2.1. Materials

DegraPol DP30 was kindly provided by Ab Medica S.p.A. (Cerro Maggiore, Italy). Silkworm cocoons derived from Bombyx mori were generously supplied by Trudel Inc. (Zurich, Switzerland). Pellethane was procured from Dow Chemicals (Durban, South Africa).

Sodium chloride (NaCl), sodium carbonate (Na₂CO₃), lithium bromide (LiBr), 1,1,1,3,3,3-HFIP, methanol, ethanol, and glutaraldehyde were obtained from Sigma-Aldrich (St. Louis, MO, USA). Paraformaldehyde (PFA) was sourced from Kimix Chemicals (Durban, South Africa).

Phosphate-buffered saline (PBS) was prepared using NaCl, potassium chloride (KCl), and potassium dihydrogen orthophosphate (KH₂PO₄) from Saarchem (Honeydew, South Africa), along with disodium hydrogen orthophosphate dodecahydrate (Na₂HPO₄·12H₂O) supplied by Merck (Darmstadt, Germany).

For in vitro coculture experiments, endothelial growth medium (C-22010), fibroblast growth medium (C-23010), and endothelial cell supplement mix (C-39215) were purchased from PromoCell (Heidelberg, Germany) and supplemented with 1% penicillin/streptomycin (P4458, Sigma-Aldrich, Darmstadt, Germany). Dulbecco’s PBS ((D) PBS, BS.L 1825) was obtained from Bio&Sell (Feucht, Germany). Trypsin-EDTA (L2153) was sourced from Biochrom (Berlin, Germany). Medium 199 (M199, M4530) and heat-inactivated fetal bovine serum (FBS, F7524) were acquired from Sigma-Aldrich. Coculture bioreactors were custom-designed and manufactured in-house.

2.2. SF Preparation

SF was prepared as previously described [31]. In brief, Bombyx mori cocoons were boiled twice in 0.02-M sodium carbonate (Na₂CO₃) solution for 1 h to remove sericin, followed by extensive rinsing with ultrapure water. The degummed silk fibers were then dissolved in 9-M lithium bromide (LiBr) at 55°C for 1 h, yielding a 10% (w/v) solution. This solution was dialyzed using a 3500-Da molecular weight cutoff membrane (Pierce, USA) against ultrapure water for 48 h with five water exchanges. The resulting aqueous SF solution was subsequently frozen in liquid nitrogen and lyophilized using a VirTis Benchtop freeze dryer (SP Industries, Warminster, PA, USA).

2.3. Electrospinning of SF-DP30

Lyophilized SF was dissolved in HFIP at concentrations ranging from 10 to 40 weight percent (wt.%). Subsequently, DP30 was added at 60 to 90 wt.% under continuous stirring overnight to prepare a 20% (w/v) SF–DP30 solution. Control solutions containing 100 wt.% DP30 were also prepared by dissolving DP30 in either HFIP or chloroform.

Electrospinning was performed under specific conditions depending on the solvent system. HFIP-based solutions were electrospun using a 19-gauge stainless steel needle onto a rotating mandrel set at 400 rpm. The applied voltage was 20 kV, with a flow rate of 1.5 mL/h, a needle-to-collector distance of 35 cm, and relative humidity maintained between 20% and 30%.

For samples prepared using chloroform, electrospinning was conducted with an 18-gauge needle at 25 kV, a flow rate of 3 mL/h, a 35-cm working distance, and the same relative humidity (20%–30%).

Following electrospinning, all tubular scaffolds were immersed in 90 vol.% methanol for 30 min to induce the transformation of SF into its water-insoluble β-sheet crystalline conformation [32]. The same methanol treatment was applied to DP30 control samples without SF. All samples were then vacuum-dried at room temperature for 3 days.

A summary of the sample preparation conditions is provided in Table 1. For scanning electron microscopy (SEM), tensile testing, and porosity measurements, six independent measurements were conducted on separate electrospun tubes.

