Cytotoxic effects of octenidine mouth rinse on human fibroblasts and epithelial cells - an in vitro study.

Abstract Objectives: This study compared the cytotoxicity of a new octenidine mouth rinse (MR) against gingival fibroblasts and epithelial cells with different established MRs. Methods: The following MRs were used: Octenidol (OCT), Chlorhexidine 0.2% (CHX), Listerine (LIS), Meridol (MER), Betaisodona (BET); and control (medium only). Human primary gingiva fibroblasts and human primary nasal epithelial cells were cultivated in cell-specific media (2 × 105 cells/ml) and treated with MR for 1, 5, and 15 min. Each test was performed 12 times. Metabolism activity was measured using a cytotoxicity assay. A cellometer analyzed cell viability, cell number, and cell diameter. The data were analyzed by two-way analysis of variance with subsequent Dunnett’s test and additional t-tests. Results: The cytotoxic effects of all MRs on fibroblasts and epithelial cells compared to the control depended on the contact time (p < 0.001). OCT and BET showed less influence on cell metabolism in fibroblasts than other MRs. OCT also demonstrated comparable but not significant results in epithelial cells (p > 0.005). Cell numbers of both cell types at all contact times revealed that OCT showed a less negative effect (p > 0.005), especially for epithelial cells compared to CHX after 15 min (p < 0.005). OCT and BET showed the best results for viability in fibroblasts (p > 0.005), but MER showed less influence than OCT in epithelial cells (p < 0.005). Conclusions: OCT is a potential alternative to CHX regarding cytotoxicity because of its lower cell-toxic effect against fibroblasts and epithelial cells.


Introduction
Antimicrobial mouth rinses (MRs) help support oral biofilm management. MRs contain antimicrobial agents, the so-called antiseptics, which are expected to work within the oral cavity in a targeted and highly effective manner without causing side effects (McDonnell & Russell, 1999;Müller & Kramer, 2008;Schulman, 1993). Antiseptics are anti-infective substances that destroy (i.e. bactericidal effect) or inhibit (i.e. bacteriostatic effect) microorganism growth after topical application (McDonnell & Russell, 1999). Generally, MRs support mechanical plaque control within the oral biofilm (Fine et al., 2007). MRs are applied to prevent or slow biofilm formation and the multiplication of pathogenic bacteria.
In addition to the bacteriostatic, bactericidal, fungistatic, and/or fungicidal activities of MRs, a specific effect against pathogenic bacteria is intended but not necessarily achieved (Slots, 2002). MR use clearly reduces germs, but the substantivity of the MR is also important (Cummins & Creeth, 1992). Substantivity is meant to be the residence time and activity of a substance in an antimicrobiotically effective concentration. However, MRs must be toxicologically harmless even during long-term therapy and not prone to trigger allergies or absorption via oral mucous membranes or the gastrointestinal tract. The cytotoxic effects of the MR must also be considered. This aspect is vitally important to avoid the inhibition of healing processes and minimize toxic effects to host tissue.
Known MR alternatives also have cytotoxic effects, but these alternatives are less effective against bacteria. In dentistry several MRs are established using different active ingredients. One group of active MR ingredients is composed of essential oils, e.g. Listerine (LIS). The substantivity of this group is considerably lower than CHX (Otten et al., 2010). Alcohol is most commonly used as a solubilizer between the essential oils and water (21.6-26.9 Vol%) in this group. Alcohol is likely responsible for the cytotoxic effects on gingival fibroblasts (Eick et al., 2011) and stem cells (Park et al., 2014) that were described for LIS. Another group of active MR ingredients are based on amine and stannous fluorides, e.g. Meridol (MER). The organic amine fluoride stabilizes the antimicrobial stannous fluoride (Arweiler et al., 2001;Netuschil et al., 1995). Eick et al. demonstrated pronounced cytotoxic effects of MER on gingival fibroblasts, which was similar to CHX and LIS (Eick et al., 2011). Another example of oral antiseptics is povidone iodine (PVPiodine), a water-soluble complex of iodine and polyvinylpyrrolidone (trade name Betaisodona: BET). Povidone iodine is a compound from the group of iodophores. Cytotoxic effects on epithelial cells were confirmed for BET despite a low concentration of active ingredients, which was significantly lower than the clinically applied concentration (Sato et al., 2014).
Studies to develop a MR with an efficacy that is comparable to CHX but with better oral biocompatibility are ongoing. This research led to the identification of octenidine dihydrochloride, trade name Octenidol (OCT). The microbiostatic and microbiocidal effectiveness of OCT is 10 times higher than CHX (Sedlock & Bailey, 1985), and this product also shows better biocompatibility, which characterizes the cellular and bacterial toxicities of topical antimicrobials (Müller & Kramer, 2008). Neither adverse effects nor allergic reactions are described in the literature. The high effectiveness of OCT supports its potential as an alternative to CHX. However, the available data for OCT are insufficient, with a particular lack of studies on cytotoxicity.
Therefore, our investigations compared the toxic effect of the new MR OCT on fibroblasts and epithelial cells to CHX and the other MRs, such as LIS, MER, and BET, which are well established in the field of dental medicine. Cytotoxicity, cell number, viability, and mean cell diameter were determined in an in vitro investigation after treatment of the cells with each of the tested MRs.
We hypothesized that OCT would exhibit a lower cytotoxic potential than CHX.

Materials and methods
The present investigation was an experimental, controlled, six-arm in vitro study of primary human cell to investigate the cytotoxicity of several antiseptic MRs.

Cells
Cryoconserved primary human gingival fibroblasts (HGFIBs; order number 1210412, Provitro GmbH, Berlin, Germany) and primary human nasal epithelial cells (HNEPCs; order number 1210711, Provitro GmbH, Berlin, Germany) were the basic materials for cell cultivation. These primary human cells were cultivated as described below in order to obtain the cells to be used in the experiments.
Freezing was performed in aliquots (1 Â 10 6 cells, respectively) under standardized conditions in Cryo-SFM-Medium (order number 2040102, Provitro GmbH, Berlin, Germany) in liquid nitrogen. The aliquots were thawed in a water bath at 37 C for the test series.
Immunohistological analyses revealed that HGFIBs were positive for the HGFIB-specific receptor CD90/Thy-1, and HNEPCs were positive for keratin. Cells were subjected to infection serology tests for bacteria, fungi, mycoplasms, HIV-DNA, and hepatitis B/C-DNA, and all cells were classified as negative.
These media were used during initial cell cultivation.

Cell cultivation
Cell cultivation was performed using standardized methods under sterile conditions based on the above-mentioned primary cells. Cell suspensions were added to 10 ml fibroblast growth medium or epithelial cell growth medium and centrifuged for 5 min at 250 Â g at room temperature. The cells were resuspended and transferred to cell culture bottles. HGFIBs were seeded at a concentration of 4000 cells/cm 2 , and HNEPCs were seeded at a concentration of 6000 cells/cm 2 . The cells were cultivated in a constant nitrogen environment of 5% at 37 C and saturated humidity in a CO 2 incubator. Culture medium was replaced by fresh culture medium every 2-3 d. Adherent growing cells were passaged after a 70% to 80% confluency of the culture flask surface was obtained. The passage of HGFIBs was performed using a sterile trypsin-EDTA solution (Passage Kit 2, order number 2040002, Provitro GmbH, Berlin, Germany), and these cells were incubated for 4-7 min at 37 C. HNEPCs were treated with dispase II (Passage Kit 3, order number 2040003, Provitro GmbH, Berlin, Germany) and incubated for 15 min under the same conditions. The removed cells were transferred to a 15-ml centrifuge tube and centrifuged for 5 min at 250 Â g at room temperature. The supernatant was discarded, and the remaining cell pellet was resuspended in 5 ml of culture medium for cell number quantification. HGFIBs from passage 6 and 7 and HNEPCs from passage 4 were used for experiments.
Only LIS contained ethanol (21.6%). BET, MER, and OCT did not contain any ethanol. The control groups (CTRL) were treated with phosphate-buffered saline (PBS) without the addition of antiseptics. Treated cells were resuspended after the determination of the cell number in fresh culture medium and transferred to 12-well plates according to the respective group and contact time. The cell concentration was adjusted to 2 Â 10 5 cells/ml culture medium per well. The medium was exchanged after every 24 h, and light microscopy observed crystal formation. The wells were treated with 1 ml of the respective oral antiseptic after 96 h (Table 1). All tests were performed in duplicate and repeated 12 times. Each solution remained on the cells for 1, 5, or 15 min. Active ingredient solutions were aspirated. The wells were rinsed with 1 ml PBS for 30 s and refilled with 1 ml of fresh culture medium. Cells in the same culture conditions were cultivated for another 24 h. Finally, the prepared plates were analyzed using an MTT test and a computer-assisted cell analysis (CellometerÔ Auto T4).

Cytotoxicity assay (MTT)
The MTT [3-(4.5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide] test (proliferation assay) was used to analyze the metabolic activity of HGFIBs and HNEPCs. The MTT was described in detail by Mosmann (1983). The MTT solution (250 ml per well) was added to the medium. The 12well plates were incubated for another 2 h under the same conditions. Enzymes of living cells reduced the yellowcolored tetrazolium salt to an insoluble blue-colored formazan product. Crystal formation was controlled visually with the help of a light microscope. Media with the MTT solution were aspirated carefully, and 1 ml of dimethylsulfoxide (DMSO) was added to every well to dissolve the crystals. The plates were incubated for another 15 min in the incubator. The optical density within the wells was measured using a spectrophotometer and an automated reader. A test wavelength of 555 nm and a reference wavelength of 690 nm were used for the HGFIBs. A test wavelength of 550 nm and a reference wavelength of 700 nm were used for the HNEPCs. Measurement results at a reference wavelength were subtracted from the test wavelength results (mean value calculation) for analysis purposes. The results of the control wells of every plate were also averaged and set as 100% because these cells were treated with pure PBS instead of the MRs to show their optimal growth. The results of the wells treated with oral antiseptics are shown as a percentage of the control wells.

Cellometer analysis
Determinations of the cell number, viability, and mean cell diameter were performed optically using the CellometerTM Auto T4 (Nexcelom Bioscience LLC, Lawrence, MA, USA). A software-supported application allowed the simultaneous determination of the number of living cells in relation to the total cell number (viability) stained with trypan blue. All cells from two associated wells were resuspended in 1 ml of the medium after centrifugation for cell counting purposes. This cell suspension (20 ml) was pipetted into a counting chamber, which was placed in the cellometer for automatic analysis. The dilution factor was set to 1, the cell type to HGFIB or HNEPC, and the cell diameter to 10-50 mm. Cell concentration was generated automatically, and the dilution factor was taken into account. The generated data (cell number, viability, and mean diameter) were transferred using Microsoft Excel Õ .

