, insufficient penetration across the bladder mucosal barrier, and an immunosuppressive tumor microenvironment. Here, we developed a bladder microenvironment-adaptive chemo-immunotherapy system, Gel/UCeNZ/Gem/CpG, to improve local bladder cancer treatment. In this system, urease-modified CeO₂ nanozymes (UCeNZ), gemcitabine (Gem), and CpG oligodeoxynucleotide (CpG ODN) were incorporated into an injectable thermosensitive adhesive hydrogel. After intravesical administration, the hydrogel rapidly formed a mucosa-adherent depot, thereby prolonging bladder residence and enabling sustained therapeutic release. Meanwhile, UCeNZ responded to endogenous urinary urea to enhance particle transport and mucosal penetration, improving local drug exposure beyond passive instillation. Functionally, CeO₂-mediated redox regulation amplified Gem-induced oxidative stress, mitochondrial dysfunction, apoptosis, and immunogenic cell death-associated signaling, as evidenced by HMGB1 translocation/release, calreticulin exposure, and intracellular ATP dysregulation. CpG ODN further promoted dendritic cell maturation and strengthened T cell-mediated antitumor immunity. In an orthotopic bladder cancer model, Gel/UCeNZ/Gem/CpG significantly suppressed tumor progression, prolonged survival, increased CD8⁺ T cell infiltration, reduced regulatory T cells and M2-like macrophages, and enhanced memory/effector-like T cell activation without obvious systemic toxicity. Transcriptomic analysis further revealed coordinated regulation of oxidative stress, apoptosis, antigen presentation, cytokine signaling, T cell activation, PD-1/PD-L1 checkpoint signaling, and p53-related tumor-suppressive pathways. Overall, this study presents a bladder-adaptive intravesical chemo-immunotherapy strategy that integrates mucosal retention, urea-responsive penetration, redox-amplified immunogenic chemotherapy, and immune microenvironment remodeling for bladder cancer treatment.

Highlights

• •Intravesical hydrogel couples mucosal retention with urea-responsive delivery.
• •UCeNZ/Gem amplifies ROS-mediated stress and ICD-associated signaling.
• •Depot-based CpG incorporation boosts DC and CD8⁺ T-cell activation.
• •Gel/UCeNZ/Gem/CpG remodels the immunosuppressive bladder tumor milieu.
• •Transcriptomics links redox stress, antigen presentation, and T-cell activation.

1. Introduction

Bladder cancer (BC) remains a major clinical challenge because of its high recurrence rate and the risk of progression from non-muscle-invasive disease to muscle-invasive malignancy [1,2]. Non-muscle-invasive bladder cancer (NMIBC), which accounts for approximately 75% of newly diagnosed BC cases, is commonly managed by transurethral resection of bladder tumor (TURBT) followed by intravesical instillation of chemotherapeutic or immunomodulatory agents, such as gemcitabine, mitomycin, or Bacillus Calmette–Guérin (BCG) [[3], [4], [5], [6], [7]].

Although intravesical administration places therapeutic agents directly in the bladder, effective tumor exposure remains difficult to maintain. Continuous urine production and periodic voiding dilute and rapidly remove instilled drugs, while the short dwell time and glycosaminoglycan (GAG)-rich mucosal layer restrict their penetration into tumor tissue [1,4,5,8,9]. These delivery barriers are further compounded by an immunosuppressive tumor microenvironment characterized by redox imbalance, inefficient antigen presentation, limited effector immune-cell infiltration, and the accumulation of suppressive immune cells [1,10,11]. Thus, increasing the drug dose or prolonging bladder residence alone is unlikely to achieve optimal therapeutic efficacy. An effective local delivery system should also facilitate tissue penetration, intensify tumor-cell stress, and promote antitumor immunity.

Local drug delivery systems (LDDS) have emerged as promising strategies to improve the therapeutic index of intravesical treatment by prolonging local drug exposure and reducing systemic toxicity [4,12,13]. Strategies developed to overcome the bladder mucus/GAG barrier mainly include mucoadhesive formulations that resist urinary washout, mucus-penetrating nanocarriers with optimized physicochemical properties, stimulus-responsive systems, and actively propelled micro/nanomotors [14]. Among them, injectable hydrogels are particularly attractive because they can be administered in a minimally invasive manner and form local depots in situ [4,5,15,16]. Thermosensitive hydrogel systems, including chitosan/β-glycerophosphate (CS/β-GP)-based formulations, can undergo sol–gel transition at physiological temperature, while dopamine-modified polymers such as HA-DA provide catechol-mediated mucosal adhesion and improved resistance to urinary flushing [1,15,17]. These properties make adhesive thermosensitive hydrogels well suited for intravesical depot formation. However, prolonged residence does not necessarily ensure efficient tissue penetration. Therapeutic agents released from hydrogel depots still depend predominantly on passive diffusion to cross the mucus/GAG barrier and reach deeper tumor regions. Thus, a hydrogel that only prolongs retention may not fully address the spatial delivery barrier of bladder cancer therapy.

To complement hydrogel-mediated retention with active transport, urease-modified nanocarriers have recently attracted attention for intravesical delivery because they can respond to endogenous urea in urine [3,10,16]. Urease catalyzes the decomposition of urea into ammonia and carbon dioxide, generating local chemical gradients that promote autonomous particle motion without an external energy source. Compared with passive diffusion, this urea-powered motion can enhance particle dispersion, mucus traversal, and tumor penetration in the bladder. Recent dual-source-powered Au–Pt@ur nanomotors further combined urease-driven motion in the urea-rich bladder lumen with Pt-catalyzed H₂O₂ decomposition in tumor tissue, thereby improving mucus traversal and deep tumor penetration during intravesical photothermal therapy [18]. By harnessing a native constituent of urine, urease-powered nanocarriers turn the very fluid responsible for rapid drug washout into a source of propulsion. Yet overcoming the physical barriers to drug delivery is only the first step. Effective tumor control also depends on potent tumor cell killing and the activation of antitumor immunity. This calls for an intravesical system in which prolonged mucosal residence and urea-driven transport are coupled with therapeutic functions that act within the tumor.

Beyond improving intravesical retention and tumor penetration, effective treatment also requires robust tumor cell killing and subsequent immune activation. Redox-active cerium oxide (CeO₂) nanozymes provide a versatile means of regulating tumor-associated oxidative processes through the reversible Ce³⁺/Ce⁴⁺ cycle [[19], [20], [21], [22], [23], [24], [25], [26], [27]]. Their SOD-like activity catalyzes the conversion of superoxide radicals (O₂•^-) into H₂O₂, thereby reducing superoxide levels and generating H₂O₂ as a redox intermediate. In contrast, their CAT-like activity decomposes H₂O₂ into H₂O and O₂, reducing H₂O₂ accumulation while improving oxygen availability [[24], [25], [26]]. Through these complementary reactions, CeNZ can regulate ROS composition and oxygen availability, with their therapeutic function depending on the surrounding redox environment and treatment modality. In bladder cancer, this catalytic property has been exploited in ICG@R11-CeO₂ nanoparticles, in which R11 improved mucosal adhesion, mucus penetration, and tumor targeting, while CeO₂-mediated H₂O₂ decomposition supplied O₂ to alleviate hypoxia and enhance ICG-mediated photodynamic therapy [25].

CeNZ-mediated redox regulation may also be integrated with chemotherapy and immunotherapy to promote immunogenic tumor destruction. Gemcitabine (Gem), a clinically used chemotherapeutic agent for bladder cancer, induces DNA damage and can promote immunogenic cell death (ICD), accompanied by the release of tumor antigens and danger-associated molecular patterns (DAMPs). CpG oligodeoxynucleotide (CpG ODN), a Toll-like receptor 9 (TLR9) agonist, can further promote dendritic cell maturation and cytotoxic T-cell responses [1,5,15,16]. The combination of CeNZ-mediated ROS regulation, Gem-induced tumor cell damage, and CpG-mediated immune stimulation therefore provides a potential strategy for linking local tumor destruction with antitumor immune activation.

Herein, we designed a bladder microenvironment-adaptive bioactive hydrogel–nanozyme system, Gel/UCeNZ/Gem/CpG, for intravesical bladder cancer chemo-immunotherapy. The adhesive thermosensitive hydrogel forms a mucosa-retained depot to prolong local therapeutic exposure, while urease-modified CeO₂ nanozymes (UCeNZ) use endogenous urinary urea to promote active transport and tumor penetration. Following delivery to the tumor, the UCeNZ/Gem combination enhanced intracellular ROS accumulation, mitochondrial dysfunction, and immunogenic cell death, while CpG promoted dendritic cell maturation and T cell-mediated antitumor immunity. Through this integrated design, Gel/UCeNZ/Gem/CpG coordinates bladder retention, urea-powered penetration, CeNZ/Gem-mediated tumor cell stress, and immune activation (Scheme 1). Its therapeutic efficacy and underlying mechanisms were evaluated in vitro and in an orthotopic bladder cancer model, with transcriptomic analysis further supporting treatment-associated changes in oxidative stress, apoptosis, antigen presentation, T-cell activation, and tumor-suppressive signaling.

Scheme 1.

Gel/UCeNZ/Gem/CpG enables bladder microenvironment-adaptive intravesical chemo-immunotherapy. (A) Schematic construction of Gel/UCeNZ/Gem/CpG through CeNZ amination, urease conjugation, Gem/CpG loading, and hydrogel incorporation. (B) After intravesical administration, the thermosensitive adhesive hydrogel forms a mucosa-adherent depot and releases UCeNZ/Gem/CpG locally. UCeNZ responds to urinary urea to enhance particle transport and mucosal penetration. CeO₂-mediated POD-like activity amplifies ROS generation, while Gem induces DNA damage, resulting in oxidative stress, mitochondrial dysfunction, ICD-associated signaling, CRT exposure, HMGB1 release, and intracellular ATP dysregulation. CpG further promotes dendritic cell activation and CD8⁺ T-cell activation. (C) The treatment remodels the tumor immune microenvironment by increasing immune infiltration and reducing immunosuppressive features, including decreased Tregs and CD206⁺ M2-like macrophage-associated signals. (D) The integrated intravesical chemo-immunotherapy strategy promotes tumor inhibition, immune activation, and prolonged survival.

2. Materials and methods

2.1. Materials and synthesis

All chemical reagents and cell culture supplies were purchased from commercial suppliers and used as received. CeNZ were synthesized via a hydrothermal method and subsequently surface-modified with polydopamine and urease to obtain urease-conjugated CeO₂ (UCeNZ). Gemcitabine and CpG oligonucleotides were loaded into UCeNZ via adsorption. Dopamine-modified hyaluronic acid (HA-DA) was synthesized through carbodiimide-mediated coupling. A thermosensitive injectable hydrogel was formulated by combining chitosan, β-glycerophosphate, HA-DA, and genipin under mild conditions. Detailed synthesis protocols are provided in the Supporting Information.

