Evaluation of promoting effect of a novel Cu-bearing metal stent on endothelialization process from in vitro and in vivo studies

Drug eluting stents (DES) have been extensively applied nowadays and reduce the incidence of in-stent restenosis (ISR) greatly as compared with bare metal stents (BMS). However, the development of DES is hindered by the risk of late stent thrombosis (LST) due to delayed re-endothelialization, while endothelialization is an important process related to ISR and LST after implantation. 316L is a traditional stent material without bioactivity and have a high risk of ISR. Cu is recognized for angiogenesis stimulation in these years. Hence a copper bearing 316L stainless steel (316L-Cu) was prepared and evaluated about its effect on endothelialization in this paper. Compared with traditional 316L, it was proved that 316L-Cu increased the proliferation of co-cultured human umbilical vein endothelial cells (HUVECs) at first day. Moreover, HUVECs stretched better on the surface of 316L-Cu. It also improved the expression of angiogenesis related genes and tube formation ability in vitro. 316L-Cu-BMS, DES and 316L-BMS were implanted in swine to evaluate the re-endothelialization ability in vivo. And 316L-Cu-BMS showed the best effect on endothelialization with good biosafety. Consequently, 316L-Cu is a kind of promising BMS material for coronary field.

recovers myocardial VEGF expression 20 . In addition, copper stimulates VEGF expression in cultured human keratinocytes and promotes angiogenesis and wound healing 18 . Copper ions also have been reported to enhance angiogenesis by stabilizing the expression of hypoxia-inducible factor (HIF-1α), thus mimicking hypoxia, which plays a critical role in the recruitment and differentiation of cells and in blood vessel formation 11 .
In order to reduce the incidence of ISR, a novel coronary stent used material, copper bearing 316L stainless steel (316L-Cu), was designed and manufactured in previous studies, in which appropriate amount of Cu was added in the stainless steel based upon the chemical composition of commercial 316L stainless steel during metallurgical process 21,22 . Thus, 316L-Cu is a kind of integral Cu-containing materials. Nano sized Cu-rich phases would precipitate in the matrix of 316L-Cu and release Cu ions in moist or liquid environment 21,23 . Thus, Cu will exist both surface and internal and we did lots of studies about its observations and analysis of microstructures as depicted in previous studies 24,25 . Meanwhile, most of the properties including roughness and mechanical properties are same to 316L. It is remarkable that 316L-Cu was proved to both reduce the smooth muscle cells proliferation and thrombosis, as well as inhibit inflammation after coronary stent implantation in pre-in vivo tests, which is promising for reducing ISR in the future application 22,26 . Mandinov et al. also found that ISR could be inhibited directly by oral copper chelation in porcine coronary 27 . Thus, the beneficial influence of Cu ions on reduction of ISR was well recognized in the coronary field. However, based upon the problem of unsatisfactory re-endothelialization or dysfunction of DES, the effect of 316L-Cu on endothelialization or revascularization in vitro and in vivo should be considered and studied seriously before clinical application.
Thus, the aim of this work is to study the effect of 316L-Cu on endothelialization through in vitro and in vivo tests including evaluations of the proliferation of human umbilical vein endothelial cells (HUVECs), cell morphology, migration and tube formation ability of cells as well as the expression of some endothelium differentiation genes on the surface of 316L-Cu. Common 316L were served as control group in vitro assays. In addition, 316L-Cu stents were implanted in porcine coronary model for 14 days in order to further study their effect on endothelialization. 316L-BMS and DES were served as control groups. Hope that the in vitro and in vivo tests could provide scientific data for the clinical application of the promising coronary stent material 316L-Cu.

