Self-Assembled Nanoclay Gel with Spheroid MSC-Derived Exosome Mimetics to Integrate Demineralized Bone Matrix and Noggin-Targeting miRNA for Synergistic Osteogenesis

Abstract

Demineralized bone matrix (DBM) is widely used as an alternative to autografts for repairing bone defects, but its clinical efficacy is limited by poor retention at the defect site and inadequate osteogenic capacity. Here, we developed a multifunctional carrier system aimed at improving the capacity of DBM by utilizing nanoclay-based self-assembly along with cell-derived exosome mimetics (EMs). EMs were derived from mesenchymal stem cells (MSCs) osteogenically induced in spheroids and were functionalized with bisphosphonates (BP) to assemble a stable gel network with laponite nanoclays via BP-nanoclay edge interactions. The resulting self-assembled nanoclay gel exhibited excellent injectability, moldability, and self-healing properties, even when loaded with a high concentration of DBM, and supported MSC osteogenic differentiation. We further enhanced the potency of DBM-mediated osteogenesis and bone morphogenetic protein (BMP) signaling in the nanoclay gel by incorporating BP-functionalized EMs (EM-BPs) loaded with miR-200c that can target BMP antagonist noggin. Lastly, we validated the bone regeneration efficacy of DBM-loaded nanoclay gels in comparison to commercially available DBM putties using a mouse calvarial defect model. This approach presents a versatile nanoclay gel carrier platform that overcomes the limitations of current DBM formulations and provides a promising strategy for improved bone repair.

1. Introduction

Bone defects, which are caused by trauma, tumor resection, and metabolic disorders, are very common in clinical practice. The demand for bone regeneration continues to increase due to the aging population[1–3]. Contributed from their intrinsic osteoconductive and osteoinductive properties, autografts are widely considered as the gold standard for the treatment of bone defects [4–6]. However, their clinical application is significantly restricted by limited availability, donor-site morbidity, and extended surgical time since these grafts are harvested from the patient[7]. Allografts such as demineralized bone matrix (DBM) have been extensively utilized as alternatives for autografts [8, 9]. DBM is produced by removing mineral components from donor bones while preserving the collagen network and endogenous bioactive factors, including osteoinductive bone morphogenetic protein-2 (BMP-2). The porous three-dimensional (3D) structure of DBM generated from the demineralization process enables cellular infiltration, proliferation, and osteogenic differentiation, which confers DBM with osteoconductive capacity. Most commercially available DBM is provided in particulate form. When implanted directly, these particles may disperse or be washed out by blood or interstitial fluids, limiting their retention at the defect site[10]. Clinical products often combine DBM with binding agents such as glycerol or hyaluronic acid to form a putty-like material for improved handling characteristics. However, these formulations lack physical stability and rapidly disintegrate in dynamic biological settings, resulting in a suboptimal concentration of DBM and reduced osteogenic potential at the target site. Self-healing hydrogels have emerged as promising carriers for drug delivery and tissue engineering due to their dynamic crosslinking networks that prevent premature leakage of bioactives and facilitate conformal filling of irregular defects with enhanced longevity[10–14]. Self-healing materials possess capability to rapidly restore their structural integrity after mechanical damage, demonstrating significant potential as a moldable vehicle for DBM.

The osteogenic efficacy of DBM products sometimes is inadequate in clinical practice, especially when the defect size is critical and the patients have impaired bone metabolism. Approaches to enhance the efficacy of DBM function include incorporating exogenous BMP-2 and transplanting living cells. However, the incorporation of exogenous BMP-2 may increase the risks of ectopic ossification, adipogenesis, and dysregulated feedback inhibition[15, 16]. Meanwhile, cell transplantation also faces challenges such as low viability of the transplanted cells, poor integration with host tissue, and the risks of immune rejection and tumorigenicity[17, 18]. Many studies indicated that the therapeutic effects of these transplanted cells are greatly achieved via paracrine signaling, especially via small vesicles such as exosomes[19, 20]. The exosomes contain biological factors such as bioactive proteins and RNAs that can modulate osteogenic differentiation of host cells[21]. Compared to transplanted cells, the exosomes have improved stability, lower immunogenicity, and are free of tumorigenic risk[22]. Moreover, miRNAs incorporated in exosomes provide a possibility to further improve their therapeutic potential. While DBM naturally contains BMP-2, the associated signaling cascade triggers compensatory upregulation of extracellular antagonists such as noggin, which binds BMPs with high affinity and robustly suppresses downstream signaling[23, 24]. This antagonistic effect significantly limits the osteogenic potential of DBM, thus strategies to attenuate BMP antagonism are needed. miRNAs such as miR-148 and miR-200c are widely reported to downregulate noggin expression and therefore enhance BMP signaling[25–29]. Exosomes provide an ideal carrier for these regulatory miRNAs, which enables targeted inhibition of BMP antagonists and thereby improves DBM-mediated osteogenesis.

Based on these insights, we hypothesized that the integration of DBM and noggin-targeting miRNA-loaded exosomes in a self-healing nanoclay gel carrier would achieve synergistic osteogenesis. To test this concept, we engineered a triphasic hybrid nanoclay gel composed of laponite (LAP) nanoclay, bisphosphonate-functionalized exosome mimetics (EM-BPs) loaded with noggin-targeting miRNA (miR-200c), and DBM. We produced EMs derived from osteogenically induced spheroid MSCs and modified them using 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-poly(ethylene glycol)-bisphosphonate (DSPE-PEG-BP) to enable coordination with cations such as Ca²⁺ and Mg²⁺ ions. This coordination chemistry drives the spontaneous formation of a strong dynamic coordination network between BP groups on EM-BPs and Mg²⁺ ions on the edge of LAP nanosheets, yielding a stable nanoclay/EM-BP gel with excellent injectability and self-healing behavior. As illustrated in Scheme 1, this integrated system is designed to enhance BMP signaling by suppressing noggin through miR-200c delivery, leverage the inherent osteogenic potential of DBM and EMs, and provide a mechanically stable, self-healing, and injectable platform for localization of DBM particles. We systematically investigated the physicochemical properties of the nanoclay gel, its effects on MSC proliferation and osteogenic differentiation, the interplay between DBM-derived cues and miRNA-enhanced BMP signaling, and its therapeutic performance in a mouse calvarial defect model. Our results demonstrate that integrating miRNA-mediated molecular regulation with DBM-derived biochemical cues within an EM-BP/nanoclay-based gel offers a powerful and synergistic strategy for bone regeneration.

Scheme 1.

Schematic illustration of the synergistic osteogenic strategy achieved by integrating DBM with noggin-targeting miRNA-loaded EM-BPs within a self-assembled LAP/EM-BP nanoclay gel. (a) Donor cells were preconditioned through spheroid culture under osteogenic induction to enhance their osteogenesis. (b) EMs were generated by serial extrusion of the preconditioned spheroids through polycarbonate membranes with pore sizes of 5 μm and 1 μm. During extrusion, DSPE-PEG-BP was incorporated into the EM membranes, and noggin-targeting miR-200c was loaded into the vesicles. (c) Upon mixing LAP nanoparticles with EM-BPs, a robust nanoclay gel self-assembled via coordination interactions between BP groups on EM-BPs and Mg²⁺ ions exposed at the edges of LAP nanosheets. BMP-2 released from DBM naturally induces compensatory upregulation of extracellular antagonists such as noggin, which binds BMPs with high affinity and strongly attenuates BMP signaling. The released miR-200c suppresses noggin mRNA expression, thereby relieving BMP inhibition and amplifying DBM-mediated osteogenesis.

2. Results and discussion

2.1. Preparation and characterization of LAP/EM-BP nanoclay gels

To enhance the osteogenic potential of EMs, donor MSCs were preconditioned via spheroid culture under osteogenic induction. Compared with traditional 2D monolayer culture, 3D spheroid culture promotes markedly stronger osteogenic differentiation[30, 31]. Although the underlying mechanisms remain complex and somewhat controversial, the enhanced osteogenic capacity of 3D-cultured MSCs is generally attributed to strengthened cell-cell interactions, increased extracellular matrix (ECM) deposition, augmented paracrine signaling, and more physiologically relevant mechano-transduction[32, 33]. Previous studies have shown that osteogenesis-associated pathways such as Wnt/β-catenin and Notch signaling are upregulated in 3D culture systems, likely due to the elevated intercellular communication and microenvironmental cues that more closely mimic native tissue architecture[34, 35]. Numerous studies have demonstrated that the osteogenic efficacy of exosomes is closely linked to the differentiation status of their donor MSCs[36, 37]. This may be attributed to the fact that MSCs release exosomes with distinct molecular signatures as they progress through osteogenic differentiation. For instance, comparative profiling has shown that the top abundant miRNAs within exosomes strongly mirror those of their originating cells, indicating a high degree of cargo concordance between exosomes and their donor MSCs. Our previous work also demonstrated that EMs derived from osteogenically induced 3D spheroids exhibit significantly superior osteoinductive effects compared with EMs obtained from 2D cultures treated with osteogenic supplements[38]. Mechanistic analyses further indicated that these 3D-derived EMs possess a distinct miRNA profile enriched for regulators of the Wnt/β-catenin and Notch signaling pathways.

Because traditional ultracentrifugation-based methods typically produce low exosome yields, we utilized a serial extrusion strategy, which has been demonstrated to achieve substantially higher yields[38–40]. To enable formation of a stable hybrid gel with LAP nanoclay, DSPE-PEG-BP was inserted into the EM membrane during extrusion. As shown in Figures 1a and 1b, BP modification did not disrupt vesicle size or structural integrity. Both samples exhibited a narrow size distribution with relatively low polydispersity index (PDI) value, and cryo-transmission electron microscopy (cryo-EM) imaging confirmed preserved spherical morphology (Figures 1a–1c). The increase in zeta potential after BP insertion indicated successful surface functionalization (Figure 1d). Fourier transform infrared spectroscopy-attenuated total reflectance (FTIR-ATR) analysis of EM-BP further confirmed the successful incorporation of BP groups, as evidenced by the appearance of characteristic DSPE-PEG-BP peaks within the 800–1200 cm⁻¹ and 1250–1500 cm⁻¹ regions (Figure 1e). The incorporation of BP into EMs did not significantly alter the bioactivity of EM-BPs (Figure S1). In this study, EM preparation was precisely controlled through serial extrusion, which forces cells to pass through filters to generate large-scale vesicles with high uniformity, thereby addressing the issues of low yield and poor reproducibility associated with natural exosomes. The resulting EMs consistently exhibited highly reproducible size distributions and spherical morphology, as confirmed by DLS and cryo-EM characterization. In addition, EMs contain natural cargo directly derived from the donor cells in spheroids. Donor MSCs were cultured using agarose microwell templates to ensure consistent spheroid size and viability. This spheroid culture system consistently produced EM populations with highly reproducible bioactivity and molecular cargo profiles in our previous study [38].

Figure 1.

Preparation and characterization of EM-BPs and LAP/EM-BP nanoclay gels. (a-b) Size distribution of (a) EMs and (b) EMs-BP. Insets show representative cryo-EM images of each sample. (c-d) PDI and zeta potential of EMs with and without BP modification. (e) FTIR-ATR spectra of EMs, EM-BP, and DSPE-PEG-BP. (f) Schematic illustration of the self-assembly process between LAP nanoparticles and EM-BP. (g) Photographs demonstrating the structural stability of LAP solution versus LAP/EM-BP gels, accompanied by cryo-EM images showing the microstructure of LAP and its interactions with EM-BP. (h-k) Photographs demonstrating (h) injectability, (i-j) moldability, and (k) self-healing capability of nanoclay gels. (l) Compressive stress-strain curves of nanoclay gels containing different EM-BP concentrations. (m) Compressive modulus of the corresponding gels. (n-o) Frequency-sweep and strain-sweep rheological analyses of nanoclay gels. (p-r) Cyclic strain tests highlighting the self-recovery behavior of the gels. (s) Viscosity measurements showing shear-thinning properties of the nanoclay gels. Mean ± SD, n = 6, *p < 0.05, **p < 0.01.

When EM-BP was mixed with LAP nanoparticles, the two components rapidly self-assembled into a stable gel network via LAP-LAP electrostatic interactions and reversible coordination bonding between BP groups on the vesicle membrane and Mg²⁺ ions exposed on the positively charged edges of LAP (Figure 1f). LAP is a synthetic layered silicate presenting negatively charged basal surfaces which originate from substitution of Mg²⁺ with Li⁺ in the octahedral layer and positively charged particle edges resulting from protonated -OH groups under physiological pH. This intrinsic face-negative/edge-positive charge enables LAP-LAP electrostatic interactions, driving LAP gelation in water (Figure 1g). Previous studies have shown that LAP edges can form coordination complexes with BP-functionalized biopolymers[41]. Consistent with this, EM-BP produced a markedly more stable nanoclay gel than unmodified EMs (Figure S2). The Incorporation of EM-BP significantly enhanced network integrity, and cryo-EM imaging revealed close interactions between LAP nanosheets and EM-BP surfaces (Figure 1g). In contrast, EMs lacking BP functionalization showed minimal reinforcement (Figure S2). Figure S3 demonstrated that nanoclay gels with > 100 μg/mL EM-BP retained their structural integrity while the gels with low EM-BP loading dissolved in culture medium or phosphate-buffered saline (PBS) after 4 h of immersion. The EM-BP concentration-dependent stability of the gels further confirmed the critical role of BP-LAP coordination in stabilizing the gel structure.

Due to the combined enhancement effects of reversible electrostatic interactions and dynamic coordination bonds, the nanoclay gels exhibited excellent injectability, moldability, and self-healing capability (Figures 1h–1k). Figure 1h shows that the nanoclay gel was injected using a 23-gauge needle smoothly and retain filamentous structure in PBS, which indicates great potential for application in minimally invasive injection. Figures 1i–1j indicated that the gels can be injected to fill molds with different shapes and self-healed to form integrated shapes. To further demonstrate the self-healing capability, red- and blue-stained nanoclay gels were cut and recombined. The rejoined pieces self-healed within minutes to form an integrated gel with red and blue colors. The healed gels retained their structural integrity in simulated body fluid (SBF) and fetal bovine serum (FBS) after 24 hours of immersion. This indicates excellent self-healing capability of the gel contributed from strong dynamic BP-LAP coordination and LAP-LAP electrostatic interactions. These results indicate that the LAP/EM-BP nanoclay gel with excellent injectability, moldability, and self-healing capability showed great potential for minimally invasive treatment of bone defects with irregular shapes.

Mechanical and rheological properties of the nanoclay gels further confirmed the reinforcement effects of BP-LAP coordination. An increase in the concentration of EM-BP from 300 μg/mL to 3000 μg/mL led to a substantial increase in the compressive modulus of the gel, rising from 1.03 ± 0.25 kPa (300 μg/mL) to 1.30 ± 0.18 kPa (1000 μg/mL) and 1.84 ± 0.29 kPa (3000 μg/mL), respectively. This result indicates that EM-BP allows concentration-dependent crosslinking in nanoclay gel networks, forming stable coordination complexes (Figures 1l–1m). Frequency-sweep testing revealed that storage modulus was greater than loss modulus (G′ > G″) at 0.1–100 Hz of frequency and at 1% of strain, confirming solid states of the formed gels (Figure 1n). Strain-sweep results showed crossover of G′ and G″ at a stain of 20–30%, which indicated the gel-sol transition of a typical physically crosslinked gel (Figure 1o). In cyclic strain tests, the nanoclay gel repeatedly transitioned from solid-like states at low strain (G′ > G″) to liquid-like states at high strain (G″ > G′) and rapidly recovered upon strain reduction (Figures 1p–1r), demonstrating excellent self-healing capability. Furthermore, the gels displayed pronounced shear-thinning behavior (Figure 1s), a desirable feature for syringe injection and minimally invasive application. The degradation behavior of the LAP/EM-BP nanoclay gels was evaluated with nanoclay gels prepared with varied EM-BP concentrations. Nanoclay gels with 300 μg/mL EM-BPs rapidly lost their structural integrity after 21 days of incubation in PBS (Figure S4). The increase of EM-BP concentration to 1000 and 3000 μg/mL resulted in a significant mass loss of 72% and 43% after 4 weeks, respectively, indicating enhanced crosslinking of nanoclay gel networks. The release profile of EM-BPs was evaluated with LAP/EM-BP nanoclay gels prepared with 3000 μg/mL EM-BP. Approximately 43% of EM-BPs were slowly released in a sustained manner up to 21 days (Figure S5).

2.2. Biocompatibility and osteogenic potential of LAP/EM-BP nanoclay gels

To investigate the effects of EM-BP concentration on the biocompatibility and osteogenic potential of nanoclay gels, we first utilized EM-BPs derived from donor MSCs cultured under standard 2D conditions (Figure 2a). MSCs were encapsulated in nanoclay gels with varied EM-BP concentrations, and cell viability and proliferation were evaluated using Live/Dead and alamarBlue assays. Similar EMs demonstrated dose-dependent osteogenic effects up to 3000 μg/mL while maintaining high cell viability [38]. Given the stable formation of the nanoclay gel network at a minimum concentration of 300 μg/mL, we evaluated EM-BP concentrations ranging from 300 to 3000 μg/mL for subsequent studies. LAP/G-HA gels without EM loading were utilized as the control group[38]. Confocal Live/Dead images revealed consistently high cell viability across all the gel formulations over 1–7 days, with only a minimal number of dead cells observed (Figure 2b). The continuous increase in the live cell numbers over time suggests cell proliferation in the nanoclay gel. Quantitative results indicated that the cells proliferate around 2.5 times by day 4 and nearly 3.5 times by day 7, and no significant difference was detected among different groups (Figure 2c). These results suggest that dynamic nanoclay gels provide a supportive microenvironment for cell growth and the EM-BP incorporation does not compromise biocompatibility of the composite gel. We next evaluate the osteogenic potential of nanoclay gels with 300–3000 μg/mL EM-BPs. qRT-PCR results showed that encapsulated MSCs exhibited significantly higher osteogenic gene expression in EM-loaded gels than the EM-free control group (Figures 2d–2h). The expressions of alkaline phosphatase (ALP), runt-related transcription factor 2 (Runx2), and osteocalcin (OCN) remained similar among experimental groups with different EM-BP concentrations, whereas osteopontin (OPN) and alpha-1 type I collagen (COL1A1) expression increased with more EM-BP incorporation.

Figure 2.

Biocompatibility and osteogenic potential of nanoclay gels. (a) Schematic illustration of EM-BPs generated from 2D-cultured MSCs. (b) Live/dead staining of MSCs encapsulated in nanoclay gels and cultured for different time periods. Live cells are shown in green, and dead cells are shown in red. Scale bar = 100 μm. (c) Quantification of MSC proliferation within nanoclay gels. (d-h) Expression levels of osteogenic genes ALP, Runx2, OCN, OPN, and COL1A1 in MSCs cultured within nanoclay gels containing EM-BPs derived from 2D-cultured MSCs. (i) Schematic illustration of EM-BPs prepared from spheroid MSCs preconditioned under osteogenic induction. (j) ALP and ARS staining of MSC-laden gels across different treatment groups. Scale bar = 1 mm. (k-l) Quantitative analyses of ALP activity and ARS staining to assess osteogenic differentiation. (m-q) Expression of osteogenic genes ALP, Runx2, OCN, OPN, and COL1A1 in MSCs cultured within nanoclay gels incorporating EM-BPs derived from osteogenically induced spheroid MSCs. Mean ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

To further elucidate the osteogenic potential of the nanoclay gels, MSCs were encapsulated in nanoclay gels incorporated with 300–3000 μg/mL EM-BPs derived from MSC spheroids under osteogenic induction (Figure 2i). Nanoclay gels containing 300 μg/mL EM-BPs from 2D MSCs were used as the control. ALP, as an early osteogenic marker, was assessed on day 4, whereas mineral deposition, a late-stage indicator, was evaluated by alizarin red S (ARS) staining on day 14. As shown in Figure 2j, nanoclay gels with EM-BPs derived from osteogenically induced spheroid MSCs exhibited significantly enhanced ALP activity and mineralization, and effects increased with the increasing of EM-BP incorporation. Quantitative ALP and ARS results corroborated the findings (Figures 2k and 2l). In agreement with staining results, qRT-PCR analysis (Figures 2m–2q) showed significantly upregulated expression of osteogenic markers in gels loaded with spheroid-derived EM-BPs compared with the 2D EM-BP control. Consistent with previous findings in Figures 2d–2h, ALP, Runx2, and OCN exhibited comparable levels across nanoclay gel with varied EM-BP concentrations, while OPN and COL1A1 were upregulated in an EM-BP dose-dependent manner. This trend also aligns with the expression patterns of osteogenic genes in cells cultured within LAP/G-HA gels containing different EM concentrations, as published in our previous work[38]. The superior osteoinductive effect of spheroid-derived EMs strongly supports the concept that spheroid-derived EMs likely carry a more potent repertoire of osteogenic factors, translating into improved cell differentiation outcomes within the nanoclay gel.

2.3. Preparation and characterization of DBM-loaded LAP/EM-BP nanoclay gels

To evaluate the capacity of LAP/EM-BP nanoclay gels as carriers for DBM, we incorporated 7.5%, 15%, and 31% (w/v) DBM into the gels (Figure 3a). As shown in Figures 3b–3e, even at the highest DBM loading of 31%, the nanoclay gel retained excellent injectability, moldability, and self-healing properties. The DBM-laden gel could be smoothly extruded through a 14-gauge needle into PBS, forming continuous filaments that maintained structural stability. When injected into molds with different shapes, the filaments self-healed to form well-integrated geometries, indicating good moldability. Moreover, the cut blue- and red-stained DBM-loaded gels self-healed into integrated gels after being recombined, and the rejoined constructs retained their structural integrity in SBF and PBS. These results suggest that the dynamic and reversible crosslinking of the nanoclay network provides sufficient mechanical resilience to incorporate high concentrations of particulate DBM. Preservation of injectability, moldability, and self-healing capability is critical for DBM carriers to fill the irregular defect geometries and to localize DBM in the defect site.

Figure 3.

Preparation and characterization of LAP/EM-BP nanoclay gels incorporating DBM. (a) Schematic illustration of DBM incorporation into the self-assembled LAP/EM-BP nanoclay gel network. (b-e) Photographs demonstrating the (b) injectability, (c-d) moldability, and (e) self-healing capability of nanoclay gels containing 31% DBM. (f) Compressive stress-strain curves of nanoclay gels formulated with varying DBM concentrations. (g) Compressive modulus of the corresponding gels. (h-i) Frequency-sweep and strain-sweep rheological analyses of nanoclay gels incorporating DBM. (j-l) Cyclic strain tests illustrating the self-recovery behavior of DBM-containing gels. (m) Viscosity measurements demonstrate the shear-thinning properties of the LAP/EM-BP nanoclay gels with DBM. Mean ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

As expected, the incorporation of DBM altered the mechanical and rheological properties of composite gels. Upon DBM loading ranging from 7.5% to 31%, the compressive modulus of the nanoclay gel significantly increased from 1.84 ± 0.29 kPa to 5.59 ± 0.49 kPa (7.5%), 9.32 ± 1.28 kPa (15%), and 18.36 ± 1.06 kPa (31%), respectively. This result indicates that DBM particles act as reinforcing fillers in the gel, leading to a substantial increase in structural rigidity (Figures 3f and 3g). Frequency-sweep testing (Figure 3h) revealed that G′ was greater than G″ at a frequency of 0.1–100 Hz and strain of 1%, which indicates solid-like states. DBM incorporation markedly increased the dynamic modulus of the composite gels, suggesting enhancement effects of DBM. Strain-sweep measurements showed a shift in the G′/G″ crossover point from 20–30% strain in DBM-free gels (Figure 1o) to approximately 5% strain in DBM-loaded gels (Figure 3i), which suggests the reduced deformability of the composite gels due to the incorporation of DBM particles. The decrease in G′/G″ crossover point is consistent with the stiffening effect of DBM, demonstrating an advantageous balance between mechanical reinforcement and network reversibility. Cyclic strain testing at 1% and 200% further confirmed that DBM-loaded gels exhibited solid-like states (G′ > G″) under low strain and showed liquid-like states (G″ > G′) under high strain, and rapidly recovered upon strain reduction (Figures 3j–3l). Moreover, shear-thinning property of the nanoclay gels was preserved even at 31% DBM loading (Figure 3m). The retention of shear-thinning behavior and reversible gel-sol transitions even at high DBM loading ensures that the composite gel remains highly feasible for clinical use and can effectively localize DBM within bone defects.

2.4. Osteogenic effects of EM-BPs in DBM-loaded nanoclay gels

To further examine the osteogenic efficacy of EM-BPs in a cell-free therapeutic context, we compared LAP/EM-BP/DBM nanoclay gels containing low (300 μg/mL) and high (3000 μg/mL) concentrations of EM-BP, with a constant 31% DBM content. LAP/G-HA and LAP/G-HA with 31% DBM (without EMs) served as control groups[10]. As shown in Figures 4a–4f, qRT-PCR analysis revealed pronounced upregulation of osteogenic genes upon DBM incorporation, confirming the intrinsic osteoinductive activity of DBM. The nanoclay gel groups with EM-BP incorporation exhibited significantly upregulated osteogenic gene expression compared to the EM-free controls, and the high-dose EM-BP group showed the strongest osteogenic differentiation among all the samples. These results indicate the excellent osteogenic effects of EM-BPs and highlight their great potential for cell-free therapy to treat bone defects.

Figure 4.

Osteogenic effects of EM-BPs in DBM-loaded nanoclay gels. (a-f) Quantitative expression of osteogenic markers ALP, Runx2, OCN, BMP-2, OPN, and COL1A1 in MSCs cultured within the different nanocomposite gels. (g-h) Expression analysis of key genes in the Wnt/β-catenin signaling pathway. (i) Expression levels of the BMP antagonist noggin across the various treatment groups. Mean ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Interestingly, the expression of β-catenin, LEF-1, and noggin closely mirrored the trends of classical osteogenic markers described above (Figures 4g–4i). These results may be explained by our previous work, in which mechanistic studies demonstrated that EM-BPs derived from spheroid MSCs under osteogenic induction exhibit a distinct miRNA profile characterized by the downregulation of several Wnt/β-catenin-related miRNAs[38]. The qRT-PCR data obtained in the present study further supports the notion that EM-BPs promote osteogenic differentiation at least in part through the activation of the Wnt/β-catenin signaling pathway, consistent with the upregulation of β-catenin and LEF-1. Moreover, DBM-containing groups exhibited elevated noggin mRNA levels, suggesting activation of BMP-2 antagonism. Intriguingly, noggin expression was higher in the high-dose EM-BP condition than in the low-dose EM-BP group. This may reflect the complex crosstalk between Wnt/β-catenin and BMP signaling pathways[42, 43], in which activation of Wnt/β-catenin signaling can modulate BMP signaling and thus may regulate noggin expression as part of a compensatory regulatory feedback loop.

2.5. Osteogenic effects of nanoclay gels with DBM/noggin-targeting miRNA in vitro

The osteoinductive effect of DBM primarily arises from residual growth factors, most notably BMP-2, retained within its collagenous matrix[44, 45]. As illustrated in Figure 5a, BMP-2 released from DBM activates downstream osteogenic pathways but simultaneously induces the expression of endogenous BMP antagonists such as noggin. Noggin competitively binds BMP-2 and attenuates BMP receptor signaling[46–48], thereby diminishing the overall osteogenic efficacy of DBM. This antagonistic effect was confirmed in MSCs cultured with different concentrations of DBM. As shown in Figure S6, MSCs exhibited increasingly enhanced ALP activity at the DBM concentration of 0–0.5 mg/mL. However, the ALP activity decreased when a higher DBM concentrations of 1–2.5 mg/mL were used. qRT-PCR analysis of ALP expression corroborated the staining results. More importantly, noggin mRNA expression peaked at the DBM concentration of 0.5 mg/mL, at which the osteogenic enhancement began to decline. These results suggest that DBM exerts a dose-dependent dual effect, promoting osteogenesis at lower concentrations but becoming self-limited at higher doses. This may be attributed to BMP signaling antagonism mediated by excessive antagonist regulation.

Figure 5.

Osteogenic effects of nanoclay gels incorporating DBM and noggin-targeting miRNA. (a) Schematic illustration showing how miR-200c suppresses noggin expression, thereby relieving BMP antagonism and enhancing downstream BMP signaling. (b) ALP and ARS staining of MSCs cultured with 10 μg/mL EM-BPs containing varying amounts of miR-200c. Scale bar = 100 μm. (c-d) Quantitative analysis of ALP activity and noggin expression in MSCs treated with EM-BPs carrying different miR-200c loads. (e-i) Gene expression of osteogenic markers (ALP, Runx2, OCN, BMP-2) and the BMP antagonist noggin in MSCs cultured within nanoclay gels incorporating DBM and miR-200c-loaded EM-BPs. (j-l) Western blot analysis of pSmad1/5 and noggin in different treatment groups. Mean ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001.

Given the bottleneck from antagonistic effect, inhibition of BMP-2 antagonists such as noggin represents a potential strategy to enhance DBM-mediated osteogenesis. miRNAs offer high specificity in targeting mRNAs and thus downregulating protein expression. Among them, miR-200c has been reported to directly target noggin mRNA and reduce noggin protein levels[28, 29, 49]. In this study, we encapsulated miR-200c into EM-BPs during the extrusion process. To confirm the influence of miR-200c on osteogenesis, MSCs were cultured with 10 μg/mL of EM-BPs which contain different amounts of miR-200c. As shown in Figure 5b, ALP and ARS staining revealed dose-dependent effects of miR-200c on ALP activity and mineral deposition. qRT-PCR analysis confirmed that ALP expressions of different groups followed the same trend (Figure 5c). More importantly, noggin expression was downregulated when increasing miR-200c concentration in the cell culture (Figure 5d), suggesting the potent inhibitory effect of miR-200c on this BMP antagonist.

We further investigated the synergistic effects of DBM and miR-200c-loaded EM-BPs on osteogenesis (Figures 5e–5l). qRT-PCR results indicated that both LAP/EM/DBM and LAP/EM-BP/miRNA groups exhibited significantly higher osteogenic gene expression than the LAP/EM-BP control group, which highlights the independent osteogenic effects of DBM and miR-200c. Notably, the LAP/EM-BP/DBM/miRNA group showed the highest expression levels of key osteogenic markers such as ALP, Runx2, OCN, and BMP-2 among all the groups, demonstrating a strong synergistic effect achieved via the combined use of DBM and miR-200c-loaded EM-BPs within the nanoclay gel. As expected, the LAP/EM-BP/DBM group exhibited the highest noggin expression while the LAP/EM/miRNA group had the lowest value. This confirmed the DBM-induced BMP antagonist expression and downregulation effects of miR-200c on noggin. The noggin expression level in the LAP/EM-BP/DBM/miRNA group further confirmed that miR-200c effectively suppressed DBM-induced noggin upregulation. Western blot analysis (Figures 5i–5l) demonstrated that the loading of DBM or miR-200c into LAP/EM-BP nanoclay gels increased the levels of Smad1/5 phosphorylation up to 2.02- or 2.21-fold, respectively. The nanoclay gels loaded with both DBM and miR-200c led to a 5.87-fold increase in pSamd1/5 level. This was accompanied by a decreased expression of noggin. Wnt/β-catenin signaling is known to act as an upstream activator of BMP expression to regulate osteoblast differentiation [42, 43]. While Wnt signaling (from EM-BP) can cooperate with BMPs (from DBM) to promote osteogenesis in DBM-loaded nanoclay gels, excessive BMP signaling may trigger expression of BMP antagonists such as noggin in an auto-inhibitory manner. Thus, the inhibition of BMP antagonism by miR-200c could play an important role in the amplified BMP signaling observed in the LAP/EM-BP/DBM/miRNA group. These results provide mechanistic evidence that the integration of DBM with noggin-targeting miRNA-loaded EM-BPs not only enhances osteogenic signaling but also overcomes the biological limitations of DBM. The strategy offers a rational and highly effective approach to maximizing the therapeutic outcome of DBM grafts by modulating endogenous inhibitory pathways, thereby establishing a promising cell-free platform to treat bone defects.

2.6. Osteogenic effects of nanoclay gels with DBM/noggin-targeting miRNA in vivo

To evaluate the in vivo osteogenic potential of nanoclay gels incorporating DBM and noggin-targeting miRNA, we employed a mouse calvarial model with non-healing bone defects. Five groups were compared, that are LAP/EM-BP nanoclay gel alone, LAP/EM-BP gel with DBM, LAP/EM-BP gel with both DBM and miRNA, the commercial DBX graft control (31% DBM in HA), and a defect-only control. As shown in Figure 6a, micro-computed tomography (micro-CT) reconstruction after 8 weeks revealed clear new bone formation in all experimental groups. Among these, the LAP/EM-BP gel alone induced modest bone regeneration, whereas the addition of DBM markedly enhanced osteogenesis. Notably, the LAP/EM-BP/DBM/miRNA group exhibited the most extensive new bone formation. Although the commercial DBX graft promoted more bone formation than the defect-only control, its regenerative effect was inferior to that of our nanoclay gel systems. Quantitative analyses (Figures 6b–6e) of relative new bone formation, BV/TV, trabecular number, and bone mineral density (BMD) were consistent with the micro-CT observations. Specifically, the LAP/EM-BP/DBM/miRNA group achieved the highest relative new bone formation of 64.01%, BV/TV of 33.75%, trabecular number of 2.27, and BMD of 0.37, while the values of the defect-only control group were 7.88%, 5.82%, 0.53, and 0.10, respectively. These results highlight that the integration of noggin-targeting miRNA-loaded EM-BPs and DBM synergistically enhanced bone regeneration in vivo, surpassing conventional DBM-based grafts.

Figure 6.

Osteogenic performance of nanoclay gels incorporating DBM and noggin-targeting miRNA in vivo. (a) Representative 3D reconstructed micro-CT images of calvarial defects 8 weeks after implantation. (b-e) Quantitative micro-CT analyses of (b) relative new bone area, (c) bone volume fraction (BV/TV), (d) trabecular number, and (e) bone mineral density (BMD). (f) Representative images of H&E staining, Masson’s trichrome staining, and OCN IHC staining demonstrating new bone formation and matrix maturation (Green arrows: new bone; black arrows: DBM). Mean ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001.

Histological evaluation using hematoxylin and eosin (H&E), Masson’s trichrome, and OCN immunohistochemistry (IHC) staining provided complementary insights (Figure 6f). H&E staining demonstrated new bone formation in the groups of Gel, Gel/DBM, and Gel/DBM/miRNA, with the latter exhibiting the most extensive bridging of the bone defect. In contrast, the groups of defect-only control and DBX showed only minimal bone formation near the edges of defects. DBM particles were retained at the defect site when delivered through the LAP/EM-BP nanoclay gel, whereas fewer DBM particles were observed in the commercial DBX group administered using liquid carriers. This suggests that the nanoclay network provides stable suspension of DBM particles for reliable localized effects, preventing the premature dispersion of the implanted graft materials. DBM particles can induce inflammatory response if they wash away from the implantation site [50]. Histological evaluation showed normal tissue integration and progressive new bone formation without evidence of excessive inflammatory infiltration. DBM is characterized by its faster resorption rate compared to mineralized bones, typically occurring over 10 to 12 weeks [51]. The lack of mineral content makes the collagen matrix more susceptible to enzymatic degradation, enhancing its osteoinductive potential for fast bone remodeling. In addition, LAP nanoclays have demonstrated excellent biocompatibility without triggering adverse inflammatory response. Similar LAP gels have been shown to support MSC proliferation and osteogenic differentiation while maintaining favorable biocompatibility in bone-regenerative environments [10, 52]. Masson’s trichrome staining further demonstrated the maturation of newly formed bone in the Gel/DBM/miRNA group, as indicated by the development of red staining. IHC staining revealed strong OCN expression in all experimental groups, with the highest levels observed in the Gel/DBM/miRNA group. The enhanced OCN expression reflects accelerated mineralization and matrix maturation, further supporting the synergistic effects of miRNA-loaded EM-BPs and DBM in promoting efficient bone regeneration.

The current use of DBM is limited to relatively small defects, primarily due to its insufficient osteogenic potential with no cells and the inconsistent osteoinductive properties following processing, along with the possibility of graft migration. The LAP/EM-BP gel platform developed in this study may help address these challenges in multiple ways. First, the dynamic nanoclay gel serves as a moldable and self-healing carrier that enhances the localization and retention of DBM particles at the defect site, which is crucial for ensuring adequate osteoinductive stimulus during early stages of healing. Second, the addition of EMs derived from MSCs enables the delivery of concentrated natural cargos to promote bone healing, offering similar therapeutic benefits to stem cells without the risks of using living cells. Lastly, the loading of miRNAs amplifies BMP signaling in DBM-mediated bone repair without the exogenous application of worrisome BMP growth factors.

Taken together, these features suggest that the LAP/EM-BP gel could serve as a promising carrier platform for DBM, creating an appropriate osteogenic microenvironment conducive to successful bone regeneration. These strategies could broaden the clinical use of DBM and other bone graft products for improved skeletal tissue repair.

3. Conclusion

In this study, we engineered a novel LAP/EM-BP gel that self-assembled via coordination chemistry and electrostatic interactions as an effective carrier for DBM. The resulting nanoclay gels exhibited excellent injectability, moldability, and self-healing capability, even in the presence of high content of DBM. This nanoclay gel carrier effectively localized DBM particles in the bone defect site and prevented their dispersion. By integrating miR-200c-loaded EM-BPs and DBM, the composite nanoclay gel platform demonstrated synergistic enhancement of BMP signaling and bone regeneration, which is attributed to the inhibition of DBM-associated BMP antagonism and amplifying the osteogenic potential of the composite gels. Our design provides a bioactive, injectable, and easily handled carrier that overcomes the limitations of conventional DBM formulations and offers a promising approach for enhancing bone repair.

4. Experimental Section

Preparation of MSC spheroids

MSC spheroids were generated via the spontaneous aggregation of MSCs in the non-adhesive agarose microwells, following a protocol established in our previous work[38, 53]. Briefly, polydimethylsiloxane (PDMS; Changtikeji) was cast onto AggreWell^™800 templates (STEMCELL^™ Technologies) and cured for 24 hours to create precise negative molds. These PDMS molds were subsequently positioned over 5 mL of agarose solution (1.5–2 w/v%) in standard 12-well plates. After allowing the agarose to solidify completely, the PDMS molds were carefully removed, producing uniform non-adhesive microwells suitable for spheroid formation. MSCs suspended in osteogenic differentiation medium (DMEM supplemented with 10% FBS, 1% P/S, 50 μg/mL L-ascorbic acid, 10 mM β-glycerophosphate, and 100 nM dexamethasone) were then seeded to the prepared agarose microwells. The seeded cells aggregated to form spheroids within about 12 hours and were cultured for seven days before collection for EM production.

Preparation and characterization of EM-BPs

To increase the yield of EM-BP production, we utilized a serial extrusion approach, as reported in our previous studies[38, 40, 54]. Briefly, donor cells were collected and suspended in PBS. MSC spheroids were first collected by pipetting and the culture medium was removed by centrifugation at 1000 rpm for 3 min. The spheroids were then dissociated using trypsin–EDTA for 3 min, followed by the addition of culture medium to neutralize trypsin. The suspension was centrifuged again to remove trypsin and medium, and the resulting cells were resuspended in PBS at a density of 1 × 10⁶ - 5 × 10⁶ cells/mL. The cell suspension was then transferred to a 1 mL syringe and sequentially extruded through polycarbonate membrane with pore sizes of 5 μm and 1 μm (Whatman) using a mini-extruder system (Avanti Polar Lipids). To ensure that the produced EMs have uniform size and narrow size distribution, the cell suspension was extruded through each filter for at least 15 to 20 cycles. The extruded suspension was centrifuged at 4 °C at 10,000 rpm for 10 minutes to remove large cellular debris. The supernatant was collected and then centrifuged using a 100 kDa molecular weight centrifugal filter (EMD Millipore) at 2,500 rpm for 15 minutes and added with fresh PBS. This washing and concentration process was repeated three times to thoroughly remove residual components generated during the extrusion process. The prepared EMs were then stored at −80 °C for future use. The concentration of EMs was measured using a Pierce BCA protein assay kit (Thermo Fisher), and the EMs were diluted with PBS as needed. The particle size, zeta potential, and PDI of EMs were measured using a dynamic light scattering (DLS) instrument (Malvern). The morphology of EMs was imaged utilizing a cryo-EM (FEI Tecnai TF20) with an accelerating voltage of 200 kV. To insert DSPE-PEG-BP into the EM membrane or load miRNA in the vesicles, the DSPE-PEG-BP or miRNA was added to the cell suspension prior to extrusion. The functionalization of DSPE-PEG-BP on the EM membrane was characterized using the FTIR-ATR (Thermo Fisher).

Preparation of LAP/G-HA control hydrogel

Guanidine-modified hyaluronic acid (G-HA) was synthesized following a previously established method[38]. Specifically, 200 mg sodium hyaluronate solution (HA, 700 kDa, Lifecore Biomedical) was dissolved in 40 mL deionized (DI) water to form a uniform solution. After that, 206.6 mg 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDC, Sigma) and 207.0 mg aminoguanidine (Sigma) were dissolved in the HA solution, followed by a pH adjustment to 4.7. The mixed solution was reacted at room temperature overnight on a shaking bed. The resulting solution was then transferred to a dialysis bad (Spectrum^™ Spectra/Por^™ 7, MWCO 3500) and dialyzed in DI water for two days. The water was refreshed every 12 hours. After dialysis, the solution was freeze-dried for three days and stored at −20 °C for future use. To prepare LAP/G-HA control hydrogel, 300 μL 5% LAP (Best of Chemicals, NY), 50 μL 5% HA-G solution, 50 μL sodium polyacrylate (ASAP, Sigma), and 50 μL DI water were uniformly mixed to obtain stable hydrogel.

Preparation and characterization of LAP/EM-BP nanoclay gel

To prepare the LAP/EM-BP nanoclay gel, 100 μL 8% LAP was mixed with 300 μL EM-BP. The two components self-assembled to stable gel within a few minutes, and the gelation time varied with the different concentrations of EM-BP incorporated. The formation of the gel was verified using a vial-tilting method, and the vials were photographed to demonstrate the gel stability. The microstructure of both LAP and LAP/EM-BP nanoclay gels was imaged using a Cryo-EM as described above. To demonstrate injectability and moldability, the nanoclay gels with (31%) and without DBM were stained with methylene blue. DBM was provided by MTF Biologics in particulate form. To incorporate DBM into the nanoclay gels, the DBM particles were added to the LAP/EM-BP nanoclay gel and uniformly mixed using a spatula to form a composite gel. The stained gels were then injected into PBS using a 23-gauge or 14-gauge needle, and the injection process was photographed using a camera. Moldability was further demonstrated by injecting the gels into different molds, which indicates the ability of the gel to fill molds with complex shapes and form integrated structures. To evaluate self-healing capability of the gels, two pieces of gels were stained with methylene blue and rhodamine, respectively. Both stained gels were cut into two pieces, and gels with different colors were combined to form two pieces of gel with both blue and red regions. The rejoined gels were further cut and rejoined to form a blue-red-blue-red color pattern. The self-healed gels were then immersed in simulated body fluid (SBF) and fetal bovine serum (FBS) for 24 hours to confirm the structural stability and integrity of the rejoined nanoclay gels.

The compressive properties of LAP/EM-BP nanoclay gels with and without DBM were tested using a UniVert kN Mechanical Test System (CellScale) equipped with a 20 N load cell. Nanoclay gel samples with 10 mm in diameter and 5 mm in height were subjected to compression at a rate of 60 mm/min. Stress-strain curves were plotted from the compressive force, displacement, and sample dimensions. The compressive modulus was calculated using the slope of the linear region (e.g., 0–50% strain region of the compressive stress curve). Rheological behavior of the nanoclay gels was characterized by using a PP25 rheometer (Anton Paar). At a constant shear strain of 1%, the storage modulus (G′) and loss modulus (G″) were measured at a frequency range of 0.1–100 Hz. The strain sweep tests were done at a shear strain of 1% to 1000% with a fixed frequency of 1 Hz. To evaluate the self-healing capability of the gels, cyclic shear strain experiments were conducted at cyclic low (1%) and high (100% or 200%) strain levels. The viscosity of the gels was measured at a shear rate of 0.1–100 s-¹ to demonstrate shear-thinning properties.

Evaluation of biocompatibility and osteogenic effects of LAP/EM-BP nanoclay gel

To evaluate the biocompatibility of nanoclay gels with different EM-BP concentrations, MSCs were encapsulated in the gels and cultured in growth medium (DMEM with 10% FBS and 1% P/S) for 1, 4, and 7 days. Before cell encapsulation, the gels were adjusted with 10× PBS to reach a final concentration of 1× PBS. Cell viability at each time point was evaluated using a live/dead assay (Invitrogen). Briefly, the cell-loaded gels were immersed in the staining solutions in incubator, washed with PBS, and imaged using a confocal laser scanning microscope (Zeiss LSM 880). Cell proliferation was quantified using a 10% (v/v) AlamarBlue assay. In brief, the gels were immersed in AlamarBlue solution and fluorescence intensity was measured using a microplate reader.

To evaluate osteogenic differentiation, MSCs encapsulated in the gels with or without DBM were cultured in osteogenic differentiation medium, as mentioned above. For qRT-PCR experiments, total RNA was extracted using the RNeasy Mini Kit (Qiagen) following the manufacturer’s manual, and RNA concentration was measured with a NanoDrop spectrophotometer. Complementary DNA (cDNA) synthesis was conducted using a SuperScript III First-Strand Synthesis Kit (Invitrogen), followed by gene amplification on a LightCycler 480 system (Roche). To evaluate ALP activity, cells in gels were fixed in 10% formalin for 15 minutes and immersed in ALP buffer solution for 10 minutes. The cells were then incubated with an ALP staining solution which contains 94% ALP buffer, 5% NBT 20X solution, and 1% 20 mg/mL 5-bromo-4-chloro-3-indolyl phosphate (BCIP) for 30–60 minutes. To quantify ALP activity, cells were lysed and then centrifuged at 12,000 rpm for 5 minutes to remove debris. The resulting supernatant was incubated with ALP substrate solution (Sigma) for 2 hours and the absorbance was measured at 405 nm. The values were normalized to total protein content determined via BCA assay. Mineralization was evaluated using ARS staining. Cells were fixed in 10% formalin for 30 minutes, incubated in 2% ARS solution (Sigma) for 5 minutes, and then imaged using an Olympus IX71 microscope (Olympus, Japan). To quantify, ARS-stained cells were dispersed in 10% (v/v) acetic acid, and absorbance was measured at 405 nm.

The osteogenic effect of nanoclay gel loaded with DBM and noggin-targeting miRNA was further investigated by Western blotting. Cells loaded in gel were lysed to extract total protein. Lysates were separated via SDS-PAGE and transferred onto PVDF membranes. Membranes were incubated with primary antibodies specific to noggin, pSMAD1/5, or GAPDH (Santa Cruz Biotechnology, USA), followed by HRP-conjugated secondary antibodies. Protein bands were visualized using SuperSignal West Pico chemiluminescent substrate and imaged with a ChemiDoc^™ Touch system (Bio-Rad, USA). Band intensities were quantified using AIWBwell software (Servicebio, China) to assess relative protein expression levels.

Evaluation of bone regeneration capacity of nanoclay gels with DBM/noggin-targeting miRNA in vivo

The in vivo bone regenerative effect of nanoclay gels incorporating DBM and noggin-targeting miRNA was assessed using a calvarial defect model in mice. All animal procedures were conducted in accordance with protocols approved by the Chancellor’s Animal Research Committee (ARC) at UCLA and complied with the guidelines of the UCLA Office of Animal Research Oversight. In brief, male nude mice (8 weeks old, Charles River) were anesthetized using isoflurane, and bilateral 3 mm full thickness calvarial defects were created using our previously established protocol[10]. Experimental groups treated with LAP/EM-BP nanoclay gel alone, LAP/EM-BP gel with DBM, or LAP/EM-BP gel incorporating both DBM and miRNA. Additional groups included defects treated with the commercial DBX graft (31% DBM in HA) and untreated defects serving as controls. Analgesia was provided via buprenorphine for three days, and trimethoprim-sulfamethoxazole was provided in the drinking water for seven days to prevent infection.

After eight weeks of surgery, all the animals were sacrificed to collect the calvarial samples. The calvarial samples were first fixed utilizing 4% formaldehyde for two days. The fixed samples were then scanned with a micro-CT instrument (SkyScan, Bruker). A 0.25 mm filter and 20 μm resolution were selected during the scanning. Three-dimensional reconstruction was conducted using NRecon, DataViewer, and Dragonfly. The relative area of new bone formation was calculated using the reconstructed images via ImageJ. The other microstructural parameters were obtained using CTAn software (Bruker).

After micro-CT analysis, all the calvarial specimens were soaked in 10% ethylenediaminetetraacetic acid (EDTA) solution for one week to remove the minerals. The decalcified samples were embedded in paraffin and then sectioned. H&E and Masson’s trichrome staining were conducted to assess tissue morphology and collagen deposition. For OCN IHC staining, sections were deparaffinized using xylene, treated with hydrogen peroxide for 10 minutes, and immersed in blocking solution for one hour at room temperature. The samples were then treated using OCN primary antibody (Santa Cruz Biotechnology) for one hour in humidified environment (room temperature). The samples were then incubated using biotinylated goat anti-mouse secondary antibody for 10 minutes and streptavidin-HRP for 10 minutes. After that, the samples were treated in 3,3′-diaminobenzidine (DAB) for color development. The samples were then treated with counterstaining, washed using water, mounted, and imaged using a microscope.

Statistical analysis

All results are presented as mean ± standard deviation (SD). Statistical analysis among groups were performed using one-way analysis of variance (ANOVA) with Tukey’s post hoc test. The data differences were considered significant for *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Data Availability Statement

The data that support the findings of this study are available from https://doi.org/10.5061/dryad.8gtht772w.