Nanomaterials for full-thickness cartilage real-time mr imaging and preparation method and application thereof

Abstract

This application relates to the biomedical field and discloses a nanomaterial for real-time MR imaging of full-thickness cartilage, its preparation method, and its application. This stepwise targeted nanomaterial comprises a methacrylamide hyaluronic acid hydrogel matrix; and Fe dispersed within the hydrogel matrix. 3+ -PKP complex; the PKP comprises an octamercaptopropyl polyhedral oligomeric silsesquioxane backbone and functional molecules grafted onto the backbone; the functional molecules include PEG, KGN, and EDA. The PKPF@HAMA of this application synergistically delivers iron ions and KGN, achieving precise enrichment from the superficial cartilage layer to the calcified cartilage layer. By driving lipid metabolism reprogramming in the calcified cartilage layer, it overcomes the limitations of traditional superficial signal enhancement, providing a new strategy for the diagnosis, efficacy evaluation, and precise repair of full-thickness cartilage lesions.

Description

Technical Field

[0001] This application relates to the field of biomedicine, and more specifically, to a nanomaterial for real-time MR imaging of full-thickness cartilage, its preparation method, and its application. Background Technology

[0002] Cartilage is a highly layered, avascular, and sparsely cellular specialized tissue. Its structure and composition vary significantly from the superficial and transitional layers to the deep calcified cartilage layer, playing an irreplaceable role in joint lubrication, biomechanical transmission, and the stability of the osteochondral interface. The calcified cartilage layer, in particular, with its low metabolic activity, high matrix mineralization, and limited substance exchange, is a key pathological initiation zone for cartilage degeneration and osteoarthritis. However, due to the unique structure, permeability, and composition of this region, current imaging techniques struggle to provide accurate, non-invasive, spatially resolved, and function-related imaging, leaving changes in the entire cartilage layer, especially the calcified layer, in a long-standing "blind spot" for clinical assessment.

[0003] MRI is widely used for soft tissue assessment due to its advantages of high soft tissue resolution and radiation-free operation. However, MRI's imaging capability of calcified cartilage layers is severely limited by the physicochemical properties of contrast agents and the metabolic characteristics of the cartilage itself. Traditional contrast agents mostly rely on diffusion or electrostatic interactions to enter cartilage tissue, but they struggle to penetrate the dense and highly ordered deep matrix. Even if they can penetrate deeper layers, the low activity of iron ion metabolism in cartilage tissue cannot provide a sufficient coordination environment, allowing exogenous iron ions to diffuse rapidly and fail to be fixed, making it difficult to form stable iron coordination structures sufficient to alter the local magnetic environment. Therefore, iron ions are difficult to deliver effectively or utilize efficiently in deep cartilage, resulting in significantly limited deep signal strength, which has become a fundamental bottleneck restricting full-thickness cartilage MRI imaging.

[0004] To address the imaging challenges of calcified cartilage layers, previous studies have attempted to improve delivery efficiency by increasing contrast agent particle size, enhancing charge density, or introducing targeting strategies specific to cartilage matrix components. However, these methods all suffer from inherent contradictions: enhanced penetration often reduces targeting accuracy and residence time, while improved targeting significantly increases resistance, hindering deep diffusion. More importantly, these strategies only focus on the "distribution of the contrast agent" and do not address the core problem of "low iron metabolism activity and inability to form a stable MRI signal source" in calcified cartilage layers. Therefore, to overcome the current technical limitations of MRI, simply optimizing the physicochemical properties of exogenous contrast agents is insufficient for effective imaging from superficial to deep layers; it is necessary to activate the utilization of iron ions by deep cartilage at the internal metabolic level. Based on this understanding, a crucial but long-standing scientific question arises: can metabolic reprogramming be induced in the calcified cartilage layer to create a stable coordination environment capable of actively utilizing iron ions, thereby providing readable MRI signals? This concept elevates MRI from the traditional "external contrast agent-dependent mode" to an "internal metabolism-driven mode," but its realization highly depends on two technical prerequisites: (1) achieving deep and precise delivery of iron ions and metabolic regulatory molecules; and (2) successfully reshaping lipid metabolism and iron metabolism pathways in the calcified cartilage layer to improve the fixation, conversion, and coordination efficiency of iron ions. Therefore, the field needs a multifunctional material system that can simultaneously overcome structural barriers and drive metabolic changes. Summary of the Invention

[0005] The purpose of this invention is to simultaneously address the two long-standing technical problems hindering cartilage MRI imaging: difficulty in deep delivery and weak metabolic activation. To this end, this application provides a nanomaterial for real-time MR imaging of full-thickness cartilage, its preparation method, and its application.

[0006] To achieve the above-mentioned objectives, this application provides the following technical solutions:

[0007] In a first aspect, this application provides a nanomaterial for real-time MR imaging of full-thickness cartilage, comprising: Methacrylamide hyaluronic acid hydrogel matrix; and Fe dispersed within the hydrogel matrix. 3+ -PKP complexes; The PKP comprises an octamercaptopropyl polyhedral oligomeric silsesquioxane backbone and functional molecules grafted onto the backbone. The functional molecules include PEG, KGN, and EDA.

[0008] Secondly, this application provides a method for preparing the nanomaterials for real-time MR imaging of full-thickness cartilage, comprising the following steps: Step 1: Acrylamide treatment of PEG, KGN and EDA; Step 2, prepare PKP by grafting acrylated PEG, KGN and EDA onto an octamercaptopropyl polyhedral oligomeric silsesquioxane backbone to obtain PKP; Step 3: Add FeCl3 solution to the PKP solution to allow iron ions to coordinate with PKP and form Fe. 3+ -PKP complex, named PKPF; Step 4: Prepare methacrylamide hyaluronic acid, named HAMA; Step 5: Mix the PKPF obtained in Step 3 and the HAMA obtained in Step 4 as the aqueous phase, and use an oily medium and emulsifier as the oil phase. Prepare the nanomaterial by microfluidic interface solidification method and name it PKPF@HAMA.

[0009] Further, in step 1, the acrylation treatment step of PEG, KGN, and EDA is as follows: PEG was dissolved in dichloromethane, triethylamine was added, and acryloyl chloride was slowly added dropwise under ice bath conditions. The reaction system was stirred at 0°C. After the reaction was completed, the reaction solution was washed, dried, filtered, and evaporated to dryness. KGN was dissolved in dichloromethane, triethylamine was added, and acryloyl chloride was slowly added dropwise under ice bath conditions while stirring the reaction. After the reaction was completed, the mixture was washed, dried, filtered, and evaporated to dryness. EDA was dissolved in DCM, triethylamine was added, and acryloyl chloride was slowly added dropwise under ice bath conditions. The reaction was stirred and the mixture was washed, dried, filtered, and evaporated to dryness.

[0010] Furthermore, in step 2, the molar ratio of octamercaptopropyl polyhedral oligomeric silsesquioxane, PEG, KGN, and EDA is 1:4:0.25:3.

[0011] Furthermore, step 2 consists of the following steps: Octadecylpropyl polyhedral oligomeric silsesquioxane was dissolved in dry tetrahydrofuran and stirred at room temperature. Acrylamide-treated KGN, PEG and EDA were then added sequentially, followed by the addition of photoinitiator DMPA. The reaction was stirred at room temperature. After the reaction was completed, dialysis was performed to remove unreacted small molecule impurities to obtain the PKP.

[0012] Furthermore, step 3 consists of the following steps: PKP was dissolved in deionized water to prepare a homogeneous solution, then FeCl3 solution was added and stirring was continued to produce Fe... 3+ -PKP complex, named PKPF.

[0013] Furthermore, step 4 is as follows: Hyaluronic acid was dissolved in preheated PBS solution and stirred continuously until a clear and homogeneous solution was formed. Then, methacrylic anhydride was slowly added dropwise, and the pH of the reaction system was maintained at 8.0-8.5. The reaction system temperature was controlled at 50°C and the reaction was continued. The reaction was terminated with PBS. Then, after dialyzing, freezing and drying, the methacrylamide hyaluronic acid was obtained and named HAMA.

[0014] Furthermore, step 5 is as follows: PKPF and HAMA were dissolved in PBS solution and stirred to form a homogeneous and stable aqueous phase. Paraffin oil is mixed with an emulsifier to form an oil phase; During the microfluidic preparation process, the aqueous phase and oil phase are injected through the internal and external phase channels of the microfluidic chip, respectively, where the internal phase is the aqueous phase and the external phase is the oil phase. The flow rates are set to 0.8 mL·h for the aqueous phase. -1 oil phase 16 mL·h -1 The generated droplets further cross-linked and solidified at 80°C to form microspheres. The obtained microspheres were collected by centrifugation and washed repeatedly with petroleum ether and deionized water to finally obtain the nanomaterial, named PKPF@HAMA.

[0015] Thirdly, this application provides the application of the nanomaterials described above for real-time MR imaging of full-thickness cartilage in the preparation of products for early screening, diagnosis, efficacy evaluation, or cartilage repair monitoring of osteoarthritis.

