Heparin- and Rosuvastatin- loaded Poly(L-lactide-co-caprolactone) Nanofiber Aneurysm Covered Stent Inhibits Inflammatory Smooth Muscle Cell in Reducing in-stent Stenosis and Thrombosis


 Background: Endovascular covered-stent has unique advantages in treating complex intracranial aneurysms. However, in-stent stenosis and late thrombosis have become the main factors affecting the efficacy of covered-stent treatment. Smooth-muscle-cell phenotypic modulation plays an important role in late in-stent stenosis and thrombosis. Thus, covered stents loaded with drugs that can inhibit smooth-muscle-cell phenotypic modulation may lower the incidence of long-term complications. Methods: Nanofiber covered stents were prepared using coaxial electrospinning. A rabbit-carotid-artery aneurysm model was established and treated with covered stents. Angiography and histology were performed to evaluate the therapeutic efficacy and incidence rate of in-stent stenosis and thrombosis. Phenotype, function, and inflammatory factors of smooth-muscle cells were studied to explore the mechanism of rosuvastatin in smooth-muscle cells. Result: Heparin–rosuvastatin-loaded nanofiber scaffolds inhibited the proliferation of synthetic smooth-muscle cells. Heparin–rosuvastatin-loaded nanofiber covered stent effectively treated aneurysms without showing any notable in-stent stenosis. In vitro experiments showed that rosuvastatin could inhibit the smooth muscle cell phenotypic modulation of platelet-derived growth factor-BB induction. The inflammatory cytokines secretion and cell viability were inhibited after rosuvastatin treatment.Conclusion: Rosuvastatin could inhibit the abnormal proliferation of synthetic smooth-muscle cells. Heparin–rosuvastatin-loaded covered stents could reduce the incidence of stenosis and late thrombosis and improve the healing rates of stents used for aneurysm treatment.


Background
Intracranial aneurysm (IA) is a cerebrovascular disease with an annual incidence of approximately 3% (1) and an annual rupture rate of approximately 1.4% (2). The mortality rate after aneurysm rupture can be as high as 60% (3). The International Subarachnoid Aneurysm Trial (ISAT) (4,5) and the International Study of Unruptured Intracranial Aneurysms (ISUIA) (6,7) proved that interventional therapy results in a lower mortality rate than a craniotomy (8). Although coil implantation is the basic method of IA treatment (9), it is unsuitable for complex IAs. Typical stents implanted to treat IAs currently include ow-diverter devices (10) and the Willis covered stent (11,12). Long-term stenosis and late thrombosis can substantially affect the therapeutic effect of the stent. Although rst-generation drug-eluting stents (DES) appeared to improve short-term in-stent stenosis, the ability of drugs to inhibit cell proliferation may cause delayed reendothelialization. Second-generation DES employed improved drug loadings to unilaterally inhibit the abnormal proliferation of smooth-muscle cells (SMCs) without affecting endothelial-cell (EC) functions.
DES polymers such as polytetra uoroethylene (PTFE) and polyethylene terephthalate (PET) can aggravate the human body's natural in ammatory immune response (13) by promoting the activation of in ammatory SMCs, thereby leading to in-stent stenosis. Although PTFE shows excellent performance in wide arteries, it is unsuitable for arteries narrower than 6 mm in diameter (14). To address this problem, Stack et al. developed a biodegradable scaffold with poly-L-lactide (PLLA) (15,16), wherein the biodegradable membrane decomposes into harmless molecules (17). Poly(L-lactide-co-caprolactone) [P(LLA-CL)] is a biodegradable material that can combine drugs stably and shows excellent mechanical properties (18,19). In addition, the biodegradation rate of P(LLA-CL) can be controlled by adjusting the molar ratio of PLLA in the copolymer, which makes it more suitable as a stent coating material.
Although previous studies have found that stent-induced platelet and EC activations can increase the risk of stenosis (20), the anticoagulant heparin can reduce thrombosis and prolong blood clotting after endothelial injury. Because rosuvastatin can promote EC proliferation and inhibit and in ammatory-SMC proliferation (21), the combination of heparin and rosuvastatin may emerge as an effective drug coating for application to next-generation stents. Therefore, in this study, a nano ber-covered stent was prepared by coaxial-electrospinning a stable, biodegradable copolymer and was loaded with both heparin and rosuvastatin.

