Fabrication and biomechanical characterization of a spider silk reinforced fibrin-based vascular prosthesis.

With fibrin-based vascular prostheses, vascular tissue engineering offers a promising approach for the fabrication of biologically active regenerative vascular grafts. As a potentially autologous biomaterial, fibrin exhibits excellent hemo- and biocompatibility. However, the major problem in the use of fibrin constructs in vascular tissue engineering, which has so far prevented their widespread clinical application, is the insufficient biomechanical stability of unprocessed fibrin matrices. In this proof-of-concept study, we investigated to what extent the addition of a spider silk network into the wall structure of fibrin-based vascular prostheses leads to an increase in biomechanical stability and an improvement in the biomimetic elastic behavior of the grafts. For the fabrication of hybrid prostheses composed of fibrin and spider silk, a statically cast tubular fibrin matrix was surrounded with an envelope layer of Trichonephila edulis silk using a custom built coiling machine. The fibrin matrix was then compacted and pressed into the spider silk network by transluminal balloon compression. This manufacturing process resulted in a hybrid prosthesis with a luminal diameter of 4 mm. Biomechanical characterization revealed a significant increase in biomechanical stability of spider silk reinforced grafts compared to exclusively compacted fibrin segments with a mean burst pressure of 362 ± 74 mmHg vs. 213 ± 14 mmHg (p < 0.05). Dynamic elastic behavior of the spider silk reinforced grafts was similar to native arteries. In addition, the coiling with spider silk allowed a significant increase in suture retention strength and resistance to external compression without compromising the endothelialization capacity of the grafts. Thus, spider silk reinforcement using the abluminal coiling technique represents an efficient and reproducible technique to optimize the biomechanical behavior of small-diameter fibrin-based vascular grafts.

Pressure-compacted and spider silk-reinforced fibrin demonstrates sufficient biomechanical stability as cardiac patch in vitro.

OBJECTIVE: The generation of bio-/hemocompatible cardiovascular patches with sufficient stability and regenerative potential remains an unmet goal. Thus, the aim of this study was the generation and in vitro biomechanical evaluation of a novel cardiovascular patch composed of pressure-compacted fibrin with embedded spider silk cocoons. METHODS: Fibrin-based patches were cast in a customized circular mold. One cocoon of Nephila odulis spider silk was embedded per patch during the casting process. After polymerization, the fibrin clot was compacted by 2 kg weight for 30 min resulting in thickness reduction from up to 2 cm to <1 mm. Tensile strength and burst pressure was determined after 0 weeks and 14 weeks of storage. A sewing strength test and a long-term load test were performed using a customized device to exert physiological pulsatile stretching of a silicon surface on which the patch had been sutured. RESULTS: Fibrin patches resisted supraphysiological pressures of well over 2000 mmHg. Embedding of spider silk increased tensile force 1.8-fold and tensile strength 1.45-fold (p < .001), resulting in a final strength of 1.07 MPa and increased sewing strength. Storage for 14 weeks decreased tensile strength, but not significantly and suturing properties of the spider silk patches were satisfactory. The long-term load test indicated that the patches were stable for 4 weeks although slight reduction in patch material was observed. CONCLUSION: The combination of compacted fibrin matrices and spider silk cocoons may represent a feasible concept to generate stable and biocompatible cardiovascular patches with regenerative potential.

Chemically induced hypoxia by dimethyloxalylglycine (DMOG)-loaded nanoporous silica nanoparticles supports endothelial tube formation by sustained VEGF release from adipose tissue-derived stem cells.

Inadequate vascularization leading to insufficient oxygen and nutrient supply in deeper layers of bioartificial tissues remains a limitation in current tissue engineering approaches to which pre-vascularization offers a promising solution. Hypoxia triggering pre-vascularization by enhanced vascular endothelial growth factor (VEGF) expression can be induced chemically by dimethyloxalylglycine (DMOG). Nanoporous silica nanoparticles (NPSNPs, or mesoporous silica nanoparticles, MSNs) enable sustained delivery of molecules and potentially release DMOG allowing a durable capillarization of a construct. Here we evaluated the effects of soluble DMOG and DMOG-loaded NPSNPs on VEGF secretion of adipose tissue-derived stem cells (ASC) and on tube formation by human umbilical vein endothelial cells (HUVEC)-ASC co-cultures. Repeated doses of 100 µM and 500 µM soluble DMOG on ASC resulted in 3- to 7-fold increased VEGF levels on day 9 (P < 0.0001). Same doses of DMOG-NPSNPs enhanced VEGF secretion 7.7-fold (P < 0.0001) which could be maintained until day 12 with 500 µM DMOG-NPSNPs. In fibrin-based tube formation assays, 100 µM DMOG-NPSNPs had inhibitory effects whereas 50 µM significantly increased tube length, area and number of junctions transiently for 4 days. Thus, DMOG-NPSNPs supported endothelial tube formation by upregulated VEGF secretion from ASC and thus display a promising tool for pre-vascularization of tissue-engineered constructs. Further studies will evaluate their effect in hydrogels under perfusion.

A 3-Layered Bioartificial Blood Vessel with Physiological Wall Architecture Generated by Mechanical Stimulation.

The generation of cellularized bioartificial blood vessels resembling all three layers of the natural vessel wall with physiological morphology and cell alignment is a long pursued goal in vascular tissue engineering. Simultaneous culture of all three layers under physiological mechanical conditions requires highly sophisticated perfusion techniques and still today remains a key challenge. Here, three-layered bioartificial vessels based on fibrin matrices were generated using a stepwise molding technique. Adipose-derived stem cells (ASC) were differentiated to smooth muscle cells (SMC) and integrated in a compacted tubular fibrin matrix to resemble the tunica media. The tunica adventitia-equivalent containing human umbilical vein endothelial cells (HUVEC) and ASC in a low concentration fibrin matrix was molded around it. Luminal seeding with HUVEC resembled the tunica intima. Subsequently, constructs were exposed to physiological mechanical stimulation in a pulsatile bioreactor for 72 h. Compared to statically incubated controls, mechanical stimulation induced physiological cell alignment in each layer: Luminal endothelial cells showed longitudinal alignment, cells in the media-layer were aligned circumferentially and expressed characteristic SMC marker proteins. HUVEC in the adventitia-layer formed longitudinally aligned microvascular tubes resembling vasa vasorum capillaries. Thus, physiologically organized three-layered bioartificial vessels were successfully manufactured by stepwise fibrin molding with subsequent mechanical stimulation.

Perfusion promotes endothelialized pore formation in high concentration fibrin gels otherwise unsuitable for tube development.

Vascularization of tissue engineered implants is crucial for their survival and integration in the recipient's body. Pre-vascularized, fibrin-based implants offer a solution since low concentration fibrin hydrogels (1 mg/mL) have been shown to promote tube formation of endothelial cells in co-culture with adipogenic stem cells. However, higher fibrinogen concentrations (> 20 mg/mL) enabling the fabrication of stable implants are necessary.We here characterized fibrin gels of 1-30 mg/mL for their rheological properties and whether they support tube formation of endothelial cell-adipogenic stem cell co-cultures for up to 7 days. Moreover, 20 mg/mL gels containing preformed channels and endothelial cell-adipogenic stem cell co-culture were perfused continuously in a customized flow chamber with 3.9 dyn/cm2 for 12 days and analyzed for capillary formation.Rheology of fibrin gels showed increasing stability proportional to fibrinogen concentration with 20 mg/mL gels having a storage module of 465 Pa. Complex tube networks stable for 7 days were observed at 1-5 mg/mL gels whereas higher concentrations showed initial sprouting only. However, perfusion of 20 mg/mL fibrin gels resulted in endothelialized pore formation in several layers of the gel with endothelial cell-adipogenic stem cell co-culture.Thus, perfusion supports the formation of capillary-like structures in fibrin gels that are too dense for spontaneous tube formation under static conditions. Future studies are necessary to further increase pore density and to investigate proper nutrition of tissue-specific target cells in the scaffold.

Complete Myogenic Differentiation of Adipogenic Stem Cells Requires Both Biochemical and Mechanical Stimulation.