Results
Proliferation, vWF synthesis and morphology of HUVECs. CCK-8 was performed to compare the proliferation of HUVECs on different substrates for 1 day and 3 days. It could be seen that cells in both groups grew with increasing days (Fig. 1). Compared with 316L, 316L-Cu stimulated cell proliferation at first day significantly and no statistically significant difference existed between them at day 3. In general, 316L-Cu showed no cytotoxicity on HUVECs.
von Willebrand factor (vWF) is a protein synthesized and secreted by endothelial cells and megakaryocytes, which is often used to identify endothelial cells. As shown in Fig. 2A, HUVECs on the surface of 316L and 316L-Cu synthesized vWF normally. Cytoskeleton is an important part for maintaining the cell shape and many physiological activities including attachment, migration, material transport and mitosis. Cytoskeleton staining could show cell morphology on the materials. As shown in Fig. 2B, F-actin filament was stained green, while nuclei was stained blue. Green fluorescence intensity was increasing significantly with culture time, indicating more expression of F-actin. There were larger amount of filopodia extending on 316L-Cu, revealing that 316L-Cu supplied a better surface for HUVECs to survive on.
Effect of ionic production on the migration and tube formation of HUVECs. In transwell migration assay, ionic production attracted cells towards lower compartment. As shown in Fig. 3A, it seemed that more cells moved through the membrane due to the effect of 316L-Cu extracts revealing that 316L-Cu activated cell migration. However, there was no statistically significant difference between the quantitative data of two groups (Fig. 3B). For tube formation assay, HUVECs began to form tube-like structures after incubation on the matrigel. As shown in Fig. 4A, cells cultured with 316L-Cu extracts were connected together to construct network shape structures while tube in control group was not formed almost. The quantitative value of tube length (Fig. 4B) and the number of branching points (Fig. 4C) were higher in 316L-Cu group as compared to 316L group with significant difference. This result indicated that 316L-Cu enhanced migration and tube formation activity of HUVECs in 3D structure.
The gene expression of VEGF and eNOS. VEGF and endothelial nitric oxide synthase (eNOS) is an important factor and enzyme in angiogenesis, respectively. VEGF is shown to promote the proliferation and migration of endothelial cells. eNOS could synthesize NO that regulates vasodilatation in angiogenesis. After cultured on the surface of 316L-Cu for 3 days, the relative mRNA expression of VEGF was enhanced by 1.6 times (Fig. 5A) while the relative mRNA expression of eNOS was increased by 1.5 times (Fig. 5B). Therefore, 316L-Cu improved the angiogenesis-related gene expression of VEGF and eNOS significantly.
QCA and OCT results. Based on the above promising in vitro results, in vivo studies were conducted to evaluate the re-endothelialization influence of 316L-Cu-BMS. A total of 24 stents of three types, including 316L-BMS,  DES, 316L-Cu-BMS were successfully implanted in coronary arteries of eight pigs. All of the animals survived in the intended study interval with no angiographic stent thrombosis or other clinical complications. The body weight and body temperature of the animals remained well throughout the study. Quantitative image analysis of coronary angiography (QCA) and optical coherence tomography (OCT) were listed in Tables 1 and 2, respectively. During 14 days, there were no significant differences in all the data statistically among the three groups (P > 0.05), indicating that 316L-Cu-BMS had good biosafety similar to the clinical used 316L-BMS and DES. SEM evaluation. Figure 6 displayed the observed endothelial coverage of the animals' arteries (more figures were shown in Supplementary Fig. S1). It was reported that plenty of endothelial cells emerged on struts just at 2 days after implant, and its coverage was nearly complete for BMS at 10 days 28 . So there was enough time for the endothelial cell to bind on the surface of stents in this paper. However, re-endothelialization onto the surfaces of struts was varied among the three types stents at 14 days. 316L-Cu-BMS exhibited most complete coverage following interventions in this period (Fig. 6A). The surfaces of the struts were completely covered   Fig. 6a1, indicating well re-endothelialization. However, incomplete endothelial coverage was observed on stent struts in DES group (Fig. 6B). In Fig. 6b1,b2, the exposed strut was visible, demonstrating poor re-endothelialization ability of DES. Although complete endothelial coverage for 316L-BMS (Fig. 6C), a numerous of particles (white arrow) were observed (Fig. 6c1), which maybe were red cells, white cells and platelets indicating a stronger acute inflammatory and thrombotic response as compared to 316L-Cu-BMS 29 .
