Co‐Treatment With rhBMP‐2 and Rapamycin Modulates Osteogenesis–Adipogenesis Balance to Enhance Aged Bone Regeneration

ABSTRACT

Recombinant human bone morphogenetic protein 2 (rhBMP‐2) is a well‐established osteoinductive agent used in clinical practice. In this study, rhBMP‐2 is found to exacerbate the imbalance between osteogenesis and adipogenesis in senescent bone marrow stromal cells (BMSCs), resulting in excess adipocytes (eADs) accumulation and a diminished osteogenic response. However, the role of eADs in age‐related bone repair deficits remains unclear. Our findings indicate that eADs within the aged microenvironment contribute to impaired bone regeneration by promoting BMSC senescence and suppressing osteogenic differentiation. To address this issue, we investigated the feasibility of regulating the abnormal differentiation of senescent BMSCs to enhance aged bone regeneration. Based on this, a novel energy‐supplying hydrogel system (PEGSN‐PGA/rhBMP‐2/Rapa, PBR) suitable for the aged regenerative microenvironment and with excellent bone integration performance is designed for local minimally invasive treatment of aged bone defects. This system effectively regulates the abnormal differentiation of senescent BMSCs, maintains the cell cycle process, and retains the regenerative potential for bone repair in the senescent microenvironment. This study presents a novel strategy for the treatment of rhBMP‐2‐mediated bone degenerative diseases and offers a pioneering perspective on the interplay among adipogenesis, cellular senescence, and bone regeneration during the aging process

This study demonstrates that the imbalance between osteogenic and adipogenic differentiation in senescent BMSCs, leading to excessive adipocyte accumulation, which subsequently impairs bone regeneration in aged mice. To address this pathological dysregulation, a novel energy‐supplying hydrogel system (PBR) has been developed to restore balanced cellular differentiation, presenting a promising therapeutic approach for age‐related bone defects and degenerative skeletal disorders.

1. Introduction

With the progression of global aging, the incidence of bone fractures attributed to age‐related osteoporosis has significantly increased [1, 2, 3]. The deterioration of the bone microenvironment and cellular function significantly impairs bone regeneration efficiency, consequently limiting the efficacy of conventional treatment methods such as autologous bone grafts, synthetic bone grafts, and internal fixation [4, 5, 6, 7]. Recombinant human bone morphogenetic protein‐2 (rhBMP‐2), which possesses the capability to recruit stem cells for proliferation and differentiation, is a commonly employed osteogenic inducer in clinical applications and serves as a therapeutic strategy to enhance the regenerative potential of stem cells [8, 9]. However, with the deterioration of the regenerative microenvironment caused by aging, such as the accumulation of reactive oxygen species (ROS) and the senescence‐associated secretory phenotype (SASP), the proliferation and differentiation capacity of bone marrow mesenchymal stem cells (BMSCs) decline with aging [4, 10, 11, 12, 13, 14]. This decline poses a significant threat to the efficacy of rhBMP‐2‐mediated bone healing [15]. Therefore, it is imperative to investigate the underlying mechanisms of rhBMP‐2‐induced bone repair failure within the aged microenvironment and to propose promising therapeutic strategies to improve rhBMP‐2‐mediated bone repair in aged individuals.

Enhanced adipogenic differentiation is a hallmark of senescent BMSCs [16]. Under physiological conditions, adipocytes are crucial for bone homeostasis, functioning as energy reservoirs and endocrine cells that secrete various signaling molecules, including adipokines, which significantly influence bone remodeling [17, 18]. The balance between osteogenesis and adipogenesis within the bone marrow is tightly regulated to maintain skeletal integrity and proper bone microenvironment function [19, 20]. However, aging disrupts this balance, leading to an increase in bone marrow adiposity [21]. The accumulation of excessive adipocytes (eADs) not only physically displaces osteoprogenitor cells by occupying space and depleting essential resources but also promotes the secretion of pro‐inflammatory factors, further impairing bone formation and regeneration [22, 23]. This age‐related shift toward adipogenesis may be a key factor underlying the reduced efficacy of rhBMP‐2‐induced bone formation in the aged microenvironment [24, 25]. Despite this, it remains unclear whether the efficiency of rhBMP‐2 is influenced by the aberrant differentiation of senescent BMSCs, and the specific role of adipose‐derived cells in aged bone regeneration has not been fully elucidated [26]. Recent findings suggested that reducing adipocyte accumulation within bone tissue can enhance bone mass and improve skeletal health [27]. Therefore, inventing the tendency of differentiation in a compromised bone microenvironment may offer a promising therapeutic strategy to restore the regenerative capacity of senescent BMSCs and enhance rhBMP‐2‐mediated bone repair in elderly patients.

In this study, we examined the impact of aberrant differentiation in senescent BMSCs on the osteogenic efficacy of rhBMP‐2 and investigated the impact of eADs accumulation in an aged microenvironment on the regenerative potential of BMSCs. The accumulation of eADs in an aged condition exacerbates the detrimental bone regeneration microenvironment, thereby adversely affecting the osteogenic response of BMSCs to rhBMP‐2 treatment. To address this challenge, we proposed and validated a therapeutic strategy that intervenes in the aberrant differentiation of senescent BMSCs to mitigate eADs accumulation, which may ameliorate the aged microenvironment and preserve the regenerative capacity of BMSCs. The mammalian target of rapamycin (mTOR) pathway plays a crucial role in adipogenic differentiation and serves as an essential mechanism for delaying cellular senescence [28, 29, 30]. In our study, we employed rapamycin (Rapa), an mTOR pathway inhibitor, to counteract the aberrant differentiation of senescent BMSCs. Building on these insights, we developed an osteoid matrix‐mimicking injectable hydrogel composed of amino‐modified PEGylated poly(glycerol sebacate) (PEGSN) and poly(γ‐glutamic acid) (PGA), loaded with rhBMP‐2 and Rapa (PEGSN‐PGA/rhBMP‐2/Rapa, PBR), specifically designed to modulate the impaired microenvironment for aged bone regeneration (Scheme 1). The PEGSN‐PGA hydrogel exhibits excellent bone integration and robust adhesion to bone tissue, while degrading to release glycerol—a metabolically active molecule that can substitute for adipocyte‐derived energy sources—thereby helping to maintain the differentiation balance of senescent BMSCs. The in situ sustained‐release of rhBMP‐2 and Rapa from PBR hydrogel effectively attenuated the age‐induced aberrant differentiation of BMSCs, preserved the regenerative capacity of BMSCs within an aged microenvironment via cell cycle regulation and bone regeneration‐related pathways, thereby enhancing the quality of bone repair in aged mice. Our findings highlight a new perspective on the potential mechanisms underlying the inefficacy of rhBMP‐2 in aged conditions and propose a novel approach to modulate the differentiation propensity of senescent BMSCs, providing a promising strategy for addressing age‐related osteoporosis‐induced bone damage.

SCHEME 1.

Schematics of the PEGSN‐PGA/rhBMP‐2/Rapa (PBR) hydrogel facilitating bone regeneration and rejuvenation in an aged microenvironment. Following injection into the site of aged bone defects, the amino‐modified PEGylated poly (glycerol sebacate) (PEGSN)/poly (γ‐glutamic acid) (PGA) hydrogel loaded with rhBMP‐2 and rapamycin (PBR) facilitates local sustained drug release, thereby promoting aged bone regeneration and rejuvenation.

