Formation of hydroxyapatite on tooth surfaces by calcium hypophosphite in saliva

Abstract

Aim

This in vitro study aimed to investigate whether calcium hypophosphite (CaP) mediates hydroxyapatite (HAP) formation in artificial and human saliva and, in a proof-of-principle toothbrushing model, promotes mineral deposition and occlusion of dentinal tubules.

Methods

A CaP solution was added to artificial saliva and human saliva, respectively. The resulting precipitate was analyzed using X-ray powder diffraction (XRD). For the toothbrushing simulations, human dentin specimens were allocated to a test group treated with 2 wt% CaP solution and a control group treated with deionized water (n = 5 per group). The specimens were immersed in artificial saliva and subjected to five consecutive 30 s brushing treatments. Confocal laser scanning microscopy (LSM) was used to assess changes in dentinal tubule occlusion before and after treatment, scanning electron microscopy (SEM) was used to examine the morphology of the deposits, and energy-dispersive X-ray spectroscopy (EDX) was used to analyze their elemental composition.

Results

In artificial saliva and human saliva, CaP led to the formation of HAP. XRD analysis of the precipitates showed characteristic HAP peaks, with a mean crystallite size of (5.7 ± 0.6) nm and crystallographic parameters of a = (9.484 ± 0.003) Å, c = (6.868 ± 0.002) Å and V = (535.0 ± 0.4) ų for the formation in human saliva. LSM and SEM demonstrated mineral deposition within open dentinal tubules following toothbrushing simulations with CaP in artificial saliva, while EDX showed that the deposits contained calcium, phosphorus, and oxygen.

Conclusion

CaP mediated HAP formation in artificial and human saliva. The proof-of-concept dentin experiment demonstrated mineral deposition within open dentinal tubules. Although firm conclusions regarding efficacy cannot be drawn, these promising findings merit further exploration in properly powered studies. With further confirmatory studies, CaP has potential as a desensitizing agent for oral care formulations.

Introduction

Due to their high biocompatibility and chemical similarity to the mineral phase of teeth and bones, synthetic calcium phosphates have versatile applications in medicine [1–4]. In dentistry, calcium phosphate-based active ingredients are used, for example, for caries prevention and enamel remineralization, as they can e.g., shift the equilibrium between tooth demineralization and remineralization towards remineralization [5–8].

Different calcium phosphate-based remineralization systems are used in oral care, including casein phosphopeptide-amorphous calcium phosphate (CPP-ACP; Ca_x(PO₄)_y ∙ nH₂O), hydroxyapatite (HAP; Ca₅(PO₄)₃(OH)), and β-tricalcium phosphate (β-TCP; β-Ca₃(PO₄)₂) [5]. Among this group of active ingredients, HAP shows the highest similarity to human enamel crystallites and can therefore be regarded as a biomimetic active ingredient for the use in various oral care fields [9–15]. In general, these calcium phosphate compounds are typically used in particulate form with particle sizes, ranging from nano- to microcrystalline dimensions. Preformed calcium phosphate particles are characterized by defined particle size distributions and have demonstrated beneficial effects in enamel remineralization. Complementary to such particulate systems, in situ calcium phosphate formation may enable mineral deposition directly at the tooth surface and within local defect structures.

The biomineralization of enamel relies on Ca²⁺ and PO₄^3- ions, which finally form HAP through highly complex biological processes [16–18]. Calcium is one of the major constituents of enamel mineral (36.5 wt% [1]), and, in combination with phosphate, it is essential for maintaining the supersaturation of saliva with respect to HAP [19, 20].

The formation of HAP in biological environments is highly complex and depends on various parameters [20]. Typically, apatite formation occurs after several formation steps by immersion in water from amorphous calcium phosphate (ACP; Ca_x(PO₄)_y(H₂O)_z) to dicalcium phosphate dihydrate (DCPD; CaHPO₄ · 2 H₂O) to octacalcium phosphate (OCP; Ca₈H₂(PO₄)₆ · 5 H₂O) to calcium deficient hydroxyapatite (CDHA; Ca_5-x(PO₄)_3-2x(HPO₄)_2x(OH)) [20, 21]. This highlights the complexity behind the formation process. Nevertheless, the whole process relies on an equilibrium between dissolved Ca²⁺, PO₄^3-, and OH^- ions on one side and the solid calcium phosphate phase on the other side.

