- Email: [email protected]
Development of a method to quantify in vitro the synergistic activity of “natural” antimicrobials
International Journal of Food Microbiology 85 (2003) 249 – 258 www.elsevier.com/locate/ijfoodmicro
Development of a method to quantify in vitro the synergistic activity of ‘‘natural’’ antimicrobials M. Dufour *, R.S. Simmonds, P.J. Bremer Microbiology Department, University of Otago, PO Box 56, 9001 Dunedin, New Zealand Received 8 July 2002; received in revised form 6 November 2002; accepted 10 November 2002
Abstract Despite numerous papers being published on the use of hurdle technology to control food-borne pathogens or spoilage organisms, there is no commonly accepted methodology to quantify the level of synergistic activity. This paper describes a method to quantify in vitro the synergistic activity of antibacterial agents against bacteria. Initially, a microtiter plate growth assay was used to determine the inhibitory concentrations of four ‘‘natural’’ antimicrobials (nisin, lauricidink, totarol, and the lactoperoxidase system (LPS)) against a panel of eight bacteria. Using the same microtiter system, the impact of various combinations of antimicrobials was assessed. The degree of synergy was based on the analysis of three criteria: (1) increase in lag phase, (2) reduction in culture density after 24 h, (3) and residual viability at 24 h. Only the lactoperoxidase system was active against all the Gram-positive and Gram-negative bacteria tested. Nisin, lauricidink, and totarol were only effective against the Gram-positive bacteria. The method successfully identified three combinations (nisin – lauricidink, LPS – nisin, and LPS – lauricidink) previously reported to have synergistic activity and highlighted the synergistic activity of two novel combinations (nisin – totarol and LPS – totarol). The development of a quick and reliable method to identify and quantify synergistic activity is a useful screening tool to establish preservative techniques that could have potential antimicrobial synergy in food-based systems. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Lactoperoxidase system; Nisin; Lauricidink; Monolaurin; Totarol; Natural antimicrobials; Synergism
1. Introduction Food-borne illness remains a major concern in industrialised countries. In the United States, 6 to 81 millions cases of food-borne illnesses each year are estimated to result in up to 9000 deaths (Gould et al.,
* Corresponding author. Tel.: +643-479-5411; fax: +643-4798540. E-mail address: [email protected] (M. Dufour).
1995; Roberts, 2000). In many foods, preservatives are required to maintain their quality, extend shelf life, and ensure safety (Wang, 1992). However, consumer preferences are moving towards foods that contain lower levels of chemical preservatives and exhibit more of the characteristics of fresh or natural products. Nisin, monolaurin, and lactoperoxidase are examples of ‘‘natural’’ preservatives currently in limited use. Although numerous studies have shown these ‘‘natural’’ preservatives to be effective against microorganisms, they have limitations, which include
0168-1605/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. doi:10.1016/S0168-1605(02)00544-5
250
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
the potential emergence of resistant strains, limited spectrum of activity, application costs, and their impact on the organoleptic quality of foods (Mazzotta and Montville, 1999). These limitations can, to an extent, be overcome by application of the hurdle technology concept, in which combinations of preservative factors (hurdles) are used to improve the microbial stability and the sensory quality of foods as well as their nutritional and economic properties (Leistner, 2000; Mansour and Milliere, 2001). The antibacterial activity of inhibitory compounds, such as nisin, monolaurin, and the lactoperoxidase system (LPS), can be enhanced if these compounds are combined with each other (Mansour et al., 1999; Mansour and Milliere, 2001; McLay et al., 2002), with chelating agents (Nikaido and Vaara, 1987; Stevens et al., 1991) or with preservative treatments such as high hydrostatic pressure, low pH, or freeze/ thaw cycles (Oh and Marshall, 1993; Roberts and Hoover, 1996; Garcia-Graelis et al., 2000; Cressy et al., 2003). Although there have been numerous papers published on the use of combinations of preservative methods to control food-borne pathogens or spoilage organisms, there is no commonly accepted methodology to detect or quantify synergistic interactions. Thus, there is a clear need to develop simple, effective, and widely applicable in vitro systems to identify and quantify synergistic combinations of ‘‘natural’’ antibacterial compounds. However, it should be noted that in vivo testing of synergistic combinations so identified would still be required. Methods, such as the epsilometer test, time-kill, checkerboard (Bonapace et al., 2000), and microgel diffusion assays (du Toit and Rautenbach, 2000), have been used to quantify synergistic interactions between clinical antibiotics, but are not always well suited for use with ‘‘natural’’ antimicrobials in food-based media. In recent years, the application of predictive microbiology to food-based applications has been greatly aided by the availability of affordable computers and automated analytical equipment (LopezMalo et al., 2000; Garcia-Gimeno et al., 2002). Unfortunately, while algorithms, such as the Gompertz equation, are useful in defining microbial growth under ideal conditions, they are not well suited to defining growth in foods (Shimoni and Labuza, 2000), when little or no growth has occurred (Sutherland et al., 1994), or in the presence of inhibitory
agents (Giannuzzi et al., 1999). Thus, the rate-limiting steps in the development of suitable synergistic combinations of ‘‘natural’’ antimicrobials remains twofold. First, the initial difficulties of isolating and characterising the bioactive compounds; and second, difficulties of testing and quantifying their synergistic interactions. The aim of the present study was to determine the activity of four ‘‘natural’’ antimicrobial agents used in combination and to develop a method suitable for defining the degree of their synergistic interactions.
