Dental caries, caused by a breakdown of dental hard tissue, is among the world's most common health problems, affecting the vast majority of children and adults [1], [2]. Dentin hypersensitivity is also a common dental problem with prevalence rates ranging from 4% to 74% [3]. Unlike enamel, dentin is more susceptible to caries progression due to the less mineral phase and more organic matrix compared to enamel [4]. Generally, a physiological equilibrium between demineralization and remineralization is kept to protect dental hard tissue [5]. Once the equilibrium was broken, dentin demineralization, dentinal tubule enlargement, and collagen exposure would arise, resulting in dental caries or dentin hypersensitivity. Thus, an effective remineralization strategy is necessary to prevent dentin caries and treat dentin hypersensitivity.

Dentin is a mineralized collagen matrix mainly comprised of 30% organic matrix, 50% mineral, and 20% water, while collagen fibrils make up the primary part of demineralized dentin [6], [7]. Though the mineralization ability of the dentinal collagen matrix remains controversial, most evidence tends to support its ineffectiveness in initializing HA nucleation and growth [8]. Collagen fibrils themselves fail to induce effective nucleation in the absence of remineralization agents [9]. For immature dentin in vivo, the non-collagenous proteins (NCPs) existing in the dentin organic matrix serve as nucleation templates, indispensable for hydroxyapatite (HAp) growth and biological mineralization [10], [11]. For mature dentin, however, NCPs are unable to play similar roles [12]. Hence, the assistance of remineralization agents with nucleation capacity is imperative for dentin remineralization.

The topical application of remineralization agents is a representative non-invasive clinical approach, and various nucleation templates have been studied. As a well-established strategy, fluoride decelerates demineralization and accelerates the remineralization process by coprecipitating calcium (Ca) and phosphate (P) ions and enhancing the precipitation of fluoridated apatite [13], [14]. Other ion-based remineralization agents, including Ca/P-releasing materials and bioactive glass, can induce epitaxial deposition of Ca and P ions over existing apatite seed crystallites of partially demineralized dentin to achieve remineralization [15], [16], [17]. Besides, it has been reported that the peptide containing 8 aspartate–serine–serine repeats (8DSS) could promote nano-minerals regeneration, inducing effective dentinal tubule occlusion [18]. In addition, PAMAM dendrimers are recognized as valid nuclear templates for facilitating mineral growth and organization owing to their self-assembly behavior and protein-like structure [19]. Nevertheless, the remineralization agents available attach to the dental tissue weakly, prone to removal by the constantly flowing saliva. Thus they are unable to stay on the demineralization lesion for a period long enough, adverse to the remineralization effect. Therefore, enhancing the adsorption force between the agents and dentin is mandatory for achieving desirable remineralization efficacy under oral conditions.

Given that the collagen fibrils are the primary components of demineralized dentin, the collagen-binding ability is crucial for the therapeutic agents remaining on demineralized dentin to promote the remineralization effect. Non-collagenous proteins (NCPs) provide attachment sites for collagen type I, which have been recognized to possess high affinity to dentin fibrils [20]. Besides, NCPs can offer preferential nucleation sites for Ca/P and apatite, steering and regulating dentin mineralization [21], [22]. Their abundant acid residues of phosphorylated serine and threonine could serve as nucleators or anchoring molecules for mineral deposition [23]. Considering the great challenge of extracting and purifying natural NCPs from dentin, peptides derived from NCPs come out as alternatives to natural NCPs [24]. The sequence DSESSEEDR derived from dentin matrix protein 1 (DMP-1), one kind of NCPs, was proved to have high binding affinity with immobilized collagen [25]. Therefore, DSESSEEDR is used as part and parcel of the novel peptide in our study. In addition to collagen binding, adsorption of hydroxyapatite is also requisite for mineral precipitation to facilitate dentin remineralization. Statherin, an active inhibitor of Ca and P precipitation in the oral cavity, is in charge of hydroxyapatite (HAp) adsorption owing to the polyacidic regions [26]. DpSpSEEKFLRRIGRFG (SN15) derived from statherin has been reported to play an essential role in controlling hydroxyapatite growth and it can be absorbed onto the surface of HAp through electrostatic interaction and conformational change [27]. Hence, it is promising to combine the two sequences aforementioned to reinforce the dentin remineralization effectiveness.

Accordingly, we adopted the sequence DSESSEEDR as N-terminal and SN15 as C-terminal to fabricate a novel peptide DSESSEEDRDpSpSEEKFLRRIGRFG (DGP) by Fmoc solid-phase synthesis (Fig. 1a). DGP is endowed with dual high affinity to both collagen fibrils and hydroxyapatite, thereby adsorbing strongly to demineralized dentin and grabbing mineral crystals via interaction with type I collagen and HAp (Fig. 1b). The novel peptide is supposed to defend against continuously flowing saliva, serving as a nucleation template and enhance dentin remineralization efficacy both in vitro and in vivo.