Abstract Chemotherapeutic drugs often fail to provide long-term efficacy due to their lack of specificity and high toxicity. To enhance the biosafety and reduce the side effects of these drugs, various nanocarrier delivery systems have been developed. In this study, we loaded the anticancer drug doxorubicin (DOX) and an MRI contrast agent into silica nanoparticles, coating them with pH-responsive and tumor cell-targeting polymers. These polymers enable the carrier to achieve targeted delivery and controlled drug release in acidic environments. This integrated diagnostic and therapeutic strategy successfully achieved both the diagnosis and treatment of liver cancer. Additionally, we demonstrated that the nanocarrier inhibits autophagic flux in liver cancer cells by targeting the autophagy-lysosome pathway and regulating the nuclear translocation of TFEB, thereby promoting tumor cell death. This novel diagnostic-integrated nanocarrier is expected to be a promising tool for targeted liver cancer treatment. Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03107-5. Keywords: Autophagic flux, Hepatocellular carcinoma, Lysosome, MRI imaging, Nanodrug Introduction Primary liver cancer ranks as the sixth most common malignant tumor globally and is the fourth leading cause of cancer-related deaths, posing a significant threat to public health [[40]1]. Transcatheter arterial chemoembolization (TACE) is the preferred treatment for stage IIb to IIIa liver cancer. However, the high toxicity and lack of specificity of chemotherapy drugs, such as DOX, result in suboptimal long-term effectiveness [[41]2]. Developing drugs that enhance chemotherapy targeting and improve drug permeability to liver cancer cells is crucial for reducing toxicity and increasing efficacy in infusion chemotherapy. Additionally, accurate diagnosis of lesions and imaging tracking during treatment are essential due to the metastatic nature of liver cancer. Currently, techniques such as ultrasound imaging, computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography-computed tomography (PET-CT) aid in detecting clinical changes in liver cancer [[42]3–[43]5]. Among these, MRI is widely used due to its high soft tissue contrast, deep tissue penetration, and lack of ionizing radiation. Developing integrated diagnostic and therapeutic strategies to track lesion changes and treat liver cancer has become a research hotspots [[44]6–[45]8]. Adriamycin (DOX) is a traditional broad-spectrum antitumor agent, primarily inhibiting RNA and DNA synthesis as a topoisomerase II inhibitor. In addition to these effects, DOX exerts its anticancer action by disrupting lysosomal acidification and function, thereby inhibiting autophagic flux. [[46]9, [47]10] However, severe toxic side effects and multidrug resistance (MDR) have hindered their clinical application. Additionally, the distribution of these chemotherapeutic drugs in the body is non-specific, and achieving therapeutic efficacy often requires high doses, which further exacerbates their cytotoxicity [[48]11, [49]12]. To address this challenge, the FDA approved Doxil^® [[50]13], a nanodrug encapsulated in polyethylene glycol-coated liposomes, which alleviates the conflict between the therapeutic efficacy and toxicity of DOX drugs. Nanocarrier drug delivery systems offer several significant advantages, including extended circulation time, improved targeting capabilities, controlled drug release, and reduced toxic side effects [[51]14]. The organic-inorganic hybrid nanodrug delivery systems prepared by combining mesoporous silica with responsive polymers have been extensively studied due to their advantages of easy preparation, tunable pore size, and facile surface modification. The porosity of mesoporous silica provides sites for drug loading, while polymer modifications on the surface endow the nanocarriers with targeting and responsive release capabilities. Therefore, this nanomaterial serves as an excellent DOX delivery vehicle, potentially enhancing DOX targeting, controlled release, and reducing toxic side effects. [[52]15, [53]16] pH-responsive polymers can react to environmental changes at pathological sites, demonstrating significant potential in drug delivery [[54]17]. Nanocarriers that respond to the acidic environment of lysosomes can significantly enhance the uptake of these carriers, thereby facilitating the localized delivery of therapeutic agents and effectively reducing systemic toxicity. [[55]18, [56]19] Lysosomes are membrane-enclosed dense granular structures involved in lysosomal reorganisation, maintenance and degradation pathways following autophagy and cell death and are closely associated with autophagic behaviour [[57]20]. Autophagy occurs at a low basal level in all cells and is upregulated in response to different stressors. (e.g., hypoxia, oxidative stress, endoplasmic reticulum stress, and drug therapy) [[58]21]. For maintaining the homeostasis of