The effects and mechanisms underlying collagen-binding peptides MDPs on dentin remineralization: an in vitro study

Abstract

Background

Currently, many biomimetic mineralization materials fail to bind effectively to collagen, which significantly limits their remineralization efficacy. To address this limitation, we designed MMP2-derived peptides (MDPs) on the basis of the strong affinity domain of collagen: MDP-a and MDP-3DSS.

Methods

Surface plasmon resonance (SPR) was used to measure the binding force of MDPs to collagen. Inductively coupled plasma-mass spectrometer (ICP-MS) was used to evaluate the effects of peptides on the adsorption of calcium ions by collagen. Transmission electron microscopy (TEM) and thermogravimetric analysis (TGA) were employed to analyze the stabilizing effect of MDPs on amorphous calcium phosphate (ACP) and the mineralization effect on collagen. Scanning electron microscopy (SEM), mechanical tests, and the CCK-8 assay were employed to evaluate the remineralization effect of MDPs on demineralized dentin and their biocompatibility.

Results

The results of the present study demonstrated that MDPs strongly bind to collagen (18 µM for MDP-a and 18.9 µM for MDP-3DSS) and can promote calcium ion adsorption by collagen (p < 0.0001). Furthermore, MDPs have been shown to stabilize ACP and induce intrafibrillar mineralization. In vitro studies revealed that both surface and cross-sectional dentin slices treated with MDPs presented dense mineral deposition. This deposition effectively occluded the dentinal tubules and resulted in enhanced mechanical properties. The tubule plugging rate, the elastic modulus and microhardness of the MDPs group were greater than those of the control group (p < 0.01). Moreover, MDP-3DSS demonstrated the most significant remineralization potential. Additionally, MDPs treatment did not affect the activity of human dental pulp stem cells (hDPSCs) (p > 0.05).

Conclusions

The present study successfully developed and evaluated two novel biomimetic mineralized peptides, MDP-a and MDP-3DSS. It is necessary to conduct further research to explore the possibility of clinical transformation of MDPs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-06680-1.

Background

As mineralized tissues formed by organisms under strict biological control, vertebrate teeth are hierarchically ordered at multiple scales and possess excellent mechanical properties [1]. Under physiological conditions, tooth demineralization and remineralization constitute a dynamic balance process. However, demineralization exceeds remineralization under certain pathological conditions (such as caries and acid erosion), which may lead to hard tissue lesions in teeth. As the most prevalent demineralization disease in the oral cavity, dental caries significantly impact chewing function and quality of life in a substantial proportion of the global population [2]. This gives rise to the explicit and extensive clinical demands for the remineralization repair treatment for demineralized dentin.

On the basis of the classical nucleation theory of ion-mediated crystallization, early remineralization treatment focused on supplementing calcium and phosphorus ions on the demineralization surface of dentin [3]. Nevertheless, it results in the formation of large disordered deposited crystals on the collagen surface, which are far from the ordered crystals in normal dentin. These remineralization treatments overlook the crucial role of non-collagen proteins (NCPs) in the biomineralization process [4]. The mineralization mechanism of non-classical nucleation theory is closer to the natural mineralization process and does not depend on the limitations of crystal nuclei. On the basis of the non-classical nucleation theory, we can use NCPs and their analogs to induce the formation of ACP and complete fiber mineralization through the gradual transformation and orderly arrangement of ACP. Hence, scholars have begun to explore the mineralization regulation of NCPs and their analogs during the process of crystal nucleation and growth, achieving biomimetic remineralization of demineralized dentin through different mineralization strategies [5]. This method has significant advantages in terms of biocompatibility, mineralization speed and mineralization depth.

Currently, NCPs have been demonstrated to be effective at achieving dentin mineralization because of their ability to stabilize prenucleation clusters and form ACP nano-precursors, such as dentin matrix protein 1 (DMP-1) and dentin phosphoprotein (DPP) [6]. Although NCPs play a crucial role in mineralization regulation, it is costly to repair dentin via natural NCPs because of their limited accessibility. In recent years, synthesized polypeptides have emerged as breakthrough points in the research of biomimetic remineralization materials because of their ability to mimic the functional region of NCPs and their excellent biocompatibility, simple synthesis, and controllable cost [7]. Thus, peptides derived from NCPs represent economically beneficial treatment strategies. Many scholars have developed short peptides, such as amelogenin-derived peptides, which contain functional sequences that facilitate remineralization, by screening the significant NCPs involved in dentin mineralization. These short peptides have been confirmed to exert this effect [8]. These polypeptides usually perform the following functions. Many acidic amino acids attract calcium ions, stabilize calcium and phosphorus ions to form ACP nano-precursors, bind to type I collagen, and guide the nucleation and growth of hydroxyapatite. Therefore, the modified design of biomimetic mineralized peptides aims mainly to screen and combine functional regions in NCPs, such as the amino acid sequences involved in mineral interactions [9]. In addition, amphiphilic peptides have been developed for dentin remineralization, which has also achieved an intrafibrillar mineralization effect [10]. Nevertheless, a significant gap remains between the degree achieved through biomimetic methods and that of natural tooth hard tissue [11]. Therefore, further exploration of the development of novel biomimetic remineralization materials, which may contribute novel ideas for the clinical treatment of biomimetic mineralization regulation in minimally invasive dentistry, is worthwhile [12].