Pathology analysis. In order to further evaluate the effect of re-endothelialization for 316L-Cu-BMS, hematoxylin-eosin (HE) staining was conducted. By day 14, animals were euthanized and the arteries were isolated. Figure 7 showed the typical HE stained images of cross-sections of arteries with stents. Consistent with the result of SEM, 316L-Cu-BMS showed better re-endothelialization as compared to DES. An anatomically intact endothelium had been re-constituted as depicted in Fig. 7A, and the zoomed-in image Fig. 7a reflected that the endothelial just covered on stent struts without neointimal hyperplasia. Whereas, most of the struts of DES were exposed without the coverage of endothelium (Fig. 7B) and the detailed structure could be observed in Fig. 7b, indicating poor endothelialization. For 316L-BMS, the re-endothelialization ability is between 316L-Cu-BMS and DES (Fig. 7C). Some struts are covered by endothelium, and some are exposed (Fig. 7c). Quantitative morphometric analysis of endothelialization scores also revealed that 316L-Cu-BMS showed the best re-endothelialization ability among the three types of stents (Fig. 7D).

Discussion
DES have reduced rates of restenosis and target lesion revascularization compared with BMS and launched a revolution in the interventional treatment of symptomatic coronary artery disease 1-6 . However, enthusiasm for this technology has been dampened by concerns about delay in arterial healing characterized by persistent fibrin deposition and delayed re-endothelialization, which increase the risk of LST, often with catastrophic consequences 1-10 . T.H. Kim investigated long-term (≥2 years) results of neointimal coverage following sirolimus-eluting stent (SES) and paclitaxel-eluting stent (PES) implantation using OCT. A follow-up angiography with OCT examination was performed in 29 patients with 32 lesions for more than 2 years. They found that incomplete coverage of the stent strut was still observed in a one-fourth of stents and considerable struts were not covered with neointima even 2 years after DES implantation 7 . The majority of DES used in clinical practice are designed to elute pharmacologic agents which could disrupt smooth muscle cell proliferation, such as sirolimus that inhibits the mammalian target of rapamycin (mTOR), a member of the phosphatidylinositol kinase-related family of serine/threonine kinase. However, these inhibitors of mTOR delay the growth and recovery of endothelial cells 1,3,5,6,30 . The first generation of DES, including polymer-based sirolimus and paclitaxel, is modified by lower drug concentrations, which may improve endothelialization and lessen drug toxicity when compared to the prototype. However, the lower drug concentration might enhance the restenosis rate again 1,3,5,30 . Therefore, a simple reduction in drug concentration does not mimic the extended-release stent. Second generation DES such as everolimus eluting stents (EES), Endeavor-zotarolimus eluting stents (E-ZES), and Resolute Integrity zotarolimus eluting stents (R-ZES) are designed with thinner strut backbone stents, less polymer and drug loading, and eluted analogues of sirolimus Consequently, the aim of our group is to research and develop a novel kind of BMS without drug coating, which can promote re-endothelialization and then enhance arterial healing process. In our previous studies, we successfully designed and manufactured a novel kind of 316L-Cu, which could be used as BMS material. It is believed that 316L-Cu could release copper ions sustainably and reduce the smooth muscle cells proliferation, thrombosis, inflammation, and promote HUVECs proliferation in vitro tests, which is expected to reduce in-stent restenosis in vivo 21,22 . It is commonly accepted that Cu element promotes angiogenesis by stimulation of endothelial cell growth and differentiation. However, it is not clear whether 316L-Cu-BMS could promote endothelialization process caused by Cu ions release from the material. Therefore, the influence of 316L-Cu on endothelialization in vitro and in vivo was conducted systematically and extensively in the present research in order to provide scientific basement for the clinic application of 316L-Cu that promoting endothelialization and healing process. The results of CCK-8 and F-actin filament staining assays have demonstrated that 316L-Cu promoted the proliferation and spread of HUVECs. Meanwhile, tube formation assays revealed that 316L-Cu was of great benefit to the migration and tube formation activity of HUVECs. All these above in vitro findings indicated that 316L-Cu could enhance endothelialization in contrast to common 316L. It's no doubt that Cu play a role in promoting migration of endothelial cells. There were lots of studies having the same results 13,17,31 . For example, C. Gerard et al. also found that CuSO 4 significantly induced the formation of tubular structures resembling to an angiogenic process 17 . In summary, Cu release indeed has a great beneficial effect on angiogenesis.