Statistical analysis
The influence of the different MR solutions, contact time, and the interaction effect of these two factors on the measurements were investigated separately according to cell types using two-way analysis of variance plus interaction term (p values are shown in Supplementary Table 1). In the case of a significant effect of the solution factor, each MR solution was compared to the control using Dunnett tests. Additionally, all MR solutions were compared to each other using paired ttests. The p values for the comparisons between the different MR for each of the examined parameters are shown the Supplementary Tables 2-5.
The significance level was set to a ¼ 5% for each test. A significant difference for the paired comparisons of the MR solutions was only assumed when the p value was below the Bonferroni-adjusted significance level of 0.05/10 ¼ 0.005. All testing procedures were conducted using the software SAS (version 9.1, SAS Institute), all figures were built using the software R (version 2.8, www.r-project.org). HGFIBs: The MR (CHX, MER, OCT, LIS, BET) showed a perceptible reduction in the metabolic activity of the HGFIBs depending on the contact time (p50.0001). There were no significant differences between groups in relation to the contact time (p40.005) ( Table 2). The test series for OCT showed the lowest cytotoxic effect after 1 min. Comparable results were obtained for BET. The lowest cytotoxicity of the MR on HGFIBs was ranked as follows: OCT5BET5CHX ¼ MER5LIS. The rankings for the 5min contact time corresponded to the contact time of 1 min, except for BET and OCT: BET5OCT5CHX ¼ MER ¼ LIS. There was a change in the ranking according to the lowest cytotoxicity after a contact time of 15 min: BET5OCT5LIS ¼ MER5CHX. The highest cytotoxicity for CHX was observed at the longest contact time. Overall, the metabolic activity was the highest level after OCT and BET treatment (Table 2).  DOI: 10.3109/01480545.2015.1121274 HNEPC: A reduction in metabolic activity depended on the contact time of the MR solution (p50.0001). Overall, only small differences were observed between groups in relation to the contact time (Table 2). Except for the differences between CHX versus MER and CHX versus LIS (p ¼ 0.0000) for the 15 min (p ¼ 0.0000 and 0.0008, respectively) and 5-min contact times (p ¼ 0.0000) and CHX versus OCT for the 5-min contact time (p ¼ 0.0000) no statistical significances were found (p40.005). OCT exhibited the highest cytotoxicity in the HNEPCs within the test series with  a 1-min and 5-min contact time. The lowest cytotoxicity rankings were as follows: CHX5BET ¼ LIS5OCT ¼ MER (after 1 min), CHX5BET5LIS ¼ MER5OCT (at the 5-min contact time) and at the longest contact time: CHX ¼ BET5OCT5LIS5MER. A slight increase in cell activity was observed in the OCT group after a contact time of 15 min (Table 2).

Cell number
The cell number was significantly reduced (p50.0001) after MR treatment (CHX, MER, LIS, BET, and OCT) in all test series (1, 5, and 15 min) compared to the control cells (overall mean ± SE for control HGFIB: 6.53 ± 0.18 [10 5 cells/ml]; HNEPC: 5.43 ± 0.11 [10 5 cells/ml]) ( Figures 1B and 2B). The cell counts of HGFIBs and HNEPCs were significantly reduced without any dependence on the contact time when CHX, MER, and LIS were used. The negative effect of antiseptic therapy with OCT and BET was less pronounced. More cells (HGFIB and HNEPC) were present in the OCT series. Variance analyses showed a significant influence of the MR used and the contact time. However, a significant interaction between the type of the solution and the contact time on cell number was not observed for HGFIBs (p ¼ 0.7337) or HNEPCs (p ¼ 0.4826). These results suggest that the solution effect was equally strong regardless of the contact time. HGFIB: All test series reduced the total cell number depending on the MR contact time ( Figure 1B). The most cells were counted in the OCT test series after 1 min. The ranking for 1 min contact time was as follows: OCT4BET4MER4CHX4LIS. Slightly reduced cell numbers were measured after a contact time of 5 min, and the ranking corresponded to the ranking after 1 min contact time: OCT4BET4MER ¼ CHX4LIS. The most HGFIBs were counted in the BET group after 15 min. The following ranking was observed: BET4OCT4MER4LIS4CHX. A longer contact time with OCT negatively influenced the cell number. A significantly lower influence of OCT compared to CHX, MER, and LIS at the contact times of 1 and 5 min was observed in HGFIBs (p50.005). No significant differences were observed after 15 min and compared to BET (p40.005) ( Table 3).
HNEPC: All test series reduced the total cell number depending on the MR contact time ( Figure 2B). The following ranking for 1 min contact time was observed: OCT ¼ BET4CHX ¼ LIS4MER. The most cells were counted in the BET group after a contact time of 5 min: BET4OCT ¼ CHX4MER ¼ LIS. A further reduction of the number of epithelial cells was observed at the 15-min contact time: OCT4BET ¼ MER4LIS4CHX. The epithelial cell number showed no differences at the contact times of 1 min and 5 min between OCT and the other MRs (p40.005). A higher epithelial cell number was observed after 15 min of treatment with OCT compared to CHX (p50.005) ( Table 3).

