Effect of erythritol and xylitol on Streptococcus pyogenes causing peritonsillar abscesses

Polyols are effective against caries-causing streptococci but the effect on oropharynx-derived pyogenic streptococci is not well characterised. We aimed to study the effect of erythritol (ERY) and xylitol (XYL) against Streptococcus pyogenes isolated from peritonsillar abscesses (PTA). We used 31 clinical isolates and 5 throat culture collection strains. Inhibition of bacterial growth by polyols at 2.5%, 5% and 10% concentrations was studied and the results were scored. Amylase levels in PTA pus were compared to polyol effectivity scores (PES). Growth curves of four S. pyogenes isolates were analysed. Our study showed that XYL was more effective than ERY inhibiting 71–97% and 48–84% of isolates, respectively, depending of concentrations. 48% of clinical and all throat strains were inhibited by polyols in all concentrations (PES 3). PES was negative or zero in 26% of the isolates in the presence of ERY and in 19% of XYL. ERY enhanced the growth of S. pyogenes isolated from pus with high amylase levels. Polyols in all concentrations inhibited the growth in exponential phase. In conclusion, ERY and XYL are potent growth inhibitors of S. pyogenes isolated from PTA. Therefore, ERY and XYL may have potential in preventing PTA in the patients with frequent tonsillitis episodes.


Results
Growth characteristics of S. pyogenes PTA isolates in the presence of polyols after 24-h incubation. The impact of ERY and XYL in different concentrations (2.5%, 5% and 10%) was studied on 31 S. pyogenes isolates from PTA pus and five type collection strains with throat origin and the results were compared with the growth in BHI. Overall, xylitol was more effective and inhibited the growth in 71-97% of investigated PTA isolates, while erythritol was less active inhibiting 48-84% of investigated isolates depending of concentrations (Fig. 1). 10% of polyol solution was the most active, followed by 5% and 2.5%. (Fig. 2). All tested polyol concentrations, except ERY 2.5%, showed statistically relevant (p < 0.0001-0.009) inhibitory effect against PTA isolates compared to BHI. 10% xylitol was more effective than 10% erythritol (p = 0.0005). The growth of all throat derived type collection strains was inhibited by both polyols in all studied concentrations.
As there were inter-isolate differences in the impact of the polyols, polyol effect was analysed for each S. pyogenes isolate. The impact of different concentrations of ERY and XYL, on the growth of studied S. pyogenes isolates compared to BHI, was calculated as the difference between the polyol growth data and BHI growth data for each isolate. Respective results are presented in Fig. 3. Both polyols showed a high inhibitory effect against most of the isolates. The growth inhibition was detected in nine tenth ( Polyol inhibition effectivity scores. Polyol effectivity scores (PES) were calculated to summarize the inhibitory effect of polyols in different concentrations (2.5%, 5% and 10%) for each isolate. PES was calculated as a summary effect of three polyol concentrations (2.5%, 5% and 10%), where each polyol concentration contributed − 1, 0 or + 1 points to the PES depending on whether the S. pyogenes isolate growth increased, remained unchanged or decreased in the presence of the polyols, respectively (Fig. 4). For example, if for an isolate X, all the three concentrations of XYL led to decrease in growth, XYL would have a PES score of 3 (+ 1 points for each concentration leading to decreased growth), while if two concentrations of XYL led to decrease and one led to increase in growth, XYL would have a PES score of 1 (which is a cumulation of + 2 points for two concentrations leading to decreased growth and -1 point for one leading to increased growth). Almost half (13/31; 42%) of S. pyogenes PTA isolates and all throat strains from the culture collection were inhibited by ERY and XYL in all the studied concentrations (as indicated by polyol effectivity scores 3). The polyol effectivity score was negative or zero for around a quarter (8/31; 26%) of the isolates in the presence of ERY and for one fifth (6/31; 19%) of the isolates in the presence of XYL. For 24 isolates, XYL and ERY impact as assessed by the PES was similar (Fig. 4), i.e. positive PES for both XYL and ERY or negative for both XYL and ERY. For 6 isolates, PES scores indicate potentially opposing impacts of ERY and XYL (Fig. 4).
Polyol effectivity score against S. pyogenes isolates from PTA pus depending on amylase levels in pus. High amylase level in PTA pus indicates association with Weber's salivary gland infection. Among the patients from where the S. pyogenes was isolated, corresponding data about the respective amylase levels in PTA pus was available for 18 patients. High amylase levels (25-10,019, median 860 U/L) were detected in 6 patients and low amylase levels (< 3 U/L) were detected in 12 patients. ERY inhibitory score of isolated S. pyogenes isolates was negative or zero in two third of high pus amylase cases compared to 17% of low pus amylase cases (4/6 vs 2/12, respectively; p = 0.034). XYL inhibitory score was negative in half of high pus amylase cases (3/6) compared to 17% of low pus amylase cases (3/6 vs 2/12, respectively; p = 0.139) (   Figure 3. Effect of erythritol and xylitol on S. pyogenes PTA isolates and type strains compared to BHI. For each condition (e.g. for polyol ERY at 2.5%), the difference between the polyol growth data and BHI growth data (OD value of BHI minus OD value of polyol solution) is represented. The colored circles above the value "0" in Y axis are "inhibited growth", values below the value "0" are enhanced growth, and on the value "0" is "no change". The growth inhibition was detected in 28/31 of the isolates in the presence of at least one concentration of ERY and in 30/31 isolates of XYL.    www.nature.com/scientificreports/ growth curve study. Both the polyols, consistently inhibited the growth of all studied isolates in all studied concentrations in exponential growth phase.

