Activity of taurolidine on species associated with periodontitis CURRENT STATUS: POSTED

Background Taurolidine is thought to be an alternative antimicrobial in periodontal therapy. The purpose of this follow-up taurolidine study was to determine in more detail the mode of action of taurolidine against bacterial species being associated with periodontal disease. Further, a potential development of resistance against taurolidine in comparison with minocycline was to evaluate. Results Visualizing the mode of action of taurolidine to Porphyromonas gingivalis by scanning and transmission electron micrographs showed in part pores and release of constituents from the cell wall. The interaction of taurolidin with bacterial cell wall is also supported by the finding that taurolidine inhibited in a concentration-dependent manner the activities of LPS and of the P. gingivalis arginine-specific gingipains. However, an effect on A. actinomycetemcomitans leukotoxin was not found. When transferring 14 clinical isolates from subgingival biofilm samples (4 P. gingivalis, 2 A. actinomycetemcomitans, 2 Tannerella forsythia, 2 Fusobacterium nucleatum, 4 oral streptococci) on agar plates containing subinhibitory concentrations of taurolidine up to 50 passages, one P. gingivalis strain developed a resistance against taurolidine which was probably linked with efflux mechanisms. When antimicrobial pressure was removed, MIC reverted to baseline value. Testing development of resistance to minocycline in a similar way, an increase of MIC values occurred in five of the 14 included strains after exposure to subinhibitory concentrations of the antibiotic. Efflux might play a role in one A. actinomycetemcomitans strains, but obviously not in the other four strains. Removing antimicrobial pressure for a few passages did not revert the increased MIC values. Conclusion Taurolidine interacts with LPS and gingipains. Development of resistance seems to be a rare event when applying taurolidine. A potential development of resistance might be associated with efflux mechanisms.


Abstract
Background Taurolidine is thought to be an alternative antimicrobial in periodontal therapy. The purpose of this follow-up taurolidine study was to determine in more detail the mode of action of taurolidine against bacterial species being associated with periodontal disease. Further, a potential development of resistance against taurolidine in comparison with minocycline was to evaluate.
Results Visualizing the mode of action of taurolidine to Porphyromonas gingivalis by scanning and transmission electron micrographs showed in part pores and release of constituents from the cell wall.
The interaction of taurolidin with bacterial cell wall is also supported by the finding that taurolidine inhibited in a concentration-dependent manner the activities of LPS and of the P. gingivalis argininespecific gingipains. However, an effect on A. actinomycetemcomitans leukotoxin was not found. When transferring 14 clinical isolates from subgingival biofilm samples (4 P. gingivalis, 2 A. actinomycetemcomitans, 2 Tannerella forsythia, 2 Fusobacterium nucleatum, 4 oral streptococci) on agar plates containing subinhibitory concentrations of taurolidine up to 50 passages, one P. gingivalis strain developed a resistance against taurolidine which was probably linked with efflux mechanisms.
When antimicrobial pressure was removed, MIC reverted to baseline value. Testing development of resistance to minocycline in a similar way, an increase of MIC values occurred in five of the 14 included strains after exposure to subinhibitory concentrations of the antibiotic. Efflux might play a role in one A. actinomycetemcomitans strains, but obviously not in the other four strains. Removing antimicrobial pressure for a few passages did not revert the increased MIC values. Conclusion Taurolidine interacts with LPS and gingipains. Development of resistance seems to be a rare event when applying taurolidine. A potential development of resistance might be associated with efflux mechanisms.