In addition to the predominant regulatory proteins within NCPs, scholars have discovered that certain intracellular proteins, such as GRP78 and TRIP1, in the bone and dentin matrix can likewise promote biomineralization, suggesting their ability to repair enamel and dentin through various proteins inspired by non-classical nucleation theory [13]. Through a literature review, Lei et al. discovered that CBP1495, a matrix metallopeptidase 2 (MMP2) precursor, can bind to type I collagen with strong affinity via isothermal titration calorimetry and enzyme-linked immunosorbent assays, confirming that CBP1495 is a novel and primordial collagen strong affinity binding peptide [14]. Currently, the design of mineralized short peptides aims mainly to increase their binding with calcium ions by increasing the number of acidic proteins [15]. Although collagen fibers are the main components of demineralized dentin, there is still little research on the design of derived peptides inspired by the collagen-binding functional domain. Many biomimetic mineralization materials fall short in effectively binding to collagen, which significantly limits their remineralization efficacy. To bridge this gap, this study focused on developing novel peptides based on the strong affinity of collagen. Therefore, MDPs were developed from the strong affinity domain of collagen in CBP1495, including MDP-a (enhanced amphiphilicity) and MDP-3DSS (combined with DSSDSSDSS amino acid sequences from DPP). Through in vitro experiments, we assessed their collagen affinity, ability to promote collagen adsorption of calcium ions, ability to stabilize ACP, and ability to promote collagen and dentin remineralization, and biocompatibility with hDPSCs. This work aims to provide a scientific basis for the application of MDPs in dentin repair and guide future in vivo experiments and clinical trials.

Materials and methods

This study was conducted according to the guidelines for human research subjects established by the Ethics Committee of Shen-Zhen Stomatology Hospital: LLWYH-PJ-20240829-001. Informed written consent was obtained from patients according to the guidelines.

Peptides synthesis

All peptides were synthesized, purified and identified by Nanjing Top-Peptide Biotechnology Co. Ltd. of China. MDPs were synthesized via the standard Fmoc solid-phase peptide synthesis method. After synthesis, the product was purified via high-performance liquid chromatography (HPLC). A Kromasil C18-5 column (4.6 mm × 150 mm) was employed for the purification process. The mobile phase consisted of two solvents: Solvent A was 100% acetonitrile containing 0.1% trifluoroacetic acid (TFA), and Solvent B was 100% water containing 0.1% TFA. The purification was carried out via an HPLC system (Waters, USA). The peptide concentration was determined at a detection wavelength of 214 nm. The purified peptides were converted into gaseous ions via mass spectrometer (MS) for data acquisition to determine the amino acid sequence of the peptides and their molecular weights. The freeze-dried of the peptide powder was packed and stored at −20 °C.

Affinity analysis of MDPs and type I collagen binding

Through the detection of the Biacore 8 K high-throughput intermolecular interaction meter (GE, US), rat type I collagen (Corning, US) was pre-enriched with pH = 5.0 sodium acetate buffer (Aladdin, China), and the highest response value was selected. Under these pH conditions, the type I collagen solution is formally coupled with the ligand and the CM5 chip via an amino covalent coupling agent. After the freeze-dried powder of MDPs peptide was dissolved, the peptide mixture was prepared and filtered. It was matched with sodium acetate buffer into five concentration gradients. The contact time was 100 s, the dissociation time was 80 s, and the flow rate was 30 µL/min. The regeneration conditions were glycine 2.0, the contact binding time was 30 s, and the flow rate was 30 µL/min. We then repeated the run sampling and regeneration programs in 3 cycles, and conducted multicycle kinetics through multicycle dynamics. Finally, the dissociation constant (K_D) value was calculated.

The stabilization effects of MDPs on calcium and phosphorus solution

The remineralization solution was composed of 3.90 mM CaCl₂ and 3.83 mM Na₂HPO₄. After the pH of the 4-(2-hydroxyethyl)−1-piperazineethanesulfonic acid (HEPE) buffer was adjusted to 7.0 with NaOH, HEPE buffer was used to adjust the pH of the remineralization solution to 7.0. MDPs freeze-dried powder were added to the remineralization solution such that the peptide concentration was 0.1 mM. A total of 10 µL of the mixed liquid was dropped into a copper net. After air-drying, the morphology, properties of the particles, and distribution of the elements were observed via TEM, selected area electron diffraction (SAED) and energy dispersive X-ray spectroscopy (EDS). The diameter of each group of 30 particles was measured via ImageJ software version 1.53a. Calcium and phosphorus solution of the same volume were mixed to constitute the control group.