2.2. Characterization of nanomaterials and hydrogel

The physicochemical characteristics of nanoparticles and hydrogels were analyzed using transmission electron microscopy (TEM), dynamic light scattering (DLS), zeta potential measurements, Fourier-transform infrared spectroscopy (FTIR), and ¹H nuclear magnetic resonance (NMR). Rheological behavior, injectability, sol-gel transition, swelling, degradation, and drug release profiles were also evaluated. Enzymatic-like peroxidase (POD) activity of UCeNZ was assessed using TMB substrates, and hemocompatibility was tested via hemolysis assays and red blood cell morphology. Detailed methods are provided in the Supporting Information.

2.3. In vitro biological evaluation

Cytotoxicity and biocompatibility of CeNZ, UCeNZ, and urease were tested on L929 fibroblasts and MBT2 bladder cancer cells via CCK-8 assay. Cellular uptake and trans-barrier penetration were evaluated using doxorubicin-labeled nanoparticles in Transwell and confocal systems. The antitumor efficacy of hydrogel extracts was assessed by cell viability, live/dead staining, apoptosis, migration, ROS generation, mitochondrial membrane potential, and ICD-associated markers [calreticulin (CRT), high mobility group box 1 (HMGB1), adenosine triphosphate (ATP)]. Bone marrow-derived dendritic cells (BMDCs) maturation and cytokine release (TNF-α, IFN-γ) were also measured following indirect co-culture with treated MBT2 cells. Detailed experimental protocols are presented in the Supporting Information.

2.4. In vivo therapeutic evaluation

An orthotopic bladder tumor model was established in female C57BL/6 mice using luciferase-expressing MBT2 cells. Tumor growth was monitored using IVIS imaging. Mice received weekly intravesical instillations of different hydrogel formulations. Antitumor efficacy was evaluated by tumor burden, survival, histological analysis (H&E, TUNEL, CD3, CD4, CD8, CD86, CD206, Ki67, CD31), and flow cytometric profiling of immune cell subsets. Intravesical retention, systemic biosafety (blood biochemistry, histology), and dose optimization were also assessed. Methods are described in detail in the Supporting Information.

Flow cytometry data were analyzed using FlowJo software. To evaluate dendritic cell maturation in mouse lymph nodes, the main cell population was first gated according to FSC-A/SSC-A profiles, followed by doublet exclusion using FSC-A versus FSC-H. CD11c⁺ dendritic cell-like populations were then gated, and the expression of the costimulatory molecules CD80 and CD86 was analyzed within the CD11c⁺ population. Mature DC-like cells were defined as CD11c⁺CD80⁺CD86⁺ cells.

For T-cell subset analysis in lymph nodes or spleen, lymphocytes were first gated based on FSC-A/SSC-A profiles, and single cells were subsequently selected using FSC-A/FSC-H. CD3⁺CD4⁺ T cells and CD3⁺CD8⁺ T cells were quantified using CD3/CD4 or CD3/CD8 gating. For regulatory T-cell analysis, CD3⁺CD4⁺ T cells were further gated, and Treg cells were identified as CD4⁺Foxp3⁺ cells. For the analysis of T-cell activation and memory phenotypes, CD3⁺ T cells were first selected and then divided into CD4⁺ and CD8⁺ T-cell subsets. Further analysis of CD44 and CD62L expression in CD4⁺ or CD8⁺ T cells was conducted to determine the proportions of the CD44⁺CD62L⁺, CD44⁺CD62L⁻, CD44⁻CD62L⁺, and CD44⁻CD62L⁻ populations. Detailed gating strategies for all flow cytometry panels are shown in Figures S1-S4.

2.5. Combination therapy and transcriptomic analysis

In order to evaluate the synergistic effect of chemotherapy-immunotherapy, mice received Gel/UCeNZ/Gem/CpG or CpG combined preparations. Analyze the drainage of lymph nodes and spleen through flow cytometry to characterize the immune activation. Transcriptome analysis of tumor tissue is carried out to study its potential molecular mechanism, including differential gene expression and pathway enrichment [Kyoto Gene and Genome Encyclopedia (KEGG), Gene Ontology (GO)]. For more details, please refer to the Supporting Information.

2.6. Statistical analysis

Unless otherwise stated, all quantitative data are expressed as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 9.5 software. One-way ANOVA was used for comparisons among multiple groups, followed by Tukey's post-hoc multiple comparison test. Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. The number of biological replicates or the number of animals for each experiment are indicated in the corresponding legend. In in vitro experiments, n represents the number of independent biological replicates; in in vivo experiments, n represents the number of mice. The statistical test used for each dataset is specified in the corresponding figure legend. Exact p values are provided in the figure legends; p values smaller than 0.0001 are reported as p < 0.0001. Statistical significance was defined as p < 0.05. Significance levels are denoted as follows: ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; and ****p < 0.0001.

3. Results and discussion

3.1. Functional construction and validation of UCeNZ as a urea-responsive redox nanozyme module

To address the limitations of traditional intravesical formulations in tumor permeability and insufficient redox-mediated tumor stress, we constructed UCeNZ as the core of a Gel/UCeNZ/Gem/CpG drug delivery system. In this design, the CeNZ core is intended to provide redox-modulating activity, while the introduction of urease coupling is to leverage the urea-rich environment of the bladder to enhance particle transport (Fig. 1A). Therefore, the characterization of UCeNZ aims not only to verify the success of surface grafting modification but also to verify whether the nanoparticles can meet the functional requirements of intravesical chemoimmunotherapy, including colloidal stability, drug/adjuvant loading, catalytic redox activity, and urea-responsive motility.

Fig. 1.

Construction and functional validation of UCeNZ for redox regulation and urea-responsive transport. (A) Schematic illustration of the preparation of UCeNZ formulations and subsequent Gem/CpG loading. (B) Representative TEM images showing the morphology of CeNZ at different modification stages, including CeNZ, CeNZ-NH₂, and UCeNZ. Scale bar: 100 nm. (C) Hydrodynamic size distribution and (D) zeta potential changes confirming sequential surface modification. (E) Colloidal stability of UCeNZ in PBS and urine, as evaluated by time-dependent changes in particle size and PDI. (F) FTIR spectra confirming surface functionalization and Gem loading. (G) Loading capacity and encapsulation efficiency of Gem and CpG in UCeNZ. (H) POD-like activity of UCeNZ nanoparticles at various concentrations, monitored by absorbance changes after 10 min incubation. (I) Time-dependent POD-like catalytic activity of UCeNZ at different concentrations. (J) UV–vis absorption spectra recorded for the evaluation of SOD-like activity at different UCeNZ concentrations. (K) Corresponding SOD inhibition rates. (L) Representative motion trajectories of nanoparticles with (blue) and without (red) urease modification in urea-containing medium. (M) Mean square displacement (MSD) curves derived from the particle trajectories. (N) Average particle speeds and (O) diffusion coefficients calculated from trajectory analysis.

TEM images showed that CeNZ maintained its spherical morphology after a series of surface modifications, and a thin, relatively transparent modified layer was observed around UCeNZ, indicating that urease was successfully immobilized on the nanoparticle surface (Fig. 1B). Consistent with this, DLS analysis showed that the hydrodynamic diameter of CeNZ gradually increased from 145.3 ± 1.4 nm in the original CeNZ to 150.5 ± 1.5 nm after amino modification, and further increased to 175.4 ± 4.9 nm after urease grafting (Fig. 1C). Zeta potential measurements further confirmed the success of the surface grafting modification, with the surface charge changing from +19.9 mV in the original CeNZ to −13.5 mV after surface modification, and remaining negative (−8.88 mV) after urease grafting, consistent with the introduction of organic and protein surface layers (Fig. 1D). These results demonstrate that UCeNZ possesses stable nanoscale dimensions and modified surface chemistry, making it suitable for subsequent drug loading and biological applications.

Because intravesically administered formulations are directly exposed to urine, maintaining colloidal stability under bladder-relevant conditions is essential for nanoparticle dispersion and transport. The modest difference in the initial hydrodynamic diameter of UCeNZ between artificial urine and PBS likely reflects the medium-dependent effects on particle hydration and colloidal interactions. Nevertheless, UCeNZ showed only minor size variations in both media throughout the observation period, while the PDI remained below 0.2, indicating good colloidal stability (Fig. 1E). This stability may minimize premature aggregation during bladder residence and help maintain nanoparticle dispersion following release from the hydrogel depot, thereby supporting subsequent local distribution and tumor penetration. FTIR spectra further verified the sequential surface modification and Gem loading. The characteristic Ce–O vibration was observed near 500 cm⁻¹ in all CeNZ-based formulations. After surface modification, broad absorption bands corresponding to –OH/–NH₂ stretching appeared in the 3300–3500 cm⁻¹ region, while urease conjugation introduced amide-related peaks at 1533 and 1489 cm⁻¹, confirming successful enzyme immobilization. After Gem loading, a new absorption peak at 785 cm⁻¹, attributable to pyrimidine ring vibrations, was detected in UCeNZ/Gem, supporting the successful incorporation of Gem (Fig. 1F). The amount of conjugated urease was further estimated by thermogravimetric analysis. The final residual masses of CeNZ, CeNZ–NH₂, and UCeNZ were 85.08%, 80.90%, and 65.69%, respectively. After correction for the intrinsic thermal weight loss of the CeNZ–NH₂ carrier, the urease content in UCeNZ was estimated to be 18.80 wt%, corresponding to approximately 231.5 mg of urease per gram of CeNZ–NH₂ (Fig. S5).

The full UV–vis spectrum and calibration curve of Gem are shown in Fig. S6. Quantitative analysis showed that the Gem loading capacity and encapsulation efficiency of UCeNZ/Gem were 13.73 ± 1.06% and 47.78 ± 4.27%, respectively (Fig. 1G). In addition to chemotherapy loading, efficient CpG adsorption is essential for coupling tumor cell killing with immune activation. The CpG-loading ability of UCeNZ was first evaluated independently. The PDA-derived surface provides catechol and amine groups capable of forming multiple mild non-covalent interactions, including hydrogen bonding and π–π stacking, with nucleic acid bases, thereby allowing CpG-ODN to associate with the nanoparticle surface without additional chemical modification. Based on UV absorbance at 595 nm using a CpG-ODN–Texas Red calibration curve (Fig. S7), the loading capacity and encapsulation efficiency of UCeNZ/CpG were 1.11 ± 0.05% and 88.79 ± 4.23%, respectively (Fig. 1G).

