2023 Volume 71 Issue 7 Pages 552-557
Benzalkonium chloride (BAC) is a useful preservative for ophthalmic solutions but has some disadvantageous effects on corneal epithelium, especially keratinocytes. Therefore, patients requiring the chronic administration of ophthalmic solutions may suffer from damage due to BAC, and ophthalmic solutions with a new preservative instead of BAC are desired. To resolve the above situation, we focused on 1,3-didecyl-2-methyl imidazolium chloride (DiMI). As a preservative for ophthalmic solutions, we evaluated the physical and chemical properties (absorption to a sterile filter, solubility, heat stress stability, and light/UV stress stability), and also the anti-microbial activity. The results indicated that DiMI was soluble enough to prepare ophthalmic solutions, and was stable under severe heat and light/UV conditions. In addition, the anti-microbial effect of DiMI as a preservative was considered to be stronger than BAC. Moreover, our in vitro toxicity tests suggested that DiMI is safer to humans than BAC. Considering the test results, DiMI may be an excellent candidate for a new preservative to replace BAC. If we can overcome manufacturing process issues (soluble time and flushing volume) and the insufficiency of toxicological information, DiMI may be widely adopted as a safe preservative, and immediately contribute to the increased well-being of all patients.
Benzalkonium chloride (BAC) is a quaternary ammonium cation and the most well-known preservative for ophthalmic solutions because it has a wide anti-microbial spectrum and the ability to promote drug-delivery through the cornea.1) The ammonium cation in BAC links to one benzyl group, two methyl groups, and one alkyl chain. The alkyl chain is a mixture of C12, C14, and C16 chains, and the mixed proportion of the three chains depends on the manufacturers and manufacturing processes. Among BAC with C12, C14, or the C16 alkyl chain, BAC with the C16 chain is known to exhibit the strongest anti-microbial activity because a longer chain is more effective against microorganisms. However, the cornea is more easily injured by a longer alkyl chain than a short chain.2) Considerations of both the anti-microbial activity and toxicity are important to use BAC as a preservative for ophthalmic solutions. Generally, in BAC products, the proportion of the C16 chain is very small as compared with C12 and C14 chains and the C16 chain may be absent, so a high concentration of BAC is required to achieve a sufficient anti-microbial effect.3)
As mentioned above, there is some information on the damage to the corneal epithelium caused by BAC, although it is a useful preservative. Hence, patients such as those with glaucoma and dry eye syndrome, who require the chronic administration of ophthalmic solutions, may suffer from BAC-induced damage of their corneal epithelium.3–5) To avoid this, ophthalmic solutions not containing BAC are desired. Currently, some containers are used to negate the need for BAC in ophthalmic solutions. These containers have a sterile filter at the tip, and the filter is able to prevent contamination of the container by outside microorganisms. The use of a filter on containers is an excellent idea to realize BAC-free solutions, but there are two issues regarding actual use. One is the pressure of pushing. The filter’s hole diameter is microscopic, and so a strong force is required to push the content through the filter. Therefore, some patients may feel difficulty using the filter-fitted container. The other is the risk of secondary contamination. After the first drop, the container will aspirate outside air. If some microorganisms exist in the air, they will adhere to the surface of the filter. Then, the adhered bacteria and molds may contaminate the second drop. Therefore, the usage of filter-fitted containers has not become widespread. Actually, more than 60% of ophthalmic solutions, except for anti-bacterial ophthalmic solutions, on sale in Japan still include BAC.6)
To resolve the above situation, new preservatives which are safer than BAC and have an adequate anti-microbial effect equivalent to BAC should be developed. In addition, any new preservative must be stable under heat stress, light stress, and on interaction with other additives. In this study, 1,3-didecyl-2-methyl imidazolium chloride (DiMI) was evaluated as a candidate new preservative because it was reported that it exhibits high anti-microbial activity against both Gram-positive and Gram-negative bacteria.7,8) We established a method for measuring DiMI, and evaluated the physical and chemical properties. We also evaluated the anti-microbial effect of DiMI with a minimum inhibitory concentration (MIC) and conducted a preservative effectiveness test (PET). Considering the results, DiMI may be an excellent candidate for a new preservative to replace BAC. At present, DiMI is not used as a food, cosmetic, or medicine additive because there is little safety information available. However, if DiMI overcomes these hurdles, it may be widely adopted as a preservative. Then, all patients will be able to benefit from ophthalmic solutions containing DiMI.
