Development of porcine skeletal muscle extracellular matrix–derived hydrogels with improved properties and low immunogenicity

Significance

This research addresses a critical gap in regenerative medicine by investigating the impact of decellularization protocols on hydrogels derived from animal ECM. These hydrogels, essential for tissue engineering, exhibit promising biomimetic properties. The study’s innovation lies in the comparison of decellularization methods and their influence on hydrogel properties and immunogenicity, focusing on hydrogels derived from porcine skeletal muscle ECM as a model. The mechanical disruption decellularization protocol, coupled with α-galactosidase treatment, emerges as a superior method, efficiently removing cells and xenoantigens, preserving essential biochemical components, and promoting cytocompatibility and low immunogenicity. This advancement is pivotal for developing low-immunogenic ECM-derived hydrogels, with enhanced properties crucial for regenerative engineering and clinical applications, bringing us closer to effective and safe tissue regeneration.

Abstract

Hydrogels derived from decellularized extracellular matrices (ECM) of animal origin show immense potential for regenerative applications due to their excellent cytocompatibility and biomimetic properties. Despite these benefits, the impact of decellularization protocols on the properties and immunogenicity of these hydrogels remains relatively unexplored. In this study, porcine skeletal muscle ECM (smECM) underwent decellularization using mechanical disruption (MD) and two commonly employed decellularization detergents, sodium deoxycholate (SDC) or Triton X-100. To mitigate immunogenicity associated with animal-derived ECM, all decellularized tissues were enzymatically treated with α-galactosidase to cleave the primary xenoantigen—the α-Gal antigen. Subsequently, the impact of the different decellularization protocols on the resultant hydrogels was thoroughly investigated. All methods significantly reduced total DNA content in hydrogels. Moreover, α-galactosidase treatment was crucial for cleaving α-Gal antigens, suggesting that conventional decellularization methods alone are insufficient. MD preserved total protein, collagen, sulfated glycosaminoglycan, laminin, fibronectin, and growth factors more efficiently than other protocols. The decellularization method impacted hydrogel gelation kinetics and ultrastructure, as confirmed by turbidimetric and scanning electron microscopy analyses. MD hydrogels demonstrated high cytocompatibility, supporting satellite stem cell recruitment, growth, and differentiation into multinucleated myofibers. In contrast, the SDC and Triton X-100 protocols exhibited cytotoxicity. Comprehensive in vivo immunogenicity assessments in a subcutaneous xenotransplantation model revealed MD hydrogels’ biocompatibility and low immunogenicity. These findings highlight the significant influence of the decellularization protocol on hydrogel properties. Our results suggest that combining MD with α-galactosidase treatment is an efficient method for preparing low-immunogenic smECM-derived hydrogels with enhanced properties for skeletal muscle regenerative engineering and clinical applications.

The extracellular matrix (ECM) is the acellular component of any tissue or organ, produced and maintained by resident cells (1). Numerous cellular functions, such as attachment, proliferation, migration, and differentiation, rely on the ECM (1). Comprising both structural and nonstructural proteins, the ECM exhibits significant variation across diverse tissues (1). In recent years, extensive efforts have been devoted to understanding the distinctive properties of each tissue’s ECM, resulting in a comprehensive knowledge of its structural, mechanical, and compositional characteristics in a tissue-specific manner (2–5).

This knowledge plays a pivotal role in regenerative engineering, where the objective is often to emulate the native ECM’s composition, structure, and mechanical properties in the scaffold design to facilitate the regeneration of damaged tissues or organs (6–15). Currently, various natural and synthetic scaffold materials have been developed using the biochemical, mechanical, and structural cues from the native ECM as a blueprint (16, 17). Despite these advancements, existing natural or synthetic materials remain inadequate in fully replicating the intricate complexity inherent in the natural ECM.

Given that the native ECM provides a bioactive milieu with all the essential components necessary for proper cell function, utilizing the native ECM devoid of cells presents an optimal cellular niche for tissue regeneration (18). Achieving this state can seamlessly be accomplished through decellularization. Decellularization can broadly be defined as the process of removing the cellular content from living tissues while retaining the bioactivity of the ECM (19). Therefore, decellularization emerges as a valuable tool for obtaining acellular bioactive scaffolds that robustly support cellular functions both in vitro and in vivo (19).

Various processes can be employed for ECM decellularization, including physical methods (such as high hydrostatic pressure, supercritical CO₂, or repeated freezing and thawing), chemical substances (including detergents or chemical solvents), and biological reagents (such as enzymes) (20). As the primary objective of decellularization is to facilitate efficient cell removal with minimal ECM damage, extensive research has been dedicated to refining these procedures in a tissue-specific manner (20).

However, a limitation associated with the use of decellularized ECM (dECM) scaffolds in their original form is their confinement to the geometrical features of the tissue of origin (18). Indeed, the development of an injectable form of dECM introduces adaptability to diverse 3D shapes upon injection (21). In this context, gradual advancements in decellularization technology have prompted the transformation of dECM scaffolds into hydrogels, significantly expanding their potential applications in tissue regeneration and repair. These applications include serving as injectable materials to fill irregular defects or as bioinks for 3D bioprinting (22). Despite the growing interest in this exceptional class of hydrogels, the impact of the decellularization protocol on the characteristics of the resulting hydrogels has not been extensively explored. Additionally, these hydrogels are frequently derived from xenogeneic tissues, which inherently contain xenoantigens. The presence of xenoantigens in animal-derived materials has been identified as a primary cause of xenograft rejection in humans (23). Surprisingly, the effectiveness of various decellularization protocols in eliminating these xenoantigens from the resulting hydrogels, and whether decellularization alone is sufficient for their cleavage, remains largely unexplored to date.

The aim of this study was to assess the impact of various decellularization protocols on the properties and immunogenicity of hydrogels derived from porcine skeletal muscle ECM (smECM), with the overarching objective of achieving cell and xenoantigen-free smECM hydrogels possessing enhanced overall properties for skeletal muscle regenerative engineering and clinical applications. Toward this goal, detergent-based techniques [sodium deoxycholate (SDC) and Triton X-100] and a MD (mechanical disruption) technique (freeze–thaw cycling) were employed to decellularize the smECM, and the resulting dECM was utilized in hydrogel fabrication. Subsequent evaluations focused on the efficacy of different decellularization protocols in terms of DNA and xenoantigen removal, preservation of biochemical composition, gelation kinetics, cytocompatibility, and myoinductivity of the resulting hydrogels. Additionally, an in vivo xenotransplantation subcutaneous animal model was employed to assess the immunogenicity of these hydrogels.

