Tetrandrine-based in situ vaccination via multifunctional Mn²⁺-hyaluronic acid hydrogel loading gelatinase-responsive nanoparticles Authors: Wanru Wang, Shujun Song, Dongliang Han, Yuling Qiu, Dan Lei, Xuxuan Gu, Haoyue Qi, Hanxi Zhang, Lanqi Cen, Jie Shao, Xiaoping Qian, Wenxiu Chen, Lixia Yu, Baorui Liu, Li Xie, Rutian Li Published: 2026 ABSTRACT AbstractIn situ tumor vaccination (ISV) has emerged as a promising strategy for converting tumors into endogenous vaccine depots, but its application in breast cancer remains limited by poor intratumoral drug retention and insufficient microenvironment remodeling. Here, we developed an injectable Mn²⁺-coordinated hyaluronic acid hydrogel embedding gelatinase-responsive tetrandrine nanoparticles (TETNPs@Gel) as an MRI-trackable local ISV platform. TETNPs@Gel exhibited biphasic, pH-responsive drug release, accelerating tetrandrine liberation under mildly acidic conditions while limiting premature leakage at physiological pH. Mn incorporation provided positive T1-weighted contrast and enabled longitudinal monitoring of depot retention in vivo. In orthotopic 4T1 tumors, TETNPs@Gel significantly inhibited tumor growth, prolonged survival, and caused no obvious systemic toxicity. In a bilateral 4T1 model, local treatment also suppressed the growth of untreated distant tumors, reduced pulmonary metastatic burden, and improved survival, indicating a systemic antitumor effect. Mechanistically, TETNPs@Gel enhanced dendritic-cell maturation and intratumoral CD8⁺ T-cell infiltration, increased C3a and C5a levels, and induced vascular/perivascular remodeling. Bulk and single-cell transcriptomic analyses further revealed complement-associated activation, extracellular-matrix remodeling, and enrichment of antigen-presenting and inflammatory macrophage programs. Thus, TETNPs@Gel functions as an in situ nanovaccine rather than a conventional cytotoxic formulation, in which controlled tetrandrine release couples direct tumor-cell killing with vascular and immune re-education, offering a potentially generalizable platform for localized immunochemotherapy in solid tumors.Graphical abstract Supplementary InformationThe online version contains supplementary material available at 10.1186/s12951-026-04545-5.Keywords: Tetrandrine nanoparticles, Immunotherapy, Injectable hydrogel, Sustained release, Dual function Introduction Breast cancer remains one of the most prevalent malignancies among women worldwide and continues to account for a substantial proportion of cancer-related mortality [1–3]. Although advances in surgery, radiotherapy, and systemic therapies have improved outcomes, the benefit of immunotherapy in breast cancer—particularly in aggressive subtypes—has been modest [4]. A central reason is the profoundly immunosuppressive tumor microenvironment (TME), which restricts antigen-presenting cell activation and T-cell infiltration, thereby limiting durable responses [5]. In situ tumor vaccination (ISV) provides a conceptual route to overcome this barrier by converting the tumor itself into a source of antigens and adjuvant signals through localized intervention [6]. By coupling local drug exposure with the release of tumor antigens and danger-associated molecular patterns, ISV can remodel the TME and prime systemic immunity [7]. However, most traditional ISV practices and proof-of-concept demonstrations have been concentrated in other tumor types, and translation to breast cancer has been hindered by challenges in sustaining local exposure, reversing stromal/vascular abnormalities, and monitoring treatment noninvasively. Natural small-molecule compounds from traditional medicine are attractive for ISV because of their pleiotropic pharmacology and relatively low systemic toxicity [8]. Among them, tetrandrine (TET), a bis-benzylisoquinoline alkaloid, is notable for its dual functionality: it exerts direct antitumor effects (antiproliferation and apoptosis induction) while also enhancing antitumor immunity by promoting dendritic-cell activation and T-cell–mediated responses [9–12]. Thus, TET is a rational single-agent “drug-as-adjuvant” candidate for ISV, provided that delivery systems can create the concentration–time window needed to couple cytotoxicity with immune activation in situ. Nanotechnology-enabled carriers and injectable hydrogels offer complementary strengths to address these needs [13–14]. Nanoparticles can improve solubility and tumor localization of small molecules, while hydrogels furnish a biocompatible depot for sustained intratumoral release [15]. Embedding nanoparticles within hydrogels further enables spatiotemporally controlled delivery and reduces systemic exposure [16–17]. In addition, incorporating bioactive metal ions can add real-time imaging and immune-stimulatory functions. In particular, Mn²⁺ can endow the depot with magnetic resonance imaging (MRI) visibility and has been reported to participate in innate immune activation, thereby aligning image guidance with pharmacodynamic effects [18–19]. Our group previously explored the concept of ISV using a nanocomposite hydrogel that sustained release of gelatinase-responsive nanoparticles to reshape the immunosuppressive TME and enhance dendritic-cell activation and CD8⁺ T-cell infiltration, thereby functioning as an ISV rather than conventional cytotoxic drugs [20]. Building directly on that foundation, the present work introduces a Mn²⁺–hyaluronic-acid (HA-DOPA/Mn) hydrogel that embeds gelatinase-responsive tetrandrine nanoparticles (TETNPs@Gel) to establish a concentration-programmed, biphasic intratumoral drug delivery system: an early high-dose pulse provoking immunogenic tumor cell death with antigen/DAMPs release, followed by a prolonged low-dose drug release phase that suppresses angiogenesis and normalizes stroma to permit immune trafficking—thereby operationalizing ISV with a single medicine drug. Beyond delivery, the platform is MRI-visible through paramagnetic Mn²⁺, enabling noninvasive, longitudinal tracking of the depot and aligning imaging readouts with pharmacodynamics. At the systems level, RNA-seq demonstrates enrichment of complement–coagulation (lectin pathway) and ECM-organization programs, providing a mechanistic bridge from concentration-controlled exposure to dendritic-cell maturation and heightened CD8⁺ T-cell infiltration—hallmark features of effective ISV (Fig. 1). Collectively, these advances reposition tetrandrine from a cytotoxic agent to the active principle of an image-guided in situ nanovaccine, achieving localized immunochemotherapy while minimizing systemic liabilities. Schematic representation of the TETNPs@Gel in-situ vaccination platform , achieving sustained and potent anti-cancer immune responses through the controlled-release mechanism of TETNPs and HA-DOPA/Mn hydrogels. In-situ gel vaccine encapsulating tetrandrine nanoparticles forms a local reservoir for controlled sustained release of immunotherapeutic agents. This vaccine effectively triggers tumor vascular normalization, activates dendritic cells and effector CD8+ T cells within tumor tissue, thereby inhibiting progression and preventing recurrence in immunologically cold tumors Materials and methods Cell culture medium and supplements were purchased from Gibco (NY, USA). IR-1061, Bovine serum albumin (BSA), and Hydroxypropyl cellulose (HPC) were purchased from Macklin (Shanghai, China). Tetrandrine (TET) reagents were purchased from MCE (Shanghai, China). Mouse antibodies were purchased from Biolegend (CA, USA). 