Reconstitution of the collagen fibers

A total of 8.33 µL of type I collagen (3 mg/mL, Corning, US) was added to 0.5 mL of self-assembly solution (50 mM glycine and 200 mM KCl, pH = 9.2, Corning, US), and incubated at room temperature for 20 min. Then, 3 µL of self-assembled collagen solution was added to the nickel net, which was left at 37 °C for 12 h. The nickel mesh was soaked in 0.05% glutaraldehyde solution, fixed for 1 h, and rinsed in water for 1 min.

Mineralization of the collagen fibers

The nickel mesh assembled with collagen was immersed in a mineralized solution (3.90 mM CaCl₂, 3.83 mM Na₂HPO₄, pH = 7.0) containing 0.1 mM MDP-a or MDP-3DSS and incubated at 37 °C. The mineralization solution was replaced every 24 h. After 7 d, the nickel mesh was rinsed with deionized water for 1 min to remove unbound peptides, and then soaked in 50% ethanol and 100% ethanol for 10 min for subsequent characterization. The control group was a mineralized liquid without peptides. The morphology and the crystallinity of mineralized collagen were examined via TEM and SAED. EDS was used to analyze the distribution of calcium and phosphorus.

Calcium ion adsorption and mineralization of the collagen membrane

The collagen membrane (Hai Ao, China) was immersed in a mineralized solution (3.90 mM CaCl₂, 3.83 mM Na₂HPO₄, pH = 7.0) containing 0.1 mM MDP-a or MDP-3DSS and incubated at 37 °C. The mineralization solution was replaced every 24 h. After 7 d, the collagen membrane was rinsed with deionized water for 1 min to remove unbound peptides, and then soaked in 50% ethanol and 100% ethanol for 10 min and then freeze-dried overnight. The dried collagen membrane (n = 9) was placed into a sample digestion vial, concentrated nitric acid was added for dissolution, the mixture was stirred thoroughly, and the mixture was digested at 600 °C until the volume reached 1 mL. Once complete dissolution was achieved, the mixture was diluted with deionized water to a final volume of 50 mL. The digested mixture was subsequently diluted and introduced into an ICP-MS 7900 instrument (Agilent, US) for analysis. The dried samples (n = 9) were placed in a thermogravimetric analyzer (Waters TGA550, USA), where they were heated gradually from room temperature to 600 °C at a rate of 10 °C/min under an airflow of 100 mm³/min, with the weight% of the samples in the crucible being recorded in real time. The control group consisted of a mineralization solution without peptides. The measured data were tested via one-way ANOVA, followed by Tukey’s post hoc test.

Preparation of dentin slices

The G*Power software tool was used to calculate the sample size. The sample size was calculated via the following criteria: 95% confidence level and 80% power. Healthy third molars were collected following the guidelines approved by the Ethics Committee of Shen-Zhen Stomatology Hospital: LLWYH-PJ-20240829-001. Informed written consent was obtained from patients according to the guidelines. A hard tissue grinding and cutting system was used to grind 1/3 of the tooth sample in the crown and make it into a 5 × 5 × 2 mm slice (n = 72), followed by grinding and polishing. All samples had the surface of the dentin eroded by 35% phosphate gel acid (Gluma, Germany) for 30 s, were rinsed with 5.25% sodium hypochlorite for 1 min, and were placed in ultrasonic water baths for 10 min to remove the minerals escaping from the surface.

Observation of the remineralization effects of dentin slices

The demineralized dentin slices were divided into a control group (treated without MDPs), a 0.1 mM MDP-a treatment group, and a 0.1 mM MDP-3DSS treatment group. After the dentin slices were soaked in MDPs solution for 12 h, they were rinsed with deionized water for 1 min to remove unbound peptides and transferred to an Eppendorf tube containing artificial saliva (Solarbio, China). The artificial saliva product was composed of NaCl, KCl, CaCl₂, NaH₂PO₄, urea, and Na₂S, and was maintained at a pH of 7.0. To ensure a stable mineralization environment, the artificial saliva was refreshed daily. After 2 w of remineralization, the dentin slices were fixed with pentyl diol, dehydrated with an ethanol gradien, air dried and sprayed with gold. Finally, the remineralization effect on the surface and cross-section of the dentin was observed via SEM. The number of open tubules in surface images taken at a magnification of 3,000× was measured via ImageJ software version 1.53a to calculate the plugging rate before and after remineralization for each group (n = 9). The measured data were tested via one-way ANOVA, followed by Tukey’s post hoc test.

Mechanical properties

The mechanical properties of the new mineral layer on the surface of the dentin slices after 2 w of mineralization were measured via a Bruker HYSITRON TI 980 instrument, along with sound and etched slices. The loading duration was 10 s, with a holding time of 2 s, a Poisson’s ratio of 0.28, and an applied force of 10 mN. Each sample was tested at five points, with nine samples in each group. The resulting data were analyzed to generate load-displacement curves, and the elastic modulus and microhardness of each group were calculated. The measured data were tested via one-way ANOVA, followed by Tukey’s post hoc test.