[0016] In summary, this application has the following beneficial effects: This application is the first to construct an ion-molecule hierarchical targeting nanoplatform (PKP) based on POSS / KGN. PKP can highly specifically chelate iron ions to form positively charged PKPF, which then targets negatively charged components such as chondroitin sulfate in the cartilage matrix. Furthermore, using microfluidic technology, methacrylated hyaluronic acid hydrogel microspheres loaded with PKPF (PKPF@HAMA) were prepared for in-situ intra-articular injection. PKPF@HAMA synergistically delivers iron ions and KGN, achieving precise enrichment from the superficial cartilage layer to the calcified cartilage layer, with the enrichment of iron ions in the deeper layers reaching 2.3 times that of the surface layer. The delivered KGN induces lipid metabolism reprogramming in the calcified layer, not only promoting cartilage differentiation of BMSCs but also enhancing iron transmembrane transport by upregulating proteins such as CLC-2 and ATPase. Simultaneously, the upregulation of Aco1 protein indicates that iron ions are efficiently utilized, resulting in abundant Fe. 3+The coordinated environment significantly enhances MRI signals, achieving a breakthrough in layered imaging of full-thickness cartilage. Combined with artificial intelligence segmentation algorithms, this application non-invasively quantifies morphological parameters such as thickness, specific surface area, and volume of full-thickness cartilage in real time. By driving the reprogramming of lipid metabolism in the calcified cartilage layer, this application overcomes the limitations of traditional superficial signal enhancement, providing a new strategy for the diagnosis, efficacy evaluation, and precise repair of full-thickness cartilage lesions. Attached Figure Description

[0017] Figure 1 Synthesis, characterization, and molecular computational simulation of PKPF. (a) Schematic diagram of one-step synthesis of PKP via click chemistry. (b) Relationship between POSS, KGN, and PKP. 1 1H NMR spectrum, DMSO-d6 (600 MHz). (c) POSS, KGN, PKP and Fe 3+ UV-Vis absorption spectrum. (d) PKP and Fe 3+ UV-Vis absorption spectra of mixtures with different equivalents (0.25–1.75 eq). (e) Nanoparticle sizes of POSS-PEG and PKPF, with average particle sizes of 143 nm and 221 nm, respectively. (f) Zeta potential determination of POSS-PEG and PKPF. (g) DFT theoretical calculations of PKPF and Fe. 3+ Spin density variation. (h) DFT calculation of PKP-Fe binding energy. (i) DFT calculation of binding energies of different composite systems PKPF-CS, PKPF-KS, and PKPF-CS / KS. (j) HOMO and LUMO distributions and energy differences of PKP-Fe. (k) Molecular electrostatic potential distribution of PKPF, with red indicating low potential and blue indicating high potential.

[0018] Figure 2 : MRI visualization of PKPF@HAMA targeting calcified cartilage layer. (a) Morphological characterization of the material, including SEM images and EDX elemental distribution maps (Si, S, Fe). (b) In vitro cumulative release curve (grey) and degradation curve (red) of the nanomaterial over 8 weeks. (c) Pore size distribution curve obtained from nitrogen adsorption-desorption test. (d) T1WI and T2WI MRI images at different Fe concentrations (0–0.09 mg / mL). (e, f) Fitting curves of R1 (1 / T1) and R2 (1 / T2) relaxation rates as a function of Fe concentration. (g) Top: MRI monitoring of the delivery capacity of nanomaterials at different concentrations (0, 0.05, 0.1 mg / mL) to the calcified cartilage layer at 2 h; Bottom: MRI monitoring of the dynamic changes in the delivery of the material to the calcified cartilage layer at different time points (0–4 h) with a fixed concentration of 0.05 mg / mL. (h) Schematic diagram of the structure of the cartilage surface, intermediate layer and deep layer, showing the stepwise targeted delivery pathway of the nanomaterial. (i) Fe in cartilage at different time points 3+ICP-MS quantitative results of content. Quantitative analysis results of (j,k)G plots, showing the changes in CNR between the superficial-intermediate layer (Sup–Int) and the deep-intermediate layer (Deep–Int) of the material under different concentrations (j) and time (k) conditions.

[0019] Figure 3 Biocompatibility and chondrogenic induction properties of PKPF@HAMA. (a) Representative fluorescence microscopy images of BMSCs after incubation with culture medium, HAMA, PKPF@HAMA, PKP, or PKPF (0.05 mg / ml) for 24, 36, and 48 hours. Scale bar 500 μm. (b) Immunofluorescence staining of Aggrecan, Collagen I, Collagen II, and MMP13 in BMSCs after incubation with culture medium, PKPF@HAMA, PKP@HAMA, PKPF, or TGF-β (0.05 mg / ml) for 24 hours. Scale bar 50 μm. (c) CCK-8 cell viability (OD value) at 24, 36, and 48 hours (n=3 per group). (d) Viable cell counts at different time points under different treatments (n=3 per group). (e) Quantitative analysis of fluorescence intensity of ACAN, COLI, COLII, and MMP13 in Figure b (n=3 per group). (f) Viable cell counts at 24, 36, and 48 h of different concentrations (0.05, 0.07, 0.1 mg / mL) of PKPF@HAMA (n=3 per group). (g) Representative laser confocal microscopy images of chondrocyte clusters of BMSCs after TGF-β or PKPF@HAMA induction. Scale bar 200 μm. (h) Gene expression levels of BMSCs in each group were analyzed by real-time quantitative PCR (qPCR). Genes detected included ACAN, COLI, COLII, and MMP13. The expression of all genes was normalized using GAPDH (or β-actin) as an internal reference gene, and relative quantification was performed using the 2^-ΔΔCt method (n=3 per group). Data are expressed as mean ± standard deviation. Statistical analysis was performed using one-way ANOVA and Tukey's multiple comparison test. P<0.05, P<0.01, P<0.001 and P<0.0001.

[0020] Figure 4: PKPF@HAMA Enhanced Cartilage MRI Imaging and its Promotion of Repair. (a) High-field T2WI images of knee cartilage before injection and at 3 days, 1 week, 2 weeks, and 4 weeks after injection in the PKPF@HAMA, PKPF, PKP@HAMA, HAMA, or PBS groups. (b) High-field T2WI images of knee cartilage at 1, 2, 3, and 4 months after PKPF@HAMA or PBS intervention. (c) Curves of contrast-to-noise ratio (CNR) of cartilage relative to subchondral bone over time in different treatment groups based on Figure a (n=3 per group). (d–f) Quantitative MRI analysis based on Figure b, including curves of cartilage thickness (d), cartilage area (e), and cartilage volume (f) over time (n=3 per group).

[0021] Figure 5 : Automatic cartilage segmentation and assessment of cartilage repair based on artificial intelligence after PKPF@HAMA intervention. (a) Automatic segmentation process of articular cartilage based on deep learning model: First, the raw MRI data is preprocessed and then input into the U-Net architecture for automatic segmentation of the cartilage region. Then, boundary extraction and restrictive dilatation (S) are performed. oute ) and surface closure (S inne Algorithms such as [list of algorithms] were used to complete the three-dimensional reconstruction of cartilage, and quantitative parameters such as cartilage surface area, volume, and thickness were automatically calculated. (b) The degree of agreement between model segmentation and manual segmentation: left is the model prediction result (red), middle is the manual annotation (green), and right is the overlap between the two (yellow), visually demonstrating the degree of agreement between model segmentation and manual segmentation. (c) Three-dimensional structural images of cartilage obtained after automatic segmentation and three-dimensional reconstruction of the knee joint of OA rats and normal rats using MRI imaging. (d–h) Quantitative parameters of cartilage in different intervention groups obtained based on artificial intelligence automatic segmentation (n=3 per group). (d) Maximum cartilage thickness, (r) average cartilage thickness, (f) Dice coefficient of segmentation model performance, (g) cartilage area, and (h) cartilage volume. Data are expressed as mean ± standard deviation. Statistical analysis was performed using one-way ANOVA and Tukey multiple comparison test. P<0.05, P<0.01, P<0.001 and P<0.0001.

[0022] Figure 6Multidimensional efficacy assessment of knee joint after PKPF@HAMA intervention. (a) Three-dimensional reconstruction images (top) and coronal CT images (bottom) of the knee joint after PKPF@HAMA, PKPF, PKP@HAMA, HAMA, or PBS intervention. (b) Quantitative analysis of osteophyte volume based on CT images. (c) Pathological staining / immunomarking results of knee cartilage in different groups, including Alcian blue staining, Safranin O-Fix Green staining, Toluidine Blue staining, MMP-13 immunohistochemistry, RUNX2 immunohistochemistry, and type II collagen immunofluorescence (Scalebar=100μm). (d–f) Quantitative analysis results of each group based on Figure C (n=3 per group). (d) Comparison of cartilage thickness, (e) Comparison of OARSI scores for OA severity, (f) Comparison of relative expression levels of type II collagen. Data are expressed as mean ± standard deviation. Statistical analysis was performed using one-way ANOVA and Tukey multiple comparison test. P<0.0001. Detailed Implementation

[0023] The technical solutions and effects of this application will be further described in detail below with reference to the embodiments and accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining the invention and are not intended to limit the invention.