There are several factors involved in the regulation of endothelial cell growth and function, and most studies are about VEGF and eNOS 15,18,19,32 . VEGF has been shown to play a pivotal role in the initiation and formation of blood vessels, which involve endothelial cell growth and differentiation. D.C. Rigiracciolo et al. provided novel evidence regarding the molecular mechanisms by which copper may trigger the expression and function of VEGF toward angiogenesis. In particular, they had shown that copper activated the EGFR/ERK/c-fos transduction pathway leading to the expression of HIF-1α, G protein-coupled estrogen receptor (GPER) and VEGF 33 19 . Consequently, the effect of Cu 2+ on endothelialization is well recognized and explained. In this present work, 316L-Cu also improved expression levels of angiogenesis-related genes of VEGF and eNOS significantly in HUVECs, indicating that Cu 2+ release from the steel material could also play the role of endothelialization. Li et al. also found that the released Cu 2+ ions from a copper containing bioactive glass nanocoating on natural eggshell membrane successfully stimulated angiogenesis of endothelial cells by the improved expression levels of VEGF, HIF-1α, VEGF receptor 2 and eNOS as well as increased VEGF and HIF-1α protein secretion 15 . All these results well explain why Cu-bearing biomaterials with Cu 2+ release have the effect of promoting endothelialization 15 .
Based on the above promising results, in vivo studies were conducted to further evaluate the re-endothelialization ability of 316L-Cu. For QCA and OCT, there was no significant difference in all the related data and thus it could not indicate the reduction role of 316L-Cu on ISR with these current in vivo tests. It is because that we just focused on the evaluation of promoting re-endothelialization of 316L-Cu rather than on ISR in this study, and we did not establish an animal model for ISR. For this reason, no neointimal hyperplasia could be observed for the three types of stents. Although we could not get information about the effect of 316L-Cu on ISR reduction in this study, our previous in vitro tests have shown that it can reduce the smooth muscle cells proliferation and migration, thrombosis, and even inflammation and hence it was suggested to reduce ISR 22,26 . It is notable that no significant difference of QCA and OCT data proved that this 316L-Cu-BMS was biosafety, which was equal to the other two clinical application stents.
Further, SEM observation and HE evaluation were performed to measure the re-endothelialization ability of 316L-Cu-BMS. 316L-Cu-BMS showed the best re-endothelialization ability, whereas DES was the worst due to the drug coating, and 316L-BMS was between the first two. Consequently, 316L-Cu-BMS without drug coating could not only reduce ISR but also promote re-endothelialization. Thus, it is possible that patient may reduce the medicine intake for inhibiting thrombosis, such as clopidogrel, and thus decrease the side effects of medicine to human body, if 316L-Cu-BMS could get clinical application.
