Synthetic ditempolphosphatidylcholine liposome-like nanoparticles for anti-oxidative therapy of atherosclerosis

Atherosclerosis (AS), a chronic inflammatory disease, is the leading cause of death worldwide. Anti-oxidative therapy has been developed for AS therapy in light of the critical role of ROS in pathogenesis of AS, but current anti-oxidants have exhibited limited outcomes in the clinic. Herein, new ROS-eliminating liposome-like NPs (Tempol-Lips) were assembled from synthetic lipids that covalently conjugated two Tempol molecules with phosphatidylcholine by esterification reaction. The obtained Tempol-Lips can be efficiently internalized into inflammatory macrophages and attenuated inflammation via scavenging overproduced intracellular ROS. After i.v. administration, Tempol-Lips with nanoscale character accumulated in the plaques of ApoE−/− mice through passive targeting and significantly inhibited the pathogenesis of AS, compared with those treated with control drugs. The therapeutic benefits of Tempol-Lips primarily are ascribed to the reduced local and systematic oxidative stress and inflammation. Preliminary studies in vivo further demonstrated Tempol-Lips were safe and biocompatible after long-term i.v. injection. Conclusively, Tempol-Lips can be developed as a novel anti-AS nanotherapy with potential translation in the clinic.


Introduction
Atherosclerosis (AS) is a chronic inammatory disease distinguished by lipid and inammatory cell accumulation in arterial walls. [1][2][3] The pathogenesis of AS always involves inammatory reactions and increased oxidative stress. 4,5 Oxidative stress is the imbalance in favor of reactive oxygen species (ROS) overproduction and/or the body's innate anti-oxidant capability decreases. ROS plays a critical role in inammation, apoptosis and oxidation of low-density lipoprotein cholesterol (LDL-C), which is closely associated with the pathogenesis of AS. 6,7 Oxidized LDL (oxLDL) possesses several proatherogenic activities, including foam cell formation, 8 ROS generation 9 and adhesion molecule and scavenger receptor expression. [10][11][12] Moreover, ROS can also interrupt the redox-dependent pathways in the walls of blood vessels to promote the development of AS, which concerns regulatory genes associated with vascular function, signal transduction pathways and inammatory components of AS. 5,13,14 Therefore, reducing ROS generation and attenuating systemic oxidative stress in AS plaques represent reasonable strategies for AS treatment.
Various kinds of anti-oxidants have been studied in preclinical, such as probucol, 15 vitamins E, 16 Tempol, 17 coenzyme Q 18 and NADPH oxidase. 19 Although determined with protective effects of anti-oxidants on AS, clinical trials showed no positive effects. 20 To some extent, this mainly is due to the rapid elimination, nonspecic distribution and low delivery efficiency at targeted plaque of AS in vivo. 21 In addition, the limited ROS-scavenging capability of most single anti-oxidant also contributes to the ineffective outcomes against inammatory diseases. 22 Accordingly, additional anti-oxidative stress strategies remain to be designed.
Much evidence demonstrated that nanoparticle (NPs)-based approaches are promising and effective for targeted therapy of AS. [23][24][25][26] NPs can target plaques by direct inltration or endocytosed by circulating phagocytes, following translocation to AS lesions by cellular inltration and recruitment. Except that NPs in most studies were just used as vehicles for targeted delivery of therapeutics to AS plaques, recent progress indicated NPs with intrinsic anti-inammation and anti-oxidative activity are potential next-generation therapy for AS and other inammatory illness. 27,28 For instance, NPs assembled from Tempolcontaining amphiphilic copolymers were effective to treat drug-induced intestinal inammation. 29 A biodegradable polymer-based NPs containing an antioxidant p-hydroxybenzyl alcohol can signicantly suppressed the inammation in ischemic tissues of C57BL/6J mice. 30 However, it should be noted that these developed NPs with complicated polymeric structures are challengeable for reproducible synthesis, structural tailoring and quality control in large-scale production. Besides, in vivo clearance performance, degradation, metabolism and safety prole of NPs are still unknown, which facilities the further development of intrinsically active and translational NPs in targeted therapy of AS.
Herein, a newly kind of synthetic liposomes-like NPs with ROS-eliminated capability was developed, where the fatty acid chains of typical lipids were replaced by two Tempol molecules. Compared with the studied polymeric NPs assembled from complicated amphiphilic polymers, our work is purely relied on well-dened ditempolphosphatidylcholine (Tempol-PC, Scheme 1). By eliminating ROS, Tempol-PC-based NPs, abbreviated as Tempol-Lips can signicantly reduce oxidative stressinduced inammation in AS lesion. In vivo therapeutic investigations in ApoE −/− mice proved that Tempol-Lips inhibited the progression of AS by decreasing local and systemic oxidative stress and inammation. This strategy provides an efficacious and safe nanotherapy for treatment of AS.