Cell viability
Cell viability was significantly reduced (p50.0001) after MR treatment (CHX, MER, LIS, BET, and OCT) in all test series (1, 5, and 15 min) compared to the controls (overall mean ± SE for control HGFIB: 95.5 ± 0.3%; HNEPC: 96.2 ± 0.3%). The variance analyses showed a significant influence of the MR used and the contact time, and a significant interaction of the MR and the contact time on HGFIB and HNEPC viability (p50.001).
HGFIB: A reduction of viability depending on MR contact time was observed in all test series ( Figure 1C). The highest viability was measured in the OCT test series after a contact time of 1 min. The following ranking for 1 min contact time was observed: OCT4BET4MER ¼ CHX4LIS (Table 4). The ranking changed after a contact time of 5 min because of a significant reduction of viability in the MER test series: OCT4BET4CHX4LIS4MER (Table 4). The highest viability was observed in the BET group after 15 min. The following ranking was observed: BET4OCT4LIS4 CHX4MER (Table 4). The longer contact time of OCT negatively influenced the viability of the HGFIBs. A noticeably lower influence of OCT and BET on HGFIB viability was observed compared to CHX, MER, and LIS at the contact times of 1 and 5 min. However, this difference was not statistically significant (p40.005). HNEPC: All test series reduced the viability depending on the MR contact time, but the values after 1 and 5 min were quite similar ( Figure 2C). The following ranking for contact times of 1 and 5 min was observed: MER4BET4OCT4 LIS4CHX (Table 4). A further reduction of viability was observed at a contact time of 15 min. MER had the highest viability, and LIS had the lowest viability: MER4OCT4CHX4BET4LIS (Table 4). Epithelial cell viability was significantly lower for OCT than MER at all contact times, and it was significantly higher than LIS (p50.005). However, there were no significant differences compared to BET. The viability after treatment with OCT, BET, and MER was significantly higher than CHX after a contact time of 1 min (p50.005).
Variance analyses revealed a significant influence of the MR used and the contact time, and a significant interaction of the MR and the contact time on the diameter of the HGFIBs and HNEPCs (p50.001).
HGFIB: The reduction in cell diameter in all test series depended on the MR contact time ( Figure 1D). The largest diameter was measured in the OCT test series after 1 min of contact time. The smallest cell diameter was measured in CHX treatment. The following ranking for 1 min contact time was observed: OCT4BET4MER4CHX ¼ LIS (Table 5). The ranking changed slightly after a contact time of 5 min: OCT ¼ BET4LIS4MER4CHX (Table 5). The largest cell diameter was measured for the BET group after 15 min: BET4OCT4LIS4MER4CHX (Table 5). A significantly lower influence of OCT and BET was observed on HGFIB diameter compared to CHX, MER, and LIS at all contact times (p50.005). There were no significant differences between OCT and BET (p40.005).
HNEPC: All test series, except MER, showed a reduction of the cell diameter depending on the MR contact time ( Figure 2D). The following ranking for a contact time of 1 min was observed: CHX4MER4BET4OCT4LIS (Table 5). The following ranking was observed at a contact time 5 min: CHX4MER4OCT4BET4LIS. The cell diameter decreased for CHX at a contact time of 15 min, but it increased for MER. The following ranking was observed: MER4CHX4OCT4BET4LIS (Table 5). LIS showed the strongest influence on cell diameter compared to the other MRs at all contact times. MER showed a significantly lower influence on cell diameter compared to OCT, BET, and LIS at all contact times, and after 15 min compared to CHX (p50.005). CHX showed a lower influence on cell diameter than OCT, BET, and LIS at 1-and 5-min contact times (p50.005).