the intracellular environment is physiologically important for preventing cells from damage that may lead to abnormal cell function and malignant transformation of cells. Autophagy plays a dual role in cancer by inhibiting the growth of benign tumours on one hand and promoting the growth of advanced tumours on the other hand; therefore, over the past decade, many researchers have identified autophagy as a potential therapeutic target in cancer [[59]22]. A large amount of preclinical data confirms the feasibility of the idea of improving the clinical outcome of cancer patients by inhibiting autophagic flow. [[60]23, [61]24] In this study, we loaded the anticancer drug DOX and MRI contrast agent Gd₂O₃ nanoparticles onto mesoporous silica. To prevent drug leakage, we further combined these with pH-responsive and tumor cell-targeting polymers using a layer-by-layer method, resulting in the preparation of an integrated theranostic nanocarrier FA-Gd[2]O[3]@MSN-DOX. This nanoplatform targets hepatocellular carcinoma cells via folate receptors, accumulates in lysosomes, disrupts lysosomal function, and inhibits the nuclear translocation of TFEB through the mTOR-TFEB pathway, ultimately inducing apoptosis. Additionally, the Gd₂O₃ nanoparticles loaded onto the mesopores of MSN provide T1-weighted imaging (T1WI) enhancement, enabling dynamic monitoring of the distribution and release of MSN-carried drugs within the tumor via MRI. This approach allows for real-time evaluation of the therapeutic efficacy of MSN and facilitates the adjustment of drug dosage based on therapeutic response, thereby maximizing dose reduction and minimizing adverse drug reactions. (Scheme [62]1). Scheme 1. [63]Scheme 1 [64]Open in a new tab Schematic illustration for the preparation and inhibiting tumor progression through mTOR-TFEB signaling pathway of FA-Gd[2]O[3]@MSN-DOX. FA-Gd[2]O[3]@MSN-DOX can be internalized by cancer cells via FA receptor-mediated endocytosis. Within acidic organelles, the pH-responsive polyelectrolytes undergo charge reversal, leading to the release of DOX. Subsequently, FA-Gd[2]O[3]@MSN-DOX inhibits TFEB nuclear translocation by activating mTOR activity, thereby reducing the expression of autophagy-lysosomal-related genes and blocking the autophagic flux in cancer cells. Additionally, this pH-responsive theranostic nanocarrier can serve as a new MRI contrast agent, leveraging the paramagnetic properties of Gd[2]O[3] nanoparticles Materials and methods Materials and reagents FA-Gd[2]O[3]@MSN-DOX was synthesized according to literature methods [[65]25]. Adriamycin (DOX) was purchased from Shanghai Hohong Biopharmaceutical Technology Corporation. All antibodies were purchased from Hefei Ruijie Biotechnology Corporation. Mice were purchased from Jiangsu Jicui Pharmachem Biotechnology Corporation. Other reagents and consumables were provided by the Institute of Biological Medicine, Anhui Medical University. All cells were provided by the Institute of Biological Drugs, Anhui Medical University, and were cultured using DMEM complete medium containing 10% FBS (neonatal fetal bovine serum) and DMEM containing 1% penicillin-streptomycin in a constant-temperature incubator at 37 °C and 5% CO[2]. Cytotoxicity assay HepG2 cells, Hela cells, and A549 cells were seeded uniformly in 96-well plates with about 5000 cells per well, and three replicate wells in each group were added with different concentrations of Gd[2]O[3]@MSN, FA-Gd[2]O[3]@MSN, Gd[2]O[3]@MSN-DOX, and FA-Gd[2]O[3]@MSN-DOX after incubating for 48 h, respectively. The absorbance value A of each well was measured at 570 nm by adding MTT solution after treatment was completed by the enzyme marker. Flow cytometry HepG2 cells were seeded in 6-well plates, after cell attachment, the wells were spiked with 62.5 µg/mL of Gd[2]O[3]@MSN, FA-Gd[2]O[3]@MSN, Gd[2]O[3]@MSN-DOX, and FA-Gd[2]O[3]@MSN-DOX, respectively, A control group was set up and incubated for 48 h. The cells were digested with trypsin free of EDTA, collected, processed annexin V-FITC and PI staining, and the apoptosis rate was detected on a flow cytometer. Western blot Proteins were extracted by lysing cells using RIPA containing protease inhibitors. The extracted proteins were separated by SDS-PAGE and transferred to the PVDF membrane. The membrane was blocked with 10% skimmed milk and incubated overnight at 4 °C with antibodies including LC3I/II, p62, Akt/p-Akt, mTOR/p-mTOR, Bax, Bcl-2, caspase-3, cleaved caspase-3, TFEB, CTSB, LAMP1, ACTB, Lamin B1, and GAPDH. Subsequently, the secondary antibody was incubated for 1 h. Protein bands were visualized with ECL developer and quantified using Image J. Nuclear and cytoplasmic extraction HepG2 cells were co-incubated with the nanomedicine and then extracted from nuclear and plasma proteins using the Nuclear Plasma Separation Kit according to the instructions. Transmission Electron Microscopy HepG2 cells were co-cultured with the