Poly (l–lactide-co-ε-caprolactone) (PLCL (50:50), with 50 mol% l-lactide, M_W: 330,000 Da) was provided by Jinan Daigang (Jinan, China). 1,1,1,3,3,3-Hexafluoro-2-isopropanol (HFIP) was obtained from Shanghai Macklin Biochemical Technology (Shanghai, China). 3-(4,5-dimethylthiazole-2-yl)-2,5-dibenzotriazole ammonium bromide (MTT) was acquired from Beijing Biotopped Technology (Beijing, China). A Cell Counting Kit-8 (CCK-8) was purchased from Aladdin Reagents (Shanghai, China). Sodium tanshinone IIA sulfonate (STS) was provided by Dalian Meilun Biotechnology (Dalian, China). Poly-ethylene oxide (PEO, M_W: 100,000 Da) was purchased from Shanghai Macklin Biochemical Technology (Shanghai, China). All reagents used for cell culture were purchased from Gibco Life Technologies (Shanghai, China) unless stated otherwise. All cells were purchased from iCell (SAIBAIKANG, Shanghai, China).

PLCL was dissolved in HFIP at a concentration of 10% (W/V) under room temperature and stirred for 24 h (84-1, Shanghai Meiyingpu Instrumentation Manufacturing Co., Ltd. Shanghai, China) to obtain core–shell structured fibers, with PLCL as the shell and PEO as the core. In this study, to obtain more core–shell structured fibers and to determine the physical properties of these fibers, PEO [6%, (w/v)], as a temporary substitute for various water-soluble functional substances, was used as the core material, while specific active ingredients (STS) will be added during the biological testing. For coaxial electrospinning, the PLCL solution was loaded into a 20-mL syringe with a 15-gauge needle at an injection speed of 1.3 mL h⁻¹, and the PEO solution was loaded into a 10-mL syringe with a 19-gauge needle at an injection speed of .8 mL h⁻¹. The needle tip was subjected to positive voltage at 11.5 kV and negative voltage at 2.5 kV (TK129, Shanghai Taco Company, Shanghai, China). By adjusting the fiber collector roller speed, random and oriented fiber membranes and tube grafts were fabricated.

Fiber morphology was observed using scanning electron microscopy (SEM, Apreo S, Thermo Fisher Scientific Technologies, Massachusetts, US), and the fiber diameter distribution was analyzed with Image-J software. The internal structure of the coaxial electrospinning fibers was observed through transmission electron microscopy (TEM, Talos F200X, Thermo Fisher Scientific Technologies, Massachusetts, US).

For mechanical property testing, the tensile strength of the fiber membranes was tested, and the suture retention strength of the tube grafts also was measured by means of a universal tensile extensibility instrument (UTM2503, Shenzhen Sansi Zongheng Technology Co., Ltd., Shenzhen, China). The fiber membranes were cut into dumbbell-shaped samples with an effective width of 4 mm by punching, and the thickness of each sample was measured. Before testing, the two ends of the sample were clamped, a sensor with a load capacity of 50 N was used, and the tensile extensibility instrument—with a fixture spacing of 3.5 cm—was stretched longitudinally until it ruptured at a speed of 5 mm min⁻¹. The prepared tube grafts were cut into 1-cm sections, and then the thickness of graft well was measured. Before testing, one end of the sample was clamped to one arm of the equipment and the other end, with 5–0 polyester suture (Shanghai Pudong Jinhuan Medical Products Co., Ltd., Shanghai, China) placed 2 mm from the edge, was clamped to the other arm of the equipment at a speed of 120 mm min⁻¹ until rupture (UTM2503, Shenzhen Sansi Zongheng Technology Co., Ltd. Shenzhen, China).

Fiber membranes were set into glass slides, and to test the hydrophilicity, a video optical contact angle measuring instrument (DSA30, Kruss, Germany) was applied. A total of 2 μL of water was dropped on a flat place of the fiber membrane to avoid air bubble formation when placing samples on glass slides, and then the video, image, and contact angle data were recorded using the instrument. After the test, the data were compiled and plotted to analyze the hydrophilicity of the membranes, which affects cell attachment and platelet adhesion.

The prepared samples were first sterilized and then placed in a 5-mL EP tube. A total of 3 mL PBS and 60 μL antibiotic-antimycotic solution were then added to the EP tube, after which they were placed into a constant temperature shaker (ZH-S) at 37°C to simulate the in vivo environment. Every week, all of the supernatant was removed and replaced with an equal volume of fresh medium, and the pH value of the supernatant was measured using a pH meter (FE28, Shanghai Mettler Toledo Instrument Co., Ltd., Shanghai, China). Every month, the samples were removed and freeze-dried (SCIENTZ-18N, Ningbo Xinzhi Biotechnology Co., Ltd., Ningbo, China), and then the samples were weighed with an electronic balance (BSA124S, Sartorius Scientific Instruments (Beijing) Co., Ltd.). The fiber membrane microstructure was also then observed using SEM (Apreo S, Thermo Fisher Scientific Technologies, Massachusetts, US).

In the blood compatibility experiment, two methods were applied: a lactate dehydrogenase (LDH) activity assay and SEM morphology observation (S-4800, Japanese High-tech Co., Ltd.). Platelets obtained after centrifugation were used for LDH detection, and rabbit whole blood was used for SEM observation. First, the prepared sample was placed into a 24-well plate. For LDH testing, rabbit blood (purchased by Nanjing Semberga Biotechnology Co., Ltd., Nanjing, China) was centrifuged (TGL-15B, Shanghai Anting Scientific Instrument Factory, Shanghai, China) at 2,500 rpm for 10 min to obtain platelet rich plasma (prp). After that, 500 μL prp was added to each sample, which were then shaken at 37°C for 3 h, and then the samples were washed with PBS, after which 1 mL of 2% Triton was added to each. After shaking for 20 min, we extracted the solution and then centrifuged it for 10 min, taking the supernatant for further testing with an LDH release kit according to the manufacturer’s instructions. We recorded the absorbance of the reaction solution at 450 nm on a fully automatic microplate reader (DNM-9606, Beijing Pulun Xin Technology Co., Ltd. Beijing, China) for the relative quantification of platelets.

For SEM observation, 1 mL of rabbit blood was added to each sample and shaken at 37°C for 2 h. After that, the samples were washed gently with sterile water, and then 1 mL of 2.5% glutaraldehyde solution was added to each well and allowed to sit for 1 h, following dehydration with a gradient alcohol solution (30%, 50%, 70%, 80%, 90%, 95%, and 100%). Finally, the processed sample was placed in a fume hood and blow-dried, and was then immediately sprayed with gold and was observed using a scanning electron microscope.

The vacuum-dried functional fiber membranes were punched to the same size discs and placed into a 24-well plate, and then cell experiments were performed for 1, 3, and 5 days. Samples were first sterilized under alcohol vapor for 4 h, followed by rinsing with PBS twice, and mouse fibroblasts (3T3) were then seeded at a concentration of 1×10⁴ cells•well⁻¹. At the corresponding setting time point, the medium was extracted, and 40 μL MTT and 360 μL DMEM medium were then added, incubating with the samples for 4 h. After that, the supernatant was aspirated by adding 200 μL DMSO to dissolve the sediment for 20 min in the shaker. At last, 100 μL supernatant was placed into a 96-well plate for further analysis with a microplate reader at 492 nm. Absorbance was found to have a good linear relationship with the number of viable cells.

The cell seeding procedures for human umbilical artery smooth muscle cells (HUASMCs) and macrophages (RAW264.7) were same as those for the 3T3 cells mentioned above, with the only difference being the number of RAW264.7 were seeded (8×10³ cells•well⁻¹). At the specific time point, 40 μL of CCK-8 was added to every well, and then allowed to incubate for 4 h. Finally, the supernatant (100 μL) was placed into 96-well plate for further analysis with a microplate reader at 450 nm.

After cells were cultured on different fiber membranes for 5 days, the cell morphologies were observed using a laser scanning confocal microscope. The cell membranes were washed with PBS. A total of 200 μL of 2.5% glutaraldehyde fixation was then added, and after 1 h the cells were washed 3 times with PBS, after which the cell membranes were treated with .1% Triton X-100 (PBS) for 3 min and then washed again with PBS. Then cell-membranes were blocked with 1% BSA at 37°C for 80 min, followed by washing with PBS. Finally, 200 μL of the pre-prepared Rhodamine B and DAPI solution were added to each well. After 30 min of incubation, the cell-membrane was washed in preparation for forward laser scanning confocal microscopy observation (LSM 880 with fast AiryScan, Carl Zeiss AG, Dresden, Germany).

All quantitative results were obtained at least three samples. Statistical ANOVA analysis was performed in Origin Pro 2018 software. Results were expressed as the mean ± standard deviation (SD), with significant differences at *p < .05, **p < .01, ***p < .001.

Scanning electron microscopy (SEM) images of the random- and oriented-fiber membranes are shown in Figure 1A. The images clearly show the differences between the random and oriented fibers; Figure 1B shows that the diameters of random fibers were about 1.79 ± 1.03 μm, the thickness was relatively uniform, and the distribution was random in all directions. The diameters of the oriented fibers were about 2.14 ± 1.40 μm, which were slightly larger than those of the random fibers, and most of the fibers were arranged in the same direction. Overall, all the fibers were connected without any beads or gross defects. The different arrangements of the fibers likely provided different physical, mechanical, and biocompatibility properties (Dehghan and Khajeh Mehrizi, 2022; Ozdemir et al., 2022; Yuan et al., 2022). The subsequent experiments were conducted with random- and oriented-fiber membranes, and it was expected that the oriented-fiber membranes would be more helpful to the growth and proliferation of cells and promote the regeneration of new tissue.

Figure 1C shows the transmission electron microscopy (TEM) morphology of the coaxial electrospun fiber. It is clear that the core–shell structure was successfully obtained, demonstrating that the PEO was successfully wrapped into the PLCL by coaxial electrospinning. Thus, the functional composition could be loaded inside or outside the fibers, which could improve drug delivery systems with sustained release, enhance the hydrophilicity and biocompatibility, likely reduce the rejection reaction between the fiber graft and human tissue, and improve the cell attachment. In addition, the oriented-fiber membranes containing STS are expected to be able not only to provide sustained release of STS over time and prevent platelet adhesion but also to induce cells to arrange orderly.

The excellent characteristics of PLCL provide it with great potential to be applied for vascular grafts, and the mechanical properties are among the most important properties for grafts. The tensile strengths of fiber membranes and the suture retention of tube grafts were tested. Figure 2A shows that the maximum stress of the random-fiber membranes was nearly 4 MPa, while that for the oriented-fiber membranes parallel to the fiber direction could reach more than 11 MPa, which was much higher than that of the random-fiber membranes. The corresponding strain for oriented-fiber membranes was 190%, which was lower than that of the random-fiber membranes (470%). This phenomenon may be explained as the orientation fibers being stretched during the spinning process such that the orientation stress was enhanced in the parallel direction. The stress for the oriented-fiber membranes perpendicular to the fiber direction was as low as .8 MPa. This was because the fibers with high order had a lower tangle degree between the fibers than that of the knotted random fibers. According to the literature, the stress range of the human aorta is .2–1.6 MPa (Famaey et al., 2014; Backman et al., 2017; Maleki and Kalantarnia, 2022). The stresses of the fiber membranes were much higher than the stress requirements of the human aorta, indicating that they met the mechanical strength requirements for implantation in the body. Therefore, the excellent mechanical properties of the fiber membranes could be applied for blood vessel grafts and good mechanical support promote the smooth flow of blood in vascular grafts, without causing an aneurysm.

The suture retention strength is very important for vascular grafts, and it should be sufficient for use in a surgical procedure. The result for suture retention testing is shown in Figure 2B. It was found that the suture retention strength of the oriented-fiber tube grafts was higher than that of random-fiber tube grafts. Overall, both tube grafts had good suture retention strengths of 5 N or higher. Based on study (Yin et al., 2020b), the suture force of animal autologous vascular tissue is 1.7 N. Therefore, both materials were completely comparable to animal autologous vascular tissue in terms of mechanical strength, which could ensure that rupture and bleeding would not occur when implanted in the body, and they could be safely used in the body.

According to previous research (Ehtesabi, 2021; Manivasagam and Ketul, 2021; Al-Azzam and Alazzam, 2022), the graft surface hydrophilicity is closely related to the adhesion and cohesion of platelets, and the hydrophilicity can also affect the migration and proliferation of cells. In this study, whether the fiber structure with random and oriented fibers affected the hydrophobicity, hemocompatibility, and cell attachment ability were explored in follow-up testing. The hydrophilicity of the random- and oriented-fiber membranes were measured by contact angle testing, and results are shown in Figure 3. The contact angles of the random-fiber membranes were larger than those of the oriented-fiber membrane. The main reason was that the distribution of random fibers was chaotic and staggered, and water molecules could not easily penetrate the material. In contrast, the oriented-fiber membranes obtained smaller contact angles due to the high degree of fiber order. When the water molecules contacted the fibers, they quickly penetrate through the fiber gaps. According to the data, the membrane contact angles were smaller after 15 s than after 5 s, and water molecules could slowly penetrate the fibers over time. Therefore, a hydrophobic fiber membrane could be shifted to relatively hydrophilicity through microstructural changes, which is likely induce cell migration, growth, and proliferation.

The degradation behaviors of the membranes are closely related to the membranes’ physical and chemical properties, and appropriate degradation is important, especially for implanted grafts, which could match the tissue regeneration rate (Wang et al., 2020; Alamán-Díez et al., 2022). In this experiment, the pH and mass changes of the membranes at various time points were examined, and the microscopic morphological changes in the membranes were observed every month. Figure 4A shows the pH changes in the random- and oriented-fiber membranes. The pH fluctuated slightly between 6.5 and 7.5, and thus, it remained basically stable. The pH of the oriented-fiber membranes fluctuated between 6 and 7.5, and the fluctuation range interval was larger than that of the random-fiber membranes. However, neither membrane type had strong acidity in the degradation process, which indicated that inflammatory responses during the fiber membrane grafts implanted in the body would be reduced. In this study, the PEO loaded inside the fibers released into aqueous solutions as the fibers degraded, and they always remained neutral or weakly basic in the aqueous solutions. This caused the pH to undergo only slight changes (Hu, 2000). Figure 4B shows the mass changes of the random- and oriented-fiber membranes during in vitro simulated degradation. The mass decreased rapidly in the first month, and it slowly declined after the first month. Overall, the mass degradation rates of the random-fiber membranes were slower than those of the oriented-fiber membranes. After 6 months, the oriented-fiber membranes had mass losses of more than 40%, while the random-fiber membranes had mass losses of about 25%. Even in the ninth month, the random-fiber membranes retained 72% of their masses.

Figure 4C shows that the fiber structures of the random-fiber membranes remained intact in the first 6 months. In the ninth month, a partial area of fiber fracture occurred, and the fiber membranes ruptured or cracked. As shown in Figure 4B, the masses were maintained at about 72% when the random-fiber membranes were broken. While the degradation rates of the oriented-fiber membranes were relatively fast, the fiber form was intact in the third month. Fiber rupture occurred in the sixth month, and the fiber morphology was not clear. As shown in Figure 4B, the masses remaining were less than 60% when the oriented-fiber membranes were broken.

The phenomenon can be explained by the results of the contact angle tests. The contact angles of the oriented-fiber membranes were smaller than those of random-fiber membranes, indicating the oriented-fiber membranes had better hydrophilicity. Thus, water molecules could more easily penetrate through the fiber gaps for the oriented-fiber membranes, and stronger interactions between the fibers and water molecules would accelerate the degradation rate. However, it can be seen from Figure 4A that the pH values of the oriented-fiber membranes were slightly lower than those of the random-fiber membranes in degradation process, indicating that weakly acidic environment also promoted the degradation of the membranes. For the random-fiber membranes, the fibers were intertwined, and the diameters of fibers were relatively evenly distributed. As a result, water had more difficulty penetrating to the insides of membranes, and therefore, the degradation for the random-fiber membranes was slow. If both random- and oriented-fiber membranes could maintain sufficient mechanical support in the early stage when implanted in vivo, there would be sufficient time to ensure new tissue regeneration. According to the literature (De Valence et al., 2012; Zhou et al., 2014; Wang et al., 2017; Sevostyanova et al., 2018; Yin et al., 2020b; Wen et al., 2020; Yang et al., 2021), when a graft was implanted in vivo, the endothelial cells gradually formed an intima, and smooth muscle cells and fibroblasts were also arranged along the axial direction of the graft, thus forming new blood vessel tissue. After 3 months, the vascular development was almost complete. Thus, the degradation rate of the oriented-fiber graft was likely suitable for the process of new tissue regeneration, while the random-fiber graft with a low degradation rate may cause the material to reside in the tissue too long, which would impede new tissue reconstitution.

The blood compatibility of the functional fiber graft is important, and the blood must flow smoothly when the grafts are implanted in the body. In this experiment, platelet adhesion on different membranes was analyzed by the LDH activity assay and SEM. The results in Figure 5A show that when platelet adhesion was measured by the LDH kit, the absorption values for the random-fiber membranes were slightly smaller than those of the oriented-fiber membranes. The data showed that the oriented-fiber membranes containing STS had a good antiplatelet adhesion effect, with lower absorbance than the others, and the effect was significantly different from those of the oriented-fiber membranes. The excellent antiplatelet effect likely was caused by STS, which possessed a good antithrombotic effect (Zhou et al., 2019; Li et al., 2020). Through the results of SEM observations in Figure 5B, the random- and oriented-fiber membranes without STS had more platelets, while there were fewer platelets on the oriented-fiber membranes containing STS. Overall, the platelets on the random fibers were relatively dispersed, and more plasma composite was coated on the random-fiber membranes. Platelets on the oriented-fiber membranes without STS could easily accumulate in the gaps of fibers, especially those that were highly ordered, which was mutually corroborated with the LDH result. Therefore, the oriented-fiber membranes containing STS showed good prospects for application for cardiovascular tissue repair with an antiplatelet adhesion effect.

Cytotoxicity tests can confirm whether materials can be safely in contact with tissue, and materials with low cytotoxicity can reduce the risks during implantation in the body. In this study, we analyzed the toxicities of the membranes through the MTT assay, and cell morphology was observed by laser scanning confocal microscopy. Figure 6A clearly demonstrates that 3T3 could continue to grow and proliferate on all kind of membranes, and the absorbance values for the orientated-fiber membranes exceeded those of the control slightly. Therefore, all these membranes had good cellular biocompatibility. From Figure 6B, it can be seen that 3T3 could be arranged in one direction on oriented-fiber membranes with or without STS.

Cell compatibility of RAW264.7 and HUASMCs on materials was tested via the CCK-8 assay. From Figure 7A, it can be seen that the growth of RAW264.7 in the control group was more rapid than that of the other three groups, and the cell growth statuses in the random-fiber membranes, oriented-fiber membranes, and oriented-fiber membranes containing STS were similar, without significant differences. Figure 7B showed that HUASMCs in the control group proliferated significantly faster than in the other three groups, and the HUASMCs on the three kinds of membranes had different growth rates. The random-fiber membranes had the slowest proliferation rate, and HUASMCs on the oriented-fiber membrane containing STS had the fastest proliferation rate. On day 7, cells on the oriented-fiber membrane with STS had higher proliferation than those on the other two membranes. This indicated that the oriented structure was effective to the proliferation of HUASMCs (Wang et al., 2013; Chen et al., 2017). Figure 8A shows that the growth of RAW264.7 was disordered on the random- and oriented-fiber membranes. Figure 8B shows that HUASMCs were disordered on the random-fiber membranes, while cells could be arranged in the same direction on the oriented-fiber membranes and oriented-fiber membranes containing STS. This confirmed that the oriented-fiber structure could effectively guide the arrangement of HUASMCs but not RAW264.7.

The distribution of cells on the random-fiber membranes showed an anisotropic disorderly arrangement. The behavior of 3T3 and HUASMCs was affected by the fiber microstructure, the parallel fibers induced cell migration along the fiber orientation. In summary, the membranes prepared were non-cytotoxic, and cells were in a good growth state on the membranes. Furthermore, the cells could be arranged in an orderly manner in the direction of the oriented fibers. Hence, this kind of fiber membrane, especially the oriented-fiber membranes consisting of STS, has high potential for in vivo implantation to repair tissue.