Zn(NO₃)₂•6H₂O (0.1 M), 2-MeIM (C₄H₆N₂, 99%), Cu(NO₃)₂•3H₂O (AR), CaCl₂ (97%), and HA (10 k) were purchased from Shanghai Aladdin Technology Co., Ltd. Ce6, DAPI, MTT, calcein-AM, and PI were supplied by Sigma-Aldrich. The annexin V-FITC/PI apoptosis kit was obtained from MultiScience Biotech Co., Ltd. All liquid chemical reagents were used without further purification.

CaO₂ NPs were obtained by a hydrolysis–precipitation process. A specific amount of CaCl₂ (1 g) was sent into the HA (50 mL, 0.1 M) solution at room temperature under continuous stirring for 30 min. After that, NH₃•H₂O (5 mL, 1 M) and H₂O₂ (1.5 mL, 30%) were sequentially injected and synthesized for 3 h. Afterward, NaOH (1.0 mL, 1 M) was added under ultrasound. Finally, the CaO₂ NPs were purified by centrifugation (13,000 rpm, 10 min) and sequentially washed with NaOH solution, pure water, and anhydrous ethanol three times.

The HA/CaO₂-Ce6@Cu-ZIF was prepared via an unfussy one-step method. A specific Zn(NO₃)₂•6H₂O (300 mg) and Cu(NO₃)₂•3H₂O (50 mg) were dissolved in methanol (100 mL) and formed an uniform solution. Then, 2-MeIM (190 mg), HA-stabilized CaO₂ NPs (50 mg), and Ce6 (20 mg) dissolved in the methanol solution (100 mL) were added drip by drip and reacted for at least 30 min under N₂ atmosphere. Finally, the HA/CaO₂-Ce6@Cu-ZIF was collected by centrifugation (13,000 rpm, 10 min) and washed with methanol three times.

TME images and corresponding elemental mapping were collected from Tecnai T20 at an accelerating voltage of 200 kV. The size of nanoparticles was calculated using Image J for 100 counting number. XRD patterns were obtained from Bruker D8 ADVANCE (Cu Kα radiation (λ = 0.154 nm) at 40 kV and 40 mA. Zeta potential and DLS measurements were gained by Zetasizer Ultra with He-Ne laser (633 nm). UV-vis absorption spectra were acquired from Shimadzu UV-1601. XPS spectra were analyzed from Rigaku DMAX-2400. FT-IR spectrum was accumulated from Nicolet Avatar 360 with the KBr wafer technique. ICP-OES measurements were surveyed from iCAP 6000 series. CLSM images were captured from Leica SP8. Flow cytometry was measured using BD accuri C6.

The generation of •OH was analyzed by TMB chromogenic reaction in pH, concentration, and time-dependent manners. The generation of ¹O₂ was determined by the DPBF chemical probe.

Cellular uptake of as-synthesized materials was operated on Panc02 cells. Cells were seeded in 6-well plates with a density of 1 × 10⁵ cells per well. The MTT cell assay was employed to evaluate the biocompatibility and toxicity of as-synthesized materials on L929 and Panc02 cells, respectively. Moreover, the live/dead cell assay was conducted to verify the cytotoxicity of the material on Panc02 cells. For intracellular ROS detection, a DCFH-DA chemical fluorescence probe was used. For the mitochondrial integrity assay, JC-1 staining kits were used to determine the J-monomer and J-aggregates separately. The intracellular fluorescence was observed by CLSM.

To investigate the biodistribution, the Panc02 tumor-bearing C57BL/6 mice were intravenously administered as-synthesized materials. For biodistribution investigation, the mice were sacrificed after 0, 2, 6, 12, 24, and 48 h. The heart, liver, lungs, spleen, kidneys, and tumors were collected for Cu contraction measurement. To estimate the anti-tumor efficacy of as-synthesized materials, the Panc02 tumor-bearing C57BL/6 mice were randomly placed into five groups (n = 5): control, CaO₂, CaO₂@Cu-ZIF, HA/CaO₂-Ce6@Cu-ZIF, and HA/CaO₂-Ce6@Cu-ZIF + Laser. During the treatment process, the tumor sizes and weights of mice were recorded once every 2 days: tumor volume = (tumor length) × (tumor width)²/2 (mm³).

After treatment process, the tumor and main organs (heart, liver, spleen, lung, and kidney) were collected for (H&E) staining according to the standard protocol for confirming caused injury.

All results were presented as mean ± S.D. Means were indicated using the student’s t-test. Statistical significance was determined at a value of *p < 0.05, **p < 0.01, ***p < 0.001.

