Remineralization of Demineralized Enamel via Calcium Phosphate Nanocomposite

Abstract

Secondary caries remains the main problem limiting the longevity of composite restorations. The objective of this study was to investigate the remineralization of demineralized human enamel in vitro via a nanocomposite containing nanoparticles of amorphous calcium phosphate (NACP). NACP were synthesized by a spray-drying technique and incorporated into a dental resin. First, caries-like subsurface enamel lesions were created via an acidic solution. Then, NACP nanocomposite or a commercial fluoride-releasing control composite was placed on the demineralized enamel, along with control enamel without a composite. These specimens were then treated with a cyclic demineralization/remineralization regimen for 30 days. Quantitative microradiography showed typical enamel subsurface demineralization before cyclic demineralization/remineralization treatment, and significant remineralization in enamel under the NACP nanocomposite after the demineralization/remineralization treatment. The NACP nanocomposite had the highest enamel remineralization (mean ± SD; n = 6) of 21.8 ± 3.7%, significantly higher than the 5.7 ± 6.9% for fluoride-releasing composite (p < 0.05). The enamel group without composite had further demineralization of −26.1 ± 16.2%. In conclusion, a novel NACP nanocomposite was effective in remineralizing enamel lesions in vitro. Its enamel remineralization was 4-fold that of a fluoride-releasing composite control. Combined with the good mechanical and acid-neutralization properties reported earlier, the new NACP nanocomposite is promising for remineralization of demineralized tooth structures.

Introduction

In the USA, 166 million dental restorations were placed in 2005 (Beazoglou et al., 2007). Composites are increasingly used because of their esthetics and direct-filling capability (Ruddell et al., 2002; Xu et al., 2006; Ferracane, 2006; Drummond, 2008; Samuel et al., 2009). Advances in fillers and polymers have improved the restorations (Watts et al., 2003; Imazato, 2009; Spencer et al., 2010; Ferracane, 2011). Nonetheless, half of restorations fail within 10 years, replacing them consumes 50 to 70% of the dentist’s time, and secondary caries is the main reason for replacement (Mjör et al., 2000; Jokstad et al., 2001; Sakaguchi 2005; Ferracane, 2011; Demarco et al., 2012).

Composites containing calcium phosphate (CaP) particles are promising for combating caries. These composites have been shown to release Ca and P ions and remineralize tooth lesions in vitro (Dickens et al., 2003; Langhorst et al., 2009). However, traditional CaP composites used particle sizes of 1 to 55 µm and had low mechanical properties that were inadequate for bulk restoratives (Dickens et al., 2003; Langhorst et al., 2009). Recent studies have reported novel nanocomposites which contained CaP and CaF₂ nanoparticles with sizes of about 50 to 100 nm (Xu et al., 2010a). Nanoparticles of amorphous calcium phosphate (NACP) were synthesized with a mean particle size of 116 nm (Xu et al., 2011). Nanocomposites containing CaP nanoparticles are advantageous because of the small size and high surface area of the nanoparticles (Xu et al., 2010a). A previous study showed that the NACP nanocomposite had mechanical properties 2-fold those of traditional CaP composites (Xu et al., 2011). The NACP nanocomposite neutralized acid attacks, while commercial controls failed to neutralize the acid (Moreau et al., 2011). In addition, composite containing CaP nanoparticles released substantially more ions than that with micrometer-sized particles at the same filler level (Xu et al., 2007), and CaP nanocomposite possessed much higher strength, fracture toughness, and wear resistance than traditional CaP composites (Xu et al., 2010a). However, tooth lesion remineralization via a NACP nanocomposite has not been reported. Therefore, the next step is to show that the new nanocomposite, while possessing improved physical/mechanical properties, can indeed also remineralize tooth lesions.

The objective of this study was to investigate human enamel lesion remineralization in vitro via a NACP nanocomposite. Extracted molars were used, and enamel demineralization was created to be microradiographically similar to subsurface lesions in vivo. The NACP nanocomposite was placed on the top of demineralized enamel, in comparison with a commercial fluoride-releasing composite control, and a second control group in which no composite was placed on enamel. It was hypothesized that: (1) the new NACP nanocomposite would successfully regenerate the mineral lost in enamel due to acid attack; (2) enamel remineralization via the NACP nanocomposite would be higher than that via a commercial fluoride-releasing composite; and (3) both composites would produce more remineralization than enamel without the NACP or fluoride-releasing composite.

