A PEG-PLGA-based polymeric capsule as theranostic nanoplatform for active targeting atherosclerosis

Background Atherosclerosis plaque is a major cause of cardiovascular diseases across the globe, which is characterized by a gradual formation of atherosclerotic plaque, hardening, and narrowing of the arteries. Nano-materials can serve as powerful delivery platforms for atherosclerosis treatment. Herein we systematically developed polymeric capsule with active targeting property encapsulating curcumin (Cur) and C-C chemokine receptor type 2 with short-interfering RNA (CCR2-siRNA) for combined treatment, which was also loaded with Gd 3+ for magnatic resonance imaging (MRI) and a near-infrared uorochrome (IR780) for uorescence imaging (FI). Results The resulting nano-capsule exhibited a favorable hydrodynamic size and negative surface charge. More importantly, the negative charge and active-targeting peptide resulted in enhanced accumulation of nano-capsule in the established atherosclerotic plaques, thereby achieving targeted drug release. The in vitro and in vivo results indicated the effective diagnostic and therapeutic sensitivity, which could promote the development of theranostics for clinical implementation. Conclusions This study demonstrated that chemotherapy combined with gene therapy can displayed synergistic effect in atherosclerosis treatment, which possessed ecient management of atherosclerosis.


Introduction
Atherosclerosis is a progressive in ammatory disease which leads to the hardening and thickening of the arterial walls [1][2][3][4] . It is initiated by the dysfunction of endothelial cells, which subsequently promotes the proliferation of vascular smooth muscle cells and reconstruction of extracellular matrix, ultimately leading to the formation of plaques, which are comprised of lipids, immune cells, and brous elements [5] .
It is a predominant cause of a variety of cardiovascular disorders including myocardial infarction, ischemic stroke, and peripheral vascular disease and accounts for signi cant morbidity and mortality rates in the aged population, especially in industrialized countries [6,7] .
There are generally no symptoms in the early stages of atherosclerosis and traditional detection techniques cannot easily, safely, and effectively detect the lesions in the early stages [8,9] . In recent years, there has been a rapid development in using nano-carrier technique as a tool for molecular imaging of 6 mL of aqueous solution containing 1% mixture of PVA and TPGS (outer aqueous phase, PVA: TPGS = 1:5), and ultrasonicated for 2 minutes (80 W) in an ice bath to form double emulsion (W/O/W). Then the emulsion was dispersed into 30 mL of 0.3% PVA solution and stirred overnight to volatilize chloroform and solidify the surface of the PLGA ball. After evaporation of the solvent, the samples were transferred to microsep (cut off: 100 KD) for ultra ltration cleaning and diluted into 5 mL of deionized water kept in 4 °C. The sample was died and weighed to calculate the concentration and solid content. PLGA-PEG-COOH capsule containing IR780, Cur, and Gd 3+ was also constructed in the similar method and denoted as PC.
Preparation of PLGA-PEG-COOH capsule modi ed with VCAM-1 polypeptide containing IR780, Cur, CCR2-siRNA and Gd 3+ (PC SV ) 1 mL of PLGA-PEG-COOH capsule (PC or PC S , 10 mg/mL) was centrifuged and dispersed into 0.5 mL of 2-morpholinoethanesulfonic acid (MES) (100 mM, pH 4.8), following adding VCAM-1 peptide (300 µg) at 37 °C. After adsorption for 1 h. 80 µL of 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC) solution was added for coupling at 37 °C overnight, which was prepared in MES solution (100 mM) and adjusted into the concentration of 240 mg/mL. The treated sample was puri ed by centrifugation with deionized water, and stored in 4 °C before use. PC and PC S conjugated with VCAM-1 peptide were denoted as PC V and PC SV , respectively.

Characterization Of Nano-capsule
The size and zeta potential of BPC, PC S and PC SV vesicles were determined using a Malvern Zetasizer Nano ZS unit (Nano ZS 90, Malvern, UK) with a He-Ne laser (λ = 633 nm) at a scattering angle of 90° at 25 °C. A drop of BPC, PC S or PC SV solution was dropped onto a copper mesh, and stained by 1% phosphotungstic acid. Subsequently, the morphology of BPC, PC S and PC SV solution were visually observed using a transmission electron microscope at 200 kV (TEM, JEM-2100F, JEOL, Japan).

