Antimicrobial efficacy of alternative compounds for use in oral care toward biofilms from caries‐associated bacteria in vitro

Abstract For caries‐active patients, antimicrobial measures may be useful in addition to mechanical biofilm removal. The aim of this study was to investigate the antimicrobial efficacy of alternative compounds for use in oral care from two main categories (i.e., preservatives and natural compounds) toward biofilms from caries‐associated bacteria as compared to oral care gold‐standards chlorhexidine digluconate (CHX), cetylpyridinium chloride (CPC), and zinc. Compounds were screened in initial Streptococcus mutans biofilms. Then, the most effective compounds were further investigated in mature S. mutans and polymicrobial biofilms comprising Actinomyces naeslundii, Actinomyces odontolyticus, and S. mutans. Here, distinct treatment periods and concentrations were evaluated. Biofilms were visualized by scanning electron microscopy and bacterial membrane damage was evaluated by means of flow cytometry and staining with SYBR Green and propidium iodide. Citrus extract was the only compound exhibiting similar antimicrobial efficacy in initial S. mutans biofilms (>5 log10) as compared to CHX and CPC, but its effect was clearly inferior in mature S. mutans and polymicrobial biofilms. Flow cytometric data suggested that the mechanism of antimicrobial action of citrus extract may be based on damage of bacterial membranes similar to CHX and CPC. From all alternative compounds investigated in this study, citrus extract exhibited the highest antimicrobial efficacy toward in vitro biofilms from caries‐associated bacteria, but still was less effective than oral care gold‐standard antiseptics CHX and CPC. Nevertheless, citrus extract may be a valuable antimicrobial compound for use in oral care for caries‐active patients.

cross-sectional study in Norway, Oscarson et al. (2017) found that living in a rural area, low socioeconomic status, less frequent tooth cleaning and sugar containing soft drinks were associated with a higher prevalence of dental caries. Therefore, the development of more selective prevention concepts for these populations with strong caries incidence remains a major goal in contemporary preventive dentistry (Edelstein, 2006;Meyer et al., 2017;Oscarson et al., 2017). In these high-risk patients, supportive antimicrobial measures may be useful in daily oral hygiene practice in addition to mechanical removal of biofilms and use of anticariogenic agents, for example, fluoride (ten Cate, 2009).
In this regard, several antiseptics such as chlorhexidine (CHX) or cetylpyridinium chloride (CPC) have been proposed, either applied as oral rinses, gels or varnishes, or incorporated into toothpastes (Haps, Slot, Berchier, & van der Weijden, 2008;Marsh, 2012;Walsh, Oliveira-Neto, & Moore, 2015). However, the actual impact of such nonfluoride antimicrobials on caries prevention is still debated controversially (Walsh et al., 2015;. For example, long-term use of CHX, which can be regarded as gold-standard antiseptic in oral care, exhibits some undesirable side effects like yellowbrown staining of enamel surfaces and tongue and altering the sense of taste (Addy, Mahdavi, & Loyn, 1995;Autio-Gold, 2008). These side effects may also result in poor compliance by patients in daily oral care (Cortellini et al., 2008;Fernandez y Mostajo et al., 2014).
Furthermore, it has been recommended to limit the use of CHX to "those applications with a clear patient benefit" (i.e., intensive care) in order to reduce the risk of inducing acquired resistance toward CHX or even cross-resistances toward antibiotics (Kampf, 2016).
Accordingly, Kitagawa et al. (2016) recently reported development of enhanced tolerance toward CHX in Enterococcus faecalis as shown by increased minimum inhibitory concentrations (MICs) after 10 passages of treatment and regrowth. Likewise,  found increased MICs in E. faecalis, Streptococcus gordonii, Fusobacterium nucleatum, and Porphyromonas gingivalis after 10 suchlike passages.
Considering the wide use of CHX in dental practice and this risk of enhanced tolerance toward CHX in oral bacteria, it may be desirable to restrict its use and search for alternative antimicrobials for daily oral care with less strategic importance in general medicine (Kampf, 2016;).
Therefore, the aim of this study was to evaluate the antimicrobial efficacy of alternative compounds for use in oral care in vitro as compared to CHX, CPC, and Zinc, which all are antimicrobial substances frequently used in oral care and therefore served as positive controls here. Two different categories of compounds were tested, that is, (1) so-called "gentle" and multifunctional preservatives and (2) natural products and essential oils. After screening in initial, 24 hr old Streptococcus mutans biofilms, the most effective compounds were then further evaluated in mature S. mutans biofilms TA B L E 1 Overview of compounds used in this study and in polymicrobial biofilms cultured from Actinomyces naeslundii, Actinomyces odontolyticus, and S. mutans.