Materials & Methods

A spray-drying technique was used to synthesize NACP as previously described (Xu et al., 2011), which yielded NACPs with a mean size of 116 nm (Xu et al., 2011). The resin consisted of ethoxylated bisphenol A dimethacrylate (EBPADMA = 62.8%; all % values are mass fractions), triethylene glycol dimethacrylate (TEGDMA = 23.2%), 2-hydroxyethyl methacrylate (HEMA = 10.4%), and methacryloyloxyethyl phthalate (MEP = 2.6%), photo-activated with 0.2% camphorquinone and 0.8% ethyl 4-N,N-dimethylaminobenzoate (Langhorst et al., 2009). The NACP nanocomposite had 40% NACP and 20% glass particles. Barium-boroaluminosilicate glass particles (median diameter = 1.4 µm, Caulk/Dentsply, Milford, DE, USA) were silanized with 4% 3-methacryloxypropyltrimethoxysilane and 2% n-propylamine. A commercial fluoride-releasing nanocomposite (Heliomolar, Ivoclar, Mississauga, ON, Canada) was used as control. Heliomolar was filled with 66.7% of silica and ytterbium-trifluoride nanoparticles of 40 to 200 nm.

Caries-free human molars were collected from the School of Dentistry clinics according to a protocol approved by the University of Maryland. Teeth were disinfected in a 0.005% promodyne solution for 4 hrs and stored at 4°C in distilled water until use. Each tooth was cut with a diamond blade (Buehler, Lake Bluff, IL, USA) into sections of approximately 200 μm in thickness, then polished to a thickness of 110 to 130 µm (Schmuck and Carey, 2010). For alignment of the “Before” and “After” microradiographic images, transmission electron microscopy (TEM) grids were adhered to the enamel sections. The sections were embedded in epoxy. The epoxy block holding the tooth section was polished until the enamel surface was exposed.

We created caries-like subsurface enamel lesions by immersing each tooth section-epoxy block, with the polished enamel surface exposed, in 2 mL of an acidic solution for 72 hrs at 37°C. This acidic solution consisted of 8.7 mmol/L CaCl₂, 8.7 mmol/L KH₂PO₄, 0.05 ppm F from NaF, and 75 mmol/L acetic acid (pH = 4.0). This solution had been demonstrated, in a previous study, of being capable of producing subsurface enamel demineralization in 3 days (Langhorst et al., 2009). This demineralized state of enamel sections was designated as the “Before” state, because it preceded the placement of a composite to cover the enamel for cyclic demineralization/remineralization treatment. The “After” state was after the 30-day cyclic demineralization/remineralization treatment, as described below.

Each enamel section was sandwiched between parafilm and glass slides (Fig. 1A) (Langhorst et al., 2009). The demineralized enamel edge was located approximately 1.5 mm below the edge of glass slides. Composite paste was placed on demineralized enamel to a thickness of 1.5 mm. Two composites were tested: NACP nanocomposite and Heliomolar. The third group had no composite on enamel (no-composite control). Six samples were tested for each group (n = 6). Each sample with composite was photo-polymerized for 60 sec (Triad-2000, Dentsply, York, PA, USA). The irradiance of Triad-2000 was previously measured at the specimen position, by means of a radiometer, to be 35 mW/cm² (Stansbury and Dickens, 2001). The depth-of-cure of NACP nanocomposite and Heliomolar was measured (Jung et al., 2001). Each paste was placed in a Teflon mold (diameter = 3.5 mm, thickness = 5 mm), clamped on one side to a glass slide, and to opaque Teflon on the other side. The glass slide was irradiated via Triad-2000 for 1 min. Specimens were demolded, and uncured paste was removed with a razor blade. The cured specimen thickness was measured, and the thicknesses of 4 specimens (n = 4) were averaged to be the depth-of-cure (Jung et al., 2001).

Figure 1.