[0024] Example I. Materials and Methods 1. Experimental Materials POSS-SH: Octadecylpropyl polysiloxane (POSS-SH, laboratory-made, purity ≥95%) PEG400: Polyethylene glycol 400 (PEG, MW 400, Sigma-Aldrich, ≥98%). KGN: Kartogenin (KGN, MCE, ≥99%); EDA: Ethylenediamine (EDA, Sigma-Aldrich, ≥99%) Acryloyl chloride: Acryloyl chloride (Acryloyl chloride, Adamas, ≥98%); DCM: Dry dichloromethane (DCM, Greagent, ≥99.9%). TEA: Triethylamine (TFA, Sigma-Aldrich, ≥99.5%) Anhydrous sodium sulfate (Sigma-Aldrich, ≥99%) DMPA: 2,2-Dimethyl-1-(4-methylphenyl)-1-hydroxyethyl ketone (DMPA, Sigma-Aldrich, ≥99%). THF: Tetrahydrofuran (THF, Sigma-Aldrich, ≥99.9%) FeCl3: Ferric chloride (FeCl3, Sigma-Aldrich, ≥98%) HA (Hyaluronic acid; Sigma-Aldrich, ≥95%); LAP photoinitiator (Lithium phenyl-2,4,6-trimethylbenzoylphosphinate, Aladdin, ≥98%) FBS: Fetal bovine serum, Sigma-Aldrich; DMEM / F-12 medium, containing antibiotics; PBS buffer, pH 7.4, Gibco; The trypsin digestion solution (0.25%, containing EDTA and phenol red) was used. All reagents did not require further purification before use. The water used for the reaction and the organic solvents were all of analytical grade.

[0025] 2. Combination of PKPF Synthesis of polyethylene glycol 400-acrylate (PEG400-Acrylate): Polyethylene glycol 400 (PEG400, 5.00 g, 12.5 mmol) was dissolved in 50 mL of dry dichloromethane (DCM), and a catalytic amount of triethylamine (TEA, 0.25 g, 5 wt%) was added as an acid-binding agent. Acryloyl chloride (Acryloyl chloride, 1.5 g, 13.8 mmol, 1.1 wt%) was slowly added dropwise under ice bath conditions (0 °C) to control the reaction rate and avoid side reactions. The reaction was carried out with stirring for 1–2 hours to ensure that the hydroxyl groups of PEG400 fully esterified with acryloyl chloride. After the reaction was completed, the system was washed several times with deionized water to remove unreacted acryloyl chloride and the generated salt impurities. The organic phase was then extracted, dried over anhydrous sodium sulfate, and the solvent was removed by rotary evaporation to obtain a yellow or transparent acrylated PEG400 product with a yield of 95%.

[0026] Synthesis of Kartogenin Acrylate (KGN-Acrylate): Kartogenin (KGN, 0.5 g, 1.7 mmol) was dissolved in dry dichloromethane (DCM, 25 mL), and a suitable amount of triethylamine (TEA, 0.025 g, 5 wt%) was added as a basic catalyst. Acryloyl chloride (0.19 g, 2.1 mmol, 1.2 wt%) was slowly added dropwise under ice bath conditions (0°C) to control the reaction rate and avoid side reactions. The reaction was carried out with stirring for 1–2 hours to ensure that the active functional groups on the KGN molecule fully esterified with acryloyl chloride. After the reaction was complete, the system was repeatedly washed with deionized water to remove unreacted acryloyl chloride and the generated salt impurities. The organic phase was then extracted, dried with anhydrous sodium sulfate, and the solvent was removed by rotary evaporation to obtain the acrylated Kartogenin product in 92% yield.

[0027] Synthesis of ethylenediamine acrylate (EDA-acrylate): Ethylenediamine (EDA, 0.93 g, 15.5 mmol) was dissolved in dry DCM (50 mL), and triethylamine (TEA, 0.047 g, 5 wt%) was added. Acryloyl chloride (1.58 g, 17.5 mmol, 1.13 wt%) was slowly added dropwise under ice bath conditions, and the reaction was stirred for 1–2 hours to ensure that the two amino groups on the EDA molecule were as completely acrylamide as possible. After the reaction was complete, the organic phase was washed with deionized water, dried, and rotary evaporated to obtain the EDA-Acrylate product.

[0028] Synthesis of octamercaptopropyl polysiloxane (POSS-SH): 20 mL of γ-mercaptopropyltrimethoxysilane (KH590), 40 mL of methanol, and 20 mL of concentrated hydrochloric acid were added to a 250 mL three-necked flask equipped with a magnetic stirrer and a spherical condenser. The mixture was heated to 90 °C with stirring and reacted for 24 hours. Heating was stopped when the solution changed from colorless to milky white, and the mixture was allowed to cool to room temperature for at least 6 hours. At this point, a white viscous substance precipitated at the bottom of the flask. The supernatant was discarded, and the product was washed three times with 30 mL of methanol. Subsequently, 60 mL of dichloromethane was added to fully dissolve the product, and the mixture was extracted four times with 10 mL × 4 saturated saline solution. The lower milky white solution was collected, and an appropriate amount of anhydrous magnesium sulfate was added and dried overnight. After removing the magnesium sulfate by filtration, the filtrate was evaporated by rotary evaporation to remove the dichloromethane, finally yielding a colorless, viscous mercaptocage-type semisiloxane (SH-POSS).

[0029] Synthesis of the target hybrid molecule PKP: Octadecylpropyl POSS (POSS-SH, 1 mmol, 1.01 g) was dissolved in dry tetrahydrofuran (THF, 20 mL) and stirred at room temperature. Acrylamide molecules were added sequentially according to the designed ratio: KGN-Acrylate (0.25 mmol, 0.073 g), PEG400-Acrylate (4.0 mmol, 1.674 g), and EDA-Acrylate (3.0 mmol, 0.27 g), followed by the photoinitiator DMPA (0.1 mmol, 0.023 g). The reaction was stirred at room temperature for 2 hours, and the small molecule was polymerized with POSS in a one-pot click polymerization via photoinitiated radical addition to obtain the functionalized POSS hybrid complex PKP. After the reaction was completed, the system was dialyzed (MWCO 1000 Da) to remove unreacted small molecule impurities. The dialyzing was repeated for 48 hours with multiple aqueous replacements, and finally lyophilized to obtain the yellow solid product PKP.

[0030] The functionalized POSS complex PKP was dissolved in deionized water to prepare a homogeneous solution of 1 mg / mL. Different equivalents of ferric chloride (FeCl3) solution were added dropwise with gentle stirring to form complexes with the coordination sites in the PKP molecule. The amount of FeCl3 used ranged from 0.25 to 1.75 equivalents, relative to the complexable carboxyl, phenolic hydroxyl, or amino sites in the PKP molecule. After adding FeCl3, gentle stirring was continued for 30 minutes to ensure sufficient coordination of iron ions with the PKP molecule to form stable Fe... 3+ -PKP complex (PKPF). The resulting PKPF is a clear, homogeneous solution that can be used for subsequent experiments. This complexation reaction is based on the strong coordination of iron ions with polyhydroxy, polyamino, or carboxyl groups, which maintains a uniform distribution of drug molecules or functional groups in the PKP molecule and enhances the stability and structural integrity of the complex through metal-ligand interactions.

[0031] 3. Synthesis of HAMA 1.0 g of hyaluronic acid (HA) was dissolved in 10 mL of PBS solution preheated to 60°C to a concentration of 10% (w / v), and the solution was stirred continuously until a clear and homogeneous solution was formed. Then, 0.8 mL of methacrylic anhydride (MA) was slowly added dropwise, and the pH of the reaction system was maintained at 8.0–8.5 using NaOH solution. The reaction temperature was controlled at 50°C, and the reaction was continued for 1 h to achieve methacrylylation modification of the hydroxyl groups in the HA molecule. The reaction was stopped five times with PBS diluent, and then dialyzed at 40°C for one week to filter impurities (14 kDa cutoff analytical tube). The HAMA aqueous solution was freeze-dried, producing a white, milky white, porous foam.

[0032] 4. Synthesis of PKPF@HAMA An aqueous phase (4 wt% HAMA, 1 wt% PKPF uniformly mixed in PBS containing 0.5 wt% light stabilizer) and an oil phase (5 wt% SPAN 80 in paraffin oil) were introduced into a microfluidic apparatus. The flow rates of the aqueous and oil phases were controlled by a syringe connected to a syringe pump. The resulting monodisperse emulsion was then optically crosslinked under ultraviolet light. The resulting microspheres were collected in test tubes, and isopropanol was added. After shaking and washing, the microspheres were collected by centrifugation at 4000 rpm.

[0033] During the microfluidic preparation process, the aqueous phase and oil phase are injected through the internal and external phase channels of the microfluidic chip, respectively, where the internal phase is the aqueous phase and the external phase is the oil phase. The flow rates are set to 0.8 mL·h for the aqueous phase. -1 oil phase 16 mL·h -1 .

