*2.12. Cytotoxicity Test*

To investigate the toxicity of blank carrier material, the prepared SA nanoparticles (SNs) and RBCM-SA nanoparticles (RSNs) without ATO were diluted with DMEM complete medium to obtain SA concentrations of 8, 12, 20, 30, 40, 50, 60 μg/mL respectively. To investigate the toxicity of the nanoparticles after drug loading, free ATO solution, SANs, and RSANs were diluted to obtain di fferent concentrations of ATO (1, 2, 4, 6, 8, 10, 12, 20 μg/mL). The above groups were administration groups (AG). 100 μL 293 cells (5 × 10<sup>4</sup> cells/mL) were placed into a 96-well culture plate for 24 h. The nutrient medium was then replaced with 100 μL AG. At 24 h time interval, all the groups were cultured with 100 μL CCK-8 for 2 h. Only DMEM medium was set as a blank group (BG) and only 293 cells were set as a control group (CG). The optical density (OD) of samples were determined at 450 nm by a microplate reader and the following formula was used to calculate the cell viability.

$$\text{Cell viability (\%)}=\text{(ODAG}-\text{ODBG)} \text{(ODCG}-\text{ODBG)} \times 100\%$$

### *2.13. In Vitro <sup>E</sup>*ffi*cacy Study*

NB4 and 7721 cells were selected to evaluate efficacy. 100 μL NB4 and 7721 cells (5 × 10<sup>4</sup> cells/mL) were inoculated in a 96-well culture plate overnight, respectively. Then free ATO solution, SANs, and RSANs were administered to the cells, respectively. The concentrations of ATO were diluted to 1, 2, 4, 6, 8, 10, 12, 20 μg/mL. After 24 h, 100 μL CCK-8 were administered to evaluate the cell viability. To further investigate the inhibitory effect, the concentration of the ATO of each group was fixed at 1 μg/mL. The OD was measured after incubation of 4, 8, 12, 24, 36, 48, and 60 h, and the cell viability was determined. The blank and control groups were same as mentioned in Section 2.12.

#### *2.14. In Vivo Toxicity and Safety Test*

Due to the potent toxicity of ATO, the administration concentration at a safe level should be determined at first. For 2 weeks, ATO with high (40 μg/mL), medium (20 μg/mL) and low concentration (10 μg/mL) was administered through the tail vein respectively once a week at a dose volume of 0.2 mL per mouse. The mental state and death of nude mice were registered during the period.

The aim of the safety trial was to investigate whether the continuous intravenous injection of RSANs would cause lesions on systemic, hematological, and major organs. Afterward, the healthy nude mice were divided into the saline group, ATO group, SANs group, and RSANs group. The drugs were administered once every 2 days at dose volume of 0.2 mL per mouse and repeated seven times. The weight of the mice was recorded at 1, 3, 5, 7, 9, 11, 13 days after administration. Meanwhile, the mental state and death of mice were observed during the process. 2 days after the last administration, orbital blood was collected into the tubes, pre-mixed with heparin sodium, and white blood cells (WBC), glutamate pyruvic transaminase (ALT), aspartate aminotransferase (AST) were analyzed. In addition, after the mice were sacrificed by CO2 asphyxiation, the principal organs were excised. The viscera coefficients were calculated after weighing. Furthermore, the tissues were fixed with paraffin solution for immunohistochemical analysis (H & E) to examine its structure and morphology.

> Visceral coefficient = weight of organ/body weight

### *2.15. In Vivo Anti-Tumor Studies*

Male nude mice 6–8 weeks old with 7721 cells were used to investigate the anti-tumor effects of our nanoparticles. First we established xenograft tumor model. For the establishment of tumor model, 7721 cell suspension (2 × 10<sup>7</sup> cells / mL) in a volume of 0.2 mL was injected subcutaneously in the armpit of the upper limb. In addition, then the formula width<sup>2</sup> × length/2 was used to calculate the tumor volume every other day. At tumor volume of 100–250 mm3, then mice were grouped into four treatment groups (*n* = 5): (1) saline, (2) ATO, (3) SANs, and (4) RSANs. The drugs were administered intravenously at a dose of 1.3 μg/g every two days and repeated seven times. The CG was given normal saline for seven times. At the same time, the body weight and the tumor volume were recorded to evaluate for tumor inhibition efficacy as well as systemic toxicity. Two days after the last administration, the mice were sacrificed by CO2 asphyxiation, and the tumors were excised from each animal. The tumors were rinsed with normal saline, dried, and photographed, and the average tumor weight of each group was determined.

### *2.16. Statistics and Data Analysis*

Data expression was shown as ± SD of the mean. Significant differences between SANs and RSANs were analyzed by Tukey Kramer multiple comparison tests, using GraphPad Prism Software, v.6.01 (GraphPad Software, Inc.). Results with *p* < 0.05 were considered significant and very significant with *p* < 0.01.

### **3. Results and Discussion**

### *3.1. Characterization and Stability Test*

The average size of RSANs was found to be 163.2 ± 4.4 nm (Figure 2A), which increased by about 15 nm as compared to SANs (147.9 ± 5.1 nm). The increment was consistent with the thickness of RBCM [35]. The PDI of SANs and RSANs was 0.24 and 0.27, respectively (Figure 2B), which indicated that NPs were well dispersed. In addition, the particle size of the nanoparticles was stable under 200 nm despite a slight increase in 15 days (Figure 2C) under the storage condition of 37 ◦C. The PDI remained below 0.3 (Figure 2D). These results revealed that NPs were stable. The structure and distribution were also uniform after 15 days. However, while the nanoparticles were stored under condition of 4 ◦C, the particle size increased significantly within 4 days and the flocculation occurred because of the obvious drug precipitation. It suggested that 37 ◦C may be a better storage condition. Moreover, after incubating the nanoparticles with serum for three days, no significant precipitation was observed, indicating that the nanoparticles can remain stable under serum conditions as well.

**Figure 2.** Particle size ( **A**) and PDI (**B**) of SANs and RSANs. Changes of particle size ( **D**) and PDI ( **C**) within 15 days. Data are shown as ± SD (*n* = 3) of mean (*n* = 3).