Compounds that readily provide Ca²⁺ ions could be a promising option for oral care applications. While particulate calcium phosphates are almost insoluble at neutral pH, water soluble calcium salts could offer an additive strategy by rapidly increasing calcium concentration in the oral cavity, i.e., at the tooth surface.

Among soluble calcium compounds for the use in oral care, calcium hypophosphite (CaP; Ca(H₂PO₂)₂), also known as calcium phosphinate (Fig. 1), has recently emerged as a particularly interesting candidate [22–24]. It is highly water-soluble (154 g/L at 25 °C [25]) and provides bioavailable calcium ions. Owing to its high calcium content, high water solubility, and neutral taste, CaP has been used as a calcium supplement, including in vitamin and mineral supplement preparations for children [23, 26]. In addition, the hypophosphite anion has previously been investigated in the context of food preservation, particularly in studies using sodium hypophosphite against pathogenic bacteria in the 1990s [27–29]. CaP has been shown to be safe and classified by the US Food and Drug Administration (FDA) as being “Generally Recognized as Safe” (GRAS) [30–32].

Fig. 1. The active ingredient analyzed in this study:

Molecular structure of calcium hypophosphite (CaP).

In the field of toothpastes, a recent in vitro surface microhardness (SMH) study demonstrated that CaP efficiently remineralizes early caries lesions in enamel [22]. Notably, the remineralization efficacy of CaP was significantly higher than that of 1450 ppm fluoride provided as sodium fluoride in a cyclic demineralization/remineralization model, with percentage remineralization values of (75.4 ± 5.5)% and (60.3 ± 7.8)%, respectively. The combination of CaP and HAP resulted in the highest remineralization efficacy ((89.7 ± 3.3)%) [22]. Similar results were obtained in a recent in vitro remineralization study using microcomputed tomography (µCT), further confirming CaP as an efficient caries remineralizing agent [24].

These previous studies did not directly investigate the interaction of CaP with salivary phosphate, identify the mineral phases formed in saliva, or assess mineral deposition within dentinal tubules. Thus, the physicochemical behavior of CaP in saliva and its effects on dentin surfaces have not yet been characterized in detail. Besides enamel remineralization, calcium-based mineral deposition may also be relevant to the management of dentin hypersensitivity, as occlusion of exposed dentinal tubules represents one of the principal approaches to reducing dentin hypersensitivity [33].

Based on these considerations, it was hypothesized that CaP as highly soluble calcium salt may interact with saliva to promote mineral precipitation on tooth surfaces.

The aim of the present in vitro study was to investigate the physicochemical behavior of CaP in saliva and to characterize the resulting precipitate. In a proof-of-principle approach, it was further evaluated whether CaP in combination with saliva can induce mineral deposition and occlusion of open dentinal tubules, thereby providing initial in vitro evidence for its potential relevance in the management of dentin hypersensitivity.

Materials and methods

Substances

Calcium hypophosphite (CaP, Ca(H₂PO₂)₂, >99.8%, pharmaceutical grade), potassium hydroxide (KOH, GPR Rectapur quality, VWR International GmbH, Germany), hydrochloric acid (1 mol/L, AppliChem GmbH, Germany), magnesium chloride hexahydrate (MgCl₂· 6 H₂O, Ph. Eur. quality, VWR International GmbH, Germany), potassium dihydrogen phosphate (KH₂PO₄, Ph. Eur. quality, VWR International GmbH, Germany), sodium chloride (NaCl, GPR Rectapur quality, VWR International GmbH, Germany), potassium chloride (KCl, GPR Rectapur quality, VWR International GmbH, Germany), dipotassium hydrogen phosphate (K₂HPO₄, VWR International GmbH, Germany), and calcium lactate pentahydrate (C₆H₁₀CaO₆ · 5 H₂O, AnalytiChem GmbH, Germany) were used without further purification.

Solution preparation

CaP solutions were prepared directly before each experiment by dissolving CaP (2 wt%) in deionized water. The pH value of the solution was adjusted to 7.2 using KOH. The solutions were used directly after preparation.

Artificial saliva was prepared based on a protocol by Amaechi et al. [34]. For the preparation of artificial saliva, MgCl₂ · 6 H₂O (0.148 mmol/L), K₂HPO₄ (4.59 mmol/L), KH₂PO₄ (2.38 mmol/L), KCl (8.39 mmol/L), NaCl (14.37 mmol/L), and calcium lactate pentahydrate (1.76 mmol/L) were dissolved in deionized water. The pH value was adjusted to 7.2 using KOH. Sodium carboxymethylcellulose and methyl-4-hydroxybenzoate, which were mentioned in the literature protocol, were not used to facilitate further investigation of the formed precipitate.

Mediated formation experiments and structural analysis

Artificial saliva (125 mL) was prepared as described above. 19.24 mL of a 2 wt% CaP solution (pH = 7.2, adjusted using KOH) was added under stirring and mixed for 2 min in a beaker (400 rpm). In the experiments, a ratio of 1:6.5 (v/v) between the 2 wt% CaP solution and saliva was chosen because literature shows that this is the mean volume ratio between the amount of applied toothpaste and the volume of saliva which is produced during 2 min of tooth brushing [35].

For comparison, the experiment was repeated with human saliva, filtered through a syringe filter (0.22 µm, minisart high flow syringe filter, Sartorius AG, Germany). Human saliva was donated by one healthy volunteer after providing informed consent. The pH value of the saliva was 7.5.

After 2 min of mixing, the dispersions were centrifuged (5000 rpm, 10 min, Megafuge 16, Thermo Scientific, Germany), the supernatant was discarded and the residue was dried overnight at 37 °C. The dried residues were ground with a pestle and mortar and further investigated using XRD.

XRD was performed on a D8 Advance powder diffractometer (Bruker) in Bragg-Brentano reflection mode with Cu K_α radiation (λ_Kα = 1.5418 Å; U = 40 kV, I = 40 mA). The powdered samples were placed on a single crystal silicon sample holder to minimize scattering and dispersed using a small amount of ethanol. After drying in air, the samples were measured from 5 to 90° (2Θ) with an increment of 0.01° and a counting time of 1.5 s per step, resulting in a total measurement time of 3.5 h. Qualitative phase analysis was performed with the software Diffrac.Suite EVA V7.1 (Bruker) using reference patterns of HAP (#09-0432) and brushite (dicalcium phosphate dihydrate (DCPD); CaHPO₄ ∙ 2 H₂O) [1] (#09-0077) from the ICDD (International Centre for Diffraction Data) database. After instrumental calibration with the reference material LaB6 from NIST (SRM 660b), quantitative Rietveld refinement was performed using TOPAS 7.0 (Bruker) to calculate the lattice parameters and the average crystallite size from diffraction peak broadening [36].

The Scherrer equation was used to calculate the crystallite size (Eq. 1) [37]:

$$D_{hkl}=\left(K\lambda \right)/\left({\beta }_{hkl}\cos \theta \right)$$ 1

Here, D_hkl describes the estimated crystallite size in the (hkl) direction of the Miller indices. K is a form factor, λ is the wavelength of measurement, β_hkl describes the peak width at half-height of the 2ϴ signal and ϴ is the diffraction angle.

SEM was carried out using an Apreo S LoVac instrument (ThermoFisher Scientific, USA). Energy-dispersive X-ray spectroscopy (EDX; UltraDry EDS detector, ThermoFisher Scientific) was used for elemental analysis and visualization of the elemental distribution. The images were recorded with a resolution of 512 by 340 and a pixel size of 0.08 µm. An acceleration voltage of 10.0 kV and a magnification of 5000 was applied.

Toothbrushing simulations on human dentin

Human teeth were collected at the University of Texas Health, San Antonio/USA (Institutional Review Board Approval number: HSC20080233N). The dentin samples were prepared by the Fraunhofer Institute for Microstructure of Materials and Systems IMWS (Halle (Saale), Germany). The teeth were cut, embedded in epoxy resin and ground to 4000 grit until the dentin was exposed. The samples were immersed in citric acid (2%, 2 min), sonicated and subsequently rinsed with deionized water. After shipping, the samples were stored at 4 °C until use.

The in vitro toothbrushing simulations were conducted in a toothbrush simulator (ZM-3, SD Mechatronik GmbH, Germany). All dentin specimens were brushed with manual toothbrushes (TePe Select medium, TePe D-A-CH GmbH, Germany) at predefined brushing parameters (duration: 30 s; force: 1.5 N; path length: 10 mm; number of cycles: 38). These parameters were selected to provide standardized and reproducible mechanical conditions for the exploratory proof-of-principle study. They were not intended to reproduce individual variations in clinical toothbrushing behavior. The dentin samples were allocated to two groups with five specimens in each group. Each specimen was brushed with deionized water before characterization with LSM to remove loose deposits from the sample surface. The characterized samples after brushing with deionized water are labeled as “untreated” in the results section. After characterization of the untreated specimens, the samples were subjected to brushing simulations using one of the following treatment solutions: (I) 2 wt% CaP solution at pH 7.2 (test group), or (II) deionized water (control group). The samples were immersed in artificial saliva (10 min, 37 °C) and subsequently installed in the brushing simulator. The sample chamber was filled with artificial saliva (8.67 mL) to cover the dentin sample. Immediately before the start of the brushing process, the test solution was applied directly to the dentin surface (1.33 mL; 1:6.5 (v/v), test solution/artificial saliva). After each brushing cycle, the samples were rinsed with deionized water, applied through a funnel (20 mL, washing angle of the sample: 45°, distance between funnel outlet and sample surface: 20 mm, outlet diameter: 5 mm).

Before treatment and after 5 cycles, the dentin samples were examined using LSM (VK-X1100, Keyence Deutschland GmbH, Germany). The laser images were recorded with the software provided (MultiFileAnalyzer, v.2.2.0.93, Keyence). The images were recorded with a 50x objective (CF Plan Apo, 50x/0.95) at brightness settings of 7038 under coaxial lighting conditions in the center of the samples.

No formal sample size or power calculation was performed because the dentin experiment was designed as an exploratory, qualitative proof-of-principle assessment. The microscopic findings were evaluated qualitatively; therefore, no statistical comparison between the groups was performed.

Institutional review board statement

Extracted human molars were used in this study with approval of the Institutional Review Board (IRB Approval #: HSC20080233N) of the University of Texas Health San Antonio, USA.

Results

Mediated formation of hydroxyapatite with calcium hypophosphite in saliva

The addition of a 2 wt% CaP solution to saliva leads to the formation of white precipitate and an increasing turbidity of the solutions. The dried and ground residues from the precipitation experiments in both artificial saliva and human saliva show characteristic diffraction peaks of HAP, Ca₅(PO₄)₃(OH) (Fig. 2).

Fig. 2. Crystallographic analysis.

(I) X-ray powder diffractogram of the ground precipitates from calcium hypophosphite (CaP) with artificial saliva shows characteristic diffraction peaks of hydroxyapatite (HAP) and brushite. (II) The X-ray powder diffractogram of the ground powder from human saliva solely shows characteristic diffraction peaks of HAP.

In the XRD diffractogram of CaP with artificial saliva, peaks that can be assigned to the lattice planes of HAP with the Miller indices (1 0 0), (2 0 0), (0 0 2), (1 1 1), (3 0 0), and (0 0 4) are highlighted (Fig. 2, (I)). Other peaks overlapping the HAP peaks can be assigned to brushite (dicalcium phosphate dihydrate (DCPD); CaHPO₄ ∙ 2H₂O) with highlighted lattice planes (0 2 0), (0 2 1), (0 4 1), and (0 2 2).

In the experiment with human saliva, only peaks that are characteristic for HAP are visible (Fig. 2, (II)). Here, additional peaks are highlighted for the lattice planes (2 1 1), (3 1 0), (2 2 2) and (2 1 3) of HAP.

Lattice parameters of the solid substances were calculated by quantitative Rietveld refinement (Table 1).

Table 1.

Crystallographic parameters of ground residues from calcium hypophosphite (CaP) with artificial saliva and human saliva, respectively.

System	Crystallographic parameters (HAP)
	a-axis/Å	c-axis/Å	V/Å3	Mean crystallite size, domain size/nm
CaP + artificial saliva (this study)	9.44 ± 0.01	6.84 ± 0.01	527 ± 1	6.4 ± 0.6
CaP + human saliva (this study)	9.484 ± 0.003	6.868 ± 0.002	535.0 ± 0.4	5.7 ± 0.6
Bulk enamel [42]	9.444 ± 0.054	6.871 ± 0.011	-	-
Powdered enamel [42]	9.440 ± 0.052	6.873 ± 0.031	-	-
Adult human enamel [4]	9.441 ± 0.003	6.880 ± 0.003	-	-
Adult human dentin [4]	9.421 ± 0.003	6.887 ± 0.003	-	-
Synthetic HAP [36]	9.4471 ± 0.0006	6.8848 ± 0.0004	532.13 ± 0.07	-
Geological HAP [36]	9.4249 ± 0.0004	6.8838 ± 0.0004	529.56	-

Additional data from literature allows comparison to bulk and powdered enamel, adult human enamel and dentin as well as synthetic and geological hydroxyapatite (HAP).

The system with 2 wt% CaP and artificial saliva shows smaller crystallographic parameters (a = (9.44 ± 0.01) Å, c = (6.84 ± 0.01) Å, V = (527 ± 1) ų) than the system with 2 wt% CaP and human saliva (a = (9.484 ± 0.003) Å, c = (6.868 ± 0.002) Å, V = (535.0 ± 0.4) ų) (Table 1). In contrast, the mean crystallite size of the system with artificial saliva ((6.4 ± 0.6) nm) is larger than the one obtained with human saliva ((5.7 ± 0.6) nm). Parameter values from literature show that bulk enamel, powdered enamel, adult human enamel, adult human dentin as well as synthetic and geological HAP also differ in their crystallographic parameters, depending on the physical condition (Table 1).

Toothbrushing simulations on human dentin

Toothbrushing simulations with 2 wt% CaP solution on human dentin in artificial saliva were carried out and LSM analysis was performed to assess changes to exposed dentinal tubules (Fig. 3).

Fig. 3. Dentin occlusion.

Laser scanning microscopy images of human dentin before treatment (left) and after five brushing cycles with calcium hypophosphite (CaP) in artificial saliva (right).

The LSM image of an untreated dentin specimen shows open dentinal tubules and after five consecutive brushing cycles, including rinsing after each brushing routine, the same area shows distinct mineral deposition on the surface (Fig. 3).

All dentin samples treated with CaP in artificial saliva show mineral deposition on the dentin surface after five brushing cycles, whereas no deposition was observed in the control group (Fig. SI1).

SEM images and subsequent EDX analyses of the treated dentinal tubules reveal accumulation of calcium, phosphorus, and oxygen within the tubules (Figs. 4, 5).

Fig. 4. Elemental mapping.

Scanning electron microscopy (SEM) image and energy-dispersive X-ray spectroscopy (EDX) elemental maps of a human dentin sample after treatment with calcium hypophosphite (CaP) in artificial saliva. The images show the distribution of carbon, nitrogen, calcium, phosphorus, and oxygen.

Fig. 5. Elemental analysis.

Energy-dispersive X-ray spectroscopy (EDX) analysis of the elemental composition of a dentinal tubule filled after treatment with calcium hypophosphite (CaP) in artificial saliva.

The EDX analysis shows that the intertubular dentin between the tubules contains increased amounts of carbon and nitrogen, compared to the deposits inside the tubules. The white deposit inside the tubules results in regions enriched in calcium, phosphorus, and oxygen, compared to the interspace (Figs. 4, 5).

Discussion

Mediated formation of hydroxyapatite with calcium hypophosphite in saliva

The key finding of this part of the study is that CaP mediates HAP formation upon contact with artificial and human saliva. This supports the proposed mechanism by which highly soluble CaP acts as a source of readily available calcium ions that interact with salivary phosphate. In simplified form, the proposed CaP-mediated formation of HAP can be represented by the following chemical equation:

$${5\mathrm{Ca}}^{2+}\left(\mathrm{aq}\right)+{3\mathrm{PO}}_4^{3-}\left(\mathrm{aq}\right)+{\mathrm{OH}}^{-}\left(\mathrm{aq}\right)\to {\mathrm{Ca}}_5{\left({\mathrm{PO}}_4\right)}_3\left(\mathrm{OH}\right)\left(\mathrm{s}\right)$$

This simplified equation illustrates the respective contributions of calcium supplied by CaP and phosphate supplied by saliva.

The additional formation of brushite in artificial saliva is consistent with the known dependence of calcium phosphate phase formation on pH, Ca/P ratio, and solution composition. In contrast, the exclusive detection of HAP in the precipitate generated with human saliva may reflect differences in pH and the presence of organic salivary components that influence calcium binding, nucleation, and crystal growth.

In the laboratory, HAP is often synthesized under artificial conditions by mixing exactly stoichiometric quantities of Ca²⁺ and PO₄³⁻ in solution at pH > 9 under boiling conditions and in the absence of CO₂ or by solid-state reactions of calcium phosphates above 1200 °C or by hydrothermal synthesis [1, 38]. Lee et al. found in precipitation experiments that the Ca/P ratio and the pH value strongly affect the formation of HAP [39]. They found that at a Ca/P ratio of 2.2, HAP powders can successfully be synthesized at 25 °C at pH 10 and pH 11. Interestingly, at pH 10 the synthesized HAP resulted in a higher calcium deficiency than at pH 11 where fewer H⁺ ions are present [39].

The results in the present study show that the pH values of the CaP solution (pH = 7.2, adjusted with KOH) and artificial saliva (pH = 7.2) provide a suitable environment to form HAP under physiological pH conditions. The experiments also show that brushite is formed as a by-product. The different outcome between artificial saliva and human saliva might stem from the different pH values: The human saliva had a slightly higher pH (pH = 7.5) than the artificial saliva. Thus, the human saliva provides a higher OH^- ion concentration for crystallization, presumably preferring the formation of HAP over brushite. As described before, the ratio between calcium from the CaP solution and saliva can also play an important role in the outcome. In this study, a ratio of 1:6.5 (v/v) between the 2 wt% CaP solution and saliva was chosen. Pulfer et al. showed that this corresponds to the mean volume ratio between the amount of toothpaste and the volume of saliva produced during 2 min of toothbrushing [35]. Translated into a Ca/P ratio, the performed experiments with artificial saliva result in an initial Ca/P ratio of 2.9. A large ratio between calcium and phosphate can, besides other environmental parameters, lead to preferred formation of HAP (Ca₅(PO₄)₃(OH)) over brushite (CaHPO₄ · 2 H₂O) or calcium deficient species [39].

The formation of brushite as a potential by-product of the interaction between CaP and saliva is also noteworthy: Brushite is known as a precursor calcium phosphate species that can be transformed into HAP through the equilibrium between dissolved Ca²⁺, PO₄³⁻ and OH⁻ ions on one side and solid calcium phosphate phases on the other side [20]. Thus, HAP as well as brushite formation is promising in terms of remineralization of dental hard tissue.

Besides parameters like stoichiometry and the pH value, organic components in saliva can have an influence on the precipitation process and thus lead to differences in mineral formation [1, 39, 40]. In particular, salivary proteins and peptides are known to create an ion-binding environment for calcium that may influence remineralization by mediating the crystallization process [20].

In addition to the qualitative information about the calcium phosphate species, the XRD measurements also allow calculations of the crystallographic parameters and compare them to literature to obtain deep insights into the crystal structure. Here, the crystallite size D_hkl is inversely proportional to the signal width β_hkl (Eq. 1). Broad peaks of HAP in the diffractogram indicate small, less crystalline apatite, which is also the case when dentin is compared to enamel: Dentin shows much broader diffraction peaks than enamel, due to more, smaller crystals located in dentin [4, 37].