2. Materials and methods 2.1. Bacterial strains and media The bacteria used were obtained from the New Zealand Reference Culture Collection: Listeria monocytogenes NZRM 44, Bacillus cereus NZRM 5, Streptococcus thermophilus TS2, Staphylococcus aureus NZRM 917, Escherichia coli O157:H7 strain NCTC 12900, Salmonella Typhimurium S411 (Ag Research MIRINZ), Yersinia enterocolitica ATCC9611 and Pseudomonas aeruginosa NZRM 2711. Bacterial strains were propagated for 24 h at 37 jC or 42 jC (St. thermophilus) in the following Difco media (Fort Richard Laboratories, New Zealand): tryptic soy broth (L. monocytogenes and Y. enterocolitica), Todd Hewitt broth (B. cereus and P. aeruginosa), brain heart infusion (S. aureus, E. coli O157, and St. typhimurium), and M17 broth (S. Thermophilus). Strains in regular use were maintained on agar plates at 4 jC and subcultured every 2 weeks; stock cultures were stored at 80 jC in skim milk glycerol. 2.2. Chemical stocks All stock solutions, unless stated otherwise, were purchased from Sigma (Missouri, USA) and filter sterilised (0.2 Am Supor Acrodisc, Gelman Sciences, USA). Nisin was prepared as a stock solution of 100 mg/ml in 0.02 M HCl. Lactoperoxidase enzyme (LPX, Tatua Biologics, Morrinsville, New Zealand) was prepared as a stock solution of 15 mg/ml in phosphate-buffered saline (pH 7). Lauricidink, a food grade glycerol monoester of lauric acid (Seolim,
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
Seoul, Korea) was prepared as a stock solution of 50 mg/ml in 95% ethanol. Totarol, or totarane diterpene (14-isopropyl-8,11,13-podocarpatrien-13-ol), was obtained from Dr. Greg Cook (Department of Microbiology, University of Otago, Dunedin, New Zealand) and prepared as a stock of 10 mg/ml in 95% ethanol. All solutions were stored at 20 jC until required. Glucose and glucose oxidase (GOD) were used to generate hydrogen peroxide in the lactoperoxidase system (LPS). Stock solutions of glucose, GOD, and sodium thiocyanate (NaSCN) were prepared in MilliQ water. 2.3. The lactoperoxidase system (LPS) The LPS comprised 1.466 ml of MilliQ water, 133 Al LPX stock (15 mg/ml), 133 Al NaSCN stock (167 mg/ml), 133 Al glucose stock (334 mg/ml), and sufficient GOD to attain an enzyme activity ratio of 1 unit GOD per 11.9 units of LPX. This gave a final LPX concentration in the system of 1 mg/ml and is expressed in this paper as a 1000 ppm concentration of the LPS. Therefore, dilutions of the LPS represent dilutions of the complete system. 2.4. Inhibition of bacterial strains by individual compounds The stock solution of nisin was diluted to concentrations of 5000, 2500, 1250, 500, 200, 100, 50, 10, 5, 1, and 0.5 ppm in MilliQ water and 100 Al volumes dispensed into the wells of a flat bottom 96-well microtiter plate (Nalgene NUNC International, Denmark). Lauricidink, totarol, and the LPS were similarly treated, but diluted to concentrations of: 250, 100, 50, 25, 10, 5, 2, 1, 0.5, and 0.1 ppm; 50, 25, 10, 5, 2, and 1 ppm; and 50, 25, 10, 5, 2.5, 1, 0.5, 0.25, and 0.1 ppm, respectively. Overnight cultures of the test organisms were diluted to approximately 1 104 cfu/ml in broth and 100 Al volumes added to each well. All tests were conducted in triplicate and controls included as appropriate. Microtiter plates were incubated at the optimum growth temperature for the test organism for 24 h in a plate reader (Multiskan Ascent Microtiterplate Reader, LabSystems, Finland) with absorbance readings (595 nm) taken every 2 h.
251
2.5. Inhibition of bacterial strains by combinations of compounds The concentrations of each compound used in combination studies are given in Table 1. Each concentration was selected on the basis of its ability to partially inhibit growth of the bacterium under study or if it had no effect on bacterial growth, the highest concentration tested was used. The appropriate combination of nisin, lauricidink, totarol, and the LPS were diluted to the desired concentrations in MilliQ water and 100 Al volumes dispensed into the wells of a flat bottom 96-well microtiter plate (Nalgene NUNC International). In all other respects, assays for combinations of compounds were performed as described for individual compounds. 2.6. Analysis of the data The absorbance readings obtained from the inhibitory assays were plotted against time to obtain the growth curves of the test organisms and to enable analysis of combinations showing synergistic activity. Individual inhibitory compounds and their combinations were scored out of a possible total score of 300 based on the analysis of three criteria: (a) the increase in lag phase, (b) the reduction in culture density at 24 h, and (c) the residual viability at 24 h. Lag phase was defined as the time required for the culture to reach a change in optical density (DOD) of 0.1. Increase in lag phase was determined as the time the test culture (treated sample) took to reach a DOD of 0.1 minus the
Table 1 Concentrations of antimicrobials selected for synergistic studies Bacterial strain
L. monocytogenes B. cereus St. thermophilus S. aureus E. coli O157 S. Typhimurium Y. enterocolitica P. aeruginosa
Selected concentration of antimicrobials Nisin (ppm)
Lauricidink (ppm)
Totarol (ppm)
LPS (ppm)
10 200 10 50 5000 5000 5000 5000
10 10 5 10 100 100 100 250
1 1 1 2 10 50 10 10
1 5 0.25 0.5 5 25 0.5 5
252
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
time required for the control culture (untreated sample) to reach a DOD of 0.1. The increase in lag phase was normalised by expressing it as a percentage of the running time of the experiment (24 h). For example, if the control and test cultures took 2 and 6 h, respectively, to reach a DOD of 0.1, the increase in lag phase of 4 h was represented as 16.7% of 24 h. The second criterion, reduction in culture density, was determined by subtracting the OD595 of the test culture (treated sample) after 24 h from the OD595 of the control culture (untreated sample). Reduction in culture den-
sity was normalised by expressing it as a percentage of the OD595 of the control culture. For example, if the ODcontrol and ODtest were 0.8 and 0.4, respectively, the reduction in culture density of 0.4 was represented as 50% of ODcontrol. The third factor, residual viability, was determined at 24 h by streaking out a loopful of each test culture (treated sample) onto appropriate bacteriological agar and incubating overnight at optimum growth temperature for the test organism. Complete loss of viability was given a score of 100% while a viable culture was given a score of 0%. The scores
Fig. 1. Effect of nisin, lauricidink, LPS and totarol on the growth of S. aureus. Panel A: Nisin added to a final concentration of (o) 0 ppm, ( ) 0.5 ppm, (D) 1 ppm, (n) 10 ppm, (E) 50 ppm, (5) 100 ppm, and (x) 200 ppm. Panel B: Lauricidink added to a final concentration of (o) 0 ppm, (5) 0.1 ppm, ( ) 0.5 ppm, (D) 1 ppm, (x) 5 ppm, (n) 10 ppm, and (E) 50 ppm. Panel C: LPS added to a final concentration of (o) 0 ppm, (w) 0.25 ppm, ( ) 0.5 ppm, (D) 1 ppm, (5) 2.5 ppm, (x) 5 ppm, (n) 10 ppm, (E) 25 ppm, and (X) 50 ppm. Panel D; totarol added to a final concentration of (o) 0 ppm, ( ) 2 ppm, (x) 5 ppm, (n) 10 ppm, (E) 25 ppm, and (X) 50 ppm. Standard error of mean indicated.