Enamel demineralization/remineralization treatment. (A) Schematic of tooth section between 2 glass slides with composite to be used in demineralization/remineralization treatment. (B) Example of enamel lesion before the demineralization/remineralization treatment, illustrating the method and the various features on the image. (C) Mineral profile. Mineral loss ΔZ was calculated as the area between the mineral profile and lines ab and bc. The leading portion of the profile is sloped rather than vertical, because the front face of the thin enamel section is usually not perfectly square with the upper and lower surfaces of the section. It is nearly impossible to cut a section that is perfectly perpendicular to the original tooth surface. Hence, the leading portion of the profile in (C) can be visualized as the front edge of the tooth section. These features are similar to those of previous studies (Chow et al., 1992; Dickens et al., 2003; Langhorst et al., 2009).

The demineralizing solution consisted of: 3.0 mmol/L CaCl₂, 1.8 mmol/L K₂HPO₄, 0.1 mol/L lactic acid, and 1% carboxymethylcellulose (pH = 4.0) (Langhorst et al., 2009). The remineralizing solution consisted of 1.2 mmol/L CaCl₂, 0.72 mmol/L K₂HPO₄, 2.6 µmol/L F, and 50 mmol/L HEPES buffer (pH = 7.0). The demineralization/remineralization solutions were used to simulate oral fluid conditions (Langhorst et al., 2009). Each day, specimens were immersed in 20 mL of fresh demineralizing solution for 1 hr, and 20 mL of remineralizing solution for 23 hrs at 37°C, according to previous studies (Chow et al., 1992; Skrtic et al., 1996; Langhorst et al., 2009). This was repeated for 30 days, which created the “After” state in enamel. The demineralization treatment in vitro was somewhat more aggressive than what is typical in vivo. The reason for the use of an aggressive acid attack was to complete the experiments in a reasonable period of time, while quickly accumulating the damaging effects of acid attacks during years of restoration service in vivo (Langhorst et al., 2009). The 1-hour duration in the demineralization solution approximates the accumulated acid challenge times in a 24-hour period (Chow et al., 1992; Langhorst et al., 2009). It should be noted that individuals may have more or fewer acid challenges than those recorded in the 1-hour total time, depending on dietary habits and biofilm compositions.

Mineral profiles of each enamel section were measured via quantitative analysis of contact microradiographs (Chow et al., 1992; Langhorst et al., 2009). Microradiographs before and after cyclic demineralization/remineralization were produced on holographic film (Integraf, Kirkland, WA, USA) exposed for 30 min to Cu-Kα radiation (Faxitron, Hewlett Packard, McMinnville, OR, USA). Images were captured with an Evolution-MP5.0 digital microscope camera (Media Cybernetics, Silver Spring, MD, USA) with 12-bit grayscale values, at an intensity resolution of 4,096 gray levels and a spatial resolution of 1.4 µm/pixel (Schmuck and Carey, 2010). Digitized images were analyzed with ImageJ software (NIH, Bethesda, MD, USA). Mineral profiles were aligned via TEM grids, and normalized by the black background being set to 0% and sound enamel to 100% of mineral density. An example of a “before” enamel lesion is shown in Fig. 1B, and the mineral profile is shown in Fig. 1C. Mineral loss, ΔZ, was calculated as the area between the mineral profile and lines ab and bc (Chow et al., 1992; Dickens et al., 2003; Langhorst et al., 2009). ΔZ was obtained for each “Before” and “After” mineral profile in the same area of the same lesion. Remineralization in the enamel lesion that occurred during cyclic demineralization/remineralization is calculated as:

$$\text{Remineralization}\text{R}=(\Delta {\text{Z}}_{\text{Before}}-\Delta {\text{Z}}_{\text{After}})/\Delta {\text{Z}}_{\text{Before}}$$

where ΔZ_Before is mineral loss in enamel section before cyclic demineralization/remineralization, and ΔZ_After is mineral loss in the same area of the lesion after cyclic demineralization/ remineralization (Dickens et al., 2003; Langhorst et al., 2009).