We then evaluated whether sequential co-loading affected the loading performance of either cargo. CpG-ODN was adsorbed onto the preformed UCeNZ/Gem nanoparticles, and both unbound CpG and Gem released during this secondary loading step were quantified. After co-loading, the final loading capacities of Gem and CpG in UCeNZ/Gem/CpG were 12.46 ± 0.20% and 1.18 ± 0.01%, respectively (Fig. S8). Compared with the individually loaded formulations, the Gem loading capacity decreased only moderately from 13.73 ± 1.06% to 12.46 ± 0.20%, with 90.63 ± 1.66% of the preloaded Gem retained after CpG adsorption. Meanwhile, the CpG loading capacity remained comparable to that obtained during individual loading, while its encapsulation efficiency was 94.27 ± 1.25% in the co-loaded formulation. These results indicate that sequential CpG adsorption caused only limited Gem displacement and that the presence of preloaded Gem did not compromise CpG adsorption. Thus, UCeNZ enables the effective co-loading of Gem and CpG, supporting the integration of Gem-mediated tumor cell killing with CpG-mediated immunostimulation.

Next, we evaluated whether UCeNZ retained the catalytic redox activity required to enhance tumor stress. In the presence of H₂O₂ and TMB, the absorbance of UCeNZ at 650 nm increased in a concentration-dependent manner, indicating its peroxidase-like catalytic activity and ROS generation capacity (Fig. 1H). Kinetic analysis further showed that absorbance increased over time, and the catalytic rate increased with higher UCeNZ concentrations (Fig. 1I). Considering that bladder tumors typically exhibit elevated oxidative stress levels and increased H₂O₂ concentrations, this POD-like activity may contribute to enhancing oxidative damage and synergistically interact with gemcitabine-induced cytotoxicity. Simultaneously, we also observed SOD-like activity, with UV spectroscopy showing full-wavelength absorption and concentration-dependent increases in superoxide radical scavenging capacity in inhibition experiments (Fig. 1J and K). These results collectively confirm the redox modulatory capacity of UCeNZ, indicating that CeNZ is not merely an inert nanocarrier, but an active catalytic component capable of modulating tumor-related oxidative stress.

To further investigate the motility of urease-modified particles, we analyzed the movement of nanoparticles in a urea environment. Compared to unmodified nanoparticles, UCeNZ exhibited longer trajectories and larger displacements, indicating enhanced responsiveness to urea (Fig. 1L). Mean square displacement analysis showed that the slope of UCeNZ was steeper than that of the unmodified group, indicating enhanced diffusivity rather than simple passive Brownian motion (Fig. 1M). Quantitative analysis also confirmed that urease modification significantly improved the average velocity and diffusion coefficient of nanoparticles (Fig. 1N and O). These results support the material design of urease modification enabling nanoparticles to utilize urea as a biochemical signal to enhance local motility. However, although these motility experiments were conducted without the addition of H₂O₂, the potential influence of CeNZ-mediated tumor-associated H₂O₂ decomposition cannot be ruled out, warranting further investigation. The resulting chemically enhanced motility may contribute to the distribution and penetration of particles in the bladder mucosa and tumor microenvironment.

The hemocompatibility of UCeNZ was examined before its incorporation into the intravesical formulation. Hemolysis remained below 5% at all tested concentrations (Fig. S9A), while erythrocytes retained their characteristic biconcave morphology without discernible membrane damage or aggregation (Fig. S9B). These findings demonstrate the favorable hemocompatibility of UCeNZ under the tested conditions and support its subsequent incorporation into the hydrogel delivery system.

3.2. Construction and functional characterization of a thermosensitive adhesive hydrogel for intravesical depot formation

After validating UCeNZ as a urea-responsive redox nanozyme, we constructed a thermosensitive adhesive hydrogel as an intravesical drug reservoir for the Gel/UCeNZ/Gem/CpG system. The hydrogel was designed to reduce rapid drug loss during urination and prolong local drug exposure. HA-DA, CS, and β-GP were combined to provide catechol-mediated wet adhesion and temperature-responsive gelation, while genipin was added as a mild secondary crosslinker to improve hydrogel stability after administration (Fig. 2A) [28,29]. In this design, the hydrogel adheres to the bladder mucosa, undergoes in situ gelation at body temperature, and sustains drug release, thereby providing sufficient local retention for UCeNZ-mediated tissue penetration.

Fig. 2.

Thermosensitive adhesive hydrogel provides intravesical retention and sustained drug release. (A) Schematic illustration of the HA-DA/CS/β-GP/genipin hydrogel network and its in situ gelation and mucosal adhesion after intravesical administration. (B) FTIR spectra of HA and HA-DA. (C) ¹H NMR spectra of HA and HA-DA. (D) FTIR spectra of HA-DA/CS/β-GP and HA-DA/CS/β-GP/genipin hydrogels, supporting genipin-mediated crosslinking. (E) Macroscopic images showing thermosensitive sol–gel transition and injectability of hydrogel formulations. (F) Ex vivo bladder mucosal adhesion and flushing resistance of the hydrogel, demonstrating its ability to maintain local attachment under a hydrated intravesical environment. (G) Temperature-dependent rheological analysis showing sol–gel transition near physiological temperature. (H) Shear-thinning behavior of hydrogel formulations, supporting their injectability for intravesical administration. (I) Swelling and (J) degradation profiles of hydrogels, indicating controlled network stability. (K) Cumulative Gem release profiles from hydrogel formulations. (L) CpG release curve from Gel/UCeNZ/Gem/CpG. (M) In vivo bladder retention behavior of Gel/CeNZ/ICG and Gel/UCeNZ/ICG hydrogel after intravesical instillation for 24 h. (N) Quantitative statistics on the depth of intravesical ICG penetration.

Having established UCeNZ as a urea-responsive, redox-regulating nanozyme, we constructed a thermosensitive adhesive hydrogel as the intravesical depot of the Gel/UCeNZ/Gem/CpG system. The hydrogel was designed to prolong local therapeutic exposure by limiting rapid urinary washout. HA-DA, CS, and β-GP were combined to couple catechol-mediated wet adhesion with temperature-responsive gelation, while genipin provided mild secondary crosslinking to gradually stabilize the network after administration (Fig. 2A) [28,29]. Together, these features were intended to anchor the formulation to the bladder mucosa, establish a sustained-release d epot, and provide sufficient local residence for nanoparticle-mediated tissue penetration.

HA-DA was characterized by FTIR and ¹H NMR spectroscopy. Compared with HA, HA-DA showed an absorption band at 1735.2 cm⁻¹ in the FTIR spectrum (Fig. 2B). In the¹H NMR spectrum, signals at δ = 6.6–6.9 ppm were assigned to the aromatic protons of the dopamine catechol group (Fig. 2C) [30,31] The appearance of these dopamine-related signals indicated that dopamine had been introduced into HA. HA-DA was subsequently used as the adhesive component of the hydrogel.

We then examined the formation of the genipin-reinforced HA-DA/CS/β-GP hydrogel network. Compared with the HA-DA/CS/β-GP formulation, the genipin-containing hydrogel displayed a new absorption peak at 1628.09 cm⁻¹, which was assigned to imine (C=N) stretching generated by Schiff base formation between genipin and amino groups on CS (Fig. 2D) [29]. Meanwhile, the broad –OH/–NH stretching band around 3174.05 cm⁻¹ decreased in intensity, and an additional band at 1047.06 cm⁻¹ corresponding to C–N stretching appeared. These spectral changes indicate that genipin-mediated covalent crosslinking was successfully introduced into the hydrogel network. Such secondary crosslinking is expected to improve the cohesiveness and structural integrity of the hydrogel after intravesical administration.

The temperature-responsive gelation behavior was then evaluated macroscopically. The HA-DA/CS/β-GP precursor underwent sol–gel transition at 37 °C and returned to the sol state upon cooling to 4 °C, confirming the reversible thermosensitive behavior of the CS/β-GP network. In contrast, after genipin-mediated secondary crosslinking, the gel state was largely maintained even after cooling, indicating the formation of a more stable covalent network. During gelation, the genipin-containing hydrogel gradually turned pale blue-green, a characteristic color change of the genipin–amine reaction that visually indicated ongoing crosslinking [29]. Before secondary crosslinking was complete, the hydrogel precursor remained smoothly injectable through a syringe, supporting its use for intravesical instillation (Fig. 2E). At the selected concentration of 0.1%, genipin reacts gradually with chitosan under mild conditions. The CS/β-GP system therefore governs the initial injectability and rapid sol–gel transition, while genipin progressively strengthens the hydrogel network after gel formation at 37 °C [32,33].

Because the bladder mucosa is continuously exposed to urine, the hydrogel must remain attached under wet and periodically flushed conditions. In the ex vivo adhesion assay, the hydrogel readily adhered to the luminal surface of the bladder and remained in place after simulated saline flushing (Fig. 2F). This finding is notable because the highly hydrated, GAG-rich mucosal layer generally limits the adhesion of foreign materials and promotes the clearance of conventional instillation solutions. The catechol groups of HA-DA likely contributed to this adhesion through hydrogen bonding, electrostatic interactions, and π–π stacking with the mucosal surface [30,31]. Lap-shear measurements further confirmed that both HA-DA/CS/β-GP and genipin-reinforced HA-DA/CS/β-GP hydrogels adhered to wet tissue, while genipin reinforcement improved the mechanical stability of the hydrogel (Table S1). Together, these results show that the hydrogel can maintain contact with the bladder mucosa under simulated flushing conditions.

Rheological measurements showed how thermosensitive gelation and genipin crosslinking contributed sequentially to hydrogel formation. During the temperature sweep, both formulations began to gel at approximately 35 °C and formed complete gels between 37 and 38 °C (Fig. 2G). The hydrogel precursor could therefore remain fluid during injection and rapidly gel after warming to body temperature. The addition of genipin increased the storage modulus (G′) during heating, indicating that secondary crosslinking had already begun as the temperature increased. Genipin also produced a moderate increase in viscosity without altering the shear-thinning behavior of the hydrogel (Fig. 2H). Thus, the hydrogel remained injectable under shear, underwent rapid CS/β-GP-mediated gelation at body temperature, and was subsequently strengthened by gradual genipin crosslinking.

Swelling and degradation assays were used to examine how genipin affected hydrogel stability. At 37 °C, the hydrogel without genipin absorbed more water and showed pronounced swelling, whereas the genipin-crosslinked hydrogel exhibited a significantly lower swelling ratio (Fig. 2I). This difference is consistent with the formation of a denser network, which limits water uptake and helps the hydrogel retain its shape in the bladder. A similar effect was observed in the degradation assay. The hydrogel without genipin underwent more rapid mass loss over 28 days, while the genipin-crosslinked hydrogel degraded more slowly (Fig. 2J) [29]. Thus, gradual genipin crosslinking reduced excessive swelling and slowed hydrogel degradation after gel formation without compromising its initial injectability.