.
. .
.
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
of each criterion were summed to give a final score out of 300. For the purpose of determining the nature of the interaction between two inhibitory compounds, the following criteria were used. If the total score of the combination system was less than or equal to the highest score of either of the two individual compounds, the compounds were determined to have acted indifferently. If the total score of the combination system was greater than the sum of the total scores of the individual compounds, the compounds were determined to have acted synergistically. If the total score of the combination system lay between these two
253
breakpoints, the compounds were determined to have acted additively.
3. Results 3.1. Concentrations of antimicrobial compounds Nisin, lauricidink, totarol, and the LPS were screened against the selected test organisms to establish concentrations at which they provided a partial inhibition of growth of test strains. The curves obtained for S. aureus (Fig. 1) are representative of
Fig. 2. Untreated control (n), and the affect of lauricidink (o) and LPS (D), alone and in combination (x), on the growth of B. cereus (Panel A), S. aureus (Panel B), St. thermophilus (Panel C). Standard error of mean indicated.
254
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
the type of data obtained for all bacterial strains tested (individual graphs not shown). From Fig. 1, it can be seen that 50 ppm of nisin (Panel A), 10 ppm of lauricidink (Panel B), 0.5 ppm of LPS (Panel C),
and 2 ppm of totarol (Panel D) resulted in a partial inhibition of S. aureus growth; thus, these concentrations were selected for the subsequent combination experiments (Table 1). In a similar manner, inhibitory,
Table 2 Inhibition of bacterial growth by a combination of lauricidin and the LPS Criteria analysed for each bacteria
Score for each criteria
Highest individual score
Sum of scores for individual compounds
Score for combination system
Conclusion
92 100 100
134c
257d
292e
synergistic
17 20 0
92 100 100
39
76
292
synergistic
59 67 0
8 0 0
92 100 100
126
134
292
synergistic
L. monocytogenes Lag phase Culture density Viability
17 0 0
21 0 0
33 0 0
21
38
33
E. coli O157 Lag phase Culture density Viability
4 39 0
17 8 0
92 100 100
43
68
292
S. Typhimurium Lag phase Culture density Viability
4 31 0
13 46 0
13 53 0
59
94
66
additive
Y. enterocolitica Lag phase Culture density Viability
13 63 0
13 13 0
13 13 0
76
102
26
indifferent
P. aeruginosa Lag phase Culture density Viability
17 37 0
17 19 0
21 50 0
54
90
71
additive
La
LPSb
L + LPS
St. thermophilus Lag phase Culture density Viability
67 67 0
46 77 0
B. cereus Lag phase Culture density Viability
2 37 0
S. aureus Lag phase Culture density Viability
a
L: Lauricidink. LPS: Lactoperoxidase system. c Scores were obtained by summing scores for individual compounds, i.e., 67 + 67 + 0 = 134. d Scores were obtained by summing scores for both compounds, i.e., 67 + 67 + 0 + 46 + 77 + 0 = 257. e Scores were obtained by summing scores for the combination of compounds, i.e., 100 + 100 + 92 = 292. b
additive
synergistic
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
255
Fig. 3. Untreated control (n), and affect of nisin (o) and lauricidink (D), alone and in combination (x) on the growth of St. thermophilus (Panel A) and S. aureus (Panel B). Standard error of mean indicated.
but not lethal, concentrations were determined for all combinations of antimicrobials against each test organism (Table 1). 3.2. Inhibition of bacterial strains by combinations of compounds The graphs obtained from the inhibitory assays were analysed to quantify the synergistic activity between two inhibitory compounds. The combination of the LPS and lauricidink showed synergistic inhibition against all of the Gram-positive test strains
except L. monocytogenes (Fig. 2 and Table 2), as well as against the Gram-negative bacterium E. coli O157 (Table 2). Table 2 indicates the scores obtained by each bacterial strain tested against the LPS and lauricidink and is representative of the data analysis process conducted for each combination of inhibitory compounds. Nisin and lauricidink showed synergistic inhibition against St. thermophilus and S. aureus (Fig. 3 and Table 3) but showed no significant enhancement of inhibition against the other organisms (Table 3). Nisin and the LPS showed synergistic inhibition against B. cereus, S. aureus, St. thermophi-
Table 3 Scores obtained for combinations of inhibitory compounds Bacterial strains
Final scores for double combinations of inhibitory compounds Na + Lb
St. thermophilus B. cereus S. aureus L. monocytogenes E. coli O157 S. Typhimurium Y. enterocolitica P. aeruginosa a
S
292 29A 292S 13A 91A 62A 81A 28A
N + LPSc S
292 292S 100S 17A 94S 84A 17 15A
N + Td 8 33A 37 63A 46A 35A 114A 33S
N: Nisin. L: Lauricidink. c LPS: Lactoperoxidase system. d T: Totarol. S Combinations identified as acting synergistically are indicated with ‘‘S’’. A Combinations identified as acting additively are indicated with ‘‘A’’. b
L + LPS S
292 292S 292S 33A 292S 66A 26 71A
L+T
LPS + T
50 53 48 77 21 27 109 42
54 23 13A 63A 52A 54A 50A 119S
256
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
Fig. 4. Untreated control (n), and affect of nisin (o) and LPS (D), alone and in combination (x), on the growth of B. cereus (Panel A), S. aureus (Panel B), St. thermophilus (Panel C) and E. coli O157 (Panel D). Standard error of mean indicated.
lus, and E. coli O157 (Fig. 4 and Table 3) but had no significant effect against the remaining bacteria (Table 3). The combinations incorporating totarol showed no synergistic activity against the selected bacteria with the exception of the LPS – totarol and nisin –totarol combinations which showed synergistic inhibitory activity against P. aeruginosa (Table 3).