Background
According to the latest reported data about 42% of the non-institutionalized U.S. population had periodontitis [1]. Periodontitis is a disease which destructs the tooth-supporting tissues. In pathogenesis of this disease host inflammatory response to subgingival bacteria leading to a pathogenic microbiota is central [2]. Here, Porphyromonas gingivalis is meanwhile postulated being a key stone pathogen in this transition [3], the major virulence factors are its cysteine proteases called gingipains [4]. Other bacteria involved in pathogenesis of periodontitis are Tannerella forsythia [5] and Aggregatibacter actinomycetemcomitans synthesizing a leukotoxin [6]. A.
actinomycetemcomitans strains differ in their ability to produce leukotoxin, highly leukotoxinproducing strains (JP2-clone) have a deletion in the promotor region [7].
Antimicrobials are widely used in treatment of periodontal diseases. Adjunctive use of chlorhexidine digluconate results in better clinical outcomes [8]. The use of adjunctive systemic antibiotics appears to be beneficial in advanced and severe cases of periodontitis [9]. However, the long-term clinical benefit following the use of antibiotics is still unclear [10]. As a topical antibiotic minocycline (a tetracycline derivative) in microspheres has shown additional clinical benefits compared with nonsurgical periodontal therapy alone [11].
One potential alternative antimicrobial agent is taurolidine. Taurolidine is a derivative of the amino acid taurine, as an antimicrobial it has been proven to be safe and effective for prevention of central venous catheter infection [12]. In vitro-studies indicate an antimicrobial activity against oral microorganisms also when being organized in a biofilm. [13,14]. An application in dentistry was discussed already several years ago. Rinsing with 2% of taurolidine solution depressed growth of dental biofilm by about 50% [15]. In several studies we evaluated the potential of taurolidine in vitro.
The minimal inhibitory concentration (MIC)s of taurolidine were all below 1 mg/ml taurolidine with the exception of Candida albicans [16]. One percent of taurolidine inhibited clearly the formation of 12species biofilms; however, the effect on an established biofilm was as limited as shown as for minocycline [17]. In an ex-vivo model using subgingival incl. supragingival biofilm samples from periodontitis patients, the decrease of bacterial counts in biofilms was 0.87 log10 cfu, corresponding 86.5% following application of 3% taurolidine gel after 60 min [18].
Development of resistance against antimicrobials is meanwhile a global problem, more than half a millions deaths annually are attributed to infections caused by antibiotic-resistant micro-organisms [19]. Development of resistance is clearly associated with antibiotic consumption and it is depending on the used antibiotic; the relationship is strong when quinolones are used and rather weak when 4 beta-lactams were applied [20] In addition, a considerable number of studies report the development of resistance to commonly used antiseptics, in part cross-resistance with antibiotics was found [21].
The increasing prevalence and spread of antimicrobial resistant bacteria is the natural consequence of genetic bacterial evolution. The more an antimicrobial agent is used, the higher is the probability of resistance formation [22].
Antimicrobial agents' resistance in bacteria can be of different origin, with a distinction between intrinsic and acquired resistance mechanisms. Intrinsic resistance mechanism is a natural property of microorganisms. Acquired resistance mechanism is based on a genetic modification of the bacterium The purpose of this follow-up study was first to determine in more detail the mode of action of taurolidine against bacterial species being associated with periodontal disease and second to verify a potential development of resistance against taurolidine in comparison with minocycline.

Antimicrobials
In all experiments, taurolidine 2% solution (Geistlich Pharma AG, Wolhusen, Switzerland) was used and diluted until to the necessary dilution. As positive control minocycline (Sigma-Aldrich, Buchs, Switzerland) and as negative control dH 2 O were used.

Microorganisms
Several oral bacterial strains (mainly P. gingivalis and A. actinomycetemcomitans) were included in the experiments ( Table 1). The clinical isolates (4 P. gingivalis, 3 A. actinomycetemcomitans, 2 T. forsythia, 2 Fusobacterium nucleatum, 4 oral streptococci) were obtained from subgingival biofilm 5 samples. Culturing subgingival biofilm, and isolation of respective bacterial strains was approved by the Ethical Committee of the Canton Bern (KEK 096/15). Without removing of supragingival biofilm, sterile paper point (ISO 50) were inserted into the periodontal pocket for 30 s. Then the pooled paper points were placed into 1 ml of reduced transport fluid. After culturing, colonies typical for the respective species were isolated and subcultivated. Identity was confirmed by nucleic acid-based methods.
Bacterial strains were kept frozen at -80°C. About one week before experiments, they were subcultured and passaged 2 -3 times on tryptic-soy-agar plates with 5% of sheep blood (Oxoid, Basingstoke, GB).

Interaction with bacterial virulence factors
To determine a potential effect of lipopolysaccharide (LPS), 10 pg and 100 pg LPS originated from Escherichia coli (Sigma-Aldrich) and 10 ng LPS from P. gingivalis (InvivoGen, Toulouse, France) were used. Taurolidine in the final concentrations of 0.01%, 0.25% and 1% and as control 16 µg/ml minocycline were added to the LPS for 2 h at 37°C. Thereafter, endotoxin activity was measured by Limulus amebocyte lysate assay (LAL) QCL-1000 TM test (Lonza, Basel, Switzerland) according to the manufacturer's recommendation. Also the two bacterial strains P. gingivalis ATCC 33277 and A.
actinomycetemcomitans Y4 (about 10 8 /ml) were exposed to same concentrations of antimicrobials for 2 h before measuring endotoxin activity.
Leukotoxin was purified as described by Kachlany et al. [26], added by a final centrifugation by using 10 kDa centrifugal filter to remove proteins of lower weights. Leukotoxin in a concentration of 4 µg/ml and two A. actinomycetemcomitans strains (10 8 /ml in M199 media) were exposed to three concentrations of taurolidine (0.01%, 0.5%, 1.5%) and to 16 mg/ml minocycline for 2 h before monocytic cells were added. Vitality of MONO-MAC-6 cells was determined after 1 h and 20 h of 7 incubation at 37°C with 5% of CO 2 by using MTT assay according to Mosmann [27].
Experiments on LPS and leukotoxin activity were made in independent triplicates. One-way ANOVA with post-hoc Bonferroni compared different antimicrobials with controls each. Further, t-test was used to find differences in viability of MONO-MAC-6 cells after exposing A. actinomycetemcomitans strains and leukotoxin with non-stimulate cells (only controls). A difference with a p-value of <0.05 was considered as being statistically significant each.

Potential development of resistance
Initially all clinical isolates were included in this experiment. However, the A. actinomycetemcomitans were prepared for TEM and SEM as described above.
Strains with increasing MICs were screened for potential efflux resistance mechanisms. The efflux inhibitors 50 µg/ml NMP, 10 µg/ml CCCP, 100 µg/ml verapamil and 20 µg/ml reserpine (all inhibitors Sigma-Aldrich) were added to the nutrient media before adding antimicrobials and determining MIC again after incubation.

Visualization of taurolidine action
Differences of scanning electron microscopy (SEM) photographs of P. gingivalis ATCC 33277, T.
forsythia Be13237 and Parvimonas micra ATCC 33270 without and with 2 h exposure to 1% taurolidine were small (Fig. 1A). After exposing to taurolidine, P. gingivalis ATCC 33277 showed in part pores and release of constituents from the cell wall, similar dense structures were attached to P. micra ATCC 33270, which might be a sign of a loss of bacterial structure. Transmission electron microscopy (TEM) photographs (Fig. 1B) underline a dissolution of cell plasma constituents by taurolidine.

Interaction with most important virulence factors
Taurolidine inhibited concentration dependent the endotoxic activity of Escherichia coli LPS, P.
gingivalis LPS and of the two gramnegative strains P. gingivalis ATCC 33277 and A. actinomycetemcomitans Y4. Statistically significant differences were always found when comparing 1% taurolidine with control. In contrast, 16 µg/ml of minocycline did not have any effect (Fig. 2).
Next, viability of monocytic MONO-MAC-6 cells measured by MTT assay underlines the toxicity of purified A. actinomycetemcomitans leukotoxin. Already after 1 h of incubation, viability of the cells tended to be lower, after 20 h the difference was statistically significant (p=0.004). Among the two used bacterial strains, the Be12206 strain belonging to the JP2 clone decreased the viability of the cells after 20 h of incubation (p=0.006). However, in this assay using the cell line of MONO-MAC-6 cells, an exposure to taurolidine clearly decreased the MTT activity, the effect is visible for 0.25% and 1% of taurolidine already after 1 h and for all three used concentrations after 20 h (each p<0.01). An inhibiting effect of the added antimicrobials (both taurolidine in three concentrations and 16 µg/ml 9 minocycline) on toxicity of leukotoxin to MONO-MAC-6-cells was never measured (Fig. 3).
Taurolidine blocked the activity of the purified P. gingivalis arginine specific cysteine protease (RgpB).
Using bacterial strains of P. gingivalis, the results were different. Applying 1% taurolidine to P.
gingivalis ATCC 33277 and J374-1, the measured arginine-specific amidolytic activity clearly decreased, whereas 0.01% did not exert any effect on the enzyme activity. In contrast, when using the HG66 strain nearly no arginine-specific amidolytic activity was found already after bacterial culture was exposed to 0.01% taurolidine. Results for minocycline were similar when using the purified gingipain and the HG66 strain but there was no inhibition of gingipain activity when using the ATCC 33277 and the J374-1 strains (Fig. 4).

Potential development of resistance
Analyzing a potential development of resistance, an increase of MIC of taurolidine was found only against one P. gingivalis strains. But it is of interest to note that four strains, among them the two included T. forsythia strains became more sensitive to the antimicrobial after certain passages ( Table   2). Resistance development against minocycline was found in five of the 16 included strains after exposure to subinhibitory minocycline (Table 3).
Next, strains exhibiting an altered sensitivity after certain passages were characterized further. The increase of resistance of taurolidine against one strain (P. gingivalis J374-1) might be linked to efflux mechanisms. It seems that this strain contains several efflux pumps. After removing antimicrobial pressure for three passages, MIC values were equal or closed to baseline values (Table 4). Efflux mechanisms might play also a role in two A. actinomycetemcomitans strains showing an increase in resistance against minocycline. It is obviously not of importance in the other three strains with increased resistance. Removing antimicrobial pressure for a few passages did not revert the increased MIC values (Table 5).
SEM photographs of P. gingivalis J374-1 show more vesicles at the surface after developing a higher resistance to taurolidine. After being exposed to taurolidine and becoming more sensitive, T. forsythia Be13237 has more rounded ends and S. constellatus BeTa7-1 lost obviously the ability to form chains.
Highly dense particles are visible in the TEM photographs of P. gingivalis J374-1 after exposure to subinhibitory taurolidine. Those seems to be located in the cells and closely attached to the surface.
TEM photographs of T. forsythia Be13237 show dense structures surrounded a halo attached to the outer membrane (Fig. 5).

Discussion
In this study in more detail the mode of action of taurolidine against bacterial species being associated with periodontal disease and a potential development of resistance were analysed. Except for the A. actinomycetemcomitans leukotoxin we focused on a potential interaction with P. gingivalis gingipains. Gingipains are membrane associated proteases functioning in proteolytic processing of nutrients, adhesion to host cells, evasion of immune response and more attributes [39,40]. The arginine specific amidolytic activity of the purified cysteine protease RgpB and of different strains was clearly inhibited until blocked by taurolidine. Total blocking was found when RgpB and bacterial culture of the HG66 strain were exposed to 1% taurolidine. The HG66 strain is the only exception of all P. gingivalis strains where gingipains are not glycosyated and bound to outer membranes or outer membrane vesicles but released in a non-glycosylated soluble form into extracellular milieu [4]. This underlines that taurolidine inhibits directly gingipain activity. Also the as control used minocycline inhibited purified RgpB but there was obviously no effect on membranebound gingipains which suggests that the potential interference of taurolidine with the cell wall might be more of importance also in inhibition of gingipain activity.
In the present study, we compared the potential development of resistance of taurolidine versus minocycline on species being associated with periodontitis. Exposing microbial strains to subinhibitory Interestingly, a few strains became more sensitive to taurolidine when being exposed to subinhibitory The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions
SE supervised and designed the study. SR, SS, NA conducted the experiments. SN carried out and analyzed the SEM and TEM analyses. SR, SS, NA processed and analyzed the data. SE, JZ and AS interpreted the results. All authors wrote, reviewed and approved the final manuscript.

Ethical approval and consent to participate
Culturing subgingival biofilm, and isolation of respective bacterial strains was approved by the Ethical Committee of the Canton Bern (KEK 096/15). All study participants signed an informed consent.

Consent for publication
Consent for publication was obtained from Geistlich Pharma AG.    Table 3 Minimal inhibitory concentration (MIC) of minocycline (µg/ml) at baseline and after certain passages on agar plates containing subinhibitory concentrations of the compound