PEGylated nanoemulsions containing 1,2-distearoyl-sn-glycero-3-phosphoglycerol induced weakened accelerated blood clearance phenomenon

Injections of polyethylene glycol (PEG)-modified nanomedicines can lead to an accelerated clearance of the next dose of PEGylated nanomedicines, which is referred to as the accelerated blood clearance (ABC) phenomenon. It has been reported that anti-PEG IgM plays an important role in the induction of the ABC phenomenon, identifying the interface between the main chain of PEG and the hydrophobic segment of the repeated injections of the PEGylated nanocarriers, resulting in increased liver uptake and loss of long-cycle characteristics. In this study, we demonstrated that the 1,2-distearoyl-sn-glycero-3-phosphoglycerol (DSPG) in PEGylated nanoemulsions (PEs) may mask this interface between the main chain of PEG and the hydrophobic segment, inhibiting the recognition and binding of anti-PEG IgM to PEs, and evidently weakening the ABC phenomenon of PEs. This will provide a novel strategy to improve the curative effect of PEGylated nanocarriers. PEGylated nanoemulsions (PEs) with 1,2-distearoyl-sn-glycero-3-phosphoglycerol (DSPG) induced weakened the accelerated blood clearance (ABC) phenomenon in Wistar rats during repeated injection of PEs.


Introduction
Polyethylene glycol (PEG) is used widely with protein and nanocarriers in the pharmaceutical field because of its good water solubility and biocompatibility. The process called PEGylation refers to conjugate PEG with nanocarriers or drugs [1,2]. Use of this technology has endowed the modified drugs or nanocarriers with better hydrophilic properties, thereby lowering their immunogenicity and improving their stability; thus, its application has been widely preferred by researchers. The history of researchers developing the technology of modifying PEG on proteins could be traced back to 1977, which significantly prolonged the blood circulation time of proteins and diminished their immunogenicity [3,4]. In 1990, the first PEGylated protein product approved by the FDA, Adagen®, used to treat severe immunodeficiency diseases, was a PEGylated adenosine deaminase [5].
However, since the first PEGylated nanomedicine, Doxil®, a liposome nanomedicine approved in the USA in 1995, in addition, no other high-density PEGylated nanocarriers have been approved and the development of PEGylated drugs has stagnated [6][7][8][9]. Anti-PEG antibody (APA) response may be elicited by injection of PEGylated nanomedicines such as PEGylated liposomes [10], proteins [11], micelles [12,13], emulsions [14,15], and nanoparticles [16] according to numbers of clinical and experimental studies; moreover, repeated administration may result in the accelerated blood clearance (ABC) phenomenon. Repeated intravenous injections of complex nano-drugs, such as PEGylated protein or nanoparticles, in animals or humans showed that with the second and subsequent doses the circulation time was shortened sharply and the curative effect reduced.
IgM produced by PEGylated nanomedicines may cause serious problems in the field of drug targeting. After a single subcutaneous or intravenous injection of the PEGylated recombinant mammalian uricase, Krystexxa, approved by the FDA in 2010, 38% of patients experienced an APA response; even if repeated a year apart, antibodies can quickly build up, leading to a loss of efficacy [17]. Non-cytotoxic anticancer drugs (such as genes) and diagnostic drugs loaded with PEGylated nanocarriers can lose their ability to treat or diagnose, due to repeated injections. Nowadays, "treatment" involves the use of imaging technology where PEGylated nanomedicines are often used as diagnostic and therapeutic carriers. However, the efficiency of the second injection may be seriously affected when the immune response is activated by the PEGylated nanomedicines of the first dose, that is, the diagnostic system. Therefore, the ABC phenomenon must be avoided as it can influence the reliability of the diagnosis. The occurrence of the ABC phenomenon leaves only one chance for treatment and diagnosis with nanocarriers. However, this limitation must be overcome in clinical settings for successful diagnosis and treatment. Studies have shown that the immunogenicity of PEGylated nanomedicines poses a threat to clinical patients [7,[18][19][20], and that the change in pharmacokinetic behavior damages the advantages of PEGylation. Thus, the ABC phenomenon poses an unavoidable challenge to further research and the application of PEGylation technology.
Exploring strategies to eliminate or reduce the ABC phenomenon is an important subject for many researchers [11,21,22], such as pre-injecting of high-molecularweight PEG [23] or anti-PEG scFv, regulating the physical and chemical properties of PEGylated nanocarriers [14,24], adjusting the administration regimens [25,26], and replacing it with other polymers or co-modifications [22,27,28]. For materials such as mPEG 2000 -DSPE (Fig. 1), the binding site (amide bond) between hydrophilic PEG and hydrophobic DSPE fragment may be the key for recognizing and mediating the scavenging of PEGylated nanomedicines by anti-PEG IgM [29]. In this study, we used 1,2-distearoyl-sn-glycero-3-phosphoglycerol (DSPG) and prepared a series of PEs containing DSPG to investigate the ability of DSPG to inhibit the ABC phenomenon in Wistar rats.

