Abstract
Background/Aim: The aim of the present research is a comprehensive evaluation of a fish-collagen based wound membrane using established ex vivo, in vitro and in vivo methodologies. A porcine pericardium membrane served as control material.
Materials and Methods: Scanning electron microscopy (SEM) and attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) analysis were initially used to analyze the structure and collagen molecular structure. Also, a comparison of protein adsorption via measurement of human serum albumin (HSA) adsorption was conducted. The membrane influence on cell viability, cell proliferation as well as their cytotoxic potential were examined in vitro. Additionally, the membrane tissue integration, degradation behavior and biocompatibility were investigated using the subcutaneous implantation model.
Results: The SEM analysis showed differences in the structure and the porosity of both membranes. The analysis via FTIR spectroscopy revealed that collagen molecules are present in both membranes in their triple helical structure. The adsorption measurements showed that the surface density of HSA adsorbed to the fish collagen membrane surfaces was significantly lower compared to the values measured for the bovine pericardium membrane. Furthermore, both membranes demonstrated sufficient in vitro cytocompatibility in the indirect colorimetric XTT, LDH and BrdU assays. The in vivo study part revealed that the fish collagen membrane induced a faster biodegradation and a more pronounced pro-inflammatory tissue response, whereas the bovine pericardium membrane degrades more slowly.
Conclusion: The results of this study highlight the potential of fish collagen membranes as biocompatible wound healing materials. However, their rapid degradation presents a challenge that needs to be addressed through targeted modifications, such as optimized cross-linking.
Introduction
Skin is the largest organ in the human body. It fulfills a variety of body functions, including thermoregulation, protection against physical or chemical environmental influences and sensory perception (1, 2). As integumentum commune, it ensures the physical integrity of the organism and forms a first barrier against external pathogens.
However, human skin is neither invulnerable nor insurmountable. Various causes, e.g., mechanical, chemical or thermal, can break through the integrity of the skin, resulting in the formation of wounds. Under physiological conditions, human skin is able to autonomously regenerate itself. Thereby, “classic wound healing” is divided into four phases, overlapping and merging seamlessly into one another: i) An initial sealing of the wound from external influences by means of coagulated blood and fibrin (exudative phase), followed by ii) the macrophage-mediated degradation of the blood clots and the formation of granulation tissue (resorptive phase) and iii) an increased collagen formation by fibroblasts for additional defect filling (proliferation phase) until iv) formation of the final scar tissue and epithelialization (repair phase) (3, 4).
Various factors can decelerate and inhibit these physiological processes of wound healing. An inadequate supply of nutrients and oxygen to the connective tissue, caused by an impaired blood circulation, e.g., in peripheral arterial occlusive disease or chronic venous insufficiency, withdraws the necessary components for regeneration from the wound environment (4). Enzymes and toxins released by infectious pathogens may interrupt the physiological wound environment and shift it in favor of a pronounced inflammatory reaction (5, 6). The complex, interlocking cellular and molecular mechanisms of wound healing are disrupted and the wound can chronify (6, 7).
The basic principles of sufficient wound treatment include shielding it from further irritation, injury or external infectious agents, creating an ideal moist wound environment and promoting cell migration and proliferation (8, 9). Modern wound dressings combine these basic principles and therefore contribute as an important treatment method in chronic wound therapy. An important group of frequently used wound membranes consist of natural polymers such as chitosan, alginate or collagen, each with its individual advantages and disadvantages (10-12).
Collagen-based wound dressings derived from mammalian sources, e.g., porcine or bovine, have been widely used in wound care for decades, offering a strong structure and slower degradation time, which is particularly advantageous in chronic wound management (13-16). They are well-studied and have a proven track record in clinical practice (17-19). However, their use may be limited by religious and ethical concerns in certain populations, and due to a minimal, though present, risk of transmitting animal-borne pathogens like bovine spongiform encephalopathy (BSE) or transmissible spongiform encephalopathy (TSE) despite rigorous purification processes (20, 21).