2.4. SEM

Dried scaffolds were examined using SEM (Phenom ProX, PhenomWorld, Eindhoven, The Netherlands) at an accelerating voltage of 15 kV, a working distance of 8.2 mm, and high-resolution mode with 500× magnification. Auto-contrast and auto-brightness settings were applied. Images were captured from both the luminal and abluminal sides of the scaffold. Fiber diameters (N = 300 fibers per image) and interpore diameters (all pores per image) were quantified using the PhenomWorld FiberMetric analysis software.

2.5. Scaffold Hydrosolubility

Dog bone–shaped specimens, punched using an in-house designed tool, were individually placed in Falcon tubes containing 10 mL of PBS (pH 7.4) and incubated at 37°C. After incubation, samples were rinsed with distilled water and dried overnight under vacuum at 37°C (MRC Scientific Instruments, Harlow, UK). Once dried, specimens were weighed using an analytical balance (XS105, Mettler-Toledo, Columbus, OH, USA) and then returned to fresh PBS (10 mL) at 37°C for continued incubation. Weights were recorded weekly for the first month and biweekly over the following 5 months, with a final measurement taken at 10 months. Each group consisted of 12 specimens (N = 12).

2.6. Scaffold Tensile Testing

Samples of DP30 scaffolds, prepared in either circumferential or axial mandrel orientations and fabricated using either chloroform or HFIP as solvents, were evaluated for ultimate tensile strength, 10% secant modulus, maximal strain, and toughness following ASTM D1708 standards (Figure​ 1). The same testing protocol was applied to SF/DP30 blends for a comparative analysis (Figure 2). Dog bone–shaped specimens, punched using a custom-designed in-house puncher, were subjected to uniaxial tensile testing at room temperature in air, utilizing an Instron 5544 tensile tester (Instron, Norwood, MA, USA) equipped with a calibrated 2-kN load cell and tensile grips. Specimens were preloaded with 0.05 N and tested at a strain rate of 12.5 mm/min until failure. Stress-strain data were recorded via the Instron software for a subsequent analysis.

2.7. Scaffold Porosity

Scaffold porosity was assessed using 6-mm diameter discs punched with a biopsy punch (Mednom, Cape Town, South Africa). Disc thickness was measured at the central point using a thickness gauge (Mitutoyo, Kawasaki, Japan). Scaffold porosity was calculated based on Archimedes’ principle through weight measurements of the specimens in air and submerged in ethanol (Balance XS105, Mettler-Toledo, Columbus, OH, USA) [33].

2.8. Heart Valve Hydrodynamic Performance

A 10 wt.% SF/DP30 blend was selected for prototype heart valve leaflet fabrication due to superior tensile properties; other blends were excluded. Electrospun scaffolds approximately 500 μm thick were produced, and three leaflets were punched and sutured in an arc-shaped configuration (Ti-Cron, Medtronic, Minneapolis, MN, USA) onto a Pellethane skirted transcatheter heart valve (TAVR) stent electrospun from chloroform under identical parameters as SF/DP30 (Strait Access Technologies (SAT), Cape Town, South Africa), resulting in the prototype valve, as shown in Figure 3.

Valves were prewetted in PBS for 24 h at 37°C prior to hydrodynamic testing in a Pulse Duplicator system (Vivitro Labs, Victoria, BC, Canada) simulating pulsatile, low-pressure right ventricular blood flow at 70 bpm, mean pressures of 50/20 mmHg, and cardiac output of 5 L/min, in accordance with ISO 5840-1:2021 standards. PBS served as a blood analog. Ten consecutive valve open/close cycles were recorded and analyzed for orifice area (cm²), transvalvular mean pressure (mmHg), regurgitation fraction (%), and transvalvular forward, closing, and leakage energy loss expressed as percentage of ventricular energy (% VE) (N = 5). Comparative hydrodynamic evaluations were conducted on in-house produced porcine pericardium and polyurethane valves [9], as well as commercial Edwards Perimount 2900 heart valves (N = 2). The porcine pericardium and polyurethane valves were 26-mm SAT TAVR valves with electrospun skirts, whereas the Edwards Perimount 2900 was a standard 25-mm bovine pericardial surgical valve. A video capturing leaflet dynamics of the 10 wt.% SF/DP30 valve was recorded.

