Activity and mechanisms of action of selected biocidal agents on Gram-positive and -negative bacteria

Authors


S.E. Walsh, Leicester School of Pharmacy, De Montfort University, Hawthorn Building The Gateway, Leicester, LE1 9BH, UK (e-mail: swalsh@dmu.ac.uk).

Abstract

Aims: This study investigates the antimicrobial activity and mode of action of two natural products, eugenol and thymol, a commonly utilized biostatic agent, triclocarban (TCC), and two surfactants, didecyldimethylammonium chloride (DDDMAC) and C10–C16 alkyldimethyl amine N-oxides (ADMAO).

Methods and Results: Methods used included: determination of minimum inhibitory concentrations (MICs), lethal effect studies with suspension tests and the investigation of sub-MIC concentrations on growth of E. coli, Staph. aureus and Ps. aeruginosa using a Bioscreen microbiological analyser. Leakage of intracellular constituents and the effects of potentiating agents were also investigated. Only DDDMAC was bactericidal against all of the organisms tested. Eugenol, thymol and ADMAO showed bacteriostatic and bactericidal activity, but not against Ps. aeruginosa. TCC was only bacteristatic against Staph. aureus, but like the other agents, it did affect the growth of the other organisms in the Bioscreen experiments. All of the antimicrobial agents tested were potentiated by the permeabilizers to some extent and leakage of potassium was seen with all of the agents except TCC.

Conclusions: DDDMAC was bactericidal against all organisms tested and all compounds had some bacteriostatic action. Low level static effects on bacterial growth were seen with sub-MIC concentrations. Membrane damage may account for at least part of the mode of action of thymol, eugenol, DDDMAC and ADMAO.

Significance and Impact of the Study: The ingredients evaluated demonstrated a range of bactericidal and bacteriostatic properties against the Gram-negative and -positive organisms evaluated and the membrane (leakage of intracellular components) was implicated in the mode of action for most (except TCC). Sub-MIC levels of all ingredients did induce subtle effects on the organisms which impacted bacterial growth, even for those which had no true inhibitory effects.

Introduction

This study was initiated to investigate the antimicrobial activity and mode of action of eugenol, thymol, triclocarban (TCC), didecyldimethylammonium chloride (DDDMAC) and C10–C16 alkyldimethyl amine N-oxides (ADMAO).

Eugenol is an antimicrobial phenolic essential oil found in cloves (Blaszyk and Holley 1998; Vázquez et al. 2001). It is used in dentistry as a root canal sealer. Reinforced zinc oxide–eugenol cements have been shown to inhibit the growth of a range of organisms, including facultative anaerobes commonly isolated from infected root canals (Chong et al. 1994;Torabinejad et al. 1995; Siqueira and Gonçalves 1996; Kaplan et al. 1999). Eugenol has also been shown to be inhibitory to Echerichia coli O157:H7 (Blaszyk and Holley 1998), Penicillium citrinium (Vázquez et al. 2001) and human herpesvirus in vitro (Benencia and Courrèges 2000).

The essential oil thymol (found in thyme) has been commercially available as part of a mouthwash for more than 100 years (Scheie 1989). Thymol has inhibitory and biocidal effects on a range of bacteria including E. coli and Staphylococcus aureus (Cosentino et al. 1999), but not Pseudomonas aeruginosa (Sivropoulou et al. 1996). It is active against tobacco cutworm larvae (Hummelbrunner and Isman 2001; Isman et al. 2001) and has mosquitocidal properties (Mansour et al. 2000). Thymol has also been suggested as a replacement biocide to chloroform for the collection of rainwater samples (Gillett and Ayers 1991; Hadi and Cape 1995; Ayers et al. 1998).

Triclocarban (3,4,4′-trichlorocarbanilide; TCC) is used as a component of antibacterial soaps and has been shown to reduce the number of Staph. aureus in patients with moderately severe atopic dermatitis (Breneman et al. 2000). TCC is not active against Gram-negative organisms (Heinze and Yackovich 1988), but has bacteriostatic activity against methicillin-resistant Staph. aureus (MRSA) and vancomycin-resistant enterococcus (VRE) (Suller and Russell 1999).

The quaternary ammonium compound (QAC) DDDMAC has bactericidal activity and has been used for cattle-shed disinfection, wood preservation (Nishihara et al. 2000) and cellulosic string protection (Morrel and Smith 1995).

The surfactant ADMAO has been shown to have antimicrobial activity against Staph. aureus and Saccharomyces cerevisiae (Glover et al. 1999). The activity of ADMAOs increases with chain length up to a cut-off point of 14 (Birnie et al. 2000) and is thought to be linked to interaction with the cell membrane (Kopecká-Leitmanováet al. 1989).

Minimum inhibitory concentrations (MICs) and suspension tests were used to evaluate the activity of the biocides. The possibility of potentiating the action of the agents was investigated using ethylenediamine tetraacetic acid (EDTA), as the disodium salt and nerolidol. The integrity of the outer leaflet of the outer membrane of Gram-negative bacteria such as Ps. aeruginosa is maintained by hydrophobic lipopolysaccharide (LPS)–LPS and LPS–protein interactions. The presence of divalent cations, notably Mg2+, is essential for stabilizing the strong negative charges of the core oligosaccharide chain of the LPS molecules. EDTA binds these cations, thereby releasing up to 50% of the LPS molecules, and causing non-polar phospholipids associated with the inner membrane to be exposed at the cell surface so that hydrophobic molecules can now enter the cell (Russell and Chopra 1996). Nerolidol is an essential oil found in many flowers and plants that has been shown to have moderate activity against bacteria, yeast and fungi (Terreaux et al. 1994; Skaltsa et al. 2000).

Potassium leakage experiments were conducted to investigate the effect of the biocides on cell membranes (potassium leakage is an early indicator of membrane damage: Lambert and Hammond 1973).

The effect of sub-MIC levels of the compounds on bacterial growth may be important as low and/or residual concentrations may be present during some of their proposed uses. A Bioscreen microbial analyser (Labsystems, Helsinki, Finland) was used to monitor bacterial growth via absorbance readings associated with turbidity.

Materials and Methods

Biocides

Eugenol, thymol and triclocarban (TCC) were obtained from Sigma (Poole, UK). Biocides were obtained from Procter and Gamble (Cincinnati, USA). Due to the restricted solubility of eugenol, thymol and TCC in water, stock solutions were prepared with absolute ethanol (Fisher, Loughborough, UK). Final concentrations of ethanol in the presence of bacteria did not exceed 2·5% v/v. ADMAO and DDDMAC stock solutions were prepared with sterile deionized water. The neat ADMAO (23% aqueous solution containing 0·65% sodium bicarbonate) had a pH of 8·5.

Organisms

Escherichia coli American Type Culture Collection (ATCC) 10536, Staph. aureus ATCC 9518 and Ps. aeruginosa ATCC 15442 (European Standard BSEN 1276: 1997, suspension test strains) were inoculated into 10 ml of nutrient broth (Oxoid, London, UK) and grown for 18 h at 37°C in a water-bath shaking at 75 rev min−1, to give approximately 1 × 109 colony forming units (CFU) ml−1. An E. coli isolate (from the parent strain ATCC 10536) with an elevated agar MIC to DDDMAC was also used in some experiments (DDDMAC E. coli 1). This isolate was produced by streaking the parent strain on a nutrient agar plate and adding a filter disc that had been soaked in DDDMAC. Colonies growing within the outer inhibition zone were isolated of which one had a slightly elevated DDDMAC agar MIC (between twice and 10 times that of the parent strain).

Minimum inhibitory concentrations (MICs)

The activity of the biocides was examined using agar MICs. Bacterial numbers were verified by viable count and used to calculate the number of CFU μl−1. The mean CFU μl−1 (per spot) were as follows: E. coli 6 × 105, Staph. aureus 1·88 × 106 and Ps. aeruginosa 9·11 × 105. Petri dishes (Sterilin triple vented; Fisher) were prepared using double strength nutrient agar (Oxoid, UK), biocide and sterile deionized water to give a specific final concentration of biocide in single-strength agar. The maximum concentrations of biocide used were as follows: eugenol 0·1% v/v, thymol 1000 μg ml−1, TCC 5 μg ml−1, DDDMAC 20000 μg ml−1 and ADMAO 8% v/v. Final concentrations of ethanol in the petri-dishes did not exceed 2·5% (v/v) and control agar containing 2·5% ethanol was tested for bacteriostatic activity during each experiment. Control plates of single strength nutrient agar with no biocide were also prepared. The Petri dishes were overdried and a Denley multi-point inoculator was used to place 1-μl spots of each culture dilution on the agar surface. All plates were incubated overnight at 37°C. Any growth was recorded and compared with the growth of the controls. Between 8–10 repeat experiments were conducted.

Suspension tests

The suspension test method was based on the European Standard EN 1276: 1997. Simulated clean (0·3 g l−1 bovine albumin) and dirty (3 g l−1 bovine albumin) conditions were used with three of the four reference strains required (the fourth strain is Enterococcus hirae ATCC 10541 and was not included in this study) and the filter disc isolate DDDMAC E. coli 1. Hard water and diluent were not used. The standard of disinfection required by the test is a 5 log reduction in viable counts in 5 min at 20°C with all four strains. Owing to the nature of the test method, log reductions of <4 cannot usually be measured.

Potentiation of the activity of antimicrobial agents

EDTA.

Agar MICs were carried out in the presence of 0·5 mM EDTA to investigate whether it potentiates the activity of the antimicrobial agents. At this concentration EDTA was shown to have no inhibitory effect on the strains tested.

Nerolidol.

The possibility of potentiating the activity of the agents with 0·005% nerolidol was explored using agar MICs. Controls showed that 0·005% nerolidol in the presence of up to 5% ethanol (to improve the solubility of nerolidol) did not have a significant inhibitory effect when used in the absence of the antimicrobial agents.

Leakage of intracellular constituents

Potassium leakage.

Experiments were conducted to measure potassium (K+) release from cells after exposure to the antibacterial agents (Following MIC and suspension test results, only DDDMAC was investigated against Ps. aeruginosa). Bacteria were grown at 37°C on the surface of 12 nutrient agar plates for 18 h. Growth was removed using ultra-pure water. After centrifuging twice (2000 ×g), cells were finally resuspended in 600 ml of ultra-pure water. Antibacterial agent was added to the cultures (of final density 2–3 × 109 cfu ml−1) to give the desired final concentration in 50 ml. Samples were removed over a 30-min period, membrane-filtered and K+ measured in triplicate by means of an atomic absorption spectrophotometer (Instrumentation Laboratory, Barcelona, Spain). Controls exposed to water, 2·5% ethanol, heat treatment (100°C for 10 min) and lysostaphin (50 mg ml−1 for 30 min) were also conducted. The mean potassium (ppm) content for cells and water was subtracted from each time point of the biocide/ethanol/boiled/lysostaphin exposed results to eliminate background leakage.

Effects of sub-MIC levels on bacterial growth

The effects of sub-MIC levels of antibacterial agent on bacterial growth over 18 h at 37°C were investigated using a Bioscreen microbiological growth analyser (Labsystems). 100-well plates (Fisher) were prepared as follows: 50 μl of organism (107 CFU ml−1) were added to double strength nutrient broth, antimicrobial agent and sterile deionized water to give a range of agent concentrations in single strength broth. The plates were placed in the Bioscreen and incubated at 37°C for 18 h and absorbance measurements (540 nm) of each well were recorded every 10 min after 60 s of shaking.

Results

MICs

In the MIC experiments (Table 1), eugenol, thymol and ADMAO showed activity against E. coli and Staph. aureus, but not Ps. aeruginosa. TCC only showed activity against Staph. aureus and DDDMAC was active against all three organisms. The range of MIC values was fairly wide in some cases, although only one strain of each organism was tested (e.g. ATCC 10536 for E. coli).

Table 1.  Agar MICs of the antimicrobial agents following incubation overnight at 37°C
 Average (mode) MIC and range of MIC values
 E. coliStaph. aureusPs. aeruginosa
BiocideModeRangeModeRangeModeRange
  1. Growth of the organisms was observed on all of the control and ethanol control plates.

  2. *No variation occurred.

Eugenol (% v/v)0·050·005–0·050·10·01–0·1>0·1>0·1
Thymol (μg ml−1)500500 to >1000500500 to >1000>1000>1000
TCC (μg ml−1)>5>50·50·05–0·5>5>5
DDDMAC (μg ml−1)  55*55500100–500
ADMAO (% v/v)  21-40·0050·005>8>8

Suspension tests

Suspension test results for the three European Standard strains and the DDDMAC filter disc isolate with an elevated MIC to DDDMAC (DDDMAC E. coli 1) are shown in Table 2. Results are from three to eight repeat experiments.

Table 2.  Summary of suspension test results under simulated clean (0.3 g l−1 bovine albumin) and dirty (3 g l−1 bovine albumin) conditions at 20°C. Log reductions (± S.D.)
Antibacterial ExposureE. coliS. aureusPs. aeruginosaDDDMAC E. coli 1
agentConcentrationtime (min)CleanDirtyCleanDirtyCleanDirtyCleanDirty
Eugenol0·1% v/v 54·25 (0·91)<4 (0)<4 (0)<4 (0)4·33 (0·58)<4 (0)
Eugenol0·1% v/v304·34 (0·58)<4 (0)<4 (0)4·64 (0·35)
Eugenol0·05% v/v 5<4 (0)<4 (0)<4 (0)<4 (0)
Thymol1000 μg ml−1 55·17 (0·29)4·85 (0·26)5·04 (0·06)5·12 (0·20)<4 (0)5·26 (0·28)4·92 (0·14)
Thymol1000 μg ml−130<4 (0)
Thymol500 μg ml−1 54·31 (0·48)3·99 (0·67)4·05 (0·24)<4 (0)4·44 (0·83)4·08 (0·17)
TCCμg ml−1 5<4 (0)<4 (0)<4 (0)<4 (0)
TCCμg ml−130<4 (0)<4 (0)<4 (0)<4 (0)
DDDMAC20 000 μg ml−1 55·16 (0·27)4·90 (0·17)4·96 (0·06)5·22 (0·37)5·11 (0·19)5·07 (0·13)5·18 (0·30)5·08 (0·13)
DDDMAC500 μg ml−1 54·49 (0·52)4·99 (0·03)
DDDMACμg ml−1 54·71 (0·56)4·31 (0·62)3·92 (0·40)3·73 (0·35)4·55 (0·64)3·89 (0·29)
ADMAO8% v/v 53·69 (0·54)5·03 (0·05)4·60 (0·53)<4 (0)3·65 (0·60)
ADMAO8% v/v304·13 (0·61)<4 (0)3·75 (0·33)
ADMAO0·005% v/v 5<4 (0)<4 (0)

Eugenol.

Eugenol (0·1%, v/v) gave a 4 log reduction with E. coli under clean, but not dirty conditions. Decreasing the eugenol concentration to 0·05% v/v gave <4 log reduction. Results with Staph. aureus and Ps. aeruginosa showed <4 log reduction. Increasing exposure time to 30 min did not significantly increase measurable effectiveness.

Thymol.

Thymol (1000 μg ml−1) gave ≅5 log reduction with E. coli and Staph. aureus under clean conditions. Activity did not decrease markedly under dirty conditions and a 4 log reduction was still observed when the concentration was reduced to 500 μg ml−1. 1000 μg ml−1 of thymol were not effective against Ps. aeruginosa even after 30-min exposure.

TCC.

TCC did not cause a >4 log reduction under any of the conditions tested.

DDDMAC.

DDDMAC (20 000 μg ml−1) showed good activity against all three organisms within 5 min. A reduction of ≅ 5 log was seen with 20 000 μg ml−1 under clean and dirty conditions, and ≅4 to 5 log reductions were recorded at agar MIC concentrations (500 and 5 μg ml−1). Activity against DDDMAC-tolerant E. coli 1 was similar to that seen with the standard strain.

ADMAO.

Some activity was recorded against E. coli, with ≅4 log reduction in 5 min. Increasing the exposure time to 30 min did not significantly increase activity. Good activity (5 log reduction) was seen with Staph. aureus under clean conditions and most of this activity was retained under dirty conditions. A <4 log reduction was recorded with Ps. aeruginosa under all conditions tested.

Potentiation of the activity of antimicrobial agents

EDTA.

EDTA appeared to potentiate the activity of eugenol against Staph. aureus and Ps. aeruginosa, thymol against E. coli, Staph. aureus and Ps. aeruginosa, TCC against Staph. aureus, DDDMAC against Staph. aureus and Ps. aeruginosa and ADMAO against E. coli and Ps. aeruginosa (see Table 3).

Table 3.  Potentiation of antimicrobial agents by EDTA or nerolidol
 E. coli MIC (mode) ± potentiating agentStaph. aureus MIC (mode) ± potentiating agentPs. aeruginosa MIC (mode) ± potentiating agent
Antimicrobial agentNoneEDTANerolidolNoneEDTANerolidolNoneEDTANerolidol
Eugenol (% v/v)0·050·050·050·10·050·05>0·10·05>0·1
Thymol (μg ml−1)50010050050010050>1000500>1000
Triclocarban (μg ml−1)>5>5>50·50·050·1>5>5>5
DDDMAC (μg ml−1)5550515500100500
ADMAO (% v/v)20·140·0050·0050·005>848

Nerolidol.

The results (Table 3) indicated that 0·005% nerolidol can potentiate the activity of eugenol and thymol against Staph. aureus, but not E. coli or Ps. aeruginosa. The MIC of TCC for Staph. aureus was on average lower with nerolidol present, but was still within the range of MIC variation (0·05–0·5 μg ml−1). No potentiation was seen with TCC and E. coli or Ps. aeruginosa. Nerolidol did not potentiate the activity of DDDMAC against any of the micro-organisms tested. However, it did appear to reduce the action of DDDMAC against E. coli, increasing the average MIC from 5 to 50 μg ml−1. Nerolidol had a small potentiating effect on the activity of ADMAO against Ps. aeruginosa, but not E. coli or Staph. aureus.

Leakage of intracellular constituents

Potassium leakage.

ADMAO, thymol, eugenol and DDDMAC induced rapid K+ loss from E. coli (Fig. 1a) and Staph. aureus (Fig. 1b). No leakage was produced by TCC from Staph. aureus (Fig. 1b; E. coli was not tested). DDDMAC also induced rapid K+ loss from Ps. aeruginosa (Fig. 1c).

Figure 1.

Biocide-induced potassium release from (a) E. coli ATCC 10536, DDDMAC E. coli 1 (BE1), (b) Staph. aureus and (c) Ps. aeruginosa (symbols are the same, but DDDMAC concentrations increased from 5 and 10 μg ml−1 to 500 and 1000 μg ml−1, respectively)

Effects of sub-MIC levels on bacterial growth

Concentrations of eugenol, thymol, TCC, DDDMAC and ADMAO below the MIC slowed the growth rate of all four organisms tested (European Standard strains plus DDDMAC E. coli 1). Thymol, TCC, DDDMAC and ADMAO reduced the 18 h absorbance of all organisms tested and eugenol reduced the 18 h absorbance of all except DDDMAC E. coli 1. The highest concentration of eugenol prolonged the lag phase of Ps. aeruginosa. Thymol prolonged the lag phase of E. coli, Ps. aeruginosa and DDDMAC E. coli 1, TCC and ADMAO that of Staph. aureus and Ps. aeruginosa and DDDMAC of E. coli and DDDMAC E. coli 1.

In summary, sub-MIC levels of biocides had a marked effect on cell growth, decreasing the final absorbance achieved after 18 h and increasing the lag phase. Even TCC, which was not inhibitory (as measured by MIC determination) to Ps. aeruginosa, significantly affected its initial growth (Fig. 2a–e).

Figure 2.

Effects of antimicrobial agents on the growth of Ps. aeruginosa at 37°C. (a) Eugenol concentration (% v/v): ▪ 0, ▴ 0·1, • 0·05, × 0·01, + 0·005; (b) Thymol concentration (μg ml−1 ): ▪ 0, ▴ 1000, • 500, × 100, + 50; (c) TCC concentration (μg ml−1 ): ▪ 0, ▴ 5, • 1, × 0·5, + 0·1; (d) DDDMAC concentration (μg ml−1 ): ▪ 0, ▴ 1000, • 500, × 100, + 50; (e) ADMAO concentration (% v/v): ▪ 0, ▴ 8,• 4, × 2, + 1

The higher concentrations of DDDMAC caused a rapid increase in absorbance. This was probably because of interaction with the growth medium (interaction was also observed with agar MIC experiments; Fig. 2d).

Discussion

All five agents possessed bacteriostatic activity, but only DDDMAC was bacteriostatic to Ps. aeruginosa. The lack of activity of thymol against Ps. aeruginosa confirms previous findings (Sivropoulou et al. 1996; Cosentino et al. 1999), as does lack of activity of TCC against Gram-negative organisms (Heinze and Yackovich 1988).

During the suspension tests, thymol showed more activity than eugenol at the concentrations tested, with 5 log reductions against Staph. aureus as well as against E. coli. As expected from the MIC results, neither of these compounds showed bactericidal activity against Ps. aeruginosa. Bactericidal activity against the DDDMAC E. coli 1 isolate was not significantly different from that seen with the standard strain.

Potentiation by EDTA may be due to the easier access to target sites in the treated Gram-negative bacteria suggesting that at least one of the major target sites for all of the antimicrobial agents lies beyond the outer layers of the cell. The binding of cations by EDTA and subsequent weakening of the cell may also be responsible for the potentiation against Staph. aureus.

As nerolidol has some antibacterial activity, the potentiation may be due to a synergistic action against the cells. The effect was greatest when used with the other essential oils (eugenol and thymol) suggesting possible similarities in action and target site.

Leakage of K+ is the first indication of membrane damage in micro-organisms (Lambert and Hammond 1973). Our studies support the previous work (Kopecka-Leitmanova et al. 1989; Takasaki et al. 1994; Shapiro and Guggenheim 1995) that membrane disruption contributes to the mode of action of phenolics such as eugenol and thymol, to the QAC, DDDMAC and to ADMAO. TCC did not induce leakage.

The marked effects of sub-MIC levels of the agents on the growth of all of the bacteria tested demonstrated that even agents such as TCC can retard the growth of Ps. aeruginosa to some extent. Recent studies (Skandamis and Nychas 2001) have indicated that essential oils can alter the physico-chemical properties of meat as well as delaying microbial growth. It is possible that these effects are contributing to the increased lag phases and decreased final counts reported with eugenol and thymol in the Bioscreen experiments.

In conclusion, the activity of the compounds varied from being bacteriostatic against one organism (TCC and Staph. aureus via MIC method) to having bactericidal activity against all organisms tested (DDDMAC). All of the compounds at a sub-inhibitory concentration affected at least the lag phase of bacterial growth as measured by the Bioscreen. This might indicate an interaction of these agents with (a) primary non-vital bacterial target site(s) within the bacteria and further investigations should confirm this contention.

The E. coli isolate with a raised MIC to DDDMAC was still susceptible to higher concentrations and no cross-resistance was seen to the other agents. All of the antimicrobial agents tested were potentiated by the permeabilizers to some extent and leakage of potassium was seen with all of the agents except TCC. This suggests that the membrane is a site of action for thymol, eugenol, DDDMAC and ADMAO.

Ancillary