Discussion
This experimental, controlled, six-arm in vitro study of primary human cell lines investigated the cytotoxicity of various antiseptic MRs, with a particular focus on OCT. To the authors' knowledge, there is only one study that investigated the biocompatibility and related cytotoxicity of MRs (Müller & Kramer, 2008), and no other results are currently available. The present investigation differentiated and investigated several different parameters of cytotoxicity. Some of the parameters, such as the cell diameter, were not described previously in the literature. Overall, the available data on cytotoxicity are insufficient.
The specific effectiveness of MRs on oral pathogens was not the subject of this investigation. Therefore, the possibility for a holistic evaluation and determination of the biocompatibility index based on the data of this study is limited.
The cytotoxic effects of all MRs on HGFIBs and HNEPCs were determined compared to control cells. The cytotoxic effects depended on the MR contact time, which was established for each of the investigated parameters (cell metabolism, cell number, cell viability, and cell diameter). A stronger negative influence on the investigated parameters was observed with longer MR contact times.
Relevant differences in cell metabolism between the solutions were not observed for the HGFIBs or HNEPCs. However, OCT and BET had the lowest influence on HGFIB OCT and BET showed the best results on the cell viability of HGFIBs at all contact times (p40.005). However, a similar cytotoxic influence was observed for both of these MRs. MER showed significantly better results on the cell viability of HNEPCs than OCT (p50.005). OCT showed a slight influence on the cell diameter of HGFIBs and HNEPCs that depended on the concentration and time.
CHX is the gold standard of all MR ingredients. The most important advantage of CHX is the very high level of substantivity, which leads to a prolonged adherence of the antiseptic on hard and soft oral tissue. Therefore, the antiseptic is gradually released at an effective dose, which assures the persistence of its antimicrobial effect (Cousido et al., 2010). Nonetheless, negative effects were observed in many studies (Cabral & Fernandes, 2007;Eick et al., 2011;Park et al., 2014). Eick et al. demonstrated that commercially available CHX MRs have a very strong cytotoxic effect on the gingival fibroblasts in the MTT assay at different concentrations (0.01%, 0.06%, 1%, and 2%). The cells were exposed to the MR for 1 min and subsequently stored in a cell culture medium in this previous study (Eick et al., 2011). The results of the present investigation confirmed the cytotoxic influence of CHX and indicate a lower cytotoxic potential of OCT in the applied concentration compared to CHX. Significant differences in epithelial cell number reductions were observed between CHX and OCT. Therefore, the working hypothesis that OCT has a lower cytotoxic potential than CHX was partially confirmed. However, the differences between the different MR, investigated in this study, were minor.
The cytotoxic effect of OCT was often comparable to BET. Müller and Kramer found that OCT had the best biocompatibility index in HGFIBs in a comparison of several antiseptics, including OCT, CHX, and BET. The biocompatibility index helps provide objective evidence for the ratio between effectiveness and biological compatibility (Müller & Kramer, 2008), which is particularly important for antiseptics that are well tolerated and show low cytotoxic effects, such as BET. These antiseptics must be applied in significantly higher concentrations to achieve a sufficient antibacterial effect, and the cytotoxic effects are intensified. The cytotoxic effects of BET on fibroblasts and keratinocytes were described previously but not quantified (Sato et al., 2014). BET in MR is used in concentrations between 1.9 Â 10 3 and 3.7 Â 10 3 mM. Sato et al. used a significantly lower concentration in their study (10 À2 mM), but the contact time was much longer (1-2 d). The oral mucosa of rats showed epithelium damage at low concentrations of BET in their study (Sato et al., 2014). Furthermore, BET must be mixed in the applicable ratio because it is not commercially available as a ready-to-use solution. Therefore, the application is more complicated than OCT and other commercial MRs.
MER showed significantly better results than OCT on HNEPC viability (p50.005). MER has only been investigated previously in the framework of a clinical and histomorphometric animal study on rats to investigate the effect on epithelial cells. Excision wounds on the palate were treated with different antiseptics (1% CHX gel, LIS, MER), and the influence on the healing process was observed. None of the antimicrobial agents showed negative effects on the wound healing. The best results were achieved with a 1% CHX gel and LIS (Kozlovsky et al., 2007). Eick et al. observed a strong viability reduction when MER and LIS were applied to gingival fibroblasts in the MTT assay (Eick et al., 2011).

Conclusion
In summary, within the limitations of this study and the current literature, OCT can be recommended as an alternative to CHX because of its lower cytotoxic potential. However, the present study did not investigate the extent of a similar antiseptic effect. Therefore, further investigations are needed to verify whether OCT truly represents an alternative to CHX in clinical practice. It needs to be mentioned that the cytotoxic effect of OCT was often comparable to BET, which showed small cytotoxic effects.