| Biofilm formation
Biofilms were formed in 96-well polystyrene culture plates (Corning ® Costar ® , Corning, NY, USA). For simulation of pellicle coating, wells were incubated with 200 μl saliva for 2 hr at room temperature. After that, saliva was discarded, and wells were filled with 200 μl of BCM containing S. mutans and incubated under anaerobic conditions. For initial S. mutans biofilms, biofilms were used for the antimicrobial assay after 24 hr. For mature S. mutans biofilms, medium was carefully exchanged after 24 and 48 hr and biofilms were used for the antimicrobial assay after a total culture period of 72 hr.
Polyspecies biofilms were cultured as it has been described in detail recently . Briefly, after discarding saliva, wells were filled with 200 μl of BCM containing A. naeslundii and A. odontolyticus and incubated under anaerobic conditions. After 16 hr, medium was carefully discarded and 200 μl fresh BCM containing S. mutans was added. After 48 hr, a further medium change was performed and biofilms were incubated for another 24 hr.

| Antimicrobial assay
After the respective culture periods of 24 or 72 hr, medium was carefully discarded from the wells and the biofilms were either incubated with phosphate-buffered saline (PBS; Biochrom, Berlin, Germany; untreated control) or with the respective test substances for 1, 3, or 10 min (50 μl each). Immediately afterward, PBS or the respective compound was carefully removed and each biofilm was brought to suspension with 200 μl of PBS and transferred to an Eppendorf tube.

| Scanning electron microscopy (SEM)
Polymicrobial biofilms were prepared on Permanox ® Chamber Slides were adjusted to ensure optimum viewing and images were taken from randomly selected fields on the slides.

| Flow cytometric evaluation of membrane damage
For flow cytometry, SYBR green and propidium iodide (PI; both: Sigma-Aldrich) were used as a fluorescent dyes to evaluate integrity of cytoplasmic membranes (Joux & Lebaron, 2000 Figure 1 shows the results of the initial first screening round.

| Screening for antimicrobial efficacy in initial (24 hr) S. mutans biofilms
Here, the antimicrobial efficacy of all compounds included in this study (Table 1) was investigated in initial (24 hr) S. mutans biofilms.
Concentrations were chosen as the highest that were still recommended for use in oral care according to the respective manufacturers. The incubation period was set to 10 min.
Citrus extract 0.25%, CHX 0.2%, and CPC 0.05% were the only compounds that showed a pronounced antimicrobial efficacy reducing CFU by ≥5 log 10 below the detection limit. The other eight compounds showed <3 log 10 steps reduction in CFU: Presol 1% and

| Antimicrobial efficacy of selected compounds in mature (72 hr) S. mutans biofilms
The three most effective compounds citrus extract, CHX, and CPC were further investigated in mature (72 hr) S. mutans biofilms, applied for different treatment periods (1 min, 3 min, or 10 min) as F I G U R E 1 Screening of all compounds for their antimicrobial efficacy in initial (24 hr) Streptococcus mutans biofilms. All results are depicted as medians, 1st, and 3rd quartiles from six independent experiments, each performed in duplicate, on a log 10scaled ordinate with untreated control group UC set to 100%. Horizontal dotted and dashed lines represent CFUreductions of 3 log 10 and 5 log 10 , respectively. Incubation period was 10 min for all compounds shown in Figure 2. Untreated biofilms (UC) showed growth of 5 × 10 7 to 2 × 10 8 S. mutans cells. All compounds showed an effect of treatment period on antimicrobial efficacy. For citrus extract 0.25%, CFU-reductions were 0.8 log 10 after 1 min of incubation, 2 log 10 after 3 min and 2.8 log 10 after 10 min. Both, CHX 0.2% and CPC 0.1%, were more effective reducing CFU of S. mutans by 3 log 10 after 1 min, by 4.6 log 10 (CHX) or 2.6 log 10 (CPC) after 3 min and by >5 log 10 after 10 min.

| Antimicrobial efficacy of selected compounds in polyspecies biofilms
The antimicrobial efficacy of citrus extract, CHX, and CPC was fur- All compounds showed concentration-dependent antimicrobial effects. Citrus extract (Figure 3a) exhibited ≤0.5 log 10 reduction in CFU at 0.05% and 0.1%. At 0.25%, reductions were 0.9 log 10 for S. mutans, 1.4 log 10 for A. naeslundii, and 1.5 log 10 for

| Visualization of polyspecies biofilms by scanning electron microscopy
Treatments had no impact on overall biofilm structure. In biofilms treated with citrus extract 0.25%, cellular debris was visible, most likely resulting from dead bacterial cells. In samples treated with CHX 0.2% and in particular in samples treated with CPC 0.1%, vesicle-like structures were apparent on the cell surfaces of bacteria.

| Flow cytometric evaluation of membrane damage in mature (72 hr) S. mutans biofilms
Flow cytometry with SYBR green and PI as fluorescent dyes was employed to evaluate bacterial membrane damage in mature (72 hr) S. mutans biofilms. These dyes intercalate into DNA and exhibit a strong increase in fluorescence upon nucleic acid binding. While SYBR green stains all bacteria, PI is only able to stain bacteria with damaged membranes (Joux & Lebaron, 2000;Tawakoli et al., 2013). Figure 5a shows the chosen region of interest on dot plots FSC versus SSC, which was confirmed by SYBR green staining to discern between bacteria and cellular debris (data not shown). Figure 5b shows exemplary dot plots for all tested groups. Figure 5c shows summarized median percentages of bacterial cells stained with either SYBR green, PI or both for all groups. Untreated biofilms (UC) showed 99% bacteria stained with SYBR green. Biofilms treated with citrus extract 0.25% exhibited 22% bacteria stained with PI and 8% doublestained bacteria, whereas biofilms treated with CHX showed 68% bacteria stained with PI and 12% double-stained bacteria. In contrast, treatment with CPC 0.1% led to 100% PI-positive bacteria.

| D ISCUSS I ON
The aim of this study was to investigate the antimicrobial efficacy of alternative compounds for use in oral care toward biofilms formed in vitro from caries-associated bacteria. For this purpose, 11 compounds from two main categories, that is, (1) so-called "gentle" and multifunctional preservatives and (2) natural compounds and essential oils, were evaluated and compared to contemporary gold-standards frequently employed in oral care CHX, CPC, and Zn 2+ .
As a first step, all compounds were tested for their antimicrobial properties toward initial biofilms of S. mutans that had been cultured for 24 hr. S. mutans is known as key organism in the pathogenesis of dental caries due to its acidogenic and aciduric properties and even more due to its ability to produce high amounts of extracellular polysaccharides, which is known to be a virulence determinant of F I G U R E 2 Antimicrobial efficacy of selected compounds citrus extract, CHX, and CPC in mature (72 hr) S. mutans biofilms. All results are depicted as medians, 1st, and 3rd quartiles from six independent experiments, each performed in duplicate, on a log 10 -scaled ordinate with untreated control group UC set to 100%. Horizontal dotted and dashed lines represent CFU-reductions of 3 log 10 and 5 log 10 , respectively. Incubation periods were 1, 3, and 10 min for all selected compounds cariogenic biofilms (Klein, Hwang, Santos, Campanella, & Koo, 2015;Koo, Falsetta, & Klein, 2013). Here, all compounds were evaluated in commonly used concentrations recommended by their respective manufacturers for use in oral care. Although Zn 2+ was included as a positive control known for its antibacterial action due to inhibition of bacterial adhesion and metabolic activity (Sheng, Nguyen, & Marquis, 2005), here it only exhibited small antimicrobial efficacy toward initial S. mutans biofilms (<2 log 10 ). This is in line with recent data from the literature where it was concluded that very high concentrations of Zn 2+ were needed to reach bactericidal effects, in particular toward biofilms (Gugala, Lemire, & Turner, 2017). However, besides this limited antimicrobial efficacy Zn 2+ may still be a valuable compound for use in oral care products due to its proven efficacy in control of halitosis and its anticalculus effects (Sanz, Serrano, Iniesta, Santa Cruz, & Herrera, 2013).
So-called "gentle" (i.e., nonirritating, paraben-free) and multifunctional preservatives like the ones included in this study are usually added to cosmetic products in order to prevent bacterial growth and extend their shelf life, but mainly to offer additional benefits, for example, for skin soothing and moistening.
Consequently, antimicrobial efficacy of such compounds is usually evaluated by testing minimum inhibitory concentrations (MICs) on planktonic bacteria or potential inhibitory effects on biofilm formation, whereas effects on already established biofilms are not in focus (Al-Ahmad et al., 2008;Lalitha & Rao, 2013). Here, from all tested preservatives only two compounds (Presol 1%, SymDiol 1%) reduced CFU in S. mutans biofilms by ≥2 log 10 steps, whereas all others were less effective (<2 log 10 ).
Although there are many studies on antimicrobial efficacy of natural compounds and essential oils toward planktonic bacteria, studies on their efficacy toward biofilms are scarce (Freires, Denny, Benso, de Alencar, & Rosalen, 2015;Karygianni et al., 2015). Here, all four essential oils tested exhibited <2 log 10 steps reduction in CFU only, when applied to initial S. mutans biofilms. Higher concentrations of the essential oils might have yielded superior antimicrobial efficacies but would be not applicable in oral care due to side effects (e.g., mucosal irritations or altering the sense of taste) and thus were not tested in this study.
From all compounds tested, only citrus extract 0.25% revealed a notable antimicrobial efficacy (>5 log 10 ) toward initial S. mutans biofilms similar to the positive controls CHX 0.2% and CPC 0.05%.
This compound is a commercially available product comprising extracts from various citrus fruits (i.e., citrus reticulata, citrus aurantium amara, and citrus aurantium sinensis), which has already been shown to have pronounced antimicrobial efficacy toward planktonic cultures of Vibrio vulnificus (Cormier, Scott, & Janes, 2013). In F I G U R E 3 Antimicrobial efficacy of selected compounds citrus extract, CHX, and CPC in polyspecies biofilms from Actinomyces naeslundii, Actinomyces odontolyticus, and S. mutans. All results are depicted as medians, 1st, and 3rd quartiles from six independent experiments, each performed in duplicate, on a log 10 -scaled ordinate. Left panels show untreated control groups UC as absolute CFU values. Right panels show relative CFU data (CFU [%]) with untreated control group UC set to 100% for each bacterial strain. Horizontal dotted and dashed lines represent CFU-reductions of 3 log 10 and 5 log 10 , respectively. The antimicrobial efficacy of CHX, CPC, and citrus extract was then further evaluated in a more complex polymicrobial biofilm model cultured from A. naeslundii, A. odontolyticus, and S. mutans, whose culture conditions have been described and extensively discussed recently . While biofilms grown in vitro from defined consortia can never mimic the vast microbial complexity found in the oral cavity (Dewhirst et al., 2010;Marsh & Zaura, 2017), this may not be a crucial point for testing antimicrobial efficacy of given compounds because it is well-known that the enhanced tolerance of bacteria in sessile biofilm mode is mostly due to the protective role of the extracellular polymeric substance (EPS), which acts as a diffusion barrier for antimicrobial agents (Bowen, Burne, Wu, & Koo, 2018;Stewart & Costerton, 2001).
Here, all compounds were tested in three different concentrations and showed concentration-dependent antimicrobial effects: At the highest respective concentrations tested, CHX 0.2% was the most effective compound (>5 log 10 reduction in CFU of all species), followed by CPC 0.1% (>5 log 10 of both Actinomyces spp., 4 log 10 of F I G U R E 4 Exemplary visualization by means of scanning electron microscopy (SEM). Exemplary visualization of randomly selected fields of polyspecies biofilms (72 h), either untreated or treated with citrus extract 0.25%, CHX 0.2%, or CPC 0.1% in 3,000-fold, 24,000-fold, and 50,000-fold magnification S. mutans). Citrus extract 0.25% led to reductions of 0.9-1.5 log 10 only. However, here it must be considered that citrus extract 0.25% was more effective than CHX and CPC at concentrations that are usually recommended for long-term use in oral care, for example, in toothpastes (0.06% or 0.02%, respectively). Therefore, this compound may still be a valuable antimicrobial ingredient in oral care products. Porphyromonas gingivalis showed no suchlike vesicles, but cellular debris only after treatment with CHX 0.2% for 20 min . While Vitkov et al. (2005) found no alterations on the biofilm surfaces in SEM, they revealed from transmission electron microscopic investigations that treatment of ex vivo biofilms with CHX 0.1% for 5 min led to membrane blebbing associated with cytoplasmic vesicle-like lesions, probably due to loss of cytoplasm through membrane perforations. Interestingly, here these vesicles were mostly found on Actinomyces cells, which may be due to their unique cell wall architecture, which makes them even resistant to treatment with lysozyme (Delisle, Barcak, & Guo, 2006). Therefore, this phenomenon may be strain dependent.
Based on these findings and for further investigation of bacterial membrane damage, flow cytometry was employed for analysis of mature S. mutans biofilms. The fluorescent stains SYBR green and PI were chosen, which both increase their fluorescence emission upon nucleic acid binding. While SYBR green enters all bacterial cells, PI is only able to enter cells with damaged membranes (Joux & Lebaron, 2000;Tawakoli et al., 2013). Both, CHX and CPC, revealed high percentages of PI-positive or double-stained bacterial cells, indicating membrane damage, which is in line with the SEM visualization.
Furthermore, in the literature the mechanism of antimicrobial action of CHX and CPC is described as damage of bacterial membranes and subsequent leakage of cytoplasmic constituents Jones, 1997;McDonnell & Russell, 1999). Citrus extract also led to an increase in PI-positive cells as compared to the untreated control group and therefore may also target on bacterial membranes, as it is known from other essential oils and flavonoids (Paparella et al., 2008).
The smaller increase as compared to CHX and CPC may be explained due to the smaller antimicrobial efficacy of citrus extract when applied for 10 min in these mature S. mutans biofilms (2.8 log 10 vs. >5 log 10 ). Exemplary dot plots for untreated control (UC) and groups treated with citrus extract 0.25%, CHX 0.2%, or CPC 0.1%, respectively. Bacterial cells with intact membranes are shown in Q4 (stained with SYBR green only), whereas bacterial cells with damaged membranes (stained with PI) are shown in Q1. Q2 shows bacterial cells with damaged membranes but intact DNA (stained with SYBR green and PI) and therefore may represent damaged cells. (c) Summarized median percentages, 1st, and 3rd quartiles of bacterial cells stained with SYBR green (green) or PI (red) and double-stained with SYBR green as well as PI are shown for untreated control (UC) and groups treated with citrus extract 0.25%, CHX 0.2%, or CPC 0.1%, respectively. Membrane damage could be observed after treatment with all compounds tested

| CON CLUS IONS
In this study, several alternative antimicrobial compounds for use in oral care were evaluated for their antimicrobial efficacy toward biofilms formed from caries-associated bacteria in vitro. As compared to the gold-standards CHX and CPC, citrus extract was the only compound to be similarly effective in initial S. mutans biofilms (> 5 log 10 ) but its antimicrobial effect was clearly inferior in mature S. mutans and polymicrobial biofilms. Nevertheless, citrus extract may still be a valuable natural alternative for use as an antimicrobial ingredient in oral care products. Thereby, its mechanism of antimicrobial action may be based on damage of bacterial membranes similar to CHX and CPC.

ACK N OWLED G M ENTS
Helga Ebensberger and Gerlinde Ferstl are gratefully acknowledged for their excellent technical support in SEM experiments.

This study was funded in part by an industrial research grant from
Dr. Kurt Wolff GmbH & Co. KG (Bielefeld, Germany) and one author (J. E.) is an employee of this company. The authors declare that there are no other conflicts of interest.

DATA ACCE SS I B I LIT Y S TATE M E NT
The data presented in this manuscript are available at request from the corresponding author's institution archives.