Lactose Stabilization Prolongs In Vivo Retention of Cross-linked Fish Collagen Subcutaneous Grafts in Nude Mice

Background:

Bovine-derived collagen gel has been used in the medical field as an injection formulation, but there are concerns about cross-infection such as bovine spongiform encephalopathy. In this study, we attempted to use fish as a safe alternative to bovine collagen.

Objective:

Fish collagen has not been used in clinical settings, so we examined its potential by comparing its properties with those of bovine-derived collagen.

Methods:

Collagen was extracted from the ventral skin of flatfish. It was cross-linked with 1%, 3%, or 5% of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) and treated with 1%, 5%, or 10% of lactose. Hydroxyproline contents and Young’s modulus (elasticity) were measured. In addition, these were injected under the back of BALB/c nude mice and the amount of hydroxyproline was observed. Histological examination of the samples was also conducted.

Results:

The amount of hydroxyproline in fish collagen was 3.3 ± 0.3 μg/mg. The 3% collagen gel treated with 5% EDC and 5% lactose had the highest Young’s modulus and was closest to the bovine-derived collagen injection formulation. When injected into mice, it was retained in vivo for about 90 days.

Conclusions:

Fish collagen has a low denaturation temperature and is unstable and easily biodegrades in mammalian organisms. However, it is possible to approach the properties of conventional mammalian collagen by cross-linking and lactose treatment, suggesting that fish collagen can be used as a scaffold for cells in regenerative medicine.

Takeaways

Question: Can we use the unstable fish collagen as a substitute of the bovine derived collagen?

Findings: It was shown that the properties of fish collagen can be made closer to those of bovine collagen by crosslinking and lactose treatment.

Meaning: Fish collagen is unstable, but can be used as a substitute of the Bovine derived collagen if it’s crosslinked and lactose treated.

INTRODUCTION

Collagen is used as an extracellular matrix in regenerative medicine to regulate cell growth and differentiation, and is widely used as a biocompatible and bioabsorbable biomaterial for scaffold materials.^1,2 Traditionally, bovine-derived collagen gel has been used as an injection preparation in the medical field. However, bovine spongiform encephalopathy (BSE) has spread mainly in the United Kingdom and other countries, but has not been confirmed by the Ministry of Health, Labor and Welfare in Japan since 2003.³ Therefore, in this study, we attempted to use fish as a safe collagen material to replace bovine collagen. Unlike mammals such as cattle and pigs, fish has not yet been reported to cause zoonosis, and is widely used as an additive in health foods and cosmetics as a peptide type,^3–5 but has not yet been used in a clinical setting.

The thermal stability of collagen is well correlated with the body temperature of the animal (or the water temperature of the fish).^11–13 Fish have a lower body temperature than mammals, and the denaturation temperature of collagen is lower, making it unstable and biodegradable in the mammalian organism. Therefore, it is not sufficiently stable as a scaffold material as it is. Attempts have been made to improve the thermostability of fish skin collagen fiber gel by introducing chemical cross-linking agents between collagen molecules.^{3,4,10,12,14–17} For example, when 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) was used as a cross-linking agent, in vitro degradation studies using collagenase reported that non-cross-linked collagen disappeared in 4 days, whereas EDC cross-linking extended the half-life to 30 days⁴ To further improve this issue, consideration was first focused on lactose, which is used as a pharmaceutical excipient because of its water solubility and low reactivity to living organisms, and is used for molding, bulking, and dilution.^18,19 Since lactose is also a protein stabilizer,^20–23 we hypothesized that further lactose treatment of EDC cross-linked fish collagen would prolong its persistence period in the body.

Based on our hypothesis, the purpose of this study was to investigate the potential of purified fish collagen as a scaffold by injecting it subcutaneously into mice after EDC cross-linking and lactose treatment. The possibility of future use for wrinkle depressions will be addressed.

