NIR-programmable 3D-printed shape-memory scaffold with dual-thermal responsiveness for precision bone regeneration and bone tumor management

Abstract

Clinically, intraoperative treatment of bone tumors presents several challenges, including the effective inactivation of tumors and filling of irregular bone defects after tumor removal. In this study, intelligent thermosensitive composite materials with shape-memory properties were constructed using polylactic acid (PLA) and polycaprolactone (PCL), which have excellent biocompatibility and degradability. Additionally, beta-tricalcium phosphate (β-TCP), with its osteogenic properties, and magnesium (Mg) powder, with its photothermal and bone-promoting abilities, were incorporated to improve the osteogenic potential of the composite and enable the material to respond intelligently to near-infrared (NIR) light. Utilizing 3D printing technology, the composite material was prepared into an NIR-responsive shape-memory bone-filling implant that deforms when the scaffold temperature increases to 48 ℃ under NIR laser irradiation. Moreover, at a lower temperature of 42 ℃, mild photothermal therapy promotes macrophage polarization toward the M2 phenotype. This process regulates the secretion of interleukin (IL)-4, IL-10, tumor necrosis factor-α, IL-6, and bone morphogenetic protein (BMP)-2, reducing local inflammation, enhancing the release of pro-healing factors, and improving osteogenesis. Overall, this innovative scaffold is a promising and efficient treatment for filling irregular bone defects after bone tumor surgery.

Graphical abstract

Introduction

Bone tumors—such as osteosarcoma, fibrosarcoma, and chondrosarcoma—are common malignant neoplasms in adolescents, causing long-term suffering for patients and imposing a considerable financial burden on their families [1, 2]. A common treatment modality includes surgically removing the localized bone tumor and filling the defects or cavities with bone grafts or bone void fillers, followed by the administration of chemotherapeutic agents [3,4,5]. However, for large and irregular bone defects, considerable challenges arise in achieving complete filling of the defect cavity using the selected bone implants [6]. Moreover, incomplete resection of tumor lesions during surgery can lead to residual tumor cell proliferation, promoting tumor recurrence and metastasis [7].

Consequently, the suboptimal osseointegration performance of bone implants and challenges regarding bone regeneration require further investigation and optimization. Specifically, to improve the therapeutic outcomes for osteosarcoma, a multifunctional scaffold with bone-remodeling and tumor eradication capabilities is needed [8,9,10]. 3D printing technology has emerged as a pivotal tool in the creation of bone defect filling scaffolds following bone tumor surgery. Its precise control over scaffold dimensions has significantly impacted the field of bone tissue engineering. The 3D printed shape memory scaffold, thanks to its unique shape memory properties, exhibits superior filling performance [11, 12]. Building on this foundation, 3D-printed scaffolds that possess both thermosensitivity and shape memory properties can be implanted minimally invasively through small incisions. Subsequently, triggered by body temperature (or a specific temperature range of 37–48 °C), they expand to reduce surgical trauma while ensuring intimate contact with irregularly shaped defect surfaces [13].

Currently, intelligent memory materials with shape-memory effects include shape-memory ceramics (SMCs), shape-memory alloys (SMAs), shape-memory hydrogels (SMHs), and shape-memory polymers (SMPs) [14]. Owing to their notable biodegradability and adjustable mechanical strength, SMPs and SMHs show considerable promise in bone tissue engineering applications [15, 16]. Specifically, polylactic acid (PLA), which is widely recognized for both its shape memory properties and its biodegradability, is particularly well-suited for the fabrication of scaffolds in bone tissue engineering applications [17]. These properties allow it to reverting toward a pre-programmed shape when exposed to specific triggers, such as changes in temperature, which is highly beneficial for creating adaptive scaffolds that can respond to the dynamic environment of the body [18]. However, due to its relatively high glass transition temperature, which can be around 60℃, pure PLA may not be optimal for all applications in bone tissue engineering where a more flexible material is required [19]. To address this, poly-ε-caprolactone (PCL), which has a lower glass transition temperature and offers greater flexibility and toughness, is often blended with PLA [20]. This combination leverages the strengths of both materials, providing a balance between the rigidity and biodegradability of PLA and the flexibility and ductility of PCL, making the composite material more suitable for bone tissue engineering applications where adaptability and flexibility are crucial [21,22,23].

To enhance the osteogenic performance of the scaffold, hydroxyapatite (HA) [24], magnesium (Mg) [25], and beta-tricalcium phosphate (β-TCP) are ingeniously integrated into the materials to fabricate bone tumor postoperative filling scaffolds designed to promote bone tissue growth. Owing to its remarkable osteogenic promotion effects, HA and β-TCP has also become a vital material in the field of bone defect repair [26]. Meanwhile, incorporating β-TCP into PLA can promote osteogenesis and partially neutralize the acidic environment produced during PLA degradation [27]. The addition of Mg to the composite significantly promotes bone formation. The continuous release of Mg²⁺ can boost alkaline phosphatase (ALP) activity and upregulate the expression of osteogenic-related genes, such as osteocalcin (OCN), type I collagen (Col I), runt-related transcription factor 2 (RUNX-2), and bone morphogenetic protein 2 (BMP-2), significantly promoting osteogenesis [28]. Furthermore, Mg²⁺ in Mg oxide/poly(lactide-co-glycolide) scaffolds prepared by 3D low-temperature printing can enhance bone repair by activating the Wnt3a/GSK-3β/β-catenin signaling pathway [29]. Moreover, Mg can enhance osteogenesis by stimulating the polarization of macrophages towards the M2 phenotype [30, 31]. Additionally, due to their excellent biocompatibility and photothermal conversion efficiency, Mg holds significant value in photothermal therapy (PTT), where it can effectively achieve precise elimination of tumor cells under NIR irradiation [32].

Near-infrared (NIR) PTT has emerged as a pivotal technique in bone tumor treatment due to its exceptional tissue penetration depth, precise temperature modulation capabilities, and the ability of NIR radiation to induce deformation in shape-memory scaffolds [2, 33,34,35]. The therapeutic mechanism relies on photothermal conversion: when irradiated by NIR lasers, specific nanomaterials (e.g., iron nanoparticles, black phosphorus nanosheets, and graphene oxide) and Mg-doped metal composite efficiently transform light energy into localized heat [9, 32, 36]. This process elevates tissue temperatures to 45–48 °C, inducing rapid tumor cell necrosis through thermal ablation [37]. Notably, temperature regulation exhibits dual biological effects: temperatures exceeding 48 °C effectively eradicate malignant cells, while maintaining a mild thermal environment at 42 °C promotes macrophage polarization toward the M2 phenotype. This immunomodulation optimizes the bone microenvironment, thereby enhancing the osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) [38]. Building upon this temperature-responsive paradigm, researchers have developed multifunctional bone-repair material systems. 3D-printed composite scaffolds demonstrated dual functionality by suppressing bone tumors through photothermal effects while stimulating bone regeneration via controlled thermal release [4, 34, 39]. However, conventional 3D-printed scaffolds face limitations in adapting to extensive irregular bone defects [40]. To address this challenge, we innovatively integrated shape-memory materials with 3D printing technology to fabricate a PLA/PCL/TCP/Mg quaternary composite smart scaffold. Following minimally invasive implantation, this scaffold exhibits controllable NIR-responsive deformation: at 48 °C, it undergoes volumetric expansion to precisely fill irregular defect cavities while eliminating residual tumor tissues through thermal energy release. Postoperatively, modulating the NIR device power to maintain 42 °C sustains a pro-osteogenic microenvironment for bone regeneration (Scheme 1). This dynamic responsiveness overcomes the physical constraints of static scaffolds, establishing a novel strategy for personalized bone reconstruction.

Scheme 1

Using 3D printing technology, a composite material comprising PLA, PCL, β-TCP, and Mg was printed into a shape-memory bone filler implant that responds to near-infrared (NIR) irradiation and undergoes deformation. Under NIR laser irradiation, the temperature of the scaffold increases to 48 °C, causing it to deform and exhibit high-temperature antitumor properties. Additionally, when the temperature reaches 42 °C during NIR-induced mild photothermal therapy, the scaffold induces macrophage polarization toward the M2 phenotype and regulates the secretion of functional cytokines, enhancing osteogenesis

Results and discussion

Preparation and functional verification of PLA/PCL/TCP/Mg composites and dual-functional scaffolds

Actually, the limitations of filling irregular scaffolds in the medical field can be overcome by applying scaffolds fabricated by 3D printing techniques employing shape-memory polymers. Among the numerous shape-memory polymers, PLA and PCL have been widely employed owing to their biocompatibility and degradation performance. Moreover, the original T_g can be altered by mixing PLA and PCL. In the current study, the T_g of the composite material gradually decreased with increasing PCL content (Figure S1). The T_g of the mixed material containing 20% PCL reached approximately 45 ℃, meeting the material requirements, and was thus used in all subsequent assays. Additionally, the selected PLA containing 20% PCL, 10% TCP, and different Mg concentrations (2.5, 5, 7.5 and 10%) was incorporated to enhance the material’s osteogenic ability and endow it with photothermal conversion capabilities. Then, DSC testing was used to verify the Tg of five composite materials (Figure S2). The result validated that the incorporation of β-TCP and a modest quantity of Mg powder in our study had exerted a negligible impact on the Tg of the composite materials. This outcome was consistent with prior research finding [41].

The 3D printed scaffolds with the composite materials were shown in Fig. 1A. As the quantity of Mg increased, the color of the resulting scaffolds gradually darkened. Moreover, the super-depth microscopy and contact angle testing revealed that the scaffold surface roughness progressively increased with increasing Mg content (Figs. 1B–C). SEM results revealed that, with increasing Mg content, the scaffold surface exhibited gradual roughening (Figure S3). The line width in all five groups of scaffolds were approximately 500 μm. EDX elemental mapping (Figure S4) was employed to assess the presence of P, Ca, and Mg, confirming the successful incorporation of β-TCP and Mg. Moreover, the contact angle measurements revealed a corresponding increase in the hydrophobicity of the scaffold surface with augmented Mg content, enhancing the overall hydrophobic properties (Figure S5). For bone-filling materials, scaffolds with optimal surface roughness can enhance stem cell adhesion and promote cell proliferation and differentiation [42]. Moderate surface roughness facilitates the integration of the scaffold with the surrounding tissues and supports the bone formation process, thereby benefiting bone tissue regeneration and repair [43].

Fig. 1

Characterization of 3D-printed scaffolds. (A) Images of different samples. (B) Super-depth microscopic images. (C) Contact angles of the five groups of 3D-printed scaffolds. (D) Comparison of the deformation rates of shape memory scaffolds prepared from five different composite materials

A universal testing machine was used to compare the mechanical properties of the different scaffold groups and their compression performance. The compressive modulus of the PLA 10%Mg composite reached 3.52 ± 0.32 GPa, which was significantly higher than that of the PLA group (2.17 ± 0.36 GPa), the PLA 2.5%Mg group (2.31 ± 0.25 GPa), and the PLA 5%Mg group (2.52 ± 0.42 GPa). Although no statistically significant difference (p > 0.05) was observed between the compressive modulus of the PLA 7.5%Mg group (2.79 ± 0.15 GPa) and that of the PLA 10%Mg group, the experimental data demonstrated a clear trend of enhanced mechanical properties in the composite scaffolds with increasing Mg content. These results confirmed that the compression modulus gradually increased with increasing Mg content in the composite material (Figure S6). This was attributed to the excellent interfacial bonding between Mg powder and the PLA. Once dispersed within the PLA, Mg particles could serve as a medium for stress transfer, more evenly distributing external forces into the PLA and thereby enhancing the overall strength of the material [44]. Based on the roughness and mechanical performance results, selecting composites with slightly higher Mg content was deemed prudent. Indeed, the Mg-doped materials release Mg²⁺ during the early inflammatory phase that facilitate the recruitment and polarization of monocyte-macrophages, thereby constructing an environment conducive to bone tissue regeneration. However, in the later stages of bone reconstruction, the continuous and excessive release of Mg²⁺ may activate osteoclast activity and inhibit osteoblast activity, significantly suppressing the mineralization and maturation of bone tissue [45].

FTIR analysis was conducted to investigate the chemical structural characteristics of five composite materials (Figure S7). The results revealed several key absorption peaks: the C = O ester carbonyl group in PLA exhibited a characteristic absorption peak at 1751 cm⁻¹, while the absorption peak at 1465 cm⁻¹ was attributed to the -CH₂ group in PCL [46]. Additionally, the phosphate group (PO₄³⁻) of β-TCP displayed a significant characteristic absorption peak at 1094 cm⁻¹, confirming the successful incorporation and uniform dispersion of β-TCP within the composite materials [47]. Notably, a characteristic vibration peak corresponding to the Mg-O bond was detected at 503 cm⁻¹. This peak was observed in all Mg-containing experimental groups, except for the ACP control group that lacked added Mg powder [48]. Considering the thermodynamic properties of Mg powder, which is highly susceptible to oxidation at room temperature [49]. It can be inferred that during the processing of the composite materials, some of the Mg powder had already reacted with atmospheric oxygen to form Mg oxide.

Next, the shape-memory recovery rate of several scaffolds was evaluated (Fig. 1D). After shaping the scaffolds into the five groups at 48 °C, they were reheated to the same temperature for assessment. Marked differences were not observed in the recovery of each scaffold group, indicating that the inclusion of Mg did not exert noticeable effects on the shape-memory performance of the scaffolds.

PTT is considered a highly promising method for cancer treatment and accelerating bone repair due to its remote controllability, noninvasiveness, excellent therapeutic efficacy, and deep tissue penetration [50]. Consequently, the photothermal conversion efficiency of scaffolds under NIR laser irradiation is crucial for tumor treatment. Accordingly, in the current study, the photothermal conversion performance of the Mg-doped composite materials was assessed using a 1064-nm laser as the NIR laser irradiation source. Under NIR laser irradiation (1064 nm, 1.5 W/cm²), the temperatures of the five scaffold groups increased rapidly, reaching equilibrium within 30 s (Figs. 2A–B). Notably, the ACP/7.5Mg scaffold temperature rose to 62.8 °C within 30 s. To achieve the ideal tumor destruction temperature of 48 °C [36]. the power of the 1064-nm NIR laser was adjusted to 0.5, 1.0 and 1.5 W/cm² to evaluate the photothermal conversion efficiency of the ACP/7.5Mg scaffolds. When the laser power was set to 1.0 W/cm², the equilibrium temperature of the scaffolds reached approximately 48 °C (Fig. 2C), which was suitable for tumor ablation. In contrast, at a lower power of 0.5 W/cm², the equilibrium temperature was approximately 42 °C, a range known to promote bone repair [51].

Fig. 2

Photothermal conversion characteristics of scaffolds under NIR laser irradiation. (A) Infrared thermographic images of the five scaffold groups under NIR laser irradiation for various exposure times. (B) Temperature changes in the five scaffold groups under NIR laser irradiation (1064 nm, 1.5 W/cm²). (C) Photothermal heating curves of the ACP/7.5Mg scaffolds under NIR laser irradiation at different powers. (D) Temperature variation of the ACP/7.5Mg scaffolds under repeated on-off NIR laser irradiation (1064 nm, 1.0 W/cm²)

To further assess the photothermal stability of the ACP/7.5Mg scaffolds under NIR laser irradiation (1064 nm, 1.0 W/cm²). After four consecutive irradiation cycles under 1064-nm NIR light exposure, the scaffold surface maintained stable maximum temperatures around 48 °C, showing negligible attenuation compared with the initial irradiation cycle. This performance demonstrated exceptional photothermal cycling stability (Fig. 2D). The curves for the ACP/7.5Mg scaffolds exhibited minimal attenuation across each cycle. Mg was an effective photothermal agent [32], effectively converting light energy into heat upon exposure to NIR laser irradiation (1064 nm).

Biocompatibility verification

The biocompatibility of each scaffold group was assessed using live/dead and CCK-8 assays and FITC phalloidin staining. The cell survival rates on the surfaces of the scaffolds in each group were nearly identical (Figures S8A–B).

FITC phalloidin staining revealed that increasing the Mg content in the scaffold enhanced BMSC flattening on the surface (Figure S9A). This can be attributed to the increased surface roughness of the scaffold resulting from the higher Mg content, promoting improved BMSC adhesion and proliferation. Additionally, CCK-8 staining (Figure S9B) confirmed that the proliferation performance of BMSCs cultured with the scaffolds was not impacted by the scaffold Mg content.

Verification of the BMSCs osteogenic differentiation

ALP and Alizarin Red staining were performed to assess the osteogenic-promoting ability of scaffolds in each group (Figs. 3A). As the Mg content in the scaffolds increased, the ability to promote osteogenesis gradually increased. However, the osteogenic effect of the ACP/7.5Mg group was higher than that of the ACP/10Mg group. Figure 3B demonstrated a clear correlation between the Alizarin Red staining intensity and the Mg mass fraction. In the ACP, ACP/2.5Mg, ACP/5Mg, ACP/7.5Mg, and ACP/10Mg groups, the intensities of alizarin red staining were determined as 4.96 ± 0.65%, 9.31 ± 0.61%, 19.45 ± 0.91%, 29.70 ± 0.56%, and 21.52 ± 0.73%, respectively. As the mass fraction of Mg incrementally rose from 0 to 7.5, the intensity of alizarin red staining continuously ascended (P < 0.1). The ACP/7.5Mg group showed peak staining intensity (29.70 ± 0.56%), approximately 6-fold higher than the ACP group. However, the ACP/10Mg group exhibited a marked decline in staining intensity (70% of ACP/7.5Mg group), representing a 27.5% reduction compared to the peak level. Overall, these findings indicated that the osteogenic efficacy of the scaffolds in this study were significantly regulated by Mg²⁺ concentration [52], and the ACP/7.5Mg group exhibited the best osteogenic ability. As demonstrated by previous studies, Mg²⁺ at optimal concentrations effectively promote osteogenesis, whereas excessively high concentrations significantly inhibit the functional activity of osteoblasts [53, 54].

Fig. 3

Validation of osteogenic performance of BMSCs on scaffold surfaces. (A) ALP staining of the BMSCs on day 14 and Alizarin red staining of the BMSCs on day 21. (B) Quantitative analysis of calcium nodules. (C–E) Relative RUNX-2, ALP, and OCN mRNA expression in BMSCs on days 7 and 14 (n = 3)

RT-qPCR analysis further validated these results; the trends in RUNX-2, ALP, and OCN expression in BMSCs were consistent with those observed via Alizarin Red and ALP staining (Figs. 3C–E). Thus, moderate levels of Mg promoted bone regeneration, whereas high concentrations weakened or inhibited the osteogenic effects. In this study, Mg was uniformly mixed with a PLA and PCL composite material to slow the Mg degradation rate within the material and promote bone formation [55].

Design and selection of scaffold structures

A curved intersecting lattice structure was engineered to fully exploit the advantages of shape-memory scaffolds. This enabled the scaffold to compress unidirectionally while preserving a consistent width during in-vitro shaping. Scaffolds with shape-memory properties were printed at 40, 50 and 60% fill rates, and their surfaces were examined via SEM (Figs. 4A–B). As the fill rate increased, the scaffolds demonstrated a progressive reduction in pore size along with decreased deformation. The internal cross-linked network structure effectively maintained consistent dimensional stability in the perpendicular direction during mechanical deformation, regardless of fill rate percentage. This design refinement enhanced their applicability in minimally invasive implantation procedures. Mechanical testing confirmed that the compressive modulus improved as the fill rate increased (Figure S10). To evaluate the shape-memory performance of the scaffold, a structure with a 50% fill rate was submerged in water at 50 °C, and a 50 g weight was placed on it. It was observed that, under these temperature conditions, the scaffold was able to rapidly revert to its original shape while supporting the 50 g load at 50 °C (Video 1).

Fig. 4

Screening of scaffolds with different structures. (A) Deformation situation with different filling rates. (B) The SEM images of the scaffolds with different filling rates. (C) Live/ dead staining of BMSCs on the surface of the scaffolds with different internal structures. (D) FITC phalloidin staining of BMSCs on the surface of the scaffolds with different internal structures

Additionally, the comparative cell viability and FITC-conjugated phalloidin staining of the scaffold surfaces with linear and curved cross-linked structures (Figs. 4C–D) demonstrated that the curved internal structures of the scaffolds were more conducive to cell adhesion and growth. Indeed, incorporating cross-linked conductive networks into the internal structure of the scaffolds enhances shaping and shape recovery [56]. In addition, existing research enlighten the capacity of scaffolds with varied geometric structures to influence cell behavior and function by modulating gene expression, which subsequently affected cell adhesion and proliferation. Building on this foundation, we posit that the intricate internal geometry of the curved cross-linked scaffolds designed in this study will foster a more conducive environment for cell adhesion and proliferation compared to linear scaffolds. This hypothesis is grounded in the belief that the complex architecture of curved cross-linked scaffolds could provide a richer surface for cell interaction, potentially enhancing the cellular response [57]. Hence, these scaffolds offer enormous potential for future medical applications when combined with those that promote osteogenic properties.

Effect of NIR laser irradiation on tumor cell eradication

Due to the excellent photothermal conversion effect exhibited by the Mg-containing scaffolds in this study, in vitro and in vivo experiments were conducted to evaluate the effectiveness of NIR laser irradiation for tumor removal. Under NIR laser irradiation (1064 nm, 1.0 W/cm²), the tumor cell mortality rate in the ACP/Mg NIR + group reached ~ 85%, whereas no significant changes were observed in the other groups (Figs. 5A–B).

Fig. 5

In vivo and in vitro validation of the effects of NIR laser irradiation (1064 nm, 1.0 W/cm²) on tumor cells. (A–B) Effect of NIR laser irradiation (1064 nm, 1.0 W/cm²) on the viability of 143B cells using live/dead staining (n = 3). (C) Representative infrared thermal image of the illuminated area under 1064-nm, 1.0-W/cm² irradiation after scaffold implantation under tumors. (D) Tumor volume changes over time across different groups following various treatments. (E) Representative images showing changes in tumor volume over time. (F) Changes in body weight of tumor-bearing mice after different treatments. (G) Representative H&E staining (n = 5)

Following in vivo scaffold implantation, NIR laser irradiation was applied daily for 10 days. Under NIR laser irradiation (1064 nm, 1.0 W/cm²) for 30 s, the scaffold temperature in the CON NIR + and ACP NIR + groups increased to 37.6 °C and 38.6 °C, respectively, before stabilizing (Fig. 5C). In contrast, the temperature of the ACP/Mg NIR + scaffolds rose to 47.9 °C after 30 s, stabilizing at 48.5 °C after 35 s.

After 10 days of irradiation, the tumor volume in the ACP/Mg NIR + group experienced a considerable 73.67% reduction, whereas the tumor volumes in the other groups progressively increased over time (Figs. 5D–E). H&E staining revealed pronounced tumor cell necrosis in the ACP/Mg NIR + group compared to the other groups (Fig. 5G). This may be attributed to Mg-free scaffolds lacking photothermal conversion capabilities and failing to generate heat under NIR laser irradiation, rendering them ineffective for tumor cell eradication. In contrast, the Mg-containing scaffolds demonstrated marked efficacy under NIR laser irradiation. Throughout the treatment period, the weight of the nude mice consistently increased (Fig. 5F). H&E staining further revealed that the tumors and subsequent treatments in each group did not elicit significant effects on the visceral organs (Figure S11).

Effect of NIR laser irradiation on macrophages

Mild NIR-triggered PTT can promote osteogenesis by inducing macrophage polarization toward the M2 phenotype, regulating the bone immune microenvironment [38]. During bone regeneration, the activated M2 macrophages are crucial in clearing debris, suppressing inflammation, and promoting osteogenesis [58,59,60]. Consequently, developing functional biomaterials with effective immune-modulatory and anti-inflammatory properties is important.

RAW264.7 cells served as the in vitro macrophage model, and LPS (a component of Gram-negative bacterial cell walls) was applied to simulate an acute inflammatory response and induce the polarization of macrophages toward the M1 phenotype. Subsequently, the effects of NIR laser irradiation (1064 nm, 0.5 W/cm²) on cell viability and proliferation were assessed using live/dead staining and CCK-8 assays. Our findings, as depicted in Figs. 6A–B and S12, indicated that NIR laser irradiation (1064 nm, 0.5 W/cm²) did not significantly impact the cell survival rate or macrophage proliferation.

Fig. 6

Regulatory effects of NIR laser irradiation (1064 nm, 0.5 W/cm²) on macrophages. (A–B) Effect of mild NIR laser irradiation (1064 nm, 0.5 W/cm²) on macrophage cell viability through live/dead staining. (C–F) Verification of the effect of NIR laser irradiation (1064 nm, 0.5 W/cm²) on macrophage polarization through immunofluorescence staining (G–H) Verification of the effect of NIR laser irradiation (1064 nm, 0.5 W/cm²) on the expression of CD86 and CD206 using RT-qPCR (n = 3)

The effect of smart scaffolds on macrophage polarization under NIR laser irradiation was validated using immunofluorescence staining. The expression of CD86 was markedly lower (Figs. 6C–D), and CD206 was higher in the ACP/Mg NIR + group than in the other five groups (Figs. 6E–F). The above results were consistent with previous research findings, confirming that NIR absorbing materials could effectively regulate macrophage polarization towards the M2 phenotype under photothermal stimulation [61, 62].

RT-qPCR results further confirmed the above results. As shown in Fig. 6G, Cd86 expression analysis showed no significant differences between CON NIR + group (1.05 ± 0.11) and CON NIR- group (0.99 ± 0.04), nor between ACP NIR+ (1.16 ± 0.12) and ACP NIR- (1.15 ± 0.08) (p > 0.05), indicating that NIR stimulation exerted no significant regulatory effect on Cd86 expression in Mg-free composite materials. Notably, the ACP/Mg NIR- group demonstrated a significant downregulation of Cd86 expression (0.790 ± 0.071) compared to both ACP (NIR+/-) and CON (NIR+/-) groups (p < 0.05), suggesting that the incorporation of Mg effectively inhibited the polarization of macrophages towards the M1 phenotype [55, 63]. Particularly, under combined NIR photothermal stimulation, the ACP/Mg NIR + group exhibited markedly reduced Cd86 expression (0.403 ± 0.081) compared to other experimental groups. This phenomenon confirmed the existence of a synergistic regulatory mechanism between Mg-containing scaffolds and NIR photothermal effects that effectively suppressed M2 macrophage marker Cd86 expression.

As illustrated in Fig. 6F, the expression level of the Cd206 gene exhibited an inverse relationship to that of Cd86. Specifically, the ACP/Mg NIR- group (1.48 ± 0.06) demonstrated a significant upregulation in Cd206 expression compared to the CON NIR + group (1.11 ± 0.06), CON NIR − group (1.00 ± 0.07), ACP NIR + group (1.10 ± 0.11), and ACP NIR − group (1.14 ± 0.10; p < 0.05). Notably, the Cd206 expression level in the ACP/Mg NIR + group (2.36 ± 0.11) was significantly higher than that of all other groups (p < 0.05). The above results indicated that the ACP/Mg scaffold combined with mild thermal stimulation at 41 ± 1 °C further promoted M2 macrophage polarization and inhibited M1 differentiation under NIR laser irradiation conditions [30]. This demonstrated the enormous potential of ACP/Mg NIR + scaffolds in shortening the inflammatory phase and transitioning toward the proliferative phase during bone regeneration.

To further evaluate the role of ACP/Mg and NIR therapy in regulating the immune response, RT-qPCR was performed to assess the expression of anti-inflammatory (Il-4 and Il-10) and pro-inflammatory (Il-6 and Tnf-a) mRNA in RAW264.7 macrophages (Figs. 7A–D). Additionally, the abundance of cytokines in the cell supernatants was evaluated via ELISAs. Under mild NIR laser irradiation, the ACP/Mg scaffolds increased the release of IL-10 and IL-4 and decreased that of TNF-α and IL-6 (Figs. 7E–H). NIR laser irradiation not only promotes the polarization of macrophages towards the M2 phenotype but also modulates the BMP-2 factor, thereby fostering osteogenesis [64]. The BMP-2 content in the supernatant of the ACP/Mg NIR + group was higher than that of the other groups (Figure S13). These results suggested that ACP/Mg scaffolds can simultaneously alleviate inflammation and polarize macrophages from the M1 to M2 phenotype under the synergistic effect of thermal stimulation provided by mild NIR laser irradiation, stimulating the release of osteogenic factors.

Fig. 7

Validation of the role of ACP/Mg and NIR therapy in regulating immune response. (A-D) The relative mRNA expression levels of inflammatory and pro-healing cytokines in RAW264.7 macrophages were analyzed. (E-H) The expression levels of a series of inflammatory and pro-healing cytokines in cell supernatants were evaluated using ELISA kits

Effect of NIR laser irradiation on the osteogenic activity of BMSCs in vitro

Although short-term exposure to high-temperature radiation can harm BMSCs, it does not directly affect long-term osteogenesis. However, bone formation can be regulated by adjusting the temperature [36]. The conditioned medium mentioned in Sect. ‘Effect of NIR Laser Irradiation on Macrophages’ was utilized to cultivate BMSCs, Thus, it is possible to evaluate the effect of ACP/Mg scaffolds on the osteogenic potential of BMSCs by modulating the immune microenvironment under mild NIR laser irradiation-induced thermal stimulation. Before these experiments, the effect of mild NIR light on the BMSCs activity was validated using live/dead staining and CCK-8 assays. As shown in Figures S14A–C, cell viability measurements demonstrated remarkable consistency across all experimental groups, with viability percentages as follows: CON NIR- (99.18 ± 0.51%), CON NIR+ (98.67 ± 0.25%), ACP NIR- (99.22 ± 0.10%), ACP NIR+ (98.82 ± 0.43%), ACP/Mg NIR- (99.06 ± 0.22%), and ACP/Mg NIR+ (99.27 ± 0.12%). Statistical analysis revealed no significant intergroup differences (p > 0.05), indicating mild NIR exposure at 41 ± 1 °C did not affect cell survival or proliferation as previous study reported [30].

The activity of the early osteogenic marker ALP was assessed using ALP staining and quantification. The most intense ALP staining was observed in the ACP/Mg NIR + group on days 7 and 14, followed by the ACP/Mg NIR − group (Fig. 8A). The osteogenic effects in the ACP NIR − and ACP NIR + groups were comparable to those in the CON NIR − and CON NIR + groups. This outcome may be attributed to the osteogenic properties of β-TCP in the ACP NIR − and ACP NIR + groups [65], with Mg also contributing to osteogenesis [52]. Additionally, mild NIR exposure at 41 ± 1 °C augmented the osteogenic effects of the ACP/Mg scaffold. This was further supported by the ALP quantitative results (Fig. 8B), demonstrating that the ALP production in the ACP/Mg NIR + group increased compared to the other groups. Consequently, the deposition of calcium minerals during the late stages of osteogenic differentiation was evaluated using Alizarin Red staining and quantification (Figs. 8C–D). The result was consistent with the ALP activity findings, with macroscopic and microscopic images showing that the ACP/Mg NIR + group had the most bone mineralization nodules, indicating better osteogenic potential.

Fig. 8

Validation of osteogenic performance of BMSCs on scaffold surfaces under NIR laser irradiation (1064 nm, 0.5 W/cm²). (A–B) Validation of the effect of mild NIR laser irradiation (1064 nm, 0.5 W/cm²) on osteogenic differentiation of the BMSCs through ALP staining. (C–D) Validation of the effect of mild NIR laser irradiation (1064 nm, 0.5 W/cm²) on osteogenic differentiation of the BMSCs through Alizarin red staining. (E–G) Effects of mild NIR laser irradiation on the relative mRNA expression levels of OCN, RUNX-2, and ALP in the BMSCs after 7 and 14 d (n = 3)

OCN, a non-collagenous protein secreted by osteoblasts, is an important component of the bone extracellular matrix (ECM) and plays a pivotal role as an osteogenic factor in the process of bone formation [66]. whereas RUNX-2 is a key transcription factor that regulates the expression of osteogenesis-related genes [67]. The ALP, OCN, and RUNX-2 expression in each BMSC group was validated by RT-qPCR (Figs. 8E–G); the results were consistent with those of the ALP and Alizarin Red staining and quantification. The abundance of OCN and RUNX-2 was further validated in each group of BMSCs by immunofluorescence staining. Figures 9A-B illustrated RUNX-2 immunofluorescence intensities among experimental groups: CON NIR+ (1.00 ± 0.04), CON NIR- (1.00 ± 0.03), ACP NIR+ (1.46 ± 0.18), and ACP NIR- (1.42 ± 0.08). The above results indicated that NIR irradiation alone had no significant effect on RUNX-2 expression in either the CON or ACP groups. In contrast, β-TCP treatment substantially elevated osteogenic activity through FAK/MAPK Signaling Pathway regulation [68]. Importantly, Mg integration further amplified this effect. Specifically, the RUNX-2 intensity of the ACP/Mg NIR group and ACP/Mg NIR + group showed 1.35-fold and 1.75-fold increases compared to ACP NIR- and ACP NIR+, respectively. Photothermal activation via NIR irradiation resulted in a 1.34-fold higher RUNX-2 expression in the ACP/Mg NIR + group compared with the ACP/Mg NIR- group. These results demonstrated a synergistic interaction between Mg and NIR photothermal stimulation, with the ACP/Mg NIR + group achieving maximal RUNX-2 expression (2.55 ± 0.09).

Fig. 9

Validation of osteogenic performance of BMSCs on scaffold surfaces under NIR laser irradiation (1064 nm, 0.5 W/cm²). (A–B) Immunofluorescence analysis of the impact of mild NIR laser irradiation on the RUNX-2 expression levels in the BMSCs. (C–D) Immunofluorescence analysis of the impact of mild NIR laser irradiation on the OCN expression levels in the BMSCs (n = 3)

As depicted in Figs. 9C-D, the immunofluorescence staining results of OCN were consistent with the expression trend of RUNX-2. The ACP/Mg NIR + group (2.27 ± 0.06) exhibited significantly elevated OCN immunofluorescence intensity, which was 2.27-fold higher than that of the CON NIR − group (1.00 ± 0.15), 2.29-fold higher than the CON NIR + group (0.99 ± 0.12), 1.69-fold higher than the ACP NIR − group (1.34 ± 0.09), 1.66-fold higher than the ACP NIR + group (1.37 ± 0.08), and 1.19-fold higher than the ACP/Mg NIR − group (1.90 ± 0.12). As demonstrated in prior research, NIR responsive scaffolds could effectively promote osteogenesis when heated to approximately 42 °C under NIR irradiation [36]. The above research results further confirmed that NIR-responsive scaffolds containing Mg could significantly enhance osteogenic effects under NIR irradiation.

Given the osteogenic potential of ACP/Mg scaffolds, mild NIR laser irradiation at 41 ± 1 °C enhanced their osteogenesis-promoting effect. This was because the ACP/Mg scaffolds—under the synergistic effect of mild NIR laser irradiation and thermal stimulation— mitigated inflammation, drove macrophage polarization toward the M2 phenotype, and facilitated the release of osteogenic factors (e.g., BMP-2), ultimately enhancing osteogenic activity [69].

In-vivo tissue regeneration under NIR laser irradiation

To further verify the effect of PTT irradiation on promoting bone formation around the scaffolds, three types of scaffolds (including PLA/PCL, ACP, and ACP/Mg) were implanted into the skull defect of rabbits (Fig. 10A), subjected to NIR at 0.5 W/cm² every other day for 12 weeks post-surgery.

Fig. 10

Verification of the enhancement of osteogenic performance of scaffolds through mild NIR laser irradiation in vivo. (A) Process diagram for using shape-memory scaffolds to repair craniofacial bone injuries. (B–E) Representative 3D reconstruction images and analysis of the BMD, T_b.T_h, and BV/TV of the newly formed bone in the defect areas following scaffold implantation and mild NIR laser irradiation for 12 weeks. (F–G) Staining of newly formed bone in the cranial defect area using H&E and Masson’s trichrome staining methods following 12 weeks of shape-memory scaffold implantation and mild NIR laser irradiation (n = 5)

Bone regeneration in the extracted samples was assessed via micro-CT analysis as well as histological and immunohistochemical staining. As shown in Figs. 10B-E, micro-CT image reconstruction and subsequent statistical analysis of the bone volume/tissue volume (BV/TV), trabecular thickness (T_b. T_h), and bone mineral density (BMD) for each group revealed that in the CON NIR − and CON NIR + groups—which did not receive osteogenic-promoting materials—the amount of newly formed bone around the scaffold was minimal. Conversely, the ACP NIR − and ACP NIR + groups—which included β-TCP—exhibited a modest increase in new bone formation around the scaffold. Notably, the ACP/Mg NIR − group demonstrated a considerably larger amount of new bone formation than the CON NIR−, CON NIR+, ACP NIR−, or ACP NIR + groups. It was noteworthy that the ACP/Mg NIR + group showed the greatest amount of new bone formation. The newly formed bone exhibited favorable bone density and well-preserved bone structure.

Subsequently, the histopathological structure of the regenerated bone at the defect site was examined using H&E and Masson’s trichrome staining. The defect areas in the CON NIR − and CON NIR + groups were occupied by fibrous tissue connecting the host bone margins, resulting in poor bone regeneration (Fig. 10F–G). In contrast, the ACP NIR − and ACP NIR + groups exhibited a small amount of newly formed bone. However, the ACP/Mg NIR- group showed a slightly higher degree of new bone formation. Meanwhile, under mild NIR light conditions at 41 ± 1 °C, the ACP/Mg NIR + group exhibited a substantial amount of continuously regenerated lamellar bone.

Masson’s trichrome staining further demonstrated that the bone-defect area in other groups had only sparse deposits of collagen fibers, whereas the bone-defect area of the ACP/Mg NIR + group exhibited abundant collagen fibers with dense and continuous structures, indicating the formation of mature lamellar bone. Thus, immunofluorescence staining validated the RUNX-2, OCN, CD86, and CD206 expression in the bone-defect regions across all groups. RUNX-2 and OCN expression in the cranial bone-defect sites of all groups were consistent with the trends observed using H&E and Masson’s trichrome staining (Fig. 11A and C). Meanwhile, the ACP/Mg NIR + group exhibited the lowest CD86 expression and highest CD206 expression compared with the other groups (Fig. 11B and D). These experimental results, combined with the in vitro findings, further illustrated that the ACP/Mg scaffold under mild NIR laser irradiation at 41 ± 1 °C promoted osteogenesis by regulating macrophage polarization.

Fig. 11

Assessment of the effects of mild NIR laser irradiation on osteogenic performance and macrophage modulation through immunofluorescence staining. (A) Expression of OCN and RUNX-2 in newly formed bone after 12 weeks of scaffold repair and mild NIR laser irradiation. (B) Fluorescence staining of CD86 and CD206 in the newly formed bone at the defect site after 12 weeks of scaffold repair and mild NIR laser irradiation (C) Relative fluorescence intensity of OCN and RUNX-2 in the newly formed bone after 12 weeks. (D) Relative fluorescence intensity of CD86 and CD206 in the newly formed bone after 12 weeks (n = 5)

To elucidate the mechanism by which ACP/Mg-mediated PTT promotes osteogenesis, high-throughput transcriptome sequencing was used to analyze the differential expression of mRNAs between the ACP/Mg NIR − and ACP/Mg NIR + groups. Following laser irradiation of the ACP/Mg scaffolds, 3193 mRNAs exhibited differential expression (Fig. 12A), of which 1851 were upregulated and 1342 were downregulated (Fig. 12B).

Fig. 12

RNA sequencing and analysis. (A) Differences in mRNA expression between the ACP/Mg NIR − group and ACP/Mg NIR + group. (B) RNA-Seq analysis of differentially expressed genes in BMSCs treated with ACP/Mg NIR + and ACP/Mg NIR − scaffolds. (C) GO annotation analysis of the differentially expressed genes. (D) KEGG enrichment analysis of the differentially expressed genes

Subsequently, gene ontology (GO) analysis was performed to investigate the potential biological functions of the differentially expressed mRNAs (Fig. 12C). Under NIR laser irradiation, the ACP/Mg scaffolds participate in functions related to immune system regulation, cell growth, and metabolic processes. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis further identified 20 associated pathways, including the MAPK and osteogenic differentiation signaling pathways (Fig. 12D). The MAPK signaling pathway is crucial in regulating osteoblast proliferation and differentiation [70]. The gene sequencing results suggested that the 42 °C stimulation mediated by PTT may activate the MAPK and osteogenic differentiation pathways, promoting osteogenesis.

Conclusions

In summary, this study pioneers an intelligent NIR-responsive therapeutic platform that integrates multiple biomedical breakthroughs. Unlike conventional shape-memory polymers limited to single-function applications, our β-TCP/Mg-incorporated PLA/PCL composite achieves unprecedented multifunctional synergy through rational material engineering. The developed scaffold combined three critical advancements: (1) precision functional conversion via NIR-mediated dual-temperature regulation (48 °C for shape recovery vs. 42 °C for immunomodulation), (2) remarkable efficacy in tumor ablation, and (3) spatiotemporal immunomodulation through macrophage polarization - a feature rarely achieved in existing bone scaffolds. Particularly, the demonstrated capacity to simultaneously address post-resection challenges (irregular defects, residual tumor cells, and inflammatory microenvironment) represents a significant advancement over current single-purpose solutions. This multimodal therapeutic strategy, which synergizes shape-adaptive filling, photothermal therapy, and macrophage phenotype reprogramming, establishes a new paradigm in regenerative medicine for oncological bone reconstruction.

Materials and methods

Materials

PCL (MW = 8 × 10⁴ Da) was purchased from Jinan Daigang Biomaterial Co., Ltd. (Jinan, China). PLA (MW = 8 × 10⁴ Da) was obtained from Nantong Jiuding Bioengineering Co., Ltd. (Nantong, China). β-Tricalcium phosphate (β-TCP; MW = 310.18 Da) was purchased from Sigma-Aldrich Co., Ltd. (St. Louis, MO, USA). Mg particles were obtained from Beijing DK Nano Technology Co., Ltd. Rabbit-derived bone marrow mesenchymal stem cells (BMSCs), 143B osteosarcoma cells, and mouse macrophages (RAW 264.7) were obtained from Pricella Co., Ltd. (Wuhan, China). Dulbecco’s modified eagle medium (DMEM), fetal bovine serum (FBS), phosphate-buffered saline (PBS), and 0.25% trypsin/ethylenediaminetetraacetic acid (EDTA) were obtained from Gibco Life Technologies (Grand Island, NY, USA). The Cell Counting Kit-8 (CCK-8), Fluorescein Isothiocyanate (FITC)-Phalloidin, Calcein-AM/propidium iodide (PI), 4’,6-diamidino-2-phenylindole (DAPI), and ALP, bicinchoninic acid (BCA) protein assay kits were purchased from Beyotime Biotechnology (Shanghai, China). The rhodamine B phalloidin was obtained from Cytoskeleton (Denver, CO, USA). The antibodies against CD86 and CD206 were purchased from Abcam. The antibodies against RUNX-2 and OCN were purchased from Thermo Fisher Scientific. The secondary antibodies were obtained from Abbkine. Enzyme-linked immunosorbent assay (ELISA) kits were purchased from Abcam (Cambridge, UK). Trizol was obtained from Biosharp (Shanghai, China).

Preparation of PLA/PCL/TCP/Mg composite materials and scaffolds

Composite materials with PLA and PCL ratios of 10:0, 9.5:0.5, 9:1, 8.5:1.5, and 8:2 were prepared individually using a mixer at 180 °C. The glass transition temperature (T_g) of the composite materials were determined using a TA Instruments Q2000 differential scanning calorimeter (DSC) under a nitrogen atmosphere at a heating rate of 10 °C/min from − 50 to 200 °C. The measurements were performed to determine the optimal ratio of PCL addition.

The PLA, PCL, Mg, and β-TCP were mixed in mass ratios of 6.6:1.65:0.75:1, 7:1.75:0.25:1, 6.8:1.7:0.5:1, and 6.4:1.6:1:1, respectively, in a compounding machine at 180 ℃ to uniformly blend them. Additionally, PLA and PCL were mixed in an 8:2 ratio using an internal mixer at 180 °C as the control group. To confirm the effects of β - TCP and Mg powder on the Tg of the composite materials, we conducted DSC tests on each of the five materials mentioned above. Then, scaffolds with varying mass fractions of Mg and 10 × 10 × 10 mm dimensions were printed using a FDM 3D printer purchased from Changchun Institute of Mechanical Science Co., Ltd. During the printing process, the nozzle temperature was adjusted to 178 °C, and the printing speed was set to 500 mm/s. The build orientation was specifically configured with a cross-90° strategy, where the printing direction alternated by 90° between successive layers to enhance interlayer bonding strength and reduce anisotropic mechanical properties. We also configured the layer height at 0.5 mm, calibrated the extrusion rate to 100% for consistent deposition and defect minimization, and kept the bed temperature at 40 °C to ensure first-layer adhesion and prevent warping. Additionally, an infill density of 50% was selected to ensure adequate structural strength. The scaffolds were divided into five groups: PLA/PCL/TCP (ACP), PLA/PCL/TCP/2.5Mg (ACP/2.5Mg), PLA/PCL/TCP/5Mg (ACP/5Mg), PLA/PCL/TCP/7.5Mg (ACP/7.5Mg), and PLA/PCL/TCP/10Mg (ACP/10Mg). During printing, the fill density of the scaffolds was set to 50%. Photographs of the printed scaffolds from each group were taken using a Digital Single-Lens Reflex Camera.

Scaffold characterization

A super-depth microscope (Keyence VHX-2000) was used to assess the scaffold roughness. Scanning electron microscopy (SEM, Apreo S HiVoc, USA) and elemental mapping were used to examine scaffolds with varying Mg mass fractions. Gold was sprayed on the surface of the scaffold at a 10–20 nm thickness. Elemental analysis was performed using energy dispersive X-ray (EDX) spectroscopy. Water contact angles were measured using a contact angle goniometer (SL150E; KINO, USA). The compressive modulus of five groups of scaffolds (n = 3) was evaluated using a universal testing machine (CMT-5504, Shenzhen SANS, China) at a crosshead speed of 1 mm/min. While the scaffold components were evaluated via Fourier transform infrared (FTIR) spectroscopy (Nicolet 6700, Thermo Scientific, USA).

Photothermal and shape-memory performance verification of scaffold

Five replicates of 50 × 10 × 1 mm scaffolds for each experimental group were printed separately to verify their shape-memory performance. The printed scaffolds were placed in 50 °C water for shaping; their shape recovery was subsequently recorded after submersion in 50 °C water for 5s. Photographs of the printed scaffolds were obtained to document the shape recovery. To verify the photothermal conversion effect of the materials, five groups of scaffolds (φ1 mm × 5 mm, n = 3 per group) were printed separately: ACP, ACP/2.5%Mg, ACP/5%Mg, ACP/7.5%Mg, and ACP/10%Mg. The photothermal conversion effects of the scaffolds were validated using a 1064-nm NIR laser (Spl Laser Optoelectronics, China) under different power levels and an infrared thermal imaging camera (FLIRTM A325SC camera, USA). Five groups of scaffolds (n = 3) were irradiated with NIR light at a fixed distance of 10 cm (1064 nm, 1.5 W/cm²). Temperature data were captured every 10s using a FLIR camera. To identify the optimal power for NIR irradiation, scaffolds of the ACP/7.5Mg group were exposed to NIR light at three different power densities (1.5, 1.0 and 0.5 W/cm²) until thermal equilibrium was achieved, with the results being meticulously recorded. Subsequently, the photothermal stability of the ACP/7.5Mg scaffolds was assessed by subjecting them to four consecutive on-off cycles. Each cycle involved 30s of laser irradiation at 1.0 W/cm², followed by natural cooling to ambient temperature.

Assessment of the osteogenic differentiation of BMSCs

Cytocompatibility of the scaffold

The third-generation BMSCs were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin (without osteogenic additives). They were incubated in a humidified incubator at 37 °C with 5% CO₂, and the medium was changed every 72 h. The viability of BMSCs on the scaffolds was evaluated on day 3 using the Live/Dead cell staining assay. After sterilization, BMSCs were seeded onto each group of scaffolds at a density of 2 × 10⁴ cells/cm². After three days, the culture medium was removed, and the scaffolds were washed twice with PBS. Calcein-AM and PI stock solutions were separately diluted 1000-fold in serum-free medium, then mixed to prepare the working solution, which was used immediately. Each group of scaffolds was immersed in the working solution for staining. After incubation at room temperature in the dark for 20 min, the stained area was washed with PBS. The samples were subsequently observed under a ZEISS Axio Observer 7 microscope (Carl Zeiss Microscopy GmbH, Germany) with 200× total magnification (20× objective lens and 10× eyepiece).

The BMSCs were cultured in 24-well plates, with each well containing one scaffold from different groups. After culturing the BMSCs on the scaffold for 1 d, the samples were fixed with 4% paraformaldehyde for 30 min, washed thrice with PBS, and immersed in 0.5% Triton X-100 for 10 min. Subsequently, rhodamine B phalloidin was used to stain the actin filaments on the surface of the bone-layer scaffold, while DAPI was applied to stain the cell nuclei. The cytoskeletons were then observed using a ZEISS Axio Observer 7 microscope (Carl Zeiss Microscopy GmbH, Germany) with 200× total magnification (20× objective lens and 10× eyepiece).

To assess the proliferation rate of the BMSCs seeded onto the scaffold, the CCK-8 assay was used. BMSCs (2 × 10⁴ cells/mL) were seeded onto each group of scaffolds and cultured in 24-well plates (n = 3 per experimental group, with three independent biological replicates). The culture medium was discarded after 1, 4, and 7 days of incubation. CCK-8 working solution, prepared according to the manufacturer’s instructions, was added to each well and incubated at 37 °C with 5% CO₂ for 2 h. Finally, absorbance was measured at 450 nm using a Bio-Rad microplate reader.

ALP staining

BMSCs (2 × 10⁴ cells/mL) were seeded onto 3D-printed scaffolds in a 24-well plate (n = 3 per experimental group, with three independent biological replicates). Upon reaching 80% confluence, the cells were treated with an osteogenic induction medium for 14 days. The culture medium was discarded, and cells were washed twice with PBS. Subsequently, the cells were fixed with formaldehyde at 37 °C for 30 min. The cells and scaffolds were washed twice with PBS and submerged in an ALP working solution containing ALP staining buffer, BCIP solution (300×), and NBT solution (150×) for 1 h. The scaffolds and cells were rinsed twice with PBS. Finally, the ALP activity was measured on days 7 and 14 using commercial ALP and BCA protein assay kits from Beyotime. The purpose of BCA protein assay kit was to determine the protein content in the sample, thereby standardizing the enzyme activity of ALP into unit protein content for accurate ALP quantitative analysis.

Alizarin red staining

The degree of extracellular matrix (ECM) mineralization was assessed using alizarin red staining. The 3D-printed scaffolds were seeded with BMSCs at a density of 2 × 10⁴ cells/mL and subsequently cultured in 24-well plates (n = 3 per experimental group, with three independent biological replicates). Upon reaching 80% confluency, the cell’s differentiation was induced with osteogenic differentiation medium for 21 days before staining. During the staining process, the culture media were discarded, and the cells were fixed with 4% (w/v) formaldehyde in PBS at 37 °C for 30 min after washing twice with PBS. Subsequently, the dye solution was applied, immersing the 3D-printed scaffolds after two additional washes with PBS. Finally, the scaffolds and cells were rinsed twice with PBS; 10% (w/v) cetylpyridinium chloride in methanol was added to quantify calcium nodules, followed by incubation and centrifugation at 12,000 rpm for 15 min. Then the absorbance of the supernatant was measured at 562 nm using a UV–Vis spectrophotometer (Agilent Cary 60, Agilent Technologies, Santa Clara, CA, USA).

Real-time quantitative PCR (RT-qPCR)

Real-time quantitative polymerase chain reaction (RT-qPCR) was used to assess the effect of each scaffold on BMSCs expression of Runt-related transcription factor 2 (RUNX-2), ALP, and osteocalcin (OCN). Under osteogenic induction culture conditions, each scaffold was incubated with BMSCs for 14 days (n = 3). Subsequently, the RNA was extracted using the Foregene kit (China 1.0-1708) and quantified using an Infinite 200 PRO NanoQuant microplate reader (Tecan). The reverse transcription kit (Genestar, China) was used in the experiment. Following reverse transcription of the RNA to DNA, RT-qPCR was conducted using the StepOnePlus Real-Time PCR System (Applied Biosystems) with 2× Fast SYBR Green Master Mix (Roche Diagnostics, Basel, Switzerland). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as the internal reference gene. The relative mRNA expression was calculated using the 2^-ΔΔCt method. The primer sequences for GAPDH, ALP, RUNX-2, and OCN were listed in Table S1.

Design and selection of scaffold structures

The ACP/7.5Mg material, which demonstrated optimal osteogenic performance and photothermal conversion efficiency, was selected for further experimentation. Scaffolds with intersecting lattice structures (1 × 1 × 1 cm) were designed in SolidWorks. The scaffolds were fabricated via FDM technology under precisely controlled process parameters, including a nozzle temperature of 178 °C and a deposition rate of 500 mm/s. The manufacturing process employed a unique strategy, implementing layer-wise 90-degree rotation between successive deposition layers while simultaneously driving the print head along a pre-programmed sinusoidal trajectory during material extrusion. A layer height of 0.5 mm was employed alongside a calibrated extrusion rate of 100% for uniform material deposition. The print bed was maintained at 40 °C to secure first-layer adhesion and minimize warping. Three distinct fill densities (40, 50 and 60%) were generated to systematically evaluate the relationship between infill percentage and scaffold functionality.

To identify scaffolds with optimal filling ratios for subsequent studies, a systematic evaluation was conducted as follows: scaffolds with different filling ratios (n = 3) were irradiated at a fixed distance of 10 cm using a 1064 nm NIR laser (1.0 W/cm²). Temperature monitoring was performed using a FLIR thermal imaging camera with data acquisition at 10s intervals. When the scaffold temperature reached the critical threshold of 48 °C, mechanical compression testing was immediately initiated using a standardized 2-kg weight. The deformation characteristics of each scaffold group were subsequently analyzed through photographic documentation.

The surface morphology was examined by SEM. The compressive moduli of three types of scaffolds were assessed using a universal testing machine (CMT-5504, Shenzhen SANS, China). Three independent replicates were tested for each group (n = 3). To observe the shape-memory performance of the scaffold more intuitively under localized temperature increases, the scaffold with 50% fill densities was submerged in water at 50 °C to raise its temperature while the interference of near-infrared light on our imaging was minimized. Simultaneously, a 50 g weight was placed on top of the scaffold. The deformation process was recorded using a Canon EOS R5 C camera.

To compare the influence of curved versus straight lattice structures on cell adhesion to scaffold surfaces, BMSCs were seeded onto the surfaces of scaffolds (n = 3) and cultured for 1 and 3 d. Subsequently, live/dead, fluorescein isothiocyanate (FITC) phalloidin, and DAPI staining processes were conducted. For specific operational procedures, refer to Sect. 2.4.1.

Effect of NIR laser irradiation on tumor cell eradication

Cell viability of 143B cells under photothermal ablation in vitro

After evaluating the shape-memory performance and mechanical properties of the intersecting lattice structure scaffolds with various fill densities, the internal structure was set at a 50% fill density for subsequent experiments. Scaffolds with 50% infill densities (φ5 mm, 1 mm) were fabricated using APC and APC/7.5Mg materials.

First, 143B osteosarcoma cells were cultured at 37 °C and 5% CO₂ in DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. The 143B cells were seeded into 24-well plates at 1 × 10⁴ cells per well and allowed to adhere after 24 h. Subsequently, the two types of scaffolds were placed on the cell layer. Groups without scaffolds were designated as the control groups. The experiment was divided into six groups (n = 3) and the specific grouping details were shown in Table 1. In the photothermal stimulation group (NIR+), the cells were exposed to NIR (1064 nm, 1.0 W/cm²) for 10 min and returned to the incubator for 24 h. Subsequently, live/dead staining was performed to assess the survival rates of each group.

Table 1 Experimental group setup

In-vivo antitumor therapy using photothermal scaffolds

Four-week-old male nude BALB/c mice (n = 48) were purchased from Sibeifu Biotechnology Co., Ltd. (Beijing, China). The in vivo experiment was approved by the Animal Care and Use Ethics Committee of Jilin University (2024-07-004). Throughout the experiment, we adhered strictly to the institutional ethical use protocols (stated in the NIH Guide for Care and Use of Laboratory Animals). The 143B osteosarcoma cells were administered into the bodies of BALB/c nude mice by subcutaneous injection. When the tumor volume reached approximately 300 mm³, the tumor-bearing mice were randomly assigned to four groups: no implanted scaffold and no light exposure (CON NIR−), no implanted scaffold with light exposure (CON NIR+), ACP scaffold without light exposure (ACP NIR−), ACP scaffold with light exposure (ACP NIR+), ACP/7.5Mg scaffold without light exposure (ACP/Mg NIR−), ACP/7.5Mg scaffold with light exposure (ACP/Mg NIR+).

Next, an incision was made at the tumor periphery. ACP (φ5 mm, 1 mm) and ACP/7.5Mg (φ5 mm, 1 mm) scaffolds were respectively implanted into the tumor base (n = 5). Every day, mice in the NIR groups were exposed to 5-min irradiation using a 1064-nm NIR laser (1.0 W/cm²), and the temperature on the surface of the tumor was recorded. The first day of NIR treatment was designated Day 0, and the length and width of the tumors were measured every two days to calculate the tumor volume (V) using the following formula: [(tumor length) × (tumor width)]²/2. The size and morphology of tumor tissues were recorded and photographed on days 1, 5, and 10. Statistical analyses of the tumor volume and body weight changes were performed on days 1, 3, 5, 7, and 9. On Day 11, mice were subjected to euthanasia by means of carbon dioxide (CO₂) inhalation, and the tumor, heart, livers, spleen, and lungs were harvested from each mouse. The samples were subsequently embedded in paraffin and stained using hematoxylin and eosin (H&E). Tissue specimens were fixed in 4% paraformaldehyde (24 h, 4 °C) and processed through graded ethanol dehydration (70-100%), xylene clearing, and paraffin embedding. Serial 5 μm sections were subjected to automated H&E staining with controlled incubation times (6 min hematoxylin; 1 min eosin). The samples were dehydrated and mounted on slides, followed by observation under an Olympus BX53 light microscope (Olympus Corporation, Tokyo, Japan) configured with 40 × objective magnification.

Effect of NIR laser irradiation on macrophages

RAW 264.7 cells were cultured under 37 °C and 5% CO₂ in DMEM supplemented with 10% FBS. LPS (100ng/mL) stimulated macrophages for 2 days to simulate an inflammatory environment and induce macrophage polarization toward the M1 phenotype. Then, RAW 264.7 cells (1.5 × 10⁴ cells/well) were separately seeded onto 24 well plates containing PLA/PCL, ACP, or ACP/7.5Mg scaffolds (n = 3). The samples were irradiated using a 0.5 W/cm² NIR laser for 15 min on days 0, 2, 4, and 6 post-seeding. Subsequently, after co-culturing for 4 days, the supernatants from each group were collected, centrifuged, and filtered through a 0.22-µm filter to obtain a conditioned medium for subsequent experiments.

To evaluate the effect of light on macrophage viability and proliferation, live/dead staining was performed on day 3, and CCK-8 assays were conducted separately on days 1, 3, and 7. CD86 serves as a crucial surface marker for the M1 phenotype, with variations in its expression levels indicating the extent of macrophage polarization towards this phenotype, while CD206 is a key marker for M2 phenotype. The expression of CD86 (M1 marker) and CD206 (M2 marker) in the macrophages from each group was validated using RT-qPCR on day 7 to validate further the impact of NIR laser irradiation on macrophage polarization. Additionally, CD86 and CD206 protein abundance was assessed via immunofluorescence analysis. After 7 days of cell culture, the cells were washed with PBS, fixed with 4% paraformaldehyde for 10 min, rinsed with excess formaldehyde, and washed with PBS; the cells were then soaked in 0.1% Triton X-100 for 20 min. Next, the cells were washed thrice with PBS and were incubated with 5% bovine serum albumin (BSA) for 30 min. Primary antibodies against CD86 (1:200) and CD206 (1:200) were added to the wells, respectively, and the cells were incubated overnight at 4 °C. The secondary antibodies (1:300) were then added and incubated with rhodamine B phalloidin for 1 h. Finally, the cell nuclei were stained with DAPI and photographed using a fluorescence microscope.

The mRNA expression of inflammation-related factors, including interleukin (Il)-4, Il-10, Il-6, and tumor necrosis factor (Tnf)-α in the RAW264.7 cells was measured using RT-qPCR. The primer sequences for Gapdh, Cd86, Cd206, Il-4, Il-10, Il-6, and Tnf-α were provided in Table S1. Additionally, an ELISA kit was employed to assess the levels of anti-inflammatory (IL-4 and IL-10) and pro-inflammatory (IL-6 and TNF-α) factors in the cell supernatant.

Effect of NIR laser irradiation on the osteogenic activity of BMSCs in vitro

The cells were cultured in conditioned medium to investigate the impact of NIR laser irradiation on macrophage polarization and its subsequent effect on the osteogenesis of BMSCs. The conditioned medium mentioned in Sect. ‘Effect of NIR Laser Irradiation on Macrophages’ was mixed with fresh DMEM in a 1:1 ratio to prepare the experimental medium. After this, 10 mM β-glycerophosphate, 0.05 mM ascorbic acid, and 10 nM dexamethasone were added to formulate the osteogenic induction medium.

After 3 days of culture, live/dead staining was performed, and CCK assays were conducted on days 1, 3, and 7 to assess the effect of light exposure on the cell survival rate and proliferation of BMSCs. Alizarin Red staining was performed after 14 and 21 days of cultivation in the osteogenic induction medium, and ALP staining was performed after 7 and 14 days of cultivation (refer to Sect. ‘ALP staining’ and ‘Alizarin red staining’ for specific instructions). RT-qPCR was performed to further evaluate the expression of ALP, OCN, and RUNX-2 in the BMSCs cultured for 7 and 14 days.

Immunofluorescence staining of bone morphogenetic protein-2 (BMP-2) and RUNX-2 further assessed macrophage regulation by the scaffold under NIR laser irradiation. Following the steps outlined in Sect. ‘Effect of NIR Laser Irradiation on Macrophages’. After 14 days of cell culture, the primary antibodies against RUNX-2 (1:200) and OCN (1:200) were used, along with the corresponding secondary antibodies (1:300).

Regeneration of bone tissue under NIR laser irradiation in vivo

Male New Zealand white rabbits (3 weeks old, n = 48) were obtained from Sibeifu Biotechnology Co., Ltd. (Beijing, China). All animal protocols were approved by the Animal Care and Use Ethics Committee of Jilin University (2023-07-005). Initially, the prefabricated scaffold was reshaped by soaking it in warm water to reduce its volume before being sterilized and prepared for the subsequent procedures. After anesthetizing the New Zealand rabbits with isoflurane inhalation, defects with 8 mm diameter were generated in the rabbit skulls using a bone drill with an 8 mm diameter bit. Subsequently, three distinct types of scaffolds—PLA/PCL (designated as the CON group), ACP, and ACP/Mg—were selected and surgically implanted into the cranial defect models. The groups were categorized based on whether they were subjected to NIR laser irradiation, thus forming the following groups: CON NIR-, CON NIR+, ACP NIR-, ACP NIR+, ACP/Mg NIR-, and ACP/Mg NIR+. Antibiotics were given continuously for 3 days following the surgery. Each group comprised eight rabbits for the experiment. NIR light (1064, 1.5 W/cm²) was subsequently applied to the scaffold to restore its volume to its original size. Postoperatively, NIR laser irradiation at 0.5 W/cm² was performed at the scaffold implantation site every other day.

Micro-CT test and histological assessments

After euthanizing the New Zealand rabbits with CO₂ inhalation 12 weeks postoperatively, skull samples were collected from six groups: CON NIR-, CON NIR+, ACP NIR-, ACP NIR+, ACP/Mg NIR-, and ACP/Mg NIR+. The 3D reconstruction of samples was performed using a high-resolution µCT (SkyScan 1272, Bruker Micro CT, Kontich, Belgium). Quantitative measurements of BMD, BV/TV, and Tb. Th at the defect repair site were conducted through a multi-software workflow: image reconstruction and alignment using Data Viewer (version 1.5.4.0, Bruker Micro CT), morphometric parameter calculation via CTAn (version 1.9.0.0, Bruker Micro CT) and volumetric visualization with CT Vox (version 3.3.1.0, Bruker Micro CT).

The skull tissues were decalcified using a 10% ethylenediaminetetraacetic acid (EDTA) solution and subsequently stained for H&E, Masson’s trichrome, and immunofluorescence analysis. The following antibodies were used: RUNX-2 (#12556; CST, 1:200), anti-OCN (#PA5-96529; Thermo Fisher Scientific, 1:200), CD86 (A16805; Abclonal, 1:200), and CD206 (AB64693; Abcam, 1:300).

RNA sequencing and analysis

To evaluate the effect of NIR illumination on osteogenic properties, we performed RNA sequencing and analysis on BMSCs from the ACP/Mg NIR- and ACP/Mg NIR + groups after 14 days of culture. Total RNA was extracted using Trizol reagent and sent to Majorbio (Shanghai, China) to create six sample libraries, including three from the ACP/Mg NIR- group and three from the ACP/Mg NIR + group. The mRNA was enriched using Oligo dT, fragmented, reverse transcribed into cDNA, adaptor-ligated, size-selected, and enriched to prepare the libraries. Sequencing was then performed on the NovaSeq X Plus platform. The reference genome index was constructed using HISAT2 software (version 2.1.0). Gene expression levels were initially quantified with HTSeq (version 0.9.1) and normalized using the FPKM method. For the exploration of biological functions, we sequentially performed GO and KEGG pathway enrichment analyses.

Statistical analysis

All data were presented as mean ± standard deviation (SD) from a minimum of three independent experiments. Statistical analyses were performed using the GraphPad Prism 9 software. One-way and two-way analysis of variance (ANOVA) were used to compare the groups, followed by Tukey’s post-hoc test for multiple comparisons. Statistical significance was set at P < 0.05.