Recently, researches regarding to Cu-bearing materials, being as a novel implant material with numerous bio-functions including antimicrobial activity, promoting osteogenesis and even angiogenesis due to the release of Cu ions, are springing up vigorously 21,26,[36][37][38][39][40][41][42] . However, some researchers especially doctors worried about the biocompatibility of Cu-bearing materials because that Cu is a kind of heavy metallic element. It is true that the biocompatibility of novel biomaterial should be seriously considered before clinical application. For this reason, a large number of evaluations of in vitro and in vivo tests have done for this Cu-bearing stainless steel in our previous work and parts of the present work. The amount of Cu ions released from 316L-Cu has been measured before, approximately 0.0076 μg (0.12 μM)/day/cm 2 , which is much lower than the cytotoxicity limit of Cu against vascular endothelium cells of 500 μM 21 . For another instance, the approximately surface area of a 316L-Cu-BMS is around 0.34 cm 2 . According to this, a 316L-Cu-BMS stent would release 0.0026 μg Cu 2+ , which is much lower than the acceptable daily intake of Cu (2-3 mg per day per person) recommended by the World Health Organization 21 . Thus, from the above data analysis, 316L-Cu is biosafety for human body. Although the release amount of Cu is trace grade (ppb), what we measured is just the concentration of "planktonic" Cu ions in extracts. The concentration of Cu ions "trapped" on the surface is much higher than that. Others also found the same that the ion concentration was higher significantly in the near implant vicinity, which was decreased quickly along the distance away from the implanted alloys 43,44 . For this reason, 316L-Cu-BMS could play promoting effect on endothelialization in situ. It means that only when tissue contact 316L-Cu closely, it can play the role directly.
All of the above results indicated that the 316L-Cu exhibited satisfactory effect on re-endothelialization and good biocompatibility both from in vitro and in vivo tests. However, there are remaining some limitations in this present study, which should be seriously considered in the future research work for the final application in clinical. The promoting re-endothelialization of 316L-Cu for longer implantation (such as 1, 6, and 12 months) and more comprehensive data analysis should be seriously evaluated. Meanwhile in the future animal test evaluation of reduction role on ISR of 316L-Cu should be focused together. Metabolism of copper from 316L-Cu in human body was still unclear. All these should be assessed in the following step.

Methods
Preparation of copper-bearing 316L stainless steels. Based on the composition of commercial 316L, 316L-Cu was designed and fabricated by the vacuum induction melting furnace, whose chemical composition was analyzed as (wt.%): Cu 3.77, Cr 17.29, Mo 3.21, Ni 14.63, and Fe in balance. The cast ingot was reheated to 1100 °C, held for 1 h and then hot forged into plates. The forged plates were solution treated at 1000 °C for 0.5 h followed by a water quench, and then aged at 700 °C for 6 h. Commercial 316L was served as control group, and its composition was (wt.%): Cr 18.49, Mo 3.26, Ni 14.16, and Fe in balance. These two kinds of samples were machined into disks of 10 mm and 5 mm in diameters and 1 mm in thickness, respectively. They were mechanically grinded with silicon carbide papers down to 2000 grade and ultrasonically washed in acetone and ethanol. All the samples were sterilized prior to experiments.
Preparation of the extracts. Samples (Φ10 mm × 1 mm) were soaked in serum free RPMI.1640 medium at the ratio of 1.25 cm 2 /mL, according to the International Standard Organization (ISO 10993-5). They were stored at 37 °C for 72 h. The culture media incubated with the steel samples were collected as extracts.
Cell proliferation. The proliferation of HUVECs on different substrates was examined with Cell Counting Kit 8 (CCK-8, Dojindo). The smaller disks were put into 96-well plates firstly. A suspension consisting of 3 × 10 3 cells was seeded onto the samples. After co-culture for 1 day or 3 days, the solution was replaced by 110 μL mixture of fresh medium (100 μL) and CCK-8 solution (10 μL), followed by an incubation at 37 °C for 1.5 h. Then 100 μL liquid from each well were transferred to a fresh plate. The absorbance at λ = 450 nm was measured by a microplate reader.
