Packaging of nanoparticles into cell membranes and characterization of MNPs
We prepared the 50:50 Poly (DL-lactide-co-glycolide) Carboxylate End Group (PLGA) nanoparticles, co-incubated them with purified cell membranes, and extruded the result from polycarbonate membranes to generate cell-membrane coated PLGA nanoparticles (Fig. 1). Transmission electron microscopy (TEM) was used to observe three types of nanoparticles: bare nanoparticles (bare NPs), coated with RBC cell membrane (RBC-NPs), and coated with MDCK cell membrane (MDCK-NPs) (Fig. 2A-C). All nanoparticles exhibited spherical structures. When RBC-NPs and MDCK-NPs are compared to bare NPs, it can be clearly seen that the surface is covered with a monolayer film. A Flow NanoAnalyzer was used to measure the diameter (DH) of the three kinds of nanoparticles; bare NPs were ~ 95 nm, and RBC-NPs and MDCK-NPs were ~ 123 nm and ~ 124 nm, respectively, or ~ 30 nm larger than bare NPs (Fig. 2D). Zeta potentials of the three nanoparticles measured using dynamic light scattering (DLS), indicated that the zeta potentials of RBC-NPs and MDCK-NPs were ~ 15 mV greater than bare NPs (Fig. 2E). To assess the dispersion of nanoparticles in liquids, the polydispersity index (PDI) of nanoparticles dispersed in PBS was measured; PDI was less than 0.2, indicating that the three nanoparticles have good dispersion (Fig. 2F). No change in PDI a week later indicated that nanoparticles can be stably dispersed in PBS over the short term (Fig. 2G).
In vitro evaluations of MNPs
Neutralization capacity in vitro is an important indicator of in vivo therapeutic efficacy. We therefore assessed the neutralization capacity of the two kinds of MNPs (RBC-NPs and MDCK-NPs) to recombinant ETX with Glutathione-S-transferase (GST) tags (GST-ETX) in vitro by 3-(4,5-Dimethylthiazol-2-yl)-5(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt (MTS) assay with cells. MDCK cells are the most sensitive cells to ETX and are commonly used in ETX cytotoxicity. ETX with different tags (GST and 6×His) did not significantly differ in toxicities (Fig. S1). Therefore, MDCK cells were used in all evaluations of GST-ETX and MNPs in vitro. The concentration of GST-ETX that killed 50% of MDCK cells (denoted by CT50) was 0.8813 nM (Fig. 3A). When the toxin reached a concentration of 20 nM, maximum cell mortality was achieved. To investigate whether the three kinds of nanoparticles could damage MDCK cells, MDCK cells were treated with a series of concentrations of nanoparticles. A measure of nanoparticles as high as 2.4 mg showed no cytotoxicity to MDCK cells (Fig. 3B).
Cell death was examined using a confocal high-content imaging system (Fig. 3C, S2). MDCK cells incubated with the two kinds of MNPs maintained a high survival rate under ETX challenge. MDCK cells were incubated with GST-ETX and a series of concentrations of MNPs. MNPs significantly reduced the toxicity of GST-ETX in a dose-dependent manner, with 2 mg MNPs nearly abolishing the toxicity of GST-ETX to MDCK cells (Fig. 3D). MDCK-NPs reduced the toxicity of GST-ETX more than RBC-NPs, suggesting that MDCK-NPs can neutralize GST-ETX faster. We next treated a series of concentrations of GST-ETX with MNPs, to assess the neutralization ability of MNPs mixed with toxin. Both MNPs neutralized ETX, reducing the toxicity of the mixture to MDCK cells (Fig. 3E). The two types of MNPs created mixtures that differed in CT50, with 2 mg of RBC-NPs neutralizing 3.793 pmol GST-ETX, and 2 mg of MDCK-NPs neutralizing 11.75 pmol GST-ETX. Thus, while both kinds of MNPs can effectively neutralize ETX, the MDCK-NPs neutralized more GST-ETX faster when cells were challenged by GST-ETX. Therefore, MDCK-NPs have a better performance against GST-ETX than RBC-NPs.