Results and Discussion

smECM Decellularization and Hydrogel Formation.

Since the notion that dECM components could be liquified and further processed to form hydrogels (22), numerous studies have detailed the development of hydrogels from dECMs originating from various tissues (24). To date, successful synthesis of dECM-derived hydrogels has been achieved for a wide range of tissues, including heart (25), lung (26), pancreas (27), skin (28), urinary bladder (28), bone (29), blood vessels (30), brain (31), fat (32), and nerves (33). However, as ECM components vary among tissues, hydrogels derived from tissue-specific ECMs exhibit superior capacity to elicit physiologically relevant cellular responses compared to those from nonmatching tissue sources (34). Therefore, for skeletal muscle regenerative engineering and clinical indications, an ideal scenario involves deriving hydrogels from the ECM of muscle origin. However, to harness the potential benefits of such tissue-specific hydrogels, it is crucial to employ a decellularization protocol that maximizes cell removal while preserving intrinsic tissue-specific cues.

Given the distinct compositional properties across various tissues (2), the decellularization protocol must be tailored to each tissue, efficiently removing cellular components while preserving ECM molecules relevant to the intended use (35). For instance, tissues like cartilage and bone tolerate relatively harsh treatment protocols, whereas more delicate tissues such as lungs and brain necessitate gentler decellularization methods to maintain ECM composition (36, 37). In this study, porcine smECM (SI Appendix, Fig. S1) underwent decellularization via three distinct protocols, each possessing a different mechanism of action (SI Appendix, Fig. S2). The first protocol, devoid of chemicals or detergents, relied on MD of the ECM through exposure to multiple freeze–thaw cycles, creating crystals that disrupted cell membranes and led to cell lysis (20). The second protocol utilized Triton X-100, a nonionic detergent disrupting cell cytoplasm and causing the release of intracellular components (e.g., DNA and organelles) (20). The third protocol involved treatment with SDC, an ionic detergent disrupting the ECM and causing cell detachment from the matrix (20).

Following decellularization, muscle tissues from the various protocols underwent incubation with isopropyl alcohol to remove any residual lipids in the sample that might potentially interfere with the gelation process (38). Additionally, along with the isopropyl alcohol treatment, the decellularized muscle tissues underwent treatment with deoxyribonuclease I (DNase) and ribonuclease (RNase). These enzymes catalyze the hydrolysis of DNA and RNA, respectively, thereby enhancing the overall efficiency of decellularization.

In contrast to the conventional mincing with a pair of scalpels, the muscle tissue in this study underwent a dual-grinding process using a manual meat grinder before decellularization (SI Appendix, Fig. S3). This approach aimed to reduce the size of the muscle tissue significantly, thereby enhancing the probability of effective cell detachment and removal. Additionally, it increased the action area for decellularization detergents and other reagents, optimizing their efficiency in the process.

Postdecellularization, all tissues exhibited a white appearance, a morphological confirmation of the removed cellular component, consistent with previous reports (27, 39, 40) (SI Appendix, Fig. S3A). In contrast, the native nondecellularized control (e.g., hydrogels from native skeletal muscles) appeared red. The tissues subjected to different decellularization protocols were then freeze-dried, cryomilled into a powder, and solubilized in an acidic-based pepsin solution, resulting in the formation of homogeneous and viscous solutions (SI Appendix, Fig. S3 B–D). After solubilization, the pH was neutralized to approximately 7.4, forming pregels with varying transparency (SI Appendix, Fig. S3E). Subsequently, the pregels were incubated at 37 °C to induce gelation. Regardless of the decellularization protocol employed, all pregels successfully formed hydrogels (SI Appendix, Fig. S3F).

Given that the decellularization process can substantially alter the composition of the ECM (41), potentially influencing the characteristics of the hydrogel and the behavior of encapsulated cells, all characterizations were conducted on the finally formed hydrogels. To exclusively assess the impact of the decellularization protocol on their overall characteristics, the resulting hydrogels were prepared at a consistent final ECM concentration of 4 mg/mL.

dsDNA and Xenoantigen Removal Confirmation.

The assessment of cellular removal is crucial for hydrogels derived from xenogeneic tissues, such as porcine skeletal muscles, to ensure biological safety and prevent immunological responses. Therefore, we meticulously examined the efficacy of different decellularization protocols for DNA removal using PicoGreen. DNA quantification demonstrated that all decellularization protocols significantly decreased the total DNA content in the resulting smECM hydrogels compared to the native nondecellularized control (Fig. 1A). Importantly, the DNA content in the decellularized smECM hydrogels fell below the established threshold of 50 ng/mg of dry ECM (40, 42, 43), indicating the effectiveness of all decellularization protocols in eliminating the cellular component. Furthermore, histological examinations provided additional validation. Staining with hematoxylin & eosin-Y (H&E) and DAPI confirmed the substantial reduction of nuclei after decellularization in all groups compared to the native nondecellularized control (Fig. 1B).

Fig. 1.

DNA quantitative and qualitative assessments. (A) PicoGreen was used to quantify the total DNA content in the finally formed hydrogels. (B) The efficacy of the different decellularization protocols in removing the DNA was qualitatively assessed via H&E and DAPI staining to detect the presence of any nuclei in the formed hydrogels. Images are presented at low (10×) and high (20×) magnifications. (Scale bars, 100 µm.)

