Inherently bioactive iron-chelating Poly (N-acryloyl 2-glycine)/chitosan hydrogel scaffolds orchestrating dual hypoxic-immune microenvironment for functional meniscus regeneration

Abstract

Despite progress in tissue-engineered meniscus (TEM) as alternatives to meniscectomy, challenges remain in inflammatory regulation, oxidative resistance, and mechanical stability under pathological microenvironments. Innovatively, we combined personalized meniscus scaffold, hydrogel ion crosslinking network technology, and microenvironment regulation function to prepare a multifunctional poly (N-acryloyl 2-glycine)/chitosan (PACG/CS) composite hydrogel meniscus scaffold featuring heterogeneous bionic structure, high strength and toughness, hypoxic inducing activity, and anti-inflammatory and antioxidant effects. Crucially, the inherently bioactive hydrogel networks crucially leveraged their carboxyl groups to orchestrate iron ion chelation, establishing a hypoxia-mediated microenvironment that dynamically modulated pro-/anti-inflammatory equilibrium, which in turn supported the chondrocyte survival, facilitated the development of a cartilage matrix, and ultimately promoted the meniscus regeneration. Notably, peripheral blood mesenchymal stem cells (PBMSCs) exhibited superior meniscus regeneration efficiency in low-oxygen conditions compared to bone marrow mesenchymal stem cells (BMSCs). After evaluating the effects of hypoxia environment induced by highly efficient iron chelation of PACG/CS hydrogel scaffolds on the activation of HIF-1α signaling pathway, anti-inflammatory and antioxidant regulation, the regulatory mechanism of immune microenvironment on the growth and cultivation quality of TEM were elucidated in vivo and in vitro. Overall, our have important implications for comprehending the biological impacts of biomaterials and developing novel approaches for meniscus regeneration.

Graphical abstract

A multifunctional poly (N-acryloyl 2-glycine)/chitosan (PACG/CS) composite hydrogel meniscus scaffold is fabricated for the meniscus regeneration, which possess the heterogeneous bionic structure, high strength and toughness, hypoxic inducing activity, and anti-inflammatory and antioxidant effects.

Highlights

• •A personalized PACG/CS hydrogel meniscus scaffold was innovatively synthesized by multistrategy-integrated bionic design.
• •An inherently bioactive iron-chelating scaffold enables dual hypoxic-immune regulation via HIF-1α activation.
• •First demonstration of PBMSCs outperforming BMSCs in hypoxia-adaptive meniscus regeneration.
• •This hydrogel meniscus scaffold possesses biomimetic, biomechanics, hypoxic, anti-inflammatory and antioxidant effects.

1. Introduction

The meniscus is a critical heterogeneous structure within the knee joint, essential for load distribution, shock absorption, joint stability, lubrication, and proprioception [1]. However, its avascular inner region severely limits healing capacity following injury, often leading to pain, swelling, functional impairment, and an elevated risk of debilitating osteoarthritis (OA) [2,3]. Current surgical treatments, primarily partial or total meniscectomy, fail to restore long-term biomechanical function and accelerate cartilage degeneration [3]. Consequently, developing effective tissue engineering meniscus (TEM) replacements that prevent OA and alleviate persistent symptoms is paramount. Existing biomaterial strategies for TEM face significant limitations: 1) absorbable scaffolds, which rely on their ability to support tissue while gradually degrading and facilitating the formation of new meniscal tissue; and 2) non-absorbable permanent implants, which aim to mimic the biomechanical function of the natural meniscus. However, until now, no replacement meniscus has been found to completely prevent articular cartilage injury after meniscectomy [4].

The clinical outcomes of implants have been less than ideal, primarily because they fail to replicate the structure-function relationships of the tissue [5]. The natural meniscus is composed of heterogeneous connective tissue cells, and the biomimetic heterogeneous structure of the natural meniscus plays a crucial role in the fabrication of meniscus scaffolds. 3D printing techniques using biodegradable polymers can be applied to the regeneration of various tissues or organs [6]. With advancements in new technologies, novel hydrogels with more advanced structures and diverse components are receiving increasing research attention. These new hydrogels offer new research directions towards meniscus repair strategies in tissue engineering and intrinsic cartilage regeneration. Medical imaging advances enable personalized hydrogel implant design by customizing geometry for 3D printing [7].

In tissue engineering strategies, it is crucial to control the inflammation present in diseased tissues and the inflammatory stimuli induced by TEM. The regulation of immune response through immune microenvironment, especially affecting the phenotype of macrophages [8,9], has become an important link in the field of tissue engineering. Pro-inflammatory factors cause macrophages to differentiate into the M1 phenotype, leading to the secretion of pro-inflammatory cytokines such as IL-1, IL-6, and TNF-α, which trigger a strong inflammatory response [10]. In contrast, anti-inflammatory factors such as IL-4 or IL-13 drive macrophages to differentiate into the M2 phenotype, which secretes IL-10 and TGF-β1, aiding in the integration of biomaterials and tissue regeneration. Consequently, researchers are inclined to design innovative biomaterials to promote macrophage polarization towards the M2 phenotype [11]. Precise control of the M1/M2 balance is critical for healing success. Beyond inflammation, maintaining cellular redox homeostasis and optimizing the oxidative microenvironment are also essential for regeneration [12]. Antioxidant supplementation can reduce oxidative stress and improve outcomes [[13], [14], [15]]. This combined strategy of regulating inflammation and optimizing the redox microenvironment offers a promising approach to enhance cellular integrity, minimize oxidative damage, and improve regeneration efficacy in TEM [16,17].

Hypoxia-inducible factor 1α (HIF-1α) is also essential in promoting cartilage regeneration. Under hypoxic conditions, HIF-1α stabilizes in the cytoplasm and dimerizes with HIF-1β in the nucleus. HIF-1α then mediates cellular adaptation to hypoxia by regulating the expression of genes involved in angiogenesis, glucose transport, glycolysis, and iron metabolism [18,19]. Mesenchymal HIF-1α inactivation in developing mouse limbs impairs cartilage/joint development and reduces expression of chondrogenic markers SOX9, collagen II, and aggrecan [20]. Since both oxygen and Fe³⁺ inhibit HIF-1α, iron chelators like deferoxamine (DFO) can stabilize it by mimicking hypoxia and improve cartilage mechanical properties [21,22]. Due to the toxicity of DFO chelators, their clinical application has been hindered. Thus, utilizing safer alternatives to deferoxamine chelators to achieve sustained activation of HIF-1α is of considerable importance. Sun et al. developed an injectable poly (decyl glycolate)-co-poly (ethylene glycol)/poly (acrylic acid) (PEGS/PAA) hydrogel that stabilizes HIF-1α via iron chelation, creating hypoxic conditions. This approach suppresses inflammatory responses, thereby improving early cartilage formation and later promoting angiogenesis [23]. Therefore, tissue engineering strategies using safe biomaterials to activate HIF-1α signaling are crucial for regulating the local microenvironment and promoting intrinsic cartilage regeneration.

Conventional repair approaches dependent on exogenous biologics/drug delivery face critical limitations: uncontrolled release profiles and carrier toxicity, resulting in transient efficacy and complex regimens. Addressing these challenges, we engineered inherently bioactive scaffolds integrating structural support with autonomous chondrogenic capacity. This material-driven paradigm enables spatiotemporal repair through intrinsic bioactivity, bypassing exogenous dependency while ensuring tissue-specific regeneration. In this work, a multifunctional inherently bioactive iron-chelating PACG/CS hydrogel meniscus scaffold was designed featuring a heterogeneous biomimetic structure, high strength and effective anti-inflammatory and antioxidant properties (Fig. 1). The scaffold was created using personalized customization and precise regulation of the anti-inflammatory and antioxidant microenvironment. Crucially, the carboxyl groups within the hydrogel network could effectively chelate iron ions, creating a hypoxic microenvironment that balanced the pro- and anti-inflammatory responses, as well as immune cell subsets. This balance supported the chondrocyte survival, facilitated the development of a cartilage matrix, and ultimately promoted meniscus regeneration. By exploring different concentrations of PACG in the hydrogel, the synthesis conditions were optimized to achieve a balance between the structural architecture and the cellular activities critical for meniscus regeneration. The effects of hypoxia induced by high-efficiency iron chelation of PACG/CS hydrogel scaffolds on activation of HIF-1α signaling pathway, inhibition of inflammatory response, regulation of equilibrium expression of oxidation generation and elimination were investigated in vivo and in vitro. We analyzed and discussed the regulatory mechanism of the improvement of immune microenvironment on the growth and cultivation quality of TEM, with a view to cultivating high-quality TEM with diversified, personalized and clinical value in terms of microstructure, composition, mechanics and biological functions.