Genipin crosslinking also affected the release of Gem from the hydrogel. Without genipin, cumulative Gem release reached approximately 90% by day 14, whereas the genipin-crosslinked hydrogel released approximately 80% over the same period (Fig. 2K). This slower release was consistent with the reduced swelling and slower degradation of the more densely crosslinked hydrogel. This slower release is likely attributable to the denser hydrogel network, which restricts molecular diffusion and reduces burst release. When evaluated separately, CpG also exhibited sustained release from the corresponding hydrogel formulation (Fig. 2L), confirming that the hydrogel matrix could regulate the release of both therapeutic components.

We further examined the simultaneous release of Gem and CpG from the final co-loaded Gel/UCeNZ/Gem/CpG formulation. As shown in Fig. S10, both Gem and CpG were released progressively over 14 days without complete burst release. Gem showed a relatively faster release profile, increasing from approximately 26% on day 1 to 76% on day 14. In comparison, CpG release increased from approximately 10% on day 1 to 52% on day 14. Thus, the sustained-release behavior was preserved after co-loading, while the two drugs exhibited distinct release rates. The Gem release level from the co-loaded formulation was also close to that observed from the individually loaded genipin-reinforced hydrogel, indicating that CpG incorporation did not markedly disrupt hydrogel-mediated Gem release under the tested conditions.

The different release rates of Gem and CpG are consistent with their distinct molecular properties and modes of association with UCeNZ. Their release from Gel/UCeNZ/Gem/CpG is expected to proceed through a two-step matrix-mediated process rather than a single burst-release mechanism. First, Gem and CpG must desorb from or diffuse out of UCeNZ. Gem is a small-molecule chemotherapeutic agent physically associated with UCeNZ mainly through adsorption-related interactions and can therefore diffuse outward relatively rapidly along the concentration gradient. In contrast, CpG-ODN is a larger oligonucleotide adsorbed onto the PDA/amine-modified UCeNZ surface through multiple non-covalent interactions, including hydrogen bonding, electrostatic interactions, and π–π stacking with nucleic acid bases. These multivalent interactions provide stronger surface retention and are consistent with the slower CpG release observed in the co-loaded formulation.

Second, after dissociation from UCeNZ, both Gem and CpG must diffuse through the HA-DA/CS/β-GP/genipin hydrogel network. The hydrated hydrogel matrix, together with genipin-mediated secondary crosslinking, restricts molecular diffusion and reduces burst release, whereas hydrogel swelling, network relaxation, and gradual degradation permit continued release over time. Therefore, the release profiles of the co-loaded formulation reflect the combined regulation of nanoparticle-associated desorption and hydrogel-mediated diffusion and erosion. This hierarchical delivery design enables relatively faster Gem availability for tumor cell killing while maintaining more prolonged CpG release for subsequent immunostimulation.

Agarose gel electrophoresis further confirmed that CpG remained structurally stable after loading and storage. Freshly prepared UCeNZ/CpG completely inhibited CpG migration, indicating successful nanoparticle association, whereas the 14-day incubated sample showed both nanoparticle-associated CpG and an intact CpG band, suggesting partial release without obvious CpG degradation (Fig. S11). These results support the ability of the hydrogel depot to maintain local drug/adjuvant availability while preserving CpG integrity.

To determine whether UCeNZ could promote tissue penetration while the hydrogel prolonged bladder retention, fluorescence imaging of bladder sections was performed 24 h after intravesical instillation. Fluorescence from Gel/CeNZ/ICG remained largely confined to the mucosal layer, whereas Gel/UCeNZ/ICG extended beyond the mucosa and into the deeper bladder wall (Fig. 2M). Quantitative analysis showed that urease modification increased the penetration depth by approximately 150 μm relative to the non-urease group (Fig. 2N). These findings show that the hydrogel and UCeNZ perform complementary functions: the hydrogel prolongs local retention and drug release, while UCeNZ promotes transport from the mucosal surface into the underlying bladder tissue. By combining these effects, Gel/UCeNZ/Gem/CpG can reduce urinary washout while improving drug penetration into the bladder wall.

3.3. Biocompatibility and urea-responsive cellular delivery of UCeNZ/DOX

Before examining UCeNZ-mediated cellular delivery, we evaluated the cytocompatibility of the nanoparticles and urease. CeNZ and UCeNZ had little effect on the viability of L929 fibroblasts at concentrations ranging from 1 to 100 μg/mL, while UCeNZ caused a moderate reduction only at 200 μg/mL (Fig. 3A). Free urease also showed minimal cytotoxicity toward L929 fibroblasts and MBT2 bladder cancer cells at concentrations up to 800 μg/mL (Fig. 3B). These results supported the use of CeNZ, UCeNZ, and urease at the selected working concentrations in the subsequent delivery experiments.

Fig. 3.

Biocompatibility and penetration behavior of UCeNZ in vitro. (A) Cell viability of L929 fibroblasts after 24 and 48 h exposure to varying concentrations of CeNZ and UCeNZ nanoparticles, assessed by CCK-8 assay. (B) Cell viability of L929 and MBT2 cells after treatment with increasing concentrations of urease for 24 and 48 h. (C) Schematic illustration of UCeNZ/DOX penetration through the Transwell barrier in the presence (+) or absence (−) of 0.1 M urea stimulation. (D) Fluorescence images of the lower chamber after 4 h incubation with UCeNZ/DOX under different urea conditions (red fluorescence: DOX; scale bar: 50 μm). (E) Quantitative analysis of fluorescence intensity in the lower chamber, indicating urea-triggered enhancement of penetration. (F) Flow cytometry histograms of MBT2 cells incubated with UCeNZ/DOX at different time points, and (G) corresponding quantification of intracellular mean fluorescence intensity (MFI). (H) Confocal fluorescence images of MBT2 cells at 0.5, 1, 2, 4, and 6 h post-treatment with UCeNZ/DOX (scale bar: 50 μm), and (I) fluorescence intensity quantification of intracellular DOX accumulation over time.

We next examined whether the increased motility of UCeNZ in urea, as observed in Fig. 1L–O, improved the delivery of nanoparticle-associated cargo. A Transwell uptake model was used, with DOX serving as a fluorescent model cargo because its red fluorescence allows intracellular accumulation to be directly visualized and quantified. In MBT2 cells, the addition of 0.1 M urea markedly increased intracellular fluorescence from UCeNZ/DOX compared with the urea-free condition (Fig. 3C and D). Quantitative analysis showed a more than sixfold increase in intracellular DOX accumulation (Fig. 3E). Thus, the urea-dependent motility observed in the particle-tracking experiments was accompanied by greater cargo delivery to tumor cells.

The uptake of UCeNZ/DOX by MBT2 cells was then monitored from 0.5 to 6 h. Flow cytometry showed a steady increase in intracellular fluorescence, with the highest signal detected at 6 h (Fig. 3F and G). Confocal microscopy revealed the same trend, as intracellular DOX fluorescence became progressively stronger over time (Fig. 3H); this increase was also confirmed by quantitative image analysis (Fig. 3I). These results show that UCeNZ/DOX accumulates in bladder cancer cells in an exposure time-dependent manner, which may increase the intracellular availability of the loaded drug during intravesical treatment.

This urea-responsive delivery behavior is particularly relevant to bladder cancer therapy. Intravesically administered drugs and nanoparticles are rapidly diluted or eliminated by urine and must overcome the GAG-rich mucosal barrier and tight urothelial architecture to reach tumor lesions [4]. In this context, urease modification provides a bladder-specific strategy by utilizing endogenous urinary urea as a biochemical cue. As demonstrated by the motion trajectory, MSD, average speed, and diffusion coefficient analyses in Fig. 1L–O, UCeNZ displayed enhanced motility under urea-containing conditions. The present cellular uptake results further show that this enhanced transport behavior can translate into improved cargo accumulation in MBT2 cells. Therefore, UCeNZ functions not merely as a passive nanocarrier, but as a urea-responsive delivery module that helps address the penetration and uptake barriers of intravesical therapy.

In vivo fluorescence imaging was used to determine whether the adhesive hydrogel could convert UCeNZ from a readily cleared suspension into a locally retained delivery system. Free ICG was almost completely cleared from the bladder within 30 min, whereas CeNZ/ICG showed a more persistent but gradually decreasing signal (Fig. S12A and B). Gel/UCeNZ/ICG exhibited the longest retention, maintaining approximately 80% of its initial bladder fluorescence at 24 h. This result is consistent with the hydrogel forming an adhesive depot that protects the encapsulated nanoparticles from rapid urinary washout.

Ex vivo imaging showed that ICG-associated fluorescence was detected mainly in the bladder, liver, and kidneys (Fig. S12C and D). Gel/UCeNZ/ICG produced a stronger bladder signal and lower renal fluorescence than CeNZ/ICG, consistent with slower loss of the fluorescent cargo from the bladder. Because ICG may dissociate from the nanoparticles, however, the organ fluorescence cannot be interpreted as the distribution of intact CeNZ. Nevertheless, when considered together with the motility, cellular uptake, and bladder-wall penetration results, these findings suggest that prolonged bladder retention gives UCeNZ more time to contact the mucosa and penetrate the underlying tissue in the presence of urinary urea.

3.4. Redox-driven mitochondrial dysfunction mediates the enhanced antitumor activity of Gel/CeNZ/Gem/CpG in vitro

We hypothesized that the enhanced antitumor efficacy of Gel/CeNZ/Gem/CpG is mediated by redox imbalance–driven mitochondrial dysfunction, in which CeNZ modulates intracellular redox homeostasis and amplifies Gem-induced cytotoxic stress. To test this hypothesis, MBT2 bladder cancer cells were treated with the indicated formulations, followed by systematic evaluation of cell viability, apoptosis, migration, intracellular ROS levels, and mitochondrial membrane potential.

Gel alone showed negligible cytotoxicity, whereas Gel/CeNZ induced a moderate reduction in cell viability (Fig. 4A). The incorporation of Gem significantly enhanced antitumor activity, and the strongest effect was observed in the Gel/CeNZ/Gem/CpG group, in which cell viability decreased to below 20%. Live/dead staining further confirmed this trend, with progressively increased dead-cell signals in Gem-containing formulations and the most extensive cell death in the Gel/CeNZ/Gem/CpG group (Fig. 4B and C). Flow cytometric analysis of apoptosis showed a consistent pattern, with both early and late apoptotic populations significantly increased in Gem-containing groups (Fig. 4D and E). Gel/CeNZ/Gem/CpG induced the highest apoptosis rate (68.28%), indicating a synergistic cytotoxic effect of redox modulation and chemotherapeutic stress.

Fig. 4.