2.9. Endothelial Cell and Fibroblast Isolation

Primary human endothelial cells were isolated from segments of the greater saphenous vein using established protocols [34, 35]. Veins were washed with M199 supplemented with 10 I.U. heparin (Ratiopharm, Ulm, Germany) and 0.1 mg/mL gentamycin (Bio&Sell, Feucht/Nürnberg, Germany), then filled with 210 U/mL collagenase type II (Worthington Biochemical Corp., St. Katharinen, Germany) in 1% human serum albumin (ZLB Behring, Bern, Switzerland) for 15 min at 37°C. Digestion was halted by addition of M199 with 20% FBS, and cells were cultured in endothelial growth medium. Cells were used up to passage three and maintained at 37°C in 95% humidified air with 5% CO₂ until reaching 80% confluency.

Human fibroblasts were isolated by outgrowth culture. Vein segments were sectioned into small squares, positioned lumen-side down between coverslips and Petri dishes, and incubated in fibroblast growth medium at 37°C under 95% humidified air with 5% CO₂.

2.10. Endothelial Cell Culture on SF/DP30 Blends

SF/DP30 scaffolds were sterilized in 70% ethanol and rinsed thrice with PBS before cell seeding. Endothelial cells were cultured in endothelial growth medium (PromoCell), with 5 × 10⁴ cells seeded onto punched, PBS-prewetted scaffolds (1 h in 96-well plates at room temperature). Cells were cultured in 96-well plates and tissue culture plastic (TCP) for control staining. After 4 days, cells were fixed with 4% formalin and stained for PECAM-1 (CD31) using antibody ab76533 (1:50; Abcam, Waltham, MA, USA) and counterstained with DAPI. Imaging was performed using confocal microscopy (LSM 880 Airy Scan, Carl Zeiss, Oberkochen, Germany).

2.11. Cell Cytotoxicity Assessment

To evaluate potential cytotoxic effects of SF-DP30 hybrid scaffolds, a WST-1 assay (Roche Diagnostics, Mannheim, Germany) quantified the metabolic activity of endothelial cells exposed to scaffold degradation products over 14 days. The assay measures mitochondrial dehydrogenase activity in viable cells. Briefly, 10 wt.% SF/DP30 scaffolds were sterilized in 70% ethanol and washed thrice in PBS. Endothelial cells (6.5 × 10⁴ per well) were seeded in 24-well plates and incubated in endothelial cell growth medium (ECGM) supplemented with 5% fetal calf serum and supplement mix (PromoCell) at 37°C and 5% CO₂. Conditioned media, collected from scaffolds incubated in growth medium for 1, 3, 6, and 14 days (N = 3), were applied to cell cultures 24 h postseeding. Controls included nonconditioned medium and blanks. WST-1 assay was performed after an additional 24 h, with optical density measured in triplicate. Data are presented as mean ± SD from three endothelial cell donors.

2.12. Coculture of Endothelial Cells and Fibroblasts on a 10 wt.% SF/DP30 Heart Valve

Sterilized 10 wt.% SF/DP30 heart valves (70% ethanol, followed by PBS rinses) were seeded with 7.5 × 10⁵ fibroblasts/cm² in a static seeding bioreactor and incubated for 24 h at 37°C in fibroblast growth medium (PromoCell). Valves were then transferred to a glass beaker with fresh medium; 50% medium changes were performed after 2 days. One week post fibroblast seeding, 7.5 × 10⁵ endothelial cells/cm² were seeded atop the fibroblast layer and incubated for 24 h under static conditions. Following this, valves were moved to a dynamic conditioning bioreactor (as described in reference [36]) and cultured for 10 days with pulsatile medium perfusion increasing from 20% to 30% over one week to simulate shear stress. Medium was changed every 2-3 days. Samples for histological analysis were fixed in 4% formalin; those for SEM were fixed in FIX II solution.

2.13. Immunofluorescent Analysis of Endothelial and Fibroblast Cell Markers

Postcoculture cells on scaffolds were processed for immunocytochemical staining. Endothelial cells were identified using von Willebrand Factor (vWF) and PECAM-1 (CD31) markers. Samples embedded in paraffin were deparaffinized through graded alcohols and permeabilized with 0.5% Triton X-100 in PBS for 30 min at room temperature. Antigen retrieval was performed using Target Retrieval Buffer (1:10) for vWF and Tris-EDTA buffer (pH 9.0) for CD31, followed by 15 min steam cooking. Blocking was done with 10% donkey serum in PBS + 0.1% Triton for 15 min. Primary antibodies—vWF (HPA001815, 1:100, Sigma) and CD31 (ab76533, 1:50, Abcam)—were incubated overnight at room temperature. Secondary antibodies (ab150066, Abcam) were applied at 1:200 for vWF and 1:400 for CD31 for 90 min in the dark. Negative controls omitted primary antibodies. Nuclei were counterstained with DAPI (1:1000). Washes (3 × 5 min) in PBS followed each step.

Fibroblasts were stained for TE-7 and collagen type I (Col-1). After similar paraffin embedding, deparaffinization, and permeabilization, antigen retrieval used 0.1-mM EDTA buffer (pH 8.0) for TE-7 and Target Retrieval Buffer (1:10) for Col-1 with 15-min steam cooking. Blocking was with 10% goat/donkey serum for 15 min. Primary antibody for Col-1 (ab34710, 1:50) was incubated overnight at 4°C. Secondary antibody (ab150079, 1:150) was incubated for 90 min in darkness. Controls without primary antibody were included. Nuclei were counterstained with DAPI. Samples were mounted (Ibidi, Gräfelfing, Germany) and imaged using fluorescence microscopy (AxioObserver, Carl Zeiss).

Histological analysis involved fixation in 4% formalin (24 h), dehydration through graded ethanol and xylene, paraffin embedding, and sectioning (5 μm). Sections were stained with hematoxylin and eosin (H&E).

2.14. Cellular SEM Imaging

Fibroblasts and endothelial cells on 10 wt.% SF/DP30 valve leaflets were rinsed in PBS, fixed in 2.5% glutaraldehyde in 0.1-M sodium cacodylate buffer for 30 min at room temperature, then postfixed in 1% osmium tetroxide for 1 h. Samples were dehydrated in graded ethanol series, critical point dried (CPD 030, BAL-TEC, Balzers, Liechtenstein), and sputter-coated with gold (SCD 050, BAL-TEC). SEM imaging (EVO LS10, Carl Zeiss, Oberkochen, Germany) was performed at 100× and 1000× magnifications on both leaflet surfaces.

2.15. Statistical Analysis

Data were tested for normality and expressed as mean ± standard deviation. Comparisons were performed using ANOVA with post hoc Tukey HSD test. A p value < 0.05 was considered statistically significant. Sample sizes ranged from N = 6 to 12 when possible.

3.1. Fiber and Pore Size Evaluation

Representative SEM images of electrospun SF/DP30 scaffolds are presented in Figure 4.

The average fiber diameters ranged from a minimum of 6 μm (range: 0.6—30 μm) observed in the 10 wt.% SF/DP30 scaffold on the lumenal side to a maximum of 13 μm (range: 0.6—42 μm) in the 40 wt.% SF/DP30 scaffold on the ablumenal side. Interfiber distances averaged between 3 μm (range: 0.6—12.42 μm) for 10 wt.% SF/DP30 (both lumenal and ablumenal sides) and 10 μm (range: 0.66—48.43 μm) for 40 wt.% SF/DP30 on the ablumenal side. Increasing SF content not only correlated with larger interfiber distances but also resulted in more heterogeneous fiber morphology. SF concentrations above 40 wt.% were not spinnable due to elevated viscosity and fiber discontinuity during electrospinning; therefore, these blends were excluded from further analysis. All scaffold variants exhibited some degree of fiber merging. Quantitative measurements demonstrated good agreement between lumenal and ablumenal sides, with both average fiber diameter and interfiber spacing increasing in tandem with SF content (Figure 5).

