Multifunctional hyaluronic acid-modified hollow MnO2 nanoplatform co-delivering with L-Arginine and Regorafenib for synergistic therapy against colorectal cancer Authors: Chong Li, Wei Liu, Ningfu Li, Yanyan Zhang, Ke Wu, Shaohong Yu Published: 2026 ABSTRACT AbstractThe high rates of recurrence and metastasis in colorectal cancer (CRC), coupled with its immunosuppressive tumor microenvironment (TME), severely constrain patient survival. To overcome the limitations of monotherapies, this study developed a multifunctional targeted nanoplatform (RLHMnH). This system consists of a hyaluronic acid (HA)-modified hollow manganese dioxide (H-MnO2) nanocarrier co-loaded with the anti-angiogenic drug Regorafenib (Reg) and the nitric oxide precursor L-Arginine (L-Arg) for synergistic triple therapy against CRC. The platform leverages HA-mediated CD44 receptor targeting to achieve enhanced tumor accumulation. Within the TME, the H-MnO2 framework undergoes rapid responsive decomposition. On one hand, Within the acidic and GSH-rich TME, H-MnO2 undergoes reductive degradation to release Mn2+, concomitantly depleting GSH and disrupting redox homeostasis. The released Mn2+ further catalyzes Fenton-like conversion of endogenous H2O2 into highly cytotoxic ·OH, thereby inducing oxidative damage and tumor cell apoptosis. On the other hand, the concomitantly released Reg effectively blocks tumor angiogenesis by inhibiting the VEGFR2 signaling pathway, while L-Arg generates nitric oxide, synergistically reversing drug resistance, alleviating hypoxia, and activating anti-tumor immunity. Both in vitro and in vivo experiments demonstrated that RLHMnH, through integrating the triple synergistic mechanisms of CDT, anti-angiogenesis, and gas therapy, achieved a remarkable tumor inhibition rate of ∼75% in a CT26 xenograft model and potently suppressed CT26 cell viability to below 20% at 72 h, showcasing significant suppression of CRC. This study provides an innovative nano-carrier-based synergistic strategy to overcome the current therapeutic bottlenecks in CRC treatment.Keywords: Colorectal cancer, Hollow manganese dioxide, Regorafenib, L-arginine, Synergistic therapy Introduction Colorectal cancer (CRC) ranks as the third most commonly diagnosed malignancy globally and represents the second leading cause of cancer-related mortality [1,2]. Radical surgical resection combined with lymph node dissection remains the gold standard for treating localized CRC. However, postoperative residual micrometastases and the presence of an immunosuppressive tumor microenvironment (TME) often lead to local recurrence and distant metastasis, posing a critical bottleneck for long-term patient survival [3,4]. Current clinical interventions encompass strategies such as chemotherapy, molecular targeted therapy, radiotherapy, immune checkpoint inhibitor therapy, and emerging nano-drug delivery systems. Nonetheless, monotherapies are often limited by suboptimal response rates or acquired resistance [5,6]. Notably, a triple-therapy system that integrates chemosensitization, targeted regulation, and immune activation through multimodal synergistic mechanisms can not only overcome the biological barriers of conventional treatments but also achieve enhanced efficacy and reduced side effects via spatiotemporally precise drug delivery, offering an innovative solution to improve the predicament of CRC treatment [1,7]. Regorafenib (Reg), a multi-target tyrosine kinase inhibitor, effectively inhibits angiogenesis and tumor proliferation in CRC by precisely modulating key signaling pathways such as the vascular endothelial growth factor receptor (VEGFR) [8,9]. Compared to traditional chemotherapeutic agents, its mechanism of multi-dimensionally blocking tumor progression significantly enhances treatment specificity while reducing the risk of systemic toxicity. Addressing the high heterogeneity and immune evasion characteristics of metastatic CRC (mCRC), Regorafenib can inhibit neovascularization in metastatic foci by downregulating the HIF-1α/VEGF axis in tumor cells and enhance the anti-tumor activity of the immune microenvironment by modulating the polarization state of tumor-associated macrophages (TAMs) [10]. The CORRECT clinical trial confirmed that Regorafenib extends the median overall survival by 1.4 months in patients with standard chemotherapy-resistant mCRC and significantly increases the disease control rate to 41% [11]. However, this drug still faces the common bottlenecks of targeted therapy, particularly regarding drug delivery and resistance [12,13]. Therefore, optimizing the spatiotemporal distribution of the drug through nano-drug delivery systems has become a vital research direction for breaking through the current therapeutic impasse. On the other hand, although Regorafenib exerts potent anti-angiogenic effects in mCRC, its therapeutic efficacy is considerably hampered by the immunosuppressive and hypoxic tumor microenvironment, which promotes drug resistance. Therefore, there is an urgent need to combine anti-angiogenic therapy with synergistic modalities capable of modulating the TME, such as chemodynamic and gas therapy. Chemodynamic therapy (CDT) induces tumor cell damage or apoptosis by generating highly reactive hydroxyl radicals (·OH) from the abundant hydrogen peroxide (H2O2) in the TME via Fenton or Fenton-like reactions [14,15]. However, the high expression of glutathione (GSH), a key antioxidant in tumor cells, scavenges reactive oxygen species (ROS), thereby compromising the efficacy of CDT [16]. Consequently, selecting CDT nano-agents capable of effectively intervening in the high-GSH TME is crucial. Developing CDT nano-systems that can efficiently and thoroughly deplete GSH holds significant value for improving treatment specificity and effectiveness, yet it remains a substantial challenge. Certain metal-based nanomaterials can disrupt the redox balance of the TME by specifically binding to GSH, while the released metal ions can produce reactive radicals via Fenton-like reactions. These unique properties make them promising candidates for developing nano-agents for tumor CDT [17]. Among various candidate materials, manganese dioxide (MnO2) has attracted considerable attention due to its excellent biocompatibility and ability to metabolize into water-soluble Mn2+ ions [18,19]. Notably, the dissociation of nanostructures during the CDT process opens new possibilities for their use as drug carriers. However, traditional MnO2 nanostructures (e.g., nanoparticles, nanosheets) are less ideal carriers due to their limited drug-loading capacity. In contrast, hollow manganese dioxide (H-MnO2, HMn) materials, leveraging their cavity structure, not only achieve high drug loading but also enable the construction of precise and controllable drug release systems [20,21]. Developing novel HMn-based drug delivery systems based on this characteristic is of significant research value. The application of such multifunctional nanotheranostic agents is expected to break through the technical bottlenecks of traditional drug delivery, providing innovative solutions to enhance the efficiency of tumor therapy. Despite the individual promises of these therapies, no study has yet integrated Regorafenib, L-Arg, and H-MnO2 into a single nanoplatform to achieve triple-synergistic therapy combining anti-angiogenesis, chemodynamic therapy, and nitric oxide (NO) gas therapy. Based on these considerations, we constructed Reg-loaded H-MnO2 nanoparticles modified with hyaluronic acid (HA) and L-Arginine (L-Arg) to generate ·OH irrespective of TME limitations. HA can target CRC cells via the CD44 receptor, enhancing the cellular internalization efficiency of the nanoparticles and promoting their accumulation within the tumor [22]. Simultaneously, L-Arginine serves as a substrate for NOS-associated NO biosynthesis in tumor cells and the inflammatory TME, which can reverse tumor drug resistance and activate anti-tumor immunity [23,24]. As illustrated in Fig. 1, the multifunctional nanoplatform (Reg & L-Arg@H-MnO2-HA, RLHMnH) can accurately target tumor sites through the enhanced tumor retention and cellular internalization after local administration. HA modification improves biocompatibility and enables binding to CD44 receptors on tumor cells. Subsequently, RLHMnH can not only deplete intracellular GSH to prevent ROS clearance but also trigger a Fenton-like reaction between Mn2+ and local H2O2, thereby achieving effective enhancement of tumor CDT. As the HMn framework decomposes, the loaded Reg is controllably released into tumor tissue, blocking tumor angiogenesis by inhibiting the VEGFR2 signaling pathway and reducing hypoxia-induced immune escape, while synergizing with NO to remodel the TME and activate anti-tumor immune responses. To the best of our knowledge, this study is the first to report a HA-modified hollow MnO2 nanoplatform for the co-delivery of Regorafenib and L-Arginine, enabling the synergistic combination of chemodynamic therapy, anti-angiogenesis, and gas therapy for CRC treatment. In summary, this novel multifunctional nanoparticle provides a new perspective for the treatment of colorectal tumors. Schematic Illustration of preparation of Reg & L-Arg@H-MnO2-HA(RLHMnH). Materials and methods Materials Regorafenib (Reg) was purchased from MedChemExpress (USA). L-Arginine (L-Arg), Potassium permanganate (KMnO4), Manganese sulfate (MnSO4), and Hyaluronic acid (HA, MW ≈ 200 kDa) were obtained from Sigma-Aldrich (USA). All other chemicals and solvents were of analytical grade and used without further purification. The mouse colon carcinoma cell line (CT26) and the mouse embryonic fibroblast cell line (NIH-3T3) were acquired from the American Type Culture Collection (ATCC). BALB/c mice (female, 6–8 weeks old) were supplied by Vital River Laboratory (Beijing, China). All animal experiments were approved by the Animal Ethics Committee of Chongqing Dazu District People's Hospital and conducted in accordance with institutional guidelines. Synthesis of SiO2 nanoparticles SiO2 nanoparticles were synthesized via the Stöber method with slight modifications [25]. Briefly, a reaction mixture was prepared by vigorously stirring a solution containing 25 mL of anhydrous ethanol, 0.5 mL of ultrapure water, and 1.6 mL of ammonia solution in a constant temperature oil bath at 45 °C for 30 min. Subsequently, 0.75 mL of tetraethyl orthosilicate (TEOS) was rapidly introduced into the mixture under stirring at 1000 rpm for 1 min. The stirring speed was then reduced to 500 rpm, and the reaction was allowed to proceed overnight. The resulting SiO2 nanospheres were collected and washed twice each with water and ethanol. Synthesis of SiO2@MnO2 core-shell nanoparticles The as-synthesized SiO2 nanoparticles were used as a template to fabricate the core-shell structure. Specifically, the SiO2 nanoparticles were ultrasonically dispersed in 10 mL of pure water. Then, 40 mL of an aqueous solution containing 250 mg of potassium permanganate (KMnO4) was added dropwise to the dispersion under continuous ultrasonication. The ultrasonication was maintained for 6 h at an initial room temperature, ensuring the temperature did not exceed 50 °C throughout the process. The resulting deep brown precipitate, indicative of MnO2 coating, was collected by centrifugation at 14,800 rpm for 30 min and washed three times with water to obtain the SiO2@MnO2 core-shell nanoparticles. Synthesis of H-MnO2 (HMn) nanoparticles The HMnO2 were obtained by etching the SiO2 core from the SiO2@MnO2 nanoparticles. Briefly, the SiO2@MnO2 nanoparticles were centrifuged, and the pellet was ultrasonically redispersed in 5 mL of a 2 M sodium carbonate (Na2CO3) solution (Solution A). Solution A was then poured into 20 mL of a preheated 2 M Na2CO3 solution at 60 °C and stirred overnight at the same temperature. The following day, the product was collected by centrifugation at 14,800 rpm and washed three times with water to remove sodium silicate, yielding the final HMnO2 nanospheres. Drug loading of regorafenib (Reg@H-MnO2, RHMn) Regorafenib (Reg) was loaded onto the HMnO2 nanocarrier via a simple adsorption method. Specifically, 125 mL of an HMnO2 dispersion (2 mg/mL) was mixed with 25 mL of a Rg solution (5 mg/mL, dissolved in DMSO) and stirred in the dark for 24 h. The mixture was then centrifuged at 12,000 rpm for 15 min to remove unloaded drug molecules. The resulting pellet, denoted as RHMn, was collected and freeze-dried for subsequent use. L-Arginine modification (Reg&L-Arg@H-MnO2, RLHMn) To load L-Arginine (L-Arg), RHMn nanoparticles (250 mg) and L-Arg (2.5 g) were co-dispersed in 100 mL of deionized water and stirred for 12 h. The L-Arg-loaded nanoparticles, designated as RLHMn, were harvested by centrifugation at 12,000 rpm for 25 min. Determination of drug loading capacity and encapsulation efficiency The unencapsulated Regorafenib (Reg) and L-Arginine (L-Arg) in the supernatant were separated from the drug-loaded nanoparticles via centrifugation (12,000 rpm, 15 min) using ultrafiltration tubes. The amounts of free drugs were determined using specific spectroscopic methods. Specifically, to ensure the complete dissolution of the hydrophobic Regorafenib, the supernatant was first mixed with an equal volume of organic solvent prior to measurement. The concentration of Regorafenib in this clarified solution was then determined using a UV-Vis spectrophotometer at λmax of 265 nm, based on a pre-established standard curve prepared in the corresponding solvent mixture. To further verify the accuracy of the loading data, the drug loading was additionally cross-validated by dissolving the lyophilized drug-loaded nanoparticles in DMSO/HCl to fully release the encapsulated Regorafenib, and the total drug content was quantified using the same UV-Vis method. Because L-Arginine lacks strong natural UV absorption, its concentration was quantified using the classic ninhydrin colorimetric assay. Briefly, the supernatant was mixed with a ninhydrin reagent solution and heated at 95 °C for 15 min. The reaction yielded a blue-purple complex, which was then measured by UV-Vis spectrophotometry at a λmax​ of 570 nm against a standard curve of L-Arg. To eliminate potential optical interference from the nanoparticle components, the supernatant of blank nanoparticles (without drug loading) subjected to the same procedures was used as the blank control for background subtraction. The encapsulation efficiency (EE) and loading capacity (LC) were calculated according to the following formulas: where Mtotal is the initial mass of the drug added, Mfree is the mass of the unencapsulated drug in the supernatant, and Mnanoparticles is the total mass of the recovered drug-loaded nanoparticles. Hyaluronic acid coating (Reg & L-Arg@H-MnO2-HA, RLHMnH) The coating with HA was achieved via a carbodiimide-catalyzed coupling reaction. First, a reaction solution (Solution A) was prepared by dissolving 240 mg of EDC and 240 mg of NHS in 90 mL of ultrapure water, followed by the addition of 60 mg of HA. Solution A was stirred at room temperature for 3 h to activate the carboxyl groups. For the synthesis of RHMnH, 250 mg of RHMn nanoparticles were ultrasonically dispersed in 30 mL of deionized water and then gradually added dropwise to Solution A. The reaction was allowed to proceed for 12 h under light-protected stirring. The final product was collected by washing twice with deionized water and freeze-drying. The synthesis of RHMnH followed an identical procedure, except that RLHMn nanoparticles (250 mg) were used as the starting material instead of RHMn. Characterization of nanoparticles The morphology and structure of H-MnO2 and RLHMnH were characterized using transmission electron microscopy (TEM, JEOL JEM-2100) and scanning electron microscopy (SEM, Hitachi SU8010). Elemental mapping was performed using energy-dispersive X-ray spectroscopy (EDS) attached to the SEM. Hydrodynamic size distribution and zeta potential were measured by dynamic light scattering (DLS) using a Malvern Zetasizer Nano ZS. Chemical composition and surface modification were analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Scientific Escalab 250Xi) and Fourier transform infrared spectroscopy (FTIR, Nicolet iS50). All measurements were performed in triplicate at 25 °C. To assess colloidal stability, the RLHMnH nanoparticles were dispersed in deionized water, PBS, and DMEM containing 10% fetal bovine serum (FBS). Their hydrodynamic diameters were continuously monitored via DLS over a 7-day incubation period at room temperature. Drug release profile The pH-responsive release behavior of Reg from RLHMnH was investigated using dialysis bags (MWCO: 3.5 kDa). RLHMnH dispersions were placed in PBS at different pH values (7.4 and 6.5) and shaken at 37 °C. At predetermined time intervals, samples were withdrawn and replaced with fresh buffer. The amount of released Reg was quantified by HPLC. In vitro cellular uptake and targeting CT26 and NIH-3T3 cells were used to evaluate the cellular uptake and CD44-mediated targeting ability of RLHMnH. Cy5-labeled RLHMnH was incubated with cells for 2 h and 4 h. Cells were then washed, fixed, and observed under a confocal laser scanning microscope (CLSM, Zeiss LSM 880). For lysosomal escape analysis, LysoTracker Green was used to stain lysosomes, and co-localization with Cy5 signals was analyzed. In vitro cytotoxicity and antitumor efficacy The cytotoxicity of various formulations (HMn, RHMn, RHMnH, RLHMnH) against CT26 and NIH-3T3 cells was assessed using the Cell Counting Kit-8 (CCK-8) assay. Cells were seeded in 96-well plates and treated with different nanoparticles for 24, 48, and 72 h. Absorbance was measured at 450 nm using a microplate reader. Intracellular ROS levels were detected using the DCFH-DA probe, and NO production was measured using the DAX-J2 Orange probe. GSH levels were quantified using a commercial GSH assay kit. Western blotting was performed to analyze the expression of VEGFR2, Cleaved Caspase-3, Notch1, p-AKT, and p-ERK. To evaluate the induction of ICD, CT26 cells were seeded in multi-well plates and subjected to treatments. For Calreticulin (CRT) exposure analysis, treated cells were fixed without permeabilization, followed by incubation with a primary anti-CRT antibody and a fluorescently labeled secondary antibody. The membrane-localized CRT signals were captured using a fluorescence microscope. To quantify ATP release, the culture supernatants from the treated cells were collected, and the extracellular ATP concentration was measured utilizing an ATP Bioluminescence Assay Kit according to the manufacturer's instructions. In vivo antitumor study BALB/c mice were subcutaneously inoculated with CT26 cells (1 × 106 cells per mouse) to establish a colorectal cancer xenograft model. When tumor volumes reached ∼100 mm3, mice were randomly divided into five groups (n = 5): 1. Control (PBS) 2. HMn 3. RHMn 4. RHMnH 5. RLHMnH. Mice received intratumoral injections every other day for 14 days. Tumor volumes and body weights were recorded every two days. After treatment, tumors and major organs were harvested for histological analysis (H&E staining, TUNEL, IHC for Ki67 and CD31). Immune cell infiltration (CD4+, CD8+ T cells, M1/M2 macrophages) was analyzed by flow cytometry. Transcriptomic analysis by RNA-seq Total RNA was extracted from harvested CT26 tumors using TRIzol reagent. RNA quality was assessed for concentration and integrity. Sequencing libraries were constructed from qualified RNA samples following the standard mRNA enrichment, fragmentation, cDNA synthesis, and adapter ligation workflow. The final libraries were quantified, and size distribution was verified before being subjected to paired-end sequencing on a DNBSEQ platform. Raw sequencing reads were processed with quality control to obtain clean reads. These reads were aligned to the mouse reference genome (GRCm39.109) for gene expression quantification. Differential expression analysis between RLHMnH-treated and control groups was performed, with genes meeting the thresholds of |log2FC| ≥ 1 and FDR <0.05 defined as significant. Functional enrichment analysis (GO and KEGG) and Gene Set Enrichment Analysis (GSEA) were conducted on the resulting gene sets to identify altered biological pathways. Biosafety evaluation Serum biochemical parameters (AST, ALT, BUN, CREA, CK, etc.) were measured using commercial assay kits. Major organs (heart, liver, spleen, lung, kidney) were sectioned and stained with H&E for histopathological examination. Statistical analysis All data are presented as mean ± standard deviation (SD). Statistical significance was determined by one-way ANOVA followed by Tukey's post-hoc test using GraphPad Prism 9.0. A p-value <0.05 was considered statistically significant. Results and discussion Characterization of HMn and RLHMnH nanoparticles The nanostructures of the synthesized hollow manganese dioxide (H-MnO2) and its functionalized complex (Reg & L-Arg@H-MnO2-HA, RLHMnH) were systematically characterized by transmission electron microscopy (TEM) and scanning electron microscopy (SEM). As shown in Fig. 2A, the pristine H-MnO2 exhibited a regular spherical hollow structure with uniform diameter (120 ± 15 nm) and a shell thickness of approximately 20 nm. The distinct contrast between the dark edges and the bright center in the TEM images confirmed the successful construction of the hollow cavity. The modified RLHMnH nanoparticles maintained spherical integrity while displaying a clear core-shell structure, with an amorphous coating layer visible on the surface, formed by the sequential deposition of the L-Arg complex and HA polymer. SEM analysis (Fig. 2B) further indicated that RLHMnH was uniformly dispersed with good monodispersity, and the particle size slightly increased (150 ± 20 nm) due to surface functionalization. Energy-dispersive X-ray spectroscopy (EDS) elemental mapping (Fig. 2C) showed a strong Mn signal (green) uniformly distributed throughout the nanostructure, confirming the integrity of the H-MnO2 framework. The detected N element (blue) was associated with the presence of Regorafenib (C21H15ClF4N4O3) and L-Arginine (C6H14N4O2), indicating successful drug loading and surface modification. The spatial co-localization of Mn (from H-MnO2), C/N (from Reg/L-Arg), and S (from HA) confirmed the hierarchical integration of the therapeutic components within the nanoparticle. The drug loading performance of the RLHMnH nanoplatform was systematically quantified. Benefiting from the hollow mesoporous structure and hydrophobic interactions, the loading capacity (LC) and encapsulation efficiency (EE) for Regorafenib were determined to be 20.8% and 52.8%, respectively. Furthermore, by employing a concentration-gradient adsorption strategy, a satisfactory L-Arginine loading capacity of 11.5% was achieved (with an EE of 32.6%), ensuring sufficient substrate availability for subsequent nitric oxide generation. Characterization of HMn and RLHMnH. (A) TEM of H-MnO2 and RLHMnH (scale bar: 200 nm); (B) SEM of HMn and RLHMnH (scale bar: 200 nm). (C) EDS elemental mapping of RLHMnH (Mn: green, C: red, N: blue, S: yellow, scale bar: 50 nm). (D) Particle size analysis of HMn and RLHMnH. (E) Zeta potential analysis of HMn and RLHMnH. (F) XPS analysis of RLHMnH. (G) FTIR analysis of RLHMnH. (H) Cumulative Reg release from RLHMnH. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) The hydrodynamic diameter and surface charge characteristics of HMn and its functionalized nanoparticle RLHMnH were systematically evaluated by dynamic light scattering (DLS). As shown in Fig. 2D, the pristine HMn nanoparticles exhibited a monomodal distribution with an average hydrodynamic diameter of 164.3 ± 3.7 nm (PDI = 0.18). This value was slightly larger than the dry-state size observed by TEM (120 nm), attributable to the solvation layer effect and the sensitivity of DLS to the hydrated state. After drug loading and surface modification, the size of RLHMnH increased significantly to 190.5 ± 4.2 nm (PDI = 0.21). Surface charge analysis (Fig. 2E) further corroborated the modification process: HMn exhibited strong negative electronegativity (Zeta potential = -29.1 ± 0.9 mV) due to surface hydroxyl groups. The negative value of the RLHMnH potential was weakened (-26.8 ± 0.7 mV), originating from the carboxyl groups (-COO-) of HA partially neutralizing the surface charge of HMn, while the amino groups (-NH3+) of L-Arg contributed positive charge. The full XPS spectrum of the RLHMnH nanocomplex (Fig. 2F) clearly revealed its surface elemental composition and chemical modification characteristics. The coexistence of Mn 2p and O 1s indicated that the MnO2 carrier maintained chemical stability after modification, and its metal oxide properties provided active sites for CDT. The high-intensity C 1s signal (accounting for 62.3%) originated from the sugar ring backbone of HA and the carbon chain structure of L-Arg, indicating the formation of an effective organic layer coating the nanoparticle surface. The appearance of the N 1s peak directly verified the successful modification of L-Arginine (L-Arg). To further verify the chemical structure of HMn and the successful surface modification of the functionalized nanocarrier RLHMnH, characterization was performed using FTIR spectroscopy (Fig. 2G). For HMn nanoparticles, a strong characteristic absorption peak appeared near 580 cm-1, attributed to the stretching vibration of the Mn–O bond, confirming the successful formation of the MnO2 crystal lattice structure. This peak position was consistent with the standard spectrum of manganese dioxide, indicating that the hollow MnO2 carrier possessed typical metal oxide characteristics. The RLHMnH nanocomplex after modification with HA and L-Arg exhibited significant spectral changes. The FTIR spectrum of RLHMnH exhibited new absorption bands attributable to HA and L-Arg, particularly the characteristic peak at 1650 cm-1 for the amide bond, which confirmed the successful surface modification. Concurrently, the appearance of absorption in the 1250-1000 cm-1 region, characteristic of C-F stretching vibrations, verified the successful loading of Reg. Using Terephthalic acid (TA) as a probe, the ability of RLHMnH to generate ·OH was investigated. As the concentration of RLHMnH increased, the fluorescence intensity at 435 nm also increased, indicating a concentration-dependent characteristic in ·OH generation (Fig. S1). Stability experiments for RLHMnH showed that the particle size of RLHMnH did not change significantly over time in either PBS or 10% FBS-DMEM medium, indicating that the prepared nanoparticles possessed good stability (Fig. S2). HMn nanoparticles exhibit significant pH responsiveness; their structure specifically degrades under the weakly acidic conditions of the tumor microenvironment (pH ≈ 6.5), releasing the loaded drug and producing Mn2+ and oxygen (O2). This responsive mechanism stems from the dissolution of manganese dioxide in an acidic environment upon reaction with H+, while concurrently consuming the overexpressed H2O2 at the tumor site, thereby ameliorating tumor hypoxia and achieving targeted controlled release of the drug. This study aimed to determine whether RLHMnH could exhibit pH-responsive release of Reg. The pH of tumor tissue (5.5−6.0) is typically lower than the pH of normal tissue (7.4). As shown in Fig. 2H, compared to the total release of approximately 16.1% of Reg at pH 7.4, RLHMnH released up to 55.6% of Reg at pH 6.5. Therefore, the anticancer drug Reg loaded in RLHMnH can exhibit pH-responsive release, particularly in the tumor tissue regions of colorectal cancer. In vitro cellular uptake and tumor cell targeting While the enhanced permeability and retention (EPR) effect in solid tumors facilitates passive targeting, it often limits the efficient internalization of nanoparticles by tumor cells [26]. To address this, we constructed an active targeting strategy reliant on a receptor-ligand mechanism by modifying RLHMn with hyaluronic acid (HA, MW ≈ 200 kDa). CD44 is a complex transmembrane adhesion glycoprotein expressed in various human cells, including embryonic stem cells, differentiated cells, and cancer cells. It is frequently overexpressed in tumors and, as a recognized cancer stem cell marker, serves as a key regulator of epithelial-mesenchymal transition (EMT), participating in tumorigenesis, progression, and metastasis. First, we measured the mRNA expression levels of endogenous CD44 in CT26 cells and the mouse embryonic fibroblast cell line (NIH-3T3) (Fig. S3). As expected, CD44 expression levels were significantly higher in CT26 cells than in NIH-3T3 cells, indicating the potential of CD44 as a target molecule. The tumor cell targeting capability of the nanoparticles was investigated using Cy5-labeled RLHMnH. Accordingly, NIH-3T3 cells were used to verify whether the enhanced uptake of RLHMnH was specific to tumor cells rather than normal cells. As shown in Fig. 3A, CT26 cells treated with Cy5-labeled RLHMnH displayed intense red fluorescent signals in the cytoplasm. In contrast, NIH-3T3 cells treated with RLHMnH showed nearly undetectable Cy5 fluorescence. Quantitative analysis using ImageJ software confirmed that the intracellular mean fluorescence intensity of the targeted group was significantly higher than that of the non-targeted group, as shown in Fig. 3B. These phenomena indicate that RLHMnH can significantly increase the uptake of RLHMn by CT26 cells through HA-CD44 targeted binding. Furthermore, the enhanced uptake was specific to tumor cells and not observed in normal cells. To validate the lysosomal escape capability of RLHMnH, lysosomal co-localization was performed by incubating CT26 cells with RLHMnH for different durations (Fig. 3C). At 2 h, the red fluorescence of RLHMnH largely overlapped with the green fluorescence of lysosomes (stained with LysoTracker Green), indicating that RLHMnH was internalized by CT26 cells and co-localized with lysosomes. The overlapping area gradually decreased over time and was significantly separated at 4 h, demonstrating the lysosomal escape ability of RLHMnH. This dynamic escape process was further corroborated by quantitative co-localization analysis (Fig. 3D). (A) Cellular uptake study of Cy5-labeled RLHMnH in CT26 cells and NIH-3T3 cells (scale bar: 20 μm). (B) Quantitative analysis of the cellular uptake. (C) Representative co-localization images of lysosome in CT26 cells after the incubation with RLHMnH and the staining of lyso-tracker Green probe (scale bar: 10 μm). (D) Quantitative co-localization analysis of the RLHMnH nanoparticles (red) and lysosomes (green) at different incubation time intervals (1, 2, 3, and 4 h) using the Pearson's correlation coefficient (PCC). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) In vitro antitumor efficacy and mechanistic studies To investigate the therapeutic effect of RLHMnH on colorectal cancer, we performed in vitro cell assays. In the in vitro experiments, a total of five subgroups were set up: 1. Control, 2. H-MnO2 (HMn), 3. Reg@H-MnO2 (RHMn), 4. Reg@H-MnO2-HA (RHMnH), 5. Reg & L-Arg@H-MnO2-HA (RLHMnH). To verify the synergistic therapeutic effect of the various components within the nanomaterial against colon cancer, we co-cultured CT26 cells and NIH-3T3 cells with the different nanoparticle formulations for 24, 48, and 72 h. At all three time points, the cell viability of CT26 cells in the RLHMnH group was significantly lower than that in the other precursor material groups (Fig. 4A). Due to the HA-mediated CD44 targeting, the cytotoxicity of the modified nanoplatform was significantly enhanced. The cell viability of CT26 cells treated with the targeted formulation dropped to approximately 20% at 72 h, which was notably lower than the 52% viability observed in the RHMn treatment group. Concurrently, the low expression of the CD44 receptor rendered HA targeting ineffective in NIH-3T3 cells, and none of the treatment groups showed significant toxicity towards NIH-3T3 cells (viability >80% at 72h) (Fig. 4B). These results preliminarily indicate that the effective combination of components in RLHMnH can exert a combined antitumor effect, maximizing tumor cell killing and inhibiting cell proliferation. Next, we sought to investigate the impact of RLHMnH on CT26 cell proliferation. Compared to the control group, RLHMnH significantly inhibited the cell proliferation rate. Accordingly, the inhibitory effects on cell proliferation in the RHMnH and RHMn groups were higher than that in the HMn group but weaker than that in the RLHMnH group (Fig. S4). As shown in Fig. 4C, the GSH levels in the RHMnH and RLHMnH groups were significantly lower compared to the other groups. This profound depletion can be attributed to a synergistic mechanism: the MnO2 cavity structure efficiently consumes GSH, and this process is further amplified by the NO generated from L-Arg, which synergistically intensifies oxidative stress. The collective action of these multiple mechanisms leads to near-complete GSH exhaustion, which effectively hinders ROS clearance and thereby enhances the antitumor effect. Subsequently, flow cytometry analysis was performed to evaluate the apoptotic status of the treated cells. At the 24 h time point, the RLHMnH-treated group showed a higher apoptotic cell ratio compared with the control group and the precursor material groups (Fig. 4D and E). (A) Cell viability of CT26 cells and (B) NIH-3T3 cells measured by MTT assay after treatment with various formulations. (C) Detection of GSH depletion in CT26 cells treated with different samples. (D) Flow cytometric analysis was utilized to assess the apoptotic status of CT26 cells at 24 h post-treatment, employing the One-step TUNEL Flow Cytometry Apoptosis Kit for cell staining. (E) Statistical analysis of flow cytometry data for the apoptotic status of CT26 cells following the corresponding treatments. (F) Intracellular ROS levels in CT26 cells detected using the DCFH-DA probe after treatment with RHMnH; scale bar: 20 μm. (G) Intracellular NO generation evaluated by the DAF-FM DA fluorescent probe in CT26 cells under different treatments; scale bar: 20 μm. (H) Western blot analysis of VEGFR2, cleaved caspase-3, NICD, Hes1, and Hey1 expression in CT26 cells across the five experimental groups. (I) Western blot analysis of p-AKT and p-ERK expression in CT26 cells across the five experimental groups. **P < 0.01. In the CT26 cell model, intracellular ROS levels showed significant differences among the different treatment groups, reflecting the functional contributions of the various components in the nano-system (Fig. 4F). The control group displayed only basal ROS levels. The HMn treatment group induced a moderate increase in ROS due to partial GSH depletion and the triggering of Mn2+-mediated Fenton-like reactions. The RHMn group further elevated ROS levels by inhibiting antioxidant pathways and exacerbating intracellular stress [27,28]. Notably, the RHMnH group exhibited significantly higher ROS generation, benefiting from the HA-mediated CD44 targeting which enhanced cellular endocytosis efficiency, thereby amplifying Mn2+-driven ·OH production. The RLHMnH group presented the most substantial ROS accumulation effect. This mechanism stems from the continuous release of NO from L-Arg, which powerfully depletes GSH, and maximizes oxidative damage effects. L-Arg can facilitate the generation of NO gas. Meanwhile, intracellular ROS production may also promote NO release from L-Arg. DAF-FM is a fluorescent probe that specifically binds to NO to produce fluorescence, enabling the detection of intracellular NO generation. As shown in Fig. 4G after modification with L-Arg, the RLHMnH group exhibited significant red fluorescence, indicating the presence of NO production, with a fluorescence intensity approximately 8.9 times higher than that of the control group. These results demonstrate the critical role of the combinatorial design, which jointly regulates ROS generation and antioxidant defense mechanisms, in enhancing chemodynamic efficacy. Given the intracellular oxidative stress and damage caused by the synergistic generation of ROS and NO, we hypothesized that the targeted nanoplatform could efficiently trigger immunogenic cell death (ICD), serving as an endogenous vaccine to prime subsequent immune responses. To verify this, key ICD biomarkers, namely the surface exposure of calreticulin (CRT) and the release of ATP, were evaluated in vitro. As depicted in Fig. S5A, prominent green fluorescence corresponding to membrane-bound CRT was distinctly observed in the targeted treatment group. By contrast, negligible CRT exposure was detected in the PBS control. Furthermore, the quantitative analysis of the culture medium (Fig. S5B) confirmed a massive surge in extracellular ATP secretion following the combination treatment. The emission of these critical damage-associated molecular patterns (DAMPs) confirms the potent ICD-inducing capability of the nanoplatform, which underpins the extensive in vivo immune cell activation and T-cell infiltration observed in the tumor-bearing models. Western Blot analysis (Fig. 4H) revealed the mechanism by which RLHMnH inhibits tumor cell growth through ROS/NO synergy. Compared to the Control group, the HMn group only slightly downregulated VEGFR2 expression, indicating that CDT alone has a limited effect on angiogenesis regulation. In contrast, the RHMn and RHMnH groups significantly reduced VEGFR2 levels, confirming that the targeted delivery of Reg effectively blocks VEGFR2 signaling. Notably, the RLHMnH group exhibited the strongest suppression of VEGFR2, originating from L-Arg-derived NO dilating tumor blood vessels and alleviating hypoxia, thereby attenuating the HIF-1α-driven feedback upregulation of VEGFR2. In the apoptosis pathway, RLHMnH induced a significantly higher expression of Cleaved Caspase-3 compared to the RHMnH and HMn groups, a mechanism closely related to the ROS/NO synergy-triggered mitochondrial apoptosis pathway. Another interesting finding was the profound inhibition of the Notch signaling axis: RLHMnH treatment significantly decreased the protein expression of the Notch1 intracellular domain (NICD) and its downstream target genes Hes1 and Hey1 compared to the control group, with an inhibitory intensity far exceeding that of other groups. The above results prove that RLHMnH not only directly activates the apoptosis execution protein Cleaved Caspase-3 but also significantly impacts the Notch signaling axis through the synergy between its components. Further analysis showed that high basal levels of p-AKT and p-ERK1/2 (Thr202/Tyr204) were detected in the Control group, reflecting the inherent activation state of proliferative signaling in CT26 cells(Fig. 4I). The HMn group only slightly downregulated p-AKT and did not significantly affect p-ERK, indicating limited intervention in these two pathways by CDT alone. Notably, the inhibitory effect was further enhanced in the HA-modified RHMnH group compared to the non-targeted RHMn group, attributable to CD44-mediated targeted internalization increasing the intracellular concentration of Reg. The RLHMnH group exhibited the most significant pathway inhibition, effectively blocking the pro-survival signaling axis through multi-target synergy. In summary, through the synergistic action of its components, RLHMnH targets and blocks the Notch and P-AKT signaling axes, and ultimately achieves a significant enhancement of Caspase-dependent apoptosis. In vivo antitumor evaluation In previous experiments, we confirmed the significant targeted killing effect of RLHMnH on CT26 cells in vitro. In this section, based on our in vitro results, we further established a CT26 cell xenograft model in BALB/c mice. Tumor-bearing BALB/c mice were randomly divided into five groups receiving different treatments, namely the control, HMn, RHMn, RHMnH, and RLHMnH groups. Mice in the four nanoparticle treatment groups were intratumorally administered with their respective formulations at a consistent dosage of 15 mg/kg. The control group received intravenous injections of an equal volume of PBS. After 14 days of treatment with the respective formulations, tumor tissues were collected from each group of mice (Fig. 5A). The body weights of the mice across the five groups showed no statistically significant difference after the 14-day treatment period (Fig. 5B). Furthermore, the different nanoparticle formulations did not affect the survival of the mice in any group. Further analysis of tumor volume and weight revealed that the RLHMnH group had the lowest average tumor volume and tumor weight among the five groups (Fig. 5C and D), indicating that the synergy between Reg/L-Arg and HA in RLHMnH enhanced the therapeutic efficacy of the nanoparticles and provided superior tumor suppression. To assess the In vivo tumor retention and biodistribution following intratumoral injection, mice were treated with Cy5-labeled RLHMnH. In vivo imaging over 72 h demonstrated persistent fluorescence strictly confined to the tumor region (Fig. S6A). Subsequent ex vivo imaging and quantitative analysis (Fig. S6B) confirmed this prolonged tumor retention, with negligible signal leakage to major organs. Such stable localized accumulation, enhanced by HA-CD44 mediated cellular uptake, ensures robust local therapeutic efficacy while minimizing systemic toxicity.” (A) Tumors were collected from five different groups 14 days after different treatments. The images of the tumors are displayed. (B) Body weight of CT26 tumor-bearing BALB/c mice in each group over 14 days after treatment. (C) Tumor weights of different groups in CT26 tumor bearing BALB/c mice. (D) Tumor volumes of different groups in CT26 tumor bearing BALB/c mice. (E) Representative hematoxylin-eosin staining in CRC tissue in five groups. (F-G) Representative Ki67 and CD31 immunohistochemical staining were observed in the 5 groups of tumor tissues. **P < 0.01. To investigate the mechanism of tumor apoptosis, we performed histological analysis via H&E staining. As shown in Fig. 5E, compared to the control group, tumors in the RHMn, RHMnH, and RLHMnH groups all showed significant nuclear condensation and fragmentation, with the most pronounced changes observed in the RLHMnH group. The HMn group also showed slight damage, but it was less evident than in the other groups. To further validate the amplification effect of our system in vivo, the actual NO levels within the excised tumor tissues were quantified using a Griess assay kit (Fig. S7). Consistent with in vitro findings, tumor tissues treated with the intact RLHMnH nanoplatform exhibited significantly higher NO concentrations than the control group. This significant enhancement provides direct evidence that our combination design can effectively amplify NO generation within the complex tumor microenvironment. To evaluate the proliferative state of tumor cells in the tissues, we performed immunohistochemistry (IHC) to detect the expression level of Ki67 in the tumor tissues of each group (Fig. 5F). The results showed the highest Ki67 expression in the control group. Expression was reduced to varying degrees in the precursor nanoparticle groups, with the least Ki67 expression found in the RLHMnH group, indicating that the synergistic effect of the components in RLHMnH significantly inhibited tumor cell proliferation. CD31 (also known as PECAM-1) is a specific surface marker of vascular endothelial cells. Detecting CD31 expression clearly reveals the distribution and morphology of vascular endothelial cells in tissue sections. To verify the inhibitory effect of REG within the nanoparticles on tumor tissue angiogenesis, we further examined CD31 expression in the tumor tissues(Fig. 5G). The results showed that in the HMn group, CD31 expression showed almost no significant change compared to the control group. In contrast, the addition of Reg significantly reduced CD31 expression. Furthermore, due to HA modification, the RLHMnH group exhibited better targeting towards tumor tissue, resulting in a significant decrease in CD31 expression in this treatment group. RNA-seq analysis reveals transcriptomic alterations induced by RLHMnH treatment To further elucidate the molecular mechanisms underlying the antitumor efficacy of RLHMnH, bulk RNA sequencing was performed on CT26 tumor tissues after RLHMnH treatment. Comparative analysis identified a total of 877 differentially expressed genes (DEGs) between the RLHMnH-treated and control groups (FDR <0.05, |log2 fold change| ≥ 1), comprising 303 upregulated and 574 downregulated genes (Fig. 6A). KEGG pathway enrichment analysis revealed significant alterations in multiple signaling cascades crucial for tumor progression and immune modulation (Fig. 6B). Notably, pathways associated with immune response and inflammation—such as the NOD-like receptor signaling pathway, cytokine-cytokine receptor interaction, and IL-17 signaling pathway—were prominently enriched, consistent with the observed immune cell infiltration within the tumor microenvironment. Moreover, key pathways involved in cell proliferation, survival, and angiogenesis were significantly modulated, corroborating the anti-angiogenic and pro-apoptotic effects mediated by regorafenib and L-Arg-derived nitric oxide (NO). Gene set enrichment analysis (GSEA) further demonstrated that RLHMnH treatment significantly activated gene sets related to anti-tumor immune pathways (Fig. 6C), aligning with the ROS/NO synergy and GSH depletion induced by RLHMnH. Cluster analysis of the DEGs also showed a clear transcriptional distinction between the treated and control groups (Fig. 6D). Collectively, these transcriptomic insights underscore the multimodal mechanistic synergy of RLHMnH in reshaping the tumor microenvironment and suppressing colorectal cancer progression. Transcriptomic profiling of CT26 tumors after RLHMnH treatment. (A) Volcano plot of differentially expressed genes (DEGs) between treated and control groups (|log2FC| ≥ 1, FDR <0.05). Red: upregulated genes; blue: downregulated genes. (B) KEGG pathway enrichment analysis of DEGs. Bars indicate the number of DEGs mapped to each significantly enriched pathway. (C) Enrichment score of the indicated term in the RLHMnH group versus control group, analyzed by GSEA based on the RNA-seq data. (D) Heatmap of clustered DEGs across samples. Red: high expression; blue: low expression. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Differences in TUNEL staining intensity directly reflect the ability of the different nanomaterials to induce tumor cell apoptosis. The TUNEL results were consistent with the H&E findings (Fig. 7A), with the most prominent green fluorescence observed in the RLHMnH group. The HMn, RHMn, and RHMnH groups also showed green fluorescence, indicating apoptosis, but to a lesser extent. The results demonstrate that apoptosis occurred in the mouse tumor tissues, with the RLHMnH group showing the most significant antitumor effect. (A) TUNEL, and (B) CD8 staining images of tumor sections of CT26 tumor-bearing BALB/c mice in different treatment groups were collected. Scale bar: 50 μm. Representative flow cytometry plots showing CD4+ T cells and CD8+ T cells(C), CD86+ and CD206+ macrophages (F) populations in tumors. Statistical analysis of flow cytometry data for CD4+ T cells(D), CD8+ T cells(E), M1 macrophages (G) andM2 macrophages (H). (n = 5); **p < 0.01. The TME, including a dense stromal barrier rich in hyaluronic acid and heterogeneous distribution of blood vessel growth, severely hinders intratumoral drug accumulation, reduces the infiltration of immune cells (such as T cells and M1 macrophages), and leads to immune tolerance in tumor cells. A hypoxic TME promotes resistance to therapy in rectal cancer cells. For these reasons, it is necessary to modulate the cancer TME by disrupting the stromal barrier to promote drug penetration and regulating the immune environment to improve the efficacy of comprehensive tumor therapy. Immunofluorescence experiments confirmed that RLHMnH increased the infiltration of CD8+ T cells and M1 macrophages into the tumors (Fig. 7B, Fig. S8). This finding prompted us to further investigate whether RLHMnH affects the function of immune cells within the TME. Flow cytometry analysis showed that the enrichment ratios of CD4+ T cells (Fig. 7C and D) and CD8+ T cells (Fig. 7C–E) were significantly increased in the tumors of the RLHMnH group. Accordingly, the RLHMnH group also promoted the infiltration of M1 macrophages within the TME. We found that the RLHMnH group increased the proportion of M1 macrophages more significantly than in the other groups (Fig. 7F and G). We further analyzed the proportion of M2 tumor-associated macrophages (Fig. 7F and H) and found that while RLHMnH increased the M1 ratio, the proportion of M2 macrophages correspondingly decreased, reflecting better immune activation. This shift in macrophage polarization underscores the multi-modal synergy of the RLHMnH platform. The released Regorafenib inhibits angiogenesis via VEGFR2 blockade, alleviating hypoxia and facilitating immune cell infiltration. Concurrently, NO generated from L-Arg promotes M1 polarization and synergizes with ROS from the Mn2+-mediated Fenton-like reaction to enhance immunogenic cell death. The hollow MnO2 nanostructure not provides a high-capacity for co-loading Reg and L-Arg but also depletes GSH and catalyzes ·OH generation, amplifying oxidative stress and TME remodeling. Thus, the combination of anti-angiogenesis, NO therapy, and CDT within a single targeted nanoparticle effectively reverses immunosuppression and augments anti-tumor immunity. Biosafety assessment Evaluating the safety of nanoparticles for biomedical applications is crucial. We used H&E staining to assess the potential organ toxicity and deposition of the nanoparticles in major organs. Mice were sacrificed 14 days after nanoparticle injection, and heart, liver, spleen, lung, and kidney tissues were collected for histological observation and analysis. As shown in Fig. 8, histopathological analysis of the major organs indicated that RLHMnH did not cause significant pathological changes compared to the other treatment groups, demonstrating good biocompatibility. In the histopathological assessment, none of the four treatment groups induced pathological alterations compared to the control group. Furthermore, clinically relevant serum biochemical parameters, such as CK, CK-MB, LDH1 for cardiac function, BUN and CREA for renal function, and ALT and AST for liver function, were analyzed in mice from the RLHMnH group and the control group. The results showed that all indicators were within normal ranges, indicating no significant toxicity was detected in the RLHMnH group mice (Table S1). Notably, the excellent biosafety profile of RLHMnH is fundamentally underpinned by the biocompatibility of its components. Hyaluronic acid (HA), Regorafenib, and L-Arginine are all U.S. FDA-approved substances, which have been extensively used in various pharmaceutical and clinical applications. Meanwhile, manganese dioxide (MnO2)-based nanomaterials have also been widely investigated and demonstrated to be biodegradable and well-tolerated in vivo. The combination of these safe and reliable components into a single nanoplatform not only ensures minimal systemic toxicity but also greatly enhances the potential for clinical translation and future broad application of this strategy. Therefore, our findings suggest its potential for in colorectal cancer tumor therapy. Ultimately, RLHMnH can not only improve local tumor control but also reduce the dosage of therapeutic drugs without compromising the tumor suppression effect, leading to lower toxicity and better protection of the immune system, thereby increasing the potential for anti-tumor immunity. H&E staining of organs in different treatment groups. Scale:50 μm. Conclusions In summary, this study is the first to successfully integrate regorafenib-induced antiangiogenesis, MnO2-based CDT and L-Arg-derived NO gas therapy into a single, multifunctional nanoplatform (RLHMnH) for the synergistic treatment of colorectal cancer. By leveraging the hollow structure of manganese dioxide for high drug loading and the pH-responsive decomposition behavior within the TME, this system achieves spatiotemporally controlled release of therapeutic agents, thereby maximizing therapeutic efficacy while minimizing off-target effects. The triple-combination strategy effectively overcomes several key limitations of conventional monotherapies: (1) the GSH-scavenging effect and self-supplying Mn2+ enhance ·OH generation via Fenton-like reactions, significantly augmenting CDT efficacy; (2) the released regorafenib potently inhibits tumor angiogenesis and proliferation through VEGFR2 suppression; and (3) NO gas not only reverses multidrug resistance and alleviates tumor hypoxia, but also synergizes with CDT to promote apoptosis and remodel the immunosuppressive TME, leading to enhanced infiltration of cytotoxic T cells and M1 macrophages. Both in vitro and in vivo experiments demonstrated that RLHMnH exhibits superior tumor growth suppression and a compelling safety profile, with no significant systemic toxicity observed. The active targeting mediated by HA further enhances tumor-specific accumulation and cellular internalization, underscoring the potential of receptor-ligand strategies in improving nanotherapeutic precision. Looking forward, this work provides a robust nanoplatform for future translational studies. Our subsequent efforts will be directed towards validating RLHMnH in orthotopic CRC models, investigating its immune memory effects, and addressing the challenges of scalable production and long-term biosafety of manganese components. Consent for publication All authors agree to be published. Ethics approval and consent to participate Animal experiment is in accordance with the regulations of the Animal Ethical and Welfare Committee of The Affiliated Dazu's Hospital of Chongqing Medical University. Funding This study was supported by the Natural Science Foundation of Chongqing (CSTB2023NSCQ-MSX0502). CRediT authorship contribution statement Chong Li: Conceptualization, Formal analysis, Methodology, Writing – review & editing. Wei Liu: Investigation, Validation, Writing – original draft. Ningfu Li: Formal analysis, Investigation. Yanyan Zhang: Conceptualization, Project administration, Supervision, Writing – review & editing. Ke Wu: Formal analysis, Investigation. Shaohong Yu: Investigation. 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.