In vitro evaluation of antitumor efficacy and intracellular responses induced by hydrogel formulations. (A) Cell viability of MBT2 cells treated with different formulations, assessed by CCK-8 assay. (B) Live/dead staining images of treated MBT2 cells (green: Calcein-AM, red: PI). Scale bar: 100 μm. (C) Quantification of live/dead cell ratio based on staining results. (D) Flow cytometric analysis of apoptosis in MBT2 cells after treatment, using Annexin V-FITC/PI double staining. (E) Statistical analysis of early and late apoptosis proportions. (F) Transwell migration assay showing the inhibitory effects of various formulations on MBT2 cell migration. (G) Quantification of the migration ratio relative to control. (H) Intracellular ROS generation detected by DCFH-DA staining and 3D surface fluorescence plots. Scale bar: 100 μm. (I) Flow cytometry histograms of DCFH-DA fluorescence intensity in different groups. (J) Quantitative analysis of mean fluorescence intensity (MFI) from flow cytometry results. (K) JC-1 staining of mitochondrial membrane potential (ΔΨm) in MBT2 cells; red: JC-1 aggregates (healthy mitochondria), green: JC-1 monomers (depolarized mitochondria). Scale bar: 50 μm. (L) Quantification of JC-1 monomer fluorescence intensity to evaluate ΔΨm disruption.

Given the cytotoxicity of Gel/CeNZ/Gem and Gel/CeNZ/Gem/CpG, the cell migration shown in the transwell results may be associated with a decrease in cell number. Gel/CeNZ/Gem/CpG resulted in the lowest number of migrated cells, with a relative migration rate of 4.13% (Fig. 4F and G), which primarily reflected the reduced viable cell population rather than intrinsic changes in migratory capacity.

We next investigated whether redox imbalance contributes to the observed cytotoxicity. Intracellular ROS levels were assessed using DCFH-DA staining. CeNZ- and Gem-containing formulations significantly increased ROS production, with the highest fluorescence intensity observed in the Gel/CeNZ/Gem/CpG group (Fig. 4H–J). Importantly, NAC pretreatment markedly reduced ROS accumulation induced by Gel/CeNZ/Gem/CpG, confirming that the system induces a strong oxidative stress response [27,34].

To determine whether ROS acts upstream of mitochondrial dysfunction, mitochondrial membrane potential (ΔΨm) was evaluated using JC-1 staining. Gel/CeNZ/Gem/CpG caused a pronounced shift from red aggregates to green monomers, indicating mitochondrial depolarization (Fig. 4K and L). Notably, NAC pretreatment partially restored ΔΨm, as evidenced by reduced monomer fluorescence and recovery of aggregate signals. This rescue effect demonstrates that ROS acts upstream of mitochondrial dysfunction in this system, and that mitochondrial damage is at least partially ROS-dependent [35,36].

Gel/CeNZ/Gem/CpG reduced MBT2 cell viability to below 20% and increased the apoptosis rate to 68.28%. These changes were accompanied by increased intracellular ROS and marked mitochondrial depolarization. NAC lowered the DCFH-DA fluorescence and partially restored mitochondrial membrane potential, suggesting that the mitochondrial damage arose, at least in part, from the elevated intracellular ROS.

The marked reduction in cell viability appeared after Gem was introduced, while the further increase in ROS following CeNZ incorporation suggests that CeNZ strengthened the oxidative component of Gem-induced cellular stress. The cytotoxicity of Gel/CeNZ/Gem/CpG may therefore result from the combined effects of chemotherapy and redox perturbation rather than from either process alone. We next examined whether this oxidative tumor cell damage was accompanied by immunogenic changes and whether CpG could further promote the response of dendritic cells.

3.5. Gel/CeNZ/Gem/CpG induces ROS-dependent ICD-associated changes and promotes dendritic cell maturation

Gel/CeNZ/Gem/CpG increased intracellular ROS, disrupted mitochondrial function, and induced apoptosis in MBT2 cells. Based on these findings, we further investigated whether such oxidative and mitochondrial stress was accompanied by immunogenic cell death (ICD)-related molecular alterations and subsequent dendritic cell (DC) activation. ICD is characterized by the exposure or release of damage-associated molecular patterns (DAMPs), including HMGB1 translocation and calreticulin (CRT) surface exposure, which facilitate antigen uptake and presentation by DCs [37,38]. Accordingly, HMGB1 and CRT were first assessed in treated MBT2 cells, followed by evaluation of BMDC maturation and cytokine secretion in an indirect co-culture system. Intracellular ATP levels were also measured as an indicator of mitochondrial metabolic status.

HMGB1 immunofluorescence showed strong nuclear localization in untreated MBT2 cells, consistent with its physiological role as a nuclear chromatin-associated protein (Fig. 5A) [36]. Gem-containing treatments reduced nuclear HMGB1 staining, with the most pronounced change observed in the Gel/CeNZ/Gem/CpG group. Quantitative analysis confirmed a significant decrease in nuclear HMGB1 fluorescence in this group (Fig. 5B). NAC pretreatment attenuated this decrease, indicating that ROS contributed to HMGB1 redistribution and reduced nuclear retention.

Fig. 5.

Gel/CeNZ/Gem/CpG induce ICD and promote dendritic cell maturation. (A) Immunofluorescence staining of HMGB1 release in MBT2 cells after treatment with various nanoparticle formulations. (B) Quantification of HMGB1 fluorescence intensity, indicating extracellular translocation. (C) Immunofluorescence detection of CRT surface exposure on MBT2 cells following different treatments. (D) Quantitative analysis of CRT fluorescence intensity. (E) Intracellular ATP levels in MBT2 cells after nanoparticle treatment, determined by luminescence-based assay. (F) Flow cytometry analysis of BMDCs co-cultured with nanoparticle-treated MBT2 cells, showing expression levels of CD11c, CD80, and CD86. (G) Schematic representation of the Transwell-based co-culture system and experimental workflow. (H) Quantitative analysis of mature BMDCs (CD11c⁺CD80⁺CD86⁺) proportions. ELISA quantification of BMDC-secreted cytokines IFN-γ (I) and TNF-α (J), reflecting immune activation in response to different treatments.

CRT exposure was then evaluated as another hallmark of ICD. During ICD, CRT translocates from the endoplasmic reticulum to the cell surface and serves as an “eat-me” signal to facilitate the recognition and phagocytosis of dying tumor cells by dendritic cells [24,26]. In control cells, CRT signal was weak and predominantly intracellular, whereas Gem-containing formulations promoted its redistribution toward the cell membrane (Fig. 5C). Gel/CeNZ/Gem/CpG induced the strongest surface-associated CRT exposure among all groups, and quantitative analysis confirmed a significant increase compared with other treatments (Fig. 5D). NAC pretreatment significantly attenuated CRT exposure induced by Gel/CeNZ/Gem/CpG, further supporting the ROS-dependent regulation of this ICD-associated event.

Intracellular ATP levels were then measured to further evaluate mitochondrial energy metabolism after treatment. Because ATP production is closely associated with mitochondrial function, changes in intracellular ATP can reflect mitochondrial metabolic stress. Compared with the control and Gel, Gem-containing formulations disrupted intracellular ATP homeostasis, and the Gel/CeNZ/Gem/CpG group induced the most pronounced alteration in intracellular ATP levels (Fig. 5E). This result is consistent with the ROS accumulation and mitochondrial membrane potential loss observed in Section 3.4, suggesting that Gel/CeNZ/Gem/CpG causes mitochondrial metabolic dysfunction.

The reduced nuclear retention of HMGB1, together with increased surface-associated CRT, supports the induction of an ICD-associated phenotype in MBT2 cells after Gel/CeNZ/Gem/CpG treatment. This response is consistent with the preceding ROS and mitochondrial results, suggesting that CeNZ-associated redox regulation acted together with Gem-induced cellular damage to promote immunogenic tumor cell stress. Previous studies have shown that CeO₂ nanozymes can alter intracellular ROS through their redox catalytic activity and thereby influence chemotherapy-induced tumor cell damage [[37], [38], [39]]. NAC attenuated both the change in HMGB1 localization and CRT exposure, indicating that ROS was involved in these ICD-associated events rather than being only a consequence of cell death.

To determine whether the ICD signals generated by treated tumor cells could promote antigen-presenting cell activation, immature BMDCs were exposed to supernatants from differently treated MBT2 cells using an indirect co-culture system (Fig. 5G). Flow cytometry analysis showed that free Gem and Gel alone resulted in relatively low proportions of CD80⁺CD86⁺ mature DCs (8.50% and 12.71%, respectively), with no significant difference between the two groups (Fig. 5F–H). The modest response observed in the Gel group may be related to the reported immunomodulatory activity of the chitosan-based matrix [[40], [41], [42]]. In contrast, Gel/CeNZ/Gem significantly increased the mature DC population to 23.88%, consistent with the enhanced ICD-associated signals observed in MBT2 cells. The incorporation of CpG further increased this proportion to 48.71%, nearly sixfold higher than that of the untreated control. This enhancement is attributable to the established TLR9-mediated immunostimulatory activity of CpG, which promotes DC activation and antigen-presenting function and thereby supports downstream T-cell-mediated antitumor immunity [[43], [44], [45]].

The enhanced DC maturation was further supported by cytokine secretion analysis. The Gel/CeNZ/Gem/CpG group significantly increased the secretion of IFN-γ and TNF-α compared with the other groups (Fig. 5I and J), indicating activation of a pro-inflammatory immune response. IFN-γ and TNF-α are important cytokines involved in innate and adaptive antitumor immunity, and their elevation is consistent with enhanced DC activation and subsequent T cell immune priming [11,35]. Therefore, the combination of ROS-dependent ICD-associated DAMP signaling and CpG-mediated TLR9 stimulation effectively converted dying bladder cancer cells into immunostimulatory substrates capable of promoting DC maturation and inflammatory cytokine production.

Consistent with the flow-cytometry results, Gel/CeNZ/Gem/CpG induced the highest secretion of IFN-γ and TNF-α (Fig. 5I and J). The elevation of these cytokines further supports BMDC activation after exposure to the treated MBT2 cells [11,35]. When considered alongside the NAC-sensitive changes in HMGB1 localization and CRT exposure, these findings outline a sequence in which CeNZ/Gem-induced oxidative and mitochondrial stress is accompanied by ICD-associated signaling, while CpG further strengthens the response of DCs. This connection between tumor cell damage and DC activation provided the rationale for subsequently examining the antitumor immune response in vivo.

3.6. Safe intravesical depot formation and enhanced orthotopic bladder tumor suppression by Gel/UCeNZ/Gem

Before evaluating antitumor efficacy in vivo, we optimized the volume of hydrogel administered into the bladder, as an insufficient volume might be cleared too rapidly, whereas an excessive volume could occupy too much of the bladder lumen. Volumes of 50, 75, and 100 μL were tested for gel formation, intravesical retention, and short-term voiding tolerability (Fig. S13A). All three volumes formed visible gels in the bladder. However, the 50 μL hydrogel was cleared relatively rapidly, while the 100 μL hydrogel left considerable residual material at 24 h (Fig. S13B). The 75 μL hydrogel remained in the bladder for an appreciable period without the excessive residual occupancy observed with 100 μL. Urine output was comparable among the groups, suggesting that none of the tested volumes caused detectable short-term impairment of urine passage (Fig. S13C). Body weight remained stable, and H&E staining showed no evident abnormalities in the major organs (Fig. S13D and E), providing additional evidence of short-term tolerability. Because 75 μL provided longer retention than 50 μL while leaving less residual gel than 100 μL, it was selected for subsequent intravesical treatment.

