Antimicrobial efficacy of chlorine agents against selected oral pathogens

Active chlorine was determined, if applicable, spectrophotometrically (Spectroquant^®, chlorine test, 1.00599.0001, Merck, Darmstadt, Germany) at 557 nm. The utilised standard curve gave a molar absorptivity, ε: value at 23,000 (mol/L)^-1*cm^-1, further used to calculate the total of free and combined chlorine (only active chlorine) in agents using the Beer-Lambert law. The chloramine-based products hold an excess of free chlorine ≈ 80 mmol/L after subtracting the stoichiometric component A (140 ± 7 mmol/L) and component B (53 ± 1 mmol/L amino acids) respectively before mixing. Irrigants containing active chlorine were adjusted with water to the same concentration (70 mmol/L of free or combined chlorine, representing 0.5% w/v of the active substance in DAK, PER and CAR); detailed information is presented in Table 1. Sodium hypochlorite, chlorhexidine and hydrogen peroxide were handled according to EU legalisation from a safety and biocompatibility perspective (the MDR directive in ISO 10993). Furthermore, preparation and handling followed the laboratory’s local restrictions, the material safety data sheet (MSDS) and instructions for use (IFU) when using commercial products.

The bacterial strains obtained from ATCC (LGC standards, SE) were kept in −80°C at the Department of Cariology, University of Gothenburg, Sweden, before use. Isolated colonies from overnight growth on agar were transferred to phosphate buffer and washed twice. The optical density of each suspension was as assessed at 500 nm and adjusted to 0.5. These standardized suspensions were used for subsequent experiments. Viable bacterial counts (CFU; colony forming units) were obtained after 10-fold dilutions were plated on to respective agar and incubated for 2 days at 37^°C under anaerobic conditions. Detailed information about bacterial strains is presented in Table 2.

Agar plates were inoculated with the test specimen by swabbing the surface with 3–4 mL of the bacterial suspension. The excess suspension was removed with sterile cotton and plates were dried at 37°C for 30 min to ensure the agar surface was dry. Filter paper discs (Ø, 6 mm) were impregnated with 10 μL of irrigants (70 mmol/L), placed on the agar surface and incubated at 37°C for 24 h in anaerobic conditions (85% N₂, 10% CO₂, 5% H₂). The size of the diameters of the inhibition zones was measured after incubation; all the series were performed in triplicate.

Single volumes (2.5, 5.0, 7.5, 10.0, 20.0, 50.0 and 100.0 μL) of the commercial agents (70 mmol/L) were added to test tubes containing 1 mL of bacterial suspension using a Hamilton syringe (0-50 μL). A droplet of the agent was placed on the neck of the tube until all seven volumes had been added. At an exact time, all the tubes were vortexed for 5 s. After five and 10 min of reaction, bacteria were sampled (25 μL in duplicates) and inoculated on blood agar plates. The plates were incubated anaerobically at 37°C for one to three days. The results were confirmed under the microscope and registered as (-) 100% lethality, total inhibition and (+) where one or more colonies were detected. The reported values represented the minimum volume required to obtain the complete lethality of bacterial cells (see Table 3).

The ten tested micro-organism cultures, suspended in Mueller Hinton Broth, supplemented with 20.25 mg Ca²⁺ and 11.5 mg Mg²⁺ per litre, adjusted to 0.5 at OD₅₅₀, were distributed (100 μL) into 96-well microtiter plates (ThermoFisher Scientific, UK). The 2-fold dilutions (from 70 to 0.03 mmol/L) of the commercial agents (CAR, PER, DAK, CHX, H₂O₂) were prepared and then added to the 96-well microtiter plate in aliquots of 100 μL. After mixing with bacterial suspension, the final concentration ranged from 0.02 to 35 mmol/L. Positive (100 μL Triton, X-100, Sigma-Aldrich) and negative controls (no agent) were included. The plates were then sealed with an adhesive plastic film and incubated for 2 h at 36 ± 1°C. After incubation, 10 μL of each sample was plated on corresponding agar plates (Table 2). The plates were incubated anaerobically for 24 h at 37°C and the results were evaluated. The tests were run in duplicate.

For CAR and PER, the reactivity of active chlorine in relation to the number of bacterial cells was calculated. The active chlorine (Cl⁺) was calculated from the molecular concentration (mol/L) in set volumes (L) using Avogadro’s constant, numbers of particles in reciprocal mole (NA=6.022×10²³/mol). These numbers were compared with the total number of bacterial cells from the established (CFU) resulting in the ratio (Cl⁺/bacterial cell), see Table 3.