To verify the difference in sensitivity between RBCs and MDCK cells, we assessed the relative hemolysis damage in RBCs caused by GST-ETX. Hemolysis increased with the concentration of GST-ETX (Fig. 3F). The concentration of GST-ETX required to cause hemolysis in RBCs is close to 20 nM, and the maximum relative hemolysis was achieved at a toxin concentration of approximately 150 nM. In addition, RBC hemolysis by GST-ETX was not complete within 1 h (Fig. 3G). In contrast, 20 nM GST-ETX caused maximum cell mortality to MDCK cells within 1 h. This indicates that RBCs and MDCK cells differ in sensitivity to ETX, and this difference is reflected in the differences between the two kinds of MNPs. We estimated the concentration of 30 nM of GST-ETX that would cause 50% relative hemolysis to verify the neutralization capacity of MNPs. We calculated that a complete neutralization of 30 nM GST-ETX 1 ml would require about 5.1 mg of MDCK-NPs and about 15.8 mg of RBC-NPs. The results of co-incubation with RBCs showed that MDCK-NPs could reduce the toxicity by 85%, and RBC-NPs could reduce the toxicity by 60% (Fig. 3H). None of the three kinds of nanoparticles caused any damage to RBCs. This result is consistent with the difference in protective ability of the two MNPs in the MDCK cell protection experiment above. When MNPs and cells competitively bind to GST-ETX, MDCK-NPs can neutralize more toxin faster. This may be because there are more receptors on the membranes of MDCK cells, given that MDCK cells are more sensitive than RBCs to GST-ETX.
In vivo safety assessment of MNPs
Safety assessment in vivo is an important precondition of in vivo therapeutic efficacy. We therefore assessed the safety of the three kinds of nanoparticles to mice in vivo. Following injection of one of the three kinds of the nanoparticles to three groups of mice intravenously, mice were observed for 7 consecutive days (Fig. 4A). Bare NPs and RBC-NPs were safe to animals, while MDCK-NPs caused death in half of the mice (Fig. 4B). We examined multiple blood parameters, including a comprehensive serum chemistry panel and blood cell counts. Cells count in mice injected with bare NPs and RBC-NPs were consistent with baseline levels, indicating these two nanoparticles did not cause any toxic or immune response (Fig. 4C). In contrast, mice injected with MDCK-NPs, displayed significant anomalies in white blood cell (WBC), eosinophil (EOS), basophil (BAS) and platelet (PLT) counts, indicating that MDCK cell membranes could trigger an immune response, which is presumably related to MDCK-NP-induced death in mice. In addition, we assessed blood markers of liver and kidney damage; alkaline phosphatase (ALP) [21] and glucose (GLU) in the blood are thought to reflect damage to the liver [22], while serum calcium (Ca), serum sodium (Na), and urea in blood are thought to reflect damage to the kidney [23–27]. However, these indicators were normal in all mice (Fig. 4D), indicating the short-term safety of MNPs.
Immune response triggered by MDCK-NPs, which include increased EOS, BAS and decreased PLT, are presumed to be related to hypersensitivity [28–30]. To verify whether the anomalies of MDCK-NPs were related to hypersensitivity, we measured IgE in mouse serum. The results of serum IgE detection showed that IgE in the serum of mice injected with MDCK cells membrane was significantly increased (Fig. 4E), while no obvious IgE reaction was detected in the serum of the other mice. These results suggest that MDCK cell membranes induced strong hypersensitivity in mice, which is likely the main cause of death in mice injected with MDCK-NPs. MDCK cells originate from normal kidney cells of a cocker spaniel dog [31], which is distantly related to murine species, and we infer that the injection of membranes from a distantly related species are likely to induce strong immune response and hypersensitivity. To test our hypothesis, we injected membranes from N2a cells, which originate from a murine species. Membrane from N2a cells induced slight immune response but did not trigger hypersensitivity, in keeping with our hypothesis.
Given the unstable safety profile of MDCK-NPs, we abandoned the use of MDCK-NPs in further animal experiments, despite their excellent ability to compete and bind ETX.
In vivo therapeutic efficacy evaluations of MNPs
After assessing neutralization capacity in vitro and the safety in vivo, we next assessed therapeutic efficacy of RBC-NPs in vivo. We first intravenously injected GST-ETX into mice of each group, and then 10 min later, we intravenously injected RBC-NPs or PBS into mice of the treatment and positive control group, respectively (Fig. 5A). The dose of 20 ng GST-ETX led to 100% death of mice in the positive control group within 24 h (Fig. S3), while mice given RBC-NPs had 100% survival for 7 days (Fig. 5B). Histopathological analysis showed that the kidneys, brains, livers and lungs of mice in the positive control group had different degrees of damage, mainly manifested as congestion or edema; however, these symptoms were significantly relieved by RBC-NPs (Fig. 5C).
In another three groups of mice, 6 h after completion of the two injections, blood for analysis was done (Fig. 5A). Blood biochemical analysis revealed that, compared to the negative control group, WBC count in the blood of ETX-infected mice in the positive control group was significantly elevated 300%, and the change of WBC was caused mainly by 900% elevation of neutrophils (NEU) (Fig. 5D). WBC count in the blood of mice treated with RBC-NPs was closer to that of the negative control group of mice. This indicates that RBC-NPs effectively attenuated the inflammatory response triggered in vivo by the ETX challenge [32]. In addition, we assessed blood markers of liver and kidney damage. The livers and kidneys of mice were damaged to varying degrees during the ETX challenge, and RBC-NPs can play a role in protecting these organs, which was consistent with the results of histopathology analysis (Fig. 5E, S4). Thus, RBC-NPs were sufficiently able to competitively bind ETX in vivo to protect host organs from damage and can effectively treat GST-ETX infection in vivo.
The interaction between RBC-NPs and ETX and its metabolism in vivo
We further investigated the interaction between in metabolisms of RBC-NPs and GST-ETX in vivo to verify how the RBC-NPs protect the host. We injected mice intravenously with fluorescently-labeled RBC-NPs and GST-ETX, and the distribution of radiant efficiency in blood and tissues reflected their interactions and metabolisms. To display the metabolism of ETX toxin and RBC-NPs in mouse tissues, the NIR dye DiOC18(7) (DiR) was used to label RBC-NPs (DiR-RNPs), the NIR dye cyanine 5.5 (Cy5.5) was used to label GST-ETX (Cy5.5-ETX). We intravenously injected Cy5.5-ETX into mice of each group and 10 min later, intravenously injected DiR-RNPs or PBS into mice of the treatment group and the positive control group, respectively. At several time points post-injection (after 5 min, 24 h, 48h and 72 h), blood of random mice in each group was taken for quantitative analysis of fluorescence intensities and its major organs were taken at the same time for fluorescence images in vitro (Fig. 6A). The quantitative analysis of fluorescence intensities in blood showed that DiR-RNPs in the blood of mice in the treatment group decreased over time (Fig. 6B). It should be noted that Cy5.5-ETX in the blood of mice in the treatment group was not completely cleared after three days, but consistently decreased along with DiR-RNPs (Fig. 6C). Levels of Cy5.5-ETX in the blood of mice in the treatment group was lower than that of the positive control group all time points and was reduced to a very low level 24 h after injection. Thus, within 5 min of injection, RBC-NPs neutralized the majority ETX, preventing ETX from spreading in vivo. In addition, the RBC-NPs that neutralized ETX remained stable in the blood and did not release toxins into the blood again.
Quantitative data of fluorescence images in vitro of DiR in tissues indicated that DiR-RNPs were captured by the liver and spleen (Fig. 6D-E, S5, S6). The Cy5.5 signal of mice in the positive control group suggested Cy5.5-ETX gradually decreased in the liver, and no fluorescence signal was observed in the spleen (Fig. 6F-G, S7, S8). In the treatment group, Cy5.5-ETX was much higher in the liver and also had a high signal in the spleen. As with the fluorescence signal in blood of the treatment group, the distribution signal of GST-ETX in organs is consistent with that of RBC-NPs. This was especially evident in the spleen, in which GST-ETX was not detected in the spleen in the positive control group, but because RBC-NPs had been captured by the spleen, it was detected in the spleen of the treatment group. The difference is that the two fluorescence signals tend to decrease over time in the blood, while fluorescent signals in the spleen and liver show a tendency to stabilize or increase. The results show that as immune and detoxification organs of animals, the spleen and liver captured RBC-NPs that had neutralized GST-ETX from the blood. This is how RBC-NPs treat GST-ETX infection in vivo.
In vivo therapeutic efficacy evaluations of nebulized pulmonary inhalation
Toxic aerosols can be used as weapons in terrorist attacks. For ETX, which was classified as a potential biological weapon, whether RBC-NPs can play a protective role in ETX challenges from pulmonary inhalation is of key import. We therefore assessed the therapeutic efficacy of the RBC-NPs during nebulized pulmonary inhalation of mice in vivo. Liquid aerosol devices were used to administer GST-ETX and RBC-NPs from the mouse trachea by quantitative nebulization, thereby simulating ETX aerosol challenges and lung drug delivery. The advantage of using liquid aerosol devices is that the drug can be evenly distributed into the lungs of mice (Fig. S9) and ensures that each mouse receives the drug at the same distribution location and reduces the risk of pulmonary edema caused by lung administration. This not only allows us to simulate an ETX aerosol weapons challenge, but also minimizes possible adverse effects during lung delivery of RBC-NPs. We first administered 50 ng GST-ETX into the tracheas of mice in each group, and 10 min later, administered 2 mg RBC-NPs by trachea or intravenously to mice in the two treatment groups; PBS was administered by trachea to mice in the positive control group. Mice in the negative control group had PBS administered by tracheas each time (Fig. 7A). Mice were observed for 14 days post-infection. Introduction of 50 ng GST-ETX into mouse lungs resulted in 100% mortality of the mice in the positive control group within 8 days (Fig. S10). Treatment mice that had RBC-NPs introduced by aerosol into the lungs had 100% survival (Fig. 7D). However, RBC-NPs given intravenously did not play a protective role to mice, mice in this group all died within 8 days.