[0034] 5. Material Characterization 1 H nuclear magnetic resonance (NMR) 1 H NMR: Chemical shifts of the intermediate and target product PKP were determined in DMSO-d6 solution using a Bruker Avance III 600 MHz NMR spectrometer. Comparison of chemical shifts and peak shapes confirmed the successful grafting of the acrylonitrified small molecule onto the POSS molecular framework and verified the successful preparation of the final material. UV-Vis: The light absorption characteristics of POSS, KGN, PKP, and PKPF solutions were measured using a UV-3600plus spectrometer. Sample concentrations were approximately 0.1 mg / mL, and quartz cuvettes (1 cm path length) were used. The scanning wavelength range was 200–600 nm. The Fe content was evaluated by comparing the absorption peak changes of PKP and PKPF. 3+ The effect of complexation on the light absorption of the system was investigated, and complex formation was confirmed. The particle size distribution and surface charge of PKP and PKPF were determined using a Malvern Zetasizer Nano ZS. The sample concentration was 0.2 mg / mL, and measurements were performed at room temperature. DLS measurements provided the average hydrodynamic diameter and particle size distribution of the complexes in the aqueous phase; Zeta potential measurements reflected the surface charge and system stability. DFT theoretical calculations: Molecular models of PKP-Fe, PKPF-CS, PKPF-KS, and the complex system were constructed using the D Mol3 module of Materials Studio software. Optimization calculations were performed using the B3LYP / 6-31G(d,p) basis set to analyze the binding energy (Eb), HOMO-LUMO energy level distribution, and molecular charge distribution, theoretically verifying the Fe... 3+The coordination sites and complex stability provide computational basis for experimental characterization. A molecular model of the PKPF complex system was constructed using the CASTEP module of Materials Studio software, and the changes in the magnetic properties of the system (mainly spin integral density) were calculated and analyzed. The structure and properties of PKPF@HAMA microspheres were systematically analyzed using various characterization methods. First, the surface and cross-sectional morphology of the microspheres were observed using a field emission scanning electron microscope (Zeiss GeminiSEM 500), and the mapping distribution of elements such as Fe, Si, and S was obtained using a matching energy dispersive spectroscopy (EDS) instrument to verify the uniform dispersion of PKPF in the microspheres. The pore structure of the microspheres was determined using a mercury porosimeter (Micromeritics AutoPore IV 9500), and the high-pressure mercury intrusion-extrusion curves were recorded to obtain the overall porosity, pore volume, and pore size distribution. The in vitro degradation performance of the microspheres was evaluated in PBS (pH 7.4) under shaking conditions at 37°C. Samples were periodically taken out, lyophilized, and weighed to calculate the mass loss ratio, and the supernatant was collected simultaneously for PKPF / Fe release testing. Release kinetics were quantitatively analyzed using the UV-Vis absorption method, and the cumulative release rate was calculated by combining fluid replenishment correction. This yielded the sustained degradation and slow-release behavior of the microspheres under physiological conditions, providing systematic and reliable material performance data for subsequent in vivo application studies.

[0035] 6. Isolation and culture of bone marrow mesenchymal stem cells Bone marrow mesenchymal stem cells (BMSCs) were obtained from 4-week-old male SD rats and isolated and cultured under sterile conditions. First, the femoral shafts were harvested from both sides and washed and soaked in sterile PBS containing 500 U / mL penicillin / streptomycin. After removing the metaphysis, the bone marrow cavity was exposed, and cells were collected by flushing the cavity with serum-free DMEM / F12 medium using a 5 mL syringe. The resulting cell suspension was filtered through a 200-mesh sieve and centrifuged at 1000 rpm for 5 min. The cell pellet was then resuspended in DMEM / F12 medium containing 10% fetal bovine serum (FBS) and 1% penicillin / streptomycin and cultured at 37°C and 5% CO2. BMSCs were isolated from the mixed cells using the differential adhesion method. When the BMSCs reached approximately 90% confluence, they were digested with an EDTA-trypsin mixture and passaged at a 1:3 ratio for further culture. The bone marrow mesenchymal stem cells, after three generations of culture and purification, were used in experiments, and the third-generation cells were selected for subsequent studies.

[0036] 7. Cytotoxicity The cytotoxicity of the material to BMSCs was assessed using a cell counting kit (CCK-8, Beyotime, China). BMSCs were counted at 0.8 × 10⁻⁶. 4Cells were seeded at a density of 0.05 mg / mL in 96-well plates, with DMEM, HAMA, PKPF@HAMA, PKP, and PKPF (0.05 mg / mL) added to each well. The plates were incubated at 37°C in a humidified cell culture incubator with 5% CO2, and the culture medium was changed every two days. At 24, 36, and 48 hours post-treatment, 90 μL of DMEM and 10 μL of CCK-8 solution were added to each well, and incubation continued for 2 hours. The absorbance was then measured at 450 nm using a microplate reader (FlexStation3, Japan).

[0037] BMSCs viability was assessed using an apoptosis detection kit (Beyotime, China). Live cells were stained green with Annexin V, and dead cells were stained red with PI. BMSCs were sputtered at 6 × 10⁻⁶ cells / year. 4 Cells were seeded at a density of 0.05 mg / mL in 24-well plates, and DMEM, HAMA, PKPF@HAMA, PKP, and PKPF (0.05 mg / mL) were added to each well. After incubation at 37°C and 5% CO2 for 24, 36, and 48 h, the culture medium was discarded and the cells were thoroughly washed with PBS. Staining working solution was then added to each well, and the cells were incubated at room temperature in the dark for 20 min. After staining, cell morphology and viability (live cells in green, dead cells in red) were observed using a fluorescence microscope (PCOM, Nikon, Japan). The number of live cells was calculated using an automated image analysis method. The obtained images were imported into the automated cell analysis software ImageJ (v1.53c, NIH, USA), and live cells were automatically counted using threshold segmentation and particle recognition algorithms. The number of live cells was recorded using the following formula: Live cell count = number of fluorescent cells counted in each image × area correction factor.

[0038] 8. In vitro induction of chondrogenic differentiation Prepare an induction premix (97 ml basal medium, 1 ml ITS supplement, 10 μl dexamethasone, 300 μl ascorbic acid, 100 μl sodium pyruvate, and 100 μl proline mixed thoroughly, Cyagen Biosciences Inc, China). Soak 1 mg PKPF@MAMA in 5 ml of the premix for 3 weeks. Then, collect 1 ml of supernatant every 3 days for cell culture, simultaneously replenishing with the same volume of premix.

[0039] Third-generation BMSCs were selected for the induction experiment. Cells were digested and counted before differentiation induction. 4 × 10⁶ cells were used. 5Each cell was transferred to a 15ml centrifuge tube, centrifuged at 250g for 4 min, resuspended in premixed buffer, and washed twice by centrifugation at 150g for 5 min. 1ml of the above cell culture medium was added, and the cells were centrifuged again at 150g for 5 min. The cells were then cultured at 37℃ with 5% CO2. The culture medium was changed every 3 days thereafter. The control group consisted of 1ml of premixed buffer with 10μl TGF-β added; all other procedures were the same as the experimental group.

[0040] Markers of cartilage formation and degradation—Aggrecan, Col1, Col2, and MMP-13—were evaluated by qRT-PCR. Total RNA was extracted from cell clusters after 3 weeks of induction and reverse transcribed into cDNA using the PrimeScript™ RT Kit (TaKaRa, Japan) according to the manufacturer's instructions. Then, RT-PCR was performed using a SYBR® Premix Ex Taq™ II Kit (TaKaRa, Japan) with a 20 μl reaction mixture containing 2 μl of cDNA in an ABI PRISM® 7500 rapid real-time RCR system (Applied Biosystems, Foster City, USA) (primer sequences are listed in Table S2). Finally, the relative expression of the specified genes was calculated using the 2^-ddCt method, and the expression of all genes was normalized to the TGF-β-treated control group. Each experiment was repeated at least three times.

[0041] 9. Laser confocal imaging After 3 weeks of BMSc induction, cell clusters were fixed with 4% paraformaldehyde and then incubated overnight at 4°C with rabbit anti-type II collagen primary antibody (1:500) (Affinity Biosciences, America). After washing three times with PBS, the cells were incubated at room temperature for 1 hour with FITC-labeled donkey anti-rabbit antibody (1:800). After washing three times with PBS, the cytoskeleton and nuclei were stained with phalloidin (1:2000) and DAPI (1:1000), respectively. Finally, the cell clusters were observed under a laser confocal microscope. Images were acquired using a confocal microscope (excitation 488nm, detection 500–550nm) with the same imaging parameters to avoid photobleaching. Images were analyzed using ImageJ (v1.53c, NIH, USA) to select regions of interest (ROIs), measuring mean fluorescence intensity or integrated density, and performing background correction. At least 3–5 images from each group were used for statistical analysis.

[0042] 10. Animal experiments Rat model of osteoarthritis and surgical treatment. This experiment was approved by the Animal Research Committee of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine. All surgical procedures were performed in accordance with the guidelines of the National Institutes of Health.

[0043] An osteoarthritis model was established in rats via anterior cruciate ligament (ACL) transection. Twelve-week-old male SD rats (350–400 g; n=24) were anesthetized with isoflurane and subcutaneously injected with meloxicam (1 mg / kg) and buprenorphine sustained-release (Bup-SR). The right hind limb was prepared, disinfected with povidone-iodine, and then sterilized with 70% (v / v) ethanol solution. The rats were transferred to the operating table, rested on heated circulating water pads, and covered aseptically. The skin of the right knee joint was incised from the distal patella to the proximal tibial plateau. The joint capsule medial to the patellar tendon was further incised and opened with scissors. The patellar tendon was moved laterally, and the fat pad was bluntly dissected to expose the ACL. The ACL was transected with microsurgical scissors, and the medial joint capsule and skin wound were sutured sequentially. After the surgery, the joint was thoroughly rinsed with sterile saline, and the patella was repositioned. The joint capsule and skin were sutured with 5-0 Polyglecaprone absorbable sutures. Postoperatively, the patient was given a subcutaneous injection of warm saline (20 ml / kg), kept alone with free access to food and water, and allowed free movement. The patient was given subcutaneous injections of meloxicam (1 mg / kg) every 24 hours and buprenorphine sustained-release (1 mg / kg) every 48 hours for 3 days.