Biocompatibility

HDPSCs were purchased from iCell Bioscience Inc, Shanghai, and were cultured in the corresponding medium. The biocompatibility of 0.1 mM MDP-a and MDP-3DSS were measured via a CCK-8 kit (Beyotime, China) according to the manufacturer’s protocol. In brief, 1000 cells/well were seeded in a 96-well plate, cultured in a 5% CO₂ incubator at 37 °C for 12 h. The original culture medium was replaced with 0.1 mM peptide-containing medium, and the mixture was cultured for an additional 1, 3, or 7 days. Then, the culture medium was removed, and 100 µL of CCK-8 solution (diluted with medium at a ratio of 1:10) was added. The optical density (450 nm) value was tested on a microplate reader (Bio-Tek) after 2 h of incubation in a CO₂ incubator at 37 °C. The significant influences of variables among the three groups under the same culture time were examined via one-way ANOVA and the difference between two groups was analyzed via Tukey’s post hoc test.

Statistical analysis

Data were analyzed via GraphPad Prism software. The normality of the data distribution was evaluated via the Shapiro-Wilk test. We used one-way ANOVA analysis for intergroup comparisons and Tukey’s HSD test to correct for multiple tests. The sample size of each group comprises nine samples. A significance level of 0.05 was considered.

Results

Peptide design and synthesis

The design, synthesis, and physicochemical properties of the peptides, as well as their amino acid sequences, are presented in Fig. 1; Table 1. In the design of MDP-a, we simply substitute amino acids without altering the total number of amino acids to increase protein stability and amphipathicity. In line with the dual simulation of the NCPs strategy and given that DPP is the most abundant NCPs in dentin, we combined CBP1495 (collagen binding sequence) and DSSSDSSDSS (remineralization sequence) from DPP to devise MDP-3DSS. After purification, the purity of the polypeptides exceeded 95% (Figure S1−2). MS analysis confirmed that the synthesized peptides were the desired target products (Figure S3−4). The molecular weights were as follows: 1668.91 Da (MDP-a) and 2363.52 Da (MDP-3DSS). The purity and structure of these peptides met the experimental requirements and are suitable for further experimental studies.

Fig. 1.

Design drawing of MDP-a and MDP-3DSS

Table 1.

Molecular characteristics of the peptides

Peptide.	Amino acid sequence	The number of amino acids	MW	PI	GRAVY
CBP1495	CPKESCNLFVLKD	13	1495.8	6.05	−0.08
MDP-a	CDWWSCWLDVLKD	13	1668.9	3.93	−0.08
MDP-3DSS	CPKESCNLFVLKDDSSDSSDSS	22	2363.5	4.04	−0.75

Compared with those in CBP1495, the thickened labeled amino acids have changed

MW average molecular weight, pl isoelectric point, GRAVY grand average hydropathicity

Affinity analysis of MDPs and type I collagen binding

The K_D values for MDPs and CBP1495 are within the 10⁻³−10⁻⁶ M range (Table 2), indicating high affinity for type I collagen. Lower K_D values indicate stronger binding. The K_D values of MDPs are lower than that of CBP1495 (59 µM), at 18 µM for MDP-a and 18.9 µM for MDP-3DSS, indicating an increase in the binding force with collagen after the modification by the derived peptides.

Table 2.

SPR analyzes the interaction between peptides and type I collagen

Injection variables Analyte 1 Solution	Quality Kinetics Chi² (RU²)	1:1 binding ka (1/Ms)	Kd (1/s)	KD(M)	free energy (kJ/mol)	Rmax (RU)
CBP1495	19.3	1.06*104	6.24*10−1	5.90*10−5	−27.9	190942.7
MDP-a	15.1	3.29*104	5.94*10−1	1.80*10−5	−30.8	53152.0
MDP-3DSS	15.6	3.08*104	5.80*10−1	1.89*10−5	−30.5	65526.7

The stabilization effects of MDPs on calcium and phosphorus solutions

Figure 2A shows TEM images of calcium and phosphorus solutions with or without MDPs. The control group exhibited irregularly shaped aggregates with diameters of 233.10 ± 81.46 nm. MDP-a group exhibited uneven, dense, spherical aggregated particles with diameters of 48.51 ± 13.17 nm. In contrast, MDP-3DSS groups presented aggregated particles arranged in loose chains, with smaller single-particle diameters of 18.61 ± 3.19 nm. The results showed that the nanoparticles formed by the mixture of MDPs and calcium and phosphorus ions were substantially smaller than those formed by the control group. EDS analysis of the nanoparticles confirmed that they were aggregates of calcium and phosphorus ions (Fig. 2B). The SAED analysis results (Fig. 2A) indicated that all of them were amorphous, which confirmed that they were ACP nano-precursors. These results suggest that MDP-a and MDP-3DSS can act as stabilizers by effectively restricting the size of ACP nano-precursors, with MDP-3DSS being particularly effective.

Fig. 2.

