Bactericidal activity and recovery effect of hydroxyl radicals generated by ultraviolet irradiation and silver ion application on an infected titanium surface

This study investigated the bactericidal effect, the underlying mechanisms of treatment, and recovery of biocompatibility of the infected titanium surface using a combination treatment of silver ion application and ultraviolet-A (UV-A) light irradiation. Streptococcus mutans and Aggregatibacter actinomycetemcomitans were used in suspension and as a biofilm on a titanium surface to test for the bactericidal effect. The bactericidal effect of the combination treatment was significantly higher than that of silver ion application or UV-A light irradiation alone. The bactericidal effect of the combination treatment was attributable to hydroxyl radicals, which generated from the bacterial cell wall and whose yield increased with the silver concentration. To assess the biocompatibility, proliferation and calcification of MC3T3E1 cells were evaluated on the treated titanium surface. The treated titanium screws were implanted into rat tibias and the removal torques were measured 28 days post-surgery. The titanium surface that underwent the combination treatment exhibited recovery of biocompatibility by allowing cellular proliferation or calcification at levels observed in the non-infected titanium surfaces. The removal torque 28 days after surgery was also comparable to the control values. This approach is a novel treatment option for peri-implantitis.

. Bactericidal tests using different diluent solutions. The reaction solution obtained after UV-A irradiation was diluted with ultrapure water (UW), 10% NaCl, 10% albumin solution, or MHB solution. Significant differences (p < 0.05) between the groups are denoted by different superscript letters (i.e., bars with the same letter are not significantly different). Ag(+)L(+): Mixture of silver nitrate solution (40 μM)and bacterial suspension (10 7 CFU/mL) with UV-A irradiation for 1 min, Ag(−)L(−):Mixture of ultrapure water and bacterial suspension (10 7 CFU/mL) without UV-A irradiation.
The SEM images show that the sandblasting and acid-etching treatments resulted in a rough titanium surface (Fig. 3i). Biofilms of both bacterial species formed successfully on the titanium surfaces (Fig. 3a,e). The biofilms of both bacterial species remained on the titanium surface after application of silver nitrate solution and UV-A irradiation (Ag(+)L(+)), without any ultrasound scaling (Fig. 3b,f). In the Ag(+)L(−) group, the biofilm was not observed on the titanium surface, however, particle-like amorphous deposits were observed (Fig. 3c,g). The biofilm was not observed in the Ag(+)L(+) group (Fig. 3d,h).
Based on the SEM results, the bactericidal test for biofilm was performed in the Ag(+)L(−) and Ag(+)L(+) groups after the ultrasound scaling. For the bactericidal effect on the biofilm, no CFUs were observed with a silver concentration of 1 mM in the Ag(+)L(+) group for S. mutans and A. actinomycetemcomitans, while the corresponding values for Ag(+)L(−) were 1.5 ± 0.5 and 1.6 ± 0.1 log CFU, respectively. This indicates a significant difference between the Ag(+)L(−) and Ag(+)L(+) treatments.

Analysis of hydroxyl radical generation.
