Development of a dual-component infection-resistant arterial replacement for small-caliber reconstructions: A proof-of-concept study

Abstract

Introduction: Synthetic vascular grafts perform poorly in small-caliber (<6mm) anastomoses, due to intimal hyperplasia and thrombosis, whereas homografts are associated with limited availability and immunogenicity, and bioprostheses are prone to aneurysmal degeneration and calcification. Infection is another important limitation with vascular grafting. This study developed a dual-component graft for small-caliber reconstructions, comprising a decellularized tibial artery scaffold and an antibiotic-releasing, electrospun polycaprolactone (PCL)/polyethylene glycol (PEG) blend sleeve. Methods: The study investigated the effect of nucleases, as part of the decellularization technique, and two sterilization methods (peracetic acid and γ-irradiation), on the scaffold's biological and biomechanical integrity. It also investigated the effect of different PCL/PEG ratios on the antimicrobial, biological and biomechanical properties of the sleeves. Tibial arteries were decellularized using Triton X-100 and sodium-dodecyl-sulfate. Results: The scaffolds retained the general native histoarchitecture and biomechanics but were depleted of glycosaminoglycans. Sterilization with peracetic acid depleted collagen IV and produced ultrastructural changes in the collagen and elastic fibers. The two PCL/PEG ratios used (150:50 and 100:50) demonstrated differences in the structural, biomechanical and antimicrobial properties of the sleeves. Differences in the antimicrobial activity were also found between sleeves fabricated with antibiotics supplemented in the electrospinning solution, and sleeves soaked in antibiotics. Discussion: The study demonstrated the feasibility of fabricating a dual-component small-caliber graft, comprising a scaffold with sufficient biological and biomechanical functionality, and an electrospun PCL/PEG sleeve with tailored biomechanics and antibiotic release.

Keywords: antimicrobial activity; biomechanics; decellularized scaffold; polymeric sleeve; small-caliber vascular graft; tibial artery.

FIGURE 1

Alcian Blue/PAS, Elastica Van Gieson, Masson’s Trichrome, DAPI and H&E staining of native PTA and decellularized PTA scaffolds treated with (M2) or without (M1) RNase/DNase and sterilized with either PAA or γ–irradiation. Alcian Blue/PAS stained nuclei dark blue, acid mucosubstances (GAGs) and proteoglycans blue, and neutral polysaccharides magenta; Elastica Van Gieson stained nuclei black/brown, cytoplasm yellow/pink, elastic fibers purple and collagen light purple; Masson’s Trichrome stained nuclei dark blue/black, cytoplasm bright red, elastic fibers red and collagen fibers blue; DAPI stained nuclei blue and fluorescent under UV light; H&E stained nuclei blue/beep purple, cytoplasm red and collagen pale pink. Media: tunicae media; Intima: tunicae interna; Externa: tunicae externa; L: lumen; IEL: internal elastic lamina; EEL: external elastic lamina. All decellularized scaffolds appeared to be void of cell and cellular debris and preserved the trilaminar structure of the native tissue. It also appeared to be depleted of acid mucosubstances whilst its ECM fibers conserved their sugar moieties (Alcian Blue/PAS). The decellularized tissue also presented a well-preserved network of collagen and elastic fibers, as well as IEL and EEL. Scale bars indicate 500 μm.

FIGURE 2

Immunohistochemical staining against α-gal and collagen IV of native PTA and decellularized PTA scaffolds treated with (M2) or without (M1) RNase/DNase and sterilized with either PAA or γ-irradiation, also showing the HRP and isotype controls. Tissue positive against the α-gal epitope was stained light brown, whereas tissue positive to collagen IV was stained dark brown. All decellularized scaffolds expressed α-gal, albeit in a reduced amount compared to the native control. Collagen IV was expressed throughout the media, intima and basement membrane of the native PTA and the decellularized PTA scaffolds that were sterilized with γ-irradiation. No staining against collagen IV could be observed in the decellularized scaffolds that were sterilized with PAA. Scale bars indicate 500 μm.

FIGURE 3

TEM imaging at low (A1–B5), medium (C1–D5) and high (E1–F5) magnification of native PTA (A1, B1, C1, D1, E1, F1) and decellularized PTA scaffolds treated with (M2; A4, B4, C4, D4, E4, F4, and A5, B5, C5, D5, E5, F5) or without (M1; A2, B2, C2, D2, E2, F2, and A3, B3, C3, D3, E3, F3) RNase/DNase and sterilized with either PAA (A2, B2, C2, D2, E2, F2, and A4, B4, C4, D4, E4, F4) or γ–irradiation (A3, B3, C3, D3, E3, F3, and A5, B5, C5, D5, E5, F5). Media: tunicae media; Intima: tunicae interna; È: elastic fibers; ▷ collagen fibers; ✶ interna elastic lamina. All scaffolds appeared to be completely decellularized, containing collagen and elastic fibers. Although no differences could be observed between M1 and M2, the elastic fibers of the PAA-treated scaffolds (E2, E4 nested) appeared to be cleaved of the thin fibrils that decorated the native tissue (E1 nested) and the γ-irradiated scaffolds (E3, E5 nested). In all samples, tropocollagen and collagen fibrils were evident between and around the collagen fibers (F1, F2, F3, F4, F5 nested), which were rather reduced in the case of the PAA-treated scaffolds (F2, F4, nested).