The structural parameters of the investigated HAP from CaP show values that are in good accordance with structural parameters of enamel and dentin as well as synthetic HAP from literature. The HAP from CaP with artificial saliva (a = (9.44 ± 0.01) Å, c = (6.84 ± 0.01) Å, V = (527 ± 1) ų) shows slightly smaller lengths of the a-axis and the c-axis than the HAP from CaP with human saliva (a = (9.484 ± 0.003) Å, c = (6.868 ± 0.002) Å, V = (535.0 ± 0.5) ų). A smaller a-axis results in a smaller basal plane of the hexagonal unit cell, while the c-axis leads to a shorter direction length, perpendicular to the basal plane [41]. A shorter a-axis indicates changes affecting the arrangement of ions, for example smaller ions that are incorporated in the lattice sites when carbonate substitutes for PO₄^3-. Anions like chloride or HPO₄^2- can increase the size of the a-axis [42]. A shorter c-axis can be interpreted as a change in the OH^- ion sites like substitutions by small anions. In the experiments, CaP in artificial saliva results in HAP with a shorter basal plane than in human saliva. Artificial saliva contains e.g., magnesium ions that can lead to ion substitution of Ca²⁺. Additionally, carbonate might have been incorporated differently between the systems. The shorter c-axis compared to HAP from human saliva can also stem from fewer OH^- ions provided from the artificial system (pH = 7.2) compared to the human saliva (pH = 7.5) and lead to vacancies in the OH^- ion lattice sites. The dimensions of the a-axis from HAP from artificial saliva are in good accordance with adult human enamel and powdered HAP from literature. The c-axis on the other hand is smaller than all the systems from literature. Incorporated impurities like Cl⁻ for OH⁻, HPO₄²⁻ for PO₄³⁻, or CO₃²⁻ for PO₄³⁻ substitution can influence the crystalline structure, commonly noticeable in an increased solubility [41]. The HAP generated in human saliva shows a slightly larger a-axis and a shorter c-axis, compared to HAP, enamel and dentin from literature. This indicates an increase of the a-axis by a HPO₄-for-PO₄ substitution or “lattice” H₂O [41]. Additionally, such impurities can suppress crystal growth, leading to a small domain size.

A highly soluble agent forming HAP crystals with a small domain size has potential over particulate agents in terms of tailored, size-independent remineralization of smallest irregularities in the tooth surface and in subsurface lesions.

Beyond this tailored mineralization approach, the formation of HAP is advantageous per se, because HAP is a biomimetic calcium phosphate known to induce enamel and dentin remineralization [6], and reduce initial bacterial colonization on tooth surfaces [12, 43]. Moreover, HAP acts as pH buffering agent [44], and as tooth whitening agent [15]. Due to the formation of HAP from CaP in saliva, it is plausible that the in situ-formed HAP exhibits modes of action comparable to those described for preformed particulate HAP, including reduced initial bacterial adhesion [12, 43] and pH buffering [44].

In addition to these established modes of action of HAP, the ion-based nature of the present approach provides a further mechanistic advantage: because mineral formation starts from dissolved calcium and phosphate ions rather than from preformed particles, the generation of sub-nanometer calcium phosphate structures is principally possible.

In this context, early calcium phosphate precursor units such as Posner’s clusters, with a diameter of approximately 0.95 nm (9.5 Å) [20, 45, 46], are substantially smaller than commercially used nano-HAP particles and may serve as starting points for subsequent HAP mineralization. For comparison, particle diameters of approximately 173 nm have been reported for nano-HAP used in oral care [47], although the exact dimensions of nano-HAP depend on the specific material, manufacturing process, and formulation.

The experiments show that CaP mediates the formation of HAP in saliva under physiological pH conditions due to provision of large amounts of bioavailable calcium ions. In principle, other soluble calcium salts can also act as calcium donors. However, CaP offers various properties that are favorable for the use in oral care formulations such as a high water solubility, a neutral taste, and a favorable safety profile. Also antibacterial effects of sodium hypophosphite in the field of food preservation have been investigated [23, 27–29].

Taking together, these findings provide the physicochemical basis for investigating whether CaP-mediated mineral formation can also promote deposition within open dentinal tubules (see section “Toothbrushing simulations on human dentin”).

Toothbrushing simulations on human dentin

The key observation from the proof-of-concept toothbrushing simulations was mineral deposition within open dentinal tubules after treatment with CaP in artificial saliva, consistent with tubule occlusion under the investigated conditions.

Toothbrushing simulations and subsequent imaging with LSM were conducted to investigate the potential application of CaP for dentinal tubule occlusion during toothbrushing. The images show that exposed dentinal tubules, which are the cause for dentin hypersensitivity, have a size of a few micrometers in diameter. In a study by Lopes et al., superficial dentinal tubules showed diameters of 2.42 µm [48]. Gevkaliuk et al. showed that the tubules are widest near the pulp with 1.7 to 2.8 µm diameter [49]. Besides nerve desensitization, for example by potassium salts such as potassium nitrate, the occlusion of open dentinal tubules is one core strategy for treatment of dentin hypersensitivity [33].

Previous in vitro studies have shown that CaP is an efficient remineralizing agent for early enamel caries lesions, as demonstrated by SMH and µCT analyses [22, 24]. The present findings extend the previously reported caries-related CaP data by demonstrating that CaP acts as a calcium donor in saliva, leading to saliva-mediated HAP formation. The exploratory dentin findings further suggest that this mineralization process can promote deposition within open dentinal tubules, indicating potential relevance to dentin-related applications. However, adequately powered quantitative and functional studies are required before conclusions regarding the management of dentin hypersensitivity can be drawn.

The toothbrushing experiments provide initial proof-of-principle evidence that CaP-mediated mineral formation can result in deposition within open dentinal tubules. The calcium-, phosphorus-, and oxygen-containing deposits remained visible after repeated brushing and rinsing cycles, supporting the potential relevance of this mineralization mechanism for tubule occlusion. However, these findings represent morphological and elemental evidence; functional efficacy and long-term durability were not assessed in the present study.

The study shows that CaP mediates the formation of HAP in saliva by acting as a highly soluble calcium donor. The mineral formation occurs within 30 s and shows occlusion of dentinal tubules in toothbrushing simulations. Further in vitro focusing on dentin permeability [50] may give more insight into the resistance of the occluding effects of CaP and compare it to desensitizing agents such as nano-HAP [51, 52].

Limitations and outlook

The present study has several limitations. First, it was conducted under controlled in vitro conditions that cannot fully reproduce the complexity of the oral environment in vivo. Second, the dentin study had a small sample size and was designed as a qualitative proof-of-principle assessment without quantitative analysis of tubule occlusion. Third, the study did not assess functional changes in dentin permeability or the long-term resistance of the deposits to acidic and abrasive challenges and repeated toothbrushing. While HAP formation in artificial and human saliva was comprehensively characterized by XRD, the dentin experiments therefore provide initial in vitro evidence of mineral deposition and dentinal tubule occlusion. Future studies should include larger sample sizes and quantitative assessment of tubule occlusion, and measurements of dentin permeability. The resistance and durability of the deposits could also be evaluated following acidic or abrasive challenges and repeated toothbrushing. Direct comparisons with established tubule-occluding desensitizing agents, followed by application-oriented in situ and in vivo studies, would further clarify the practical relevance of CaP.

Conclusions

This in vitro study demonstrates that CaP mediates the formation of HAP in artificial and human saliva. The crystallographic analyses provide direct evidence of HAP formation and support the proposed mechanism by which highly soluble CaP supplies readily available calcium ions that interact with salivary phosphate. Thus, the study establishes a physicochemical basis for using CaP to promote saliva-mediated mineral formation directly at tooth surfaces.

The dentin component was designed as an exploratory proof-of-concept experiment and demonstrated calcium-, phosphorus-, and oxygen-containing mineral deposition within open dentinal tubules following repeated brushing with CaP in artificial saliva. Because of the qualitative design and limited sample size, firm conclusions regarding the magnitude, reproducibility, durability, or functional efficacy of tubule occlusion cannot be drawn from this experiment. Nevertheless, the findings suggest that promising results may be obtained in an adequately powered study and therefore merit further exploration through quantitative tubule-occlusion analyses, dentin-permeability measurements, durability testing, and comparisons with established desensitizing agents.

Overall, CaP represents a promising active ingredient for oral care formulations because it enables biomimetic, saliva-mediated HAP formation. With further confirmatory studies, CaP has potential as a desensitizing agent for oral care formulations.

Author contributions

Conceptualization and methodology, ESzW, PF, and JE; validation, PF and JE; investigation, formal analysis and data curation, PF; writing—original draft preparation, PF; writing—review and editing, JE; writing—review, ESzW and BTA. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Dr. Kurt Wolff GmbH & Co. KG, Bielefeld, Germany.

Competing interests

The authors PF, JE, and ESzW are employees of Dr. Kurt Wolff GmbH & Co. KG, Germany.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41405-026-00502-4.