Abstract The complex dual role of autophagy provides new insights for enhancing tumor treatment efficacy. However, effectively regulating this process is key to enhancing therapeutic efficacy. To address this challenge, this study designed a gambogic acid-iron nanozyme (GAFe) as a novel carrier to enhance the effectiveness of chemotherapy drugs such as doxorubicin (DOX) by inducing excessive autophagy and oxidative stress. The synthesized nanoparticle (GAFe@DOX) is capable of slowly releasing its active components over a prolonged period within tumor tissues. Gambogic acid can induce excessive autophagy, while the multi-enzyme activity of GAFe and the activation of ferroptosis amplify and sustain excessive autophagy, thereby enhancing the chemotherapy effect of DOX. Meanwhile, ferroptosis activated via the GPX4 pathway by GAFe can synergize with excessive autophagy, amplifying oxidative stress and consequently enhancing the overall therapeutic efficacy. Characterization experiments confirmed the successful synthesis of GAFe@DOX and probe assays demonstrated its superior multi-enzyme activity. In vitro cell studies showed that GAFe@DOX effectively kills tumor cells, while in vivo animal experiments revealed its excellent biocompatibility and significant tumor growth inhibition. This study demonstrates a promising strategy to improve tumor therapeutic efficacy by modulating excessive autophagy and oxidative stress. This provides a novel and effective approach to improve the treatment of refractory tumors such as glioblastoma. Graphical abstract [38]graphic file with name 12951_2025_3519_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03519-3. Keywords: Chemotherapy, Autophagy, Nanozymes, Multi-enzyme activity, Ferroptosis Introduction Autophagy is a process in which cells use autophagosomes and lysosomes to degrade their own organelles and macromolecules under the regulation of autophagy-related gene (ATG) [[39]1–[40]3]. As a major cellular degradation mechanism, autophagy plays a complex dual role in tumorigenesis and development [[41]4]. In general, the autophagy process can remove intracellular metabolic wastes and damaged organelles, and help tumor cells to adapt to the harsh environment such as hypoxia and drug toxicity, thus enhancing their survival ability [[42]5]. However, when autophagy exceeds the cell’s tolerance threshold, excessive autophagy can exacerbate oxidative stress in tumor cells, trigger self-destruction, and ultimately lead to cell death, known as autophagic cell death [[43]6–[44]8]. Therefore, how to induce excessive autophagy in tumors is a major challenge in utilizing autophagy for cancer therapy. In recent years, natural compounds such as gambogic acid (GA) have attracted attention for their unique ability to induce excessive autophagy [[45]9–[46]11]. Studies have shown that this kind of natural compound can effectively regulate the autophagic process in tumor cells, promoting the excessive accumulation of autophagosomes and thereby triggering excessive autophagy [[47]12–[48]14]. This property offers a novel approach to enhance tumor therapy through the regulation of autophagy. In addition to its intrinsic cytotoxic effects on tumor cells, excessive autophagy can effectively enhance chemotherapy in the treatment of refractory tumors, such as glioblastoma (GBM) [[49]15]. For example, the combination of autophagy inducers with chemotherapeutic drugs like doxorubicin (DOX) can enhance the therapeutic effect of DOX through excessive autophagy [[50]16, [51]17]. DOX, a widely used anticancer drug, primarily kills tumor cells by causing DNA damage and inducing apoptosis [[52]18]. However, tumor cells often activate autophagy to cope with the stress induced by the drug, thereby reducing drug efficacy. Combined use of autophagy inducers can excessively activate autophagy, leading to excessive accumulation of damaged organelles and metabolic wastes in tumor cells, exacerbating oxidative stress and endogenous cellular damage, and ultimately promoting the death of tumor cells [[53]19–[54]21]. This mechanism provides a potential synergistic strategy to enhance chemotherapy effectiveness. However, the process of excessive autophagy is highly dependent on high levels of oxidative stress resulting from the accumulation of intracellular reactive oxygen species (ROS) levels [[55]22–[56]24]. Therefore, ensuring an adequate supply of ROS within tumor cells is crucial for achieving excessive autophagy therapy. In recent years, metal nanozymes, especially iron-based nanozymes, have attracted widespread attention due to their ability to catalyze ROS generation [[57]25, [58]26]. Iron-based nanozymes have multiple enzymatic activities similar to peroxidase (POD) and oxidase (OXD), which can generate large amounts of ROS through pathways such as the Fenton reaction, thereby enhancing intracellular oxidative stress. This approach is known as chemodynamic therapy (CDT) [[59]27]. Therefore, iron-based nanozymes can effectively address the problem of insufficient ROS supply during excessive autophagy. Furthermore, it has been shown that autophagy is closely associated with other types of programmed cell death, such as ferroptosis and apoptosis [[60]28]. The accumulation of iron ions released by iron-based nanozymes can activate ferroptosis, autophagy can promote ferroptosis by increasing the release of iron ions and enhancing lipid peroxidation. The ROS generated by ferroptosis can further exacerbate oxidative stress within the cell, amplifying the autophagic response and forming a self-amplifying feedback loop, Oxidative stress caused by this synergistic effect can also enhance apoptosis induced by chemotherapeutic drugs [[61]29–[62]31]. Thus, the combination of iron-based nanozymes with autophagy inducers can sufficiently amplify oxidative stress, induce excessive autophagy and synergize other cell death mechanisms, such as ferroptosis and apoptosis, to enhance the antitumor effect of overall chemotherapy. Based on these insights, in this study, a gambogic acid-iron nanozyme (GAFe) is designed and developed as a novel carrier to load the chemotherapeutic drug DOX, preparing a new type of nanoparticle (GAFe@DOX, GFD) for enhancing glioma chemotherapy (Scheme [63]1). The nanoparticle has several unique advantages: (1) GAFe@DOX exhibits POD-like and OXD-like multi-enzyme activities, generating abundant ROS such as O₂⁻ and ·OH, inducing oxidative stress, and effectively killing tumor cells. (2) The GA in GAFe@DOX can enhance the chemotherapy effect of DOX by inducing excessive autophagy, while GAFe generates high levels of ROS through its multi-enzyme activity and ferroptosis activation, ensuring smooth GA-induced excessive autophagy. (3) Excessive autophagy and ferroptosis promote each other, amplifying oxidative stress and synergizing with other cell death pathways, thereby maximizing the antitumor effect of GAFe@DOX. Characterization experiments confirmed that GAFe@DOX was successfully synthesized with excellent physical properties and slow-release effects. Probe assays confirmed the excellent multienzyme activity of GAFe@DOX for ROS generation and oxidative stress induction. In vitro cell experiments showed that GAFe@DOX has good biocompatibility and effectively kills tumor cells. Transcriptomic sequencing confirmed that GAFe@DOX successfully induced oxidative stress, excessive autophagy, ferroptosis and apoptosis to fully exploit its antitumor potential. In vivo results also verified its good biosafety and significant inhibition of tumor growth. The GAFe nanozyme designed in this study serves as a novel carrier, enhancing chemotherapy by inducing excessive autophagy and oxidative stress, providing a new approach for the use of autophagy strategies to enhance tumor treatment. Scheme 1. [64]Scheme 1 [65]Open in a new tab (a) Schematic diagram of the synthesis of GFD. (b) Schematic diagram of the antitumor mechanism of GFD. (c) Mechanism diagram of excessive autophagy, ferroptosis, and apoptosis induced by GFD Materials and methods Materials Hydrogen peroxide (H₂O₂), GA, and ferric chloride hexahydrate (FeCl₃·6 H₂O) were obtained from Aladdin (Shanghai, China). Methanol, Polyvinylpyrrolidone (PVP), DOX, chloroquine (CQ), methylene blue (MB), o-phenylenediamine (OPD), p-phthalic acid (PTA), cyanine 5.5 carboxylic acid (Cy5.5), reduced glutathione (GSH) and 5,5’-dithiobis (2-nitrobenzoic acid) (DTNB) were purchased from Macklin (China). Dulbecco’s modified Eagle’s medium (DMEM), phosphate-buffered saline (PBS), trypsin-EDTA solution, Cell Counting Kit-8 (CCK-8) and 4’,6-diamidino-2-phenylindole (DAPI) were obtained from SparkJade (China). Fetal bovine serum (FBS) was purchased from Wisent (China). Annexin V-FITC/PI, BODIPY 581/591 C11 and monodansylcadaverine (MDC) were acquired from Bestbio (China). 2′,7′-Dichlorofluorescin diacetate (DCFH-DA) and mitochondrial membrane potential assay kit (JC-1) were obtained from Biyotime (China). Preparation of GFD GFD was synthesized via a one-step self-assembly method (Scheme [66]1). Initially, The molar ratio of FeCl₃·6 H₂O: GA: DOX was maintained at 5: 1: 2. 20 mg of FeCl₃·6 H₂O and 100 mg of PVP were dissolved in 10 mL methanol under vigorous stirring for 10 min. Subsequently, 10 mg of GA and 18 mg of DOX were separately dissolved in 5 mL methanol each. Under continuous stirring, the GA and DOX methanol solutions were added dropwise into the FeCl₃ solution. After thorough mixing, the mixed solution was subjected to ultrasonic treatment for 10 min, then low-speed stirring was carried out at 200 rpm and a rotational speed of 37℃ for 24 h. Upon completion of stirring, the solution was dialyzed against deionized water for 24 h using a dialysis membrane with a molecular weight cutoff of 8,000 Da to obtain the GFD solution. The mixture was then centrifuged at 12,000 rpm for 20 min to collect a black precipitate, which was washed three times with purified water and dried overnight to yield the final GFD nanoparticles. Characterization of the GFD nanomaterials The synthesized GFD was analyzed via transmission electron microscopy (TEM; JEOL-JEM 2100 F, Japan) and scanning electron microscopy (SEM; Hitachi S4800, Japan) to acquire high-resolution images. The average particle diameter and zeta potential of GFD particles in different solutions were measured using dynamic light scattering instrument (Malvern Zetasizer Nano S90, UK). The elemental composition and valence states were analyzed via X-ray photoelectron spectroscopy (XPS; Thermo Scientific K-Alpha, USA). XRD patterns were recorded via X-ray diffractometer (Thermo Fisher ARL EQUINOX 3000, USA). Fourier transform infrared (FT-IR) spectroscopy was performed using Nexus 670 instrument to reveal the stretching vibrations of the characteristic peaks. Finally, the encapsulation efficiency and drug loading capacity of GA and DOX were calculated via UV‒visible spectrophotometer (Thermo Scientific Biomate 160, USA) to determine the optimal molar ratios of the components. Ion release rate It is essential to evaluate the Fe³⁺ release rate of GFD at various pH values to assess its pH sensitivity and sustained-release properties. First, GFD solution (50 mg/mL) was prepared by dissolving it in aqua regia. The component ratios were determined via ICP optical emission spectrometry (ICP‒OES; EXPEC 6500). Dialysis media at pH 6.5 and pH 7.4 were prepared, and 1 mL GFD solution (1 mg/mL) was placed into a dialysis membrane. Samples of the dialysis media were collected at various time intervals, and the Fe³⁺ concentrations were measured via ICP‒OES. The cumulative Fe³⁺ release rates over time were calculated. The prolonged release of DOX was assessed via the method similar to that outlined above. Evaluation of POD-like activity in response to GFD MB and OPD are frequently utilized as indicators for assessing POD-like activity [[67]32]. The absorbance of OPD was measured at 425 nm via UV‒visible spectrophotometer in PBS. The absorbances of PBS solutions (1 mL, pH 6.5) containing H₂O₂, OPD, GFD, GFD + OPD and GFD + OPD + H₂O₂ were measured separately. Additionally, the sensitivity of different concentrations of H₂O₂ was verified by establishing an isometric gradient of H₂O₂ concentrations (0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 mM). The sensitivity of GFD was then verified by establishing an isometric gradient of GFD (0, 10, 20, 40, 60 and 80 µg/mL). For the MB probe assay, OPD was replaced with MB, and the wavelength was set at 665 nm; the other groupings and test methods were similar. Additionally, PTA was introduced to further evaluate POD-like activity. The specific groupings and treatments were similar. The fluorescence intensity at 425 nm was measured via a spectrophotometer (F-4700 FL Spectrophotometer) to monitor changes in the characteristic peaks. Data were collected and statistically analyzed. In the Electron Spin Resonance (ESR) experiment, GFD nanoparticles were mixed with 100 µM H₂O₂ in phosphate-buffered saline (PBS, pH 6.5) and incubated at 37 °C for 20 min. Subsequently, 50 mM DMPO was added as a spin-trapping agent, and the mixture was vortexed thoroughly. The sample was immediately transferred into quartz capillaries and analyzed using ESR spectrometer (JEOL-JEM 2100 F). The generation and types of radicals were confirmed by characteristic peaks observed in the ESR spectrum. Glutathione peroxidase (GSH-Px) activity of GFD GSH-Px activity was evaluated using DTNB as a probe. In a reaction system at pH 6.5, 100 µM GSH and various concentrations of GFD (50, 100 and 200 µg/mL) were introduced. At different time intervals (0, 0.25, 0.5, 1, 2, 3, 4, 5 and 6 h), 20 µM DTNB was added as the probe. The absorbance at 412 nm was recorded via a UV‒Vis spectrophotometer. Oxygen concentration detection The oxygen concentration was assessed with a portable dissolved oxygen meter (JPB605F). Prior to the experiments, the dissolved oxygen was eliminated by flushing with nitrogen for 40 min. The oxygen production levels of different concentrations of GFD in a pH = 6.5 solution with a 100 µM H₂O₂ concentration were determined. Similarly, the deoxygenation levels in oxygen-free water with a consistent concentration of GFD and varying concentrations of H₂O₂ were measured. The data were then collected and subjected to statistical analysis. Hemolysis assay To assess the in vitro biocompatibility of the GFD, a hemolysis assay was performed. Mouse whole blood (500 µL) was diluted with PBS to a final volume of 10 mL. The samples were then centrifuged at 10,000 rpm for 10 min, and the process was repeated three times to collect the washed red blood cells. The supernatant was removed and adjusted to 10 mL with PBS. Then, 100 µL of red blood cell suspension was mixed with 900 µL of PBS solution containing different concentrations (25, 50, 100, 200 and 400 µg/mL) of GFD. Deionized water was used as the positive control, whereas PBS served as the negative control. The samples were incubated in a cell incubator for 4 h. After the samples were centrifuged at 10,000 rpm for 5 min, 100 µL of the supernatant was removed from each tube, and the absorbance at 577 nm was recorded. Cell lines The GL261 mouse glioma cell line and HUVECs (human umbilical vein endothelial cells) were obtained from the Cell Bank of the Chinese Academy of Sciences. These cells were grown in DMEM containing 10% FBS and 1% penicillin-streptomycin solution and maintained in an incubator with 20% oxygen and 5% carbon dioxide. Cellular uptake experiment GL261 and HUVECs cells were seeded in 35 mm confocal dishes at a density of 3 × 10⁵ cells per dish and incubated at 37 °C for 24 h until reaching approximately 80% confluence. For the preparation of Cy5.5-labeled GFD solution, 20 µL of Cy5.5 dye was added simultaneously with GA and DOX during the synthesis process, while maintaining all other synthesis steps and conditions unchanged. Subsequently, 800 µL of 50 µg/mL Cy5.5-labeled GFD solution was added to each dish. Cellular uptake was monitored and images were acquired at designated time points (0, 2, and 4 h). Cytotoxicity assay To evaluate the cytotoxicity of GFD, HUVECs. GL261 cells were plated in 96-well cell culture plates at a density of 1 × 10⁴ cells per well. After the cells adhered overnight, they were treated with different concentrations of GFD solution for 12 h. Subsequently, cell viability was measured using the CCK-8 assay. Next, we assessed the cell viability of GAFe (GF), GFD, and the combination of GF and GFD with H₂O₂ (100 µM) at the same concentrations of 0, 20, 40, 60, 80, and 100 µg/mL. Cell scratch assay GL261 cells were seeded in six-well plates at a density of approximately 5 × 10⁵ cells per well. After incubation for 12 h, the cells formed a confluent monolayer. A sterile needle was used to create a linear scratch on the monolayer, followed by gentle rinsing of the wells with PBS. The cells were then maintained in a serum-free medium for further culture. The following seven treatments were compared: (1) the Control group (PBS); (2) the H₂O₂ group; (3) the GF group; (4) the GF + H₂O₂ group; (5) the DOX group; (6) the GFD group; and (7) the GFD + H₂O₂ group. The concentration of H₂O₂ was set at 100 µM, while the concentrations of GF and GFD were 100 µg/mL. The cells were treated for 12 h, and cell migration was observed by taking images at consistent intervals via bright-field microscopy. Calcein-AM/PI assay A calcein-AM/PI assay was employed to assess the therapeutic impact of GFD on glioma cells. Specifically, GL261 cells were plated at a density of 5 × 10⁵ cells per dish in 35 mm confocal dishes and incubated for 24 h until they reached 80% confluence. The cells were subsequently treated for 12 h. The specific grouping was identical to that of the cell scratch assay, with the material concentration set at 100 µg/mL. The cells were treated with calcein-AM/PI for 16 min and then examined under a confocal microscope. Colony formation assay After digestion, the cells from each group were resuspended in cell suspensions in culture medium. The cell suspension was diluted stepwise, and 800 cells were plated per well. After gently shaking to ensure even cell distribution, 2 mL of complete medium was added to each well for incubation. After the cells had been allowed to adhere overnight, the materials were added at the same concentrations as those used in the scratch assay and incubated for 12 h. The original medium was then discarded and replaced with a complete medium (2 mL per well), and the cells were cultured for 13 days, with the medium changed every 2 days. When visible colonies were observed in the culture dishes, the culture was stopped. The supernatant was discarded, and the wells were washed with PBS. Each well was then fixed with 2 mL of 5% paraformaldehyde for 15 min. After fixation, the paraformaldehyde was removed, and an appropriate volume of crystal violet staining solution was added for 30 min. The staining solution was gently rinsed away with running water and then allowed to dry in the air. The plates were then inverted, and a transparent sheet with a grid was placed on top, allowing colonies to be counted visually. Detection of ROS DCFH-DA was utilized as a probe to measure intracellular ROS production. GL261 cells were plated and incubated for 24 h. Next, the cells were treated for 6 h following the procedure of the calcein-AM/PI assay, with the specific groupings and drug concentrations identical to those used in the calcein-AM/PI experiment. Subsequently, the cells were exposed to fresh medium containing DCFH-DA for 20 min. Fluorescence images of the cells were captured to observe the production of ROS. Mitochondrial membrane potential detection The culture and treatment of GL261 cells were performed as described above for live‒dead staining. The overall membrane potential was observed via confocal microscopy. The cells were treated and then stained, followed by observation and imaging under a microscope. Photographs were taken within 2 h after staining. Detection of autophagosomes MDC staining was used to directly label autophagosomes in cells, allowing observation of the number and general morphology of autophagosomes via confocal microscopy. The culture and drug treatment of GL261 cells were conducted in the same manner as in the live/dead staining assay. GL261 cells were plated in confocal dishes and incubated until they reached 80% confluence. The cells were then subjected to various treatments and stained. Fluorescence images were obtained within 2 h after staining. Detection of lipid peroxidation BODIPY 581/591 C11 was used to assess ferroptosis. In a similar procedure, GL261 cells were plated in confocal dishes and cultured until they reached approximately 80% confluence. The cells were then subjected to different treatments and stained via a lipid peroxidation staining kit following the provided instructions. Fluorescence images were subsequently captured. Western blot (WB) analysis WB was performed to evaluate protein expression. In this study, proteins involved in autophagy, apoptosis, and ferroptosis—such as GPX4, caspase-3, p62, beclin1, and LC3—were analyzed. GL261 cells were treated under different conditions for 12 h and then collected via trypsinization and centrifugation. Cellular proteins were extracted via RIPA buffer. Protein samples were resolved via SDS‒PAGE and subsequently transferred to PVDF membranes for additional analysis. The membranes were blocked with nonfat milk, followed by an overnight incubation at 4 °C with diluted primary antibodies. After brief washes, the membranes were incubated with diluted secondary antibodies at room temperature for 90 min. Chemiluminescent emissions were measured, and protein images were acquired with the help of an imaging system (Bio-Rad ChemiDoc, USA). The data were then analyzed with the corresponding software. Bio-TEM To investigate the effects of GFD on mitochondria and the induction of ferroptosis, TEM imaging was conducted on GL261 cells cultured in vitro. The cells were plated in six-well plates and exposed to various nanodrugs for 12 h, the drug treatment concentrations were identical to those used in the calcein-AM/PI staining assay. The cells were rapidly fixed, scraped, and centrifuged, followed by a second fixation. The samples were imaged via Bio-TEM. mRNA sequencing analysis To assess the gene expression profiles across the treatment groups, total RNA sequencing was performed on GL261 cells. The cells were exposed to different drugs, and mRNA was extracted via TRIzol reagent. Personalbio Co., Ltd. (Shanghai, China) conducted a series of processing steps for mRNA, including sequencing. Gene expression levels were normalized via the transcripts per kilobase million (FPKM) method. Differentially expressed genes (DEGs) were then subjected to the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) enrichment analyses to identify the functional pathways associated with these genes. In vivo therapeutic efficacy evaluation BALB/c nude mice (female, 6 weeks old, 18–20 g) were used as the animal model in this study. The mice were maintained in an SPF facility at Anhui Medical University, and all the animal procedures were approved by the Ethics Committee of Anhui Medical University (No. LLSC20220731). To establish a subcutaneous xenograft model of GL261 cells, each mouse was injected subcutaneously with 100 µL of PBS containing 5 × 10⁶ cells(n = 4). Tumor growth was monitored, and after 7 days, the tumors reached an approximate size of 100 mm³. The mice were randomly divided into four groups and treated with 100 µL of the respective injection: (1) the control group (PBS); (2) the GA group; (3) the DOX group; (4) the GFD group. The drug concentrations were referenced from the aforementioned in vitro experiments, approximately set at 20 mg/kg. Tumor size was assessed daily via calipers, and the tumor volume was determined. The body weights of the mice were measured every two days. The mice were sacrificed on day 14, followed by dissection, imaging, and histological staining, including hematoxylin and eosin (H&E), Ki67 for tumor cell proliferation, TUNEL for apoptosis, and staining for GPX4, Beclin1, p62, and LC3. The tissue sections were imaged via a fluorescence microscope (Leica DMI8, Germany). The orthotopic tumor implantation method involved stereotactic guidance for injection, with 6 mice in each group. All other procedures were performed in a similar manner as described above. Biosafety evaluation All animal experiments adhered strictly to the guidelines established by the regional animal ethics committee and received approval from the Animal Protection and Utilization Committee of Anhui Medical University. BALB/c mice (6 weeks old, 18–20 g) were administered GFD solution through intraperitoneal injection into the lower right abdomen. Hematological and biochemical parameters, including RBC count, white blood cell (WBC) count, platelet (PLT) count, neutrophil (NEU) count, mean corpuscular volume (MCV), lymphocyte (LYM) count, hemoglobin (HGB) concentration, hematocrit (HCT) level, creatinine (CRE) concentration, blood urea nitrogen (BUN), and alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities were evaluated. Furthermore, the major organs were collected and analyzed via H&E staining. Statistical analysis Quantitative data were collected from at least three separate samples. Graphs were created with Origin 2024 software. For comparisons of the means of the two groups, the data were statistically analyzed via Student’s test. Statistically significant thresholds were set at *p < 0.05, **p < 0.01, and ***p < 0.001. Results and discussion Synthesis and characterization of GFD In this design, GA, Fe³⁺, and DOX were selected as raw materials, with PVP used for modification. GA contains many benzene rings and alkyl groups, providing chelation sites for metal ions. Fe³⁺ is a common element in the human body. The structure of DOX contains multiple functional groups that can form stable covalent bonds with other molecules. PVP is a commonly used nanodrug carrier. These components are all biocompatible [[68]33, [69]34]. The materials were effectively combined through a self-assembly method, and after synthesis, material characterization tests were conducted (Fig. [70]1a). TEM revealed that the successfully prepared GFD nanoparticles were uniformly sized round particles (Fig. [71]1b and c). The average diameter of GFD particles is approximately 5 nm. with a PDI value of less than 0.4 and an average zeta potential of -24 to -32 mV (Fig. [72]1d and [73]S1). XRD analysis verified that GFD exhibited a characteristic structure, with the characteristic peaks of GFD similar to those of FeCl₃·6 H₂O (JCPDS No. 01-0475-1069) (Fig. [74]S2). XPS revealed the chemical elements and their valence states in GFD. As shown in Figs. [75]1e-f and Fig. [76]S3, the peaks at 710.2, 538.2, 401.3 and 280.4 eV correspond to Fe 2p, O 1 s, N 1 s, and C 1 s, respectively, further indicating that iron exists in the GFD nanoparticles in the states of Fe²⁺ and Fe³⁺. The FT-IR spectrum of GFD shows absorption peaks at 2958 nm, 1695 nm, 1315 nm, and 1165 nm, corresponding to the C–H, C = C, C = N, C–N, and C–O bonds in GA, GF, and GFD, respectively (Fig. [77]1g). These results confirm the successful synthesis of GFD. Fig. 1. [78]Fig. 1 [79]Open in a new tab Characterization of GFD. (a) Schematic diagram of the synthesis of GFD. (b-c) Representative TEM image of a GFD. (d) Zeta potential of the GFD. (e-f) XPS spectra of GFD. (g) FTIR spectrum of GFD. (h) UV‒vis spectra of different concentrations of DOX. (i-j) Time-dependent release profiles of Fe³⁺ and DOX from GFD at pH 6.5/7.4 According to the literature, DOX and GA have characteristic peaks at 480 nm and 360 nm, respectively [[80]35]. The characteristic peaks of different concentrations of GA and DOX were measured, and the optimal ratio of DOX to GA was calculated and adjusted accordingly (Figs. [81]1h and [82]S4). When the DOX/GA molar ratio was 2:1, the DOX loading capacity of GFD was 86.9%, and the encapsulation efficiency was 18.67%. By balancing the loading capacity and encapsulation efficiency, the DOX/GA ratio of 2:1 was selected as the optimal ratio for the following experiments. Furthermore, the sustained-release capacity of GFD in different pH environments was verified. By measuring the contents of Fe and DOX under different pH conditions, it was found that as the pH decreased, the release rates of Fe³⁺ and DOX from GFD progressively increased. At pH 7.4, after 72 h, the release rates of Fe³⁺ and DOX were 3.4% and 1.7%, respectively (Fig. [83]1i and j). At pH 6.5, after 72 h, the release rate of Fe³⁺ reached 45%, and the release rate of DOX reached 43%. These results indicate that in the acidic tumor microenvironment (TME), GFD can gradually dissociate to release Fe³⁺ and DOX, confirming the sustained release effect of GFD. Furthermore, we also observed the morphological changes of GFD under different pH conditions, as shown in the Fig [84]S5. The material remained stable in PBS buffer at pH 7.4, while significant dissociation occurred at pH 6, with nearly no intact particles observable. The above results demonstrate that sustained release is pH-responsive, which facilitates the accumulation of the drug in the mildly acidic tumor environment, thereby fully exerting its antitumor effect. In summary, high-purity, stable, and monodisperse GFD was successfully prepared. GFD has good physical properties, pH responsiveness, and excellent sustained-release capabilities. Evaluation of the multienzyme activity of GFD Building upon prior studies regarding the catalytic activity of iron-based nanozymes, GFD was hypothesized to possess similar multienzyme activities, which were evaluated experimentally (Fig. [85]2a) [[86]36]. The generation of reactive ROS was detected using MB, OPD, and PTA as substrates to verify the POD-like activity of GFD. Under acidic conditions, POD-like activity can decompose exogenous H₂O₂ into ROS. The generated ROS can alter the color of MB and OPD from blue to colorless and from colorless to yellow, respectively, and induce distinctive fluorescence emission of PTA at 425 nm. As shown in Figs. [87]2b-c and Fig. [88]S6, the sample containing GFD, OPD, and H₂O₂ presented the strongest absorbance peak at 450 nm, indicating POD-like activity. Moreover, as the concentrations of GFD and H₂O₂ increased, the absorbance at 450 nm also increased, and the solution gradually turned yellow, suggesting that the enzyme activity is related to the concentrations of the material and H₂O₂. Subsequently, ROS generation was assessed using MB as the substrate by observing the absorbance peak at 650 nm. Similar results to those with OPD were observed, as shown in Figs. [89]2d-e and Fig. [90]S7. Considering the high sensitivity of PTA in detecting low-concentration nanozymes and prolonged POD activity, PTA was used as a fluorescent probe to detect the fluorescence peak at 425 nm. Compared with the other three samples, the samples containing PTA, GFD, and H₂O₂ presented the highest fluorescence intensity (Fig. [91]S8). Additionally, the fluorescence intensity increased with increasing concentrations of GFD or H₂O₂ (Fig. [92]2f and g). Further evaluation of GFD’s ability to induce free radicals and their types was conducted using ESR. After adding GFD and H₂O₂ to the solution, the ESR spectrum exhibited a 1:2:2:1 intensity ratio, confirming the generation of ·OH (Fig. [93]S9). This indicates that GFD can react with H₂O₂ to produce ·OH, thereby increasing oxidative stress within tumors. In summary, GFD demonstrated significant POD-like activity, which was dependent on the concentrations of the material and H₂O₂. Fig. 2. [94]Fig. 2 [95]Open in a new tab Multienzyme activity of GFD. (a) Schematic diagram of the multienzyme activity of GFD. (b-c) OPD colorimetric assay to evaluate ROS generation. (d-e) MB colorimetric assay to assess ROS generation. (f-g) PTA fluorescence analysis. (h) GSH-Px activity assay of GFD. (i) Time-dependent GSH depletion rates for 50 µg/mL, 100 µg/mL and 200 µg/mL GFD. (j) Decomposition rates of H[2]O[2] with different concentrations of GFD Elevated GSH levels in tumor cells can counteract ROS, thereby mitigating cellular damage. Consequently, reducing GSH levels can lead to increased ROS production; this process is known as GSH-Px-like activity. Since previous reports have shown that iron-based nanozymes exhibit GSH-Px-like properties, DTNB, which reacts with GSH to produce a characteristic absorbance peak at 442 nm, was used as a GSH probe for confirmation [[96]37]. After coincubation with 50 µg/mL GFD, the absorbance peak gradually decreased over time, and complete depletion of GSH was noted after 6 h (Fig. [97]2h). Experiments conducted with varying concentrations of GFD provided additional evidence that the extent of GSH depletion increased in proportion to the concentration of GFD; 200 µg/mL and 100 µg/mL GFD completely consumed GSH after 2 and 4 h, respectively (Fig. [98]2i). These results indicate that GFD possesses excellent GSH-Px-like activity. Hypoxia can promote the adoption of protective autophagy by tumors, hindering the efficacy of antitumor drugs [[99]38]. Some nanomaterials can decompose high concentrations of H₂O₂ in the TME into O₂, alleviating the hypoxic environment in tumor tissues. The results demonstrated that GFD also exhibits similar properties in acidic environments. in an acidic environment. As the concentrations of GFD and hydrogen peroxide increased, the corresponding concentrations of dissolved oxygen also gradually increased (Fig. [100]2j and [101]S10). Owing to its ability to catalyze the decomposition of hydrogen peroxide to produce oxygen, a GFD can efficiently improve the hypoxic conditions in tumor tissues, promoting excessive autophagy. Overall, these experiments confirmed that GFD has multiple enzyme activities, which is expected to contribute to its excellent antitumor performance. In vitro antitumor capability evaluation Given the excellent multienzyme activity of GFD, its antitumor effects at the cellular level were further investigated. First, the cellular uptake of GFD in HUVECs and GL261 cells was examined through endocytosis experiments. In these experiments, blue fluorescence indicated cell nuclei, while red fluorescence represented Cy5.5-labeled GFD. Over time, red fluorescence surrounding blue nuclei was observed in GL261 cells (Fig. [102]3a). By 4 h, this perinuclear accumulation became particularly pronounced. In contrast, HUVECs exhibited predominantly diffuse red fluorescence with minimal nuclear envelopment, demonstrating successful uptake of GFD by GL261 cells. Subsequently, its biocompatibility was evaluated via hemolysis assays (Fig. [103]3b). Even when the GFD concentration increased to 400 µg/mL, the rate of hemolysis remained under 5%, demonstrating excellent hemocompatibility. Next, the cytotoxic selectivity of GFD was preliminarily assessed in HUVECs and the GL261 cell (Fig. [104]3c). After 12 h of incubation, even at a GFD concentration as high as 100 µg/mL, the viability of HUVECs remained nearly 100%. Moreover, due to the serum-containing culture medium, an obvious proliferation tendency was observed. In sharp contrast, the viability of the GL261 cells continued to decrease, dropping to less than 50% at a concentration of 100 µg/mL. These findings indicate that GFD has significant cytotoxicity toward GL261 cells but has almost no effect on normal cells. Further incubation of GF and GFD at different concentrations was conducted to explore their concentration- and time-dependent cytotoxicity. Considering the multienzyme activity of GFD, 0.1 mM H₂O₂ was added to the system to simulate the high concentration of H₂O₂ in the TME, with the aim of investigating the role of multienzyme activity in GFD-mediated tumor killing. As shown in Fig. [105]3d and [106]S11, the concentration- and time-dependent cytotoxicity of GF, GFD, GF + H₂O₂, GFD + H₂O₂ was confirmed. Additionally, after stimulation with a high concentration (100 µM) of H₂O₂, the cell-killing abilities of GF and GFD were further enhanced. These results indicate that multienzyme activity enhances the tumor-killing effect. Fig. 3. [107]Fig. 3 [108]Open in a new tab Evaluation of in vitro antitumor activity. (a) Cellular uptake assay in GL261 and HUVECs. (b) Hemolysis assay. (c) Viability of GL261 and HUVECs cells after treatment with different concentrations of GFD. (d) Cell viability after treatment with GF, GFD, GF + H₂O₂ and GFD + H₂O₂ at concentrations of 0, 20, 40, 60, 80 and 100 µg/mL (with 100 µM H₂O₂). (e) Analysis of the scratch wound assay results. (f) Analysis of the colony formation assay results. (g) Scratch wound assay. (h) Colony formation assay Furthermore, the concentration of 50 µg/mL was selected for follow-up experiments. After successfully verifying the toxicity of a GFD to tumors, scratch assays were conducted to evaluate its effect on cell migration via the following seven interventions: (1) the control group (PBS); (2) the H₂O₂ group; (3) the GF group; (4) the GF + H₂O₂ group; (5) the DOX group; (6) the GFD group; and (7) the GFD + H₂O₂ group. After incubation, progressive inhibition of glioma cell migration was observed (Fig. [109]3e and g). Compared with the control, GF, H₂O₂, and GF + H₂O₂ groups, the DOX, GFD, and GFD + H₂O₂ groups exhibited significant inhibition of GL261 cell migration (p < 0.05). The GFD + H₂O₂ group presented the greatest migration inhibition, demonstrating that GFD has a significant inhibitory effect on tumor migration and that high concentrations of H₂O₂ in the TME can enhance the inhibitory effect. On the basis of these findings, colony formation assays were performed to further assess the impact of GFD on the colony formation and sustained proliferative potential of GL261 cells. As shown in Fig. [110]3f and h, after two weeks of culture, significant differences were observed among the groups. The untreated control group exhibited good proliferation, whereas the colony formation and proliferative capacities of GL261 cells under the different treatment conditions were suppressed to varying extents. In particular, in the GFD + H₂O₂ group, which exhibited the most pronounced inhibition, the GL261 cells showed almost no growth. These findings indicate that GFD suppresses the colony formation and long-term proliferation of GL261 cells. The tumor-killing effect of GFD was further validated through the use of calcein-AM/PI in cells subjected to different treatments (Fig. [111]4a and d). The fluorescence images reveal patterns that are consistent with the results of earlier toxicity experiments, with GFD demonstrating the most significant cytotoxicity against GL261 cells. The cytotoxic effect was further enhanced in the GF and GFD groups after the addition of H₂O₂. Overall, these experiments indicate that GFD possesses good biosafety and excellent tumor-killing efficacy. Fig. 4. [112]Fig. 4 [113]Open in a new tab Exploration of antitumor activity and apoptosis in vitro. (a) Analysis of the results of live/dead staining results. (b) Analysis of the JC-1 staining results. (c) Analysis of the results of the apoptosis assay results. (d) Live/dead staining. (e) JC-1 staining. (f) The apoptosis assay Exploring the antitumor mechanism of excessive autophagy/ferroptosis/apoptosis After the excellent in vitro antitumor effects of GFD were verified, the underlying mechanisms by which GFD kills tumor cells were further explored. JC-1 was used to examine tumor cell apoptosis triggered by GFD. Disruption of the mitochondrial membrane potential is associated with apoptosis. These changes were observable through JC-1 staining (Fig. [114]4b and e). As shown in the figure, with the use of various materials, the red fluorescence gradually diminished, whereas the green fluorescence progressively intensified, indicating interference with mitochondrial function. In the control group and H₂O₂ group, this interference was not significant; however, it increased significantly after the addition of GF and DOX, and this effect was further enhanced in the GFD and GFD + H₂O₂ groups. These results suggest that GF, DOX, and GFD can effectively induce cell apoptosis. The proportion of apoptotic cells induced by GFD in the overall population and at different stages of apoptosis was investigated using the apoptosis detection kit (Fig. [115]4c and f). The proportion of apoptotic cells increased with different treatments, predominantly in the late apoptotic phase. The late apoptotic ratio peaked after the addition of GFD, followed by cell necrosis. These results suggest that apoptosis, especially late-stage apoptosis, plays an important role in cell death induced by GFD, and DOX has a particularly significant impact on apoptosis. After verifying the successful induction of apoptosis by GFD, the roles of oxidative stress induced by ROS in the antitumor activity of GFD were further experimentally confirmed. ROS were directly assessed via the probe DCFH-DA (Figs. [116]5a and [117]S12). Green fluorescence indicates intracellular ROS production, whereas blue fluorescence represents cell nuclei. As shown, the control and H₂O₂ groups presented minimal green fluorescence. In contrast, the GF-related groups (GF, GF + H₂O₂, GFD, GFD + H₂O₂) displayed significant green fluorescence, and fluorescence was further enhanced by the addition of H₂O₂. These results indicate that GF is crucial for the production of ROS triggered by GFD. After confirming the generation of ROS, the role of ferroptosis in the antitumor effects of GFD was investigated. Lipid peroxidation, caused by oxidative stress and a hallmark of ferroptosis, was detected via C11-BODIPY staining (Figs. [118]5b and [119]S13) [[120]39]. Similar to the ROS results, the GF-related groups (GF, GF + H₂O₂, GFD, GFD + H₂O₂) presented pronounced green fluorescence, indicating increased lipid peroxidation accumulation and indirectly confirming the successful induction of ferroptosis. Fig. 5. [121]Fig. 5 [122]Open in a new tab Exploration oxidative stress of and ferroptosis in vitro. (a) DCFH-DA staining. (b) LPO staining After verifying the oxidative stress and ferroptosis, the potential of GFD to induce excessive autophagy was further examined. Initially, the role of autophagy in antitumor activity was clarified through cytotoxicity assays. As shown in Fig. [123]6a, CQ, an autophagy inhibitor, was used to suppress cellular autophagy. After the addition of CQ, no substantial difference in cell viability was determined between the GA and GA + CQ groups when the material concentration was less than 60 µg/mL. Only when the concentration exceeded 60 µg/mL, GA exhibit significant cytotoxicity due to excessive autophagy. The replacement of GA with GF subsequently resulted in marked and universal excessive autophagy at various concentrations (Fig. [124]6b), and all these effects were statistically significant. These results strongly demonstrate that GF can induce and amplify excessive autophagy, which aligns with the initial hypothesis. To visualize excessive autophagy by GF-related groups (GF, GF + H₂O₂, GFD, GFD + H₂O₂), MDC staining was used to examine the quantity and shape of the autophagosomes in the cells. As shown in Fig. [125]6c and d, the number of autophagosomes increased progressively with different treatments, particularly in the GF-related groups(GF, GF + H₂O₂, GFD, GFD + H₂O₂), where the number of autophagosomes was significantly greater and the morphological changes were obvious. Combined with the substantial increase in ROS in these groups, the results confirmed that GFD may achieve excessive autophagy by inducing the generation of large numbers of autophagosomes, ROS played an important role in this process. Fig. 6. [126]Fig. 6 [127]Open in a new tab Exploration of excessive autophagy and ferroptosis in vitro. (a) Cell viability after treatment with GA and GA + CQ. (b) Cell viability after treatment with GF and GF + CQ. (c-d) MDC staining results and analysis. (e-f) WB results and analysis (1: control group; 2: H₂O₂ group; 3: GF group; 4: GF + H₂O₂ group; 5: DOX group; 6: GFD group; 7: GFD + H₂O₂ group). (g) Bio-TEM images of cells after different treatments (red arrow: mitochondria; yellow arrow: autophagosomes) After confirming the successful induction of ferroptosis, apoptosis, and excessive autophagy at the cellular level, these processes were further validated at the protein level via WB analysis. As shown in Fig. [128]6e and f, P62, Beclin-1, and LC3 are autophagy-related proteins, with Caspase-3 closely associated with apoptosis. GPX4 is a ferroptosis-related protein. With respect to apoptosis, elevated levels of apoptotic proteins were detected in all the groups except the control and H₂O₂ groups, which aligns with the findings from the cellular experiments. In terms of autophagy, the effects in the GFD-treated groups (GFD, GFD + H₂O₂) were greater than those in the other groups. These findings confirm that GFD exerts antitumor effects through excessive autophagy. In terms of ferroptosis, the expression of ferroptosis-related proteins had a significant effect on the GF-related groups (GF, GF + H₂O₂, GFD, GFD + H₂O₂), confirming that GFD induces ferroptosis successfully. To further substantiate these results, Bio-TEM was used to visually verify the intracellular morphological changes related to autophagy, ferroptosis, and apoptosis (Fig. [129]6g). The results demonstrated that the mitochondria showed marked swelling, membrane rupture and ruffling, which are hallmark characteristics of both ferroptosis and apoptosis. Additionally, numerous autophagosomes, which are hallmarks of autophagy, were observed in the GFD group. These observations verified the successful induction of excessive autophagy, ferroptosis, and apoptosis by GFD. Antitumor biological mechanism of GFD In our preliminary experiments, the antitumor mechanism of GFD was successfully confirmed at the cellular and protein levels. Additionally, numerous autophagosomes and morphological changes associated with the relevant mechanisms were visually observed within the cells via electron microscopy. To gain a deeper understanding and further validate the relevant mechanisms of GFD in the antitumor process (Fig. [130]7a), transcriptome sequencing was conducted on extracts from GL261 cells divided into control and GFD groups. The overall distribution and analysis of genes are shown in Fig. [131]7b and c. Compared with the control group, the GFD group presented upregulated expression of 2016 genes and downregulated expression of 2069 genes (|log2 FC| > 1, p-value < 0.05). Heatmaps showing marked differences in gene expression between the control and GFD groups. To gain deeper insights into the DEGs, enrichment analyses were conducted separately for the genes whose expression increased or decreased. KEGG pathway enrichment analysis was also performed (Fig. [132]7d) and revealed significant upregulation of pathways related to autophagy, apoptosis and ferroptosis. Notably, the autophagy pathway is related primarily to the mTOR and the PI3K/Akt pathway, the apoptosis pathway is directly related to the FoxO pathway, and the p53 pathway and the MAPK pathway are closely associated with ferroptosis and the autophagy-dependent ferroptosis pathway, GFD potentially mediates crosstalk between autophagy and ferroptosis through coordinated modulation of the mTOR/PI3K-Akt (autophagy-related) and p53/MAPK (ferroptosis-associated) signaling pathways [[133]40–[134]43]. GO enrichment analysis (Fig. [135]7e) revealed that the most enriched terms were related to cellular ferroptosis, apoptosis, and autophagy. WikiPathway enrichment analysis (Fig. [136]7f) further confirmed the high expression of related genes. Fig. 7. [137]Fig. 7 [138]Open in a new tab Exploration of the intrinsic antitumor mechanisms. (a) Scheme of intracellular apoptosis, ferroptosis and autophagy. (b) Volcano plots of different groups. (c) Hierarchical clustering analysis of different groups. (d) KEGG enrichment analysis. (e) GO enrichment analysis. (f) WikiPathways enrichment analysis In vivo anticancer activity After the mild toxicity and excellent antitumor effect of GFD were verified, in vivo antitumor ability was evaluated via animal experiments. BALB/c mice were subcutaneously implanted with GL261. In previous experiments, the optimal molar ratio of GA to DOX was determined to be 1:2. On the basis of the outcomes of the cellular experiments, this GFD ratio was deemed effective in achieving desirable tumor-killing efficiency. Therefore, to maintain consistency, this ratio was used in subsequent animal experiments. The qualified mice were assigned to four groups (n = 4): (1) the Control group; (2) the GF group; (3) the DOX group; and (4) the GFD group. The respective formulations were administered directly into the tumor tissue (Fig. [139]8a). Across the two-week treatment phase, images of the tumors were captured for each group (Fig. [140]8b), tumor volumes were monitored (Fig. [141]8c), and tumor masses were recorded during treatment (Fig. [142]8d). The GFD group presented the most significant tumor inhibition effect, with a marked reduction in tumor size and weight. H&E staining and immunofluorescence analysis of harvested tumor sections revealed a significant reduction in tumor cells in the GFD group, with markedly decreased Ki67 expression, significantly increased TUNEL staining, and a significant effect on autophagy and the ferroptosis-related markers GPX-4, p62, and LC3. These findings indicate that a GFD effectively killed tumor cells and significantly inhibited tumor proliferation. Fig. 8. [143]Fig. 8 [144]Open in a new tab Evaluation of antitumor activity in vivo. (a) Schematic diagram of the subcutaneous tumor model. (b) Photographs of the tumors. (c) Relative tumor volume. (d) Terminal tumor mass. (e) H&E staining and immunofluorescence Further evaluation of the in vivo antitumor efficacy of a GFD was conducted using an intracranial orthotopic transplantation tumor model. The mice were treated according to the procedure shown in Fig. [145]9a, the specific method of orthotopic tumor implantation referenced in Fig. [146]9b. The mice were subsequently grouped for treatment, and during the treatment period, the relative body weights of the mice were measured to assess the therapeutic effects (Fig. [147]9c). The negligible weight loss indicates high safety of the treatment, whereas the slight weight reduction in the DOX and GFD groups may be related to the toxic side effects of DOX. In addition, the survival rates of the mice during the entire treatment duration were evaluated (Fig. [148]9d), revealing that the GFD group presented the highest survival rate (83.3%). Further histological staining and immunofluorescence analysis of harvested tumor sections were performed (Fig. [149]9e). H&E staining clearly demonstrated the effects of different treatments on the tumors. There were no significant changes in the Control and GF groups, whereas tumors in the DOX and GFD groups were significantly reduced, with markedly decreased infiltration. Compared with those in the Control group, the number of tumors in the GFD group was markedly lower, with a significantly decreased number of tumor cells observed in the pathological sections. The results of the immunofluorescence staining were similar to those observed in the subcutaneous tumors; the GFD group, which presented the best effect, presented significantly downregulated proliferation markers and significantly increased apoptosis, autophagy, and ferroptosis-related marker expression. Overall, both subcutaneous transplantation tumor models and intracranial orthotopic transplantation tumor models demonstrated that a GFD possesses excellent antitumor capabilities at the animal level. This capability appears to be related to the aforementioned antitumor mechanisms. Fig. 9. [150]Fig. 9 [151]Open in a new tab Evaluation of antitumor activity in vivo. (a) Schematic diagram of the orthotopic tumor model. (b) Photographs of the orthotopic tumor inoculation. (c) Relative changes in body weight. (d) Survival curve. (e) H&E staining and immunofluorescence Biodistribution and biosafety in vivo To ensure the biocompatibility of the material in vivo, H&E staining of the heart, spleen, liver, kidneys and lungs, as well as examinations of blood biochemical parameters, were conducted. During the implantation and treatment periods of subcutaneous xenografts and orthotopic tumors, no accumulation of a GFD or harm to healthy tissues was detected in the primary organs (Fig. [152]10a). Routine blood tests and biochemical marker analysis revealed no notable abnormalities (Fig. [153]10b). These experimental results indicate that GFD has good biocompatibility. Fig. 10. [154]Fig. 10 [155]Open in a new tab Evaluation of biosafety. (a) H&E staining of major organs; (b) Hematology and biochemical analysis Conclusion In conclusion, this study developed a novel GAFe nanozyme, aimed at addressing the scientific challenge of how to modulate autophagy to enhance the efficacy of tumor therapy. The nanoparticle (GAFe@DOX) utilizes GA to induce excessive autophagy and thus enhance the chemotherapeutic effect of DOX. GAFe nanozyme generates a large amount of ROS through multienzyme activity and ferroptosis activation, which ensures that GA-induced excessive autophagy is carried out smoothly. Meanwhile, autophagy and ferroptosis promoted each other, amplifying oxidative stress, synergizing with other programmed cell death pathways and ultimately enhancing the overall chemotherapeutic efficacy. This excessive autophagy strategy successfully overcomes the technical challenge of regulating autophagy to enhance chemotherapy efficacy, offering a novel therapeutic approach for optimizing treatment regimens for refractory tumors like GBM. Electronic supplementary material Below is the link to the electronic supplementary material. [156]Supplementary Material 1^ (1.9MB, docx) Acknowledgements