Immunofluorescence staining. Immunofluorescence is a method to observe the vWF and F-acting expression of cells on the surface of different substrates. The samples of 10 mm in diameter were placed in 24-well plates. 100 μL cell suspension at a appropriate density was dropped on each disk. 1 mL culture medium supplemented with 10% FBS was added after cells were adherent. After incubation for 1 day or 3 days, the substrates were washed three times with PBS, fixed in 4% paraformaldehyde for 15 min, and then permeabilized with 0.5% (v/v) Triton X-100 for 10 min. Next was staining process for vWF and F-actin, respectively.
For vWF staining, cells were incubated with a rabbit anti-vWF primary antibody (1:100 dilution) overnight after blocking by 10% goat serum at room temperature for 30 min. Washed the samples and added FITC-labeled goat anti-rabbit second antibody (1:25 dilution) next day. After incubation for 1 h, the vWF expression was visualized using a fluorescent microscope.
For F-actin staining, HUVECs were stained with phalloidin-FITC (Sigma, USA) at a concentration of 5 ug/ml at 37 °C for 1 h. Cells were further stained with DAPI (Beyotime) to visualize the nuclei. An inverted fluorescence microscope was used to observe the results.
Transwell migration assay. The effect of different ionic extracts on cell migration was evaluated by transwell assay. HUVECs were suspended at a density of 2 × 10 5 cells/mL with serum free culture medium after 24 h starvation. 100 μL cell suspension was added in the upper chamber with a 8 μm pore. The lower chamber was filled with 600 μL extract supplemented with 5% serum. Washed the inserts with PBS after 24 h and removed the cells in the upper chamber by cotton swabs gently. Then the cells in the underside surface of the upper chamber were fixed with 95% alcohol, and stained with 0.1% crystal violet for dyeing. Finally, select five views randomly under the microscope and calculate the average cell number to represent the migration cell number.
Endothelial cell tube formation assay. The tube formation activity of HUVECs cultured with different extracts was investigated by observing capillary-like structures in matrigel. 50 μL matrigel (Corning) was transferred to the well in a 96-well plate by cold pipette tips. The plate was placed at 37 °C for 0.5 h to make matrigel solidified. 100 μL cell suspension with 5 × 10 4 cells were seeded per well and 100 μL ionic dissolution product with 10% serum was added. After incubation at 37 °C for 2 h, the tube formation of different groups was observed under a light microscope. Quantitative data including tube length and branching points were analyzed by WimasisWim Tube.

Real-time quantitative PCR (RT-qPCR).
Gene expression of VEGF and eNOS from cells cultured on different substrates was detected by RT-qPCR. The larger disks were placed in 48-well plates, and 3 × 10 4 cells were seeded per well. Three days later, total RNA of cells was extracted using a Trizol reagent (Vazyme) according to the manual. Cells in every three wells were collected as one sample. Three samples were performed for each group in each independent experiments. The RNA concentrations were measured on a nanodrop 1000 spectrophotometer (Thermo Scientific). About 400 ng of the extracted total RNA was reversely transcribed for cDNA with TaKaRa PrimeScript TM RT reagent Kit (Perfect Real Time) in 10 μL system. The following primers were designed by the information in NCBI database and the specificity was confirmed by Primer-BLAST: VEGF, forward 5′-AGTCCAACATCACCATGCAG-3′ and reverse 5′-TTCCCTTTCCTCGAACTGATTT-3′; eNOS, forward 5′-CATCTTCAGCCCCAAACGGA-3′ and reverse 5′-AGCGGATTGTAGCCTGGAAC-3′; GAPDH, forward 5′-GGAGCGAGATCCCTCCAAAAT-3′ and reverse 5′-GGCTGTTGTCATACTTCTCATGG-3′. 2 μL of cDNA, which was diluted with 5 μL RNase free dH 2 O, was mixed with 10 μL SYBR Premix, 6.4 μL dH 2 O, and Scientific REPORtS | 7: 17394 | DOI:10.1038/s41598-017-17737-9 1.6 μL primers at the concentration of 10 μM to prepare reaction solution according to TaKaRa SYBR ® Premix Ex Taq TM II (TliRnaseH Plus). Real-time PCR assays were performed on the LightCycler480. The mean cycle threshold (Ct) value of each target gene was normalized against Ct value of GAPDH and the relative expression was analyzed using the 2 −ΔΔCT method.