Each enamel section generated 12 sets of “Before/After” curves at 12 different areas of the lesion, to provide the average value for that enamel lesion. Each set of “Before/After” curves generated one ΔZ_Before − ΔZ_After value. There were 12 ΔZ_Before − ΔZ_After values for each enamel section, yielding the mean ΔZ_Before − ΔZ_After value which in turn yielded the mean R value to represent that enamel section. Each group had 6 such R values (n = 6), which were used in one-way analysis of variance (ANOVA) to detect the significant effects of the groups. Tukey’s multiple comparison was used at p = 0.05.

Results

The depth-of-cure (mean ± SD; n = 4) was 3.2 ± 0.2 mm for the NACP nanocomposite, and 3.0 ± 0.2 mm for Heliomolar. They did not differ significantly from each other (p > 0.1).

Microradiographs showed typical enamel lesions before the cyclic demineralization/remineralization treatment (Fig. 2). There was significant remineralization in enamel under the NACP nanocomposite after demineralization/remineralization treatment. In contrast, enamel lesions under fluoride-releasing composite and for no-composite control had no noticeable remineralization (Figs. 2B, 2C).

Figure 2.

Representative examples of microradiographs of enamel lesions. “Before” refers to the initial enamel demineralization created in the acidic solution. “After” means after the 30-day cyclic demineralization/remineralization regimen. (A) There was successful remineralization in enamel under the NACP nanocomposite after 30 days. (B) There was little remineralization in enamel with fluoride-releasing commercial composite control. (C) There was no remineralization in the control group without a composite. Arrows in (A) point to a twist in the TEM grid, indicating that the “Before” and “After” images were taken in the same area. Arrows in (B) point to the same wedge for “Before” and “After”. Arrows in (C) indicate the same specimen edge for “Before” and “After”, confirming that the “Before” and “After” images were taken in the same area.

Demineralized enamel without composite showed further mineral loss during cyclic demineralization/remineralization (Fig. 3A). In contrast, enamel with the NACP nanocomposite showed a significant increase in mineral content after cyclic demineralization/remineralization (Fig. 3B). There was a much smaller increase in mineral content after demineralization/ remineralization for the fluoride-releasing group (Fig. 3C).

Figure 3.

Representative mineral profiles of human enamel lesions before and after the 30-day cyclic demineralization/remineralization for: (A) no-composite control, (B) the NACP nanocomposite on demineralized enamel, and (C) the fluoride-releasing commercial composite on demineralized enamel. Enamel lesions without a composite showed further mineral loss. Enamel lesions with the NACP nanocomposite had a greater increase in mineral content, while those with fluoride composite had a smaller increase in mineral content.

The remineralization values (Fig. 4) showed that enamel lesions with the NACP nanocomposite had the highest remineralization (mean ± SD; n = 6) of 21.8 ± 3.7%, much higher than the 5.7 ± 6.9% for the fluoride-releasing composite. Enamel without composite had further demineralization of −26.1 ± 16.2%. These values are different from each other (p < 0.05).

Figure 4.

Remineralization (mean ± SD; n = 6) of human enamel lesions in the 30-day cyclic demineralization/remineralization regimen: R = (ΔZ_Before − ΔZ_After)/ΔZ_Before. Enamel lesions with the NACP nanocomposite had the highest remineralization, 21.8 ± 3.7%; the fluoride-releasing commercial composite produced 5.7 ± 6.9% remineralization; enamel sections without a composite had −26.1 ± 16.2%, which means further demineralization during the 30-day cyclic demineralization/remineralization treatment. These 3 values are significantly different from each other (p < 0.05).

Discussion

Traditional CaP composites had relatively low mechanical properties that were “inadequate to make these composites acceptable as bulk restoratives” (Skrtic et al., 2000). It was suggested that CaP composites “could serve as a restoration-supporting lining materials” (Dickens et al., 2003), and that ACP composites could be “useful as pit and fissure sealants” (Skrtic et al., 2000). Currently available posterior/hybrid composites can be used in stress-bearing restorations, but they do not release Ca and P ions; conversely, composites that release Ca and P ions cannot be used in stress-bearing restorations. Therefore, there is a need to develop new composites with a combination of load-bearing properties and CaP ion release and remineralization capability. Nanotechnology presents a promising route to the development of a new generation of composites with a combination of load-bearing and remineralization capabilities. Recent studies focused on the first half of this picture and synthesized new nanocomposites that possessed greatly improved strength, fracture toughness, and wear resistance compared with those of traditional CaP composites (Xu et al., 2007, 2010a,b). The present study focused on the second half of this picture, remineralization capability, and showed that human enamel lesions were indeed remineralized in vitro, and that the lost mineral was regenerated via the NACP nanocomposite.

Caries is a dietary carbohydrate-modified bacterial infectious disease (Featherstone, 2004; ten Cate, 2006). Acidogenic bacteria ferment carbohydrates and produce organic acids, with biofilm pH dropping into the cariogenic region (Deng and ten Cate, 2004). To simulate in vivo pH cycles, previous studies tested cyclic demineralization/remineralization in vitro (Chow et al., 1992; Langhorst et al., 2009), with demineralization solution at pH 4, and remineralization solution at pH 7. Three points should be noted. First, pH 4 for the demineralization solution has been used in many studies as an accelerated model (e.g., Chow et al., 1992; Skrtic et al., 1996; Langhorst et al., 2009). While pH 4 is near the lower end of plaque pH, it is not unrealistic. For example, a previous study showed that S. mutans plaque pH reached 3.7 ± 0.3 (Kashket et al., 1989). Another study showed that oral biofilms with lactic acid could reach pH 4 (Wijeyeweera et al., 1989). Second, the 1-hour immersion in pH 4 demineralization solution thermodynamically favors demineralization; remineralization of enamel occurs during the 23-hour immersion in the remineralization solution at pH 7. Therefore, the fact that the NACP nanocomposite could release more ions at pH 4 (Xu et al., 2011) indicates that it could reduce enamel demineralization; it does not mean that it caused remineralization at pH 4. Third, the present study had a positive control (fluoride-releasing composite) tested under the same accelerated demineralization/remineralization regimen. Therefore, the continued release of ions from the nanocomposite contributed to enamel remineralization, consistent with the fact that the control without CaP ions achieved little remineralization. However, the acid neutralization properties of the nanocomposite (Moreau et al., 2011) likely also played a role in protection from demineralization. During the demineralization steps at pH 4, a localized higher pH near the enamel-nanocomposite interface due to acid neutralization would significantly reduce enamel mineral loss. While CaP ions and acid neutralization are both beneficial for remineralization and caries inhibition, further study is needed to separate the effect of CaP ions from the effect of local pH due to acid neutralization.

Calcium phosphate ions can help prevent enamel demineralization. For example, dentifrice with calcium was more effective than SiO₂ in reducing enamel demineralization (Cury et al., 2003). Calcium in mouthrinses or chewing-gums remineralized enamel lesions (Reynolds et al., 2003). Dentin remineralization was achieved via solutions containing calcium and phosphate (Hara et al., 2008). A previous study showed that CaP ion release from the NACP nanocomposite was greatly increased when pH was decreased from 7 to 4 (Xu et al., 2011). The NACP nanocomposite was responsive and increased the ion release at cariogenic pH4, when these ions were most needed to inhibit caries. Similarly, Heliomolar had a cumulative fluoride release at 84 days of 7 µg/cm² at pH 4, significantly higher than 2 µg/cm² at pH 7 (Xu et al., 2010b). Heliomolar remineralized enamel lesions by 5.7% in the present study, while control enamel without a composite had further demineralization of -26%. During the time period of 30 days, the NACP nanocomposite had enamel remineralization that was 4-fold that of the fluoride-releasing control. The effective remineralization of this study, coupled with the acid neutralization and good mechanical properties reported earlier (Moreau et al., 2011; Xu et al., 2011), indicates that the NACP nanocomposite is promising for caries-inhibiting restorations. Further studies should incorporate antibacterial agents into the NACP nanocomposite, with biofilm inoculation, to investigate its dual remineralizing and antibacterial effects on caries inhibition.

In summary, a NACP nanocomposite was shown, for the first time, to effectively remineralize demineralized human enamel in vitro. Its remineralization was 4-fold that of a commercial fluoride-releasing composite in a 30-day cyclic demineralization/remineralization regimen. Combined with its previously reported strong mechanical and acid-neutralizing properties, the new NACP nanocomposite is promising for restorations that can remineralize demineralized tooth structures.