Fabrication and structural characteristics of nano ber mats and covered stent
Using coaxial electrospinning, nano ber mats were fabricated (Fig. 1Aa). Nano ber mats wrapped the stent grafts completely to form covered stents (Fig. 1Ab). The magni ed images clearly show that the nano ber mats are composed of brous structures and do not show any structural dissolution at 37°C (Fig. 1Ac). The morphology of the nano ber scaffolds was observed under TEM. The shell is light and thin and shows clear edges, and the drug-loaded core is dark. Neither shows any aggregation or discontinuity (Fig. 1Ad). Figure 1B shows that the mean diameters of the Rosu 50, Rosu 75, and Rosu 100 core-shell bers were slightly thicker than that of the PBS ber. However, there were no statistical differences between their diameters.

Effects of heparin-rosuvastatin-loaded nano ber scaffolds on synthetic SMCs
SMCs are the key contributor to late in-stent stenosis, and SMC morphology is highly correlated with SMC phenotypic functions. Under SEM observation, the SMCs appeared as at triangles and as lesscontracted spindles in the PBS group. With increasing rosuvastatin ratio, the cell morphology was spindle-like in the Rosu 50 and Rosu 75 groups, and the number of at triangular cells decreased. In the Rosu 100 group, cells were slender spindle-like, and their morphological changes were more obvious than those of the cells in the PBS group ( Fig. 2A). Phalloidin staining showed that with increasing rosuvastatin concentration, the cell morphology gradually became spindle-like (Fig. 2B). The Hoechst-33342 staining showed that with increasing rosuvastatin ratio, the number of attached synthetic SMCs decreased (Fig. 2C). Using CCK-8, we quanti ed the number of synthetic SMCs attached to the nano ber scaffolds. After 24 and 48 h of culturing, the synthetic SMCs proliferated the most in the PBS group, and the proliferation activity decreased with increasing rosuvastatin ratio, showing a negative correlation (Fig. 2D). The in ammatory factors secreted by cells can be detected using the MILLIPLEX® Map kit. The in ammatory factors decreased signi cantly in the Rosu 100 group compared with the PBS one. (Fig. 3).

Balloon-expansion assay
After the attached-cell experiment, we preliminarily chose Rosu 100 as the most suitable volume ratio and fabricated nano bers on stents for 3, 5, and 10 min to test different thicknesses. Figure 4 shows the results of the balloon-expansion experiment. For the nano bers spun for 10 min, the proximal and distal ends of the stent began to expand at 4 bar. However, the expanded balloon could not drive the middle segment to expand, leading to a dog-bone-like stent graft. At 6.5 bar, the stent opened in pulsations. At 9 bar, no translucent phenomenon could be observed. The covered nano ber mats still retained a certain thickness and remained wrapped around the stent, thereby limiting stent expansion. Simultaneously, tiny cracks could be observed on the covered mats. At 9.5 bar, the balloon ruptured and leaked. For the nano bers spun for 5 min, the proximal and distal ends of the stent began to expand at 4 bar. When the lling pressure reached 6 bar, the stent expanded in pulsations and when the pressure reached 9.5 and 10 bar, although the nano ber mats appeared transparent and still covered the stent, ne cracks could be seen. For the nano bers spun for 3 min, the proximal and distal ends of the stent began to expand at 3.5 bar, the dog-bone shape disappeared at 5 bar, and the stent expanded completely, which was different from the other two nano ber mats. When the stent expanded completely, the mats clearly were wrapped around the stent, and the stent expanded smoothly without sudden expansion. When the lling pressure reached 6 bar, the stent was completely expanded, and the mat remained intact. The scatter chart prepared according to data obtained for the stent outer diameter shows that the outer diameter of the stent graft can be expanded with changing pressure. The optimal pad thickness was obtained for the nano ber scaffolds spun for 3 min; therefore, we carried out the remaining experiments on those stents.

Rabbit aneurysm model development, stent implantation, and short-and long-term follow-ups
The rabbit aneurysms were induced by injecting porcine pancreatic elastase into the right common carotid arteries (CCAs) of New Zealand white rabbits (Fig. 5Aa). After induction, the right CCA was larger than a normal blood vessel, forming a longitudinal aneurysm body. The right subclavian artery was the parent artery (Fig. 5Ab). DSA was performed 30 d after aneurysm induction. Angiography showed that aneurysms were induced (Fig. 5Ac), and the covered stent was delivered to the parent artery to cover the aneurysm neck. With the expansion of the covered stent, the aneurysm body completely disappeared in the DSA images (Fig. 5Ad).
After stent implantation, aneurysms were divided into three grades and used to evaluate the therapeutic effect during follow-up. Table 1 shows the grading standards. Figure 5B shows the angiography images of the 3 grades. Most of the stents achieved good coverage on immediate postoperative angiography after stent implantation, and there was no signi cant difference between the PBS and Rosu 100 groups.
However, at the 3-month (short-term) follow-up in the PBS group, the number of type A grade decreased by 40%, while types B and C grade both increased by 50%. Meanwhile, the number of each grade in the Rosu 100 group did not change, and the stent did not further promote thrombosis in the aneurysms. At the 12-month (long-term) follow-up, the number of type A grade in the PBS group decreased by 33%, and the number of types B and C grade both increased by 1. The Rosu 100 group showed one case of slight blood leakage into the aneurysm cavity after treatment, which had been rated as type B grade after the operation. At the 12-month follow-up, the aneurysm with the original type B grade was no longer developed, and all three aneurysms were cured, reaching type A grade (Fig. 5B). We further analyzed the changes in the parent artery after stent implantation and found that in both the short-and long-term follow-ups, there was no obvious stenosis in the parent arteries (Fig. 5C). In the toxicity experiment, no in ammatory reaction could be seen in H&E staining at either 1 or 3 months. (Fig. 6C).

PDGF-BB-induced synthetic SMC model and viability of rosuvastatin-treated SMCs
Figure S1A demonstrates that 10 ng/mL PDGF-BB can signi cantly increase SMC viability in 24 h (p < 0.0001). Meanwhile, the proliferation activity of SMCs increased with increasing rosuvastatin concentration. However, at 1000 ng/mL, the proliferation activity decreased, suggesting adverse reactions. Figure S1B shows the results for the rosuvastatin treatment of contractile SMCs. Clearly, although contractile-SMC proliferation was not signi cantly inhibited at lower rosuvastatin concentrations (p > 0.05), it was signi cantly inhibited at 100 µM (p < 0.0001). To determine the effective inhibitory rosuvastatin concentration for PDGF-BB, we treated SMCs with 10 ng/mL PDGF-BB and varied rosuvastatin concentration for 24 h. The CCK-8 assay showed that 5 µM rosuvastatin inhibited the PDGF-BB-induced contractile-SMC proliferation (p < 0.05). The inhibitory effect was signi cant (p < 0.0001) at 10 µM and strengthened with increasing rosuvastatin concentration ( Figure S1C). Flow cytometry results showed that 10 µM rosuvastatin calcium did not cause abnormal apoptosis in contractile SMCs ( Figure  S2). Therefore, we used 10 µM rosuvastatin calcium for the subsequent experiments.

SMC morphology, viability, and function
Phalloidin staining revealed the cell morphology. In the PBS and Rosu groups, the cells remained spindleshaped, while the cells in the PDGF group appeared atter and triangular. However, in the PDGF + Rosu group, because the ability of the PDGF-BB to change the SMC morphology was signi cantly inhibited, the cell morphology was more spindle-like (Fig. 7A).
EdU staining revealed that after stimulation with 10 ng/mL PDGF-BB for 24 h, the number of SMCs labeled with EdU increased signi cantly (green, p < 0.0001). When the cells were cotreated with rosuvastatin, the number of SMCs labeled with EdU decreased, which was signi cantly different from the trend for the PDGF group (p < 0.0001) (Fig. 7B).
In the scratch test, after 24 h of intervention, 67% of the remaining area in the control group remained unhealed while only 53% in the PDGF group did, which was signi cantly different from the control group (p < 0.001). Although the residual area in the PDGF + Rosu group was 66%, which was signi cantly different from that in the PDGF group (p < 0.01), the residual area in the Rosu group was not notably different from that in the control group (Fig. 7C).
In the transwell assay, although the cell penetrabilities of the control and Rosu groups were weak compared with that of the control group, the number of cells that had penetrated the PDGF group increased signi cantly (p < 0.01). Although SMC migrability was substantially improved after PDGF-BB stimulation, the number of penetrating cells decreased after treatment with rosuvastatin, which was remarkably different from the trend in the PDGF group (Fig. 7D).

Effects of rosuvastatin on SMC phenotype and in ammatory factors
The PCR results showed that the expression of contractile phenotypic markers α-SMA and SM22-α was lower in the PDGF group than in the control group. After rosuvastatin treatment, the expression of α-SMA and SM22-α both increased; therefore, OPN was used for reverse veri cation. OPN was originally upregulated in the PDGF group and showed obvious attenuation after the rosuvastatin treatment. The expression of TNF-α, MCP-1, MMP-2, and MMP-9 also decreased substantially with the addition of rosuvastatin (Fig. 8A) The protein expression levels of phenotype markers and in ammatory factors in PDGF-BB-induced SMCs were detected by western blotting. The results showed that SM22-α and OPN were notably downregulated and upregulated after PDGF-BB stimulation, respectively, and that the protein expression of MMP-9 was also notably increased after PDGF-BB stimulation. With the addition of rosuvastatin, the SM22-α and OPN expressions increased and decreased, respectively (Fig. 8B).
The in ammatory factors secreted by cells can be detected using the MILLIPLEX® Map. After PDGF-BB stimulation, the in ammatory factors increased, and the secretion of IL-1β was up to 4 times that of the control group. Compared with the PDGF group, the ability of the rosuvastatin-treated SMCs to secrete in ammatory factors decreased, and there was a signi cant difference compared with the PDGF-BB group (Fig. 8C).

Discussion
To prevent stent stenosis and thrombosis, drugs loaded on covered stent grafts should be able to regulate cell proliferation, in ammatory reactions, and thrombosis. After stent implantation, effective anticoagulation and inhibition of platelet adhesion are the rst steps. Heparin plays a major role in anticoagulation therapy. Although the heparin coating reduced the thrombosis rate in animal experiments (22,23), it could not effectively improve late vascular patency and neointimal hyperplasia (24), suggesting that promoting the proliferation ability of ECs is crucial. In our previous study, we found that rosuvastatin-and heparin-coated stents could effectively promote early endothelialization and provide a basis for reducing long-term complications (25). Early endothelialization is the key to preventing in-stent stenosis and late thrombosis (26,27). Willis covered stents are widely used in the treatment of complex aneurysms. However, although ePTFE reduces the incidence of early stent-graft stenosis, it is also prone to delayed reendothelialization (28). Nondegradable polymers that remain in arteries can continue to cause local in ammatory reactions (29). In patients treated with PTFE-covered stents, the incidence of nonfatal myocardial infarction is higher than that in patients treated with a bare metal stent (30). Although PLLA is a biodegradable synthetic polymer, it shows poor exibility. Blending PLLA with a moreelastic polymer is an effective method of improving its mechanical properties (31,32). Previous studies have shown that the P(LLA-CL) nano ber membrane shows good EC adhesion (33); therefore, we chose P(LLA-CL) to lay the foundation for the development of the nano ber scaffolds. Because the cover mats of the Willis stent graft are loosely sutured to the stent (34), the nano bers may be damaged during placement (35). Studies have con rmed that nano ber scaffolds prepared by coaxial electrospinning are less likely to be damaged, and coaxial electrospinning can not only load drugs evenly on polymers but also facilitate drug release. Furthermore, coaxial electrospinning can prepare shell-core structures (33,36), allowing for polymer drug loading, which is bene cial to the combination of drugs and polymers. Uniform drug loading can provide a good foundation for drug release, and sustained drug release can provide a stable therapeutic effect. In addition, in cerebral arteries, the stent-graft must be miniaturized and show high exibility and maneuverability to smoothly navigate to the target artery (37). Therefore, we used an Apollo stent made of 316L stainless steel, which shows su cient supporting force, and its special spatial structure can provide effective toughness when the stent expands, which is critical because excessive pressure during balloon expansion can lead to thromboembolism, vascular dissection, or rupture (38). The expansion pressure of a bare Apollo stent is only 6 bar. When the nano ber was spun for 3 min, the membrane did not break or fall off at 6 bar, indicating that a 3-min-thick membrane is more suitable so that the stent can be completely opened under lower pressures without affecting the pressure of the Apollo stent itself. Therefore, we adopted the balloon catheter expandable stent and set the membrane thickness at 3 min.
The elastase-induced rabbit aneurysm is a mature aneurysm model (39). Different animal models have different research objectives. In the cardiovascular system, the porcine coronary artery model can be used to evaluate the risk of stent stenosis. The tendency of thrombosis and neointimal formation in the porcine coronary system is similar to that in humans, which is especially helpful in evaluating arterial stenosis after stent implantation (40,41). However, the porcine coronary artery is unsuitable for simulating the environment of small arteries. In addition to porcine and rodent models, the atherosclerosis model of New Zealand white rabbits can also be used to evaluate stent stenosis. Stents were implanted in New Zealand white rabbits fed a high-fat diet to evaluate in-stent stenosis. Compared with the porcine model, the in-stent stenosis model of New Zealand white rabbits is more suitable for observing the drug mechanism (42,43). The use of a rabbit CCA model to evaluate in-stent stenosis is also an emerging method, and studies have found that it is not different from traditional methods (44).
The traditional ow-diverter device shows only 56% occlusion at the 3-month follow-up, and it takes 12 months to reach 95% occlusion (45). In our study, the aneurysm was immediately excluded from the blood ow after the covered stent was placed, and the aneurysm was no longer visible. In addition, studies have shown that if there is a mild endoleak after the covered stent has been implanted, leakage can eventually occluded (35). In our long-term follow-up, a type C aneurysm sprang a leak, and the aneurysm was still visible after stent implantation. After treatment with the Rosu 100-covered stent for 12 months, the aneurysm was no longer visible. SEM observation showed that the endothelial coverage of the aneurysm neck was intact. However, in the PBS group that was originally satisfactory after the treatment, only one case did not leak or recanalize at the 12-month follow-up, suggesting that the heparin and rosuvastatin loads showed a certain therapeutic effect and a stable aneurysm cure rate.
After vascular injury, SMCs can transform from a resting contractile type to a pro-in ammatory, dedifferentiated one. The phenotypic regulation of VSMCs, which describes SMC differentiation, was rst conceptualized by Chamley-Campbell et al. (46). When the differentiated type is modulated to the dedifferentiated one, in ammatory cytokines are produced and in ammatory cell markers are expressed (47,48), involving the participation of multiple cytokines (49). PDGF-BB is one of the key factors affecting the phenotypic modulation of SMCs (50,51), which promotes the SMC morphology to change from spindle-shaped to at triangular or oblique squares (52). When PDGF-BB is inhibited, the concentration of contractile-type markers α-SMA and SM-22α begin to increase (53). Studies have found that rosuvastatin has pleiotropic effects and can reduce SMC phenotypic modulation (54). In our study, PDGF-BB was used to construct a model of in ammatory SMCs to simulate the cytokines released by endothelial injury and the affected SMC phenotype modulation. We found that PDGF-BB stimulated SMC proliferation and migration. Once the PDGF-BB concentration increased, the expression of OPN was substantially increased, and the expression of phenotypic marker molecules of contractile SMCs was reduced, suggesting that the cytokines released after endothelial damage could regulate the SMC phenotype. On the nano ber mats, we observed that the increase in rosuvastatin concentration reduced the number of PDGF-BB-stimulated SMCs and gradually restored their shape to spindle-like. Our results con rmed that the rosuvastatin-loaded nano ber functioned by inhibiting the abnormal proliferation of SMCs. Studies have found that statins can inhibit the transduction of PDGF-BB on SMCs by blocking the G0/G1 cell cycle and PDGFRβ-Akt signaling cascade (55), thereby inhibiting the pathological proliferation and migration of SMCs (56). Simvastatin can reduce the secretion and mRNA expression of PDGF/IL-1induced MMP-9 and reduce MMP-9 secretion in vascular smooth-muscle cells by inhibiting the RhoA/ROCK signaling pathway (57), thereby reducing the transduction of the SMC in ammatory response. After we constructed a PDGF-BB cell model, rosuvastatin was used to treat SMCs, and we found that rosuvastatin could inhibit PDGF-BB to promote the proliferation and migration of in ammatory SMCs. Analysis of the secreted supernatant revealed that the expression of in ammatory factors in rosuvastatin-treated SMCs decreased, which con rmed that rosuvastatin calcium can effectively regulate the phenotype of smooth-muscle cells.

Conclusions
This study explored the application of a nano ber-covered stent in vivo and its mechanism in vitro. In vitro experiments demonstrated that rosuvastatin can inhibit the proliferation and migration of in ammatory SMCs and inhibit the PDGF-BB-induction of SMCs. In-vivo studies revealed that the covered stent did not cause persistent in ammatory reactions in tissues. The therapeutic effect of the Rosu 100covered stent was ideal for treating rabbit RCCA aneurysms. The results suggest that drug-loaded covered stents may reduce the risk of late in-stent thrombosis and stenosis.

Nano ber diameter measurement
Four groups of sterilized nano ber scaffolds were soaked in Dulbecco's Modi ed Eagle Medium (DMEM, Hyclone,Utah) for 24 h in triplicate and were observed using scanning electron microscopy (SEM; Phenom XL, the Netherlands). The diameters of 100 bers, as shown in the SEM images, were measured and recorded.

Characterization of nano ber scaffolds
PBS, Rosu 50, Rosu 75, and Rosu 100 were fumigated with 75% alcohol for 3 d and then sterilized by ultraviolet radiation for 30 min to prepare sterile scaffold mats. SMCs were stimulated with 10 ng/mL platelet-derived growth factor-BB (PDGF-BB; PeproTech, USA) for 24 h before seeding. SMCs were seeded into four groups of nano ber mats at a density of 2 × 10 4 cells/well in triplicate. After 24 and 48 h, the intervention was terminated, glutaraldehyde was added to x the cells, and the samples were observed by SEM. Three regions of each sample were selected for observation.

Hoechst staining of cells attached to nano ber mats
SMCs were stimulated with 10 ng/mL PDGF-BB for 24 h and were seeded on nano ber mats at a density of 2 × 10 4 cells/well in triplicate and then cultured for 48 h. After xing the cells with 4% paraformaldehyde (PFA; Sinopharm, China) overnight, nano ber mats were stained with Hoechst 33342 (Beyotime, China) for 5 min, following the manufacturer's instructions. Samples were photographed using a confocal uorescence microscope (Carl Zeiss, Germany). The Hoechst-33342-labeled nuclei were stained blue.

Analysis of attached-SMC viability
SMCs were pretreated with 10 ng/mL PDGF-BB for 24 h and seeded into 4 groups of nano ber mats at a density of 10 4 /well in triplicate. A cell counting Kit-8 assay (CCK-8; Dojindo, Japan) was conducted according to the manufacturer's instructions to study the viability of the cells on the scaffold mats. The seeded mats were cultured for either 24 or 48 h, 500 µL of CCK-8 staining solution was then added to each well, and the mats were incubated for 2-3 h. The absorbance at 450 nm was measured with a microspectrophotometer (ThermoFisher Scienti c™, Waltham, Massachusetts).

Balloon-expansion experiment
The covered stent was connected to a balloon catheter (MicroPort Co., Ltd. Shanghai, China) and was gradually lled with 0-10 bar (0-1000 kPa) of air for expansion. The outer diameters of the proximal, distal, and middle segments of the covered stent were measured and statistically graphed using GraphPad Prism 8.0 software.

Development of rabbit aneurysm model and stent implantation
The New Zealand white rabbit common-carotid-aneurysm model has been explained in detail in our previous paper (58). Thirty (30) d after modeling, which was su cient for the aneurysm to mature, digital subtraction angiography (DSA) was used to observe the aneurysm formation. Covered stents were placed into the right subclavian artery and were maintained at equal lengths on either side of the aneurysm ostium. According to the results of the in vitro assay, Rosu 100 and PBS were used in the experiment and control groups, respectively. The stents were randomly selected from the PBS and Rosu 100 groups.
Immediately after the procedure, DSA was used to evaluate blood ow in the aneurysm. Based on the DSA characteristics, therapeutic e cacy was divided into 3 grades, as shown in Table 1.

Rabbit aneurysm short-and long-term follow-ups
The treated animals were administered aspirin (20 mg/d) 7 d prior to and 14 d after stenting. Therapeutic e cacy was evaluated with DSA at 3 and 12 months after stenting and was divided into 3 grades. Stent grafts were collected at 3 and 12 months after stent implantation, and SEM and histology were used to observe the endothelialization of the parent artery.

Histology
Specimens were xed with 4% PFA overnight and dehydrated with gradient alcohol (75, 85, 90, 95, and 100%) for 24 h. Then, the specimens were soaked in xylene for 4 h and treated with curing monomers I, II, and III for 24 h, after which the specimens were encased in monomers 1-1.5 cm above the tops of the tissue specimens themselves by adding the curing monomer and using a gas pump in a dryer to prevent bubble formation. The tissue specimens were deposited at 4°C for one week, further deposited at room temperature until the monomer thickened, and then transferred into an oven (37°C) until the monomer hardened. After the methyl methacrylate (MMA) was completely hardened, a LeicaSP600 hard tissue slicer (Leica) was used to prepare 100-µm-thick tissue sections, which were further polished to 50 µm, sealed, and repolished. We observed the sections and calculated the coverage ratio for all the stents in each section. A t-test was used for statistical analysis.

Toxicity of nano ber mats
The nano ber mats were implanted under the abdominal skin of C57 male mice. Tissue specimens were collected 1 and 3 months after implantation and were stained with hematoxylin and eosin (H&E) to observe the in ammatory reaction of the mats on the tissue.

Viabilities of PDGF-BB-induced in ammatory, rosuvastatin-treated, and rosuvastatin-treated in ammatory SMCs
PDGF-BB was used to establish an in ammatory-SMC model. PDGF-BB was dissolved in sterilized H 2 O and diluted with DMEM to 0, 1, 10, 20, 50, 100, and 1000 ng/mL. Rat aortic SMCs were seeded onto 96well plates at a density of 3 × 10 3 cells/well in triplicate and cultured in DMEM supplemented with 10% fetal bovine serum (FBS; Epizyme, China). When the cells reached 50-60% con uence, SMCs were incubated in the serum-starvation condition for 24 h. Cell viability was measured using the CCK-8 assay.
The absorbance was measured at 450 nm using a spectrometer.

Cell proliferation assay
SMCs were seeded into 24-well plates at a density of 2 × 10 4 /well in triplicate. The cells were starved prior to the intervention and were divided into groups as follows: the control and SMCs stimulated with 10 ng/mL PDGF-BB (PDGF), treated with 10 µM Rosuvastatin (Rosu), and cotreated with 10 ng/mL PDGF-BB and 10 µM rosuvastatin (PDGF + Rosu). Each group was cultured for 24 h, after which an EdU cell proliferation detection kit (Beyotime, China) was used to evaluate the proliferation rate, according to the manufacturer's instructions. The cells were counterstained with Hoechst 33342 (Beyotime, China) for 10 min and then observed with a focused uorescence microscope. Hoechst 33342 and EdU showed blue and green uorescences, respectively. ImageJ software used to count the Hoechst-33342-and EdUlabeled cells, and the EdU/Hoechst-33342 ratio was then calculated.

Scratch test
Rat aortic SMCs were seeded into six-well plates at a density of 1 × 10 5 /well in triplicate. When reaching 80% density, the cells were cultured in serum-free DMEM for 24 h. Pipette tips were used to scratch the cell monolayer across the center of each well. The detached cells were washed away with sterile PBS.
The cells were divided into control, PDGF, Rosu, and PDGF + Rosu groups. Each group was cultured for 24 h. Three images were photographed for each well and were processed with ImageJ software. The areas in the images where cells remained after 24 h were compared.

Transwell tests
SMCs were seeded into a 12-well plate at a density of 5 × 10 4 /well in triplicate. The cells were starved prior to the intervention. The cells were then digested and seeded into 8-µm-pore transwell chambers

RNA extraction and real-time polymerase chain reaction
The cellular ribonucleic acid (RNA) was isolated using an RNA puri cation kit (Yishan, China) according to the manufacturer's instructions. The RNA integrity was quanti ed using a NanoDrop™ 1000 spectrophotometer (ThermoFisher Scienti c™, UT, USA). The reverse transcription (RT) reaction was performed using a fast all-in-one RT kit (Yishan, China) according to the manufacturer's instructions. Real-time polymerase chain reaction (PCR) was performed using the Hieff® qPCR SYBR® Green master mix (Yeasen, China) and was detected with a real-time PCR system (7900HT, ABI). No nonspeci c ampli cation was observed based on the dissociation curve. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH; Sangon biotech, China) was used as an internal control. The data were further analyzed by the comparison Ct (2 −ΔΔCT ) method and were expressed as a fold change relative to the respective control.
The sequences used for the qPCR primers are listed in Table S1.

Western blot
Proteins were lysed from the rat aortic SMCs after the intervention. The cells were divided into 4 groups, as described in above experiments, and 40 µg of protein per lane was separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; Epizyme, China). The proteins were electrotransferred onto a polyvinylidene di uoride (PVDF; Millipore Sigma, Billerica, MA, USA) membrane and blocked with a western quick-blocking buffer (Beyotime, China) for 15 min at room temperature. The blocked PVDF membranes were incubated overnight with primary antibodies including goat anti-SM22α (1:500 dilution; Abcam, England), rabbit anti-OPN (1:1000 dilution; Abcam, England), and rabbit anti-MMP9 (1:1000 dilution; Abcam, England). After 12-14 h, the PVDF membranes were incubated with a horseradish-peroxidase-conjugated secondary antibody(Huabio, Chia) for 1 h at room temperature. The immunoblots were probed using an enhanced chemiluminescence (ECL; Thermo, Rockford, IL, USA) substrate. An imaging system (Bio-Rad, Hercules, CA, USA) was used to detect the blots, and the chemiluminescence levels were recorded. The results were normalized to that of GAPDH, and the experiments were replicated three times.