4T1 breast cancer cells were obtained from the Cell Bank of Shanghai Institute of Biochemistry and Cell Biology and cultured with RPMI 1640 medium containing 10% FBS and 1% PBS at 37℃ and 5% CO2. BALB/c female mice aged 5–6 weeks were purchased from Shanghai Sippr-BK Laboratory Animal Co. Ltd. (Shanghai, China). All animal experimental protocols were approved by the Laboratory Animal Care and Use Committee of the Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School. Preparation and characterization of TETNPs and TETNPs@Gel The nanoparticle carrier mPEG-Pep-PCL was synthesized following a previously reported protocol [21]. Tetrandrine (TET) nanoparticles (TETNPs) were prepared using the emulsification–solvent evaporation method. Briefly, TET and the mPEG-Pep-PCL carrier (5 mg) were ultrasonically dissolved in 0.5 mL of dichloromethane. Subsequently, 1.5 mL of 5% PVA solution was added for ultrasonic emulsification for 2 min, followed by the addition of 4 mL of 1% PVA solution for further dispersion. The mixture was stirred at room temperature for 3 h under a fume hood to remove residual solvents. The solution was then filtered through a 0.45 μm membrane and centrifuged at 15,000 rpm at 4 ℃ for 30 min to collect the resulting TETNPs precipitate. The particle size, zeta potential, and polydispersity index (PDI) of TETNPs were measured using a laser diffraction particle size analyzer (Mastersizer 3000, Malvern Instruments, UK). Drug encapsulation efficiency was determined by ultraviolet absorbance at 360 nm using a multifunctional microplate reader. According to our previous synthesis scheme [22], the HA-DOPA/Mn hydrogel was fabricated by mixing HA-DOPA with varying concentrations of manganese ions (14, 16, 18, 20, 30 mM). After incubation for one week at room temperature, gel formation was confirmed using the vial inversion method, with 18 mM Mn²⁺ identified as optimal. For subsequent experiments, HA-DOPA (50 mg/mL), Mn²⁺ (18 mM), and TETNPs were co-blended in distilled water to prepare TETNPs@Gel. The cross-sectional morphology of freeze-dried hydrogels was examined by field-emission scanning electron microscopy (FE-SEM, SUPRA 55VP, Carl Zeiss, Germany). Injectability was evaluated using a 22-gauge syringe to draw and expel the hydrogel. Self-healing was assessed by qualitative tensile testing of rejoined gel pieces. Additionally, hydrogels incorporating CY5 and CY7 dyes were prepared, and their self-repairing capabilities were evaluated after physical attachment. In vitro evaluations In vitro release of TET from TETNPs and TETNPs@Gel The release behavior of TET from TETNPs and TETNPs@Gel was evaluated under sink conditions by a dialysis-based method. Formulations containing an equivalent amount of TET were transferred into dialysis bags and immersed in phosphate-buffered saline (PBS) adjusted to pH 7.4 or pH 6.8 to mimic physiological and mildly acidic tumor-relevant environments, respectively. The release systems were maintained at 37 °C with gentle shaking throughout the experiment. At predetermined time points (0.5, 1, 2, 4, 6, 8, 24, 48, 72, and 96 h), aliquots of the external release medium were collected and immediately replaced with an equal volume of fresh prewarmed buffer to maintain sink conditions. The concentration of released TET was quantified by UV absorbance at 280 nm using the standard curve established for TET quantification. The cumulative release percentage was calculated after correction for the sampled volume and expressed as the percentage of the total loaded drug. Unless otherwise indicated, each condition was tested in triplicate. Cell apoptosis Apoptosis of 4T1 cells was evaluated using the Annexin V-FITC/PI assay. Cells were treated with TETNPs at concentrations of 25, 50, and 100 µg/mL in triplicate. After harvesting 1 × 106 cells, they were washed with precooled saline and resuspended in 100 µL of binding buffer. Annexin V-FITC (5 µL) and PI (10 µL) were added, gently mixed, and incubated for 15 min at room temperature in the dark. The mixture was then diluted with 400 µL of binding buffer and maintained on ice. Samples were acquired within 1 h using a flow cytometer. Fluorescence staining of tumor cells was also examined under a fluorescence microscope. Cell viability assay Cells were seeded at a density of 1 × 10⁴ cells per well in 96-well plates and incubated at 37 ℃ with 5% CO₂ in a humidified atmosphere for 24 h. Cells were treated with hydrogel components at 1–100 µg/mL (1, 2, 5, 10, 25, 50, 100 µg/mL) for 24 h. Subsequently, CCK-8 solution was added to the medium at a final concentration of 10%, and the cells were incubated for an additional 1 h. Absorbance was measured at 450 nm using a microplate reader (Tecan Infinite F200). Cell viability was evaluated for 4T1 cells treated with NPs, TET, TETNPs, and TETNPs@Gel. Cellular uptake assay Cellular uptake was assessed using coumarin-6-oaded nanoparticles as a fluorescent probe. Coumarin-6-loaded nanoparticles were prepared by the emulsification–solvent evaporation method and co-cultured with 4T1 cells. After 24 h of incubation, cells were rinsed twice with saline to remove excess nanoparticles and imaged using a confocal laser scanning microscope (LSM780, Zeiss). Bone marrow–derived dendritic cells (BMDCs) treated with either free coumarin-6 or coumarin-6 nanoparticles were fixed in acetone at − 20 ℃ for 10 min and washed with PBS. Nuclei were stained with DAPI (Vector Laboratories, USA) and imaged by confocal microscopy. Images were quantified from random fields, and expression levels were calculated by normalizing the stained area of interest to the DAPI-stained nuclei area, expressed as the ratio of green to blue (FITC/DAPI) relative fluorescence units. BMDCs maturation To assess in-vitro DC maturation, BMDCs exposed to free coumarin-6 or coumarin-6 nanoparticles (25, 50, 100 µg/mL) were stained with CD45-FITC, CD11c-APC, CD86-PE-Cy7, and MHC-II APC-eFluor 780. The activation of DCs was determined by analyzing the co-expression of CD86 and MHC II, markers associated with DC activation, within the CD45+/CD11c+ DC population. In vivo evaluations Biodegradation experiment in vivo To monitor the in vivo retention and imaging behavior of the Mn²⁺-containing hydrogel, small-animal magnetic resonance imaging (MRI) was performed using T1-weighted imaging (T1WI) as the primary readout. Because Mn²⁺ is a paramagnetic ion that mainly shortens longitudinal relaxation, T1-weighted sequences were used to visualize the hydrogel depot and to track its persistence over time after local administration. T2-weighted imaging (T2WI) was additionally acquired as an exploratory supplementary readout to assess whether local Mn²⁺ confinement within the hydrogel network influenced transverse-relaxation behavior; these data were used as supportive information rather than the principal imaging endpoint. For subcutaneous imaging, HA-DOPA and HA-DOPA/Mn hydrogels were injected into the abdominal subcutaneous tissue of mice. At the indicated time points, mice were anesthetized and scanned on a 7.0 T small-animal MRI system. T1WI were collected serially to monitor signal changes at the injection site over time. Regions of interest encompassing the hydrogel depot were delineated on each image, and temporal changes in signal intensity were used to evaluate in vivo retention and degradation behavior. To assess the imaging performance of the hydrogel in a tumor setting, 4T1 tumor-bearing mice were established by subcutaneous inoculation of 4T1 cells. When tumor volumes reached 100 ~ 150 mm³, hydrogel formulations were administered locally at the tumor site, and longitudinal MRI was performed using the same