Drug loading and in vitro drug release study
To calculate the loading e ciency of PC SV , lyophilized powder was dissolved and measured by UV-vis spectrophotometer (DU730, Beckman Coulter). Brie y, lyophilized powder was dissolved in DMSO, and the absorbance measured at 440 nm to obtain Cur loading e ciency. Likewise, lyophilized powder was dissolved in DEPC aqueous solution, and the absorbance measured at 260 nm and 280 nm to obtain CCR2-siRNA loading e ciency. According to the pre-established standard curve of Cur and CCR2-siRNA ( Figure S1, S2), the drug-loading content (DLC) and drug-loading e ciency (DLE) of nanoparticles were calculated by the following equations.
To study the drug release pro le in vitro, PC S and PC SV nanoparticle solutions (1 mL) were added to disposable dialysis cups (Slide-A-Lyzer MINI Dialysis Units, MWCO:3500 Da, Thermo Scientifc) in PBS (10 mL). At different time points, the external drug and gene release buffers were collected and an equivalent amount of PBS was added. The cumulative amount of Cur and siRNA released was quanti ed by UV-vis spectrophotometer.

Cellular Binding And Uptake Tests
The murine bEnd.3 cells were seeded in 12-well plates at a density of 1 × 10 5 cells per well in 1 mL of DMEM medium containing 10% FBS, and cultured at 37 °C with 5% CO 2 for 24 h. When the cells reached 60-70% con uence, 3 ng/mL tumor necrosis factor-α (TNF-α) was added and further incubated for 12 h.

Distribution of nanoparticles in vivo
ApoE-/-mice were fed with a high fat diet (HFD) for 14 weeks, mice in the HFD group were randomly divided into three groups with four mice in each group. They were separately injected with PBS (0.01 M, pH = 7.4), PC V (0.6 mg Cur/kg, 5 mg Gd 3+ /kg) and PC SV (0.6 mg Cur/kg, 5 mg Gd 3+ /kg) via tail vein injections on days 1, 3, and 5. On day 4, the atherosclerosis in the thoracoabdominal aorta of the mice was examined by MRI coronal reformatted imaging using the 3.0 T GE MRI system using a fast spin echo sequence. For quantitative analysis, we calculated the signal-to-noise ratio (SNR) in the same 3 regions of interest placed in the aortic wall of the lower intrathoracic aorta on 3 different coronal reformatted images. For in vivo whole-body NIRF imaging, the IR780 uorescence from the anesthetized mice were collected by using a Bruker In Vivo Xtreme imaging system with 704 nm excitation and 740 nm emission lters. After in vivo imaging, the mice were sacri ced and perfused with 10 mL of PBS by injection through the left ventricle. The aorta, heart, liver, spleen, pancreas, and kidney were imaged ex vivo to obtain quantitative photon counts of their NIRF intensities.

Histology And Immunohistochemistry
ApoE-/-mice were randomized into 4 groups (four mice per group), and given the HFD for 14 weeks.
Then, the mice were subjected to the different treatments for one month. The mice injected with 5% sucrose served as the control group, while the other three groups were treated with either free drug, PC V or PC SV at a dose of 0.7 mg/kg of Cur every two days via tail vein injection. The body weight of mice was monitored every two days during the treatment to evaluate the systemic toxicity. After treatment for 30 day, the aortas, from the heart to the iliac bifurcation, were harvested. The aortic roots were xed with paraformaldehyde (4% in PBS) for 1 h, and then prepared to para n sections. After depara nizing, Masson's trichrome and Toluidine blue staining were used to quantify the content of collagen and the necrotic core, respectively. For immunohistochemistry analysis, Sections of the main organs including aortas, heart, liver, spleen, lung, and kidney were also analyzed by hematoxylin-eosin (HE) staining.

Complete Blood Biochemistry And Routine Analysis
Blood was collected in EDTA2K spray-coated tubes after treatment for one month, and immediately analyzed using an automated hematology analyzer (Sysmex KX-21, Sysmex Co., Japan). The concentrations of alanine aminotransferase and aspartate aminotransferase in plasma from different treatments were quanti ed by an automated analyzer platform (Roche Cobas C501, Roche Co., Switzerland).

Statistical analysis
The results were presented as mean ± standard deviation (SD). Statistical signi cance was analyzed using Student's t-test.

