Tissue engineered bovine saphenous vein extracellular matrix scaffolds produced via antigen removal achieve high in vivo patency rates

Abstract

Diseases of small diameter blood vessels encompass the largest portion of cardiovascular diseases, with over 4.2 million people undergoing autologous vascular grafting every year. However, approximately one third of patients are ineligible for autologous vascular grafting due to lack of suitable donor vasculature. Acellular extracellular matrix (ECM) scaffolds derived from xenogeneic vascular tissue have potential to serve as ideal biomaterials for production of off-the-shelf vascular grafts capable of eliminating the need for autologous vessel harvest. A modified antigen removal (AR) tissue process, employing aminosulfabetaine-16 (ASB-16) was used to create off-the-shelf small diameter (< 3 mm) vascular graft from bovine saphenous vein ECM scaffolds with significantly reduced antigenic content, while retaining native vascular ECM protein structure and function. Elimination of native tissue antigen content conferred graft-specific adaptive immune avoidance, while retention of native ECM protein macromolecular structure resulted in pro-regenerative cellular infiltration, ECM turnover and innate immune self-recognition in a rabbit subpannicular model. Finally, retention of the delicate vascular basement membrane protein integrity conferred endothelial cell repopulation and 100% patency rate in a rabbit jugular interposition model, comparable only to Autograft implants. Alternatively, the lack of these important basement membrane proteins in otherwise identical scaffolds yielded a patency rate of only 20%. We conclude that acellular antigen removed bovine saphenous vein ECM scaffolds have potential to serve as ideal off-the-shelf small diameter vascular scaffolds with high in vivo patency rates due to their low antigen content, retained native tissue basement membrane integrity and preserved native ECM structure, composition and functional properties. STATEMENT OF SIGNIFICANCE: The use of autologous vessels for the treatment of small diameter vascular diseases is common practice. However, the use of autologous tissue poses significant complications due to tissue harvest and limited availability. Developing an alternative vessel for use for the treatment of small diameter vessel diseases can potentially increase the success rate of autologous vascular grafting by eliminating complications related to the use of autologous vessel and increased availability. This manuscript demonstrates the potential of non-antigenic extracellular matrix (ECM) scaffolds derived from xenogeneic vascular tissue as off-the-shelf vascular grafts for the treatment of small diameter vascular diseases.

Keywords: Extracellular matrix; Small diameter vessel; Tissue engineering.

Fig. 1.

Cellularity and content of unknown minor histocompatibility antigens in saphenous vein ECM scaffolds. (A) Both AR and SDS-decellularization processing methods were successful at decellularizing SV scaffolds determined by the lack of visible nuclei. Scale bar 200 μm. (B) Similarly, both processing methods resulted in scaffolds with significantly low levels of DNA content compared to untreated SV. (C) Via western blot, AR (Lane-2) was capable of significantly reducing both hydrophilic (D) and lipophilic (E) unknown minor histocompatibility antigens when compared to untreated SV (Lane–1). SDS (Lane–3) decellularization also reduced both hydrophilic and lipophilic antigens, however only lipophilic reduction was statistically different from untreated SV. Lack of statistical significance of hydrophilic antigen reduction for SDS scaffolds may be related to the nature of the non-parametric statistical tests utilized (Wilcoxon/Kruskal-Wallis Test with Dunn post-hoc analysis on non-parametric medians). * = p <0.05, ** = p < 0.01, *** = p < 0.001.

Fig. 2.

Macromolecular structure and retention of basement membrane integrity in SV ECM scaffolds. (A) AR preserved SV collagen organization (scale bar 100 μm), collagen IV (B, scale bar 200 μm) and laminin (C, scale bar 200 μm) content and organization, when compared to untreated SV. (A) Conversely, SDS-decellularization resulted in the disruption of abluminal collagen organization and a decrease in collagen IV (B) and laminin (C) content. (D, E, scale bar 200 μm). Similarly, AR retained the collagen alignment of untreated SV, while SDS scaffolds significantly decreased collagen alignment. Wilcoxon/Kruskal-Wallis Test with Dunn post-hoc analysis on non-parametric medians. * = p < 0.05.

Fig. 3.

In vivo immune and pro-regenerative response towards subpannicularly implanted 1 × 1 cm scaffolds. Allografts implants were used as negative immunogenic controls. (A) Representative histological images of the cut edges of explanted scaffolds. Qualitatively, lymphocyte presence was comparable between AR and Allograft groups, whereas greater lymphocyte presence was identified in untreated or SDS SV scaffolds. (CD3 positive lymphocytes = red, DAPI stained nuclei = blue). Scale bar 100 μm.(B) H&E stained images demonstrate a higher number of non-lymphocytic, predominantly spindle shape cell infiltration into AR-scaffold when compared to untreated SV and SDS-scaffolds (marked with an * in AR-scaffold, B, scale bar 200 μm. Dotted lines indicate location of scaffold), C (scale bar 100 μm), D (scale bar 50 μm), E (scale bar25μm). (F) Presence of these non-immune spindle shaped cells in AR scaffolds was associated with collagen reabsorption and scaffold turnover (newly formed collagen does not polarize-area within dashed lines in AR-scaffold) (Small dashed line represents original scaffold location (all groups), large dash line (AR group only) represents scaffoldcore which has yet to be turned over, continuing to polarize at pre-implantation levels). No evidence of scaffold turnover was seen in the other groups as evidenced by nochange in polarization compared to pre-implantation levels. Scale bar 500 μm. (G) Quantification of lymphocyte infiltration demonstrated that AR scaffolds significantly reduced graft-specific CD3 positive cell presence when compared to untreated SV. (H) The fibrotic encapsulation present in AR-scaffolds and untreated SV scaffolds was comparable to that of Allografts, while that found around the SDS-decellularized scaffolds was statistically higher than Allografts. (I) Quantification of infiltrating non-immunogenic spindle cells demonstrated AR scaffolds were not significantly different from allograft, whereas both untreated SV and SDS scaffolds had significantly lower cell infiltration. Wilcoxon/Kruskal-Wallis Test with Dunn post-hoc analysis on non-parametric medians. * = p < 0.05, ** = p < 0.01

Fig. 4.