4. Discussion For use in food-based applications, it is desirable to identify combinations of inhibitory compounds that
have a wide rather than a narrow spectrum of inhibitory action and a persistent rather than a transient inhibitory effect (Gould, 2000). We examined these requirements by screening each compound and combination of compounds against a panel of four Grampositive and four Gram-negative bacteria selected on the basis of their significance as either food spoilage bacteria or food pathogens, and by testing for inhibition of bacterial growth over a 24-h period under conditions favourable for optimum growth. Testing each combination against a panel of strains of widely differing susceptibilities was also important to ensure that a potentially synergistic combination was not
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
missed. For any one organism, if the concentration of either of the antimicrobial agents being tested was a little too high or low, the ability of the test method to identify synergy was compromised (unpublished data). Although this could be overcome by testing a greater number of intermediate concentrations of each compound, we feel that it was adequately compensated for by the widely differing susceptibility of the strains used in the test panel. In order to quantify our results in terms of the synergistic potential of the combination systems, we chose to measure fundamental parameters relevant to the organisms potential to cause spoilage or disease, i.e., how long was the lag phase, what cell numbers were reached after 24 h, and was the population viable? By summing the unweighted normalised scores of three critical growth parameters, we generated a single numerical value that allowed the relative effectiveness of different combinations of inhibitory compounds to be evaluated. Although some authors have been careful to define their usage of the term synergy and to distinguish between synergistic and additive responses (ter Steeg et al., 1999), there is no clear agreement in the literature with regard to quantifiable definitions for these terms. We chose to define the nature of each interaction as either: indifferent, additive, or synergistic in a manner similar to that proposed by Bonapace et al. (2000) for evaluation of the synergistic interaction of classical antibiotics against Acinetobacter baumannii. We chose to distinguish combinations of compounds acting in an additive manner (where the degree of inhibition was greater than could be accounted for by the action of compound 1, or the action of compound 2, acting independently) from those showing a synergistic interaction (where the degree of inhibition was greater than could be accounted for by the action of compound 1 plus the action of compound 2) because we believe it is the truly synergistic interactions that are of principal interest in food preservation applications. In the present study, the application of this method clearly identified the previously established synergistic interactions between nisin and monolaurin, nisin and LPS, and monolaurin and LPS. Mansour et al. (1999) and Mansour and Milliere (2001) showed that while nisin or monolaurin were not effective against Bacillus strains, the combination of the two was synergistic and resulted in complete inhibition of
257
spore outgrowth in milk. In the present study, nisin and lauricidink were shown to have synergistic or additive effect against both Gram-positive and Gramnegative bacteria. Combinations of nisin and LPS have been reported to give synergistic activity against L. monocytogenes, both when these agents were added simultaneously (Zapico et al., 1998) and when added in two steps (Zapico et al., 1998; Boussouel et al., 2000). In the present study, this combination was identified as yielding additive or synergistic interactions against a wide range of organisms, confirming its potential for use in a number of food preservative applications. McLay et al. (2002) have previously reported that combinations of monolaurin and LPS resulted in synergistic inhibitory activity against S. aureus and E. coli O157, a result confirmed in the present study and extended to include a number of food pathogens and food spoilage organisms not previously examined. Totarol, a diterpenoid isolated from the root bark of Podocarpus nagi (Ying and Kubo, 1991), has been reported to have bactericidal activity against a diverse range of bacterial species, including methicillin-resistant strains of S. aureus (MRSA) (Muroi and Kubo, 1996), but has not previously been considered for use in food preservative applications. It was included in the present study for comparison against the established synergistic combinations. Although not as potent as the other combinations tested, combined with either nisin or the LPS, it did show a modest enhancement of activity against a number of organisms, particularly the Gram-negative organisms. By contrast, totarol showed no enhanced activity in combination with lauricidink against any organism.
5. Conclusion The use of a microtiter-based assay and automated data recording and processing enabled us to rapidly and effectively test the inhibitory potential of combinations of antimicrobial agents against a panel of eight bacteria. The differentiation of response into categories labelled as indifferent, additive, or synergistic allowed the readily identification of combinations having potential for use in multihurdle food preservation processes. The method successfully identified
258
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
three previously established effective combinations and was successful in identifying two novel inhibitory combinations (nisin – totarol and LPS – totarol) that may be useful in food-based applications.
Acknowledgements This work was supported in part by Tatua Cooperative Dairy (Morrinsville, New Zealand).