The chemical DiMI was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). BAC was purchased from Nichiyu Co., Ltd. (Tokyo, Japan). All other chemicals were of analytical grade.
Staphylococcus aureus ATCC 6538, Escherichia coli ATCC 8739, Pseudomonas aeruginosa ATCC 9027, Candida albicans ATCC 10231, and Aspergillus brasiliensis ATCC 16404 were used in all experiments. Before inoculation, these strains were pre-incubated.
Human corneal epithelial (HCE-T) cells9) were obtained from Dr. Kaoru Sasaki, Riken BioResource Research Center Cell Bank (Ibaraki, Japan).
HPLC AnalysisTo evaluate several features of DiMI, we developed a method to measure DiMI. The method was expected to have an easy procedure and show accuracy, and so we developed an HPLC measurement method. The molecular weight of DiMI is 399.10, and there are a methyl imidazole ring and two alkyl chains (C10 chains) (Fig. 1). Therefore, we considered that DiMI can be detected with a UV-visual detector, but a mobile phase or diluent must be selected to obtain a sharp peak shape because of the two hydrophobic alkyl chains. For the UV detection wavelength, we referred to the measurement method of imidazole peptides.10)
The HPLC system was comprised of the autosampler SIL-20AC, two LC-20AD pumps, the degasser DGU-20A3, UV-vis detector SPD-20A, and column oven CTO-20AC (Shimadzu Corporation, Kyoto, Japan). A column was adopted, YMC-Pack C8 (YMC Co., Ltd., Kyoto, Japan). The mobile phase was 1 M hydrochloric acid : acetonitrile = 2 : 3. The analysis conditions were as follows: a column oven at 40 °C, flow rate of 1.5 mL/min, and detection wavelength of 210 nm. All injected solutions were stored under room temperature in the autosampler, and the injection volume was set as 25 µL. The retention time of DiMI was about 13 min.
The HPLC method was validated for accuracy, linearity, limit of detection (LOD), and limit of quantification (LOQ). Accuracy was evaluated with three standard solutions (0.010, 0.013, and 0.015 mM), and the recovery rate was calculated in each sample. Linearity, LOD, and LOQ were evaluated with five standard solutions (0.0063, 0.010, 0.013, 0.015, and 0.019 mM).
Physical and Chemical PropertiesThe physical and chemical properties of DiMI were evaluated with absorption to the sterile filter, solubility, heat stress stability, and photostability.
Absorption to the sterile filter (4.5 cm2) was evaluated with 0.13 mM DiMI solutions (30, 50, and 70 mL) after filtration through the sterile filter (0.22 µm). DiMI contents in the filtrates were measured by HPLC, and compared with DiMI contents of standard solution.
Solubility to water was evaluated with 0.1, 0.5, or 2 g of DiMI in 100 mL of water. Each sample was mixed with a stirrer overnight. In addition, solubility time was evaluated by mixing 0.01 g of DiMI in 100 mL of water for 30 min, 1, and 2 h. After stirring each time, DiMI contents in the sample solutions were measured by HPLC. Since high concentration solutions more than 0.01% (0.25 mM) are not needed in an actual preparation of DiMI, we used 0.25 mM solution for solving time testing.
Heat stress stability was evaluated with 0.25 mM DiMI solution in three glass ampoules. The first ampoule was stored in a chamber set at 4 °C for 4 weeks, the second ampoule was stored in a chamber set at 60 °C for 2 weeks, and the third ampoule was stored in a chamber set at 60 °C for 4 weeks. After storage, all ampoules were stored in the 4 °C chamber until measurement by HPLC. DiMI contents in heat stress samples were compared with that of the first ampoule (stored in a cool chamber for 4 weeks).
Photostability was evaluated with 0.25 mM DiMI solution. The DiMI solution (5 mL) was dispensed into five glass ampoules. One ampoule was stored in a 4 °C dark chamber, and the other four samples were stored at 25 °C and exposed to 3000 lx of white light for about 20 d or UV for about three days until total illuminances achieved 120 k lux/h and 200 W h/m2, respectively. Further, one of the four samples was covered with thin aluminum foil during the period. After storage, DiMI contents of all samples were measured by HPLC, and compared with the stored sample in the 4 °C dark chamber.
Interaction between DiMI and Medicine AdditivesTo investigate the interaction between DiMI and additives, DiMI contents in the mixtures with some additives were measured by HPLC. Additives used in this examination were buffer components, isotonic agents, surfactants, and divalent metal ions. Actually, acetate, borate, citrate, phosphate, and trometamol (Tris) were adopted as buffer components as a 25 mM final concentration. Sodium chloride, glycerin, and D-mannitol were adopted as isotonic agents (0.15, 0.28, and 0.31 M, respectively). Polysorbate 80 and tyloxapol were adopted as surfactants at 1.5 and 0.45 mM. Divalent metal ions, such as 0.1 M of magnesium chloride and calcium chloride, were adopted to evaluate the interaction because our past experiment indicated that the multivalent metal ions form complexes with other compounds (data not shown). In addition, we evaluated the effect of interaction with additives on the minimum inhibitory concentration (MIC) of DiMI. In this evaluation, buffer components and surfactants were used.
MIC MethodThe MIC values for DiMI and BAC were obtained by the method recommended by the Japanese Society of Chemotherapy,11) with some modifications to decrease deviation on procedure of the inoculation. The modifications were the changes of sample volume (from 100 to 200 µL), inoculated microbial suspension volume (5 to 10 µL), and incubate temperature and time (from 35 °C, 18–24 h to 32.5 °C, 24–48 h). The molar concentration of BAC was calculated from one with C12 chain (Mw 339.99).
Preservative Effectiveness Test (PET) with Simple FormulationSince DiMI is expected to be used in ophthalmic solutions as a preservative, PET according to the Japanese pharmacopeia (JP), United States pharmacopeia (USP), and European pharmacopeia (EP) should be performed to evaluate the anti-microbial effect.12–14) Some DiMI formulations were prepared with buffer components and isotonic agents. In the formulations, DiMI was added at 0.075, 0.025, 0.013, and 0.0025 mM, and trometamol (buffer component) was added at 0.2 M. Also, D-mannitol or glycerin was added to be isotonic, respectively. The pH of formulations was adjusted to 7.5 with hydrochloric acid. As same as DiMI formulation, we evaluated anti-microbial effect of formulated BAC solution (10 mM phosphate buffer, sodium chloride, pH 7.0) with PET method. One formulation (0.025 mM DiMI, 0.2 M trometamol, and 0.31 M D-mannitol) was also prepared at pH 6.5 or 8.5 to evaluate the effect of pH on PET.
Measurement of Cell Viability and CytotoxicityHCE-T cells were cultured in Dulbecco’s modified eagle medium/Ham’s F-12 with 5% fetal bovine serum, 10 ng/mL epidermal growth factor, 5 µg/mL insulin, and 40 µg/mL gentamicin. The viability was analyzed using Cell counting kit-8 (CCK-8) (Dojindo Laboratories, Kumamoto, Japan). The cytotoxicity was evaluated by measuring the activity of lactate dehydrogenate (LDH) released from damaged cells using LDH Cytotoxicity Detection Kit (TaKaRa Bio Inc., Shiga, Japan).
Accuracy evaluated with three concentrations of DiMI solutions, linearity evaluated with five concentrations of DiMI solutions, LOD, and LOQ are shown in Table 1. Figures 2, 3 and 4 show a chromatogram of DiMI, a linearity graph and residual plot, respectively. As shown in the Table and Figures, the recovery ratio of accuracy was between 90 and 110%, correction coefficient was more than 0.990, 95% confidence interval of y-intercept included a point of origin, and residual plot was appropriately scattered across zero. In addition, LOD and LOQ, which were calculated with linearity, indicated that our method by HPLC was appropriate to measure the DiMI content of all samples in this study.
| Accuracy | |||
|---|---|---|---|
| DiMI content | |||
| 0.010 mM | 0.013 mM | 0.015 mM | |
| Recovery rate | 102.1% | 105.2% | 100.6% |
| Linearity | |||
| Calculated items | Results | ||
| Correlation coefficient of regression line | 1.000 | ||
| Slope of regression line | 4592 | ||
| y-Intercept of regression line | 821.1 | ||
| Residual sum of squares | 224800 | ||
| 95% confidence interval of y-intercept | −387.30≦ y0 ≦ 2030 | ||
| LOD and LOQ | |||
| LOD (µg/mL) | 0.3102 (Approx. 0.00078 mM) | ||
| LOQ (µg/mL) | 0.9399 (Approx. 0.0023 mM) | ||
DiMI solution (0.013 mM) was analyzed.
Data show mean of two experiments.
The results of absorption to sterile filter (4.5 cm2) are shown in Fig. 5. The DiMI content of 0.13 mM solution in the filtrate was increased depending on the flushing volume. Recovery rates of the DiMI content in filtrates at 30, 50, and 70 mL were 96.4, 97.6, and 98.1%, respectively. Therefore, we decided that the appropriate flushing volume should be 70 mL for 0.13 mM solution.
Data show mean ± standard error of the mean (S.E.M.) of three experiments.
The results of solubility of DiMI in water are shown in Fig. 6, indicating that DiMI contents in 2.5 and 13 mM solution against standard solution were 100% or over 100%, but 50 mM DiMI solution showed about 96% of DiMI content against the standard. From the result of 50 mM solution, the solubility of DiMI was calculated to be 1.92 g/100 mL. In addition, the results of solution time with 0.01 g of DiMI in 100 mL of water are shown in Fig. 7. The DiMI content in the solution with a mixing time of 2 h was 99.7% against the standard solution, being sufficient time to dissolve DiMI.
Data show mean ± S.E.M. of three experiments.
Data show mean ± S.E.M. of three experiments with 0.25 mM DiMI solutions.
Heat stress stability and photostability with 0.25 mM DiMI solution are shown in Table 2. In the heat stress stability test, the DiMI contents of the storage solution for two and four weeks did not change; therefore, DiMI in water was stable and not decomposed by heat stress (60 °C). In the photostability test, DiMI did not change, the same as heat stress stability.
| Heat stress stability | |||
|---|---|---|---|
| Initial | 2 weeks | 4 weeks | |
| DiMI contents | 96.7% | 96.5% | 96.5% |
| DiMI contents after heat stress against initial | — | 99.8% | 99.8% |
| Photostability | |||
| Shading | Cooling | Exposing | |
| DiMI contents | 97.3% | 99.3% | 98.7% |
| Mean | — | — | 97.5% |
| — | — | 98.8% | |
| — | — | 98.2% | |
| DiMI contents after light and UV stress against shading standard | — | — | 101.0% |
| DiMI contents after light and UV stress against shading and cooling standard | — | — | 98.9% |
DiMI solutions (0.25 mM) were tested in both stability tests.
We investigated the interaction between DiMI and additives because DiMI may be used as a preservative for ophthalmic solutions. We examined buffer components, isotonic agents, and surfactants often used in eye drops. The changes of DiMI contents and MIC values due to the coexistence of additives are shown in Fig. 8, Table 3, and Table 4, respectively. In addition, MIC values of DiMI and BAC were showed in Table 5. Regarding the samples mixed with buffer components, isotonic agents, and surfactants, DiMI contents almost showed no change, but a decrease in DiMI contents was observed in magnesium chloride and calcium chloride solutions. The MIC values of samples did not significantly change compared with those of DiMI alone. It is known that the MIC test has a deviation and MIC values from 1/2 to 2-fold are regarded as equivalent.15) On the other hand, we obtained some specific results. When DiMI was mixed with borate, phosphate, or surfactants, these MIC values were lower than those of DiMI alone against some microorganisms. In the comparison with MIC values of DiMI and BAC, both MIC values were almost equivalent or slightly higher than BAC.
Data shows mean ± S.E.M. of three experiments with 0.025 mM DiMI solutions.
| Additives | Strains | ||||
|---|---|---|---|---|---|
| S. aureus | E. coli | P. aeruginosa | C. albicans | A. brasiliensis | |
| Acetate | 0.008 | 0.016 | 0.016 | 0.008 | 0.063 |
| Borate | 0.004 | 0.016 | 0.016 | 0.008 | <0.001 |
| Citrate | 0.004 | 0.008 | 0.016 | 0.008 | 0.063 |
| Phosphate | 0.008 | 0.016 | 0.016 | 0.008 | >0.063 |
| Trometamol | 0.004 | 0.016 | 0.016 | 0.008 | 0.031 |
| Without additive | 0.008 | 0.016 | 0.016 | 0.008 | 0.063 |
| Additives | Strains | ||||
|---|---|---|---|---|---|
| S. aureus | E. coli | P. aeruginosa | C. albicans | A. brasiliensis | |
| Polysorbate 80 | <0.004 | 0.031 | 0.031 | 0.008 | 0.13 |
| Tyloxapol | <0.004 | 0.016 | 0.031 | 0.008 | 0.13 |
| Without additive | 0.016 | 0.031 | 0.031 | 0.031 | 0.13 |
| Additives | Strains | ||||
|---|---|---|---|---|---|
| S. aureus | E. coli | P. aeruginosa | C. albicans | A. brasiliensis | |
| DiMI | 0.016 | 0.031 | 0.031 | 0.031 | 0.13 |
| BAC | 0.009 | 0.037 | 0.074 | 0.019 | >0.15 |
The results of PET with 0.075, 0.025, 0.013, and 0.0025 mM DiMI solutions according to JP, USP, and EP are shown in Table 6. As there are two criteria, EP-A and EP-B, in PET of the EP method, we used both criteria to evaluate anti-microbial effect of DiMI. As the results, more than a 0.013 mM DiMI concentration in the formulations conformed to PET criteria in JP and USP, but a 0.075 mM of DiMI concentration was required to conform to EP criteria (both EP-A and EP-B). PET results of BAC solution are shown in Table 7. As the results, a 0.074 mM BAC concentration conformed to PET criteria in JP, USP, and EP-B, but a 0.029 mM of BAC concentration did not conform to all four criteria.
| Formulation | Criteria | |||
|---|---|---|---|---|
| JP | USP | EP-B | EP-A | |
| 0.075 mM | Pass | Pass | Pass | Pass |
| 0.025 mM | Pass | Pass | fail | fail |
| 0.013 mM | Pass | Pass | fail | fail |
| 0.0025 mM | fail | fail | fail | fail |
| Formulation | Criteria | |||
|---|---|---|---|---|
| JP | USP | EP-B | EP-A | |
| 0.15 mM | Pass | Pass | Pass | Pass |
| 0.074 mM | Pass | Pass | Pass | fail |
| 0.029 mM | fail | fail | fail | fail |
The influence of pH on the anti-microbial effect of DiMI formulation was evaluated, but the results of PET were not affected by the change of pH (data not shown), which corresponded to other reports of DiMI solution.16)
Cytotoxicity of DiMI and BACTo investigate the impact of DiMI and BAC on corneal epithelial cells, cell viability and cytotoxicity were assessed for HCE-T cells. The cells treated with 0.025 mM DiMI and 0.029 mM BAC solutions showed similar cell viability (Fig. 9A). The cytotoxicity levels increased by 10–25% were observed for 30 min after the administration of 0.025 mM DiMI, and the 0.029 mM BAC solution also caused some cytotoxicity. However, the injury level of BAC was lower than that of DiMI for 45 or 60 min after the administration (Fig. 9B).
(A) Cell viability. (B) Cytotoxicity. Data show mean ± S.E.M. of three experiments. 
: 0.025 mM DiMI, 
: 0.029 mM BAC.
In this study, we focused on DiMI in order to develop a new preservative for ophthalmic solutions instead of BAC. As the reason, DiMI has a nitrogen cation, the same as BAC, and the chemical reactivity of the cation included in the imidazole ring is considered to be lower than that of quaternary ammonium in BAC. At first, we investigated the physical and chemical properties of this compound. The solubility of DiMI at room temperature was 1.92 g/100 mL, categorized as “Sparingly soluble” in JP. The category indicated that it is relatively easy to prepare DiMI for ophthalmic solution as a preservative. From the heat stress experiment, DiMI may not be hydrolyzed under 60 °C because hydrolysis is generally enhanced as the temperature increases. As for the photostability, DiMI did not decompose under our experimental conditions, and this indicates that DiMI can be used as a preservative in natural light environments. Since the storage temperature and light condition should not be considered, DiMI can be applied to a large number of ophthalmic solutions. In addition, there is almost no interaction between DiMI and major additives of ophthalmic solutions (buffer components, isotonic agents, and surfactants), since neither the DiMI content nor anti-microbial activity was decreased. The results indicate that DiMI is easy to formulate as a preservative for ophthalmic solutions and has almost no restrictions on use with other additives. Exceptionally, DiMI and divalent metal ion should not coexist because the DiMI content was decreased in our experiments. Both DiMI and divalent metal ions have plus charges in solution; therefore, it is not considered that complexes of DiMI with metal ions are constructed. Although we do not understand the mechanism of the decrease at present, we need to pay attention on using them together.
We investigated the anti-microbial activity of DiMI by determining MIC. The activity of DiMI without other additives indicated that it was same as MIC value of BAC. On the other hand, the PET results with simple formulations showed that DiMI is more effective than BAC. We cannot mention a plausible reason for the difference of antimicrobial effect between DiMI and BAC in PET, but both compounds seem to express the antimicrobial effects with same mechanism that the nitrogen cation attacks cell membrane. Therefore, two alkyl chains or imidazole ring in DiMI may lead better effect than BAC. MIC values of DiMI did not change by coexistence with the above additives except for phosphate (Table 3). From these results about the physical and chemical properties plus the anti-microbial effect, DiMI may be a valuable preservative for ophthalmic use.
As shown in Fig. 9B, 0.029 mM BAC solution was safer than 0.025 mM DiMI solution after 30 min in terms of cytotoxicity. However, this toxicity test over 60 min was designed to observe differences between compounds. Generally, aqueous ophthalmic solutions flow out via the nasolacrimal duct within 5 min. In addition, we conducted a statistic evaluation with t-test. As the result, there is no statistically significant difference between DiMI and BAC cytotoxicity values at 45 and 60 min (p = 0.156 at 45 min, p = 0.159 at 60 min). We, therefore, considered that the toxicity of DiMI is the same as that of BAC. Considering the concentrations to pass JP, USP, and EP criteria for PET, the content of DiMI required for formulations would be lower than that of BAC. Therefore, DiMI is considered to actually be safer than BAC in prescriptions.
On the other hand, there are two shortcomings regarding the manufacture of products containing DiMI. One is that the solving time is too long for actual manufacturing because 0.25 mM DiMI solution needs two hours to dissolve in water. Although the solving time may be shortened depending on the concentration of DiMI, we considered that a marked reduction of the solving time would be difficult because of the slight wettability of DiMI. In the preparation of DiMI solution, DiMI powder drifted in water or floated on the surface. To overcome this, we tried an improvement with an ultra-sonic equipment, but it had no effect (data not shown). Regarding the stability of DiMI against heat stress, heat treatment may help to dissolve DiMI in water. Another defect is the flushing volume required for filtration. Since the flushing volume of 0.13 mM DiMI solution required for an effective filtration area (4.5 cm2) was 70 mL, about 0.0008 g of DiMI was needed for 1 cm2. If a manufacture process uses a filter with an about 6000-cm2 effective filtration area, 4.7 g of DiMI will be needed. Thus, about 94 L of 0.13 mM DiMI solution is required to perform flushing. In ophthalmic solutions, all pharmacopeia (JP, USP, and EP) recommend that the content of a preservative should be minimized to prevent negative actions. According to these recommendations, the DiMI content should be lower than 0.13 mM (for EP: 0.075 mM, for JP and USP: 0.013 mM, Table 6). The flushing volume, therefore, will need to be about 160 L (for EP) or 940 L (for JP and USP), which is very impractical.
In addition, there is little information on the toxicity of DiMI,17) although we showed some toxicological data in this study. To develop DiMI as a preservative for ophthalmic solutions, it will be necessary to obtain data on eye irritation, allergenicity, genotoxicity, and carcinogenicity etc. After assessment of the safety, DiMI may be widely adopted as a safe preservative and immediately contribute to the well-being of all patients.
We are grateful to Dr. Kaoru Sasaki (Riken BioResource Research Center) for kindly providing HCE-T cells. The authors thank E. Kashiwabara for in-vitro study and M. Ishikawa for laboratory assistance. This study was supported by members of Senju Pharmaceutical Co., Ltd.: T. Sekiya, A. Isowaki, H. Tokushige, T. Terai, A. Kawamura, and K. Ueda.
The authors declare no conflict of interest.