FIGURE 4

Mean DNA (A), sGAG (B), HYP (C) and dHYP (D) content per dry weight of native PTA and decellularized PTA scaffolds treated with (M2) or without (M1) RNase/DNase and sterilized with either PAA or γ-irradiation. Data was expressed as means (n = 6); error bars indicate 95% C.I.; MSD: minimum significant difference; indicates significant difference between the native and the decellularized groups. Connectors indicate significant difference between the originator and end arrow group. (E–G) Characteristic en face multiphoton images of the lumen of decellularized PTAs, treated with M1 and sterilized with either PAA (E,F) or γ-irradiation (G), showing the elastic fiber network of internal elastic lamina (E, G; imaging plane was approximately 20 μm below lumen surface) and the collagen fibers of tunica media (F; imaging plane was about 60 μm below lumen surface). Black arrow heads (►) show elastic fibers. Scale bars indicate 100 μm (E,F) or 50 μm (G).

FIGURE 5

Contact (A–H) and extract (I,J) cytotoxicity results of the decellularized PTA scaffolds treated with (M2; (B,D,I) or without (M1; (A,C,I) RNase/DNase and sterilized with either PAA (A,B,I) or γ-irradiation (C,D,I), together with the corresponding results for the 150:50 (E,J) and 100:50 (F,J) electrospun polymeric sleeves sterilized with PAA. The contact cytotoxicity (A–H) was performed with L929 murine fibroblasts, which were stained with Giemsa. Cyanoacrylate glue and collagen gel served as the positive (+ve) and negative (-ve) control, respectively. Scale bars indicate 100 μm. No contact cytotoxicity effects were observed with either the decellularized scaffolds or polymeric sleeves. The extract cytotoxicity assay (I,J) was also performed with L929 murine fibroblasts with 80% (v/v) DMSO and RPMI 1640 medium serving as positive (C+) and negative (C−) controls, respectively. Luminescence was expressed as means (n = 6); error bars indicate 95% C.I.; MSD: minimum significant difference; indicates significant difference between the native and decellularized groups. Connectors indicate significant difference between originator and end arrow group.

FIGURE 6

Biomechanics of native PTA and decellularized PTA scaffolds treated with (M2) or without (M1) RNase/DNase and sterilized with either PAA or γ-irradiation, tested axially. (A) Elastic and collagen phase slope; (B) transition stress and ultimate tensile strength; (C) transition and failure strain; (D) sample thickness. Data shows means (n = 6); error bars indicate 95% C.I.; MSD: minimum significant difference; indicates significant difference between the native and the decellularized groups. Connectors indicate significant difference between originator and end arrow group.

FIGURE 7

(A–C) Biomechanics of non-sterilized and PAA-sterilized 100:50 and 150:50 polymeric sleeves, tested axially and circumferentially. (A) Primary and secondary phase slope; (B) failure strain and ultimate tensile strength (UTS); (C) sample thickness. Data shows means (n = 6); error bars indicate 95% C.I.; MSD: minimum significant difference. Connectors indicate significant difference between originator and end arrow group. (D) Sample stress-strain behavior of sleeve, showing primary, secondary and failure phases. (E,F) SEM of 100:50 (E) and 150:50 (F) sleeves (non-sterilized). Note the increased fiber thickness and pore size of 150:50 blend. Scale bars indicate 5 μm. (G) Burst pressure of PTA scaffolds treated with (M2) or without (M1) RNase/DNase and sterilized with either PAA or γ-irradiation, with and without PAA-sterilized 100:50 and 150:50 polymeric sleeves. Data shows means (n = 6); error bars indicate 95% C.I.; MSD: minimum significant difference. Connector indicates significant difference between originator and end arrow group.

FIGURE 8

Bacteria growth curves for the polymer blends (A–I) and decellularized PTA (J), inoculated with S. Aureus, S. Epidermidis or E. Coli. (A) 100:50A (5 mg/mL VAN and 5 mg/mL GEN supplemented in electrospinning solution); (B) 150:50A (5 mg/mL VAN and 5 mg/mL GEN supplemented in electrospinning solution); (C) 100:50-5V/5G-1H (100:50 soaked in 5 mg/mL VAN and 5 mg/mL GEN for 1 h); (D) 150:50-5V/5G-1H (150:50 soaked in 5 mg/mL VAN and 5 mg/mL GEN for 1 h); (E) 100:50-5V/5G-24H (100:50 soaked in 5 mg/mL VAN and 5 mg/mL GEN for 24 h); (F) 150:50-5V/5G-24H (150:50 soaked in 5 mg/mL VAN and 5 mg/mL GEN for 24 h); (G) 100:50-5V/8G-24H (100:50 soaked in 5 mg/mL VAN and 8 mg/mL GEN for 24 h); (H) 150:50-5V/8G-24H (150:50 soaked in 5 mg/mL VAN and 8 mg/mL GEN for 24 h); (I) 15:50-W/O (150:50 without antibiotics); (J) Decel-PTA-5V/8G-24H (M2-PAA decellularized PTA soaked in 5 mg/mL VAN and 8 mg/mL GEN for 24 h). (K) Mean bacteria inhibition area for the 100:50-5V/8G-24H, 150:50-5V/8G-24H and Decel-PTA-5V/8G-24H groups, inoculated with S. Aureus, S. Epidermidis or E. Coli, and cultured for 24 h in the disk diffusion assay. Data expressed as means (n = 3); error bars indicate 95% C.I.; MSD: minimum significant difference. Connectors indicate significant difference between originator and end arrow group.