References Bonapace, C.R., White, R.L., Friedrich, L.V., Bosso, J.A., 2000. Evaluation of antibiotic synergy against Acinetobacter baumannii: a comparison with Etest, time-kill, and checkerboard methods. Diagn. Microbiol. Infect. Dis. 38, 43 – 50. Boussouel, N., Mathieu, F., Revol-Junelles, A.-M., Milliere, J.-B., 2000. Effects of combinations of lactoperoxidase system and nisin on the behaviour of Listeria monocytogenes ATCC 15313 in skim milk. Int. J. Food Microbiol. 61, 169 – 175. Cressy, H.K., Jerrett, A.R., Osborne, C.M., Bremer, P.J., 2003. A novel method for the reduction of cell numbers of Listeria monocytogenes by freezing in combination with an essential oil in bacteriological media. J. Food Prot. (in press). du Toit, E.A., Rautenbach, M., 2000. A sensitive standardised micro-gel well diffusion assay for the determination of antimicrobial activity. J. Microbiol. Methods 42, 159 – 165. Garcia-Gimeno, R.M., Hervas-Martinez, C., de Siloniz, M.I., 2002. Improving artificial neural networks with a pruning methodology and genetic algorithms for their application in microbial growth prediction in food. Int. J. Food Microbiol. 72, 19 – 30. Garcia-Graelis, C., Valckx, C., Michiels, C.W., 2000. Inactivation of Escherichia coli and Listeria innocua in milk by combined treatment of high hydrostatic pressure and the lactoperoxidase system. Appl. Environ. Microbiol. 66, 4173 – 4179. Giannuzzi, L., Contreras, E., Zaritzky, N., 1999. Modeling the aerobic growth and decline of Staphylococcus aureus as affected by pH and potassium sorbate concentration. J. Food Prot. 62, 356 – 362. Gould, G.W., 2000. Preservation: past, present and future. Br. Med. Bull. 56, 84 – 96. Gould, G.W., Abee, T., Granum, P.E., Jones, M.V., 1995. Physiology of food poisoning microorganisms and the major problems in food poisoning control. Int. J. Food Microbiol. 28, 121 – 128. Leistner, L., 2000. Basic aspects of food preservation by hurdle technology. Int. J. Food Microbiol. 55, 181 – 186. Lopez-Malo, A., Guerrero, S., Alzamora, S.M., 2000. Probabilistic modelling of Saccharomyces cerevisiae inhibition under the ef-
fects of water activity, pH, and potassium sorbate concentration. J. Food Prot. 63, 91 – 95. Mansour, M., Milliere, J.B., 2001. An inhibitory synergistic effect of a nisin – monolaurin combination on Bacillus sp. vegetative cells in milk. Food Microbiol. 18, 87 – 94. Mansour, M., Amri, D., Bouttefroy, A., Linder, M., Milliere, J.B., 1999. Inhibition of Bacillus licheniformis spore growth in milk by nisin, monolaurin, and pH combinations. J. Appl. Microbiol. 86, 311 – 324. Mazzotta, A.S., Montville, T.J., 1999. Characterisation of fatty acid composition, spore germination, and thermal resistance in a nisin-resistant mutant of Clostridium botulinum 169B and in the wild-type strain. Appl. Environ. Microbiol. 65, 659 – 664. McLay, J.C., Kennedy, M.J., O’Rourke, A.-L., Elliot, R.M., Simmonds, R.S., 2002. Inhibition of bacterial foodborne pathogens by the lactoperoxidase system in combination with monolaurin. Int. J. Food Microbiol. 73, 1 – 9. Muroi, H., Kubo, I., 1996. Antibacterial activity of anacardic acid and totarol, alone and in combination with methicillin, against methicillin-resistant Staphylococcus aureus. J. Appl. Bacteriol. 80, 387 – 394. Nikaido, H., Vaara, M., 1987. Outer membrane. In: Neidhardt, F.C. (Ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. American Society for Microbiology, Washington, DC, pp. 7 – 22. Oh, D.H., Marshall, D.L., 1993. Influence of temperature, pH, and glycerol monolaurate on growth and survival of Listeria monocytogenes. J. Food Prot. 56, 740 – 749. Roberts, J.A., 2000. Economic aspects of food-borne outbreaks and their control. Br. Med. Bull. 56, 133 – 141. Roberts, C.M., Hoover, D.G., 1996. Sensitivity of Bacillus coagulans spores to combinations of high hydrostatic pressure, heat, acidity and nisin. J. Appl. Bacteriol. 81, 363 – 368. Shimoni, E., Labuza, T.P., 2000. Modeling pathogen growth in meat products: future challenges. Trends Food Sci. Technol. 11, 394 – 402. Stevens, K.A., Sheldon, B.W., Klapes, N.A., Klaenhammer, T.R., 1991. Nisin treatment for inactivation of Salmonella species and other Gram-negative bacteria. Appl. Environ. Microbiol. 57, 3613 – 3615. Sutherland, J.P., Baylis, A.J., Roberts, T.A., 1994. Predictive modelling of growth of Staphylococcus aureus: the effects of temperature, pH and sodium chloride. Int. J. Food Microbiol. 21, 217 – 236. ter Steeg, P.F., Hellemons, J.C., Kok, A.E., 1999. Synergistic actions of nisin, sublethal ultrahigh pressure, and reduced temperature on bacteria and yeast. Appl. Environ. Microbiol. 65, 4148 – 4154. Wang, G.-H., 1992. Inhibition and inactivation of five species of foodborne pathogens by chitosan. J. Food Prot. 55, 916 – 919. Ying, B.-P., Kubo, I., 1991. Complete 1H and 13C NMR assignments of totarol and its derivatives. Phytochemistry 30, 1951 – 1955. Zapico, P., Medina, M., Gaya, P., Nunez, M., 1998. Synergistic effect of nisin and the lactoperoxidase system on Listeria monocytogenes in skim milk. Int. J. Food Microbiol. 40, 35 – 42.
Development of a method to quantify in vitro the synergistic activity of ‘‘natural’’ antimicrobials M. Dufour *, R.S. Simmonds, P.J. Bremer Microbiology Department, University of Otago, PO Box 56, 9001 Dunedin, New Zealand Received 8 July 2002; received in revised form 6 November 2002; accepted 10 November 2002
Abstract Despite numerous papers being published on the use of hurdle technology to control food-borne pathogens or spoilage organisms, there is no commonly accepted methodology to quantify the level of synergistic activity. This paper describes a method to quantify in vitro the synergistic activity of antibacterial agents against bacteria. Initially, a microtiter plate growth assay was used to determine the inhibitory concentrations of four ‘‘natural’’ antimicrobials (nisin, lauricidink, totarol, and the lactoperoxidase system (LPS)) against a panel of eight bacteria. Using the same microtiter system, the impact of various combinations of antimicrobials was assessed. The degree of synergy was based on the analysis of three criteria: (1) increase in lag phase, (2) reduction in culture density after 24 h, (3) and residual viability at 24 h. Only the lactoperoxidase system was active against all the Gram-positive and Gram-negative bacteria tested. Nisin, lauricidink, and totarol were only effective against the Gram-positive bacteria. The method successfully identified three combinations (nisin – lauricidink, LPS – nisin, and LPS – lauricidink) previously reported to have synergistic activity and highlighted the synergistic activity of two novel combinations (nisin – totarol and LPS – totarol). The development of a quick and reliable method to identify and quantify synergistic activity is a useful screening tool to establish preservative techniques that could have potential antimicrobial synergy in food-based systems. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Lactoperoxidase system; Nisin; Lauricidink; Monolaurin; Totarol; Natural antimicrobials; Synergism
1. Introduction Food-borne illness remains a major concern in industrialised countries. In the United States, 6 to 81 millions cases of food-borne illnesses each year are estimated to result in up to 9000 deaths (Gould et al.,
* Corresponding author. Tel.: +643-479-5411; fax: +643-4798540. E-mail address: [email protected] (M. Dufour).
