Action of Disinfectant Quaternary Ammonium Compounds against Staphylococcus aureus

N-Alkyldimethylbenzylammonium chloride (consisting of a blend of C₁₄ [50%], C₁₂ [40%], and C₁₆ [10%] homologues) and didecyldimethylammonium chloride were obtained from the Stepan Company, IL. Hexadecyltrimethylammonium bromide (Cetrimide) was obtained from Sigma, Poole, Dorset, England. Biocides were prepared in sterile 0.05 M HEPES buffer (adjusted to pH 7.4) to give a final stock concentration of 200 ppm. Biocide preparations were maintained in a dark, cool storage cabinet and prepared daily. All glassware was pretreated with Decon⁹⁰ (BDH, Poole, Dorset, England) using the following procedure: immersion in 5% (vol/vol) Decon⁹⁰ for 24 h, washing in deionized water, and immersion for a further 24 h in 0.1 M HCl followed by a wash in deionized water.

Staphylococcus aureus (ATCC 6538) was obtained as a freeze-dried ampoule, and cells were rehydrated in nutrient broth (ICN Biomedicals, Aurora, OH). Stock S. aureus cells were conditioned for freezing (55) and maintained at −70°C (model no. MDF-U71V; Sanyo Electric Company).

Frozen bacterial samples were reconstituted monthly in 100 ml sterile nutrient broth and incubated at 37°C for 18 h. From the suspension, tryptone soya agar (TSA) (Oxoid, Basingstoke, Hampshire, England) isolation plates were incubated at 37°C for 24 h and stored at 4°C. The purity of the culture was verified by microscopic examination and Api STAPH (bioMerieux, Marcy l'Etoile, France). Single colonies removed from the isolation plates were used to inoculate 10 ml of nutrient broth, which was incubated at 37°C for 18 h. After further subculturing, the suspension was used to inoculate an AOAC (Becton Dickinson, Sparks, MD) agar plate (supplemented with 1.2% [wt/vol] agar granules [Oxoid technical agar no. 3]), which was maintained at 37°C for 18 h.

Following incubation, 20 ml sterile HEPES buffer was added to the surface of the AOAC plate, and the lawn of bacterial cells was gently scraped off. The cell suspension was then diluted further with HEPES buffer and washed three times by centrifugation at 1,508 × g for 20 min (Sorvall RT6000B; DuPont, Stevenage, Hertfordshire, England) before finally being resuspended in 30 ml HEPES buffer. Stock suspensions of S. aureus at 2.0 × 10¹⁰ CFU/ml were prepared by reference to a standardized calibration graph.

Final biocide concentrations ranging from 0 to 3.2 μg/ml (increasing in 0.1-μg/ml increments) in HEPES buffer were added to separate sterile petri dishes to which 19 ml of molten TSA (Oxoid, Basingstoke, Hampshire, England) was added at 50°C, mixed, and allowed to set.

The surface of each agar plate was inoculated in triplicate with 5 μl of challenge inoculum containing 10-fold dilutions from 1 × 10⁹ to 1 × 10⁵ CFU/ml in HEPES buffer. The plates were incubated at 35°C for 48 h. At 24 and 48 h, the presence of growth was examined visually on each plate, and the MIC was recorded. The MIC was determined as the concentration where no growth or <10 colonies were detected for each inoculation spot within a triplicate. Those cells growing in an inoculation spot where the MIC was recorded were subcultured and retested for MIC determinations. They exhibited no further resistance.

Final biocide concentrations studied were 35, 45, and 55 μg/ml, and these were prepared in HEPES buffer. Reaction vessels (100-ml wide-necked conical flasks) contained 36 ml biocide, and 4 ml of bacterial stock suspension was added (2 × 10¹⁰ CFU/ml). Experiments were conducted at 25°C and 35°C (shaking water bath at 100 oscillations min⁻¹ and an amplitude of 10 cm). Upon exposure of cells to biocide, 1-ml samples were removed and transferred to 9 ml AOAC letheen neutralizer broth (Becton Dickinson, Cockeysville, MD) every 30 s for up to 5 min inclusive. At the end of the experiment, a 10-fold dilution of the stock suspension into HEPES buffer followed by a further 10-fold dilution into neutralizer was performed to confirm the original viable count. A Bioscreen (Labsystems, Helsinki, Finland) was employed to estimate viable counts of all samples by recording the time taken for an unknown inoculum to grow to the threshold optical density, subsequently converting this value to a defined inoculum size by reference to a calibration curve. A sterile 100-well plate consisting of 10 columns, each containing 10 wells filled with 360 μl of nutrient broth into which samples were inoculated, was used according to the method described previously by Lambert et al. (27).

The spectrophotometric determination of quaternary ammonium surfactants with orange II dye (44) was modified to observe uptake isotherms for ADBAC and DDAC. Final concentrations of 4.5, 9, 12, 15, 25, and 35 μg/ml ADBAC and DDAC were prepared in 9 ml HEPES buffer. In addition, a 9-ml volume containing HEPES buffer only was subjected to the test. One milliliter of S. aureus stock suspension was added to each 9-ml preparation to give a final concentration of 2 × 10⁹ CFU/ml. The mixture was vortexed every 20 min for a total contact time of 1 h. After the contact period, the suspension was centrifuged (using glass centrifuge tubes) at 1,430 × g (Sorvall RC-5B; Dupont, Stevenage, England) for 15 min. Four milliliters of supernatant liquid was then removed and mixed with 1 ml orange II dye (4 × 10⁻⁴ M prepared in 0.1 M NaCl), followed by the addition of 5 ml chloroform. The mixture was vortexed for 30 s to ensure that the chloroform and dye were adequately mixed. Three milliliters from the bottom layer containing the dye-biocide complex (orange color) was removed into a UV silica cuvette (Hellma precision synthetic far-UV quartz, 200 to 2,500 nm; light path, 10 mm), and the absorbance was measured at 485 nm. A HEPES buffer control extracted in a similar way was used to blank the spectrophotometer (lambda 2; Perkin-Elmer, CT). Variables such as orange II concentration, absorption maxima for dye salts, extractability of dye salt using chloroform, and pH were validated by methods described previously by Scott (44). The assay was unaffected by changes in pH, buffer, and the presence of cell exudates.

Biocide uptake was calculated using calibration graphs of biocide concentrations versus the absorbance at 485 nm. Isotherm profiles comparing the equilibrium concentration (μg/ml) with the uptake concentration (μg/2 × 10⁹ cells) were plotted.

Final biocide concentrations of 4.5, 9, 12, 15, and 18 μg/ml in addition to a HEPES control were investigated against a final S. aureus challenge suspension of 2 × 10⁹ CFU/ml. Reaction vessels (glass 100-ml wide-necked conical flasks) contained 45 ml biocide plus 5 ml bacterial cells. Experiments were conducted at 25°C and 35°C in a shaking water bath (100 oscillations min⁻¹, with an amplitude of 10 cm). Before the introduction of cells to the biocide, a 10-fold dilution of the stock inoculum in HEPES buffer was made, and 5 ml of that suspension was passed through a 0.45-μm filter (Acrodisc [32 mm] with a Supor membrane; Pall Gelman, MI) into a 15-ml glass Bijou bottle. This filtrate represented the level of leakage from S. aureus at time zero; this procedure was performed before the inoculation of each reaction vessel.

After the addition of bacterial cells to the reaction vessel, 4-ml samples were removed and filtered into 15-ml glass screw-cap Bijou bottles at 0.5, 1.5, 3, 6, 9, 15, 30, 45, 120, and 240 min (the time points refer to the midpoint of filtration, which could take up to 20 s). Each sample was removed using a 10-ml sterile plastic syringe attached to a sterile plastic filling tube to enable easy access to the reaction mixture suspension. After each sample removal, the syringe was washed in sterile deionized water and reused. A separate syringe was employed for each reaction vessel. All filtrates were stored at −70°C until analysis.

Two additional vessels containing 2 × 10⁹ CFU/ml S. aureus suspended in 1 mM cetrimide (prewarmed to 37°C) and HEPES buffer were prepared and maintained for 1 h shaken (120 rpm) at 37°C (Innova incubator 4230 and platform shaker 2000; New Brunswick Scientific, Edison, NJ). These reaction vessels were established to determine the maximum intracellular potassium pool and to monitor leakage from cells in the absence of biocide, respectively.

To determine the soluble 260-nm-absorbing pool, treatment of cells with trichloroacetic acid was employed according to the protocol described previously by Gale and Folkes (15). The UV/VIS spectrophotometer (lambda 2; Perkin-Elmer, CT) was blanked with a 5% (wt/vol) trichloroacetic acid sample (exposed to the same conditions as the test samples). In the following experiments, material that leaked following biocide treatment was recorded as a percentage of the intracellular pool.

The potassium ion concentration in filtrate samples was determined using a Perkin-Elmer 1100B atomic absorption instrument in flame emission mode (wavelength, 766.5 nm; slit, 0.7 nm high; air-acetylene flame) (21). Before calibration and measurement of samples, the instrument was autozeroed with HEPES buffer, and this was repeated periodically throughout the analysis. The instrument was calibrated using potassium standards (analytical grade; Sigma-Aldrich, Poole, United Kingdom) of 100, 200, 300, 400, 500, and 600 μg/liter (final concentration) prepared in HEPES buffer. There was a linear relationship between emission and potassium concentration.

The filtrate samples were diluted in HEPES buffer to give potassium levels that could be detected at the midpoint on the calibration graph. The potassium standards were remeasured periodically during the experiment to verify instrument accuracy. Filtration and the presence of ADBAC, DDAC, and HEPES buffer did not interfere with potassium analysis.

For each filtrate sample, 260-nm-absorbing material was detected using a Perkin-Elmer lambda 2 UV/VIS spectrophotometer at a wavelength of 260 nm (42). A matched pair of UV silica cuvettes (Hellma precision synthetic far-UV quartz, 200 to 2,500 nm; light path, 10 mm) was used to measure each filtrate sample against HEPES buffer. After each measurement, the cuvette was washed in HEPES buffer.

Three separate reaction vessels containing an 18-ml volume of ADBAC, DDAC, and HEPES buffer were prepared. The final reaction concentration of each biocide was 9 μg/ml, and both reaction vessel and stock suspension (2 × 10¹⁰ CFU/ml) were maintained at 35°C. Two milliliters of the stock suspension was added to each reaction vessel to give a final concentration of 2 × 10⁹ CFU/ml. The reaction vessels were then maintained at 35°C for 4 h (Mickle shaking water bath at 100 oscillations min⁻¹, with an amplitude of 10 cm) to initiate autolysis. After 4 h, 20 ml of double-strength neutralizer (HEPES buffer consisting of 0.14% [wt/vol] lecithin and 1.0% [wt/vol] Tween 80) was added to each reaction vessel for 10 min. The reaction vessel contents were then centrifuged at 1,508 × g for 20 min (Sorvell RT6000B; DuPont, Stevenage, Hertfordshire, England) at room temperature. The pellet was then resuspended in 40 ml HEPES buffer, centrifuged, and finally resuspended in 20 ml HEPES buffer. The harvested cells were then held at 35°C for the duration of the experiment. Sample volumes of 3 ml were removed at 0, 2, 4, 18, 22, and 24 h and filtered (Acrodisc [32-mm, 0.45 μm] filter with a Supor membrane; Pall Gelman, MI), and the absorbance at 260 nm was measured.

To assess survivor levels, the procedure described above was followed, except that at 0, 2, 4, 6, 18, 22, 24, and 48 h, a 1-ml sample was removed from each reaction vessel and serially diluted in HEPES buffer, and 100 μl was plated onto TSA and incubated at 37°C. After 18 h of incubation, colony counts were recorded and converted to CFU/ml survivors.

The MIC remained stable after 48 h of incubation, and the MIC versus inoculum concentration was recorded (Table 1) to evaluate trends. The growth recorded within the triplicate inoculation spots was consistent under the defined conditions.

As shown in Table 1, MICs for ADBAC and DDAC appeared to be unchanged between 1 × 10⁵ and 1 × 10⁷ CFU/ml. Upon increasing the inoculum to 1 × 10⁸ CFU/ml, the MIC rose by 0.2 to 0.3 μg/ml. A sharp rise by almost double the MIC (1 × 10⁸ CFU/ml) occurred for ADBAC when the inoculum was increased to 1 × 10⁹ CFU/ml, and more than double the MIC was witnessed for DDAC.

Concentrations of 35, 45, and 55 μg/ml, when challenged with 2 × 10⁹ CFU/ml S. aureus, provided kill profiles for each biocide at 25°C and 35°C. A general overview shown in Fig. 3A and B indicates that DDAC was less temperature sensitive than ADBAC. ADBAC and DDAC profiles suggested an initial rapid interaction of biocide with cells inflicting killing, which eventually slowed.

After 5 min of contact time, the ADBAC concentrations appeared to have profoundly different endpoints, which were also affected by temperature. In comparison, the dialkyl quaternary ammonium compound (DDAC) suggested rapid killing in a short contact time at both 25°C and 35°C (Fig. 3B).

The two-sample t test assuming equal variance analysis at 25°C [t STAT = 8.58; P (T⇐t) two tail = 0.00035] and 35°C [t STAT = 2.51; P (T⇐t) two tail = 0.046] suggested that at both temperatures, a significant difference in killing between ADBAC and DDAC occurred, with DDAC being the more potent agent.

The concentration exponent values derived from a 2-log₁₀ (99%) reduction endpoint (23) were 1.6 and 1.8 for ADBAC and 1.6 and 1.2 for DDAC at 25 and 35°C, respectively (for the values at 25°C for ADBAC and DDAC, the r² values were 0.05). Hugo and Denyer (23) previously suggested that the concentration exponent should be largely independent of the level of killing. The concentration exponents for both ADBAC and DDAC were the same at 25°C. At 35°C, the concentration exponent for DDAC decreased, and that for ADBAC increased.

The Q₁₀ (Table 2) values suggested that upon increasing the temperature from 25°C to 35°C, the level of killing inflicted could be almost four times greater. The Q₁₀ appeared to be influenced by the biocide challenge concentration at 25°C and 35°C. The Q₁₀ value for DDAC confirms that it was less sensitive to temperature changes than ADBAC.

Fig. 4 indicates the uptake isotherms for ADBAC and DDAC when exposed to 2 × 10⁹ CFU/ml S. aureus. Upon applying the uptake classification scheme developed previously by Giles et al. (19), the ADBAC and DDAC isotherm profiles obey the L4 and H3/4 patterns, respectively.

ADBAC (Fig. 4A) showed Langmurian uptake to the cells with a possible subtle inflection point between 6 and 7 μg/2 × 10⁹ cells (marked A and B) and a final plateau (marked C) with a maximum uptake of around 14 μg/2 × 10⁹ cells. Upon increasing the biocide concentration beyond point C, little additional biocide was bound to cells. DDAC (Fig. 4B) possessed a very strong initial uptake, representing high-affinity binding, and over the biocide concentration range studied, the majority of biocide was bound to cells. An inflection point was clearly observed between 9 and 14 μg/2 × 10⁹ cells (marked A and B). Beyond point B, a further rapid uptake of DDAC by cells was observed. The inflection points for both ADBAC and DDAC may be due to additional binding sites arising by cell fragmentation or critical membrane damage exposing additional inner target regions.

Hugo and Longworth (24) previously described a procedure for calculating the approximate number of molecules required to form a biocide monolayer on the surface of a bacterial cell. The number of ADBAC monolayers per cell was calculated to be 1.2 at the completion of primary uptake (marked B in Fig. 4A). It appears that the ADBAC biocide could exercise its effect through monolayer or supramonolayer absorption. The ADBAC plateau sorption at point C (Fig. 4A) would represent approximately a two-monolayer uptake. The completion of primary uptake for DDAC occurred at the binding of 14 μg/2 × 10⁹ cells (marked B in Fig. 4B). The number of DDAC monolayers at the completion of primary uptake was calculated to be 1.9, and the sorption plateau at point C (Fig. 4B) represented approximately a 4.4-monolayer uptake.

Potassium leakage for both ADBAC (Fig. 5A) and DDAC (Fig. 5B) at concentrations of 15 μg/ml (0.0463 mM and 0.0460 mM, respectively) and above reached maximum pool levels within approximately 5 min of contact at 25°C. For DDAC, a 9-μg/ml (0.0276 mM) dose appeared to release potassium levels to completion over 45 min, compared to 9 μg/ml (0.0278 mM) ADBAC, which depleted the internal potassium pool by only 30%. Based on the concentrations studied, 9 μg/ml may be an important concentration differentiator for both ADBAC and DDAC; concentrations above 9 μg/ml (0.0278 mM and 0.0276 mM) for both agents appeared to rapidly release the entire internal potassium pool, and a 4.5-μg/ml (0.0139 mM and 0.0138 mM) dose induced leakage similar to that induced by the HEPES control. For ADBAC and DDAC at 35°C (data not shown), the leakage profiles were very similar to those obtained at 25°C, although leakage appeared more rapid at 35°C.

ADBAC- and DDAC-generated 260-nm-absorbing material leakage was monitored over 45 min at 25°C (Fig. 6A and B); it was rapid and reached maximum pool levels at concentrations above 9 μg/ml. Leakage at 9 μg/ml was gradual, and leakage at 4.5 μg/ml followed the leakage profile of the HEPES control. Upon increasing the ADBAC and DDAC reaction temperature to 35°C (data not shown), there was a rapid leakage of 260-nm-absorbing material compared to that at 25°C, and there was only modest leakage above the 100% pool level. This suggested that only largely unbound 260-nm-soluble pool material was detected at both temperatures.

The mean percent leakage of intracellular material at 30 min for both ADBAC and DDAC at 25°C (Fig. 7A and B) was compared to the biocide concentration (μg/ml). This approach was used in an attempt to determine the approximate concentrations where the initial membrane integrity is lost and subsequent catastrophic membrane collapse then occurs (11). A loss of membrane integrity occurs when the membrane loses conformity and leakage exceeds that of untreated cells. Catastrophic membrane collapse occurs when a complete loss of membrane integrity occurs upon exposure to biocide, and high levels of intracellular material are rapidly released. Figure 7 indicates concentrations where membrane damage of various severities may be assumed to occur. Upon examining this figure, the critical difference between the ADBAC and DDAC agents is the release of material induced at a concentration of 9 μg/ml (0.0278 mM and 0.0276 mM). For ADBAC, a loss of membrane integrity was induced at approximately 9 μg/ml (0.0278 mM), as opposed to DDAC, where the range lies just above 4.5 μg/ml (0.0138 mM). The catastrophic membrane collapse for both ADBAC and DDAC occurred above concentrations of 15 μg/ml (0.0463 mM) and 12 μg/ml (0.0368 mM), respectively. Figure 7 suggests that the ADBAC and DDAC concentrations studied do not cause major intracellular coagulation (39), although this may become evident at higher, uninvestigated concentrations.

Schematics representing the percent leakage of potassium (at 45 min) and 260-nm material (at 240 min) at 25°C and 35°C for both ADBAC and DDAC are presented in Fig. 8A to D. Figure 8A and B focus on data for ADBAC and DDAC at 25°C, respectively. Leakage above the cells' internal 260-nm pool material (100%) (indicated by a dotted line in the figures), suggestive of autolysis, was not convincing for either agent. A marked difference in 260-nm leakage was induced by a 9-μg/ml dose of either agent compared to 4.5 μg/ml, and this concentration boundary also had a mirrored effect on potassium leakage.

Figure 8C and D show that a temperature of 35°C encourages the autolytic processes. Figure 8C indicates that greater levels of autolysis are induced by ADBAC than by DDAC over a concentration range of 9 to 18 μg/ml. This also coincided with a release of >60% of intracellular potassium. Indeed, concentrations between 9 and 18 μg/ml for both ADBAC and DDAC inflicted almost maximum potassium leakage compared to concentrations of 4.5 μg/ml and the HEPES control. This observation is consistent with a relationship between potassium leakage and the onset of autolysis.

For ADBAC and DDAC at 35°C (Fig. 8C and D), statistical analyses were conducted to identify whether the biocide concentration was linked to the levels of autolysis recorded. For ADBAC and DDAC, a one-way analysis of variance was conducted to identify differences in the amounts of material that leaked at 4 h for each biocide concentration and the HEPES buffer control cells. Test analyses suggested that overall, a difference between biocide concentration existed for both ADBAC and DDAC, supported by the following outcomes: an F of 7.72 and a P value of 0.001 and an F of 6.19 and a P value of 0.002, respectively. Further analyses of the confidence interval plots and Tukey's pairwise comparison tests for ADBAC and DDAC suggested that two distinct response groups existed, HEPES buffer/4.5 μg/ml and the biocide range of 9 to 18 μg/ml, corresponding to no autolysis or induced autolysis, respectively, although within each group, there were no major differences in leakage performance.

Control cells maintained in HEPES buffer after exposure to biocide decreased in viability by approximately 30% over 24 h (Fig. 9A). This suggests that HEPES buffer causes only limited stress on cells. A survivor count at 48 h for control cells decreased by approximately 90%, so data comparisons were made over 24 h only. Upon the removal of biocide, the cell numbers for both ADBAC- and DDAC-exposed cells decreased. Gilbert and Brown (17) previously reported that centrifugation procedures used in the standardization of bacterial suspensions killed a significant proportion of cells depending on the strains and also injured cells, increasing their susceptibility to bactericidal activity.

Over 24 h, ADBAC-exposed cells in HEPES buffer were reduced by just over 90%, and the effect of autolysis on DDAC-pretreated cells was 99% killing. It should be noted that over the first 6 h, the reduction in the cell number for DDAC-pretreated cells appeared to be quicker than that for cells treated with ADBAC. Between the 6-h and 18-h time points, both groups of cells were reduced by almost 90%. At 22 or 24 h, the decrease in cell number was minor and possibly suggested no further reduction. At the point where no further loss of cells was observed, cell death may be attributed to bactericidal action, with sublethally injured cells being unable to recover on the nutrient-rich medium and cells committed to death via autolysis. It is difficult to allocate a quantitative (CFU/ml) value to each of the previously named causes of cell death. The remaining viable cells were able to survive exposure to a 9-μg/ml biocide dose. It was observed for ADBAC and DDAC that colonies at early time points varied in size, being either large (similar to HEPES control cells), which predominated, or small colonies. As the reaction time progressed, only large colonies were observed.

As shown in Fig. 9B, the HEPES control profile suggested that the breakdown of 260-nm-absorbing material was evident over 24 h, and this has been reported to occur in buffer environments (5). Leakage of 260-nm-absorbing material for cells exposed to ADBAC and DDAC exceeded that of the HEPES control cells, and this was largely due to the 4-h biocide contact period, prior to the removal of biocide. The leakage rates after biocide removal appeared to be the same irrespective of pretreatment conditions.

The agar dilution (MIC) technique indicated that ADBAC and DDAC appeared to inhibit bacterial growth at 48 h when challenged with S. aureus densities ranging between 1 × 10⁵ CFU/ml and 1 × 10⁹ CFU/ml, with MICs between 0.4 and 1.8 μg/ml. These values were within the MIC range quoted for commercial QACs, 0.5 to 5.0 μg/ml (35). Acheampong (1) previously investigated the action of pure ADBAC homologues with carbon chain lengths of C₉ and C₁₄ against Escherichia coli. Acheampong (1) suggested that when only small amounts of biocide agent were taken up by cells, increasing the cell concentration had little effect on the inhibition since sufficient free biocide was present in the original challenge dose. This observation was evident for ADBAC and DDAC over the concentration ranges of 1 × 10⁵ CFU/ml to 1 × 10⁷ CFU/ml, where the MICs were consistent for each agent. The increase in MICs at cell concentrations of >1 × 10⁸ CFU/ml could be attributed to an increasing reservoir of cellular negatively charged teichoic acids, which could interact with the cationic biocide molecules, reducing the available dose in solution.

Profiles of kill versus time at concentrations of 35, 45, and 55 μg/ml for ADBAC and DDAC at 25°C suggested an initial rapid interaction of biocide with cells, resulting in a 2- to 3-log₁₀ reduction within 3 min. Such a quick reduction in cell number has been previously reported for QACs (30, 31). Other authors (8, 53) have commented on the shape of kill curves, referring to them as sigmoidal, terminating at a point of no further killing. This trend was generally more apparent at 25°C than at 35°C in this investigation.

In this study, regarding ADBAC, an increase in temperature from 25°C to 35°C produced faster killing and eradicated the initial delay in killing observed at 25°C. A similar trend was observed in a previous study of QACs by Gélinas et al. (16), where the activity between 4°C and 50°C increased with rising temperature and longer contact times. Ahlström et al. (3) previously observed that amphiphilic betaine esters (QACs) at 0°C were less active than at 30°C even though binding affinities remained unchanged. The reduction in killing was believed to be due to the reduced fluidity of the outer membrane of Salmonella enterica serovar Typhimurium. A reduction in temperature and pH resulted in a more tightly packed lipopolysaccharide, thus reducing activity. Gilbert et al. (18) previously reported that as membranes pass through the transition from high to low temperatures, they become more crystalline. In this investigation, DDAC exhibited the greatest biocidal activity and appeared not to be influenced by temperature (47). Temperature trends could be described by Q₁₀ values, approximately 1 to 2 for DDAC. For ADBAC, the Q₁₀ values suggested that approximately three to four times the activity occurred at 35°C compared to that at 25°C. Gélinas et al. (16) previously reported a twofold increase in activity for ADBAC against S. aureus between 20°C and 37°C. Sundheim et al. (45) previously reported that increased activity was observed when the temperature of a QAC disinfectant used in the food industry was raised to between 30 and 40°C.

During a disinfection treatment procedure, the concentration exponent may be subject to minor fluctuations. Hugo and Denyer (23) presented η values for phenol, which showed that η decreased with higher endpoint values.

Upon increasing the temperature from 25°C to 35°C, the η values for DDAC decreased (1.6 and 1.8 for ADBAC and 1.6 and 1.2 for DDAC at 25 and 35°C, respectively [r² values of 0.05]). Tilley (49) previously determined η values for some biocides and reported that upon increasing the temperature across the range of 10°C to 40°C, η values decreased. As temperature was increased, the kill time to reach specific endpoints decreased. This could possibly suggest that upon increasing the temperature, a difference in the cell/biocide interaction occurred that assisted killing, resulting in lower concentration exponents. The concentration exponent values for ADBAC marginally increased with temperature.

Uptake isotherm studies for ADBAC and DDAC obeyed the L4 and H3/4 patterns, respectively. DDAC possessed a very strong binding affinity for S. aureus cells, and low residual levels in the supernatant liquid (equilibrium concentration) were recorded. This could explain the greater levels of killing experienced using the DDAC agent, and this could perhaps also be responsible for the significant slowing of killing after 5 min. The uptake isotherms and concentrations used in this study were also reported previously by Takasaki et al. (47, 48), who reported that DDAC was taken up rapidly, independently of the treatment temperature, and that the cytoplasmic membrane was the first uptake site, as the majority of the agent was distributed in the membrane and cytoplasmic fractions. The fact that temperature did not influence uptake might be inferred from the kill-survivor curves at 25°C and 35°C obtained in this study.

ADBAC at low concentrations exhibited moderate binding with the challenge inoculum, but at high concentrations, greater levels of biocide remained unbound. From the ADBAC uptake isotherm profile, there appeared to be a limit to the amount of ADBAC that could be adsorbed by cells, possibly due to the exhaustion of binding sites. This observation may be linked to the ADBAC kill curves, where a plateau effect was observed, suggesting no further killing of cells. Unbound ADBAC levels in the supernatant liquid could possibly be dominated by one single homologue chain length. Acheampong (1) previously observed that for ADBAC homologues, the C₁₄ had a higher affinity than the C₉ against E. coli under identical conditions.

The inflection points observed in the isotherms were not marked but could disclose a concentration level where additional target sites become exposed and are made available to ADBAC and DDAC due to critical membrane damage (11, 26). Diphasic isotherms with both primary and secondary saturation plateaus have been classified for S, L, and H isotherms (2, 41). This was true for ADBAC and DDAC, whose primary plateau uptakes (marked A and B) were between 6 and 7 μg/2 × 10⁹ cells and 9 and 14 μg/2 × 10⁹ cells, respectively, approximating the threshold concentrations required for bactericidal activity where the leakage of intracellular materials in excess of control levels became apparent.

Leakage for both ADBAC and DDAC occurred at bactericidal concentrations (>9 μg/ml) rather than at concentrations that appeared to inhibit bacterial growth (<4.5 μg/ml). Greater levels of leakage at 9 μg/ml were induced by DDAC (0.0276 mM) than by ADBAC (0.0278 mM), and this concentration potentially marked the boundary between a cell releasing internal pool material to completion or resisting biocide-induced membrane changes. Potassium leaked rapidly from cells compared to 260-nm-absorbing material at any given biocide concentration. In general, DDAC inflicted quicker leakage, and at low concentrations, greater amounts of material leaked than did ADBAC. This was evident when concentrations that are considered to achieve membrane permeabilization (ADBAC, >9 μg/ml [>0.0278 mM]; DDAC, >4.5 μg/ml [>0.0138 mM]) and catastrophic membrane collapse (ADBAC, 15 μg/ml [0.0463 mM]; DDAC, 12 μg/ml [0.0368 mM]) were achieved. It was also evident that cells exposed to bactericidal concentrations of ADBAC and DDAC immediately released leakage markers, albeit in different quantities; internal pool material was released to completion within short exposure times, exhibiting actions similar to those of chlorhexidine (24). Concentrations that induced potassium leakage only were not evident, even at concentrations that appeared to inhibit bacterial growth for ADBAC and DDAC.

Autolysis was evident for cells exposed to biocide when leakage of 260-nm-absorbing material into the exogenous environment exceeded the metabolic pool. This was thought to be due to the breakdown of RNA material, including rRNA, by activated RNases. Measurements of 260-nm-absorbing material released at 35°C confirmed that both ADBAC and DDAC initiated autolysis, with autolytic levels increasing with longer contact times (42).

A greater amount of 260-nm-absorbing material leaked at 35°C than at 25°C, suggesting a temperature-dependent autolytic process and evidence of an enzymatic reaction (42). Hugo and Bloomfield (22) previously reported that experiments at 4°C inhibited autolytic enzyme activity, with only metabolic pool leakage being recorded. For both ADBAC and DDAC, the concentration appeared to dictate whether autolysis would be initiated in cells. Autolysis was observed at concentrations of 9 μg/ml and above, and this coincided with an intracellular potassium depletion from cells of more than 30% (26). The HEPES control and a concentration of 4.5 μg/ml failed to initiate autolysis in the experimental time scale, with a corresponding low level of potassium leakage. Sykes (46) and Lamikanra and Allwood (28) previously reported the importance of potassium for maintaining ribosome stability and the internal ionic environment.

At 35°C, levels of autolysis were greater for ADBAC than for DDAC. However, for ADBAC, autolysis peaked at 12 μg/ml and then appeared to decrease with higher biocide concentrations, which is suggestive of biocide retarding the action of RNase enzymes or possibly impeding the further release of material. For DDAC, the level of autolysis appeared constant, but lower, over the concentration range studied. Salton (42) previously observed a concentration relationship between cetyltrimethylammonium bromide (CTAB) and 260-nm release at several contact time points. Although Salton (42) examined concentrations greater (up to 900 ppm) than those used in this study, it was observed that a reduction in 260-nm release occurred above 270 ppm. This observation appeared in the region of maximum CTAB uptake by the cells. Salton (42) observed that 260-nm-absorbing material on the CTAB-saturated cells partly accounted for the fall in 260-nm detection. Takasaki et al. (48) previously reported that intracellular macromolecules such as phospholipids, proteins, and nucleic acids leaked from S. aureus cells and could be precipitated with cationic surfactants. In this study, the ADBAC uptake isotherm revealed that unbound ADBAC at challenge concentrations of 12 μg/ml and above would have been present in the reaction vessel and may have interacted with free 260-nm-absorbing material that leaked from cells. DDAC possessed an H (high) uptake isotherm, and only very low levels of residual biocide would be free to interact with 260-nm-absorbing material; hence, no concentration effects were observed. Hugo and Longworth (24) previously found that secondary leakage from cells was reduced due to the inhibition of autolytic enzymes by chlorhexidine. It was also believed that the cell surface or the cytoplasmic membrane congealed, preventing leakage.

Cell survivors monitored over 24 h after the removal of ADBAC and DDAC suggested that autolysis makes a limited contribution to bactericidal activity. During exposure of cells to 9 μg/ml ADBAC or DDAC for 4 h, a proportion of cells would have been killed due to bactericidal activity. Other cells would have received a sublethal dose, although it would have been sufficient to induce autolysis. Some cells would have avoided bactericidal or autolytic inducing activities and would therefore have been able to survive, provided that a suitable recovery medium was employed. Upon the removal of the biocide, cells that experienced autolysis or that were too badly damaged to recover would not be able to replicate and were committed to death over the 24-h detection period, and those cells unaffected by biocide would survive. The surviving cells were possibly responsible for the stabilized survivor counts observed at around 18 to 24 h.

Cells preexposed to 9 μg/ml ADBAC or DDAC indicated an accelerated release of 260-nm-absorbing material above the metabolic pool, in comparison to cells maintained in HEPES buffer. This suggested that ADBAC and DDAC induce autolysis, although upon the removal of biocide, the leakage rate appeared to decrease. Over 24 h, irrespective of cell pretreatment, autolysis was observed and appeared to be an irreversible event, as observed previously by Hugo and Bloomfield (22).

We conclude that both ADBAC and DDAC are membrane-active agents that interacted with the cytoplasmic membrane in S. aureus, inducing the immediate leakage of intracellular constituents, confirming the primary lesion for both. Over the range of concentrations studied for leakage, MIC, and bactericidal investigations, DDAC was marginally more potent than ADBAC, but no outstanding differences in their properties were observed.

ADBAC and DDAC were both initiators of autolysis at low biocide concentrations (9 to 18 μg/ml), which, together with bactericidal activity, contributed to cell death. DDAC was less sensitive to temperature changes from 25°C to 35°C than ADBAC, and this was reflected in the Q₁₀ values.

The protocols employed and data observations reported in this study together with the calculated numerical descriptors of biocide action (Q₁₀ values and concentration exponents) will allow formulators and microbiologists using antiseptics and disinfectants to make a judicious selection of quaternary ammonium compound to achieve broad-spectrum activity and curb the threat of resistant outbreaks, especially in hospital settings. This will lead to improved antiseptic and disinfectant design while also considering the ultimate treatment of abiotic and biotic surfaces. These numerical efficacy parameters can also be used to better understand quaternary ammonium compound performance under a given set of conditions and advise on appropriate application whether for use as preservatives in pharmaceutical products or for general use in cleaning clinical or domestic and food environments.