In recent years, decellularized fish skin and marine sourced collagen have emerged as promising biomaterials for wound healing applications. Fish-derived collagen, particularly from species such as Atlantic cod and salmon, has demonstrated significant biocompatibility and favorable properties that enhance tissue regeneration (22, 23). Furthermore, fish collagen has been shown to promote cellular migration and proliferation as well as angiogenesis and therefore fundamental components of wound healing (24-26). Additionally, marine sourced collagen is derived from the fishing industry’s byproducts, making it a sustainable and cost-effective alternative to mammalian sources (27). As the natural collagen is fully biodegradable, fish skin grafts and membranes are naturally resorbed over time, thereby reducing the need for painful dressing changes and lowering the risk of secondary infections. An important key advantage of fish collagen over mammalian sources is its lower risk of immunogenic reactions, given its relatively low antigenicity, as well as the absence of risk of transmission of animal diseases like BSE or TSE (28). Its widespread acceptance in diverse patient populations, due to the lack of religious restrictions, further enhances its applicability. Furthermore, the porous structure of fish collagen supports moisture retention and cell migration, facilitating an ideal healing environment (29, 30). However, a potential drawback of marine collagen is its faster degradation rate in mammalian wounds compared to bovine or porcine collagen, due to its lower degradation temperature. This is caused by the lower body and external temperature of the fish compared to mammals and can be through chemical crosslinking of the collagen to enhance its durability (31).
The following research focuses on the multifactorial characterization of a wound membrane made from decellularized skin of salmon for the treatment of chronic wounds. The marine collagen was further cross-linked after decellularization, using 1-ethyl-3-(3-dimethylaminopropyl) Carbodiimide (EDC) and N-hydroxysuccinimide (NHS), to further increase the membranes mechanical stability, its degradation time in the wound bed as well as its protective function from external noxae, resulting in a longer regeneration-promoting effect (32, 33).
The aim of the present research is a comprehensive evaluation of the fish-derived wound membrane using established ex vivo, in vitro and in vivo methodologies according to the standardized DIN EN ISO 10996 −5/−6 and −12 (34, 35). An already commercially established collagen-based wound membrane derived from bovine pericardium served as a control material in all tests conducted.
Initially, the membranes’ surface as well as their inner structures were analyzed using scanning electron microscopy (SEM). Further analyses focused on a comparison of the collagen molecular structure of both membranes via attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) analysis and a comparison of protein adsorption via measurement of human serum albumin (HSA) adsorption. The membranes influence on cell viability, cell proliferation as well as their cytotoxic potential were examined in vitro on L929 mouse fibroblasts. Additionally, the membranes tissue integration, degradation behavior and biocompatibility were investigated under histopathological examination methods in a subcutaneous implantation model using Wistar rats. The aim was to comprehensively characterize the suitability of the fish collagen membrane for use in the treatment of chronic wounds.
Materials and Methods
Biomaterials. For this study, a decellularized fish skin graft (Kerecis® Omega3 Wound, Kerecis Ltd., Isafjordur, Iceland) and a porcine pericardium collagen membrane (Jason membrane®, botiss biomaterials GmbH, Zossen, Germany) were utilized.
Scanning electron microscopy (SEM). Microscopic analysis was conducted using a focused ion beam electron scanning microscope XB550L (Zeiss, Oberkochen, Germany). Membranes were carefully exercised with a sharp scalpel and mounted on specimen holders to examine both their surface and cross-section. The membranes were analyzed in their untreated state, without any coating or additional treatment. The conducted SEM images were performed using a voltage of 1.5 kV combined with a sample current of 6 pA at a working distance of 2.7 mm. All images and investigations were conducted using a secondary electron detector (SE) for enhanced surface and topography characterization of the membranes.
Attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) analysis. ATR-FTIR spectral analysis of both examined membranes was performed on a PerkinElmer Frontier spectrometer with a diamond ATR. For each spectrum of dry membranes, 16 scans were collected with a spectral resolution of 4 cm−1 over the wavenumber range 4,000-650 cm−1, at a temperature of 25°C.
Adsorption of proteins. The propensity of collagen membranes to adsorb surrounding proteins was examined with human serum albumin (HSA, Sigma-Aldrich Chemie GmbH, Steinheim, Germany). Therefore, the membranes were incubated in a solution of HSA (1 mg/ml) in Tris buffer at pH 7.4 (Sigma-Aldrich, Steinheim, Germany) for 2 h. The residual concentration of HSA in the incubation solution was then analyzed following the Bradford assay (36). Each type of membrane (fish membrane and bovine pericardium membrane) was tested in triplicate.
In vitro analysis. To assess cytocompatibility, indirect colorimetric assays were conducted according to DIN EN ISO 10993-5:2009/−12 2012 standards, as detailed in previous publications (34, 35, 37). To carry out the assays, obtaining extracts from the materials to be examined is essential.