Vascular tissue engineering of the middle layer of natural arteries requires contractile smooth muscle cells (SMC) which can be differentiated from adipose-derived mesenchymal stem cells (ASC) by treatment with transforming growth factor-ß, sphingosylphosphorylcholine and bone morphogenetic protein-4 (TSB). Since mechanical stimulation may support or replace TSB-driven differentiation, we investigated its effect plus TSB-treatment on SMC orientation and contractile protein expression. Tubular fibrin scaffolds with incorporated ASC or pre-differentiated SMC were exposed to pulsatile perfusion for 10 days with or without TSB. Statically incubated scaffolds served as controls. Pulsatile incubation resulted in collagen-I expression and orientation of either cell type circumferentially around the lumen as shown by alpha smooth muscle actin (αSMA), calponin and smoothelin staining as early, intermediate and late marker proteins. Semi-quantitative Westernblot analyses revealed strongly increased αSMA and calponin expression by either pulsatile (12.48-fold; p < 0.01 and 38.15-fold; p = 0.07) or static incubation plus TSB pre-treatment (8.91-fold; p < 0.05 and 37.69-fold; p < 0.05). In contrast, contractility and smoothelin expression required both mechanical and TSB stimulation since it was 2.57-fold increased (p < 0.05) only by combining pulsatile perfusion and TSB. Moreover, pre-differentiation of ASC prior to pulsatile perfusion was not necessary since it could not further increase the expression level of any marker.

Biochemical Myogenic Differentiation of Adipogenic Stem Cells Is Donor Dependent and Requires Sound Characterization.

IMPACT STATEMENT: We here showed that even under optimized conditions for biochemical differentiation of adipose-derived stem cells (with respect to a pronounced marker protein expression for a reasonable period of time) it was not possible to obtain functional smooth muscle cells from all donors. Moreover, an underestimated role may play the effect of the scaffold material on smooth muscle cell functionality. Both aspects are crucial for the successful tissue engineering of the vascular medial layer combining autologous cells with a suitable scaffold material and thus should be thoroughly addressed in each individualized therapeutic approach.

Biocompatibility of Intensified Decellularized Equine Carotid Arteries in a Rat Subcutaneous Implantation Model and in a Human In Vitro Model.

Limited biocompatibility of decellularized scaffolds is an ongoing challenge in tissue engineering. We recently demonstrated that intensified detergent-based decellularization of equine carotid artery (dEACintens) removed residual cellular molecules from the scaffold more efficiently than a conventional decellularization (dEACcon), although this approach did not eliminate its immunogenicity entirely. CCN1 has been shown to improve biocompatibility of dEACcon in a sheep model. In this study, we tested the biocompatibility of dEACintens and dEACcon with or without CCN1 coating after subcutaneous implantation in rats for up to 12 weeks. Explants were assessed by conventional histopathology and immunostaining for infiltrating M2 macrophages. Moreover, human macrophages derived from monocytes (MDM) or THP-1 cells (THP-derived macrophages [TDM]) were seeded onto dEACcon and dEACintens, and activation was assessed either by cytokine expression or matrix metalloprotease 2 and 7 staining. dEACintens showed a significantly reduced inflammatory infiltration (52%; p < 0.0001), as well as an earlier and denser neovascularization (1.4-fold, p < 0.0001) independent of CCN1 coating, which, however, reduced fibrosis exclusively with dEACintens (26-53%; p < 0.05). Human MDM seeded for 48 h onto dEACintens showed higher transcript levels for anti-inflammatory IL-10 (2.3-fold), proinflammatory TNFα (2.2-fold), and macrophage/monocyte recruiting MIP1α (3.5-fold; all p < 0.05) and MCP (2.7-fold; p < 0.01), whereas 1.92-fold more TDM on dEACintens showed staining for MMP2 (p > 0.001). Thus, although being advantageous in regard to fibrosis, CCN1 coating of dEACintens does not appear to be necessary for further improving dEACintens excellent biocompatibility in rats. In humans, the unspecific cellular immune response toward dEACintens seemed to be more complex, but generally comparable to the mild acute inflammatory tissue reaction with high remodeling activity as observed after rat subcutaneous implantation.

Antibody formation towards porcine tissue in patients implanted with crosslinked heart valves is directed to antigenic tissue proteins and αGal epitopes and is reduced in healthy vegetarian subjects.

BACKGROUND: Glutaraldehyde-fixed porcine heart valves (ga-pV) are one of the most frequently used substitutes for insufficient aortic and pulmonary heart valves which, however, degenerate after 10-15 years. Yet, xeno-immunogenicity of ga-pV in humans including identification of immunogens still needs to be investigated. We here determined the immunogenicity of ga-pV in patients with respect to antibody formation, identity of immunogens and potential options to reduce antibody levels. METHODS: Levels of tissue-specific and anti-αGal antibodies were determined retrospectively in patients who received ga-pV for 51 months (n=4), 25 months (n=6) or 5 months (n=4) and compared to age-matched untreated subjects (n=10) or younger subjects with or without vegetarian diet (n=12/15). Immunogenic proteins were investigated by Western blot approaches. RESULTS: Tissue-specific antibodies in patients were elevated after 5 (1.73-fold) and 25 (1.46-fold, both P<.0001) months but not after 51 months, whereas anti-Gal antibodies were induced 4.75-fold and 3.66-fold after 5 and 25 months (both P<.0001) and still were significantly elevated after 51 months (2.85-fold, P<.05). Western blots of porcine valve extracts with and without enzymatic deglycosylation revealed strong specific staining at ≈65 and ≈140 kDa by patient sera in either group which were identified by 2D Western blots and mass spectrometry as serum albumin and collagen 6A1. Vegetarian diet reduced significantly (0.63-fold, P<.01) the level of pre-formed αGal but not of tissue-specific antibodies. CONCLUSION: Immune response in patients towards ga-pV is induced by the porcine proteins albumin and collagen 6A1 as well as αGal epitopes, which seemed to be more sustained. In contrast, in healthy young subjects pre-formed anti-Gal antibodies were reduced by a meat-free nutrition.

Effect of Intensified Decellularization of Equine Carotid Arteries on Scaffold Biomechanics and Cytotoxicity.

Decellularized equine carotid arteries (dEAC) are suggested to represent an alternative for alloplastic vascular grafts in haemodialysis patients to achieve vascular access. Recently it was shown that intensified detergent treatment completely removed cellular components from dEAC and thereby significantly reduced matrix immunogenicity. However, detergents may also affect matrix composition and stability and render scaffolds cytotoxic. Therefore, intensively decellularized carotids (int-dEAC) were now evaluated for their biomechanical characteristics (suture retention strength, burst pressure and circumferential compliance at arterial and venous systolic and diastolic pressure), matrix components (collagen and glycosaminoglycan content) and indirect and direct cytotoxicity (WST-8 assay and endothelial cell seeding) and compared with native (n-EAC) and conventionally decellularized carotids (con-dEAC). Both decellularization protocols comparably reduced matrix compliance (venous pressure compliance: 32.2 and 27.4% of n-EAC; p < 0.01 and arterial pressure compliance: 26.8 and 23.7% of n-EAC, p < 0.01) but had no effect on suture retention strength and burst pressure. Matrix characterization revealed unchanged collagen contents but a 39.0% (con-dEAC) and 26.4% (int-dEAC, p < 0.01) reduction of glycosaminoglycans, respectively. Cytotoxicity was not observed in either dEAC matrix which was also displayed by an intact endothelial lining after seeding. Thus, even intensified decellularization generates matrix scaffolds highly suitable for vascular tissue engineering purposes, e.g., the generation of haemodialysis shunts.

The immune response to crosslinked tissue is reduced in decellularized xenogeneic and absent in decellularized allogeneic heart valves.