To investigate whether the bactericidal effect of the combination treatment was caused by hydroxyl radicals, bactericidal tests were performed with or without DMSO as a reactive oxygen species scavenger. Figure 4 shows the results of the bactericidal tests with and without DMSO at a silver ion concentration of 400 μM. The colony-forming unit value in group Ag(+)L(+) without DMSO was 0 ± 0 log for S. mutans and 0.8 ± 0.7 log for A. actinomycetemcomitans, while the corresponding values with DMSO were 4.2 ± 0.1 and 5.1 ± 0.5 log, respectively. A significant difference was observed between the values obtained with and without DMSO in the Ag(+)L(+) group. The corresponding values in the Ag(+)L(−) group without DMSO were 3.5 ± 1.4 log for S. mutans and 5.0 ± 0.2 log for A. actinomycetemcomitans, and the values with DMSO were 4.2 ± 0.1 log and 5.1 ± 0.5 log, respectively. These values were closer to those of Ag(+)L(+) with DMSO.     5A shows the representative ESR spectra of 5,5-dimethyl-1-pyrroline N-oxide (DMPO)-OH and the spin adduct of the hydroxyl radical generated by 2 mM silver nitrate solution and UV-A (365 nm) irradiation with dispersions or biofilms of different bacteria. The presence of DMPO-OH was confirmed by hyperfine coupling constants (hfcc) of aN = aH = 1.49 mT22. The patterns of ESR spectra in all groups were consistent with that of DMPO-OH. The ESR patterns in all groups were similar to that of hydroxyl radicals. Figure 5B shows the yield of DMPO-OH generated by different concentrations of silver nitrate application with or without UV-A light irradiation to bacterial suspensions or biofilms. The generated DMPO-OH in the Ag(+)L(+) group with the bacterial dispersion ranged from 11 to 48 μM for S. mutans and from 4 to 42 μM for A. actinomycetemcomitans, while the corresponding values without bacteria were approximately 0.3-0.9 µM. The values in the Ag(−)L(−) group for S. mutans and A. actinomycetemcomitans were 1.0 and 0.7 μM, respectively. For the biofilms of S. mutans and A. actinomycetemcomitans, the generated DMPO-OH levels in the Ag(+)L(+) group ranged from 5.3 to 23.5 μM and from 7.3 to 25.7 μM, respectively. The corresponding values for the Ag(−) To further investigate the part of bacteria that reacted with silver nitrate solution and UV-A light irradiation, the hydroxyl radical levels generated from each bacterial cellular fraction were measured (Table 1). Hydroxyl radicals were generated from all cellular fractions but the yields of the generated hydroxyl radical from each bacterial cellular fraction were different. The hydroxyl radical levels generated from the cell wall were 2.3 ± 1.0 μM in S. mutans and 5.3 ± 0.7 μM in A. actinomycetemcomitans, which were higher than those in the other cellular fractions. Figure 6 shows the representative ESR spectra pattern from each cellular fraction. The ESR patterns for hydroxyl radicals in the cell wall, cell membrane, and cytoplasm were identical. Furthermore, the hydroxyl radical levels generated by LPS (1 μg/mL) and peptidoglycan (1 μg/mL), which are components of the cell wall, were 87.7 ± 5.1 μM and 2.0 ± 0.7 μM, respectively. The ESR patterns for hydroxyl radicals in LPS and peptidoglycan were identical. Figure 7 shows the cell viability of the combination treatment. The cell viability with silver ion treatment decreased depending on the silver ion concentration. On the other hand, UV-A light irradiation alone reduced the cell viability to 94% ± 16%. The viability in group Ag(+)L(+) was comparable to that in group Ag(+)L(−), except at 50 μM. No significant difference was observed between Ag(+)L(+) and Ag(+)L(−) groups from 100 to 1000 μM.  Table 1. Hydroxyl radical levels generated by the silver application and UV-A light irradiation with each bacterial cellular fraction. Since the ESR spectra generated from the cell membrane, cytoplasmic fraction and peptidoglycan were smaller than those generated from the cell wall, they are presented at higher magnification. Note that the difference in the MnO marker intensity. ○ and ▴ shows the MnO marker and DMPO-OH, respectively. The presence of DMPO-OH was confirmed by hyperfine coupling constants (hfcc) of aN = aH = 1.49 mT22.

Biocompatibility of the combination treatment in vitro and in vivo.
The in vitro findings for cell biocompatibility of the titanium surface after the combination treatment following ultrasound scaling on biofilm of A. actinomycetemcomitans are shown in Fig. 8. The proliferation of MC3T3E1 cells on the titanium surface treated by Ag(+)L(+) recovered to the levels of the control (Fig. 8A). The corresponding value for Ag(+)L(−) was significantly lower than that of the control and Ag(+)L(+) groups. The value for Ag(−)L(−) was significantly lower than that of the other groups. On the other hand, calcification in the Ag(+) L(+) and Ag(+)L(−) groups at 28 days was similar to that in the control group (Fig. 8B). No significant differences were observed among the groups.
For the in vivo investigation of tissue biocompatibility, a biofilm-infected titanium screw was prepared. After the prepared titanium screw received the combination treatment following ultrasound scaling, the treated titanium screw was implanted into rat tibia and the removal torque was measured at 28 days post-surgery. The results of the removal torque test are shown in Table 2. The removal torque value in the Ag(+)L(+) group was comparable to that in the normal group (sandblasted and acid-etched titanium screw without bacterial infection implanted into the tibia). Figure 9 shows the non-decalcified tissue around the implant at 28 days after surgery. No severe inflammation, necrosis, or drainage was observed in both groups. Partial osseointegration at the bone-implant interface was observed in both groups.  www.nature.com/scientificreports www.nature.com/scientificreports/

Discussion
The main causative organism behind peri-implantitis has not yet been identified. Aggregatibacter. actinomycetemcomitans, Porphyromonas gingivalis, Treponnema denticola, and Tannerella forsythia have all been detected at high ratios in peri-implantitis 38 . On the other hand, S. mutans has been detected in dental plaque 39 . While the biofilm is comprised of several bacterial species, S. mutans is a key contributor for the formation of biofilms 39,40 . S. mutans synthesises adhesive extracellular matrix, which allows the assembled biofilms to firmly attach onto the tooth and titanium surfaces 41,42 . Furthermore, we have previously demonstrated that S. mutans is more resistant to the antimicrobial activity of hydroxyl radicals 43 . The S. mutans in this study, therefore, provides more insight into the full potential of the bactericidal effect of this treatment.
Finally, A. actinomycetemcomitans is a gram-negative bacterium, whereas S. mutans is a gram-positive bacterium. In this study, we opted to study the two different gram stain bacterium to gain further insights into different outcomes of each strain when subjected to a combination treatment of silver ion application and UV-A irradiation.
Bactericidal effect of the combination of silver application and UV-A light irradiation. The 1 mM (169.8 ppm) silver ion concentration inhibited bacterial proliferation to log CFU 0 and this concentration is substantially higher than those reported previously. This difference could be attributed to the exposure time of silver ion and UV-A irradiation. Since shorter treatment times are better for disinfection treatment, the silver exposure time was set at 1 min in this study. In other reports, the complete inhibitory concentrations of silver ion ranged from 0.08 to 18.9 μM 22,44-47 . However, the exposure times of silver application were longer than that used in this study and no previous report has used a silver ions exposure time smaller than 1 min. For example, Choi et al. reported that the complete eradication concentration of silver ion against Escherichia coli was approximately 4.2 μM but their exposure time was 24 hours 45 .
The application of silver ions exhibited bactericidal effects in a silver ion concentration dependence manner, which is consistent with the previous reports 22 . UV-A irradiation is also known to possess a bactericidal effect and the resistance to UV-A irradiation differs across bacterial species 28 . A UV-A irradiation for 1 min exhibited bactericidal effects against A. actinomycetemcomitans but not S. mutans in this study. Nakamura et al. reported that UV-A irradiation for 4 min on an S. mutans suspension exhibited a 1−log reduction of CFU 36 . This may be due to the short irradiation time. However, the combination treatment with silver ion application and UV-A light irradiation exhibited a higher bactericidal effect than that of silver ion application or UV-A irradiation alone. This result   www.nature.com/scientificreports www.nature.com/scientificreports/ indicates that the addition of UV-A irradiation to silver ion's application enhanced the bactericidal effect. On the other hand, the biofilm remained on the titanium surface after the combination treatment with no ultrasound scaling, whereas no biofilm remained in SEM observations obtained following ultrasound scaling. In bactericidal activity's assessment, the combination treatment followed by ultrasound scaling showed a significantly higher bactericidal effect than the effect of silver ion application alone after ultrasound scaling. These results suggest that the use of the combination treatment alone was not enough to remove the biofilm but was useful as an adjunctive treatment for peri-implantitis.
Bactericidal mechanism of the combination of silver application and UV-A irradiation. Next, we investigated the mechanism underlying the bactericidal effect of this combination treatment. Park et al. reported that silver ions generate hydroxyl radicals 25 . Meanwhile, Song et al. noted that UV-A irradiation generates reactive oxygen species in water 28 . The generated hydroxyl radical induces oxidative damage in DNA, proteins, and cell membranes and leads to delays in growth 48,49 . Based on these reports, we hypothesised that the enhanced bactericidal effect of the combination treatment might be attributable to active oxygen generation. Therefore, assessments of bactericidal activity with or without the reactive oxygen species scavenger, DMSO, were performed in this study. The bactericidal effect of the combination treatment decreased with DMSO and the ESR investigations showed that generated active oxygen molecules were hydroxyl radicals. Thus, the bactericidal effect of the combination treatment may be attributable to hydroxyl radical generation.
The yield of hydroxyl radicals generated by the combination treatment increased with the silver ion concentration and was approximately 10-40-fold of that obtained with the silver application alone. However, the yield of hydroxyl radicals generated with the combination of a silver application and UV-A irradiation without bacteria was significantly smaller than that generated in the presence of bacteria. This suggests that all the 3 elements of this combination-bacteria, silver ions, and UV-A irradiation-were essential for a high yield of hydroxyl radical generation. Furthermore, the yield of the generated hydroxyl radicals from the bacterial cell wall was higher than that from the cell membrane or cytoplasm in both bacterial species. This suggests that the silver ion attaches to the bacterial cell wall and generates hydroxyl radicals under UV-A light irradiation. Furthermore, the combination treatment generated a high yield of hydroxyl radicals with LPS which suggests that the combination treatment captures free LPS and removes it via hydroxyl radicals, and may be a useful treatment for periodontitis.

Biocompatibility of the combination treatment with the silver application and UV-A irradiation.
ROS are generated by the inflammatory cells such as the macrophage or neutrophil in chronic inflammation and the generated ROS are related to the progress of periodontitis 50 . In general, the generated hydroxyl radicals were promptly disappeared 51 , whereas the ROS generation continued due to inflammatory cells in periodontitis. The results of the cell-cytotoxicity test show that the combination treatment did not enhance the cytotoxicity of silver ions. Miyayama et al. reported that exposure of the cells to 2.5 μM silver nitrate for 24 h reduced the cell viability by half 51 . However, the IC 50 of the combination treatment and silver ion application with 1 min exposure was within the range of 200-400 μM in this study. This difference is attributable to the exposure time. Foldjerg et al. reported that the cytotoxicity of silver ions was dependent on the exposure time and concentration 52 . Therefore, the cytotoxicity of ROS generated in this combination treatment is believed to be limited. Treatments using the mechanical or chemical approach have been used for cleaning infected implant surfaces; however, these methods induce surface alternation and result in a decrease of the implant surface's biocompatibility 7 . Therefore, the recovery of the titanium surface's biocompatibility after disinfection is an important factor for peri-implantitis treatment. The proliferation and calcification of MC3T3E1 cells recovered to a level comparable with that of the non-infected titanium surface. This indicates that the biocompatibility of the titanium surface was recovered by the combination treatment and potentially induced bone re-osseointegration. In vivo, the removal torque force of the titanium screw that underwent the combination treatment was consistent with that of a non-infected titanium screw. Histological observations showed that the newly formed bone attached directly to the titanium surface in both groups and the osseointegration was observed after silver application and UV-A light irradiation treatment, similar to the non-infected titanium group. These results indicate that silver ion application and UV-A light irradiation allowed recovery of the infected titanium implant surface and induced re-osseointegration. Yamada et al. reported that UV light irradiation recovered the hydrophilic titanium surface and improved cell compatibility 33 . We also report that UV-A irradiation with H 2 O 2 recovered the hydrophilic titanium surface and improved cell compatibility 37 . Thus, UV-A irradiation might have changed the surface condition hydrophilicity, which may have contributed to the recovery of biocompatibility in this study.
In this study, UV-A irradiation intensity was set at 1000 mW/cm 2 , in line with our previous report 53 . In dental treatment, 465 nm LED at an irradiance of up to 4000 mW/cm 2 has been used to cure the light-curing composite resin in oral cavities 54 . Cell viability of the combination of UV-A at 1000 mW/cm 2 and silver nitrate is statistically comparable to that without UV-A irradiation in this study. This indicates that UV-A irradiation at 1000 mW/cm 2 for 1 min is not high enough to affect cell proliferation.
UV-A has been used with an exogenous photosensitiser such as Photochemotherapy for not only skin diseases including atopic dermatitis or psoriasis 55 but also for oral mucosa manifestations of chronic graft-versus-host disease 56,57 . Therefore, we believe that the use of an adequate dose of UV-A is acceptable if the treatment efficacy prevails the risk of an adverse effect. However, when the combination treatment of UV-A irradiation and silver is applied to peri-implantitis, the bone around the implant is also exposed to UV-A. DNA damage of UV-A irradiation is different when in the chromatin conformation. Moreover, the safety range of the dose of UV-A to the bone and the effect of the combination treatment should be further investigated in a peri-implantitis model. (2020) 10:8553 | https://doi.org/10.1038/s41598-020-65411-4 www.nature.com/scientificreports www.nature.com/scientificreports/

Materials and methods
Bacterial cultures. Two species of oral bacteria, Streptococcus mutans JCM5705 (RIKEN BioResource Center, Wako, Japan) and Aggregatibacter actinomycetemcomitans JCM2434 (RIKEN BioResource Center, Wako, Japan) were used in this study. Cultures were grown anaerobically using AneroPack (Mitsubishi Gas Chemical Company, Tokyo, Japan) in a brain heart infusion (BHI) broth (Becton Dickinson Labware, Franklin Lakes, NJ, USA) or BHI broth with 10 g/L of yeast extract (BHI-YE) (Oxoid, Hampshire, UK) at 37 °C for 24 h. The prepared bacterial suspension was diluted in sterile saline and contained approximately 1 × 10 8 colony-forming units (CFUs)/mL, in line with our previous report 35 . For the biofilm-related test, the bacterial suspension was seeded on the titanium surface or plastic well and incubated in BHI-YE for Aggregatibacter actinomycetemcomitans or BHI with 1% sucrose (Kanto chemical, Japan) (BHI-S) for Streptococcus mutans at 37 °C for 24 h in anaerobic conditions. Titanium surface treatment. Commercially pure titanium disks (diameter, 5 mm: thickness, 2 mm: Grade 4; TB550, Nishimura, Fukui, Japan) were sandblasted and acid-etched, according to our previous report 37 , to obtain a rough surface similar to the commonly used commercial dental implants. Briefly, sandblasting was performed with 250 µm alumina particles at a blasting-pressure of 0.4 MPa. The specimens were acid-etched in 49% sulphuric acid for 1 h at 60 °C and were the cleaned by ultrasound-cleaning in ultrapure water (10 min), acetone (10 min), and finally ultrapure water (10 min). After cleaning, the specimens were autoclaved for 15 min at 121 °C. The samples were prepared 1 day before use and stored in ultrapure water.
Bactericidal assay. Establishment of a protocol for determination of the treatment effect. The prepared bacterial suspension (100 μL; final concentration, 10 7 CFU/mL) was mixed with 100 μL of silver nitrate solution (40 μM) in a 96-well cell culture plate. After mixing, the sample was irradiated with UV-A light (Omnicure LX400 + ; Lumen dynamics, Japan) at an irradiation intensity of 1000 mW/cm 2 for 1 min. The reacted solution was diluted 2-fold with ultrapure water, 10% sodium chloride (NaCl) (Wako, Japan), 10% albumin solution (Wako, Japan), or 10-fold Mueller Hinton Broth (MHB) (BBL Brain Heart Infusion, BD, USA) solution to stop the bactericidal effect of the remnant silver ions. The MHB solution was prepared as follows: 370 g of BHI powder was dissolved into 1 L of ultrapure water and autoclaved for 15 min at 121 °C and then kept for 1 h at room temperature. These groups were designated as "Ag(+)L(+) UW, " "Ag(+)L(+) 10% NaCl, " "Ag(+)L(+) 10% albumin, " and "Ag(+)L(+) MHB, " respectively. As a control, the bacterial suspension was mixed with ultrapure water and incubated in a black box for 1 min and diluted 2-fold with ultrapure water and designated as "Ag(−)L(−) UW. " Next, the treated solution was diluted 10-10 5 -fold with saline, after which 10 μL of the diluted solution was seeded on BHI agar (Oxoid, Basingstoke, UK) for determination of the treatment effect. The agar plates were anaerobically cultured at 37 °C for 48 h followed by colony enumeration for the determination of CFU/mL. Bacterial suspension. In this step, 100 μL of the prepared bacterial suspension was mixed with 100 μL of different concentrations of silver nitrate solution (100, 200, 400, 800, 1200, 1600, and 2000 μM) or ultrapure water in a 96-well cell culture plate. The final concentrations of silver nitrate solutions were 50, 100, 200, 400, 600, 800, and 1000 μM (8.5, 17.0, 33.97, 67.9, 101.9, 135.9, and 169.8 ppm, respectively). The final concentration of the bacterial suspension was 10 7 CFU/mL. After mixing, the sample was irradiated with UV-A light (365 nm) (Omnicure LX400 + ; Lumen Dynamics, Japan) at an irradiation intensity of 1000 mW/cm 2 or kept in a light-shielding box for 1 min. The samples were classified into 4 groups depending on the treatment method as follows; Ag(+)L(+): mixing silver nitrate solution and bacterial suspension followed by UV-A light irradiation, Ag(+)L(−): mixing silver nitrate solution and bacterial suspension followed by incubation in a light-shielding box, Ag(−)L(+): mixing ultrapure water and bacterial suspension followed by UV-A light irradiation, Ag(−)L(−): mixing ultrapure water and bacterial suspension followed by incubation in a light-shielding box. The reacted solution was diluted 2-fold with 10-fold MHB solution. CFU/mL was finally evaluated by colony enumeration after 48 h as described above. The initial bacterial count (inoculum size) was also evaluated.
To determine the mechanism underlying the bactericidal effect of the combination treatment with silver nitrate and UV-A irradiation, 100 μL of the prepared bacterial suspension was mixed with 100 μL of silver nitrate solution (final concentration; 400 μM [33.97 ppm]) with 1.4 M Dimethyl sulfoxide (DMSO) as a reactive oxygen species scavenger or ultrapure water in a 96-well cell culture plate. After mixing, the sample was irradiated with UV-A light at 1000 mW/cm 2 irradiation intensity or kept in the light-shielding box for 1 min. Next, the reacted solution was subjected to the same protocol. All tests were performed as 3 independent assays.
Biofilm on the titanium surface. The biofilm on the titanium surface was prepared as above described. The culture solution of the prepared biofilm was removed, and the titanium was washed with PBS twice. After removing PBS, the titanium plate was cleaned with an ultrasound scaler plastic tip (Varios V-P10; Nakanishi, Japan) using an ultrasound scaler device (Varios 970; Nakanishi, Japan) for 1 min. The titanium plate was then immersed in 500 μL of 1000 μM (169.8 ppm) silver nitrate solution and irrigated with UV-A light at an irradiance of 1000 mW/ cm 2 or kept in a light-shielding box for 1 min. These groups were designated as "Ag(+)L(+)" or "Ag(−)L(+), " respectively. The treated titanium plate was washed with PBS twice and incubated in 300 µL of a mixture of type I collagenase (4 mg/mL; Thermo Fisher Scientific, Japan) and Disperse (2 mg/mL; Thermo Fisher Scientific, Japan) for 2 h under shaking culture at 37 °C to extract the remaining bacteria from the titanium surface. Subsequently, 250 µL of the reacted solution was seeded onto BHI agar and cultured at 37 °C for 48 h, followed by colony enumeration for determination of the CFU/mL value. As a control group, the infected titanium was immersed into ultrapure water for 1 min and subjected to the same procedure. This group was designated as "Ag(−)L(−). " The extracted solution was diluted 10-fold with sterile saline, and then, 10 μL of the diluted solution was seeded onto BHI agar. www.nature.com/scientificreports www.nature.com/scientificreports/ Scanning electron microscopic observation. The titanium disks and the biofilm were prepared as described above. The prepared biofilm on the titanium desk was subjected to ultrasound scaling with an ultrasound scaler plastic tip for 1 min, which was followed by immersion into 1 mM silver nitrate solution with either irradiation with UV-A light an irradiance of 1000 mW/cm 2 or incubation in a light-shielding box for 1 min; these groups were designated as "Ag(+)L(+)" and"Ag(+)L(−), " respectively. As a control group, the prepared biofilm on the titanium desk was immersed into 1 mM silver nitrate solution and irradiated with UV-A light for 1 min and designated as an "Ag(+)L(+) without scaling." To observe the titanium surface, the treated titanium disk was examined using a scanning electron microscopy (SEM; SU-5000; Hitachi, Japan) at 10 Kv following carbon-palladium alloy sputtering.
Analysis of hydroxyl radical generation. Bacterial suspension. The yield of hydroxyl radicals generated by the combination treatment with silver ion application and UV-A light irradiation was analysed by an electron spin resonance (ESR) spin-trapping technique following a previously published study 58  Bacterial cellular fraction. The bacterial cellular fraction was prepared following the report by Kumode et al. 59 . Briefly, bacterial suspensions of S. mutans or A. actinomycetemcomitans were incubated in BHI broth or BHI-YE at 37 °C for 24 h. The prepared bacterial suspension was centrifuged at 13,000 × g for 10 min at 4 °C. The supernatant was removed, and the precipitate was re-suspended in protoplast buffer (50 mM Tris-HCl, 5 mM EDTA, 5 mM NaCl, and 25% sucrose at pH 7.5) followed by addition of lysozyme (final concentration, 1 mg/mL) and incubated at 37 °C for 16 h. The obtained bacterial suspension was centrifuged at 13,000 × g for 10 min at 4 °C, and the supernatant was collected as the cell wall component of bacteria. The remaining precipitate was re-suspended in ultrapure water and centrifuged at 13,000 × g for 10 min at 4 °C, and the supernatant was collected as the bacterial cell membrane component while the precipitate was collected as the cytoplasmic component. All obtained components were lyophilised and resuspended in ultrapure water. A mixture of 75 μL of each cellular fraction (10 µg/mL), 75 µL of 2 mM (339.6 ppm) silver nitrate solution, and 50 µL of 75 mM DMPO was irradiated with UV-A light (irradiation intensity: 1000 mW/cm 2 ) for 1 min. The ESR spectrum of each prepared suspension was analysed.

Biocompatibility. Cellular cytotoxicity of the combination of silver application and UV-A light irradia-
tion. The MC3T3-E1 mouse osteoblastic cell line provided by RIKEN Cell Bank (Tsukuba, Japan) was used. Cells were cultured in a cell culture flask using Dulbecco's (D) modified Eagle's medium (α-MEM; Nakalai Tesque, Kyoto, Japan) supplemented with 10% foetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, MA, USA), 100 U/mL penicillin, and 0.1 mg/mL streptomycin (1% P/S; Wako Pure Chemicals Industries, Osaka, Japan) at 37 °C under a 5% CO 2 atmosphere. The MC3T3E1 cells were trypsinised and seeded on the treated titanium disk in 96-well plates at a density of 5 × 10 3 cells per well. The plates were incubated overnight at 37 °C under a 5% CO 2 atmosphere. The culture medium was removed and washed with PBS. Next, 100 μL of each silver nitrate solution with a different concentration (100, 200, 400, 800, 1200, 1600, and 2000 μM) was added into the wells and irradiated with UV-A light for 1 min. The final concentrations of silver nitrate solutions were 50, 100, 200, 400, 600, 800, and 1000 μM. The reaction solution was then removed and 100 μL of fresh D-Medium was added, and the mixture was incubated for 3 h at 37 °C under a 5% CO 2 atmosphere. Cell viability was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT; Sigma, Japan) assay according to our previously reported protocol 60 . The culture medium was replaced with MTT solution (final concentration 1 mg/ mL), followed by incubation for 1 h at 37 °C under a 5% CO 2 atmosphere. The MTT solution was replaced with DMSO, followed by incubation for 30 min at room temperature. The absorbance of the reacted solution at 570 nm was measured using a microplate reader (Spectra MAX 190; Molecular Devices, Japan). As a control, untreated MC3T3E1 cells at a density of 5×10 3 cells per well were immersed into ultrapure water in a light-shielding box for 1 minute. The ultrapure water was removed and 100 μL of fresh D-Medium was added, and the cells were incubated for 3 h at 37 °C under a 5% CO 2 atmosphere. The absorption of the cells was normalised to that of the control. www.nature.com/scientificreports www.nature.com/scientificreports/ Cell proliferation assay and cell calcification assay. The titanium disks and the biofilm were prepared using A. actinomycetemcomitans as above described. The infected titanium desk was subjected to ultrasound scaling for 1 min, followed by immersion into 2 mM (339.6 ppm) silver nitrate solution with either irradiation with UV-A light at 1000 mW/cm 2 or incubation in a light-shielding box for 1 min. The treated titanium samples were washed with PBS and designated as the "Ag(+)L(+)" and "Ag(+)L(−)" groups, respectively. For the negative control group, the titanium desk was cleaned with ultrasound scaling for 1 min, immersed into ultrapure water, incubated in a light-shielding box for 1 min, and named "Ag(−)L(−). " For the positive control group, the non-infected titanium desk treated without application of silver nitrate and UV-A irradiation was prepared and named the control group.
The MC3T3E1 cells were trypsinised and seeded on the treated titanium disk in 48-well plates at a density of 3 × 10 4 cells per well. The plates were incubated at 37 °C under a 5% CO 2 atmosphere for 3 days for cell proliferation analysis or 28 days for cell calcification analysis. Cell proliferation was determined by the MTT assay. Cell calcification was determined by Alizarin red staining. One gram of Alizarin red S (Nacalai tesque, Japan) and 0.1 mL of 28% ammonium solution (Nacalai tesque, Japan) were diluted with 100 mL of ultrapure water. Both prepared solutions were mixed, and the pH was adjusted to 6.4. The cells were washed with PBS and fixed with methanol for 20 min. After removing the methanol, the cells were sufficiently washed with PBS, followed by addition of 250 μL of the prepared Alizarin red solution in room temperature for 30 min. After removing the Alizarin red solution, the cells were sufficiently washed with PBS and 250 μL of 5% formic acid at room temperature was added, followed by shaking for 10 min. Subsequently, 100 μL aliquots were taken for spectrophotometric analysis with an absorbance microplate reader at 415 nm.
In vivo removal torque test. A titanium screw (diameter, 1.6 mm: length, 6.12 mm: Fukuoka Seimitu, Japan) was sandblasted, acid-etched, and autoclaved as described above. After the formation of the biofilm of A. actinomycetemcomitans on the titanium screw, the infected titanium screw was subjected to ultrasound scaling for 1 min, which was followed by immersion into 2 mM (339.6 ppm) silver nitrate solution. Subsequently, the mixed solution was irradiated with UV-A light at an irradiance of 1000 mW/cm 2 . In the control group, after the titanium screw was sandblasted, acid-etched and autoclaved, the screw was stored in ultrapure water without any bacterial infection.
Wistar rats (8 weeks old) were used in this experiment in accordance with the guidelines for the care and use of laboratory animals of Tohoku University. The rats were anaesthetised with an intraperitoneal injection of medetomidine (Domitor; 0.375 mg·kg −1 bodyweight; Nippon Zenyaku Kogyo, Japan), midazolam (Sandoz; 2 mg·kg −1 bodyweight; Sandoz, Japan), and butorphanol tartrate (Vetorphale, 2.5 mg·kg −1 bodyweight; Meiji Seika Co., Japan). After clipping the leg of the rat, the skin surrounding the tibia was cut and the muscle was peeled. After exposing the tibia, the access hole (diameter: 1 mm) was prepared using VIVA ace (GC, Japan). The prepared titanium screw was implanted into the tibia of the rats with an average torque of 1.1 ± 0.2 cN·m. After implantation, the skin was rigidly sutured to prevent infection. The removal torque test of the implanted titanium screw was performed using Dejirache (No. GLK060, KTC, Japan) 28 days after surgery. All animal experimental protocols were reviewed and approved by the Institutional Animal Experiment Committee of Tohoku University (Approval number: 2018DnA-044) before beginning the animal experiments.
Histological observation. The tissue surrounding the implants was extracted and fixed using 4% glutaraldehyde. The samples were embedded into methylacrylate resin, cut into 5-7 μm sections after standardised dehydration and infiltration procedures, and stained with haematoxylin and eosin (H-E). The prepared sections were observed under a light microscope.
Statistical analyses. All data are presented as means ± standard deviation (SD). Normal distribution of the data was verified using the Shapiro-Wilk test. Since data were normally distributed, statistical significance (p < 0.05) of each parameter for the surface characterisation of titanium specimens and the cell proliferation assay was assessed by either the Student's t-test for pairwise comparisons or one-way analysis of variance (ANOVA), followed by the Tukey-Kramer honestly significant difference test for multiple comparisons using SPSS (IBM, USA).