nanomedicine for 48 h and collected by centrifugation, immediately fixed by adding glutaraldehyde, and then fixed with 1% osmium acid for 2 h. Then they were stained, embedded, sectioned, and observed by transmission electron microscopy. RNA sequencing Cells were lysed with trizol reagent to extract RNA, which was then processed by BGI for sequencing. The raw data obtained from sequencing was filtered by SOAPnuke (v1.5.6) to obtain a clean date, which was then analyzed and plotted by Dr. Tom’s Multi-Organomics Data Mining System ([66]https://biosys.bgi.com). Immunofluorescence staining assay The cell crawls were fixed with 4% paraformaldehyde, goat serum was used for sealing for 1 h, 0.1% triton was used for permeabilization, primary antibody (TFEB) was used overnight, fluorescent secondary antibody was used for 1 h, the nuclei were stained with DAPI for 10 min, and the slices were sealed with anti-fluorescent extractant, and the fluorescence microscope was used for observation. TUNEL staining Tumor tissues were fixed with 4% paraformaldehyde, paraffin-embedded, sectioned, and subsequently stained with Tunel staining kit, sealed, and photographed for observation using a fluorescence microscope. In vivo drug metabolism in mice All animal experiments involved in this work were approved by the Ethics Committee (Anhui Medical University, Hefei, China). C57 male mice were injected with FA-Gd[2]O[3]@MSN-DOX, and then killed by cervical vertebral dissection before and 0.5 h, 6 h, 12 h, 24 h, 48 h, 72 h, 1 w, and 2 w after injection, and the heart, lungs, livers, kidneys, spleens and digestive tracts were dissected and scanned by IVIS^® Lumina Series III imaging system. Blood was collected at 3 and 7 days after injection for routine blood tests and blood biochemistry. Mouse subcutaneous tumor model Twenty 4-week-old BALB/Nude mice were divided into 4 groups, HepG2 cells were centrifuged and resuspended to adjust the concentration to 5 × 10^6 cell/mL, and each mouse was injected with 200 µL of cell suspension and when the tumor grew to 100 mm3, they were injected with FA-Gd[2]O[3]@MSN-DOX, DOX, FA-Gd[2]O[3]@MSN, and PBS respectively, and then they were executed after 1 week. Tumor photography and weighing were performed. The tumor tissues and heart, liver, spleen, lung, and kidney were subjected to HE staining. Mouse orthotopic liver cancer model Ten 4-week-old C57 mice were anesthetized by intraperitoneal injection of 3% sodium pentobarbital, cut along the midline of the abdomen to expose the liver, logarithmically grown Hepa1-6 cells were adjusted to a concentration of 5 × 10^6 cell/mL, centrifuged and 10 µL of matrix gel was added to resuspend the cells, and 20 µL of cell suspension was slowly injected into the left lobe of the liver by using a micro sampler until a white punctate protrusion was seen, and slowly retracted the needle. Silk suture was used to close the mice. After waiting for tumor formation, FA-Gd[2]O[3]@MSN-DOX was injected via the tail vein for treatment, and changes in signal intensity and tumor size were observed by magnetic resonance scanning. MRI scanning HepG2 cells and Hela cells were seeded in 6-well plates, and after the cells were attached to the wall, FA-Gd[2]O[3]@MSN and Gd-DTPA solutions containing Gd^3+ concentrations of 0.2 mM, 0.1 mM, 0.05 mM, 0.02 mM, 0.01 mM, 0.005 mM, and 0 mM were added and co-incubated for 48 h, addition of 2% agarose dispersed cells in 2 mL centrifuge tubes and scanned by MRI. Siemens 3.0T MRI machine, human shoulder coil, and conventional cross-sectional scanning. SET1WI sequence was used. Specific parameters: T1WI (TR/TE 500 ms/14 ms), echo chain length: 20–40, layer thickness: 2 mm, layer spacing: 2 mm; matrix: 384 256, FOV 14 cm x 14 cm, Nexus: 2 times. Mice were anesthetized with isoflurane, placed in the experimental chamber, fixed with tape, and scanned with Philips 3.0T magnetic resonance, mouse coil, and cross-sectional conventional scanning SET1WI and SET2WI sequences (TR/TE 800 ms/30 ms), layer thickness: 1 mm, FOV 50 mm x 60 mm. Sante DICOM Viewer Free software measured lesion size and signal intensity. Statistical analysis All data analyses were statistically analyzed using SPSS 26.0. Statistical graphs were plotted using GraphPad Prism 9. Values are expressed as Mean ± SD. and one-way ANOVA was used to evaluate significance, *P < 0.05; **P < 0.01 ***P < 0.001. Three independent experiments were conducted for all experiments. Results and discussion Characterization of MSNs FA-Gd[2]O[3]@MSN-DOX was synthesized according to literature methods [[67]25]. The successful preparation of naked MSN nano-particles was confirmed by the TEM observation and the diameter is approximately 50 nm (Fig. [68]S1). The pH-responsive polymer P(DMA-co-APMA) was synthesized via reversible addition-fragmentation chain transfer (RAFT) polymerization. The resulting amine groups were subsequently modified with THPA to form the negatively