The stabilizing effect of MDPs on ACP. TEM images of each group A. The upper right corner is the SAED diffraction pattern of the sample. EDS mapping of the calcium and phosphorus of the above samples B

Effect of MDPs on the adsorption of calcium ions by collagen

The ranking of the calcium ion content from highest to lowest was as follows: MDP-3DSS group, MDP-a group, and the control group (Fig. 3A). Collagen membranes treated with MDPs demonstrated significantly greater calcium ion adsorption than the control group did (p < 0.0001), indicating that MDPs can enhance the ability of the collagen membrane to adsorb calcium ions. In addition, MDP-3DSS promoted the adsorption of calcium ions by collagen more effectively than MDP-a, with a statistically significant difference (p < 0.0001).

Fig. 3.

The content of calcium ions (A) adsorbed on the collagen membrane after 7 days of mineralization and the weights of minerals (B)

Mineralization effect of MDPs on collagen fiber

After 7 d of mineralization, TEM and EDS analyses revealed distinct differences among the groups (Fig. 4). Compared with the disordered electron deposition outside the fiber of the control group, the collagen fiber in MDP-3DSS group exhibited a uniform and dense electron distribution, with significantly increased electron density, indicating a highly mineralized structure. While MDP-a group also presented an increase in electron density, the distribution along the collagen fiber was notably uneven, with a substantial amount of electrons present outside the fiber (Fig. 4A). EDS analysis further demonstrated that in MDP-3DSS group, calcium and phosphate ions were highly concentrated along the fiber axis, with a minimal presence of these ions outside the fiber. Conversely, the calcium and phosphate ions in MDP-a group were more scattered along the fiber axis. Additionally, many free calcium and phosphate ions were still found outside the fibers. In the control group, calcium and phosphate ions were not aligned along the collagen fiber axis but instead formed disordered calcium-phosphate crystal aggregates outside the fiber (Fig. 4B). However, both MDP-a and MDP-3DSS groups exhibited evident fiber mineralization, as indicated by the red arrow, with black minerals deposited along the longitudinal axis of the collagen fiber. The SAED pattern (Fig. 4d-f) shows a characteristic diffraction ring of hydroxyapatite and arc-shaped patterns: MDP-a (002,300) and MDP-3DSS (002,112,401,122), indicating that minerals are deposited along the longitudinal axis of the collagen fiber, whereas only one diffraction ring (300) was observed in the control.

Fig. 4.

Mineralization effect of MDPs on collagen. TEM images of collagen fiber after 7 d of mineralization A. (d-f) is the SAED diffraction patterns of the fiber indicated by the red arrow in the sample above. EDS mapping of the calcium and phosphorus of the above samples B

According to the TGA results (Fig. 3B), compared with the control group, both MDP-3DSS and MDP-a groups showed statistically significant mineralization effects. MDP-3DSS group exhibited the highest residual mineral content, whereas the control group presented the lowest residual mineral content (p < 0.0001).

Biocompatibility and remineralization test of dentin induced by MDPs

SEM imaging revealed that after 35% phosphoric acid etching, the smear layer on the dentin surface was effectively removed, collagen fiber exposed, and the dentinal tubule remained open (Fig. 5A, a). In the cross-sectional observations, the dentinal tubule orifices were open, the diameter was widened to a trumpet shape, and collagen fiber exposure can be seen in the dentin tubules under high magnification. Finally, at least a 8 μm thick artificial demineralized dentin layer was produced (Fig. 5B, b). After remineralization, most of the dentinal tubules in the control group remained in an open state (Fig. 6A, a). In MDP-a and MDP-3DSS groups, dense mineral growth was observed on the surface and inside the dentinal tubules (Fig. 6B, C). Most dentinal tubule diameters in MDP-a and MDP-3DSS groups were significantly reduced, as indicated by the red box in Fig. 6b-c, or even completely closed.

Fig. 5.

The surface (A) and cross-sectional (B) of morphological characteristics of demineralized tooth samples through SEM. (a-b) An enlarged view of the red box content of (A-B), respectively

Fig. 6.

The surface morphological characteristics of the remineralized tooth samples were observed via SEM. a-c An enlarged view of the red box content of (A-C), respectively

The results of the cross section are consistent with those of the surface. In MDP-a and MDP-3DSS groups, the dentin tubules orifice was blocked by dense and deep minerals, and the mineral depth reached 7 μm (Fig. 7b, c). Thus, the surface and deep layers of the tubules had newly deposited mineral blockages. Moreover, no significant mineral deposits were observed in the tubules of the control group (Fig. 7A, a). The plugging rate of each group after remineralization is shown in Fig. 8A. Compared with those of the control group, MDP-a and MDP-3DSS groups increased significantly (p < 0.0001). These results indicate that both MDP-a and MDP-3DSS are applicable for promoting the remineralization of dentin. Compared with group MDP-a, MDP-3DSS group present a greater plugging rate of dentinal tubules (p = 0.0019).

Fig. 7.

SEM was used to observe the cross-sectional morphological characteristics of the remineralized dentin slices subjected to different treatments. The yellow areas point to the remineralization area in the tubules. a-c Magnified pictures of the yellow box content of (A-C), respectively

Fig. 8.

Biocompatibility and remineralization tests of dentin induced by MDPs. A Plugging rates of each group after remineralization (**p < 0.01, ****p < 0.0001). B Analysis of the cell activity of hDPSCs via the CCK-8 assay (p > 0.05)

As shown in Table 3, the means and standard deviations (SDs) of each group revealed the changes in the mechanical properties of the dentin under different treatment conditions. Following the erosive challenge, a significant decrease in the elastic modulus and microhardness of dentin was observed. The remineralization treatment improved the mechanical properties of the dentate surface. Compared with those of the acid etching group, the elastic modulus (p < 0.0001) and microhardness (p = 0.0013) of MDP-a group and the elastic modulus (p < 0.0001) and microhardness (p < 0.0001) of MDP-3DSS group improved. Compared with those of MDP-a group, the elastic modulus and microhardness values of MDP-3DSS group were closer to those of sound dentin group.

Table 3.

Means and standard deviations of the elastic modulus and microhardness in all the tested group

	elastic modulus (GPa)	microhardness (GPa)
Etched	5.40 ± 1.48a	0.26 ± 0.10A
MDP-a	15.14 ± 2.74b	0.49 ± 0.08B
MDP-3DSS	22.26 ± 2.14c	0.72 ± 0.12C
Sound dentin	25.85 ± 3.24d	0.81 ± 0.16C

Different lowercase letters indicate statistically significant differences in the elastic modulus. Different uppercase letters indicate statistically significant differences in microhardness (p < 0.05)

Figure 8B shows the effects of MDPs on the cell activity of hDPSCs. No significant reduction in the viability of hDPSCs was observed on days 1 (p = 0.7847), 3 (p = 0.6749), or 7 (p = 0.9451) after co-culture with MDPs. In summary, the preliminary results indicate that MDPs exhibits satisfactory short-term biocompatibility with hDPSCs.

Discussion

On the basis of non-classical nucleation theory, biomimetic mineralization mediated by NCPs and their analogs represents one of the hot spots of research on the treatment of caries. Aspartate-serine-serine (DSS), a common repeating unit in DPP, has been shown to significantly promote mineralization, with at least 3DSSs in this process [8]. The dual simulation of NCPs strategy is promising for the synthesize of a novel peptide (MDP-3DSS) through CBP1495 (collagen-binding sequence) and DSSDSSDSS (remineralization sequence) for dentin remineralization. Furthermore, amphiphilic peptides have gained attention in various biomedical applications in recent years, including mineralization [16]. Therefore, a novel amphiphilic peptide (MDP-a) was designed in this study on the basis of the known ability of CBP1495 to bind strongly to collagen. When designing MDP-a, we pay attention mainly to the balance of the hydrophilic and hydrophobic parts and the impact of specific amino acids on peptide properties. In our study, we substituted three hydrophilic amino acids in CBP1495 with the hydrophobic amino acid tryptophan. This substitution enhanced the amphiphilicity of the peptide, which is crucial for improving its stability, which is a key factor for the application of mineralized peptides in complex biological environments. The acidic properties of aspartic acid enable it to form stable coordinate bonds with calcium ions, which is highly important for stabilizing ACP and promoting the mineralization process. Thus, we also introduced acidic aspartic acid through replacement to improve the binding ability of MDP-a with calcium ions.

In additional, the short peptide MDPs have an amino acid length of 13–22 and a molecular weight of less than 6000 Da. This is an advantageous because they can penetrate into the gap region of collagen to complete intrafibrillar mineralization while maintaining a lower synthesis cost. In nature, the pivotal role of NCPs in regulating mineralization is attributed to their acidic amino acids or negatively charged properties, which enable them to bind with the positively charged regions in the gaps of type I collagen fibers and subsequently promote further crystal growth within these collagen fibers gaps [17]. According to predictions via the ProtParam website, the isoelectric points of MDP-a and MDP-3DSS are lower than the pH of the oral saliva environment. These peptides carry a negative charge in the oral environment. This negative charge characteristic further promotes their specific combination with calcium ions, thus playing a key role in mineralization. According to the ICP-MS and TEM results, MDPs can increase the ability of the collagen membrane to adsorb calcium ions. On the basis of the aforementioned characteristics, we reasonably speculate that MDPs possess a mineralization potential similar to that of NCPs.

The binding properties of peptides to collagen are crucial for promoting the bottom-up mineralization of dentin [18, 19]. The SPR used in this experiment can be used to monitor the interactions between molecules. It provides data on binding affinity and offers the advantages of being label-free, highly sensitive, and quantitative. Currently, only one study indicating that NCPs mimetic peptides derived from osteopontin rely on SPR to detect their affinity for collagen, with a K_D value of approximately 5 mM [20]. Notably, the K_D values of MDPs with collagen developed in this study were 18 µM and 18.9 µM (Table 2), demonstrating a stronger binding force than the peptides described in a previous study. The K_D value of MDP-3DSS was lower than that of the parent peptide CBP1495, indicating that its calcium-binding functional region in DPP can adsorb collagen, which is consistent with previous observation [21]. In addition, we observed that the binding affinity of the amphiphilic peptide MDP-a for collagen was greater than that of CBP495. We speculate that this may be due to their amphiphilic nature: amphiphilic polymers can generate charges through ionization and adsorb onto the surface of collagen with opposite charges or enhance the interaction with collagen through hydrogen bond formation, hydrophobic group interactions, and other mechanisms. According to Table 1, the isoelectric points of MDP-a and MDP-3DSS are below the pH level of the oral environment, so they are acidic in the mouth. As the overlapping areas of collagen are rich in positively charged amino acid residues, peptides may also bind to collagen through electrostatic interactions [22].