In another groups of mice, blood samples were taken for analysis 6 h post-infection (Fig. 7A). Compared to the negative control group of mice, WBC counts in ETX-infected mice were elevated 300%. The WBC counts of mice treated with RBC-NPs in lungs were closer to the negative control group of mice, but the WBC counts of mice treated with RBC-NPs intravenously were closer to those of mice in the positive control group. The increase in WBCs was mainly caused by 600% increases in NEU (Fig. 7E-F). This result showed that only RBC-NPs in lungs effectively attenuate the inflammatory response triggered in vivo by the ETX challenge in lungs. As before, we also tested blood markers of liver and kidney damage. The livers of mice were not damaged during the ETX challenge in lungs, but kidneys, the most sensitive organ to ETX, were slightly damaged with ETX challenge to lungs. Compared with injection of MNPs via veins, RBC-NPs in lungs appear to play a critical role in protecting kidneys of mice (Fig. 7G).
As with our experiment testing therapeutic efficacy injecting intravenously, we again dissected the main organs and stained tissue with H&E (Fig. 7B) for aerosol-challenged mice. Histopathological analysis showed that the brains of mice had no obvious pathological changes (Fig. 7H, S11). The lungs of mice in the positive control group had serious damage, mainly manifested as congestion and edema, while the liver and kidneys of these mice had mild damage, mainly manifested as edema. However, these symptoms were significantly relieved by RBC-NPs in lungs.
To verify why pulmonary infection with ETX causes less damage to organs other than the lungs, and how RBC-NPs in lungs can protect mice, Cy5.5-ETX and DiR-RNPs were used to measure their interactions and metabolism in lungs (Fig. 7C). Quantitative data of fluorescence images in vitro of the DiR indicated that DiR-RNPs cannot escape the lungs (Fig. 7I). Quantitative data of fluorescence images in vitro of the Cy5.5 of the mice in the positive control group indicated that most Cy5.5-ETX remains in the lungs after 24 h, but after 48 h, some Cy5.5-ETX was detected in the livers and a smaller amount of Cy5.5-ETX was detected in the kidneys (Fig. 7J). This result showed why GST-ETX causes little damage to other organs when administered via the lung. However, for mice in the treatment group all Cy5.5-ETX was in the lungs and the distribution of the signal of GST-ETX was consistent with that of RBC-NPs. Thus, GST-ETX was completely neutralized in the lungs by RBC-NPs, preventing it from causing damage to the lungs and also preventing it from spreading beyond the lungs. This is how RBC-NPs protects the host against an ETX challenge to the lungs. This experiment demonstrated the efficacy of using RBC-NPs in lungs to treat of GST-ETX infection in lungs in vivo.
Sustained protection of MNPs in vivo
Previous tissue in vitro imaging experiments had shown that RBC-NPs have a significantly longer circulation time than expected in vivo. This suggests that the stability of RBC-NPs can provides long-term protection in vivo. To test this, we delivered RBC-NPs to mice 1–3 days in advance of a toxin challenge via intravenous and aerosol, and then delivered GST-ETX to mice using the same delivery system. Mice were then observed for survival (Fig. 8A, C). RBC-NPs given intravenously 3 days in advance still provided protection for mice against ETX, and RBC-NPs in lungs 2 days in advance provided partial protection for mice (Fig. 8B, D). In other groups of mice, 6 h after completing toxin delivery, blood was taken for analysis (Fig. 8A, C). Blood biochemical analysis shows that, according to the two standards of WBC and NEU counts, appropriate early injection of RBC-NPs can provide protection to the host, although the protective effect decreases with time (Fig. 8E, 8F). This experiment demonstrated the efficacy of using RBC-NPs in advance to treat of GST-ETX infection in vivo.