2. Results

2.1. The eADs in Aged Microenvironment Accelerate the Functional Decline of BMSCs

Given that the biological activity of rhBMP‐2 may differ between young and aged microenvironments, we initially evaluated the effects of rhBMP‐2 on the osteoblastic differentiation of senescent bone marrow mesenchymal stem cells (BMSCs). To evaluate the osteogenic efficacy of rhBMP‐2 in a senescent case, in vitro models of senescent BMSCs (S) and young BMSCs (Y) induced by conditioned medium (CM) were established (Figure S1). Senescent BMSCs demonstrated significantly elevated expression levels of senescence‐associated beta‐galactosidase (SA‐β‐gal) and the cellular senescence marker p53 (Figure S2A,B). Meanwhile, the expression of Ki67, a protein associated with cellular proliferative activity, was downregulated in senescent BMSCs, indicating a reduced capacity for cell proliferation (Figure S2C–E). Additionally, the observed decrease in adenosine triphosphate (ATP) levels suggests a decline in the metabolic activity of senescent BMSCs (Figure S2F). On this basis, we further investigated the effects of rhBMP‐2 on the heterogeneous differentiation of BMSCs in an aged microenvironment (Figure 1A). A rhBMP‐2 concentration of 2000 ng mL⁻¹ was utilized for subsequent investigations (Figure S3). Alkaline phosphatase (ALP) and alizarin red S (ARS) staining were used to evaluate the osteogenesis properties, and Oil‐red staining was used to evaluate the adipogenesis properties. As shown in Figure 1B–D, the osteogenic response of senescent BMSCs to rhBMP‐2 was significantly decreased, whereas the accumulation of lipid droplets was substantially elevated in the culture environment, compared with young BMSCs. To further investigate the differentiation propensity of senescent BMSCs under the influence of rhBMP‐2, messenger ribonucleic acid (mRNA) was collected from BMSCs and subjected to quantitative real‐time polymerase chain reaction (RT‐qPCR) analysis to determine the levels of osteogenesis‐related genes (osteopontin (Opn), runt‐related transcription factor 2 (Runx2), alkaline phosphatase (Alpl)) and adipogenesis‐related genes (fatty acid synthase (Fasn), fatty acid binding protein 4 (Fabp4), peroxisome proliferator‐activated receptor gamma (Pparγ)) (Figure 1E,F). The results demonstrated that senescent BMSCs exhibited elevated expression levels of adipogenesis‐related genes, and rhBMP‐2 further potentiated the adipogenic differentiation propensity of senescent BMSCs.

FIGURE 1.

Senescent BMSCs exhibited significant adipogenic differentiation in the presence of rhBMP‐2. (A) Schematic illustration of rhBMP‐2 combined with Rapa in the treatment of senescent BMSCs with aberrant differentiation. (B) ALP, ARS, and Oil‐Red staining of BMSCs. (C) Quantification of relative ALP activity. (D) Quantification of relative ARS activity. (E,F) Quantitative real‐time polymerase chain reaction (RT‐qPCR) analysis of Opn, Runx2, Alpl, Fasn, Fabp4, and Pparγ. (G) Immunofluorescence staining of F‐actin (cytoskeleton, red), DAPI (cell nucleus, blue), p53 (marker of cellular senescence, purple), Perilipin1 (marker of adipocytes, green). (H,I) Relative expression of Perilipin1 and p53 in BMSCs. (J) Colocalization analysis of the region of interest (ROI). (Y: young BMSCs; YB: young BMSCs treated with rhBMP‐2; S: senescent BMSCs; SB, senescent BMSCs treated with rhBMP‐2).

Next, we evaluated the distribution of senescent cells and adipocytes in BMSCs through immunofluorescence staining using markers specific to senescent cells (p53, visualized with purple fluorescence) and adipocytes (Perilipin1, visualized with green fluorescence). Immunofluorescence staining revealed that adipocytes were detected in both senescent (SB) and young BMSCs (YB) following rhBMP‐2 treatment (Figure 1G). Uncoupling Protein 1 (UCP‐1) serves a critical function in modulating energy expenditure and preserving metabolic homeostasis [31]. However, neither senescent nor young BMSCs‐derived adipocytes expressed UCP‐1 (visualized with yellow fluorescence) (Figure S4). The proportion of adipocytes (Perilipin1‐positive cells) in senescent BMSCs was approximately 4.6‐fold higher than that in young BMSCs (Figure 1H), indicating that senescent BMSCs had a greater propensity to differentiate into adipocytes in response to rhBMP‐2. Furthermore, a substantial number of senescent cells were observed in the vicinity of these adipocytes within the aged microenvironment (Figure 1G,I,J). Therefore, we hypothesized that the eADs derived from senescent BMSCs may contribute to the reduced osteogenic efficacy of rhBMP‐2 under an aged condition.

To further investigate the impact of eADs accumulation in an aged microenvironment compared to a youthful condition on BMSCs, bone marrow adipocytes (BMADs) differentiated from BMSCs were co‐cultured with non‐senescent BMSCs after being induced by CM (Figure 2A; Figure S5). After 3 days of rhBMP‐2 treatment, BMSCs co‐cultured with young CM‐induced BMADs (CMAD‐Y) exhibited significantly higher osteogenic‐related ALP activity than BMSCs co‐cultured with senescent CM‐induced BMADs (CMAD‐S) (Figure 2B,C). After 3 days of co‐culture, BMSCs in CMAD‐S group exhibited elevated expression levels of senescence‐associated markers (SA‐β‐gal and p53) (Figure 2D,E) and increased expression of senescence‐related genes (p16, p21, p53) (Figure 2F). These findings suggested that BMADs within an aged microenvironment accelerate the cellular senescence process in neighboring BMSCs. BMSCs in group CMAD‐S showed higher senescence‐associated secretory phenotypes (SASP) expression, such as Mmp3, Il1α, Il1β and Il6, indicating that BMADs in the aged microenvironment adversely impact the regenerative capacity of BMSCs (Figure 2G,H). These findings indicate that adipocytes accumulating in an aged microenvironment are more prone to compromise the regenerative milieu and cellular function, posing a threat to the therapeutic efficacy of rhBMP‐2 in the degraded bone regeneration microenvironment. Thus, reducing the accumulation of eADs in aged microenvironment represents a promising therapeutic strategy to mitigate cellular senescence and enhance the repair of aged bone.

FIGURE 2.

The accumulation of adipocytes within the senescent microenvironment is detrimental to the regenerative capacity of BMSCs. (A) BMSCs were induced to differentiate into BMADs and subsequently induced in young or aged CM for co‐culture with non‐senescent BMSCs. (B) ALP staining of BMSCs co‐cultured with BMADs and rhBMP‐2. (C) Quantification of relative ALP activity. (D) SA‐β‐gal staining of BMSCs co‐cultured with BMADs. Black arrow: SA‐β‐gal positive cells. (E) Immunofluorescence staining of F‐actin (cytoskeleton, red), DAPI (cell nucleus, blue), p53 (marker of cellular senescence, purple). (F) RT‐qPCR analysis of cellular senescence marker gene expression: p16, p21, and p53. G, H) RT‐qPCR analysis of SASP gene expression: Mmp3, Mmp9, Mmp12, Il1α, Il1β, and Il6. (CMAD‐Y: BMSCs co‐cultured with BMADs induced by young CM; CMAD‐S: BMSCs co‐cultured with BMADs induced by aged CM).

2.2. Enhance the Regeneration Potential of Senescent BMSCs by Mitigating the Accumulation of eADs Under the Treatment of rhBMP‐2

The accumulation of adipocytes in an aged microenvironment impairs tissue regeneration and function, promotes senescence progression through SASP secretion, and accelerates cellular senescence via SASP accumulation [32, 33, 34, 35]. To investigate the effect of eADs on the microenvironment in the aged microenvironment, GW9662 (GW) and Rapamycin (Rapa) were used to mitigate the accumulation of eADs. GW, a known antagonist of the PPAR pathway, was utilized to investigate whether inhibiting excessive adipogenic differentiation could enhance bone regeneration potential in senescent BMSCs. By inhibiting the mammalian target of rapamycin (mTOR) pathway, Rapa not only activates cellular autophagy, which is beneficial for delaying cellular senescence, but also interferes with the expression of PPARγ, and has been utilized to develop a strategy for interfering with the differentiation of senescent BMSCs [36, 37]. Rapa effectively restricted the adipogenic differentiation ability of BMSCs in adipogenic induction medium (Figures S6 and S7). Subsequently, the supernatant solution (SS) derived from BMSCs cultures was collected for further analysis (Figure 3A). Cytokine analysis of supernatant solution derived from BMSCs. Supernatant solution of senescent BMSCs treated with Rapa (SBR) and GW (SBGW) in the presence of rhBMP‐2 confirmed a down‐regulation of numerous cytokines such as CXCL1, CXCL2, CXCL3, CCL20, CXCL17, TIMP‐1, and VEGF (Figure 3B,C). Rapa has a more significant inhibitory effect on CXCL1, CXCL2, CXCL3, and LIX. CXCL1 (C–X–C Motif Chemokine Ligand 1), CXCL2, and CXCL3 are proteins that possess the capability to activate immune cells and induce inflammatory responses [38]. CCL20 (C–C Motif Chemokine Ligand 20), also referred to as macrophage inflammatory protein 3α (MIP‐3α), exhibits a potent chemotactic effect on various immune cells [39]. TIMP‐1 (Tissue Inhibitor of Metalloproteinases‐1) modulates the degradation of the extracellular matrix (ECM) through the inhibition of matrix metalloproteinases (MMPs). These factors are not only implicated in the inflammatory response but are also intricately associated with SASP, exhibiting elevated expression levels during cellular senescence [13, 40]. These results suggested that the elimination of eADs accumulation in the aged microenvironment may alleviate the accumulation of SASP factors in the microenvironment.

FIGURE 3.

Reducing the eADs under a senescent condition is beneficial to delay the cellular senescence of BMSCs. (A) The supernatants obtained from the samples were collected for the detection of secretion factors and subsequent co‐culture with non‐senescent BMSCs. (B, C) Qualitative cytokine array images used to analyze the supernatants, and corresponding quantification of relative cytokine expression carried out by densitometry. (D) ALP staining of BMSCs cultured with supernatants and rhBMP‐2. (E, F) RT‐qPCR analysis of Alpl and Runx2. (G) SA‐β‐gal staining of BMSCs cultured with supernatants for. Black arrow: SA‐β‐gal positive cells. (H) Immunofluorescence staining of F‐actin (cytoskeleton, red), DAPI (cell nucleus, blue), p53 (marker of cellular senescence, purple). (I) RT‐qPCR analysis of cellular senescence marker gene expression: p16, p21, and p53. (J, K) RT‐qPCR analysis of SASP gene expression: Mmp3, Mmp9, Mmp12, Il1α, Il1β, Il6. (SS‐YB: BMSCs co‐cultured with supernatant solution of young BMSCs treated with rhBMP‐2; SS‐SB: BMSCs co‐cultured with supernatant solution of senescent BMSCs treated with rhBMP‐2; SS‐SBR: BMSCs co‐cultured with supernatant solution of senescent BMSCs treated with rhBMP‐2 and Rapa; SS‐SBGW: BMSCs co‐cultured with supernatant solution of senescent BMSCs treated with rhBMP‐2 and GW).

Our previous study has demonstrated that the presence of senescent cells significantly impairs the functionality of non‐senescent BMSCs, while senescent BMSCs (p16⁺LepR⁺ cells) comprised merely 30% of the total BMSC population in aged bone tissue [16, 41]. To investigate the impact of microenvironmental changes on the regenerative behavior of non‐senescent BMSCs, the collected supernatant solution was subsequently co‐cultured with non‐senescent BMSCs (Figure 3A). BMSCs cultured in senescent conditions after mitigating eADs had higher osteogenic activity under the effect of rhBMP‐2 than those cultured in senescent conditions without intervention (Figure 3D). After 3 days of rhBMP‐2 treatment, the expression of Alpl in BMSCs increased by 13.5% ± 3.4% in group SS‐SBR and 7.3% ± 3.3% in group SS‐SBGW, and the expression of Runx2 increased by 23.5% ± 4.0% in group SS‐SBR and 9.5% ± 2.1% in group SS‐SBGW (Figure 3E,F). As illustrated in Figure 3G–I, the delayed senescence of BMSCs, marked by decreased expression levels of age‐related markers (SA‐β‐gal and p53) and age‐related genes (p16 and p21), was associated with the mitigation of eADs. Compared to the SS‐SB group, the SS‐SBGW group exhibited significantly reduced SASP expression levels, including Mmp9, Mmp12, Il1α, and Il6 (Figure 3J,K), which may contribute to maintaining a favorable microenvironment. Compared to the SS‐SBGW group, the expression levels of senescence‐related markers (p16, p21, and p53) in BMSCs were significantly reduced in the SS‐SBR group, indicating that Rapa treatment exerts a more pronounced effect on delaying cellular senescence and preserving cell function in BMSCs. Therefore, Rapa can interfere with the regeneration efficiency of BMSCs in an aged microenvironment by reducing the accumulation of eADs. Together, these results suggested that senescent BMSCs exhibited aberrant differentiation patterns, characterized by enhanced adipogenic differentiation and diminished osteogenic differentiation, in response to rhBMP‐2. The excessive accumulation of adipocytes within an aged condition results in the accumulation of SASP in the microenvironment, thereby accelerating the senescence of BMSCs. Improving the regenerative environment of senescent BMSCs with Rapa represents a promising strategy to augment the efficacy of rhBMP‐2 within aged conditions.

2.3. Rapa Modulates the Cell Cycle and Bone Regeneration‐Related Pathway of Senescent BMSCs

The physiological condition of the cells plays a crucial role in ensuring effective bone repair [42]. To elucidate the underlying cellular mechanisms of Rapa on the osteogenesis of senescent BMSCs in the presence of rhBMP‐2, we investigated the differences in gene expression between senescent BMSCs treated with rhBMP‐2 (SB) and senescent BMSCs treated with rhBMP‐2 and Rapa (SBR) using RNA sequencing. The principal component analysis (PCA) demonstrated consistent reproducibility across samples within each group (Figure S8). The volcano plot of differentially expressed genes (DEGs) revealed a total of 381 downregulated DEGs and 406 upregulated DEGs in SB group, as compared to the SBR group (Figure S9A). Heatmap analysis (Figure S9B) revealed substantial differences in gene expression between the two groups, indicating that Rapa treatment significantly influenced the gene expression profile of senescent BMSCs. Gene Ontology (GO) analysis was performed on differentially expressed genes, including biological process (BP), cellular component (CC), and molecular function (MF) classification. Enrichment analysis revealed alterations in a multitude of BP, such as immune response, cell cycle related terms (including cell cycle pathway, mitotic cell cycle pathway, mitotic cell cytokinesis, G2/M transition of mitotic cell cycle, regulation of G1/S transition of mitotic cell cycle and regulation of G2/M transition of mitotic cell cycle), skeletal system development, bone mineralization and DNA replication related terms (including DNA replication initiation and mitotic DNA replication checkpoint signaling), indicating the comprehensive regulatory effects of rhBMP‐2 and Rapa on senescent BMSCs (Figure 4A).

FIGURE 4.

Transcriptomic profiling of senescent BMSCs. (A) GO analysis categorizes DEGs by biological process, cellular component, and molecular function. (B, C) GSEA plots comparing BMSCs from the SBR and SB groups in bone mineralization and bone morphogenesis pathways. (D) Heat map representing the DEGs expression values for SBR versus SB in pathways associated with bone regeneration. (E–H) GSEA plots and DEGs expression for SBR versus SB in cell cycle and mitotic cell cycle pathways. (I) Schematic illustration summarizing the differentially expressed genes (DEGs) associated with cell cycle regulation during the S‐G2‐M phase, comparing SBR and SB.

Furthermore, gene set enrichment analysis (GSEA) was conducted. Senescent BMSCs treated with rhBMP‐2 and Rapa exhibited significantly upregulated expression of genes associated with bone mineralization and bone morphogenesis pathways (Figure 4B,C). A series of DEGs associated with bone regeneration exhibited significant up‐regulation in the SBR group, including Aspn11, Apsn, Mmp13, Ibsp, Bmp4, Omd, and Loc103690116 (related to bone mineralization), Sfrp1 and Wnt10b (related to bone trabecula formation), Postn, Mmp13 and Sfrp4 (related to bone morphogenesis), Cfth, Ttc9, and Bmp4 (related to bone development) (Figure 4D). These data indicated that the co‐treatment of rhBMP‐2 and Rapa exerts a beneficial effect on the osteogenic differentiation of senescent BMSCs. Furthermore, the SBR group exhibited a significant upregulation of genes associated with the cell cycle pathway and mitotic cell cycle pathway (Figure 4E,F), including Cks2, Mcm5, Cdca3, Cdc20, Mki67, Cep55, Cdkn3, Aurkb, Ncaph, Oip5, Ndc80, Aurka, Cenpw, Ncapg and Cenpt (related to cell cycle), Ska3, Cenpw, Mastl, Ska1, Cenpf, Cenpt, Cenpe, Espl1, Cdc6, Kif18b, Plk1 and Aurka (related to mitotic cell cyle) (Figure 4G,H). These results indicated that Rapa preserved the proliferative capacity of senescent BMSCs in a compromised condition. Furthermore, BMSCs exhibited significantly elevated Lpl expression following co‐treatment with rhBMP‐2 and Rapa, suggesting that the incorporation of Rapa enhances lipid metabolic activity in senescent BMSCs (Figure S10).

The cell cycle is intricately associated with the process of cellular senescence. Cell cycle arrest at G1 to S phase is a hallmark characteristic of senescent cells [43]. To further elucidate the regulatory effects of rhBMP‐2 and Rapa on the cellular senescence process of BMSCs, a Kyoto encyclopedia of genes and genomes (KEGG) map of the cell cycle pathway was presented (Figure 4I). Mcm, Plk1, Mpsl, Bubr1, and Esp1 genes were significantly up‐regulated in the SPBR group. Minichromosome maintenance proteins (MCM proteins) constitute a family of proteins that are activated during the DNA replication phase (S phase) of the cell cycle and play an indispensable role in ensuring accurate DNA replication [44]. PLK1 proteins (Pickle) are involved in cell cycle regulation, playing a crucial role in specific phases of the cell cycle, particularly during cell division and phase transitions [45]. As a dual‐specificity kinase, MPS1 primarily functions in regulating the activation of the spindle assembly checkpoint (SAC), ensuring accurate separation of sister chromatids [46]. The BUBR1 protein is a key molecule in maintaining the normal progression of the cell cycle and facilitating DNA damage repair [47]. Compared to the senescent BMSCs in the SB group, the BMSCs in the SBR group exhibited significantly enhanced regenerative activity, suggesting that the co‐treatment of rhBMP‐2 and Rapa effectively mitigated the senescence process of BMSCs. These data indicated that BMSCs in an aged microenvironment exhibit a rejuvenated niche when co‐treated with Rapa. Younger BMSCs are more responsive to the osteogenic differentiation induced by rhBMP‐2, preventing the production and accumulation of eADs, and demonstrating enhanced bone regeneration potential. Therefore, the synergistic treatment of rhBMP‐2 and Rapa is beneficial for preserving the regenerative capacity of BMSCs in aged conditions.

2.4. Dual Release System of rhBMP‐2 and Rapa Constructed Based on an Amino Modified PEGylated Poly (Glycerol Sebacate) (PEGSN)/poly (γ‐glutamic acid) (PGA) Hydrogel

Next, we aimed to identify a strategy to address the issue of rhBMP‐2‐mediated adipogenic differentiation of BMSCs, inhibiting bone repair in an aged microenvironment. The application of growth factors is frequently constrained by their short half‐life and uncontrolled dispersion [48]. Therefore, drug‐loaded scaffolds capable of enabling sustained release of bioactive factors and adapting to the microenvironment associated with aged bone regeneration are essential for effective in vivo therapeutic applications. Our previous research demonstrated that PEGS‐based hydrogels, owing to their viscoelastic properties, exhibit an elastic modulus comparable to that of collagenous bone, enabling them to effectively mimic the mechanical rigidity of the bone tissue extracellular matrix (ECM) and thereby promote osteogenic differentiation [49]. To achieve efficient loading and sustained release of rhBMP‐2 and Rapa, as well as to accommodate the regenerative microenvironment characteristics of senescent BMSCs, a PEGSN‐PGA hydrogel was engineered (Figure 5A). To facilitate the condensation reaction of PEGS with PGA, PEGS was functionalized with an amino group (Figures S11 and S12). The peak g (8.34 ppm) of ¹H NMR analysis indicated the amino group (Figure S11B). Following injection, the amino‐functionalized PEGS (PEGSN) was reacted with PGA in the presence of 1‐ethyl‐3‐(3‐dimethylaminopropyl)‐carbodiimide hydrochloride (EDC) and N‐hydroxysuccinimide (NHS) as coupling agents, leading to cross‐linking and the formation of a hydrogel. To investigate the contribution of hydrogel viscoelasticity to bone integration, we prepared a series of viscoelastic PEGSN‐PGA hydrogels (P‐1, P‐2, and P‐3) and characterized their mechanical properties under identical conditions (Table S1). In the tensile performance test of hydrogels, the P‐2 hydrogel exhibited the highest modulus of 130.225 KPa (Figure S13A) and demonstrated a superior toughness compared to the other hydrogels (Figure S13B). To further evaluate the adhesive strength of the hydrogel to bone tissue, bone slices were immersed in phosphate‐buffered saline (PBS, pH 7.4) to simulate a physiologically relevant moist environment (Figure S13C). With the increase of crosslinking degree, the adhesive strength of the hydrogel initially increased, reaching a maximum value before subsequently decreasing (Figure S13D). Among them, P‐2 exhibited the highest adhesive strength to bone tissue, reaching 36.67 ± 1.895 KPa (Figure S13E), which may be attributed to the presence of surface carboxyl groups on the P‐2 hydrogel that can form covalent bonds with the bone tissue interface while also providing strong internal cohesion, thereby resulting in superior adhesion performance. Achieving stable adhesion to bone tissue in a moist environment is conducive to enhancing the integration ability of the hydrogel with bone tissue and promoting bone repair. Therefore, the formulation corresponding to P‐2 hydrogel was selected for subsequent experiments.

FIGURE 5.

Preparation and characterization of rhBMP‐2 and Rapa dual‐release systems based on PEGSN‐PGA hydrogel (PBR). (A) Schematic representation of the preparation of the PEGSN‐PGA hydrogel. (B) Photographs of the shape adaptation of injectable PBR hydrogels to complex contours and adhesion to bone defects. (C) Rheological analysis of the hydrogels at 37°C in oscillatory time sweep mode. Circle dots represent the storage moduli (G′), and square dots represent the loss moduli (G″). (D) P hydrogel was degraded in conditioned medium at 37°C to produce glycerol molecules, and the quantitative analysis of glycerol yield from the degradation of the P hydrogel. (E) Energy dispersive spectrometer (EDS) mapping analysis of the P hydrogel. (F) XPS spectra of Ca 2p_3/2 and Ca 2p_1/2 peaks. (G) Oil‐Red staining of senescent BMSCs cultured in adipogenic induction medium. (H) Quantitative analysis of relative ALP activity of senescent BMSCs cultured in osteogenic induction medium. (I) The degradation behavior of PBR hydrogel at 37°C in PBS. (J) Cumulative release profiles of rhBMP‐2 and Rapa from PBR hydrogel at 37°C in PBS. (K) ALP, ARS, and Oil‐Red staining of senescent BMSCs treated with PB or PBR hydrogel. (L, M) Quantification of relative ALP and ARS activity. (N) RT‐qPCR analysis of Runx2 and Pparγ. (O) Cell cycle analysis of BMSCs. (P) Quantification of the proportions of BMSCs in the G1 phase. (Q, R) Immunofluorescence staining and quantification of Mcm5 expression.

The rhBMP‐2 solution and Rapa solution were mixed with the PEGSN precursors before crosslinking. The PBR hydrogel formed rapidly upon injection into the mold and bone injury samples (less than 60 s), demonstrating its capability to conform to complex injury topographies and achieve stable adhesion (Figure 5B). The scanning electron microscopy (SEM) image of the PBR hydrogel is presented in Figure S14. The storage modulus (G') of the PEGSN‐PGA hydrogel (P) and PEGSN‐PGA/rhBMP‐2/Rapa hydrogel (PBR) increased over time, ultimately exceeding the loss modulus (G''), and stabilized after 5 min (Figure 5C). The mechanical stability of the hydrogels was assessed through oscillatory frequency scanning experiments. The G' values of both P and PBR hydrogels exceeded their corresponding G'' values across the entire frequency testing range, indicating that these hydrogels exhibited stable viscoelastic solid behavior (Figure S15). These results demonstrate that the PEGSN precursors formed stable hydrogels in the presence of coupling reagents, and the incorporation of rhBMP‐2 and Rapa solution did not adversely affect the gelation properties of the hydrogel. As derivatives of polysebacate glyceryl ester (PGS), PEGSN possess the potential to generate and release glycerol molecules during the degradation process. As derivatives of polysebacate glyceryl ester (PGS), PEGSN possess the potential to generate and release glycerol molecules during the degradation process (Figure 5D). Glycerol is one of the energy substrates that adipocytes provide to surrounding cells [50]. The sustained release of glycerol during the degradation of the P hydrogel suggests that the PBR hydrogel can maintain an energy supply to senescent BMSCs to compensate for the positive functions of adipocytes in the microenvironment. Furthermore, the carboxyl groups (─COOH) abundantly present in the P hydrogel possess the capability to chelate Ca²⁺ from the surrounding microenvironment (Figure 5E). After incubation with Ca²⁺ solution, the P‐hydrogel was characterized by X‐ray photoelectron spectroscopy (XPS), and the doublet peaks observed at 345.8 and 349.3 eV for Ca_2p were assigned accordingly (Figure 5F). These results suggested that the P hydrogel provides an essential mineral support for the regeneration of aged bone. In vitro induction differentiation experiments further confirmed that P hydrogel could partially mitigate adipocyte aggregation during adipogenic induction of senescent BMSCs (Figure 5G) and significantly enhance their osteogenic differentiation capacity (Figure 5H), thereby helping to alleviate the imbalanced differentiation potential of BMSCs in an aged microenvironment.

Excellent biodegradability and biocompatibility are crucial prerequisites for hydrogels to be effectively utilized in the treatment of bone injuries [51]. Senescent BMSCs were observed to adhere to the surface of PBR hydrogel (Figure S16A) and exhibited no significant cytotoxic effects after 24 h of co‐culture (Figure S16B), suggesting that the PBR hydrogel possesses excellent biocompatibility. The PBR hydrogel was observed to be completely degraded within 21 days when incubated in PBS at 37°C (Figure 5I). To investigate the release profiles of rhBMP‐2 and Rapa from the PBR hydrogel, enzyme linked immunosorbent assay (Elisa) and high‐performance liquid chromatography (HPLC) were employed to quantify the respective release of rhBMP‐2 and Rapa were released to 58.95% ± 4.25% and 74.50% ± 5.10%, respectively, within the initial 5‐day period, after which the release rate decelerated (Figure 5J). The treatment of PBR hydrogel significantly suppressed the adipogenic differentiation of senescent BMSCs, and lipid droplets were virtually absent in the culture environment (Figure 5K). Compared with the PB group, the PBR group enhanced the osteogenic differentiation ability of senescent BMSCs. (Figure 5K). The osteogenic activity (ALP activity) and mineralization ability (ARS activity) were enhanced by 31.0% ± 6.7% and 33.6% ± 5.9%, respectively (Figure 5L,M). The expression of the key osteogenic gene Runx2 was significantly up‐regulated, while that of the key adipogenic gene Pparγ was significantly down‐regulated (Figure 5N). These results indicated that Rapa modulated rhBMP‐2‐mediated differentiation of senescent BMSCs. Cytokine analysis of the supernatant solution from senescent BMSCs co‐treated with rhBMP‐2 and Rapa (SPBR) revealed a significant down‐regulation of multiple cytokines, such as Acrp30, CXCL7, IGFBP‐3, Lipocalin‐2, CXCL6, IGFBP‐9, RBP4, Serpin E1, TWEAK (Figure S17), and the change trend paralleled that observed following the inhibition of adipogenic differentiation (Figure 3B,C). Among these factors, CXCL17 (C‐X‐C Motif Chemokine Ligand 17), CXCL6 (C‐X‐C Motif Chemokine Ligand 6), IGFBP (Insulin‐like growth factor binding protein), RBP4 (Retinol‐Binding Protein 4), Serpin E1 (Serpin Family E Member 1), and TWEAK (Tumor necrosis factor‐like weak inducer of apoptosis) are intricately associated with the development of inflammatory environments and aging‐related diseases [52]. The application of P hydrogel and Rapa mitigated the accumulation of eADs and prevented further deterioration of the regenerative microenvironment. The expression profiles of these factors in the microenvironment of the senescent BMSCs following Rapa treatment (SPBR) exhibited greater similarity to those observed in rhBMP‐2‐treated young BMSCs (YPB). These findings suggest that Rapa enhances the bone regeneration capability of senescent BMSCs by inhibiting excessive adipogenic differentiation, thereby preventing an overabundance of adipocytes from compromising the regenerative microenvironment and preventing the vicious cycle between cellular senescence, aberrant differentiation, and deterioration of the microenvironment.

To further elucidate the role of the cell cycle pathway in rhBMP‐2‐mediated regulation of senescent BMSCs, the hydrogel was co‐cultured with BMSCs in vitro for cell cycle analysis. Compared to young BMSCs co‐cultured with PB hydrogel (YPB), senescent BMSCs co‐cultured with PB hydrogel (SPB) exhibited a more pronounced cell cycle arrest (Figure 5O; Figure S18). About 51.57% ± 4.89% of the cells in YPB group were in G1 phase, 81.97% ± 1.10% in SPB group, and 73.01% ± 5.83% in SPBR group (Figure 5P). About 31.82% ± 8.09% cells in YPB group, 14.12% ± 1.01% in SPB group, and 15.11% ± 1.88% in S+PBR group were in S phase (Figure S19A). About 16.67% ± 9.40% cells in YPB group, 3.24% ± 1.55% in SPB group, and 14.38% ± 1.10% in SPBR group were in G2 phase (Figure S19B). The Mcm5 gene exhibited significant up‐regulation in senescent BMSCs following co‐treatment with rhBMP‐2 and Rapa (Figure 4I). To validate the differential expression of the Mcm5 protein, immunofluorescence analysis was conducted on BMSCs treated with both PB and PBR hydrogels (Figure 5Q). The addition of Rapa significantly increased the expression of Mcm5 in senescent BMSCs (Figure 5R). These findings suggested that the proliferative capacity of senescent BMSCs is markedly diminished compared to that of young BMSCs. The incorporation of Rapa attenuates the excessive and aberrant differentiation of BMSCs and partially restores the regenerative potential of senescent BMSCs.

2.5. PBR Hydrogel Facilitates the Repair of Bone Defects in Aged Mice

To assess the bone regeneration efficacy of PBR hydrogel in aged mice, a distal femoral defect model was established in 20‐month‐old male C57BL/6 mice. Micro‐CT analysis demonstrated that, in aged bone, the optimal osteogenic effect was achieved with a local rhBMP‐2 loading dose of 1 µg (Figure S20). Higher doses did not significantly enhance bone regeneration at the defect site and instead led to the formation of an enlarged callus. Therefore, 1 µg of rhBMP‐2 and the corresponding dose of Rapa were selected for subsequent in vivo treatment experiments. To evaluate the potential systemic side effects of PBR hydrogel in mice, we conducted histopathological analyses on major visceral organs from both postoperative aged mice and healthy aged mice (Figure S21A). Compared with the aged mice without any treatment, no significant tissue damage was observed in the major organs of aged mice following in vivo hydrogel implantation. Blood physiological and biochemical parameters further demonstrated low systemic toxicity, with standard hepatic and renal function indicators and inflammatory factorsall falling within normal reference ranges (Figures S21B–G and S22). These results revealed that the PBR hydrogel exhibits excellent biocompatibility and negligible systemic toxicity in aged mice with bone defects.

Following the injection of the hydrogels into the defect sites, samples were collected at the specified time points illustrated in Figure 6A to evaluate their osteogenic properties. The rapid ingrowth of blood vessels plays a crucial role in guiding bone tissue regeneration [53]. The PBR hydrogel significantly enhanced the expression of endothelial cells (CD31⁺ cells) in the bone defect region of aged mice, reaching levels comparable to those observed with the PB hydrogel in young mice (Figure S23). This indicates that the incorporation of Rapa facilitates rhBMP‐2‐mediated regeneration of surrounding tissues. Sequential fluorescent staining was employed to quantify the early osteogenesis rate. Calcein (CA) and alizarin red (AL) were utilized to label newly formed bone at day 3 and day 17 post‐treatment, respectively (Figure 6B). Compared with the SCtrl group, the bone formation rate was increased by 81.61% ± 16.73% in SP group, 146.14% ± 33.71% in SPB group, and 276.60% ± 35.56% in SPBR group (Figure 6C). These results indicated that the P hydrogel was conducive to early bone formation, and Rapa enhanced the osteogenic effects of rhBMP‐2. At week 4 and week 8 post‐surgery, the regenerated bone was evaluated using high‐resolution micro‐computed tomography (micro‐CT). Compared to the SCtrl group, the experimental group exhibited superior osteogenesis outcomes at the 4‐week post‐surgery (Figure 6D). Based on the morphometric analysis, the SPBR group exhibited superior osteogenic properties compared to other groups with higher bone volume fraction (BV/TV: 37.21% ± 2.93%), trabecular bone number (Tb. N:2.69 mm⁻¹ ± 0.19 mm⁻¹), and bone mineral density (BMD: 0.54 g cm⁻¹ ± 0.03 g cm⁻¹) (Figure 6E–G). Compared to the SPB group, the addition of rapamycin led to a 24.8% ± 9.8% increase in BV/TV, a 20.4% ± 8.7% increase in Tb. N, and a 20.9% ± 5.7% enhancement in BMD. These findings suggested that the incorporation of Rapa significantly enhances the capability of rhBMP‐2 to promote new bone formation in a compromised microenvironment. To further investigate the histological characteristics of the newly formed bone, decalcified specimens were subjected to sectioning and staining with hematoxylin–eosin (H&E) and Masson's trichrome. The SPBR group exhibited a significantly higher amount of trabecular bone (TB) and demonstrated the densest new bone (NB) structures, as illustrated in Figure 6H. At week 8 post‐surgery, the defects in the SPBR group were nearly completely repaired in comparison to those in the SPB group (Figure 6I). In histological analysis using H&E staining, haversian systems (HS) were clearly evident within the lamellar bone of the SPBR group, indicating a high degree of maturation of the newly formed bone (Figure 6J). These findings indicated that the incorporation of Rapa enhances the efficacy of rhBMP‐2 in promoting aged bone regeneration.

FIGURE 6.

The pro‐osteogenic efficacy of PBR hydrogel on aged bone was evaluated in vivo. (A) Experimental design time flowchart. (B, C) Fluorescence observation and quantification of new bone mineralization during orthotopic bone regeneration. (D) Micro‐CT 3D reconstruction images of samples at week 4 post‐treatment, the yellow circles highlight the defects in the femur. (E–G) Quantification of bone volume fraction (BV/TV), trabecular number (Tb. N), and bone mineral density (BMD) obtained from micro‐CT at week 4 post‐treatment. (H) Hematoxylin‐eosin (H&E) and Masson's trichrome staining images of the defects at week 4 post‐treatment. (I) Micro‐CT 3D reconstruction images of samples at week 8 post‐treatment, the yellow circles and arrows highlight the defects in the femur. (J) H&E and Masson staining images of the defects at week 8 post‐treatment. (SCtrl, control group; SP, PEGSN‐PGA hydrogel group; SPB, PEGSN‐PGA/rhBMP‐2 hydrogel group; SPBR, PEGSN‐PGA/rhBMP‐2/Rapa hydrogel group; TB, trabecular bone; NB, new bone; HS, haversian systems).

To further verify that modulating eADs accumulation is crucial for rhBMP‐2‐mediated repair of aged bone, immunofluorescence analyses were performed on samples collected at day 7 and day 14 post‐surgery. Compared to immune cells such as macrophages (CD68⁺ cells) and T cells (CD3⁺ cells), the PBR hydrogel exerts a more pronounced regulatory effect on adipocytes (Perilipin1⁺ cells) within the bone defect region of aged mice (Figures S24 and S25). The Preilipin1⁺ cell content in the SPB group was increased by 55.75% ± 27.31% and 97.01% ± 45.34% compared to the YPB group at day 7 and day 14 post‐surgery, respectively (Figure S25). However, this significant difference was abolished following Rapa administration. These findings indicate that the senescent microenvironment is prone to excessive adipocyte accumulation, which may serve as a critical factor impairing rhBMP‐2‐mediated bone regeneration in elderly individuals, and co‐treatment with Rapa can mitigate this detrimental effect.

The diminished proliferative capacity and functional impairment of stem cells within a compromised environment constitute primary factors contributing to the reduced efficiency of bone regeneration in elderly individuals [10, 54, 55]. To investigate the effect of PBR hydrogel on BMSCs recruitment and cellular senescence in vivo, samples were harvested at day 3 post‐surgery for immunofluorescence analysis. Senescent cells were labeled with p16 (visualized in green fluorescence), and BMSCs were labeled with Leptin receptor (LepR, visualized in red fluorescence) (Figure 7A). As illustrated in Figure 7A,B, rhBMP‐2 demonstrated a substantial recruitment effect on BMSCs (LepR‐positive cells) at the defect site, with a 628.84% ± 59.55% increase compared to the control group, and Rapa did not significantly affect the recruitment efficiency of BMSCs by rhBMP‐2. In the control group, senescent BMSCs (p16⁺LepR⁺ cells) comprised 30.0% ± 6.26% of the total BMSCs (LepR⁺ cells) (Figure 7C). Following rhBMP‐2 treatment, this proportion decreased to 9.59% ± 2.62%. The combination of Rapa and rhBMP‐2 resulted in the lowest proportion of senescent BMSCs at 6.56% ± 2.27%. These findings indicated that rhBMP‐2 facilitates the recruitment of a substantial number of young BMSCs to the injury site during the early phase of aged bone defect repair, thereby establishing a foundation for subsequent bone regeneration.

FIGURE 7.

Immunofluorescence and immunohistochemistry analysis of the defects post‐treatment. (A) Immunofluorescence images of p16 and LepR staining of the samples (Blue fluorescence: cell nucleus; red fluorescence: LepR‐positive (LepR⁺) cells; green fluorescence: p16‐positive (p16⁺) cells. (B) Quantification of LepR⁺ cells of the samples; (C) Quantification of p16 expression in LepR⁺ cells of the samples. (D) Immunohistochemistry images of Perilipin1, p16, and OCN staining of the samples. (E–G) Quantification of Perilipin1, p16, and OCN expression of the samples. (SCtrl: control group; SP: PEGSN‐PGA hydrogel group; SPB: PEGSN‐PGA/rhBMP‐2 hydrogel group; SPBR:e78496. PEGSN‐PGA/rhBMP‐2/Rapa hydrogel group).

To evaluate the impact of combined rhBMP‐2 and Rapa treatment on aged bone defect regeneration, samples were collected at week 4 post‐surgery for immunohistochemical analysis. The proteins Perilipin1, associated with adipogenesis; p16, linked to cellular senescence processes; and OCN, involved in osteogenesis, were utilized to assess newly formed bone (Figure 7D). A significant number of adipocytes were observed surrounding the newly formed bone in the SPB group. Notably, the incorporation of rapamycin resulted in a 51.60% ± 15.10% reduction in the proportion of adipocytes in response to rhBMP‐2 treatment (Figure 7E). Consistent with the expression trend of Perilipin1, a significant number of p16‐positive cells were observed surrounding the newly formed bone in the SPB group. The SPBR group exhibited a 59.70% ± 9.30% reduction in the number of p16‐positive cells compared to the SPB group (Figure 7F). The SPBR group exhibited the highest OCN expression level, with a 55.4% ± 17.0% increase compared to the SCtrl group and a 35.2% ± 14.8% increase compared to the SPB group (Figure 7G). These data indicated that the accumulation of adipocytes in a compromised environment is associated with cellular senescence, thereby diminishing the efficacy of rhBMP‐2 in promoting bone regeneration in aged mice. Co‐treatment with Rapa can mitigate cellular senescence and the reduced osteogenesis resulting from excessive adipocyte accumulation, thus accelerating the aged bone regeneration. These findings revealed that, although rhBMP‐2 has the potential of recruiting a considerable number of younger BMSCs during the early stages of aged bone regeneration, excessive and aberrant differentiation within the aged microenvironment substantially influences cellular senescence and compromises aged bone regeneration. Therefore, intervening in the differentiation behavior of senescent BMSCs holds significant guiding implications for the treatment of bone degenerative diseases.

3. Discussion

Senile osteoporosis, a prevalent degenerative bone disorder, is a leading cause of disability and mortality in elderly patients [56]. Recombinant human bone morphogenetic protein‐2 (rhBMP‐2) has been extensively utilized in the clinical management of bone injury‐related conditions owing to its superior osteoinductive properties [57]. However, we observed that the osteogenic differentiation efficiency of senescent bone mesenchymal stem cells (BMSCs) was markedly reduced compared to young BMSCs, and senescent BMSCs exhibited an exaggerated adipogenic differentiation response following rhBMP‐2 treatment, leading to an accumulation of excess adipocytes (eADs) (Figure 2B). Previous research has demonstrated that rhBMP‐2 modulates the expression of the key osteogenic gene Runx2 via the canonical Smad signaling pathway and regulates the key adipogenic gene Pparγ through the non‐canonical Smad signaling pathway [58, 59]. In this study, we observed that senescent BMSCs exhibited reduced Runx2 expression and elevated Pparγ expression compared to their younger counterparts in response to rhBMP‐2, which contributes to the increased accumulation of adipocytes (Figure 2E,F). This shift impairs bone regeneration efficiency in aged individuals. We suggested that the aberrant differentiation of BMSCs within an aged microenvironment is a critical factor influencing the osteogenic efficiency of rhBMP‐2 (Figure 3). The accumulation of eADs also significantly exacerbates the deterioration of the regenerative microenvironment (Figure 4B,C). Our findings indicate that the intervention of eADs represents a promising target for enhancing the efficiency of aged bone regeneration.

In addition to the degradation of the bone microenvironment, the accumulation of adipocytes significantly exacerbates the decline in regenerative potential within aged bone [22, 27]. Here, we investigated the impact of the eADs differentiated from senescent BMSCs within an aged microenvironment and identified the eADs as a critical factor influencing the regenerative efficacy of BMSCs (Figure 3B,C). The eADs exacerbate the accumulation of SASP within the aged microenvironment, thereby impacting the stem cell niche, and accelerating cellular senescence and functional decline in BMSCs (Figure 3). In light of these findings, we propose that mitigating the production of eADs will have considerable benefits for maintaining the regenerative capacity of senescent BMSCs. Senescent BMSCs exhibited elevated expression of adipogenic genes and demonstrated significant adipogenic differentiation, which was further enhanced by rhBMP‐2 (Figure 2). This phenomenon exacerbates the vicious cycle between cellular senescence, increased adipogenic differentiation, and diminished osteogenic differentiation in the aged microenvironment, which ultimately leads to the limited therapeutic effect of rhBMP‐2 under aged conditions. To address this issue, rapamycin (Rapa), an inhibitor of the mTOR pathway that interferes with Pparγ expression, was utilized in this study to mitigate the aberrant differentiation of senescent BMSCs [29]. We found that co‐treatment with rhBMP‐2 and Rapa can effectively delay cellular senescence and restore the regenerative capacity of BMSCs by maintaining cell cycle progression in an aged microenvironment (Figure 4). These findings suggest that the incorporation of Rapa mitigated the progressive functional decline of BMSCs in an aged microenvironment. We attributed this phenomenon to the fact that the addition of Rapa hindered the sustained effects of eADs on the surrounding BMSCs, thereby fostering a more conducive microenvironment for aged bone regeneration. This approach provides a promising strategy to promote the regeneration of senescent bone. Given the complexity of the aged microenvironment, it is imperative to further investigate these mechanisms that promote aged bone regeneration within the intricate interplay between senescent cells and aging‐related signaling pathways.

Adipocytes are key cells responsible for maintaining metabolic homeostasis and supplying energy for tissue regeneration [50]. Here, guided by the aberrant differentiation of BMSCs during bone regeneration in the elderly, we have considered the underlying mechanisms that may drive the accumulation of eADs. The buildup of eADs reflects an adaptive response to meet the heightened demand for energy reserves in aged tissues undergoing regeneration. The PEGSN‐PGA hydrogel developed in this study generates glycerol molecules upon degradation, preventing adipocyte accumulation and providing an alternative energy source for aged bone tissue in lieu of adipocytes (Figure 5D). At the same time, considering clinical practice, simple and easily implementable treatment strategies are essential for the effective management of bone injury conditions. This hydrogel is engineered to rapidly form a stable gel matrix within 60s post‐injection, accurately conforming to the anatomical contours of the injured site (Figure 5B). This minimally invasive treatment approach significantly alleviates patient discomfort while substantially reducing the associated procedural burden. Morever, the hydrogel demonstrates superior biocompatibility and degradability, rendering it highly suitable for aged bone regeneration and an effective delivery vehicle for rhBMP‐2 and Rapa (Figure 5I,J). In vivo studies have confirmed that PEGSN‐PGA/rhBMP‐2/Rapa hydrogel (PBR) intervention can ameliorate the aberrant differentiation of senescent BMSCs, thereby enhancing the regenerative capacity of aged bone tissue and promoting the rejuvenation of regenerated bone (Figures 6 and 7). This delivery system enhances the clinical feasibility of rhBMP‐2 and Rapa for treating age‐related bone injuries and provides novel insights into implant materials designed to target degenerative diseases.

4. Conclusion

In summary, our findings indicated that the accumulation of eADs in a senescent condition impairs the regenerative potential of BMSCs and accelerates their senescence, and these effects can be mitigated by reducing eADs accumulation. Mechanistically, co‐treatment with rhBMP‐2 and Rapa can delay cellular senescence and restore the regenerative capacity of BMSCs by safeguarding cell cycle progression in an aged microenvironment. Furthermore, we have developed an injectable PEGSN‐PGA hydrogel loaded with rhBMP‐2 and Rapa (PBR), specifically designed to align with the unique aged bone regeneration characteristics, for the treatment of age‐related bone defects. We believe this study provides innovative insights and strategic approaches for the clinical application of rhBMP‐2 in the treatment of bone degenerative diseases.

5. Experimental Section

5.1. Preparation and Characterization of PEGSN

PEGSN was synthesized through an esterification reaction followed by an acid‐induced deprotection step as previously described [60]. The chemical structures of PEGS and PEGSN were determined by ¹H Nuclear Magnetic Resonance Spectra (¹H NMR, 600 MHz) and deuterated dimethyl sulfoxide (DMSO‐d6) was used as a deuterated reagent. Fourier transform infrared spectroscopy (FTIR) was used to further verify the chemical structures of PEGS and PEGSN.

5.2. Preparation of the PBR Hydrogel

PEGSN (160 mg) was dissolved and mixed with γPGA (80 mg, average Mn: 500 kDa) in a rhBMP‐2 and Rapa solution (0.8 mL). The coupling reagents were prepared separately and the molar ratio of EDC to NHS was fixed at 1:1. The three main hydrogels were prepared with different crosslinking degrees by changing the molar ratio of the theoretical number of amino groups on the PEGSN backbone to EDC. Nomenclature and its constituent components of the hydrogels are listed in Table S1.

5.3. Mechanical Studies of the Hydrogels

To evaluate the mechanical properties of the hydrogels with different crosslinking degrees and, an electronic mechanical testing machine (SANS CMT2503) equipped with a 20 N load cell was used to measure the tensile properties of the completely gelled hydrogels with different crosslinking degrees. All tests were conducted at a constant displacement rate of 5 mm min⁻¹. Rectangular hydrogels with 35 mm length × 6.5 mm width × 1 mm thickness were prepared for the tensile tests.

5.4. Adhesive Properties of the Hydrogels

An electronic mechanical testing machine (SANS CMT2503) with a 20 N load cell was used to measure the lap–shear strength of hydrogels with varying cross‐linking degrees. The displacement rate was fixed at 5 mm min⁻¹. 300 µL of sample was injected into the adhesion area, which was fixed at 25 mm width × 10 mm length.

5.5. Characterization of the PBR Hydrogel

The fully dried PBR hydrogel was placed on the conductive glue, sputter‐coated with Au and observed under an SEM (Hitachi S‐3400, Tokyo, Japan) at an accelerating voltage of 15 kV. Following a 24‐h incubation in PBS solution containing Ca²⁺ (2.5 mm), Ca²⁺ adsorbed on the hydrogel surface was detected by X‐ray photon energy spectrometry (XPS, ESCALAB 250Xi) and energy dispersive spectrometer (EDS, S‐3400N, HITACHI). A rheometer (MARS3, Thermo HAAKE, USA) was utilized to conduct rheological tests.

5.6. Rheological Studies of the PBR Hydrogel

The gap between parallel plates (25 mm in diameter) was set to 1.0 mm. A 1 mL volume of liquid hydrogel was deposited onto a parallel plate, and the reaction proceeded for 10 min at 37°C, with a strain of 0.1% and a frequency of 1.0 Hz. The degradation behavior of the hydrogel was evaluated by measuring the residual mass in a PBS solution at 37°C under shaking condition.

5.7. Analysis of rhBMP‐2 and Rapa Release

The rhBMP‐2 used in this study was expressed in a prokaryotic system and purchased from Shanghai Ruibang Biomaterials Co., Ltd. RhBMP‐2 content released from PBR hydrogel was measured by a rhBMP‐2 ELISA kit according to the manufacturer's instructions. Rapa content released from PBR hydrogel was measured by high‐performance liquid chromatography (HPLC, Agilent 1200, USA) equipped with a GraceSMart RP18 column (250 mm × 4.6 mm, 5 µm). The mobile phase consisted of a methanol: water mixture at a ratio of 80:20, with an injection volume of 10 µL, a flow rate of 1 mL min⁻¹, a detection wavelength of 278 nm, and a column temperature of 45°C.

5.8. Determination of Glycerol Content

The PEGSN‐PGA hydrogel (P) was immersed in conditioned medium at a concentration of 0.1 g mL⁻¹, and samples were collected at day 3, 7, and 14. Following centrifugation to remove material fragments, the glycerol content in the supernatant was quantified using the Free Glycerol Assay Kit (Abcam, ab65337), according to the manufacturer's protocol.

5.9. Cell Culture in Conditioned Medium

Senescent and young BMSCs models were established as described previously [41]. Blood samples from aged (24‐month‐old) and young (1‐month‐old) Sprague Dawley (SD) male rats were collected through multiple, low‐volume tail vein punctures. All procedures were conducted under the guidance of professionals and with appropriate analgesic and anesthetic measures to minimize the animals' pain and stress. The isolated serum was pooled, aliquoted, and stored at −80°C for all subsequent in vitro experiments. All experimental protocols were approved by East China University of Science and Technology (Approval number: ECUST‐2024‐048). Conditioned medium (CM) was formed with 10% serum and 90% α‐MEM, without penicillin–streptomycin. BMSCs were isolated from femurs of 1‐month‐old SD male rats. The primary BMSCs were expanded to the third passage (P3) in a complete medium (α‐MEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin). Subsequently, the P3 BMSCs were induced with conditioned medium (CM) for 3 days to generate both young and senescent BMSCs for further experimentation.

5.10. Analysis of BMSCs Differentiation Properties

BMSCs were seeded at a density of 20 000 cells mL⁻¹ and induced with CM for 3 days. After induction, BMSCs were cultured in different experimental groups for osteogenic differentiation. The alkaline phosphatase activity of the BMSCs was quantified using a BCIP/NBT Alkaline Phosphatase Color Development Kit (Beytime, C3206) and an Alkaline Phosphatase Assay Kit (Beyotime, P0321) following the manufacturer's instructions. The osteogenic mineralization ability of BMSCs was detected by an Alizarin Red S staining kit (Beyotime, C0148) following the manufacturer's instructions. The expression of osteogenic differentiation‐related genes in BMSCs was assessed 3 days after induction, ALP activity was measured at 7 days post‐induction, and ARS staining was quantified at day 14 post‐induction.

5.11. Real‐Time Quantitative Polymerase Chain Reaction (RT‐qPCR)

To investigate gene expression in BMSCs, RT‐qPCR was performed. Total mRNA was extracted from BMSCs with RNAiso Plus Kit (TAKARA, 9109, Japan) according to the manufacturer's instructions. The purity and concentration of mRNA were quantified using NanoDrop 2000 (Invitrogen, USA). Reverse transcription was conducted using PrimeScript RT reagent kit (TAKARA, RR037A) as per the manufacturer's instructions. qPCR was performed using TB Green Premix Ex Taq kit (TAKARA, RR420A) on Bio‐Rad's CFX96 real‐time PCR system with a total of 39 cycles, following the manufacturer's protocol. Glyceraldehyde‐3‐phosphate dehydrogenase (Gapdh) was used as the reference genes. The primer sequences are detailed in Table S2 and synthesized by Sangon Biotech (Shanghai).

5.12. Detection of Factors

BMSCs were seeded at a density of 25,000 cells mL⁻¹ in six‐well culture plates and induced with CM for 3 days. After treatment, BMSCs from various groups were incubated in 1 mL medium per well (α‐MEM) for 24 h, and the supernatant was collected for the detection of secreted factors. Three parallel samples from each group were collected and subsequently pooled for analysis. Profiler Rat XL Cytokine Array (R&D, ARY030) was used for cytokine array, following the manufacturer's instructions. Membranes were visualized for up to 5 min using a chemiluminescence imaging system (Tanon, China) and each matrix point was quantified by Image J software (V.1.52a).

5.13. Senescence‐Associated‐𝛽‐Galactosidase Staining

BMSCs were seeded at a density of 10 000 cells mL⁻¹ in 24‐well culture plates. After treatment, the cells were stained using a Senescence β‐Galactosidase Staining Kit (Beytime, C0602) in accordance with the manufacturer's instructions and subsequently incubated overnight at 37°C in a CO₂‐free culture environment. Images were captured using an inverted fluorescence microscopy (Leica, Dmi8).

5.14. mRNA Sequencing and Analysis

The total mRNA of the senescent BMSCs after treatment were extracted using RNAiso Plus Kit (TAKARA, 9109, Japan), and the senescent BMSCs treated with rhBMP‐2were used as control. mRNA purification, library construction, and sequencing were performed by Hangzhou LC‐Bio Technology Co., LTD. PE150 sequencing mode was selected, and paired‐end sequencing was performed using Illumina Novaseq 6000 according to the standard procedures. In the analysis phase, CleanData was aligned to the Rattus genome using HISAT2 software, followed mRNA expression level analysis using StringTie and Ballgown software packages. Genes with an absolute log fold change greater than 1 and a p‐value less than 0.05 were identified as differentially expressed genes (DEGs). DAVID software was utilized to conduct Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of the DEGs. All aligned genes from the MSigDB version 7.3 database were considered, and gene sets with an absolute Normalized Enrichment Score (NES) greater than 1, a nominal p‐value less than 0.05, and a False Discovery Rate (FDR) less than 0.25 were deemed statistically significant.

5.15. Detection of Cell Cycle

50,000 BMSCs were seeded in each well of sixwell culture plates and induced with CM for 3 days. After overnight starvation treatment, the cells were co‐cultured with the hydrogel for 48 h, and the BMSCs was stained by a Cell Cycle Staining kit (Multi Science, CCS012) according to the manufacturer's procedure and subsequently analyzed by flow cytometry (Beckman, CytoFLEX s, USA) and ModFit software (V. 5.0).

5.16. In Vivo Study of Aged Bone Regeneration

All procedures were conducted in compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All experimental protocols were approved by East China University of Science and Technology (Approval number: ECUST‐2024‐048). A defect of 1.2 mm in diameter was made in the distal femur of aged male C57BL/6 mice (20‐month‐old). Hydrogels were administered in situ to evaluate their osteogenic properties in vivo. Femoral samples obtained post‐surgery were subjected to micro‐CT scanning and sectioned for histological examination.

5.17. In Vitro Immunofluorescence Analysis

BMSCs were seeded at a density of 20,000 cells mL⁻¹ and induced with CM for 3 days. After treatment, the cells were fixed to assess the extent of cellular senescence and the adipocyte content. The cells were permeabilized with 0.1% Triton X‐100, blocked with 5% goat serum, and incubated overnight at 4°C with primary antibodies against p53 (CST, 2524), Perilipin1 (Abcam, ab3526), UCP‐1 (Abcam, ab234430), Mcm5 (SantaCruz, sc‐165994), or Ki67 (CST, 9129). Fluorescein‐labeled secondary antibodies (goat anti‐mouse IgG conjugated with Alexa Fluor 647 and goat anti‐rabbit IgG conjugated with Alexa Fluor 488) were subsequently added. The cell nuclei and cytoskeleton were counterstained with DAPI (Alexa Fluro 405) and F‐actin (Alexa Fluro 555 and Alexa Fluro 555). All procedure according to the manufacturer's instructions. The cells were imaged using a laser confocal microscope (Leica TCS SP8) and quantified using Photoshop software (V. 2020).

5.18. In Vivo Histological and Immunological Analysis

For histological analysis, samples were fixed in 4% neutral paraformaldehyde buffer, decalcified in ethylenediaminetetraacetic acid (EDTA, 0.5 m), dehydrated in ethanol, embedded in paraffin, and sectioned at 4.5 µm thickness. Sections were stained with H&E and Masson, and the microstructure of the newly formed bone tissue was analyzed utilizing a light microscope. For immunohistochemical analysis, sections were immersed in a 3% H₂O₂ solution, washed and subjected to antigen retrieval using sodium citrate, permeabilized with 0.1% Triton X‐100, blocked with 5% goat serum, incubated with the primary antibody against p16 (Abcam, ab241543), Perilipin1 (Abcam, ab3526), or OCN (Abcam, ab93876) overnight at 4°C, and then incubated with the HRP‐conjugated secondary antibody at 37°C, followed by washing for chromogenic analysis. For immunofluorescence analysis, samples were collected and fixed in 4% ice‐cold poly‐formaldehyde solution for, decalcified in EDTA (0.5 m), dehydrated in 20% sucrose and 2% polyvinyl pyrrolidone, embedded and frozen in OCT compounds (Leica, 14020108926), and cryosectioned at 10 um thickness on a Leica CM1950 cryometer. The sections were permeabilized with 0.3% Triton X‐100 and subsequently blocked with 5% goat serum. The sections stained with primary antibodies LepR (R&D, BAF497), p16 (CST, 29271), CD68 (CST, 97778), CD3 (Abcam, ab135372), CD31 (Abcam, ab222783) or Perilipin1 (Abcam, ab3526) overnight at 4°C. Fluorescein‐labeled secondary antibodies biotinylated antibodies (Alexa Fluro 555) and goat anti‐rabbit immunoglobulin G (Alexa Fluro 488) were subsequently added, and then sealed with DAPI (CST, 8961). Images were captured using a laser confocal microscopy (Leica, TCS SP8, Germany).

5.19. Statistical Analysis

Unless otherwise specified, each in vitro experiment should include a minimum of three replicate samples, and each in vivo experiment should include a minimum of four replicate samples. All data were represented as mean ± SD. GraphPad Prism software (V. 8) was used for statistical analyses. One‐way ANOVA was employed for the comparison of multiple experimental groups as appropriate. Unpaired, two‐tailed Student's t‐test was utilized for comparisons of two independent groups. A p value less than 0.05 was deemed statistically significant, with the following notations: *p < 0.05; **p < 0.01; ***p < 0.005; ****p < 0.001.

Conflicts of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.;