A novel nanomedicine for osteosarcoma treatment: triggering ferroptosis through GSH depletion and inhibition for enhanced synergistic PDT/PTT therapy | Journal of Nanobiotechnology

maio 4, 2025

2

Synthesis and characterization of CSIR

Scheme 1 illustrates the design of a system for depleting intracellular GSH and inhibiting its synthesis to trigger ferroptosis, thereby enhancing the therapeutic efficacy of PDT/PTT. The optimal proportional composition of the system referred to as CSIR, was determined by investigating the mass ratio of sorafenib, Cu²⁺, and IR780. Particle size and polydispersity index (PDI) were characterized for various feed ratios with different amounts (0.5, 1, 1.5, and 2.0 μg) of IR780 (10 mg mL^− 1 in DMSO) dissolved in 1 mL of ultrapure water. As shown in Fig. 1a and b, the CSIR prepared at a feed mass ratio of 1:2:1 exhibited the most favorable nanoparticle size and PDI distribution. This ratio was consequently chosen for further experiments. Figures 1c and d demonstrate that CSIR displayed acceptable changes in hydrodynamic size and PDI after incubation in PBS at different pH values for 9 days, indicating its stability. Further characterization of the CSIR nanomedicine’s composition was achieved using energy-dispersive X-ray spectroscopy (EDS) analysis (Figure S1a, SI) and elemental mapping (Fig. 1e). Additionally, X-ray photoelectron spectroscopy (XPS) (Figure S1b, SI) and Fourier-transform infrared spectroscopy (FTIR) were employed to evaluate the functional groups present within the CSIR nanomedicine. As depicted in Fig. 1f, the absorption peaks observed at 1202 cm⁻¹ and 1155 cm⁻¹ corresponded to the S = O stretching vibration mode of the sulfate group. Similarly, the peaks at 659 cm⁻¹ and 604 cm⁻¹ originated from the S = O bending vibration mode of the same group. The peaks at 1724 cm⁻¹ and 1690 cm⁻¹ were characteristic of the C = O stretching vibration. Notably, the CSIR spectrum primarily resembled that of IR780, although additional peaks were present. The peak near 618 cm⁻¹ indicated the S = O bending vibration of copper sulfate pentahydrate, while the peak near 1700 cm⁻¹ corresponded to the C = O stretching vibration of SRF. These observations suggest the successful incorporation of copper sulfate pentahydrate and SRF into the IR780 matrix. Furthermore, the sample exhibited a noticeable change in the peak shape of the N-H and O-H absorption peaks around 3429 cm⁻¹, potentially indicating the presence of hydrogen bonding interactions between the various components of the CSIR [41]. In addition, the original sharp peaks of each sample disappeared and stretched into broad peak shapes, which was due to the interaction between nanomaterial particles (and quantum size effect), indicating that the synthesized nanomaterials have a large dispersion and small particle size [42]. To gain deeper insight into the formation of CSIR, XPS analysis was conducted (Figure S2, Supporting Information). The C1s spectrum of LICN revealed six distinct peaks corresponding to various bonding configurations: 284.79 eV (C-C/C-H), 285.90 eV (C-O), 287.11 eV (C = O), 288.72 eV (CF), 291.19 eV (CF2), and 292.76 eV (CF3). The F1s spectrum of CSIR displayed a single peak at 688.10 eV (C-F). Additionally, the presence of chlorine (Cl), iodine (I), and copper (Cu) was confirmed by the XPS survey spectrum (Figure S2).