The extreme instability of ACP can hinder both internal and external mineralization of the fibers, and its stability can be affected by particle size, pH, temperature and other factors [23]. The critical step in regulating hydroxyapatite deposition in collagen fibers is the control of ACP size by NCPs, which prevents premature crystallization and maintains stability. This allows ACP to permeate the gaps in collagen and initiate mineralization [24]. In this study, MDPs were confirmed by TEM, SAED and EDS to contribute to stabilizing ACP and forming the nano-precursors with smaller diameters. The particle diameters in the MDPs group were far smaller than those in the control group, and the molecular weights of MDPs were much lower than 6000 Da, these features facilitate the accumulation of ACP carried by MDPs within collagen fibers [25]. The particles formed by MDP-a group were marginally larger than those formed by MDP-3DSS group. This difference might be attributed to the ability of amphiphilic peptides to self-organize into various structures, such as nanovesicles, in aqueous solutions or the susceptibility of hydrophobic side chains to aggregation, which results in an adsorption effect [26, 27].

The main organic component of dentin is type I collagen fibers, so intrafibrillar mineralization is a crucial link in the biomineralization of dentin. In recent years, intrafibrillar mineralization has become a research hotspot for exploring the restoration of dentin demineralization caused by caries and other factors [28, 29]. The two-dimensional collagen mineralization model used in this study revealed that MDPs can induce intrafibrillar collagen mineralization. This is closely related to the ability of MDPs to stabilize ACP, promote the adsorption of calcium ions and their strong affinity for collagen. These characteristics enable them to permeate into the internal water-filled compartments of collagen fibers and induce intrafibrillar mineralization [30]. In contrast, the control group without MDPs stabilization only formed large, disordered crystals outside the collagen, restricting the collagen mineralization process. Interestingly, the collagen of MDP-3DSS group was obviously denser than that of MDP-a group. MDP-a group exhibited pronounced external mineral deposition, with calcium and phosphate aggregation along the fiber axis being less pronounced than in MDP-3DSS group. These findings suggest that MDP-3DSS is more effective in guiding calcium and phosphate ions into collagen fibers and ensuring uniform deposition. MDP-3DSS excels in promoting collagen fibers mineralization, likely because of its ability to form smaller, more stable ACP and facilitate calcium-ion adsorption and guidance for their aggregation on collagen fibers. This enables the ACP carried by MDP-3DSS to accumulate within collagen fibers gaps and progressively achieve uniform mineralization. The mineral thermal analysis of the collagen membrane also confirmed the better mineralization effect of MDP-3DSS group. Commercially available remineralizing agents, such as fluoride and bioactive glass, are cost effective but cannot facilitate remineralization in the deep demineralized layer of dentin. This incomplete remineralization restricts collagen repair, making the restored dentin prone to dissolution and insufficient mineralization, and ultimately compromising its mechanical properties. In contrast, bio-inspired peptides, which mimic the function of NCPs, can achieve intrafibrillar mineralization, thus offering the advantage of deep-layer dentin restoration [31]. Unfortunately, the current TEM data are not sufficient to definitively confirm the presence of interfibrillar mineralization. Future studies should incorporate TEM observations of collagen fibers membranes and dentin to better elucidate the precise locations of fibers mineralization.

This study was conducted on biomimetic mineralized peptides in the form of an aqueous solution similar to most studies [32, 33]. However, when the oral cavity is full of flowing liquid, continuous saliva tends to remove the therapeutic agent. Therefore, the binding ability of collagen plays a crucial role in enhancing the remineralization effect [34]. However, SPR can reflect only the binding stability between peptides and collagen. Further research should use a fluorescence microscopy and molecular dynamics simulations to visualize their interactions during the dentin remineralization process. In addition, compared with existing remineralizing agents such as fluoride, ACP stabilizers (such as CPP-ACP), and NCPs-derived peptides (amelogenin-derived peptides), whether peptides with strong collagen-binding properties can form a better integrated interface and show stronger acid resistance and wear resistance are not explored in this study. In the future, in-depth studies on the permeabilty and chemical stability of the mineral layer are imperative to comprehensively evaluate the advantages of MDPs in the remineralization process.

