Abstract Enhanced tendon‒bone healing is of critically importance for achieving optimal postoperative recovery following a rotator cuff tendon tear (rotator cuff tears, RCTs). Although RCTs patch-augmented scaffolds demonstrate clinical potential, there is a paucity of reports on biodegradable scaffolds that effectively integrate high strength and bioactivity. Inspired by the composition and aligned nanofibrous structure of the natural fish bladder matrix (fish swim bladder, FSB), we employed a gallium (Ga)‒tannic acid (TA) metal‒polyphenol network (MPN)-modified decellularized fish bladder matrix (GaPP@FSB) as a novel biomaterial to address this problem. Ga-TA MPN represents a “two birds with one stone” modification strategy that allows GaPP@FSB to demonstrate commendable mechanical strength alongside multiple biological activities, including antibacterial, antioxidant, anti-inflammatory and osteogenic differentiation promotion. Furthermore, GaPP@FSB regulates the focal adhesion-based mechanical signal transduction pathway in tendon stem/progenitor cells (TSPCs), thereby activating the α5β1/Akt/PI3K pathway to induce tenogenic differentiation. Additionally, this scaffold exhibits remarkable anti-inflammatory and antibacterial activities. In a rat RCTs model, GaPP@FSB promoted regeneration at the tendon‒bone interface while restoring both rotator cuff biomechanics and joint movement function. Consequently, this biomaterial derived from natural FSB has outstanding biosafety and biological activity, making it a highly promising candidate for clinical applications in both tendon repair and the restoration of the tendon‒bone interface. Graphical abstract [38]graphic file with name 12951_2025_3580_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03580-y. Keywords: Fish swim bladder, Rotator cuff tears, Tendon–bone healing, Metal‒polyphenol network, ECM-responsive signalling Introduction Rotator cuff tendon tears (RCTs) is a prevalent sports injury that can result in severe pain and shoulder dysfunction [[39]1]. The prognosis of RCTs following surgical treatment is often unsatisfactory, primarily due to the challenge of poor healing at the tendon‒bone interface post-surgery [[40]2]. Given the avascular and acellular nature of this gradual interface, the self-healing capacity of the reconstructed interface after surgery is limited, leading to suboptimal scar tissue formation and compromised biomechanical properties. This inadequate tissue integration significantly contributes to re-tearing, high failure rates, and other clinical complications associated with RCTs [[41]3, [42]4]. Consequently, achieving functional tendon–bone healing remains a critical hurdle in RCTs treatment. To address these challenges, research on enthesis regeneration has explored innovative techniques such as the biomimetic design of biological scaffolds that mimic the structure and composition of tendons and entheses [[43]5]. These biomimetic bioactive scaffolds are aimed to replicate the extracellular matrix (ECM) found in tendons or enthesis tissues, thereby facilitating specific signalling interactions with cell surface receptors to regulate cellular behaviour and differentiation. Examples include 3D bioprinted scaffolds resembling enthesis structures and composite scaffolds composed of decellularized ECM and biopolymers [[44]6–[45]9]. However, the mechanical properties of the current bioactive scaffolds remain unsatisfactory. Tension at the tendon‒bone interface is an undeniable factor, particularly in numerous pathologies associated with RCTs and frequently accompanied by substantial tendon defects. Consequently, RCTs patch augmentation-scaffolds exhibiting exceptional mechanical properties are more conducive to practical clinical operations [[46]10, [47]11]. The natural fish swim bladder (FSB) is a promising candidate high-strength bioactive scaffold for tendon/tendon–bone interface repair (Fig. [48]1). The matrix of FSB is abundant in collagen, elastin and glycosaminoglycan exhibiting high similarity to the extracellular matrix of the natural tendon-bone interface [[49]12]. This characteristic may endow FSB with the potential to modulate cell behavior. Additionally, this material can be readily sourced from kitchen waste, highlighting its sustainability and accessibility. Currently, few studies on FSB-based biomaterials exist, with most focusing on tissue-engineered blood vessels [[50]13]. There are almost no reports exist on FSB-based tendon repair scaffolds. Further research is needed to develop FSB as a biomaterial for tendon repair and tendon-bone healing, especially through the development of a novel strategy that simultaneously enhances its mechanical strength and biological activity. In the past decade, metal‒polyphenol networks (MPNs) have garnered significant attention in various fields because of their facile preparation conditions, straightforward methodologies, and cost effectiveness. In the realm of biomedicine, MPNs have found extensive applications in modifying biological scaffolds and fabricating nanocoatings and nanoenzymes in a weakly alkaline environment [[51]14–[52]18]. Polyphenols, such as tannic acid (TA), can chelate with multiple galloyl groups under neutral or weakly alkaline liquid environments to form MPNs; in turn, these MPNs exhibit omnidirectional growth, leading to the formation of nanoscale particles or surface coatings [[53]18, [54]19]. The modification of MPNs can be considered a programmable strategy, as diverse types of metal ions, including Cu^2+, Zn^2+, Mg^2+, Fe^3+, Ti^4+, and Zr^4+, have been reported to bond with polyphenols to generate MPNs [[55]14–[56]16]. In the ROS environment, the catechol groups in polyphenols are oxidized into quinone structures, leading to the disruption of the MPN structure and triggering the responsive release of metal ions [[57]18]. Additionally, TA can cross-link collagen through hydrogen bonding [[58]20, [59]21]. Specifically, under alkaline conditions, the catechol structure in TA is susceptible to oxidation into a quinone state, which can subsequently undergo Michael addition or Schiff base reactions with amine groups, thereby providing an opportunity for covalent bonding with amine or thiol groups and the formation of stable connections. Owing to the exceptional biocompatibility of TA, its use to replace conventional chemical cross-linking agents such as glutaraldehyde can mitigate concerns regarding biomaterial toxicity. This concept has inspired the notion that FSBs could be crosslinked with polyphenols to confer the mechanical properties necessary for tendon repair. Moreover, polyphenol-bound FSBs provide binding sites for metal ions, resulting in the formation of MPNs from which released metal ions can stimulate stem cell differentiation and expedite healing processes. Under acidic and reactive oxygen species (ROS) liquid environments, complexation between polyphenols and metal ions within MPNs gradually transitions from triple complexes to monocomplexes, followed by the gradual release of metal ions [[60]14]. This property suggests that MPN-modified FSBs could serve as ion‒polyphenol controlled-release platforms that exhibit dual pH/ROS response capabilities and could thereby regulate physiological microenvironments and enhance surrounding tissue regeneration. Fig. 1. [61]Fig. 1 [62]Open in a new tab Schematic summarizing the preparation of Ga-TA MPN-modified fish swim bladder (GaPP@FSB) for enhancing tendon–bone healing. This patch-like scaffold is designed for the repair of RCTs, with one end anchored to the tendon stump and the other end secured to the humeral head, thereby facilitating tendon-bone healing. Specifically, TA can crosslink and toughen FSB while chelating gallium ions, forming Ga-TA MPNs on the FSB. Ga-TA MPNs are capable of responding to acidic and ROS microenvironments by releasing Ga ions, which in turn promotes the regeneration of the transitional tissue at the tendon-bone interface. GaPP@FSB also emulates the ECM, thereby activating the α5β1/Akt/PI3K pathway to induce tenogenic differentiation. Additionally, GaPP@FSB exhibits remarkable anti-inflammatory, antiapoptotic and antibacterial activities Here, we propose a gallium-tannic acid (Ga-TA) MPN network-modified FSB scaffold (GaPP@FSB), which effectively combines mechanical strength with bioactivity (Fig. [63]1). Specifically, GaPP@FSB is designed to withstand tension at the tendon‒bone interface while promoting fibrocartilage regeneration. The fish swim bladder’s aligned nanofiber structure mimics the extracellular matrix at the tendon‒bone interface [[64]13], showcasing natural nanobiomimicry. Gallium has been extensively utilized in various biomedical applications, such as bone regeneration, sterilization, and haemostasis [[65]22]. Given its ability to enhance bone regeneration, we hypothesized that gallium ions could promote fibrocartilage layer regeneration. However, further confirmation is needed due to limited existing reports. In this study, we investigated the microstructure and mechanical properties of GaPP@FSB, evaluated its biocompatibility, antimicrobial and anti-inflammatory activities, and explored its effects on tendon‒bone healing using a rat RCTs model. Biomechanical and gait analyses confirmed GaPP@FSB’s role in restoring biomechanical function. Additionally, considering potential interactions between decellularized FSB matrix components and cell surface receptors, we hypothesized that FSB scaffolds could modulate cell behaviour and differentiation. Bioinformatics approaches were used to predict these interactions and analyze their influence on tendon stem/progenitor cells (TSPCs) differentiation. Results Characterization of gapp@fsb After an initial decellularization process (Fig. [66]2A), we compared the compositional disparities between the natural FSB (N-FSB) and the decellularized FSB matrix. As depicted in Fig. [67]S1, the processed FSB exhibited minimal residual DNA while retaining a significant portion of collagen and elastin from the native swim bladder; however, there was a relatively substantial loss of GAG. Following the completion of FSB preparation, we further modified the matrix with TA and Ga-TA MPNs to obtain PP@FSB and GaPP@FSB (Fig. [68]2A). 1GaPP@FSB, 2GaPP@FSB and 3GaPP@FSB, corresponding to different concentrations of gallium ions in the MPN construction system (0.1 mg/ml, 0.2 mg/ml, and 0.5 mg/ml), respectively, while maintaining a constant concentration of tannic acid. As observed through SEM and TEM images (Fig. [69]2B), all groups presented aligned protein nanofibers similar to the extracellular matrix of natural tendons. Their cross-sectional SEM images (Fig. [70]S2) revealed that the FSBs consisted of multiple layers of aligned fibrous membranes arranged longitudinally and latitudinally. Additionally, increasing the concentration of gallium ions in the MPN construction system led to an increase in the protein nanofiber width (Fig. [71]2B-C). Fig. 2. [72]Fig. 2 [73]Open in a new tab Characterization of GaPP@FSB. (A) Preparation process diagram of GaPP@FSB; (B) microstructure and morphology of GaPP@FSB, including scanning electron microscopy (SEM) images, transmission electron microscopy (TEM) images, and size distribution of nanofibers; (C) average diameter of protein nanofibers in GaPP@FSB scaffolds; (D) degradation rate of GaPP@FSB within 30 days; (E) FT-IR data of GaPP@FSB scaffolds; (F) XPS data of GaPP@FSB scaffolds; (G) release curves of gallium ions released by 2GaPP@FSB within 30 days in different microenvironments; (H) schematic diagram of the response of Ga-TA to ROS and hydrogen ions Furthermore, the assessed swelling rates demonstrated a 20.1% increase in the FSB weight from its dry to wet state (Fig. [74]S3). Notably, as the soaking time increased to 24 h, the swelling rate of FSB remained consistently stable, indicating remarkable scaffold stability. Conversely, crosslinking with MPNs yielded a decreased and sustained swelling rate of the scaffold. We subsequently evaluated the degradation behaviour of the GaPP@FSB scaffolds using type I collagenase solution for 30 days (Figs. [75]2D & S4). FSBs modified with Ga-TA MPNs and TA exhibited slow degradation with less than a 10% mass loss, whereas unmodified FSBs degraded at a faster rate, with a quality loss of 19.2%. The 3GaPP@FSB group exhibited a mass loss of up to 13.0%, indicating that high concentrations of gallium ions hindered the crosslinking effect between tannic acid and FSBs due to binding interactions between gallium ions and ortho-diphenol groups. On the basis of metal‒polyphenol interactions in a weakly alkaline environment, we hypothesized that tannins, when bound to collagen fibres in FSB, further combine with Ga ions to form a Ga(III)-phenolic network (Ga-TA MPNs). The Fourier transform infrared spectroscopy (FTIR) data revealed characteristic peaks of amide I, amide II, and amide III bonds from collagen/elastin at 1655 cm^− 1, 1560 cm^− 1, and 1240 cm^− 1, respectively, in the FSB scaffold (Fig. [76]2E). A characteristic peak at 1217 cm^− 1 corresponding to the tannic acid ortho-diphenol group was observed in the PP@FSB group, and this finding indicated successful binding of tannic acid within the FSBs. EDS analyses confirmed the presence of a significant amount of elemental Ga in GaPP@FSB, whereas the FSB and PP@FSB groups did not contain Ga (Fig. [77]S5). Mapping analysis also demonstrated that Ga was uniformly distributed between the fibres of FSB, implying the formation of Ga-TA MPNs (Fig. [78]S6). Furthermore, XPS analysis was performed to investigate the surface elemental composition of each group of FSB scaffolds. As shown in Fig. [79]2F, since proteins such as collagen are the main components of FBS, the signals of oxygen (O), carbon (C) and nitrogen (N) were present as major elements. After Ga-TA MPN modification, the characteristic peaks of Ga (Ga2p, Ga3d) appeared in the XPS spectra of the 1GaPP@FSB, 2GaPP@FSB and 3GaPP@FSB groups. The characteristic peaks of Ga3p are difficult to observe, possibly because of the relatively low gallium content on the material surface. Overall, these data demonstrated that Ga-TA MPNs successfully combined with FSBs. We subsequently employed the 2GaPP@FSB group to investigate the ion release behaviour in different microenvironments. Figure [80]2G shows that 2GaPP@FSB released Ga ions at a low rate in a liquid environment at pH = 7.4, and the ion release period lasted more than 30 days. During the first 7 days of the ion release period, the Ga ions were released at a higher rate. After the 7th day, the ion release stabilized at a relatively moderate rate. In contrast, in an acidic liquid environment (i.e., at pH = 4.5) and an ROS-containing liquid environment (i.e., containing 200 µM H[2]O[2]), 2GaPP@FSB showed a pronounced burst release of Ga ions during the first 7 days, after which it stabilized. A similar trend was shown in the ROS-containing liquid environment, where GaPP@FSB also exhibited a significant burst release of ions for the first 7 days, followed by a stabilization of Ga ion release. These results showed that the rate of Ga ion release from GaPP@FSB is positively correlated with the concentration of hydrogen ions and free radicals in the microenvironment, as MPNs are more sensitive to hydrogen ions and ROS. The mechanism is as follows: when Ga-TA MPNs are formed in the swim bladder, a free Ga ion will form a covalent bond via ligand interactions with three galloyl groups from tannins in a neutral or alkaline liquid environment. As shown in Fig. [81]2F, this interaction can be disrupted by ROS and hydrogen ions in the microenvironment. Based on this, we used the 2GaPP@FSB scaffold as an example and analyzed its FT-IR spectrum after immersion in the aforementioned ROS-containing medium for 3 days. The results revealed that treatment with ROS, induced the appearance of a new absorption peak at 1708 cm^− 1, which was attributed to the carbonyl group (C = O) of quinone (Fig. [82]S7). Therefore, Ga-TA MPNs respond to scavenge excess ROS and release Ga ions substantially in acidic or ROS microenvironments, providing antimicrobial effects and enhancing bone regeneration. The DPPH method was subsequently used to evaluate the antioxidant activities of the FSBs. As shown in Fig. [83]4a, the PP@FSB group exhibited the best antioxidant activity. GaPP@FSB also exhibited favourable total antioxidant activity, but this activity was significantly lower than that of the PP@FSB group. This may have been due to the different states of the gallic acid/galloyl groups in the MPNs of the scaffolds. The total antioxidant activity in the FSB group was poor and markedly lower than that in the other two groups. A DPPH scavenging assay was also conducted to evaluate the ROS scavenging ability of swim bladder scaffolds from different groups. As depicted in Fig. [84]S8, the purple colour of the DPPH solution in the PP@FSB and all GaPP@FSB groups faded almost completely, whereas the FSB group exhibited minimal fading of the purple colour associated with DPPH radicals. A total antioxidant capacity test kit (FRAP method) was subsequently used to assess the overall antioxidant rate of each scaffold group. The results demonstrated that PP@FSB, 1Ga@FSB, 2Ga@FSB, and 3GaPP@FSB exhibited exceptional antioxidant properties, whereas FSB displayed negligible antioxidative capability (Fig. [85]S8B). These findings demonstrate that modification with MPNs significantly enhances the antioxidant activity of FSB scaffolds, thereby improving their efficacy in mitigating oxidative stress and inflammation in shoulder joints. Fig. 4. [86]Fig. 4 [87]Open in a new tab The impact of GaPP@FSB on the cell adhesion, migration, and proliferation of Multiple types of cells derived from the tendon bone interface (BMSCs, Chondrocytes and TSPCs). The fluorescence images of live‒dead staining (A) revealed the presence of BMSCs, Chondrocytes and TSPCs within the GaPP@FSB scaffold, demonstrating its directional adhesion capability (Bar = 200 μm). Moreover, the CCK-8 data (B) indicated the ability of this scaffold to support long-term cell proliferation. The role of scaffold components in cell migration was investigated: (C) Schematic illustration of stimulating cell migration using a solution of enzyme-degraded scaffold; (D) Quantitative data from the 24-hour cell scratch assay; (E) Observation of cell migration under an optical microscope (Bar = 200 μm). Data are presented as mean values ± SD (n = 4), *p < 0.05, **p < 0.01, ***p < 0.001 Mechanical properties of the gapp@fsb scaffolds We postulated that the incorporation of Ga-TA MPNs could increase the mechanical strength of FSBs by leveraging the strong affinity of polyphenols for collagen-based materials. Tensile‒strain testing was conducted to evaluate the mechanical properties of FSBs modified with MPNs, aiming to elucidate the role of Ga-TA MPNs in toughening the FSBs. Figure [88]3A shows a schematic diagram depicting the use of the GaPP@FSB scaffold as a rotator cuff patch for repairing RCTs, thereby establishing a bridging effect at the tendon‒bone interface. Considering the substantial dimensions of the swim bladder, it can be tailored to accommodate diverse types of RCTs. As shown in Fig. [89]3B-D, unmodified decellularized scaffolds derived from FSBs (FSB group) presented significantly inferior mechanical properties (measured at 3.1 ± 0.8 MPa for maximum failure force, 15.2 ± 0.9 MPa for elastic modulus, and 20.0 ± 5.1% for destructive strain). These findings suggest that solely utilizing a decellularized scaffold fabricated from FSBs would hardly meet the clinical requirements for RCT repairs because of its subpar mechanical properties and potential risk of surgical failure. Interestingly, after TA modification, both the PP@FSB and all GaPP@FSB groups exhibited significantly improved mechanical properties (maximum failure force: PP@FSB: 26.6 ± 2.3 MPa; 1GaPP@FSB: 26.0 ± 0.9 MPa; 2GaPP@FSB: 26.4 ± 0.6 MPa; 3GaPP@FSB: 22.4 ± 0.7 MPa). Elastic modulus: PP@FSB: 87.7 ± 19.0 MPa; 1GaPP@FSB: 107.6 ± 9.7 MPa; 2GaPP@FSB: 102.0 ± 11.0 MPa; 3GaPP@FSB: 102.2 ± 5.0 MPa) As depicted in Fig. [90]3D, GaPP@FSB can support a weight of 500 g, regardless of whether it is shaped into fragmented or linear forms, indicating the potential of GaPP@FSB as an ideal candidate for tendon repair in clinical settings. By possessing mechanical properties similar to those of natural tendons, GaPP@FSB mitigates the risk of surgical failure due to scaffold rupture or stress shielding caused by excessive mechanical strength in the repaired tendon. Furthermore, this result highlights the significant role played by the covalent bonds between Ga ions and polyphenols in enhancing the material’s mechanical properties. Notably, GaPP@FSB exhibited superior mechanical properties to those of glutaraldehyde-crosslinked FSB vascular scaffolds (7.9 ± 0.8 MPa), suggesting that MPN-based modification could serve as an alternative to conventional glutaraldehyde cross-linking and has potential applications in tissue engineering. Additionally, the tensile strength of the GaPP@FSB scaffold surpassed that of the majority of the scaffolds used for tendon‒bone interface repair mentioned in the Ashby diagram (Fig. [91]3E) [[92]23–[93]30]. Overall, the exceptional mechanical performance of the GaPP@FSB scaffold renders it highly suitable for rotator cuff repair. Fig. 3. [94]Fig. 3 [95]Open in a new tab Mechanical properties of the GaPP@FSB scaffolds. (A) FSB and GaPP@FSB possess ample dimensions to meet a wide range of clinical requirements while functioning akin to a patch that serves as an intermediary and enhancer at the tendon‒bone interface. Tensile stress‒strain curves (B) of FSB, PP@FSB and each GaPP@FSB scaffold. Comparison of (C) the maximum tensile strength, elastic modulus, and strain at failure of FSB, PP@FSB and each GaPP@FSB scaffold. (D) The scaffold can withstand a weight of 500 g, and (E) Ashby diagram of the maximum tensile strength of reported scaffolds for tendon-to-bone repair [[96]23–[97]30]. These findings indicate that the mechanical strength of the GaPP@FSB scaffolds is completely suitable for RCT repair. Data are presented as mean values ± SD (n = 4), *p < 0.05, **p < 0.01, ***p < 0.001 Influence of gapp@fsb on cell viability and differentiation Although few studies have demonstrated the advantages of FSB application for vascular tissue engineering, the biocompatibility of FSB-derived biomaterials, especially MPN-modified FSB decellularized scaffolds, remains unknown. In light of this knowledge gap, we evaluated the viability, adhesion and differentiation of BMSCs and TSPCs seeded onto GaPP@FSB scaffolds. Furthermore, we investigated the potential risk of GaPP@FSB-induced inflammation after implantation by examining its effect on macrophage polarization. As depicted in Fig. [98]4A&B, the results from the CCK-8 assays revealed that cell proliferation increased over time in all the groups tested; notably, 2GaPP@FSB increased the proliferation capacity of both BMSCs and TSPCs. With respect to the cellular morphology of each scaffold group after 24 h of inoculation, excellent dispersion and adhesion were observed across all the fish bladder scaffold groups, with the cells predominantly exhibiting an extended morphology accompanied by numerous pseudopods. These findings suggest that GaPP@FSB scaffolds effectively support cell adhesion and proliferation, particularly for cell types originating from the tendon‒bone interface, indicating exceptional biocompatibility properties. Interestingly, the cells arranged on the FSB scaffold displayed a distinct directional morphology, which was potentially attributable to the microstructure of the FSBs—an indirect indication of remarkable resemblance between the FSBs and the natural tendon matrix. Overall, 2GaPP@FSB had the most pronounced effect on promoting BMSC and TSPC proliferation. Therefore, for subsequent experiments, 2GaPP@FSB was selected for the experimental group of scaffolds, which was referred to as the GaPP@FSB group. This study further probed into the role of scaffolds in the migration of various cells. As depicted in Fig. [99]4C, after the scaffolds were digested and dissolved by papain and added to the culture medium, the effect of scaffold components on cell migration was explored through the cell scratch assay. The results, as shown in Fig. [100]4D&E, indicated that for different types of cells, whether they were TSPCs, BMSCs or chondrocytes, all groups of scaffolds could significantly promote cell migration. GaPP@FSB demonstrated the most significant function in promoting cell migration. Evidently, the FSB scaffold exhibited broad-spectrum biocompatibility. Currently, the effect of gallium ions on tendon–bone healing is unclear, although gallium ions have been shown to regulate the osteogenic differentiation of osteoblasts and inhibit the activation of osteoclasts. Therefore, in this study, the effects of each group of FSB scaffolds on the differentiation of BMSCs and TSPCs, as well as the role of FSB scaffolds on maintaining chondrocyte phenotype, were first verified via Q‒PCR. As shown in Fig. [101]5A&B, the FSB decellularized scaffolds without MPN modification (FSB group) presented favourable osteogenesis-promoting bioactivities. Compared with the blank control, FSB alone promoted the expression of genes related to the osteogenic differentiation of BMSCs (Runx2, Opn, Ocn and Col1). There was no significant difference between the PP@FSB and FSB groups. In contrast, the BMSCs cultured on GaPP@FSB exhibited the highest level of expression of osteogenic differentiation-related genes. As shown in Fig. [102]5B, BMSCs cultured on GaPP@FSB also showed stronger osteopontin (Opn) fluorescence, indicating increased intracellular Opn protein levels. Additionally, the results of ARS and ALP staining revealed that the GaPP@FSB-treated BMSCs presented significantly increased levels of mineralization deposition (Fig. [103]S9). The results of Alcian blue (AB) staining also showed the matrix secretion level of GaPP@FSB-treated chondrocyte, which is consistent with the results of Q-PCR. These results are also consistent with the previous findings that Ga^3+ ions can regulate the osteogenic differentiation of osteoblasts [[104]31, [105]32], suggesting that gallium-modified FSB scaffolds promote satisfactory levels of bone regeneration and mineralization. Fig. 5. [106]Fig. 5 [107]Open in a new tab The impact of GaPP@FSB on the differentiation of BMSCs, Chondrocytes and TSPCs. Confocal laser scanning microscopy (CLSM) images showing fluorescence-labelled (A) OPN in the BMSCs (scale bar = 200 μm), and the Q-PCR data (B) revealed the expression of osteogenic differentiation-related genes in BMSCs cultured on the GaPP@FSB scaffold. Additionally, the CLSM images also showed the fluorescence-labelled (C) Tnmd in TSPCs (bar 100 μm), and Q-PCR data (D) revealed the expression of tenogenic differentiation-related genes in TSPCs cultured on the GaPP@FSB scaffold. The investigation approach for the influence of scaffolds on chondrocytes is uniform. (E) CLSM reveals the level of ACAN, a chondrocyte matrix-related proteoglycan, and (F) the Q-PCR results demonstrate the expression of genes related to chondrogenic differentiation. Data are presented as mean values ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 Interestingly, FSBs had a positive effect on the tendonogenic differentiation of TSPCs (Fig. [108]5C&D), with TSPCs cultured on FSB, PP@FSB and GaPP@FSB showing significantly higher gene expression levels of Scx, Mkx, Tnmd, and Col 1 than those in the blank group. In addition, these three groups also presented relatively high levels of Tnmd protein fluorescence (Fig. [109]5C). These findings suggest that FSBs themselves possess the inherent ability to enhance tendinous differentiation in TSPCs, as evidenced by the comparable outcomes observed among the FSB, PP@FSB and GaPP@FSB groups. These findings suggest that tannic acid and gallium ions within GaPP@FSB may not significantly influence TSPC differentiation. This observation indicates that FSBs exert a favourable influence on tendonogenic differentiation, which prompted us to delve deeper into the underlying regulatory mechanisms and assess the potential usefulness of FSBs in the field of tendon tissue engineering. FSBs also exhibit a positive effect on maintaining the gene expression related to the chondrocyte chondrogenic phenotype. The results presented in Fig. [110]S5E&F reveal that GaPP@FSB and PP@FSB show the highest levels of chondrogenic-related gene expression and fluorescent labeled cartilage matrix proteoglycan Acan. In general, this FSB scaffold demonstrates diverse cellular activities, which contribute to stimulating the directed differentiation of different types of cells (that is, osteogenic differentiation of BMSCs, tenogenic differentiation of TSPCs, and phenotype maintenance of chondrocytes) and achieving the regeneration of this enthesis transition tissue. Mechanism of the effects of gapp@fsb on tenogenic differentiation To elucidate the impact of GaPP@FSB on cell differentiation, we conducted an RNA sequencing analysis on TSPCs treated with FSB, PP@FSB, or GaPP@FSB. As shown in Fig. [111]6, Gene Ontology (GO) enrichment analysis revealed that the significantly expressed genes were associated with cell adhesion, cellular spatial morphology, and the ECM response. These findings suggest that FSB scaffolds can mimic ECM properties and regulate cellular behaviour through interactions between the scaffold and ECM-cell receptors. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed significant expression of genes related to ECM-responsive signalling, the Akt-PI3K signalling pathway, and the HIF-1α signalling pathway. Previous studies have shown that collagen and elastin in the ECM regulate Akt-PI3K expression via an integrin signalling mechanism [[112]33, [113]34], which subsequently influences cellular behaviour, including tendon differentiation and regeneration. Particularly, KEGG analysis demonstrated the enrichment of Focal Adherence and cell adhesion molecules, indicating that FSBs are capable of transmitting mechanical stimulation signals to cells through the adhesion - integrin signaling system. Within the focal adhesion, integrins can bind to components of the extracellular matrix such as collagen and fibronectin, attaching cells to the matrix. Integrins are interconnected with the actin cytoskeleton, and this interaction is stabilized by the binding of adhesion proteins [[114]35, [115]36]. Considering that the major components of FSBs are collagen and fibronectin, this outcome implies that FSBs serve as a bridge between the scaffold matrix stress and intracellular forces through the adhesion protein - integrin, thereby regulating cell differentiation. Fig. 6. [116]Fig. 6 [117]Open in a new tab Transcriptomic analyses. (A) Volcano plot of GaPP@FSB. Red dots indicate upregulated genes, and blue dots indicate downregulated genes. (B) KEGG enrichment pathway analysis. The blue box indicates enrichment of the ECM receptor and activation of the Akt-PI3K signalling pathway. (C&D) Western blot analysis of proteins related to integrin α5β1 and the Akt-PI3K signalling pathway, including (C) protein band and related quantitative analysis. (E) GO analysis showing the cellular components, biological processes and molecular functions of GaPP@FSB. The blue box indicates that GaPP@FSB can affect multiple physiological processes related to the ECM. (E) Schematic diagram of GaPP@FSB-induced activation of the Akt‒PI3K signalling pathway in TSPCs. Upon treatment with GRGDNP, the activation of the Akt-PI3K pathway and the expression of tendon differentiation-related genes in TSPCs were examined by (G) Western blot and (H) Q-PCR analysis. This indicates that GRGDNP blocks the signal transmission between FSBs and cells, resulting in the downregulation of tendon differentiation-related gene expression. Data are presented as mean values ± SD (n = 4), *p < 0.05, **p < 0.01, ***p < 0.001 Therefore, we hypothesized that FSB scaffolds induce TSPC differentiation by mimicking the natural ECM and regulating adult tendon differentiation through the ECM-ECM receptor-Akt-PI3K signalling cascade. To validate this hypothesis, Western blot analysis was performed, which revealed significantly increased phosphorylation levels of Akt and PI3K in TSPCs treated with FSB, PP@FSB or GaPP@FSB compared with those in the control group, thereby indicating the activation of relevant signalling pathways. Notably, there was no difference among these three groups, suggesting that the natural tendon-like composition and microstructure of FSB played crucial roles, whereas MPNs had no significant effect on Akt-PI3K activation. Moreover, the FSB, PP@FSB, and GaPP@FSB groups exhibited high expression of integrins α5 and β1, which aligns with previous studies indicating that the signalling modality mediated by integrin α5β1 is a crucial pathway for Akt-PI3K activation through the ECM. As shown in Fig. [118]6G&H, we subsequently introduced an integrin α5β1 inhibitor (RGD peptide (GRGDNP)) into the scaffold cultured with cells system of TSPCs and GaPP@FSB. The Q-PCR results demonstrated that the GRGDNP inhibitor counteracted the promotional effect of GaPP@FSB. This means that GRGDNP blocks the signal transmission between FSBs and cells, resulting in downregulation of tendon differentiation related gene expression. Consequently, FSB scaffolds may activate the Focal adhesion-based mechanical signal transduction pathway, that is, to form cell surface receptor-integrin α5β1 complex due to their natural ECM-like structure and composition. This activation subsequently triggers the Akt-PI3K signalling pathway to regulate tendonogenic differentiation [[119]37]. Furthermore, the utilization of GaPP@FSB in RCT repair can potentially enhance tendon healing, improve tendon quality, and mitigate the risk of recurrent RCT tears. Immunomodulatory effects and biosafety of gapp@fsb Furthermore, in this study, we investigated the immunogenicity and potential anti-inflammatory effects of the GaPP@FSB scaffold using Raw264.7 macrophages as an in vitro model, laying the foundation for further investigation. As shown in Fig. [120]S10, after 3 days of culture, the Q-PCR results revealed that the FSB group exhibited favourable anti-inflammatory bioactivity under the induction of LPS. Compared with those in the blank group, genes associated with M1 macrophage polarization and proinflammatory factors (iL-1β, iNOs, and CCR7) were significantly downregulated; moreover, genes related to M2 polarization (Arg-1, CD206, iL-10, iL-1ra and IL-6) were significantly upregulated. These findings indicate that FSB itself possesses immunomodulatory functionality. After MPN modification, the GaPP@FSB exhibited greater anti-inflammatory bioactivity than FSB. The genes associated with M1 polarization (TNF-a, iL-1β, iNOs, and CCR7) were further downregulated in this group. Conversely, genes related to M2 polarization (Arg-1, CD206, iL-10, iL-1ra and IL-6) were further upregulated in the GaPP@FSB group. This superior anti-inflammatory bioactivity can be attributed to the presence of polyphenols in the MPNs. Immunofluorescence staining corroborated these findings, as macrophages seeded onto GaPP@FSB presented stronger Arg-1 and weaker iNOS fluorescence signals than the other groups did (Fig. [121]S10A-B). These results also suggest that MPNs primarily exert early anti-inflammatory effects during immunomodulation processes. Overall, our results confirm that as a tendon implant material, GaPP@FSB does not induce localized chronic inflammation but instead has favourable anti-inflammatory properties that facilitate the removal of negative factors detrimental to rotator cuff healing. To further investigate the biosafety of FSBs, we conducted haemolysis experiments as a validation method. As depicted in Fig. [122]S11, after mixing with blood, the FSB PP@FSB and GaPP@FSB stents allowed easy collection of blood cells through centrifugation while leaving the supernatant transparent in colour, with minimal haemolysis observed. These results provide additional evidence supporting the clinical potential of the GaPP@FSB stents proposed in this study. Antibacterial, antioxidant stress, and antiapoptotic property testing As depicted in Fig. [123]7A-B, cultured with GaPP@FSB scaffolds led to a significant reduction in the bacterial colony counts for both E. coli and S. aureus compared with those in the other scaffold groups. Surprisingly, the ‘FSB’ group exhibited limited antibacterial activity; instead, it significantly promoted S. aureus proliferation. As shown in Fig. [124]7B, statistical analysis revealed relative bacterial survival rates of 123.28% ± 8.60%, 0.11% ± 0.03%, and 0.01% ± 0.02% against S. aureus in the FSB, PP@FSB, and GaPP@FSB groups, respectively. Similarly, for E. coli, these three groups presented relative bacterial survival rates of 19.63% ± 7.12%, 2.47% ± 0.85%, and 0.42% ± 0.05%, respectively. Overall, GaPP@FSB demonstrated superior antibacterial efficacy, as evidenced by minimal observation of S. aureus colonies on the culture plates, indicating the nearly complete eradication of bacteria; similar results were also observed for E. coli. SEM images were used to observe micromorphological changes in the bacteria. As shown in Fig. [125]7C, GaPP@FSB caused substantial damage to the morphological integrity of the bacteria compared with that in both the control group and the PP@FSB group. This finding is consistent with previous studies suggesting that gallium (Ga) can induce direct physical or chemical damage to bacterial membranes, leading to increased permeability. Moreover, owing to the chemical similarity between gallium ions and iron (Fe^3+) ions, Ga ions compete with Fe^3+ upon entering the bacterial cytosol, binding essential proteins and enzymes required for maintaining physiological homeostasis [[126]38]. This result further confirms that the incorporation of gallium ions into materials can enhance the antibacterial activity of FSB scaffolds. Fig. 7. [127]Fig. 7 [128]Open in a new tab Antibacterial, antioxidant stress, and antiapoptotic properties. (A) Colony formation on bacterial culture plates was assessed following 24 h of coculturing of Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) with various scaffolds. (B) The antibacterial efficacy of each scaffold group was evaluated on the basis of colony quantification. (C) SEM images of the E. coli morphology after cultured with various scaffolds (Bar = 2 μm). The impact of GaPP@FSB on TSPC apoptosis under hydrogen peroxide stimulation was assessed via (D) CLSM and (E-F) flow cytometry to quantify the population of apoptotic cells. The impact of GaPP@FSB on the ROS level in TSPCs stimulated with hydrogen peroxide was assessed via (G-H) flow cytometry and (I) CLSM. (Bar = 200 μm). Data are presented as mean values ± SD (n = 4), *p < 0.05, **p < 0.01, ***p < 0.001 As previously mentioned, MPNs are responsive to ROS; thus, we investigated the effects of GaPP@FSB in clearing oxidative stress and inhibiting cell apoptosis. In a Transwell co-culture system involving TSPCs treated with hydrogen peroxide (200 mM), as depicted in Fig. [129]7D-F, the GaPP@FSB group presented the lowest early/late cell apoptosis ratio, with no significant difference observed between the PP@FSB group and the GaPP@FSB group. Furthermore, as shown in Fig. [130]7H-I, both the PP@FSB group and the GaPP@FSB group presented significantly reduced ROS levels in TPSCs under H[2]O[2] stimulation, indicating that tannic acid plays a crucial role in this process. Oxidative stress is an important factor that induces tendon cell apoptosis after the occurrence of an RCT and contributes to the limited self-healing ability of RCTs. The exceptional antioxidant properties of GaPP@FSB can help mitigate this risk factor that hinders RCT healing. In vivo experiments Restoration of the natural enthesis structure at the tendon‒bone interface, specifically the fibrocartilage layer with a four-layer structure, is crucial for restoring the biomechanical properties of the rotator cuff after surgery. In clinical practice, fibrous scar tissue often fills the surgically reconstructed tendon‒bone interface of the rotator cuff, resulting in suboptimal integration strength. Therefore, biomaterials used for rotator cuff repair must possess regenerative capabilities to promote fibrocartilage layer regeneration at this interface. Consequently, this study utilized a rat RCT model to investigate the therapeutic efficacy of GaPP@FSB scaffolds for repairing rotator cuff injuries (Fig. [131]8A). As depicted in Fig. [132]8B, FSB, PP@FSB and GaPP@FSB nearly completely integrated with the host tissue eight weeks after surgery, and no significant inflammatory infiltration was observed. These findings are consistent with previous data indicating that FSB-based scaffolds do not induce local chronic inflammation and serve as biocompatible implants. The safranin O/fast green staining results demonstrated that, compared with the other groups, the GaPP@FSB group presented significantly greater levels of fibrocartilage tissue (indicated by red staining). This finding was further confirmed through quantitative analysis, which revealed a significantly larger area of fibrocartilage in the GaPP@FSB group than in the other three groups. The results from HE and toluidine blue staining also supported this finding, particularly highlighting the enhanced maturation of enthesis tissue in the GaPP@FSB group, with more mature fibrochondrocytes at the regenerated tendon‒bone interface. In contrast, the blank group displayed a smaller area of regenerated fibrocartilage layer along with some disordered fibrous tissue. Moreover, both the FSB and PP@FSB groups displayed positive therapeutic effects, indicating that the FSB itself contributes to improved tendon–bone healing; however, these two groups were not as effective as the GaPP@FSB group. The histological scores of the GaPP@FSB group were also greater than those of the other three groups (Fig. [133]S13). Overall, these findings collectively demonstrate that GaPP@FSB significantly promotes tendon‒bone healing and facilitates regeneration of the fibrocartilage layer, thereby facilitating the restoration of soft tissue and biomechanical balance. Furthermore, this study highlights promising outcomes associated with the use of gallium ions for enhancing tendon–bone healing and provides inspiration for the development of novel biomaterials for repairing tendon–bone interfaces. Furthermore, biomechanical tests were conducted to evaluate the biomechanical properties of the rotator cuff of the repaired rats at 8 weeks post-surgery (Fig. [134]8D-E). In terms of failure mode analysis, the GaPP@FSB group exhibited a significantly greater maximum failure tension than the other three groups did (GaPP@FSB vs. PP@FSB vs. FSB vs. blank = 8.3 ± 0.2 MPa, 7.5 ± 0.5 MPa, 4.2 ± 0.3 MPa; 2.6 ± 0.2 MPa). Additionally, the GaPP@FSB group demonstrated significantly greater stiffness (15.2 ± 1.4 N/mm) than both the FSB and blank groups did; however, there was no statistically significant difference in the stiffness value between the GaPP@FSB group and the PP@FSB group. Importantly, the rotator cuff repaired with GaPP@FSB achieved a maximum failure tension equivalent to that of native rotator cuff tendon tissue (8.3 ± 0.2 MPa vs. 7.5 ± 0.1 MPa), demonstrating comparable strength to that of the normal rotator cuff tendon. Fig. 8. [135]Fig. 8 [136]Open in a new tab Histopathological analysis and biomechanical analysis of the therapeutic effect of FSB scaffolds on RCTs in a rat model. (A) Schematic summarizing the in vivo experiments. (B) Images of safranin-O-fast green staining, H&E staining and toluidine blue staining showing the healing of the tendon‒bone interface and the regeneration of fibrocartilage in each group at 4 and 8 weeks after surgery. Biomarkers related to inflammation and tissue regeneration at the tendon‒bone interface were also detected via fluorescence labelling (blue: DAPI; red: collagen II (Col-II); green: TNF‒α; yellow: Arg-1) (bar = 200 μm). (C) Quantitative analysis of the relative area of fibrocartilage tissue at the regenerated tendon‒bone interface, which was executed on the basis of images of safranin‒fast green staining and toluidine blue staining. The relative fluorescence intensity of Col-II at the interface was analysed via ImageJ (n = 5). (D-E) Biomechanical analysis of the repaired rotator cuff tendons in each group 8 weeks after surgery. (D) Digital images of biomechanical testing. The detection items are (E) maximum failure tension, ultimate failure load and stiffness. Data are presented as mean values ± SD (n = 5), *p < 0.05, **p < 0.01, ***p < 0.001 As shown in Fig. [137]9, the footprint analysis data consistently aligned with the biomechanics data. At 8 weeks post-RCT repair, rats in the blank and FSB groups exhibited hind limb dragging and crawling only. In contrast, rats implanted with PP@FSB demonstrated relatively coordinated movements of their anterior and posterior limbs. Furthermore, GaPP@FSB-implanted rats displayed the best gait trajectory, exhibiting not only synchronized front and rear limb movements but also consistent trajectories closest to those of normal rat paw prints (Fig. [138]9A). The statistical analysis further revealed significant differences between the GaPP@FSB group and the other groups, indicating superior coordination in gait trajectory (Fig. [139]9B-D). These findings provide additional evidence that implanting GaPP@FSB into rotator-cuff tendon defects restores not only the mechanical strength but also the motor function of the shoulder joint. These findings confirm that GaPP@FSB has positive effects on the treatment of rotator cuff injuries and suggest that our proposed fish bladder-based scaffold holds great promise as a material for repairing tendon/tendon–bone interfaces. Fig. 9. [140]Fig. 9 [141]Open in a new tab Footprint analysis of the rat models at 8 weeks post-surgery. (A) Morphology, length, and force intensity of the rat footprints. (B-C) Quantitative analysis of upper limb paw prints focuses on three key parameters: (B) the average area of paw prints, (C) the length of paw prints, and (D) the mean force intensity exerted by the rat’s paws on the sensor during locomotion. Data are presented as mean values ± SD (n = 5), *p < 0.05, **p < 0.01, ***p < 0.001 To further evaluate the clinical potential of GaPP@FSB, we employed a rat RCTs model to compare its therapeutic effects with those of commercialized PET (polyethylene terephthalate) patches. As shown in Fig. [142]S14, photographs depict the repair of rat RCTs using these two scaffolds. Additionally, an animal model without patch use but only sutured with the clinically common eight- suture technique was designated as the Blank group. After 4 and 8 weeks of repair, pathological histological staining and quantitative analysis (Fig. [143]S15&[144]S16) revealed that the tendon-bone interface in the GaPP@FSB group exhibited a significantly higher level of regenerated fibrocartilage compared to the PET and Blank groups. Notably, the PET group displayed more fibrous scar tissue formation, consistent with prior studies. Furthermore, the GaPP@FSB group demonstrated enhanced type II collagen regeneration, corroborated by safranin O-fast green staining results (Fig. [145]S15A). Additionally, GaPP@FSB showed superior inhibition of shoulder joint inflammation (Fig. [146]S16). Biomechanical testing (Fig. [147]S17) indicated that, GaPP@FSB exhibited significantly greater maximum failure tension compared to the other groups (GaPP@FSB: 12.3 ± 0.9 MPa; PET: 9.3 ± 1.2 MPa; Blank: 5.6 ± 2.1 MPa). Moreover, the stiffness of the GaPP@FSB group (17.0 ± 0.8 N/mm) was markedly higher than that of the PET and Blank groups. Gait analysis (Figure [148]S18) revealed that rats implanted with GaPP@FSB exhibited relatively coordinated upper limb movement, indicating the best gait trajectory. Collectively, these findings demonstrate that GaPP@FSB offers superior therapeutic efficacy for RCTs compared to commercialized patches, underscoring its significant clinical application potential. Discussion Arthroscopic surgery serves as the primary clinical intervention for RCTs; however, statistics indicate a significantly high failure rate in RCT repair surgery [[149]39–[150]43]. It is widely recognized that optimizing tendon-bone healing, in particular, achieving optimal regeneration of the enthesis fibrocartilage layer, plays a crucial role in improving postoperative RCT healing [[151]2]. In light of extensive research in tissue engineering, numerous strategies have been proposed to enhance tendon–bone healing. Previous studies have focused predominantly on designing RCT bioscaffolds via ECM-mimicking approaches [[152]5]. Samavedi et al. developed an electrospun biomimetic nanofibrous membrane for a tendon‒bone matrix utilizing electrostatic spinning technology. The microstructure of this nanofibrous membrane gradually transitioned from aligned to random, thereby promoting cell differentiation and enhancing tendon–bone healing [[153]43]. Ji et al. presented a cocktail-like hydrogel biograft loaded with BMSCs. This bioactive graft exhibited a nanoclay gradient and successfully facilitated regeneration of the fibrocartilage layer in a rat model of rotator cuff injury [[154]44]. However, it is crucial to acknowledge the evident limitations in current strategies, as most reported biodegradable bioscaffolds exhibit insufficient mechanical properties [[155]45]. Given that human tendons naturally experience tension in both the tense and relaxed states, if the scaffold does not match the mechanical strength of the tendon, there is a risk of in vivo rupture, which could lead to undesired interface regeneration or even surgical failure [[156]10]. For rotator cuff repair, the natural supraspinatus tendon typically exhibits a maximum failure force ranging from 4 to 24 MPa and an elastic modulus ranging from 10 to 170 MPa [[157]45]. The scaffold proposed in this study, designed as a patch-like reinforcement scaffold for rotator cuff repair, has a mechanical strength comparable to that of natural rotator cuff tendons (maximum failure force: 26.4 ± 0.6 MPa; elastic modulus: 102.0 ± 11.0 MPa). This avoids unexpected rupture of the patch after RCTs surgery. Available evidence has confirmed that RCTs are frequently accompanied by the upregulation of inflammatory cytokines and oxidative stress, which exacerbates disease progression [[158]46–[159]49]. Furthermore, localized inflammation plays a significant role in the development of pain and dysfunction in rotator cuff tendon injuries. GaPP@FSB also exhibited beneficial effects on anti-inflammation and the alleviation of oxidative stress, thereby contributing to the maintenance of seed cell populations involved in tissue regeneration. Importantly, we observed remarkable similarities between the structure and composition of the scaffold (FSBs) and those of the native tendon, bone, and inserted ECM. In particular, GaPP@FSB exhibited aligned protein nanofibers similar to the extracellular matrix of natural tendons. Furthermore, the controlled release of gallium ions facilitated tissue regeneration at the termination point of the RCT model, enhanced the tendon‒bone healing process, and restored shoulder joint function. Additionally, our findings indicate that FSBs can regulate tendon differentiation through the integrin α5β1-Akt-PI3K signalling pathway, similar to the natural tendon ECM. Therefore, this study highlights the advantages of utilizing decellularized matrices derived from fish bladders in tendon tissue engineering applications, which not only enhance tendon–bone healing but also create an optimal ECM microenvironment for postinjury tendon regeneration. Natural biomaterials, particularly those exhibiting exceptional mechanical properties, offer promising solutions for tendon‒bone repair. Some researchers have identified decellularized scaffolds sourced from bovine pericardium or pig small intestinal epithelium as potential candidates for RCT repair [[160]50]. However, these terrestrial biomaterials may harbour zoonotic pathogens or induce chronic inflammation [[161]51]. Recognizing the advantages of marine biomaterials, Han et al. proposed a nanocalcium silicate-fish scale (CS-FS) scaffold with superior strength that demonstrated successful tendon–bone healing in both rat and rabbit RCT models [[162]52]. Nevertheless, the size limitation of natural fish scales (2–3 cm) hinders their application in repairing extensive RCTs (> 5 cm). In this study, we introduced an alternative and more suitable solution by exploring the use of the natural fish bladder as a novel nonterrestrial biomaterial. First, on the basis of the aforementioned results, it was observed that the mechanical strength of the FSB was suboptimal and significantly lower than that of the natural supraspinatus tendon. However, Ga-MPN modification significantly enhanced the mechanical properties of GaPP@FSB scaffolds, making these properties comparable to those of the natural supraspinatus tendon. This enhancement can be attributed to the binding of TA with collagen through hydrogen bond formation and Michael addition for cross-linking. Moreover, our findings validate Ga-MPN modification as a multifaceted strategy that not only improves mechanical properties but also enables responsiveness to ROS in the microenvironment. The oxidation of the phenolic hydroxyl group of TA by ROS leads to scavenging effects on ROS (The catechol group was oxidized to form a quinone structure) and the subsequent release of Ga ions [[163]18]. The released Ga ions further participated in the regulation of cell differentiation. In previous studies, gallium Ga ions have been shown to promote the osteogenic differentiation of cells [[164]38]. Herein, for the first time, this study reports the beneficial effects of gallium ions on enthesis regeneration. Importantly, due to their antibacterial activities, both TA and Ga in GaPP@FSB contribute to the achievement of an almost 100% bactericidal rate against E. coli and S. aureus. Additionally, our results confirm the outstanding antioxidant and anti-inflammatory activities of GaPP@FSB. Furthermore, this novel tendon grafting biomaterial can induce osteogenic differentiation of BMSCs as well as tendonogenic differentiation of TSPCs while promoting regeneration within the fibrocartilaginous layer in an RCT animal model. Collectively, these outcomes highlight how GaPP@FSB addresses various challenges faced by clinicians. From the perspective of clinical applications in tendon injury repair, existing clinical grafts can be broadly categorized into three types: autografts, allografts, and artificial grafts [[165]53, [166]54]. Allografts and autografts elegantly retain the intricate architecture and essential components of natural tendons, along with their bioactive molecules. Nevertheless, the shadow of adverse effects—ranging from donor site morbidity and hostile immune reactions to the insidious onset of synovitis—casts a pall over these approaches. Such challenges not only elevate the surgical failure rate but also curtail the broader clinical adoption of these otherwise promising strategies [[167]55–[168]58]. Artificial grafts, predominantly biologically inert materials (e.g., commercialized PET grafts), demonstrate excellent mechanical strength but suffer from issues such as poor degradability and insufficient tissue ingrowth, which may lead to surgical failure [[169]59–[170]61]. Therefore, developing a rotator cuff tendon repair material that combines mechanical strength with biological activity remains a significant challenge [[171]37]. Although recent advancements have led to the development of scaffolds with high biological activity, their mechanical strength often falls short of matching that of human rotator cuff tendons, increasing the risk of re-tearing post-implantation. Hence, the fish bladder-based scaffold introduced in this study represents a promising clinical solution. This novel, degradable rotator cuff implant integrates appropriate mechanical strength with high biological activity. Furthermore, the size of the fish bladder is sufficiently large (Fig. [172]3A) to accommodate various types of rotator cuff tears—small (< 1 cm), medium(1–3 cm), large (3–5 cm), or massive (> 5 cm) [[173]62]—ensuring defect filling without inducing stress shielding caused by overly tight sutures. Not limited to the aforementioned application, we anticipate broader future applications of this fish bladder-based scaffold. Given that MPN can serve as a programmable strategy, organ repair of various types may potentially be realized by modulating the combination of metal ions. Moreover, owing to its robust and patch-like structure, this scaffold could also find potential applications in periosteum regeneration and anterior cruciate ligament reconstruction (by rolling the patch into a tubular shape). In summary, our findings confirm that GaPP@FSB promotes tissue ingrowth and enthesis regeneration while restoring motor function in the rat model. Overall, GaPP@FSB addresses the limitations of existing tendon repair grafts and demonstrates substantial clinical application potential. Conclusion In this study, we utilized FSB as a natural source to develop a novel patch-like reinforcement biomaterial with high strength and multiple biological activities for tendon tissue engineering. On the one hand, Ga-MPNs were incorporated into GaPP@FSB to enhance its mechanical properties, matching those of native rotator cuff tendons. On the other hand, GaPP@FSB exhibited exceptional antioxidant, anti-inflammatory, and antibacterial activities, effectively addressing detrimental factors such as inflammation and bacterial infection that impede RCT repair. Furthermore, GaPP@FSB demonstrated the responsive release of gallium ions to stimulate osteogenic differentiation and regulate tendon differentiation through the ECM-focal adhersion-Akt-PI3K mechanical signal transduction pathway. Ultimately, in a rat RCT model, GaPP@FSB facilitated healing at the rotator cuff tendon‒bone interface and enthesis regeneration while restoring the physiological function of the rotator cuff. Overall, our findings highlight that combining gallium ions with polyphenols has a comprehensive positive effect on promoting tendon-to-bone repair at the interface. Considering the fish swim bladder’s inherent composition of aligned nanofibers, this unique nanostructure exquisitely mirrors the intricate arrangement of the extracellular matrix found at the tendon-bone interface, thus presenting a remarkable example of natural nanobiomimicry. The modification of the natural swim bladder via MPNs represents an innovative class of high-performance and multibioactive materials that hold promise as potential candidates for clinical tendon repair or improvement in tendon–bone healing. Materials and methods Preparation of gapp@fsb scaffolds First, the swim bladders of freshwater carp were selected from food waste. After being thorough washed with ample water, the samples were meticulously stripped of fascia, residual blood, and other soft tissues on the surface. The FSBs that exhibited directional structures when observed with the naked eye were chosen. The cleaned swim bladders were subsequently immersed in 75% ethanol and sonicated for 30 min to eliminate potential pathogenic microorganisms. Following this step, decellularized scaffolds derived from the swim bladders were prepared via a previously reported method [[174]13]. Briefly, the treated swim bladders were uniformly cut into pieces measuring 5 × 5 cm in size and immersed in a mixed solution containing sodium dodecyl sulfate (SDS) and Triton X-100 (both at a concentration of 1%, 500 ml). After soaking for 24 h at room temperature, the swim bladder samples were then exposed to DNase I solution (0.3 mg/ml) at 37 °C for six hours. Subsequently, the swim bladder samples were thoroughly washed with deionized water before being freeze-dried. The freeze-dried samples were then rehydrated by soaking them again in deionized water to obtain decellularized fish bladder scaffolds called FSBs. Furthermore, the FSB scaffold material was used to prepare PP@FSB and GaPP@FSB. The freeze-dried FSB (10 g) was immersed in a 100 mM TA solution (500 ml) for 6 h, followed by washing with deionized water to obtain the crosslinked FSB, designated PP@FSB. GaPP@FSB was prepared by configuring Ga-TA mixed solutions. Specifically, three mixed solutions containing gallium nitrate and TA were prepared with a constant concentration of 100 mM TA and varying concentrations of gallium nitrate (10 mM, 20 mM, and 30 mM). After centrifugation at 10,000 rpm, each sample of FSB (10 g) was soaked in the respective mixed solution (500 ml) for 6 h. The samples were subsequently transferred to Tris-HCl solution (pH = 9.0) for an additional soaking period of 12 h. Finally, the samples were washed with deionized water to obtain GaPP@FSB loaded with different amounts of gallium. On the basis of the concentrations of gallium nitrate used in the mixed solutions (i.e., 10 mM, 20 mM, and 30 mM), the resulting scaffolds were named as follows: 1GaPP@FSB, 2GaPP@FSB, and 3GaPP@FSB. Additionally, the DNA, GAG, collagen and elastin concentrations (n = 3 for each group in every test) in both fresh and FSB samples were evaluated as previously [[175]63] described by the DNeasy Blood & Tissue Kit (Qiagen, Milan, Italy), Glycosaminoglycan Assay Blyscan, Insoluble Collagen Assay and Fastin Elastin (B1000, S1000 and F2000, respectively, Biocolour, UK) following the manufacturer’s instructions. Characterization of gapp@fsb scaffolds The microstructural differences among FSB, PP@FSB, and GaPP@FSB were observed via scanning electron microscopy (SEM, SU8220, Hitachi, Japan). Energy-dispersive X-ray spectroscopy (EDS) and mapping analysis were employed to verify the presence of Ga and the elemental changes in each sample group. The microscopic structure of the protein fibre composition in the FSB scaffolds was examined via resin embedding, sectioning, and transmission electron microscopy (TEM, JEM-2100 F, JEOL, Japan). On the basis of the TEM images, 50 fibres were selected from each group of samples via ImageJ software, and their fibre widths were subsequently calculated. Furthermore, X-ray photoelectron spectroscopy (XPS; ESCALAB 250Xi, Thermo Fisher Scientific, USA) equipped with a monochromatic Al Kα X-ray source (1486.6 eV, 15 kV) was utilized for surface chemistry characterization of the support materials. Additionally, Fourier transform infrared spectroscopy (ATR-FTIR) was used to analyse the chemical composition of each scaffold group. The release kinetics of Ga ions from the GaPP@FSB scaffolds in ROS-rich and acidic microenvironments were measured via plasma atomic emission spectroscopy (ICP; 715-ES Varian Inc., USA). Briefly, 1 cm diameter discs cut from GaPP@FSB scaffolds were immersed in centrifuge tubes containing 10 ml of deionized water (n = 4). The centrifuge tubes were placed on a shaker at a constant temperature of 37 °C, and the supernatant was collected at different time points for ICP analysis to plot a Ga release curve. A total antioxidant capacity test kit (FRAP method) was used to assess the overall antioxidant rate of each scaffold group. In brief, 10 mg samples were extracted from each group of scaffolds and subjected to testing in accordance with the manufacturer’s instructions. The final assessment of antioxidant efficiency was based on the corresponding FeSO[4] concentration. For the swelling rate of the scaffolds, first, the initial dry and wet weights of each group were measured, and the wet weight was recorded as W[0]. The hydrogels were subsequently immersed in PBS at 37 °C, and the changes in scaffold weight were recorded at different time points. The percentage swelling was calculated according to Eq. ([176]1), where Wt is the weight of the hydrogel that dissolves at a predetermined time interval. graphic file with name d33e1168.gif 1 Mechanical analysis of GaPP@FSB scaffolds: Sample preparation for mechanical examination: In accordance with the previous method, the samples of FSB, PP@FSB and each group of GaPP@FSBs (n > 5) were cut into a dogbone shape with a total length of 24 mm and width of 5 mm. Both ends of the samples were rectangular with dimensions of 5 mm × 7 mm, and the middle portion had a length of 10 mm and a narrowest width of 2 mm. The tensile strength of each group of scaffolds was measured by a universal mechanical testing machine (Instron-5566, Instron, Canton, USA). The individual parameters were as follows: transducer accuracy of 0.0001 N (model 5566, Instron, Canton, MA) and a maximum load of 10 kN. Stress‒strain curves were collected for each group of samples via Fast-Track software (Instron, Canton, USA). In addition, the ends of the samples were clamped with a pneumatic fixture in a dry state. Next, the samples were pretreated five times with a preloading force of 0.5 N and then loaded at a rate of 10 mm/min (strain rate of 1.7%/s) until complete fracture. The load and displacement of each patch were recorded at a sampling rate of 20 Hz, and the stress‒strain curves, maximum tensile strengths, and stresses were recorded for each set of samples during the test. In vitro cytocompatibility Rat mesenchymal stem cells (BMSCs) were obtained from Cyagen Biotechnology Co. Primary rat tendon stem/progenitor cells (TSPCs) were extracted from rat Achilles tendon tissue as previously reported [[177]52]. The medium used for cell culture was complete alpha minimal essential medium (α-MEM, Gibco, USA) (containing 10% foetal bovine serum (FBS, Gibco, USA), 1% IU/ml penicillin and 1 µg/ml streptomycin (HyClone, USA)). The cells were cultured in an incubator (37 °C, 5% CO2) and kept moist. In this study, the effect of each group of GaPP@FSB scaffolds on the viability of BMSCs and TSPCs was preliminarily verified via a CCK-8 assay. Briefly, all samples were cut into 1.2 cm diameter discs, which were soaked in 75% ethanol, sterilized by UV irradiation, and then attached to the bottom of the wells of a 24-well plate (Corning, USA). The viability of the BMSCs and TSPCs was analysed at 1, 3, and 7 days after they were seeded onto each sample with use of a CCK-8 kit (Dojindo, Japan). Additionally, BMSCs or TSPCs were seeded onto the scaffolds at a density of 2 × 10^4. After 24 h, the cells were fluorescently labelled with a live/dead staining kit (Dojindo, Japan), and cell adhesion was observed via fluorescence microscopy. The effect of scaffolds on cell differentiation According to the data from the CCK-8 analysis, 2GaPP@FSB was selected as the experimental group for subsequent experiments. To investigate the bioactivity of GaPP@FSB in inducing the differentiation of BMSCs and TSPCs, the levels of cell differentiation-related genes and proteins were evaluated via Q‒PCR and immunofluorescence staining. Briefly, the scaffolds were cut into 1.2 cm diameter discs, soaked in 75% ethanol, sterilized by UV irradiation, and attached to the bottom of the wells of a 24-well plate (Corning, USA). BMSCs or TSPCs were then seeded onto the scaffolds at a density of 2 × 10^4. After 7 days of culture, the cells were collected by digestion and centrifugation, and then, mRNA was extracted with TRIzol reagent (Invitrogen, USA) according to the manufacturer’s instructions. The extracted mRNA was reverse transcribed into cDNA by using an mRNA reverse transcription kit (PrimeScript™ RT reagent Kit, Takara, Japan). Subsequent RT‒PCR-based analysis (TB Green^® Premix Ex Taq, Takara, Japan™) was performed to detect the expression levels of relevant genes in each group of cells (n = 4). The osteogenic differentiation-related genes in the BMSCs that were examined were Runx2, Ocn, Opn, and BMP-2. The tendonogenic differentiation-related genes in the TSPCs that were examined were Scx, Mkx, Col-1, and Tnmd. The sequence information of the primers used is shown in Table [178]S1. For immunofluorescence staining, to avoid nonspecific antibody binding to the scaffold, this experiment was performed via a Transwell system. Briefly, cells were cultured in the lower chamber of the Transwell system, and each group of scaffolds was placed in the upper chamber. After 7 days of culture, the cells were fixed with 4% paraformaldehyde solution. OPN (antibody purchased from Affinity, China, #AF0227) in BMSCs and Tnmd (antibody purchased from Affinity, China, #DF13715) in TSPCs were then fluorescently labelled, and the expression levels of osteogenic differentiation markers and tendon differentiation markers were analysed. In addition, the cell membranes and nuclei of the cells were labelled with Plasm dye and Dapi dye (Dojindo, Japan), respectively, for fluorescent labelling. Additionally, the BMSCs cultured via this method were subjected to alkaline phosphatase (ALP) staining (Solarbio, China) and alizarin red (ARS) staining (Solarbio, China) after 7 and 28 days of cultivation, respectively. In vitro antioxidant and anti-inflammatory activities of gapp@fsb To investigate the ability of GaPP@FSB to scavenge excessive intracellular ROS, a DCFH-DA fluorescent probe (Dojindo, Japan) was used. Briefly, TSPCs were seeded onto a Transwell plate and cultured with each group of scaffolds as previously described. Next, the cells were treated with α-MEM containing 200 µM hydrogen peroxide for 6 h, followed by replacement with complete medium. After 3 days of culture, the cells were treated with DCFH-DA fluorescent probe solution, followed by observation and analysis via fluorescence microscopy and flow cytometry to assess changes in intracellular ROS levels. Hydrogen peroxide-induced cell apoptosis was assessed via an Annexin V-FITC Apoptosis Detection Kit (Dojindo, Japan), which incorporates flow cytometry and fluorescence microscopy for analysis. Anti-inflammatory activity of the scaffolds Next, we explored the anti-inflammatory activity of each group of scaffolds based on Raw264.7 cells. Briefly, Raw264.7 cells (Cyagen Biotechnology Co., Ltd.) were seeded onto each group of scaffolds at a density of 1 × 10^4. After cell attachment, all groups of cells were subsequently stimulated with LPS (10 ng/ml, Solarbio, China) for 12 h to promote the polarization of Raw264.7 cells towards the proinflammatory phenotype of M1. After 7 days of culture, the levels of cell polarization-related genes and proteins in Raw264.7 cells were evaluated via Q-PCR and immunofluorescence staining as previously described. The genes related to M1 polarization and the proinflammatory factors detected were CCR7, iNOS, iL-1β, and TNF-α, whereas the genes related to M2 polarization and the anti-inflammatory factors detected were Arg-1, CD206, iL-10, and iL-1ra. For immunofluorescence staining, the cells were fixed with 4% paraformaldehyde solution after being cultured in the groups of scaffolds for 3 days. Then, Arg-1 in Raw264.7 cells was fluorescently labelled to analyse the effect of GaPP@FSB on M2 polarization. In addition, the cell membrane and nucleus were fluorescently labelled with Plasm dye and Dapi dye, respectively. In vitro haemolysis assay The haemolysis assay is a widely used method for assessing the blood compatibility of materials, and the haemolysis rate in each sample was determined following the protocol described in the literature [[179]64]. The samples were fragmented into particles, followed by incubation with 500 µl of 10% (v/v) blood cell suspension at 37 °C for 1 h. Subsequently, the samples were centrifuged at 1000 rpm for 10 min, and the supernatants from each group were collected to measure the absorbance at 540 nm. Deionized water and PBS were used as negative and positive controls, respectively. The haemolysis rate was calculated via Eq. ([180]2) graphic file with name d33e1228.gif 2 where Bs represents the absorbance at 540 nm of the sample suspension, Bp denotes that of PBS, and Bw corresponds to deionized water. In vitro antibacterial activity of gapp@fsb The antibacterial activity of GaPP@FSB was investigated against both Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus). Briefly, scaffold samples (200 mg) from each group were taken and immersed in a bacterial suspension (2 × 10^6 CFU/mL, 10 ml) and subsequently cultured for 24 h in a 37 °C incubator. Then, the bacterial suspension of each group was diluted 1000 times, and the bacteria were homogeneously inoculated on blood agar plates via the scratch method. The incubation was continued for 24 h, and the number of colonies on the blood agar plates was recorded. In addition, the scaffolds of each group after cultured with the bacterial suspension were fixed with 2.5% glutaraldehyde solution and dehydrated with an ethanol gradient. After all the samples were freeze-dried, the morphology of the bacteria in each group was observed via SEM. In vivo experiments Animal modelling of rotator cuff injuries All animal experiments were performed in accordance with the procedures approved by the Experimental Animal Ethics Committee of the Laboratory Animal Center of Nantong University (S20240116-004). All experimental procedures complied with the ARRIVE guidelines and were performed in accordance with the British Animals (Scientific Procedures) Act 1986 and related guidelines. A rat RCT was modeled as shown in Fig. [181]S12. Briefly, 20 mature male SD rats (260 ± 22 g) were used. After anaesthesia with isoflurane gas, the rats were fixed in a supine position, and the skin was sterilized after the shoulder joint was shaved. The skin of the acromion was incised with a scalpel to expose the supraspinatus tendon of the rotator cuff, and the supraspinatus tendon was separated with a Gram needle. After the supraspinatus tendon was severed, it was repaired using each group of scaffolds (3 mm × 6 mm), i.e., one end of the scaffold was sutured to the stump of the supraspinatus tendon, and the other end was fixed to the humeral head. The suture used in the procedure was a 4–0 Ethibond suture (Johnson & Johnson). A 1-mm-wide hole was made in the humeral head via Gram’s needle so that the suture could pass through the hole and securely fix the scaffold to the humeral head. In the control group, the rotator cuff stump was fixed to the humeral head with surgical sutures. Next, the wound was sutured and sterilized. The rats were kept in large cages and allowed to move freely to prevent postoperative joint stiffness. Subsequently, the rats were sacrificed at 4 and 8 weeks after surgery for harvesting of the humeral head-supraspinatus tendon or for biomechanical testing (n = 5). Histopathological analysis A histopathological analysis was performed at 4 and 8 weeks after surgery, in which the supraspinatus tendon attached to the humeral head of each rat was harvested and other soft tissues were removed from the humeral head. All the samples were subsequently fixed in a 10% neutral formalin solution for three days and then immersed in EDTA decalcification solution (pH = 7.2; Solarbio, China) for one month. After complete decalcification, the samples were dehydrated via gradient alcohol solutions (30-100%). Following treatment with xylene, the samples were embedded in paraffin and sectioned along the tendon-to-bone direction at a thickness of 5 μm. Safranin-O-fast green staining, H&E staining, and toluidine blue staining were performed according to the manufacturer’s instructions (Solarbio, China). The stained sections were observed under a light microscope (eclipse 80i; Nikon), and histopathological images were captured with a camera (ds-5 m; Nikon). The proportion of fibrocartilage tissue at the tendon‒bone interface was quantified via ImageJ software on the basis of safranin-O-fast green staining and toluidine blue results. Additionally, each group received a histological score [[182]65]; details of the modified histological scoring system used for evaluating the regenerated tendon‒bone interface are provided in Table [183]S2. Biomechanical test and footprint analysis: Biomechanical tests were conducted via an Instron-5566 material testing system (USA). Four specimens of the rat supraspinatus brachii tendon were used in each group. The upper limb bone and supraspinatus muscle were kept intact on the specimens to facilitate clamping, while other soft tissues were removed. To minimize slippage, the humeral head and upper limb bone were secured in a 1.5 cm diameter mould with denture powder before the specimen was fixed. Preconditioning of the rotator cuff sample involved applying a preload of 1 N at a test speed of 5 mm/min (strain rate: 0.8%/s) for a total of 20 cycles. The ultimate failure load (N), ultimate failure force (MPa), and load deformation curve were recorded. The stiffness (N/mm) was calculated from the slope of the linear region in the load‒deformation curve at the point of maximal load‒failure. A footprint analysis was conducted to assess shoulder function recovery in the rats via the CatWalk XT system (Noldus Information Technology, Wageningen, Netherlands). Additionally, this work also carried out additional animal experiments, employing commercially available polyethylene terephthalate (PET) tendon grafts (Ligament Advanced Reinforcement System, LARS^®, France) as the control group to evaluate and highlight the advantages of GaPP@FSB in repairing rotator cuff injuries (Fig. [184]S14). Furthermore, the clinically established eight-shape suture technique was utilized for rotator cuff repair as an additional control group(n = 5) [[185]66]. Statistical analysis All the quantitative data were analysed via SPSS version 18.0 (Chicago, IL, USA) and are expressed as the means ± standard deviations (SDs). Statistical significance among three or more groups was determined using a one-way analysis of variance (ANOVA) followed by the Tukey’s post-hoc test. Statistical comparisons between two groups were performed using Student’s t-test. Statistical significance was indicated as ∗∗∗p < 0.001, ∗∗ p < 0.01, and ∗ p < 0.05. Electronic Supplementary Material Below is the link to the electronic supplementary material. [186]Supplementary Material 1^ (9.9MB, docx) Acknowledgements