The synthesis of HA/CaO₂-Ce6@Cu-ZIF was done through a two-step process. At first, CaO₂ NPs were synthesized through a hydrolysis-precipitation process. Then, the HA/CaO₂-Ce6@Cu-ZIF was synthesized through a simple one-step method. In detail, a specific Zn(NO₃)₂•6H₂O and Cu(NO₃)₂•3H₂O were dissolved in methanol and formed a uniform solution. Following this, 2-MeIM, HA-stabilized CaO₂ NPs, and Ce6 dissolved in the methanol solution were added drop by drop and reacted for 30 min to obtain HA/CaO₂-Ce6@Cu-ZIF. As revealed by transmission electron microscopy (TEM), the CaO₂ NPs are about 90 ± 2.3 nm, demonstrating the uniform size distribution. (Figures 1A, B). X-ray diffraction (XRD) pattern reveals that the synthesized CaO₂ NPs show obvious peaks at 30.1°, 35.6°, and 47.3° (Figure 1F), which is consistent with the JCPDS, No. 03-0865 according to previous literature for CaO₂. (Sun et al., 2021a). After this, Cu-ZIF was utilized to encapsulate the CaO₂ NPs and Ce6 via a self-assembly method to obtain the HA/CaO₂-Ce6@Cu-ZIF nanoplatform. The TEM image shows that the HA/CaO₂-Ce6@Cu-ZIF presents a regular octahedral shape with a particle size of around 110 ± 3.8 nm (Figures 1C, D). The homogeneous distributions of Zn, Cu, Ca, N, and O elements in HA/CaO₂-Ce6@Cu-ZIF are revealed by the elemental mapping, which demonstrates the successful loading of CaO₂ NPs (Figure 1E). Moreover, the XRD pattern of HA/CaO₂-Ce6@Cu-ZIF is consistent with that of ZIF-8, indicating that the as-synthesized materials are well held in the crystal structure of ZIF-8 (Figure 1F). (Li et al., 2021b) To endow the CaO₂-Ce6@Cu-ZIF with higher hydrophilicity for further biological application, HA with superior biocompatibility and targeted ability was employed for surface modification. As displayed in Figure 1G, the zeta potentials of CaO₂, CaO₂@Cu-ZIF, CaO₂-Ce6@Cu-ZIF, and HA/CaO₂-Ce6@Cu-ZIF are −20.03, +10.12, +23.9, and −25.6 mV, respectively, indicating that the CaO₂ NPs and Ce6 are successfully introduced into the Cu-ZIF and HA are effectively modified on the surface of as-synthesized materials. Meanwhile, the size distribution of CaO₂, CaO₂-Ce6@Cu-ZIF, and HA/CaO₂-Ce6@Cu-ZIF was obtained from the dynamic light scattering (DLS) measurements, the polydispersity index of which was 0.18, 0.19, and 0.17, respectively, demonstrating the good stable ability of HA modification. Figure 1I shows the hydrodynamic diameter is 142, 164, and 220 nm, respectively. The Fourier transform infrared (FT-IR) spectrum was recorded in the wavelength range of 500–4,000 cm^–1 (Figure 1H), also suggesting the sequential addition of CaO₂, Ce6, and HA, finally forming HA/CaO₂-Ce6@Cu-ZIF. (Li et al., 2021a). As shown in X-ray photoelectron spectroscopy (XPS), HA/CaO₂-Ce6@Cu-ZIF was also performed to evaluate the valence electron distribution, and the spectra are presented in which the coexistence of Zn, Cu, Ca, N, and O signals appear (Figure 1J). The high-resolution XPS of Zn, Cu, and Ca are shown in Figures 1K–M. In the high-resolution XPS of Cu spectrum, 933.3 and 953.6 eV peaks are assigned to Cu 2P3/2 and Cu 2p1/2, respectively. In addition, the satellite peaks at around 943.1 eV demonstrate the presence of Cu²⁺. (Li et al., 2021b). All the above materials’ characterizations imply the rational design and synthesis of H₂O₂/O₂ self-supply and Ca²⁺ overloading MOF-based nanoplatform.