MATERIALS AND METHODS

Collagen Extraction Method (Acid/Enzymatic Method)

The white belly skin of flatfish was used as a material for fish collagen. The collagen extraction method from flatfish was modified from a previously reported method.^{3,4,5,20,24–27} (See figure, Supplemental Digital Content 1, which displays fish-derived collagen extraction procedure, http://links.lww.com/PRSGO/C189.) We added 2.5 mL of each collagen solution prepared by the method shown in Figure 1. The collagen was added to a gel filtration chromatography column (PD-10 Columns, GE Healthcare Japan Corporation, Tokyo, Japan) and desalted. The collected solution was filtered through a 0.22-µm membrane filter, and the resulting solutions were designated as 1% collagen solution and 3% collagen solution, respectively.

Fig. 1.

Type determination of fish-derived collagen. A, A peak appeared at 60 mL of elution from the FPLC chromatogram. B, From HPLC reversed-phase chromatogram, a peak appeared in 10 minutes. C, SDS-PAGE figure is (1) molecular weight marker, (2) no dilution, and (3) 10× dilution SDS acrylamide concentration is 10% 10 μL sample electrophoresis at 20 mA for 2 hours.

Amino Acid Analysis Method

Amino acid analysis was performed using an amino acid analyzer (Model L-8800A‚ Hitachi High-Technologies Corporation, Tokyo, Japan). The measurement method is shown in Supplemental Digital Content 2. (See figure, Supplemental Digital Content 2, which displays amino acid analysis method, http://links.lww.com/PRSGO/C190.)²⁸ Analytical conditions for ion exchange chromatography (a) in first protein liquid chromatography (FPLC) and reversed-phase chromachromatography (b) in HPLC (a) AKTA purifier (GE Healthcare Japan Corporation) was used. Starting buffer was 0.02 M (Na⁺) sodium acetate containing 1.0 M urea PH4.8 as mobile phase and elution buffer was 0.02 M (Na⁺) sodium acetate containing 1.0 M urea and 0.8 M NaCl, pH 4.8. The column was a CM-cellulose column (Hiprep 16/10 CM FF; GE Healthcare Japan Corporation) with a bed length of 100 mm, i.d. of 16 mm, detection wavelength of 254 nm, and flow rate of 2 mL/min. The sample was measured at a flow rate of 2 mL/min. We dissolved 2.5 mg of the sample in 200 µL of starting buffer and 20 µL of it was injected. The eluted fraction was dialyzed with Milli-Q water and lyophilized to be used for Western blotting. (b) Reversed-phase HPLC (Prominence. SHIMADZU CORPORATION, Kyoto, Japan) was performed using mobile phase A. 0.1% TFA/H₂O, B. 0.1% TFA in 90% CH₃Cl in H₂O, and a gradient method with a linear increase in organic solvent was used. A. 0.1% TFA/H₂O in 90% CH₃Cl in H₂O. The column was Inertsil Peptides C18 4 µm 2.1 × 150 mm (GL Sciences Inc., Tokyo, Japan), the detection wavelength was 230 nm, the flow rate was 0.3mL/min; 4.5 mg of the measurement sample was dissolved in 300 µL of solution A‚ and 40 µL was injected for measurement.²⁹

Western Blotting

The Mini-PROTEAN Tetra System (Bio-Rad Laboratories, Inc., Calif.) was used for the measurement.

Collagen samples obtained by FPLC were diluted in electrophoresis sample buffer (nonreducing) (Bio-Rad Laboratories, Inc.) to a concentration of approximately 4 mg/mL and heated at 95°C for 5 minutes to make SDS. Next, the electrophoresis gel plate was opened, the electrophoresis apparatus was assembled, 10 µL of sample and 5 µL of molecular weight marker were spotted on a lane of 10% SDS gel, and electrophoresis was performed at 20 mA for about 2 hours per gel. Separated protein bands were stained with 0.1% Coomassie Brilliant Blue G-250 and decolorized with a solution containing methanol, acetic acid, and pure water. Primary antibody anti-type I collagen (bovine) and secondary antibody antibody II biotinylated goat anti-rabbit: IgG(H+L) (Bio-Rad Laboratories, Inc.) was used to analyze the gels using the gel documentation system (Trance Blot SD Bio-Rad Laboratories, Inc.).²⁸