To determine whether repeated intravesical administration caused local bladder irritation, healthy mice received four weekly instillations of PBS or the indicated hydrogel formulations. H&E staining on day 32 showed intact urothelial architecture in all groups, with no evident mucosal erosion, tissue disruption, or marked inflammatory cell infiltration (Fig. S14A). Urinary IL-6 and IL-1β levels varied slightly among the groups, but none of these differences reached statistical significance (Fig. S14B and C). Neither the incorporation of UCeNZ and Gem nor the addition of CpG increased either cytokine relative to PBS. Thus, under the tested dosing schedule, repeated hydrogel administration produced no detectable histological injury or sustained elevation of urinary inflammatory cytokines at the day-32 endpoint. Because the cytokines were measured only at this endpoint, short-lived inflammatory responses after individual instillations cannot be excluded.

After selecting the administration volume, the antitumor efficacy of Gel/UCeNZ/Gem was evaluated in an orthotopic bladder cancer model (Fig. 6A and B). Tumor-bearing mice received one intravesical treatment per week according to the experimental schedule. The 7-day interval allowed repeated local treatment while avoiding more frequent catheterization. Importantly, this interval should not be interpreted as evidence that the hydrogel remained in the bladder for 7 days. The degradation and Gem release assays were performed under static in vitro conditions and reflect the intrinsic degradation of the hydrogel and diffusion of the drug through its network. In the bladder, urine production, periodic voiding, bladder contraction, mucosal movement, and fluid exchange are expected to accelerate hydrogel erosion and drug clearance. Each administration was therefore intended to prolong mucosal contact for a limited period rather than maintain the hydrogel in the bladder until the next treatment.

Fig. 6.

In vivo therapeutic efficacy of Gel/UCeNZ/Gem in an orthotopic bladder cancer model. (A) Schematic illustration of the intravesical hydrogel instillation procedure. (B) Experimental timeline for orthotopic tumor establishment and intravesical treatment. (C) Representative bioluminescence images of tumor-bearing mice in each treatment group at the indicated time points. (D) Quantification of tumor bioluminescence intensity. (E) Kaplan–Meier survival curves of mice in the different treatment groups. (F) Changes in body weight during treatment; gray boxes indicate the time points at which mice died. (G) Representative H&E, Ki67, CD31, and TUNEL staining of bladder tumor sections from each treatment group. (H–J) Quantification of Ki67⁺ proliferating cells (H), CD31⁺ microvessel density (I), and TUNEL⁺ apoptotic cells (J).

Plasma Gem concentrations were then measured after intravesical administration of free Gem and the two CpG-containing hydrogel formulations (Fig. S15). Free Gem entered the circulation rapidly, reaching a peak concentration of 68.75 μg/mL at 0.25 h before declining sharply. In comparison, both hydrogel formulations delayed the peak to 0.5 h and substantially lowered its magnitude. The peak concentrations were 15.81 μg/mL for Gel/CeNZ/Gem/CpG and 23.81 μg/mL for Gel/UCeNZ/Gem/CpG. After the peak, Gel/UCeNZ/Gem/CpG maintained moderately higher plasma Gem concentrations than the non-urease formulation, which may reflect the greater drug transport associated with urease modification. Nevertheless, its early plasma concentration remained markedly lower than that of free Gem, and all three groups approached baseline by 24 h. These profiles show that the hydrogel slowed the transfer of Gem from the bladder into the circulation and reduced the sharp early plasma peak. Urease modification modestly increased Gem transport relative to the non-urease hydrogel without reproducing the rapid systemic peak observed with free Gem.

Bioluminescence imaging showed rapid tumor progression in the control and Gel/CeNZ groups, indicating that Gel/CeNZ without Gem was insufficient to suppress tumor growth under the tested conditions (Fig. 6C). Free Gem and CeNZ/Gem produced only partial tumor suppression.

Compared with CeNZ/Gem, Gel/CeNZ/Gem produced a greater reduction in tumor bioluminescence. Together with the preceding intravesical retention results, this matched comparison supports the therapeutic benefit of incorporating CeNZ/Gem into the adhesive hydrogel depot, which prolonged local residence and increased therapeutic exposure in the bladder.

Gel/UCeNZ/Gem further reduced tumor bioluminescence compared with Gel/CeNZ/Gem and showed the strongest tumor suppression among the tested formulations (Fig. 6C and D). Because these two hydrogel formulations differed in urease functionalization, their comparison supports the additional therapeutic benefit associated with urease modification. This result is consistent with the motility and bladder-wall penetration experiments, which showed that urease functionalization enhanced nanoparticle transport in the urea-containing bladder environment. Photographs of the excised bladder tumors are shown in Fig. S16.

Survival analysis provided further evidence of the therapeutic effect of Gel/UCeNZ/Gem. Mice receiving Gel/UCeNZ/Gem survived longer than those in the control and free Gem groups (Fig. 6E), without substantial loss of body weight during treatment (Fig. 6F). H&E staining revealed no evident histopathological abnormalities in the heart, liver, spleen, lung, or kidney (Fig. S17), and the measured serum indicators of hepatic and renal function, including ALT, AST, BUN, and CREA, showed no marked abnormalities across the treatment groups (Fig. S18). Thus, under the tested dosing schedule, the improvement in tumor control was not accompanied by detectable changes in body weight, major-organ histology, or the measured hepatic and renal function indicators.

Histological examination of the bladder tumors showed dense and disorganized tumor tissue in the control and Gel/CeNZ groups, whereas Gem-containing treatments reduced tumor cellularity. The most extensive tissue disruption was observed after Gel/UCeNZ/Gem treatment (Fig. 6G). Ki67 and CD31 staining was also markedly reduced in this group, while TUNEL staining revealed a corresponding increase in apoptotic tumor cells. Quantitative analysis confirmed significantly lower Ki67 and CD31 signals and a higher proportion of TUNEL-positive cells than in the other groups (Fig. 6H–J). These tissue-level changes were consistent with the bioluminescence results and associated the stronger tumor suppression produced by Gel/UCeNZ/Gem with reduced proliferation and vascularization and increased apoptosis.

The matched in vivo comparisons demonstrated the therapeutic benefit of hydrogel incorporation and the additional effect of urease functionalization [5,17]. The catalytic assays and cellular ROS/NAC rescue experiments provided complementary evidence for the redox-regulating activity of CeNZ, while the motility and bladder-wall penetration assays supported the transport advantage conferred by urease modification [35]. Considered together, these findings support the design of Gel/UCeNZ/Gem, in which Gem chemotherapy is combined with hydrogel-mediated retention, CeNZ-associated redox regulation, and urease-enhanced delivery to improve local tumor control.

3.7. Material-enabled immunogenic chemotherapy promotes dendritic cell maturation and T-cell immune remodeling in vivo

Gel/UCeNZ/Gem markedly suppressed orthotopic bladder tumor progression, reduced tumor cell proliferation and angiogenesis, and increased apoptosis. We next investigated whether this local antitumor response was accompanied by immune activation in vivo. Based on the ICD-associated changes and BMDC maturation observed in vitro, dendritic cell maturation and T cell responses were examined in tumor-draining lymph nodes and tumor tissues.

Dendritic cell maturation in tumor-draining lymph nodes was first evaluated by flow cytometry. Gel/UCeNZ/Gem increased the proportion of CD11c⁺CD80⁺CD86⁺ mature DCs to 27.65%, compared with 11.89% in the control group and 10.68% in the free Gem group (Fig. 7A and B). These results show that the complete Gel/UCeNZ/Gem formulation induced a stronger DC response than free Gem. Together with the HMGB1 and CRT changes and the BMDC maturation results observed in vitro, the increased DC maturation was consistent with enhanced immunogenic tumor damage after Gel/UCeNZ/Gem treatment.

Fig. 7.

Immune-cell profiles in tumor-draining lymph nodes and bladder tumors after treatment. (A) Representative flow-cytometry plots of CD11c⁺CD80⁺CD86⁺ mature dendritic cells in tumor-draining lymph nodes. (B) Quantification of mature dendritic cells. (C, D) Representative flow-cytometry plots and quantification of CD3⁺CD4⁺ T cells. (E, F) Representative flow-cytometry plots and quantification of CD3⁺CD8⁺ cytotoxic T lymphocytes (CTLs). (G) Representative gating plots of CD3⁺CD4⁺Foxp3⁺ regulatory T cells (Tregs) and CD3⁺CD4⁺Foxp3^- effector T cells (Teffs). (H, I) Quantification of Tregs (H) and Teffs (I). (J, K) Ratios of CD8⁺ CTLs to Tregs (J) and CD4⁺ Teffs to Tregs (K). (L) Representative immunofluorescence images of bladder tumor sections stained for CD3, CD4, and CD8. Scale bar: 100 μm. (M − O) Quantification of intratumoral CD3⁺ (M), CD8⁺ (N), and CD4⁺ (O) T-cell infiltration.

We next examined whether the increased DC maturation was accompanied by changes in adaptive T cell populations. Flow cytometric analysis showed increased proportions of both CD3⁺CD4⁺ and CD3⁺CD8⁺ T cells after Gel/UCeNZ/Gem treatment (Fig. 7C–F). In particular, the proportion of CD8⁺ T cells increased from 12.45% in the control group to 29.08% in the Gel/UCeNZ/Gem group (Fig. 7F). Thus, the increase in mature DCs was accompanied by higher proportions of both CD4⁺ and CD8⁺ T cells.

In parallel, the immunosuppressive T cell compartment was analyzed. The proportion of CD4⁺Foxp3⁺ regulatory T cells gradually decreased across treatment groups, with the lowest level observed in the Gel/UCeNZ/Gem group (Fig. 7G and H). Meanwhile, CD4⁺Foxp3^- effector T cells were significantly increased (Fig. 7I). Consequently, both the CD8⁺ CTL/Treg ratio and the CD4⁺ Teff/Treg ratio were markedly elevated after Gel/UCeNZ/Gem treatment (Fig. 7J and K). These ratio-based indicators are particularly important because antitumor immunity depends not only on increasing effector T cell populations, but also on relieving Treg-mediated immune suppression. Therefore, the increased CTL/Treg and Teff/Treg ratios suggest that the material-enabled local therapy shifted the tumor-associated immune balance from an immunosuppressive state toward a more tumoricidal immune phenotype.

