Caries is one of the most widespread diseases, occupying first place among chronic diseases in children [1]. Ammonium hexafluorosilicate (AHFS) has been under active study in recent years as a potential anticaries substance [2,3,4,5,6]. The features of the action of AHFS are due to its hydrolysis in saliva to generate soluble forms of silicon dioxide, which catalyze the precipitation of calcium phosphate [7] and lead to prolonged occlusion of dentin tubules. The anticaries action of the fluoride component of AHFS could probably be increased by replacing the ammonium cation with organic “onium” cations with specific types of pharmacological activity, particularly antibacterial. These compounds include the recently synthesized cetylpyridinium hexafluorosilicates [8], guanidium derivatives [9], and pyridinium derivatives [10], whose cations have antibacterial activity and stimulate salivation. Animal experiments have demonstrated [11, 12] that the use of “onium” hexafluorosilicates leads to significant improvements in the biochemical parameters of the dental pulp and periodontium, decreasing the number and depth of caries lesions and providing a high level of prophylactic efficacy against caries. The aim of the present work was to synthesize and study the structure, physicochemical properties, and anticaries activity of a novel potential anticaries drug – 2-amino-4,6-dihydroxypyrimidinium hexafluorosilicate.

Experimental Chemical Section

Synthesis was carried out using commercial hexafluorosilicic acid (45%, analytical grade) and 2-amino-4,6-dihydroxypyrimidine (Acros). Experimental studies used sodium fluoride NaF (Khimzavod Ftorsolei, Russian Federation) and ammonium hexafluorosilicate (NH4)2SiF6(Reachim, Russian Federation, pure grade). IR absorption spectra were taken on a Spectrum BX II FT-IR System spectrophotometer (Perkin Elmer) (4000 – 350 cm-1, samples in KBr tablets). EI mass spectra were recorded on an MX-1321 spectrometer (direct sample injection into the source, electron ionization energy 70 eV). 19F NMR spectra were recorded on a Varian Gemini-200 spectrometer (188.14 MHz, solvent D2O, standard CFCl3).

2-Amino-4,6-dihydroxypyrimidinium hexafluorosilicate (L1H)2SiF6(I). Portions of 1.27 g (0.01 mol) of 2-amino-4,6-dihydroxypyrimidine (L1) were dissolved in 700 ml of boiling methanol and the resulting solution was supplemented with 15 ml of 45% hexafluorosicilic acid and the reaction mix was left to complete evaporation of solvents. Elemental analysis data for the resulting colorless crystalline product I were consistent with calculated values and the yield was 76%.

The mass spectrum, m/z, (Irel), was: 127 (7.9), [ML1–H2CN]+ 99 (1.45), [SiF3]+ 85 (100).

The IR spectrum, vmax, cm-1, was: 3581, 3411, 3282, 3147, 3041, 2988 ((NH2), N+H), (OH)), 2858 (CH), 740 (SiF).

The 19F NMR spectrum (D2O), δ, ppm, was: – 133.0. The calculated structure of complex I was determined using an Xcalbur E diffractometer (room temperature, double-coordinate CCD detector, graphite monochromator, MoKα radiation).

Crystals of I (C8H12F6N6O4Si, M = 398.33) were monoclinic: a = 18.1871(10), b = 6.0848(3), c = 12.5977(6) Å, β = 91.333(5), V = 1393.74(12) Å3, space group 12/a, Z = 4, ρcalc = 1.898 mg/m3, μ = 0.277 mm-1, F(000) = 808. A total of 2015 reflections were recorded, of which 1221 were independent. Final refinement results: Rl= 0.0352, wR2 = 0.0840 using 1023 reflections with I > 2σ (I); R1 = 0.0443, wR2 = 0.0894 for all reflections. Structures were calculated and refined using SHELX97 [13]. All non-hydrogen atoms were refined in anisotropic approximation. Hydrogen atoms in amino and imino groups were found by difference Fourier synthesis and refined in isotropic approximation. The lengths of some bonds and valence angles for I are given in Table 1 and the geometrical parameters of H bonds are given in Table 2. Crystallographic data for I were deposited in the Cambridge Structural Database (SSCD No. 1485315).

Table 1. Bond lengths (d) and valence angles (ω) for compound I
Full size table
Table 2. Parameters of H Bonds in Compound I
Full size table

2-Amino-4,6-dihydroxypyrimidine was identified in salt I in the diketo-tautomeric form. The ionic composition of I was constructed on the basis of L1H+ pyrimidinium cations and SiF62- anions at a ratio of 2:1, linked by the NH···F H-bond system and CH···F contacts (Table 2, Fig. 1). The protonation center in L1H+ cations, as in the case of the pyrimidine cations previously studied by x-ray diffraction analysis (L2H)2SiF6(L2 = 2-aminopyrimidine) and (L3H)2SiF6(L3 = 5-cyanocytosine) [14], is one of the nitrogen atoms of the pyrimidine ring. Protonation of the nitrogen atom in L1H+ increases the C(1)-N(1)-C(2) angle in the heterocycle to 125.18(19)° as compared with the 119.3° for nonprotonated L1 [15]. The structure of the pyrimidinium cation in I is identical to that in (L1H)2(MoBr6)(H3O·H2O) [16]. The geometry of the SiF62- anion in I is a distorted octahedron and the length of the Si-F bond is in the range 1.6479(18) – 1.6984(12) Å. Redistribution of Si-F bond lengths is due to inclusion of the fluorine atoms of the anion in H bonds of different strengths with the H donor fragments of the cations. An excess of proton acceptors in the system is responsible for the bifurcated nature of the hydrogen bonds with the involvement of NH groups and the existence in the crystal of homomeric motifs, linear chains of SiF62- anions connected by short contacts with F(2)···F(4) = 2.771 Å, along with cation layers stabilized by one short CH···O contact and two NH···O hydrogen bonds; these motifs are combined in the three-dimensional structure by NH···F hydrogen bonds, in which all the fluorine atoms take part (Table 2, Fig. 2).

Fig. 1.
figure 1

Structure of compound I showing thermal ellipsoids, hydrogen bonds, and partial numbering of atoms.

Full size image
Fig. 2.
figure 2

A fragment of the packing in the structure of I.

Full size image

The purity of I was not determined; however, given that the study substance consists mainly of monocrystals, it can be assumed to have relatively high purity.

Experimental Biological Section

Gels containing the fluorinated compounds NaF, AHFS, and I were prepared on the basis of carboxymethylcellulose (sodium salt) gel [11]. Substance concentrations in the gel were selected such that the fluorine dose was 1.88 mg/kg.

Animal experiments were carried out in accordance with the “European Convention on the Protection of Vertebrates used for Experimental and Other Scientific Purposes” (Strasbourg, 1986) and the Ukrainian Law “Protection of Animals from Cruelty” (Ukraine, 2006). Experiments were run using 35 white Wistar rats (females, aged one month, mean live weight 40 ± 1.5 g), which were divided into five groups (Table 3). Rats of groups 2 – 5 were kept on the Stephan caries-inducing diet (CID) (50% sucrose) [17]. All rats of the experimental groups (groups 3 – 5) and the control group (group 2) received daily oral application of gel containing fluoride substances at a dose of 0.3 ml per application, covering the teeth and gums, for 30 days (except Sundays). Rats received no food for 1 h after applications.

Table 3. Caries Prophylactic Actions of Fluorinated Compounds
Full size table

Animals were subjected to euthanasia on experimental day 31 under thiopental (OAO Kievmedpreparat) anesthesia (20 mg/kg) by total cardiac exsanguination.

Pulp was extracted from the incisors and homogenates were used for assay of alkaline (AlkP) and acid (AcP) phosphatases (μkat/kg) [18] and the mineralization index (MI) was calculated as the AlkP/AcP ratio [19].

AlkP and alanine aminotransferase (ALT) were assayed in serum [20].

Jaws were harvested and the number and depth of caries lesions to the teeth were determined [17]. Caries prophylactic efficacy (CPE) was calculated as:

$$ \mathrm{CPE}=\left[\left(\mathrm{A}-\mathrm{B}\right)/\mathrm{A}\right]\Big(100\%, $$

where A is the number of caries lesions to the teeth in rats receiving the CID and B is the number of caries lesions in rats receiving the CID + a fluorinated compound.

Study results were subjected to standard statistical processing with calculation of the arithmetic mean (M) and the error of the arithmetic mean (± m). Values in groups were compared using Student’s t test. Significant differences were identified at p < 0.05 [21].

CPE results for fluorinated compounds are shown in Table 3. This shows that AHFS and I significantly decreased the numbers of caries lesions, by 22.7% and 45.5% respectively, i.e., I had a high CPE, five times greater than that of sodium fluoride.

Table 4 shows results on phosphatase activity and MI in the dental pulp of rats receiving the CID and fluorinated substances. These show that caries was associated with a significant decrease in AlkP activity and a significant increase in AcP activity, with a 40% decrease in pulp MI. Gel containing NaF significantly increased AlkP activity and significantly decreased AcP activity. This resulted in a twofold increase in MI as compared with the group of rats receiving placebo. In the case of compound I, MI in this group of rats was just as low as in rats receiving the CID and placebo, which may be evidence for differences in the mechanisms of the anticaries actions of sodium fluoride and AHFS on the one hand and compound I on the other. We note that the actions of previously studied hexafluorosilicates with cations which are substituted guanidinium and pyridinium derivatives were in all cases accompanied by increases in pulp AlkP activity and normalization of MI [11].

Table 4. Activity of Phosphatases and Mineralization Index (MI) of the Dental Pulp in Rats Receiving Fluorinated Agents
Full size table

Figure 3 shows serum AlkP levels. These data show that the CID induced a very significant (by 57%) increase in the AlkP level, which significantly decreased in response to all the fluorinated substances tested. The mechanism of the increase in serum AlkP activity in rats receiving the CID and the decrease in the level in response to fluorinated compounds remains unknown. For compound I, in contrast to NaF and AHFS, there was a significant increase in ALT activity (by 39%, Fig. 4), which may be evidence of a hepatotoxic action [20]. Pyrimidine derivatives are known to have a wide spectrum of biological activity [23], though results of PASS prediction for 2-amino-4,6-dihydroxypyrimidine did not confirm the probability of this type of activity. The presumptive source is active metabolites of L1.

Fig. 3.
figure 3

Effects of fluorinated substances on serum alkaline phosphatase (AlkP) activity in rats receiving the CID. 1) Normal; 2) placebo; 3) NaF; 4) AHFS; 5) compound I. Significant differences compared with (p < 0.05): * group 1; ** group 2.

Full size image
Fig. 4.
figure 4

Effects of fluorinated substances on serum ALT activity in rats receiving the CID. 1) Normal; 2) placebo; 3) NaF; 4) AHFS; 5) compound I. Significant differences compared with (p < 0.05): * group 1; ** group 2.

Full size image

Thus, the results obtained here provide evidence that compound I has a quite high CPE, notably greater than the CPE of both sodium fluoride and AHFS. In terms of its mechanism of biological activity, this substance has significant differences from fluorides, which to some extent act via a pulp activation mechanism [22]. An apparent disadvantage of compound I as a potential candidate caries prophylactic agent is the presence of a hepatotoxic action, which appears to be linked with the specific effects of the pyrimidine cation, which will stimulate the search for novel compounds among the hexafluorosilicates of pyrimidine derivatives with minimal toxic effects.