Synthesis of ditempolphosphatidylcholine lipids
Lipids ditempolphosphatidylcholine (Tempol-PC lipids) were synthesized through conjugating Tempol onto GPC, as shown in Fig. 1a. Tempol reacted with succinic anhydride (Tempol-COOH) was rstly prepared. Briey, succinic anhydride (1.76 g/17.43 mmol) was mixed with a solution of Tempol (1 g/5.81 mmol) and DMAP (0.36 g/2.91 mmol) in anhydrous CH 2 Cl 2 and allowed to react for 0.5 h at 35°C. Aer washed with 0.1 M HCl three times, the synthesized product was puried by silica gel column chromatography eluting with ethyl acetate/hexane Next, Tempol-COOH (0.5 g/1.84 mmol) was added to a solution of CDI (0.45 g/2.76 mmol) in dry CH 2 Cl 2 and reacted for 4 h at room temperature. Without further purication, GPC (0.22 g/ 0.84 mmol) and DBU (0.28 g/1.84 mmol) dissolved in DMSO were added into the system, heated and stirred at 45°C overnight. The resulting products were precipitated in diethyl ether (40 mL × 3) acidied with 2 mL of glacial CH 3 COOH and then, puried by silica gel column chromatography using CH 2

Tempol-Lips formulation
Tempol-Lips were prepared by using conventional thin lm technique, with some modications. 31 An aqueous solution of Tempol-PC lipids (10 mg) in CHCl 3 was evaporated to form a thin lm cling to the walls of ask. This lm was hydrated with 5 mL PBS (pH 7.4) solution during rotation for 0.5 h at 50°C. Liposomes-like NPs were nally collected by homogenization with mini-extruder set (Avanti Polar Lipids Inc., Alabaster, AL) and ltration through 220 nm aseptic membrane, prior to different characterizations. The concentration of Tempol-PC lipids was determined by UPLC (ACQUITY Arc System, Waters, MA, USA).

Particle size and zeta-potential measurements
Hydrodynamic diameter and zeta-potential of Tempol-Lips were measured using dynamic light scattering (DLS) technique equipped with Zetasizer NanoZS90 and zeta-potential analyzer (Malvern Instruments Ltd, Worchestershire, UK). Average particle size and zeta-potential of liposomal formulations were determined in distilled water or serum-containing growth medium.

Morphological characterization
Morphology of Tempol-Lips with a concentration of 200 mg mL −1 in PBS (pH 7.4) was analyzed by transmission electron microscopy (TEM). Sample was placed on a carbon lm copper grid, air-dried and negatively stained with phosphomolybdic acid (2%, w/v). The images of TEM were visualized by JEM-2100 system (JEOL, Japan), which was operated at an acceleration voltage of 80 kV.

ROS-scavenging capability
ROS-scavenging capability of Tempol-Lips was measured using a previously established protocol. 32 Briey, various concentrations of Tempol-Lips ranged from 0 mg mL −1 to 200 mg mL −1 were incubated with 2 mL of PBS (pH 7.4) containing 500 mM H 2 O 2 for 48 h. The remaining H 2 O 2 was quantied by a uorometric hydrogen peroxide assay kit (Sigma-Aldrich, MO, USA), and eliminated H 2 O 2 was calculated.
The superoxide anion-scavenging capability of Tempol-Lips was evaluated by incubation with superoxide anion at 37°C for 40 min, using a commercially available test kit (Beyotime, Shanghai). Moreover, the free radical scavenging capability of Tempol-Lips was further measured. To this end, 1 mL of DPPHc (100 mg mL −1 ) was treated with different amount of Tempol-Lips at 37°C for 0.5 h, following recording the absorbance at 517 nm by UV-visible spectroscopy.