The UV–Vis absorption spectra indicate that both CSIR and IR780 exhibit a maximum absorption peak at approximately 780 nm, with the absorbance of CSIR being slightly lower than that of IR780 (Figure S4). The molecular structure of sorafenib includes a pyridine ring and ether bonds, where the nitrogen and oxygen atoms possess lone pair electrons that can coordinate with copper ions (Cu²⁺). As a metal center, Cu²⁺ forms coordination bonds with the pyridine ring of sorafenib and with specific groups (such as nitrogen or oxygen atoms) in IR-780 iodide, thereby linking the two compounds together. IR-780 iodide is a lipophilic cationic compound. Its hydrophobic segment, such as the long-chain conjugated structure, can interact with other hydrophobic regions (for example, the hydrophobic domains of sorafenib) via hydrophobic interactions. These interactions promote the close packing of molecules, leading to the formation of the nanoparticle core structure. In particular, the above-mentioned spectral range contains peaks characteristic of the Cu²⁺ [31]. These findings collectively validate the successful synthesis of CSIR. The content of IR780, SRF, and Cu²⁺ within CSIR were quantified using UV-visible absorption spectroscopy and inductively coupled plasma mass spectrometry (ICP-MS) for each component, respectively. The encapsulation efficiencies were 82.41% for IR780, 73.60% for SRF, and 78.53% for Cu²⁺.

Figures 1g and h depict the results of infrared thermal imaging and temperature change curves for PBS, free IR780, and CSIR solutions irradiated with an 808 nm NIR laser at 1.2 W cm^− 2 for 5 min. Notably, the PBS group exhibited minimal temperature change under laser exposure. In contrast, the temperature of both the free IR780 and CSIR groups rose significantly, exceeding 80 °C, which was well above the 42 °C threshold required for thermal tumor cell ablation [43]. These findings, corroborated by the similar infrared thermographic images of free IR780 and CSIR (Fig. 1g), demonstrated that CSIR effectively retained the superior photothermal properties of IR780. Furthermore, Fig. 1i and j illustrate that the temperature increase of CSIR accelerated with higher laser power density and concentration. Based on the results presented in Fig. 1g h, we chose to characterize the photothermal conversion efficiency under the conditions of a 40 μg/mL concentration and a laser power density of 1.2 W/cm². Under irradiation with an 808 nm laser at 1.2 W/cm², we recorded the temperature increase of the 40 μg/mL CSIR solution over 600 s. After switching off the laser, the solution was allowed to cool naturally, and the temperature change was similarly recorded over 600 s. Using the method described in reference, we calculated the photothermal conversion efficiency of the CSIR nanomaterials. As shown in Figure S3 illustrates a heating and cooling cycle, and fitting the data to produced a t‑-lnθ plot with a time constant (τs) of 235.6 s. According to the formula provided in the literature, the photothermal conversion efficiency of the CSIR solution was determined to be 68.9%. This observation suggests that the photothermal conversion efficiency of CSIR exhibited a dependence on both laser power and concentration.

Cellular uptake and biodistribution of CSIR

Given the critical role of intracellular uptake in tumor eradication, we employed CLSM to investigate the cellular internalization of CSIR in K7M2 tumor cells. Cellswere incubated with either “free Cu²⁺+SRF + IR780” (a mixture of unbound components) or CSIR. Hoechst 33,342 fluorescent stain (blue) was used to visualize cell nuclei, while IR780 emitted red fluorescence. As depicted in Figure S7a, K7M2 cells treated with CSIR exhibited a significantly stronger red fluorescence intensity than those treated with “free Cu²⁺+SRF + IR780” or PBS alone. This observation suggests a higher degree of cellular uptake for CSIR. These findings were further corroborated by flow cytometry analysis (Figure S7b). Figure 2a further demonstrates a time-dependent increase in red fluorescence intensity, indicating a gradual accumulation of CSIR within the cells over time (0 to 12 h). Notably, the CSIR group displayed a markedly stronger red fluorescence signal than the free drug group (“free Cu²⁺+SRF + IR780”). This observation confirms the effective uptake of CSIR by K7M2 cells. Flow cytometry analysis (Fig. 2b and c) further validated these findings, supporting the feasibility and potential of CSIR for in vivo applications.

