**4. Discussion**

Nano-hydroxyapatite is considered one of the most promising bioactive materials with a broad spectrum of applications in preventive and restorative dentistry. N-HAp particles have been demonstrated to have a similar structure to natural enamel apatite crystals and can remineralize dental hard tissues [15,23–25]. Non-substituted n-HAp is also used for desensitizing and anti-erosion purposes [26,27]. The synthesis of nano-scaled zinc carbonate or magnesium-strontium carbonate hydroxyapatite (ZnCO3/n-HAp, Mg-SrCO3/n-HAp) represented significant progress in this field, allowing the introduction of n-HAp-based toothpastes with remineralization and antibacterial potential [28]. Different substituted n-Hap, coupled to other compounds with antibacterial activity, such as Zn PCA, chitosan, and fluoride, have been developed to achieve the goal of the remineralization of dental hard tissues [4,29–32]. Nevertheless, very few data are available in the literature about the antibacterial activity of these compounds, which is important in caries prevention because of the impact of oral biofilms in the dynamics of carious lesion formation.

Biofilms produce organic acids that cause the demineralization of the hard tissues. Moreover, biofilms can act as a reservoir of calcium and phosphate for remineralization, but also as a diffusion barrier for other compounds, such as fluoride, thus reducing the efficacy of some caries-preventive agents [16]. From a microbiological point of view, n-HAp-containing products might allow adsorbed nanoparticles to interact with bacterial cells, reducing adherence to hard tissue surfaces, thereby reducing biofilm formation.

The current in vitro study compared the effect of two substituted n-HAp-based toothpastes on early colonization and biofilm formation. Teeth specimens treated with the α toothpaste showed no significant differences in EC and BF compared to the control and β group, whereas specimens treated with β toothpaste showed a significantly lower EC and BF compared to the control (Figures 4 and 5). Both toothpaste formulations contained doping agents having antibacterial activity, namely zinc in α toothpaste and fluoride and strontium in β toothpaste, suggesting that n-HAp doping agents, rather than n-HAp itself, play a crucial role in determining the antibacterial properties of these compounds. Furthermore, fluoride and strontium have been shown to exhibit synergistic antibacterial performance [33,34]. EDS data (Table 2) showed that the zinc and fluoride concentrations were in the same range, while strontium in the β toothpaste was four times higher. A possible explanation of these

results could be that the association of fluoride and strontium was more e ffective in reducing EC and BF than zinc alone at the tested concentrations. Nevertheless, fluoride release (Figure 7), which was only found on brushed enamel surfaces, did not help in explaining the reduced EC and BF by β toothpaste on RBC surfaces. Furthermore, despite a significant initial release of fluoride, the amount of this ion was not increased in biofilms grown over enamel and RBC surfaces treated with the β toothpaste, confirming the fact that biofilms do not tend to accumulate fluoride [16].

EDS data also showed the presence of calcium and phosphate on the surfaces of RBC specimens treated with both toothpastes, as well as traces of zinc on specimens treated with α toothpaste and magnesium on specimens treated with β toothpaste (Figure 2 and Table 1). It is noteworthy that toothpaste remnants could be associated with alumina and barium glass fillers rather than silica particles. This observation, needing further confirmation, opens the possibility for such compounds to actively interact with RBC surfaces with the e ffect of reducing or preventing secondary caries occurrence.

Regarding the influence of the adherence surface on the antibacterial activity of substituted n-HAp, as the surface features can deeply influence biological behavior, the roughness of all specimens was reduced below 0.2 μm Ra [35] to exclude the possible influence of this parameter on EC and BF. The two tested surfaces were selected considering their bu ffering capacity: a bu ffering surface (enamel) and a non-bu ffering one (RBC). The non-bu ffering surface was selected because it has been demonstrated to promote the growth of biofilms with a high concentration of cariogenic bacteria [17]. Our data showed that the adherence surfaces did not influence the e ffect of the tested toothpastes on EC and BF. This result was observed for both *S. mutans* and the oral microcosm, indicating that, unexpectedly, both substituted n-HAps had the same e ffect on bu ffering and non-bu ffering surfaces independent of their chemical composition, suggesting a similar interaction of the active components with the specimen surface. EDS results confirm the presence of active compounds from both toothpastes on both adherence surfaces. This presence was seen to a larger extent on enamel compared to RBC. However, the presence of such active compounds on the tested RBC surface, as seen in the present study, is very interesting from the perspective of controlling EC and BF on these surfaces, with the final aim of reducing secondary caries occurrence. Indeed, small but accumulative e ffects can prevent dysbiotic changes in dental biofilms and help maintain a beneficial oral microbiome [36]. In this sense, the use of this kind of active principles might be an e fficient means of reducing secondary caries occurrence by actively promoting our natural microbiota and reducing the impact of the drivers of dysbiosis, such as, for instance, the presence of a non-bu ffering surface.

Two test setups were used, shaking multiwell plates and a drip-flow bioreactor to simulate clinical situations as close as possible, avoiding the variabilities that characterize in vivo studies. The first test setup was designed to evaluate EC and BF under low hydrodynamic stress conditions, for instance, during the night. Moreover, the first setup favors a progressive increase in microorganism catabolites and antimicrobial agents released from the surfaces. The second one simulated relatively high hydrodynamic stress conditions that occur during the daytime. *S. mutans* showed lower EC and BF in the first test setup compared to the second one, probably due to the limited amount of nutrients available and the accumulation of metabolic acids in the culture medium.

From the microbiological model point of view, the approach simulated as much as possible the clinical situation by testing an oral microcosm in addition to a monospecific *S. mutans* biofilm. The oral microcosm was used to evaluate the mixed oral flora developing on hard tissues, while the monospecific *S. mutans* model was used to assess the behavior of the cariogenic part of flora when treated with the tested compounds. In our study, *S. mutans* showed a higher a ffinity for RBC surfaces than for enamel, which is probably related to the bu ffering capacity of the tested surfaces being a crucial factor for both EC and BF.

Still, the microbiological models and incubation conditions may not be an accurate replica of all the complex interactions taking place in the oral environment. However, in vitro simulations are useful since they can push microenvironmental conditions to levels that might be di fficult to control, or even unethical if replicated in vivo. The culture medium was made to have a high content in sucrose, therefore promoting the growth of fermentative Streptococcus as well as the acidogenic species in the mixed inoculum. The aim was to provide a shift in microbial composition towards a pathogenic biofilm, to test the activity of the toothpastes in a harsh environment. An active principle such as substituted n-HAp is especially useful if it can provide a controlled release of ions over time and, especially, in response to defined environmental conditions. A fermentative biofilm can reach pH values lower than 4.0, promoting the demineralization of enamel (critical pH ≈ 5.5). n-HAp is known to be very stable at neutral pH. Therefore, an acidogenic microbial challenge as the one produced in the present setups helped in understanding the effect of the toothpastes when challenged by pathogenic biofilms.

Considering incubation time, 12 h was selected as toothbrushing twice a day is the most frequently advised routine, while a 24 h incubation can provide a sufficient amount of time to evaluate BF. The results showed that the most relevant differences between the tested toothpaste groups were found on *S. mutans* EC, which means after 12 h of incubation. In a translational sense, this may sugges<sup>t</sup> that the tested toothpastes provided the best effect when used twice a day, rather than a single time.