Cell culture
Murine macrophage RAW264.7 cells provided by Cell Bank, Chinese Academy of Science (CAS, Shanghai, China) were cultured in RPMI 1640 media with FBS (10%, v/v), streptomycin (100 mg mL −1 ) and penicillin (100 U mL −1 ) in a humidied environment of 5% CO 2 in air at 37°C. Cells were harvested for the following experiments as the pre-cultured cells reached 80% conuence.

Cytotoxicity evaluation
1.0 × 10 4 cells per well of RAW264.7 cells were cultured with 96well plates and allowed to attach for overnight. Cells were treated with Tempol-Lips at different doses for predetermined 24 h. The cell viability was quantied by MTT assay per manufacturer instructions and analyzed at 490 nm using a Model 680 Microplate Reader (Bio-Rad, CA, USA). Cells without treatment was used as controls. The percentage of cell viability was presented as comparison with the cells incubated in DMEM medium while tested in sixtuplicates and performed in quartets.

Cellular uptake
RAW264.7 cells were cultured in 3.5 cm-confocal dished containing 2 mL of growth RPMI-1640 medium for 12 h. Aer treated with 1 mg mL −1 LPS in free medium, equivalent 5 mM of Cy5.5-loaded Tempol-Lips was added for further 2 h incubation. RAW264.7 cells were rinsed and stained with DAPI (100 mM in PBS). Finally, confocal laser scanning microscopy (CLSM) was conducted to acquire uorescence photos.
For ow cytometric analysis, RAW264.7 cells with a density of 1 × 10 6 cells per well were cultured in 6-well plates for 12 h attachment. Then the cells were treated with 1 mg mL −1 LPS and further incubated with 5 mM of Cy5.5-loaded Tempol-Lips. At predened time points, cells were collected for ow cytometric analysis (BD FACSCanto, NJ, USA).

Anti-inammatory effects in vitro
Briey, RAW264.7 cells were seeded in 6-well plate at 1 × 10 6 cells per well. Aer 12 h, cells were treated with Tempol-Lips or free Tempol for 2 h and then stimulated with 1 mg mL −1 LPS for 24 h. The inammatory cytokines including TNF-a and IL-1b in the culture supernatant were determined by ELISA assays, where the levels of total protein were analyzed by BCA (Beyotime, Shanghai, China).

Intracellular ROS generation
RAW264.7 cells were pretreated with free Tempol and Tempol-Lips (10 mM) for 2 h and stimulated with LPS (1 mg mL −1 ). The normal control group was treated with fresh medium, while model group was only stimulated with LPS at equivalent dosage. Subsequently, cells were incubated with 10 mM of DCFH-DA in serum-free RPMI 1640 for 0.5 h in the dark at 37°C. The intracellular ROS change was observed by parallel CLSM (FV3000, Olympus, Japan), and the uorescence intensity of green DCF was determined using FCM (BD FACSCanto, NJ, USA).

Animals
Animals tests and care were performed in consistence with the Provision and General Recommendation of Chinese Experimental Animal Administration Legislation and approved by the Institutional Animal Care and Use Committee of the Affiliated Yantai Yuhuangding Hospital of Qingdao University. Male ApoE −/− mice (6-8 week) were supplied by the Pengyue Pharmaceutical Co., Ltd (Jinan, China), preserved in a humidity standard (60 ± 5%) and controlled temperature (22 ± 2°C) animal room and allowed free access to food and water.

Pharmacokinetic and plaque targeting study
The pharmacokinetic prole of Tempol-Lips was studied using Cy5.5-loaded liposomes aer i.v. injection. Cy5.5-loaded Tempol-Lips was administrated to ApoE −/− mice at 1.2 mg kg −1 . At pre-dened time points, 100 mL of whole blood samples was obtained in 96-well black plate. For plaque targeting investigation, male ApoE −/− mice administrated with Cy5.5-loaded Tempol-Lips were perfused with 4% paraformaldehyde, euthanized, and the aortas were isolated. Ex vivo imaging was observed by an IVIS spectrum system (Lumina 3, PerkinElmer, CA).

