Sono-activable and biocatalytic 3D-printed scaffolds for intelligently sequential therapies in osteosarcoma eradication and defect regeneration

Abstract

To mitigate the necessity for multiple invasive procedures in treating malignant osteosarcoma, an innovative therapeutic approach is imperative to achieve controllable tumor-killing effects and subsequent bone repair. Here, we propose the de novo design of sono-activable and biocatalytic nanoparticles-modified 3D-printed hydroxyapatite (HA) scaffold (HS-ICTO) for intelligently sequential therapies in osteosarcoma eradication and bone defect regeneration. The engineered HS-ICTO scaffold displays superior, spatiotemporally controllable H₂O₂-catalytic performances, which promptly generate massive reactive oxygen species via multienzyme-like mechanisms coupled with sono-activation, thus augmenting tumor cell apoptosis. Furthermore, HS-ICTO can intelligently switch to catalyze H₂O₂ to O₂ within the inflammatory bone defect microenvironment, effectively blocking endogenous H₂O₂-mediated oxidative stress, which positively modulates the osteogenic differentiation of stem cells and ultimately facilitates defect regeneration. We validate that this multifaceted HS-ICTO scaffold possesses robust and on-demand abilities to prevent neoplastic recurrence and promote anti-inflammatory osseous tissue repair, representing a promising platform for precision oncological intervention and regenerative medicine.

Fig. 1. Engineering sono-activable and biocatalytic 3D-printed scaffolds for limb salvage therapy of osteosarcoma.

a Synthesis of the ICTO and ICTO-modified 3D-printed HA scaffolds (HS-ICTO). b Sono-activable and versatile biocatalytic ROS generation in intelligent and controllable osteosarcoma eradication. TME indicates tumor microenvironment. c Excess endogenous H₂O₂ scavenging and redox homeostasis remodeling from the defect area for superior bone reconstruction. BMP-2 bone morphogenetic protein-2, Runx2 Runt-related transcription factor-2, COLI type I collagen, RANKL receptor activator of nuclear factor κB ligand, HIF-1α hypoxia-inducible factor 1-alpha, MMP matrix metalloproteinase, NFATc1 nuclear factor of activated T-cells, cytoplasmic 1, ACP5 acid phosphatase 5.

Fig. 2. Morphology and structural characterization.

a Structure illustration of HS-ICTO scaffold at different scales from millimeter to atomic scale (color codes: gray, O; cyan, Ti; yellow, Ir). b Representative SEM images and c, d magnified SEM images of HS-ICTO. e TEM images of ICTO. f The corresponding EDS elemental mapping of ICTO. g HRTEM images of ICTO. h Magnified HRTEM images of ICTO. i HAADF-STEM image of ICTO with the corresponding FFT pattern. j XPS spectra of ICTO and Ir/C in Ir 4f regions. k XPS spectra of ICTO and TO in Ti 2p regions. l Normalized XANES spectra at Ir L₃-edge. m FT k³-weighted FT spectra in (k)-function of the EXAFS spectra for the Ir L-edge. n WT images for the k³-weighted EXAFS signals. R represents the distance between the adsorbing atoms and neighboring atoms, and χ(k) denotes the amplitude of the EXAFS oscillations as a function of photoelectron wavenumber k. The color gradient delineates the transition from high signal intensity (orange) to low signal intensity (cyan). In (j–l), a.u. indicates the arbitrary units. Experiments were repeated independently (b–i) three times with similar results. Source data are provided as a Source Data file.

Fig. 3. Enzyme-like biocatalytic activities and theoretical calculations of ICTO.

a A comprehensive schematic representation of the advantageous structural characteristics of ICTO and the resultant versatile enzyme-like biocatalytic activities. b POD-like and OXD-like activities utilizing TMB assay at a wavelength of 652 nm after incubation with ICTO and TO (n = 3 independent experiments, data are presented as mean ± SD). c Comparison of kinetic parameters in ICTO and TO. d GSH depleting abilities with DTNB as the trapping agent of -SH in GSH. e The EPR spectra of the in situ •O₂⁻ radical detection by DMPO and ¹O₂ radical detection by TEMP. f The H₂O₂ elimination and O₂ generation capacities of ICTO and TO under neutral conditions (n = 3 independent experiments, data are presented as mean ± SD). g A comparative analysis of the V_max and K_m parameters for ICTO and TO. h Proposed reaction pathways and i corresponding Gibbs free energy diagram of POD-like and CAT-like catalytic pathways on ICTO (color codes: gray, O; green, Ti; orange, Ir; red, H). j Computed PDOS of O 2p orbital of *OH (Ir) and *OH (Ti) in ICTO. k Differential charge density analysis of ICTO*2OH (cyan and yellow are employed to denote charge depletion and accumulation, respectively; the cut-off of the density-difference isosurface is 0.01 e·Bohr⁻³). l Calculated Bader charge of *OH (Ti) and *OH (Ir) in ICTO. The assessment in (b, f) of P-values was performed by a two-tailed Student’s t-test; all tests were two-sided. V_max is the maximal reaction velocity, K_m is the Michaelis constant, TON is the turnover number. In (b–e), a.u. indicates the arbitrary units. Source data are provided as a Source Data file.

