MD simulations
MD simulations were performed for RHE and DOX to help us understand how the molecules interact with each other in aqueous solutions. The RHE and DOX molecules were initially in a dispersed state. The π-π stacking interactions formed between RHE and DOX molecules at 0.5 ns; then these small-size clusters gradually accumulated and formed relatively large clusters at 2.0 ns. A stable aggregate was ultimately formed within 11 ns (Additional file 1: Figure S1).
Furthermore, the number of intermolecular hydrogen bonds and π-π stacking interactions between DOX and RHE molecules increased, indicating that RHE and DOX self-assembled in water to form RD NPs through π-π stacking and hydrogen interactions (Additional file 1: Figure S2). By analyzing the solvent-accessible surface areas (SASA) and the number of hydrogen bonds between the co-assembled structures and solvent water, hydrophobic interactions helped DOX and RHE molecules form the co-assembled structures (Additional file 1: Figure S3). Hence, hydrogen bond interactions, hydrophobic interactions and π-π stacking interactions would contribute to the co-assembly process.
In addition, the comparison to MD simulations of pure RHE and DOX alone in aqueous solution indicated that more π-π stacking and intermolecular hydrogen bonds formed between DOX and RHE molecules in RD NPs than in pure RHE and DOX alone (Additional file 1: Figure S4 and S5), suggesting that the driving force for self-assembly in individual drug molecules would be weaker than in the mixture. In the mixed system, DOX and RHE preferentially interacted to form RD NPs.
Preparation and characterization of RD NPs
The assembled RHE particles formed irregular sheet-like structures, which were characterized using TEM (Fig. 2A). When the DOX was added to the RHE suspension (RHE/DOX molar ratio = 1:1), DOX molecules coassembled with RHE via interaction forces and formed rod-like RD NPs (Fig. 2B) with diameters of approximately 240 nm (Additional file 1: Figure S6). DOX was clear and transparent in aqueous solution, and the solution of RD NPs showed obvious Tyndall light scattering (Fig. 2D). The DLS analysis of RD NPs revealed a narrow monomodal distribution with a small mean hydrodynamic diameter of 249.90 ± 5.20 nm and a PDI of 0.14 ± 0.03, which was within the accepted range for efficient EPR and ensured a passive tumor targeting effect. The surface charge of RD NPs was determined to be 25.67 ± 1.03 mV (Additional file 1: Figure S7). The EEs of RHE and DOX were 96.23 ± 0.22% and 50.98 ± 7.72%, respectively (Table 1). For comparison, the EEs of erlotinib and DOX were 50% and 93% in erlotinib/DOX codelivering nanoparticles, indicating that compared with other codelivering nanoparticles, the RD NPs exhibited an improved drug-carrying capacity [44]. Furthermore, the emission intensity of RD NPs was lower than the aqueous DOX solution due to the aggregation induced by π-π stacking interactions (Fig. 2C). RD NPs showed an increased emission intensity compared to RHE, indicating that the aggregation of RHE was partially inhibited by the interactions between RHE and DOX. This inhibition may be attributed to the encapsulation of RHE in the hydrophobic domains of DOX.
Table 1. The encapsulation efficiencies (EEs) of RHE and DOX in their self-assembled nanoparticles (NPs).
Molar ratio
|
RHE (%)
|
DOX (%)
|
RHE/DOX=1:0.5
|
98.29 ± 0.58
|
70.26 ± 3.47
|
RHE/DOX=1:1
|
96.23 ± 0.22
|
50.98 ± 7.72
|
RHE/DOX=1:2
|
94.14 ± 2.51
|
30.54 ± 5.81
|
Colloid stability plays an important role in the biomedical application of nanoscale drug-delivery systems. No significant changes in the hydrodynamic diameter or the PDI of RD NPs were observed for up to 8 days at 4 °C (Fig. 2E). The diameter measured using DLS remained unchanged over 8 days, suggesting the excellent stability of RD NPs at room temperature (Additional file 1: Figure S8). Additionally, the coassembled NPs retained good colloidal stability in 10% FBS (Fig. 2F). Although no surfactants or excipients were applied, the RD NPs displayed desirable stability without any precipitation and phase separation.
RHE, DOX, and RD NPs were dispersed in buffer solutions at pH 7.4 and 5.0 to measure the release profile at 37 °C for 24 h. In Fig. 2G and H, a significant difference in the release properties of free RHE and DOX was not observed at different pH values, and approximately 80% of the drugs was released during the first 2 h. In contrast, the release rate was slower in the RD NPs than in the free RHE and DOX solutions. Approximately 25% of RHE and 20% of DOX were released from RD NPs (pH = 7.4) in 2 h. A total of 73.95±8.93% of RHE and 92.07±1.78% of DOX were released from NPs at pH 5.0 within 24 h; both release rates would improve the therapeutic efficacy. In contrast, approximately 62.70% of ursolic acid is released from ursolic acid NPs under the same conditions [45]. Moreover, RHE and DOX release from RD NPs were best modeled using a first-order kinetic model with R2>0.99 (Table S1). The relatively slower RHE release rate might result from the carboxyl groups of RHE, which are more readily protonated under acidic conditions, decreasing its solubility and decelerating its diffusion. Similarly, the slightly faster release of DOX is potentially due to the increased solubility associated with the protonated amino group under acidic conditions [46].
Cellular uptake analysis
We compared the cellular uptake of RHE, DOX, RHE/DOX and RD NPs by 4T1 breast cancer cells to verify that RD NPs were efficiently internalized by cancer cells. According to the CLSM images (Fig. 3A), RD NPs noticeably increased the cellular uptake of RHE and DOX, leading to much higher intensities of green and red fluorescence in 4T1 cells compared with the free RHE, free DOX and RHE/DOX groups. These data were also qualitatively corroborated by the line-scanning profiles of fluorescence intensity over selected cells (Fig. 3B). The green and red lines represent RHE and DOX fluorescence intensities, respectively. The fluorescence intensity of RHE and DOX in the RD NPs was much higher than the RHE, DOX or RHE/DOX groups, and the colocalization of red and green fluorescence showed that DOX and RHE were codelivered to the cells. The increased accumulation of these drugs in 4T1 cells following treatment with the nanosuspension was expected to improve their antitumor effect.
In vitro antitumor efficacy
The in vitro cytotoxicity of RHE, DOX, RHE/DOX and RD NPs was assessed in 4T1 cells using MTT assays to evaluate the antitumor efficacy of RD NPs. As shown in Additional file 1: Figure S9A, B and C all formulations exerted dose-dependent effects on 4T1 cells. Free RHE and free DOX had IC50 values of 73.86 ± 4.40 and 1.30 ± 0.22 μM, respectively (Table 2). Compared with the single-drug treatments, combination therapy with RHE and DOX resulted in relatively lower cell viability. The IC50 of DOX was reduced to 0.84 ± 0.16 μM when combined with RHE in the mixture. Furthermore, the cell viability in the RD NP group was further reduced, with an IC50 of 0.63 ± 0.23 μM. The IC50 of PTPNs in 4T1 cells was 63-fold lower than that of free paclitaxel, and the IC50 of RD NPs was 73-fold lower than that of free RHE, indicating that RD NPs showed better antitumor efficiency than PTPNs [47]. The increased antitumor efficacy of RD NPs was potentially attributed to several factors. RHE and DOX were simultaneously internalized in the cells, which allowed both drugs to support the individual antitumor effects of the other drug. In addition, RD NPs were taken up more efficiently by cells than free drugs. In addition, we calculated the CI values of RHE/DOX and RD NPs. As shown in Table 2, the CI value of RHE/DOX was 0.68, indicating a strong synergistic antiproliferative effect. Moreover, RD NPs had a CI value of 0.51; therefore, the relatively profound synergistic effect of RD NPs was consistent with the cytotoxicity results.
Table 2. The IC50 values of RHE/DOX and RD NPs in 4T1 cells.
|
IC50 (μM)
|
Samples
|
RHE
|
DOX
|
RHE
|
73.86 ± 4.40
|
-
|
DOX
|
-
|
1.30 ± 0.22
|
RHE/DOX
|
2.53 ± 0.43
|
0.84 ± 0.16
|
RD NPs
|
1.88 ± 0.60
|
0.63 ± 0.23
|
Apoptosis refers to a programmed cell death pathway controlled by genes that maintain the stability of the internal milieu. The apoptosis rate is an important index for evaluating the therapeutic effects of antineoplastic agents. We measured the apoptosis induced by RD NPs by performing double staining with annexin V-FITC and PI. As displayed in Fig. 3C, RHE/DOX and RD NPs showed a stronger ability to induce cell apoptosis (31.30 ± 4.69% and 37.40 ± 5.88%, respectively) than free DOX (21.27 ± 2.08%). Meanwhile, the RD NP group induced an even higher level of apoptosis than the RHE/DOX group, which was probably due to the increased cellular uptake of the NPs compared to the free drugs.
In vitro anti-metastatic effects
Next, the ability of RD NPs to inhibit the metastasis of 4T1 cells in vitro was detected. First, a cell scratch assay was applied to investigate cell migration, and the images of cells preincubated with various formulations were evaluated after scratching to evaluate the inhibitory effect of RD NPs on cell migration (Fig. 4A). As displayed in Fig. 4A and Additional file 1: Figure S10, the scratch healing rates of cells treated with RHE, DOX, RHE/DOX and RD NPs were 39.64 ± 5.84%, 53.66 ± 2.57%, 35.29 ± 3.22% and 28.32 ± 3.84%, respectively. Thus, RHE/DOX and RD NPs exerted a stronger inhibitory effect on the scratch healing rates than DOX. Moreover, migration and invasion assays (Fig. 4D) were performed to further determine the inhibitory effects of RHE, DOX, RHE/DOX and RD NPs on 4T1 cell migration. As exhibited in Fig. 4B and Additional file 1: Figure S11A, the cells treated with RHE, DOX, RHE/DOX and RD NPs exhibited decreased migration compared to the control, with migration rates of 53.05 ± 8.47%, 61.22 ± 2.08%, 47.42 ± 3.21% and 40.08 ± 8.54%, respectively, indicating that RHE/DOX and RD NPs exerted stronger inhibitory effects on migration than free DOX. Furthermore, the results of the invasion assay were consistent with the data obtained from the scratch healing and migration assays (Fig. 4C and Additional file 1: Figure S11B).
The inhibition of cell migration by RHE/DOX and RD NPs might be attributed to a decrease in the level of the NF-κB protein, which is an essential contributor to metastasis. NF-κB is a critical intermediate involved in the progression of cell migration, invasion and proliferation; thus, the NF-κB protein plays relatively important roles in cell metastasis and apoptosis. Therefore, the expression of a variety of proteins, including NF-κB P65, the metastasis-related protein MMP-9, the proapoptotic protein Bax and the antiapoptotic protein Bcl-2, were analyzed in 4T1 cells treated with RD NPs using Western blotting to detect the mechanisms by which RHE/DOX and RD NPs inhibit cell migration. As shown in Fig. 4E and Additional file 1: Figure S12, after treatment with RHE/DOX or RD NPs, the levels of NF-κB P65 and MMP-9 were significantly decreased compared with those of cells treated with DOX alone (p < 0.01). Furthermore, compared with the DOX group, the RD NP group presented significantly decreased Bcl-2 levels (p<0.001). In addition, Bax levels were significantly increased in the RHE/DOX and DOX groups (p<0.01). Thus, the anti-metastasis and apoptosis mechanism of the combination of RHE and DOX included the inhibition of NF-κB P65. Moreover, the RD NP group exhibited the lowest level of NF-κB P65 due to the increased cellular uptake of the NPs in this group compared to the uptake of drugs in the other groups.
Pharmacokinetic and biodistribution analyses
The pharmacokinetic behaviors of RHE and DOX were investigated after the intravenous injection of the three formulations. As shown in Fig. 5A and 5B, RD NPs exhibited a prolonged circulation time compared to RHE, DOX and RHE/DOX after an i.v. injection. The plasma concentrations of free RHE, free DOX or RHE/DOX decreased rapidly within 12 h, while RD NPs exhibited a significantly delayed blood clearance. The analysis of the pharmacokinetic parameters revealed that RD NPs increased the half-life of RHE from 1.81 h to 6.87 h (Table 3). Similarly, the RD NPs extended the half-life of DOX from 2.92 h to 7.14 h. While erlotinib/DOX codelivering nanoparticles extended the half-life of DOX from 1.81 h to 3.19 h [44], the use of RD NPs as anticancer codelivering nanoparticles is advantageous over erlotinib/DOX codelivering nanoparticles in prolonging the circulation time. In addition, the area under the curve (AUC0-∞) of RHE increased by ~11.36-fold for RD NPs compared to free RHE. Similarly, the AUC0-∞ of DOX increased by ~9.44-fold for RD NPs compared to free DOX. Based on these results, RD NPs showed improved pharmacokinetic profiles and might exhibit superior synergistic antitumor efficacy in vivo.
Table 3. Main pharmacokinetic parameters of drugs in rats after the intravenous injection of different drugs (n = 5).
Drugs
|
Formulations
|
Cmax (μg/mL)
|
T1/2 (h)
|
AUC0-∞ (mg/mL·h)
|
CL (L/h/kg)
|
MRT0-∞ (h)
|
RHE
|
Free RHE
|
0.35±0.10***
|
1.81±0.57*
|
0.94±0.52***
|
6.25±2.65*
|
2.51±0.98
|
RHE/DOX
|
0.62±0.29***
|
2.15±0.10*
|
1.25±0.52***
|
4.49±1.86*
|
2.59±1.07
|
RD NPs
|
5.46±0.92
|
6.87±2.11
|
10.68±1.04
|
0.47±0.04
|
5.20±2.31
|
DOX
|
Free DOX
|
0.36±0.04*
|
2.92±0.14*
|
1.59±0.06*
|
3.15±0.12***
|
5.07±0.29*
|
RHE/DOX
|
0.52±0.07
|
3.21±0.05*
|
1.61±0.15*
|
3.11±0.30***
|
5.24±0.60*
|
RD NPs
|
2.09±0.48
|
7.14±1.42
|
15.01±5.55
|
0.37±0.14
|
9.63±2.33
|
*p < 0.05, **p < 0.01, and ***p < 0.001 compared to RD NPs.
Encouraged by the increased therapeutic efficiency of RD NPs in vitro, the biodistribution of RD NPs was detected using fluorescence imaging. Since the DOX fluorescence was weak in vivo, the mice were euthanized and tissues (heart, liver, spleen, lung, kidney and tumor) were removed for ex vivo imaging, which is a more accurate qualitative biodistribution analysis. As shown in Fig. 5C and Additional file 1: Figure S13, the quantitative analysis of DOX fluorescence in regions of interest ex vivo substantiated the superior tumor accumulation of RD NPs, as a higher DOX intensity was observed in the tumor at levels 1.68- and 3.38-fold higher than free DOX at 12 and 24 h, respectively. The intratumor distribution of DOX was also observed in frozen tissue sections at 24 h postinjection (Fig. 5D), which showed that treatment with RD NPs led to a higher DOX intensity in the tumor than the free DOX treatment. The drastically higher accumulation of RD NPs was probably due to the long-term circulation and the EPR effect of the NPs.
We conducted biodistribution experiments and performed a quantitative analysis to more intuitively and accurately observe the accumulation of DOX and RHE. Biodistribution assays were performed with RHE, DOX and RD NPs to evaluate the tumor accumulation of RHE and DOX. Free RHE accumulated at lower levels in tumor tissues than RD NPs (Additional file 1: Figure S14). Similarly, lower levels of DOX accumulated in tumor tissues from mice treated with free DOX and RHE/DOX than in mice treated with RD NPs. Moreover, 1.45-fold more RHE accumulated in tumors after treatment with RD NPs than free RHE at 24 h postinjection. Similarly, 3.42-fold more DOX accumulated in the tumors after treatment with RD NPs than free DOX at 24 h postinjection. In addition, the highest DOX concentration that accumulated at the tumor site in mice treated with RD NPs (10.94±0.91 μg/g) was observed at 12 h, consistent with the data from the imaging of isolated tumors. Accordingly, RD NPs displayed higher tumor accumulation than the free drugs. Thus, RD NPs were efficiently delivered to the tumor via EPR effects. Moreover, RD NPs resulted in a low level of DOX accumulation in the heart. However, few NPs accumulated in other organs. This accumulation is a common biological challenge, as most NPs become rapidly sequestered from the blood, followed by their accumulation in organs of the reticuloendothelial system (RES), such as the spleen or lung [48].
In vivo antitumor efficacy
A metastatic orthotropic 4T1 mammary adenocarcinoma model was established to evaluate the benefits of combination therapy with RD NPs. As shown in Fig. 6A, rapid tumor growth was detected in the saline group, while moderately restricted tumor growth was observed in animals treated with DOX or RHE/DOX. Tumor growth was significantly inhibited in the RD NP group, and the TIR was 55.90±2.58% on day 10 (Fig. 6B), which was further confirmed by images of tumor xenografts in mice (Fig. 6C). The RD NP group displayed the lowest tumor volume after the final injection, indicating that the combination of RHE and DOX exerted a synergistic antitumor effect [49]. H&E staining of the tumor tissues harvested at the end of the study revealed the greatest cancer cell clearance in the RD NP group, which included coagulative necrosis and empty intercellular spaces, further validating the high antitumor activity of this treatment. TUNEL staining revealed the highest level of induced cell apoptosis in animals treated with this combination (Fig. 6D).
The orthotropic 4T1 tumors established in this therapeutic study spontaneously form lung metastases. We therefore evaluated whether RD NPs exerted an antimetastatic effect. Photographs of lung metastases and the H&E staining of lung sections were examined to assess the antimetastatic efficacy. As shown in Fig. 6E and Additional file 1: Figure S15, the RD NP treatment led to a significantly decreased number of metastatic lesions on the lung surface compared to the control treatment. In addition, the analysis of the lung tissue sections further supported these results. Thus, the RD NPs not only effectively inhibited primary tumor growth but also successfully suppressed tumor metastasis, consistent with the results of in vitro experiments showing antimetastatic effects [50].
Immunohistochemical staining was performed to determine whether tumor growth and aggressiveness were associated with decreased NF-κB expression (Additional file 1: Figure S16). Excised tumor sections from the RHE/DOX or RD NP groups displayed lower expression of the NF-κB P65 and MMP-9 proteins than the DOX group. Additionally, the highest expression of the proapoptotic protein Bax and the lowest expression of the antiapoptotic protein Bcl-2 was observed in the RD NPs group. Compared with free DOX, the efficient silencing of NF-κB P65 and the increased apoptosis induced by RHE/DOX were probably due to the presence of RHE. The RD NPs likely caused the greatest downregulation of NF-κB P65 and the highest apoptosis among the treatments because of the EPR effect on the RD NPs group. In addition, the improved pharmacokinetic profiles in the RD NPs group may have prolonged their circulation time and increased the accumulation of the drugs in the tumor.
NF-κB has been reported to play an important role in cancer cell migration and invasion. Thus, we investigated the effects of RD NPs on cell migration and invasion. The RD NP treatment groups exhibited significantly inhibited cell migration compared to the DOX treatment group (p < 0.001). As displayed in Fig. 6F and Additional file 1: Figure S17, the expression of MMP-9, which regulates cell invasion, was reduced in the RD NP treatment group compared with the DOX group (p < 0.001). Therefore, the increased inhibition of cell migration and invasion likely resulted from the greater inhibition of NF-κB activity and the reduced expression of MMP-9. Additionally, the highest level of the proapoptotic protein Bax was observed in the RD NPs group. Meanwhile, the lowest level of the antiapoptotic protein Bcl-2 was observed in the RD NPs group. Therefore, the inhibition of NF-κB or anti-NF-κB therapy might be applied as a possible therapeutic approach to control tumor metastasis. Therefore, the RD NPs would efficiently inhibit metastatic breast cancer.
Safety profiles
Body weight, biochemical functions and histopathological changes were evaluated and compared with saline and free DOX, which served as the negative and positive controls, respectively, to evaluate whether RD NPs induced any adverse effects during treatment. None of the treatments led to substantial body weight losses (Fig. 7A), indicating that RD NPs did not induce severe systemic toxicity. As shown in Fig. 7B, RD NPs did not exert a measurable adverse effect on blood cells or on the heart, liver and renal functions, based on the safety profiles. The numbers of peripheral blood cells were all within the normal ranges, indicating that no illnesses occurred, including hemolytic anemia and acute infection. In the blood chemistry analysis, the levels of cardiac troponin I (cTnI), the liver function biomarkers, e.g., alanine aminotransferase (ALT), aminotransferase, total protein and albumin, and the renal function biomarkers, e.g., blood urea nitrogen (BUN), creatinine (CRE), glutamic acid and uric acid (UA), were all normal, indicating that the RD NPs induced negligible hepatotoxicity and nephrotoxicity (Fig. 7C). In contrast, the levels of cTnI, ALT, BUN, CRE and UA were significantly increased in the free DOX group compared to the control group, indicating the presence of acute inflammation in the heart, liver and kidney. The histopathological results also verified these conclusions (Fig. 7D). The mice in the RD NPs and saline groups did not exhibit toxicity in the major organs, while an abnormal architecture was observed in the heart, liver and kidney tissues of animals pretreated with DOX, such as cavities in the heart, cytoplasmic degeneration of hepatocytes in the liver, and focal tubular necrosis in the kidney, indicating the apparent cardiotoxicity, hepatotoxicity and nephrotoxicity of DOX. In summary, RD NPs displayed superior therapeutic efficacy when administered at a safe level.