Broth and agar dilution methods have generally been developed for use with watersoluble preparations and require modification for use with essential oils of low water solubility. To ensure contact between the test organism and tea tree oil for the duration of the assay it is necessary to use a solubilising or emulsifying agent. The agents most commonly used are Tween 80 and Tween 20 at concentrations ranging from 0.001 to 20% (Beylier 1979; Walsh and Longstaff 1987; Chand et al. 1994; Carson et al. 1995b; Griffin et al. 1998). DMSO (Scortichini and Rossi 1991; Aboutabl et al. 1995), DMF (Kubo et al. 1991), ethanol (Morris et al. 1979; Deans and Svoboda 1988; Biondi et al. 1993) and 0.15-0.2% agar have also been used (Remmal et al. 1993; Mann and Markham 1998).
Use of emulsifying and solubilising agents may result in changes in the physicochemical properties of the test system, even though they have no antimicrobial activity when tested on their own. It has been reported that nonionic surfactants, such as the Tween compounds, form micelles above a concentration known as the critical micelle concentration. Lipophilic molecules, such as the components of tea tree oil, may become solubilised within the micelles and thus partitioned out of the aqueous phase of the suspension (Schmolka 1973). Kazmi and Mitchell (1978) have shown that antimicrobials solubilised within the surfactant do not contribute to the activity as they do not come into direct contact with the microorganisms. The amount of material solubilised increases at higher concentrations of surfactant (Van Doorne 1990). Results from our laboratory support this view: the antimicrobial activity of tea tree oil decreased as the concentration of Tween 20 in the test medium was increased from 0.1% to 5% (Mann and Markham 1997). Premixing of the oil with Tween before addition to the test medium would be expected to exacerbate this problem. The impact of such effects must also be considered in the formulation of pharmaceutical and cosmetic products, as the interaction between components may reduce the activity of the active ingredients.
Other studies suggest that low concentrations of surface-active agents in the mixture may actually enhance the activity of the antimicrobial through causing changes to the permeability of the cell membrane of the microorganism (Denyer and Baird 1990). This raises doubts as to the suitability of agents such as Tween 80 and 20 in the assay procedures. Van Doorne (1990) reports that ethanol, in concentrations as low as 5%, can have a marked potentiating effect on the activity of antimicrobial agents and these authors question its use as a solubilising agent.
To obtain consistent, reproducible results it is important that contact between the oil and the microorganism is maintained throughout the test period. Allegrini et al. (1973) reported that emulsions of essential oils in water containing 1% Tween 80 or Tween 20 disintegrated within one hour but that emulsions with 5% Tween were stable for 24 hours, the incubation period of many assays. However, as stated above, Tweens at this concentration exert an inhibitory effect. A more suitable dispersing agent for tests carried out in broth is bacteriological agar at concentrations of 0.15-0.2% (w/v) (Mann and Markham 1998;
Remmal et al. 1993). Emulsions containing 0.15% bacteriological agar were stable for 19 hours (Mann and Markham 1998). As well as the stability of the emulsion, the use of agar as a stabiliser has the advantages of its lack of chemical reactivity and, as reported in the study of Remmal et al. (1993), MIC values are lower than when Tween 20, Tween 80, Triton X100 or ethanol are included as emulsifiers or solubilisers.
To conclude this section, several aspects of the methods which have been described should be highlighted. Firstly, test conditions do not generally reflect the actual conditions of use of the preparation. Thus, an agent may be active when in direct contact with an organism in a liquid medium, but may be inactivated in the presence of blood or pus in a wound, or be unable to penetrate unbroken skin and hence not reach the contaminated site in sufficient concentration to be effective. Exposure time, particularly at concentrations near the MIC, may not be sufficient, particularly in external use, unless the agent has residual activity.
Secondly, some tests measure only inhibition of growth of the test organisms, not a lethal effect. Whilst the former are the tests most commonly reported in the literature, the link between these measurements, which generally involve prolonged contact (18 hours or more) between the organism and the agent, and in vivo use, are questionable. Even where such methods are adapted to also measure MBC no indication is given of the death kinetics (rate of kill).
Thirdly, the test methods which have been published in the literature have been developed for water-soluble compounds, and they do not always give reliable results with non-water-soluble oils and may underestimate the true antimicrobial activity. Adaptations are needed to ensure adequate and consistent contact between the oil and the test organism throughout the period of the test. The importance of standardisation of methods must be stressed.
Lastly, the method of formulation of the product can profoundly affect the physical and biological properties of the active agent. Thus, it cannot be assumed that a formulation will be effective simply because a specific amount of active ingredient has been included. This also highlights the need for microbiological testing of products to ensure that they are effective at inhibiting or eliminating microorganisms: chemical tests to measure concentrations of active ingredients are more precise, but do not, on their own, provide evidence of the efficacy of the product. Laboratory tests are important, but results of these tests can still only be regarded as a useful preliminary to clinical trials.
Was this article helpful?