[0044] Rats were randomly divided into 5 groups (n=5 per group). Every two weeks, rats were injected intra-articularly with an equal volume of PBS, PKPF@HAMA, PKPF, PKP@HAMA or HAMA suspension (all materials were at a concentration of 0.05 mg / mL). Every two days, rats were trained to run on a horizontal treadmill for 20 minutes at a speed of 8 m / min to induce knee osteoarthritis.

[0045] 11. MRI and CT imaging (1) Calculation of material relaxation rate The MRI relaxation rate of PKPF@HAMA was evaluated using a 9.4T MRI scanner (BAP 94 / 31, Bruker, Germany). Materials were placed in skirtless PCR plates (Thermo Fisher Scientific, USA) and equilibrated at a constant temperature (25 ± 0.5°C) for 30 min before measurement. First, a localization phase was scanned, followed by T1WI, T2WI, T1mapping, and T2mapping. T1WI used a T1 Rapid Acquisition with Relaxation Enhancement (T1 RARE) sequence with a repetition time (TR) of 800 ms, echo time (TE) of 6 ms, matrix size of 512 × 512, pixel spacing of 0.156 mm × 0.156 mm, slice thickness of 0.75 mm, and gap of 1 mm. T2WI uses a T2-Turbo Rapid Acquisition with Relaxation Enhancement (T2-TurboRARE sequence) with TR=2500ms; TE=33ms; matrix size=512×512; pixel spacing=0.156mm×0.156mm; layer thickness=0.75mm; and gap=1mm. T1mapping uses a T1map Rapid Acquisition with Relaxation Enhancement (T1map RARE sequence) with multiple TR acquisition modes, TR set to 200, 400, 800, 1500, 3000, and 5500ms respectively; TE=7ms; matrix size=512×512; pixel spacing=0.156mm×0.156mm; layer thickness=0.75mm; and gap=1mm. T2mapping employs a T2map Multi-Slice Multi-Echo (T2map MSME) sequence, using a multi-TE acquisition method with TE values ​​of 7.5, 15, 22.5, 30, 37.5, 45, 52.5, 60, 67.5, 75, 82.5, and 90 ms; TR = 2200 ms; matrix size = 512 × 512; pixel spacing = 0.156 mm × 0.156 mm; slice thickness = 0.75 mm; and gap = 1 mm.

[0046] After generating quantitative parametric maps from the original images of the PKPF@HAMA T1 and T2 mapping sequences using the Paravision 360 software included with Bruker, the parametric maps were plotted using the largest slice and its two adjacent slices. The average of the obtained parameter values ​​was then used as the relaxation times (T1 and T2). The relaxation rates R1 and R2 were calculated using the following formulas: R = 1 / T; Where R is the relaxation rate R1 or R2, and T is the relaxation time T1 or T2.

[0047] (2) Animal magnetic resonance imaging and image post-processing All model animals were examined using a 9.4T magnetic resonance imaging (MRI) scanner (BAP 94 / 31, Bruker, Germany), with imaging performed using a mouse head coil. Mice were anesthetized with isoflurane in 100% oxygen (3-4% at initial oxygen, 1-2% during scanning). Respiration was controlled and maintained for 60–100 minutes throughout the experiment. -1 The body temperature was 37±1°C. First, a localization phase was scanned for anatomical reference, followed by T2WI sequences using a T2-Turbo Rapid Acquisition with Relaxation Enhancement (T2-TurboRARE) sequence. TR=2500ms; TE=23ms; matrix=512×512; pixel pitch=0.138mm×0.138mm; slice thickness=0.7mm; slice interval=1mm.

[0048] The contrast-to-noise ratio (CNR) of knee articular cartilage relative to subchondral bone was obtained by delineating regions of interest (ROIs). Three representative slices were selected from the MRI image, and ROIs were manually delineated in the cartilage layer and adjacent subchondral bone, as well as in the air region of the image. The signal intensity of the ROIs in the cartilage layer and adjacent subchondral bone was extracted separately, and the standard deviation of the air region ROIs was extracted to represent image noise. The CNR calculation formula is as follows: CNR=(S cart S sub ) / σ noise ; Where S cart S represents the signal intensity of the cartilage layer. sub σ represents the signal intensity of the subchondral bone. noise This represents image noise. The average CNR of each slice is taken as the final value.

[0049] (3) Animal CT imaging and image post-processing Sixteen weeks after ACL amputation, rats were sacrificed, and the isolated knee joints were wrapped in saline-moistened gauze for imaging (Skyscan1178, BrukerMicroCT, Belgium). Scanning parameters were: tube voltage 65kV, tube current 615µA, rotation step 0.54°, and voxel size 4.4µm×4.4µm×4.4µm. Subsequently, 3D reconstruction and quantitative analysis of the CT images were performed using 3D Slicer software (v5.10.0, NA-MIC, USA). DICOM format data was imported into 3D Slicer, and the extent of osteophytes was delineated layer by layer in the images. The volume of osteophytes was calculated after generating VOI from the ROI.

[0050] 12. Histological staining and immunohistochemical staining (1) Histological staining Sixteen weeks after ACL transection, all rats were sacrificed. The genu joints were then fixed in 4% paraformaldehyde, decalcified with 10% EDTA, dehydrated with ethanol, and embedded in paraffin. Serial paraffin sections were prepared to a thickness of 5 μm. Histopathological evaluation was performed using hematoxylin-eosin (HE), alicin blue (AB), safranin / fast green (SF), and toluidine blue (TB) staining. ImageJ software was used to determine the depth of cartilage wear and the content of glycosaminoglycans in the sections. Furthermore, the degree of cartilage degeneration was assessed using the OARSI (Osteoarthritis Research Society International) semi-quantitative scoring system established by Pritzker et al. The score comprehensively considered the depth of the lesion (0–6 points) and the extent of the lesion (0–4 points), with the total score being the product of the depth and extent scores.

[0051] (2) Immunohistochemical staining Paraffin sections were first treated with 0.5% pepsin (Life Tech), then with 3% H2O2 (Life Tech) and 1% BSA (Sigma), followed by blocking with PBS containing 5% bovine serum albumin for 30 minutes. Primary antibodies MMP-13, RUNX2, and Collagen II were then added and incubated overnight (4°C). The next day, sections were incubated with the corresponding secondary antibodies for 30 minutes. MMP-13 and RUNX2 were developed using DAB, and Collagen II was fluorescently labeled and developed. Finally, hematoxylin was used to counterstain cell nuclei, and sections were mounted. Staining results were observed under a fluorescence microscope (PCOM, Nikon, Japan), and quantitative analysis of MMP-13, RUNX2, and Collagen II was performed using ImageJ software.

[0052] 13. Construction of an automatic segmentation model for knee articular cartilage The nnU-Net framework was selected as the core segmentation model, and a full-resolution 3DU-Net model was trained on this framework. During model training, knee joint MR images and their corresponding manually segmented labels were used as input data. The manually segmented labels covered two types of cartilage tissue: femoral cartilage and tibial cartilage, forming the training set. The training objective was to output the corresponding tissue segmentation mask after inputting the image; the training process was achieved by minimizing the loss function. Furthermore, the automatic post-processing module built into the nnU-Net framework was used to apply connectivity post-processing to both the merged mask and individual masks of the segmentation output. By retaining post-processing combinations that improved segmentation performance, the accuracy of the segmentation results was further optimized. For the robustness verification of the segmentation model, stratified sampling was used to divide the dataset containing samples with different degrees of knee osteoarthritis severity into training and test sets.

[0053] 14. Statistical Analysis The statistical analysis experiment was repeated three times to verify the results, and at least three copies of the test were performed. Data are expressed as mean ± standard deviation. One-way ANOVA and Tukey's multiple comparison test were used for comparisons between groups. Statistically significant differences were expressed as follows: P<0.05, P<0.01, P<0.001 and P < 0.0001. Statistical analysis was performed using SPSS software (v26.0, IBM, USA).

[0054] II. Experimental Results 1. Preparation, characterization, and cartilage matrix targeting verification of stepwise targeted nanomaterials To achieve the core strategy of progressively targeting the calcified cartilage layer, this application first requires the construction of a molecular structure capable of simultaneously pre-targeting iron ions and precisely anchoring them to the cartilage matrix. Therefore, this application designs a nanohybrid molecule PKP (with POSS as the framework and PEG, KGN, and EDA as functional side chains). Figure 1 (a) in the text. This structure provides a basis for future Fe... 3+ The coordinating sites of multiple hydroxyl, carbonyl, and amino groups required for complexation are the first step in achieving stepwise targeting. Firstly, the intermediates and final products were characterized by proton nuclear magnetic resonance spectroscopy, such as... Figure 1As shown in b, the unacrylated sample exhibits a distinct allyl double bond signal peak in the 5.8–6.4 ppm region. The mercaptohydrogen ion in POSS-SH is a broad peak at approximately 1.6 ppm. After photoinitiated catalytic addition of mercapto-olefin, both the olefin peak (5.8–6.4 ppm) and the SH peak disappear, while a new multiplet appears in the 2.7–3.3 ppm region, attributed to the newly formed thioether (–CH2–S– / –CH–(S–)) structure. These peak changes indicate that PEG, KGN, and EDA have been covalently grafted onto the POSS framework via mercapto-olefin click covalently, thus completing the construction of PKP. Meanwhile, our DFT-based simulations show that the energy difference between the HOMO and LUMO orbitals of PKP is 1.8009 eV, further demonstrating the structural stability of PKP. Figure 1 In c, for single-component POSS, KGN, Fe 3+ The UV-Vis absorption spectra of the final assembled product, PKP, were compared. POSS itself exhibits almost no significant absorption, with only a weak band appearing in the 210–230 nm range; KGN shows typical aromatic ring π→π transitions around 280 nm. Absorption peak; Fe 3+ In aqueous solution, it exhibits a broad and gentle ligand field absorption. In contrast, the spectrum of PKP shows a series of superimposed characteristic absorption peaks, retaining not only the aromatic absorption of KGN but also new shoulder peaks and redshifts in the 250–320 nm range, indicating that the multi-components have successfully covalently coupled and formed a chemical bonding network with an electronic structure different from that of the physical mixture. To demonstrate that PKP can chelate Fe in physiological fluids... 3+ A stable complex is formed by gradually adding different equivalents of Fe to the PKP solution. 3+ (0.25–1.75 eq), significant spectral changes were observed. Figure 1 (d) in Fe. 3+ With increasing dosage, the absorption intensity of PKP gradually increases in the 280–350 nm range and exhibits a slight red shift, while forming a sharper and more regular absorption band. This dose-dependent enhancement trend is typical of metal-ligand charge transfer (MLCT) and coordination-induced electron cloud rearrangement effects, indicating that multiple phenolic hydroxyl, carbonyl, and amino sites in the PKP molecule can interact with Fe³⁺. + Co-coordination, and the ability to gradually form stable Fe 3+ -PKP complex (PKPF). Furthermore, at higher equivalence Fe... 3+ The time-spectral density stabilized and no longer changed significantly, indicating that the system had reached coordination saturation. Next, to verify Fe... 3+ To investigate whether PKP can be induced to fold conformationally and form nanocomposites, we analyzed the relationship between particle size and zeta potential (e.g., ...). Figure 1 e–f in Fe3+ After chelation, the particle size increased from ~143 nm to ~221 nm, and the surface potential changed from negative to positive (-21.7 eV - 33.6 eV), indicating that the cross-chain coordination of iron ions induced folding and structural rearrangement of the nanocomposite, forming an iron ion-rich layer on the surface, thereby further enhancing the aqueous stability of the composite. This positively charged surface structure is beneficial for the next step of targeting the negatively charged ECM of cartilage. To further verify its strong coordination properties, we analyzed its electronic structure magnetism based on DFT calculations. The DFT calculation results are as follows: Figure 1 The g in the figure shows that free Fe 3+ The spin density is approximately 5 / 2, but after coordination with PKP, it significantly decreased to 3.89. / 2 indicates that the multi-coordination structure of PKP can partially passivate Fe through the electron cloud. 3+ The unpaired electrons reduce paramagnetism. Considering the correlation between T2-weighted MRI signal and the spin density of the paramagnetic center, this decrease in paramagnetism can reduce local magnetic field disturbances, thereby enhancing T2 signal brightness and providing a theoretical basis for MRI imaging enhancement. Simultaneously, the initial spin polarization of PKP is significantly increased, indicating a more consistent spin orientation of the remaining unpaired electrons, exhibiting higher order. Although the initial magnetic moment of the system increases numerically, this change mainly stems from the redistribution of spin density in the coordination network and enhanced cooperative magnetic response, rather than an increase in the number of unpaired electrons. DFT calculations also show that the PKP–Fe binding energy is as high as... 10.58eV ( Figure 1 (h) To verify the second-level targeting, PKPF was used to simulate chondroitin sulfate (CS) and keratin sulfate (KS). Figure 1 The data in Figure i show that the binding energy between PKPF and CS / KS is significantly enhanced, with the PKPF–CS / KS binding energy reaching [value missing]. The 17.75 eV value indicates that PKPF has extremely high affinity and strong targeting ability for cartilage ECM, and can achieve active enrichment by binding to CS / KS-mediated proteins (such as aggrecan and collagen II). Simultaneously, the HOMO / LUMO distribution and its band gap difference (ΔE = 0.5124 eV) indicate that PKPF has high stability. Figure 1 j in the equation. Finally, the electrostatic potential diagram obtained through calculation is ( Figure 1 The k in the figure shows that there are continuous negatively charged regions on the PKPF surface, which can actively adsorb positively charged Fe. 3+ In summary, PKP can not only efficiently capture and passivate Fe, but also... 3+ PKPF also possesses cartilage-targeting properties that strongly bind to CS / KS.

[0055] 2. PKPF microsphere loading strategy improves stability and enables deep MRI-visualized delivery to cartilage. Given that free PKPF in the joint cavity is prone to short residence time and difficulty in maintaining local concentration due to dilution by synovial fluid, mechanical flushing, and exudate removal, this application uses HAMA as a biocompatible hydrogel matrix and constructs PKPF@HAMA microspheres using microfluidic methods to achieve tissue-level retention, improved structural stability, and Fe 3+ Continuous delivery of passivation and metabolic remodeling signals. First, the microstructure and elemental composition of PKPF@HAMA were systematically characterized. Scanning electron microscopy (SEM) showed that the obtained PKPF@HAMA microspheres were generally regular spherical, with a particle size distribution concentrated at approximately 100 μm. The surface was smooth and possessed a typical hierarchical porous network structure, suggesting the existence of continuous hydrogel channels within the microspheres. EDSM approximation results further confirmed the uniform distribution of elements such as Fe, Si, and S within the microspheres, indicating that PKPF did not exhibit significant aggregation within the hydrogel network. Microspherization successfully achieved structurally stable embedding of PKPF. Figure 2 (a) In vitro degradation experiments showed that PKPF@HAMA microspheres exhibited slow linear degradation characteristics in PBS (pH=7.4, 37°C) environment, maintaining a stable cumulative release over 8 weeks. Figure 2 (b) The release behavior of both iron ions and PKPF exhibited obvious sustained-release kinetics, indicating that microsphere formation effectively overcame the problems of rapid diffusion and short-term effect of free PKPF. Next, mercury intrusion porosimetry further revealed that the microspheres possess a typical multi-scale porous structure (~400 nm) with a porosity of 68.75%. This is conducive to iron ion swelling and diffusion, gradual precipitation of PKPF, chondrocyte migration, and ECM exchange, providing conditions for metabolic reprogramming of the calcified cartilage layer and the establishment of an iron ion coordination environment. Figure 2 (c in the text)

[0056] The MRI visualization performance of PKPF@HAMA was further verified, and the results showed that the signal intensity of PKPF@HAMA on both T1WI and T2WI was concentration-dependent. Figure 2 (d in the text), and exhibits high longitudinal and transverse relaxation rates (R1=0.813s). -1 R² = 2.035s -1 ; Figure 2 (e–f in the text). This indicates that the stable iron coordination structure formed by PKPF can significantly alter the local magnetic field environment, satisfying the physical basis for MRI imaging of cartilage calcification layers.

[0057] To evaluate the targeting ability of PKPF@HAMA to the calcified cartilage layer, we further applied it to a cartilage model. The results showed that PKPF progressively penetrated from the superficial (Super) layer to the intermediate (Inter) and deep (Deep) layers, including the calcified cartilage layer and subchondral bone, increasing the image differences between different cartilage layers. Figure 2 The diffusion behavior of PKPF (g–h) is consistent with its positive charge and porous structure-promoted diffusion behavior. Particularly in the 0.05 mg / mL group, PKPF clearly penetrated the calcified layer within 1 hour, reaching a deep enrichment peak after 2 hours. ICP-MS measured Fe... 3+ The content also showed that the enrichment of deep iron ions increased significantly with time, reaching the highest level at 2 hours. Figure 2 (j) The imaging differences between different cartilage layers are measured by calculating the contrast-to-noise ratio (CNR), which is calculated using the following formula: CNR=|Si super or Deep -Si Inter | / SD noise ; Among them, Si super Indicates the signal intensity of the superficial cartilage; Si Deep Indicates the signal intensity of deep cartilage; Si Inter Indicates the signal intensity of the intermediate layer of cartilage; SD noise This indicates background noise.

[0058] The results showed that, at different concentrations, Sup–Int and Deep–Int had the highest CNR values ​​at 0.05 mg / mL. Figure 2 The concentration of k in the figure achieved the strongest stratification contrast. Meanwhile, time-varying dynamics showed that the CNR increased rapidly within 1–2 hours, then stabilized after 2 hours. Figure 2 l), and the aforementioned Fe³ + Consistent enrichment peaks were observed in deeper layers. These results collectively demonstrate that PKPF@HAMA can target the cartilage calcification layer, achieving delivery from the surface to the deeper layers. Furthermore, by constructing a stable iron coordination environment, PKPF@HAMA overcomes the problem of weak MRI signal caused by slow iron metabolism and low iron content in the cartilage calcification layer.

[0059] 3. PKPF@HAMA calcified cartilage layer lipid metabolism reprogramming promotes BMSCs chondrogenesis. This application further investigated the safety and chondrogenic induction potential of lipid metabolism reprogramming in calcified cartilage layer on bone marrow mesenchymal stem cells (BMSCs) in vitro. First, the extract of PKPF@HAMA was prepared at concentrations of 0.05, 0.07, and 0.1 mg / mL according to the material concentration, and combined with basal culture medium as a blank control for short-term culture of BMSCs.

[0060] To further clarify the effect of lipid metabolism reprogramming on BMSCs, empty vectors HAMA, PKP, PKPF, and PKPF@HAMA were co-cultured with BMSCs. After 24, 36, and 48 hours of culture, we performed live / dead staining on each group of cells. Figure 3 a) CCK-8 detection ( Figure 3 c) and automated cell counting based on Calcein-AM ( Figure 3 (d) In the figure, all four groups of cells showed an increasing trend over time, with no obvious cell death, indicating that the material has excellent biocompatibility. PKPF and PKPF@HAMA showed the most significant promoting effect on cell proliferation, especially PKPF@HAMA. This may be due to microsphere formation and Fe... 3+ The PKPF network structure allows for controlled release of KGN, maintaining a stable and effective concentration within the cell microenvironment, and, in conjunction with the hydrogel, provides a more suitable space for cell adhesion and growth. Furthermore, PKPF@HAMA at concentrations of 0.07–0.1 mg / mL significantly increased cell proliferation at 36 h and 48 h, with cell viability significantly superior to the control group. Figure 3 (f in the text). Therefore, we used 0.07 mg / mL as the optimal range for PKPF@HAMA on chondrocytes in subsequent experiments.

[0061] Based on the material's demonstrated good biocompatibility, we further investigated its ability to induce chondrogenesis through lipid metabolism reprogramming. BMSCs were induced and cultured using PKPF@HAMA extract, with TGF-β serving as a positive control. Immunofluorescence results ( Figure 3 b) shows that PKPF@HAMA significantly enhanced Aggrecan and Collagen II expression, while significantly inhibiting degeneration-related Collagen I and MMP-13, even outperforming the induction effect of TGF-β. Quantitative analysis ( Figure 3 e) further confirms that the lipid metabolism reprogramming effect of PKPF@HAMA can promote the enhanced expression of typical cartilage matrix proteins and the downregulation of fibrosis and degenerative factors, thereby maintaining chondrocyte phenotypic homeostasis.

[0062] The material's ability to induce chondrocyte aggregation was further investigated. Figure 3 The g-index showed that the PKPF@HAMA group exhibited strong and uniform Collagen II signaling in chondrocyte clusters, while three-dimensional reconstruction further highlighted its abundant cartilage matrix deposition. Gene-level detection ( Figure 3The results also showed that PKPF@HAMA significantly increased the expression of Col2a1 (COL I) and Aggrecan (ACAN), demonstrating its stable effect in promoting chondrocyte differentiation, and its performance was no weaker than that of the traditional TGF-β induction system.

[0063] 4. PKPF@HAMA enhances Fe 3+ Coordination environment enhances MRI visualization of cartilage Long-term stable drug release within the joint cavity is a key factor in improving in vivo cartilage regeneration efficiency; therefore, non-invasive visualization of cartilage is crucial. This study aimed to demonstrate the deep cartilage Fe production induced by PKPF@HAMA stepwise targeting and metabolic reprogramming. 3+ The enhanced coordination environment further validated the contrast effect of MRI imaging on articular cartilage. PKPF@HAMA was injected intra-articularly once starting in the second week post-surgery, and MRI examinations were performed before injection, on day 3, 1 week, 2 weeks, and 4 weeks post-injection. In the PKPF@HAMA group, the articular cartilage was clearly visible and remained at a relatively stable level for most of the time after injection. As the material gradually degraded, the signal intensity of the cartilage slowly decreased, and a slightly high T2WI signal was still visible in the rat articular cartilage until week 4. Figure 4 a) reflects Fe 3+ The enhanced imaging contrast was improved by coordination enhancement. The PKPF group showed clear high signal intensity on T2WI of cartilage on day 3, but the rapid decline in T2WI signal later led to reduced cartilage clarity. The cartilage clarity in the PKP@HAMA, HAMA, and PBS groups was poor. This indicates that PKPF@HAMA has excellent drug sustained-release properties and joint cavity retention effect. To further verify this conclusion, the contrast-to-noise ratio (CNR) of articular cartilage to subchondral bone at different time points was calculated. Figure 4 (c) The CNR of PKPF@HAMA gradually increased over time, then slowly decreased in the later stages, while the CNR of the PKPF group peaked on day 3 and then rapidly decreased. The CNR and SNR of the PKP@HAMA, HAMA, and PBS groups were all at low levels. These results collectively indicate that enhancing Fe... 3+ In its coordinated environment, PKPF@HAMA achieved superior MRI contrast and longer intra-articular retention. This further confirms the sustained-release properties and intra-articular retention effect of PKPF@HAMA.

[0064] Meanwhile, MRI monitoring of the cartilage repair process based on the PKPF@HAMA system was also validated. PKPF@HAMA, PKP@HAMA, PKPF, HAMA, and PBS were injected into the joint cavity every two weeks post-surgery, and MRI scans were performed monthly. Figure 4 (b) was scanned four times in total. It was observed that there was no significant difference in articular cartilage among the rat groups in the first month, but the articular cartilage in the PBS group was significantly thinner in the fourth month. The PKPF@HAMA group showed better therapeutic effects compared to other groups. To further confirm this conclusion, the cartilage thickness of the PKPF@HAMA and PBS groups was calculated. Figure 4 d) area ( Figure 4 e) and volume ( Figure 4 (f) In the first month, there was no significant difference between the two groups, but over time, the cartilage thickness, area, and volume in the PBS group gradually decreased, significantly lower than those in the PKPF@HAMA group. This indicates that PKPF@HAMA targets and enriches itself in the calcified cartilage layer through a stepwise "ion-molecule" targeting mechanism, and enhances iron utilization through lipid metabolism reprogramming, thereby synergistically promoting cartilage structure repair. It also enhances cartilage MRI signals for easier monitoring of repair.

[0065] 5. Achieving accurate and real-time visualization of cartilage based on a deep learning segmentation model. Traditional manual segmentation methods are inefficient and highly subjective, making them unsuitable for large-sample studies. They also fail to accurately capture sub-millimeter-level subtle changes in cartilage during osteoarthritis, which can easily lead to statistical bias and the omission of early lesions.

[0066] This application provides a key solution to the above problems by using an automatic segmentation model based on the nnU-Net deep learning framework. Figure 5 (a) First, the rat knee joint MR images were standardized preprocessed (pixel value normalization, 0.5×0.5×0.5mm). 3 Spatial resampling and axial effective region cropping were performed, and the rats were randomly divided into training and testing sets in a 7:3 ratio. The training set consisted of 70 rats, and the testing set consisted of 30 rats. The training set images were manually annotated by two senior radiologists for femoral and tibial cartilage and then converted into NIfTI format training samples. The 3DU-Net cascaded model was initialized based on the nnU-Net automatic configuration function. The model training used the SGD optimizer, with Dice loss and cross-entropy loss (weight 1:1) as the optimization objectives, and a differentiated batch size and 1000 training epochs were set. After training, the nnU-Net connectivity post-processing module was called to optimize the segmentation mask, finally forming a complete segmentation model including forward inference and post-processing.

[0067] The visualization results, including segmentation model predictions (red area), manual annotations (green area), and the overlap of the two (yellow area), visually demonstrate that the segmentation boundary of the segmentation model closely matches the gold standard. Figure 5 (b) in the text. Combined with the Dice coefficient results ( Figure 5In the MRI images, the Dice values ​​of the PKPF@HAMA, PKP@HAMA, and PKPF groups were all higher than 0.90, while the Dice values ​​of the HAMA and PBS groups were relatively lower. This indicates that the targeted modified probe can improve the signal discrimination of cartilage in MRI images, thereby helping the model achieve better segmentation consistency. After precise segmentation, the cartilage thickness, cartilage area, and cartilage volume of each group were obtained. Figure 5 (d, e, g, h in the original text). The maximum / mean cartilage thickness of the PKPF@HAMA, PKP@HAMA, and PKPF groups was significantly higher than that of the HAMA and PBS groups, and the cartilage area and volume were also higher. This verifies the cartilage repair effect of PKPF@HAMA on OA and suggests that molecular probe-enhanced MRI combined with AI segmentation can more sensitively capture subtle changes in early cartilage thickness. In summary, MRI-based imaging and artificial intelligence can achieve real-time visualization of cartilage microstructures, providing a new diagnostic strategy for early screening of micro-OA.