Mechanical Properties of Fish Collagen

Young’s modulus was measured using the fish-derived collagen obtained. The method of operation is shown in figure, Supplemental Digital Content 3, which displays mechanical measurement method venoustron III (AXIOM Co., Ltd.), http://links.lww.com/PRSGO/C192).³⁰

EDC Cross-linking of Collagen

For cross-linking of collagen‚ 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan, cat.no348-03631) was dissolved in 3% collagen solution at the ratio of 1%, 3%, and 5% and reacted at 4°C for 4 days. The method of operation is shown in figure A, Supplemental Digital Content 4, which displays (A) cross-linking of fish-derived collagen and (B) lactose treatment method, http://links.lww.com/PRSGO/C193.^17,20

Lactose Treatment of EDC Cross-linked Collagen

Five percent aqueous lactose solution (Mylan EPD G.K., Tokyo, Japan) was added to the EDC cross-linked collagen and shaken overnight at room temperature. The method of operation is shown in figure B, Supplemental Digital Content 4, which displays (A) cross-linking of fish-derived collagen and (B) Lactose treatment method, http://links.lww.com/PRSGO/C193.^20,21

Histological Examination of Fish-derived Collagen Gels

The obtained fish collagen was observed using gross findings (PENTAX Optio RZ10) and a scanning electron microscope (SU-8010, HITACHI High-Technologies Corporation). Detailed operating procedures are shown in figure, Supplemental Digital Content 5, which displays scanning electron microscope sample preparation method, http://links.lww.com/PRSGO/C194.^31,32

The days of retention of 1% EDC cross-linked collagen injected subcutaneously into nude mice at each lactose concentration. The prepared 1% EDC cross-linked collagen gel was mixed with 1%, 5%, and 10% lactose solution, and 300 mg of the gel was injected under the skin of a nude mouse and the swelling was observed by the naked eye (Patent: JP4609897). The operation method is shown in figure, Supplemental Digital Content 6, which displays the days of retention of 1% EDC cross-linked collagen injected subcutaneously into nude mice at each lactose concentration, http://links.lww.com/PRSGO/C195.

Measurement of Hydroxyproline Amount

Various types of fish collagen (3% fish collagen, 1%, 3%, and 5% EDC, and 5% lactose) were injected into mice, and the excised tissues after the course of the injection were used as samples. The method for measuring hydroxyproline content is shown in figure, Supplemental Digital Content 7, which displays the measurement of hydroxyproline (OHP) in tissues, http://links.lww.com/PRSGO/C196.³³

Subcutaneous Injection of Nude Mice and Various Types of Retrieval of Harvested Specimens

Animal care and experimental procedures were conducted in accordance with the Teikyo University Animal Experimentation Guidelines and with the approval of the Animal Care and Use Committee (Approval No.: Teikyo University Animal Care and Use Committee No. 11-016). After 0.3 g of various fish collagen preparations were injected subcutaneously into the back of each of eighteen 7-week-old male balb/c nude mice at two locations, the period until the disappearance of the bulge on the back of 36 mouse specimens was observed grossly. At the same time, after 1, 3, 6, 9, 12, and 15 weeks, the injected specimens were removed and weighed, and the amount of hydroxyproline was measured.

Histological Evaluation of Excised Specimens

Tissue specimens from the injection site were removed immediately after and 90 days after subcutaneous injection of lactose-treated collagen gel into the back of nude mice. The specimens were immersed 4% paraformaldehyde and sent to pathological laboratory (Biopathology Institute Co. Ltd. Ooita, Japan) for outsourcing and blinding of histological examination. The specimens were stained with Azan stain, Hematoxylin and Eosin stain, and immunostained for histological examination. Immunostaining was processed using Envision/HRP DAKO (Agilent Technologies, Ink., Calif.). The primary antibody for collagen type I staining was collagen type 1 antibody (GTX41285, GeneTex, Inc., Calif.), followed by biotin-labeled secondary antibody.³⁴ The primary antibody for collagen type III staining was anti (mouse) type III collagen (LSL, LB-1393), followed by a biotin-labeled secondary antibody‚ expressed as negative (−), weakly positive (+), or strongly positive (++) for type I and type III.

Statistical Analysis

Microsoft Excel 2012 was used. All experimental results are presented as mean ± SD.

The t test was used when comparing two groups. The Dunnett test was used to compare with controls for multiple comparisons. Values of P less than 0.01 and P less than 0.05 were considered statistically significant, respectively.

RESULTS

Confirmation of Prepared Collagen as Fish Collagen Type I

The first peak of the chromatogram on the ion exchange column of FPLC is pepsin. The arrowed peak is consistent with the single measurement of calf collagen type I standard (Purecol Advanced BioMatrix Inc., Calif.) (Fig. 1A). When this peak was applied to the reversed-phase chromatography column of HPLC, it was a single peak (Fig. 1B). Furthermore, type I was confirmed by Western blotting of SDS-PAGE (Fig. 1C).

No type III peak was detected. Hydroxyproline, an amino acid unique to collagen, is 65.2 μg/mg and the mol% is 6.4%. Compared to the human hydroxyproline content of 10%, this is a low value. Young’s modulus of type I bovine collagen (Zyderm collagen implant) and fish collagen was measured: type I bovine collagen had a minimum value of 1.9 kPa and a maximum value of 29.6 kPa, 13.5 ± 9.1 kPa (n = 16). Fish collagen had a minimum value of 0.7 kPa and a maximum value of 1.9 kPa with 1.0 ± 0.3 kPa (n= 16). Comparison of Young’s modulus between type I bovine collagen and fish collagen showed that fish collagen was significantly more fragile (t test, **P < 0.01) (Fig. 2). Gross appearance of the fish collagen was translucent and gelatinous (Fig. 3A). The convexity and concavity of the sample can be clearly observed due to metal deposition (Fig. 3B). Under strong magnification, the EDC cross-linked collagen had a fibrous structure consisting of nanofibers of about 70 nm. A unique stripe-like structure was observed (Fig. 3C). Furthermore, when lactose was added, the stripe-like structure was thicker and more stable (Fig. 3D).

Fig. 2.

Comparison of Young’s modulus between Zyderm and fish collagen. The average Young’s modulus of Zyderm is about 13.5 kPa ± 9.1 (n = 11). The average Young’s modulus of fish collagen is around 1.02 kPa ± 0.3 (n = 16). **P < 0.01.

Fig. 3.

Macroscopic findings and scanning electron micrographs (SEMs). A, Macroscopic findings of fish collagen type I (PENTAX Optio RZ10). B, Fish collagen microscope image at ×2500. C, EDC-treated collagen microscopic image at ×20,000. D, Microscopic image of EDC+lactose-treated collagen at ×20,000.

Five percent EDC cross-linking made fish collagen the most biologically stable after lactose treatment.

The amount of hydroxyproline in 3% collagen only was 3.3 ± 0.3 μg/mg, whereas that in 1% EDC was 2.8 ± 0.2 µg/mg, 3% EDC was 2.9 ± 0.3 µg/mg, and 5% EDC was 3.5 ± 0.1 µg/mg. There was no significant difference compared to 3% fish collagen only (Dunnett test) (Fig. 4).

Fig. 4.

Comparison of hydroxyproline content of fish collagen gel treated with three different EDC concentrations. There was no significant difference between 3% collagen, 5% lactose, 1% EDC and 3% collagen, 5% lactose, 5% EDC compared to 3% fish collagen only. The statistical analysis carried out is the Dunnett test. Data value represents the mean ± SD (n = 3).

Although lactose-untreated EDC cross-linked collagen was injected and retained as subcutaneous swelling for 4.5 ± 0.4 days, EDC cross-linked fish collagen treated with 1%, 4.7%, and 10% lactose were retained in mice for 19 ± 2.2, 93 ± 8.0, and 6.5 ± 0.4 days, respectively. There was a significant difference compared to untreated (Dunnett test **P < 0.01). Therefore, 5% lactose treatment is the most effective in inhibiting biodegradation (Fig. 5). We have obtained a national patent (4609897)²⁰ for this discovery.

Fig. 5.

Retention days of 1% EDC cross-linked collagen at each lactose. (n = 3) **P < 0.01.

Next, the mechanical properties of the collagen gels were examined by comparing the three gels at the above ratios. Young’s modulus (n = 3 for each) was EDC1% 4.7 ± 1.1 kPa, EDC3% 5.8 ± 1.6 kPa, and EDC5% 10.3 ± 0.3 kPa. There were significant differences at 3% and 5% compared to 1% (*P < 0.05). The retention days of the swelling on the body surface of mice were 17.5 ± 0.5, 27.5 ± 3.8, and 89.8 ± 0.9 µg/mg days, respectively, indicating that the EDC concentration of 5% was retained for the longest period of time (**P < 0.01) (Fig. 6).

Fig. 6.

Mechanical properties of extracted specimen and retention days in subcutaneous grafts. A, Young’s modulus of fish collagen gel after subcutaneous injection in mice, and (B) retention days after implantation. Young’s modulus was measured using water as the solvent. Significant differences were observed in the 3% and 1% EDC concentrations compared to the 5% EDC concentration (*P < 0.05). Significant difference in 3% and 1% compared to 5% EDC (**P < 0.01). The statistical analysis carried out is the Dunnett test. Data value represents the mean ± SD (n = 3).

The biostability of lactose-treated 5% EDC cross-linked fish collagen was comparable to that of type I bovine collagen. The hydroxyproline content of the lactose-treated 5% EDC cross-linked fish collagen excised at 1, 3, 6, 9, 12, and 15 weeks after injection were 3.6 ± 0.2, 3.9 ± 0.7, 3.9 ± 0.7, 2.9 ± 0.3, 0.8 ± 0.4, and 0.8 ± 0.6 µg/mg, respectively. After 6 weeks, the amount of hydroxyproline showed a decreasing trend, and after 12 weeks, only a small amount of hydroxyproline remained. There was a significant difference at 12 and 15 weeks compared to 1 week (Dunnett’s test **P < 0.01)‚ although the hydroxyproline content of type I bovine collagen had been demonstrated to be 3.4 ± 0.2, 2.7 ± 0.5, 1.8 ± 0.2, 0.8 ± 0.1, 1.7 ± 0.3, and 1.2 ± 1.2 µg/mg, respectively. The hydroxyproline content of Zyderm at 12 weeks increased from that at 9 weeks significantly (P < 0.001) (Fig. 7A). In comparison with Zyderm, fish collagen did not increase its hydroxyproline content at 12 weeks.

Fig. 7.

Amount and weight of hydroxyproline after injection in mice. Measurements were taken after 7 days (1 week), 21 days (3 weeks), 42 days (6 weeks), 63 days (9 weeks), 84 days (12 weeks), and 105 days (15 weeks). A, There was a significant difference at 12 and 15 weeks compared to 1 week (**P < 0.01). There was a significant difference at 6, 9, 12, and 15 weeks compared to 1 week (**P < 0.01). B, Same trend, with significant differences at 6, 9, 12, and 15 weeks compared to 1 week (**P < 0.01). The statistical analysis carried out is the Dunnett test. Data value represents the mean ± SD (n = 3).

The weight of the excised tissue of the lactose-treated 5% EDC cross-linked fish collagen decreased gradually from 300 mg at the time of injection to 121.7 ± 0.3, 97.0 ± 1.0, 61.1 ± 0.7, 32.2 ± 0.3, 6.8 ± 0.4, and 6.2 ± 0.6 mg, respectively (Fig. 7B). Significant differences were observed at weeks 6, 9, 12, and 15 compared to 1 week (Dunnett’s test **P < 0.01).

Findings of Excised Specimens

Light microscopic findings of HE staining showed that lactose-untreated collagen gel tended to lose its shape immediately after injection (Fig. 8A), whereas lactose-treated collagen gel remained in a clumped state (Fig. 8D). Immunostained images (collagen type I- and III-specific staining) of lactose-untreated collagen gels showed very mild expression of collagen type I in the dermis (Fig. 8B) and moderate expression of collagen type III in the epidermis and very mild expression in the dermis (Fig. 8C). In the immunostained image of 5% lactose treatment, collagen type I was very mildly expressed in the dermis (Fig. 8E), while collagen type III was moderately expressed in the epidermis and very mildly expressed in the dermis (Fig. 8F). Specimens with fibrosis showed weakly positive (+) and strongly positive (++) results for both type I collagen and type III collagen, while those without fibrosis were negative (−). The specimens were strongly positive at 90 days for type I bovine collagen, but weakly positive at 180 days for type I bovine collagen, suggesting that fibrosis may have peaked at 90 days for type I bovine collagen and then decreased.

Fig. 8.

Azan staining and HE staining after subcutaneous injection of fish collagen immediately after lactose treatment. Hematoxylin is oxidized to hematein, which binds to the metal in the stain. The nucleus is stained purple and the cytoplasm is stained red. A, Fish collagen without EDC cross-linking treatment. B, Immunostained collagen type I. C, Immunostained collagen type III. D, Collagen treated with lactose. E, Immunostained collagen type I. F, Immunostained collagen type III.

We see more fibrous growth than in our experiment. Observations from 1 to 4 days after fish collagen injection showed a mild infiltration of inflammatory cells, mainly macrophages, around the EDC-only collagen clumps, whereas lactose-initial and lactose-free EDC-only-initial were negative (Table 1). Sugiura et al¹⁶ found that there was a significant difference in the inflammatory response at 4 weeks compared to 1 week after transplantation of EDC-only cross-linked terapia collagen in rabbits, with a significant reduction in the inflammatory response at 4 weeks. Regarding fibrogenesis, our experiments showed very little increase in fibrosis around the collagen after 3 days of injection (Table 1). After 90 days, the maximum retention period, lactose-treated collagen remained in clumps in the tissue specimens by Azan staining (Fig. 9B). Compared to control (no EDC or lactose treatment) and fish collagen (EDC + lactose treatment), immunostaining showed the expression of type I and type III collagen (detection method shown in SDC8) in the pericutaneous muscle after 90 days of EDC + lactose-treated fish collagen and increased collagen fibers were observed by Azan staining. (See figure, Supplemental Digital Content 8, which displays the collagen types I and III staining protocol, http://links.lww.com/PRSGO/C197). Figure 9D shows a comparison of fish collagen and type I bovine collagen after 90 days of treatment, showing a higher degree of fibrogenesis in the pericutaneous region of type I bovine collagen than in fish collagen. All images were measured at n = 3, and typical tissue findings were selected (Fig. 9C and G).

Table 1.

Immunostaining Test Result of Mouse Skin (Collagen I and III)

			5% Lactose EDC	EDC Only	EDC Only	EDC Only	EDC Only	EDC Only
			Fish Collagen	Fish Collagen	Fish Collagen	Fish Collagen	Fish Collagen	Fish Collagen
					1 d Later	2 d Later	3 d Later	4 d Later
Collagen type I	Epidermis		–	–	–	–	–	–
	Dermis	The dermis just below the epidermis	±	±	++	+++	++	+++
		Hypodermis	±	±	±	±	±	±
Collagen type III	Epidermis		++	++	++	++	++	++
	Dermis	The dermis just below the epidermis	±	±	±	±	±	±
		Hypodermis	±	±	±	±	±	±
HE stain	Inflammatory cell infiltration		–	–	+	+	+	+
	Fish collagen mass perimeter
	Fibrous growth		–	–	–	–	±	–
	Fish collagen mass perimeter

Gross observation and judgment was made by a pathology researcher in a blinded manner.

–, no change; ±, very mild expression; +, mild expression; ++, moderate expression; +++, high level of expression.

Fig. 9.

Azan staining and immunostaining of lactose-treated fish collagen 90 days after subcutaneous injection. Azan staining (A–D) is used to stain connective tissue. Collagen fibers: dark blue nuclei: dark red cytoplasm: light red cell secretory granules: blue for basophilic, red for acidophilic. Immunostaining (E-H) is an effective method for detecting specific antigens in tissues and is a staining method that utilizes an antigen-antibody reaction in which an antibody binds to an antigen. Therefore, it stains collagen fibers (collagen type I and III). A, Fish collagen only. B, EDC cross-linked and lactose-treated fish collagen (40×). C, 200× magnification of (B). D, Azan stain for bovine collagen (×200). E, Immunostained image of fish collagen only (×40). F, Immunostained image of (B) (×40). G, Immunostained image of (C) (×200). H, Immunostained image of bovine collagen (D) (×200).

DISCUSSION

Suzuki³⁵ and Miyata³⁶ reported hydroxyproline contents of 10 mol/% in humans, 9.1 mol/% in cattle, 9.8 mol/% in pigs, and 6.4 mol/% in marine collagen. Our experimental results also showed that the hydroxyproline content of fish collagen was 6.4 mol/%. There were no significant differences in amino acid composition among mammalian species, but the hydroxyproline content of marine atelocollagen was lower than that of mammals.

Burke et al^.37,38 injected type I bovine collagen (Zyderm collagen implant) into porcine and human skin and conducted histological studies. When injected into human skin, mononuclear cell infiltration was observed around the transplanted collagen for up to 1 week, and this inflammatory cell response lasted up to 4 months. At 35 days, degradation and resorption of the injected collagen began to be observed, and at 112 days, the intradermally injected collagen was reported to have disappeared, and in its place, elastin-free, neoplastic, fine-fiber collagen was observed. This mechanism is based on the fact that human types I, II, and III collagen and collagen-derived peptides bind specifically to fibroblasts. Therefore, bovine-derived collagen is thought to directly interact with fibroblast receptors and promote their activation. After 90 days of injection, the injected implant is clearly degraded and contains no elastic fibers (elastin). It is replaced by the host’s nascent collagen. In our previous study‚³⁹ in which type I bovine collagen was injected into mice, a decreasing trend was also observed from 3.4 ± 0.2 μg/mg on day 0, 2.7 ± 0.5 μg/mg on day 30, and 1.8 ± 0.2 μg/mg after 45 days. After 60 days, the number of collagen fibers further decreased to 0.8 ± 0.1 μg/mg, and after 90 days, at 1.7 ± 0.3 μg/mg, and a high increase of collagen I was also observed around the skin muscle. At 1.2 ± 0.2 μg/mg after 120 days, the number of collagen fibers decreased (Fig. 7B). Comparing Figure 7A and B, it remains to be seen whether the fact that fish collagen does not increase the amount of hydroxyproline after 90 days as type I bovine collagen does not indicate the formation of new collagen. Our experiments with fish collagen also showed that the amount of hydroxyproline in the collagen gel injected into mice began to decrease from week 9, and by week 12, the amount of hydroxyproline was minimal, and there was little swelling in the mouse body as seen by the naked eye. Although there is a difference in Young’s modulus between type I bovine collagen and fish collagen (Fig. 2), there is no significant difference between bovine-derived collagen at 13.5 ± 9.1 kPa and fish-derived collagen at 10.3 ± 0.3 kPa when EDC is cross-linked and wrapped with lactose, and the amount of hydroxyproline contained in the collagen is almost the same between 1164 µg/0.3 g and 1159 µg/0.3 g after 90 days. Even after 90 days, the concentrations are 1.7 ± 0.3 and 0.82 ± 0.4 µg, respectively. Although fish is cheaper (about ¥5600) and more readily available in large quantities than mammals,^3,4 it has the problem that the denaturation temperature of collagen is low,^3,12 making it unstable and easily biodegradable in the mammalian organism. To solve this problem, it is possible to approach the properties of conventional mammalian-derived collagen by cross-linking and lactose treatment (Fig. 10), suggesting that it can be used as an injectable collagen preparation and as a cell scaffold in regenerative medicine. There are many study limitations: this is a murine and immunodeficient animal study; this is a substantial qualitative analysis and does not have quantitative outcomes. The findings obtained are preliminary, and future research should be extended to large animals for use as scaffolds.

Fig. 10.

Illustration of biology/mechanism of action. Collagen fibers are shown with EDC cross-linking and added sugars. A collagen molecule is composed of three α-chains, polypeptide chains with a molecular weight of approximately 100,000. In the case of collagen type I, two of the three chains are α₁ chains, and the other one is an α₂ chain. The three chains are bundled together to form a bundle.