Scientific Publications - Work Done by Microbiology Reader Bioscreen C

Journal of Applied Microbiology, 2000, Aug, 89(2), 275-279

Susceptibility testing inoculum size dependency  of inhibition using  the Colworth MIC technique

Lambert RJ
 

ABSTRACT

The minimum inhibitory concentration, MIC, is an accepted and well used criterion for measuring the susceptibility of organisms to inhibitors. Many factors influence the MIC value obtained, including temperature, inoculum size and type of organism. A modification of the method developed in this laboratory to obtain inhibition profiles of antimicrobials was used to examine the effect of inoculum size on the degree of inhibition observed with respect to inhibitor concentration. The data obtained enabled the production of an empirical model of inhibition, based on a Gompertz function, relating the level of growth observed to both the inoculum size and concentration of the inhibitor. The inoculum size dependencies of phenethyl alcohol, phenoxyethanol, p-chloro-m-cresol, trichloro-phenol, thymol and dodecyltrimethylammmonium bromide against Staphylococcus aureus were obtained.

INTRODUCTION

The minimum inhibitory concentration, MIC, is an accepted and well used criterion for measuring the susceptibility of organisms to inhibitors. In general, however, MIC is a qualitative indicator of the antimicrobial levels required to inhibit the growth of a particular organism. It is qualitative in the sense that the MIC is dependent on many factors, such as the temperature of incubation and the size of the test inoculum. Yet, many of these factors have not been actively researched; rather, the method has evolved to reduce their influences in an attempt to rationalize comparisons between inhibitors. Brown (1988) reported that using the agar dilution procedure, an inoculum size dependency on the MIC was observed when certain antibiotics and organisms were used. As the inoculum test size increased, so did the observed MIC. Riha and Solberg (1975), Eley and Greenwood (1981) and Borobio et al. (1986) have also reported inoculum size dependencies on the MIC value, amongst many others.

In current work on biocides, it has been possible to show that the standard European bacterial suspension tests suffer from an inoculum size dependency which the standard itself ignores ( Johnston et al. 2000). It has also been suggested that the non-linear tailing of log survivor-time graphs are caused either by a soil effect or by the level of microbes in solution ( Lambert and Johnston 2000). It was suggested that such effects should also be seen with preservatives. If this is the case, then it may be possible to show, as discussed by Denyer 1990, that a continuum exists between preservatives and biocides, i.e., there is no real demarcation line between one and the otherit is a matter of where you place the flag.

In addition to the determination of the MICs for biocides and disinfectants, other forms of susceptibility testing are also used to identify and describe effective preservation systems, and to validate preservation systems in products, e.g. challenge tests. A number of different approaches are used for susceptibility testing, with no internationally agreed protocols for the standardization of such procedures. The most commonly used approaches do not generally take account of effects of inoculum size. Many of these methods use relatively large numbers of cells. The methods adopted as standard allow the operator little freedom to vary the test in order to obtain other types of data. If the inoculum size has an effect on the value of the MIC, the standard tests will not show this. If such information was available, it may be possible to produce products with lower levels of preservative, if it is known, from other sources, the maximum level of possible contamination.

The purpose of the work described here was to explore the area between the non-inhibitory concentration, NIC, and the MIC ( Lambert and Pearson 2000) in the hope of developing a mathematical description of any inoculum size effect present.

MATERIALS AND METHODS

Phenethyl alcohol (PeA), phenoxy ethanol (PoE), p-chloro-m-cresol (pCmC), trichlorophenol (TCP), thymol and dodecyltrimethylammmonium bromide (C12QAC) were obtained from Sigma-Aldrich and used as received.

Preparation of bacterial suspensions

Staphylococcus aureus ATCC 6538 was grown overnight in a flask containing 80 ml Tryptone Soya Broth (TSB; Oxoid), shaking at 30 °C. The culture was centrifuged at 1660  g (Rotina 48R; Hettich, Tuttlingen, Germany) for 10 min. The resulting cell pellets were pooled and resuspended in 0·1% peptone water.

Method

A typical procedure is outlined below for the investigation of inoculum size dependency on the inhibition of Staph. aureus by phenethyl alcohol, PeA.

PeA (1·357 g) was made up to 100 ml with TSB (0·111 mol l ¹). A series of dilutions was performed, giving a range from 0 to 45 mmol l ¹ PeA in 5 mmol l ¹ steps. From an initial inoculum, I ₀ , of 6·3  10⁸ ml ¹, nine inocula dilutions were prepared: I ₀ , I _0/10, I _0/100, 6I _0/10, 6I _0/100, 6I _0/1000, I _0/4, I _{0/40 and} I _0/400. To each column of the Bioscreen plate was added a level of PeA, and each row was given a level of Staph. aureus inoculum. This gave a 10  9 matrix with the remaining row (row 10) left with broth/PeA only as the background controls for each column. The plate was incubated at 30 °C for 20 h. The analysis of the data was done in terms of fractional areas ( Lambert and Pearson 2000). In short, the area under the O.D./time curve of the test was calculated and compared with the area obtained by the control (zero antimicrobial added).