1995; Roberts, 2000). In many foods, preservatives are required to maintain their quality, extend shelf life, and ensure safety (Wang, 1992). However, consumer preferences are moving towards foods that contain lower levels of chemical preservatives and exhibit more of the characteristics of fresh or natural products. Nisin, monolaurin, and lactoperoxidase are examples of ‘‘natural’’ preservatives currently in limited use. Although numerous studies have shown these ‘‘natural’’ preservatives to be effective against microorganisms, they have limitations, which include
0168-1605/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. doi:10.1016/S0168-1605(02)00544-5
250
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
the potential emergence of resistant strains, limited spectrum of activity, application costs, and their impact on the organoleptic quality of foods (Mazzotta and Montville, 1999). These limitations can, to an extent, be overcome by application of the hurdle technology concept, in which combinations of preservative factors (hurdles) are used to improve the microbial stability and the sensory quality of foods as well as their nutritional and economic properties (Leistner, 2000; Mansour and Milliere, 2001). The antibacterial activity of inhibitory compounds, such as nisin, monolaurin, and the lactoperoxidase system (LPS), can be enhanced if these compounds are combined with each other (Mansour et al., 1999; Mansour and Milliere, 2001; McLay et al., 2002), with chelating agents (Nikaido and Vaara, 1987; Stevens et al., 1991) or with preservative treatments such as high hydrostatic pressure, low pH, or freeze/ thaw cycles (Oh and Marshall, 1993; Roberts and Hoover, 1996; Garcia-Graelis et al., 2000; Cressy et al., 2003). Although there have been numerous papers published on the use of combinations of preservative methods to control food-borne pathogens or spoilage organisms, there is no commonly accepted methodology to detect or quantify synergistic interactions. Thus, there is a clear need to develop simple, effective, and widely applicable in vitro systems to identify and quantify synergistic combinations of ‘‘natural’’ antibacterial compounds. However, it should be noted that in vivo testing of synergistic combinations so identified would still be required. Methods, such as the epsilometer test, time-kill, checkerboard (Bonapace et al., 2000), and microgel diffusion assays (du Toit and Rautenbach, 2000), have been used to quantify synergistic interactions between clinical antibiotics, but are not always well suited for use with ‘‘natural’’ antimicrobials in food-based media. In recent years, the application of predictive microbiology to food-based applications has been greatly aided by the availability of affordable computers and automated analytical equipment (LopezMalo et al., 2000; Garcia-Gimeno et al., 2002). Unfortunately, while algorithms, such as the Gompertz equation, are useful in defining microbial growth under ideal conditions, they are not well suited to defining growth in foods (Shimoni and Labuza, 2000), when little or no growth has occurred (Sutherland et al., 1994), or in the presence of inhibitory
agents (Giannuzzi et al., 1999). Thus, the rate-limiting steps in the development of suitable synergistic combinations of ‘‘natural’’ antimicrobials remains twofold. First, the initial difficulties of isolating and characterising the bioactive compounds; and second, difficulties of testing and quantifying their synergistic interactions. The aim of the present study was to determine the activity of four ‘‘natural’’ antimicrobial agents used in combination and to develop a method suitable for defining the degree of their synergistic interactions.
2. Materials and methods 2.1. Bacterial strains and media The bacteria used were obtained from the New Zealand Reference Culture Collection: Listeria monocytogenes NZRM 44, Bacillus cereus NZRM 5, Streptococcus thermophilus TS2, Staphylococcus aureus NZRM 917, Escherichia coli O157:H7 strain NCTC 12900, Salmonella Typhimurium S411 (Ag Research MIRINZ), Yersinia enterocolitica ATCC9611 and Pseudomonas aeruginosa NZRM 2711. Bacterial strains were propagated for 24 h at 37 jC or 42 jC (St. thermophilus) in the following Difco media (Fort Richard Laboratories, New Zealand): tryptic soy broth (L. monocytogenes and Y. enterocolitica), Todd Hewitt broth (B. cereus and P. aeruginosa), brain heart infusion (S. aureus, E. coli O157, and St. typhimurium), and M17 broth (S. Thermophilus). Strains in regular use were maintained on agar plates at 4 jC and subcultured every 2 weeks; stock cultures were stored at 80 jC in skim milk glycerol. 2.2. Chemical stocks All stock solutions, unless stated otherwise, were purchased from Sigma (Missouri, USA) and filter sterilised (0.2 Am Supor Acrodisc, Gelman Sciences, USA). Nisin was prepared as a stock solution of 100 mg/ml in 0.02 M HCl. Lactoperoxidase enzyme (LPX, Tatua Biologics, Morrinsville, New Zealand) was prepared as a stock solution of 15 mg/ml in phosphate-buffered saline (pH 7). Lauricidink, a food grade glycerol monoester of lauric acid (Seolim,
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
Seoul, Korea) was prepared as a stock solution of 50 mg/ml in 95% ethanol. Totarol, or totarane diterpene (14-isopropyl-8,11,13-podocarpatrien-13-ol), was obtained from Dr. Greg Cook (Department of Microbiology, University of Otago, Dunedin, New Zealand) and prepared as a stock of 10 mg/ml in 95% ethanol. All solutions were stored at 20 jC until required. Glucose and glucose oxidase (GOD) were used to generate hydrogen peroxide in the lactoperoxidase system (LPS). Stock solutions of glucose, GOD, and sodium thiocyanate (NaSCN) were prepared in MilliQ water. 2.3. The lactoperoxidase system (LPS) The LPS comprised 1.466 ml of MilliQ water, 133 Al LPX stock (15 mg/ml), 133 Al NaSCN stock (167 mg/ml), 133 Al glucose stock (334 mg/ml), and sufficient GOD to attain an enzyme activity ratio of 1 unit GOD per 11.9 units of LPX. This gave a final LPX concentration in the system of 1 mg/ml and is expressed in this paper as a 1000 ppm concentration of the LPS. Therefore, dilutions of the LPS represent dilutions of the complete system. 2.4. Inhibition of bacterial strains by individual compounds The stock solution of nisin was diluted to concentrations of 5000, 2500, 1250, 500, 200, 100, 50, 10, 5, 1, and 0.5 ppm in MilliQ water and 100 Al volumes dispensed into the wells of a flat bottom 96-well microtiter plate (Nalgene NUNC International, Denmark). Lauricidink, totarol, and the LPS were similarly treated, but diluted to concentrations of: 250, 100, 50, 25, 10, 5, 2, 1, 0.5, and 0.1 ppm; 50, 25, 10, 5, 2, and 1 ppm; and 50, 25, 10, 5, 2.5, 1, 0.5, 0.25, and 0.1 ppm, respectively. Overnight cultures of the test organisms were diluted to approximately 1 104 cfu/ml in broth and 100 Al volumes added to each well. All tests were conducted in triplicate and controls included as appropriate. Microtiter plates were incubated at the optimum growth temperature for the test organism for 24 h in a plate reader (Multiskan Ascent Microtiterplate Reader, LabSystems, Finland) with absorbance readings (595 nm) taken every 2 h.
251
2.5. Inhibition of bacterial strains by combinations of compounds The concentrations of each compound used in combination studies are given in Table 1. Each concentration was selected on the basis of its ability to partially inhibit growth of the bacterium under study or if it had no effect on bacterial growth, the highest concentration tested was used. The appropriate combination of nisin, lauricidink, totarol, and the LPS were diluted to the desired concentrations in MilliQ water and 100 Al volumes dispensed into the wells of a flat bottom 96-well microtiter plate (Nalgene NUNC International). In all other respects, assays for combinations of compounds were performed as described for individual compounds. 2.6. Analysis of the data The absorbance readings obtained from the inhibitory assays were plotted against time to obtain the growth curves of the test organisms and to enable analysis of combinations showing synergistic activity. Individual inhibitory compounds and their combinations were scored out of a possible total score of 300 based on the analysis of three criteria: (a) the increase in lag phase, (b) the reduction in culture density at 24 h, and (c) the residual viability at 24 h. Lag phase was defined as the time required for the culture to reach a change in optical density (DOD) of 0.1. Increase in lag phase was determined as the time the test culture (treated sample) took to reach a DOD of 0.1 minus the
Table 1 Concentrations of antimicrobials selected for synergistic studies Bacterial strain
L. monocytogenes B. cereus St. thermophilus S. aureus E. coli O157 S. Typhimurium Y. enterocolitica P. aeruginosa
Selected concentration of antimicrobials Nisin (ppm)
Lauricidink (ppm)
Totarol (ppm)
LPS (ppm)
10 200 10 50 5000 5000 5000 5000
10 10 5 10 100 100 100 250
1 1 1 2 10 50 10 10
1 5 0.25 0.5 5 25 0.5 5
252
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
time required for the control culture (untreated sample) to reach a DOD of 0.1. The increase in lag phase was normalised by expressing it as a percentage of the running time of the experiment (24 h). For example, if the control and test cultures took 2 and 6 h, respectively, to reach a DOD of 0.1, the increase in lag phase of 4 h was represented as 16.7% of 24 h. The second criterion, reduction in culture density, was determined by subtracting the OD595 of the test culture (treated sample) after 24 h from the OD595 of the control culture (untreated sample). Reduction in culture den-
sity was normalised by expressing it as a percentage of the OD595 of the control culture. For example, if the ODcontrol and ODtest were 0.8 and 0.4, respectively, the reduction in culture density of 0.4 was represented as 50% of ODcontrol. The third factor, residual viability, was determined at 24 h by streaking out a loopful of each test culture (treated sample) onto appropriate bacteriological agar and incubating overnight at optimum growth temperature for the test organism. Complete loss of viability was given a score of 100% while a viable culture was given a score of 0%. The scores
Fig. 1. Effect of nisin, lauricidink, LPS and totarol on the growth of S. aureus. Panel A: Nisin added to a final concentration of (o) 0 ppm, ( ) 0.5 ppm, (D) 1 ppm, (n) 10 ppm, (E) 50 ppm, (5) 100 ppm, and (x) 200 ppm. Panel B: Lauricidink added to a final concentration of (o) 0 ppm, (5) 0.1 ppm, ( ) 0.5 ppm, (D) 1 ppm, (x) 5 ppm, (n) 10 ppm, and (E) 50 ppm. Panel C: LPS added to a final concentration of (o) 0 ppm, (w) 0.25 ppm, ( ) 0.5 ppm, (D) 1 ppm, (5) 2.5 ppm, (x) 5 ppm, (n) 10 ppm, (E) 25 ppm, and (X) 50 ppm. Panel D; totarol added to a final concentration of (o) 0 ppm, ( ) 2 ppm, (x) 5 ppm, (n) 10 ppm, (E) 25 ppm, and (X) 50 ppm. Standard error of mean indicated.
.
. .
.