Animals
Wistar rats (male, 180-200 g) were provided by the Experimental Animal Center of the Shenyang Pharmaceutical University (Shenyang, China). All the animal experiments were conducted according to guidelines of the animal welfare committee of Shenyang Pharmaceutical University (NIH publication #85-23, revised in 1985).

Preparation of PEs
The different formulations of the PEs used are shown in Table 1. The different PEs were prepared using the DiR as the marker at a concentration of 1.8 mg/mL with the oil phase MCT prescription amount being 30 mg/mL, and the lipid prescription composition as shown in Table 1. All the PEs were prepared according to the following method. The S75, mPEG 2000 -DSPE, MCT, and DiR (Table 1) were mixed in sterile water and heated to 55 °C. Sterile water was added to the oil phase at the same temperature and stirred quickly. DSPG at n mol% PE-DSPGs without any S75 was diluted in sterilized water. The mixture was stirred at 55 °C for 20 min and sonicated in an ice bath for 8 min using a laboratory ultrasonic cell pulverize to prepare the emulsions. Then, nanoemulsions were prepared by extrusion through 0.8-, 0.45-, and 0.22-μm polycarbonate membranes. Using different PEG modification densities, n mol % PE and n mol % PE-DSPG (n = 10, 30, 50) were prepared with 5 μmol phospholipid/mL. Different DSPG concentrations, PE-DSPG-n (n = 2, 5, 9, 15, 30) with a fixed mPEG 2000 -DSPE concentration of 0.5 μmol/mL, were prepared. Particle size distribution and zeta potentials of the nanoemulsions were determined by Nicomp 380. Entire process was performed without light exposure.

Pharmacokinetics of PEs
Rats were randomly divided into 28 groups (n = 3 in each group). The injection schemes with 7-day interval between the two PE injections are shown in Table 2. The rats were administered PEs intravenously at a rate of 5 μmol phospholipid/kg. The DiR was used to demonstrate the pharmacokinetics of the nanoemulsions at 0.3 mg/kg due to the high encapsulation rate of nanoemulsions. At 1, 5, 15, 30 min, 1, 4, and 8 h after the first and second injections, 0.5-mL plasma was collected from the ophthalmic venous plexus and placed in a heparinized tube. The plasma was separated and the DiR was measured as described previously [14].

Determination of anti-PEG IgM in serum
Anti-PEG IgM in serum was determined by improved enzyme-linked immunosorbent assay (ELISA) as described previously [14]. Blood samples were collected and incubated at 25 °C for 2 h. Serum was separated by centrifugation at 800 g for 10 min, and the anti-PEG IgM was determined.

Statistics
Comparisons between the two groups were conducted using the Student's t-test with SPSS software. Data in all the tables and figures were represented as the means ± standard deviations (SDs), and values at P < 0.05 were considered statistically significant.

Characteristics of the PEs
The particle size of the PEs was controlled between 119 and 131.2 nm, the polydispersity index (PDI) of each nanomedicine less than 0.2, and the zeta potential below − 20 mV, indicating that the nanomedicines showed good stability (Table S1).
As everyone knows, the phase transition temperature of SPC is − 20 °C, while that of DSPG is 55 °C. In order to eliminate the influence of the two emulsifiers on the release of emulsions in blood, we compared the release behavior of different emulsions in vitro simulated in vivo environment.
Of the PEs, 1.0 mL was precisely aspirate and added to a dialysis bag (with a molecular weight cutoff of 10 kDa), clamped the two ends and placed it in 200 mL of PBS buffer (500 mmol/L, pH = 7.4, containing appropriate penicillin), and stirred under constant temperature at 37.0 ± 0.5 °C and   3.0 mL of dialysate was drawn, and an equal amount of release medium was added. The dialysate was filtered with a 0.45-μm microporous filter membrane and the fluorescence intensity was measured at λ ex = 750 nm, λ em = 790 nm, and the concentration was calculated by substituting it into the standard curve equation; the cumulative release Rn of the drug is calculated, the formula is as follows, and the results are as follows (Fig. 2).
Among them, Cn is the concentration at the nth sampling, V0 is the volume of the release medium, V is the volume of each sampling, and Mt is the total drug concentration. Results showed that there was no significant difference in the release behavior among different PE.

Effect of the PEG modified densities on the pharmacokinetics of n mol % PEs and PE-DSPGs of single and repeated intravenous injection
The 8-h pharmacokinetic behavior of a single injection of n mol % PEs and PE-DSPGs was evaluated by measuring the plasma concentration of the DiR. As indicated by the area under the curve (AUC) and T 1/2 , the circulation time of n mol % PEs and PE-DSPGs in the Wistar rats increased gradually with the increase in PEG density on the surface of n mol % PEs and PE-DSPGs (Table S2). Seven days after the first injection, the rats were injected repeatedly. The strength of the ABC phenomenon was evaluated using the ABC index , which indicated the ratio of AUC of the second dose to that of the first dose; that is, the ABC index = AUC (0-60 min) of the second injection/AUC (0-60 min) of the first injection. The higher the ABC index value is, the smaller the difference in pharmacokinetic behavior between the second and first injections is, which meant that a weaker ABC phenomenon is occurring. Here, the ABC index(0-60 min) was used as the criterion to evaluate the strength of the ABC phenomenon.
The repeated administration of n mol % PEs in each group induced the ABC phenomenon. When the modification density of PEG increased from 10 to 30 mol %, the ABC phenomenon decreased; however, when the PEG modification density continued to increase, the ABC phenomenon was enhanced. Compared with n mol % PEs, there were no significant differences in the level of anti-PEG IgM induced by n mol % PE-DSPGs with the same PEG modification density (Fig. 3), but the ABC index of each group was significantly higher than that of n mol % PEs, with the same PEG modification density, indicating that the ABC phenomenon induced by each group was Fig. 2 The rate of the DiR cumulative releases from PEs in 48 h in vitro. The release kinetics follows Ritger-Peppas equation. Data are shown as means ± standard, n = 3 Fig. 3 Determination of anti-PEG IgM titer in rats 7 days after the first dose of PEs. Significant difference compared with control is indicated by asterisk (*), and number sign (#) means significant difference between groups. Data are shown as means ± SDs, n = 3. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 remarkably weakened. When the ABC index of 30 mol % PE-DSPG reached 0.89, almost no ABC phenomenon was observed (Fig. 4). These results suggest that both the modification density of PEG and the addition of DSPG could affect the intensity of the ABC phenomenon. The modification density of PEG affected the secretion levels of anti-PEG IgM, whereas while DSPG did not affect the secretion level of antibodies, it weakened the ABC phenomenon.

Effect of DSPG concentration on the pharmacokinetics of single and repeated injections of PE-DSPG-ns
To further investigate the influence of DSPG on the ABC phenomenon, we prepared a series of PEs with different DSPG concentrations, PE-DSPG-ns, using the same procedure as that for n mol % PEs and PE-DSPGs (Fig. 5). No significant differences were observed in the pharmacokinetic  (Table S2). Seven days after the first injection, the PE-DSPE-ns containing DiR were injected repeatedly with the same phospholipid dose. The ABC index showed that the addition of DSPG could indeed weaken the ABC phenomenon. When the modification concentration of PEG was fixed, i.e., PE-DSPG-ns groups, the ABC phenomenon gradually weakened with the gradual increase of DSPG concentration and when the molar ratio of the DSPG to PEG increased to 30:1 in the PE-DSPG-30 group, the ABC index reached approximately 0.94, and the ABC phenomenon almost disappeared.
No significant differences were observed in the level of anti-PEG IgM induced by the first injection of PEs with the same PEG modification density; however, the addition of DSPG weakened the ABC phenomenon in each group, significantly. Therefore, we selected 5 μmol phospholipid/kg 10 mol % PE with the smallest ABC index for the first injection, and the rats with high levels of anti-PEG IgM after the first injection of 10 mol % PE were termed as ABC ( +) rats. (F) Accumulation of liver and spleen 8 h after single (red) and repeated (black) injection of DiR-PEs in rats. S and L represent accumulation in liver and spleen after single injection; s and l represent accumulation in liver and spleen after repeated injection. Significant differences between the accumulation of single and repeated injection of the same PEs in the same tissue are indicated by asterisk (*). *P < 0.05, **P < 0.01, and ****P < 0.0001. Data are shown as means ± SDs, n = 3 As shown in Fig. 6(A)-(C), the ABC ( +) rats produced a high level of anti-PEG-IgM 7 days after the first injection while the second 30 mol % PE injection produced a strong ABC phenomenon, with the ABC index reaching 0.25, while ABC index of repeated 30 mol % PE injection was 0.7. It was suggested that the reason for the weak ABC phenomenon after the repeated injection of 30 mol % PE may be that the higher density of PEG modification inhibited the production of anti-PEG antibodies. When the level of anti-PEG IgM reached a higher level, the 30 mol % PE would still be cleared quickly with an accumulation in the liver and spleen (Fig. 6D).
The ABC index of the second injection of 30 mol % PE-DSPG in the ABC ( +) rats reached 0.82, which was slightly injection; s and l represent accumulation in liver and spleen after cross-injection. Significant differences between the accumulation of repeated and cross injection of the same PEs in the same tissue are indicated by asterisk (*). *P < 0.05, **P < 0.01, ***P < 0.001. Data are shown as means ± SDs, n = 3 lower than the repeated injection of 30 mol % PE-DSPG and further demonstrated the attenuation effect of DSPG on the ABC phenomenon; although there was a high level of anti-PEG IgM in rats, a strong ABC phenomenon did not occur. The ABC ( +) rats injected with PE-DSPG-30 did not cause the ABC phenomenon. This also showed that higher DSPG concentrations could effectively inhibit the occurrence of the ABC phenomenon.

IgM binding of 30 mol % PE and 30 mol % PE-DSPG in the ABC ( +) rats
To determine the influence of DSPG on the intensity of the ABC phenomenon of PEs, we measured the changes in anti-PEG IgM before and 6 h after the second injection of 30 mol % PE, 30 mol % PE-DSPG, and PE-DSPG-30 in the ABC ( +) rats. As shown in Fig. 7, no significant change was observed in the plasma anti-PEG IgM levels of the ABC ( +) rats 6 h after the injection of 30 mol % PE-DSPG and PE-DSPG-30, while the level of anti-PEG IgM in the ABC ( +) rats decreased significantly after the intravenous injection of 30 mol % PE, indicating that the second injection of 30 mol % PEs could be combined with anti-PEG IgM of the ABC ( +) rats to produce the ABC phenomenon. In contrast, 30 mol % PE-DSPG and PE-DSPG-30 combination did not enable the anti-PEG IgM recognition in the ABC ( +) rats; therefore, the ABC phenomenon did not occur.

