Polycaprolactone (PCL, Mn = 80,000) and Span 80 were purchased from Sigma-Aldrich (United States). Hexafluoroisopropanol (HFIP) was purchased from Shanghai Darui Fine Chemicals Co., Ltd. (Shanghai, China). Heparin (Hep), dichloromethane (DCM), and toluidine blue were purchased from Macklin (Shanghai, China). Anticoagulated sheep blood, Calcein AM/PI detection working solution, 4,6-diamidino-2-phenylindole (DAPI), phosphate-buffered saline (PBS), and 4% paraformaldehyde were purchased from Solarbio (Beijing, China). Phalloidin-iFluor 488 (FITC 488) was purchased from Abcam (United Kingdom). Penicillin/streptomycin solution and 0.25% trypsin were obtained from Biosharp (Beijing, China). The Cell Counting Kit-8 (CCK-8) was purchased from Glpbio (United States). Fetal bovine serum (FBS) was purchased from PAN Biotech (Germany). Dulbecco’s Modified Eagle Medium (DMEM) was purchased from Procell (Wuhan, China).

Cell proliferation of HUVECs seeded on P-0% E, P-10% E, P-20% E, P-30% E, and P-40% E was assessed by CCK-8 assay. Briefly, the obtained nanofiber membranes were adhered to glass slides and placed in a 24-well plate. The plate was sterilized with 75% ethanol vapor for 4 h, followed by exposure to UV irradiation for 30 min. The blank glass slides were set as a control. HUVECs were seeded at 7 × 10³ cells/well in 24-well plates and cultured in 400 µL of DMEM supplemented with 10% FBS for 1, 3, and 5 days. Then, the medium was removed and replaced with a medium containing 10% CCK-8 (v/v), and the plate was incubated in an incubator for 2 h. Afterward, 100 µL of medium was transferred to a 96-well plate for measuring the absorbance at 450 nm using a microplate reader (Thermo Scientific Varioskan™ LUX).

The morphology of HUVECs on P-0% E, P-10% E, P-20% E, P-30% E, and P-40% E was observed by an inverted fluorescence microscope on days 3 and 5. Each well was washed three times with PBS, followed by the application of 4% tissue cell fixative to fix HUVECs at 4 °C overnight. Consequently, the fixative was washed with PBS, and 0.1% Triton was added to incubate for 5 min. This was followed by PBS washing and the addition of 1% BSA for 1 h. The F-actin staining solution was diluted with BSA at a 1:2000 volume ratio. This solution was added for 30 min and then washed with PBS. Subsequently, DAPI was added for 5 min at room temperature. The morphology of stained HUVECs was observed and photographed using an inverted fluorescence microscope.

The cell proliferation and morphology of HUVECs for P-E/H nanofibers were assessed using the aforementioned methods.

The cytotoxicity of P-E/H on HUVECs was assessed through live-dead staining. Briefly, P-E/H was adhered to glass slides and placed in a 24-well plate. The plate was sterilized with 75% ethanol vapor for 4 h, followed by exposure to UV irradiation for 30 min. The blank glass slides were set as a control. HUVECs were seeded at 2 × 10⁴ cells/well in 24-well plates and cultured in 400 µL of DMEM supplemented with 10% FBS for 24 h. Then the medium was removed and washed with PBS, followed by the addition of 200 µL Calcein AM/PI detection working solution to incubate in the dark at 37 °C for 30 min. Observations of live and dead HUVECs were performed under a fluorescence microscope, and the ratio of live to dead cells was determined using ImageJ software.

The preparation of P-E/H vascular scaffolds was conducted by modifying the methods in Section 2.5.1. Specifically, the flat plate receiver was replaced with a stainless steel metal rod receiver with diameters of 1 mm, 2 mm, 3 mm, 4 mm, 5 mm, and 6 mm. The rolling speed of the receiver was set at 600 rpm. The morphology of the cross-section and longitudinal section of a 4 mm diameter P-E/H vascular scaffold was observed using SEM.