imaging workflow. T1-weighted sequences were used to determine whether the Mn²⁺ hosting hydrogel could provide positive contrast enhancement and permit noninvasive tracking of depot persistence within the tumor microenvironment. Exploratory T2WI were also acquired under the same positioning conditions. To further evaluate in vivo biodegradation, the IVIS spectral imaging system (CRi Maestro, USA) was used for near-infrared fluorescence (NIRF) imaging. When preparing HA-DOPA/Mn/TETNPs-CY5, first connect CY5 (0.5 mg/mL) to the nanoparticles, and then fuse them with the hydrogel; for preparing HA-DOPA/CY5 and HA-DOPA/Mn/CY5, integrate the CY5 into the hydrogel (0.5 mg/mL). Mice received subcutaneous injections (100 µL) of CY5 solution, HA-DOPA/CY5, HA-DOPA/Mn/CY5, or HA-DOPA/Mn/TETNPs-CY5. Animals were anesthetized with isoflurane, and NIRF images were acquired at 0, 2, 4, 6, 8, 24, 48, 72, 96, 120, and 144 h. On day 7, mice were euthanized; tumors and major organs (heart, liver, spleen, lung, kidney) were harvested for ex-vivo NIRF quantification. In vivo antitumor activity To assess antitumor efficacy, 1 × 10⁶ 4T1 cells suspended in 100 µL PBS were injected subcutaneously into the right inguinal region of 6-week-old female BALB/c mice. When tumors reached 75 ~ 100 mm³ (day 7), mice were randomized into six groups (n = 6): NS (control), NPs, TET (1 mg/kg), TETNPs, HA-DOPA/Mn, and TETNPs@Gel. Treatments (100 µL) were administered peritumorally on days 7 and 14. Tumor volume and body weight were recorded every 2 days until day 21. Tumor volume (V) was calculated using the formula: V = 0.5 × L × W². Mice were sacrificed on day 21, and tumors were excised and weighed. Bilateral 4T1 distant-tumor model, survival, and pulmonary metastasis evaluation To determine whether local treatment could elicit a systemic antitumor effect, a bilateral 4T1 tumor model was established. Primary tumors were first generated by subcutaneous inoculation of 1 × 10⁶ 4T1 cells into one inguinal region. When tumors reached 75 ~ 100 mm³ (day 0), mice were randomized into six groups (n = 6): NS (control), TET (1 mg/kg), TETNPs, HA-DOPA/Mn, TET@Gel and TETNPs@Gel. Treatments (100 µL) were administered peritumorally on days 0 and 4. On day 8, a second inoculation of 1 × 10⁶ 4T1 cells in 100 µL PBS was performed subcutaneously at the contralateral inguinal site to establish an untreated distant tumor. The growth of the distant tumors was monitored every 2 days using caliper measurements, and tumor volume was calculated as described above. Body weight was recorded throughout the experiment. On day 21, mice were euthanized, and both primary and distant tumors were harvested and weighed. Lungs were collected for gross observation of metastatic nodules and for histopathological examination by hematoxylin and eosin (H&E) staining. Major organs, including the liver, spleen, kidney, and heart, were harvested in parallel for safety-related histological assessment. For survival analysis, a separate cohort of tumor-bearing mice treated under the same schedule was followed until reaching the animal ethics standards, and survival time was recorded for Kaplan-Meier analysis. Antitumor immune response analysis Fresh tumors, tumor-draining lymph nodes (TDLNs), and spleens were collected for flow-cytometric immune profiling. Tumor tissues were minced and digested with type IV collagenase at 37 ℃ for 3 h, filtered, and resuspended in PBS. Samples were stained with CD3-FITC, CD4-PE/Cy7, CD8-APC, CD11c-FITC, CD80-APC, and CD86-PE/Cy7 (30 min, 4 °C, dark), washed, and analyzed in FlowJo. Bulk RNA sequencing of tumor tissues Tumor samples from TETNPs@Gel and NS groups were snap-frozen in liquid nitrogen and stored at − 80 °C. RNA sequencing was performed by GENEWIZ (Suzhou, China). The workflow included RNA extraction, quality assessment, library construction, purification, quantification, clustering, and sequencing, with stringent quality control at each step. Libraries were pooled according to effective concentration and sequencing depth and subjected to Illumina sequencing. Single-cell RNA sequencing and bioinformatic analysis Fresh tumor tissues from the NS and TETNPs@Gel groups were dissociated into single-cell suspensions and subjected to single-cell RNA sequencing using the SeekOne DD single-cell 3′ transcriptome platform (Seekgene, China). Libraries were constructed according to the manufacturer’s protocol, quality-checked with an Agilent 4200 TapeStation, and sequenced at a target depth of at least 50,000 reads per cell. Raw reads were filtered with fastp and processed with SeekSoul Tools to generate the gene-cell expression matrix. Reads were aligned to the mouse reference genome (mm10) with STAR, and gene counts were obtained with featureCounts after UMI collapsing. Low-quality droplets were removed using the provider’s standard workflow, and cells with 200 ~ 100,000 UMIs and 100 ~ 30,000 detected genes were retained for downstream analysis. After quality control, 7,464 cells from the NS group and 10,043 cells from the TETNPs@Gel group were included. Highly variable genes were selected using the vst method (2,000 genes), batch effects were corrected with Harmony, and clustering was performed in Seurat based on 30 principal components at a resolution of 0.5. UMAP and t-SNE were used for dimensionality reduction and visualization. Cell-type annotation was performed using the SeekSoul reference-based workflow and further interpreted with canonical marker genes. Differential expression and enrichment analyses were carried out on the integrated dataset using the SeekSoul Online platform. In vivo safety evaluation One week after treatment, blood samples were collected via retro-orbital puncture to assess liver and kidney function. After 14 days, mice were sacrificed, and tumors and major organs (heart, liver, spleen, kidney, lung) were harvested for histological analysis. Tissues were fixed in 4% formaldehyde, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). Statistical analysis Statistical analyses were performed using GraphPad Prism 10. Data are presented as mean ± SEM. Comparisons were made using unpaired two-tailed t-tests or one-way ANOVA. Differences were considered significant at P < 0.05 (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). Flow cytometry data were analyzed with FlowJo, and graphs were prepared with Prism, Adobe Illustrator 2025 and R studio. Results Synthesis and characterization of TETNPs Since gelatinase is widely expressed in malignant tumors, we developed gelatinase-responsive nanoparticles (NPs) by inserting the gelatinase-cleavable peptide Pro-Val-Gly-Leu-Iso-Gly (PVGLIG) between polyethylene glycol (mPEG) and polycaprolactone (PCL) segments, resulting in the formation of mPEG-Pep-PCL NPs [20]. These NPs enable passive tumor targeting via microenvironment-responsive mechanisms. Tetrandrine (TET) was encapsulated in this copolymer using the emulsification–solvent evaporation method to generate TETNPs. The mean particle diameter was 208.6 ± 1.6 nm, the zeta potential was − 9.63 ± 0.02 mV, and the polydispersity index (PDI) was 0.21 ± 0.01 across three independent batches (Fig. 2A-B). Transmission electron microscopy (TEM) further confirmed the spherical morphology and uniform size distribution of TETNPs, with diameters consistent with dynamic light scattering results (Fig. 2C). Characterization of TETNPs. (A) Particle size distribution of TETNPs (n = 3). (B) Zeta potential of TETNPs (n = 3). (C) Representative TEM image of TETNPs. Scale bar, 500 nm. (D–F) Stability testing of TETNPs particle size, zeta