A solution of the H₂O₂-responsive probe TPFP was prepared based on previous studies. MSNs were synthesized as follows: 0.12 g cetrimonium tosylate (CTAT), 0.03 g triethanolamine (TEAH₃), and 10 mL deionized water were mixed and stirred at 80°C for 30 min. Subsequently, a solution of 1.0 g tetraethyl orthosilicate (TEO) was added dropwise to the surfactant solution. The resulting mixture was stirred at 80°C for another 4 h at 1,000 rpm. The products were collected by centrifugation (10,000 rpm 30 min), washed three times with ethanol, and subsequently refluxed in an ethanol solution of NH₄NO₃ (1% w/v) for 12 h. Briefly, we used VP-16 and TPFP (1:20 mass ratio) dissolved in DMSO for the co-loading of the probes. To configure probes containing different ratios of chemotherapeutic components, the mass ratio of VP-16 to TPFP was adjusted to 0.5:20, 1:20, and 2:20 to prepare VP-16-MSN@TPFP (L), VP-16-MSN@TPFP (M), and VP-16-MSN@TPFP (H). The nanoparticles were collected by high-speed centrifugation and resuspended in the aqueous phase. Subsequently, VP-16-MSN@TPFP was mixed with CM derived from 4T1 cells, sonicated with heating for 5 min, and subsequently extruded through 200 nm polycarbonate membranes to obtain VP-16-MSN@TPFP@CMs.

The morphology of the MSNs was characterized using a JEM-2100F transmission electron microscope (TEM; JEOL, Ltd., Japan) and a scanning electron microscope (SEM; FEI Quanta 200F). The hydrodynamic diameter and zeta potential of the nanoparticles in water and PBS were characterized using a Nano-ZS 90 Nanosizer (Malvern Instruments Ltd., Worcestershire, United Kingdom). UV-vis absorption spectra were recorded using a U-3310 spectrophotometer (Hitachi, Japan). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used to characterize the protein composition of nanoparticles. Stability experiments were performed by measuring the nanoprobes in Dulbecco’s Modified Eagle’s medium (DMEM) plus 10% FBS for 7 days using dynamic light scattering (DLS).

The mouse breast cancer 4T1 cell line was cultured in DMEM supplemented with 10% FBS along with penicillin and streptomycin at 100–100 U/mL, respectively. Cells were incubated at 37°C in 5% CO₂-95% air atmosphere.

For the detection of exogenous H₂O₂, 4T1 cells were incubated with nanoprobes at 37°C for 30 min, images of VP-16-MSN@TPFP@CMs in live cells were investigated via spectral confocal multiphoton microscopy (Olympus FV1000-IX81) with a high-performance model titanium-sapphire laser source (Maitai, Spectra-Physics, United States), with the emission ranging from 575 to 630 nm. The excitation wavelength was 860 nm with a constant intensity.

Student’s t-test was used to analyze the differences between two groups. Differences between more than two groups were analyzed using a one-way analysis of variance. Simple linear regression was used to analyze the correlation between H₂O₂ signals and the expression level of COX2, PEG2 and their ratios. Sample sizes (n) and p-values (p) for all statistical analyses are indicated in figures and figure legends. Data were analyzed using statistical software OriginPro 2021b. In all cases, p < 0.05 represents a statistically significant difference.

Mesoporous silica materials (MSNs) were prepared as previously reported. The SEM and TEM images revealed monodisperse spherical MSNs with a diameter of 80 nm (Figure 1A; Figure 1B). The prepared MSNs were loaded with the ROS-responsive fluorescent probe TPFP and chemotherapeutic VP-16 to form VP-16-MSN@TPFP. Release results show that simultaneous release of the two components can be achieved (Supplementary Figure S2). Subsequently, 4T1 breast cancer CM were isolated and coated with VP-16-MSN@TPFP to prepare VP-16-MSN@TPFP@CMs. As shown in Figure 1C, the bioactive materials possess a core–shell structure with an MSN core enclosed in a smooth membrane shell. Consistently, a slight increase in hydrodynamic diameter and a decrease in surface charge were observed after CM coating (Figure 1D; Figure 1E). Protein electrophoresis demonstrated the presence of membrane proteins on VP-16-MSN@TPFP@CMs (Figure 1F), suggesting the successful integration of the cell membrane-coated nanoplatforms. Additionally, stability experiments in 10% FBS-containing medium confirmed that VP-16-MSN@TPFP@CMs exhibited little aggregation after 7 days of incubation (Supplementary Figure S1).

Before the in vitro and in vivo experiments, a series of optical experiments were performed to evaluate VP-16-MSN@TPFP@CM for ROS-responsive imaging properties. The absorption spectra showed a characteristic absorption of VP-16-MSN@TPFP@CM in phosphate buffer saline at 428 nm and no significant absorption at 560 nm. After adding H₂O₂ (100 μM) to the solution for 30 min, the absorption spectra showed a decrease at 428 nm and concomitant absorption at 560 nm (Figure 2A). In contrast, the fluorescence emission spectra showed that excitation at 560 nm after the addition of H₂O₂ resulted in an emission peak at 699 nm (Figure 2B). Quantitative studies have shown that the fluorescence intensity of VP-16-MSN@TPFP@CM can be used for the determination of in vitro and in vivo hydrogen peroxide concentration (Supplementary Figure S3, S4). Subsequently, we evaluated the imaging function of the prepared VP-16-MSN@TPFP@CM in 4T1 cells. As shown in Figure 2C, the fluorescence of TPFP and VP-16-MSN@TPFP@CM was found in the cytoplasm and represented H₂O₂ levels, indicating that VP-16-MSN@TPFP@CM was suitable for intracellular H₂O₂ analysis.

To ascertain the correlation between H₂O₂ levels and repopulation/metastasis in vitro in VP-16-treated cells, we conducted the chemotherapy-exacerbated repopulation/metastasis models in vitro by appropriate VP-16 stimulation (Figure 3A). Cell viability results showed that co-culturing 4T1 cells in six-well plates 12 h post-chemostimulation with all three materials caused significant cell death at 72 h (Figure 3B). The mean fluorescence intensity (MFI) of the VP-16-MSN@TPFP@CMs was measured using flow cytometry 12 h after the chemotherapeutic challenge (Supplementary Figure S5). The levels of intracellular COX2 and PGE2 in the culture medium were assayed on day 2. Increased levels of both the components were observed with increasing levels of VP-16 (Supplementary Figure S6, S7). Cell repopulation and cell migration assays showed that VP-16-stimulated cells shaped a microenvironment conducive to cell repopulation and metastasis by increasing components such as the inflammatory protein COX2 and cytokine PGE2 (Figures 3E–G). The intracellular fluorescence signal after treatment with nanoprobes was measured using flow cytometry. These data are consistent with the changes in COX2 protein expression and PGE2 secretion during chemotherapy. Correlation analysis showed that the MFI of VP-16-MSN@TPFP@CMs-treated cells was highly positively correlated with COX2 expression levels (Pearson r:0.9210, p < 0.001) (Figure 3D) and PGE2 levels (Pearson r:0.9821, p < 0.001) (Figure 3C), confirming that VP-16-MSN@TPFP@CMs might provide valuable information for the evaluation of repopulation/metastasis after chemotherapy.

We established a 4T1 tumor-bearing mouse model to evaluate H₂O₂ detection using VP-16-MSN@TPFP@CM. We examined the H₂O₂ signals at the tumor site after intravenous treatment with VP-16-MSN@TPFP@CM or a mixture of VP-16 and TPFP. The results showed that, compared to free small molecules, VP-16-MSN@TPFP@CM could achieve up to 12 h of H₂O₂ monitoring at the tumor site owing to better tumor targeting and retention (Supplementary Figure S8). To further validate the advantages of the cell membrane-cloaked nanoprobe, we compared the distribution characteristics of VP-16-MSN@TPFP and VP-16-MSN@TPFP@CM. The results showed that the cell membrane-cloaked nanoprobes accumulated more at the tumor site after 12 h of i.v. administration, thus facilitating better in vivo monitoring (Supplementary Figure S8). We treated the tumor-bearing mice with a single dose of VP-16-MSN@TPFP@CM (Figure 4A). At the non-therapeutic stimulation dose, the growth rate of the tumor was higher than that of cells treated with MSN@TPFP@CM, and more metastatic nodules were found (Figure 4B; Figure 4C). Subsequently, the relationship between the H₂O₂ levels at the tumor site port-chemo with umor growth rate (TGR) or metastatic nodules was compared between MSN@TPFP@CM and VP-16-MSN@TPFP@CM. We observed more pulmonary metastatic nodules (PMN) in the lungs of VP-16-stimulated mice, and the TGRwas higher than in mice without chemotherapy (Figure 4D; Figure 4E). Both PMN and TGR fitted well when conducting correlation analysis with the H₂O₂ signal of the orthotopic tumor (Figure 4F; Figure 4G).

Biosafety is a prerequisite for the clinical translation of biomaterials. Therefore, we examined the body weight, serum biochemical index, and histology of the major organs to evaluate the systemic toxicity of VP-16-MSN@TPFP@CM. No weight loss was observed in the VP-16-MSN@TPFP@CM group compared to the saline control group (Figure 5A). The levels of serum chemistry indices, including albumin (ALB), alanine transaminase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and Serum creatinine (CRE), did not change remarkably in any of the VP-16-MSN@TPFP@CM groups (Figure 5B). Furthermore, H&E staining of the liver, spleen, kidney, lung, and heart indicated the absence of pathological damage to the major organs during exposure (Figure 5C). Collectively, these results indicated that VP-16-MSN@TPFP@CM is biologically safe.