To determine whether the immune changes observed in tumor-draining lymph nodes were accompanied by increased T cell accumulation within the tumor, immunofluorescence staining was performed. Consistent with the flow cytometry data, Gel/UCeNZ/Gem markedly enhanced intratumoral infiltration of CD3⁺, CD4⁺, and CD8⁺ T cells compared with the other treatment groups (Fig. 7L–O). The increased intratumoral T cell infiltration, together with the lymph-node flow cytometry results, was consistent with the development of a local adaptive immune response after Gel/UCeNZ/Gem treatment.

Overall, Gel/UCeNZ/Gem increased DC maturation, increased the proportions of CD4⁺ and CD8⁺ T cells, enhanced intratumoral T cell infiltration, and shifted the effector-to-Treg balance toward an effector-favored profile. These findings show that the local tumor damage induced by Gel/UCeNZ/Gem was accompanied by antigen-presenting cell activation and adaptive immune remodeling. The intravesical retention experiments, together with the comparison between CeNZ/Gem and Gel/CeNZ/Gem, support the contribution of the hydrogel depot. The additional effect of urease functionalization was supported by the matched comparison between Gel/CeNZ/Gem and Gel/UCeNZ/Gem, as well as the motility and bladder-wall penetration assays. The catalytic assays and cellular ROS/NAC rescue experiments provided complementary evidence for the redox-regulating activity of CeNZ. However, because a Gel/Gem group was not included, the additional in vivo contribution of CeNZ relative to hydrogel-retained Gem could not be directly quantified. Including this comparison in future studies would provide a more complete component-level analysis of the antitumor and immune responses.

3.8. CpG incorporation further amplifies Gel/UCeNZ/Gem-mediated immune remodeling and tumor regression in an orthotopic bladder cancer model

Animal study results have demonstrated that Gel/UCeNZ/Gem inhibits the growth of orthotopic bladder tumors and induces immune remodeling through material-mediated immunogenic chemotherapy. However, UCeNZ/Gem-induced tumor damage alone may not be sufficient to fully overcome the immunosuppressive bladder tumor microenvironment. To further enhance antitumor immune activation, CpG ODN, a TLR9 agonist, was introduced to promote dendritic cell activation and T cell-mediated immunity. Importantly, two CpG administration strategies were compared: separately administered CpG combined with Gel/UCeNZ/Gem (Gel/UCeNZ/Gem + CpG) and CpG incorporated within the hydrogel–nanozyme formulation (Gel/UCeNZ/Gem/CpG). This comparison was designed to evaluate whether incorporating CpG into the same intravesical depot could provide additional therapeutic benefit over simple separate administration by improving local CpG availability during UCeNZ/Gem-induced tumor damage.

Bioluminescence imaging showed a progressive increase in the mean tumor signal in the Control and free CpG groups, indicating that CpG alone was insufficient to suppress orthotopic bladder tumor growth under the tested conditions (Fig. 8A and B). Relatively large within-group variability was observed in these two groups from day 11 onward, which may reflect inter-animal differences in tumor establishment and subsequent growth kinetics. In contrast, Gel/UCeNZ/Gem markedly reduced tumor bioluminescence, consistent with the antitumor effects observed in the preceding experiments. Compared with Gel/UCeNZ/Gem, the two CpG-containing treatment groups exhibited further reductions in tumor bioluminescence. Among them, Gel/UCeNZ/Gem/CpG produced the lowest and most sustained tumor signal, indicating greater tumor suppression than that achieved by the separate administration of Gel/UCeNZ/Gem and CpG. This difference may be related to the improved local retention and sustained availability of CpG within the hydrogel–nanozyme depot, thereby facilitating the temporal coordination of CpG-mediated immune stimulation with UCeNZ/Gem-induced immunogenic tumor damage.

Fig. 8.

CpG incorporation enhances Gel/UCeNZ/Gem-mediated immune remodeling and tumor regression in an orthotopic bladder cancer model. (A) Representative bioluminescence images of tumor-bearing mice treated with the indicated formulations; gray boxes denote mice that died during the study. (B) Quantification of tumor bioluminescence intensity at the indicated time points. Identical criteria for region-of-interest delineation and background subtraction were applied to all animals, groups, and time points. Data are presented as the mean ± SD (n = 5). (C, D) Representative flow-cytometry plots and quantification of CD3⁺ T cells in tumor-draining lymph nodes. (E) Representative flow-cytometry plots of CD3⁺CD8⁺ cytotoxic T lymphocytes and CD3⁺CD4⁺ helper T cells in tumor-draining lymph nodes. (F, G) Quantification of CD3⁺CD8⁺ (F) and CD3⁺CD4⁺ (G) T-cell populations. (H, I) Representative gating plots of CD44⁺CD62L⁻ effector/memory-like CD4⁺ (H) and CD8⁺ (I) T cells in the spleen. (J, K) Corresponding quantification of CD4⁺ (J) and CD8⁺ (K) effector/memory-like T-cell populations. (L) Representative immunofluorescence images of bladder tumor sections stained for CD3, CD8, CD86, and CD206. Scale bar: 100 μm. (M − P) Quantification of intratumoral CD3 (M), CD8 (N), CD86 (O), and CD206 (P) fluorescence intensity. (Q–S) Serum levels of IFN-γ (Q), TNF-α (R), and IL-12p70 (S) in the different treatment groups.

Survival analysis further supported the therapeutic benefit of CpG incorporation. During the 32-day observation period, survival in the control group decreased to 40%, whereas all mice in the CpG, Gel/UCeNZ/Gem, Gel/UCeNZ/Gem + CpG, and Gel/UCeNZ/Gem/CpG groups remained alive until the scheduled endpoint (Fig. S19). Body weight remained stable or gradually increased in the Gel/UCeNZ/Gem/CpG group, indicating that the enhanced antitumor efficacy was not accompanied by obvious systemic toxicity (Fig. S20). Gross tumor images further supported the superior tumor regression achieved by Gel/UCeNZ/Gem/CpG (Fig. S21). These results indicate that CpG incorporation enhanced the therapeutic performance of Gel/UCeNZ/Gem while maintaining good treatment tolerability.

Total cerium levels were measured in the bladder and major organs to examine the tissue distribution of CeNZ after intravesical administration (Fig. S22). Gel/UCeNZ/Gem/CpG treatment markedly increased the cerium content in bladder tissue, reaching 15.9 ng/g tissue compared with 2.8 ng/g in the PBS group. In contrast, cerium levels in the heart, liver, spleen, lung, and kidney did not differ significantly between the two groups. The selective increase in bladder tissue is consistent with local retention of the CeNZ-containing formulation after intravesical administration, while the absence of a detectable increase in the major organs suggests limited systemic distribution at the experimental endpoint. Together with the serum biochemical and histological results, this distribution profile supports the local-delivery characteristics and systemic tolerability of the formulation under the tested dosing schedule. Because cerium was measured at a single endpoint, these data do not describe the complete clearance kinetics or exclude longer-term accumulation.

We next investigated whether the improved tumor suppression was associated with enhanced immune activation. Flow cytometric analysis of lymph node cells showed that Gel/UCeNZ/Gem/CpG significantly increased the proportion of CD3⁺ T cells, including both CD8⁺ cytotoxic T lymphocytes and CD4⁺ helper T cells, compared with the other groups (Fig. 8C–G). Although Gel/UCeNZ/Gem + CpG also promoted T cell activation relative to Gel/UCeNZ/Gem alone, the CpG-incorporated Gel/UCeNZ/Gem/CpG formulation showed a stronger immune-activating effect. This result suggests that depot-based CpG delivery may more effectively enhance antigen-presenting cell activation and downstream T cell priming during Gel/UCeNZ/Gem-induced tumor destruction.

The effect of CpG incorporation on broader immune activation was further evaluated by analyzing memory/effector-like T cell populations in the spleen. Both CD4⁺CD44⁺CD62L⁻ and CD8⁺CD44⁺CD62L⁻ T cell subsets were increased after CpG-containing treatment, with the highest levels observed in the Gel/UCeNZ/Gem/CpG group (Fig. 8H–K). These results indicate that incorporating CpG into the Gel/UCeNZ/Gem formulation promoted a stronger systemic memory/effector-like T cell phenotype than separate CpG administration. However, because a tumor re-challenge experiment was not performed in this section, these CD44⁺CD62L⁻ populations should be interpreted as memory/effector-like immune activation rather than definitive evidence of long-term protective immune memory.

Immunofluorescence staining of tumor tissues further confirmed the superior immune remodeling induced by Gel/UCeNZ/Gem/CpG. Compared with separate CpG administration, the CpG-incorporated formulation induced stronger intratumoral infiltration of CD3⁺ and CD8⁺ T cells and increased CD86 expression, indicating enhanced local T cell infiltration and antigen-presenting cell activation (Fig. 8L–O). In parallel, CD206⁺ M2-like tumor-associated macrophages were markedly reduced in the Gel/UCeNZ/Gem/CpG group (Fig. 8P), suggesting that CpG incorporation helped shift the tumor immune microenvironment away from an immunosuppressive macrophage phenotype. These findings demonstrate that CpG incorporation not only enhanced immune activation in the tumor-draining lymph nodes but also promoted effector immune-cell infiltration and immune remodeling within the tumor. Nevertheless, the spatial organization and maturation of tumor-associated tertiary lymphoid structures were not evaluated in the present study and warrant further investigation.

Consistent with the immune cell profiling results, serum cytokine analysis showed that Gel/UCeNZ/Gem/CpG significantly increased IFN-γ, TNF-α, and IL-12p70 levels compared with the other groups (Fig. 8Q–S). IFN-γ and TNF-α are important mediators of cytotoxic antitumor immunity, while IL-12p70 promotes Th1-skewed immune responses and supports cytotoxic T cell activation. The stronger cytokine response in the Gel/UCeNZ/Gem/CpG group further suggests that CpG incorporation strengthened systemic pro-inflammatory antitumor immunity. Mechanistically, this result is consistent with the proposed therapeutic cascade: UCeNZ/Gem induces ROS-amplified tumor stress and ICD-associated antigen/DAMP signaling, while CpG activates TLR9 signaling in antigen-presenting cells, thereby promoting DC maturation, cytokine secretion, and T cell-mediated tumor killing [1,15,43,44,46].

Histological and immunofluorescence analyses further confirmed that CpG incorporation enhanced tumor regression at the tissue level. H&E staining showed extensive tumor tissue disruption in the Gel/UCeNZ/Gem/CpG group (Fig. S23A). Ki67 staining revealed the lowest proliferative activity after CpG-incorporated treatment, while CD31 staining showed markedly reduced tumor-associated angiogenesis (Fig. S23B and C). TUNEL staining further demonstrated the highest level of apoptotic tumor cell death in the Gel/UCeNZ/Gem/CpG group (Fig. S23). These pathological findings are consistent with the bioluminescence, survival, and immune profiling results, indicating that CpG incorporation enhances both direct tumor suppression and immune-associated tumor destruction.