Stents manufacture. In order to investigate whether 316L-Cu have the function of re-endothelialization, 316L-Cu-BMS were successfully fabricated by our group (Fig. 8). 316L-Cu was processed into tube and then laser cut to fabricate stent. Commercial available 316L-BMS (SINO Medical Sciences and Technology INC, China) and Three types of stents were randomly assigned and placed in the left anterior descending (LAD), circumflex (LCX), or right coronary artery (RCA) in one animal. In order to decrease the incidence of acute thrombosis, premedication with 300 mg aspirin (Bayer, Germany) and 75 mg clopidogrel (Xin Li Tai Pharmaceutical Co. Ltd, China) was administered before stent implantation for one day. Aspirin (100 mg/d) and clopidogrel (75 mg/d) were then given until sacrificed. Subcutaneous injection of 0.3 mg/kg ketamine (Double-Crane Pharmaceutical Company, China) was administered to all animals prior operation for basic anesthesia, and then intravenous injection of 0.05 g/L sodium pentobarbital (Peking University First Hospital Animal Center, China) was given to keep the status of anesthesia and to extend the experimental time. Onwards, the groin of the pig was fixed and sterilized, and the femoral artery was punctured with Seldinger method. Then, 6F sheath was inserted with intravenous injection of heparin (100 U/kg). With 0.035 inch guide wire, 6F JL 3.5 and JR 3.5 guiding catheters were used for coronary angiography. According to the target vessel diameter, the stent size was selected (stent expanded diameter: vascular diameter, 1.1:1-1.2:1). Guide wire (0.014 inch) was sent to the distal coronary artery. Along the guide wire, the stent was pushed to the target vessel site and released by using 8 atm-12 atm single expansion for 5 s-10 s. The angiography was repeated to make sure that no serious dissection and thrombosis occurred. Then the sheath tube was withdrawn, and the puncture site was compressed to stop bleeding for 20 min-30 min, with intramuscular injection of 8 × 10 4 U/kg penicillin for preventing infection.  post-fixed in 1% osmium tetroxide for about 30 min. They were then dehydrated in a graded series (50%, 60%, 70%, 80%, 90%, and 100%) of ethanol. After they had been dried at the critical point, the tissue samples were mounted, sputter-coated with gold and observed under an SEM (Hitachi S-3000N, Japan).
Histopathological evaluation. At 14 day of implantation, all the animals were sacrificed one after another. The heart was removed and perfused with heparin saline for 5 min at the pressure of 100 mmHg (1 mmHg = 0.0133 kPa). The stent vessel segment was separated rapidly and fixed with 10% formaldehyde solution. The samples with stent were embedded by methyl methacrylate (Prepared by Peking University School of Stomatology), and sliced (50 μm) with a hard tissue microtome (LEICA, Germany). All sections were stained with HE, and specimens were prepared for light microscopy (LEICA, Germany) examination. Stent endothelialization score was defined as the extent of the circumference of the arterial lumen covered by the endothelial cells and graded from 1 to 3 (1 = 0-25%, 2 = 25-75%, 3 = 75-100%).

Statistical analysis.
Values in normal distribution were expressed as mean ± standard deviation (SD).
Independent-sample T test was used to analyze the statistical difference in vitro. Group imaging and integral data in vivo were analyzed with one-way ANOVA test. n representing the number of stents of different types. The value of P < 0.05 was considered statistically significant. All statistical analysis were performed by commercially available SPSS 14.0 system software (IBM, USA).