Therefore, each material was immersed in cell culture medium (minimum essential medium with 10% fetal calf serum, 1% penicillin/streptomycin, and 2 mM L-glutamine) at a ratio of 1 ml medium to 0.1 g of the analyzed membrane. The mixtures were incubated for 24±2 h under standard cell culture conditions (37°C, 5% CO2, 95% humidity). Sodium dodecyl sulfate (SDS) (Sigma-Aldrich, Steinheim, Germany) served as positive control with verifiable cytotoxic qualities and was incubated similarly with an extraction rate of 0.2 mg to 1 ml medium. The cell culture medium as described above was utilized as both negative and blind control.
L-929 mouse fibroblasts (Cell Lines Service GmbH, Eppelheim, Germany) were seeded in 96-well plates (105 cells/well) and incubated for 24±2 h under standard cell culture conditions as described above. After incubation, extracts were separated from the materials and applied to the fibroblasts (100 μl extract/well). Cells and extracts were incubated for an additional 24±2 h. Subsequently, a bromodeoxyuridine (BrdU) cell proliferation assay (Roche Diagnostics GmbH, Mannheim, Germany), a lactate dehydrogenase (LDH) cytotoxicity assay (PromoCell, Heidelberg, Germany), and a 2,3-bis-(2-methoxy-4-nitro-5-sulphenyl)-(2H)-tetrazolium-5-carboxanilide (XTT) cell viability assay (Thermo Fisher Scientific, Waltham, MA, USA) were performed according to the individual test protocols as supplied by the manufacturers.
In vivo analysis. The in vivo implantation procedure as well as the histopathological analysis were performed based on previously published protocols established by Barbeck et al. (38-40). In brief, this analysis step was performed in cooperation with the Faculty of Medicine of the University of Niš, Serbia, based on the approval by the Local Ethical Commission of the Faculty of Medicine Niš and the Veterinary Directorate of the Ministry of Agriculture, Forestry and Water Management of the Republic of Serbia (number 323-07-09101/2020-05/5, date of approval: 26/08/2020). According to the DIN EN ISO/IEC 10993-6, the fish collagen membrane and the bovine pericardium membrane were subcutaneously implanted into the rostral subscapular areas of Wistar rats. In total, 20 male Wistar rats, 12 to 14 weeks old with an average weight of 250±20 g were used to conduct the study. Five animals per group (i.e., membrane) and per time point were used. At predetermined time points (10- and 30-days post-implantation), the animals were euthanized, and the biomaterials in combination with the adjacent tissue was removed, fixed in 4% formalin, and histologically processed. The tissue samples were dehydrated through a graded alcohol series, treated with xylene, embedded in paraffin, and sectioned into 3-5 μm slices. These sections were stained with hematoxylin and eosin (HE). To assess the tissue compatibility and tissue responses, the slides were histologically analyzed focusing on membrane integration behavior in the implant beds was conducted as well as histopathologically scored in accordance with the appropriate DIN EN ISO standard 10993-6 (39). Thereby, a variety of criteria were assessed for the safety of the histology sections. Tissue analysis focused on the presence of granulocytes, lymphocytes, plasma cells, macrophages and multinucleated giant cells in combination with tissue parameters like fibrosis, vascularization and fatty tissue infiltration. Based on the numbers or occurrence of the cell type and tissue responses, the sections were examined and ranked following the irritancy/reactivity grading system from the Annex E of ISO 10993-6. Finally, the irritancy score for each test or control treatment was then calculated by averaging the irritancy scores of all test or control defect sites for each treatment, respectively. Each irritancy/reactivity score was calculated as follows: Test Article Irritancy Score-Control Article Irritancy Score= Irritancy/Reactivity score.
Statistical analysis. Raw scoring data were statistically analyzed using the Mann-Whitney U test to compare two independent groups. The in vitro data were analyzed using an analysis of variance (ANOVA) followed by an LSD post hoc test. Both calculations were conducted with GraphPad Prism version 10.2 (GraphPad Software Inc., La Jolla, CA, USA). Statistical significance was defined as p≤0.05 (*p≤0.05), high significance as p≤0.01 (**p≤0.01), and very high significance as p≤0.001 (***p≤0.001). Results are expressed as mean±standard deviation.
Results
Scanning electron microscopy (SEM). Scanning electron microscopy facilitates the initial assessment of membrane surfaces and their ultrastructure. At 100×magnification, the fish collagen membrane exhibits a fissured, irregular, and partially porous surface (Figure 1A). In contrast, the bovine pericardium collagen membrane surface appears more homogeneous, smoother, and significantly less porous (Figure 1B).