Co-injection of 10 mol % PE with 30 mol % PE, 30 mol % PE-DSPG, and DSPG-PE-30
Here, we investigated the effect of the co-injection of 10 mol % PE containing DiR with blank 30 mol % PE or DSPG-PE-30 into the ABC ( +) rats. As shown in Fig. 8, the coinjection of 30 mol % PE and DSPG-PE-30 and 10 mol % PE did not affect the ABC phenomenon of 10 mol % PE.

Discussion
The induction of the immune reaction by the first dose of PEGylated nanocarriers, eliminating the long-cycle characteristics of the second-dose PEGylated nanocarriers, is called the ABC phenomenon. Many studies have reported that the main reason for the occurrence of the ABC phenomenon is the recognition of the PEGylated nanocarriers of the secondary injection by the anti-PEG IgM leading to accelerated clearance of the second dose of PEGylated nanocarriers. In previous reports, it was confirmed that the binding site between PEG and the hydrophobic segment of PEG-lipid derivatives such as mPEG 2000 -DSPE was the main binding site recognized by anti-PEG IgM. The findings of a previous study [29] showed that recognition binding could be eliminated by adding a hydrophilic segment between the PEG and the hydrophobic segment. Therefore, masking the binding site between PEG and the hydrophobic segment of mPEG-DSPE may weaken or eliminate the ABC phenomenon. Here, by controlling the particle size of the emulsions, excluding the influence of the release behavior, and only investigating the influence of the addition of DSPG on ABC phenomenon, we further proved the importance of the interface confrontation between PEG and the hydrophobic segment, recognized by anti-PEG IgM to induce the ABC phenomenon.
DSPG is an anionic phospholipid with the same saturated fat chain as mPEG-DSPE. The addition of DSPG to PEs weakened the ABC phenomenon significantly; however, the antibody-induced ability of the APA response did not decrease, and it still induced the production of anti-PEG IgM. Using cross-injections and co-injections, we confirmed that the recognition and binding abilities of anti-PEG IgM to PEs containing DSPG were weakened significantly, and that the ABC phenomenon disappeared when the molar ratio of DSPG to mPEG 2000 -DSPE was 7:3 and 30:1. In comparing the structures of DSPG and SPC (Fig. 1), it was found that the two fatty chains of SPC were unsaturated and could not form a neat emulsion layer with the two saturated fat chains of mPEG-DSPE; however, DSPG had the same Fig. 7 Changes of anti-PEG IgM titer 7 days after the first dose (black) and 6 h after the second dose (gray) of 10 mol % PE, 30 mol % PE, 30 mol % PE-DSPG, and PE-DSPG-30. Significant difference between the titer is indicated by asterisk (*). **P < 0.01. Data are shown as means ± SDs, n = 3 fatty chain as DSPE, and the two could form a more neat and uniform emulsion layer. From the perspective of spatial structure, the two free hydroxyl groups of DSPG being adjacent to the amide bond in mPEG2000-DSPE may, to some extent, mask the amide bond.

Conclusions
In this study, in order to weaken or eliminate the ABC phenomenon, we investigated the ability of DSPG to mask the recognition sites of anti-PEG IgM to PEs. We verified the DiR-PEs in rats. S and L represent accumulation of liver and spleen after repeated injection; s and l represent accumulation of liver and spleen after repeated dose and co-injection. Data are shown as means ± SDs, n = 3 potential mechanisms of the ABC phenomenon. By using co-injections or cross-injections with common PEs, the ability of PEs containing DSPG to bind to IgM was significantly weakened. When the content of DSPG reached a specific value, the ABC phenomenon disappeared and did not change the titer of anti-PEG IgM in the plasma of the ABC ( +) rats. Therefore, this study further demonstrated that the hydrophobic interface between the PEG and the hydrophobic segment is an important site for the recognition and binding of anti-PEG IgM, and thus, by masking this site, the ABC phenomenon can be effectively reduced.