N-hydroxysuccinimide (NHS), N-(3-dimethylaminopropyl)-N-ethylcarbodiimide hydrochloride (EDC), and folic acid (FA) were obtained from Sigma-Aldrich (St. Louis, MO, United States). BBR (purity >98%) was purchased from Solarbio (Beijing, China) and dissolved with dimethylsulfoxide (DMSO) (Sigma). Other chemicals were commercially available and used without further purification.

The design and synthesis of GO-PEG-FA/BD were illustrated in Scheme 1. Firstly, GO was synthesized by a modified Hummers method described elsewhere (Park and Ruoff, 2009; Huang et al., 2011). Subsequently, the GO-PEG-NH₂ was prepared by an acetylation reaction between carboxyl groups in GO and amine groups in PEG-NH₂ (Kim et al., 2016). And then, GO-PEG-NH₂ was conjugated with FA according to the literature with mild modification (Huang et al., 2020). Briefly, EDC (100 mg) and NHS (50 mg) were dissolved in the DMSO solution of FA (2 mg/mL, 2 mL) and stirred for 2 h to activate the carboxyl groups of FA. Subsequently, the mixed solution was added into the DMSO solution of GO-PEG-NH₂ (2 mg mL⁻¹, 2 mL) and stirred overnight at room temperature, followed by dialysis against DI-water to remove the unreacted reactants. The product (GO-PEG-FA) was washed three times with DI-water and then freeze dried.

For BD loading, the BD was firstly prepared according to the method described in the previous study (Fu et al., 2015). After that, 10 mg GO-PEG-FA were first dispersed in BD (2 mg/mL in DMSO, 10 mL) by ultrasonication and then gently stirred at room temperature for 24 h. Subsequently, the BD-loaded GO-PEG-FA (GO-PEG-FA/BD) was collected by centrifugation at 12,000 rpm for 60 min and washed 3 times with DMSO to remove the unbound drug. The yielded NPs were then washed with PBS for 3 times and stored at 4°C before further use.

Scanning electron microscope (SEM) images were obtained with a transmission electron microscope (ZEISS Gemini 300). The surface charge of samples was evaluated by using a Zetasizer Nano ZS90 equipment (Malvern Instruments, United Kingdom).

The hemolysis experiments were performed by detecting the released hemoglobin. In brief, 2.0 mL blood samples of healthy mice were extracted and anticoagulated. The red blood cells (RBCs) were firstly obtained after 5 min of centrifugation at 8,000 rpm under 4°C and 3 times of rinses with PBS and then diluted by adding 4.0 mL PBS. Afterwards, 0.8 mL PBS with various amounts of NPs was added into 0.2 mL the above RBCs suspension with the final concentration of NPs was 50, 100, 200, 400 and 800 μg mL⁻¹, respectively. RBC suspension plus PBS or pure water without NPs were used as negative and positive control, respectively. After 2 h of incubation at 37°C under shaking, the mixed systems were centrifuged at 8,000 rpm and the absorbance at 541 nm of the obtained supernatant was determined with a UV-vis spectrophotometer. The hemolysis percentage of RBCs was calculated according to the protocol described elsewhere.

The MNNG/HOS cells (human OS cells) used in the experiment were purchased from Cell Bank of Shanghai Institute of Biochemistry. For incubation of the MNNG/HOS cells, α-modified essential medium (MEM) with 1% antibiotics (penicillin and streptomycin) and 10% fetal bovine serum (FBS) was used. The bone marrow stromal cells (BMSCs) were kindly provided by Dr. Song Gong (Tongji Medical College, Huazhong University of Science and Technology). The BMSCs were cultured with Dulbecco’s MEM (DMEM)/F12 containing 15% FBS and 1% penicillin–streptomycin. All the cells were maintained in a humidified incubator with 5% CO₂ at 37°C.

To assess cancer cell uptake ability of the nanoparticles, cells were planted in a 6-well plate (1 × 10⁵ cells/well) and cultured at 37°C for 24 h. After 4 h of incubation with various samples (BD, GO-PEG/BD and GO-PEG-FA/BD) at equivalent concentration of BD (10 μg mL⁻¹), the cells were incubated with Hoechst 33342 (10 μg mL⁻¹) for 10 min and rinsed for three times with PBS. Subsequently, the fluorescence of cells was recorded by a fluorescence microscope.

In brief, the MNNG/HOS cells were firstly seeded (1 × 10⁵/well) in six-well cell culture plates and incubated overnight. Subsequently, the cells were incubated with various samples for 6 h and then stained with Mito-Tracker Red (Beyotime Biotechnology, China) as well as Hoechst 33432 (10 μg mL⁻¹) for 10 min. After 3 times rinses with PBS, the cells were observed and imaged by a fluorescence microscope.

Cell counting kit-8 (CCK-8, Dojindo, Kyushu Island, Japan) assay was employed to evaluate the cytotoxicity in vitro. In brief, cells were seed at a density of 5,000 cells/well in 96-well plates and culture at 37°C for 24 h. Then, the cells were maintained in dark for another 24 h. After the corresponding treatment, the culture medium was then replaced with 100 μL of the corresponding culture medium containing 10% CCK-8 solution and cultured at 37°C for 2 h. Finally, the absorbance was measure at 450 nm using a microplate reader (Biotek, Winooski, VT, United States).

The apoptosis-inducing effect was determined by Annexin V-FITC/propidium iodide (PI) Apoptosis Kit (Nanjing Keygen Biotech, Nanjing, China). Briefly, the cells were seeded in a 6-well plate at a density of 1 × 10⁵ cells/well and cultured overnight. After the indicated treatment and subsequent 24 h of incubation, the cells were collected and rinsed twice with PBS. The Annexin V-fluorescein isothiocyanate (FITC)/PI kit was employed for staining of the above cells according to the manufacturer’s protocol. After another 20 min incubation in the dark, the cells were analyzed by flow cytometry (Becton Dickinson, Franklin Lakes, New Jersey, United States).

The intracellular ROS production was detected by a ROS assay kit (Beyotime Company, Shanghai, China). In brief, the MNNG/HOS cells were seeded in a six-well cell culture plate (1 × 10⁵/well) and incubated for 24 h. After the corresponding treatment, cells were washed twice with PBS, incubated with DCFH-DA reagent (10 µM) in medium without FBS at 37°C for 30 min, and then washed with PBS three times. After that, the fluorescence intensity of the cells was detected by a fluorescence microscope.

The MNNG/HOS cells were firstly seeded and cultured for 24 h in six-well cell culture plates. Following 6 h of corresponding treatment, the cells were treated by an NIR laser for 10 min.

For detection of the change of MMP, the 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolocarbocyanine iodide (JC-1) detection Kit was employed. Briefly, the cells were incubated with JC-1 (5 μg mL⁻¹) for 20 min, washed by PBS and then observed by fluorescence. For detection of the cellular ATP, the ATP determination kit (Beyotime, China) was used. In brief, the treated cells were harvested and thoroughly lysed with the lysis buffer. Afterwards, the supernatant was obtained by 5 min of centrifugation at 12,000 rpm under 4°C and the absorption was recorded. Finally, the ATP levels of the samples were determined according to the standard curve plotted based on the manufacturer’s protocol.

All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at Tongji Medical College, Huazhong University of Science and Technology.

MNNG/HOS tumor-bearing mice were obtained by subcutaneous injection with 200 μL of MNNG/HOS cell suspension at the density of 1 × 10⁷ cells/mL. When tumors grew to about 150 mm³, the mice were administered GO-PEG-FA/ICG via tail vein injection. After administration, the mice were scanned at the predetermined time and the tumor imaging was obtained by IVIS small animal imaging system (PerkinElmer Inc., Waltham, United States).

MNNG/HOS tumor-bearing mice were obtained as above mentioned. When the tumor reached about 50 mm³, the mice were randomly divided into four groups: 1) PBS + NIR; 2) free BD + NIR; 3) GO-PEG-FA/BD; and 4) GO-PEG-FA/BD + NIR (n = 3 per group). 100 μL PBS or free BD (10 mg kg⁻¹) was injected intraperitoneally every 3 days. Other therapeutic agents were intravenously injected into the mice via the tail vein every 3 days. Six hours after injection, the tumor of all groups were irradiated with/without 808 nm laser for 5 min. The mice were weighted and measured with caliper every other day. Tumor volumes were calculated as length × width²/2 (mm³).

At the end of the treatment, the tumors were collected. Hematoxylin and eosin (H&E) staining assay and terminal deoxynucleotidyl transferase-mediated dUTP Nick-End Labeling (TUNEL) assay were performed to assess the anti-tumor efficacy of different groups. To evaluate the safety of all the groups, the major organs (heart, liver, spleen, lung, and kidney) were harvested at the end of treatment and stained with H&E staining. Furthermore, the plasma was also collect to perform blood chemistry analysis.

All of the experimental data were analyzed at least three independent experiments under the same experimental conditions. The results were expressed as the mean values ±standard deviation (SD). All the statistical analyses were tested by using Student’s t-test and one-way analysis of variance (ANOVA) conducted on the software GraphPad Prism version 6.01 for Windows. A p-value <0.05 was considered as statistically significant.

The detailed design of GO-PEG-FA/BBRG was showed in Figure 1A. Firstly, GO was synthesized using the modified Hummers method. As shown in Figure 1B, the morphology of as-prepared GO-PEG-FA/BD was characterized by SEM. As demonstrated in Figure 1C, these altered zeta potentials provided evidence for the successful synthesis of the NPs.

Due the good photothermal conversion efficiency, GO was usually used as a photothermal agent for PTT against cancer (Guo et al., 2022). In order to investigate the photothermal properties of GO-PEG-FA/BD in vitro, an infrared thermal imaging camera was used for real time measurement of the temperature changes of NPs after NIR irradiation (808 nm). As displayed in Supplementary Figure S1A, the temperature of GO-PEG-FA/BD NPs exhibited a promising concentration-, time-, and laser power intensity-dependent manner. However, it could be found the temperature of PBS rose slightly under the same conditions. In addition, the photothermal stability of the NPs upon the NIR irradiation was further evaluated. As shown in Supplementary Figure S1B, the temperature variation curves and peak shape were of no significant change after five cycles of irradiation with an 808 nm laser. Overall, all these results suggested that the nanocomposites exhibited superior photothermal conversion property and excellent photothermal stability.

The blood hemolysis assay was first applied to investigate the in vitro hemocompatibility of GO-PEG-FA NPs. As demonstrated in Figure 1D, no significant hemolysis of RBCs was observed in the GO-PEG-FA NPs even at the highest concentration of 800 μg mL⁻¹. In marked contrast, severe RBCs hemolysis happened to the positive control (pure water). These results demonstrated that the GO-PEG-FA NPs possessed an excellent hemocompatibility for peripheral blood circulation. In addition, the cytotoxicity of GO-PEG-FA NPs was further assessed by CCK8 assay. As displayed in Supplementary Figure S2, GO-PEG-FA NPs showed less toxicity to MNNG/HOS cells and normal human cells (BMSCs) after 24 h treatment.

All these findings suggested that the as-synthesized GO-PEG-FA/BD could be utilized as an extraordinarily biocompatible nanoplatform for drug delivery.

Numerous studies have shown that FA modified nanoparticles could target various cancer cells including OS (Hu et al., 2021; Mi et al., 2021). Accordingly, we modified FA molecules on the surface of the NPs (GO-PEG) hoping to achieve active tumor targeting by the specific interaction with FA receptor. In order to investigate the cellular uptake behavior of the NPs, cells were incubated with various drug formulations and then examined by using a fluorescence microscope. As shown in Figure 2A, the fluorescence intensity in MNNG/HOS cells treated with GO-PEG-FA/BD was dramatically stronger than that of the cells treated with GO-PEG/BD and free BD. In addition, the FA molecule competing assay illustrated that the uptake of GO-PEG-FA/BD was dramatically reduced by the pretreated free FA molecules. These results suggested that the enhanced uptake of GO-PEG-FA/BD might be partly mediated by FA-mediated endocytosis, which provided the potential for tumor-targeted drug delivery.

To further investigate the mitochondrial targeting effects of the nanocomposites, the sub-cellular localization of the NPs was evaluated by a fluorescence microscope. As shown in Figure 2B, free BD (emitted green fluorescence) could partially accumulate into the mitochondria stained as red fluorescence by Mito Tracker Red. Comparatively, higher green fluorescence in mitochondria was observed in cells treated with GO-PEG-FA/BD. This result further confirmed that the NPs exhibited the appreciable tumor target ability and excellent mitochondria-targeting ability.

After demonstrating the excellent mitochondrial targeting ability of the NPs, the synergistic antitumor efficacy was first investigated in vitro. MNNG/HOS cells were treated with various formulations and determined with the CCK8 assay for cytotoxicity testing. As shown in Figure 2C, the formulation of GO-PEG-FA/BD showed higher cytotoxicity to MNNG/HOS cells than free BD under the same condition, which might be attributed to the enhanced internalization of nanodrugs through FA-receptor mediated endocytosis. Furthermore, GO-PEG-FA/BD couple with a low temperature laser irradiation displayed an enhanced cell-killing effect, indicating the nanocomposite with NIR laser irradiation achieved a prominent synergistic antitumor effect. Additionally, the apoptosis-inducing effect was quantitatively investigated by using flow cytometry with Annexin-V- FITC/PI double staining. As expect, GO-PEG-FA/BD under a low temperature laser irradiation could induce a dramatically higher level of cell apoptosis (Figure 2D). Collectively, all these results confirmed the prominent synergistic therapeutic effect of mitochondria-targeting LPTT.

Mitochondria are vital metabolic organelles playing an essential role in regulating the adenosine triphosphate (ATP) production and intrinsic cell apoptosis (Sood et al., 2021). Therefore, these subcellular organelle is certainly to be considered as an excellent therapeutic target for cancer treatment. Increasing evidence has suggested that the mitochondria damage might directly activate the intrinsic mitochondrial apoptosis pathways, ultimately inducing cell apoptosis (Tan et al., 2018; Wang et al., 2019). It is hypothesized that mitochondria-targeting LPTT using GO-PEG-FA/BD could eliminate membrane potential (ΔΨm), deprive the ATP of cancer cells, and increase the levels of ROS, which ultimately induce cell apoptosis.

The mitochondrial membrane potential changes were firstly investigated by using mitochondrial membrane potential assay kit with JC-1. As shown in Figure 3A, GO-PEG-FA/BD couple with a low temperature laser irradiation exhibited the strongest green fluorescence, indicating the destruction of membrane integrity. Meanwhile, ATP levels were significantly reduced in the cells treated with GO-PEG-FA/BD couple with a low temperature laser irradiation (Figure 3B). Consistent with these results, we found that GO-PEG-FA/BD + NIR showed the greatest potency in stimulating the production of ROS (Figure 3B). In brief, all the results above exhibited that MNNG/HOS cells treated with GO-PEG-FA/BD couple with a low temperature laser irradiation underwent the mitochondrial depolarization, ATP deprivation, and substantial ROS generation, ultimately inducing mitochondrial associated apoptosis in cancer cells.

To evaluate the in vivo distribution of the nanocomposites, we firstly established the MNNG/HOS tumor-bearing mouse model. Subsequently, the mice were intravenously injected with ICG labeled GO-PEG-FA (GO-PEG-FA/ICG) and then observed at the predetermined time by an IVIS small animal imaging system. As illustrated in Figure 4A, the fluorescence signals in the tumors gradually increased over time and the fluorescence intensity reached a maximum at 6 h post-injection. It was worth noting that the tumor site still remained the strongest fluorescence after 24 h administration, indicating the long retention of the NPs. The efficient tumor enrichment effect might be ascribed to the enhanced permeability and retention (EPR) effect-induced passive targeting and FA-mediated active targeting delivery (Park et al., 2019; Zhao et al., 2019; Hu et al., 2021).

Based on the above findings, we chose 6 h post-injection as the optimal time point for photothermal treatment. After 6 h post-injection, photothermal images from the tumor were taken to monitor the changes of temperature. As displayed Figure 4B, the local tumor temperature of the mice treated with GO-PEG-FA/BD could reach about 45°C after irradiation at 1.0 W cm⁻² for 5 min. However, the temperature in the PBS treated group increased slightly under the same treatment conditions. These findings demonstrated that the nanocomposites could be employed to realize the quick increase of the temperature in vivo.

Encouraged by the effective tumor accumulation of GO-PEG-FA/BD and the satisfying anticancer effect in vitro, the in vivo synergistic combination antitumor therapeutic efficiency was further evaluated by a tumor-bearing mouse model established using MNNG/HOS cells. As indicated in Figures 4C, D, the mice treated with GO-PEG-FA/BD + 808 nm laser irradiation showed remarkable inhibition on tumor growth, which verified the outstanding synergistic therapeutic effect. Of note, there was no significant change in the weight of the mice in all groups during the 14 days of monitoring (Figure 4E). In addition, no significant tissue damage was observed in all the groups, confirming the favorable biological safety of all the treatments (Figure 5).

To further confirm the synergistic therapeutic efficacy in vivo, the tumors were harvested for HE staining and TUNEL assay (Figure 4F). In HE staining, the tumor tissue from the mice treated with GO-PEG-FA/BD + NIR showed the largest area of tumor necrosis compared to other groups, which was in good agreement with the anticancer results. As exhibited in the TUNEL staining images, the significant cell apoptosis was observed in tumor tissues treated with GO-PEG-FA/BD + NIR. Meanwhile, the liver and kidney function was evaluated by determining the serum level of alanine aminotransferase (ALT) and blood urea nitrogen (BUN), respectively. As shown in Supplementary Figure S3, the levels of ALT and BUN were all within normal range. Over all, all these results collectively certified the excellent synergistic antitumor effects and remarkable systemic safety of the mitochondria-targeting LPTT.