Recellularization capacity and cell modulatory capacity of SV ECM scaffolds. Both BM (A, B) and NBM (C,D) AR scaffolds ret ained their recellularization capacity determined by the presence of healthy endothelial cell layer at day 8 post seeding. Alternatively, the presence of endothelial cells in both BM (E, F) and NBM (G, H) SDS scaffolds was significantly diminished by day 8 post-seeding. Additionally, the presence of basement membrane proteins in AR BM (A, B) scaffolds modulated endothelial cells to spread more when compared to the cells on the other groups (original seeding shape demonstrate with white circle). Scale bar 2000 μm.

Fig. 5.

Gross pathology of in vivo jugular interposition grafts and associated humoral adaptive immune response. (A) Representative intraoperative surgical images of the jugular vein (left) and interposition graft (right) immediately after implantation. (B) Representative images of the interposition sites immediately before explantation (day 35), showing the graft location (arrows) and development of collateral vessels (arrow heads). (C) Representative cross-sectional image of the graft lumen in the mid-graft region for each group (ruler marks = 1 mm). Organized pre-mortem intraluminal thrombus is evident in the untreated SV and non-basement membrane groups (arrows), whereas the allograft and basement membrane groups have no evidence of thrombus formation. (D) Scaffolds in the Autograft and BM groups resulted in high patency rates, whereas untreated SV scaffolds and NBM scaffolds resulted in low patency rates. (E) Graft specific antibody production towards all bovine scaffolds is first significantly different from autograft on day 21. Antibody production towards AR-scaffolds starts to plateau by day 28, while the antibody production towards untreated SV scaffolds continues to increase. Both AR groups stimulate significantly less graft-specific antibody production than that toward untreated SV scaffolds by day 35. Groups not connected by lower case letters are statistically different. Repeated measures two-way ANOVA and Tukey-Kramer HSD post-hoc analyses on standard least squares means.

Fig. 6.

Rabbit in vivo response towards jugular interposition scaffolds. (A) All groups were evaluated for luminal area via histological assessment of H&E stained images. Scale bar 10 0 0 μm. (B) Presence of BM facing the vascular lumen maintains mid graft lumen area at levels equivalent to those of autograft. Conversely, absence of basement membrane components (NBM) resulted in significantly reduced mid-graft lumen area compared to basement membrane presence (BM) and Autograft groups. (C,D) Immune cell infiltration was assessed in all grafts via PCR. AR-scaffolds in both groups (BM and NBM) resulted with significant lymphocyte (C) and macrophage infiltration (D)when compared to Autografts. No significant difference in the number of infiltrating lymphocytes or macrophages was identified between AR-scaffolds (BM and non-BM) and untreated SV tissue. (E–K) Subtype of lymphocytes and macrophages infiltrating the scaffolds was determined by real-time qPCR. CD3 (E), CD4 (F) and CD8 (G) genes were run to determine the amount of lymphocytes, while FOXP3 (H) was run to determine the presence of regulatory T cells (T_reg). Gene CD163 (I) was used to determine the total amount of macrophages, while TNF a (J) was used to determine M1 macrophage polarization and MRC1 (K) the presence of M2 polarized macrophages. Although no significant differences were found between groups for any of the immune cellular subtypes, definitive conclusions regarding the specific type of infiltrating immune cell could not be reached due to high data variability. * = p < 0.05, ** = p < 0.01.

Fig. 7.

In vivo recellularization of scaffolds implanted in jugular interposition model. Scaffolds recellularization capability was determined by the degree of cell repopulation of scaffolds as seen in H&E histological images, platelet endothelial cell adhesion molecule (PECAM-1) staining and its quantification. (A) BM scaffolds experienced higher number of cell repopulation on the luminal surface and within the scaffold tunica media, whereas untreated SV scaffolds and NBM scaffolds resulted in very low overall cellular repopulation. It is possible some cell repopulation occurred in the Autograft scaffolds, however the majority of the cells were likely present before the surgical procedure. Scale bar 100 μm. (B) Positive PECAM-1 staining indicates some of the cells present in the luminal surface of BM scaffolds are endothelial cells. The difference in cells morphology and PECAM-1 secretion pattern of endothelial cell in the BM scaffolds from those in the Autograft scaffolds may be due to the short timeframe in vivo implantation. Scale bar 10 μm. (C) BM scaffolds experienced the highest number of endothelial cell repopulation in the lumen of the scaffold when compared to the other bovine groups. Due to the non-parametric statistics the number of endothelial cells found in the lumen of the BM scaffolds was not statistically different than that of the Autograft group although the absolute number is noticeably lower. Wilcoxon/Kruskal-Wallis Test with Dunn post-hoc analysis on non-parametric medians. * = p < 0.05, ** = p < 0.01, *** = p < 0.001.