BACKGROUND: The degeneration and failure of xenogeneic heart valves, such as the Matrix P Plus valve (MP-V) consisting of decellularized porcine valves (dec-pV) and equine glutaraldehyde-fixed conduits (ga-eC) have been linked to tissue immunogenicity accompanied by antibody formation. In contrast, decellularized allograft valves (dec-aV) are well-tolerated. Here, we determined tissue-specific antibody levels in patients after implantation of MP-V or dec-aV and related them to valve failure or time period after implantation. METHODS AND RESULTS: Specific antibodies toward whole tissue-homogenates or alphaGal were determined retrospectively by ELISA analyses from patients who received MP-V with an uneventful course of 56.1 ± 5.1 months (n = 15), or with valve failure after 25.3 ± 14.6 months (n = 3), dec-aV for various times from 4 to 46 months (n = 14, uneventful) and from healthy controls (n = 4). All explanted valves were assessed histopathologically.MP-V induced antibodies toward both tissue components with significantly higher levels toward ga-eC than toward dec-pV (68.7 and 26.65 µg/ml IgG). In patients with valve failure, levels were not significantly higher and were related to inflammatory tissue infiltration. Anti-Gal antibodies in MP-V patients were significantly increased in both, the uneventful and the failure group. In contrast, in dec-aV patients only a slight tissue-specific antibody formation was observed after 4 months (6.24 µg/ml) that normalized to control levels after 1 year. CONCLUSIONS: The strong humoral immune response to glutaraldehyde-fixed tissues is reduced in decellularized xenogeneic valves and almost absent in decellularized allogeneic tissue up to 4.5 years after implantation.

Immunogenicity of intensively decellularized equine carotid arteries is conferred by the extracellular matrix protein collagen type VI.

The limited biocompatibility of decellularized scaffolds is an ongoing challenge in tissue engineering. Here, we demonstrate the residual immunogenicity of an extensively decellularized equine carotid artery (dEAC(intens)) and identify the involved immunogenic components. EAC were submitted to an elaborated intensified decellularization protocol with SDS/sodium desoxycholate for 72 h using increased processing volumes (dEAC(intens)), and compared to dEAC(ord) prepared by an ordinary protocol (40 h, normal volumes). Matrix integrity was checked via correlative volumetric visualization which revealed only minor structural changes in the arterial wall. In dEAC(intens), a substantial depletion of cellular components was obvious for smooth muscle actin (100%), MHC I complexes (97.8%), alphaGal epitops (98.4% and 91.3%) and for DNA (final concentration of 0.34 ± 0.16 ng/mg tissue). However, dEAC(intens) still evoked antibody formation in mice after immunization with dEAC(intens) extracts, although to a lower extent than dEAC(ord). Mouse plasma antibodies recognized a 140 kDa band which was revealed to contain collagen VI alpha1 and alpha2 chains via mass spectrometry of both 2D electrophoretically separated and immunoprecipitated proteins. Thus, even the complete removal of cellular proteins did not yield non-immunogenic dEAC as the extracellular matrix still conferred immunogenicity by collagen VI. However, as lower antibody levels were achieved by the intensified decellularization protocol, this seems to be a promising basis for further development.

Coating decellularized equine carotid arteries with CCN1 improves cellular repopulation, local biocompatibility, and immune response in sheep.

Decellularized equine carotid arteries (dEAC) are potential alternatives to alloplastic vascular grafts although there are certain limitations in biocompatibility and immunogenicity. Here, dEAC were coated with the matricellular protein CCN1 and evaluated in vitro for its cytotoxic and angiogenic effects and in vivo for cellular repopulation, local biocompatibility, neovascularization, and immunogenicity in a sheep model. CCN1 coating resulted in nontoxic matrices not compromising viability of L929 fibroblasts and endothelial cells (ECs) assessed by WST-8 assay. Functionality of CCN1 was maintained as it induced typical changes in fibroblast morphology and MMP3 secretion. For in vivo testing, dEAC±CCN1 (n=3 each) and polytetrafluoroethylene (PTFE) protheses serving as controls (n=6) were implanted as cervical arteriovenous shunts. After 14 weeks, grafts were harvested and evaluated immunohistologically. PTFE grafts showed a patency rate of only 33% and lacked cellular repopulation. Both groups of bioartificial grafts were completely patent and repopulated with ECs and smooth muscle cells (SMCs). However, whereas dEAC contained only patch-like aggregates of SMCs and a partial luminal lining with ECs, CCN1-coated grafts showed multiple layers of SMCs and a complete endothelialization. Likewise, CCN1 coating reduced leukocyte infiltration and fibrosis and supported neovascularization. In addition, in a three-dimensional assay, CCN1 coating increased vascular tube formation in apposition to the matrix 1.6-fold. Graft-specific serum antibodies were increased by CCN1 up to 6 weeks after implantation (0.89±0.03 vs. 1.08±0.04), but were significantly reduced after 14 weeks (0.85±0.04 vs. 0.69±0.02). Likewise, restimulated lymphocyte proliferation was significantly lower after 14 weeks (1.78±0.09 vs. 1.32±0.09-fold of unstimulated). Thus, CCN1 coating of biological scaffolds improves local biocompatibility and accelerates scaffold remodeling by enhancing cellular repopulation and immunologic tolerance, making it a promising tool for generation of bioartificial vascular prostheses.