Scanning electron microscopy (SEM) images of the surface (A, B) and inner (C-F) morphology of the analyzed membranes. Left column: fish collagen membrane. Right column: bovine pericardium membrane. A-D: 100× magnification, E, F: 1,000× magnification. Scale bars: 200 μm (A–D) resp. 20 μm (E and F).
Regarding the inner structure, the fish collagen membrane sections reveal densely stacked layers interrupted by few irregularly arranged and variably sized pores, ranging between 80 and 200 μm. Furthermore, the cross-section shows highly compact opposing surfaces with minimal to no porosity, aside from a few truncated artifacts on the upper membrane surface (Figure 1C and E). In comparison, the bovine pericardium membrane lacks a layered structure and appears highly porous on the upper half of the membrane, characterized by an uneven distribution and varying sizes of pores, ranging between 50 and 150 μm (Figure 1D and F). In contrast, the lower half of the membrane appears highly compact and completely free of pores.
ATR-FTIR analysis. FTIR spectroscopy was applied to reveal structural information about the collagen molecules in both collagen membranes. The spectra of both the fish collagen membrane and the pericardium membrane exhibit a high degree of similarity (Figure 2), with both displaying all characteristic protein bands. NH stretching vibrations of side chains gave rise to the amide A band (centered at 3,288 cm−1 for fish collagen membrane and at 3,303 cm−1 for the pericardium material) and amide B band (located at 3,075 cm−1 for both membranes). The downshift of the amide A band in case of the fish collagen membrane to lower wavenumber (red shift) is commonly accepted as evidence of increased hydrogen bonding with water molecules (41). The amide I band primarily arising from the stretching vibrations of the amide carbonyl groups in the polypeptide backbone, was found at 1,634 cm−1 for both membranes. The amide II band, related to the NH in-plane as well as CN stretching vibrations, appeared at 1,542 cm−1 for the fish collagen membrane and 1,546 cm−1 for the pericardium material, demonstrating again a red shift and a higher degree of hydrogen bonding in the fish membrane. Amide III band, attributed to CN stretching and NH bending, was found at the same wavenumber 1,237 cm−1 in both membranes.
Comparative ATR-FTIR spectra of the fish collagen membrane (cyan) and the bovine pericardium membrane (grey).
The presence of the peak at 1,452 cm−1, associated with CH bending, indicated that collagen retains its triple helical structure in both membranes. Furthermore, the A1237/A1452 ratio, a very sensitive indicator of the preservation of the collagen’s tertiary helical structure (42), was calculated as 1.15±0.06 for the fish membrane and 1.15±0.07 for the pericardium membrane. These values closely align with those previously reported for collagen in a fully triple helical conformation (43, 44).
Human serum albumin (HAS) adsorption measurement. The surface density of HSA adsorbed to the fish collagen membrane was 152±18 μg/cm2, while on the bovine pericardium membrane it was significantly higher (201±11 μg/cm2, p<0.05) (Figure 3).
Results of human serum albumin (HSA) surface density measurements as indicator for protein adsorption tendency of both membranes. Means with error bars indicating standard deviations. Significant differences are indicated accordingly (*p<0.05 using the Mann–Whitney U test). BoviPer: Bovine pericardium membrane; FishMem: fish membrane.
In vitro analysis. Cytocompatibility is defined in DIN EN ISO 10993-5/−12 for materials that do not fall below a measured cell viability (XTT assay) or cell proliferation (BrdU assay) of 70% compared to a negative control or do not exceed a cytotoxicity (LDH assay) of 70% compared to a positive control in the colorimetric assay (34, 37). In the present study, the fish collagen membrane was compared with a membrane based on bovine pericardium collagen. For the XTT and BrdU assays, pure cell medium was used as negative control and as reference, while SDS served as positive control and reference material in the LDH assay.
In the XTT assay, both membranes reached the range defined as cytocompatible and differed significantly (p <0.001) both from the negative control and from each other, with better cell viability being demonstrated for the fish collagen membrane (Figure 4A).
Cytocompatibility assay results using L929 fibroblasts. (A) Viability measured by 2,3-bis-(2-methoxy-4-nitro-5-sulphenyl)-(2H)-tetrazolium-5-carboxanil (XTT) assay. (B) Cytotoxicity measured by lactate dehydrogenase (LDH) assay. (C) Proliferation measured by bromodeoxyuridine (BrdU) assay. Values were matched against the medium control (MC) for XTT and BrdU arrays, and against the positive control (PC) for the LDH assay. Means with error bars indicating standard deviations. Dotted lines define thresholds that should be met (XTT, BrdU) or not exceeded (LDH). Significant differences are indicated accordingly (***p<0.001 using ANOVA followed by the LSD post hoc test). FCM: Fish collagen membrane; BCM: bovine collagen membrane.
In the LDH assay, neither the fish collagen membrane nor the bovine pericardium membrane exceeded the limit range defined as cytotoxic (Figure 4B). Both membranes differed significantly (p<0.001) from the SDS-positive control and from each other, with higher values measured for the fish collagen membrane compared to the bovine pericardium membrane. In the BrdU assay, both membranes reached the cytocompatible limit and differed significantly (p<0.001) from the medium control and from each other (Figure 4C). Thereby, higher proliferation was measured for the bovine pericardium membrane compared to the fish collagen membrane.
In vivo analysis. The in vivo analysis showed that both the fish collagen membrane and the pericardium membrane were detectable within the implantation beds at day 10 post-implantation (Figure 5 and Figure 6). In case of the fish collagen membrane the central part of the material was found free of migrated cells (Figure 5A), while even in the surface regions of the material a higher cell migration in concert with moderate numbers of blood vessels was found (Figure 5A and B). Within the surface regions of the fish collagen membranes, a moderate cell influx and tissue reactions mainly associated with materials fragments were observed (Figure 5C and D). Thereby, high numbers of lymphocytes and macrophages were visible between moderate numbers of polymorphonuclear cells, plasma cells and multinucleated giant cells. Also, moderate correlations of necrosis and vascularization as well as no signs of fatty tissue infiltration were observable.
Representative histological microphotographs from the implantation bed of fish collagen membrane (FCM) at (A-D) 10 and (E-G) 30 days post-implantation within the subcutaneous connective tissue (CT). GT: Granulation tissue; white arrows: material fragments; blue arrow: cell migration, red arrows: vessels; black arrows: macrophages; yellow arrows: lymphocytes; green arrows: eosinophils; white triangles: necrotic enlarged macrophages (HE-stainings, A and E: 10× magnifications, scalebars=50 μm; B and F: 20× magnifications, scale bars=20 μm; C, D and G: 40× magnifications, scalebars=20 μm).
Histopathological images of the bovine pericardium collagen membrane (BCM) at (A and B) 10 and (C and D) 30 days post-implantation within the subcutaneous connective tissue (CT). Red arrows: Vessels; black arrows: macrophages; yellow arrows: lymphocytes; green arrows: granulocytes (HE-stainings, A and C: 10× magnifications, scalebars=50 μm; B and D: 40× magnifications, scalebars=20 μm).
The analysis furthermore revealed that different integration patterns as well as degradation pattern were found in both study groups at day 30 post-implantation (Figure 5 and Figure 6). In the group of the fish collagen membrane an almost complete degradation was observable and only minimal band-shaped remnants of the former membrane were found (Figure 5E-G). The remnants of the fish collagen membrane were surrounded by a cell- and vessel-rich granulation tissue (Figure 5E and F). Mainly macrophages, lymphocytes and granulocytes were involved in this tissue reaction pattern (Figure 5G). Moreover, moderate numbers of macrophages with an increased cytosol volume and a brownish staining of the cytosol were found as components of the granulation tissue (Figure 5G).
In contrast, the bovine pericardium membrane induced a moderate inflammatory tissue reaction both at day 10 and 30 post-implantation (Figure 6). Mainly macrophages were found at the material surfaces beside lower numbers of lymphocytes and granulocytes in concert with moderate vessel numbers that were mainly located within the surrounding connective tissue (Figure 6B and D). The material center was cell-free at both study time points, while a beginning cell infiltration within the superficial membrane regions was especially detectable at day 30 post-implantation (Figure 6).
The histopathological scoring resulted in the irritancy scores that revealed that the fish collagen membrane had an average treatment irritancy score of 18.3, and the bovine pericardium membrane as control article had an average treatment irritancy score of 14.3 at 10 days post-implantation. Thus, the overall irritancy score for the fish collagen membrane was 4.0 so that this material was considered to be slightly irritant. At 30 days post-implantation the fish collagen membrane had an average treatment irritancy score of 18.1, and the bovine pericardium membrane as control article had an average treatment irritancy score of 14.53. Thus, the overall irritancy score for the fish collagen membrane was 3.5 so that this material was still considered to be slightly irritant.
The statistical analysis of the scoring data revealed that the fish membrane induced a statistically significant (p <0.0001) lower number of polymorphonuclear cells, but a significantly (p<0.0001) higher number of lymphocytes in comparison to the occurrence in the group of the bovine pericardium membrane at day 10 post-implantation (Figure 7A and B). There were no differences between the scoring values in the groups of both membranes regarding all other cell or tissue parameters (Figure 7C-G). At day 30 post-implantation, only a significantly (p<0.001) higher number of multinucleated giant cells was observed in the group of the bovine pericardium membrane (Figure 7E), while no other statistically significant differences were measured in the cell numbers of tissue parameters (Figure 7).
Results of the scoring analysis at 10- and 30-days post-implantation according to the DIN EN ISO 10993-6 scoring system. Significant differences are indicated accordingly (***p<0.01 and ****p<0.001 using the Mann–Whitney U test).
Discussion
The aim of the present study was to comprehensively characterize the cyto- and biocompatibility of a fish collagen membrane derived from cross-linked and decellularized fish skin tissue through established ex vivo, in vitro, and in vivo methodologies (34, 35, 45). Therefore, scanning electron microscopy (SEM) as well as Fourier transform infrared spectroscopy (FTIR) were initially used to examine the membranes’ microstructure and collagen molecular structure in comparison to a commercially available wound membrane made from bovine pericardium that was utilized as control material. This data was supplemented by well-established biological analysis methodologies, i.e., human serum albumin (HSA) adsorption measurement in combination with cytocompatibility assays as well as detailed in vivo analysis of the tissue integration and degradation behavior using the subcutaneous animal model (45).
Initially, the SEM analysis revealed a rough, porous and heterogeneous surface of the fish collagen membrane, as well as a compact, layered structure with sporadically large and unevenly distributed pores. In contrast, the bovine pericardium membrane showed a smoother, more homogeneous surface, with a significantly more porous upper half and a very compact lower half. In this context, the relationship between the structure of a wound membrane and its influence on cell migration and proliferation plays an important role in biomaterial research (46-48). Altogether, it is stated that an optimal collagen-based biomaterial for wound healing should exhibit a standing time, which allows cellular remigration, ECM and collagen formation in concert with a sufficient revascularization but also for re-epithelialization (49, 50).
In view of fish collagen membrane, its homogeneous porosity might facilitate cellular migration but also efficient gas and fluid exchange within the wound area, ensuring adequate oxygen supply and moisture control to the wound bed, which should combinatorially promoting wound healing (51, 52). Thereby, a material porosity with pore sizes ranging from 100 to 200 μm has shown to promote cell adhesion and migration into skin wounds, thus further facilitating the wound healing process (53, 54). The SEM analysis revealed pore sizes ranging from approximately 80 to 200 micrometers in case of the fish collagen membrane. Thus, this porous framework may serve as a scaffold for wound healing based on a migration support of fibroblasts and inflammatory cells, further improving their colonization and migration, as well as facilitating the angiogenesis in the wound bed (55, 56). These data are similar to data obtained by Kangning et al., who developed a wound membrane based on decellularized tilapia skin (23). The researchers observed a densely packed outer layer alongside an overall porous and loosely structured inner layer of the tilapia-based membrane. They concluded that the porous inner structure promotes cellular migration, while a denser outer layer aids in wound closure, protecting against external influences. Thus, the fish collagen membrane could optimally serve to promote cell and vessel migration into the wound bed, which is a crucial factor of wound healing. In contrast, the bovine pericardium membrane, which has a denser structure, should not allow for a fast cellular migration and no early material vascularization. However, the results of the SEM analysis may also lead to a different conclusion - based on various other studies on the biodegradation of biomaterials based on collagen and also bone substitutes. Further results of the phagocytosis process of biomaterials mediated by macrophages (and multinucleated giant cells) have revealed that a fast and high cellular migration in combination especially with induction of pro-inflammatory M1-macrophages might led to a fast or premature biodegradation (57, 58). Thus, further analyses especially focusing on the chemical characterization and the in vivo response were conducted in this study.
The analysis via FTIR spectroscopy revealed that collagen molecules are present in both membranes in their triple helical structure in a comparable fully helical conformation. Thus, these results showed that the purification and/or decellularization processes of both analyzed materials have been successfully applied to ensure optimal collagen nativity, which is favored for the associated tissue reactions and desired biocompatibility (59, 60). In this context, it has shown that the tissue responses to two of the most widely used GBR membranes, which are both based on native collagen, include mainly fibroblasts, macrophages, and eosinophils (61). These cell types seem to mediate the breakdown of natural collagen membranes, and it was claimed that this resorption mechanism indicates complete biocompatibility (61). Thus, further analysis steps have to prove this assumption.
The next analysis step included the measurement of the HSA attachment onto the surfaces of the collagen membranes. Regardless of the type of the implanted biomaterial and its place of implantation the initial stage taking place at the implant surface within minutes post implantation is the adsorption of proteins from blood and other biological fluids - known as the “Vroman effect” (62). The layer of adsorbed proteins and also the conformation of the proteins is directly dependent on the physicochemical characteristics of the material surface (63, 64). This protein layer, the specifically bound proteins, their conformation and the resulting material-specific exposed cell binding sites then also cause surface-dependent cellular (inflammatory) tissue reactions (65, 66). In view of this reaction cascade, the adsorption of human serum albumin (HSA) to the surfaces of the analyzed membranes was analyzed as albumin is the most abundant serum protein and thus the main component of the most biomaterial protein coronae (67). In this context, Visalakshan et al. analyzed the biomaterial surface hydrophobicity-mediated serum protein adsorption and the associated immune responses, and the data showed that hydrophilic surfaces had a higher affinity to albumin being 45% of the total protein corona (68). Moreover, this study interestingly revealed that this surface characteristics regulated the response of macrophages towards an anti-inflammatory M2-phenotype. This knowledge can be related to different studies that showed the anti-inflammatory tissue reactions to collagen biomaterials (61, 69). The adsorption measurements of the present study showed that the surface density of HSA adsorbed to the fish collagen membrane surfaces was significantly lower compared to the values measured for the bovine pericardium membrane. Based on the afore-mentioned relationship it could thus be concluded that the bovine pericardium membrane is more likely to induce a M2-response and slower biodegradation, as biodegradation is more likely to be linked to the presence of M1 macrophages (70). In contrast, it could be expected that the fish collagen membrane underlies a faster biodegradation due to higher M1-macrophage levels.
Furthermore, both membranes demonstrated sufficient in vitro cytocompatibility in the indirect colorimetric XTT, LDH and BrdU assays. The fish collagen membrane performed marginally better in the XTT viability assay but exhibited slightly more pronounced cell irritative properties in the LDH assay. Also, cell proliferation in the BrdU assay was slightly lower for the fish collagen membrane compared to the bovine pericardium membrane. In this context, it has been shown by Zhang et. al. that there is significantly higher cell viability in the presence of fish collagen compared to the bovine collagen control material in the colorimetric MTT assay using human foreskin fibroblast (HFF-1) cells (71). However, Tilapia was used as the source of fish collagen instead of salmon. Nevertheless, these results are consistent with the XTT results obtained in the present study, suggesting a significant improvement in cell viability due to fish collagen. This positive effect on connective tissue cells and wound healing has also been confirmed for various other marine collagen sources (72).
The in vivo biocompatibility analysis could interestingly confirm the results of the HSA adsorption study part as the fish collagen membrane induced a significantly higher proinflammatory tissue reaction related to a much faster biodegradation that was nearly completed up to 30 days post-implantation. Interestingly, the histopathological analysis also revealed other significant differences. Thus, significantly fewer polymorphonuclear cells but more lymphocytes were measured within the implantation beds of the fish collagen membrane compared to the bovine pericardium membrane at 10 days post-implantation. At 30 days post-implantation, the cellular environments of both wound beds were nearly identical, except for a higher number of multinucleated giant cells surrounding the bovine pericardium membrane and higher numbers of plasma cells in case of the fish collagen membrane. Overall, these findings suggest different immunological interactions of both membranes with the host tissue. The high occurrence of granulocytes in the implantation beds of the bovine pericardium membrane could be related with early neutrophil immune responses to biomaterials that have been associated with modulation of M2 macrophages towards a material-related tissue healing. In this context, it has also been shown that a bovine pericardium-based barrier membrane for guided bone regeneration (GBR) was found to have ossified via osteoconductive properties but also induced a tissue response mainly dominated by M2-macrophages (73). The lower presence of polymorphonuclear cells or eosinophils within the implantation beds of the fish membrane may indicate a reduced “recognition” of the fish collagen in contrast to the mammalian collagen in case of the pericardium membrane. In this context, it has been shown that the physiological collagen turnover and the biodegradation of native collagen biomaterials involves eosinophils, macrophages and fibroblasts (61). This reaction pattern might be related to the chemical difference or the differences in the amino acid sequences between fish and mammalian collagen and a related recognition of the fish collagen as a foreign body (71). This fact is also supported by the higher number of lymphocytes, whose occurrence might suggest a more biomaterial-associated specific immune response. Another explanation can be found focusing on the foreign body reaction cascade to biomaterials as it has been described that cytokines modulating both the pro- and anti-inflammatory responses are mainly produced by (T-) lymphocytes (74-76). Based on the fast cell- or macrophage-mediated biodegradation the high occurrence of lymphocytes could have been linked with this process as it is known that this cell type also enhances macrophage adhesion and activation via the aforementioned cytokines (74, 76).
Altogether, fish collagen membranes have shown to promote wound tissue healing in preclinical and clinical studies (77, 78). The present study could explain the mode of action of this material type. In this context, the material-induced vessel-rich granulation tissue might contribute to tissue regeneration in skin wound healing but also in different other indications. However, the rapid degradation of fish skin is a crucial consideration for future studies. Generally, marine collagens are known for their fast degradation rate compared to mammalian collagens (79). Therefore, additional cross-linking of fish skin is necessary for prolonging its degradation time in the wound bed (80). Future studies should explore alternative natural cross-linking agents to address this challenge.
Conclusion
The present study has demonstrated that fish collagen membranes exhibit promising properties for wound healing. In particular, their porous structure facilitates cell migration and gas exchange, creating an optimal environment for tissue regeneration. The chemical analysis using FTIR further confirmed the successful decellularization and preservation of the native collagen structure in both tested membranes, which favors good biocompatibility.
However, it was observed that the fish collagen membrane induces a faster biodegradation and a more pronounced pro-inflammatory tissue response, whereas the bovine pericardium membrane degrades more slowly and promotes a healing response predominantly mediated by M2 macrophages. These differences in immune response could be ascribed to the specific structural and chemical properties of fish and mammalian collagen.
In summary, the results of this study highlight the potential of fish collagen membranes as biocompatible wound healing materials. However, their rapid degradation presents a challenge that needs to be addressed through targeted modifications, such as optimized cross-linking. Future studies should explore alternative natural cross-linking methods to enhance the clinical applicability of this material.
Acknowledgements
The Authors gratefully acknowledge the funding by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) for the research unit 5250 “Mechanism-based characterization and modeling of permanent and bioresorbable implants with tailored functionality based on innovative in-vivo, in-vitro and in-silico methods” (project no. 449916462).
Footnotes
Authors’ Contributions
Conceptualization: O.J. and M.B.; methodology: O.J., S.S., S.S., S.N., R.K., F.W. and M.B.; SEM: S.S. and M.M.B.; ATR-FTIR and HAS adsorption: T.A. and R.K.; in vitro analyses: O.J., C.R. and S.P.; in vivo analyses: O.J., K.B., M.R.S.; S.S., S.N. and M.B.; software: M.B.; validation, O.J. and M.B.; formal analysis: O.J., K.B., S.S., M.M.B., C.R., T.A., M.R.S., R.K., S.P. and M.B.; investigation: O.J., K.B., S.S., M.M.B., C.R., T.A., M.R.S., R.K., S.P. and M.B.; resources: M.B., S.S., S.N., R.K., F.W. and M.B.; data curation: O.J., K.B., S.S., M.M.B., C.R., T.A., M.R.S., R.K., S.P. and M.B.; writing - original draft preparation: O.J., K.B., S.S., M.M.B., T.A., R.K., S.P. and M.B.; writing - review and editing: O.J., K.B., S.S., T.A., R.K., S.P. and M.B.; visualization: O.J., K.B., S.S., M.M.B., T.A., R.K., S.P. and M.B.; supervision: O.J., S.S., S.N., R.K., F.W. and M.B.; project administration: M.B.; funding acquisition: O.J., S.S., S.N., R.K., F.W. and M.B. All Authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
All the Authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
- Received April 8, 2025.
- Revision received May 20, 2025.
- Accepted May 22, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
Jump to section
Related Articles
Cited By...
- Poly-lactic Acid Dermal Filler Enriched With InCube(R)-encapsulated Vitamin C and Glutathione: Preclinical Evidence of Superior Collagen Induction and Tissue Response Compared With Conventional PLA
- Granulocytes and their Involvement in the Foreign Body Response to Biomaterials and Tissue Repair
- Comparative Evaluation of Biocompatibility Data from the Subcutaneous and the Calvaria Implantation Model of a Novel Bone Substitute Block