Discussion
Our study showed that ERY and XYL influences the growth of S. pyogenes isolated from PTA in a strain-specific manner. In the exponential growth phase, both the polyols, in all the tested concentrations, were effective against S. pyogenes, however in stationary phase the effect of polyols was type and concentration dependent. The growth inhibition was detected in nine tenth of the isolates in the presence of at least one concentration of ERY and in all except one isolate of XYL. S. pyogenes from the patients with high amylase levels in the PTA pus, indicating small salivary glands as possible source of PTA, were less responsive to growth inhibitory effect of ERY as opposed to XYL. ERY and XYL have been proven to inhibit the growth of oral cariogenic bacteria and widely used in caries prophylactics 11,17 . Strain-specific effect of polyols on cariogenic bacteria such as S. mutans, S. sobrinus and S. wiggsiae has been demonstrated in previous studies 10,18 . As there are inter-strain differences among bacterial species, research conclusions based on observations from a large number of isolates should be more reliable and translatable. The results presented in this study is a cumulation of experiments on 36 S. pyogenes isolates, including 31 clinical S. pyogenes isolates and five throat derived strains from the culture collections. To our best knowledge, this is the first study to address polyol efficacy on the oropharynx derived pathogen, S. pyogenes, which is able to cause acute tonsillitis and potentially life-threatening PTA. ERY has been shown to be more effective than XYL against caries causing oral mutans streptococci 11,12,17 . Though both polyols are generally similar, they differ in size and in the effectivity to scavenge hydroxyl radicals. Accordingly, superiority of ERY against cariogenic streptococci is associated with smaller molecule size and higher biological activity 19 . In our present study, we found XYL to be more effective against S. pyogenes than ERY, even though S. pyogenes and mutans streptococci are genetically closely related. Further research is needed to clarify the exact mechanisms behind these findings.
Erythritol and xylitol are generally considered non-toxic and safe for human consumption 11 . At the same time, they inhibit the growth of bacteria. Antibacterial effect of xylitol has been associated with accumulation of toxic nonmetabolizable xylitol phosphate in the bacterial cells inhibiting glucose uptake and glycolytic enzymes of bacteria 20,21 . Inactivation of the fxpC gene coding for the fructose-phosphoenolpyruvate:sugar www.nature.com/scientificreports/ phosphotransferase system gives resistance to xylitol in S. mutans 21 . Erythritol mechanism of action is considered similar to xylitol 22, 23 , however exact reasons for non-susceptibility are not clarified yet. Previous investigations have shown that polyols, and also triclosan, may in certain concentrations enhance the biofilm formation of bacteria instead of inhibition 10,24 . In our study we observed both, growth inhibition and enhancement depending on polyol concentration and specificity of the S. pyogenes isolate. In order to compare the isolates and to generalize the results, we scored and summed three polyol concentration results according to the effect to the growth (positive effect, zero effect or negative effect). The growth of all throat-derived type collection strains and half of the PTA isolates was inhibited in the presence of all polyol concentrations used in our study (2.5%, 5% and 10%) in a 24 h challenge. The growth of another half of clinical PTA isolates was enhanced by all or some polyol concentrations. The matter of differences in polyol concentrations in various oropharynx locations must be considered. Though the safety of polyol consumption in the acute phase of infection with high rate of S. pyogenes replication, the possibility that under certain conditions the polyols may promote growth of potential pyogenic pathogen, should not be overlooked.
We were interested in linking the differences of the clinical attributes of the patients and the corresponding behaviours of S. pyogenes isolated from the patients. Amylase is an enzyme found in human saliva and increased amylase levels in PTA pus are associated with involvement of minor salivary glands in infectious processes 3,8 . High amylase in PTA abscess indicates inflammation of the salivary gland as the onset of infection. We tried to reveal possible differences between the isolates originating from the patients with high and low amylase levels. The amylase levels in pus were available from 18 patients. Erythritol had limited growth inhibitory effects (growth either increased or remained unchanged) on S. pyogenes isolated from pus with high amylase levels. The similar effect of xylitol was not statistically significant.
In the rapid growth (exponential growth) phase of S. pyogenes, polyols were effective in all concentrations against all studied isolates. Loimaranta et al. 18 have also found the inhibitory effect of polyols on S. mutans biofilm formation to be more noticeable in the early stages of growth. The highest concentrations (10% ERY and 10% XYL) were the most effective in the rapid growth phase. We found that the growth of S. pyogenes may be promoted by polyols at certain concentrations. For peritonitis caused by E. coli, animal experiments have shown that in the stationary phase of bacterial growth, slow bacterial growth actually occurs in parallel with the non-multiplying bacteria and the destruction of bacteria by the body's defences 26 . Laboratory experiments have shown that S. pyogenes strains can remain in saliva for a long time in the stationary phase 27 . When using polyols, phase, strain and concentration specific effects would be relevant to take into account.
One limitation of our study is the absence of amylase test results about all PTA cases. However, comparative data of 18 patient's pus amylase levels and growth characteristics of S. pyogenes isolates from the respective pus in the presence of polyols indicate possible differences between PTA cases depending on pathogenesis.

Conclusion
In conclusion, ERY and XYL are potent growth inhibitors of S. pyogenes isolated from PTA patients, especially in the exponential growth phase, though the activity is strain-specific. Therefore, the ERY and XYL-containing candies and chewing gums may have potential in preventing PTA in the patients with frequent tonsillitis episodes.
Altogether 36 Streptococcus pyogenes isolates, 31 from pus of PTA patients and 5 throat strains from culture collections (CCUG 23117, CCUG 53553, DSM 25932, DSM 25943 and DSM 11728) were included in the study.
The clinical S. pyogenes isolates and pus amylase data of respective patients were acquired during prospective clinical study of etiopathogenesis of PTA. The informed consent was obtained from all study subject or, if subjects are under 18, from a parent and/or legal guardian. All samples and data from patients were collected in accordance with relevant guidelines and regulations and all study protocols were approved by an institutional committee (approved by Ethics Review Committee on Human Research of the University of Tartu, protocol No. 255/T-1). All. S. pyogenes isolates are deposited in the Human Microbiota Biobank of Tartu (HUMB; http:// www. eemb. ut. ee/ eng/). Details of the study are described elsewhere 25 .

Bacterial growth inhibition after 24 h of incubation by different polyol concentrations.
The effect of polyols was studied using a previously described method 22,23 with slight modifications. Briefly, the tested substances were sterilized by filtration at desired concentration and added to the brain-heart infusion (BHI, Oxoid Limited, Basingstoke, Hampshire, UK) medium (sterilization by autoclaving at 121 °C for 15 min). The bacteria were incubated at 37 °C in 10% CO 2 for 24 h on blood agar (Oxoid Limited, Basingstoke, Hampshire, UK).
The microtiter plate wells (CELLSTAR 96 well polystyrene suspension culture microplates, F-bottom, Greiner Bio-One GmbH, Kremsmünster, Austria) were inoculated with equal amounts (200 µl) of bacterial (the final test-concentration of 10 5 CFU/ml) and polyol solution. The bacterial cells were grown at 37 °C in 10% CO 2 on the microtiter plate. The density of bacterial growth was detected spectrophotometrically in an absorbance microplate reader (Labsystems Multiskan Mcc/340, Fisher Scientific, Pittsburgh, PA, USA) at 620 nm. The measurement time points included 0 and 24 h.
The tested polyol concentrations (weight/volume) included 2.5%, 5% and 10%. The polyol effect was calculated as difference between average BHI and polyol test values. Polyol effectivity scores were calculated as a summary effect of three polyol concentrations (2.5%, 5% and 10%) where each polyol concentration gave − 1, 0 or + 1 points