The stability experiments of HA/CaO₂-Ce6@Cu-ZIF show that the as-synthesized materials maintain good dispersion within 7 days in cell medium (Figure 2A).The ultraviolet-visible (UV-vis) absorption spectra of Ce6, CaO₂, CaO₂@Cu-ZIF, and HA/CaO₂-Ce6@Cu-ZIF was shown in Figure 2B. Compared with the broad peak of CaO₂ and CaO₂@Cu-ZIF ranging from 450 to 800 nm, the absorption band of HA/CaO₂-Ce6@Cu-ZIF not only has the broad peak of CaO₂@Cu-ZIF but also exhibits the typical characteristic peak of Ce6 around 650 nm. Encouraged by the results from the photo-properties of HA/CaO₂-Ce6@Cu-ZIF, the ¹O₂ generation of PDT effect was explored by the UV-vis spectrum, where the 1,3-diphenylisobenzofuran (DPBF) was used as a real-time probe. The HA/CaO₂-Ce6@Cu-ZIF and PBS solutions were irradiated by 650 nm laser (0.5 W/cm²), respectively. At first, the HA/CaO₂-Ce6@Cu-ZIF could release Ce6 under acidic conditions. Then, DPBF could be oxidized by ¹O₂ which was generated from the combination of the released Ce6, light, and self-supplying O₂, so that the absorption peak of the DPBF (the specific absorption wavelength was at 410 nm) gradually decreased along with time increase (Figure 2C). However, the absorption peak of the DPBF solution that was treated with PBS was almost unchanged (Figure 2D). The relative intensity value of the UV-vis absorption peak at 410 nm for DPBF mixed with HA/CaO₂-Ce6@Cu-ZIF and PBS, respectively, further demonstrates the apparent decrease of DPBF absorption intensity (Figure 2E). To further confirm the production of ¹O₂, the 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) was also used (Figure 2F). And the results are consistent with the above. For •OH detection, a typical colorimetric analysis based on 3,3′,5,5′-tetramethyl-benzidine (TMB) was utilized to investigate the CDT effect of HA/CaO₂-Ce6@Cu-ZIF. HA/CaO₂-Ce6@Cu-ZIF can catalyze the oxidation of TMB to yield blue-colored oxTMB with typical absorbances at 370 and 652 nm. Considering the biodegradable properties related to the pH value of HA/CaO₂-Ce6@Cu-ZIF, the influence of the pH on •OH generation was first analyzed (pH = 4.5, 5.5, 6.5, and 7.4). The result shows that the pH has a significant influence on the •OH generation (Figure 2G). There is no evident •OH generation at pH 7.4, while the ability of •OH generation remarkably increases with the downregulation of pH. Then the concentration effect of HA/CaO₂-Ce6@Cu-ZIF for •OH generation was also investigated (Figure 2H). It shows an advanced ability of •OH generation along with the increased concentration (5, 15, and 20 μg/mL under pH 6.5). The •OH generation ability of HA/CaO₂-Ce6@Cu-ZIF related to time was also investigated (Figure 2I).

Given the successful construction of HA/CaO₂-Ce6@Cu-ZIF and advanced ROS generation capacity, the therapeutic effect of HA/CaO₂-Ce6@Cu-ZIF against Panc02 cells in vitro was further investigated. The therapeutic performance was first examined through the calcein-AM and propidium iodide (PI) double-staining assay (Figure 3A). The confocal laser scanning microscopy (CLSM) images show that the HA/CaO₂-Ce6@Cu-ZIF + Laser group exhibits the highest red-green ratio, where the red represents dead cells and green represents living cells, indicating the excellent anti-cancer effect of HA/CaO₂-Ce6@Cu-ZIF. Meanwhile, the flow cytometric apoptosis assay with Annexin V-FITC and PI staining was used to calculate the apoptotic cell death mediated by HA/CaO₂-Ce6@Cu-ZIF. The apoptotic ratio induced by HA/CaO₂-Ce6@Cu-ZIF under irradiation was 51.83% (the sum of Q2 and Q3), which was markedly higher than other groups under the same condition. This is mainly attributed to synergistic H₂O₂/O₂ self-supplying CDT/PDT synergistic effect. The intracellular ROS triggered by HA/CaO₂-Ce6@Cu-ZIF under laser irradiation was further investigated using a 2,7-dichlorofluorescein diacetate (DCFH-DA) probe, which can be hydrolyzed to DCFH. This can be rapidly oxidized by the generated ROS and form DCF with green-fluorescent (excited by 488 nm). The CLSM images exhibit that there is almost no green fluorescence in the control and CaO₂ groups. On the contrary, weak green fluorescence is exhibited in CaO₂@Cu-ZIF and HA/CaO₂-Ce6@Cu-ZIF groups. The strongest green fluorescence in the HA/CaO₂-Ce6@Cu-ZIF + Lase group indicates that HA/CaO₂-Ce6@Cu-ZIF under laser irradiation could generate more toxic ROS to induce tumor cell death (Figure 3B). The cytocompatibility of HA/CaO₂-Ce6@Cu-ZIF on L929 normal cells was evaluated by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) method (tetramethylazole salt microenzyme reaction colorimetric assay). As shown in Supplementary Figure S1, HA/CaO₂-Ce6@Cu-ZIF does not exhibit significant cytotoxicity to L929 cells, and the viability of cells treated with as-synthesized material for 24 h was 92.5% even at a concentration of 500 μg/mL, demonstrating the “silent” HA in the normal cellular microenvironment. Afterward, MTT assay was also used to estimate the cytotoxicity on Panc02 cells. Compared with others, the inhibition rate of HA/CaO₂-Ce6@Cu-ZIF under laser irradiation is as high as 53.5%, where the concentration of HA/CaO₂-Ce6@Cu-ZIF is 200 μg/mL (Figure 3C). Given that the Ca²⁺ overloading originating from CaO₂ could further enhance the oxidative stress and result in mitochondrial dysfunction, the mitochondrial integrities of different treatment groups were examined through JC-1 staining flow cytometry (Figure 3D). The qualitative comparison of J-monomer (green) and J-aggregates (red) following various treatments shows that the group treated with HA/CaO₂-Ce6@Cu-ZIF under laser irradiation exhibits abundant mitochondria damage. The endocytosis process of HA/CaO₂-Ce6@Cu-ZIF in Panc02 cells was evaluated using specific fluorescence properties of Ce6. As is known, when excited with 488 nm light, the loaded Ce6 can radiate green fluorescence. As shown in Figure 3E, the results suggest that HA/CaO₂-Ce6@Cu-ZIF could be effectively endocytosed by Panc02 cells and the internalization amount increased with prolonged time.