Therapeutic efficiency
Fiy ApoE −/− mice aer fed with high-fat diet for 2 months were randomly assigned into 5 groups (n = 10). Then mice were treated with 0.9% saline, probucol (5.2 mg kg −1 ), free Tempol (17.2 mg kg −1 ) and Tempol-Lips (20 mg kg −1 ). Probucol was dissolved in saline with 30% ethanol and free Tempol was in saline. All formulations were i.v. injected every other day for continuous 1 month.

ORO staining
Aer various treatments, ApoE −/− mice were sacriced. The aorta was resected and then perfused with 10% formalin for 1 h. The aorta was opened longitudinally and stained with ORO. Moreover, cryo-sections of aortic root was further stained with ORO. Plaque area analysis was carried out with the Image-Pro Plus 6.0 soware.

Dihydroethidium staining
Aortic roots of different groups were embedded in Tissue-Tek O. C. T. compound, and 6 mm-thickness sections were achieved on a Leica cryostat. Dihydroethidium (DHE) as uorescent dye was used to probe the generation of superoxide anion in situ, according to description reported previously. The sections were incubated with 2% triton X-100 and blocked with 5% BSA in PBS. Aerwards, slides were stained with DHE (2 mM) in Krebs solution and nally observed by CLSM.

Inammatory cytokines determination
The excised aortas were homogenized and centrifuged in saline. Supernatant samples were collected and the levels of TNF-a and IL-1b inammatory cytokines were measured using ELISA assays (Beyotime, Shanghai, China). Similarly, TNF-a and IL-1b cytokines in serum were determined.

Serum lipids measurements
Aer different treatments, the levels of triglycerides (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C) in serum were measured using commercially available kits (Solarbio, Beijing, China).

Safety evaluation
During the treatment period, the body weight of ApoE −/− mice were weighted every other week for possible adverse effects. Aortas and major tissues were further dissected and analyzed for H&E staining, as described previously.

Statistical analysis
Results were expressed as mean ± SD, and statistical analysis between groups was analyzed by GraphPad Prism Soware (Version 8.0, La Jolla, CA, USA) using one-way ANOVA. Statistical signicance was assessed at *P < 0.05 and **P < 0.01.

Synthesis, preparation and characterization of Tempol-Lips
Amphiphilic Tempol-PC lipids was synthesized by a two-step reaction, according to the scheme in Fig. 1a. As shown, Tempol-COOH intermediate was rst prepared by esterication of Tempol and SA under the catalytic condition of DMAP and TEA, following covalently conjugated to GPC. The synthesized Tempol-PC lipids were successfully by 1 H NMR and 13 C NMR spectra (Fig. 1b and c). Compared with 1 H NMR spectrum of the intermediate Tempol-COOH, newly appeared peaks at 3.25 ppm is belong to the methyl protons (-N + (CH 3 ) 3 ) from GPC molecule. In addition, the peaks at 54.3 ppm are the characteristic carbon signals of -N + (CH 3 ) 3 in 13 C NMR spectrum, which further conrmed the esterication between Tempol-COOH and GPC.
Liposomes-like NPs (Tempol-Lips) assembled from Tempol-PC lipids were prepared by a conventional thin lm technique. The average size and morphology of Tempol-Lips were studied by using DLS and TEM. As displayed in Fig. 2a, the hydrodynamic diameter of Tempol-Lips was 112.07 ± 1.78 nm with a negatively charged zeta-potential of −23.56 mV in PBS (pH 7.4), indicating the formation of NPs with a narrow size distribution (PDI = 0.132). Characterization by TEM revealed the spherical morphology for the desired Tempol-Lips (Fig. 2b). These results collectively demonstrated the self-assembly of Tempol-PC lipids, while the dual Tempol molecules play a critical role in the assembly of Tempol-Lips. Based on the well-dened structure of Tempol-PC, the content of Tempol loaded in NPs is higher upto 44.9% calculated by molecular weight aer a simple calculation.
Tempol is a free radical scavenger with equivalent efficacity of superoxide dismutase (SOD), which has been demonstrated effective in treatment of oxidative-stress-related diseases. 17,33 Thus, in vitro ROS-eliminating capability of assembled Tempol-Lips was studied and results were shown in Fig. 2c. As expected, Tempol-Lips scavenged H 2 O 2 in a dose-dependent manner aer 48 h post-incubation. Similarly, it was also found that the eliminated DPPH radical was remarkably increased along with the improved concentration of Tempol-Lips, conrming their signicant ROS-scavenging ability. These results potentiated Tempol-Lips as anti-AS platform in in vivo administration. Moreover, different simulated pHs (7.4 or 5.0) with 10% FBS or not were used to investigate the kinetic release of Tempol from Tempol-Lis by dialysis (MWCO 1000) method at 37°C (Fig. 2d). Appropriately 40.5% of parent Tempol was released form Tempol-Lips aer 72 h of incubation in PBS (pH 7.4) solution with FBS (10%, v/v), whereas Tempol-Lips showed an accelerated dissociation and acid-responsive Tempol release upto 92.1% at pH 5.0. Accordingly, the breakage of ester bonds incorporated in Tempol-Lips at low pH would lead to the disintegration and payload release under acidic microenvironment of inammatory macrophages in AS lesions, 34,35 therefore contributing their possibilities of enhancing anti-AS activity in vivo.