The anticoagulant efficacy of tubular scaffolds was determined by an in vitro blood circulation test. A closed-loop system, comprising a peristaltic pump, centrifuge tubes, silicone tubing, and P-E/H vascular scaffolds, was utilized to investigate the anticoagulant properties of the P-E/H vascular scaffold by simulating the in vitro blood circulation process. A peristaltic pump was used to simulate venous blood flow, with a flow rate of 20 mL/min. The P-E/H vascular scaffolds were removed after 3 and 5 days of blood circulation, respectively. The residual blood from the inner surface of P-E/H vascular scaffolds was washed with PBS, and the cross-section and longitudinal sections of the scaffolds were photographed.

The multiple comparison procedures between groups were performed using one-way ANOVA with Origin, and each group was repeated at least three times. Statistical results were expressed as means ± standard deviation (SD). To observe the significance of differences between the test groups. Student’s t-test was used for all two-by-two comparisons. A value of ^* P < 0.05 represents the lowest significance level, ^** P < 0.01 represents a moderate significance level, and ^*** P < 0.001 represents the highest significance level.

Endothelial cells secrete various functional proteins, such as fibronectin and collagen, which constitute the primary regulatory network of the angiogenic microenvironment, along with VEGF. As a proangiogenic factor, VEGF modulates endothelialization by regulating the proliferation and migration of endothelial cells through activation of the PI3K/AKT signaling pathway (Krüger-Genge et al., 2019). The extracted ECd as crude conditioned medium lysate of HUVECs contains 0.35 ± 0.019 mg endothelial cell-secreted proteins and 41.96 ± 20.39 pg VEGF in ECd to promote endothelial proliferation (Supplementary Figure S1). In this study, coaxial electrospinning technology was employed to fabricate PCL-ECd nanofibers with a core-shell structure. As demonstrated in SI, the core layer and shell layer flow rates were set to 0.1–1 mL/h, 0.1–2 mL/h, and 0.05–2 mL/h, named as PE1, PE2, and PE3, respectively. The morphology of resultant nanofibers was displayed in Supplementary Figure S2, suggesting the spinnability of PE1, PE2, and PE3. Although the obtained fibers exhibited a random distribution, PE2 and PE3 showed more uniform thickness, smoother surfaces, and a more satisfying morphology compared to PE1. The loading efficiency of ECd in PE1, PE2 and PE3 was measured as 50.68% ± 8.05%, 74.44% ± 12.68%, and 81.68% ± 10.75%, respectively (Supplementary Figure S3). The superior loading efficiency of PE2 and PE3 may have contributed to the faster flow rate of the shell layer, allowing it to stretch and trap nanofibers, thereby facilitating the encapsulation of the core layer. The highest loading efficiency of PE3 may be relative to the lower flow rate of the core layer compared to PE2, which allows the core layer solution to fill into the interior fiber with the fast flow rate of the shell layer, resulting in a reduction of uneven distribution of nanofibers for the core layer and improvement of loading efficiency for the core layer.

Subsequently, PCL-ECd nanofibers with a core-shell structure were prepared following the parameters of PE3. The morphology and fiber diameter distribution of PCL-ECd nanofibers are shown in Figure 1A. The produced random nanofibers with different amounts of ECd featured smooth surfaces. Figure 1B indicated that the average diameter of P-0% E, P-10% E, P-20% E, P-30% E, and P-40% E was 0.89 ± 0.29 µm, 1.10 ± 0.43 µm, 1.63 ± 0.38 µm, 1.70 ± 0.44 µm, and 2.17 ± 0.38 µm, respectively. The increase in ECd concentration is correlated with the diameter of the fiber. FITC could emit green fluorescence at specific wavelengths by binding with functional groups of proteins (Restan et al., 2019). FITC was selected to confirm the core-shell structure for PCL-ECd nanofibers. As shown in Figure 1C, the core layer displayed green fluorescence as FITC binding proteins within ECd, indicating the successful preparation of a core-shell structure for PCL-ECd nanofibers. The addition of ECd improved the hydrophilicity of fibrous membranes compared to the pure PCL nanofiber membrane (Figure 1D). It could be explained by the formation of hydrogen bonds between the hydroxy or carboxyl group within ECd and the hydroxy group within water and the van der Waals force. The parameters for producing P-30% E were selected to conduct the following experiments, which were described in detail below. The tensile properties were evaluated through stress-strain curves, which compared the tensile strength of P-30% E and pure PCL nanofibers (Figure 1E). The tensile strength of P-30% E was 5.44 ± 0.37 MPa, which was lower than PCL nanofibers at 8.21 ± 0.48 MPa. Though the presence of ECd weakened the mechanical properties of the prepared fibrous membrane, the tensile strength of P-30% E is over the tensile modulus for human coronary arteries (0.7–2.1 MPa) (Norouzi and Shamloo, 2019). It could meet the mechanical performance requirements for small-caliber artificial vessels. The loading efficiency of P-30% E was measured as 75.14% ± 1.58%. P-30% E nanofibers exhibited a fast release in 24 h at 79.29% ± 3.63% and cumulative release up to 84.22% ± 2.62% after 30 days (Figure 1F). The release profile of P-30% E failed to meet the long-term sustained-release requirements. Then, coaxial-emulsion electrospinning was employed to prepare nanofibers for subsequent experiments, aiming to achieve sustained-release behavior.

The biocompatibility of PCL-ECd nanofibers with different amounts of ECd was assessed using HUVECs. P-0% E, P-10% E, P-20% E, P-30% E and P-40% E indicated good biocompatibility over 5 days, where P-30% E and P-40% E showed significant difference with P-0% E, P-10% E and P-20% E (Figure 2A). P-30% E exhibited a faster proliferation behavior than P-40% E, suggesting that an appropriate ECd concentration is favorable for HUVECs proliferation.

Artificial vessels, as medical devices that come into contact with blood, require superior hemocompatibility. Otherwise, they may lead to severe damage to red blood cells, hemolysis, or even the formation of thrombus (Xiang et al., 2024). The hemolysis test is used to evaluate damage to red blood cells caused by material, and the hemolysis rate demonstrates damage severity (Lu et al., 2021). In this test, double-distilled water served as the positive control, and saline solution served as the negative control. As shown in Figure 2C, the supernatant of P-0% E, P-10% E, P-20% E, P-30% E, P-40% E, and the negative control was colorless and transparent with red blood cells precipitated at the bottom. In comparison, sedimentation was not observed after centrifugation, as red blood cells were ruptured in the positive control. The quantitative analysis indicated the hemolysis rate in material groups was below 2%, indicating mild hemolysis. GB/T 16,886.4–2022 requires the hemolysis rate of medical devices to be below 5%, which suggests the promising potential of PCL-ECd nanofibers in clinical settings (Ge et al., 2019). The plasma recalcification time was determined to assess anticoagulant properties of P-0% E, P-10% E, P-20% E, P-30% E, and P-40% E. Calcium ions were added into plasma to induce coagulation, and the coagulation rate was measured by observing the half-coagulation time and the slope of the recalcification curve (Ma et al., 2024; Xu et al., 2023a). Comparing with the negative control, the absorbance of other groups was increased, suggesting the appearance of coagulation. However, material groups failed to prolong coagulation time, demonstrating poor anticoagulant efficacy of PCL-ECd nanofibers.

To improve the anticoagulant and anti-thrombotic properties of PCL-ECd nanofibers, Hep was involved in fabricating P-E/H nanofibers with a core-shell structure. The morphology of P-E/H, P-E, P-H and pure PCL was displayed in Figure 4A, illustrating a smooth surface and uniform diameter for P-E/H. As shown in Figure 4B, the average diameters of P-E/H, P-E, P-H, and PCL were 631.3 ± 198.4 nm, 550.6 ± 156.57 nm, 459.7 ± 106.2 nm, and 1.074 ± 0.435 µm, respectively. Although nanofibers introduced two different components, the average diameters of P-E/H were smaller than those of pure PCL. The introduction of Hep and oil-in-water emulsion narrowed the diameter of PCL-ECd nanofibers. The addition of Span 80 enhanced the interface compatibility between the aqueous phase and the oil phase in the emulsion, resulting in a more uniform bonding of the interface between the core layer and the shell layer during the electrospinning process and stretched jet under the electric field. Furthermore, ECd and Hep were encapsulated within the aqueous phase by PCL in the oil phase to form dispersed droplets, rather than a continuous aqueous phase solution, which significantly reduced the viscosity of the core layer, allowing for the stretching of nanofibers with smaller diameters.

The successful preparation of a core-shell structure for P-E/H nanofibers was confirmed following the aforementioned approach. Briefly, FITC mixed with ECd and Hep in the core layer showed green fluorescence, while the shell layer was uncolored (Figure 3C). The tensile properties of P-E/H, P-E, P-H, and PCL were examined through stress-strain curves (Figure 4A). Due to the reduction of PCL within the core layer, the tensile strength of P-E/H, P-E, and P-H was lower than that of PCL at 9.04 ± 0.79 MPa. And the ultimate stress of P-E/H, P-E, and P-H reached 4.39 ± 0.41 MPa, 4.69 ± 0.11 MPa, and 5.56 ± 0.67 MPa, respectively, which was weaker than PCL at 9.04 ± 0.79 MPa (Figure 4B). The elongation at break of P-E/H and P-E was 218.10% ± 28.27% and 257.14% ± 35.18%, respectively (Figure 4C). It was lower than P-H (361.12% ± 27.96%) and PCL (380.94% ± 25.32%). Though the tensile strength of P-E/H lowered to 4.6 ± 0.2 MPa, it is still greater than that strength desired for human coronary arteries at 0.7–2.1 MPa. Herein, P-E/H exhibited ideal mechanical performance to meet the requirements for small-caliber artificial vessels. The loading efficiency of ECd and Hep in P-E/H was assessed as 78.51% ± 5.77% and 65.14% ± 1.58%, respectively. The combination of coaxial-emulsion electrospinning prolonged the release time of ECd and Hep, achieving a sustained release of up to 60 days (Figure 4D). The cumulative release of ECd in P-E/H reached 78.41% ± 0.03% on day 30, and Hep in P-E/H was cumulated to 78.46% ± 2.42% for 20 days, followed by a slow-release behavior. It may be assigned with a water-in-oil emulsion, where the aqueous phase is dispersed as droplets in the oil phase to hinder the release of ECd and Hep from the aqueous phase. Moreover, the core-shell structure of nanofibers exerted a synergetic effect in impeding the release of ECd and Hep, as they were encapsulated in the core layer. Additionally, the shell layer of nanofibers may contain tiny pores, allowing the oil phase of the emulsion to fill them and form a denser shell layer, thereby reducing the leakage of ECd and Hep.