Inspired by the promising in vitro CDT/PDT synergistic effect of HA/CaO₂-Ce6@Cu-ZIF nanoplatform, the in vivo therapeutic assay in Panc02 tumor-bearing C57BL/6 mice model was conducted. When the tumor sizes reached about 100 mm³, twenty-five Panc02 tumor-bearing mice were randomly divided into five groups, followed by treatments: control, CaO2, CaO₂@Cu-ZIF, HA/CaO₂-Ce6@Cu-ZIF, and v) HA/CaO₂-Ce6@Cu-ZIF + Laser. As depicted in Figure 4A, Panc02 tumor-bearing mice were treated by intravenous administration on 1 and 7 days with injection doses of 15 mg/kg of mouse body weight. The body weight (Figure 4B) and tumor volume (Figure 4C) of mice were measured every 2 days during the treatment process. Moreover, the time-dependent Cu biodistribution of HA/CaO₂-Ce6@Cu-ZIF at the tumor and major organs were evaluated (Figure 4D). The results indicate an effective accumulation of HA/CaO₂-Ce6@Cu-ZIF at the tumor site, ensuring the following synergistic CDT/PDT therapeutics. In Figure 4B, during the treatment period, all the mice feature slight weight increases, demonstrating the negligible negative impacts of these treatments on the health of mice. As exhibited in Figure 4C, the relative tumor volume was notably suppressed in HA/CaO₂-Ce6@Cu-ZIF + Laser group in comparison with the other groups. Specifically, the suppression rate of the HA/CaO₂-Ce6@Cu-ZIF + Laser group was determined to be 60.8%, calculated from the variation in the relative tumor volume. This high suppression is attributed to the HA/CaO₂-Ce6@Cu-ZIF induced cascade-amplified CDT/PDT therapy as follows: 1) CaO₂ decomposed to generate H₂O₂ and O₂, which alleviated the insufficiency of intracellular H₂O₂ and relieved hypoxia conditions in TME; 2) H₂O₂/O₂ self-supplying effectively enhanced the production of •OH and ¹O₂ in Cu²⁺-mediated CDT and Ce6-induced PDT, respectively and 3) Ca²⁺ originated from CaO₂ could further enhance the oxidative stress and result in mitochondrial dysfunction induced by Ca²⁺ overloading. Intensive therapeutic efficacy was also confirmed by hematoxylin and eosin (H&E) staining of tumor sections from each group (Figure 4E). The results were consistent with the above tumor growth data. Additionally, the histological observations of major organs (heart, liver, spleen, lung, and kidney) present negligible acute pathological toxicities and adverse effects during the treatment duration for the control or treated groups (Figure 5). These results demonstrate that HA/CaO₂-Ce6@Cu-ZIF is of high biocompatibility.