Live/dead staining was used to detect the antimicrobial efficacy of the commercial agents in damaging or killing the bacterial cells in a concentration- and time-dependent manner, a method known as time-kill kinetics. Standardized bacterial suspensions were prepared in the manner mentioned above. The live/dead tracer, Syto 9 and propidium iodide were prepared following the protocol for Invitrogen (Filmtracer live/dead biofilm viability kit L 10316, Invitrogen) and 160 μL of the film tracer mixture was added to 16 mL of each bacterial suspension. Mixed bacterial suspensions were next added to the 96-well microtiter plate in aliquots of 100 μL. The antimicrobial agents were also prepared in the same manner as 2-fold dilutions (70 mmol/L), with the final concentration range from 0.02 to 35 mmol/L in aliquots of 100 μL. Before the test, standard curves for each tested specimen stained with Syto 9 were created. Positive (100 μL Triton, X-100, Sigma-Aldrich) and negative controls (no agent) were included. Kinetics were measured in a microplate reader (CLARIOstar, BMG Labtech, GmbH, Germany, equipped with the CLARIOstar software, 2013) using the Syto 9 dye with an excitation/emission maximum at 480/500 nm for intact cells (green) and 490/635 nm for lysed cells (red) respectively. Readings started directly after adding agents and data were collected every 10 min up to 1 h, at 37°C with rate 2 shaking in between measurements. Tests were performed in triplicate.

The irrigants displayed distinct antimicrobial characteristics, apart from H₂O₂ and AAP (both ranking 7), where no growth inhibition was observed (Table 3). The results were analysed for differences between the agents and between the two groups of bacteria: cariogenic and periodontal. The strongest, statistically significant effect on all bacteria was observed for CHX (ranking 1), followed by OCl and Dakin’s solution. The weakest inhibitory effect was observed for NOG (ranking 6). Overall, CAR exhibited a more substantial inhibitory effect than PER. Moreover, OCL (ranking 2) and PER (ranking 5) were more effective against periodontal, gram-negative bacteria, whilst DAK (ranking 3) and CAR (ranking 4) inhibited cariogenic, gram-positive bacteria more efficiently. In all, CHX was the preferred antimicrobial agent to use between both groups of bacteria, with a significant difference in comparison to all the other agents, p < 0.05.

The broth dilution method revealed that the lowest instant killing volume, after 5-min exposure, for DAK was 2.5 μL (all bacterial strains except PN and PG), followed by PER (for FN) and CAR (for AA) 5 μL and for CHX (for LBC and AA) 20 μL (Table 3). Hydrogen peroxide did not inhibit the growth of any tested bacteria. The lowest instant killing effect was more pronounced after 10 min; however, the lethality pattern was similar, regardless of the exposure time. Bacterial groups, caries and periodontal, only differed significantly in response to CAR treatment (two-way ANOVA with Sidak’s multiple comparisons test, p<0.0001). Additionally, a significant difference in CAR treatment was observed in cariogenic bacteria between five and 10 min (p<0.05), Fig. 2. Significant differences between agents were observed; Table 4 presents details in supplementary material.

All the agents inoculated in microbial suspensions showed an antimicrobial effect observed on agar plates. Figure 3 presents the representative results obtained for two strains of Streptococcus mutans. The MBC values for each agent on each bacterial strain are shown in Table 3. Generally, the lowest MBC value was recorded for CHX (0.03 mmol/L), followed by Dakin’s (2.2 mmol/L) and chloramines (PER, CAR) (2.2–8.8 mmol/L), with H₂O₂ being least effective. As expected, due to bacterial structure, the gram-negative bacteria within the periodontal group were more susceptible than the gram-positive, cariogenic bacteria.

To calculate the number of single chlorine molecules contributing to the MBC effect of an agent, the stoichiometric ratio of chlorine vs the bacterial cells (1:1) per specimen was calculated. The impact of chloramines (CAR and PER) was similar on SM-IB (11.96), PI (12.20) and PN (11.95), whilst a variation was observed for the remaining bacterial specimens (Table 3). PER emerged as more efficacious, with a lower chlorine/bacteria ratio (≤ 12) than CAR for most tested bacterial species.

A standard curve of bacterial growth was produced for each tested bacterial species for the live/dead analysis with Syto 9 (490:650 nm), with the absorbance of the living cells at 490 nm that increased over time and the absorbance at 650 nm for lysed cells (Fig. 4).

The time-kill curves for five agents (CHX, H₂O₂, DAK, PER, CAR) using A. actinomycetemcomitans and S. mutans IB are shown in Fig. 5. CHX induced a bacteriostatic effect in all analysed bacteria; the activity depended on the concentration of the agent and differed between bacterial strains. Bactericidal activity, greater than a three log₁₀-fold decrease, was obtained with 8.9 mmol/L for AA and 1.15 mmol/L for SM-IB as early as the tenth minute of the test. A similar kinetic pattern was obtained for DAK and PER, with the inhibition of bacterial growth up to 1.5 log₁₀. CAR inhibited the growth of Streptococcus mutans after 10 min in 17.5 mmol/L. H₂O₂ induced minimal growth inhibition in SM-IB and did not affect the growth of AA.

The disc diffusion test established that CHX was the most effective agent, with the diameter of inhibition zones significantly larger than that of other tested agents. These results also agree with other studies showing the more significant impact of CHX on inhibition zones in lower concentrations than NaOCl [42]. On the other hand, CHX and hypochlorite irrigants are reported to produce a similar reduction in bacterial levels during root canal therapy [30], but in vitro testing is highly dependent on its concentration [43, 44]. However, the reaction with CHX is much slower than with the oxidative sodium hypochlorite, which reacts readily in forming short-lived intermediaries [17, 40]. This means that sodium hypochlorite compound is consumed at a higher rate and may explain the better result obtained by CHX in the disc diffusion method that does not deteriorate at the same rate and would therefore be more beneficial under long-term exposure. Moreover, it is vital to consider the molecular weight of an active substance when validating its antibacterial activity. This was found by Müller et al. in 2008 comparing the cytotoxicity and antibacterial activity of commercial agents, where the ranking order was considerably different based on molar concentration (mol/L), apart from mass concentration (w/v) [43]. This was especially evident for reagents with a relatively high molecular weight, such as PHMB and CHX. This was also the basis of this study and the reason why all the concentrations were equaled. Furthermore, CHX was slightly more efficacious on periodontitis bacteria than caries bacteria, which is consistent with the agent being used more frequently in the periodontitis context [24]. However, apart from discolouration and taste disturbances, there are a few cases reporting adverse side effects from CHX. It should be noted that CHX is considered a hidden allergen and might be involved in more cases than recognised [45].

The efficacy of component A in PER, OCL and DAK surpass CHX. These oxidative forms of chlorine act as oxidative agents and gain electrons in the reaction with bacterial proteins, carbohydrates and lipids, thereby disturbing the bacterial cell membrane, most often into lysis [17, 46]. This is a process that resembles the MPO (myeloperoxidase) reaction from the radical oxygen system (ROS) of the immune system, turning HOCl into potent chloramines for the killing of bacterial pathogens in an alkaline milieu [40, 47]. These reactions are considered rapid or short lived [46], as they involve both redox and radical systems [17, 46]. However, rapid agent decomposition may facilitate cell survival during analysis in an agar disc diffusion test, making it less suitable. On the other hand, according to the collision theory in thermodynamics and as stated in the Arrhenius Equation, the rate of a chemical reaction is proportional to the number of collisions between reactant molecules [48]. Moreover, it will increase in a volume (increased surface size) compared with a solid state (disc), as both molecular movement and contact are increased in a solution [48], which is why rapid reaction mechanisms might be more favoured in methods that involve solvents or suspensions.

Regarding the bacterial groups, DAK was more effective in the caries group, whilst OCL (comp A) was more effective in the periodontal group, where there is no marked difference other than the pH value. Comparable results for chloramines, PER and CAR, were observed, with CAR being more effective, especially on caries bacteria, and thus in line with the intent of use. Chloramines have slightly lower oxidative power (oxidation state +½ of chlorine) than both OCL (component A of Perisolv) (+I, Cl⁺) and Dakin’s solution (+I, Cl⁺) and this could explain the lowered inhibitory effect of CAR and PER in comparison with the other aforementioned irrigants. In addition, chloramines are reported to be very unstable and are turned into radicals from the homolysis of N-Cl, thus being consumed rapidly and not effective for a long exposure time [40, 46].

Hydrogen peroxide and AAP (component B of Perisolv) were not effective. AAP (amino acids, pH 10.5) is not a potent antimicrobial agent and, as a result, no growth inhibition was observed. Hydrogen peroxide (oxidation state -I of oxygen, pH 4) is readily decomposed into water if not handled properly and, if so, when it has perhaps evaporated [49], it will have little inhibitory effect. Furthermore, lower concentrations of H₂O₂, through bacterial enzymes such as peroxidases, may induce tolerance to lower concentrations of H₂O₂ [17]. Even though the concentrations were equalled, the pH of the irrigants was different; OCL (pH 11) and DAK (pH 9), which may influence the results.

It is worth noting that commercially available dental solutions containing CHX and H₂O₂ exhibit significantly lower and higher concentrations, respectively, when compared to 70 mmol/L concentrations used in this study. Specifically, CHX solution typically range from 1 to 2 mmol/L (1–2 mg/ml), whilst H₂O₂ have concentrations as high as 800 mmol/L (3%). These findings suggest that CHX demonstrates a strong antibacterial efficacy, even at very low concentrations with regard to the commercially available products. On the other hand, and in line with the commercial concentrations, H₂O₂ requires relatively high concentrations to achieve a similar effect. This indicates the importance of understanding the mechanism behind the active ingredients, when evaluating the efficacy of dental products which can only be compared if concentrations are equaled.

DAK, with its significantly smallest bactericidal volume at both 5 and 10 min of treatment, was followed by PER, CHX and CAR, with H₂O₂ the least effective (p<0.001). The low reactivity of H₂O₂ could be attributed to its decomposition into water [49]. The molecular size and redox reactions of the microbicidal agents may play a role in their effectiveness. Smaller molecules like DAK and PER, which are less sterically hindered than CHX, may disperse more easily in a closed container, leading to a more potent antimicrobial effect. The capacity of chlorine atoms to ‘jump’ between reactants before reaching equilibrium [40] in a hypochlorite solution may also contribute to the longer antimicrobial effect. Taken as a whole, this will favour the oxidative power of hypochlorite solution, for example, compared with CHX. On the other hand, CHX overcame the efficacy of the oxidative H₂O₂, which was suspected of being a potent antimicrobial agent after reaction with amines and sulphur components from bacterial cells [17, 47, 50], but this was not observed. Again, the unwanted decomposition of H₂O₂ or an effect of the slightly acidic pH of the H₂O₂ solution could have hindered these outcomes. Chlorine is perhaps a more potent oxidising agent in bacterial suspensions [51]. Even though CAR was the only agent showing a significant difference in sensitivity between caries and periodontal bacteria and time, significantly smaller volumes of DAK, PER and CHX are needed to obtain lethality.

Like previous methods, CHX emerged as most effective, with MBC values below 0.15 mmol/L, followed by Dakin’s solution (2–4 mmol/L) and chloramines (8–17 mmol/L). Hydrogen peroxide was again ineffective. Perisolv had a slightly stronger bactericidal effect than Carisolv, possibly due to added titanium dioxide [51]. Dakin’s solution, which is a buffered solution (NaHCO₃, pH 9) without amine functions, has a higher oxidative number of the chlorine than the chloramines and this perhaps contributes to the improved effect. In another study, sodium hypochlorite was tested with well-known antibiotics with both a lower antimicrobial effect (MBC; 20mmol/L) and less effect on gram-negative bacteria [52]. In addition, Carisolv and Perisolv are formulated in cellulose gels, which might also be seen as a physical diffusion barrier. Due to their inability to maintain stable anaerobic conditions during preparation and treatment, we did not obtain results for three periodontal bacteria: PG, PI and PN. The remaining AA and FN followed earlier trends of treatment, with CHX and DAK being the most effective.

Previous studies have investigated the kinetics of CHX on skin infection bacteria using live/dead staining. The results demonstrated that CHX inhibited bacterial count by two logarithmic units at low concentrations (0.125 mmol/L) over the course of several days [53].

As a result, our study monitored bacterial growth exposed to agents over time using Syto 9 dye. We found that the bacteriostatic effect or inhibition using a similar technique with the Syto 9 dye was dependent on their concentration and varied between the tested bacteria. A three-log inhibition change in AA growth was observed during CHX treatment (35-8.9 mmol/L) after 10 min; however, after 30 min, a total recovery was observed. The bactericidal effect of CHX on SM was obtained after 10 min with concentrations of ≥1.15 mmol, whilst lower concentrations resulted in a 1.5-log inhibition change. The periodontal group of bacteria was less sensitive than the caries group and in line with our calculation of single chlorine atoms needed to kill a single bacterial cell at a ratio of 1:1.

Exposure to DAK and PER resulted in a maximum 1.5-log inhibition of S. mutans, with no recovery phase. A. actinomycetemcomitans responded with a one-log inhibition to the highest DAK concentration and was not affected by PER treatment. Once again, the periodontal group was more resistant to treatments than the caries group. Thus, one conclusion is that chlorine and not chloramines have the same effect on bacteria. A bactericidal effect was observed using Carisolv (chloramine without titanium, ≥17.5 mmol/L) in S. mutans after just 10 min. Carisolv appeared to affect the caries bacteria in the same way as CHX. A. actinomycetemcomitans growth was not influenced by any CAR concentration.

Hydrogen peroxide treatment resulted in less than a one-log inhibition of S. mutans and no inhibitory changes in AA growth pattern. These results may again depend on the decomposition of the molecule into H₂O [17, 49]. The reactivity of hydrogen peroxide is highly restricted by pH, where alkalinity increases its oxidative effect [46]. In this study, the hydrogen peroxide had an acidic pH.

The initial aim of this study was to evaluate the efficacy of different agents and their performance by diverse antimicrobial methodologies. For this purpose, individual bacterial strains were used. These may not reflect the true situation of a complex oral biofilm as a dysbiotic biofilm would involve many more parameters. It is in future projects a plan to analyse the response of dysbiotic biofilms to examined irrigants using most suitable method.