The biocompatibility of P-E/H, P-E, P-H, and PCL was assessed using HUVECs. Figure 5A shows the staining results of HUVECs co-cultured on different fibers for 3 days, where cell nuclei were stained blue, and the cytoskeleton was stained green. HUVECs cultured on P-E/H, P-E, P-H, and PCL exhibited a pebble-like morphology, overlapping with that of the control. Cell compatibility was further evaluated using live/dead cell staining. Live cells exhibited green fluorescence, while dead cells exhibited red fluorescence (Figure 5B). The fluorescence images of HUVECs co-cultured with P-E/H, P-E, P-H, and PCL for 24 h demonstrated that P-E/H possessed good cell compatibility. The number of live cells for produced nanofibers exceeded 96%, demonstrating no adverse effects on cell viability for P-E/H, P-E, P-H, and PCL (Figure 5D). The absorbance of HUVECs co-cultured with P-E/H, P-E, P-H and PCL was measured for 1, 3, and 5 days (Figure 5C). There was a cell proliferation of HUVECs co-cultured with obtained fibrous membranes over 5 days, suggesting a desired cell compatibility for P-E/H, P-E, P-H and PCL. P-E/H showed superior cell proliferation ability compared with PCL on day 3 (P = 0.025) and day 5 (P = 0.00021). P-E/H showed a pronounced cell proliferation profile compared to P-E (P = 0.0025 on day 3 and P = 0.0251 on day 5) and P-H (P = 0.0017 on day 1, P = 0.000003 on day 3 and P = 0.023 on day 5), which may contribute to ECd and Hep synergistically exerting promotion effects for HUVECs proliferation (Wang et al., 2015). PCL also showed acceptable cell proliferation that may be attributed to the high sensitivity of HUVECs to their local microenvironment. HUVECs responded to topographical cues of micron-scale PCL, facilitating cell spreading and proliferation (Ahmed et al., 2018).

The hemolysis test was employed to assess the damage of red blood cells caused by P-E/H, P-E, P-H and PCL. In this test, double-distilled water served as the positive control, and saline solution served as the negative control. As shown in Figure 6A, P-E/H, P-E, P-H, PCL, and the negative control did not exhibit significant changes, as the supernatant was transparent with red blood cells precipitated at the bottom. In contrast, the red blood cells of the positive control were ruptured, with the absence of sedimentation after centrifugation. Based on the quantitative analysis of supernatant, the hemolysis rates in the material groups and the negative control were below 2% (Figure 6A). It demonstrated P-E/H, P-E, P-H and PCL lead to mild hemolysis, being classified as non-hemolytic biomaterials.

The anticoagulant effect of P-E/H, P-E, P-H, and PCL was evaluated through a dynamic coagulation experiment to determine the coagulation of anticoagulated whole blood on the obtained nanofibers. Specifically, the red blood cells that were not involved in the thrombus were lysed in double-distilled water, and the absorbance change was then determined (Li G. et al., 2024; Xu et al., 2023a). Figure 6C shows the absorbance change of P-E/H, P-E, P-H and PCL contacting with whole blood at different time points. At 180 min, the absorbance ranked as P-E/H > P-H > P-E > PCL. The highest absorbance for P-E/H and a slower decline in the coagulation curve, illustrating that the addition of hep improved the anticoagulant effect of P-E/H. The plasma recalcification time was determined to further verify the anticoagulant performance of P-E/H, P-E, P-H and PCL. As shown in Figure 6B, P–E, PCL, and positive control exhibited calcification with the addition of CaCl₂. In comparison, P-E/H and P-H calcified until 240 min, suggesting the presence of Hep improved the anticoagulant properties of P-E/H and P-H.

As shown in Figure 7A, P-E/H vascular scaffolds with different diameters were successfully fabricated. Figure 7B reveals that longitudinal and transverse sections of P-E/H with 4 mm diameters show a disordered arrangement without bead-like morphology. This structure enhances flexibility for vascular scaffolds and promotes cell proliferation. A closed-loop system was used to evaluate the anticoagulant properties of P-E/H vascular scaffolds by simulating the in vitro blood circulation process (Figure 7C). Figure 7D shows images of the 4 mm P-E/H vascular scaffold in vitro circulation test on days 3 and 5. Results indicate that after 3 and 5 days of circulation testing, the outer surface of the vascular scaffold remained smooth, and no thrombus was observed on the inner wall of the scaffold. Therefore, the P-E/H vascular scaffold exhibits anticoagulant properties over a certain period, preliminarily mimicking the functionality of natural vessels (Xu et al., 2023b). However, our study lacks in vivo patency data to verify the long-term performance of P-E/H vascular scaffolds. Future animal experiments will be conducted to further compare the performance of P-E/H vascular scaffolds with natural vessels.