Comparison of collagens extracted from swim bladder and bovine Achilles tendon

Collagen is a type of natural biopolymer material, which is widely used in tissue engineering and medicine owing to its exceptional properties such as biodegradability, biocompatibility, hemostatic properties, and low immunogenicity. Collagens from different sources can differ in type, structure, and function. In this study, collagen was extracted from swim bladder and bovine Achilles tendon by acid-enzyme binding method at low temperature. UV spectrum, Fourier transform infrared spectrum, sodium dodecyl sulfate–polyacrylamide gel electrophoresis, scanning electron microscope, and differential scanning calorimetry were used to characterize these two collagens. The blood compatibility and cytotoxicity of the two kinds of collagen were studied.The results showed that the collagens from the two sources belong to the characteristics of type I collagen and had biological safety. Their differences in structure and thermal stability can provide a theoretical basis for the selection of collagen in practical application.


Introduction
Collagen is a biomaterial with a long history of applications [1]. It is the most abundant structural protein in the extracellular matrix of many connective tissues in the body, accounting for approximately 30% of total protein in various connective tissues [2,3]. The molecular weight, diameter, and length of collagen are about 300 kDa, 14-15 Å, and 2800 Å, respectively. The right-handed triple helix structure is formed by the intertwining of three left-handed α-peptide chains, where the α-peptide chain is mainly a glycine-proline-hydroxyproline repeat sequence [4]. Hydroxyproline is the characteristic amino acid of collagen and is crucial for collagen biosynthesis and the formation of a stable triple helix structure. It can be used for the qualitative and quantitative analysis of collagen. Twenty-nine different types of collagens have been discovered and named so far, and they are distributed in different tissues [5,6].
Collagen, as a type of multifunctional protein, has biodegradability and biocompatibility properties that are unmatched by other polymer materials [1,7]. In addition, it has unique physicochemical properties such as low immunogenicity, easy absorption by the human body, promoting survival and growth of cells, and promoting platelet aggregation [8,9]. Therefore, collagen-based biomaterials have been widely used in biomedicine and tissue repair engineering. Based on different clinical requirements, collagen is processed into various forms such as hemostatic powder, burn dressing, drug carrier, cardiac membrane, blood vessel, esophagus, trachea, and surgical suture. Thus, it can be used for different purposes, showing a good application potential in clinical medicine [10][11][12][13][14][15]. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
As far as mammalian collagens are concerned, bovines and pigs are the main sources of collagen in industrial applications. However, due to the outbreak of bovine spongiform encephalopathy and foot-and-mouth disease and the religious sentiment of people, the applications and trade of collagen have been limited [16]. Therefore, many studies have attempted to discover new sources of collagen. Recently, the study of fish-derived collagens has become a substitute because they do not contain the risk of animal diseases and pathogens. Furthermore, fish-derived collagens are not subjected to religious and ethical restrictions and are more resourceful than other animals [17][18][19]. Previous studies have shown that collagens from different animal sources have different thermal denaturation temperatures due to slight differences in their amino acid composition and content [20].
Owing to the effect of differences in species, mammalian bovine collagens, and marine fish collagens are bound to have differences in structure and function.
Presently, the main methods of collagen extraction are hot water extraction, alkaline extraction, acid extraction [21], enzyme extraction [22], and acid-enzyme combination extraction [23]. Among these different extraction methods, alkaline extraction is rapid. In this method, all amino acids containing hydroxyl and sulfhydryl groups are destroyed, and structural variation is generated. The extraction process of the acid method is rapid and complete; the structure of tryptophan is destroyed, the structure of serine and tyrosine is partially destroyed, and the extraction rate of the production is very low. The enzyme extraction method causes less harm to collagen, which can extract better and more complete active protein, but the hydrolysis is not sufficient enough. Presently, the acid-enzyme binding method is often used to extract collagen, which has mild extraction conditions and fast reaction speed. Nevertheless, the extracted collagen has a complete three-strand helix structure, high protein purity, and stable physical and chemical properties.
In this study, bovine Achilles tendon collagen (BATC, representing mammals) and swim bladder collagen (SWC, representing marine organisms) were extracted by the acid-enzyme binding method. The structure and function of the two were compared by ultraviolet spectroscopy, sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE), infrared spectroscopy, differential scanning calorimetry, scanning electron microscopy (SEM), blood compatibility test, and cytotoxicity test.