To evaluate the in vivo tumor tropism of CSIR, we employed a small animal fluorescence imaging system to track the distribution of CSIR formulations within tumor-bearing Balb/c nude mice. The mice were injected with CSIR samples via their tail veins. As shown in Fig. 2d and e, the experimental data demonstrated a higher fluorescence intensity in the tumors of the CSIR group compared to the group receiving “free Cu²⁺+SRF + IR780”. The CSIR group exhibited the most pronounced fluorescence intensity at 36 h post-injection, which remained detectable within the tumor even 48 h later. This significant tumor accumulation of CSIR can be attributed to the EPR effect [44]. In contrast, the group receiving the free drugs (free Cu²⁺+SRF + IR780) displayed weak fluorescence. In conclusion, these findings convincingly demonstrate the tumor targeting and accumulation capabilities of CSIR. Based on the observed intratumoral accumulation profile, the NIR laser irradiation time was set to 24 h post-injection for the subsequent in vivo antitumor efficacy studies.

Following a 48-hour observation period, the ex vivo fluorescence imaging analysis encompassed the tumor and key organs such as the liver, heart, kidneys, lungs, and spleen (Fig. 2f). Notably, most of the fluorescent signal was localized within the tumor region (Fig. 2g), with a secondary but significant accumulation observed in the lung tissue. This pulmonary uptake can be attributed to the non-specific sequestration of IR780 [45]. Overall, these results suggest the promising tumor-targeting potential of CSIR. In tumor therapy, the selective enrichment of therapeutic agents within tumor tissue and cells is a critical determinant of treatment efficacy. Having established the effective accumulation of CSIR, our subsequent investigation focused on elucidating the potential mechanisms of tumor inhibition at both the cellular and in vivo levels.

To further assess the suitability of CSIR for in vivo applications, we evaluated its temperature-modulating capabilities following administration to tumor-bearing mice. Twenty-four hours post-injection, we employed an infrared (NIR) thermal camera to record thermal changes within the tumor region following 5 min of laser irradiation at 808 nm and a power density of 1.20 W cm^− 2. This approach facilitated real-time temperature monitoring. Following 5 min of laser irradiation, Balb/c nude mice treated with CSIR developed tumor temperatures higher than those of the “free Cu²⁺+SRF + IR780” group and PBS group (Fig. 2h, i), exhibiting its photo-thermal conversion capabilities in vivo.

In vitro GSH depletion and synergistic therapeutic performance in vitro

Glutathione, the primary cellular antioxidant, safeguards cells against oxidative stress. Elevated GSH levels have been documented in various cancer cells, contributing to their survival. Inhibiting GSH synthesis and depleting GSH stores could diminish the adaptive antioxidant capacity, thereby inducing significant oxidative stress within cancer cells [46]. Furthermore, heightened oxidative stress can render cancer cells more susceptible to GSH depletion. As illustrated in the schematic diagram of Fig. 3a, a subsequent experimental investigation was conducted to elucidate the antitumor effect of CSIR and its mechanism of GSH biosynthesis inhibition and intracellular GSH depletion. Each component of our formulated carrier-free nanomedicine plays a distinct role. SRF, as established, hinders GSH biosynthesis by impeding the Xc- transporter system’s Xc- transport activity [28, 29]. We employed 5,5ʹ-dithiobis-(2-nitrobenzoic acid) (DTNB) to assess GSH depletion. As shown in Fig. 3b, a peak absorption was observed at 407 nm, indicative of a substantial quantity of GSH within the cells. Conversely, a progressive decline in the absorption peak intensity at 407 nm was observed with increasing incubation time, potentially attributable to the reaction between GSH and Cu²⁺. As depicted in Fig. 3c, a rise in Cu²⁺ concentration corresponded to a gradual decrease in absorption at 407 nm. These experiments convincingly demonstrate CSIR’s remarkable capacity for GSH depletion.

The generation of reactive oxygen species (ROS) by CSIR in vitro was assessed using a singlet oxygen sensor green fluorescent probe. As shown in Figs. 3d and e, minimal green fluorescence was observed in the PBS control groups. Similarly, negligible fluorescence signals were detected in the “Cu²⁺” and “SRF” groups, indicating insufficient ROS production via the Cu²⁺-mediated Fenton-like reaction and a minimal effect of sorafenib on ROS levels. In contrast, cells treated with samples containing IR780 and exposed to subsequent laser irradiation exhibited a marked increase in green fluorescence, suggesting significant ROS generation by IR780 under laser irradiation. This effect compensates for the limited ROS production from Cu²⁺ and SRF. Notably, the CSIR group displayed the most pronounced fluorescence intensity among all groups (Figure S5c). These results demonstrate the efficient cellular uptake of our carrier-free nanomedicine formulation, which leverages the synergistic effects of its individual components to induce substantial ROS production.

Subsequently, cell viability was assessed by staining the cells with calcein-AM and propidium iodide (Calcein-AM/PI). As shown in Fig. 3g, cells treated with individual drugs (Cu²⁺, SRF, or IR780) primarily displayed live cells (green fluorescence), similar to the PBS control group, indicating that these agents alone do not exhibit significant antitumor effects. Notably, only a few dead cells (red fluorescence) were observed in the SRF + IR780 and Cu²⁺+SRF + IR780 groups, while a substantial number of dead cells were detected in the CSIR group, demonstrating its superior antitumor activity. These observations were further supported by the MTT assay results (Fig. 3f), which indicated a 98% antitumor rate for the CSIR group. As shown in Figure S5a and b, under NIR laser irradiation (1.2 W/cm² for 3 min), cell viability in the PBS + NIR group showed no significant difference compared to the PBS group, suggesting that NIR irradiation alone does not affect cell viability. However, in the absence of NIR irradiation, the survival rate of osteosarcoma cells in the CSIR group was 32.3%, with live/dead fluorescence staining revealing only a small number of dead cells. In contrast, under NIR irradiation, the survival rate of osteosarcoma cells in the CSIR group significantly decreased to 2%. These findings suggest that CSIR, when combined with NIR irradiation, induces a remarkably enhanced antitumor effect via multimodal therapy. Even without the photothermal effect, CSIR exhibits a certain degree of osteosarcoma inhibition by releasing Sorafenib and Cu²⁺, which effectively deplete intracellular GSH and significantly reduce its levels. When exposed to NIR irradiation, the photothermal effect synergizes with other therapeutic modalities, substantially enhancing its antitumor efficacy. We have also included data on the release levels of Sorafenib, Cu²⁺, and IR780 from CSIR in PBS, as shown in Figure S6. After 6 h, the release of Sorafenib from CSIR reached a plateau, with a cumulative release of approximately 30%. Similarly, after 8 h, CSIR released 0.97 mg/L of Cu²⁺ and 35.6% of IR780.

Ability of CSIR to induce ferroptosis

Ferroptosis is a regulated form of cell death characterized by reduced intracellular GSH levels, elevated ROS levels, and accumulation of lipid peroxides. Notably, mitochondrial shrinkage is also a hallmark of ferroptosis. We further investigated whether combining these effects could synergistically induce potent ferroptosis. MMP (mitochondrial membrane potential) serves as an indicator of mitochondrial health, structural integrity, and function. In the “Cu²⁺+SRF + IR780” and “CSIR” treated groups, a significant shift in the cell population from the PE channel to the FITC channel was observed by flow cytometry (FCM) analysis. This shift was most pronounced in the “CSIR” group (Fig. 4a), suggesting severe mitochondrial damage. Furthermore, transmission electron microscopy (TEM) was employed to assess changes in K7M2 cell mitochondrial morphology following treatment with different samples. As shown in Fig. 4b, mitochondria in the mixed drugs treated group exhibited damage, while those in the “Cu²⁺+SRF + IR780” and “CSIR” groups displayed severe disruption, characterized by edema, membrane rupture, and cristae breakdown. These TEM findings corroborate the results obtained by FCM analysis. Collectively, these results indicate that CSIR can amplify mitochondrial damage through ROS accumulation and GSH depletion.

As lipid peroxidation is a hallmark of ferroptosis, we measured the amount of peroxidation in K7M2 cells by staining them with BODIPY-C11 581/591. BODIPY-C11 exhibits a red fluorescence signal that transforms to green upon lipid oxidation. As depicted in Fig. 4c, the PBS and Cu²⁺ treatment groups displayed the strongest red fluorescence intensity and the weakest green fluorescence intensity. Conversely, a mild decrease in red fluorescence intensity and a moderate increase in green fluorescence intensity were observed following treatment with either SRF or IR780 alone. Notably, these changes were further amplified in the combination treatments of IR780 with Cu²⁺ or sorafenib. As expected, the combined treatment of IR780 with sorafenib and Cu²⁺ mostly exhibited green fluorescence, suggesting that the combination of three drugs could significantly promote the accumulation of lipid peroxides. Importantly, a synergistic interaction among the components within our formulated carrier-free nanomedicine effectively mediated a substantial decrease in reduced GSH levels in tumorigenic cells.

Ferroptosis is characterized by the inactivation of GPX4, either through a decrease in GSH levels or direct inhibition during ferroptosis. Additionally, the inhibition of xCT by sorafenib leads to a reduction in GSH synthesis. Western blot analysis of K7M2 cells treated with various interventions revealed alterations in the expression of GPX4 and xCT (Fig. 4d). Notably, the SRF group and groups containing SRF exhibited a reduction in xCT expression compared to the other groups, with the CSIR group showing the lowest expression level (Fig. 4e). Moreover, Fig. 4f demonstrates a significant downregulation of GPX4 expression in the CSIR-treated group upon NIR irradiation. This suggests that CSIR, when exposed to NIR irradiation, can downregulate GPX4 expression via the GSH/GPX4 axis [23], thereby enhancing lipid peroxidation and promoting ferroptosis. Collectively, these results indicate that CSIR induces ferroptosis by depleting GSH and inhibiting GSH biosynthesis.

Potential mechanisms of CSIR mediated ferroptosis

To elucidate the molecular mechanisms underlying the anti-tumor activity of CSIR, a transcriptome sequencing analysis was employed to compare the gene expression profiles of K7M2 tumors treated with CSIR to those of a control group. To elucidate the molecular signaling pathways associated with ferroptosis, we employed DAVID software for differential expression analysis. Principal component analysis (PCA) revealed a substantial divergence in gene expression profiles between the two K7M2 cell groups (Fig. 5a). Figure 5b presents a cluster analysis heatmap of differentially expressed genes (DEGs), where the bottom axis represents samples and the top axis.

represents sample clusters. Furthermore, a volcano plot (Fig. 5c) visually depicts the differentially expressed genes between the two groups, with red signifying upregulated DEGs and green signifying downregulated DEGs. Functional enrichment analyses using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases were performed on the DEGs identified in CSIR-treated K7M2 tumors (Fig. 5d and e). These analyses revealed a spectrum of functions and pathways associated with the DEGs. GO analysis indicated that the DEGs participate in various biological processes (BP), including small GTPase-mediated signal transduction, chromatin organization, and regulation of cell morphogenesis. Regarding molecular function (MF), the DEGs were enriched for activities such as GTPase regulation, nucleoside-triphosphatase regulation, and GTPase activation. Intriguingly, KEGG pathway enrichment analysis highlighted the intersection between differentially expressed genes, iron death-related genes, and ferroptosis signaling as the most significant pathway (Fig. 5f). Within the enriched ferroptosis.

signaling pathway genes, GPX4 exhibited the most pronounced difference between the CSIR-treated and PBS-treated groups. Furthermore, SLC genes were also ranked among the top 10 most differentially expressed genes (see Figure S8a and b).

In vivo antitumor efficiency of CSIR

To assess the in vivo antitumor efficacy, nude mice harboring K7M2 tumors were randomly assigned to eight groups upon reaching a predetermined tumor volume of approximately 200 mm³. A schematic of the experimental scheme can be found in Fig. 6a. The laser parameters used in the in vivo experiments for CSIR were as follows: laser power of 1.2 W/cm², treatment duration of 5 min, administered once every other day. Tumor volume measurements were obtained every other day, and the resulting data points were used to construct tumor growth curves. Neither the PBS nor the Cu²⁺ group exhibited a significant effect on tumor growth over the course of 12 days, as shown in Fig. 6b and c. Conversely, the treatment groups receiving SRF and IR780 demonstrated a moderate inhibitory effect on tumor growth. Tumor inhibition was somewhat weaker in the “Cu²⁺+IR780”, in contrast to the “SRF + IR780” group, which had a partial tumor-suppressing effect. The tumor volume in the “Cu²⁺+SRF + IR780” group was considerably reduced after treatment, attributed to Cu²⁺, sorafenib, and IR780 synergistically inducing ferroptosis to enhance the effect of PDT/PTT. Crucially, nude mice in the CSIR group exhibited the most significant tumor growth inhibition following therapy compared to the combined treatment with Cu²⁺, SRF, and IR780 mixtures. This finding underscores the remarkable potential and therapeutic value of these self-assembled nanomedicines in enhancing the efficacy of combination therapy. Twelve days after treatment initiation, tumors were excised, photographed, and measured. Notably, ex vivo tumor size and weight in the CSIR group were demonstrably reduced compared to the unassembled drug groups (Fig. 6d and e). In three mice, complete tumor disappearance was observed, highlighting the exceptional tumor ablation efficacy of CSIR. As shown in Supplementary Figure S9, no significant weight loss was observed in any treatment group throughout the study.

As indicated by the tumor growth curves, there was no significant difference in tumor volume between the single free drug groups (Cu²⁺, SRF, IR780) and the PBS group, which may be attributed to the poor targeting ability of Cu²⁺ and SRF in vivo, as well as the poor hydrophilicity of IR780, ultimately affecting their anti-tumor efficacy. Although the tumor volume in the Cu²⁺+SRF + IR780 group was reduced after treatment, it remained about 31.1% of that in the PBS group, demonstrating some anti-tumor activity. This is mainly attributed to the synergistic induction of ferroptosis by Cu²⁺, Sorafenib, and IR780, which enhances the effects of PDT/PTT. Notably, the CSIR nanocomposite drug, self-assembled from Cu²⁺, Sorafenib, and IR780, exhibited the most impressive anti-tumor effect, with tumor volume significantly different from that of the other seven groups. Furthermore, tumor tissue weight measurements also showed that the CSIR group displayed the best anti-tumor effect. These results indicate that the CSIR nanocomposite drug not only possesses the functions of the individual free drugs but also demonstrates tumor targeting ability, enhancing anti-osteosarcoma therapeutic effects through the induction of ferroptosis combined with PDT/PTT treatment.

Furthermore, analysis of key blood indicators, including uric acid (UA), serum creatinine (CREA), aspartate aminotransferase (AST), alanine transaminase (ALT), blood urea nitrogen (BUN), albumin (ALB), total bile acid (TBA), and direct bilirubin (DBil), revealed values similar to those of the PBS group (Fig. 6f and Figure S12). This suggests the excellent safety profile of CSIR. Additionally, H&E staining of major organs (heart, liver, spleen, lung, and kidney) revealed no discernible pathological changes (Supplementary Figure S10). Moreover, as demonstrated in Supplementary Figure S11, CSIR exhibited good blood compatibility, with hemolysis rates below 8% at various concentrations. Following therapy, tumors were excised for TUNEL, H&E, Ki-67, SLC, and GPX4 staining (Fig. 6g). The results revealed substantial tumor tissue damage and necrosis induction in the CSIR group, whereas the PBS group displayed intact cellular morphology. Furthermore, immunohistochemistry (IHC) analysis was performed to assess the expression of Ki-67 proliferation proteins, along with GPX4 and xCT, which are representative proteins associated with tumor ferroptosis and amino acid transport, respectively. Ki-67, GPX4, and xCT expression levels were significantly lower in the CSIR group compared to other treatment groups. These findings collectively demonstrate the remarkable potential of CSIR nanomedicine for antitumor therapy by triggering ferroptosis in synergy with PDT/PTT therapy.