While all the tested protocols demonstrated efficient removal of the cellular component, it is essential to note that this alone does not ensure the biological safety of hydrogels derived from animal ECM. Previous studies have indicated that decellularized animal grafts may elicit adverse immunological responses in humans due to the presence of other immunogenic elements, particularly xenoantigens (23). The primary xenoantigen responsible for hyperacute rejection of animal ECM-derived materials in humans is α-Gal (Galactose-α-1,3-galactose) (44–47). Remnants of the α-Gal antigen have been identified as a major factor in triggering adverse immunological responses in humans, potentially affecting tissue remodeling outcomes (48). The α-Gal antigen is prevalent in tissues of all nonprimate mammals, including pigs, rodents, bovines, rabbits, horses, prosimians, and New World monkeys (49). Therefore, to further mitigate the immunogenicity of materials derived from decellularized animal matrices, the decellularization protocol should ideally eliminate the α-Gal from the ECM (47). In this context, we assessed the efficacy of different decellularization protocols in removing α-Gal in the resultant hydrogels. Histological sections from each hydrogel group were subjected to immunohistochemical (IHC) staining with the M86 antibody, specifically targeting the α-Gal epitope (47). High expression of α-Gal epitope was detected in the native nondecellularized control group, confirming the presence of the α-Gal antigen in the animal ECM-derived hydrogel (Fig. 2A). Although much lower in expression intensity, the α-Gal antigen was still detected in all decellularized hydrogel groups (Fig. 2B), suggesting that decellularization alone is insufficient to completely remove the α-Gal antigen.

Fig. 2.

α-Gal epitope assessment. M86 antibody-stained sections from the (A) native nondecellularized hydrogel group and (B) MD, Triton X-100, and SDC hydrogel groups before and (C) after treatment with the enzyme α-Galactosidase. (Scale bar, 100 µm.)

Previous works have demonstrated several protocols for the removal of the α-Gal antigen from xenotissues, which primarily focus on treating the ECM postdecellularization with α-galactosidase. This enzyme specifically cleaves the α-Gal antigen from tissues without compromising their regenerative potential (47, 50). In order to further mitigate the immunogenicity of the resulting hydrogels, all decellularized tissues from every group underwent enzymatic treatment with α-galactosidase. This straightforward enzymatic approach efficiently cleaved the α-Gal antigen from the resulting hydrogels in all groups, as evidenced by the negative expression of α-Gal epitope posttreatment (Fig. 2C). Collectively, these findings highlight the necessity of an α-galactosidase treatment following decellularization for the effective removal of the α-Gal antigen from xenotissues. Moreover, this same treatment could be extended to animal tissues of diverse origins to further reduce their immunogenicity postdecellularization, thereby enhancing their potential for clinical acceptance.

Biochemical Characterizations.

An inherent advantage of utilizing dECM-derived hydrogels lies in the presence of various ECM components, often not fully replicated in conventional hydrogels (e.g., natural or synthetic hydrogels). Therefore, maintaining the integrity of these ECM components postdecellularization is crucial to create an optimal microenvironment with appropriate biochemical cues, fostering enhanced cell–matrix interactions. This study quantitatively and qualitatively assessed the biochemical composition of the resulting hydrogels to elucidate the impact of different decellularization protocols on their ultimate biochemical blueprint. Emphasis was placed on investigating the preservation of growth factors, total protein content, and other ECM molecules integral to muscle tissue differentiation, development, and structural organization, such as collagens, sulfated glycosaminoglycan (sGAG), laminin, and fibronectin (5).

The Quantabody growth factor array facilitated the quantification of growth factors in the constructed smECM hydrogels. In the native nondecellularized hydrogels, more than 38 growth factors, with varying functions such as myogenesis, neurogenesis, angiogenesis, embryogenesis, morphogenesis, tissue development, growth, repair, maturation, and homeostasis (51), were identified (Fig. 3). These growth factors were also detected in hydrogels prepared using the MD method, in quantities comparable to those observed in the native nondecellularized control. In contrast, growth factors were either completely denatured or significantly reduced in hydrogels prepared using the Triton X-100 and SDC protocols. These outcomes align with previous findings, indicating that MD via freeze–thaw cycling does not substantially increase the loss of growth factors from the tissue, in contrast to detergent-based decellularization methods (52).

Fig. 3.

Heat map illustrating the growth factors content in the different smECM hydrogels as assessed via the Quantibody^® Growth Factor Array Kit. Each growth factor is highlighted with a rectangular box colored with its corresponding function.

Total protein was assessed using the bicinchoninic acid (BCA) assay, sGAG was quantified through the dimethylmethylene blue assay, and collagen content was determined via hydroxyproline quantification. The total protein and sGAG contents were maintained at levels similar to the native nondecellularized control when the MD protocol was employed for decellularization (Fig. 4 A and B). In contrast, the Triton X-100 and SDC protocols significantly diminished the total protein and sGAG contents in the final hydrogels compared to the native and MD groups. However, the collagen content remained consistent across all treatment groups, showing no significant differences, although a lower trend was observed in the Triton X-100 and SDC groups (Fig. 4C). These findings align with prior research where the use of ionic and nonionic detergents for decellularization in tissues from other sources significantly reduced sGAG content postdecellularization, while collagen content remained relatively unchanged (53, 54).

Fig. 4.

Quantitative biochemical evaluations of the different smECM hydrogels. (A) total protein, (B) sGAG, and (C) collagen contents.

Histological examinations utilizing alcian blue and picrosirius red staining validated the presence of sGAG and collagen in all hydrogels. However, a noticeable decrease in staining intensity was observed in the Triton X-100 and SDC-treated groups, consistent with the quantification assessments (Fig. 5 A and B). Subsequently, picrosirius red–polarized imaging was conducted on the same picrosirius red–stained sections to further elucidate the impact of different decellularization protocols on the nativity of collagen in the formed hydrogels. This imaging technique highlights the natural birefringence of collagen fibers when exposed to polarized light, providing precise identification of collagen phenotypes in tissue sections, where type I collagen fibers look yellow/red, while type III collagen fibers look green (55). In the native nondecellularized hydrogels, collagen type I predominated over type III (SI Appendix, Fig. S4), mimicking their ratio in native skeletal muscle tissue (56). The MD hydrogels contained approximately similar amounts of both collagen types. Conversely, hydrogels prepared using the Triton X-100 and SDC methods exhibited more collagen type III with a minor presence of collagen type I. This aligns with previous reports indicating that both Triton X-100 and SDC can alter the nativity of collagen when used for decellularization of tendons (57) and urinary bladders (58).

Fig. 5.

Qualitative biochemical evaluations of the different smECM hydrogels via histological staining. (A) Alcian blue and (B) picrosirius red–stained sections were captured at low (10×) and high (20×) magnifications. (Scale bars, 100 µm.)

Further evaluations of the biochemical composition of the smECM hydrogels were conducted using IHC against two crucial smECM proteins: laminin and fibronectin. IHC staining confirmed the presence of both laminin and fibronectin in all hydrogels (Fig. 6 A and B). However, a noticeable reduction in the expression of these proteins was observed only in the Triton X-100 and SDC-treated groups. Collectively, these results indicate that hydrogels prepared using the MD protocol retain structural ECM components abundant in native skeletal muscle, including collagens, sGAG, laminin, and fibronectin, as well as nonstructural ECM proteins (e.g., growth factors) that can support muscle differentiation and development (5). Consequently, it is concluded that the MD decellularization protocol could yield smECM hydrogels with a well-preserved biochemical blueprint compared to the Triton X-100 and SDC protocols.

Fig. 6.

Qualitative biochemical evaluations of the different smECM hydrogels via immunohistochemistry staining. (A) laminin and (B) fibronectin-stained sections captured at low (10×) and high (20×) magnifications. (Scale bars, 100 µm.)

Gelation Kinetics.

The rapid gelation of hydrogels is often crucial for both in vitro and in vivo regenerative engineering applications (40, 59). This is particularly significant for cell-populated hydrogels, as rapid gelation prevents gravity-induced cell sedimentation, ensuring a uniform cell distribution within the hydrogel matrix and minimizing cell loss at the application site (40, 59). To assess the impact of different decellularization protocols on overall gelation speed, the gelation kinetics of the various smECM hydrogels were spectrophotometrically analyzed using turbidimetric assessment. This method quantifies the increase in turbidity within the hydrogels during incubation at 37 °C (21, 29, 53) (Fig. 7 A and B). All hydrogels exhibited a lag in gelling characteristics, with gelation occurring after a lag period ($t_{\mathit{lag}}$) (Fig. 7C). This delay in fibrillogenesis (lag phase) has been reported in other ECM-derived hydrogels from diverse sources (21, 29, 53). However, compared to the native nondecellularized control, only Triton X-100 and SDC produced hydrogels that gelled after a significantly longer lag phase ($t_{\mathit{lag}}$ 10.5 ± 0.5, 14.25 ± 0.82, and 16 ± 0.70 min, respectively). In contrast, MD hydrogels had a lag phase relatively similar to that observed in the native control without any significant differences ($t_{\mathit{lag}}$ 10.6 ± 0.74 min). Further turbidimetric analysis revealed that Triton X-100 and SDC-treated hydrogels had significantly longer $t_{1/2}$ (50% gelation) and $t_{95}$ (95% gelation) values than the native control and MD hydrogel groups (Fig. 7 D and E). The MD hydrogel group exhibited a higher gelation velocity compared to Triton X-100 and SDC hydrogel groups (Fig. 7F). However, no significant differences were found between the groups. Compared to the native hydrogel group, significantly lower gelation velocities were observed in the MD (*P < 0.05) and Triton X-100 and SDC hydrogel groups (*P < 0.001). Together, these data imply that only the MD decellularization protocol could produce hydrogels with improved overall gelation kinetics.

Fig. 7.

Gelation kinetics of the different smECM hydrogels assessed by measuring the increase in their turbidity upon incubation at 37 °C. (A) Raw and (B) normalized turbidimetric values. Normalized values were used to calculate (C) $t_{\mathit{lag}}$ (time elapsed before the hydrogel starts gelling), (D) $t_{1/2}$ (time elapsed before the hydrogel reaches 50% gelation), (E) $t_{95}$ (time elapsed before the hydrogel reaches 95% gelation), and (F) S (gelation velocity).

The superior gelation kinetics observed in hydrogels produced by the MD protocol may be attributed to the preservation of smECM components known to facilitate the self-assembly process of collagen molecules, as previously reported (40). In theory, dECM hydrogels are formed through a self-assembly process of collagen molecules, partially controlled by the preservation of glycol proteoglycans, proteoglycans, and other essential ECM components (60). Glycol proteoglycans act as bridging molecules for the assembly of collagen fibrils, and proteoglycans regulate the collagen self-assembly process (40). Additionally, glycoproteins like fibronectin facilitate the organization of collagen fibrils by acting as a bridging molecule during the collagen self-assembly process (61, 62). Although no significant differences were observed in collagen content quantification across the groups (Fig. 4C), Triton X-100 and SDC treatments significantly reduced other components crucial for regulating the collagen self-assembly process. Thus, the preservation of collagen and other components, such as glycol proteoglycans, proteoglycans, and fibronectin in MD hydrogels, likely facilitated their faster gelation. Moreover, the presence of remnants of detergents may have contributed to the decreased gelation time in Triton X-100 and SDC hydrogels. Fernández-Pérez et al. utilized a detergent-based decellularization protocol to create cornea ECM-derived hydrogels (63). The hydrogel could be formed at a lower detergent concentration, but the hydrogel formation was impeded by an increased detergent concentration. They also confirmed the presence of detergent remnants in the hydrogel, aligning with these findings.

Ultrastructure.

The different hydrogels (Fig. 8A) were examined under scanning electron microscopy (SEM) to assess their overall ultrastructure and the influence of different decellularization protocols on physical properties such as fibril density. SEM images confirmed that all hydrogels exhibited high porosity, a favorable attribute facilitating efficient transport of oxygen and nutrients throughout the matrix (Fig. 8B). Furthermore, all hydrogels demonstrated a similar ultrastructure characterized by an array of sheet-like structures with some spacing in-between. These spaces were observed to be filled with tiny fibrils, and the density of these fibrils in the intersheet spaces decreased in the Triton X-100 and SDC groups compared to the native control and MD hydrogels. Quantitative analysis validated that hydrogel fibril density was significantly higher in native hydrogels than in hydrogels produced through Triton X-100 and SDC treatments (Fig. 8C). In contrast, no significant differences were found between MD and native hydrogels, indicating well-preserved physical properties in hydrogels produced through MD.

Fig. 8.

Gross morphology and ultrastructure of the different smECM hydrogels. (A) The gross morphology and (B) ultrastructure of the different smECM hydrogels under SEM captured at low (1,000×) and high (3,000×) magnifications. (C) SEM images were used to quantify the fibril density in the different smECM hydrogels.

In this study, consistent gelation parameters, including temperature, ionic strength, pH, and ECM concentration, were employed to induce gelation in all hydrogels. Gelation was initiated at 37 °C, a temperature previously established as essential, as ECM-derived pregels failed to gel at 4 °C or 22 °C, only transitioning to a gel state upon incubation at 37 °C (64). An ionic strength of 0.5× phosphate buffered saline (PBS) was used when diluting the stock to a final concentration of 4 mg/mL. While many studies on ECM-derived hydrogels use an ionic strength of 1× PBS (27, 65, 66), other studies have clearly shown the striking positive effects of reducing the ionic strength to 0.5× PBS on the hydrogel fiber density and gelation speed (63, 67). Finally, a pH of ~7.4 was used as a neutral pH significantly accelerates the collagen self-assembly process compared to acidic or alkaline pH values (67). Consequently, since all hydrogels were fabricated following the same parameters, we attribute the significant reduction in the hydrogel fibril density observed in the Triton X-100 and SDC groups to the considerable depletion of the self-assembling ECM molecules in these hydrogels, which may have attenuated the degree of the collagen self-assembly, leading to an overall decrease in fibril density.

Migration.

Recruitment of resident muscle progenitor cells, such as satellite stem cells (SSCs), to the injury site can significantly accelerate the repair process (68). Previous reports have demonstrated that decellularized skeletal muscle smECM scaffolds can serve as a conducive environment for SSC recruitment. This is attributed to the presence of preserved growth factors known for guiding cell recruitment, including basic fibroblast growth factor, bone morphogenetic proteins-7, epidermal growth factor (EGF), EGF receptor (EGFR), hepatocyte growth factor, insulin-like growth factor binding protein-4, insulin-like growth factor-1, platelet-derived growth factor-AA, transforming growth factor beta-1, and transforming growth factor beta-3 (69–71). In this study, these growth factors were identified in the smECM hydrogels (Fig. 3), suggesting their preservation. Therefore, we sought to assess whether the smECM hydrogels could act as chemoattractants for the migration of SSCs and if the decellularization protocol influenced these chemoattractive properties. To do so, a transwell-based cell migration assay was conducted, with primary SSCs on the top porous membrane and different hydrogels in the lower compartment (Fig. 9A). Besides evaluating the different smECM hydrogels, 10% FBS, and 1 mg/mL pepsin in 30 mg/ mL acetic acid solution in H₂O, were added as positive and negative controls, respectively. In addition, the commercially available collagen type I hydrogel was added as a third control for comparison. Regardless of the decellularization protocol, SSCs exhibited significantly higher migration toward all smECM hydrogels compared to controls, including commercially available collagen (Fig. 9B). However, SSC migration was significantly higher toward native and MD hydrogels than Triton X-100 and SDC hydrogels. No significant differences were observed in the magnitude of SSCs migration toward the native and MD hydrogels. The diminished SSCs migration toward the Triton X-100 and SDC hydrogels compared to the native and MD may be ascribed to the poor preservation of growth factors in these hydrogels due to the harsh decellularization protocols applied, thus, rendering them less effective chemoattractants. These data highlight the chemoattractive property of the smECM hydrogels and demonstrate the influence that the decellularization protocol can have on this feature.

Fig. 9.

Migration of SSCs toward the different chemoattractants. (A) A schematic illustrating the transwell-based system used to assess the migration of SSCs toward the different chemoattractants. (B) The degree of SSCs migration toward the different chemoattractants.

Cytocompatibility.

Next, we assessed the impact of the decellularization protocol on the cytocompatibility of the finally formed hydrogels. Primary SSCs were embedded within the different hydrogels, and their viability and growth over time were evaluated. Collagen type I hydrogel served as a control in all biological evaluations for comparison, given its recognized ideal cytocompatibility and widespread use as a microenvironmental niche for SSC growth (72, 73). Live/dead viability confocal images revealed high cell viability within the collagen, native, and MD hydrogel groups at both 3 and 7 d of culture (Fig. 10A). In contrast, the Triton X-100 and the SDC rendered completely cytotoxic hydrogels, with cells appearing completely round and dead within these hydrogels at both culture timepoints.

Fig. 10.

Cytocompatibility of the different smECM hydrogels. (A) Confocal images of SSCs embedded within the different smECM hydrogels and labeled with the green stain for live cells (Calcein AM) and the red stain for dead cells (Ethidium Homodimer-1) after days 3 and 7 of culture. (B) Growth of SSCs within the different smECM hydrogels at days 3 and 7 of culture as measured by MTS (^@P < 0.0001 and ^%P < 0.001compared to day 3). (Scale bar, 100 µm.)

Next, we utilized the MTS assay to evaluate the growth of SSCs in the different hydrogels. The MTS assay demonstrated significantly higher growth rates of SSCs in the native and MD hydrogel groups at both culture timepoints, surpassing all other hydrogels, including the commercially available collagen type I (Fig. 10B). No statistically significant differences were noted in growth rates between the native control and MD groups. These findings indicate that, among the decellularization protocols employed, only the MD method yielded highly cytocompatible hydrogels. Additionally, the data suggest that cells exhibited greater metabolic activity and significantly higher growth rates when surrounded by their original ECM environment compared to the pure collagen type I. This suggests that the smECM hydrogel provided cells with a more conducive microenvironment, rich in various tissue-specific components, thereby enhancing the observed cell growth.

During the culture period, a significant cell-mediated contraction in both the native and MD hydrogels was observed (SI Appendix, Fig. S5A). This cell-mediated hydrogel contraction phenomenon has been reported in various studies using dermal, amnion, and cornea ECM-derived hydrogels (28, 40, 63). Conversely, in cases where cells died, such as with Triton X-100 and SDC treatments, the hydrogels retained their initial shape and size, showing no contraction. Similarly, the commercially available collagen type I hydrogel did not undergo any contraction over the culture period. Quantitative analysis of the change in hydrogel area revealed that this cell-mediated contraction started as early as day one of culture, reaching approximately 85% contraction by day 7 in both the native and MD hydrogel groups (SI Appendix, Fig. S5B). These findings align with the Live/Dead and MTS data, further confirming that cells within these two hydrogel groups were highly viable and metabolically active. Their ability to effectively attach, proliferate, and migrate within their 3D space caused them to contract by pulling in the matrix’s fibrils. Although cells within the collagen type I hydrogel were highly viable, they did not exhibit the same level of metabolic activity as in the native and MD hydrogels. This difference could be attributed to the absence of essential instructive tissue-specific ECM components, preventing them from contracting the matrix. While rapid hydrogel contraction could be advantageous for inducing guided cell alignment in engineered muscles (74–76), it may not be desirable for certain applications, particularly when the hydrogel is intended for use as an injectable system. It is important to note that the hydrogels in this study were maintained in a free-floating culture system. It is expected that their ability to contract would be less pronounced when they adhere to the tissue matrix in vivo.

Myogenic Inductivity.

A primary objective in developing dECM-derived hydrogels is to create a tissue-specific microenvironment for enhanced cell functions. Therefore, it was crucial in this study to identify a suitable decellularization protocol that would not compromise the myogenic-inductive cues present in the skeletal muscle matrix. This consideration was vital to achieve the desired benefits of using such a tissue-specific material. To assess the impact of the decellularization protocol on myogenic-inductivity, we encapsulated SSCs within different smECM hydrogels, including commercially available collagen as a control for comparison—a material extensively investigated in skeletal muscle regenerative engineering (72, 73).

Immunofluorescence-stained hydrogels for myosin heavy chain (MHC) revealed that SSCs differentiated and rapidly fused into multinucleated myofibers specifically in the native and MD hydrogel groups at day 7 postinduction, in contrast to the other groups, including the commercially available collagen (Fig. 11). While SSCs in the collagen control differentiated into myocytes with positive MHC expression, no evidence of myofiber formation was observed. Triton X-100 hydrogels contained a small population of MHC-positive cells, whereas MHC expression was entirely negative in the SDC hydrogels, possibly due to their cytotoxicity. This observation highlights the superior myogenic potency of smECM hydrogel over commercially available collagen and sheds light on the influence of the decellularization method on the myogenic induction properties of the smECM hydrogels.

Fig. 11.

Myoinductivity of the different smECM hydrogels. Immunofluorescence staining images for MHC at day 7 postinduction showing the myogenic differentiation of SSCs in the different smECM hydrogels. (Scale bar, 50 µm.)

Collectively, these data suggest that smECM hydrogels produced using the MD protocol exhibited the most favorable properties overall, indicating the suitability of this protocol for deriving smECM hydrogels with enhanced characteristics for skeletal muscle regenerative engineering. However, since the smECM hydrogel is derived from xenogeneic tissue, its biological safety must be evaluated in vivo using an appropriate animal model to examine immunogenicity and its ability to integrate with surrounding tissues. Due to the efficacy of the MD decellularization protocol in producing smECM hydrogels with improved overall properties, this protocol was selected for preparing smECM hydrogels for the in vivo immunogenicity assessment.

In Vivo Immunogenicity Assessments.

Materials sourced from allogeneic origins address certain concerns related to immunogenicity but are limited to older donors, exhibit considerable batch-to-batch variability, and lack ready availability (77). In contrast, xenogeneic materials derived from porcine tissues, for instance, are abundantly accessible and can be obtained from younger tissue sources, which is advantageous for regenerative medicine therapies (78). However, xenogeneic materials may pose potential immunogenic challenges and encounter regulatory hurdles (77). Given these considerations, it was imperative to thoroughly investigate the in vivo biological safety of our porcine skeletal muscle-derived hydrogels. This examination aimed to gauge the effectiveness of our decellularization protocol in mitigating their overall immunogenicity.

The Balb/c mouse was selected as our xenotransplantation model due to its immunocompetence and fully functional immune system. This model has been widely utilized for evaluating the biocompatibility and immunogenicity of various xenogeneic tissues (77). In addition to assessing the immunogenicity of the MD hydrogels, the native nondecellularized and collagen type I hydrogels were included as proinflammatory and proremodeling controls, respectively. Each hydrogel was injected subcutaneously—a common delivery route for ISO standard biocompatibility tests (SI Appendix, Fig. S6 A and B). The injected hydrogels were collected for histological, IHC, and flow cytometry assessments on days 3, 7, and 14 postinjection (SI Appendix, Fig. S6 C and D). These specific time points were chosen to evaluate the early (innate), mid, and late-phase immune response, respectively, given that previous studies have indicated that the macrophage response and polarization during this timeframe are indicative of long-term biocompatibility and tissue repair (79, 80). Longer time points were not investigated because ECM-derived hydrogels are known to degrade completely by 3 to 4 wk in vivo (28).

The harvested biomaterials underwent hematoxylin & eosin-y (H&E) staining to evaluate cellular infiltration and density. Generally, greater cellular infiltration was observed in the native nondecellularized hydrogels compared to the collagen type I control and MD hydrogels at all time points (Fig. 12A). Moreover, foreign body giant immune cells (FBGIC), a collection of fused macrophages generated in response to the presence of unfavorable materials (81), were clearly present in the native nondecellularized hydrogels at all time points. In contrast, fewer infiltrated cells and no FBGIC were observed in the collagen type I and MD hydrogels throughout the experimental duration. For a quantitative assessment of infiltrated cell number, the harvested hydrogels were digested, and the number of released cells was counted by flow cytometery. The number of infiltrated cells was significantly higher in the native nondecellularized hydrogels compared to the collagen type I control and MD hydrogels at all time points, consistent with the H&E images (Fig. 12B). The MD hydrogels exhibited a significantly higher number of infiltrated cells at day 3 compared to the collagen type I control. However, no significant differences were observed between these two groups at later time points. The number of infiltrated cells significantly increased on day 7 compared to day 3 in all groups, reaching its peak value before declining by day 14. This mid-phase increase of infiltrated cell numbers is consistent with previous reports (82) and is part of the wound healing immune response.

Fig. 12.

Cellular infiltration assessment. (A) H&E images showing the local tissue immune response at days 3, 7, and 14 postsubcutaneous injections of the different hydrogels in the xenotransplantation animal model. (B) Quantification of the total number of infiltrated cells by flow cytometry. The dotted black lines at the middle of each image separate the subcutaneous tissue (referred to as “ST”) from the injected hydrogels (referred to as “H”). The dotted blue squares inserted on the H&E images of the native group highlight the presence of FBGICs. The solid blue squares are the higher magnification of the dotted blue squares. The scale bar for the H&E images is 100 µm, whereas the scale bar for the magnified H&E images is 50 µm.

The pronounced cell infiltration observed in the native nondecellularized control group implies an unfavorable immune response to the biomaterial, as increased cell infiltration is recognized as an initial sign of acute graft rejection (83, 84). Moreover, the evident presence of FBGIC further indicates graft rejection of the native nondecellularized hydrogels. In contrast, the significantly lower number of infiltrated cells in the collagen type I control and MD hydrogels compared to the native nondecellularized control at all time points, along with the absence of FBGIC, suggests the acceptance of these biomaterials, consistent with previous reports (77).

While assessments like H&E staining and quantifying infiltrated cell numbers offer initial insights into biomaterial’s biocompatibility and immunogenicity, they lack detailed information about the host immune response. A more precise measure for evaluating these aspects involves identifying the phenotype of infiltrated cells and assessing their predominance. Macrophages and T cells are pivotal immune cells that commonly infiltrate biomaterials postadministration in vivo, irrespective of the biomaterial’s biocompatibility (84, 85). Evaluating the predominance of infiltrating macrophages and T cell phenotypes can indicate biomaterial rejection or acceptance postadministration, as previously reported (84, 85). For instance, macrophages can be polarized into M1 (proinflammatory) and M2 (anti-inflammatory) subtypes (84, 85). Predominance of M1 macrophages, associated with a proinflammatory immune response, is often linked to graft rejection (84, 85). Conversely, M2 macrophages stimulate an anti-inflammatory immune response resembling the immune response during wound healing and tissue remodeling (84, 85). Hence, their predominance often indicates graft acceptance (84, 85). Similarly, T cells are subcategorized into T-helper (T-Help) and T-cytotoxic (T-Cyto) subtypes (84, 85). T-Help T cells elicit a proremodeling immune response, whereas T-Cyto T cells elicit a proinflammatory immune response (84, 85). Therefore, it can be speculated that the predominance of T-Help T cells in the biomaterial is favorable.

In this study, comprehensive IHC and flow cytometry analyses were conducted to distinguish the phenotype and predominance of infiltrated immune cells within the different hydrogels. CD86 (M1 macrophage marker) and CD163 (M2 macrophage marker) were utilized for the IHC and flow cytometry analysis, and the percentages of CD86+ and CD163+ cells measured by the flow cytometry assessment were employed to calculate the M2/M1 ratio, estimating the magnitude of macrophage predominance. IHC analysis revealed the presence of polarized macrophages toward both the M1 and M2 subtypes in all groups at all timepoints, but in distinct densities (Fig. 13A).

Fig. 13.

Polarized macrophage infiltration. (A) Representative IHC images for CD86 (M1 macrophage marker) and CD163 (M2 macrophage marker) showing macrophage infiltration and the phenotype of the infiltrated macrophages in the different hydrogels at days 3, 7, and 14 postsubcutaneous injections. DAPI was used to label the cellular nuclei (blue). Flow cytometry was used to quantify the percentages of the infiltrated M1 and M2 macrophages, and the acquired percentages were used to calculate the ratio of M2/M1. M2/M1 ratio calculations using (B) CD163 and CD86 and (C) Arginase (M2 macrophage marker) and iNOS (M1 macrophage marker). (Scale bar, 50 µm.)

At day 3, the ratio of M2/M1 macrophages for all three hydrogels was below 1, suggesting an early-phase M1-dominant macrophage response (Fig. 13B). By day 7, the native nondecellularized hydrogels remained M1-dominant, while the collagen type I and MD hydrogels transitioned to an M2-dominant response. By day 14, the native nondecellularized hydrogels continued to be M1-dominant, while the M2 predominance further increased in the collagen type I and MD hydrogels. This demonstrated a shift from a proinflammatory to a proremodeling macrophage polarization response for the collagen and MD hydrogels, a characteristic of wound healing (86), suggesting graft acceptance. However, compared to the collagen type I control, the MD hydrogels exhibited a more favorable immune response, as evidenced by significantly higher values in the M2/M1 ratio analysis on days 7 and 14. This observation is consistent with previous studies demonstrating that ECM-derived materials can stimulate a robust proremodeling response (87–90), which is beneficial for the host integration of tissue-engineered materials. To further confirm the subtype of polarized macrophages, flow cytometry analysis against additional M1(iNOS) and M2 (Arginase) markers was also conducted. The ratio of Arginase/iNOS was calculated using the percentages of Arginase+ and iNOS+ cells measured by the flow cytometry assessment. Identical ratio trends were observed using these additional markers, further emphasizing the above findings (Fig. 13C).

Next, a similar assessment was performed to calculate the ratio of T-Help/T-Cyto in the different hydrogels using CD4 (T-help marker) and CD8 (T-Cyto marker). IHC analysis showed that T-Help and T-Cyto T cells were present in the different hydrogels but in distinct densities and expression magnitudes (Fig. 14A). Specifically, T-Help T cells were found to be present at all timepoints in the different hydrogels. In contrast, T-Cyto T cells were only present at days 3 and 7 in the collagen type I and MD hydrogels, whereas their presence was detected at all timepoints in the native nondecellularized hydrogels. Comparing the ratio of T-help to T-Cyto, we found that the native nondecellularized hydrogels were consistently T-Cyto T cell dominant at all timepoints (Fig. 14B). In contrast, the collagen type I and MD hydrogels were always T-Help T cell dominant, and the T-help T cell predominancy was further increasing over time. This transition in immune response toward the collagen type I and MD hydrogels from a lesser to a higher T-Help T cell dominant environment has been previously reported (91), suggesting that only the native nondecellularized hydrogels were being rejected, as expected. However, compared to the collagen type I control, the MD hydrogels showed significantly higher values in the T-Help/T-Cyto ratio analysis at all timepoints, suggesting a more favorable immune response toward the MD hydrogels.

Fig. 14.

T-help and T-Cyto T cell infiltration. (A) Representative IHC images for CD4 (T-help T cell marker) and CD8 (T-Cyto T cell marker) showing T cell infiltration and the phenotype of the infiltrated T cells in the different hydrogels at days 3, 7, and 14 postsubcutaneous injections. DAPI was used to label the cellular nuclei (blue). (B) Ratio calculations of T-help/T-Cyto using the percentages of CD4+ (T-help) and CD8+ (T-Cyto) T cells as measured by flow cytometry. (C) Ratio calculations of Th2/Th1 using the percentages of IL-4Rα+ (Th2) and IFN-γ+ (Th1) T cells as measured by flow cytometry. (Scale bar, 50 µm.)

It should be noted that the immune response is a highly complex process. In this sense, T-Help T cell (CD4+) predominance does not always indicate a favorable immune response. For instance, T-help T cells can be polarized into Th1 and Th2 subtypes, which are both CD4+ (SI Appendix, Fig. S7) (85). Th1 T cells produce a set of cytokines, including IFN-γ, IL-12, IL-18, and IL-27, which stimulate macrophage polarization toward the M1 proinflammatory subtype (85). In contrast, Th2 T cells stimulate macrophage polarization toward the M2 proremolding subtype by producing IL-4Rα, IL17RB, CCR3, and CCR4 (85). Therefore, the polarization of macrophages toward the M1 and M2 subtypes is mainly determined by the subtype into which T-help T cells polarize. To distinguish the subtype of the T-help T cells in the different hydrogels, as well as the predominancy of each subtype, the ratio of IL-4Rα/IFN-γ was calculated using the percentages of IL-4Rα+ and IFN-γ+ cells measured by the flow cytometry assessment. We found that the Th2 T cells were the most predominant T-help T cell subtype in the collagen type I and MD hydrogels at all timepoints. In contrast, the native nondecellularized hydrogels were mostly Th1 dominant at all timepoints (Fig. 14C). These data were consistent with the findings in our macrophage ratio analysis (Fig. 13 B and C).

Collectively, these findings indicate that xenogeneic MD hydrogels triggered a robust proremodeling immune response in the xenotransplantation animal model, suggesting their biocompatibility and the efficacy of the MD decellularization protocol in reducing their overall immunogenicity. However, while the utilized animal model allowed for extensive exploration of the hydrogel’s immunogenicity, it was unable to assess the role of the α-Gal antigen, crucial in xenotissue rejection in humans (44–46). Addressing concerns about this antigen could involve postdecellularization cleavage and evaluation, as performed in this study, or additional tests in nonhuman primate models, knockout mice, or pigs lacking this antigen (92–94). Although these models are not readily available, they offer predictive insights into the immune response to such an antigen. Thus, future studies should prioritize evaluating the immunogenicity of xenogeneic tissues in these models to expedite their clinical integration.

Conclusion

In conclusion, this study demonstrated the significant impact of the decellularization protocol on the properties of ECM-derived hydrogels, utilizing a tissue model based on porcine skeletal muscles. Among all tested protocols, the MD decellularization protocol emerged as highly effective, substantially reducing total DNA content while preserving the ECM’s biochemical composition. This resulted in smECM hydrogels that are not only highly cytocompatible but also myogenic inductive, displaying enhanced overall properties suitable for skeletal muscle regenerative engineering. The study also demonstrated that decellularization alone inadequately removes xenoantigens, like the α-Gal antigen, from animal ECM-derived materials. It suggests that an additional postdecellularization treatment regimen may be necessary for efficient cleavage, thereby reducing overall immunogenicity. Comprehensive in vivo assessments demonstrate the biocompatibility and low immunogenicity of smECM hydrogels decellularized by the MD protocol. This is evident through the polarization of macrophages and T cells toward the M2 and T-help phenotype, respectively, effectively inhibiting rejection in a xenotransplantation model.

Materials and Methods

Detailed information is provided in SI Appendix, Materials and Methods.

smECM Preparation and smECM Hydrogel Characterization.

The decellularization process of porcine hindlimb skeletal muscles was conducted to obtain cell-free smECM hydrogels for potential use in regenerative engineering. Skeletal muscle tissues were harvested, purified, and subjected to three different decellularization protocols: freeze–thaw, Triton X-100, and SDC. The resultant muscle fragments were treated with DNase-I, RNase, and α-galactosidase to eliminate DNA/RNA and cleave α-Gal antigens. The muscle fragments underwent freeze-drying and cryomilling to obtain smECM in powder form, which was stored for later use.

To form smECM hydrogels, smECM powders were digested and neutralized, and gelation was induced by incubation at 37 °C. Bovine corium collagen type I hydrogel was prepared as a control. Quantitative biochemical characterizations were performed to assess DNA, total protein, sGAG, and collagen content. Growth factors content was analyzed using a Human Growth Factor Array Kit. Qualitative biochemical characterizations were conducted through histological and IHC assessments to assess DNA, sGAG, collagen, α-Gal antigens, laminin, and fibronectin contents within the finally formed smECM hydrogels.

Gelation kinetics were determined by turbidimetric spectrophotometric analysis, measuring absorbance at 405 nm. SEM was employed to visualize hydrogel fibril density.

Primary SSCs were isolated from the calf muscles of New Zealand white rabbits. The migration of SSCs toward hydrogels was assessed using a cell migration assay kit. Cytocompatibility assessments were conducted through LIVE/DEAD cytotoxicity assay and MTS assay at days 3 and 7 of culture. Myogenic-inductivity assessments involved inducing differentiation of SSCs within hydrogels were evaluated by myofiber formation at day 7 of differentiation through immunofluorescence staining.

In Vivo Study Design.

For in vivo studies, subcutaneous injections of smECM hydrogels were performed in Balb/c mice, with native nondecellularized smECM hydrogels and collagen I hydrogels as controls. Samples were harvested at days 3, 7, and 14 postinjection and underwent histology, IHC, and flow cytometry analyses to assess the immune response to the injected materials.

Statistical Analysis.

All quantitative data were expressed as mean ± SD. All statistical analyses were performed using the statistical software Prism GraphPad version 8 (GraphPad, USA). Statistical analyses were performed using the one-way ANOVA and two-way ANOVA with Tukey’s post hoc test. Statistical significance was evaluated at *P <0.05, **P <0.01, ***P < 0.001, and ****P < 0.0001.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included either in the main text and/or in SI Appendix.