Purified granular collagen (Col) derived from porcine skin was purchased from Nippon Meat Packers (Osaka, Japan). Terror peptides of collagen molecule that are antigenic determinants have been removed by enzyme degradation. Col (21.6 g) was dissolved in distilled water (1800 mL) under stirring with a screw propeller below 25˚C overnight to obtain Col aqueous solution (1.2 w/v%, pH 2.8). Aside from this, Col (21.6 g) was dissolved in distilled water (1800 mL) under stirring with a screw propeller below 25˚C overnight and then heated at 60˚C for 1 hour to obtain heat-denatured collagen aqueous solution (1.2 w/v%, pH 2.8). This heat-denatured collagen is referred to as Col’. The Col aqueous solution was mixed with the Col’ aqueous solution by stirring and then adjusted to the conditions of pH 4.5 by dropping sodium hydroxide aqueous solution to obtain a clear mixture.

First, the Col aqueous solution (800 ml) was mixed with the Col’ aqueous solution (400 mL). This mixture (2 mL) was poured into each of the 10 partitions (18 mm × 18 mm) of a plastic tray and refrigerated at 4˚C for 6 hours or more, and then was frozen at −85˚C overnight and followed by freeze-vacuum drying to obtain the sheet-shaped sponge. This sheet-shaped sponge was named as Sponge-CC’21. The composition ratio of Col and Col’ is 2/1. Second, the Col aqueous solution (600 mL) was mixed with the Col’ aqueous solution (600 mL). The sheet-shaped sponge was manufactured in a similar way. This sponge was named as Sponge-CC’11. The composition ratio of Col and Col’ is 1/1. Finally, the Col aqueous solution (400 ml) was mixed with the Col’ aqueous solution (800 mL). The sheet-shaped sponge was manufactured in a similar way. This sponge was named as Sponge-CC’12. The composition ratio of Col and Col’ is 1/2.

Both surfaces of the sheet-shaped sponge were irradiated using a 15 W ultraviolet lamp with a wavelength of 253.7 nm at a distance of 20 cm to introduce intermolecular crosslinks between collagen molecules. Both surfaces of Sponge-CC’21 were treated by UV irradiation for 5, 10, 15 minutes. They were named as Sponge-CC’21-5x2, Sponge-CC’21-10x2 and Sponge-CC’21-15x2, respectively. Both surfaces of Sponge-CC’11 were treated by UV irradiation for 5, 10, 15 minutes. They were named as Sponge-CC’11-5x2, Sponge-CC’11-10x2 and Sponge-CC’11-15x2, respectively. Both surfaces of Sponge-CC’12 were treated by UV irradiation for 5, 10, 15 minutes. They were named as Sponge-CC’12-5x2, Sponge-CC’12-10x2 and Sponge-CC’12-15x2, respectively.

The cylindrical sponge was manufactured in the same way as the sheet-shaped sponge. Col (8.0 g) was dissolved in distilled water (400 mL) to obtain Col aqueous solution (2.0 w/v%, pH 2.8). Aside from this, Col (4.0 g) was dissolved in distilled water (200 mL) and then heated at 60˚C for 1 hour to obtain Col’ aqueous solution (2.0 w/v%, pH 2.8). The Col aqueous solution was mixed with the Col’ aqueous solution by stirring and then adjusted to the conditions of pH 4.5 by dropping sodium hydroxide aqueous solution to obtain a clear mixture. This mixture (2 mL) was poured into each of the cylindrical shaped plastic bottle with a diameter of 10 mm and a depth of 30 mm and refrigerated at 4˚C for 6 hours or more, and then was frozen at −85˚C overnight and followed by freeze-vacuum drying to obtain the cylindrical sponge. This sponge was named as cy-Sponge-CC’21. The composition ratio of Col and Col’ is 2/1. Both sides of the cylindrical sponge were treated by UV irradiation for 15 minutes. This sponge was named as cy-Sponge-CC’21-15x2.

The water absorption of each sheet-shaped sponge was measured by immersing 10 sheet-shaped sponges in a plastic container filled with distilled water (200 mL) at 37˚C. After 10 minutes, the hydrated sponges were collected using a 10-mesh size stainless steel mesh and then placed on this mesh for 30 minutes to remove excess water and then weighed. The weight of 10 hydrated sponges was measured and compared with the original weight before immersion in water. Each experiment was performed using 8 containers. The average value was calculated by measuring the weight of the hydrated 10 sponges collected from each container. In a similar way, the water absorption of the cylindrical sponge was measured by immersing 5 cylindrical sponges in a plastic container filled with distilled water (200 mL) at 37˚C. After 10 minutes, the hydrated sponges were collected using a 10-mesh size stainless steel mesh and then placed on this mesh for 30 minutes to remove excess water and then weighed. The weight of 5 hydrated sponges was measured and compared with the original weight before immersion in water. This experiment was performed using 8 containers. The average value was calculated by measuring the weight of the hydrated 5 sponges collected from each container.

The weight change of sheet-shaped sponge was measured by immersing 10 sheet-shaped sponges in a plastic container filled with distilled water (200 mL) and placing it in an incubator at 37˚C for a predetermined time. The hydrated 10 sponges were collected using a 10-mesh size stainless steel mesh after immersion in water for 3 hour, 6 hour, 12 hour, 1 day, 2 days, 3 days, 5 days and 7 days and then returned to a plastic tray (80 mm × 50 mm). These hydrated sponges were frozen at −85˚C overnight and followed by freeze-vacuum drying to obtain dry sponges. The weight of 10 dry sponges was measured and compared with the original weight before immersion in water. Each experiment was performed using 8 containers. The average value was calculated by measuring the weight of 10 dry sponges collected from each container.

The weight change of cylindrical sponge was measured by immersing 5 cylindrical sponges in a plastic container filled with distilled water (200 mL) and placing it in an incubator at 37˚C for a predetermined time. The hydrated 5 sponges were collected using a 10-mesh size stainless steel mesh after immersion in water for 3 hour, 6 hour, 12 hour, 1 day, 3 days and 7 days and then returned to a plastic tray (80 mm × 50 mm). These hydrated sponges were frozen at −85˚C overnight, and followed by freeze-vacuum drying to obtain dry sponges. The weight of 5 dry sponges was measured and compared with the original weight before immersion in water. This experiment was performed using 4 containers. The average value was calculated by measuring the weight of 5 dry sponges collected from each container.

Collagenase powder was purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). The enzymatic degradation of sheet-shaped sponge by collagenase was investigated using Sponge-CC’21-15x2, Sponge-CC’11-15x2 and Sponge-CC’12-15x2 as typical samples. Collagenase (32 mg) was dissolved in distilled water (1600 mL). The enzyme concentration was adjusted to 0.002 w/v%. The weight change of each sheet-shaped sponge was measured by immersing 10 sheet-shaped sponges in each plastic container filled with distilled water containing collagenase (200 mL) and placing it in an incubator at 37˚C. The hydrated sponges were collected using a 10-mesh size stainless steel mesh after immersion for 1 day and then returned to a plastic tray (80 mm × 50 mm). These hydrated sponges were frozen at −85˚C overnight and followed by freeze-vacuum drying to obtain dry sponges. The weight of 10 dry sponges was measured and compared with the original weight. Each experiment was performed using 8 containers. The average value was calculated by measuring the weight of 10 dry sponges collected from each container.

In addition, the enzymatic degradation of cylindrical sponge by collagenase was measured using cy-Sponge-CC’21-15x2 in a similar way. The weight change of this cylindrical sponge was measured by immersing 5 cylindrical sponges in each container filled with distilled water containing collagenase (200 mL) and placing it in an incubator at 37˚C. The hydrated sponges were collected using a 10-mesh size stainless steel mesh after immersion for 1 day and then returned to a plastic tray (80 mm × 50 mm). These hydrated sponges were frozen at −85˚C overnight and followed by freeze-vacuum drying to obtain dry sponges. The weight of 5 dry sponges was measured and compared with the original weight. This experiment was performed using 8 containers. The average value was calculated by measuring the weight of 5 dry sponges collected from each container.

Data were expressed as means ± standard error. Statistical analysis was performed using Student’s t-test for comparison between two groups. Each experiment was performed 8 times to examine statistically significant differences in the measured values (n = 8).

As a preliminary experiment, the change in viscosity of Col aqueous solution due to temperature change was measured. Col dissolves in water under acidic conditions of pH 3 or less. However, precipitation occurs when pH value reaches 4 or higher under 25˚C conditions. Such a phenomenon can be considered as follows. The positively charged state of the side chains of Col molecules weakens at pH 4 or higher. As a result, it is thought that the binding force between the side chains of the Col molecules and the surrounding water molecules is weakened. Such conditions promote lateral alignment between rod-shaped Col molecules having a triple helical structure, thereby causing aggregation and precipitation. Once the acidic Col aqueous solution is heated to 40˚C or higher, Col molecules undergo a conformational change from triple helical structure to random structure. This phenomenon reflects a decrease in viscosity. The viscosity of Col aqueous solution was measured at the range from 20˚C to 60˚C using a standard viscometer. This viscometer rotates the cylindrical part in the solution and measures the rotating top surface resistance. The results of preliminary experiments showed that the viscosity of Col aqueous solution (1.2 w/v%) was 10.0, 9.2, 0.1 and 0.1 dPa·s at 20˚C, 30˚C, 40˚C and 60˚C, respectively. Once the Col aqueous solution was heated to 60˚C and then cooled to 20˚C, the viscosity was 0.4 dPa·s (Figure 1). This phenomenon indicates that the random structure does not return to the original helical structure. Such conditions inhibit lateral alignment between random-shaped molecules due to mutual steric hindrance. As a result, this heat-denatured Col’ aqueous solution does not precipitate even

in the neutral pH range. The mixed aqueous solution of Col and Col’ with a composition ratio of 1/2 causes partial precipitation at pH 5.5 or higher. The mixed aqueous solution of Col and Col’ with a composition ratio of 1/1 or 2/1 causes partial precipitation at pH 5.0 or higher. For this reason, the pH of these mixed aqueous solutions was set to 4.5 to obtain a clear mixture.

The sheet-shaped sponges were manufactured by freeze-vacuum drying this clear mixed aqueous solution of Col and Col’ under the condition of 1.2 w/v% collagen concentration at pH 4.5. The both surfaces of sheet-shaped sponge were irradiated by UV lamp to introduce crosslinks between collagen molecules. Sponge-CC’21 and Sponge-CC’11 are very soft and are suitable for application to soft mucosal and gingival defect wounds. On the other hand, Sponge-CC’12 is harder than other two sponges and is not suitable for application to soft mucosal and gingival defect wounds. The sheet-shaped sponge has a size of 17 mm × 17 mm and a thickness of 3 mm (Figure 2(a)).

The cylindrical sponge was manufactured by freeze-vacuum drying the mixed aqueous solution of Col and Col’ under the condition of 2.0 w/v% collagen concentration at pH 4.5. The cylindrical sponge shrank slightly after freeze-vacuum drying. The sponge could be easily removed from the cylindrical shaped plastic bottle container. This cy-Sponge-CC’21 showed sufficient physical properties to be used as filler for tooth extraction sockets. A small mark was placed on one part of the cylindrical sponge and then UV irradiation was performed on both sides to introduce crosslinks between collagen molecules. The cylindrical sponge has a diameter of 8 mm and a length of 25 mm (Figure 2(b)). The diameter of

the cylindrical sponge is matched to the diameter of tooth extraction socket. This sponge is inserted into the tooth extraction socket and compressed to match both size.

It is important that the sponge has high water absorption properties. It is a great merit to be able to absorb a large amount of exudates containing various cell growth factors. For this reason, the water absorption properties of collagen sponges were examined. Sheet-shaped sponges showed high water absorption when immersed in water at 37˚C for 10 minutes (Figure 3). Sponge-CC’12-15x2 absorbed 45 times more water than its own weight. Sponge-CC’11-15x2 absorbed 48 times more water than its own weight. Sponge-CC’21-15x2 absorbed 46 times more water than its own weight. On the other hand, cy-Sponge-CC’21-15x2 absorbed 30 times more water than its own weight. The sheet-shaped sponge was manufactured by freeze-vacuum drying the collagen aqueous solution at the concentration of 1.2 w/v%. On the contrary, the cylindrical sponge was manufactured by freeze-vacuum drying the collagen aqueous solution at the concentration of 2.0 w/v%. The cylindrical sponge has a denser structure than the sheet-shaped sponge. As a result, cy-Sponge-CC’21-15x2 showed a lower water absorption value than Sponge-CC’21-15x2.

It is important that collagen sponge is biodegraded and absorbed by the time

new tissue formation is completed. The weight change of the sponge when immersed in water is related to the degree of crosslinks between collagen molecules. The degree of crosslinks is related to UV irradiation time. For this reason, the weight changes of collagen sponges treated by different UV irradiation times were examined. The weight change of the sheet-shaped sponge was measured by immersing 10 sheet-shaped sponges in water at 37˚C (Figure 4(a)). The weight change rate of Sponge-CC’12 when immersed in water was measured. The weight change rate of Sponge-CC’12-15x2 was 70% of the original weight when immersed in water for 7 days. On the contrary, the weight change rate of Sponge-CC’12-5x2 was 32% of the original weight when immersed in water for 3 days (Figure 5). Next, the weight change of Sponge-CC’11 when immersed in water was measured. The weight change rate of Sponge-CC’11-15x2 was 71% of the original weight when immersed in water for 7 days. On the contrary, the weigh change rate of Sponge-CC’11-5x2 was 25% of the original weight when immersed in water for 3 days (Figure 6). Next, the weight change of Sponge-CC’21 when immersed in water was measured. The weight change rate of Sponge-CC’21-15x2 was 76% of the original weight when immersed in water for 7 days. On the contrary, the weight change rate of Sponge-CC’21-5x2 was 51% of the original weight when immersed in water for 3 days (Figure 7). These results indicate that the degree of crosslinks between collagen molecules depends on the UV irradiation time.

The degree of crosslinks between collagen molecules also depended on the composition ratio of Col and Col’. The intermolecular crosslinks was inadequate under the condition of UV irradiation time of 5 minutes, and increased in the

order of Sponge-CC’11-5x2, Sponge-CC’12-5x2, and Sponge-CC’21-5x2 (Figure 8). The intermolecular crosslinks was successfully introduced under the condition of UV irradiation time of 10 minutes, and increased in the order of Sponge-CC’11-10x2, Sponge-CC’12-10x2, and Sponge-CC’21-10x2 (Figure 9). There was a slight difference in weight change among 3 groups due to sufficient intermolecular crosslinks when UV irradiation time was 15 minutes (Figure 10).

The weight change of cylindrical sponge was measured by immersing 5 cylindrical sponges in water at 37˚C (Figure 4(b)). The weight change rate of cy-Sponge-CC’-21-15x2 measured 3 hours, 6 hours, 12 hours, 1 day, 3 days and 7 days after immersion was 88%, 72%, 49%, 44%, 36% and 34% of the original

weight, respectively (Figure 11). The cylindrical sponge showed a marked weight loss 12 hours after immersion in water, thereafter a slight weight lost.

It is important that collagen sponge is biodegraded and absorbed by the time new tissue formation is completed. As anin vitro experiment, the enzymatic degradation properties of sponge by collagenase were investigated. The enzymatic degradation of sheet-shaped sponge and cylindrical sponge was determined (Figure 12). As a result of comparative experiment, the weight change rate of Sponge-CC’12-15x2, Sponge-CC’11-15x2 and Sponge-CC’21-15x2 when immersed in water for 1 day was 79%, 79% and 80% of the original weight, respectively. The weight change rate of Sponge-CC’12-15x2, Sponge-CC’11-15x2 and Sponge-CC’21-15x2 when immersed in water containing collagenase (0.002

w/v%) for 1 day was 70%, 77% and 77% of the original weight, respectively. This result suggests that the crosslinks between collagen molecules in these sheet-shaped sponges can effectively delay enzymatic degradation by collagenase. In addition, the weight change rate of cy-Sponge-CC’21-15x2 when immersed in water for 1 day was 44% of the original weight. On the other hand, the weight change rate of cy-Sponge-CC’21-15x2 when immersed in water containing collagenase (0.002 w/v%) was 34% of the original weight. This result suggests that the crosslinks between collagen molecules in this cylindrical sponge also can effectively delay enzymatic degradation by collagenase. It is expected that collagen sponge is biodegraded and absorbed by the time new tissue formation is completed.