Materials and reagents
The bovine Achilles tendon was purchased from Zhenqidao Biotechnology Co., Ltd., and the swim bladder was purchased from Zhejiang Chongshan Biological Products Co., Ltd. They were quickly transported to the laboratory, frozen, and then immediately stored at −20°C for future use. All the chemicals used were of the analytical grade.

Extraction of collagen
Of note, 15 g each of swim bladder and bovine achilles tendon were weighed and soaked in 0.1 mol l −1 NaOH solution for 12 h. The NaOH solution in the mixture was changed every 6 h to remove the fat and miscellaneous protein in the sample. Later, the solution was washed with ddH 2 O until the wash water became neutral. After soaking it in 0.5 mol L −1 acetic acid solution for 30 min, the sample was washed thrice with distilled water. The sample was again soaked in 0.5 mol L −1 acetic acid solution overnight for about 12 h, where the sample was observed to be semitransparent and swollen. An appropriate amount of 0.5 mol L −1 acetic acid solution was added to the swollen sample that was further crushed to obtain homogenate. After dissolving pepsin powder in proper 0.5 mol L −1 acetic acid solution, the pepsin powder was mixed with homogenate and stirred fully. The mixture was hydrolyzed for 5 days at 4°C and stirred every 8 h to make the reaction sufficient. The enzymatic hydrolysate was centrifuged for 30 min at 10°C at 7000 rpm. The supernatant was collected, and the solution was adjusted to alkaline using NaOH. The solution was placed for some time, and the collagen suspension was obtained when the collagen was fully precipitated. The collagen suspension was centrifuged at 10°C and 7000 rpm for 10 min, and the precipitate obtained was crude collagen. The crude collagen was purified by washing, centrifuging, and using dialysis, and finally, the targeted collagen was obtained. This collagen was stored at 4°C and set aside until further use. The objective to prepare collagen sponge by vacuum freeze-drying after prefreezing collagen solution at −20°C was thus achieved.

Ultraviolet spectrometric determination
The SBC and BATC were dissolved in 0.5 mol L −1 acetic acid solution to obtain 1 mg ml −1 collagen solution. The acetic acid standard solution was used as a blank control and placed in a quartz cuvette at the reference end. An ultraviolet-visible light spectrophotometer was used to perform absorption spectrum scanning, where the scanning wavelength range was 210-800 nm, the scanning speed was medium, and the scanning interval was 1 nm.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
The protein profile of collagen was determined according to the method of Laemmli [24]. The electrophoresis gels contained 10% separating gel and 5% stacking gel. To prepare 0.3 mg ml −1 SBC solution and 1 mg ml −1 BATC solution, the collagen solutions were mixed with 5× loading buffer at a ratio of 4:1. The loading volume of each well was 16 μl. Electrophoresis was performed at a constant voltage of 100 Vfor 4h. After electrophoresis, the cells were stained with 0.1% Coomassie brilliant blue R-250 in 25% (v/v) isopropyl alcohol and 10% (v/v) acetic acid for 40 min, and then with 5% (v/v) ethanoland 10% (v/v) acetic acid. The gel was destained with acetic acid until the band in the electrophoresis gel was clear. The isolated collagen bands were compared with standard molecular weight markers.
Fourier transform infrared spectroscopy Samples were prepared by grinding collagen with potassium bromide and then pressing them into translucent particles. The infrared spectrum of lyophilized collagen was measured by a Fourier infrared spectrometer in the absorbance mode of 4000-400 cm -1 .
Determination of the denaturation temperature Of note, 5 mg of fish SWC sponge and BATC sponge was taken after vacuum freeze-drying. After accurate weighing, they were added into a sealed aluminum tray, and the empty aluminum tray was used as a reference. Under the flow of nitrogen, the heat flow information of the sponge in the range of 20°C to 130°C was recorded at a heating rate of 10°C min −1 .

Scanning electron microscopy
The morphological characteristics of collagen samples were observed by SEM. We prepared 6 mg ml −1 of SBC and BATC solutions and allowed them to stand at 4°C for 24 h to eliminate air bubbles. The sponge was prepared after pre-freezing at −20°C for 12 h and vacuum freeze-drying for 36 h. A thin layer of the collagen sponge was carefully cut out with a tissue slicer, and it was trimmed into pieces of regular shape. The surface of the sample was covered with a layer of gold-palladium conductive layer by the glow system, and it was observed under a scanning electron microscope.
In vitro hemolysis test 4 ml of fresh blood was collected from male New Zealand rabbits (weight: 2.0-2.5 kg). Sodium citrate (3.8% sodium citrate:rabbit blood = 1:9), an anticoagulant, was added to the fresh blood, and 5 ml of normal saline was added to dilute the blood.
We added 20 mg of fish SWC or BATC to 10 ml of normal saline as the sample group. The positive control group and negative control group were replaced with 10 ml ultrapure water and 10 ml normal saline, respectively. The samples of each group were incubated at 37°C for 30 min, then 0.2 ml of blood was slowly added to each tube. The samples were again incubated at 37°C for 1 h. The samples of each group were centrifuged at 1000 rpm for 5 min, and the supernatant was removed and added to the wells of a 96-well plate. The absorbance value was determined at 540 nm, and four parallel tests were performed to calculate the hemolysis rate using the formula (1).
where HR is the hemolysis rate, and OD0, OD1, and OD2 represent the absorbance values at 540 nm in the negative control group, positive control group, and sample group, respectively.

In vitro cytotoxicity test
According to the provisions of the Chinese standard GB/T 16886.12, the sponge samples of SWC and BATC were cut into pieces of size 10 mm × 50 mm. After sterilizing them under UV irradiation, the two collagens were added to a complete medium (containing 5% fetal bovine serum, 1% Penicillin streptomycin double antibody, and RPMI 1640 medium) with a surface area/volume ratio of 3 cm 2 ml −1 . The solutions of the blank group and negative control group were added to complete media (containing 5% fetal bovine serum, 1% double-antibody, and RPMI 1640 medium), and the positive control group solution was a complete medium with 0.64% phenol (containing 5% fetal bovine serum, 1% double-antibody, and RPMI 1640 medium). All the samples were incubated at 37°C for 24 h. The L929 cells in the logarithmic growth phase were collected in a T25 culture flask. After digestion and resuspension, the cell concentration was adjusted to 5 × 10 4 cells ml −1 , and 100 μl of the cell suspension was seeded in each well of a 96-well plate. The blank group was added with the same amount of complete medium, and six similar wells were set in each group. After the adherence of the cells to the walls, the culture medium was discarded, and 100 μl of the solutions of the blank group, negative control group, positive control group, SWC extraction, and BATC extraction were added into the wells of each group. After incubating the culture for 24 h in a 5% incubator at 37°C, the cell morphology of each group was observed under an inverted microscope, and the images of the cells were obtained. The residual solution was then washed twice with phosphate-buffered saline. The CCK8 kit was used to detect the absorbance of each hole and calculate the cell survival rate.

Statistical analysis
GraphPad Prism 8.0.1 was used for performing all statistical analyses. The results are presented as mean ±standard deviation (SD), and a P value of <0.05 is considered statistically significant.

Results and discussions
Comparative analysis of the preparation process of collagen The preparation of fish SWC and BATC is shown in the flow chart of figure 1. After acid immersion, the swim bladder became swollen and crystal clear, whereas the bovine Achilles tendon was slightly swollen and remained light yellow. This may be because of the different densities of the tissue. The tissue structure of the swim bladder is loose, and the acid solution can easily penetrate. However, the tissue structure of the bovine Achilles tendon is packed, and the acid solution is difficult to penetrate. The SWC, after alkali analysis, was a uniform and stable suspension. The collagen was loose, and it resembled gelatin after letting it stand for a while. The BATC was not stable after the alkali analysis. The BATC suspension did not become stable after letting it stand for a while. Therefore, the two types of collagen molecules behave differently in a solution. The agglomeration of fish collagen molecular size is small. It can stably remain in the solution until several hydrogen bonds are crosslinked. The molecular aggregates of the BATC settle after a short time owing to their large size. The BATC is relatively white after centrifugation. At the same concentration, the fish SWC and BATC solutions were significantly different. The fish SWC solution was less viscous, whereas the BATC solution was more viscous. Both collagen sponges were white and soft.

Comparative analysis of the morphology of collagens
The structure of the SWC and BATC sponges is shown in figure 2. Both collagen sponges are white and have a loose fibrous porous structure ( figure 2(A)). Pepsin breaks down the terminal structure of collagen, causing the fibers to twist into a loose network [25]. As shown in the apparent structure diagram, the SWC sponge is fluffy and soft, whereas the BATC sponge has a regular shape. As shown in the microscopic images, both types of collagen had partial wrinkles on the surface ( figure 2(B)), which could be caused by moisture sublimation during freeze-drying [26,27].
The two types of collagens have high porosity at low magnification, but the porosity of the BATC sponge is higher than that of the SWC sponge. The parallel arrangement of collagen molecules and the aggregation of collagen fiber bundles were observed at high magnification. Thus, the morphology, molecular configuration, and arrangement of the two types of collagens are different (figures 2(B), (C)). Collagen is an important biomaterial and plays an important role in biomedical engineering. Collagen sponge has hygroscopic properties and extremely high porosity, which can be used as a hemostatic dressing. Swim bladder collagen sponge is soft and may have a good skin-friendly feeling. Bovine Achilles tendon collagen sponge has a regular arrangement of porous mesh structure, which may be more conducive to cell adsorption and growth and oxygen and nutrient delivery.

Comparative analysis of the characteristics of collagens
The electrophoretic patterns of collagen from different sources, showing the composition of collagen subunits, are shown in ( figure 3(A)). α-chain, β-chain, and γ-chain bands were found in SWC and BATC, indicating that the main components of the two collagens were type I collagen, which is consistent with the results of a previous study [28]. No collagen fragments with molecular weight lower than the chain were observed in the electrophoresis map, indicating that all the extracted collagen proteins maintain their structural integrity. Notably, the positions of the α1 and α2 chains of SWC are close, indicating that the molecular weights of the two peptide chains are similar, whereas there is a certain distance between the positions of BATC α1 and α2 chains. There is a certain difference in the molecular weight of the two peptide chains. Moreover, collagen from the bovine Achilles tendon contains two different β chains, whereas only one major β chain is found in SWC [29].
In the electrophoretic map of SWC, the color of the band near 130 kDa is darker than that near 250 kDa and above, which may be because most of the polypeptide chains in fish SWC are present as a single chain or it could be due to the low cohesion between peptide chains. Furthermore, 300 kDa fish SWC is easy to dissociate into a single peptide chain in the process of heating. However, in the electrophoretic map of BATC, the color of 250 kDa and above bands is similar to that of the bands near 130 kDa, which may be because of the similar content of single chain and polymer in BATC or because the aggregation between peptide chains is strong, and the polymer is not easy to dissociate in the process of heating.
The ultraviolet spectrum of collagen is shown in figure 3(B). Both SWC and BATC produce light absorption in the 210 nm-400 nm band, and the maximum ultraviolet absorption peak appears near 230 nm, indicating that the main components of the two collagens are type I collagen, which is consistent with the results of a previous study [30]. No distinct absorption peak was found near 280 nm, which may be owing to the low content of aromatic amino acids, such as tryptophan and phenylalanine, in collagen [30]. The UV spectra of the two types of collagen are slightly different, indicating that there are some differences in the type and number of amino acids between the two types of collagen.
The infrared spectrum of collagen is shown in (figure 3(C)). The infrared spectra of SWC, BATC, and rat tail collagen (RTC) are similar, but there are some differences, indicating that their secondary structures are different. The three types of collagens contained five characteristic absorption peaks, which are amide A, amide B, amide I, amide II, and amide III. Amide A is mainly related to the frequency of N-H stretching [31]. When N-H in the peptide chain is involved in forming hydrogen bonds, the position of the amide A band is red-shifted. Statistical analysis showed that BATC might contain more hydrogen bonds (table 1). Amide B is related to the asymmetric stretching vibration of CH2, which is the characteristic group of collagen tertiary structure, and its appearance implies that both SWC and RTC contain tertiary structure [32]. Amide I, amide II, and amide III are closely related to the structure of the peptide chain skeleton, indicating the triple helix structure of collagen. Among them, amide I is related to the tensile vibration of Coluo and is a marker of the secondary structure of collagen and can form interchain hydrogen bonds [30]. Amide II is mainly related to the N-H bending vibration caused by C-N stretching vibration, which indicates the number of N-H groups bound by hydrogen bonds between adjacent α chains [33]. Amide III is mainly related to C-N stretching vibration, N-H bending vibration, and C-H stretching vibration [34]. Some absorption peaks around 1402 cm -1 -1454 cm -1 are mainly related to the vibration of pyrrolidine rings in proline and hydroxyproline [35]. Statistical analysis showed that the wavenumber positions of the five characteristic absorption peaks of the three types of collagens are different, which may be owing to the different folding modes of collagen molecules from different sources, resulting in different intramolecular and intermolecular hydrogen bond interactions. The strength of the hydrogen bond affects the wavenumber position of the absorption peak. Furthermore, our results also showed that if the ratio of  the intensity of the absorption peak near 1240 cm -1 to that near 1450 cm -1 is close to 1, it means that collagen has a complete triple helix structure [36]. The intensity ratios of SWC, BATC, and RTC are 1.14,1.50 and 1.02, respectively, indicating that the triple helix structure of the three types of collagen is intact, and the structural compactness of BATC is higher than that of SWC. Collagen is stable at low temperatures and denatures at high temperatures. The thermal change curves of the fish SWC and BATC sponges are shown in figure 3(D). The two curves have distinct endothermic peaks around 100°C, and the corresponding temperature is the thermal denaturation temperature of collagen sponges. The table shows that the thermal denaturation temperature of the SWC sponge is lower than that of the BATC sponge, indicating that the cross-linking degree of SWC is lower and the arrangement is looser (table 2).

Comparative analysis of the biological activity of collagen
The cytocompatibility analysis of collagens is shown in figure 4(A). After 24 h of culture with SWC and BATC extract, L929 cells proliferated well and showed a fusiform shape, and the growing state was the same as that of the negative control group (which was the complete culture medium group). The phenol group was cultured with L929 cells for 24 h. Cell proliferation was not active, the cell morphology was shrunk, and the cell surface was rough. The results of the CCK8 test showed that the cell survival rate of the SWC group and BATC group was  higher than 75%, and there was no distinct cytotoxicity, whereas the phenol group showed a low cell proliferation rate and cytotoxicity. The statistical chart of the cytotoxicity test is shown in figure 4(B).
The blood compatibility analysis of collagen is shown in figure 4(C). Hemolysis refers to the phenomenon of lysis and dissolution of red blood cells and the release of internal cellular substances affected by adverse factors. The degree of hemolysis is one of the important evaluation indexes of blood compatibility of biomaterials. The hemolysis rates of the SWC sponge and BATC sponge are less than 5%, which is consistent with the standard requirements of biosafety (table 3).

Conclusion
Collagen from different sources has different physical and chemical properties, and determining their differences is helpful in selecting appropriate collagen for production and application. In this study, the physicochemical and biological properties of collagen from swim bladder and bovine Achilles tendon were investigated and compared. We found that fewer aromatic amino acids are present in SWC and BATC; therefore, no strong absorption peak was found near 280 nm in the UV spectrum, which is the major difference between collagen and other proteins. SDS-PAGE showed more monomers in SWC and more dimers and trimers in BATC. Both SWC and BATC have characteristic bands of type I collagen, which is consistent with the results of a previous study [37]. Infrared spectroscopy showed that the triple helical structure of the two collagens was complete, which is crucial for the corresponding biological activities of collagens. The helical structure of BATC may be similar to that of SWC. The thermal denaturation temperature of BATC of this study was higher than that of a previous study [38]. Hydroxyproline forms interchain hydrogen bonds through hydroxyl groups to stabilize the triple helix structure of collagen, thereby increasing the thermal denaturation temperature. In this study, the thermal stability of BATC is higher than that of fish SWC. Therefore, we speculated that the content of amino acids in BATC is higher than that of fish SWC [39]. The morphology of fish SWC sponge and BATC sponge is slightly different. The white and soft appearance and highly porous network structure of these two collagens can be used in the field of biomedical materials for preparing moisturizers, wound dressings, and cell proliferation matrix [23,40]. Collagen sponges should be biosafe when used as a dressing for wound repair in vitro or scaffold implants in organisms. Herein, we explored the blood compatibility and cell compatibility of the two collagens and found that they fulfill the requirements of biosafety.