Model for Inoculum size dependency

The Gompertz function used to examine the inhibition profile models the observed data using two principle parameters: the inflexion point of the function, M, and the slope, B. It was hypothesized that any inoculum size effect would be manifested in these two parameters. By fitting the model to the data obtained for a number of initial starting inocula, the variation in M and B with inoculum size was obtained.

RESULTS

Phenethyl alcohol, PeA and Staph. aureus

Using the above method, the fractional areas were calculated (see Table 1 and Fig. 1). The results were analysed using the Gompertz function, Eq. 1 ( Lambert and Pearson 2000).

where fa= fractional area, B= slope parameter, x = log inhibitor concentration and M= inflexion point.The variation of M was fitted to a quadratic equation, Eq.2, where a, b and c are empirically derived constants.

For all inhibitors examined, the value of M increased with an increase in the initial inoculum size. The quadratic equation given above gave a good fit to all the inhibitors examined.

Two types of phenomenon were observed in relation to the magnitude of B. Either the value remained approximately constant, or B decreased with increasing initial inoculum size as was found for TCP and C12QAC.

Using the values obtained, Equation 1 was altered to accommodate the effect of inoculum size on the level of inhibition, Eq. 3.

M has been replaced by the quadratic expression and B is either single valued (the mean value obtained) or is another function expressing B in terms of logI ₀ . Figure 2 gives a plot of the calculated fa against logI ₀ and concentration, and allows a comparison with the observed values, Fig. 1. Table 2 summarizes the results obtained for examinations of PoE, thymol, pClmC, TCP and C12QAC against Staph. aureus.

Calculation of MIC

MIC was defined ( Lambert and Pearson 2000) as the interception of the line tangential to the point (M (A + Ce ¹)) with the fa = 0 line;

Replacing M by the quadratic expression obtained above allows the calculation of the MIC for any initial inoculum size within the boundaries of the experiment. Figure 3 plots the calculated MIC against logI ₀ for PeA and Staph. aureus. Although the model is constrained between the inoculum levels used to obtain the model, the model predicts that almost three times more PeA is required to inhibit 1  10⁸ Staph. aureus ml ¹ than the 39 mmol l ¹ quoted in texts ( Paulus 1993). The calculation does indicate that 39 mmol l ¹ will inhibit an inoculum of 3  10⁶ Staph. aureus ml ¹. This calculation simply suggests that knowing the inoculum size dependency of the inhibition may be necessary for a fuller description of the inhibitory effect itself. It should be noted that this model assumes that the inhibition is a continuous function of concentration, and that no change in the mechanism of inhibition occurs at higher concentrations.

Calculation of NIC

NIC was defined ( Lambert and Pearson 2000) as the interception of the line tangential to the point (M (A + Ce ¹)) with the fa = 1 line;

Again, the quadratic expression for M can be used to calculate the NIC for any initial inoculum size within the boundaries of the experiment.

Figure 3 shows the change in the MIC and NIC with the log of the inoculum challenge. On a logarithmic scale, the change in slope of the MIC with logI ₀ is very similar to that of the NIC with logI ₀ . This shows that as the inoculum challenge increases, the inhibition profile moves to the right of the log concentration/fa plot but retains its overall shape. This suggests that there is no apparent change in the mechanism of action of the preservative with increasing inoculum size.

Comparison of PeA and POE preservation of Staph. aureus

Zero growth curves can be constructed using Eq.4 for the inoculum size dependency of the inhibition of Staph. aureus by PeA and PoE. These curves describe the growth/no growth boundary which relates the inoculum challenge to the preservative concentration required to achieve zero growth relative to a control. The models suggest that on a molar basis, PeA in TSB is a better inhibitor than PoE against Staph. aureus. The literature ( Paulus 1993) suggests PoE to be the better inhibitor of the two. It is possible that differences in the initial inoculum size employed could have resulted in this conclusion, but the numbers used were not published.

DISCUSSION

The method used to obtain information on the inhibition of micro-organisms has been discussed previously ( Lambert and Pearson 2000). The method, in essence, is based on the comparison of the growth rate of a control inoculum with the test which contains an amount of a preservative or inhibiting substance. Modelling of the data was achieved using a Gompertz function. From the model, two principle concentrations were identified: the concentration where no growth occurred relative to the control, termed the MIC, and a concentration at which inhibition of growth begins, termed the NIC. Between the NIC and the MIC, there is an increasing level of inhibition as the level of inhibiting substance increases.

The establishment of an inoculum size dependence on the outcome of a disinfection test ( Johnston et al. 2000) suggested that an inoculum size effect should be seen with MIC experiments. It is also suggested that the results given here on the effect of inoculum size on the inhibition of growth have a similar underlying cause. If an inoculum size is decreases by half but retains the same level of inhibitor, there is now twice as much inhibitor per cell. If the inhibitor is not in a vast excess over the cellular contents, then this must have an effect on the level of inhibition observed.

Using the MIC test method developed, it was reasoned that the parameter used for the inflexion point of the Gompertz fit would be a sensitive indicator of any inoculum size dependency. The slope parameter, B, dictates the overall shape of the inhibition profile and was not expected, at first, to show any appreciable inoculum size dependency. In general, a large change in M with smaller changes in B was observed. This suggested that the inhibition profile was remaining constant, but was being shifted to the right of the log concentration/fa plot as the inoculum size increased. In the case of C12QAC, the slope parameter was also dependent on the inoculum size, with the slope decreasing with increasing inoculum challenge. Although the reason for this change is uncertain, it may be related to a combination of a biocidal and an inhibitory phenomenon, i.e., at lower inoculum challenges, the C12QAC can elicit a greater biocidal effect (or produce severe injury), whereas at higher inoculum challenges, there is a greater degree of quenching of the biocide, causing it to act more as a simple (sublethal) inhibitor. This method may therefore be quantifying the effect in the region between reversible and irreversible damage, or sublethal injury, to cell death, as discussed by Denyer (1990, 1995), Denyer and Stewart (1998) and Lambert and van der Ouderra (1999).

Knowledge of the inoculum-MIC effect could allow levels of preservative to be reduced in preserved products if, through other means, it was established that the product would never receive an inoculum challenge above a certain level. For example, let us consider a product which, due to good hygienic practice etc., will rarely, if ever, have a level of Staph. aureus in excess of 1  10⁴ ml ¹. If PeA was used as a preservative, a concentration 39 mmol l ¹ would be expected, as this is the published MIC. In this case, because we have knowledge of the inoculum size-PeA concentration dependency relationship, less than 20 mmol l ¹ of PeA would be required.

FIGURES

Fig. 1  Observed fractional areas, fa, for the inhibition of Staphylococcus aureus by PeA. ( ), 0-0·2 fa;...

Fig. 2  Calculated fractional areas, fa, for the inhibition of Staphylococcus aureus by PeA. ( ), 0-0·2 ...f...

Fig. 3  Calculated MIC and NIC of PeA (mmol l ¹) with respect to log of the initial inoculum size of Sta...

 
Table 1  PeA/(mmol l ¹) fractional areas

 
Table 2  Inhibition parameters

REFERENCES

• Borobio, M.V., Pascual, A., Dominguez, M.C., Perea, E.J. (1986) Effect of medium, pH and inoculum size on activity of ceftizoxime and Sch-34343 against anaerobic bacteria. Antimicrobial Agents and Chemotherapy 30, 626 627.

• Brown, W.J. (1988) National Committee for Clinical Laboratory Standards agar dilution susceptibility testing of anaerobic Gram-negative bacteria. Antimicrobial Agents and Chemotherapy 32, 385 390.

• Denyer, S.P. (1990) Mechanisms of action of biocides. International Biodeterioraton 26, 89 100.

• Denyer, S.P. (1995) Mechanisms of action of antibacterial biocides. International Biodeterioration and Biodegradation 36, 227 244.

• Denyer, S.P. & Stewart, G.S.A.B. (1998) Mechanisms of action of disinfectants. International Biodeterioration and Biodegradation 41, 261 268.

• Eley, A. & Greenwood, D. (1981) In vitro activity of ceftizoxime against Bacteroides fragilis: comparison with benzylpenicillin, cephalothin, and cefoxitin . Antimicrobial Agents and Chemotherapy 20, 332 335.

• Johnston, M.D., Simons, E.-A., Lambert, R.J.W. (2000) One explanation for the variability of the bacterial suspension test. Journal of Applied Microbiology 88, 237 242.

• Lambert, R.J.W. & Johnston, M.D. (2000) Disinfection kinetics: a new hypothesis and model for the tailing of log-survivor/time curves. Journal of Applied Microbiology 88, 907 913.

• Lambert, R.J.W. & Pearson, J. (2000) Susceptibility testing: accurate and reproducible Minimum Inhibitory Concentration, MIC, and Non-Inhibitory Concentration, NIC. Journal of Applied Microbiology 88, 784 790.

• Lambert, R.J.W. & van der Ouderra, M.L.-H. (1999) An investigation into the differences between the Bioscreen and the traditional plate count disinfectant test methods. Journal of Applied Microbiology 86, 689 694.

• Paulus, W. (1993) Microbicides for the Protection of Materials. pp. 31 32. London: Chapman & Hall.

• Riha, W.E. & Solberg, M. (1975) Clostridium perfringens inhibition by sodium nitrite as a function of pH, inoculum size and heat . Journal of Food Science 40, 439 442.

------- 8< -------
 

(Full Text online)

 

© 2005 Transgalactic Ltd (manufacturer of Bioscreen C software) | Privacy Statement | P.O. Box 1393, 00101 Helsinki, Finland, phone: +358 9 85172920, fax: +358 9 8749481, e-mail: microbiology@bionewsonline.com
 

Last modified: May 25, 2005