Results And Discussion
Fabrication and characterization of vesicles E cient solubilization is critical for hydrophobic Cur to reach a su cient dose in the atherosclerotic lesion [15,16] . We, therefore, generated PC SV by encapsulating the hydrophobic Cur into the hydrophobic capsule of PLGA using a micro-emulsion method. In contrast to the saturation solubility of Cur in water (only ≈ 6 µg / mL), we were able to load 117 µg of Cur into 1 mL of PLGA aqueous solution (117 µg / mL), thereby dramatically increasing its solubilization (more than 15 times greater than in water). As shown in Table S2, DLE and DLC of PC SV were 0.9% and 58.5%, respectively. For CCR2-siRNAs, DLE and DLC of PC SV were 0.228% and 100%, respectively.
Next, we rst examined the hydrodynamic diameters (D h ) of BPC, PC S and PC SV by DLS. As shown in Moreover, PC SV showed a relatively constant hydrodynamic diameter after long-term storage at room temperature, indicating its favorable stability properties (Fig.S3). This good stability of nano-carrier in physiological condition was important and bene cial for their in vivo application and intravenous injection [17][18][19] . In addition, surface zeta potential analysis showed that BPC had a zeta potential of − 20.8 ± 0.8 mV, whereas PC S (− 25.4 ± 1.5 mV) and PC SV (− 38.8 ± 1.7 mV) have lower values (Fig. 1d, e, f).
The targeting polypeptide conjugated nano-carriers further decreased the zeta potential, enabling long circulation to the plaque sessions. Since the physiochemical properties of all the nano-carriers in this study were similar, it was reasonable to presume that their ability to target atherosclerotic plaque would be dominated by the targeting ligand.
We further analyzed the morphologies of BPC, PC S and PC SV using TEM. Both nanoparticles showed a uniform sphere morphology. Moreover, they displayed a clear capsule nanostructure without conspicuous interconnection (Fig. 1g, h, i), which con rmed that the W/O/W emulsion nanostructure was successfully prepared. To verify the IR780 encapsulated in the nano-carrier, uorescent images of BPC, PC S and PC SV was displayed as Fig. 2a. This uorescent intensity suggested that IR780 were strongly retained on the inside of PC S and PC SV .
We designed drug and siRNA codelivery system, loading drug and siRNA by emulsion methods, which was a classic method for the preparation of nanoparticles [20,21] . The in vitro release of Cur from the nanoparticles was investigated in PBS (pH 7.4) solution to simulate the physiological environment (Fig. 2b). A fast release of Cur from both nanoparticles was observed in the initial 24 h (> 20%). After 60 h incubation in PBS, 44.2% of Cur was released from PC S nanoparticle, while 39.8% of Cur was released from the PC SV nanoparticle. When compared with the PC S nanoparticles, PC SV nanoparticles showed a slightly slower Cur release pro le, which may be ascribed to the additional polypeptide in the outer sphere acting as a diffusion barrier [21] .
The release curves of siRNA from PC S and PC SV nanoparticles were similar, and these were because the siRNA were encapsulated in the inner layer of the capsules, which could be extended released readily [22] . In contrast to Cur release pro les, there was a considerable delay in the release of siRNA from the PC S and PC SV , especially during the rst 4 h, during which almost none of the loaded siRNA was released (Fig. 2c). After 60 h, only 18.9% and 17.8% of siRNA loaded in the PC S and PC SV had been released, respectively. This delayed release of siRNA was likely due to the fact that siRNA was loaded only in inner aqueous phase. These results demonstrated that the design of nanoparticles with spatially segregated Cur in the outer oil layer and siRNA in the aqueous core was veri ed for successfully prepared.

