A senescent metabolism-modulating hierarchical scaffold restores NAD⁺ homeostasis and redox balance for aged bone repair

Abstract

Senescence-impaired bone repair poses a significant clinical challenge, primarily due to the metabolic microenvironment imbalance driven by senescent cells. Current biomaterial strategies are largely confined to providing passive mechanical support and/or delivering therapeutic factors, failing to fundamentally reverse this specific pathological microenvironment. Here, we propose a senescent metabolism-reprogramming therapeutic strategy by developing a hierarchical bioceramic scaffold, comprising an outer 3D-printed β-tricalcium phosphate (β-TCP) hollow tube and an inner freeze-casting MnTCP ceramic rod with the loading of nicotinamide mononucleotide (NMN). During aged bone repair, the scaffold enables sustained release of both NMN and manganese (Mn) ions, targeting two core problems of the senescent bone microenvironment: NAD⁺ metabolic imbalance and accumulated oxidative stress. The scaffold was demonstrated to effectively restore intracellular NAD⁺ levels for reversing senescence-induced energy metabolism disorder by improving mitochondrial structure and function, and to scavenge excessive reactive oxygen species (ROS) for regulating cellular redox homeostasis. Therefore, it achieved cellular senescence alleviation by restoring energy metabolism and redox homeostasis. This anti-senescence action together with the inhibition of inflammatory responses collectively promoted osteogenic differentiation and enhanced bone regeneration. Based on the dual regulatory mechanism, the scaffold implantation significantly promoted new bone formation and maturation in aged rats with femoral condyle defects. This work not only provides an efficient biomaterial solution for bone defects in the elderly but, more importantly, pioneers a “material-mediated metabolic modulation” strategy to target tissue senescence, offering a new direction for treating senescence-related diseases in regenerative medicine.

Graphical abstract

Highlights

• •A hierarchical scaffold enables sustained co-release of NMN and Mn ions.
• •The scaffold targets NAD ⁺ deficiency and oxidative stress in the senescent pathology.
• •NMN restores energy metabolism while Mn ions regulate redox homeostasis in senescent BMSCs.
• •The scaffold alleviates senescence and inflammation, thus promoting osteogenesis in vitro/in vivo.
• •A material-mediated metabolic modulation strategy for aged bone repair is proposed.

1. Introduction

The accelerating global population aging has emerged as a critical public health challenge [1,2]. Incidence of senescence-related bone metabolic disorders, such as osteoporotic fractures and osteoarthritis in the elderly, continues to rise annually, leading to an increasing demand for bone defect repair and significantly compromising patients’ quality of life [3,4]. The underlying cause of the decline in bone repair capacity with aging lies in the dysregulation of metabolic networks. On one hand, nicotinamide adenine dinucleotide (NAD⁺), a key coenzyme in cellular energy metabolism and DNA repair as well as an essential signaling molecule for epigenetic regulation and stress response through proteins like sirtuins, undergoes a senescence-dependent decline. This decline directly leads to mitochondrial dysfunction, inhibition of osteoblast proliferation and differentiation, and senescence of bone marrow-derived mesenchymal stem cells (BMSCs), thereby impairing bone formation potential [5,6]. On the other hand, the senescence-associated redox imbalance results in the accumulation of reactive oxygen species (ROS). Excessive ROS deteriorates the bone marrow microenvironment, inhibits osteoblast function while over-activating osteoclasts, thus disrupting the bone remodeling balance and severely compromising bone defect repair [7,8]. Together, these interconnected metabolic imbalances create a hostile microenvironment that impedes bone regeneration. Therefore, precisely targeting these core metabolic mechanisms represents a promising therapeutic strategy to fundamentally build a pro-regenerative microenvironment in aged bone defects and restore bone repair capacity in the aging population [9,10].

Among numerous metabolic modulators, nicotinamide mononucleotide (NMN), a key NAD⁺ precursor, is prominent due to its efficient NAD⁺-boosting ability. NMN can be rapidly converted to NAD⁺ via the salvage pathway, thereby restoring mitochondrial function, enhancing cellular energy metabolism efficiency, and activating longevity-associated proteins such as sirtuins[[11], [12], [13]]. This process inhibits osteoblast senescence and promotes bone matrix mineralization, providing core metabolic support for aged bone repair. Previous studies have demonstrated that NMN alleviates BMSCs senescence and promotes osteogenesis via NAP1L2/SIRT1-related regulation [14], and the NMN-based delivery systems can promote bone regeneration by restoring NAD⁺-dependent energy metabolism, reducing oxidative stress, and improving mitochondrial function [15]. Concurrently, manganese (Mn), as an essential trace element, plays a complex yet important role in bone metabolism [16]. It serves not only as the core cofactor for manganese superoxide dismutase (MnSOD), which is localized to mitochondria and crucial for scavenging ROS and alleviating oxidative stress in bone cells [17], but also as a cofactor for various key enzymes such as glycosyltransferases, directly participating in bone matrix metabolism including protein glycosylation and collagen synthesis [[18], [19], [20]]. Therefore, the delivery of Mn ions through biomaterials holds promise for reversing oxidative stress-mediated impairment and improving bone regeneration ability [21]. The co-delivery of NMN and Mn ions is expected to effectively reverse the hostile metabolic microenvironment, including NAD⁺ deficiency-induced mitochondrial dysfunction and accumulated oxidative stress, during the senescence-associated bone defect repair. NMN replenishes NAD⁺ to enhance energy metabolism and alleviate osteoblast senescence, while Mn ions scavenge ROS via their antioxidant capacity and serve as essential cofactors for MnSOD, thereby enhancing mitochondrial antioxidant defense, mitigating oxidative damage and fostering a favorable osteogenic microenvironment. By addressing these key aspects of aging metabolism, this combination strategy offers the potential for optimizing bone repair. However, effective co-delivery remains challenging, as both agents require sustained, long-term release to ensure efficient intracellular NAD⁺ conversion and persistent antioxidant and osteogenic properties [22]. Therefore, developing a delivery system capable of prolonged co-release of NMN and Mn ions is of significant importance for advancing age-related bone defect repair.

Herein, we designed a senescent metabolism-modulating porous composite scaffold, comprising a 3D-printed hollow β-tricalcium phosphate (β-TCP) outer wall and an inner MnTCP ceramic rod formed via freeze-casting. This spatially organized hollow-porous scaffold offers structural and functional advantages for aged bone repair. The 3D-printed β-TCP outer framework provides mechanical support and preserves defect space, while the freeze-cast porous MnTCP inner rod enables NMN loading. Furthermore, the scaffold enables sustained release of both NMN and Mn ions during aged bone repair, targeting two core senescence-associated pathological aspects: not only supplementing NAD⁺ precursors to restore the cellular energy metabolic hub, but also utilizing Mn ions to coordinately regulate redox balance. Together, these actions effectively reverse senescence-related energy metabolism imbalance by improving mitochondrial structure and function (i.e., the tricarboxylic acid (TCA) cycle and oxidative phosphorylation (OXPHOS) processes) for anti-senescence, and reduce accumulation of excessive ROS for anti-inflammation in senescent cells, thereby reprogramming the senescent bone microenvironment and improving osteogenic differentiation and bone regeneration ability (Fig. 1). Mechanistic studies indicated that the scaffold could activate FoxO-related signaling pathway, and thereby enhance antioxidant defense, alleviate cellular senescence, and promote osteogenic regeneration during aged bone repair. In an aged rat femoral condyle defect model, the scaffold demonstrated a significantly enhanced ability to accelerate osteogenesis. This work offers an innovative therapeutic strategy of biomaterial-mediated senescent metabolism-reprogramming for treating aged bone defects.

Fig. 1.

Schematic illustration of the construction of HT-MnT/NMN scaffold system and its ability to reprogram the senescent bone metabolic microenvironment for promoting aged bone defect repair.

2. Results and discussion

2.1. Fabrication, characterization, and optimization of HT-MnT/NMN scaffolds

MnTCP powder was synthesized following a previously reported method [23]. X-ray diffraction analysis (XRD) analysis confirmed that the crystal structure of β-TCP remained unaltered upon Mn ion doping (Fig. S1a). Scanning electron microscope (SEM) images showed that the powder particles had a size of approximately 800 nm, and elemental mapping further verified the homogeneous distribution of Ca, P, O, and Mn within the powder (Fig. S1b). Inductively coupled plasma optical emission spectrometry (ICP-OES) results indicated an actual Mn doping concentration of 5.1%, demonstrating the successful preparation of MnTCP powder (Table S1). The antioxidant capacity of the MnTCP powder was evaluated through 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), 1,1-diphenyl-2-picrylhydrazyl (DPPH), superoxide anion (O₂•^-), and hydrogen peroxide (H₂O₂) scavenging experiments, which confirmed its effective scavenging of multiple ROS (Fig. S2).

Further, a series of scaffolds with different MnTCP contents were fabricated by freeze-casting MnTCP in the hollow tubes of TCP scaffolds (the MnTCP slurry was composed of MnTCP (0∼40 wt%), polyvinyl pyrrolidone (PVP, 2 wt%), polyacrylic acid sodium (PAAS, 0.02 wt%), and phosphate buffer saline (PBS, 97.98 wt%-(0∼40 wt%))) (Fig. 2a). The antioxidant results demonstrated that the scaffolds with Mn-TCP rods exhibit ROS/radical-scavenging activity in vitro, which was gradually enhanced with increasing MnTCP slurry concentrations (Fig. 2c). Subsequently, the NMN loading on the scaffolds indicated that the maximum NMN loading was reached when the NMN concentration was over 40 mg/mL (Fig. 2e). Notably, as the MnTCP content increased, the NMN loading capacity of the scaffolds showed a gradually decreasing trend. This phenomenon might be attributed to the changes in the scaffold porosity. A higher MnTCP slurry concentration led to a decrease in porosity [24] (Fig. 2d), and this microstructural evolution clearly corroborated by SEM (Fig. 2b). Mechanical testing revealed that the compressive strength of the scaffolds ranged from approximately 2 to 6 MPa, and the compressive strength increased progressively with higher MnTCP content (Fig. S3). The scaffolds exhibited favorable sustained release behavior for both NMN and Mn ions in all cases (Fig. 2f,g, and Fig. S4). The release of NMN continued for more than 56 days, with a cumulative release rate exceeding 60%. Meanwhile, the Mn ion release profile showed steady release throughout the 56-day period. The initial release of NMN and Mn ions during the first week may help treat acute oxidative stress, mitochondrial dysfunction, and NAD⁺ depletion during the early inflammatory microenvironment. The sustained release from approximately day 7 to day 28 may support BMSCs survival, mitochondrial energy metabolism, and osteogenic differentiation during the reparative phase. The continued low-level release up to 56 days may further contribute to redox homeostasis, matrix maturation, and early remodeling, which are particularly important in aged bone defects where repair is often delayed.

Fig. 2.

Characterization and performances of the 3D-printed scaffolds. (a) Overview of different scaffolds. (b) SEM images and high magnification SEM images of scaffold structures. (c) Antioxidant properties of the scaffolds. (d) Porosity of the scaffolds. (e) NMN loading curve of the scaffolds. (f-g) NMN (20 mg/mL) and Mn ions release of the scaffolds. Data are represented as mean ± SD. n = 3. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, as determined by one-way ANOVA analyses with Tukey's post-hoc test for all comparisons.

2.2. HT-MnT/NMN scaffolds restore mitochondrial metabolism and redox homeostasis in senescent BMSCs

BMSCs isolated from New Zealand white rabbits at passage 2 (P2) were utilized as the cellular model, and senescent BMSCs were induced through D-galactose (D-gal) treatment [25]. A BMSC senescence model was successfully constructed using a D-gal concentration of 20 g/L, selected based on maximal senescence-associated β-galactosidase (SA-β-gal) positivity and preserved cell viability (Figs. S5 and S6). Compared to young cells (YCs), senescent cells (SCs) displayed hallmark features of cellular senescence, including enlarged cell morphology, elevated ROS, NAD⁺ depletion, fragmented and condensed mitochondria with increased mitochondrial ROS (mtROS), upregulation of related proteins (p16 and γ-H2AX), and altered gene expression profiles characterized by downregulated osteogenic-related genes and upregulated senescence/inflammatory-related genes (Fig. S7–S10). The successful construction of the stable senescence model provides a reliable cellular foundation for subsequent systematic evaluation of scaffold-mediated anti-senescence effects.

To optimize the scaffold composition, we simultaneously evaluated the effects of the NMN loading concentration (5∼40 mg/mL) and the embedded MnTCP content (10%∼40%) on senescent BMSCs proliferation after 7 days of culture (Fig. S11). Both NMN concentration and MnTCP content significantly influenced cell proliferation. It can be observed that the HT-20% MnT scaffolds with NMN loading at 20 mg/mL and 40 mg/mL exhibited a better pro-proliferative effect across all of scaffolds. This optimal performance might be attributed to a balanced release profile of the NMN and Mn ions from the scaffolds (Fig. 2f and g). The scaffolds with lower MnTCP content (e.g., 10%) released adequate NMN but insufficient Mn ions, whereas those with higher MnTCP content (e.g., 40%) delivered more Mn ions but exhibited limited NMN loading and release. The results from Fig. S11 suggest that NMN serves as the primary factor promoting senescent BMSCs proliferation in this system, while the combined action of Mn ions and NMN is essential for maximizing the cell-promoting effect. Based on these parallel evaluations, the HT-20% MnT scaffold loaded with 20 mg/mL NMN (termed HT-MnT/NMN) was selected for all subsequent experiments. Meanwhile, the control groups were set as follows: the hollow TCP scaffold (HT), and the composite scaffolds with an outer wall composed of TCP and inner layers composed of TCP or MnTCP ceramic rods (HT-T or HT-MnT).

To delineate the respective contributions of NMN and Mn ions to the observed pro-proliferative effect, cell proliferation of senescent BMSCs was further assessed on the HT, HT-T, HT-MnT, and HT-MnT/NMN scaffolds, which were denoted as HT(sen), HT-T(sen), HT-MnT(sen), and HT-MnT/NMN(sen), respectively. Meanwhile, normal BMSCs cultured on HT scaffolds were included as a reference control, named as HT(nor), to indicate the baseline proliferative capacity of non-senescent cells. The HT-MnT(sen) group exhibited enhanced proliferation compared to the HT(sen) and HT-T(sen) groups, indicating that the embedded MnTCP ceramic rods effectively promoted cell proliferation (Fig. 3a). Notably, the HT-MnT/NMN(sen) group showed the highest proliferation capacity, which was significantly superior to that of the HT-MnT(sen) group and close to the proliferation level observed in the HT(nor) group, confirming that the combination of NMN delivery and Mn ion release further enhanced the pro-proliferative effect of the scaffold. The proliferation results were further corroborated by EdU staining, which also showed the highest percentage of EdU-positive cells in the HT-MnT/NMN(sen) group compared with other senescent treatment groups (Fig. S12). To explore the underlying mechanism, the intracellular NAD⁺ levels were determined using an NAD⁺/NADH detection kit. A markedly reduced NAD⁺ level was found in senescent BMSCs compared to normal BMSCs, while a significant increase in the NAD⁺ levels for senescent BMSCs was observed for the HT-MnT/NMN group compared to other groups, confirming the effective release of NMN from the scaffolds and its efficient conversion into NAD⁺ within the cells (Fig. 3b). These results indicate that the sustained release of Mn ions and NMN collectively promotes senescent BMSCs proliferation.

Fig. 3.

Scaffolds restore cellular energy metabolism by regulating mitochondrial structure and function. Normal BMSCs were included as the control group in all experiments. (a) Proliferation of BMSCs cultured on different scaffolds for 1, 3, and 7 days. (b) Relative NAD⁺/NADH ratio after different scaffold treatments. (c) Mitochondrial OCR analysis under different scaffold-treated conditions. (d-e) SIRT1 levels and relative MnSOD activity after different scaffold treatments. (f, g) Representative protein expression and corresponding quantification of energy metabolism-related markers in BMSCs cultured on different scaffolds for 7 days. (h) Mitochondrial regulation, including mitochondrial morphology labeled with MitoTracker, mitochondrial ultrastructure observed by Bio-TEM, mitochondrial ROS levels, and mitochondrial membrane potential. (i) Expression of metabolism-related genes after different scaffold treatments. (j) ATP levels and metabolism-related enzyme activity in BMSCs cultured on different scaffolds for 7 days. For proliferation assay, n = 4; for other assays, n = 3. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns, not significant, as determined by one-way ANOVA analyses with Tukey's post-hoc test for all comparisons.

To assess the impact on mitochondrial energy metabolism specifically the TCA cycle and OXPHOS processes [26], immunofluorescence analysis was carried out on oxoglutarate dehydrogenase (OGDH, a TCA cycle marker) and NADH-ubiquinone oxidoreductase MLRQ subunit homolog (NDUFA4L2, an OXPHOS marker) (Fig. 3f and g). The HT(nor) group served as a baseline, exhibiting the highest fluorescence intensities for both OGDH and NDUFA4L2. Compared to the HT(sen) and HT-T(sen) groups, the HT-MnT(sen) group exhibited significantly increased fluorescence intensities for both OGDH and NDUFA4L2 proteins, indicating that the Mn ions released from the scaffolds can effectively promote energy metabolic pathways. Notably, the HT-MnT/NMN(sen) group showed further elevated fluorescence intensities for OGDH and NDUFA4L2 proteins relative to the HT-MnT(sen) group, suggesting a complementary effect of Mn ions and NMN in restoring cellular energy homeostasis. The SIRT1 protein (a key NAD⁺-dependent deacetylase involved in mitochondrial metabolism and cellular energy homeostasis) levels were also evaluated across different groups (Fig. 3d). The HT(sen) and HT-T(sen) groups showed markedly reduced SIRT1 protein levels, while the HT-MnT(sen) group partially restored SIRT1 expression, and the HT-MnT/NMN(sen) group induced the most pronounced increase. Additionally, real-time quantitative polymerase chain reaction (RT-qPCR) analysis revealed that the treatment with HT-MnT/NMN(sen) scaffold significantly upregulated the expression of energy metabolism-related genes, including nicotinamide mononucleotide adenylyltransferase (NMNAT) and SIRT1 (Fig. 3i). Collectively, these results demonstrate that the HT-MnT/NMN scaffold facilitates the restoration of normal physiological function in senescent BMSCs through regulation of metabolism-related proteins and genes.

Mitochondria are the primary organelles for cellular energy metabolism, and their structural and functional integrity is crucial for maintaining normal cellular activities [27]. Thus, the effects of various scaffolds on the mitochondrial morphology and function were investigated (Fig. 3h). MitoTracker Deep Red staining revealed the fragmented mitochondria in the HT(sen) and HT-T(sen) groups and intact mitochondria in the HT(nor) group, suggesting that cell senescence causes mitochondrial fragmentation and structural damage. However, the mitochondrial morphology improvement was observed in both the HT-MnT(sen) and HT-MnT/NMN(sen) groups, with mitochondria in the latter appearing continuous, elongated, and more structurally intact. These results suggest that the co-release of Mn ions and NMN collectively reverses the senescence-associated mitochondrial abnormalities. Bio-transmission electron microscopy (Bio-TEM) analysis provided further evidence of mitochondrial structural changes in BMSCs following different treatments (Fig. 3h). The mitochondria in the HT(nor) group showed clear double membranes and well-organized cristae, whereas the mitochondria in the HT(sen) and HT-T(sen) groups displayed pathological alterations such as blurred membranes and cristae breakage or disappearance. In contrast, these pathologies were partially alleviated in the HT-MnT(sen) group. Significantly, the mitochondria in the HT-MnT/NMN(sen) group exhibited clear and intact double membranes with well-arranged cristae, demonstrating the most pronounced morphological recovery through the combined action of Mn ions and NMN. At the functional level, the MnSOD activity was severely suppressed in the HT(sen) and HT-T(sen) groups. The HT-MnT(sen) group exhibited only a modest elevation, whereas the HT-MnT/NMN(sen) group restored activity to a level comparable to that of the HT(nor) group (Fig. 3e). Subsequently, MitoSOX Green assay showed significantly reduced mtROS levels in both the HT-MnT(sen) and HT-MnT/NMN(sen) groups (Fig. 3h). In parallel, the mitochondrial membrane potential measurements showed that both the HT-MnT(sen) and HT-MnT/NMN(sen) groups improved mitochondrial membrane potential, but the combined treatment restored it to a level close to that of the HT(nor) group (Fig. 3h). These results indicate that the HT-MnT/NMN scaffolds effectively attenuated mitochondrial oxidative stress and preserved mitochondrial function, likely through the enhancement of MnSOD-related mitochondrial antioxidant defense. Additionally, biochemical assays revealed that the HT-MnT/NMN(sen) group significantly enhanced cellular energy metabolism in senescent BMSCs, as evidenced by promoted ATP production and increased activities of key metabolism-related enzymes including isocitrate dehydrogenase (IDH), malate dehydrogenase (MDH), alpha-ketoglutarate dehydrogenase (α-KGDH) to the levels comparable to those of the HT(nor) group (Fig. 3j and Fig. S13). Furthermore, the oxygen consumption rate (OCR) data showed that the senescence-inducing significantly impaired OXPHOS process whereas the HT(nor) group maintained stable mitochondrial respiratory function. The HT-MnT/NMN treatment effectively improved mitochondrial functions, evidenced by the upregulation of key respiratory capacity-related indexes including basal respiration, ATP production, and maximal respiratory capacity in senescent cells (Fig. 3c and Fig. S14). These above results demonstrate that the HT-MnT/NMN scaffold, through the co-release of Mn ions and NMN, can effectively restore mitochondrial structure and function, enhance MnSOD-associated antioxidant defense, scavenge mtROS, while promoting metabolic enzyme activities and ATP synthesis, thereby correcting the energy metabolism imbalance in senescent BMSCs.

2.3. HT-MnT/NMN scaffolds alleviate senescence-associated dysfunction, oxidative stress, and inflammation

Given that improved energy metabolism is closely associated with the attenuation of cellular senescence [28], we next investigated the anti-senescence effects of the scaffolds following our confirmation that the HT-MnT/NMN scaffold could restore energy metabolism in senescent cells. The SA-β-gal staining results showed that the proportion of SA-β-gal-positive cells was significantly reduced in the HT-MnT(sen) group compared to the control group, indicating a preliminary alleviation in the senescence process of senescent BMSCs. Notably, the HT-MnT/NMN(sen) group exhibited a reduction in SA-β-gal positivity to a level similar to that of the HT(nor) group, demonstrating a more potent anti-senescence effect (Fig. 4a and Fig. S15a). Immunofluorescence analysis of the related proteins revealed consistent trends: compared with the HT(nor) group, the fluorescence intensities of the senescence-associated protein p21 and the DNA damage marker γ-H2AX were increased in the HT(sen) and HT-T(sen) group. In contrast, both markers were decreased in the HT-MnT(sen) group, with a more pronounced reduction observed in the HT-MnT/NMN group(sen) (Fig. 4b and Fig. S15b and c), indicating that co-release of Mn ions and NMN can efficiently suppress cell cycle-inhibiting protein and repair DNA damage. Flow cytometry analysis of the cell cycle showed that, compared with other senescent treatment groups, the HT-MnT/NMN(sen) group increased the proportion of cells in S phase and G2 phase, decreased the G1-phase population, with a cell cycle distribution closer to that of the normal control, reflecting improved cell cycle progression (Fig. 4e). At the gene level, the RT-qPCR results revealed that the HT-MnT(sen) group mildly downregulated key senescence-related genes including p21, and p53, whereas the HT-MnT/NMN(sen) group significantly downregulated the expression of these genes (Fig. 4f). These results further confirm the efficient regulation of senescence-associated pathways by the HT-MnT/NMN scaffold. ELISA measurements of inflammatory factors TNFα and MCP-1 showed that the HT-MnT/NMN(sen) group exhibited markedly reduced secretion levels compared to other senescent treatment groups, indicating the attenuation of the senescence-associated secretory phenotype (SASP) (Fig. 4g). Collectively, these findings indicate that the HT-MnT scaffold exerts basal anti-senescence function through the sustained Mn ion release, while the HT-MnT/NMN scaffold achieves enhanced anti-senescence effect through the combined action of Mn ions and NMN.

Fig. 4.

Scaffold-mediated anti-senescent, antioxidant, and anti-inflammatory effects in BMSCs. Normal BMSCs were included as the control group in all experiments. (a) SA-β-gal staining levels after 7 days of culture. (b, d) Representative immunofluorescence images and quantitative analysis of the senescence marker p21, DNA damage marker γ-H2AX, and inflammatory markers. (c) Intracellular ROS levels. (e) Cell cycle analysis by flow cytometry. (f, h) Expression of senescence-associated and inflammation-related genes. (g) Levels of inflammatory cytokines (TNFα and MCP-1). Data are presented as mean ± SD, n = 3. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns, not significant, as determined by one-way ANOVA analyses with Tukey's post-hoc test for all comparisons.

The senescent bone microenvironment typically exhibits redox imbalance with excessive ROS generation, which induces inflammation and impairs osteogenesis [29]. Therefore, we assessed the antioxidant properties of the scaffolds. The 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) fluorescent probe was employed to assess the ability of scavenging intracellular ROS within the D-gal-induced senescent BMSCs. The results showed that the ROS levels were significantly lower in the HT-MnT(sen) and HT-MnT/NMN(sen) groups compared to the HT(sen) and HT-T(sen) groups (Fig. 4c and Fig. S15d), indicating that the embedded MnTCP ceramic rods in the scaffolds confer excellent antioxidant activity, which could effectively mitigate oxidative stress damage in senescent cells. In addition, immunofluorescence staining showed strong IL-1β intensity in the HT(sen) and HT-T(sen) groups compared with the HT(nor) group, confirming an enhanced inflammatory response in senescent BMSCs. The IL-1β intensity was reduced in the HT-MnT(sen) group and further decreased in the HT-MnT/NMN(sen) group, suggesting that the HT-MnT/NMN scaffold effectively alleviates inflammation-associated cellular damage in senescent BMSCs (Fig. 4d and Fig. S15e). Furthermore, the qPCR analysis of inflammation-related gene expressions revealed that the HT-MnT and HT-MnT/NMN scaffolds significantly downregulated the mRNA expression levels of the pro-inflammatory factors IL-1β and TNF-α (Fig. 4h). Therefore, the designed scaffold can simultaneously mitigate cellular senescence and exert antioxidant effect to suppress the inflammatory response in senescent cells.

2.4. HT-MnT/NMN scaffolds enhance osteogenic differentiation of senescent BMSCs

Restoration of cellular metabolic/redox homeostasis and anti-senescence are critical for regulating the osteogenic differentiation potential of senescent BMSCs and achieving efficient bone repair [30]. We further systematically evaluated the effect of the HT-MnT/NMN scaffold on the osteogenic differentiation capacity of senescent BMSCs. From the staining images for the early osteogenic differentiation marker alkaline phosphatase (ALP) and late marker alizarin red S (ARS) (Fig. 5a) as well as the quantitative analyses for the ALP activity and ARS mineralization level (Fig. 5c), the HT(nor) group exhibited strong osteogenic differentiation and mineralization capacity after osteoinduction culture. Among senescent BMSCs, the HT-MnT(sen) group demonstrated a clear mineralization-promoting ability compared to the HT(sen) and HT-T(sen) groups, while the HT-MnT/NMN(sen) group exhibited the strongest mineralization-promoting capability, showing an osteogenic differentiation recovery trend toward HT(nor) group. Immunofluorescence staining further revealed that the cell senescence markedly weakened the expression of type I collagen (COL-I) and osteocalcin (OCN) (Fig. 5b–d). Among senescent treatment groups, the expression of COL-I and OCN was increased in the HT-MnT(sen) group relative to the HT(sen) and HT-T(sen) groups, and the HT-MnT/NMN(sen) group showed the most pronounced increase in the expression of these two extracellular matrix proteins (Fig. 5b–d). These results indicated that both the HT-MnT and HT-MnT/NMN scaffolds promote matrix protein expression, but the HT-MnT/NMN scaffold exhibited a significantly enhanced effect, resulting from the combined effect of the antioxidant capacity of Mn ions and the efficient conversion of NMN to NAD⁺. At the gene expression level, the RT-qPCR results showed that the HT(sen) group exhibited lower expression of key osteogenesis-related genes (OCN and BSP) than the HT(nor) group, indicating impaired osteogenic differentiation under senescent conditions. The expression of OCN and BSP was upregulated in the HT-MnT(sen) group, while these genes were significantly upregulated in the HT-MnT/NMN(sen) group (Fig. 5e), further confirming that the combination of Mn ion-mediated antioxidant and NMN-induced NAD⁺ elevation effectively promotes osteogenic differentiation. These findings demonstrate that combining energy metabolism restoration with redox modulation produces a significant effect in promoting osteogenic differentiation.

Fig. 5.

Scaffold-mediated regulation of osteogenic differentiation in BMSCs. Normal BMSCs were included as the control group in all experiments. (a, c) Representative ALP and ARS staining images (including both micrographs and smartphone-captured whole-well photos) with quantification at 7 and 14 days in osteogenic differentiation medium. (b, d) Immunofluorescence images and quantification of osteogenesis-related proteins COL-1 and OCN at 7 days. (e) Expression of osteogenesis-related genes at 7 days. Data are presented as mean ± SD, n = 3. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns, not significant, as determined by one-way ANOVA analyses with Tukey's post-hoc test for all comparisons.

2.5. Transcriptomic analysis and mechanistic verification

Transcriptomic analysis further elucidated the underlying action mechanisms. Differentially expressed genes (DEGs) were identified using the thresholds of |log₂ FC| ≥ 2 and adjusted p-value ≤0.05. Regarding comparison between the HT-T and HT-MnT/NMN groups, volcano plots displayed 1940 upregulated and 1160 downregulated DEGs (Fig. 6a). Heatmap analysis showed a clear separation of gene expression profiles between the HT-T and HT-MnT/NMN groups, indicating that the HT-MnT/NMN treatment substantially reshaped gene expression of senescent BMSCs (Fig. 6b). The Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis indicated that the DEGs were predominantly enriched in pathways related to cell cycle, cellular senescence, energy metabolism (OXPHOS process) and osteogenesis (focal adhesion, regulation of actin cytoskeleton), with several additional pathways, including HIF-1, ErbB, FoxO, and mTOR signaling pathways, also being involved in cellular metabolism, stress response, and osteogenic regulation (Fig. 6d). Gene ontology (GO) enrichment analysis revealed a significant enrichment trend of DEGs in cellular metabolism-related pathways, confirming again the energy metabolism regulation ability of the HT-MnT/NMN scaffold (Fig. 6c).

Fig. 6.

Transcriptomic analysis and mechanistic verification of scaffold-mediated regulation in senescent BMSCs. (a) Volcano plot of differentially expressed genes (DEGs) between HT-T and HT-MnT/NMN-treated senescent BMSCs. (b) Heatmap of DEGs. (c, d) GO and KEGG enrichment analyses of DEGs. (e) GSEA of the FoxO pathway. (f) KEGG analysis of FoxO pathway-related DEGs. (g) Expression of metabolism-, senescence- and osteogenesis-related proteins following scaffold treatment and after FK866 adding. (h) Effects of FoxO transcriptional activity inhibitor on upstream and downstream protein expression. (i) Schematic diagram of the underlying action mechanisms linking scaffold-mediated metabolic regulation to antioxidant defense, senescence attenuation, and osteogenic recovery.

Based on these transcriptomic findings, we proposed that the HT-MnT/NMN scaffold first promotes mitochondrial metabolic reprogramming in senescent BMSCs. Given that the transcriptomic results highlighted energy metabolism-related pathways, we next examined whether this metabolic modulation was causally dependent on NAD⁺ replenishment from NMN. To this end, FK866 was employed to block the conversion of NMN to NAD⁺ for disrupting cellular energy metabolism (Fig. 6g) [31]. FK866 markedly attenuated the HT-MnT/NMN-induced upregulation of mitochondrial metabolic markers, including OGDH and NDUFA4L2, and osteogenic markers including COL-I and OCN, while partially reduced the senescence- and DNA damage-related markers, including p21 and γ-H2AX. These results demonstrate that the disruption of NAD⁺ homeostasis weakens the metabolic, anti-senescence, and pro-osteogenic effects of the HT-MnT/NMN scaffold, supporting a causal role of NAD⁺ restoration in these biological outcomes. Notably, FK866 reduced but did not completely abolish the protective effects of the HT-MnT/NMN scaffold, suggesting that Mn ions may provide additional antioxidant, anti-senescence, and pro-osteogenic actions independent of the NAD⁺ salvage pathway.

Following FK866-based validation of NAD⁺-dependent metabolic remodeling, we further investigated the signaling mechanism linking metabolic regulation to the anti-senescent, antioxidant, and pro-osteogenic effects of the HT-MnT/NMN treatment. Given that NAD⁺-regulated energy metabolism and redox homeostasis are closely associated with FoxO signaling, and that the FoxO transcription factors are involved in cellular stress resistance, senescence regulation, mitochondrial metabolic adaptation, antioxidant defense, osteoblast differentiation, and bone homeostasis [32,33], we focused on the FoxO pathway for subsequent analysis. Gene set enrichment analysis (GSEA) revealed significant upregulation of the FoxO signaling pathway in the HT-MnT/NMN group (Fig. 6e), suggesting that the FoxO-related signaling was engaged after the scaffold treatment. To further identify the metabolism-related signaling nodes associated with FoxO pathway regulation, we performed KEGG enrichment analysis of FoxO-associated DEGs identified from the transcriptomic dataset. This analysis revealed significant enrichment of the AMPK signaling pathway (Fig. 6f). This result is consistent with previous studies showing that the AMPK/FoxO axis plays an important role in regulating oxidative stress resistance, cellular senescence, mitochondrial homeostasis, and bone metabolism [34,35]. As a metabolism-sensitive kinase, activated AMPK can phosphorylate FoxO transcription factors and thereby regulate the FoxO-dependent transcriptional responses [36]. Based on these transcriptomic findings, we further examined the AMPK/FoxO axis by Western blotting (Fig. 6h). The HT-MnT/NMN treatment increased the levels of p-AMPK and p-FoxO, indicating the activation of AMPK/FoxO-related signaling under our experimental conditions. Moreover, the increased phosphorylation of AMPK and FoxO was accompanied by altered expression of FoxO-associated downstream functional markers, including upregulation of catalase (CAT) and COL-1 and downregulation of p21.

More importantly, we introduced AS1842856, a FoxO transcriptional activity inhibitor, to determine whether the FoxO activity is functionally involved in the regulatory effects of the HT-MnT/NMN treatment. AS1842856 did not markedly alter the levels of p-AMPK or p-FoxO, which is mechanistically reasonable because AS1842856 primarily inhibits FoxO transcriptional activity rather than blocking upstream AMPK activation or FoxO phosphorylation [37]. However, AS1842856 attenuated the downstream antioxidant, anti-senescent, and osteogenic effects induced by the HT-MnT/NMN treatment, as reflected by the decreased CAT and COL-1 expression and increased p21 expression (Fig. 6h). These findings suggest that the FoxO transcriptional activity acts as an important downstream effector linking scaffold-mediated metabolic regulation to antioxidant defense, senescence attenuation, and osteogenic recovery. Accordingly, a potential regulatory mechanism is proposed here: (i) HT-MnT/NMN scaffold promotes mitochondrial metabolic reprogramming; (ii) such metabolic remodeling is associated with the activation of AMPK signaling; (iii) AMPK activation is accompanied by increased FoxO phosphorylation; (iv) FoxO transcriptional activity contributes to upregulating antioxidant and anti-senescence responses; (v) this signaling cascade finally enhances osteogenesis (Fig. 6i). Collectively, the HT-MnT/NMN-mediated improvement of mitochondrial energy metabolism may serve as an upstream regulatory event that engages the FoxO-related signaling cascade and contributes to the antioxidative, anti-senescent, and pro-osteogenic effects observed in senescent BMSCs.

2.6. HT-MnT/NMN scaffolds promote aged bone defect repair in vivo

An 18-month-old SD rat model with femoral condyle defect was established to evaluate the aged bone repair ability of the HT-MnT/NMN scaffold, with 6-week-old normal SD rats were implanted with 3D-printed TCP scaffolds as the normal control group. Rats were euthanized 8-week post-implantation, and the defect areas were analyzed using high-resolution micro-computed tomography (micro-CT). The results showed that, compared to the blank, HT-T and HT-MnT groups, the HT-MnT/NMN group exhibited significantly increased new bone volume, improved bone microstructural integrity, and enhanced bone healing (Fig. 7a). Quantitative micro-CT analysis revealed that at 8 weeks, the normal control group exhibited the highest bone volume fraction (BV/TV) and superior trabecular microarchitecture among all groups, serving as a benchmark for healthy bone regeneration. Importantly, micro-CT images showed partial degradation of the central ceramic rods within the scaffold, indicating that this outer hollow tube-inner ceramic rod architecture could maintain early structural support while progressively generating space for tissue infiltration and new bone ingrowth. In the aged rat model, the HT-MnT group showed significant improvements in parameters such as bone volume fraction (BV/TV) compared to the blank and HT-T group. In contrast, the HT-MnT/NMN group exhibited the most robust bone-regeneration outcome among all aged groups, demonstrating the highest BV/TV, the greatest trabecular number (Tb·N), the largest trabecular thickness (Tb·Th), and the smallest trabecular separation (Tb·Sp). Notably, these morphometric parameters in the HT-MnT/NMN group approached those of the normal control group, indicating that the HT-MnT/NMN scaffold restored aged bone to a near-healthy regenerative state (Fig. 7d). Histological analysis of decalcified specimens using hematoxylin-eosin (H&E) and Masson's trichrome staining further confirmed that the HT-MnT/NMN group formed denser, well-organized trabecular structures and more mature new bone tissue compared to other aged groups. At the defect margins, H&E staining showed that newly formed bone grew from the original host bone edges into the peripheral regions of the scaffold (Fig. 7b). Consistently, Masson's trichrome staining revealed collagen-rich newly formed bone matrix surrounding and extending into the scaffold region, further demonstrating marginal-to-internal bone ingrowth and matrix maturation (Fig. 7b). The newly formed bone architecture in the HT-MnT/NMN approaching that of the normal control group, although it did not fully reach the level of the normal controls, consistent with the micro-CT images. Taken together, these results demonstrate that, based on the enhanced effect of the antioxidant function of Mn ions and the NAD⁺-elevating effect of NMN, the HT-MnT/NMN scaffold effectively improved the regenerative microenvironment in aged bone defects, thereby accelerating the bone healing process, which could be close to the healthy young bone.

Fig. 7.

In vivo aged bone regeneration-promoting ability of the scaffolds. (a) Micro-CT images of rat bone defects obtained 8 weeks after implantation (the green, red and white color in 3D Micro-CT images represented newly-formed bone, scaffolds and primary bone, respectively). (b) Representative H&E staining and Masson staining images of bone defect. (HB represents host bone, S represents scaffold and NB represents new bone.) (c) Immunofluorescent images of COL-1, and p21 expression of aged bone defect tissue. (d) Quantification of CT-derived bone volume fraction (BV/TV; %), trabecular number (Tb.N; mm⁻¹), trabecular thickness (Tb.Th; mm), trabecular separation (Tb.Sp; mm), n = 4. (e) Quantification of COL-1, and p21 expression of aged bone defect tissue. Data are represented as mean ± SD. n = 3. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, as determined by one-way ANOVA analyses with Tukey's post-hoc test for all comparisons.

Immunofluorescence staining revealed significantly enhanced expression of COL-1 and NDUFA4L2 in the HT-MnT/NMN group, with quantitative analysis confirming statistically significant intergroup differences (Fig. 7c–e and Fig. S16). These findings suggest that the HT-MnT/NMN scaffold promotes recovery of energy metabolism and enhances osteogenesis in bone tissue. Conversely, the expression levels of the DNA damage marker γ-H2AX and the senescence-associated protein p21 were significantly downregulated in the HT-MnT/NMN group, indicating enhanced DNA damage repair and effective alleviation of cellular senescence in bone tissue. Collectively, the results demonstrate that the HT-MnT/NMN scaffold possesses excellent anti-senescence and osteogenesis-promoting capabilities in vivo. By effectively counteracting senescent BMSCs senescence and promoting bone regeneration under senescent conditions, this scaffold offers a potential therapeutic strategy for the repair of aged bone defects.

In addition to evaluating bone regeneration, the in vivo biosafety of the HT-MnT/NMN scaffold was further assessed. As shown in Fig. S17, complete blood count parameters showed no significant differences among the sham (surgery only, no bone defect, no scaffold), blank (bone defect building without scaffold implantation), and scaffold treatment groups, indicating that the scaffold implantation did not induce detectable hematological abnormalities. Serum biochemical analysis showed that most liver- and kidney-function-related markers, including albumin (ALB), alanine aminotransferase (ALT), creatinine (CREA), and urea (UREA), exhibited no obvious adverse changes after scaffold implantation. Although alkaline phosphatase (ALP) was affected after model establishment, this change was likely associated with surgery-induced systemic responses and bone remodeling rather than scaffold-induced hepatic or renal toxicity. Importantly, no significant differences were observed between the blank and scaffold groups for these biochemical markers, suggesting that the scaffold itself did not induce detectable liver or kidney dysfunction. H&E staining of major organs, including the heart, liver, spleen, lungs, and kidneys, showed no obvious structural abnormalities, inflammatory infiltration, fibrosis, or necrosis in the HT-MnT/NMN group (Fig. S17). In addition, the hemolysis assay showed that the hemolysis rate of the scaffold was 0.8%, markedly lower than the commonly accepted safety threshold of 5%, suggesting that the scaffold exhibited good hemocompatibility (Fig. S18). These results indicate that the HT-MnT/NMN scaffold implantation did not induce detectable hematological abnormalities, liver/kidney dysfunction, hemolytic toxicity, or major organ damage, which is favorable for clinical translation.

3. Conclusion

Metabolic imbalance and oxidative stress in the senescent microenvironment severely hinder aged bone repair—a challenge that conventional biomaterials cannot fundamentally address. In this study, a senescent metabolism-modulating hierarchical bioceramic scaffold (HT-MnT/NMN) was developed via 3D printing and freeze-casting, enabling sustained co-release of NMN and Mn ions to simultaneously target two core pathological features of aged bone: NAD⁺ deficiency and excessive ROS accumulation. NMN restores intracellular NAD⁺ levels to ameliorate senescence-impaired mitochondrial structure and function, thereby boosting energy metabolism, while Mn ions regulate redox homeostasis through reducing ROS accumulation and enhancing mitochondrial MnSOD activity. Based on these effects, the HT-MnT/NMN scaffold effectively mitigates senescent BMSCs senescence, reduces inflammation, thus improving the local senescent bone microenvironment to enhance osteogenesis. Mechanistically, FoxO-related signaling may functionally link this scaffold-mediated metabolic reprogramming to antioxidant, anti-senescent, and osteogenic responses. These benefits were validated in an aged rat femoral condyle defect model, where the scaffold improved the bone metabolic microenvironment and significantly promoted new bone formation. This work proposes a novel “material-mediated metabolic reprogramming” strategy for aged bone repair, providing an effective bioactive scaffold that targets the root causes of impaired regeneration in aging. Beyond its translational potential for clinical treatment of age-related bone defects, this study also establishes a paradigm for treating tissue senescence through metabolic regulation, opening new avenues in regenerative medicine for age-associated diseases.

Ethics approval and consent to participate

All the animal assays were approved by the Institutional Animal Care and Use Committees of Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University (approval number: DWSY2022-0018).

CRediT authorship contribution statement

Jiamin Jiang: Conceptualization, Data curation, Formal analysis, Investigation, Project administration, Validation, Visualization, Writing – original draft. Jiajie Chen: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Supervision, Writing – review & editing. Xiao Wang: Investigation, Validation. Xinxin Zhang: Investigation, Validation. Hongxu Lv: Formal analysis, Funding acquisition, Methodology, Supervision. Chengtie Wu: Funding acquisition, Resources, Supervision. Lei Han: Formal analysis, Methodology, Resources, Supervision, Writing – review & editing. Yufang Zhu: Conceptualization, Formal analysis, Funding acquisition, Methodology, Resources, Supervision, Writing – review & editing.

Declaration of competing interest

Chengtie Wu is an editorial board member for Bioactive Materials and was not involved in the editorial review or the decision to publish this article. All authors declare that there are no competing interests.

Appendix A. Supplementary data

The following is the Supplementary data to this article: