Odontogenesis by Endocytosis of Peptide Embedding Bioactive Glass Composite

Introduction

Dental caries remains one of the most prevalent chronic diseases worldwide. As the caries progresses, the destruction has effects, from the demineralization of the hard tissues to the disintegration of the deep dental tissues. Therefore, the restoration of deep caries involves not only the remineralization of hard tissues but also the restoration of the pulp–dentin complex. However, most research on caries restoration has focused either on the hard tissue remineralization or the biological repair of the pulp–dentin complex as 2 separate fields, neglecting the holistic restoration of both. Classic remineralization materials such as fluoride emphasize guiding the accumulation of calcium and phosphorus (Cochrane et al. 2010; Slayton et al. 2018). Existing bioengineered materials like Wnt and BMP proteins have mainly aimed at regenerating the pulp–dentin complex (Casagrande et al. 2010; He et al. 2019). Owing to the negative consequences of the current restorative techniques and diverse deficiencies of the restorative materials, ideal solutions should perform a “dual effect” to remineralize the hard tissues and regenerate lost dentin tissues in the meantime.

Based on biological events involved in tooth development, some bioactive extracellular matrix molecules are thought to be directly implicated in dental mineralization and indirectly affect the recruitment and differentiation of pulp progenitors (Goldberg et al. 2009). Amelogenin is the main component of enamel matrix proteins. It is reported that amelogenin-derived proteins can induce reparative dentin formation (Veis 2003; Oh et al. 2015). These proteins are also involved in initiating enamel mineralization, supporting crystal growth at desired ultrastructural locations, protecting the growing mineral phase, and arranging the crystals into discrete prism bundles (Ieong et al. 2011; Moradian-Oldak 2012). Therefore, amelogenin and its derived products cannot be considered exclusively structural enamel proteins. Massive data have now converged to the point that they play multiple roles, including participating in dentinogenesis (Hammarström 1997; Grandin et al. 2012), as well as initiating hard tissue remineralization (Lv et al. 2015; Han et al. 2017).

Focusing on the highly conserved Gln-Pro-X (QPX) sequences in amelogenin, we designed an amelogenin-derived peptide QP5 in previous studies (Lv et al. 2015; Han et al. 2017), which contained 5 Gln-Pro-X repeats and a 7-residue hydrophilic segment (C-tail). QP5 has been proved to exert a significant remineralization effect both in vitro and in vivo, mainly depending on its different functional domains. (QPX)5 was the major adsorption region with hydroxyapatite (HAP), both (QPX)5 and C-tail inhibited nucleation, and C-tail contributed more to improve the HAP orientation degree (Wang, Hu, et al. 2020). We further verified that the peptide could also induce the formation of tertiary dentin (Peng et al. 2021). Therefore, this biomimetic peptide has the potential to exert a dual effect for deep caries repair. However, the application of the single peptide should be optimized by combining with solutions containing calcium (Ca)/phosphorus (P) ions for enamel remineralization or with gelatin to shape the material and fulfill the marginal sealing for pulp capping. Obviously, the clinical application of this peptide and of all peptide or protein-like materials remains a problem that has yet to be solved. For instance, they lack of their own Ca/P sources in remineralization. In intricate body fluids and cellular environment, they could be easily diluted and degraded and thus could not maintain long-term effective concentration.

The loading of functional proteins into a drug delivery system that can supplement ion sources and long-term release of the peptide may be a promising approach to address these problems. Among the various biocompatible carriers, bioactive glass (BG) has several advantages owing to its capacity for Ca/P storage and interface bonding ability. BGs are mainly composed of SiO₂, CaO, and P₂O₅, which have a satisfactory affinity for collagen of demineralized dentin (Miguez-Pacheco et al. 2015). As a unique class of bioactive materials, BGs have demonstrated remarkable properties for both hard and soft tissue repair and regeneration (Lin et al. 2012; Sevari et al. 2020). A next-generation bioactive glass called mesoporous bioactive glass (MBG) has been developed; it has a porous microstructure, significant bioactive characteristics, and controlled drug release characteristics (Xia and Chang 2006; Feng and Chang 2011). BGs have been widely studied in the field of bone regeneration (Nommeots-Nomm et al. 2017; Ravanbakhsh et al. 2019). Owing to the similarity of bone and tooth formation, BG was reported to influence dentin regeneration (Wang, Huang, and Dong 2020). Combining it as a carrier with biomimetic functional peptides may supplement the deficiencies of the single component, and it is expected to become a new bioactive material for caries repair by remineralizing the degraded hard tissue and inducing the regeneration of reparative dentin.

Therefore, we developed a novel composite of MBG and QP5 to complement the limitations of the lack of Ca/P sources in remineralization and to investigate whether it can enhance the restoration of dentin. In this article, we 1) examined the intelligent drug delivery and pH-responsive releasing properties of this composite in the acidic environment of caries, 2) further focused on the odontogenesis ability of this composite on human dental pulp cells (hDPCs) and explored the cellular uptake mechanism of the composite, 3) evaluated its effect on the differentiation of hDPCs, and 4) determined its ability to regenerate dentin-like tissues in vivo to ultimately realize the “dual-effect” potential of this new composite for deep caries restoration.

Materials and Methods

Details are provided in the Appendix.

Results

Characterization of Highly Ordered Mesoporous Bioactive Glass

Figure 1A shows the surface morphology and microstructure of MBG particles examined by a scanning electron microscope (SEM; top panel) and a transmission electron microscope (TEM; bottom panel). MBG exhibited a smooth surface with a heterogenous size (4–20 nm) and a hexagonal homogeneous mesoporous microstructure. The differentiation peak at 15° to 35° revealed the amorphous states of the glass (Fig. 1B). MBG had 3 diffraction peaks (100, 110, and 200) in the small-angle region, confirming the hexagonal mesoporous structure. Nitrogen sorption analysis exhibited a type IV isotherm, and the pore size was mainly 4 to 8 nm (Fig. 1C). The surface area, pore volume, and pore size were 673.2 m²/g, 0.65 cm³/g, and 5.2 nm, respectively (Appendix Table 1). These observations suggest that the mesoporous structures of MBG were successfully fabricated.

OPEN IN VIEWER

Figure 1.

Characterization of mesoporous bioactive glass (MBG). (A) Surface morphology and microstructure by scanning electron microscope (top panel) and transmission electron microscope (bottom panel) exhibited a highly ordered mesoporous microstructure of MBG particles. (B) Wide-angle and small-angle X-ray diffraction patterns of MBG. Red arrow indicates the diffraction peaks of 100, 110, and 200, respectively. (C) Nitrogen sorption isotherms and pore width distribution of MBG by N₂ adsorption analysis. (D) Circular dichroism spectra of 1.5 mg/mL QP5 peptide in a different pH environment. There were no obvious changes in the secondary structure of QP5. SAXRD, small-angle X-ray diffraction pattern; SEM, scanning electron microscope; TEM, transmission electron microscope; WAXRD, wide-angle X-ray diffraction pattern.

Increased Peptide Loading of the Composite by Diverse Electrical Properties

Electrospray ionization mass spectrometry of QP5 peptide confirmed the successful synthesis of QP5 (Appendix Fig. 1). Due to the pH complexity of the oral and drug-carrying environment, we first tested the stability of the peptides. The results of circular dichroism presented similar spectra with no significant changes in the characteristic peak, indicating that QP5 is structurally stable under different pH used in this study (Fig. 1D). To determine the best formula for MBG and QP5, 50 mg MBG was soaked in various concentrations (0.2–7 mg/mL) of QP5 solutions. As shown in Figure 2A, entrapment efficiency (EE) decreased, while the loading capacity (LC) of the peptide into MBG increased (Appendix Table 2). These variables intersected at the concentration between 1.4 and 2.0 mg/mL. In consideration of effective peptide loading and economic cost, 1.5 mg/mL of peptide concentration was used for the subsequent experiments. The Fourier transform infrared spectroscopy (FTIR) spectrum of MBG/QP5 displayed the characteristic peak of the C-NH₂ around 1,550 cm⁻¹ (Fig. 2B, black frame), indicating successful peptide loading into the composite. X-ray photoelectron spectroscopy (XPS) spectrum analysis displayed that the composite had Si, Ca, P, and N elements (Fig. 2C).

OPEN IN VIEWER

Figure 2.

Loading, releasing profiles, and characterization of MBG/QP5 composite. (A) Entrapment efficiency and loading capacity of the composite. (B) The Fourier transform infrared spectra of MBG, QP5, and MBG/QP5 composite. A zoom at 1,550 cm⁻¹ is displayed on the right panel. Black frames delineate peak of amino group. (C) X-ray photoelectron spectrum of MBG/QP5 composite. A zoom of the N peak is displayed (insert). (D) Zeta potential of MBG, MBG/QP5 composite, and MBG/QP5 composite after releasing QP5. MBG (7.4), MBG at pH 7.4; MBG (5.0), MBG at pH 5.0; MBG/QP5 (5.0), MBG/QP5 composite at pH 5.0; MBG/QP5 released (5.0), MBG/QP5 composite after QP5 releasing at pH 5.0. (E) Zeta potential of QP5 under various pH. Values are presented as mean (SD). Letters (a, b, c, and d) indicate significant differences among the groups at P < 0.05. (F) pH change of artificial saliva after 1 mg/mL of MBG soaking. *P < 0.05. (G) Release profiles of QP5, Ca, and P ions from MBG/QP5 composite at pH 7.4, 5.5, and 4.0. Ca, calcium; EE, entrapment efficiency; LC, loading capacity; MBG, mesoporous bioactive glass; P, phosphorus.

Electrostatic absorption is another factor that influences the loading process. MBG was negatively charged at pH 5 and 7.4, whereas QP5 exhibited a positive charge when pH <6 (Fig. 2D, E). Therefore, the loading process was conducted under 3 pH conditions: 7.4, 5.0, and 4.0. The results demonstrated that the EE and LC synchronously increased by 9.51% and 19.01 μg/mg, respectively, when the pH decreased from 7.4 to 4.0 (Appendix Table 3). Zeta potential displayed that MBG/QP5 composite had a reduced negative charge compared to MBG alone under pH 5.0. The negative charges of the composite increased after QP5 release, and the negative charges exceeded that of MBG alone (Fig. 2D). The optimized MBG/QP5 composite was obtained as white lyophilized powder (Appendix Fig. 2) and stored at −20°C for later use.

Active Endocytosis by Human Dental Pulp Cells and Intracellular Long-Term Retention of Peptides

As shown in Figure 3A, in the observation of up to 9 d, concentrations at or below 100 μg/mL of the composite exhibited favorable cytocompatibility. To track whether the MBG/QP5 composite had been internalized by the cells, fluorescein isothiocyanate (FITC)–linked QP5 peptide was used for composite preparation. As shown in Figure 3B, higher FITC fluorescence intensity was detected in hDPCs treated with the MBG/QP5 composite compared to the blank control. The amount of fluorescence of the composite entering the cell appeared in a time- and concentration-dependent manner. When the composite and peptide were removed from the medium after 4 h of coculture, the intracellular fluorescence decreased over time. However, the fluorescence in the MBG/QP5 composite group exhibited a slighter decline compared to the peptide group (Fig. 3C). The cellular uptake behavior was also investigated with confocal laser scanning microscopy (CLSM) images. Figure 3D shows that MBG/QP5-FITC and QP5-FITC were present in the cytoplasm of the hDPCs after 24 h of incubation, confirming that the composite and the peptide were internalized by the cells.

OPEN IN VIEWER

Figure 3.

Mechanism of cellular uptake of MBG/QP5 composite. (A) Survival rate after 9 d of treatment under concentrations of 0 to 200 μg/mL MBG/QP5 composite by CCK-8 assay. (B) Flow cytometry results demonstrated the amount of intracellular fluorescence elevated with the increase of coculture time and composite concentration. (C) Changes of intracellular fluorescence after removal of MBG/QP5 composites or QP5 peptides. (D) Confocal microscopy images showing the MBG/QP5-FITC and QP5-FITC were internalized by the cells and present in the cytoplasm of the human dental pulp cells (hDPCs). Nuclei were stained with DAPI. F-actin filaments were stained with rhodamine phalloidin. (E) Effect of endocytosis inhibitors on MBG/QP5-FITC and QP5-FITC internalization by hDPCs. Figures from left to right: cellular uptake after different inhibitor treatments in hDPCs by flow cytometry analysis and the relevant quantitative results of QP5-FITC and MBG/QP5-FITC respectively. Data are presented as mean ± standard deviation. Letters (a, b, c, and d) indicate significant differences among the groups at P < 0.05. AL, amiloride; CPZ, chlorpromazine; FITC, fluorescein isothiocyanate; MBG, mesoporous bioactive glass; MβCD, methyl-β-cyclodextrin; NC, negative control; PC, positive control; 4°, culture temperature at 4°C.

To understand the mechanism by which the MBG/QP5 composite was internalized, several inhibitors of major endocytotic pathways were applied to track the cell uptake behavior: chlorpromazine (CPZ) as an inhibitor of clathrin-mediated endocytosis, methyl-β-cyclodextrin (MβCD) as a lipid raft inhibitor, and amiloride (AL) as an inhibitor of macro-pinocytosis (Adjei et al. 2014; Rennick et al. 2021). hDPCs were also incubated at 4°C to prevent adenosine triphosphate (ATP)–dependent endocytosis. After 4 h of treatment, FITC-positive cells were counted by flow cytometry, as shown in Figure 3E. The uptake of the composite and peptide was significantly decreased in ATP depletion culture condition (4°C), confirming ATP-dependent endocytosis. In the QP5 group, the internalization was inhibited by CPZ and MβCD but not by AL. In the MBG/QP5 composite group, the cell uptake was also inhibited by AL, indicating that the composite could enter the cell through macro-pinocytosis in addition to the pathway of the released peptides.

Enhancement of hDPC Migration and Odontogenic Differentiation Ability under MBG/QP5 Induction

The scratch wound assay showed a significant enhancement in the migration of hDPCs after the MBG/QP5 treatment. The wound closure reached 100%, while that of the control group reached 78.06% after 24 h (Fig. 4A, B). After 5 d of odontogenic inducing, more obvious alkaline phosphatase (ALP) staining (indicating the activity of ALP) was observed in the composite group compared with MBG or QP5 alone of an equivalent concentration (50 μg/mL) (Fig. 4C). Alizarin red staining (ARS) presented the most mineralized nodules that were generated in the composite group after 2 wk (Fig. 4C). We further measured the promineralization ability of MBG/QP5 by detecting odontogenic-related genes at the messenger RNA (mRNA) level. The results showed that after 7 d of induction, all the mineralization-related genes, including COL1, DSPP, ALP, and RUNX2, were upregulated, especially in the MBG/QP5 composite group. Furthermore, after 14 d of induction, DSPP, OPN, and OCN were significantly upregulated when treated with the composite (Fig. 4D). A schematic illustration of the design strategy and effect pattern of the composite is shown in Figure 4E.

OPEN IN VIEWER

Figure 4.

Odontogenic differentiation and mineralization of human dental pulp cells (hDPCs) after mesoporous bioactive glass (MBG)/QP5 composite treatment. (A) Images of scratch wound assays by light microscopy. (B) Quantitative analysis of migration rates of hDPCs at 12 and 24 h cultured with composite concentrations of 0, 50, and 100 μg/mL by scratch wound closure. (C) Alkaline phosphatase (ALP) staining and alizarin red staining (ARS) of hDPCs after 5 d and 2 wk of treatment. (D) Expressions of odontogenic-related genes after 7 and 14 d of treatment. Data are presented as mean ± standard deviation. *P < 0.05. Letters (a, b, c, and d) indicate significant differences among the groups at P < 0.05. (E) Schematic illustration of MBG/QP5 composite and the odontogenesis effect on hDPCs: QP5 peptides were loaded into the hexagonal MBG.

In Vivo Formation of Hard Tissues through Direct Pulp Capping

Hard tissue formation was observed and evaluated underneath the capping material to test the repair ability of the materials (Fig. 5A). Micro–computed tomography imaging revealed that tertiary dentin formation observed beneath the cavity in the mineral trioxide aggregate (MTA)-, QP5-, MBG-, and MBG/QP5 composite–treated samples, as well as the effect in MTA and MBG/QP5, was most obvious. Little hard tissue formation was observed in the phosphate-buffered saline (PBS) group (Fig. 5B). The entire calcification of the pulp cavity was more obvious in the MTA group, while diffused mineralized nodules in the pulp cavity could be observed in the QP5 and MBG-alone group. In particular, the composite group presented a regenerated dense dentin-like layer below the perforation site (Fig. 5B).

OPEN IN VIEWER

Figure 5.

Evaluation of the effects of mesoporous bioactive glass (MBG)/QP5 composite on dentin formation in vivo. (A) Establishment of the direct pulp capping model in rats. The maxillary first molars were cleaned (a) and mechanically exposed (b), and the cavity was capped with different materials (phosphate-buffered saline [PBS], mineral trioxide aggregate (MTA), QP5, MBG, and MBG/QP5 composite) (c) and sealed with glass ionomer cement (GI) for 4 wk (d). (B) X-Y axis-plane images of micro–computed tomography reconstruction displaying restorative dentin formation beneath the cavity in the PBS-, MTA-, QP5-, MBG-, and MBG/QP5 composite–treated samples.

Discussion

Based on the multiple functional potentials of the QP5 peptide, we designed and synthesized a novel composite. It is intended that the composite could provide its own source of Ca/P ions and could further restore defected dentin tissues to fulfill the “dual effect.”

The surface area was significantly increased by the highly ordered mesoporous structure, thus making the material applicable for drug loading (Xia and Chang 2006). To promote the loading capacity, grafting positive-charged groups such as -NH₂ to MBG has been a widely used solution to switch the negative charge of MBG (Lee et al. 2016; Wang et al. 2017). However, the grafting process may collapse the pores of MBG (Jiang et al. 2017). Since peptides exist as zwitterion in aqueous solvent, taking advantage of the fact that the charge of peptides is affected by the pH of the solution, we converted the negative charge of the QP5 to a positive one so that it could generate additional electrostatic adsorption with the negatively charged MBG. Therefore, by using the amphoteric property of the peptides, we elevated the peptide loading efficiency by simply adjusting the pH of the loading solution under the isoelectric point of the peptides (Fig. 2D, E).

Notably, the composite exhibited a pH-responsive release profile. In most cases, the increased dissolution of MBG in acidic environment resulted in incremental drug release (Yang et al. 2014). In our study, the ion release increased with the decrease of the pH (Fig. 2F). Accordingly, more peptides were released at pH 4.0 than pH 5.5. However, most peptides were released at pH 7.4, which was contrary to the trend of ion release. This was also influenced by the amphoteric property of the peptide. When the pH >6.0, the positive charge converted to negative, resulting in massive desorption between MBG and QP5. This unique release profile of the unsynchronized release of ions and QP5 could rhythmically function in clinical application. MBG would dissolve due to the decrease of pH caused by dental caries, resulting in the massive release of therapeutic ions and the upregulation of the environmental pH (Fig. 2G). Subsequently, QP5 would be released more in a neutral environment to initialize functioning together with the therapeutic ions.

After confirming the appropriate concentration of the composite for hDPCs, we aimed to track the endocytosis of the composite and its mechanism. Bioactive molecules or particles are internalized into cells through various pathways, including active endocytosis by membrane structural changes, which is ATP dependent, or nonspecific diffusion (Lee et al. 2017). By FITC fluorescent labeling, we found that the cell internalization of the MBG/QP5 composite and the QP5 peptide was ATP dependent, indicating the active uptake by the cells. Considering the ATP-dependent endocytosis was through either macro-pinocytosis or receptor-mediated endocytosis (Adjei et al. 2014) for screening out which pathway would be involved, specific inhibitors were applied to determine the mechanism. Noteworthily, the cell uptake of the composite is slightly different from that of QP5 alone. The pathway of the peptide was mediated by clathrin and lipid rafts of the cell membrane (inhibited by CPZ and MβCD, respectively). In contrast, the composite is larger in size and consists of MBG particles and QP5 peptide. Therefore, in addition to the pathway of the QP5 peptide described above, it could enter the cell via macro-pinocytosis (inhibited by AL). This could be explained by the fact that the endocytosis pattern is influenced by the particle size of the material, and particles with a size (diameter) above 200 nm usually activate the macro-pinocytosis pathway (Rennick et al. 2021).

Regarding the biological effects of the composite on hDPCs, the composite promoted the cell migration in a concentration- and time-dependent manner (Fig. 4B), which was consistent with the previous findings of the intracellular uptake in a similar manner (Fig. 3B), indicating that the endocytosis of the composite was firmly correlated with the biological function of the composite. Compared to the same concentration of a single component (MBG or QP5), the composite demonstrated superior performance in odontogenic differentiation of hDPCs. The facilitation may be explained by the extra bioactive Ca/P ions from the composite, which were proven to promote hard tissue repair (Gong et al. 2014; Zhang et al. 2019). Moreover, since the composite enhanced the retention of intracellular peptides (Fig. 3C), it may contribute to maintaining the effective intracellular concentration of functional peptides. This also confirmed that the composite form may possess a protective effect on peptide-based materials in clinical applications. In dealing with sophisticated fluidic and intracellular environments, the application of the composite form may be more favorable than the peptide-based materials alone.

A rat pulp capping model was chosen to further verify the in vivo effect of the composite. When applying QP5 alone, we observed a mass of diffused calcified nodules throughout the pulp chamber, which may be due to the diffusion of the peptides and thus unable to form effective sealing below the perforation site. MBG alone induced mineralized tissues that developed under the exposure site, the amount of which was relatively small. We observed a regenerated dentin-like layer below the pulp perforation in the composite group, suggesting that the composite combined the advantages of the edge sealing of MBG and the permeation of QP5. Although in the MTA group, a thicker restorative layer was formed, it was accompanied by calcification and shrinkage of the entire pulp chamber. Considering MTA has been applied clinically as the “gold-standard” therapy recently, limitations including discoloration (Esmaeili et al. 2016) and poor adhesion to the existing bonding system (Shin et al. 2014) remain challenges for this material. To further optimize the application of this composite, specific clinical factors, including color and bonding properties of the material, should be considered in future studies.

In conclusion, we successfully synthesized a novel composite in a simple and efficient way by using the porous structure of MBG and unique amphoteric properties of QP5. The intelligent pH-responsive drug release mode of this composite could neutralize the acidic environment and maintain effective therapeutic ions and peptides under the carious conditions. Furthermore, this composite could be actively transported into cells in multiple ways and function in promoting the migration, differentiation, and mineralization of hDPCs in vitro and in vivo. Thus, the MBG/QP5 composite may be considered a promising material for treating diseases related to hard tissue defect and for dentin tissue engineering.

Author Contributions

J. Lu, contributed to conception, design, data acquisition, analysis, and interpretation, drafted and critically revised the manuscript; Z. Liu, contributed to design, data acquisition, analysis, and interpretation, critically revised the manuscript; K. Wang, contributed to conception, design, and data interpretation, critically revised the manuscript; X. Peng, contributed to design, data acquisition and interpretation, critically revised the manuscript; M. Gu, Y. Zhang, X. Chen, Y. Chen, contributed to design and acquisition, critically revised the manuscript; L. Zhang, contributed to conception, design, data analysis and interpretation, critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of the work.

Supplemental Material