In vitromacrophage inhibition and endothelial cell adhesion of nanoparticles
The cytotoxicity of the nanoparticles was evaluated in the macrophages. As shown in Figure 2d, all nanoparticles showed a dose-dependent cytotoxicity in RAW264.7 cells. To evaluate the in vitro cytotoxicity quantitatively, we calculated the IC 50 of the nanoparticles, de ned as the drug concentration required to kill 50% of the incubated cells. BPC, which was not loaded with Cur, did not show any cytotoxic effect (with an IC 50 of >100 μg/mL) (Fig.S4). As expected, the cytotoxic effect of PC V (IC 50 : 6.35 μg/mL) did not differ from that of PC SV (6.87 μg/mL). These results suggested that PC V and PC SV were comparable in inhibition of macrophage proliferation. In contrast, at the same dose, the slightly more potent antiproliferative activity of free Cur (IC 50 : 4.60 μg/mL) might be ascribed to the slower Cur release from PLGA of PC V or PC SV [23] . Moreover, bEnd.3 cells were used to develop the in vitro in ammatory cell model by stimulating with appropriate concentration of TNF-α. The expression level of VCAM-1 was gradually increased with the increase in TNF-α concentration [13,24] . In order to examine the targeting adhesion of these nanoparticles to the in ammatory cells, bEnd.3 cells were pre-incubated with TNF-a at a concentration of 3 ng / mL for 12 h then further incubated with PC SV for an additional 1 h. Fig.3 showed representative bright-eld micrographs of these nanoparticles bound to cells, revealing that these targeted nanoparticles can bind the TNF-α-stimulated cells. Importantly, the number of nanoparticles bound to the stimulated cells increased with the addition of TNF-α, which indicated that the cell a nity of nanoparticles to the stimulated cells can infer the cell in ammation degree.
Organ-level distribution of PC V and PC SV As silencing of CCR2, the monocytic chemokine receptor responsible for migration of the in ammatory monocyte subset to sites of in ammation can be realized by using siRNA technology, siRNA silencing of CCR2 (siCCR2) signi cantly can attenuate receptor protein expression in circulating and splenic monocytes [14] . We investigated the in vivo MRI e cacy of PC V and PC SV via comparison of scans obtained pre-injection and at treatment with agent injected into the tail vein of atherosclerotic mice on day 4. MRI revealed pronounced and heterogeneously distributed hyper-intense areas throughout plaquerich regions in the abdominal aorta after administration of nanoparticles. A representative example of the results of imaging ApoE-/-mice with PC V and PC SV were shown in Fig.4a. A clear and signi cant enhancement of image density was observed in the aorta wall post-injection in PC V and PC SV , indicating that nanoparticles attached with VCAM1 binding peptide was successful in targeting the plaques and enhancing them in MR images [25][26][27] . By comparison, the MR signal of PC SV was lowered than that of PC V after administration of nanoparticles. It indicated that intervention of siRNA lowered the MRI signal in association with attenuated monocyte recruitment. Therefore, the aortic section of siCCR2 treated ApoE-/mice had decreased MR signal. Adoption of the medical image-analysis software package eFilm allowed the signal intensity to be ascertained and the percent change in normalized enhancement ratio postinjection to be calculated. The results of this analysis were displayed in Fig.4b. The data con rmed the visual impression that PC V was an effective agent for the enhancement of images of the wall of the aorta of ApoE-/-mice, giving an enhancement of 11%, while that of PC SV was only 6%, which showed that the PC V was targeted more effectively to atherosclerotic lesions [28] .
We also monitored the organ-level distribution of PC V and PC SV by utilizing IR780-labeled capsules.
ApoE-/-mice received treatment of PC and PC V on day 4, followed by the performance of in vivo NIRF imaging of the animals (Fig.4c) and ex vivo NIRF imaging of the excised organs (Fig.4d). For PC SV group, we observed stronger signals in the liver and lung and ex vivo imaging analysis revealed the most-intense IR780 uorescence in the liver among all of the organs. Other organs with a noticeable accumulation of PC were the lung, kidney, spleen. For the PC V group, in vivo and ex vivo NIRF imaging data also showed strong uorescence in the liver, lung, and spleen. Our NIRF imaging data suggested the more accumulation of PC V in heart and aorta than that of PC SV (Fig.4e), which recon rmed decrease of VCAM1 in the PC SV group. As VCAM1-binding peptide conjugation enabled high targeting capability of the nanoprobe to the in amed endothelium expressing VCAM1 in a high plaque regions [29] , the decreased endothelium expressing VCAM1 attenuated targeting of peptide, thus lowered the intensity of NIRF imaging. The results of MR and FI all implied that there was a more favorable therapeutic effect in PC SV compared with PC V group in the process of treatment.