potential, and polydispersity index (PDI) (n = 3). (G) Linear relationship between UV-visible absorption spectrum and gallic acid concentration (280 nm). (H) Optimal drug loading capacity and encapsulation efficiency of TETNPs. (I) Cumulative release of TET from TETNPs at pH 6.8 and pH 7.4 To assess stability, TETNPs were incubated in PBS at 37 °C under continuous agitation for seven days. Particle size, zeta potential, and PDI were measured daily, and no significant changes were detected, indicating good stability (Fig. 2D-F). A standard curve for TET was established using standard drug solutions at different concentrations (1, 2, 5, 12.5, 25, 50, 100 µg/mL) (Fig. 2G). The optimal drug loading capacity and encapsulation efficiency of TETNPs were determined to be 12.5% and 89.47%, respectively (Fig. 2H). The release behavior of TET from TETNPs exhibited clear pH responsiveness. Under mildly acidic conditions (pH 6.8), which approximate the tumor microenvironment, cumulative release rose rapidly during the first several hours and reached about 76% at 24 h, followed by a more gradual increase to approximately 83% at 96 h. In contrast, at physiological pH 7.4, release was slower, reaching about 64% at 24 h and about 76% at 96 h (Fig. 2I). These data indicate that TETNPs can accelerate tetrandrine release under tumor-relevant acidic conditions while still maintaining a sustained-release profile over time, thereby supporting local concentration build-up after administration. Synthesis and characterization of hydrogel HA-DOPA was synthesized via an EDC/NHS coupling reaction, as previously reported [22–23]. Dopamine (DOPA), recognized for its mussel-inspired adhesive properties, has been widely used in biomedical applications [24]. Successful synthesis of HA-DOPA was confirmed by ¹H-NMR. The ¹H-NMR spectrum displayed characteristic proton peaks for the HA backbone at 1.8-2.0 ppm and aromatic protons of the DOPA moiety at 6.5-7.0 ppm. The grafting ratio of HA-DOPA was calculated by comparing the integral values of the four hydrogen atoms on the two methylene groups in the DOPA structure with those of the three hydrogen atoms on the benzene ring, and with the integral values of the three hydrogen atoms on the methyl group in the HA structure. One DOPA molecule is grafted onto each HA repeating unit (Fig. 3A-B). Dopamine groups are known to coordinate with metal cations, enabling cross-linking and strengthening of hydrogel networks [25]. Prior studies have demonstrated that cations such as Fe²⁺ and K⁺ can coordinate with the catechol groups of HA-DOPA [26–27]. In this work, we aimed to fabricate a cross-linked hydrogel system through catechol-metal coordination, specifically using Mn²⁺ (from MnCl₂) to coordinate with DOPA residues on HA-DOPA chains. Hydrogel Synthesis and Characterization (A) Synthesis route for HA-DOPA. (B) ¹H-NMR spectrum of HA-DOPA in D₂O. The figure compares the integral values of four hydrogen atoms in the DOPA structure (green box) and three hydrogen atoms on the benzene ring (blue box) with those of three hydrogen atoms in the HA structure (red box). Each HA repeating unit bears one grafted DOPA molecule. (C) 18 mM HA-DOPA/Mn gelation experiment. (D) Sol-gel transition study at different Mn concentrations using the inverted method (1 week). (E) SEM images of 18 mM HA-DOPA/Mn oblique section (left) and cross-section (right). (F) Syringe injection of HA-DOPA/Mn to validate injectability. (G) Cut HA-DOPA/Mn hydrogels reconnect into a continuous structure, demonstrating self-healing properties. (H) Cumulative release of TET from TETNPs@Gel at pH 6.8 and pH 7.4 An inversion test was performed to evaluate the effect of Mn²⁺ concentration on gelation. After one week, only the sample with 18 mM Mn²⁺ formed a stable gel that adhered to the bottom of the vial (Fig. 3C-D). Other concentrations (14, 16, 20, and 30 mM) resulted in weak or no gelation, indicating that an appropriate Mn²⁺ concentration is essential for achieving rigid hydrogel formation. Based on these results, 18 mM Mn²⁺ was selected for subsequent experiments to ensure robust gelation while minimizing potential cytotoxicity. The microstructure of the freeze-dried hydrogel was examined using scanning electron microscopy (SEM), which revealed a three-dimensional honeycomb-like porous network (Fig. 3E). As a drug depot, the mechanical properties of hydrogels are critical, particularly their rheological behavior for local delivery applications. To assess injectability, the hydrogel was loaded into a 22G syringe and extruded smoothly, confirming its suitability for injection. Moreover, the hydrogel exhibited clear shear-thinning behavior when dispensed onto a Petri dish (Fig. 3F). The hydrogel also displayed self-healing capability: when separately dyed hydrogel pieces were brought into contact under mild pressure, they adhered and could be stretched, indicating rapid structural recovery (Fig. 3G). This property helps address challenges related to gelation kinetics and supports the effective application of in situ gelling drug delivery systems. After TETNPs were embedded into the HA-DOPA/Mn, the resulting TETNPs@Gel retained pH-dependent release but displayed a more selective depot effect than the nanoparticles alone. At pH 6.8, cumulative TET release from TETNPs@Gel increased quickly in the early phase and reached approximately 79% at 24 h, eventually approaching 84% at 96 h. In contrast, at pH 7.4 the release was markedly restrained, reaching only about 37% at 24 h and about 46% at 96 h (Fig. 3H). When compared with free TETNPs, incorporation into the hydrogel therefore substantially reduced neutral-pH leakage while preserving efficient drug liberation under acidic conditions. These data show that the HA-DOPA/Mn hydrogel functions as more than a passive carrier. By integrating a porous, injectable, and self-healing matrix with gelatinase-responsive nanoparticles, the TETNPs@Gel system preserves rapid release under tumor-mimicking acidic conditions but suppresses premature drug escape at physiological pH. This release behavior is consistent with the design objective of achieving localized high drug exposure at the tumor site while reducing unnecessary diffusion to normal tissues. Biodegradation test in vivo Magnetic resonance imaging (MRI) was used to monitor the in vivo fate of the hydrogel depot. In the imaging experiment, HA-DOPA/Mn produced a clear positive signal on T1-weighted MRI immediately after injection, whereas HA-DOPA alone showed markedly weaker and less durable enhancement (Fig. 4A). Longitudinal imaging further demonstrated that the signal at the HA-DOPA/Mn injection site remained readily detectable throughout the observation window and could still be visualized on day 7, while the signal from the HA-DOPA control declined much more rapidly. Quantitative analysis of the region-of-interest signal intensity showed that the HA-DOPA/Mn group consistently maintained higher values than the HA-DOPA group at all post-injection time points (Fig. 4B-C). These data indicate that, under the present formulation conditions, Mn incorporation mainly endows the hydrogel with T1-positive contrast capability and enables noninvasive visualization of depot retention in vivo. This revised interpretation is also more consistent with the physicochemical behavior of Mn2+-containing biomaterials. As a paramagnetic ion, Mn2+ primarily shortens the longitudinal relaxation time of surrounding water protons and is therefore widely used as a positive T1 contrast source [28]. In Mn-labeled hydrogels, strong T1 enhancement has likewise been reported as the dominant imaging feature. Accordingly, T1-weighted MRI is presented here as the principal readout of the HA-DOPA/Mn system. At the same time, we consider it scientifically reasonable to retain the previous T2-weighted observations as a secondary phenomenon rather than the main imaging mode. Earlier studies have noted that Mn-based agents may also exert a minor T2-shortening effect, and local confinement, coordination, and spatial heterogeneity within a cross-linked matrix can further influence transverse relaxation by increasing local susceptibility-related field perturbation [29–31]. To validate this property, we subcutaneously injected HA-DOPA and HA-DOPA/Mn hydrogels into the abdominal region of mice and performed continuous T2WI. We further performed MRI following intratumoral injection in tumor-bearing mice, using the same procedures as for subcutaneous administration. Compared to subcutaneous injection, the control group exhibited slightly prolonged retention, likely due to the tumor microenvironment (Table. S1). To further assess in vivo degradation and retention, we conducted near-infrared fluorescence (NIRF) imaging after subcutaneous injection of CY5-labeled hydrogels. Both CY5 and HA-DOPA groups showed a rapid decline in fluorescence within 4 h, with only faint residual signal by 8 h. In contrast, the HA-DOPA/Mn group exhibited detectable fluorescence up to 144 h, confirming its week-long retention in vivo (Fig. 4D-E). Incorporation of TETNPs did not significantly alter hydrogel biodegradability, and ex vivo imaging after 7 days revealed higher fluorescence intensity in tumor tissues from the TETNPs@Gel group, indicating enhanced tumor targeting (Fig. 4F). Collectively, findings crosslinked hydrogel structure significantly enhances retention capacity compared to unmodified hydrogels in vivo and in vitro. The incorporation of TETNPs did not compromise retention while conferring favorable tumor-targeting capability. In vivo imaging and retention evaluation of HA-DOPA/Mn hydrogel. (A) Schematic illustration of Mn-coordinated HA-DOPA/Mn hydrogel formation and MRI monitoring. (B) Representative T1-weighted MR images of the hydrogel depot after injection at the indicated time points; arrows indicate the injection site. (C) Quantitative analysis of T1-weighted signal intensity over time. (D) Longitudinal near-infrared fluorescence images of free CY5, HA-DOPA/CY5, HA-DOPA/Mn/CY5, and HA-DOPA/Mn/TETNPs-CY5. (E) Quantification of fluorescence intensity changes. (F) Ex vivo fluorescence imaging of tumor and major organs collected on day 7 TETNPs inhibit tumor cells and promote DC maturation in vitro We assessed the pro-apoptotic effect of TETNPs using flow cytometry (Fig. 5A-B). Moreover, TET-induced tumor cell apoptosis was observed to be dose-dependent, with TETNPs exhibiting an enhanced capacity to promote apoptosis compared to free TET. To evaluate the superior synergistic therapeutic effect of TETNPs@Gel in vitro, we monitored cytotoxicity over 24 h (Fig. 5C). The TETNPs@Gel formulation showed significantly higher antitumor activity against 4T1 cells compared to the free drug. In contrast to the TET and TETNPs treatment groups, the percentage of viable cells was notably reduced in the TETNPs@Gel group. Nanoparticle-induced apoptosis and cellular uptake and cytotoxicity of cells subjected to different treatments. (A) Representative flow cytometry images of apoptosis in 4T1 cells following treatment with different concentrations of TETNPs (n = 6). (B) Statistical analysis of apoptosis in 4T1 cells treated with varying concentrations of TETNPs. (C) In vitro cytotoxicity of NS, TET, TETNPs, and TETNPs@Gel at different concentrations on 4T1 cells after 24 h (n = 6). (D) Representative fluorescence images of apoptosis in 4T1 cells treated with different concentrations of TETNPs (n = 3). Green fluorescence indicates live cells treated with membrane-bound protein V-FITC; red fluorescence indicates dead cells treated with PI. Scale bar: 200 μm.(E) Confocal microscopy images of 4T1 cells treated with different concentrations of C6-NPs and C6 solution (n = 3). Scale bars: 10 μm and 60 μm. (F) Confocal microscopy images of BMDCs treated with varying concentrations of C6-NPs and C6 solution (n = 3). Scale bar: 10 μm.(G) Flow cytometric analysis of MHC-II and CD86 expression in CD11c+ BMDC populations 24 h post-treatment with NS, TET, NPs, and TETNPs (n = 6) *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 We also employed Calcein-AM/PI staining to assess drug-induced cytotoxicity in 4T1 cells. Fluorescence microscopy clearly elucidated the cytotoxic effects across treatment groups. The strongest green fluorescence in the NS group was observed, indicating minimal cell death (Fig. 5D). By contrast, the HA-DOPA/Mn, TETNPs, and TETNPs@Gel groups all exhibited substantial red fluorescence, with the highest intensity observed in the TETNPs@Gel group. These findings collectively indicate that the nanocomposite hydrogel significantly enhances the antitumor efficacy of TET. Concurrently, we evaluated the capacity of nanoparticles to deliver drugs to 4T1 cells. When coumarin-6-loaded nanoparticles served as fluorescent probes, 4T1 cells exhibited significantly higher uptake of these nanoparticles compared to free coumarin-6 (Fig. 5E). We evaluated the ability of nanoparticles to deliver cargo to dendritic cells. Using coumarin-6–loaded nanoparticles as a fluorescent probe, bone marrow–derived dendritic cells (BMDCs) exhibited greater uptake than with free coumarin-6, with uptake increasing in a dose-dependent manner (Fig. 5F). Flow cytometry further showed that TETNPs enhanced BMDC cytokine production and promoted maturation, as evidenced by higher proportions of CD86⁺/MHC II⁺ cells (Fig. 5G). Local TETNPs@Gel therapy inhibits breast cancer growth We administered TETNPs@Gel injections once a week and used two weeks to evaluate the tumor suppression effect (Fig. 6A). TETNPs@Gel suppressed tumor growth more effectively than saline (NS) and blank gel controls, and it also outperformed TETNPs (Fig. 6B-C, H). Representative gross images of excised tumors (Fig. 6D) showed that five mice in the TETNPs@Gel group exhibited reductions in tumor volume relative to baseline, consistent with a partial response (PR). Kaplan–Meier analysis demonstrated significantly prolonged survival in the TETNPs@Gel group (Fig. 6E). Moreover, no significant changes in body weight were observed across treatment groups, supporting the in vivo safety profile of TETNPs@Gel (Fig. 6F-G). Schematic of the experimental protocol for the antitumor effect of the nanogel delivery system (TETNPs@Gel) on the 4T1 breast cancer model (A). On days 7 and 14 post-subcutaneous implantation of 4T1 tumors in the left inguinal region of Balb/C mice, TETNPs@Gel, HA-DOPA/Mn, TETNPs, TET, NPs, and NS were administered peritumorally. Mouse body weight and tumor volume were measured every other day for 14 days post-administration. Tumors, lymph nodes, and spleens were analyzed on day 21. (B-C) Tumor growth curves representing mean tumor volumes across groups (n = 8) *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. (D) Representative gross images of excised tumors from different treatment groups (n = 5). (E) Tumor survival curves for each group. (F-H) Mouse tumor volumes curves (n = 8) Immune response induced by TETNPs@Gel To evaluate the systemic immune response elicited by TETNPs@Gel, we performed flow cytometric analysis of immune cells from tumor-draining lymph nodes (TDLNs), spleen, and tumor tissues on day 14 post-treatment. Notably, TETNPs@Gel