In the present study, mineral deposits were observed on both the dentin surface and in the deeper cross-sectional layers, and a significant reduction in the number of open dentin tubules was found. This is distinct from the simple deposition of large crystals of calcium and phosphorus on the dentin surface. This “bottom-up” remineralization method is applicable for restoring the hardness and enzyme exclusion mechanism of exposed collagen fibers in the mixed layer. This illustrates its massive potential in preventing caries and improving the durability of resin-dentin bonding. The minerals newly formed in the process of remineralization mimic the interlocking tissue junctions in dentin, which can enhance the mechanical properties of the dentin after demineralization [35]. This is one of the main goals of remineralization therapy. In our research, we found that after treating demineralized dentin with MDPs, its mechanical properties were significantly improved. Although the microhardness and elastic modulus did not fully return to healthy levels, this effective remineralization process led to dentinal tubule blockage and an increase in the overall mechanical performance of the dentin. The resulting strong and mechanically durable interface is crucial for maintaining dental structural integrity and handling mechanical stress [36].

Although the results of in vitro experiments are encouraging, there are still many challenges in turning these findings into clinical applications. For example, in the real oral environment, the surface of dentin is usually covered with a pellicle or biofilms, and there are a variety of proteases in saliva. The results of this experiment simulate only the pH value and calcium/phosphorus concentration of the oral environment. This simulation does not account for the dynamic pH fluctuations or the presence of other complex components, such as enzymes and biofilms, within the oral cavity. In addition, this experiment did not compare the efficacy with that of existing commercial remineralization treatment agents (fluoride, CPP-ACP, amelogenin-derived peptides, etc.) and should be further investigated in future experiments.

Although our experiments revealed that 12 h of peptide treatment is enough to achieve remineralization, the degradability of the peptides means that their active concentration may drop below the ideal remineralization level during this period. This could affect the sustainability and stability of the remineralization effect if not replenished in time. Therefore, before clinical transformation, we still need to explore optimization strategies and attempt different application material types (e.g., sustained-release hydrogels) to ensure that peptides can effectively penetrate and interact with dentin in dentin repair applications and consider the continuous and effective effects of peptides on dentin. Although MDPs have shown short-term remineralization and short-term biocompatibility with hDPSCs in in vitro experiments, future in vivo studies will also require long-term mineralization and biosafety observations. In fact, although mineralized peptides show potential in dentin repair, some studies have put forward different views. For example, because peptides have a cost disadvantage compared with fluoride, bioactive glass and other materials and because of their own degradability, some experts have questioned the popularization of mineralized peptides in dentin restoration [37–39]. In addition, these methods usually inhibit spontaneous precipitation by slowing the release of ions, which requires a longer processing time to promote the full mineralization of collagen to effectively cover the exposed dentin tubules. Experts also have great differences in whether these methods can achieve efficient remineralization. Nevertheless, insights from this in vitro model enhance our understanding of the mechanisms of action of biomimetic peptides in dentin mineralization. Applying knowledge from in vitro dentin models to in vivo research could provide valuable insights into the molecular mechanisms of mineralized tissue formation.

In conclusion, the present study successfully developed and evaluated two novel biomimetic mineralized peptides, MDP-a and MDP-3DSS, which show strong collagen affinity. Our results demonstrated that MDP-3DSS forms smaller ACP, promotes the adsorption of more calcium ions by collagen, and has superior mineralization effects on collagen fibers and dentin than does MDP-a. Future research should focus on developing practical application forms of these peptides and conducting in vivo experiments to further investigate their efficacy and safety.

Abbreviations

NCPs

Non-collagen proteins

SPR

Surface plasmon resonance

ACP

Amorphous calcium phosphate

HDPSCs

Human dental pulp stem cells

MDPs

MMP2-derived peptides

DPP

Dentin phosphoprotein

DMP-1

Dentin matrix protein 1

TEM

Transmission electron microscope

SEM

Scanning electron microscope

TGA

Thermogravimetric analysis

SAED

Selected area electron diffraction

EDS

Energy dispersive X-ray spectroscopy

MS

Mass spectrometry

HPLC

High-performance liquid chromatography

ICP-MS

Inductively coupled plasma mass spectrometer

SDs

Standard deviations

TFA

Trifluoroacetic acid

HEPE

The 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

Authors’ contributions

HL: Writing– review & editing, Writing– original draft, Data curation. GL: Writing– review & editing, Writing– original draft, Data curation. XL: Writing– review & editing, Validation. BW: Writing– review & editing, Supervision, Data curation, Funding acquisition, Project administration, Resources. All authors have read and confirmed the final version of the manuscript and agreed to the submission.

Funding

This work was supported by The Shenzhen Science and Technology Program, Grant/Award Number: JCYJ20220530162408019 and JCYJ20220530162408020; and The President Foundation of Southern Medical University Shenzhen Stomatology Hospital, Grant/Award Number: 2022A001 and 2022B001.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The studies involving humans were approved by the Ethics Committee of Shen-Zhen Stomatology Hospital (LLWYH-PJ-20240829-001). The studies were conducted in accordance with the local legislation and institutional requirements. The methods used in this study were in compliance with the 2013 revision of the World Medical Association’s Declaration of Helsinki. Informed consent for trial participation was obtained from all patients involved in the study.

All authors gave their final approval and agreed to be accountable for all aspects of the work.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.