In Vivo target atheroprotective effect
We detected the composition of atherosclerotic plaque in aortic root sections by immunohistochemistry staining. The results showed that the atherosclerosis plaque in the PC V and PC SV group disappeared on the whole, and no plaque rupture was observed. In contrast, other groups (control and Cur) displayed obvious atherosclerosis plaque and plaque rupture was observed in saline-treated groups (Fig.5a).
The necrotic areas in the aortic roots were detected by toluidine blue staining. As shown in Fig.5b, the control group exhibited large necrotic areas with substantial cholesterol crystals, indicating advanced lesions. After treatment with Cur, PC V and PC SV , the necrotic area was signi cantly decreased.
Quantitative analysis revealed that, compared to the control group, the average necrotic area was decreased to 14.5%, 7.3%, and 5.2% in response to free Cur, PC V , and PC SV treatment, respectively (Fig.5c).
The increased collagen produced by hyperplasia smooth muscle cells leads to the enlargement of the plaque areas, which could further narrow the vascular lumen [30,31] . We tested the content of collagen in plaque areas using Masson's trichrome staining (Fig.5d). Compared to the control, free Cur, PC V , and PC SV effectively decreased the content of collagen. Quantitative analysis revealed that, compared to the control group, the average content of collagen was decreased to 20.5%, 14.3%, and 11.2% in response to free Cur, PC V , and PC SV treatment, respectively (Fig.5e). Therefore, PC V and PC SV all exhibited preferable therapy on the plaque in ApoE-/-mice by reducing necrotic area and collagen deposition. Due to intervention of siRNA, PC SV can achieve more favorable therapeutic effect, leading to a more decreased plaque index [32] .

Biosafety Assessment
To assess biosafety, adverse effects were studied after treatment for one month. There was less 10% body weight loss in free Cur, while no signi cant difference in body weight of mice was observed in PC V and PC SV groups (Fig.6a). The mice receiving various treatments showed no signi cant change in the weight compared to that of untreated control, which showed that mice exhibited tolerance to all the treatments. Moreover, as the nano-capsules accumulated in the liver, the effect of liver was needed to be veri ed. No obvious difference in the liver index of suggested no signi cant toxicity, and the liver index of the group treated with PC V gradually returned, which was closed to normal level. The blood biochemical assays of alanine aminotransferase (ALT) and aspartate aminotransferase (AST), were at normal levels, which indicated that the functions of the liver were not impaired by the treatment (Fig.6b,c) [24,31] . Therefore, PC SV did not induce signi cant adverse effects in long-term treatment, indicating its potential as a safe candidate for chronic vascular disease therapy. The results of hematoxylin-eosin (H&E) staining showed that no noticeable pathological changes could be found in the main organs, which further con rmed their biocompatibility (Fig.6d).

Conclusions
In summary, we developed polymeric capsule with active targeting encapsulating Cur and CCR2-siRNA for combined treatment, which was loaded with Gd 3+ for MRI and a near-infrared uorochrome (IR780) for FI.     a MRI images of the abdominal aorta, indicated by arrows, of an ApoE-/-mice, before (left) and 24 h postinjection (right) with PCV or PCSV. b Quantitative analysis of MRI signal enhancement in T1-weighted images on day 4 after injection of the contrast agent in ApoE-/-mice. Data represent mean ± SD (n = 4). c In vivo NIRF imaging of ApoE-/-mice intravenously injected with PCV or PCSV after 4 days treatment. d Ex vivo NIRF imaging of the heart and aorta, and other organs collected from the injected ApoE-/-mice after 4 d post-injection. e Quantitative analysis of the heart and aorta, and other organs distribution of NIRF in ApoE-/-mice on day 4 after intravenous injection. Data represent mean ± SD (n = 4). Figure 5 a Representative micrographs of H&E-stained aorta sections from ApoE-/-mice treated with saline, free Cur, PCV, and PCSV for 30 days (scale bar = 500 μm). b The images of the necrotic areas stained by Toluidine blue (scale bar = 500 μm). c Quantitative data of the necrotic areas in the aortic root sections. d The images of collagen in the plaque areas stained by Masson's trichrome (scale bar = 500 μm). e Quantitative data of the content of collagen in aortic root sections. n = 4, mean ± SD, *p < 0.05, **p < 0.01, and ***p < 0.001. Page 20/21 Figure 6 a Effect of different treatments on body weight, Data represents mean ± SD. b The biochemical assays of aspartate aminotransferase (AST). n = 4, mean ± SD. c The biochemical assays of alanine