[0068] 6. Multidimensional verification of the feasibility of MRI-based visualization of cartilage changes using PKPF@HAMA. Because cortical bone appears as low signal or no signal on MRI, making it difficult to visualize details, we supplemented the scans with CT scans after four consecutive MRI scans, and finally sacrificed the rats to obtain the knee joints for pathological evaluation. In the HAMA and PBS groups, significant subchondral bone thickening and varying degrees of subchondral bone sclerosis were observed, while in the PKPF@HAMA group, which possesses deep cartilage-level stepwise targeting capability, such signs of degeneration were not obvious. Figure 6 (a) Further osteophyte volume measurements showed ( Figure 6 In group b), the PBS group had the largest osteophyte volume, which was 1.99 ± 0.09 mm. 3 The second group was the HAMA group (1.68±0.08mm). 3 In contrast, the osteophyte volume in the PKPF@HAMA and PKP@HAMA groups, which were able to precisely accumulate in the calcified cartilage layer, was significantly reduced, at only 0.53±0.13 mm. 3 With 0.68±0.07mm 3 This suggests that it has a significant advantage in inhibiting osteophyte formation and slowing down subosseous sclerosis. This trend is consistent with cartilage behavior monitored by MRI.

[0069] To further evaluate the cartilage repair performance of PKPF@HAMA and the accuracy of MRI visualization of cartilage degeneration, histological analysis was performed on tissue sections of the knee joints in each group. Histological evaluation of cartilage based on Alcian Blue (AB), Safranin / Fix Green (SF), and Toluidine Blue (TB) staining was conducted. Figure 6(c) In the PBS and HAMA groups, very typical pathological manifestations of osteoarthritis could be observed, such as roughness of the cartilage surface, erosion and deformation of cartilage tissue, and longitudinal cracks in the cartilage layer, which were more pronounced in the PBS group, with a visible cartilage erosion depth of 5.4±0.07 mm. Figure 6 (d) Among them, the PKPF@HAMA, PKPF, and PKP@HAMA groups best preserved the columnar structure of cartilage tissue and the integrity of the superficial cartilage, with no obvious tissue peeling or deformation, and to a certain extent increased tissue cell clones. There was no statistically significant difference in cartilage erosion depth among them. OARSI score results ( Figure 6 (e) shows that the PKPF@HAMA, PKPF, and PKP@HAMA groups were significantly lower than the HAMA and PBS groups. The PKPF@HAMA group had the lowest OARSI score, with intact cartilage structure, smooth surface, and good matrix staining, suggesting that it has the most significant repair effect on cartilage degeneration. In contrast, the scores of the PKPF and PKP@HAMA groups were slightly higher, but still significantly better than the HAMA and PBS groups. It is particularly noteworthy that there was no significant difference in OARSI scores between the HAMA and PBS groups, both showing relatively rough cartilage surface, structural disorder, and significant matrix loss, suggesting that the protective effect of HAMA alone on cartilage is limited. According to the above staining results, compared with the HAMA and PBS groups, the MHS@PPKHF, PKPF, and PKP@HAMA groups had the highest relative content of glycosaminoglycans (GAG). Figure 6 The PKPF@HAMA system (f) showed good performance in maintaining the cartilage matrix. The trends in cartilage changes assessed by CT and pathology were completely consistent with the trends visualized by MRI, indicating that our developed PKPF@HAMA system has great potential for real-time monitoring of OA progression and early OA screening.

[0070] In summary, this application is the first to construct an ion-molecule hierarchical targeting nanoplatform (PKP) based on POSS / KGN. PKP can highly specifically chelate iron ions to form positively charged PKPF, which then targets negatively charged components such as chondroitin sulfate in the cartilage matrix. Furthermore, using microfluidic technology, methacrylated hyaluronic acid hydrogel microspheres loaded with PKPF (PKPF@HAMA) were prepared for in-situ intra-articular injection. PKPF@HAMA synergistically delivers iron ions and KGN, achieving precise enrichment from the superficial cartilage layer to the calcified cartilage layer, with the enrichment of iron ions in the deeper layers reaching 2.3 times that of the surface layer. The delivered KGN induces lipid metabolism reprogramming in the calcified layer, not only promoting cartilage differentiation of BMSCs but also enhancing iron transmembrane transport by upregulating proteins such as CLC-2 and ATPase. Simultaneously, the upregulation of Aco1 protein indicates that iron ions are efficiently utilized, resulting in abundant Fe. 3+The coordinated environment significantly enhances MRI signals, achieving a breakthrough in layered imaging of full-thickness cartilage. Combined with artificial intelligence segmentation algorithms, this application non-invasively quantifies morphological parameters such as thickness, specific surface area, and volume of full-thickness cartilage in real time. By driving the reprogramming of lipid metabolism in the calcified cartilage layer, this application overcomes the limitations of traditional superficial signal enhancement, providing a new strategy for the diagnosis, efficacy evaluation, and precise repair of full-thickness cartilage lesions.

[0071] This specific embodiment is merely an explanation of this application and is not intended to limit it. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they fall within the scope of the claims of this application.

Claims

1. A nanomaterial for real-time MR imaging of full-thickness cartilage, characterized in that, include: Methacrylamide hyaluronic acid hydrogel matrix; as well as Fe dispersed within the hydrogel matrix 3+ -PKP complex; The PKP comprises an octamercaptopropyl polyhedral oligomeric silsesquioxane backbone and functional molecules grafted onto the backbone. The functional molecules include PEG, KGN, and EDA.

2. The method for preparing the nanomaterial for real-time MR imaging of full-thickness cartilage according to claim 1, characterized in that, Includes the following steps: Step 1: Acrylamide treatment of PEG, KGN and EDA; Step 2, prepare PKP by grafting acrylated PEG, KGN and EDA onto an octamercaptopropyl polyhedral oligomeric silsesquioxane backbone to obtain PKP; Step 3, FeCl3solution was added to the PKP solution to allow the iron ions to coordinate with PKP to form Fe 3+ -PKP complex, named as PKPF; Step 4: Prepare methacrylamide hyaluronic acid, named HAMA; Step 5: Mix the PKPF obtained in Step 3 and the HAMA obtained in Step 4 as the aqueous phase, and use an oily medium and emulsifier as the oil phase. Prepare the nanomaterial by microfluidic interface solidification method and name it PKPF@HAMA.

3. The preparation method according to claim 2, characterized in that, In step 1, the acrylation treatment steps of PEG, KGN, and EDA are as follows: PEG was dissolved in dichloromethane, triethylamine was added, and acryloyl chloride was slowly added dropwise under ice bath conditions. The reaction system was stirred at 0°C. After the reaction was completed, the reaction solution was washed, dried, filtered, and evaporated to dryness. KGN was dissolved in dichloromethane, triethylamine was added, and acryloyl chloride was slowly added dropwise under ice bath conditions while stirring the reaction. After the reaction was completed, the mixture was washed, dried, filtered, and evaporated to dryness. EDA was dissolved in DCM, triethylamine was added, and acryloyl chloride was slowly added dropwise under ice bath conditions. The reaction was stirred and the mixture was washed, dried, filtered, and evaporated to dryness.

4. The preparation method according to claim 2, characterized in that, In step 2, the molar ratio of octamercaptopropyl polyhedral oligomeric silsesquioxane, PEG, KGN and EDA is 1:4:0.25:

3.

5. The preparation method according to claim 2, characterized in that, The steps in step 2 are as follows: Octadecylpropyl polyhedral oligomeric silsesquioxane was dissolved in dry tetrahydrofuran and stirred at room temperature. Acrylamide-treated KGN, PEG and EDA were then added sequentially, followed by the addition of photoinitiator DMPA. The reaction was stirred at room temperature. After the reaction was completed, dialysis was performed to remove unreacted small molecule impurities to obtain the PKP.

6. The preparation method according to claim 2, characterized in that, The steps in step 3 are as follows: PKP was dissolved in deionized water to prepare a homogeneous solution, then FeCl3 solution was added and stirring was continued to produce Fe... 3+ -PKP complex, named PKPF.

7. The preparation method according to claim 2, characterized in that, The steps in step 4 are as follows: Hyaluronic acid was dissolved in preheated PBS solution and stirred continuously until a clear and homogeneous solution was formed. Then, methacrylic anhydride was slowly added dropwise, and the pH of the reaction system was maintained at 8.0-8.

5. The reaction system temperature was controlled at 50°C and the reaction was continued. The reaction was terminated with PBS. Then, after dialyzing, freezing and drying, the methacrylamide hyaluronic acid was obtained and named HAMA.

8. The preparation method according to claim 2, characterized in that, The steps in step 5 are as follows: PKPF and HAMA were dissolved in PBS solution and stirred to form a homogeneous and stable aqueous phase. Paraffin oil is mixed with an emulsifier to form an oil phase; During the microfluidic preparation process, the aqueous phase and oil phase are injected through the internal and external phase channels of the microfluidic chip, respectively, where the internal phase is the aqueous phase and the external phase is the oil phase. The flow rates are set to 0.8 mL·h for the aqueous phase. -1 oil phase 16 mL·h -1 The generated droplets further cross-linked and solidified at 80°C to form microspheres. The obtained microspheres were collected by centrifugation and washed repeatedly with petroleum ether and deionized water to finally obtain the nanomaterial, named PKPF@HAMA.

9. The application of the nanomaterial for real-time MR imaging of full-thickness cartilage as described in claim 1 in the preparation of products for early screening, diagnosis, efficacy evaluation, or cartilage repair monitoring of osteoarthritis.

Images