charged, pH-sensitive polymer P(DMA-co-TPAMA). To achieve folic acid targeting, P(DMA-co-APMA) was further modified with folic acid, yielding P(DMA-co-FAPMA). The successful synthesis of the three copolymers was confirmed by ^1H NMR analysis, with signal assignments for the corresponding protons. The molecular weights of P(DMA-co-APMA), P(DMA-co-TPAMA), and P(DMA-co-FAPMA) were 1.3 kDa, 2.7 kDa, and 3.1 kDa, respectively, with a polydispersity index (M[w]/M[n]) of 1.05 (Fig. [69]S2). After preparing the MSN and pH-responsive polymer, Gd₂O₃ nanoparticles were incorporated into the MSN structure to form Gd₂O₃@MSN nanoparticles. X-ray diffraction (XRD) analysis revealed peaks corresponding to Gd₂O₃, confirming the presence of Gd in the prepared Gd₂O₃@MSN (Fig. [70]S3A). To minimize premature release of encapsulated payloads, the surfaces of the MSNs were functionalized with APTES, resulting in the formation of amine-functionalized Gd₂O₃-loaded MSNs (denoted as Gd₂O₃@MSN-NH₂). Eventually, we loaded DOX onto the Gd[2]O[3]@MSN-NH[2] nanoparticles and further coated the nanoparticles with pH-responsive polymer and folic acid-containing polymer using a layer-by-layer assembly method (denoted as FA-Gd[2]O[3]@MSN-DOX). The successful preparation of FA-Gd[2]O[3]@MSN-DOX nano-particles was confirmed by the TEM (Fig. [71]1A). The peaks of Gd[2]O[3] were found in the energy dispersive spectroscopy (EDS) spectrum and XRD spectra (Fig. [72]1B-E and Fig. [73]S3B-C). The presence of amine groups on the surfaces of FA-Gd[2]O[3]@MSN-DOX was confirmed by FT-IR spectroscopy, as evidenced by the appearance of the broad N–H stretching vibration peak of the amine residues at 2900 cm^− 1 (Fig. [74]S3D). The Gd loading efficiency was determined to be 0.69% by ICP-MS (Fig. [75]1F), and the nanoparticle diameter was approximately 55 nm, as measured by DLS (Fig. [76]1G). Zeta potential analysis indicates that the surface of the prepared Gd₂O₃@MSN-NH₂ is positively charged. Following the layer-by-layer assembly method for coating with P(DMA-co-FAPMA), the surface potential shifts closer to neutral (Fig. [77]1H). The loading efficiency of DOX was 5.7% as measured by UV-vis spectroscopy (Fig. [78]S3E-F). We monitored the stability of FA-Gd[2]O[3]@MSN-DOX in DMEM, PBS, and water over the course of one week and found that there were no significant changes in its UV absorbance or diameter during this period (Fig. [79]S4). Fig. 1. [80]Fig. 1 [81]Open in a new tab Characterization of nanoparticles. (A-E) scanning TEM images of FA-Gd[2]O[3]@MSN-DOX and corresponding elemental mapping images. (F) Gd^3+ content measured by ICP-MS of FA-Gd[2]O[3]@MSN-DOX and MSN. (G) Intensity-average hydrodynamic diameter distributions of nanoparticles. (H) The zeta potential of nanoparticles. (I) DOX release from FA-Gd[2]O[3]@MSN-DOX nanoparticles at different pHs After fabricating FA-Gd₂O₃@MSN-DOX, the pH-triggered release of DOX was investigated. The fitted release curves show that the drug release rate at pH 6.0 (k[obs] = 9.09 h⁻¹) is faster than at pH 7.4 (k[obs] = 2.94 h⁻¹). After 48 h, the cumulative DOX release reaches 55% at pH 6.0, while at pH 7.4, the cumulative release is only 15%. These results suggest that DOX release is highly pH-dependent, due to the pH-triggered dissociation of the coated polyelectrolytes (Fig. [82]1I). In vitro cytotoxicity and cell apoptosis Cytotoxicity is a crucial consideration for in vivo biomaterial applications. The MTT assay was employed to assess the cytotoxicity of four different nanoparticles in HepG2 cells, HeLa cells, and Fibroblastic reticular cells. The DOX-unloaded nanomedicine demonstrated no significant fluctuation in cell survival with increasing concentration. Conversely, DOX-loaded Gd[2]O[3]@MSN exhibited a notable decrease in tumor cell survival with increasing concentration, indicating concentration-dependent cytotoxicity (Fig. [83]2A-B). Furthermore, we found that the hybrid materials exhibited much lower cytotoxicity against fibroblast reticular cells (FRC) at identical concentrations (Fig. [84]2C), which was beneficial for their biomedical applications with decreased systemic toxicity. Moreover, DOX-free Gd[2]O[3]@MSN demonstrated non-toxicity across all cell types, while DOX-loaded Gd[2]O[3]@MSN displayed concentration-dependent cytotoxicity. Additionally, FA-Gd[2]O[3]@MSN-DOX exhibited higher cytotoxicity than Gd[2]O[3]@MSN-DOX, likely attributed to the mediated action of folate receptors, resulting in enhanced cell internalization properties. Subsequently, apoptosis-related proteins were detected via WB assay, revealing a significant increase in the pro-apoptotic protein BAX expression in both the FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX groups, alongside a decrease in the expression of the anti-apoptotic protein Bcl-2. Cleaved caspase-3, recognized as a pivotal terminal cleavage enzyme in apoptosis, demonstrated elevated expression, indicating accelerated apoptosis when the upstream pathway is activated [[85]26]. Notably, Cleaved caspase-3 levels were significantly increased in both the FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX groups, underscoring the substantial increase in apoptosis within the drug-loaded group (Fig. [86]2D-H). Finally, apoptosis was quantified by flow cytometry, revealing that the FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX groups exhibited apoptosis rates of 87.21% and 83.7% in HepG2 cells, respectively. These rates significantly increased compared to the non-loaded group, consistent with the MTT results. This confirms that FA- Gd[2]O[3]@MSN-DOX demonstrates substantial cytotoxicity in tumor cells (Fig. [87]2I-J). Fig. 2. [88]Fig. 2 [89]Open in a new tab In vitro anti-cancer effects of the nanodrugs. (A-C) Cell viabilities of Hela cell, HepG2 cell and Fibroblastic reticular cell after being treated with Gd[2]O[3]@MSN, FA-Gd[2]O[3]@MSN, Gd[2]O[3]@MSN-DOX and FA-Gd[2]O[3]@MSN-DOX for 48 h at various concentrations. (D-H) Western blot assay shows the expression levels and quantification analysis of apoptosis-related proteins in HepG2 cell after different treatments for 48 h. (I) Statistical analysis of apoptosis for the flow cytometry analysis in different treatment groups. (J) Flow cytometry analysis shows the apoptosis of HepG2 cell after receiving different treatments for 48 h. (Data are shown as mean ± SD (n = 3); *P < 0.05; **P < 0.01 ***P < 0.001) In vitro cellular uptake and inhibition of autophagy Observing the distribution of nanomedicine within hepatocellular carcinoma cells via laser confocal microscopy, it was observed that free DOX accumulated in the nucleus, while FA-Gd[2]O[3]@MSN-DOX predominantly localized within lysosomes upon entry into hepatocellular carcinoma cells (Fig. [90]3A). However, laser confocal observation of DOX distribution in normal hepatocytes using the same treatment conditions showed that free DOX was still diffusely distributed, whereas the nanoparticle group showed a significant decrease in DOX fluorescence intensity in normal hepatocytes, which was attributed to folate targeting to tumour cells indicating folate receptors (Fig. [91]S5). This led to the dissociation of pH-sensitive polyelectrolytes and subsequent DOX release, thereby altering the acidic lysosomal environment and impairing lysosomal function. Lysosome-associated membrane protein (LAMP1) is a highly glycosylated membrane protein that is mainly enriched in lysosomal membranes and plays an important role in lysosomal biogenesis, PH regulation and autophagy. When lysosomes are disrupted, LAMP1 will decrease [[92]27]. CTSB is a lysosomally localised cysteine protein hydrolase active at ph 3.0–7.0 [[93]28]. WB experiments, assessing lysosome-associated membrane protein 1 (LAMP1) and cathepsin B (CTSB), revealed significant reductions in their levels within the FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX groups, confirming the disruption of lysosomal function following DOX release (Fig. [94]3B and D-E). The autophagy-lysosome pathway plays a critical role in the removal of intracellular protein aggregates [[95]29]. Lysosomal dysfunction effectively inhibits cellular autophagy, and research indicates that autophagy inhibition can lead to apoptosis or necrosis [[96]30]. Autophagy inhibition holds promise as a therapeutic strategy for cancer treatment across various tumor models [[97]31]. Previous studies have demonstrated that the accumulation of autophagic vesicles enhances the therapeutic efficacy of anticancer drugs. For instance, in advanced solid tumors [[98]32, [99]33], the accumulation of autophagic vesicles subsequent to treatment with the autophagy inhibitor chloroquine notably augmented the anticancer effects of histone deacetylase inhibitors. The microtubule-associated protein 1 A/1B Light chain 3B family (LC3A, LC3B, or GABARAP) undergoes lipidation during autophagy, facilitating its insertion into the membranes of autophagosomes and autophagic lysosomes [[100]34]. Various ligands can bind to LC3, with one of the most significant being sequestosome 1 (SQSTM1, also known as p62). SQSTM1 binds to LC3 via the LC3 interaction region and simultaneously recognizes (typically polyubiquitylated) proteins for degradation within autophagosomes or autophagic lysosomes modified by LC3. Consequently, an increase in autophagic flux is often accompanied by intracellular depletion of P62, whereas an elevation in P62 concentration is typically indicative of autophagy inhibition [[101]29]. Western blot assays revealed no significant difference in LC3II/LC3I and P62 levels among the control, Gd[2]O[3]@MSN, and FA-Gd[2]O[3]@MSN groups. However, a notable increase was observed in the Gd[2]O[3]@MSN-DOX and FA-Gd[2]O[3]@MSN-DOX groups, indicating blocked autophagic flow in the cells (Fig. [102]3B and F-G). In these groups, a plethora of autophagosomes accumulated within the cytoplasm of the cells, and electron microscopy illustrated deformation and partial fracture of the nuclear membranes (Fig. [103]3C). These observations align with the Western blot results, suggesting significant inhibition of autophagic flux in the drug-loaded group based on cellular morphology. A similar phenomenon was observed in HeLa cells under identical conditions (Fig. [104]S6). Fig. 3. [105]Fig. 3 [106]Open in a new tab Cellular uptake of the nanodrugs and the effect on the autophagy. (A) Confocal laser scanning microscopy (CLSM) images of HepG2 cells incubated with FA-Gd[2]O[3]@MSN-DOX, Gd[2]O[3]@MSN-DOX, and free DOX. Scale bars: 20 μm. (B) Western blot assay illustrates the expression levels of autophagy and lysosome-related proteins in HepG2 cells following different treatments for 48 h. (C) TEM image displays the effect on autophagy in HepG2 cells after various treatments for 48 h. (red arrows: autophagosome-like structures; yellow arrows: autolysosome-like structures) (D-G) Statistical analysis of autophagy and lysosome-related proteins based on the Western blot assay in different treatment groups. (Data are presented as mean ± SD (n = 3); *P < 0.05, **P < 0.01, and ***P < 0.001) Nanodrugs inhibit autophagy through mTOR signaling In order to investigate which pathway nanoparticles inhibit cellular autophagy flux, we treated cells with nanomedicines and then performed RNA sequencing. The statistical analysis of differential gene expression revealed a total of 902 up-regulated genes and 164 down-regulated genes in the FA-Gd[2]O[3]@MSN-DOX/Control group (Fig. [107]S7A). The Venn plot depicting differential gene expression demonstrated that 377 genes were specific to the FA-Gd[2]O[3]@MSN-DOX group (Fig. [108]S7B). KEGG Pathway enrichment analysis of the differential genes indicated significant enrichment in the mTOR signaling pathway (Fig. [109]4A). Furthermore, the clustering heatmap of mTOR pathway-related genes illustrated the activation of the core gene of this pathway, mTOR (Fig. [110]S8). In our subsequent Western blot experiments (Fig. [111]4C and F-G), we observed that autophagic flux was blocked in the FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX groups upon addition of the mTOR-specific inhibitor rapamycin. Conversely, the autophagy-associated protein P62 was significantly reduced, and autophagic flux resumed upon rapamycin addition. mTOR serves as a crucial regulator of cellular autophagy and plays a pivotal role in cell and cellular growth [[112]35]. In Western blot experiments, we noted that FA-Gd[2]O[3]@MSN-DOX induced noticeable phosphorylation of mTOR itself in hepatocellular carcinoma cells, while its upstream molecule Akt showed no significant phosphorylation, suggesting that the inhibition of cellular autophagic flux by FA-Gd[2]O[3]@MSN-DOX was linked to mTOR signaling rather than the Akt pathway (Fig. [113]4B and D-E). These findings collectively suggest that the mechanism underlying FA-Gd[2]O[3]@MSN-DOX-mediated autophagic flux blockade is closely associated with mTOR pathway activation. Fig. 4. [114]Fig. 4 [115]Open in a new tab Nanodrugs inhibit autophagy through the mTOR pathway. (A) KEGG Pathway Enrichment analysis reveals that the differential genes were predominantly enriched in the mTOR signaling pathway (FA-Gd[2]O[3]@MSN-DOX/Control). (B) Western blot assay illustrates the expression levels of mTOR pathway-related proteins in HepG2 cells following different treatments for 48 h. (C) Western blot assay depicts the expression levels of autophagy-related proteins in HepG2 cells after different treatments for 48 h. (D-G) Statistical analysis of the Western blot assay across different treatment groups (Data are presented as mean ± SD (n = 3); *P < 0.05, **P < 0.01, and ***P < 0.001) Nanogrugs regulate the subcellular distribution of TFEB via mTOR mTOR, as a pivotal regulator of cellular autophagy, can inhibit amino acid-dependent autophagy while stimulating protein translation and cell growth by modulating the downstream effector TFEB (transcription factor EB) [[116]36]. TFEB positively regulates the transcription of genes involved in various steps of lysosomal biogenesis, thereby promoting lysosomal proliferation, acidification, and extracellular secretion [[117]37]. Originally described as a transcriptional regulator of the lysosomal system, cytoplasmic nuclear translocation of TFEB has emerged as a major mechanism for regulating TFEB activity [[118]38]. In normal cells, TFEB is retained in the cytoplasm in a phosphorylated form, and under specific conditions, it can be dephosphorylated and translocated to the nucleus to initiate lysosomal biogenesis and the transcription of autophagy-related genes. Therefore, we initially investigated the effect of FA-Gd[2]O[3]@MSN-DOX on the phosphorylation of TFEB in hepatocellular carcinoma cells following treatment with FA-Gd[2]O[3]@MSN-DOX (Fig. [119]5A-B). Western blot analysis detected changes in TFEB protein expression levels, revealing increased phosphorylation of TFEB in the FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX groups. Subsequently, we employed nucleoplasmic separation techniques to assess TFEB protein enrichment in the cytoplasm and nucleus following cell treatment (Fig. [120]5C and E-F). Our analysis revealed a significant up-regulation of TFEB protein in the cytoplasm after treatment with FA-Gd[2]O[3]@MSN-DOX and Gd[2]O[3]@MSN-DOX. Conversely, the quantity of nuclear-localized TFEB proteins decreased, with nearly all TFEB proteins being enriched in the cytoplasm. Additionally, immunofluorescence staining analysis corroborated the Western blot results, indicating that FA-Gd[2]O[3]@MSN-DOX induced endogenous TFEB enrichment in the cytoplasm while inhibiting TFEB nuclear translocation (Fig. [121]5D). This inhibition led to the suppression of the autophagy-lysosomal pathway, resulting in autophagic flux blockade and ultimately achieving tumor cell death. Fig. 5. [122]Fig. 5 [123]Open in a new tab Nanogrugs regulate the subcellular distribution of TFEB via mTOR. (A) Western blot assay illustrating the expression levels of p-TFEB in HepG2 cells after different treatments for 48 h. (B) Statistical analysis of the Western blot assay across different treatment groups. (C) Western blot analysis of TFEB in cytoplasmic and nuclear fractions of HepG2 cells after different treatments for 48 h. LaminB1 serves as the control for the nuclear fractions, while GAPDH serves as the control for the cytoplasmic fractions. (D) Fluorescence microscopy image depicting the distribution of TFEB in HepG2 cells following various treatments for 48 h. FA-Gd[2]O[3]@MSN-DOX induces cytoplasmic enrichment of TFEB in HepG2 cells. (E-F) Statistical analysis of the Western blot assay across different treatment groups. (Data are presented as mean ± SD (n = 3); *P < 0.05, **P < 0.01, and ***P < 0.001) In vivo anti-cancer effect and biosafety of nanotubes In the mouse subcutaneous tumor model, FA- Gd[2]O[3]@MSN-DOX can be seen to be targeted and enriched in the tumor after drug injection, and the DOX fluorescence signal gradually enhances in the tumor with the change of time, and the fluorescence signal reaches the peak at 24 h, and then fades away gradually after 72 h (Fig. [124]S9). It is suggested that the FA-targeting and pH-responsiveness of FA- Gd[2]O[3]@MSN-DOX enable the nanoparticles to be targeted in tumor cells and stable drug release within the acidic environment of tumor cell lysosomes. To further study the anticancer effects of nanomedicines in animals, a subcutaneous tumor model was utilized. The timeline of tumour treatment was illustrated in Fig. [125]6A. The results indicated no significant difference in tumor volume and weight between the FA-Gd[2]O[3]@MSN group and the control group. However, the FA-Gd[2]O[3]@MSN-DOX group and the free DOX group exhibited a significant reduction in tumor volume and weight compared to both the control and FA-Gd[2]O[3]@MSN groups. Furthermore, the FA-Gd[2]O[3]@MSN-DOX group showed a more pronounced decrease in tumor volume and weight compared to the free DOX group (Fig. [126]6B and D-E). Tunel staining of tumor tissues revealed a significantly higher apoptosis rate in both the FA-Gd[2]O[3]@MSN-DOX and free DOX groups (Fig. [127]6C and F). Digital pathology section and laser confocal image of the tumour tissue similarly revealed extensive tumour tissue necrosis (Fig. [128]S14). These results indicate that FA-Gd[2]O[3]@MSN-DOX exhibits a strong tumor suppressor effect on tumor tissues. Furthermore, histopathological analysis of major organs (heart, liver, spleen, lungs, and kidneys) and tumors using HE staining showed no significant pathological changes in any organs following FA-Gd[2]O[3]@MSN-DOX injection (Fig. [129]6G). In the FA-Gd[2]O[3]@MSN-DOX group, tumor cells displayed extensive necrotic degeneration, cytoplasmic eosinophilia, and vacuolization, indicating that the nanomedicine effectively kills tumor cells. Blood routine and biochemical analyses conducted 3 and 7 days after FA-Gd[2]O[3]@MSN-DOX injection into the tail vein of mice showed no statistically significant differences in all indices compared to the control group, with values remaining within the normal range (Fig. [130]S10). This suggests that the nanomedicine is non-toxic and demonstrates good biological safety in mice. Fig. 6. [131]Fig. 6 [132]Open in a new tab In vivo anti-cancer effect and biosafety of nanodurgs. (A) Schematic illustration of the in vivo treatment process. (B) Images of tumors after different treatments for 7 days. (C) TUNEL immunofluorescence staining of tumor sections following treatment in different groups. (Scale bar: 100 μm). (D-E) Changes in tumor volume and tumor weight within 7 days after different treatments. (F) Statistical analysis of the TUNEL positive rate in different treatment groups. (G) H&E staining of the heart, liver, spleen, lungs, and