Fig. 4. In vitro therapeutic effects in osteosarcoma eradication.

a Schematic illustration of the therapeutic mechanism of HS-ICTO under TME condition. b POD-like activity by evaluating the TMB absorbance value at λ = 652 nm after incubation with HS-ICTO-x (n = 3 independent experiments, data are presented as mean ± SD). c Sono-activable biocatalytic oxidation of TMB and DPA under the co-incubation of HS-ICTO (n = 3 independent experiments, data are presented as mean ± SD). d PCA plot for different cellular samples. e Volcano plots showing differential gene expression in HS-ICTO + US vs. HS and HS-ICTO vs. HS comparison (Sig up, significantly upregulated; Sig down, significantly downregulated; Up, upregulated; Down, downregulated; Non-Sig, nonsignificance). f GO enrichment analysis of the differentially expressed genes associated with ROS-triggered damage, stress response, and apoptosis biological process in 143b cells. The data in (d–f) were representative of three biologically independent samples from each group. g Representative CLSM images of the live/dead staining of 143b cells on the scaffolds after different therapies. The images were representative of three independently repeated experiments from each group. h Flow cytometric apoptosis analysis of Annexin V-FITC/PI-stained 143b cells harvested from scaffolds with different treatments. The gate strategy is shown in Supplementary Fig. 53. The plots were representative of three independently repeated experiments from each group. i Expression of HIF-1α and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in 143b cells harvested from different scaffolds analyzed by western blot assay (n = 3 biologically independent replicates). The assessment in (c) of P-values was performed by a two-tailed Student’s t-test. In (e), P-values were obtained from two-sided DESeq2 test without multiple comparison. In (f), P-values were obtained from one-sided Hypergeometric test without multiple comparisons. In (b, c), a.u. indicates the arbitrary units. Source data are provided as a Source Data file.

Fig. 5. In vivo osteosarcoma xenograft killing effect.

a Illustration of 143b osteosarcoma xenograft establishment and the treatment procedure. b Schematic diagram of the osteosarcoma eradication mechanism of HS-ICTO. c Digital photos of the osteosarcoma xenograft after excision on day 15. d Tumor volumes are recorded every third day (n = 5 biologically independent mice per group; tumor volumes on day 15 were used for statistical analysis). e The tumor xenograft weights were measured on day 15 after excision (n = 5 biologically independent mice per group). f The tumor growth inhibition ratios (n = 5 biologically independent mice per group). g Mouse weight change presented as a heat map (n = 5 biologically independent mice per group). h The H&E, TUNEL, and Ki67 fluorescence staining images of the tumors from different groups. The fluorescence images were representative of three independently repeated experiments from each group. i Quantitative analysis of the TUNEL-positive stained cell numbers (n = 5 biologically independent replicates). j Quantitative analysis of the MFI of Ki67 staining (n = 5 biologically independent mice per group). Data are presented as mean ± SD, and ns represents no significant difference; statistical significance was calculated using one-way ANOVA followed by Tukey’s post hoc test for multiple comparisons; all tests were two-sided. In (j), a.u. indicates the arbitrary units. Source data are provided as a Source Data file.

Fig. 6. In vitro redox homeostasis modulation and stem cell protection against oxidative stress.