Taken together, these results demonstrate that CpG plays a critical role in amplifying Gel/UCeNZ/Gem-induced immunogenic chemotherapy, and that its formulation mode influences therapeutic outcome. Compared with separate CpG administration, incorporating CpG into the Gel/UCeNZ/Gem platform produced stronger tumor regression, enhanced T cell activation, increased memory/effector-like T cell populations, reduced M2-like macrophage infiltration, and promoted Th1-associated cytokine secretion. Therefore, Gel/UCeNZ/Gem/CpG functions as a depot-based intravesical chemo-immunotherapy system in which hydrogel-mediated retention, UCeNZ/Gem-induced ICD, and CpG-mediated TLR9 activation are integrated to enhance local antitumor immune remodeling.

3.9. Transcriptomic analysis reveals coordinated redox stress, immunogenic chemotherapy, and immune activation following Gel/UCeNZ/Gem/CpG treatment

To further connect the therapeutic outcomes of Gel/UCeNZ/Gem/CpG with its proposed mechanism of material-enabled immunogenic chemo-immunotherapy, transcriptomic profiling was performed to examine treatment-induced molecular reprogramming within the tumor microenvironment. This analysis was intended to verify whether the material-enabled therapeutic cascade—hydrogel-mediated local retention, UCeNZ-associated redox stress amplification, Gem-induced tumor cell damage, and CpG-mediated immune activation—could be reflected at the gene-expression level. Principal component analysis (PCA) showed clear separation between the control and Gel/UCeNZ/Gem/CpG-treated groups (Fig. 9A), indicating substantial transcriptomic remodeling after treatment. Consistently, differential expression analysis identified a broad set of upregulated and downregulated genes (Fig. 9B), suggesting that the treatment affected multiple biological programs rather than acting through a single cytotoxic pathway.

Fig. 9.

Transcriptomic analysis reveals immune activation and tumor suppression induced by Gel/UCeNZ/Gem/CpG treatment. (A) PCA showing distinct clustering between control and treated groups. (B) Volcano plot of DEGs (red, upregulated; blue, downregulated; gray, non-significant). (C) GO enrichment analysis highlighting immune activation and tumor-related biological processes. (D) KEGG pathway analysis indicating enrichment in immune- and tumor-associated signaling pathways, including PI3K-Akt, PD-1/PD-L1, NF-κB, Toll-like receptor, T cell receptor, and p53 pathways. (E–G) Heatmaps of representative DEGs associated with ROS-induced oxidative stress, ICD, and oxidative stress-mediated apoptosis. (H) GSEA showing enrichment of immune-related pathways, including immune system, cytokine signaling, innate immune system, and neutrophil degranulation. (I) PPI network of key DEGs. Node size indicates degree, and node color represents relative expression (red, upregulated; blue, downregulated), revealing coordinated regulation of immune activation and apoptosis.

Functional enrichment analysis further characterized the biological programs associated with Gel/UCeNZ/Gem/CpG treatment. GO enrichment analysis showed that the differentially expressed genes were predominantly associated with immune and tumor-regulatory processes, including responses to cytokines, inflammatory signaling, lymphocyte activation, regulation of apoptosis, and cell cycle control (Fig. 9C). These transcriptional changes were consistent with the experimentally observed enhancement of DC maturation and T cell infiltration, together with reduced tumor proliferation and increased apoptosis.

KEGG analysis further identified several pathways related to tumor cell fate and immune regulation, including PI3K–Akt signaling, PD-1/PD-L1 checkpoint signaling, antigen processing and presentation, NF-κB signaling, Toll-like receptor signaling, T cell receptor signaling, and p53 signaling (Fig. 9D). When interpreted together with the preceding experimental results, enrichment of p53-and apoptosis-related pathways was consistent with treatment-induced tumor cell stress and damage, whereas Toll-like receptor, NF-κB, antigen-presentation, and T cell receptor pathways supported the induction of coordinated innate and adaptive immune responses. These enrichment results indicate the biological processes associated with the combined treatment but do not, by themselves, establish the activation state or individual contribution of each pathway.

Given the enrichment of the PD-1/PD-L1 checkpoint pathway, we further examined PD-1 and PD-L1 expression at the protein level. Western blot analysis showed that both PD-1 and PD-L1 protein levels were significantly increased in tumor tissues following Gel/UCeNZ/Gem/CpG treatment compared with the control group (Fig. S24). These findings corroborate the checkpoint-related transcriptional signature at the protein level (Fig. 9E). The concurrent upregulation of PD-1 and PD-L1 may represent an adaptive immune-checkpoint response accompanying treatment-induced immune activation in the tumor microenvironment, suggesting a potential rationale for combining Gel/UCeNZ/Gem/CpG with PD-1/PD-L1 blockade. Because the Western blot analysis was performed using whole-tumor lysates, however, the cellular sources and spatial distribution of PD-1 and PD-L1 remain to be further determined.

To further connect these enriched pathways with the proposed therapeutic cascade, representative gene heatmaps were analyzed across three mechanism-related axes. Genes associated with ROS-induced oxidative stress and DNA damage were markedly regulated after Gel/UCeNZ/Gem/CpG treatment, supporting the role of UCeNZ-mediated redox amplification and Gem-induced genotoxic stress (Fig. 9E). Genes related to immunogenic cell death and antigen presentation were also strongly altered, consistent with the increased HMGB1 translocation, CRT exposure, DC maturation, and cytokine secretion observed in vitro and in vivo (Fig. 9F). In addition, genes involved in oxidative stress-mediated apoptosis and proliferation control exhibited coordinated changes, providing molecular support for the reduced Ki67 expression and increased TUNEL-positive tumor cell death observed in tumor tissues (Fig. 9G). Thus, the heatmap results link the material–drug functions of the system to the biological outcomes of redox stress, ICD-associated immunogenicity, and tumor growth inhibition.

Gene set enrichment analysis (GSEA) further confirmed that Gel/UCeNZ/Gem/CpG treatment activated immune-related transcriptional programs. Immune system, cytokine signaling, innate immune response, and neutrophil degranulation pathways were significantly enriched after treatment (Fig. 9H). These enriched gene sets are consistent with the elevated IFN-γ, TNF-α, and IL-12p70 levels, increased CD4⁺ and CD8⁺ T cell responses, and decreased immunosuppressive cell populations observed in previous sections. These findings suggest that CpG incorporation did not merely enhance local tumor killing, but promoted broader immune activation in the tumor-bearing host. Therefore, transcriptomic evidence supports the interpretation that the Gel/UCeNZ/Gem/CpG platform converts localized intravesical tumor damage into an immune-activating tumor microenvironment.

Finally, protein–protein interaction network analysis revealed a highly interconnected regulatory network among key differentially expressed genes (Fig. 9I). Hub genes such as Ifng, Tnf, Cd4, Cxcl10, and Tlr9 formed a core immune activation module, consistent with CpG/TLR9-mediated innate immune stimulation, inflammatory cytokine production, chemokine-mediated immune recruitment, and T cell activation. Meanwhile, tumor cell fate-related genes, including Trp53, Casp3, Cdkn1a, and Mki67, were integrated into the same network, linking immune activation with apoptosis, cell cycle regulation, and proliferation suppression. The coexistence of immune-regulatory and tumor-suppressive nodes within the same interaction network suggests that Gel/UCeNZ/Gem/CpG does not act as a simple local chemotherapy depot; rather, it induces a coordinated biological response in which tumor cell stress, immunogenic death, antigen presentation, and effector immune activation are functionally connected.

Rather than reflecting a single cytotoxic effect, the molecular profile of Gel/UCeNZ/Gem/CpG-treated tumors revealed a coupled response involving oxidative stress and DNA damage, apoptosis and cell-cycle suppression, antigen processing and presentation, cytokine signaling, and T cell activation. The accompanying increase in PD-1 and PD-L1 protein expression suggests that treatment-induced immune activation was accompanied by an adaptive checkpoint response. In conjunction with the material characterization and in vivo findings, these data link sustained local delivery and redox-amplified immunogenic tumor damage with CpG-associated remodeling of antitumor immunity, while also identifying the PD-1/PD-L1 axis as a potential constraint on the therapeutic response.

4. Conclusion

In summary, we developed a bladder microenvironment-adaptive bioactive hydrogel–nanozyme system, Gel/UCeNZ/Gem/CpG, for intravesical bladder cancer chemo-immunotherapy. This system integrates four functionally complementary modules within one local therapeutic platform: a thermosensitive adhesive hydrogel for mucosal retention, urease-modified CeO₂ nanozymes for urea-responsive transport, redox-active CeO₂ for amplification of tumor oxidative stress, and Gem/CpG co-delivery for immunogenic chemotherapy and immune activation. After intravesical instillation, the hydrogel adhered to the bladder mucosa and remained temporarily, prolonging local drug exposure without leaving excessive residual gel. In the urea-rich bladder environment, urease modification increased nanoparticle motility and allowed UCeNZ to penetrate more deeply into the bladder wall than non-urease-modified CeNZ. At the cellular level, CeO₂-mediated redox regulation amplified Gem-induced oxidative stress, mitochondrial dysfunction, apoptosis, and ICD-associated DAMP signaling, while CpG-mediated TLR9 stimulation promoted dendritic cell maturation and T cell activation. In an orthotopic bladder cancer model, Gel/UCeNZ/Gem/CpG suppressed tumor progression, prolonged survival, enhanced CD8⁺ T cell infiltration, reduced Treg and M2-like macrophage populations, and promoted a more immune-active tumor microenvironment with acceptable biosafety. Transcriptomic analysis further confirmed that the therapeutic benefit was associated with coordinated regulation of oxidative stress, DNA damage, apoptosis, antigen presentation, cytokine signaling, T cell activation, immune checkpoint-related pathways, and tumor-suppressive signaling. These findings indicate that Gel/UCeNZ/Gem/CpG acts not merely as a local drug depot, but as a functionally integrated intravesical chemo-immunotherapy system that couples bladder-specific delivery adaptation with redox-amplified immunogenic chemotherapy and CpG-driven immune remodeling. This strategy provides a promising approach for improving localized bladder cancer treatment.

CRediT authorship contribution statement

Yulin Yuan: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Linkang Yu: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Huayi Feng: Validation, Methodology, Investigation, Formal analysis. Chuanjie Li: Investigation, Data curation. Xiaogang Chen: Methodology, Investigation, Funding acquisition. Longbao Feng: Resources, Investigation. Teijun Pan: Resources. Lei Gao: Writing – review & editing, Supervision, Conceptualization. Yu Zhou: Writing – review & editing, Supervision, Project administration, Conceptualization. Ling Zhang: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.

Data availability statement

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

Ethics approval and consent to participate

All animal experiments in this study were reviewed and approved by the Medical Ethics Committee of Wuhan University of Science and Technology (Ethics approval reference: (2024) No. 207, approval date: 12 October 2024). No human‑participant experiments were involved in this work, therefore, informed consent is not applicable.

Declaration of competing interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article: