Multi-modal imaging for dynamic visualization of osteogenesis and implant degradation in 3D bioprinted scaffolds

Abstract

In situ monitoring of bone regeneration enables timely diagnosis and intervention by acquiring vital biological parameters. However, an existing gap exists in the availability of effective methodologies for continuous and dynamic monitoring of the bone tissue regeneration process, encompassing the concurrent visualization of bone formation and implant degradation. Here, we present an integrated scaffold designed to facilitate real-time monitoring of both bone formation and implant degradation during the repair of bone defects. Laponite (Lap), CyP-loaded mesoporous silica (CyP@MSNs) and ultrasmall superparamagnetic iron oxide nanoparticles (USPIO@SiO₂) were incorporated into a bioink containing bone marrow mesenchymal stem cells (BMSCs) to fabricate functional scaffolds denoted as C@M/GLU using 3D bioprinting technology. In both in vivo and in vitro experiments, the composite scaffold has demonstrated a significant enhancement of bone regeneration through the controlled release of silicon (Si) and magnesium (Mg) ions. Employing near-infrared fluorescence (NIR-FL) imaging, the composite scaffold facilitates the monitoring of alkaline phosphate (ALP) expression, providing an accurate reflection of the scaffold's initial osteogenic activity. Meanwhile, the degradation of scaffolds was monitored by tracking the changes in the magnetic resonance (MR) signals at various time points. These findings indicate that the designed scaffold holds potential as an in situ bone implant for combined visualization of osteogenesis and implant degradation throughout the bone repair process.

Graphical abstract

To achieve continuous dynamic monitoring of the bone regeneration process, an integrated scaffold was prepared here. Alkaline phosphatase-activated nanofluorescent probes and ultra-small superparamagnetic iron oxide nanoparticles were incorporated into bioink containing bone marrow mesenchymal stem cells, and subsequently, functional scaffolds were fabricated by 3D bioprinting. Early bone regeneration and scaffold degradation were effectively visualized by tracking the near-infrared fluorescence and magnetic resonance signal changes of the scaffolds at different time points.

Highlights

• •The secretion of alkaline phosphatase and the degradation of scaffolds were monitored by multi-modal imaging.
• •The C@M/GLU scaffolds enable non-destructive, continuous monitoring of bone regeneration without sacrificing animals.
• •The composite scaffold can enhance bone regeneration by releasing silicon and magnesium ions.

1. Introduction

3D bio-printed scaffolds stand out as an optimal tissue engineering technique for the treatment of bone defects [[1], [2], [3], [4], [5]]. These scaffolds provide essential mechanical support and demonstrate concurrent biological responsiveness during bone repair [6,7]. Nevertheless, this process involves a cascade reaction of multiple cells, factors, and extracellular matrix components, rendering it one of the intricate biological processes in living organisms [8]. Previous studies have indicated that achieving optimal bone regeneration necessitates continuous diagnosis and medical intervention at each stage following scaffold implantation [9]. For instance, real-time and precise tracking of scaffold degradation using imaging techniques would enhance the refinement of scaffold materials and designs [10,11]. However, current histological assessments, including hematoxylin-eosin staining and Masson staining, often necessitate sacrificing animals and are constrained by the intermittent and fragmented acquisition of information [12]. Those limitations significantly reduce the efficiency and precision of information collection. Consequently, there exists a pressing need for a non-invasive imaging technique to continuously monitor the state of bone tissue engineering scaffolds throughout the bone regeneration process.

To date, advanced non-destructive imaging techniques, including magnetic resonance (MR) imaging [13], fluorescence (FL) imaging [14], computed tomography (CT) [15], ultrasound (US) imaging [16], and photoacoustic (PA) imaging [17], have exhibited significant potential in visualizing tissue engineering. Distinguished for its non-invasiveness, high sensitivity, and resolution, near-infrared fluorescence (NIR-FL) is a widely employed method for monitoring physiological or pathological processes, such as inflammation, scaffold degradation, and bone regeneration [9,[18], [19], [20]]. Specifically, upon activation NIR fluorescent probe, responds to the molecular target of interest by emitting fluorescence. This fluorescence shift effectively diminishes the background signal and enhances the signal-to-background ratio (SBR) for real-time in vivo imaging [21].

CyP, a semicarbocyanine NIR-FL probe, interacts with alkaline phosphate (ALP) to emit fluorescence. Li and colleagues have demonstrated that CyP exhibits exceptional sensitivity and a low detection limit (0.003 U/mL) for ALP, rendering it suitable for detecting endogenous ALP [22]. Interestingly, ALP demonstrates a close association with bone formation. Throughout the bone repair process, active osteoblasts release elevated levels of ALP. This, in turn, promotes bone mineralization by binding to bone matrix proteins and stimulating pyrophosphate hydrolysis [23,24]. Hence, CyP emerges as a promising option for monitoring ALP level serving as a marker of bone formation. After the reaction of CyP with ALP secreted by BMSCs, the fluorescence intensity changed from weak to strong, and the change in fluorescence intensity indirectly reflected the osteogenic activity of the scaffold. In recent years, the utilization of nanomaterials has paved the way for extensive opportunities in the development of imaging platforms [25]. Inorganic and organic nanoparticles, frequently found in biomedicine, can be readily employed in various imaging modes due to their strong biocompatibility and biodegradability [26,27]. Mesoporous silica nanoparticles (MSNs), a notable example of inorganic nanomaterial, exhibit remarkable advantages including a high specific surface area, a large pore volume, and adjustable pore sizes. These features enable MSNs to efficiently transport chemicals for disease diagnosis and treatment [28,29]. Furthermore, silica materials exhibit a pronounced lack of optical signal absorption and do not interfere with the magnetic signals inherent to magnetic nanomaterials. Consequently, MSNs find widespread application in diagnostic and therapeutic domains, serving as efficacious carriers for pharmaceutical agents and facilitating the incorporation of molecular imaging probes in diagnostic and therapeutic modalities [30]. Additionally, numerous organic fluorescent dyes suffer from poor solubility and instability. To enhance their photophysical and photochemical properties, incorporating the fluorescent probe CyP into mesoporous silica represents an effective approach.

The degradation rate of scaffolds is a critical factor influencing bone regeneration. An optimal degradation rate provides mechanical support for tissue growth while ensuring sufficient space for tissue development [31]. However, biomaterials are believed to degrade at asynchronous rates in vivo and in vitro, presenting a challenging factor in achieving a balance between scaffold degradation and new bone regeneration [32]. For longitudinal and non-destructive tracking of substrate degradation, researchers have successfully monitored the in vivo degradation of biomaterials [18,33,34]. This has been achieved either by modifying molecular chains with NIR-FL probes or by directly doping them with biomaterials. Nonetheless, NIR-FL dyes are highly susceptible to fluorescence quenching and are associated with limitations, including concerns about photostability and the complexity of synthesis. These factors render them unsuitable for long-term tracking of biomaterial degradation. In contrast, MR imaging, a safe and non-invasive technique with deep penetration capabilities, provides an effective means to monitor material degradation and tissue regeneration in vivo. Compared to NIR-FL probes, MR contrast agents like USPIO are more readily accessible and exhibit high biocompatibility, making them ideal candidates for monitoring implant degradation [35,36].

In this study, we engineered a multi-modality NIR/MR imaging bio-scaffold for non-invasive in situ monitoring of bone repair. As illustrated in Scheme 1, a mixture containing CyP-loaded MSNs (CyP@MSNs), silica-coated USPIO (USPIO@SiO₂) magnetic nanoparticles was prepared and subsequently incorporated into a blend of gelatin methacrylate (GelMA) and laponite (Lap) to create the composite bioink. The composite scaffold (C@M/GLU) was fabricated using 3D bioprinting technology. With the secretion of ALP in early osteogenesis, CyP released from the integrated scaffold interacts with ALP, leading to the enhancement of fluorescence intensity, which in turn predicts early osteogenic activity through fluorescence image acquisition. Meanwhile, the USPIO@SiO₂ incorporated in the scaffold gives the scaffold the ability of MR imaging, which is expected to enable the observation of scaffold degradation at different periods. Moreover, the sustained release of Mg and Si ions from the integrated scaffold functioned as an effective promoter of new bone regeneration. In both in vitro and in vivo studies, the scaffold exhibited a favorable response to ALP, with fluorescence intensity increasing proportionally to ALP secretion during the early stages of osteogenesis. The results of MR imaging revealed a significant correlation between scaffold degradation and MR imaging signals, highlighting its effectiveness in monitoring scaffold degradation. Additionally, we employed BMSCs to evaluate scaffold-induced cell proliferation and osteogenic activity. The in vivo osteogenic potential of the scaffold was assessed through micro-CT analysis and histopathological staining. In summary, this study introduces an innovative approach to investigate scaffolds with the capacity to monitor bone regeneration in situ.

Scheme 1.

Schematic of combined 3D printing and imaging techniques to prepare C@M/GLU hydrogel scaffolds for in vivo monitoring of in situ bone repair. As ALP expression continues, the fluorescence of CyP released from MSNs becomes activated and progressively intensifies, enabling dynamic monitoring of early bone regeneration. USPIO@SiO₂ functions as an MR contrast agent to visualize the degradation process of the scaffolds.

2. Results and discussion

2.1. Preparation and imaging characterization of CyP@MSNs and USPIO@SiO₂

Developing signal-activated molecular imaging probes for the non-destructive and precise detection of disease markers in vivo is crucial for the early diagnosis of bone regeneration and real-time assessment of efficacy [37]. Therefore, for the effective monitoring of bone regeneration following scaffold implantation, we employed the NIR fluorescent probe CyP due to its responsiveness to ALP [22]. ALP secreted by BMSCs cleaves the phosphate group from CyP, thereby restoring the intramolecular charge transfer (ICT) effect and inducing fluorescence to “turn on” (Fig. 1A). To mitigate the fluorescence decay of the organic fluorescent dye CyP and enhance its photostability, we loaded it into MSNs. The morphology and structure of MSNs were observed by TEM, as illustrated in Fig. 1B. MSNs prepared using the oil-water biphasic layering method exhibited a uniform spherical shape with an average particle size of approximately 200 nm. Dendritic mesoporous structures are distributed in MSNs, which have large specific surface area and pore volume and can delay the quenching of CyP molecules by adsorption of CyP in the internal pores. The microstructures of CyP@MSNs showed no significant differences compared to the MSNs, and the successful loading of CyP was confirmed through elemental mapping (Fig. S1). Subsequently, the spectral property of the CyP@MSNs probe was investigated. Evident from the UV absorption spectra (Fig. 1D) CyP displayed a maximum absorption peak at 616 nm, whereas CyP @MSNs and MSNs did not exhibit a significant absorption peak. The disappearance of the original absorption peak of CyP was attributed to its encapsulation by MSNs. However, an absorption band at 702 nm reappeared after adding ALP to the CyP@MSNs solution for 12 h (Fig. 1E). The addition of ALP reinitiated the ICT process, ultimately resulting in a redshift of its maximum absorption band from 616 to 702 nm. This observation is consistent with the results of a previous study [22]. Fluorescence imaging further validated the responsiveness of the CyP@MSNs probe to ALP. In Fig. 1F, the CyP@MSNs probe emitted weak fluorescence in the absence of ALP. As ALP enzyme activity increased, the fluorescence intensity gradually elevated. At an enzyme activity level of 2 U/mL, the fluorescence intensity of the CyP@MSNs probe became 2.6 times that of the CyP@MSNs without ALP treatment (Fig. 1G). These results indicate that the developed nanoprobe CyP@MSNs can demonstrate an “on-off" response to ALP, rendering it an ideal probe for monitoring bone regeneration.

Fig. 1.

Characterization of CyP@MSNs and USPIO@SiO₂nanoparticles. (A) Schematic diagram illustrating the fluorescence imaging process after the reaction of CyP with ALP. TEM images of (B) MSNs and CyP@MSNs nanoparticles and (C) USPIO and USPIO@SiO₂ nanoparticles. (D) UV–Vis absorption spectra of CyP, CyP@MSNs, and MSNs. (E) UV–Vis absorption spectra of the probe CyP@MSNs probe before and after the reaction with ALP. (F and G) NIR-FL images of (F) and the relative FL intensity curve (G) of CyP@MSNs incubated with different enzyme activity units of ALP (λ_ex = 660 nm). (H and I) The relationship between 1/T₂ (H),1/T₁ (I) and different iron concentrations. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

The TEM images in Fig. 1C depict the morphology and size of the synthesized USPIO@SiO₂ nanoparticles. Both USPIO and USPIO@SiO₂ show spherical morphology, with an average particle size of 8.52 ± 1.20 nm for USPIO and 9.73 ± 1.41 nm for USPIO@SiO₂ (Fig. S2). Fourier transform infrared spectroscopy (FTIR) results in Fig. S3 indicate characteristic peaks at 458 cm⁻¹ and 592 cm⁻¹ correspond to the stretching vibration of the Fe–O bond, while the characteristic peaks at 798 cm⁻¹ and 1100 cm⁻¹ correspond to the stretching vibration of the Si–O bond. These findings confirm the successful synthesis of USPIO and the encapsulation of SiO₂. The relaxation rate serves as a crucial indicator of Magnetic Resonance Imaging (MRI) contrast agents. Samples with a high relaxation rate can produce optimal imaging results with a lower dosage. To demonstrate the influence of SiO₂ shells on the imaging capability of USPIO, we assessed the weighted MR effect of USPIO@SiO₂ utilizing a 0.5T NMR imager. As depicted in Fig. 1H and I, both USPIO@SiO₂ and USPIO exhibit a good linear relationship with 1/T₁ and 1/T₂. Moreover, the relaxation rate of USPIO@SiO₂ is notably high, reaching 124.70 mM⁻¹s⁻¹ with an R₂/R₁ ratio of about 19. Hence, USPIO@SiO₂ could serve as an effective magnetic resonance T₂ contrast agent for monitoring implant degradation. It is worth noting that an excessively high concentration of nanoparticles can exert adverse effects on cellular proliferation. Consequently, the biocompatibility of both USPIO@SiO₂ and CyP@MSNs was assessed through cytotoxicity experiments at 24 h and 48 h. As illustrated by the live/dead staining images and quantitative CCK-8 results presented in Fig. S4, it is evident that both types of contrast agents exhibit low toxicity within a specific concentration range. Collectively, these results affirm their suitability for the fabrication of scaffolds intended for in situ bone repair monitoring.

2.2. In vitro imaging and characterization of scaffolds

Our prior research has indicated that the inclusion of Lap in hydrogel bioinks enhances the printability of bioink. These bioinks release ions such as Mg²⁺ and Si⁴⁺, exhibiting bioactive properties, that include immune microenvironment modulation and the promotion of bone regeneration [38]. To enhance the osteogenic activity, Lap was integrated into the 3D bioprinted scaffold (C@M/GLU). It is worth noting that the concentration of Lap influences the mechanical strength of the scaffold. The correlation between Lap concentrations and the stress-strain curve was investigated. The compressive strength of the scaffolds increased gradually with higher Lap content (Fig. S5). Based on our previous research an optimal Lap concentration of 1% w/v was selected for this study due to its enhanced mechanical properties and excellent biocompatibility [38]. Consequently, the 3D bioprinted scaffold in this study was fabricated with a Lap concentration of 1% w/v. Fig. 2A presents macroscopic and optical micrographs of the 3D printed scaffolds, including GelMA/Laponite (GL), GelMA/Laponite/USPIO@SiO₂ (GLU), CyP/MSNs/GLU (C + M/GLU), and CyP@MSNs/GLU (C@M/GLU). The integration of USPIO@SiO₂ imparted a brown coloration to the scaffold, whereas the inclusion of CyP@MSNs gave it a pale green hue. Remarkably, the overall scaffold morphology was well-preserved. Scaffold degradation behavior plays a crucial role in regulating drug release rates and facilitating tissue regeneration. Achieving an appropriate degradation rate can significantly enhance the success of bone regeneration. To investigate this, the degradation behavior over 56 days was monitored by immersing the scaffolds in a phosphate-buffered saline (PBS) solution. In vitro degradation curves revealed that the bare hydrogel had the fastest degradation rate. Nevertheless, with the incorporation of USPIO@SiO₂ and CyP@MSNs, the composite hydrogels exhibited prolonged degradation times, which is more favorable for long-term bone repair (Fig. S6). The stress-strain curve of the composite hydrogel is depicted in Fig. 2B. Incremental addition of USPIO@SiO₂ and CyP@MSNs led to an increased Young's modulus of the scaffold, reaching values of 261.37 ± 11.01, 332.72 ± 9.29 and 380.04 ± 16.30 kPa, respectively (Fig. 2C). Inductively coupled plasma atomic emission spectrometry (ICP-AES) was used to study the ion release from the C@M/GLU scaffold. As shown in Fig. 2D, the cumulative percentage of Si and Mg ions released from the scaffold increased with prolongation immersion time, and both ions were released rapidly within 14 days, while the Fe ion release couldn't be detected. The sustained release of Si and Mg ions within the scaffold is advantageous as it stimulates neovascularization and prompts bone formation [39,40]. In vitro release of CyP from C + M/GLU and C@M/GLU scaffolds was assessed via UV spectrophotometry. As depicted in Fig. 2E, the release of CyP from the C + M/GLU scaffold was initially rapid within 10 days, with a cumulative release of 59.20% ± 1.53%. Subsequently, the release rate of CyP considerably slowed during the following period. In contrast, CyP release from the C@M/GLU scaffold continued until day 57. Therefore, MSNs play a pivotal role in achieving a controlled and slow release of CyP, which is conducive to the ongoing monitoring of ALP.

Fig. 2.

Characterization of scaffolds. (A) Representative images of GL, GLU, C + M/GLU and C@M/GLU scaffolds. (B and C) Corresponding strain-stress curves (B) and compressive Young's modulus (C) for different scaffolds. (D) In vitro release of Si, Mg and Fe ions from C@M/GLU scaffolds. (E) In vitro release kinetics of CyP probe from the scaffolds. (F and G) NIR-FL images (F) and FL intensity (G) of scaffolds incubated with different enzyme activity units of ALP (λ_ex = 660 nm). (H and I) MR images (H) and relative signal intensity curve (I) of C@M/GLU scaffold with different Fe concentrations. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

To evaluate the in vitro imaging performance, C + M/GLU and C@M/GLU scaffolds with equivalent CyP content were fabricated. Simulating ALP secretion within the bone defect environment, the fabricated scaffolds underwent sequential immersion in solutions with equivalent volumes but varying ALP enzyme activity units. Subsequently, images were captured using a NIR imaging system. Our investigation revealed a positive correlation between the fluorescence intensity of both C + M/GLU and C@M/GLU scaffold groups and the increase in ALP enzyme activity units (Fig. 2F and G). Notably, the fluorescence intensity of the scaffolds in the C + M/GLU group was consistently exceeded that of the C@M/GLU group. This difference can be attributed to the swift release of CyP from C + M/GLU. However, in the short term, the presence of MSNs within the scaffold contributed to a reduced rate of CyP release. Furthermore, we explored scaffolds with different iron (Fe) concentrations using MR imaging. By utilizing USPIO@SiO₂ as a negative contrast agent, we observed an inverse correlation between the MR signal and Fe concentration (Fig. 2H and I). The modification of USPIO with a SiO₂ shell remarkably augmented the MR signal intensity of USPIO. At a Fe concentration of 2.128 mM, the MR signal intensity of the USPIO@SiO₂ group was 1.33 times higher than that of the USPIO group. This enhancement can be attributed to the robust affinity of SiO₂ for water, which enhances the chemical and thermal stability of USPIO@SiO₂, thereby influencing the MR signal [41]. These findings collectively establish the fundamental capability of C@M/GLU scaffolds for in vitro detection of ALP activity and monitoring of scaffold degradation using MR imaging.

2.3. In vitro biocompatibility, ALP responsiveness and ALP expression of scaffolds

To assess the biocompatibility of GL, GLU, and C@M/GLU scaffolds, BMSCs-laden scaffolds were cultured for 1, 3, and 5 days and then examined using the Cell Counting Kit-8 (CCK-8) and live/dead cell staining. Live/dead cell staining revealed a notable increase in the number of BMSCs in all three scaffold groups with prolonged culture time (Fig. 3A). Quantitative results revealed comparable optical density (OD) values across the three scaffold groups, affirming their ability to facilitate cell proliferation (Fig. 3B). This highlights the suitability of the C@M/GLU scaffold as a reliable platform for monitoring bone regeneration, owing to its excellent in vitro biocompatibility.

Fig. 3.

In vitro cell experiments. (A) In vitro cell activity evaluation of GL, GLU, and C@M/GLU scaffolds using BMSCs at days 1, 3, and 5. (B) BMSCs proliferation in different scaffolds by CCK-8 assay at days 1, 3, and 5. (C) ALP staining at days 7 and 14. (D) ALP activity on days 3, 7, and 14. (E and F) NIR-FL images (E) and FL intensity (F) of BMSCs-loaded scaffolds cultivated to 0, 3, 7, and 14 days. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

We assessed the in vitro osteogenic potential of the scaffolds by evaluating ALP activity. ALP serves as a pivotal phenotypic marker of osteoblasts, providing a direct indication of osteoblast activity and function. To enhance the precision of NIR-FL imaging, we conducted ALP staining and performed a quantitative analysis of ALP expression in BMSCs. The results of the ALP staining are presented in Fig. 3C. Remarkably, the C@M/GLU group substantially enhanced the osteogenic differentiation of BMSCs compared to the control group. In Fig. 3D, the ALP activity in the C@M/GLU group exhibited a substantial increase compared to the control, GL, and GLU groups on day 14. These results accentuate that the C@M/GLU scaffolds, enriched with Si⁴⁺ and Mg²⁺ content, significantly promote the osteogenic differentiation of BMSCs. In summary, these experiments affirm that CyP@MSNs-labeled scaffolds are well-suited for bone tissue engineering, offering FL capabilities.

We further evaluated the capability of CyP@MSNs to detect endogenous ALP originating from BMSCs. BMSCs play a crucial role in bone regeneration research, and their osteogenic differentiation is induced by a specific medium, resulting in increased ALP expression. To monitor this, we employed a confocal laser scanning microscope (CLSM) to measure the fluorescence intensity in BMSCs at various time intervals. As depicted in Fig. S7, a progressive increase in fluorescent signal in the cells over time (3, 7, and 14 days) was observed, indicating enhanced osteoblast differentiation. The cytoskeleton remained unaffected even after nanoparticle uptake, highlighting the low toxicity of CyP@MSNs. Subsequently, NIR imaging on the bio-scaffolds cultured at various time points was conducted. As anticipated, ALP secretion by BMSCs triggered CyP fluorescence within the scaffolds, leading to the illumination of C + M/GLU and C@M/GLU scaffolds with red fluorescence. In both the C + M/GLU and C@M/GLU groups, the fluorescence intensity increased over time, with more robust fluorescence signals observed on C@M/GLU scaffolds compared to C + M/GLU when the culture time reached 14 days (Fig. 3E). Notably, the fluorescence intensity of the scaffolds in the C@M/GLU group at day 14 was 2.9 times higher than that at day 0 (Fig. 3F). In contrast, the fluorescence intensity in the C + M/GLU group exhibited a decelerated upward trajectory between day 7 and day 14, significantly lower than that in the C@M/GLU group. This disparity may be attributed to the absence of the protective effect conferred by MSNs. The interaction between CyP and the surrounding environment could account for this accelerated quenching.

2.4. In vitro osteogenic effect of scaffolds

The osteogenic differentiation of BMSCs is a crucial process in directing bone formation and ensuring the functional integrity of bone tissue [42]. To evaluate the osteogenic differentiation ability of the scaffolds at the molecular level, BMSCs were cultured in an induction medium containing scaffold extracts for 7 and 14 days, respectively. Subsequently, quantitative real-time PCR (qRT-PCR) was utilized to assess the expression of key genes and proteins critical for osteogenesis, as depicted in Fig. 4A–E. The quantitative results revealed that following 7 and 14 days in the induction medium, the expressions of runt-related transcription factor 2 (RUNX2), ALP, osteocalcin (OCN), osteopontin (OPN), and collagen type I (Col1a) were significantly enhanced in the C@M/GLU group. The elevated expression of bone-related genes in the C@M/GLU group can be attributed to the release of Si⁴⁺ and Mg²⁺ from the scaffold. This emphasizes the scaffold's ability to promote osteogenic differentiation, a pivotal factor in bone tissue engineering. Additionally, it is noteworthy that the expression of RUNX2, ALP, OCN, OPN, and Col1a by BMSCs in the other groups demonstrated a time-dependent increasing trend, except for RUNX2. As a member of the RUNX family of transcription factors, RUNX2 plays a role in osteoblast differentiation and bone morphogenesis [43]. RUNX2 is predominantly expressed during the early stages of bone healing, which accounts for its reduced expression at day 14. Notably, we specifically quantified the mRNA expression level of ALP, revealing a substantial increase in the C@M/GLU group on both day 7 and day 14. In contrast, OCN manifests during the later stages of osteoblast differentiation, playing a role in binding to Ca²⁺ to regulate calcium homeostasis and bone mineralization. OPN stands out as one of the most abundant non-collagenous proteins in the bone matrix, produced by osteoblasts and osteoclasts, effectively stimulating osteoclastogenic and resorptive activity in mature osteoblasts. Additionally, Col1a is an extracellular matrix protein that promotes osteoblast adhesion and differentiation. Consistent with ALP expression, a parallel trend in the expression of OPN, OCN, and Col1a was observed among the GL, GLU, and C@M/GLU groups. These results indicate that the C@M/GLU group had the most significant stimulatory effect on the osteogenic differentiation of BMSCs. Furthermore, a Western blot analysis was conducted to assess the expression of related proteins during the early stages of osteogenic differentiation. The expression of RUNX2, ALP, OCN, OPN, and Col1a proteins, as depicted in Fig. 4F, aligned with the trend observed in the qPCR results. Compared to the GL group, the expression levels of these bone-associated markers were significantly elevated in the GLU group and the C@M/GLU group. Quantitative results of Western blot (Fig. 4G) revealed higher levels of RUNX2, ALP, OCN, OPN, and Col1a proteins in the BMSCs treated with C@M/GLU composite scaffolds compared to those treated with scaffolds in the GL and GLU groups. Based on the above data, it was speculated that various paracrine factors including interleukin 8 (IL-8) and transforming growth factor-β1 (TGF-β1) might be released under the stimulation of Si and Mg ions. Furthermore, these factors are involved in promoting the recruitment, proliferation, and osteogenic differentiation of endogenous BMSCs through multiple signaling pathways, potentially promoting bone regeneration [44,45]. Altogether, these findings affirm that the release of Si⁴⁺ and Mg²⁺ from the scaffolds has a positive impact on the enhancement of osteogenic activity [46,47].

Fig. 4.

Evaluation of the osteogenic potential of different scaffolds. (A–E) Effect of the conditioned medium from GL, GLU, and C@M/GLU scaffolds on the osteogenic gene expression in BMSCs at 7 and 14 days, including RUNX2 (A), ALP (B), OCN (C), OPN (D), Col1a (E). (F and G) Representative Western blot images (F) and quantitative analysis (G) of RUNX2, ALP, OCN, OPN, and Col1a protein expression in BMSCs cultured with the conditioned medium at day 14. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

2.5. In vivo NIR-FL/MR imaging analysis

The reliability and biocompatibility of the C@M/GLU scaffold prompted us to further explore its potential for in situ monitoring during bone repair. Customized C + M/GLU and C@M/GLU scaffolds were implanted into bone defects with a diameter of 5 mm in each SD rat. Subsequently, NIR-FL images were captured from the rats on days 0, 3, 7, and 14, respectively (Fig. 5A). On day 0 of implantation, the scaffolds exhibited minimal fluorescence due to the low level of ALP in the bone defect and the absence of CyP fluorescent molecules released. On day 3, ALP secretion increased as the scaffold-loaded BMSCs proliferated and differentiated. The reaction of the produced ALP with CyP released from the MSNs resulted in enhanced fluorescence intensity. Notably, the fluorescence intensity of the C + M/GLU group surpassed that of the C@M/GLU group because the CyP fluorescent probe in this experimental group was released directly from the hydrogel to come in contact with ALP. The fluorescence intensity of the C@M/GLU group's scaffolds remained robust on days 7 and 14, facilitating the monitoring of ALP expression at the defect site due to the protective role of the MSNs. However, the fluorescence intensity of the C + M/GLU group of scaffolds decreased rapidly on day 7, aligning with the in vitro fluorescence imaging results of the scaffolds (Fig. 5B). The reason for this finding may be that the CyP fluorescent molecules were directly exposed to the external environment, which caused a rapid decrease in their fluorescence signals. In contrast, the fluorescence intensity of the C@M/GLU scaffolds on day 14 was 34 times higher than that on day 0 (Fig. 5C). This suggests that the encapsulation of MSNs results in a more stable fluorescent probe, thereby enhancing the accuracy of early ALP imaging. These NIR-FL images provide substantial evidence of the C@M/GLU scaffold's ability to execute continuous in vivo ALP monitoring.

Fig. 5.

In vivo NIR-FL/MR imaging study of different scaffolds by a rat calvarial defect model. (A) Schematic of NIR-FL/MR image acquisition at different time points. (B and C) NIR-FL images (B) and FL intensity curves (C) of the ALP expression in bone defect regions at different times (λ_ex = 660 nm). (D and E) Immunofluorescence staining (red) (D) and quantitative analysis (E) of ALP expression. The cell nuclei were stained with DAPI (blue). (F and G) MR images (F) and signal intensity curves (G) of scaffold degradation monitoring. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

To validate the accuracy of CyP@MSNs for monitoring ALP expression, we conducted additional validation through ALP immunofluorescence staining. As depicted in Fig. 5D, the ALP fluorescence signals exhibited a similar increasing trend in both groups for 0, 3, 7, and 14 days, indicating the progressive formation of new bone over time. This is visually evident in Fig. 5E, where the highest ALP expression was observed in the scaffolds of both the C + M/GLU and C@M/GLU groups on day 14. The sustained release of Si⁴⁺ and Mg²⁺ following implantation may contribute to the remarkable bone regeneration properties observed in the C+M/GLU and C@M/GLU scaffolds. The elevated expression of ALP in rats during the early post-implantation period signifies a more pronounced differentiation of pre-osteoblasts into mature osteoblasts [48]. In summary, in conjunction with the in vivo NIR-FL imaging results, it is reasonable to conclude that the fabricated C@M/GLU scaffolds are effective for accurately visualizing early ALP expression.

Real-time monitoring of material degradation is essential for determining in vivo retention times and for the design and evaluation of degradable biomaterials. Generally, scaffolds undergo a significantly more intricate degradation process in vivo compared to in vitro, primarily due to variations in environmental factors. Over time, the structural integrity and quality of hydrogel scaffolds often diminish [49]. MR imaging is a non-invasive imaging modality commonly utilized for continuous hydrogel scaffold degradation. To assess the in vivo degradation of these hydrogel scaffolds, GL, GLU, C + M/GLU, and C@M/GLU scaffolds with a 5 mm diameter were implanted into the skull defects of rats. Subsequently, MRI was employed to monitor these scaffold groups over 5 weeks. It is noteworthy that scaffolds typically induce an inflammatory response within the first hour to four days after implantation, followed by a gradual recovery [50,51]. As illustrated in Fig. 5F, both the control and GL groups exhibited elevated MR signals at week 0 in comparison to the GLU, C + M/GLU, and C@M/GLU groups. This increase can be attributed to the initial inflammatory edema. As the inflammatory response subsided, the scaffold began to degrade, facilitating the formation of new bone in the surrounding area. This degradation process was evident in the decreasing MR signal over time. In contrast, the GLU, C + M/GLU, and C@M/GLU scaffold groups exhibited lower MR signals at week 0, attributed to the higher USPIO@SiO₂ concentration. Subsequently, the MR signals in all groups displayed a pattern of initial increase followed by a decrease. This suggests that during the initial phase, the reduction of USPIO@SiO₂ was primarily associated with hydrogel degradation. Conversely, in the later stages, the MR signal diminished as it merged with the adjacent bone tissue, signifying the development of new bone. Quantitative analyses revealed that the peak of the MR signal was observed in the second week, with signal intensity decreasing to its minimum across all groups by the fifth week, consistent with the findings of in vivo imaging (Fig. 5G). Subsequently, analyzed the correlation between the in vitro mass retention rate of the scaffolds at 2, 4, and 5 weeks and MR signal intensity. The results in Fig. S8 indicate that within weeks 2–5, scaffold degradation is accompanied by a decrease in MR signal, demonstrating a strong correlation between them. This reaffirms the feasibility of MR imaging for monitoring scaffold degradation.

2.6. In vivo bone regeneration assessment

Considering the demonstrated osteogenic potential in vitro with C@M/GLU scaffolds, a rat cranial bone defect model was implemented to comprehensively assess in vivo osteogenic efficacy. Rats were euthanized at 5- and 10-weeks post-implantation and the bone regeneration was evaluated through micro-CT imaging, histological staining, and immunofluorescence (Fig. 6A). The reconstructed 3D micro-CT images of the cranial regions at 5 and 10 weeks display a progressive growth of new bone within the defect area. At week 5, the control group, exhibited delayed fracture healing, resulting in the largest bone defect area. Additionally, the new bone area in both the C + M/GLU and C@M/GLU groups was comparable and slightly larger than that in the GL and GLU groups (Fig. 6B). After 10 weeks of implantation, all groups displayed an increased coverage of new bone compared to the groups at 5 weeks, with the C + M/GLU and C@M/GLU groups exhibiting the smallest bone defect areas. This pattern remained consistent when analyzing the coronal view. The progressive increase in new bone tissue over time was further substantiated by evaluating bone mineral density (BMD) and the percentage of bone volume (BV/TV) (Fig. 6C and D). Quantitative analyses demonstrated that the BV/TV and BMD of the C + M/GLU and C@M/GLU groups were significantly superior to those of the GL and GLU groups at 10 weeks, underscoring their superior osteogenic performance.

Fig. 6.

In vivo evaluation of bone regeneration with different scaffolds by a rat calvarial defect model. (A) Schematic representation illustrating the design of rat calvarial bone defect repair. (B–D) Representative 3D-reconstructed images (B) obtained from micro-CT scanned calvarial bone after scaffold implantation for 5 and 10 weeks, including transverse and coronal sections. Analysis of BV/TV (C) and BMD (D) analysis of microstructural parameters of regenerated bone tissue. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

After the micro-CT analysis, the formation of new bone at the defect site was subsequently examined through hematoxylin and eosin (H&E) staining and Masson's trichrome staining (Fig. 7 A). After 10 weeks of implantation, H&E stained sections were observed, and no obvious inflammatory reaction and necrosis were observed in each group, indicating that the scaffold had good biocompatibility. The control group exhibited substantial fibrous tissue bridging the defect edge. Importantly, all scaffold-containing groups displayed varying degrees of new bone formation. Significantly greater new bone formation was observed in the GLU, C + M/GLU, and C@M/GLU groups compared to the GL group. In both the C + M/GLU and C@M/GLU groups, the defects were predominantly filled with newly formed bone, emphasizing their comparable regenerative potential. These results indicate that the C + M/GLU and C@M/GLU groups containing higher bioactive ions were more conducive to the repair of bone defects repair.

Fig. 7.

Histological and immunofluorescence staining observation of decalcified bone slices after implantation. (A) Histological analysis of cranial sections after 10 weeks of implantation, including H&E staining and Masson's trichrome staining. (B) Immunofluorescent staining of OCN (red) and OPN (red) after 5 weeks of implantation. The cell nuclei were stained with DAPI (blue). (C and D) Quantitative analysis of positive areas of OCN (C) and OPN (D) expression. Statistical analysis was performed using the one-way analysis of variance (ANOVA) followed by Tukey's method (n = 3, *p < 0.05, and **p < 0.01).

Immunofluorescence staining was employed to further assess the effect of C@M/GLU scaffolds on bone regeneration. As shown in Fig. 7B, the signals of osteoblast markers detected in the control group were significantly lower than those in the GL, C + M/GLU, and C@M/GLU groups. OPN and OCN were expressed at the highest levels in the C + M/GLU and C@M/GLU groups, suggesting that more bone tissues were formed. The sustained release of Si⁴⁺ and Mg²⁺ was responsible for the excellent bone regeneration properties of the scaffold. Five weeks after implantation, the relative fluorescence intensity of OCN in control, GL, GLU, C + M/GLU, and C@M/GLU groups was 0.31 ± 0.11%, 1.21 ± 0.34%, 1.33 ± 0.33%, 1.86 ± 0.13%, and 2.30 ± 0.26%, respectively (Fig. 7C). A consistent trend was observed for OPN expression (Fig. 7D). The expression of OCN and OPN in the C@M/GLU group was 7 and 3 times higher than the control group, respectively. In conclusion, through the analysis of the results of CT, H&E, Masson's trichrome, and immunofluorescence staining, it was demonstrated that the C@M/GLU scaffold possesses the dual function of monitoring bone regeneration in situ and promoting bone regeneration. This establishes it as an ideal therapeutic self-monitoring scaffold.

So far, many kinds of cleverly designed scaffolds have been prepared for bone defect repair, such as 3D-printed imitation flowerbed-inspired biomimetic scaffolds and lotus root-like biomimetic scaffolds [42,52]. Although these scaffolds have achieved a good bone repair effect, they cannot be followed up in real time after implantation to obtain key repair parameters, such as the progress of bone regeneration and the degradation of scaffolds, which limits the further development of tissue engineering scaffolds. In this study, the 3D-printed C@M/GLU scaffold had dual functions of monitoring and treatment, which not only promoted bone regeneration through the release of Si and Mg ions but also realized in-situ non-invasive monitoring of bone regeneration process through NIR-FL imaging and MRI. In addition, it has been shown that Si and Mg ions released from composite scaffolds can also potentially promote angiogenesis, which is essential for the growth and healing of bone tissue [53,54]. However, the emission wavelength of the CyP fluorescent molecule used in this study is in the near-infrared I region. Still, there is a need for continuous development of fluorescent probes with lower background noise and scattering effects for deeper tissue imaging. In addition, multiple fluorescent probes should be employed simultaneously to label multiple markers to help reveal more complex physiological and pathological mechanisms in organisms in the future. MRI contrast agents with higher contrast and sensitivity should be developed for more accurate monitoring of scaffold degradation. In summary, we constructed a multi-modal imaging scaffold that can dynamically monitor bone regeneration and scaffold degradation.

3. Conclusion

In summary, we have successfully fabricated an integrated C@M/GLU scaffold by leveraging the synergy of 3D printing and imaging technology. This innovative approach allows real-time monitoring of bone regeneration progress. Furthermore, the incorporation of Lap, CyP@MSNs, and USPIO@SiO₂ has significantly enhanced the mechanical strength of C@M/GLU scaffolds. Of particular significance is the effective induction of bone regeneration facilitated by the release of Si⁴⁺ and Mg²⁺ into the bone defect microenvironment. The fluorescence intensity of the C@M/GLU scaffold demonstrated a positive correlation with the expression of ALP in the bone defect region. Benefiting from the protection provided by MSNs, CyP demonstrated prolonged monitoring capabilities and robust in vivo fluorescence, enabling accurate tracking of ALP. Additionally, the incorporation of contrast agent USPIO@SiO₂ enabled non-invasive monitoring of hydrogel scaffold degradation. The continuous degradation of the hydrogel in vivo and loss of USPIO@SiO₂ were monitored by acquiring MR signals at various time points. Meanwhile, histological analysis confirmed the existence of new bone tissue at the defect site, providing additional validation of the accuracy and feasibility of MR imaging monitoring. Overall, the functional C@M/GLU scaffolds we have developed, incorporate both imaging and osteogenic enhancement functions, enabling non-destructive, continuous monitoring of bone regeneration without sacrificing animals. This approach presents a practical strategy for advancing the next generation of bone tissue engineering scaffolds, equipped with the capability to monitor bone tissue regeneration in real time.

Data availability statement

The data supporting the findings of this study can be made available by the corresponding author upon reasonable request.

Ethics approval and consent to participate

This study does not involve human participants, human data, or human tissues. All the experimental animal procedures were performed by following local animal welfare laws and guidelines and approved by the Animal Ethics Committee of Donghua University (No: DHUEC–NSFC–2020-14).

CRediT authorship contribution statement

Qian Feng: Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Kanwal Fatima: Writing – review & editing. Ai Yang: Methodology, Investigation, Formal analysis. Chenglin Li: Validation. Shuo Chen: Methodology, Conceptualization. Guang Yang: Methodology, Conceptualization. Xiaojun Zhou: Writing – review & editing, Methodology, Funding acquisition. Chuanglong He: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article.