Fig. 1.

Schematic illustration of the whole study. This study integrates hydrogel ionic crosslinking technology, personalized meniscus injection mold, and microenvironmental regulation through strong iron ion chelation to develop a multifunctional composite hydrogel meniscus scaffold, which features heterogeneous biomimetic structures, high strength and toughness, excellent self-recovery, as well as anti-inflammatory and antioxidant properties. ACG is an acrylamide monomer with a carboxylic acid functional group at its terminus. First, a mixture of short-chain chitosan (CS) and carboxyl-rich ACG monomers is placed into a customizable meniscus injection mold. Then, the mixture is exposed to UV light to form a semi-interpenetrating network and PACG/CS hydrogel meniscus. After being total transplanted into the rabbit knee joint, the customized PACG/CS composite hydrogel menisci were chelated with iron ions to create a hypoxia-mimicking microenvironment, promoting the meniscus regeneration.

2. Results

2.1. Characterizations, biomechanics and biocompatibility of PACG/CS hydrogels

Hydrogen nuclear magnetic resonance spectrum proved the successful synthesis of ACG monomer in Fig. S1. As shown in Fig. 2A, a PACG/CS composite hydrogel was synthesized through UV-initiated radical polymerization, achieving covalent/dual hydrogen bond crosslinking of PACG alongside homogeneous chitosan dispersion within the hydrogel matrix. Three composite hydrogels, PACG, PACG/CS-1.6 and PACG/CS-3.9, were prepared with various chitosan concentrations in the hydrogels (Fig. 2B). Compared to the pure PACG hydrogel, PACG/CS-1.6 and PACG/CS-3.9 hydrogels had higher deformation ductility and compressive strength that were comparable to that of the natural meniscus in human and rabbit meniscus (Fig. 2C). Scanning electron microscope (SEM) showed that single-network PACG hydrogel was compact with distributed isotropic pore structures, which may ascribe to the strong intermolecular hydrogen bond interaction. Addition of the chitosan caused a “horizontal-vertical” regular orientation of pore structures for PACG/CS-1.6 hydrogel, and as the chitosan content increased, the “horizontal-vertical” orientation arrangement structure gradually changed into a large-sized irregular pore structure for PACG/CS-3.9 hydrogel (Fig. 2D). In addition, the biocompatibility of these hydrogels was evaluated by live/dead double staining and cytoskeleton staining in vitro and subcutaneous implantation in vivo. BMSCs encapsulated within PACG/CS hydrogels exhibited a rounded, chondrocyte-like morphology, demonstrating the favorable role of these three-dimensional matrices in maintaining the characteristic shape of chondrogenic cells (Fig. 2E). The favorable MSCs activity and growth capacity within all the hydrogels (Fig. 2F and G) and the negligible inflammatory response (Fig. 2H and S2) demonstrated the good biocompatibility.

Fig. 2.

Characterizations, biomechanics and biocompatibility of PACG/CS hydrogels. (A) Schematic diagram of preparation of PACG/CS composite hydrogel. (B) Appearance of the various hydrogels. (C) Mechanical behaviors of various hydrogels, rabbit and human meniscus. (D) SEM of various hydrogels. (E) Cytoskeleton staining revealing morphology of BMSCs in hydrogels (Blue: DAPI; Orange: Phalloidin). Scale bar: 100 μm. (F) Live/Dead double staining of BMSCs in different hydrogels. Scale bar: 500 μm. (G) The effects of various hydrogels on the growth of BMSCs were detected by CCK-8. (H) In vivo biocompatibility of various hydrogels after subcutaneous implantation. Different groups were observed to be devoid of cellular deposition. The inflammatory response to the implants were evaluated by a hematoxylin-eosin (H&E) stain. Blue arrows indicate neutrophils. Scale bar: 200 μm.

2.2. Mimicking hypoxia microenvironment induced by PACG/CS hydrogels

FeCl₃ with a concentration of 0.1 mol/L was selected to study the changes in the structure and properties of various hydrogels after soaking in FeCl₃ for different time (0 h, 4 h, 24 h and 168 h). During this phase, chloride and ferric ions diffused into the PACG/CS hydrogel, acting as dual crosslinking agents to establish amino-anion and carboxyl-cations domains: (i) chloride-induced salting-out initiated polymer aggregation, prompting spontaneous chitosan chain collapse into entangled networks; (ii) Fe³⁺ coordinated with carboxyl groups through electrostatic interactions, generating ionic crosslinks within PACG (Fig. 3A). As the soaking time in FeCl₃ increased, the color of the all hydrogels deepened continuously, indicating greater incorporation of Fe³⁺ ions to crosslink with carboxyl groups within the hydrogel (Fig. 3B). Compared to the PACG hydrogel, the PACG/CS hydrogels exhibited the lower swelling ratio and stronger iron ion chelating ability along with the extension of soaking time, which indicated the above-mentioned dual crosslinking effects by the chloride and ferric ions on the structural compaction of hydrogels. In addition, we quantificationally determine the iron ion chelating ability by iron ion detection kit in Fig. 3C. As expected, the iron ion chelating ability of the PACG/CS-1.6 group was significantly higher than that of the PACG and PACG/CS-3.9 groups, which was attributed to its densest network (Fig. 2D) with optimal formulation to facilitate the efficient chelation effect. Under this circumstance, RAW264.7 cells cultured with PACG/CS-1.6 hydrogels exhibited higher levels of HIF-1α compared to those cultured with control group (Fig. 3E and F). To further determine the regulatory role of HIF-1α in hypoxia microenvironment and meniscus regeneration, two inhibitors of HIF-1α were selected for interference validation. The expression of HIF-1α protein was inhibited by different concentrations of FM19G11 (0.3 μM and 1 μM) and Oltipraz (30 μM and 50 μM). The results showed that 50 μM Oltipraz had the most obvious inhibition effect on HIF-1α protein. For further verification of HIF-1α signaling pathway, 50 μM Oltipraz was chose based on its optimal efficacy in suppressing HIF-1α protein expression (Fig. 3E and F). The expression of HIF-1α protein was overexpressed by CoCl₂ (200 μM) in Fig. S3A. The BMSCs, PBMSCs and RAW264.7 cells cultured with PACG/CS-1.6 hydrogel exhibited highest levels of HIF-1α compared to those cultured with control, PACG, PACG/CS-3.9 groups (Fig. 3G). Correspondingly, we investigated the simulation of endogenous hypoxia induced by PACG/CS hydrogels in vivo, and found that the strong HIF-1α staining and nuclear localization were detected in the synovial tissues surrounding PACG and PACG/CS hydrogels at one week postoperatively, whereas no such strong staining was observed in the Control group (Fig. 3H). HIF-1α expression was highest in the PACG/CS-1.6 group compared to the Control, PACG and PACG/CS-3.9 groups, which suggested the induction of an early hypoxic-like microenvironment in vivo (Fig. 3I).

Fig. 3.

Mimicking hypoxia microenvironments induced by PACG and PACG/CS hydrogels. (A) Schematic diagram of the transition from composite hydrogels to composite hydrogels by the soaking strategy. (B) Photographs of various hydrogels soaked in FeCl₃ solution (C_FeCl3 = 0.1 mol/L) at different times. (C) Iron ions chelating activity of PACG, PACG/CS-1.6, PACG/CS-3.9 hydrogels in FeCl₃ solution were detected by iron ion detection kit. Differences were considered significant at ∗∗p < 0.01, ∗∗∗p < 0.001. (D) Schematic diagram of the effect of PACG and PACG/CS hydrogels on PBMSCs, BMSCs and RAW264.7 cells after chelating iron ions. (E, F) Western blotting showing the protein expression of HIF-1α and the effects of FM19G11 and Oltipraz on HIF-1α protein in RAW264.7 cells cultured with PACG/CS-1.6 hydrogel. (G) Western blotting showing the protein expression of HIF-1α in (a) BMSCs, (b) PBMSCs and (c) RAW264.7. (H) Immunohistochemical analysis of HIF-1α in the synovial tissue cells surrounding hydrogels in vivo. Blue arrows indicate HIF-1α positive cells. Scale bar:100 μm.(I) Quantitative analysis of HIF-1a protein expression in the synovial tissue cells surrounding hydrogels in vivo. (∗∗p < 0.01 statistical significance compared with Control group; ^##p < 0.01 statistical significance compared with PACG group; ⁺p < 0.05, ⁺⁺p < 0.01 statistical significance compared with PACG/CS-1.6 group).

2.3. Immunomodulatory effect of PACG/CS hydrogels

Upon stimulation with LPS, RAW264.7 cells produced a significant amount of ROS. However, co-culturing RAW264.7 cells with PACG and PACG/CS hydrogels resulted in a notable ROS scavenging effect. The results demonstrated that PACG and PACG/CS hydrogels could effectively scavenge ROS by chelating iron ions (Fig. 4B). It is well known that the downstream target genes of HIF-1α, stromal cell-derived factor 1α (SDF-1α) and vascular endothelial growth factor (VEGF), play crucial roles in stem cell recruitment and angiogenesis, both of which are essential for tissue repair. Therefore, we investigated the expression of SDF-1α and VEGF in RAW264.7 cells by ELISA, and found that the secretion of SDF-1α and VEGF in the PACG and PACG/CS hydrogel group were significantly higher than that in the Control and LPS groups. The PACG/CS-1.6 group exhibited higher levels of SDF-1α and VEGF secretion compared to the PACG group. To further study the immunoregulatory effect of HIF-1α, HIF-1α inhibitor (Oltipraz) and activator (CoCl₂) were added to the PACG/CS-1.6 group to affect the expression of HIF-1α. The results demonstrated that inhibition of HIF-1α significantly reduced the levels of SDF-1α and VEGF secretion (Fig. 4C and D), and the overexpression of HIF-1α can significantly increase the secretion levels of SDF-1α and VEGF (Fig. S3B and S3C.)

Fig. 4.

Immunomodulatory effect of PACG/CS hydrogels. (A) Schematic diagram of the effects of different hydrogels on RAW264.7 cells. (B) The scavenging effect of PACG/CS hydrogels on reactive oxygen species (ROS) after chelating iron ions. ROS were detected by the DCFH-DA probein the ROS assay kit under fluorescence microscope. (C, D) The levels of SDF-1α and VEGF secreted by RAW264.7 cells in the collected culture medium were measured using ELISA (n = 5). (E, F) mRNA expression of M1 and M2 macrophage markers in Raw 264.7 cells was analyzed by qRT-PCR under various conditions on days 4 and 7 (n = 5). (i) Control group, (ii) LPS group, (iii) LPS + PACG group, (iv) LPS + PACG/CS-1.6 group, (v) LPS + PACG/CS-3.9 group, and (vi) LPS + PACG/CS-1.6 group + Oltipraz. ∗∗p < 0.01 statistical significance compared with Control group; -p < 0.05, --p < 0.01 statistical significance compared with LPS group; ^#p < 0.05, ^##p < 0.01 statistical significance compared with PACG group; +p < 0.05, ++p < 0.01 statistical significance compared with PACG/CS-1.6 group; ^p < 0.05, ^^p < 0.01 statistical significance compared with PACG/CS-3.9 group.

QPCR was used to analyze the proinflammatory M1 markers including cyclooxygenase-2 (Cox-2), inducible nitric oxide synthase (iNOS), and TNF-α, as well as the anti-inflammatory M2 markers including arginase 1 (Arg-1) and IL-10. The expression of the proinflammatory M1 markers (Cox-2, iNOS and TNF-α) was significantly decreased in the co-culturing RAW264.7 cells with PACG and PACG/CS hydrogels compared to the LPS group. In contrast, the expression of anti-inflammatory M2 markers (Arg-1 and IL-10) was significantly increased under the same conditions. In the PACG/CS-1.6-Oltipraz group, the expression of proinflammatory M1 markers (Cox-2, iNOS and TNF-α) was significantly increased, while the expression of anti-inflammatory M2 markers (Arg-1 and IL-10) was significantly decreased compared to the PACG/CS-1.6 group (Fig. 4E and F). Whereas the PACG/CS-1.6-CoCl₂ group exhibited the opposite pattern (decreased M1 markers; increased M2 markers) (Fig. S3D and S3E). These results indicate that PACG and PACG/CS hydrogels could efficiently induce a hypoxic microenvironment by chelating iron ions, thereby activating HIF-1α. Subsequently, HIF-1α effectively regulated the secretion of SDF-1α and VEGF and the polarization of macrophages from a proinflammatory M1 phenotype to a pro-regenerative M2 phenotype, thereby benefiting faster resolution of inflammation and tissue regeneration.

Through dynamic monitoring (6 h–10 days) in an LPS-induced RAW 264.7 cell inflammation model, the modulatory capacity of the PACS/CS-1.6 hydrogel has been systematically assessed on both HIF-1α (Fig. S4) and 18 key inflammatory factors (Fig. S5). Dynamic monitoring revealed that the hydrogel stabilizes HIF-1α expression in RAW 264.7 cells (Fig. S4). Under LPS-induced inflammatory conditions, the hydrogel creates a hypoxic microenvironment while exerting anti-inflammatory and antioxidant effects, resulting in a dual regulatory pattern: key pro-inflammatory mediators like TNF-α and IL-1β are significantly suppressed, while tissue repair factors such as SDF-1α/CXCL12 and IL-10 are synergistically enhanced via HIF-1α activation. Dynamic factors exhibit time-dependent modulation governed by microenvironmental interplay– MCP-1/CCL2 shows an initial transient rise due to early hypoxia but is later suppressed, whereas Fractalkine/CX3CL1 expression increases to bolster anti-inflammation. Collectively, this suppresses the inflammatory storm (TNF-α/IL-1β) and activates repair programs (CXCL12/IL-10), shifting the balance from destructive inflammation towards tissue regeneration (Fig. S5).

2.4. Effect of PACG and PACG/CS hydrogels on the recruitment and chondrogenic differentiation of BMSCs and PBMSCs

The improvement of the immune microenvironment by tissue engineering implants would promote the recruitment of endogenous stem cells, thereby fostering regeneration and the development of TEM with structural and functional heterogeneity in vivo. To further investigate whether PACG and PACG/CS hydrogels stimulate RAW264.7-mediated microenvironment changes and promote BMSCs or PBMSCs migration, we initially applied a layer of hydrogel to the bottom surface of culture dishes. Subsequently, RAW264.7 cells were seeded on PACG and PACG/CS hydrogels, while BMSCs or PBMSCs were seeded in the transwell chambers. PACG and PACG/CS hydrogels significantly enhanced the recruitment of BMSCs or PBMSCs in vitro. Compared with Blank, GelMA, PACG, and PACG/CS-3.9 hydrogels groups, the recruitment of BMSCs and PBMSCs in PACG/CS-1.6 hydrogel group was the highest. The PBMSCs collected in PACG, PACG/CS-1.6, and PACG/CS-3.9 groups were significantly higher than those in BMSCs. In the Blank and GelMA groups, a small number of BMSCs and PBMSCs were recruited, and there was no significant difference in the number of BMSCs and PBMSCs recruited (Fig. 5A and B). Preliminary experiments indicated that the hypoxic microenvironment induced by PACG/CS-1.6 hydrogel was more favorable for the recruitment of MSCs. Compared to BMSCs, PBMSCs was more easily induced and recruited by the hypoxic microenvironment. The mRNA expression of Sox9 in BMSCs and PBMSCs, a key regulator of chondrogenic differentiation, was analyzed by qRT-PCR after 3 days culture on the hydrogels. The results showed a significant increase in Sox9 mRNA expression in the PACG and PACG/CS hydrogel groups compared to the blank group. Among these, the PACG/CS-1.6 hydrogel group exhibited the highest Sox9 mRNA expression compared to the blank group, PACG group, and PACG/CS-3.9 hydrogel group. Additionally, the Sox9 mRNA expression in PBMSCs was significantly higher in the PACG, PACG/CS-1.6, and PACG/CS-3.9 groups than in BMSCs (Fig. 5C).

Fig. 5.

Effect of PACG and PACG/CS hydrogels on the recruitment and chondrogenic differentiation of BMSCs and PBMSCs. (A) The indirect migration effects of BMSCs and PBMSCs induced by PACG and PACG/CS hydrogels via co-culture with RAW264.7 cells in vitro. Representative images showing the number of BMSCs and PBMSCs that penetrated the well membrane were presented. (B) Quantification of the number of migrated MSCs (n = 6). ∗p < 0.05 between the BMSCs migration and PBMSCs migration in the same group; ^#p < 0.05, ^##p < 0.01 between the PACG/CS-1.6 group and other groups in the same kind of MSCs. (C) The mRNA expression of Sox9 in BMSCs and PBMSCs was detected by qRT-PCR after 3 days culture on the hydrogels (n = 6). ∗p < 0.05 between the BMSCs migration and PBMSCs migration in the same group; ^#p < 0.05, ^##p < 0.01 between the PACG/CS-1.6 group and other groups in the same kind of MSCs. (D)The glycosaminoglycans (GAGs) contents in MSCs after 3 weeks chondrogenic induction (n = 6). ∗p < 0.05 between the BMSCs migration and PBMSCs migration in the same group; ^#P < 0.05, ^##P<0.01 between the PACG/CS-1.6 group and other groups in the same kind of MSCs. (E) Safranin-O staining of BMSCs-laden hydrogels and PBMSCs-laden hydrogels after 3 weeks chondrogenic induction. (F) The relative analysis of Safranin-O staining positive area (n = 6). ∗p < 0.05 between the BMSCs Safranin-O positive area and PBMSCs Safranin-O positive area in the same group; ^##p < 0.01 between the PACG/CS-1.6 group and other groups in the same kind of MSCs. (G) Immunofluorescence of the expression of COL-2 and COL-X of BMSCs-laden hydrogels and PBMSCs-laden hydrogels after 3 weeks chondrogenic induction. (H) The relative analysis of COL-2 positive area. (I) The relative analysis of COL-X positive area. ∗p < 0.05 between the BMSCs COL-2 positive area and PBMSCs COL-2 positive area in the same group; ^##P < 0.01 between the PACG/CS-1.6 group and other groups in the same kind of MSCs (n = 6).

To investigate the cartilage differentiation potential of BMSCs and PBMSCs cultured in hydrogels over an extended period, BMSCs and PBMSCs were encapsulated in different hydrogel, respectively. Chondrogenesis could be verified by assessing the intensity of glycosaminoglycans (GAGs), proteoglycan (PG) and collagen-2. After three weeks of induction, BMSCs in the PACG/CS-1.6 hydrogel group exhibited higher expression of GAGs and PG, two chondrogenesis markers, compared to those in the PACG and PACG/CS-3.9 hydrogel groups. PBMSCs in the PACG/CS-1.6 hydrogel group exhibited higher GAG expression compared to those in the PACG and PACG/CS-3.9 hydrogel groups (Fig. 5D). PBMSCs in the PACG/CS-1.6 hydrogel group showed higher PG expression than those in the PACG hydrogel group (Fig. 5E and F). Additionally, PBMSCs in the both PACG and PACG/CS hydrogel groups demonstrated higher GAGs and PG expression than BMSCs in the same groups. Histological observations and quantitative analyses revealed that chondrocytes in the PACG/CS groups formed a superior collagen type II matrix compared to those in the PACG group after three weeks of induction. BMSCs in the PACG/CS-1.6 hydrogel group showed greater collagen type II matrix expression compared to those in the PACG and PACG/CS-3.9 hydrogel groups (Fig. 5G and H). Similarly, PBMSCs in the PACG/CS-1.6 hydrogel group exhibited higher levels of collagen type II matrix than those in the PACG hydrogel group. Moreover, PBMSCs in both the PACG and PACG/CS hydrogel groups had higher collagen type II matrix expression than BMSCs in the corresponding hydrogel groups. However, collagen X expression was minimal across all groups, with no statistically significant differences observed between all groups (Fig. 5G and I and S6). These results suggested that the composition of hydrogel could significantly promote the chondrogenic differentiation of MSCs, and the addition of 1.6 % concentration of chitosan to PACG hydrogels has a positive effect on the chondrogenic formation of MSCs. In addition, the hypoxia-induced microenvironment provided by PACG/CS hydrogels may offer greater chondrogenic differentiation potential for PBMSCs compared to bone marrow mesenchymal stem cells.

2.5. Anisotropic construct for meniscal regeneration in vivo

Using Mimics software, the rabbit meniscus was three-dimensionally reconstructed (Fig. 6A–a). Subsequently, a high-transparency polycarbonate (PC) meniscus injection mold (Fig. 6A–b and 6A-c) was fabricated through machining and electrochemical etching processes, resulting in a novel composite PACG/CS-1.6 hydrogel meniscus (Fig. 6A–d). The efficacy of meniscus regeneration in vivo was assessed by transplanting the personalized PACG/CS hydrogel meniscus constructs into the knee joints of rabbits. No cells seeded into PACG/CS-1.6 hydrogels group (PACG/CS-1.6 treatment), BMSCs seeded into PACG/CS-1.6 hydrogels group (PACG/CS-1.6+BMSCs treatment), PBMSCs seeded into PACG/CS-1.6 hydrogels group (PACG/CS-1.6+PBMSCs treatment), no-transplanting constructs group(blank), and native meniscus group (native) groups were chosen for in vivo evaluation.

Fig. 6.

Regeneration of PACG/CS-1.6 hydrogel meniscus resembling the native tissue after 24 weeks in vivo. (A) Construction of PACG/CS-1.6 hydrogel meniscus. (a) Mimics 3D reconstruction rabbit meniscus. (b) Design of injection mold. (c) High-transparency polycarbonate (PC) meniscus injection mold. (d) The preparation of rabbit PACG/CS-1.6 hydrogel meniscus using high-transparency PC meniscus injection mold. (B) Gross view of native or regenerated menisci at 24 weeks after in vivo implantation in rabbit knees. The yellow dotted line demarcates the synovial tissue-regenerated meniscus interface. (C, D) Zone-specific matrix phenotype analysis in engineered versus native tissue. Tissue sections were stained by immunohistochemistry for COL-2 or with toluidine blue (TB) for proteoglycans and picrosirius red (PR) for COL-1 and COL-3. Fibroblast-like and chondrocyte-like cells were both observed in the implants in the PACG/CS-1.6+PBMSCs group and native group. (E) H&E staining demonstrates zone-specific cellular organization within the regenerated meniscal tissue, with representative images derived from six biological replicates analyzed in a single construct. (F–G) Quantitative evaluation of biomechanical characteristics at 24 weeks post-implantation (n = 6). ∗p < 0.05 statistical significance compared with Native group; ∗∗p < 0.01 statistical significance compared with Native group; #p < 0.05 statistical significance compared with PACG/CS-1.6 group; +p < 0.05 statistical significance compared with PACG/CS-1.6+BMSCs group). (H, I) Biomechanical properties for different group in bidirectional tensile testing in different locations. (H) Circumferential tensile modulus (n = 6). +p < 0.05 between the inner region and outer region in the same group; ∗∗p < 0.01 between the PACG/CS-1.6, PACG/CS-1.6+BMSCs or PACG/CS-1.6+PBMSCs group and native group in the inner zone; ^#p < 0.05, ^##p < 0.01 between the PACG/CS-1.6, PACG/CS-1.6+BMSCs or PACG/CS-1.6+PBMSCs group and native group in the outer zone. (I) Radial tensile modulus (n = 6). ∗p < 0.05, ∗∗p < 0.01 statistical significance compared with Native group; #p < 0.05 statistical significance compared with PACG/CS-1.6 group; +p < 0.05 statistical significance compared with PACG/CS-1.6+BMSCs group).

Joint specimens incorporating neoformed tissues were harvested at 24-week post-implantation for gross observations and histological analyses. Synovial fluid aspiration enabled cytokine profiling (IL-6 and TNF-α) to assess immune activation. No significant inflammatory response was observed at any phase (Fig. S7). After 24 weeks, the regenerated meniscus in the PACG/CS-1.6+PBMSCs group resembled the natural rabbit meniscus and had a normal gross appearance with a shiny white color in outer synovial margin and a slightly reddish color in inner free margin, and a smooth surface. The meniscus in the PACG/CS-1.6 group and PACG/CS-1.6+BMSCs group partially regenerated, forming approximately 50 % and 60 % new meniscus tissue, respectively, with a shiny white appearance in the outer synovial margin (Fig. 6B). At 24 weeks, the histological assessment of the regenerated meniscus in the PACG/CS-1.6+PBMSCs group showed zone-specific matrix phenotypes in the regenerated meniscus, which were similar to those found in the native tissue (native group). The outer zone demonstrated a fibrous matrix with a structured alignment, primarily consisting of COL-1, while the inner zone exhibited a cartilaginous matrix enriched with COL-2 and proteoglycans (Fig. 6C and D). In contrast, the PACG/CS-1.6 group displayed a diffuse fibrous tissue, without distinct zone-specific tissue characteristics. The inner zone of the PACG/CS-1.6+PBMSCs group at 24 weeks, exhibited a dense population of round, chondrocyte-like cells encircling cartilage islands, while the outer zone was filled with fusiform, fibroblast-like cells, resembling the native meniscus structure (Fig. 6E). In contrast, the PACG/CS-1.6 group displayed an abundance of spindle-shaped fibroblast cells with elongated nuclei dispersed throughout the fibrous tissue of the entire implant. At 24 weeks, histological evaluation in the PACG/CS-1.6+BMSCs group revealed a tendency towards a heterogeneous matrix phenotype in the regenerated tissue.

The biomechanical properties of meniscus constructs were also evaluated. At 24 weeks, the PACG/CS-1.6-PBMSCs group demonstrated restored tensile and aggregate modulus comparable to native meniscus (Fig. 6F and G). Spatial heterogeneity was quantified through tensile testing in the radial and circumferential directions and localized compression measurements across distinct zones of the 24-week in vivo tissue-engineered meniscus (Fig. 6H and I). Following 24 weeks of in vivo maturation, PACG/CS-1.6-PBMSCs meniscal constructs exhibited region-specific mechanical gradients (inner versus outer zones) in circumferential tensile modulus, reduced modulus, and hardness, with spatial patterns mirroring those of native meniscal tissue. Furthermore, the PBMSCs- and BMSCs-enhanced constructs demonstrated the significantly elevated biomechanical metrics compared to the PACG/CS-1.6 group under multidirectional tensile and compressive loading conditions. These cumulative findings substantiate that PBMSCs-laden PACG/CS-1.6 constructs successfully re-established meniscal functional anisotropy and biomechanical characteristics at the 24-week post-implantation.

2.6. Gross and microscopic evaluation of the joints 24 weeks post-implantation

According to the International Cartilage Repair Society (ICRS) scoring system, at 24 weeks post-implantation, the PACG/CS + PBMSCs group exhibited a smooth cartilage surface (Fig. 7A and D). In contrast, the Blank group showed the most severe cartilage damage, which was visibly evident. Histological evaluations using the Mankin scoring system (Fig. 7B, C and 7E) revealed that at 24 weeks, the Blank group and the PACG/CS-1.6 hydrogels group exhibited significantly diminished intensity of toluidine blue (TB) and Safranin O/Fast Green (SO/FG) staining in the femoral condyle (FC) and tibial plateau (TP). Conversely, the PACG/CS + PBMSCs group showed robust TB and SO/FG staining, indicating a well-preserved cartilage surface with all structural zones intact. Cartilage damage progressively worsened over time, with the Blank group consistently showing a higher Mankin score than the other groups at all evaluated time points. This indicates that the other groups exhibited varying degrees of cartilage protective effects in comparison to the Blank group. At 24 weeks, the Blank group and PACG/CS-1.6 hydrogels group exhibited more extensive cartilage degeneration compared to the PACG/CS + PBMSCs treatment. The femoral condyles and tibial plateaus showed distinct patterns of damage: the femoral condyles experienced significant proteoglycan loss accompanied by surface fibrillation, while the tibial plateaus displayed less pronounced proteoglycan depletion but more severe surface structural destruction. Significant differences in the ICRS and Mankin scores were noted between the PACG/CS + BMSCs group and the PACG/CS + PBMSCs group, with the PACG/CS + BMSCs group exhibiting higher scores. At 24 weeks post-treatment, knee joints in the PACG/CS + PBMSCs group demonstrated minimal signs of degeneration and maintained cartilage integrity. Consequently, both gross observations and histological evaluations indicated that the regenerated meniscus in the PACG/CS + PBMSCs group provided a superior chondroprotective effect compared to the PACG/CS + BMSCs group.

Fig. 7.

Gross and microscopic evaluation of the joints 24 weeks post-implantation. (A) Macroscopic examination of the femoral condyle (FC) and tibial plateau (TP). (B) H&E, (C) TB staining, and SO/FG staining of the articular cartilage surfaces in the FC and TP. (D) ICRS scores and (E) Mankin scores for the assessment of cartilage degeneration (n = 6). ∗p < 0.05 statistical significance compared with Native group; -p < 0.05 statistical significance compared with Blank group; #p < 0.05 statistical significance compared with PACG/CS-1.6 group; +p < 0.05 statistical significance compared with PACG/CS-1.6+BMSCs group).

2.7. Analysis of transcriptome after PACG/CS hydrogel meniscus implantation in early stage in vivo

To gain comprehensive insight into the specific impact of composite hydrogels on the early in vivo microenvironment, we performed mRNA-seq analysis to assess gene expression in tissues adjacent to the blank group and the PACG/CS + PBMSCs group (Hybrid 2 group). Fig. 8A showed the 395 differential genes (DEGs) of the Hybrid 2 group based on the blank group (205 up-regulated DEGs and 190 down-regulated DEGs, FG ≥ 1.2, p ≤ 0.05). The differentially expressed genes (DEGs) in each group were visualized as a heatmap (Fig. 8B). KEGG pathway enrichment analysis of the up-regulated DEGs highlighted the activation of key pathways such as Focal adhesion, ECM-receptor interaction, and Protein digestion and absorption, which are crucial for biological processes and cartilage tissue regeneration (Fig. 8C). Conversely, the KEGG analysis of down-regulated DEGs predominantly implicated pathways associated with Primary Immunodeficiency, NF-κB Signaling Pathway, and Rheumatoid Arthritis (Fig. 8D), indicating a suppression of immune responses and inflammatory processes. GO analysis of the up-regulated DEGs revealed an enhancement in immune response functions, including chondrocyte differentiation, regulation of ion transmembrane transport, and skeletal muscle fiber development (Fig. 8E). This suggested a potential enhancement in tissue-specific differentiation and homeostatic maintenance. In contrast, the GO analysis of down-regulated DEGs indicated the alterations in functions related to inflammatory responses, T cell differentiation, and B cell activation processes (Fig. 8F), highlighting a shift towards anti-inflammatory or reparative mechanisms. Gene Set Enrichment Analysis (GSEA) demonstrated an up-regulation of the HIF-1 signaling pathway (Fig. 8G), while it showed a down-regulation of the NF-κB signaling pathway (Fig. 8H). These results pointed towards the anti-inflammatory efficacy of PACG/CS hydrogels and PBMSCs. Research has indicated that an enduring pro-inflammatory response in the initial phase of the healing process is typically correlated with impaired healing, while the subsequent activation of an anti-inflammatory signaling cascade is deemed essential for achieving a favorable healing outcome. The use of PACG/CS hydrogels in conjunction with PBMSCs may significantly improve meniscus healing after injury by enhancing the HIF-1 signaling pathway and regulating inflammatory responses within the microenvironment.

Fig. 8.

Peri-hydrogel tissue was collected after 7 days for RNA-seq analysis. (A) Upregulated and downregulated DEGS Volcano plot. (B) Upregulated and downregulated DEGS enrichment heatmap. (C) Upregulated DEGS KEGG enrichment. (D) Downregulated DEGS KEGG enrichment. (E) Upregulated DEPS GO enrichment. (F) Downregulated DEPS GO enrichment. (G) GESA for HIF-1 signaling pathway. (H) GESA for NF-κB signaling pathway.

3. Discussion

Research on TEM includes various approaches such as cell-scaffold composite methods, cytokine-induced regulation, extracellular matrix-induced modulation, and the construction of heterogeneity across different TEM regions. However, no TEM has yet been successfully cultivated that closely replicates the structure, function, and regional heterogeneity of the natural meniscus [[24], [25], [26]]. The reason may be that TEM strategies cannot accurately simulate the heterogeneous structure-characteristic relationship of meniscus, and researchers have ignored the important role of anti-inflammatory and antioxidant microenvironment in regulating TEM growth and cultivation. Recognizing the critical function of HIF-1ɑ in governing chondrocyte activity and activating downstream pathways for chondrocyte regeneration, we designed a PACG/CS hydrogel to locally establish a hypoxia-simulating niche. The selection of PACG and CS as primary hydrogel components was driven by their synergistic capacity to establish a biomimetic and mechanically stable microenvironment. PACG provides essential covalent cross-linking for structural integrity and sustained Fe³⁺ chelation by the powerful carboxyl-Fe³⁺ interaction, thus enabling the persistent hypoxia-inducible capacity in situ. In addition, CS as a native cartilage glycosaminoglycan actively can not only weaken harsh acidic environments by the carboxyl-abundant PACG, but also improve the cell growth and proliferation for facilitating chondrogenesis. Together, they create an optimal niche where PACG-mediated bioactivity synergizes with Fe³⁺ chelation-induced hypoxia, while tunable mechanics the CS component ensures the physiological relevance—advantages not achievable with conventional single-component hydrogel scaffolds. This strategy not only improved meniscus regeneration compared to traditional TEM but also addressed the limitations associated with external factors by accelerating the regeneration process in a more controlled manner. To the best of our knowledge, this is the first study to utilize a pure biomaterial with a hypoxia-mimicking design by high efficiency iron chelation for meniscus regeneration, which presents a promising alternative strategy and introduces a novel biomaterial design for advancing meniscus regeneration.

HIF-1α is well-recognized as a crucial regulator of gene expression involved in skeletal lineage specification [27,28]. While significant interactions between cellular dynamics and differentiation have been established, the effective stabilization of HIF-1α and the optimal duration of its expression remain inadequately understood and are particularly important for meniscus regeneration. A hypoxia-responsive hydrogel has been demonstrated to stimulate the HIF signaling pathway through oxygen consumption in vivo [29]. Nevertheless, controlling in vivo oxygen levels is challenging, and the inevitable fluctuations in oxygen concentration can lead to various pathophysiological conditions [30]. HIF-1α promotes M1 or M2 macrophage polarization in a context-dependent manner. Within the tumor microenvironment, it could be asserted the two opposing roles for HIF-1α in promotion of macrophage polarization in which succinate-mediated HIF-1α stabilization and further stimulation of glycolysis favors M1 polarization, whereas high lactate-mediated HIF-1α stabilization promotes M2 polarization. HIF-1α also promotes M2 polarization by inducing Zinc finger E-box binding homeobox 1 (Zeb1), which enhances glycolysis and lactate release in cancer cells [31,32]. In this study, the CS-incorporated PACG hydrogel was designed to activate the HIF-1α pathway using a hypoxia-mimicking strategy through in situ chelation of iron ions. In addition, a high-transparency PC meniscus injection mold therapy provided a customized design for complex meniscus regeneration. Post-crosslinking, PACG/CS hydrogels demonstrated tunable biomechanical profiles (Fig. 2A and B) while exhibiting pro-adhesive surfaces conducive to cellular colonization (Fig. 2E and F) and further MSCs recruitment and chondrogenic differentiation (Fig. 5). More specifically, SEM analysis revealed that pure PACG hydrogel featured compact and isotropic pore structures. With chitosan addition, a “horizontal-vertical” orientation of pores emerged, which transformed into larger irregular pores as chitosan content increased, with PACG/CS-1.6 exhibiting the most regular pore structure (Fig. 2D). The relationship between hydrogel porous features and stem cell chondrogenesis is nonlinear, requiring precise tuning within a specific optimal range to maximize cellular response. Typically, chondrogenesis is enhanced by high pore uniformity, interconnectivity, and carefully optimized pore size and porosity [33,34]. In terms of hypoxia-mimicking capacity, PACG/CS-1.6 and PACG/CS-3.9 hydrogels showed an 8-fold and 7-fold increase in iron chelation ability, respectively, compared with PACG hydrogel (Fig. 3B and C). This effectively localized the elimination of iron ions, leading to sustained activation of HIF-1α signaling axis and enhanced ROS scavenging. Cellular and molecular analyses demonstrated that PACG/CS hydrogels, particularly PACG/CS-1.6, significantly enhanced the recruitment of tissue-repairing cells by activating the SDF-1α/CXCR4 chemokine pathway (Fig. 4, Fig. 5A). Additionally, these hydrogels promoted MSC differentiation into chondrocytes. Based on the in vitro chondrogenic differentiation assessments, PACG/CS-encapsulated MSCs exhibited an enhanced secretion of extracellular matrix components (collagen II, GAGs, and PG), compared to MSCs encapsulated solely in the PACG hydrogel after three weeks of induction (Fig. 5D–H). Moreover, PBMSCs in both the PACG and PACG/CS hydrogel groups had higher collagen type II matrix expression than BMSCs in the corresponding hydrogel groups (Fig. 5G and H). Consequently, PACG/CS hydrogels, particularly PACG/CS-1.6, possess unique mechanical properties that enable stable HIF-1α expression through iron chelation, effectively influencing the behaviors of downstream cells, such as RAW264.7 cells and MSCs, and promoting meniscus regeneration.

The future of knee meniscus treatment lies in precision, intelligence, and personalization [35,36]. Once an ideal implant meets all necessary criteria, biomaterials can facilitate the regeneration of healthy, functional tissue in the affected area [37,38]. Achieving ideal biomaterials requires advanced manufacturing techniques capable of precisely fabricating complex geometries, optimizing mechanical properties, enhancing cell and nutrient permeability, and promoting both angiogenesis and bone regeneration [39,40]. In clinical practice, physicians can utilize patient-specific medical imaging data, such as computed tomography (CT) or magnetic resonance imaging (MRI), to design and fabricate custom orthopedic implants tailored to the patient's unique surgical needs [37,41]. Advancements in digital technology, computer-aided design (CAD) techniques, and surgical technology have heightened interest in patient-specific implants that closely mimic the structural and functional properties of natural tissue [41,42]. In the study, we utilize Mimics software to analyze MRI data for the three-dimensional reconstruction of rabbit menisci (Fig. 6A–a). Subsequently, a high-transparency polycarbonate (PC) meniscus injection mold was produced using machining and electrochemical etching techniques (Fig. 6A–b and 6A-c). This mold was then utilized to create an innovative composite PACG/CS-1.6 hydrogel meniscus (Fig. 6A–d). The advanced fabrication techniques of the mold ensure precise replication of meniscal geometry, enhancing the implant's integration and functionality within the natural knee joint environment. Meanwhile, the composite hydrogel structure is designed to deliver superior mechanical properties and effective microenvironment regulation, promoting optimal tissue regeneration.

To validate the anisotropic architecture and articular cartilage preservation capacity of neo-fibrocartilage, we performed translational validation spanning in vitro PACG/CS-1.6+MSCs-induced fibrochondrogenesis to 24-week in vivo implant functionality and tissue restoration assessments. Histological studies in vivo meniscus transplantation animal models confirmed that the regenerated meniscus in the PACG/CS-1.6+PBMSCs group showed a zone-specific matrix phenotype closely resembled those of native tissue after 24 weeks (Fig. 6B). The outer zone had a fibrous matrix with COL-1, while the inner zone had a cartilaginous matrix rich in COL-2 and proteoglycans (Fig. 6C and D). The PACG/CS-1.6 group, however, lacked these distinct zones and had a diffuse fibrous tissue (Fig. 6C and D). The PACG/CS-1.6+PBMSCs group showed chondrocyte-like cells in the inner zone and fibroblast-like cells in the outer zone, resembling native meniscus, while the PACG/CS-1.6 group had spindle-shaped fibroblast cells with elongated nuclei (Fig. 6E). At 24 weeks, the PACG/CS-1.6+PBMSCs group consistently demonstrated a yield of fibrochondrocytes exhibiting zone-specific matrix phenotypes within the regenerated tissue (Fig. 7). The PACG/CS constructs demonstrated compromised biomechanical performance and suboptimal histological grading (ICRS/Mankin), suggesting insufficient extracellular matrix biosynthesis with structural irregularities impaired load-bearing capacity critical for articular surface preservation. These results provided the first evidence that PACG/CG hydrogel combined with PBMSCs could initiate meniscus regeneration. Previous studies had shown that hypoxia would promote the proliferation of PBMSCs [43]. Consistent with our findings, PBMSCs exhibit biological characteristics similar to those of BM-MSCs. However, while BM-MSCs show stronger osteogenic and adipogenic differentiation, PBMSCs demonstrate greater chondrogenic potential [44]. Notably, the survival of allogeneic PBMSCs in the regenerated cartilage was confirmed through the detection of the rabbit sex-determining region Y-linked gene sequence, revealing that these transplanted PBMSCs survived for up to 3 months to promote the cartilage repair [45]. Although meniscus regeneration appears to be a complex process involving multiple biological and mechanical factors, the concept of a hypoxia-mimicking biomaterial-based model holds significant promise. This innovative approach could not only facilitate the initial stages of meniscus regeneration but also contribute to achieving effective and sustained regeneration in the long term. By simulating the low-oxygen environment typically found in injured tissues, this model could optimize the integration of transplanted PBMSCs and biomaterials with the native tissue. Meniscus regeneration typically involves three key phases: 1) immune-primed MSC recruitment initiates neo-tissue formation; 2) chondroprogenitor lineage specification establishing a cartilaginous anlagen foundation. 3) later-stage ingrowth of new meniscal tissue. The advancements in meniscus regeneration via hypoxia-mimetic hydrogel can be explained through the following molecular pathways.

The hypoxia-mimicking environment induced by the PACG/CS hydrogel significantly promoted the recruitment of MSCs and the development of chondrocytes in the initial phase of meniscus regeneration. In one respect, mimicking hypoxic microenvironment facilitated effective immunoregulation by inducing HIF-1ɑ and scavenging ROS. It is well-established that the migration of MSCs from distant sites to injured tissues, along with their interaction with the inflammatory microenvironment, plays a key role in tissue repair [[46], [47], [48], [49]]. In line with previous studies, our findings demonstrated that PACG/CS-induced HIF-1α, along with reduced ROS levels, effectively modulated immune responses by enhancing MSC migration through the CXCR4/SDF-1 pathway and shifting macrophage polarization from the pro-inflammatory M1 phenotype to the pro-regenerative M2 phenotype. In another respect, by affecting HIF-1α expression with the HIF-1α inhibitor Oltipraz and activator CoCl₂, we further confirmed that the PACG/CS hydrogel induced MSCs to differentiate into chondrocyte-like cells through activation of the HIF-1 pathway and upregulation of chondrocyte markers, including SOX9, COL-2, and VCAN.

This study integrates personalized meniscus scaffold technology with anti-inflammatory and antioxidant microenvironmental regulation to develop a multifunctional meniscus scaffold characterized by a heterogeneous biomimetic structure, high strength and toughness, as well as the ability to induce a simulated hypoxic microenvironment with anti-inflammatory and antioxidant effects. The tough hydrogel further facilitates in situ meniscus regeneration by promoting several key processes, including stem cell recruitment, promotion of cartilage differentiation, activation of the HIF-1α signaling pathway under hypoxic conditions induced by efficient iron chelation, inhibition of inflammatory responses, and regulation of the balanced expression of oxidative generation and elimination. By demonstrating that manipulating the physical and chemical properties of materials can yield optimal therapeutic outcomes, this work presents a novel strategy for clinical applications.

4. Conclusion

In summary, we have successfully engineered an inherently bioactive iron-chelating PACG/CS hydrogel scaffold with dual hypoxic-immune microenvironment regulatory capabilities, establishing a biomimetic platform for functional meniscus regeneration. By integrating advanced mold manufacturing technology, the hydrogel precisely replicates native meniscal geometry while providing robust mechanical support, ensuring seamless integration and biomechanical compatibility within the knee joint. The scaffold's distinctive iron-chelating capacity establishes a dual regulatory system through coordinated biological pathways: controlled sequestration of iron ions induces sustained HIF-1α stabilization, generating a hypoxia-mimetic microenvironment that specifically potentiates the chondrogenic differentiation advantage of PBMSCs over BMSCs; while simultaneously modulating macrophage polarization to reshape immune microenvironment, thereby synergistically facilitating cartilage formation. This synergistic regulation of hypoxic signaling and immune responses significantly amplifies extracellular matrix synthesis and cellular recruitment, effectively bridging the gap between structural reconstruction and biological functionality. Our findings not only validate HIF-1α as a pivotal mediator in meniscal repair but also establish a strategic framework for designing microenvironment-engineered biomaterials that simultaneously address mechanical demands and biological signaling in complex tissue regeneration. This dual-targeting strategy opens new avenues for developing intelligent meniscus implants with combined immunomodulatory and hypoxic preconditioning capabilities.

5. Experimental section

5.1. Cell harvesting, culture, and seeding

Bone marrow- and peripheral blood-derived MSCs (BMSCs/PBMSCs) were isolated from New Zealand white rabbits. Cell isolation, expansion, trilineage differentiation assays, and immunophenotyping followed established protocols [43], with third-passage cells used for hydrogel seeding.

RAW 264.7 macrophages (ATCC) were maintained in high-glucose DMEM supplemented with 10 % FBS under standard culture conditions (37 °C, 5 % CO₂).

5.2. Preparation of personalized hydrogel meniscus

The personalized extraction of meniscus was realized by obtaining the relevant parameters of meniscus accurately with magnetic resonance imaging (MRI). Using Mimics software, the rabbit meniscus was three-dimensionally reconstructed. Subsequently, a high-transparency polycarbonate (PC) meniscus injection mold was fabricated through machining and electrochemical etching processes, resulting in a novel PACG/CS-1.6 hydrogel meniscus.

5.3. Preparation of a novel compositehydrogel (PACG/CS) with high strength and high toughness

5.3.1. a. Synthesis of N-acryloyl glycine (ACG)

Glycine was added to a 1000 mL double-mouth bottle and dissolved in 480 mL of distilled water. Glycine solution was cooled through ice bath to 0 °C before NaOH was added. Then acryloyl chloride was slowly added to the glycine solution, and the reaction was maintained at pH 7.5–7.8 with NaOH solution at room temperature for 8 h. After the reaction, the reaction solution was extracted with ethyl acetate and the aqueous phase was retained. The pH of the reaction solution was adjusted at 2.0 by hydrochloric acid solution. Then the reaction solution was extracted with a large amount of ethyl acetate, dried with anhydrous magnesium sulfate, filtered, and then concentrated by spin evaporation, precipitated with petroleum ether, and dried in vacuum.

5.3.2. b. Preparation of PACG/CS composite hydrogels

Different concentrations of ACG monomer, CS, cross-linking agent and photoinitiator Irgacure 2959 were added to H₂O and dissolved by circular oscillator. The clear and transparent composite solution was illuminated by UV for 2 h to obtain the semi-interpenetrating network hydrogel.

5.4. Detection of Fe³⁺ ion chelating ability of PACG/CS hydrogels

PACG/CS composite hydrogels with different mass fractions were immersed in salt solution containing high-valence Fe³⁺ ions. Based on the strong ion complexation between carboxyl groups and Fe³⁺ ions in PACG polymer chains, a core-shell structure double-network hydrogel with gradient ion cross-links was constructed. The chelation ability of Fe³⁺ ion was detected by iron colorimetry detection kit. Fe³⁺ binds to the protein to form a complex, which was released from the complex in an acidic medium and then reduced to Fe²⁺ by a reducing agent and combined with ferrozine to form a purplish red compound. The absorbance was measured at a wavelength of 540–580 nm.

The 3 mM standard stock solution was serially diluted using the kit-supplied dilution buffer to generate a concentration gradient ranging from 300 μM to 4.69 μM, with a zero-concentration control included. Mix A was prepared by combining the buffer solution with 4.5 % potassium permanganate at a 1:1 volumetric ratio. Blank controls, standard series, and test samples were then aliquoted into designated reaction tubes. The mixtures were incubated at 60 °C for 1 h, followed by cooling to room temperature and centrifugation to consolidate residual droplets. Subsequently, 30 μL of iron detection reagent was added to each tube, vortex-mixed, and incubated for 30 min at ambient temperature prior to a final centrifugation step to clarify the supernatant. Aliquots (200 μL) of the processed supernatants were transferred to a 96-well microplate for absorbance measurement at 540–580 nm. Iron concentrations were determined through standard curve interpolation.

5.5. Validation of the HIF-1α signaling pathway

To explore the expression of HIF-1α in Raw 264.7 cells, varying concentrations of PACG/CS hydrogels were added to the transwell chambers of a 6-well plate seeded with Raw 264.7 cells. After a 2-h pre-treatment, lipopolysaccharide (LPS, 100 ng/mL) was added and the cells were further treated for 12 h. The PACG/CS hydrogels continued to co-culture with the Raw 264.7 cells at the bottom of the 6-well plate for an additional 3 days. The expression levels of HIF-1α protein in Raw 264.7 cells were analyzed using Western blotting.

To investigate the expression of HIF-1α in MSCs, different PACG/CS hydrogels were added to the transwell chambers of a 6-well plate seeded with MSCs, and co-cultured for 3 days. Protein expression levels of HIF-1α in Raw 264.7 cells were analyzed using Western blotting. To further determine the role of HIF-1α in anti-inflammatory and antioxidant regulation, we interfered with the expression of HIF-1α using the inhibitors FM19G11 and Oltipraz. Under the influence of these HIF-1α inhibitors, we verified whether the inhibition of HIF-1α affects macrophage phenotype, the migration of MSCs, and the expression of SDF-1α and VEGF.

5.6. Gene expression analysis

Different PACG/CS hydrogels were added to the 6-well plates containing cultured Raw 264.7 cells and incubated for 2 h, followed by treatment with LPS (100 ng/mL) for an additional 12 h. Total RNA was harvested from samples using Trizol reagent (Invitrogen). Following RNA isolation, reverse transcription was performed, and the cDNA was analyzed via quantitative real-time PCR. Primer sequences specific for Cox-2, iNOS, TNF-α, Arg-1, IL-10 and Sox-9 are detailed in Table S1. GAPDH mRNA levels were utilized as an endogenous reference for normalization. The relative quantification of the target genes was determined employing the 2^−ΔΔCT method.

5.7. Western blot

The cells were washed twice with PBS and lysed on ice in a 1 × lysis buffer (Cell Signaling) supplemented with a cocktail of protease inhibitors (Roche), followed by sonication. Subsequently, the cellular proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on gels (Fisher) based on their molecular weight. After electrophoresis, the proteins were transferred onto a polyvinylidene fluoride (PVDF) membrane and stained with specific antibodies directed against the proteins of interest (all sourced from Abcam).

5.8. The detection of antioxidant capacity

Raw 264.7 cells were pre-treated with various PACG/CS hydrogels for 2 h, followed by the addition of LPS (100 ng/mL) for an additional 4 h. Subsequently, the production of ROS was assessed using an ROS detection assay kit (Beyotime).

5.9. Inflammation-mediated migration of PBMSCs and BMSCs

The expression of SDF-1α and VEGF in RAW264.7 cells was detected using ELISA kits (Cusabio). The migratory behavior of PBMSCs is mediated by co-cultured RAW264.7 cells. Different hydrogels were synthesized across the base surface of a 24-well culture plate. RAW264.7 cells were seeded at the bottom of the wells, and PBMSCs and BMSCs were respectively inoculated into the transwell chambers for 24 h. The migration of PBMSCs and BMSCs was stained with crystal violet and observed.

5.10. Hydrogel-induced chondrogenic differentiation of PBMSCs and BMSCs

To study the chondrogenic matrix formation of PBMSCs and BMSCs, a 3D culture system was employed for PBMSCs and BMSCs seeded within hydrogels. Chondrogenic differentiation medium (Cyagen) was replaced every 2–3 days, and the culture was maintained for two weeks. After 21 days of in vitro 3D culture, the samples were subjected to histological processing. The constructs were fixed in 4 % formalin for 24 h, embedded in OCT, and processed using standard histological procedures. Tissue sections were stained with Safranin O, Collagen II, and Collagen X. The glycosaminoglycan (GAG) content was determined using the Dimethylmethylene Blue (DMMB) assay.

5.11. Immunofluorescence analyses

Hydrogel samples underwent blocking with 5 % BSA/PBST (37 °C, 1h) prior to sequential immunostaining: overnight incubation (4 °C) with anti-COL-X (ab260040, Abcam) and anti-COL-2 (ab34712, Abcam) primary antibodies (1:300), followed by 1h room-temperature exposure to species-matched secondary antibodies (AF488 anti-rabbit, Invitrogen). Nuclei were counterstained with DAPI after PBST washes.

5.12. Surgical procedure

Implantation surgery was conducted following established protocols [50]. Sixty-nine skeletally mature female New Zealand White rabbits, aged 6 months and weighing approximately 3.0 kg, were randomly assigned to experimental groups for animal studies. The study employed a double-blind protocol, ensuring that the surgical team and those responsible for postoperative care were unaware of the treatment assignments. All animals underwent a medial total meniscectomy on the left knee, and were subsequently randomized into five distinct experimental groups. 15 knees received implantation of PACG/CS hydrogel meniscus without cells (PACG/CS group), 15 knees received implantation of PACG/CS hydrogel meniscus with BMSCs (PACG/CS + BMSCs group) and 15 knees received implantation of PACG/CS hydrogel meniscus with PBMSCs (PACG/CS + PBMSCs group), 15 knees received implantation of native menisci (native group), 9 knees received no implantation (blank group). A subset of three rabbits from each experimental group was randomly selected for euthanasia at 1week post-implantation. Subsequently, three knee joints were subjected to transcriptome profiling to evaluate the early-stage in vivo response following PACG/CS hydrogel meniscus implantation. Twelve rabbits from native group, PACG/CS group, PACG/CS + BMSCs group and PACG/CS + PBMSCs group were euthanized at 24 weeks. A random selection of six knee joints was subjected to histological evaluation for the assessment of implant-induced regeneration and the progression of articular cartilage degeneration. Additionally, the biomechanical properties of the implanted constructs were evaluated in a separate cohort of six knees. Concurrently, six animals from the blank group were euthanized at 24 weeks post-implantation for histological analysis of articular cartilage degeneration.

The surgical procedure was conducted under a combined regimen of general and local anesthesia. General anesthesia was induced via intravenous administration of sodium pentobarbital solution at a dosage of 30 mg/kg. Concurrently, local anesthesia was achieved through subcutaneous injection of 1 % lidocaine hydrochloride, with a volume of 3 mL. The medial parapatellar approach was utilized to access the rabbit's knee joint. A complete medial meniscectomy was executed, with careful dissection along the meniscal periphery and detachment at the anterior and posterior horns, preserving the integrity of the medial collateral ligament. The anterior horns and periphery of the hydrogel menisci were sutured to their corresponding root attachments and adjacent synovium using absorbable No. 4-0 sutures (Ethicon). For the posterior horn, which is in proximity to the posterior cruciate ligament, a custom threading device and extracapsular knotting technique were employed to secure the hydrogel menisci to the ligamentous structures (Fig. S8). In the native group, the total medial meniscus was repaired in situ. Postoperatively, animals were housed in their home cages with free mobility. They received prophylactic antibiotics and analgesic therapy, which included opioids and nonsteroidal anti-inflammatory drugs (NSAIDs), as required. Following confirmation of unimpaired ambulation and absence of infection in the 2- to 3-week postoperative recovery period, the rabbits were permitted to engage in free-ranging activities and exercise at an agricultural setting approved by the Animal Care and Use Committee (IACUC).

5.13. Immunohistochemical analysis

To determine the expression of COL-1, COL-2, and HIF-1α in tissues, paraffin sections were deparaffinized/hydrated and antigen repaired. The slide box containing 0.01 M citrate buffer (pH 6.0) was placed in a microwave and heated to boiling before tissue retrieval. The prepared sections were then placed into the hot retrieval solution, and the microwave was set to medium power and continued to heat for 15 min. Following antigen retrieval, all sections were incubated in a 3 % hydrogen peroxide solution for 10 min. Subsequently, the sections were blocked with 5 % bovine serum albumin (BSA) for 30 min. Following this, the sections were incubated with primary antibodies specific to COL-1(goat monoclonal antibody, ARG21965, 1:200 dilution, Arigo), COL-2 (mouse monoclonal antibody, MA5-13026, 1:200 dilution, Invitrogen), and HIF-1α (rabbit monoclonal antibody, ab179483, 1:200 dilution, Abcam) overnight at 4 °C. Then, the sections containing COL-1, COL-2 and HIF-1α proteins were incubated with mouse anti-goat antibody (1:1000 dilution, Abcam), goat anti-mouse antibody (1:1000 dilution, Abcam), or goat anti-rabbit antibody (1:1000 dilution, Abcam) for 1 h at room temperature, respectively. The cell nuclei were counterstained with 50 % hematoxylin.

5.14. Mechanical analysis

The biomechanical properties of the hydrogel meniscus scaffold in vitro or regenerated meniscus were assessed using a material testing machine (AG-IS; Shimadzu). Biaxial tensile testing was implemented through sequential uniaxial loading along radial and circumferential directions in rectangular specimens (1 mm thickness) featuring standardized 3 mm gauge length within targeted tissue zones. The tensile evaluation utilized posterior segments of in vitro-engineered meniscal constructs. Circumferentially oriented samples from anatomically stratified zones (inner/outer regions) underwent displacement-controlled loading at 0.06 mm/s until structural failure, with elastic modulus derivation via linear regression analysis of stress-strain curve per established biomechanical protocols [24].

Confined compression mechanics were evaluated using cylindrical samples (2 mm diameter × 1 mm thickness) excised from the anterior segment in regenerated/hydrogel meniscal scaffolds. Constant-rate loading (0.1 mm/min) facilitated elastic modulus calculation through the linear portion of the stress-strain curve per referenced methodology [24].

5.15. RNA-seq analysis

To gain comprehensive insight into the specific impact of composite hydrogels on the early in vivo microenvironment, we performed mRNA-seq analysis at 1 week (n = 3) to assess gene expression in tissues adjacent to meniscus in each group. Rabbit tissues were sent to the laboratory of China Huada Company for RNA-seq experimental procedures. Data comparison and database construction are all completed by the company's laboratory. We used R (https://cloud.r-project.org), Dr. Tom (https://biosys.bgi.com), STRING (https://cn.string-db.org), and Cytoscape (https://github.com/cytoscape) for subsequent bioinformatics analysis.

5.16. Statistical analysis

Quantitative data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 9.0 (California, USA). Comparisons among multiple groups were assessed by one-way analysis of variance (ANOVA) with Tukey's post-hoc test for multiple comparisons. Comparisons between two groups were analyzed using two-tailed unpaired t-test. A p-value <0.05 defined statistical significance.

CRediT authorship contribution statement

Bingbing Xu: Writing – original draft, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Jing Ye: Writing – review & editing, Investigation, Formal analysis, Data curation. Shitang Song: Writing – review & editing, Methodology, Investigation. Xueyu Dou: Methodology, Data curation. Chao Li: Investigation, Data curation. Xing Wang: Writing – review & editing. Jia-Kuo Yu: Supervision, Project administration, Funding acquisition.

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Peking University Third Hospital (protocol reference #A2020128).

Declaration of competing interest

The authors declare no conflict of interest.

Appendix. ASupplementary data

The following is the Supplementary data to this article: