ABSTRACT
The spread of antibiotic resistance and the challenges associated with antiseptics such as chlorhexidine have necessitated a search for new antibacterial agents against oral bacterial pathogens. As a result of failing traditional approaches, drug repurposing has emerged as a novel paradigm to find new antibacterial agents. In this study, we examined the effects of the FDA-approved anticancer agent toremifene against the oral bacteria Porphyromonas gingivalis and Streptococcus mutans. We found that the drug was able to inhibit the growth of both pathogens, as well as prevent biofilm formation, at concentrations ranging from 12.5 to 25 μM. Moreover, toremifene was shown to eradicate preformed biofilms at concentrations ranging from 25 to 50 μM. In addition, we found that toremifene prevents P. gingivalis and S. mutans biofilm formation on titanium surfaces. A time-kill study indicated that toremifene is bactericidal against S. mutans. Macromolecular synthesis assays revealed that treatment with toremifene does not cause preferential inhibition of DNA, RNA, or protein synthesis pathways, indicating membrane-damaging activity. Biophysical studies using fluorescent probes and fluorescence microscopy further confirmed the membrane-damaging mode of action. Taken together, our results suggest that the anticancer agent toremifene is a suitable candidate for further investigation for the development of new treatment strategies for oral bacterial infections.
INTRODUCTION
Oral infections are among the most common diseases worldwide (1). These infections are typically caused by biofilm-forming bacteria present on the surfaces of both hard and soft tissues (2). The Gram-negative anaerobic bacterium Porphyromonas gingivalis and the Gram-positive bacterium Streptococcus mutans are two important causative agents of oral infections. P. gingivalis is frequently involved in chronic inflammatory diseases, such as periodontitis and peri-implantitis, resulting in the destruction of soft and hard tissues surrounding teeth and dental implants, respectively (3, 4). S. mutans is known to be the main pathogenic agent of dental caries, a chronic disease characterized by irreversible destruction of the tooth (5).
Treatment of oral infectious diseases frequently involves the use of anti-infective agents, such as chlorhexidine, or, in severe cases, antibiotics (6, 7). However, the side effects associated with chlorhexidine, such as tooth staining, calculus formation, and change of taste sensation, and the development of resistance against antibiotics necessitate a search for alternatives (7, 8). Recently, drug repurposing has gained more attention as an alternative strategy to identify new antimicrobial agents. There are several advantages to repurposing old drugs with known safety and pharmacokinetic profiles over de novo drug discovery. Examples are reductions in time, cost, and risks associated with the development of novel antibiotics (9, 10). In an effort to repurpose existing drugs as antibacterial agents, we recently screened the NIH clinical library against P. gingivalis. Three compounds were selected that showed potent activity against P. gingivalis (toremifene, zafirlukast, and N-arachidonoylaminophenol [AM404]) (11). The antibacterial activity of toremifene (Fig. 1), an FDA-approved drug used in the treatment of breast cancer (12, 13), was further characterized in this study.
Structure of toremifene (pKa = 8.0).
The first aim of the study was to assess the antibacterial and antibiofilm activities of toremifene against the oral pathogens P. gingivalis and S. mutans. Furthermore, the effect of toremifene against oral biofilms formed on titanium, a material frequently used for implant applications, was evaluated. Finally, the antibacterial mode of action of toremifene was investigated. The findings from the study will provide valuable insight into the potential therapeutic application of toremifene for the treatment of oral infectious diseases.
RESULTS
Antibacterial and antibiofilm activities of toremifene against P. gingivalis and S. mutans.In a previous screening of a drug-repositioning library, the anticancer drug toremifene was identified as a new antibacterial compound that shows activity against P. gingivalis (11). To further evaluate the antibacterial potential of toremifene against oral bacteria, its activity against the prominent oral pathogens P. gingivalis and S. mutans was investigated using MIC, minimum bactericidal concentration (MBC), minimum biofilm-inhibitory concentration (MBIC), and minimum biofilm reduction concentration (MBRC) assays. Strikingly, as evidenced in Table 1, the activities of toremifene against planktonic and biofilm cultures are similar, underlining the antibacterial potential of the compound. In addition, we found that toremifene is active against biofilms grown under shaking conditions.
MIC, MBC, MBIC, and MBRC values of toremifene against oral pathogens
Activity of toremifene against P. gingivalis and S. mutans biofilms grown on titanium disks.Titanium has a high level of biocompatibility, making it a very suitable material for dental implants (14). Therefore, we tested if toremifene remains active against P. gingivalis and S. mutans biofilms grown on titanium disks. As shown in Fig. 2A and B, toremifene concentrations of 25 and 12.5 μM significantly reduced biofilm formation on titanium disks by P. gingivalis and S. mutans, respectively. In addition, the LIVE/DEAD bacterial-viability kit was used to visualize the viability of biofilms formed on titanium surfaces (Fig. 2C). The kit contains two dyes: SYTO 9, which stains live bacteria green, and propidium iodide (PI), which stains bacteria with compromised membranes. Lower numbers of viable green cells were detected on the titanium disks incubated in a concentration of 25 μM toremifene for P. gingivalis and in a concentration of 12.5 μM toremifene for S. mutans than on untreated disks, corroborating the results of the CFU counts.
(A and B) Reduction of P. gingivalis (A) and S. mutans (B) biofilm formation on titanium disks by toremifene. Shown is the percentage of biofilm formation in the presence of toremifene relative to the untreated control. The values are means and standard deviations (SD) of the results of three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 compared with the untreated control. (C) Fluorescence microscopy images of biofilms formed on titanium disks. Live cells stained green, and cells with compromised membranes stained red. The images were processed with an unsharp mask of Zen 2.0. Scale bars, 100 μm.
Time-kill assay of toremifene and chlorhexidine against S. mutans.To investigate the bactericidal activity of toremifene and to compare it with the activity of the commonly used antiseptic chlorhexidine, time-kill assays were performed (Fig. 3). For a number of practical reasons (e.g., sampling at different time points under anaerobic conditions), we decided to assess the killing kinetics of toremifene against S. mutans instead of P. gingivalis. A clear bactericidal effect was observed when S. mutans cells were incubated for 24 h at 1× and 4× the MIC of toremifene, as could be seen by a reduction in cell counts by 6.2 log10 CFU/ml and 6.5 log10 CFU/ml, respectively. Chlorhexidine exhibited much slower bactericidal activity, with a reduction in cell counts by only 4.8 log10 CFU/ml after 24 h of incubation with 4× the MIC. Regrowth was observed after 24 h of incubation with 1× the MIC of chlorhexidine.
Time-kill kinetics of toremifene against S. mutans. (A) Exponential-phase cells of S. mutans were treated with 1× the MIC and 4× the MIC of toremifene (TOR), with 1× the MIC and 4× the MIC of chlorhexidine (CHX), or with the solvents of the drugs (DMSO and water, respectively). Samples were taken at 0, 1, 2, 3, 4, 5, and 24 h, and the numbers of CFU per milliliter were determined. The data represent means ± SD from the results of 3 independent experiments. The dashed line indicates the lower limit of detection.
Single-step resistance selection.For an antibacterial agent to remain effective during treatment, emergence of resistance should be minimal. For this reason, we attempted to determine the frequency at which mutants resistant to toremifene appear. However, no spontaneous toremifene-resistant mutants of S. mutans could be generated (mutation frequency = <4.7 × 10−9). In contrast, rifampin-resistant mutants of S. mutans were obtained with an average mutation frequency of 1.24 × 10−8 ± 0.9 × 10−8. Similarly, no spontaneous toremifene-resistant mutants of P. gingivalis could be recovered (mutation frequency = <5.93 × 10−9).
Effect of toremifene on macromolecular synthesis pathways.The effects of toremifene on three macromolecular synthesis pathways (DNA, RNA, and protein synthesis) were tested by determining the incorporation of radiolabeled precursors into macromolecules after short exposure of S. mutans to 4× the MIC of toremifene (Fig. 4). The effect of toremifene on a specific macromolecular synthesis pathway was compared to the effect after treatment with 4× the MIC of a known inhibitor of the pathway (ciprofloxacin [DNA synthesis], rifampin [RNA synthesis], or tetracycline [protein synthesis]). In addition, negative controls were included in all the assays (tetracycline for DNA and RNA synthesis and ciprofloxacin for protein synthesis). Treatment with toremifene causes moderate inhibition of the incorporation of precursors into all the tested macromolecules and does not result in preferential inhibition. These results are typical for treatment of bacterial cells with a membrane-damaging agent (15–18). Indeed, treatment of the cells with the membrane-damaging antibacterial agent triclosan caused an effect similar to that of toremifene on the incorporation of precursors into macromolecules.
Percentage of incorporation of radiolabeled precursors into macromolecules after treatment of S. mutans with 4× the MIC of toremifene (TOR) or control antibacterials (ciprofloxacin [CIP], rifampin [RIF], tetracycline [TET], and triclosan [TRI]). The data represent the means from at least three independent replicates ± SD.
Effect of toremifene on membrane permeability.To investigate the membrane-damaging effects of toremifene on the outer membrane of P. gingivalis, the hydrophobic fluorescent probe N-phenyl-1-napthylamine (NPN) (Sigma, USA) was used. Normally, NPN cannot partition into the membrane due to the presence of lipopolysaccharides (LPS). However, when the outer membrane is damaged, NPN can enter the phospholipid layer, which results in increased fluorescence (19). As shown in Fig. 5A, treatment of the bacteria with increasing concentrations of toremifene resulted in increased uptake of NPN in the membrane. These results indicate that toremifene alters outer-membrane permeability. To determine the effect of toremifene on the inner membrane of P. gingivalis and the membrane of S. mutans, the nucleic acid stain Sytox green was used. The stain does not penetrate the inner membrane of bacteria. However, when the inner membrane is permeabilized, Sytox green can enter the cell and bind to nucleic acids, thereby emitting a strong fluorescent signal (20). As seen in Fig. 5B and C, Sytox green uptake was increased with increasing concentrations of toremifene, indicating that the compound is also capable of permeabilizing the inner membrane of P. gingivalis and the membrane of S. mutans.
Effect of toremifene on membrane permeability. (A) Outer-membrane permeabilization of P. gingivalis after treatment with different concentrations of toremifene, assessed by quantifying NPN uptake. Cells treated with 1× the MIC of triclosan (TRI) were used as a positive control (see Table S1 in the supplemental material). (B) Inner-membrane permeabilization of P. gingivalis after treatment with different concentrations of toremifene, determined by measuring Sytox green uptake. Melittin (MEL) (10 μg/ml) was used as a positive control. (C) Effects of increasing concentrations of toremifene on the membrane permeability of S. mutans, monitored by the uptake of Sytox green. Cells treated with melittin (2.5 μg/ml) served as a positive control. For all the panels, cells treated with ciprofloxacin (CIP) (1× the MIC) served as a negative control. The data represent the means from three independent replicates ± SD (**, P < 0.01; ***, P < 0.001). a.u., arbitrary units.
Binding of toremifene with LPS.Next, we examined the interaction between toremifene and LPS of P. gingivalis, using the Bodipy TR cadaverine (BC) displacement assay. Bodipy TR cadaverine is a fluorescent probe that strongly binds to the lipid A moiety of LPS. When a compound is added that interacts with LPS, Bodipy TR cadaverine is displaced from the complex, which results in increased fluorescence (21). A fast increase in fluorescent signal was observed after treatment with different concentrations of toremifene, suggesting that the compound binds with high affinity (Fig. 6).
Determination of the binding affinity of toremifene for LPS of P. gingivalis using BC. The concentration-dependent displacement of BC from LPS induced by toremifene is shown. Cells treated with 1× and 4× the MIC of chlorhexidine (CHX) were used as a positive control (see Table S1 in the supplemental material). Cells treated with 1× the MIC of ciprofloxacin (CIP) were used as a negative control (see Table S1). The data represent the means from three independent replicates and SD.
Microscopic visualization of membrane damage.To further examine the effect of toremifene on the membrane, fluorescence microscopy was employed using the membrane stain N-(3-triethylammoniumpropyl)-4-(p-diethylaminophenyl-hexatrienyl) pyridinium dibromide (FM 4-64; Molecular Probes). Treatment of the cells with a solvent control (dimethyl sulfoxide [DMSO]) resulted in intact, homogeneously stained membranes (Fig. 7). On the other hand, treatment of the cells with 4× the MIC of toremifene resulted in disrupted membranes (Fig. 7). Furthermore, the latter observations are comparable to those obtained after treatment of the cells with 4× the MIC of triclosan. This phenotype was not observed after treatment of cells with antibiotics with different modes of action (ciprofloxacin, rifampin, and tetracycline) (see Fig. S1 in the supplemental material).
Microscopic visualization of toremifene-induced membrane damage using the lipophilic dye FM4-64. Cells were treated with DMSO (solvent control) or with 4× the MIC of toremifene (TOR) or triclosan (TRI). Scale bars, 2 μm. The images were processed with the unsharp mask of Zen 2.0.
Hemolytic activity and cytotoxicity.Repurposing of existing drugs offers the advantage of known safety and pharmacokinetic profiles. However, for novel applications, the cytotoxicity of these compounds remains to be investigated. We assessed the hemolytic activity of toremifene against horse red blood cells (RBCs), as well as its potential cytotoxic effect on a human oral gingival epithelial cell line (HOC18). As shown in Fig. 8A, concentrations of toremifene as high as 100 μM did not cause hemolysis, indicative of good hemocompatibility. Conversely, exposure for 24 h to toremifene concentrations exceeding 25 μM was toxic to HOC18 cells (Fig. 8B).
Effect of toremifene on mammalian cells. (A) Dose response of the hemolytic activity of toremifene against red blood cells. Red blood cells were treated with different concentrations of toremifene, and its hemolytic activity was determined in comparison with Triton X-100 (100% hemolysis) and PBS (0% hemolysis). Tests were performed in quadruplicate, and the results are presented as means and SD. (B) Dose response of the cytotoxic activity of toremifene against HOC18 cells. Cytotoxicity was determined in comparison with Triton X-100 (positive control) and supplemented αMEM (0% cytotoxicity). Tests were performed in duplicate, and the results are presented as means and SD.
DISCUSSION
Known side effects of currently used antiseptics and the rising threat of antibiotic resistance demonstrate the need for the development of novel therapies to treat oral infections. In an attempt to identify new drugs with potent activity against the oral pathogen P. gingivalis, we recently performed a screen of a repurposing library (11). From this screen, toremifene was selected for further characterization. Toremifene is an FDA-approved anticancer drug used in the treatment of breast cancer (12, 13). The compound is known to bind with the estrogen receptor, thereby interfering with the estrogen-mediated growth stimuli of tumor cells (13). Earlier studies reported the potency of toremifene in other applications. Toremifene has been reported to have antibacterial activity against Francisella novicida, a model organism for the tularemia-causing pathogen Francisella tularensis; Staphylococcus aureus; Staphylococcus epidermidis; and Pseudomonas aeruginosa (22, 23). Furthermore, we and others have reported on the antifungal effects of toremifene (23–25). In addition, toremifene has antiviral activity against Ebola virus (26). However, to our knowledge, no data on its activity and mode of action against oral bacterial pathogens exist. Likewise, no extensive study on the antibacterial mode of action of toremifene exists.
We report here that toremifene displays potent activity against the prominent oral pathogens P. gingivalis and S. mutans, making it a potential candidate for use as a new antibacterial agent. Of note, studying bacterial killing kinetics revealed fast killing by toremifene compared to the antiseptic chlorhexidine, which is likely to have a positive effect on treatment outcome. Furthermore, we found that toremifene has a low tendency for selection of spontaneous resistant mutants, adding to its potential as a novel therapeutic.
Understanding the mode of action of toremifene is crucial for its development as a potential antibacterial agent. To get a first idea about its mode of action, a macromolecular synthesis assay was conducted. We were unable to perform this assay under the strict anaerobic conditions necessary to avoid physiological changes caused by oxidative stresses in P. gingivalis cells (27). Therefore, we performed the assay using S. mutans cells, for which we found that toremifene moderately inhibits the synthesis of all tested macromolecules. These data suggest that toremifene possibly acts by disrupting the integrity of the bacterial membrane (15–18). Subsequently, we validated the notion that toremifene rapidly permeabilizes the outer and inner membranes of P. gingivalis and the membrane of S. mutans. In addition, we showed that toremifene is able to interact with the LPS of the outer membrane of P. gingivalis, which further confirms direct interaction of toremifene with bacterial membranes. Finally, we microscopically visualized the changes in bacterial membrane integrity. Nonhomogeneously stained membranes were observed after treatment of both P. gingivalis and S. mutans with toremifene. Combined, these results indicate that membrane damage likely is the primary antibacterial mode of action of toremifene, which is in accordance with previous studies. Indeed, Dean and van Hoek (22) demonstrated that toremifene at a concentration of 5 μM strongly permeabilizes the membrane of the Gram-negative bacterium F. novicida. Furthermore, Delattin et al. (24) found that toremifene at a concentration of 12.5 μM induces membrane permeabilization in Candida albicans biofilm cells. However, further work is needed to identify the molecular mechanisms behind the observed membrane damage.
Thanks to their potentially rapid bactericidal effects, activity against both growing and dormant populations, and low potential for resistance development, membrane-acting agents are believed to be good candidates for treating biofilm-related persistent infections (28). Recently, the activity of toremifene against S. aureus biofilms formed under in vivo conditions has been described, further highlighting the potential of the compound to be used in treatment of biofilm-related bacterial infections (23). However, to evaluate the potential of toremifene for application against oral infections, additional experiments should be conducted using a relevant in vivo model (29). In addition, special attention should be paid to the fact that in nature, biofilms of multiple bacterial species often exist, underscoring the need for investigating the activity of toremifene against mixed-species biofilms formed on different surfaces (30).
Usually, in treatment of breast cancer, patients receive toremifene orally at a dose of 60 mg/day. Some clinical studies even mention the use of toremifene at a dose of 680 mg/day, which lies well within the range of recommended antibiotic dosages for treatment of oral infections (13, 31). Regarding toxicity, toremifene is generally well tolerated by patients (13). The commonest side effects include hot flushes, sweating, nausea, and vaginal discharge, and serious adverse events are rare (13). This is in accordance with our data showing good hemocompatibility and limited cytotoxicity. It should be noted that, compared to the toxicity assay conditions, shorter treatments (e.g., in the case of mouthwashes) are likely to be even less detrimental. These findings further pave the way to repurposing the compound for antibacterial therapeutic uses.
In conclusion, we demonstrated that the anticancer drug toremifene displays antibacterial activity against planktonic and biofilm cells of the prominent oral bacterial pathogens P. gingivalis and S. mutans. Moreover, we showed that toremifene effectively kills these bacteria in a rapid manner by damaging the bacterial membrane. Future experiments, including in vivo studies, will be necessary to fully reveal the potential of toremifene to be used in the treatment of oral bacterial infections.
MATERIALS AND METHODS
Bacterial strains and chemicals. P. gingivalis ATCC 33277 was routinely grown on 5% horse blood agar supplemented with hemin (5 μg/ml) and menadione (1 μg/ml) at 37°C under anaerobic conditions (90% N2, 5% H2, and 5% CO2) using an Anoxomat AN2OP system (Mart Microbiology, Drachten, the Netherlands). S. mutans ATCC 25175 was routinely grown on solid Trypticase soy broth (TSB) agar (Becton Dickinson Benelux) containing 1.5% agar at 37°C. Liquid cultures of all the strains were grown in TSB.
Toremifene was purchased from TCI Europe N.V., and stock solutions of 20 mM were prepared in DMSO.
Antibacterial assays.The MIC of toremifene was evaluated in TSB as described previously (32). To determine the MBC, 10-μl aliquots were taken from the wells of the MIC assay that did not show bacterial growth and were plated onto agar plates. After incubation of the plates, the MBC was determined as the lowest concentration of toremifene for which no CFU were observed.
Antibiofilm assays.The MBIC values of toremifene were determined using crystal violet staining. P. gingivalis biofilms were grown anaerobically on the polystyrene pegs of Nunc Immuno-TSP (VWR International) lids, as described previously with minor modifications (33). Overnight cultures of P. gingivalis were diluted 1/10 in TSB. Next, 2-fold serial dilutions of toremifene in cell suspension (0 to 200 μM) were prepared at a volume of 150 μl in the polystyrene microtiter plates of the Nunc device. Subsequently, the plates were covered with a lid containing the pegs, and biofilms were allowed to grow on the pegs for 72 h at 37°C without shaking. After incubation, the pegs were washed once with phosphate-buffered saline (PBS), stained with 200 μl 0.1% (wt/vol) crystal violet in an isopropanol-methanol-PBS solution (1/1/18 [vol/vol]) for 1 h, washed with water to remove excess stain, and air dried (0.5 h). Next, the remaining crystal violet stain was removed from the pegs in 200 μl acetic acid (30%), and the intensity was measured by determining the optical density at 570 nm (OD570), using a Synergy MX multimode reader (Biotek, Winooski, VT).
S. mutans biofilms were grown on the bottoms of the wells of polystyrene microtiter plates, as they failed to grow on pegs. To this end, overnight cultures were diluted 1/200 in brain heart infusion (BHI) medium (Becton Dickinson Benelux) supplemented with 3% sucrose, and 2-fold serial dilutions (150 μl) of toremifene in the cell suspensions (0 to 200 μM) were prepared in the microtiter plate. After 24 h of biofilm formation at 37°C, the biofilm formation was assessed by crystal violet staining as described above. The lowest concentration of toremifene required to inhibit biofilm formation was defined as the MBIC.
In addition, the biofilm-inhibitory effects of toremifene against S. mutans and P. gingivalis were tested under shaking conditions. Biofilms were grown and quantified as described above, with the difference that the biofilms were grown in a shaking incubator.
To determine the effect of toremifene on preformed biofilms, 72-h-old (P. gingivalis) or 24-h-old (S. mutans) biofilms were grown on polystyrene surfaces as described above. Subsequently, the biofilms were treated with 150 μl growth medium containing toremifene (0 to 200 μM) and incubated at 37°C for 24 h. Next, the biofilms were washed with PBS and quantified with cell titer blue (CTB; Promega Benelux) by adding 200 μl of CTB diluted 1/100 in PBS to each well. After 24 h of incubation in the dark at 37°C, fluorescence was measured (excitation wavelength [λex] = 535 nm; excitation wavelength [λex] = 590 nm) using the Synergy MX multimode reader (Biotek, Winooski, VT). The MBRC was defined as the lowest concentration of toremifene able to eradicate the preformed biofilm.
Inhibition of biofilm formation on titanium disks.To evaluate the biofilm-inhibitory activity of toremifene against P. gingivalis and S. mutans biofilms grown on titanium, round titanium disks (commercially pure titanium, grade 2; height, 2 mm; width, 0.5 cm) were used. Tests with P. gingivalis were performed under anaerobic conditions. First, bacterial suspensions were prepared by diluting overnight cultures of P. gingivalis 1/10 in TSB and those of S. mutans 1/200 in BHI medium supplemented with 3% sucrose. Next, the titanium disks were placed at the bottoms of the wells of a 96-well plate and challenged with 200 μl of a bacterial suspension containing 0 to 50 μM toremifene. After 72 h (P. gingivalis) or 24 h (S. mutans) of incubation at 37°C under static conditions, the disks were removed from the wells and subsequently washed with PBS to remove nonadherent bacteria and placed in centrifuge tubes containing 1 ml PBS. Adherent bacteria were removed from the disks by sonication (45,000 Hz in a water bath sonicator [VWR USC 300-T] for 10 min), followed by vortexing (1 min). Bacterial viability was quantified by serial dilution plating (CFU counts).
In addition, the BacLight LIVE/DEAD bacterial-viability staining kit (Molecular Probes, Invitrogen) was used to microscopically evaluate the viability of the biofilms formed on titanium disks. After incubation, the disks were washed with 1× PBS and transferred to a LIVE/DEAD staining solution containing SYTO 9 and PI (prepared according to the manufacturer's instructions). After 10 min of incubation at room temperature in the dark, the disks were washed again in 1× PBS and mounted on a coverslip for imaging. The stained biofilm cells were visualized under a Zeiss Axio imager Z1 fluorescence microscope equipped with an EC Plan-Neofluar 20× objective using the SYTO 9 (λex = 483 nm; λem = 500 nm) and PI (λex = 305 nm; λem = 617 nm) channels.
Time-kill assay.Exponential-phase cells of S. mutans were incubated with 1× and 4× the MIC of toremifene or chlorhexidine at 37°C under shaking conditions (see Table S1 in the supplemental material). At periodic intervals, aliquots taken from the samples were serially diluted in MgSO4 and subsequently plated on TSB agar. After incubation for 2 days as 37°C, cell viability was determined by CFU counting.
Single-step resistance selection.The frequency at which mutants of P. gingivalis and S. mutans that are resistant to antibacterial agents emerge was determined as described previously (34). Briefly, 500 μl of an overnight culture of P. gingivalis or S. mutans was plated on agar plates containing antibacterial agents at 5× the MIC (see Table S1 in the supplemental material). In parallel, the overnight cultures were serially diluted and plated on nonselective agar. After incubation of the plates for 7 (P. gingivalis) or 2 (S. mutans) days, the MICs of the antibacterial agents for the surviving colonies on selective agar were determined to verify resistance. The spontaneous-mutation frequency was calculated by dividing the number of surviving colonies on selective plates by the total number of colonies on nonselective plates after incubation.
Macromolecular synthesis assay.The effects of toremifene on the macromolecular synthesis pathways in S. mutans were determined by monitoring the incorporation of radiolabeled precursors of macromolecules. Briefly, S. mutans exponential-phase cells (OD595, 0.2 to 0.3) were incubated with radiolabeled precursors for DNA ([3H]thymidine [1 μCi]), RNA ([3H]uridine [2.5 μCi]), and proteins ([3H]leucine [2.5 μCi]). Next, the cells were treated with 4× the MIC of toremifene or control antibacterials (ciprofloxacin, rifampin, tetracycline, and triclosan) (see Table S1 in the supplemental material). After 10 min of incubation at 37°C, 100 μl was taken from the samples and resuspended in 3 ml ice-cold 10% trichloroacetic acid to stop the reactions and to release free radiolabeled precursors from the cells. Next, the samples were filtered through Whatman 25-mm GF/C glass microfiber filters and washed three times with 3 ml ice-cold water. Subsequently, the dried filters were transferred to scintillation vials containing 3.5 ml scintillation fluid. Each vial was counted in a liquid scintillation counter (Hidex 300 SL) for 2 min. The results are expressed as the percentage of incorporation compared to the untreated control.
Membrane permeabilization assays.The ability of toremifene to permeabilize the outer membrane of P. gingivalis was determined using the fluorescent dye NPN, as previously described (34) with some modifications. Briefly, exponential-phase cells were washed and resuspended to an OD595 of 0.1 in buffer (5 mM HEPES, pH 7.4). Next, toremifene (0 to 25 μM) and NPN (10 μM) were added, and changes in fluorescence were recorded after incubation for 5 min using a Synergy MX multimode reader (Biotek, Winooski, VT; λex = 350 nm, and λem = 420 nm). Triclosan at 1× the MIC was used as a positive control because of its strong outer-membrane-permeabilizing properties. Ciprofloxacin at 1× the MIC was used as a negative control (see Table S1 in the supplemental material).
The ability of toremifene to permeabilize the inner membrane of P. gingivalis and the membrane of S. mutans was determined using the fluorescent dye Sytox green (Invitrogen, USA), as previously described (34). Briefly, exponential-phase cells were washed and resuspended to an OD595 of 0.5 in PBS. Next, the cells were incubated with toremifene (0 to 25 μM) and Sytox green (1 μM) at 37°C for 15 min. Thereafter, the increase in fluorescence was measured using a Synergy MX multimode reader (Biotek, Winooski, VT; λex = 504 nm, and λem = 523 nm). Melittin (10 μg/ml for P. gingivalis; 2.5 μg/ml for S. mutans) was used as a positive control because of its strong inner-membrane-permeabilizing properties. Ciprofloxacin at 1× the MIC was used as a negative control (see Table S1 in the supplemental material).
The fluorescence values for each condition were divided by the respective OD595 values to correct for the cell density of the culture. In addition, the ratio was corrected for background fluorescence by subtracting the fluorescence values of untreated cells.
Bodipy TR cadaverine displacement assay.To determine the ability of toremifene to bind with the lipid A part of LPS, the fluorescent probe BC (Thermo Fisher Scientific, USA) was used. In this study, exponential-phase cells of P. gingivalis were washed and resuspended to an OD595 of 0.3 in PBS. Next, the cell suspensions were transferred to the wells of a black 96-well microtiter plate and mixed with 2.5 μM Bodipy TR cadaverine. After 2 h of incubation at 37°C, a 2-fold serial dilution of toremifene (0 to 50 μM) was added to the wells. Then, fluorescence was assessed for 30 min using a Synergy MX multimode reader (Biotek, Winooski, VT; λex = 580 nm, and λem = 620 nm). Cells treated with 1× the MIC of ciprofloxacin were used as a negative control (see Table S1 in the supplemental material). Cells treated with 1× and 4× the MIC of chlorhexidine were used as a positive control (see Table S1). BC displacement from LPS was calculated using the following formula: [(F − F0)/(Fmax − F0)] × 100, where Fmax is the fluorescence intensity of BC without cells, F0 is the intensity in the presence of cells alone, and F is the intensity of the mixture of cells and BC at various concentrations of toremifene, chlorhexidine, or ciprofloxacin.
Fluorescence microscopy.Exponential-phase cells of P. gingivalis and S. mutans were treated with 4× the MIC of toremifene or control antibacterials (ciprofloxacin, rifampin, tetracycline, and triclosan) (see Table S1 in the supplemental material). After 30 min of incubation at 37°C, the cells were centrifuged and stained with 10 μg/ml FM 4-64 (Molecular Probes) for 10 min at room temperature before being imaged. The cells were visualized using a Zeiss Axio imager Z1 fluorescence microscope equipped with an EC Plan-Neofluar 100× objective, using the FM 4-64 channel (λex = 506 nm; λem = 751 nm).
Hemolysis assay.The hemolysis assay was performed as described previously, with some modifications (35). Briefly, fresh horse RBCs were rinsed three times with PBS by centrifugation for 10 min at 800 × g and diluted in PBS to achieve a final RBC concentration of 4%. The resulting suspension was incubated at 37°C for 10 min under shaking conditions. Subsequently, 200 μl of the suspensions was transferred to the wells of a microtiter plate, and the assay was initiated by addition of different concentrations of toremifene to the suspensions. Controls included RBC suspensions treated with PBS and with Triton X-100 (1%) to provide references for 0% and 100% hemolysis, respectively. The resulting suspensions were incubated for 60 min at 37°C. Following centrifugation for 10 min at 800 × g, hemolysis was assessed by measuring the absorbance of the supernatant at 540 nm. The percentage of hemolysis was calculated relative to 100% hemolysis with Triton X-100.
Cytotoxicity assay.A cytotoxicity test of toremifene was performed on a cell type relevant to the oral cavity homeostasis, with the aim of screening for concentrations that did not inhibit cell growth or induce cell death. HOC18 cells, an immortalized human oral gingival epithelial cell line, were used (36). Cells were plated in 96-well plates at 15,000 cells/well in minimum essential medium with the Eagle-alpha modification (αMEM; Sigma, Bornem, Belgium) with 0.292 g/liter l-glutamine (G7513; Sigma, Bornem, Belgium) supplemented with 10% fetal bovine serum (PAA Laboratories GmbH, Pasching, Austria) and 1% antibiotic-antimycotic (Gibco 15240; Life Technologies SAS, Saint Aubin, France). The cells were maintained overnight at 37°C in a humidified environment with 5% CO2.
At day 1 postseeding, the cells were incubated with toremifene by adding the compound to the culture medium. A 2-fold serial dilution assay of toremifene was used, starting from 50 μM. Suspensions of the same cell line under the same conditions exposed to Triton X-100 (5%) or cultured without chemicals were used as controls. The proliferation of the HOC18 cells in the presence or absence of chemicals was investigated after 1 day of compound addition. Cell viability was monitored using the XTT assay according to the manufacturer's instructions (XTT cell proliferation kit II; Roche Diagnostics GmbH, Roche Applied Science, Penzberg, Germany). Briefly, this is a colorimetric assay, performed by adding XTT solution 4 h prior to the end of toremifene (or Triton X-100) exposure. In the assay, metabolically active cells cleave the yellow tetrazolium salt to form the orange formazan dye, whose absorbance is recorded at 450 nm and 650 nm (reference wavelength) using a spectrophotometer (Multiskan Ascent 96/384; Thermo Scientific, Waltham, MA, USA) associated with Ascent software version 2.6 (FIN-01621; Thermo Electron Corporation, Vantaa, Finland). The percentage of cytotoxicity was calculated relative to 100% cytotoxicity with Triton X-100.
Statistical analysis and reproducibility of results.Statistical analysis was performed by one-way analysis of variance (ANOVA), followed by Dunnett's multiple-comparison test. P values of <0.001, <0.01, and <0.05 were considered to be statistically significant. All the experiments were repeated at least three times.
ACKNOWLEDGMENTS
The research leading to these results received funding from the Industrial Research Fund of KU Leuven through the knowledge platform IOF/KP/11/007, Research Foundation-Flanders FWO (G.0471.12N, G0B2515N), and the Interuniversity Attraction Poles Program initiated by the Belgian Science Policy Office. K.T. acknowledges the receipt of a mandate from the Industrial Research Fund (IOFm/05/022, KU Leuven).
FOOTNOTES
- Received 24 August 2016.
- Returned for modification 6 September 2016.
- Accepted 11 December 2016.
- Accepted manuscript posted online 19 December 2016.
Supplemental material for this article may be found at https://doi.org/10.1128/AAC.01846-16 .
- Copyright © 2017 American Society for Microbiology.