The antibacterial properties and mechanism of nanosilver and nanozinc incorporated mesoporous calcium-silicate nanoparticles

Background: Silver (Ag) and/or zinc (Zn) incorporated mesoporous calcium-silicate nanoparticles (MCSNs) have good physicochemical characteristics and can be advanced materials for root canal filling. This study was to analyze the antibacterial properties and mechanism of Ag/Zn-MCSNs with different percentage of Ag and Zn. Methods: The antibacterial properties and the cytotoxicity of them were evaluated. Human root canals were inoculated with E. faecalis for 4 weeks to establish bacterial biofilm model. The E. faecalis biofilms were treated with MCSNs, Ag-MCSNs, Zn-MCSNs, Ag/Zn-MCSNs, calcium hydroxide (CH) and mineral trioxide aggregate (MTA) for 7 days. The ultrastructure and distribution of viable bacteria of the specimens were evaluated using SEM and CLSM. Human root canals were pretreated with CH, MCSNs, Ag-MCSNs, Zn-MCSNs and Ag/Zn-MCSNs, then the root canals were immersed in E. faecalis suspension for 7 days. The adhesion and colonization of E. faecalis on the root canal walls were observed using SEM and CLSM. Endocytosis of E. Faecalis treated by Ag-MCSNs and Ag/Zn-MCSNs were observed using TEM. Results: The MCSNs containing Ag showed better antibacterial properties than MCSNs and Zn-MCSNs ( P <0.05). MCSNs, Ag/Zn(1:1)-MCSNs and Ag/Zn(1:9)-MCSNs showed no obvious cytotoxicity ( P >0.05), while Ag-MCSNs and Ag/Zn(9:1)-MCSNs showed cytostatic effects. Zn-MCSNs slightly promote cell proliferation ( P <0.05). The MCSNs containing Ag showed antibacterial abilities against E. faecalis biofilm in different degree, and can adhere to dentin surfaces to get a continuous antibacterial effect, but MTA, MCSNs and

7 by a spectrophotometer (Spectramax190, Molecular Devices, USA). The cells not treated by extracts were used as controls. The results of optical density (OD) were obtained in sextuplicate.

Antimicrobial activity of the nanoparticles against E. faecalis biofilms
Mature human mandibular premolars with single root were collected under the protocol approved by the Ethical Committee Department, the Affiliated Stomatological Hospital of Nanjing Medical University (PJ 2017-055-001). The crowns were removed and roots were standardized to 12 mm long from root apex. The root canals were prepared using ProTaper NiTi rotary instruments (Dentsply Maillefer, Tochigi, Japan) to F3 size according to standard processes. For sterilization, they were autoclaved at 121ºC for 20 min. Then they were placed in 3 mL of E. faecalis suspension (1×10 8 /mL) and were cultured under anaerobic conditions for 4 weeks at 37°C. The BHI broth was refreshed every second day to remove dead cells and to ensure bacterial viability. CH and MTA pastes (mixed with sterile ddH 2 O=1:1.5), MCSNs, Ag-MCSNs, Zn-MCSNs, Ag/Zn(1:1)-MCSNs, Ag/Zn(1:9)-MCSNs and Ag/Zn(9:1)-MCSNs pastes (mixed with sterile ddH 2 O=1:3) were prepared. The pastes were introduced into the root canals by lentulo-spirals (Mani Inc, Tochigiken, Japan). Then the roots were placed in sterile tubes. Five specimens were tested in each group. After 7 days, each canal was gently washed with 10 mL phosphate buffer saline (PBS) to clear the intracanal paste and dried with steriled paper points.
Then the specimens from each group were split into two halves. One root-half randomly selected from each root was scanned with a SEM (HITACHI SU3500, Tokyo, Japan) or FE-SEM (QUANTA 200F, Fei, USA) according to our previously described method [8]. The other root-half was stained with fluorescent LIVE/DEAD BacLight Bacterial Viability stain (Molecular Probes, Eugene, OR, USA) according to the instructions of manufacturer. Three randomly selected canal wall areas of each root were scanned by a confocal laser scanning microscope (CLSM, LSM 710, Carl Zeiss, Germany) with a 5 μm step size at 20×lens. The excitation/emission wavelengths were 488/525 nm for SYTO ® 9 and 561/642 nm for PI. Simultaneous dual-channel imaging was used to display the green fluorescence (live bacteria) and red fluorescence (dead bacteria) using the ZEN software (Carl Zeiss, Germany).

Endocytosis of E. faecalis
Live E. faecalis showed typical spherical structure and intact cell membrane in the TEM images (Fig 9. A). For coculture groups, the destruction of cell walls and cell membranes was observed in Ag-MCSNs and Ag/Zn(1:1)-MCSNs groups (Fig 9. B-F).

Discussion
Nano-antibacterial agents have been proposed as a choose for intracanal disinfections because they can disrupt bacterial biofilm and prevent bacterial adhesion to dentin [13][14][15]. As compared with antibiotics, microorganisms are unlikely to develop resistance against nanoparticles [16]. In this study, the synthesized Ag/Zn(1:9)-MCSNs and Ag/Zn(9:1)-MCSNs possess a representative mesoporous structure, and the Ag and Zn are distributed inside the mesoporous structures. The nanoparticles were on the nano-scale and had high surface areas and pore volumes. They release Ca 2+ and SiO 4 4− in aqueous solution, producing a weak alkaline micro-environment and maintaining a high pH value over time.
In this study, MCSNs showed a slight antimicrobial activity, which is agree with our previous study [8]. The released Ca 2+ and SiO 4 4− of MCSNs produce a weak alkaline microenvironment [7], and the nano-dimension of MCSNs may interfere with bacteria metabolism, especially when they are in direct contact with bacteria wall in a liquid environment [17]. Zn 2+ had antibacterial property that might destroy the bacteria membrane structure, intracellular enzymes and the replicate ability of DNA [18]. These results were comparable as previous studies that Ag have adverse effects on human health [16][17]19]. It was interesting to find that Zn-MCSNs has advantage to promote cell proliferation, the higher the content of Zn in Ag/Zn-MCSNs, the less cytotoxicity of them.
As a cofactor and a structural and regulatory ion, Zn takes major biological roles and involves in homeostasis, immune responses, oxidative stress and apoptosis [20]. Previous studies demonstrated that Zn takes the active part in bone metabolism and has a stimulatory effect on osteogenesis [21][22][23]. However, MTA showed no obvious antibacterial effect to E. faecalis biofilms. MCSNs showed less antibacterial effective than CH against bacterial biofilms. The possible reason maybe that the changes in interaction between MCSNs and bacteria in bacteria biofilms instead of planktonic bacteria, which compromised the antibacterial effect of MCSNs [8].
Studies reported that nano-sized ZnO exhibit distinct bactericidal mechanisms [12]. Substantivity on dentin surface is another significant feature of intracanal medications.
Fan et al reported that CH paste has very weak substantivity on dentin surface, and there was no residual antimicrobial activity after CH paste being removed [9]. The findings of this study also showed that CH had no substantivity property, and the root canal surface treated with CH seemed to have enrichment effect on E. faecalis. MCSNs had no obvious resistance to bacterial adhesion and proliferation, while Ag-MCSNs and Ag/Zn-MCSNs adhered to dentin and infiltrated into dentinal tubules, which will release Ag + continuously that get a long-term antibacterial effect. Zn-MCSNs also showed enrichment effect on E.
faecalis and seem to encourage E. faecalis to form biofilms. This phenomenon may be explained by Zn promote cell proliferation, which is consistent with our cytotoxic results.

Conclusions
In this study, we successfully synthesized MCSNs containing different proportions of Ag and Zn. They have ideal physicochemical properties and obvious antibacterial activity.        C and E: ×100,000; D and F: ×200,000.