kidneys. (Data are shown as mean ± SD (n = 3); *P < 0.05, **P < 0.01, and ***P < 0.001) In vivo drug distribution studies and MRI imaging To study the in vivo distribution of the drug, we injected FA-Gd[2]O[3]@MSN-DOX via the tail vein and subsequently excised the major organs (heart, liver, spleen, lungs, and kidneys) for fluorescence imaging. The imaging results revealed that the drug was primarily distributed in the lungs, liver, and kidneys, with nearly complete clearance by the liver and kidneys within 72 h (Fig. [133]S11). MRI is a powerful diagnostic technique that provides precise anatomical information. Paramagnetic Gd[2]O[3] nanoparticles have been previously used as MRI contrast agents, exhibiting higher relaxivity than commercially available MRI contrast agents [[134]39, [135]40]. First, we assessed the T1-enhancing effect of FA-Gd₂O₃@MSN-DOX under different pH conditions through in vitro experiments. Various concentrations of FA-Gd₂O₃@MSN-DOX were dispersed in PBS solutions with pH values of 6.0 and 7.4, respectively, followed by MRI scans. The results showed that at pH 6.0, the T1 signal value progressively increased with concentration, while at pH 7.4, the T1 signal remained unchanged with varying concentrations. This suggests that the pH-responsive polymer on the outer surface of FA-Gd₂O₃@MSN-DOX can dissociate in acidic environments (Fig. [136]S12). To further evaluate its enhancement effect in cells and within living tumors, we treated HepG2 cells with different concentrations of FA-Gd[2]O[3]@MSN and Gd-DTPA. FA-Gd[2]O[3]@MSN demonstrated the highest signal-to-noise ratio, which continued to increase with concentration, indicating a concentration dependence (Fig. [137]7A-B). The same phenomenon was observed in Hela cells under similar conditions (Fig. [138]S13). These results all suggest that FA-Gd[2]O[3]@MSN-DOX can be a good MRI contrast agent and exhibits better visualisation than the existing MRI contrast agents. In a subcutaneous tumor model in living animals, after intratumoral injection of the drug, the signal intensity of the lesion continued to increase over one week, while the size of the tumor lesion decreased. This demonstrated good intratumoral diffusion and prolonged retention in the tumor cells, resulting in a long-lasting MRI enhancement effect (Fig. [139]7C and E). In the orthotopic liver cancer model, the liver cancer lesion exhibited the highest T1 signal intensity 6 h after injection via the tail vein, with continued enhancement over time (Fig. [140]7D and F). These results confirm that nanomedicines can remain within tumor cells for extended periods, targeting and continuously killing the tumor [[141]41]. Combined with the in vivo drug distribution study, it is evident that FA-Gd[2]O[3]@MSN-DOX can serve as a contrast agent for MR imaging of targeted tumors. This enables real-time monitoring and assessment of treatment efficacy while simultaneously treating tumors. However, nephrotoxicity remains a significant concern with all GBCAs. While our experiments have demonstrated that FA-Gd₂O₃@MSN-DOX shows favorable biosafety in mice, the potential for Gd-induced NSF being a rare and delayed complication necessitates longer follow-up studies to further confirm the safety and scientific reliability of this platform in future research. Fig. 7. [142]Fig. 7 [143]Open in a new tab In vivo and in vitro T1-weighted MRI enhancement effects of nanodrugs. (A-B) T1-weighted imaging (T1WI) signal intensity and signal-to-noise ratio (SNR) of HepG2 cells treated with FA-Gd[2]O[3]@MSN and Gd-DTPA for 48 h at various concentrations. (C) Changes in MRI intensity and tumor volume of subcutaneous tumors in nude mice after intratumoral injection of FA-Gd[2]O[3]@MSN-DOX for 7 days. Scale bars: 5 mm. (D) Changes in MRI intensity and tumor volume of orthotopic hepatic carcinoma in C57 mice after intravenous administration of FA-Gd[2]O[3]@MSN-DOX for 7 days. Scale bars: 5 mm. (E-F) Statistical analysis of tumor volume and MRI intensity at different treatment times. (Data are presented as mean ± SD (n = 3); *P < 0.05, **P < 0.01, and ***P < 0.001) Conclusions FA-Gd[2]O[3]@MSN-DOX is a diagnostic-integrated drug-carrying nano platform that combines MRI imaging with DOX chemotherapy. The results demonstrated that FA-Gd[2]O[3]@MSN-DOX can affect lysosomal pH, disrupt lysosomal function, inhibit autophagic flux via the mTOR-TFEB pathway, and ultimately induce apoptosis in hepatocellular carcinoma cells, achieving tumor suppression. It has shown significant therapeutic effects both in vitro and in vivo, and its Gd[2]O[3] content enables real-time detection of therapeutic effects through MRI imaging during treatment. We believe that this platform can be a promising approach in the diagnosis and treatment of liver cancer. Electronic supplementary material Below is the link to the electronic supplementary material. [144]Supplementary Material 1^ (11.5MB, docx) Acknowledgements