**5. Biological properties of e-BC gel**

To investigate the potential of the e-BC gel for use as a cell culture scaffold, we measured the growth rates of human umbilical vein endothelial cells (HUVECs) cultured on the e-BC gel.

Mechanical and Biological Properties of

mean ± SD (n=4).

would show the same or later biodegradation profile *in vivo*.

Bio-Inspired Nano-Fibrous Elastic Materials from Collagen 269

previously described [40]. Figure 10 shows that e-BC gel was completely digested for 24 h as well as the e-SC gel. The e-BC gel degraded slightly later than the e-SC gel. This may be due to the low denaturation temperature of SC (19ºC). The physiological concentration range of collagenase is approximately 1 U/ml [41], therefore, the e-BC gels might show the slow degradation profile *in vivo* than the *in vitro* results. Additionally, e-SC gel gradually degrades 1 month after implantation in rat subcutaneous pouches [16], e-BC gel, therefore,

Fig. 9. SEM images of HUVECs cultured on e-BC gel for 1d (a) and 5d (b). Bars: 50 μm.

Fig. 10. In vitro degradation rate of e-BC gel and e-SC gel in collagenase solution. Values are

The cell number was evaluated by the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) test [13]. The MTT test is an established method to determine viable cell number by measuring the metabolic activity of cellular enzymes. Figure 8 shows steady increases in cell number with culture time on both the e-BC gel and tissue culture plate (TCP). There was a lag time before steady increase in cell number on the e-BC gel. Semler et al. reported that cells shows high growth rate on matrices with high mechanical compliance such as TCP, whereas the cells aggregate three-dimensionally on matrices with low mechanical compliance such as collagen gel [39]. Therefore, the difference in the mechanical properties of the surface of the culture substrate might affect the initial rate of cell growth.

Fig. 8. Growth curves of HUVECs cultured on e-BC gels and plastic tissue culture plate. Values are mean ± SD (n=4).

The SEM images show the distribution and spreading morphology of the HUVECs cultured on the e-BC gel at day 1 and day 5 after cultivation (Fig. 9). Apparently, the cells at 5d cultivation were confluent and showed cobblestone-like morphologies (Fig 9b). HUVECs also grow better on the e-SC gel and shows confluent at day 6 after cultivation [13]. Therefore, the e-BC gel could be used as a cell culture scaffold as well as the e-SC gel. EDC cross-links collagen molecules by the formation of isopeptides without being incorporated itself, thus precluding depolymerization and the possible release of potentially cytotoxic reagents. Furthermore, the by-product of the cross-linking reaction and un-reacted EDC in the e-BC gel should be completely removed by the drastic shrinkage in hot water. It is expected that the e-BC gel has good biocompatibility and no cytotoxicity.

To assess the degradability of the e-BC gel in collagenase solution (50 U/ml), protein content measurement was performed using a bicinchoninic acid protein assay kit as

The cell number was evaluated by the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) test [13]. The MTT test is an established method to determine viable cell number by measuring the metabolic activity of cellular enzymes. Figure 8 shows steady increases in cell number with culture time on both the e-BC gel and tissue culture plate (TCP). There was a lag time before steady increase in cell number on the e-BC gel. Semler et al. reported that cells shows high growth rate on matrices with high mechanical compliance such as TCP, whereas the cells aggregate three-dimensionally on matrices with low mechanical compliance such as collagen gel [39]. Therefore, the difference in the mechanical properties of the surface of the culture substrate might affect the initial rate of cell growth.

Fig. 8. Growth curves of HUVECs cultured on e-BC gels and plastic tissue culture plate.

expected that the e-BC gel has good biocompatibility and no cytotoxicity.

The SEM images show the distribution and spreading morphology of the HUVECs cultured on the e-BC gel at day 1 and day 5 after cultivation (Fig. 9). Apparently, the cells at 5d cultivation were confluent and showed cobblestone-like morphologies (Fig 9b). HUVECs also grow better on the e-SC gel and shows confluent at day 6 after cultivation [13]. Therefore, the e-BC gel could be used as a cell culture scaffold as well as the e-SC gel. EDC cross-links collagen molecules by the formation of isopeptides without being incorporated itself, thus precluding depolymerization and the possible release of potentially cytotoxic reagents. Furthermore, the by-product of the cross-linking reaction and un-reacted EDC in the e-BC gel should be completely removed by the drastic shrinkage in hot water. It is

To assess the degradability of the e-BC gel in collagenase solution (50 U/ml), protein content measurement was performed using a bicinchoninic acid protein assay kit as

Values are mean ± SD (n=4).

previously described [40]. Figure 10 shows that e-BC gel was completely digested for 24 h as well as the e-SC gel. The e-BC gel degraded slightly later than the e-SC gel. This may be due to the low denaturation temperature of SC (19ºC). The physiological concentration range of collagenase is approximately 1 U/ml [41], therefore, the e-BC gels might show the slow degradation profile *in vivo* than the *in vitro* results. Additionally, e-SC gel gradually degrades 1 month after implantation in rat subcutaneous pouches [16], e-BC gel, therefore, would show the same or later biodegradation profile *in vivo*.

Fig. 9. SEM images of HUVECs cultured on e-BC gel for 1d (a) and 5d (b). Bars: 50 μm.

Fig. 10. In vitro degradation rate of e-BC gel and e-SC gel in collagenase solution. Values are mean ± SD (n=4).

Mechanical and Biological Properties of

**6. Conclusion** 

**7. Acknowledgment** 

used for the fabrication of tissue-engineered vascular graft.

Bio-Inspired Nano-Fibrous Elastic Materials from Collagen 271

platelets was calculated according to the methods reported previously [14]. Figure 11 shows the number of platelets that adhered to the samples. The platelet number was estimated from the acid phosphatase reaction [47]. There was a linear relationship between the PRP concentration and the absorbance values at 405 nm, indicating that the acid phosphatase reaction of the platelets may be considered a reliable indicator of platelet number (data not shown). The results demonstrate that the platelet adhesion rates were markedly low on the e-BC gel when compared to the fibrinogen-coated or polystyrene surfaces. The e-SC gel also showed an adhesion rate as low as the e-BC gel. We are separately studying the fabrication of a vascular graft using the e-SC gel [16]. Consequently, the e-BC gel can potentially be

Considering that the platelets adhered better to the collagen-coated than to the gelatincoated surface [14], the anticoagulant ability of the e-BC gel may have been due to heat denaturation. The e-BC gel was prepared by heat treatment at 60°C resulting in collagen denaturation (gelatinization). Polanowska-Grabowska and coworkers reported that the platelet adhesion rate on a gelatin-coated surface was lower than on collagen- coated or fibrinogen-coated surfaces [48]. However, blood coagulation is known to depend on material properties, such as surface-free energy, surface charge, and wettability; these properties govern protein adsorption involving platelet adhesion [49, 50]. Experiments using human whole blood are needed to test the clinical applications of the gels. Further

In conclusion, we successfully fabricated an elastic collagen material from EDC cross-linked BC fibrillar gel by heat treatment. "Bio-inspired crosslinking" used in this study involves collagen fibril formation in the presence of EDC as a crosslinking reagent, which was developed in an attempt to mimic the in vivo simultaneous occurrence of fibril formation and crosslinking. We successfully prepared the bio-inspired crosslinked BC gels by adjusting the NaCl and EDC concentrations during collagen fibril formation. An advantage of bio-inspired crosslinking is the achievement of homogenous intrafibrillar crosslinking as well as interfibrillar one, providing higher mechanical properties compared to the traditional sequential crosslinking in which monomeric collagen initially forms fibril, then subsequently crosslinked using chemical or physical methods. Another advantage is the elastic properties of bio-inspired crosslinked BC gels after heat treatment. Although common collagen materials dissolved in water at a temperature above their denaturation temperature, we found that the bio-inspired cross-linked BC gel drastically shrank at a high temperature without remarkable dissolution. The collagen gel obtained interestingly showed rubber-like elasticity and high stretchability. The human cells showed good attachment and proliferation on this elastic material, suggesting its potential to be utilized in biomaterials for tissue engineering. Additionally, the elastic material demonstrated excellent blood compatibility. Our future work will focus on fabrication of small-caliber tubes (inner diameter < 6 mm) for small-caliber vascular grafts and preclinical animal studies to further

This study was supported by Grants-in-Aid for Young Scientists (B) (20700393) from the

examinations are necessary to ensure the blood compatibility of the e-BC gel.

assess the safety and effectiveness of the collagen-based vascular grafts.

Ministry of Education, Science, and Culture, Japan.

Owing to the mechanical, biological, and biodegradable properties of the e-BC gel, it could potentially be used to engineer blood vessels *in vivo*. Synthetic materials such as polyethylene terephthalate (Dacron) and expanded polytetrafluoroethylene (ePTFE) have been clinically applied as vascular grafts for a long time to replace or bypass large-diameter blood vessels. However, when used in small-diameter blood vessels (inner diameter < 6 mm), the patency rates are poor compared to autologous vein grafts. These failures are due to early thrombosis and gradual neointimal hyperplasia, and the pathological changes occurred due to the lack of blood or mechanical compatibility of the synthetic grafts [42]. To address this problem, tissue engineering approach is promising. A variety of biodegradable polymers and scaffolds have been evaluated to develop a tissue-engineered vascular graft [43-46]. These approaches depend on either the *in vitro* or *in vivo* cellular remodeling of a polymeric scaffold. For successful *in vivo* cellular remodeling, the biocompatibility, biodegradability, and mechanical properties of the scaffold must be suitable to the dynamic environment of the blood vessel. Therefore, the ideal scaffold should employ a biocompatible and biodegradable polymer with elastic properties that interact favorably with cells and blood. Therefore, the interaction of the e-BC gel with rat whole blood and plasma was investigated to assess their blood compatibility for use in vascular-tissue engineering.

Fig. 11. Platelet adhesion rates on the e-BC gel, e-SC gel, and the control samples. Values are mean ± SD (n=4).

After incubation of the e-BC gel with platelet-rich plasma (PRP) collected from rat blood, the colored *p*-nitrophenol produced by the acid-phosphatase reaction of the platelets was measured with a microplate reader at an absorbance of 405 nm. The percentage of adherent platelets was calculated according to the methods reported previously [14]. Figure 11 shows the number of platelets that adhered to the samples. The platelet number was estimated from the acid phosphatase reaction [47]. There was a linear relationship between the PRP concentration and the absorbance values at 405 nm, indicating that the acid phosphatase reaction of the platelets may be considered a reliable indicator of platelet number (data not shown). The results demonstrate that the platelet adhesion rates were markedly low on the e-BC gel when compared to the fibrinogen-coated or polystyrene surfaces. The e-SC gel also showed an adhesion rate as low as the e-BC gel. We are separately studying the fabrication of a vascular graft using the e-SC gel [16]. Consequently, the e-BC gel can potentially be used for the fabrication of tissue-engineered vascular graft.

Considering that the platelets adhered better to the collagen-coated than to the gelatincoated surface [14], the anticoagulant ability of the e-BC gel may have been due to heat denaturation. The e-BC gel was prepared by heat treatment at 60°C resulting in collagen denaturation (gelatinization). Polanowska-Grabowska and coworkers reported that the platelet adhesion rate on a gelatin-coated surface was lower than on collagen- coated or fibrinogen-coated surfaces [48]. However, blood coagulation is known to depend on material properties, such as surface-free energy, surface charge, and wettability; these properties govern protein adsorption involving platelet adhesion [49, 50]. Experiments using human whole blood are needed to test the clinical applications of the gels. Further examinations are necessary to ensure the blood compatibility of the e-BC gel.