Cytotoxic effects of polyhexanide on cellular repopulation and calcification of decellularized equine carotids in vitro and in vivo.

PURPOSE: Disinfection of biological implants is indispensable for clinical safety. Here, decellularized equine carotid arteries (dECAs) were disinfected by polyhexanide (PHX), an effective, well-tolerated and nontoxic wound disinfectant and evaluated as vascular grafts for their repopulation and local biocompatibility in vivo. 
 METHODS: dECAs were terminally disinfected by a combination of 0.1% PHX and 70% ethanol (dECA_PHX-ET) or exclusively ethanol (dECA-ET) and subsequently implanted as arteriovenous shunts in sheep for 14 weeks. Repopulation was determined by immunohistochemistry for endothelial- (ECs) or smooth muscle cells (SMCs) using antibodies against CD31 and smooth muscle actin. Histological evaluation was performed on HE-stained sections. Cytotoxicity of dECAs was measured directly by seeding the scaffolds with L-929 fibroblasts, which were visualized by calcein staining. Indirect cytotoxicity was determined by WST-8 viability assay by incubation of L-929 with dECA extracts. 
 RESULTS: dECA_PHX-ET completely lacked repopulation with ECs and SMCs, showed leukocyte infiltration, strong calcification and poor neovascularization indicating insufficient biocompatibility and inflammatory graft degeneration. PHX-treatment reduced cell viability to 33.2 ± 12.6% and disturbed cell growth at direct contact. In contrast, dECA_ET had no direct cytotoxic effect and only slightly influenced cell viability (82.9 ± 12.5%), showed a substantial repopulation by ECs and SMCs including neovascularization, and were only slightly calcified. 
 CONCLUSION: The disinfectant polyhexanide seems to exert severe cytotoxic effects when used for the processing of decellularized matrices and may result in degenerative graft deterioration. In contrast, dECAs exclusively disinfected with ethanol were well integrated. Thus, ethanol seems to be a more suitable tool for graft processing than polyhexanide.

The effect of detergent-based decellularization procedures on cellular proteins and immunogenicity in equine carotid artery grafts.

Decellularized equine carotid arteries (dEAC) may represent a reasonable alternative to alloplastic materials in vascular replacement therapy. Acellularity of the matrix is standardly evaluated by DNA quantification what however may not record sufficiently the degree of matrix immunogenicity. Thus, our aim was to analyze dEAC with a low DNA content for residual cellular proteins. A detergent-based decellularization protocol including endonuclease treatment resulted in dEAC with 0.6 ± 0.15 ng DNA/mg dry weight representing 0.33 ± 0.14% of native tissue DNA content. In contrast, when matrices were homogenized and extracted by high detergent concentrations westernblot analyses revealed cytosolic and cytosceleton proteins like GAPDH and smooth muscle actin which were depleted to 4.1 ± 1.9% and 13.8 ± 0.55%, resp. Also putative immunogenic MHC I complexes and the alpha-Gal epitop were reduced to only 14.8 ± 1.2% and 15.1 ± 2.05%. Mass spectrometry of matrix extracts identified 306 proteins belonging to cytosol, organelles, nucleus and cell membrane. Moreover, aqueous matrix extracts evoked a pronounced antibody formation when administered in mice and thus display high immunogenic potential. Our data indicate that an established decellularization protocol which results in acellular matrices evaluated by low DNA content reduces but not eliminates cellular components which may contribute to its immunogenic potential in vivo.