induced the highest proportion of mature dendritic cells (mDCs; CD11c⁺ CD80⁺ CD86⁺) in TDLNs (89.6%), significantly outperforming the HA-DOPA/Mn (74.3%), TETNPs (66.6%), and control groups (44.7%) (Fig. 7A-B). Treatment with TETNPs@Gel elicited systemic antitumor immune responses (n = 6). (A–B) Representative images and percentages of activated dendritic cells (CD11C⁺ CD80⁺ CD86⁺) in lymph nodes. (C–D) Representative images and percentage of CD3⁺ CD8⁺ T cells in lymph nodes. (E–F) Representative image and percentage of memory T cells (CD4+ CD44+ CD62L+) in the spleen. (G) Representative image of activated dendritic cells (CD11C+ CD80+ CD86+) in the spleen. (H–I) Representative images and percentage of activated dendritic cells (CD11C⁺ CD80⁺ CD86⁺) in tumors. (J-K) Representative images and percentage of CD3⁺ CD8⁺ T cells in tumors. Data are expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. (L) Representative images of tumor CD8+ and Ki-67 immunohistochemical staining Consistent with enhanced antigen presentation, TETNPs@Gel also boosted adaptive immune activation. We observed increased frequencies of CD3⁺ CD8⁺ T cells in TDLNs across the TETNPs@Gel, HA-DOPA/Mn, and TETNPs groups (Fig. 7C-D). Furthermore, splenic memory T cells (CD44⁺ CD62L⁺) were significantly elevated in the TETNPs@Gel group, suggesting the induction of durable immunological memory (Fig. 7E-F) [32]; there is also a slight increase in splenic dendritic cells (Fig. 7G). Within the tumor microenvironment, TETNPs@Gel treatment resulted in markedly higher infiltration of both DCs and CD3⁺ CD8⁺ T cells compared to other groups (Fig. 7H-K), corroborated by immunohistochemistry showing enhanced CD8⁺ T cell infiltration and reduced Ki-67 proliferating tumor cells (Fig. 7L). These data collectively demonstrate that TETNPs@Gel enhances antitumor immunity through a multi-modal mechanism: (i) sustained intratumoral release of TET promotes immunogenic cell death and antigen availability; (ii) the hydrogel scaffold facilitates DC maturation and antigen presentation; and (iii) the resulting T cell activation, infiltration, and memory differentiation collectively reverse the immunosuppressive tumor microenvironment. In contrast, TETNPs without hydrogel encapsulation lacked sustained release and microenvironment-modulating properties, while free TET suffered from rapid systemic clearance and insufficient tumor accumulation. The blank hydrogel alone showed no significant immune activation or antitumor effect. We propose that TET released from the hydrogel promotes tumor cell apoptosis, antigen release, and subsequent DC maturation, while the hydrogel matrix itself serves as an immunocompatible reservoir that maintains high local drug concentrations and supports long-term survival and function of immune cells. TETNPs@Gel elicited systemic antitumor activity and suppressed pulmonary metastasis in a bilateral 4T1 model To determine whether local administration of TETNPs@Gel could generate antitumor effects beyond the treated lesion, a bilateral 4T1 tumor model was established. Primary tumors were inoculated subcutaneously and treated peritumorally twice, after which a contralateral untreated tumor was implanted to evaluate systemic tumor control (Fig. 8A). In this setting, TETNPs@Gel displayed the strongest inhibitory effect on distant-tumor progression among all groups (Fig. 8B). Compared with NS and the other control formulations, the growth of the untreated contralateral tumors was delayed more effectively in the TETNPs@Gel group, and this suppressive effect was accompanied by a marked reduction in endpoint distant-tumor weight (Fig. 8C). Kaplan–Meier analysis further showed a clear survival advantage in mice receiving TETNPs@Gel (Fig. 8D), supporting that this formulation acted as more than a local depot and was able to induce a systemic antitumor response. TETNPs@Gel elicited systemic antitumor activity in a bilateral 4T1 model and TETNPs@Gel reduced pulmonary metastatic burden and did not cause obvious histological injury to major organs. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. (A) Experimental design of the bilateral tumor model. (B) Growth curves of untreated distant tumors after local therapy at the primary site (n = 5). (C) Endpoint weights of distant tumors (n = 5). (D) Kaplan–Meier survival curves (n = 8). (E) Representative gross images of lungs together with corresponding H&E staining lung sections. Arrows and circles indicate representative metastatic nodules or foci (n = 3). Scale bars: 2000 μm. (F) Representative H&E staining of lung, liver, spleen, kidney, and heart collected at the experimental endpoint (n = 3). Scale bars: 100 μm The anti-metastatic consequence of this systemic effect was further evaluated by gross observation of the lungs and histopathological examination of lung sections. Multiple visible metastatic nodules and extensive metastatic foci were observed in the lungs from the NS group. Although metastatic burden was reduced to varying degrees in the TET, TETNPs, HA-DOPA/Mn and TET@Gel groups, residual lesions remained readily detectable. In contrast, the TETNPs@Gel group showed the lowest pulmonary metastatic burden, with fewer visible surface nodules and less pronounced metastatic foci on H&E-stained sections (Fig. 8E). The alveolar architecture was also relatively better preserved in this group, indicating that local treatment with TETNPs@Gel not only restrained the growth of the distant untreated tumor but also suppressed spontaneous pulmonary dissemination in the highly metastatic 4T1 model. Major-organ H&E staining performed at the same endpoint did not reveal obvious treatment-related pathological abnormalities in the liver, spleen, kidney, or heart (Fig. 8F). Taken together, these findings indicate that TETNPs@Gel achieved systemic tumor control without overt histological toxicity under the present experimental conditions. This bilateral-tumor experiment therefore provides functional support for the view that TETNPs@Gel behaves in an in situ vaccination-like manner, coupling local therapy with a measurable whole-body antitumor benefit. TETNPs@Gel remodeled vascular and complement-associated features of the tumor microenvironment To further define tissue-level changes associated with the therapeutic response, tumor sections were stained for the vascular- and stromal-associated markers CD31 and α-SMA. Distinct differences were observed among treatment groups. In the NS, TET, TETNPs and HA-DOPA/Mn groups, CD31-positive structures were relatively sparse and less organized, and α-SMA staining remained weak or discontinuous in most fields. By contrast, tumors treated with TET@Gel and, more prominently, TETNPs@Gel showed stronger CD31-positive vascular profiles together with more evident α-SMA-associated perivascular/stromal staining (Fig. 9A). These histological changes are consistent with treatment-associated vascular and perivascular remodeling within the tumor microenvironment. Although single-marker chromogenic staining does not by itself establish perfusion or definitive pericyte coverage, the combined CD31 and α-SMA patterns support a shift toward a more structured vascular/stromal phenotype after TETNPs@Gel treatment. TETNPs@Gel remodeled vascular/perivascular features and enhanced local complement signaling in tumor tissues (n = 3). (A) Representative immunohistochemical staining of α-SMA and CD31 in tumor sections from the indicated groups. (B) Intratumoral C3a levels. (C) Intratumoral C5a levels Because our transcriptomic analyses pointed to complement-related pathways, we examined complement effector levels in tumor tissues. Quantification of intratumoral C3a and C5a revealed that both mediators were markedly elevated in the TETNPs@Gel group relative to the control groups, with C5a showing a particularly pronounced increase (Fig. 9B-C). When considered together with the enhanced dendritic-cell maturation, increased intratumoral CD8⁺ T-cell infiltration, distant-tumor suppression, and reduced pulmonary metastasis described above, these data suggest that TETNPs@Gel reshaped the treated tumor into a microenvironment that was more permissive for antitumor immune activation. Single-cell transcriptomic profiling reveals macrophage-state remodeling induced by TETNPs@Gel To resolve the cellular basis of the therapeutic response, representative tumors from the NS and TETNPs@Gel groups were subjected to single-cell RNA sequencing. After quality control, filtering, and integration, 17,507 high-quality cells were retained, including 7,464 cells from the NS group and 10,043 cells from the TETNPs@Gel group. Unsupervised analysis resolved 17 transcriptionally distinct clusters (Table. S2), which were further consolidated into six major cell populations: macrophages, granulocytes, fibroblasts, T cells, monocytes, and endothelial cells (Fig. 10A). Across the integrated atlas, macrophages constituted the dominant compartment (9,024/17,507, 51.55%). When stratified by treatment, macrophages remained the largest population in both groups, accounting for 54.22% of cells in NS tumors and 49.56% in TETNPs@Gel tumors. In parallel, TETNPs@Gel tumors contained proportionally more T cells (6.27% vs. 3.47%) and fibroblasts (17.66% vs. 15.45%), indicating broad remodeling of the tumor microenvironment rather than a simple expansion or depletion of one lineage (Fig. 10B). Single-cell atlas of NS and TETNPs@Gel tumors. (A) UMAP visualization of the six major cell populations identified after integration. (B) Relative abundance of each major cell population in the NS and TETNPs@Gel groups. (C) UMAP visualization of the 17 transcriptionally distinct clusters. (D) Dumbbell plot showing treatment-associated redistribution of clusters, with key macrophage-related clusters highlighted. Percentages are shown as the proportion of total high-quality cells within each sample At the cluster level, treatment induced a clear redistribution of macrophage-rich states. Clusters 1, 4, 5, and 16 increased after TETNPs@Gel treatment, whereas cluster 2 declined markedly; cluster 10 was largely maintained. Specifically, cluster 1 increased from 15.35% to 17.13%, cluster 4 from 5.94% to 9.34%, cluster 5 from 5.73% to 6.48%, and cluster 16 from 0.21% to 0.43%, while cluster 2 decreased from 20.35% to 10.61% (Fig. 10C-D). This pattern argues against a uniform shift along a rigid M1/M2 axis and instead supports treatment-dependent remodeling among discrete macrophage states, which is more consistent with current single-cell views of tumor-associated macrophage heterogeneity. Marker-based refinement further clarified the identity of the treatment-responsive clusters. Cluster 1 was enriched for antigen-processing and presentation genes, including Cd74, H2-Aa, H2-Ab1, H2-Eb1, Ctss, and H2-DMa, together with Il1b, Cd14, and Cybb, supporting its designation as an antigen-presenting activated macrophage population. Cluster 5 displayed a strong interferon-responsive inflammatory program marked by Rsad2, Ifit1, Ifit2, Isg15, Cxcl10, Ccl4, and Ccl5, while retaining antigen-presentation features, consistent with an inflammatory APC-like macrophage state. Cluster 4 was defined by Mrc1, Cd163, Folr2, F13a1, Selenop, and Apoe, indicating a resident-like macrophage program with overlap with the FOLR2-positive perivascular/tissue-resident states described in breast cancer. Cluster 10 expressed C1qa, C1qb, C1qc, Ifi30, Mki67, and Top2a, consistent with a proliferating C1q+ macrophage subset. Cluster 16 was characterized by Ctsk, Acp5, Mmp9, Atp6v0d2, and Nfatc1, indicating an osteoclast-like or matrix-remodeling phenotype. By contrast, cluster 2 showed high expression of Spp1 together with Krt8, Krt18, Epcam, and Tm4sf1, supporting annotation as an Spp1-high hybrid remodeling cluster rather than a canonical macrophage subtype (Fig. 11A-B). Macrophage-state annotation supports treatment-induced immune remodeling. (A) Abundance shifts in selected macrophage-related clusters between NS and TETNPs@Gel tumors. (B) Heatmap of representative cluster-enriched genes derived from cluster-specific differential analysis These data converge on a consistent interpretation: TETNPs@Gel reduced a mixed remodeling-dominant state while enriching antigen-presenting and inflammatory macrophage programs. This shift is concordant with the enhanced DCs maturation and increased intratumoral CD8+ T-cell infiltration observed in the orthogonal immune analyses, and it offers a cellular explanation for the complement/coagulation and extracellular-matrix remodeling signatures identified by bulk transcriptomic profiling. Accordingly, the single-cell dataset supports the view that TETNPs@Gel functions not merely as a local cytotoxic depot, but as a myeloid-remodeling platform that reprograms the treated tumor toward a more immunologically permissive state. Potential mechanism of complement-mediated immune activation To delineate the molecular mechanisms governing TETNPs@Gel-mediated immune remodeling, we conducted bulk RNA sequencing on tumors harvested from 4T1-bearing mice following treatment with TETNPs@Gel or vehicle control. Differential expression analysis identified 122 significantly upregulated and 40 downregulated transcripts (Fig. 12A-B), indicating substantial transcriptional reprogramming induced by the nanocomposite hydrogel system. Gene Ontology (GO) enrichment analysis revealed pronounced alterations in immune-relevant biological processes, including leukocyte cell-cell adhesion, regulation of cell migration, and actin cytoskeleton/extracellular matrix (ECM) organization (Fig. 12C). These changes were accompanied by enrichment in focal adhesion and calcium channel activity terms, collectively consistent with active stromal remodeling that facilitates immune cell trafficking. KEGG pathway analysis further demonstrated significant enrichment of the complement and coagulation cascades, PPAR signaling, arachidonic acid metabolism, and downregulation of cytochrome P450-mediated xenobiotic metabolism (Fig. 12D). RNA sequencing analysis of differentially expressed genes in 4T1 mouse tumors following TETNPs@Gel treatment. (A) Distribution of differentially transcribed transcripts (TETNPs@Gel versus control).X-axis: log₂ fold change; higher absolute values indicate greater expression difference. Y-axis: -log₁₀ p-value; higher values denote more significant expression differences. Red dots: Upregulated genes; Blue dots: Downregulated genes; Grey dots: Genes with no significant difference. Red dots: Upregulated genes; Blue dots: Downregulated genes; Grey dots: Genes without significant difference. (B) Heatmap of common gene expression profiles. (C) GO enrichment analysis of DEGs involved in biological effects induced by TETNPs@Gel (Biological Process (BP), Molecular Function (MF), and Cellular Component (CC)). (D) KEGG enrichment analysis. (E) Functional interaction network of DEGs The observed complement signature establishes a potential mechanistic link between transcriptomic alterations and immunophenotypic outcomes (Fig. 12E). Upregulation of Mbl1 and Mbl2—identified as hub genes within the protein-protein interaction network (PPI)—reflects activation of the lectin pathway, wherein MBL-MASP complexes trigger the formation of C3/C5 convertases [33–34]. Subsequent generation of anaphylatoxins C3a and C5a promotes the recruitment and functional maturation of antigen-presenting cells (APCs), providing a molecular basis for the elevated frequency of CD11c⁺CD80⁺CD86⁺ dendritic cells observed in tumor-draining lymph nodes [35–36]. Complement opsonins, including C3b and iC3b, facilitate phagocytosis of tumor-derived material, enhancing antigen cross-presentation, while assembly of the membrane attack complex (C5b-9) induces direct lysis of susceptible tumor cells [37–39], collectively amplifying antigen availability and priming robust CD8⁺ T-cell responses—consistent with the enhanced intratumoral CD3⁺CD8⁺ T-cell infiltration detected in treated mice. The coagulation cascade, enriched in concert with complement, may further reinforces antitumor immunity through multiple cooperative mechanisms. Thrombin-mediated cleavage of C5 into C5a provides an auxiliary route for anaphylatoxin production, augmenting APC recruitment. Concurrently, fibrin and fibrinogen deposition induces perivascular ECM restructuring, establishing pro-adhesive migratory tracks that support leukocyte extravasation and interstitial motility—findings that align with GO terms related to ECM organization and cell adhesion, as well as histological evidence of stromal reorganization [39]. Protease-activated receptor (PAR) signaling on myeloid populations further amplifies APC activation and trafficking, thereby bridging innate immune sensing with adaptive T-cell priming. Critically, these immunomodulatory pathways operate within a context of sustained local drug exposure enabled by the TETNPs@Gel depot. The Mn²⁺-crosslinked hyaluronic acid hydrogel ensures prolonged intratumoral retention of tetrandrine nanoparticles, establishing a biphasic pharmacodynamic profile: an initial high-concentration phase directly elicits tumor cell apoptosis and liberates antigenic substrates and damage-associated molecular patterns, priming complement-mediated opsonization and dendritic cell uptake; followed by a sustained low-concentration phase that normalizes the tumor vasculature and stromal architecture, thereby promoting immune cell infiltration and sustaining effector function. Corroborating this model, pathway enrichment related to ECM organization, cell migration/adhesion, and lipid-inflammatory signaling (PPAR/arachidonic acid) reflects broad stromal and vascular normalization, while downregulation of cytochrome P450 genes (e.g., Cyp3a25, Cyp2c23) indicates attenuated metabolic immunosuppression, preserving chemotactic gradients and antigen-presentation efficacy. Together, these transcriptomic insights unify our flow cytometric and immunohistochemical observations, establishing complement–coagulation activation coupled with APC maturation and CD8⁺ T-cell infiltration as a central mechanism through which the TETNPs@Gel system achieves durable antitumor immunity via. spatiotemporally controlled drug delivery. Biosafety assessment To assess the biosafety of the drug delivery system, 4T1 tumor-bearing mice were euthanized two weeks post-treatment, and major organs (heart, liver, spleen, lungs, and kidneys) were collected for hematoxylin and eosin (H&E) staining (Fig. 13A). As shown, no significant histopathological differences were observed among the TETNPs@Gel, HA-DOPA/Mn, HA-DOPA, TETNPs, TET, NPs, and NS groups, indicating that the nanocomposite hydrogel system did not induce additional toxicity compared to single-agent formulations. Moreover, the localized delivery of tetrandrine via the nanocomposite hydrogel may reduce off-target adverse effects associated with systemic administration. Previous studies have reported that tetrandrine exhibits a short circulation half-life and may induce cardiotoxicity or affect fetal development. Therefore, local injection of tetrandrine-loaded nanocomposite gels as an in-situ vaccine could significantly mitigate these systemic side effects. Biosafety assessment. (A) Representative H&E-stained images of lungs, livers, spleens, kidneys, and hearts from 4T1 mice following different treatments. Scale bar: 50 μm. (B) Hepatic and renal function indicators in each group of mice (n = 3) Furthermore, serum biochemical analysis confirmed the excellent biosafety profile of TETNPs@Gel, with no evidence of hepatic or renal impairment across all treatment groups (Fig. 13B). Key liver function markers—ALT, AST, and ALB—as well as renal function parameters—urea and CREA—remained within normal physiological ranges in the TETNPs@Gel group, showing fold changes comparable to those in the normal saline (NS) group. These results demonstrate that TETNPs@Gel does not elicit detectable hepatotoxicity or nephrotoxicity, underscoring its high biocompatibility and potential for safe in vivo application. Conclusion We developed an injectable, MRI-visible Mn²⁺-hyaluronic acid hydrogel encapsulating tetrandrine-loaded nanoparticles (TETNPs) as a biphasic, long-acting intratumoral depot. By maintaining local tetrandrine concentrations within a defined window, this formulation is mechanistically linked to a dual-function immunochemotherapeutic response: an initial high-dose pulse that induces tumor-cell death and enhances antigen and damage-associated molecular pattern release, followed by a sustained low-dose phase that suppresses angiogenesis, normalizes the tumor microenvironment, and facilitates antigen-presenting-cell trafficking and effector T-cell entry. Multi-omic and cellular analyses support a coherent mechanism in which complement and coagulation cascades are transcriptionally engaged (including upregulation of Mbl1/Mbl2), leading to generation of C3a/C5a to recruit and mature antigen-presenting cells, C3b/iC3b to promote antigen handling, and membrane attack complex formation to augment tumor-cell lysis, ultimately resulting in increased CD8⁺ T-cell infiltration and antitumor activity. Concurrent modulation of PPAR/arachidonic-acid signaling and xenobiotic/CYP450 pathways indicates metabolic reprogramming that relieves immunosuppression and stabilizes chemotactic cues. Collectively, these findings demonstrate that the Mn²⁺-HA/TETNP hydrogel functions as an MRI-trackable in situ nanovaccine that integrates direct cytotoxicity with durable immune and stromal remodeling in solid tumors. Electronic Supplementary Material Below is the link to the electronic supplementary material. Supplementary Material 1. Author contributions W.W is in charge of the overall research and the main manuscript text writing. S.J is responsible for the article revision and experimental guidance. D.H, Y.Q, D.L, X.G, H.Q, H.Z, L.C, J.S, X.Q, W.C and L.Y assist with experiments and experimental guidance. B.L, L.X and R.L provide research ideas and experimental guidance. All authors reviewed the manuscript. Funding This study was supported by National Natural Science Foundation of China (82272852), Science Fund for Distinguished Young Scholars of Jiangsu Province (BK20240017). Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval Analysis was conducted on mice whose samples were obtained in accordance with a protocol approved by the Nanjing Gulou Hospital Laboratory Animal Welfare Ethics Review Committee. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Baorui Liu, Email: baoruiliu@nju.edu.cn. Li Xie, Email: sherry01130@163.com. Rutian Li, Email: rutianli@nju.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement No datasets were generated or analysed during the current study.