Anti-oxidation and anti-inammation of Tempol-Lips in vitro
Overproduced ROS and sustained oxidative stress can induce cell and tissue damage, which in turn triggers inammatory circulation and results in amplication of oxidative stress. 36 The antioxidation of Tempol-Lips was checked against inammatory RAW264.7 macrophages. The model group that RAW264.7 cells were merely stimulated by LPS for 4 h displayed a considerably high level of ROS, as probed by a DCF-DA uorescent dye with green uorescence (Fig. 3a). By contract, aer pretreated with free Tempol and Tempol-Lips for 2 h, the intensity of uorescent signals of DCFH-DA was largely reduced, particularly in the case of Tempol-Lips-treated cells. Further quantitative analysis by ow cytometry also revealed that intracellular ROS generation in LPS-activated RAW264.7 cells was maximally inhibited by Tempol-Lips. This promising anti-oxidation of Tempol-Lips can be attributed to the nanostructure of cell membrane-like NPs that is benet to the enhanced uptake in a manner through adsorption-mediated endocytosis (Fig. 3b) and as a result, more Tempol accumulated into these cells, thereby endowing its anti-inammation. To investigate this hypothesis, the cytotoxicity of Tempol-Lips was evaluated against RAW264.7 cells using MTT assay (Fig. 3c). Aer incubation for 24 h, RAW264.7 cells showed >80% cell viability ranged from 10 mg mL −1 to 320 mg mL −1 , disclosing good cytocompatibility of Tempol-PC lipids. Based on the above-mentioned feature, in vitro anti-inammation of Tempol-Lips was nally determined by checking their inammatory response in macrophages (Fig. 3d and e). Treatment of RAW264.7 cells with LPS stimulation promoted the production of TNF-a and IL-1b pro-inammatory cytokines, whereas cells pretreated with free Tempol or Tempol-Lips at equivalent dose of 10 mg mL −1 were found with inhibitive expression of these cytokines. In particulate, treatment with same doses of Tempol- Lips exhibited more positive on anti-inammatory effects, compared to that of free Tempol, which is in consistence with the results of enhanced internalization in vitro. Previous ndings have shown that ROS can activate multiple signal transduction cascades, which in turn modulate inammation in atherosclerosis. 4,9 Consequently, Tempol-Lips can attenuate inammation in macrophages by decreasing intracellular ROS production.

Therapeutic efficacy
In vivo pharmacokinetic prole of Cy5.5-loaded Tempol-Lips was rst analyzed in atherosclerotic ApoE −/− mice aer 2 month fed with high-fat diet. Following i.v. administration of free Cy5.5 and Cy5.5-loaded Tempol-Lips, blood samples from each group were collected and evaluated by measuring the relative uorescence intensity at various time points using an in vivo uorescence imaging system (IVIS, PerkinElmer, MA). As shown in Fig. 4a, uorescence imaging indicated that Tempol-Lips signicantly extended blood circulation with a span of 24 h, while free Cy5.5 was almost completely removed from the blood. This phenomenon would be benet for the accumulation of Tempol in AS plaques. Then the uorescence in the isolated entire aortas from each group was further recorded to evaluate in vivo targeting capability of i.v. injected Tempol-Lips. At 24 h, more signicant accumulation of Cy5.5-loaded Tempol-Lips with almost 3-fold higher than free Cy5.5 group was observed (Fig. 4b), suggested that an i.v.-injected Tempol-Lips was able to accumulate in AS lesions.  Together with above potential results, therapeutic efficacy of Tempol-Lips in vivo was assessed. Aer receiving a high-fat diet for 2 months, ApoE −/− mice assigned randomly into 5 groups that were i.v.-treated with different formulations every other day. Apart from free Tempol, probucol, which was demonstrated to be effective as a small-molecule anti-oxidative drug for AS treatment in different animal models, was also used as positive group. At the end of the experiment, the entire aortas were obtained and stained with ORO (Fig. 4c). The saline group with high ORO positive area was seen, clearly indicating the form of atherosclerotic plaques. Treatment with Tempol-Lips afford considerable effects on the decrease of inammatory plaque at 20 mg kg −1 and quantication analysis of the average plaque area was down to 8.35% of total aorta tissue, while the data was 20.42% and 15.38% for mice treated with free Tempol and probucol, respectively. Much better outcomes of Tempol-Lips were signicantly achieved to attenuate the development of atherosclerosis. Consistent with this result, analysis on serum levels of TC, TG, LDL-C and HDL-C also revealed the most signicant anti-AS activity for Tempol-Lips (Fig. 4e-h). This might be ascribed to their favorable pharmacokinetic prole and targeting capabilities in vivo.
The mechanism responsible for in vivo anti-AS activity of Tempol-Lips was preliminarily studied by staining aortic root with DHE. As shown in Fig. 5a, sections of aortic roots from saline-treated ApoE −/− were observed with obvious uorescence, due to the reaction of DHE and superoxide anions that yields ethidium with red uorescence. 37 However, observation and quantitative analysis discovered that oxidative stress was drastically suppressed aer i.v-treated Tempol-Lips (Fig. 5b). Combined with similar result that lowest levels of ROS were tested (Fig. 5c), it can be concluded both lesional and systematic oxidative stress was mitigated by i.v.-delivered Tempol-Lips. Furthermore, two typical inammatory cytokines of TNF-a and IL-1b in the serum and aorta from ApoE −/− mice received Tempol-Lips treatment was found with lowest expression, compared with free Tempol and protocol ( Fig. 5d-g). Signicantly, Tempol-Lips were able to reduce systematic oxidative stress and inammation effectively as well as decrease inammation and oxidative stress in AS plaques.

Safety evaluation
The possible toxic side effects of NPs are a great matter of safety concern for biological applications in clinic. To evaluate biosafety in vivo, H&E-stained histological sections of the major organs (heart, liver, and kidney) were performed on AS model mice aer 1 month treatment with different Tempol formulations. No distinct pharmacological lesions were detected in sections of all treated groups (Fig. 6a). In addition, as shown in Fig. 6b, these formulations had little effect on the body weight changes in AS mice model during drug administration, collectively supporting the acceptable biocompatibility of Tempol-Lips at the examined dose.

Conclusions
A novel kind of ditempolphosphatidylcholine lipids (Tempol-PC) with ROS-scavenging capability was successfully synthesized by replacing fatty acid chains of typical lipids with two Tempol. Through a well-established thin lm evaporation approach, Tempol-PC lipids can assemble into a nanotherapy of liposomes-like NPs (Tempol-Lips) that efficiently internalized and eliminated overproduced intracellular ROS in inammatory macrophages. Importantly, treatment with Tempol-Lips resulted in signicant therapeutic outcomes with pathogenesis inhibition of AS compared with those treated with control groups., which was realized by reducing local and systematic oxidative stress and inammation in plaques. Combined with the safe prole aer i.v. long-term administration, Tempol-Lips is efficacious and promising as new nanotherapy for AS therapy.

Conflicts of interest
The authors declare no conict of interest.