3.2. Porosity and Hydrosolubility

Porosity, defined as the percentage of air volume within the porous scaffold, increased progressively from 41% ± 5% in 0 wt.% SF/DP30 scaffolds to 75% ± 2% in 40 wt.% SF/DP30 scaffolds. The 0 wt.% SF/DP30 scaffold electrospun in chloroform exhibited an intermediate porosity of 65% ± 3%. Hydrosolubility studies in PBS at 37°C revealed a dry weight loss initiating after 4 weeks, followed by an exponential degradation profile. After 40 weeks, no significant differences in degradation rates were observed between groups, except for 10 wt.% SF/DP30 and 20 wt.% SF/DP30 samples, which showed statistically significant differences (p < 0.05). DP30 scaffolds electrospun in chloroform demonstrated higher hydrosolubility compared to those prepared in HFIP (Figure 6).

3.3. Tensile Properties

Tensile testing comparing plain DP30 scaffolds electrospun in chloroform versus HFIP revealed no significant differences in ultimate tensile strength (chloroform: 0.5 ± 0.1 to 0.7 ± 0.1 MPa; HFIP: 1.0 ± 0.2 to 1.1 ± 0.2 MPa), maximum strain (chloroform: 41 ± 8% to 67 ± 14%; HFIP: 43 ± 12% to 52 ± 13%), toughness (chloroform: 0.2 ± 0.1 MJ/m³; HFIP: 0.3 ± 0.1 to 0.4 ± 0.2 MJ/m³), and 10% secant modulus (chloroform: 1.6 ± 0.2 to 3.3 ± 0.3 N/mm²; HFIP: 4.7 ± 1.1 to 5.7 ± 1.1 N/mm²) (Figures 1(a) and 1(b)). No significant differences were detected between samples punched in the circumferential versus axial orientation (data not shown).

Figure 2 summarizes tensile test data for all SF/DP30 blends, highlighting parameters critical for selecting materials for heart valve prototypes. Ultimate tensile strength ranged from 0.4 ± 0.04 MPa (30 wt.% SF/DP30, circumferential) to 1.1 ± 0.4 MPa (40 wt.% SF/DP30, axial). The 10% secant modulus spanned 1.6 ± 0.2 N/mm² (DP30 + chloroform, circumferential) to 7.6 ± 1.9 N/mm² (40 wt.% SF/DP30, circumferential). Maximum strain varied between 11.6 ± 4.4% (40 wt.% SF/DP30, axial) and 89.6 ± 10.0% (10 wt.% SF/DP30, circumferential). Toughness ranged from 0.04 ± 0.02 MJ/m³ (40 wt.% SF/DP30, axial) to 0.6 ± 0.1 MJ/m³ (10 wt.% SF/DP30, circumferential). Circumferentially punched samples exhibited slightly higher toughness and maximum strain compared to axially punched samples, which showed relatively higher ultimate tensile strength.

3.4. Hydrodynamic Properties

Hydrodynamic pulse duplicator testing of heart valve prototypes fabricated with 10 wt.% SF/DP30 leaflets demonstrated smooth opening and closing dynamics (see supporting video, cranial view (available here)). Occasional leaflet delamination was observed in samples exceeding 500 μm thickness. Averaged data from five heart valves are summarized as follows: transvalvular mean pressure gradient (ΔP) of 5.0 ± 0.6 mmHg, regurgitation fraction of 10.4 ± 2.8%, effective orifice area (EOA) of 1.65 ± 0.14 cm², transvalvular forward energy loss of 12.2 ± 3.6% ventricular energy (VE), closing energy loss of 1.9 ± 0.8% VE, and leakage energy loss of 3.4 ± 1.5% VE. Comparative hydrodynamic data for SF/DP30 valves, porcine pericardium, polyurethane, and a commercial Edwards Perimount 2900 valve are summarized in Table 2 [9].

3.5. WST-1 Cytotoxicity Assessment

No significant cytotoxicity was detected at any time point (T1, T3, T6, and T14) relative to 48-h controls (Figure 7). All groups exposed to 10 wt.% SF/DP30, with or without direct material incubation, maintained optical density (OD450) values comparable to controls. Metabolic activity remained stable or slightly increased over time, indicating that degradation products are nontoxic and may even promote cell viability. The 24-h baseline control exhibited significantly lower metabolic activity than all subsequent time points, consistent with expected cell proliferation. These results confirm scaffold biocompatibility, corroborated by favorable endothelial and fibroblast coculture outcomes.