a Schematic illustration of the H₂O₂ redox homeostasis restore and O₂-induced hypoxia relief of HS-ICTO under bone defect condition. b Time-dependent H₂O₂ elimination of HS and HS-ICTO-x (0.5, 1.0, and 2.0). c The produced O₂ concentration was measured in the presence of HS and HS-ICTO-x (0.5, 1.0, and 2.0) and H₂O₂. d Representative CLSM images of live/dead staining of BMSCs seeded on 3D scaffolds with different treatments on day 5 and the corresponding CCK-8 assay result (n = 3 biologically independent replicates). e Representative CLSM images of phalloidin-Rhodamine stained BMSCs with different treatments and the corresponding cell size evaluation (n = 20, the size of twenty random cells from each group was calculated by ImageJ software). f TEM observation of the BMSCs after different treatments. The red arrow indicates damaged mitochondria, and the red star indicates swollen endoplasmic reticulum. g ALP staining of the BMSCs stimulated by osteogenic medium for 14 days with different treatments. h ARS staining of the BMSCs stimulated by osteogenic medium for 21 days with different treatments. i Quantitative analysis of the ALP and j ARS positive staining area (n = 3 biologically independent replicates). k Immunofluorescence staining of Runx2, BMP-2, and COLI expression of BMSCs stimulated by osteogenic medium for 14 days with different treatments. The images in (d–h, k) were representative of three independently repeated experiments from each group. l Quantitative analysis of the MFI of Runx2, BMP-2, and COLI immunofluorescence staining (n = 3 biologically independent replicates). m The Runx2, BMP-2, COLI, and ALP gene expression of BMSCs after different stimulation for 14 days measured by RT-qPCR (n = 3 biologically independent replicates). n Schematic illustration of HS-ICTO regulating H₂O₂ redox homeostasis and protecting BMSCs. Data are presented as mean ± SD, and ns represents no significant difference; statistical significance was calculated using one-way ANOVA followed by Tukey’s post hoc test for multiple comparisons; all tests were two-sided. In (l), a.u. indicates the arbitrary units. Source data are provided as a Source Data file.

Fig. 7. Suppression of osteoclastogenesis of HS-ICTO.

a Illustration of the transwell co-incubation experiment of K7M2 and RAW264.7 cells. The K7M2 cells were seeded in the upper chamber, and the RAW264.7 cells were seeded in the lower chamber. b The NFATc1, MMP-9, ACP5, and c-Fos gene expression of RAW264.7 cells co-incubated with K7M2 for 7 days measured by RT-qPCR (n = 3 biologically independent replicates). c Schematic depiction of the HS-ICTO inhibition mechanism affecting osteoclastogenesis. d Representative images of phalloidin-FITC stained RAW264.7 cells after different treatments. The large cells with green actin ring indicate osteoclasts. Positive control group (PC, only stimulated by 50 ng/mL RANKL). e Quantitative analysis of the size of osteoclasts (n = 30 cellular replicates). f Representative images of TRAP-stained RAW264.7 cells after different treatments. The fluorescence images in (d, f) were representative of three independently repeated experiments from each group. g Quantitative analysis of the TRAP-positive cell counts (n = 5 biologically independent replicates). h The NFATc1, MMP-9, ACP5, and c-Fos gene expression of RAW264.7 cells after different stimulations for 7 days measured by RT-qPCR (n = 3 biologically independent replicates). i The expressions of HIF-1α and α-tubulin of RAW264.7 were analyzed by western blot assay. Western blot experiments were independently repeated three times with similar results. Data are presented as mean ± SD, and ns represents no significant difference; in experiment (b), statistical significance was calculated using a two-tailed Student’s t-test and in experiment (e, g, h), statistical significance was calculated using one-way ANOVA followed by Tukey’s post hoc test for multiple comparisons; all tests were two-sided. Source data are provided as a Source Data file.

Fig. 8. In vivo bone regeneration in rat cranial defect model.

a Process illustration of the in vivo bone regeneration assessments. b PCA plots for different samples. c GO enrichment analysis of the differentially expressed genes associated with ROS damage, bone destruction, and inflammation in harvested cranial bone tissue. d Heat maps illustrating the differential expression profiles. The data in (b–d) were representative of three biologically independent samples from each group. e H&E staining and f Masson’s trichrome staining of regenerated bones induced by different scaffolds. g Coronal views of the defect areas from micro-CT images. h Quantitative analysis of bone volume (BV/TV, n = 3 biologically independent mice per group) and i trabecular number (Tb.N, n = 3 biologically independent mice per group) induced by different scaffolds. j Immunofluorescence staining of Runx2, BMP-2, and COLI from different scaffolds. The images in (e–g, j) were representative of three biologically independent samples from each group. k Quantitative analysis of the Runx2 positive cell numbers (n = 3 biologically independent mice per group). l Quantitative analysis of the BMP-2 positive cell numbers (n = 3 biologically independent mice per group). Data are presented as mean ± SD, and ns represents no significant difference; In (c), P-values were obtained from one-sided Hypergeometric test without multiple comparisons. The assessment in (h, i, k, l) of P-values was performed by a two-tailed Student’s t-test. Source data are provided as a Source Data file.