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
of each criterion were summed to give a final score out of 300. For the purpose of determining the nature of the interaction between two inhibitory compounds, the following criteria were used. If the total score of the combination system was less than or equal to the highest score of either of the two individual compounds, the compounds were determined to have acted indifferently. If the total score of the combination system was greater than the sum of the total scores of the individual compounds, the compounds were determined to have acted synergistically. If the total score of the combination system lay between these two
253
breakpoints, the compounds were determined to have acted additively.
3. Results 3.1. Concentrations of antimicrobial compounds Nisin, lauricidink, totarol, and the LPS were screened against the selected test organisms to establish concentrations at which they provided a partial inhibition of growth of test strains. The curves obtained for S. aureus (Fig. 1) are representative of
Fig. 2. Untreated control (n), and the affect of lauricidink (o) and LPS (D), alone and in combination (x), on the growth of B. cereus (Panel A), S. aureus (Panel B), St. thermophilus (Panel C). Standard error of mean indicated.
254
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
the type of data obtained for all bacterial strains tested (individual graphs not shown). From Fig. 1, it can be seen that 50 ppm of nisin (Panel A), 10 ppm of lauricidink (Panel B), 0.5 ppm of LPS (Panel C),
and 2 ppm of totarol (Panel D) resulted in a partial inhibition of S. aureus growth; thus, these concentrations were selected for the subsequent combination experiments (Table 1). In a similar manner, inhibitory,
Table 2 Inhibition of bacterial growth by a combination of lauricidin and the LPS Criteria analysed for each bacteria
Score for each criteria
Highest individual score
Sum of scores for individual compounds
Score for combination system
Conclusion
92 100 100
134c
257d
292e
synergistic
17 20 0
92 100 100
39
76
292
synergistic
59 67 0
8 0 0
92 100 100
126
134
292
synergistic
L. monocytogenes Lag phase Culture density Viability
17 0 0
21 0 0
33 0 0
21
38
33
E. coli O157 Lag phase Culture density Viability
4 39 0
17 8 0
92 100 100
43
68
292
S. Typhimurium Lag phase Culture density Viability
4 31 0
13 46 0
13 53 0
59
94
66
additive
Y. enterocolitica Lag phase Culture density Viability
13 63 0
13 13 0
13 13 0
76
102
26
indifferent
P. aeruginosa Lag phase Culture density Viability
17 37 0
17 19 0
21 50 0
54
90
71
additive
La
LPSb
L + LPS
St. thermophilus Lag phase Culture density Viability
67 67 0
46 77 0
B. cereus Lag phase Culture density Viability
2 37 0
S. aureus Lag phase Culture density Viability
a
L: Lauricidink. LPS: Lactoperoxidase system. c Scores were obtained by summing scores for individual compounds, i.e., 67 + 67 + 0 = 134. d Scores were obtained by summing scores for both compounds, i.e., 67 + 67 + 0 + 46 + 77 + 0 = 257. e Scores were obtained by summing scores for the combination of compounds, i.e., 100 + 100 + 92 = 292. b
additive
synergistic
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
255
Fig. 3. Untreated control (n), and affect of nisin (o) and lauricidink (D), alone and in combination (x) on the growth of St. thermophilus (Panel A) and S. aureus (Panel B). Standard error of mean indicated.
but not lethal, concentrations were determined for all combinations of antimicrobials against each test organism (Table 1). 3.2. Inhibition of bacterial strains by combinations of compounds The graphs obtained from the inhibitory assays were analysed to quantify the synergistic activity between two inhibitory compounds. The combination of the LPS and lauricidink showed synergistic inhibition against all of the Gram-positive test strains
except L. monocytogenes (Fig. 2 and Table 2), as well as against the Gram-negative bacterium E. coli O157 (Table 2). Table 2 indicates the scores obtained by each bacterial strain tested against the LPS and lauricidink and is representative of the data analysis process conducted for each combination of inhibitory compounds. Nisin and lauricidink showed synergistic inhibition against St. thermophilus and S. aureus (Fig. 3 and Table 3) but showed no significant enhancement of inhibition against the other organisms (Table 3). Nisin and the LPS showed synergistic inhibition against B. cereus, S. aureus, St. thermophi-
Table 3 Scores obtained for combinations of inhibitory compounds Bacterial strains
Final scores for double combinations of inhibitory compounds Na + Lb
St. thermophilus B. cereus S. aureus L. monocytogenes E. coli O157 S. Typhimurium Y. enterocolitica P. aeruginosa a
S
292 29A 292S 13A 91A 62A 81A 28A
N + LPSc S
292 292S 100S 17A 94S 84A 17 15A
N + Td 8 33A 37 63A 46A 35A 114A 33S
N: Nisin. L: Lauricidink. c LPS: Lactoperoxidase system. d T: Totarol. S Combinations identified as acting synergistically are indicated with ‘‘S’’. A Combinations identified as acting additively are indicated with ‘‘A’’. b
L + LPS S
292 292S 292S 33A 292S 66A 26 71A
L+T
LPS + T
50 53 48 77 21 27 109 42
54 23 13A 63A 52A 54A 50A 119S
256
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
Fig. 4. Untreated control (n), and affect of nisin (o) and LPS (D), alone and in combination (x), on the growth of B. cereus (Panel A), S. aureus (Panel B), St. thermophilus (Panel C) and E. coli O157 (Panel D). Standard error of mean indicated.
lus, and E. coli O157 (Fig. 4 and Table 3) but had no significant effect against the remaining bacteria (Table 3). The combinations incorporating totarol showed no synergistic activity against the selected bacteria with the exception of the LPS – totarol and nisin –totarol combinations which showed synergistic inhibitory activity against P. aeruginosa (Table 3).
4. Discussion For use in food-based applications, it is desirable to identify combinations of inhibitory compounds that
have a wide rather than a narrow spectrum of inhibitory action and a persistent rather than a transient inhibitory effect (Gould, 2000). We examined these requirements by screening each compound and combination of compounds against a panel of four Grampositive and four Gram-negative bacteria selected on the basis of their significance as either food spoilage bacteria or food pathogens, and by testing for inhibition of bacterial growth over a 24-h period under conditions favourable for optimum growth. Testing each combination against a panel of strains of widely differing susceptibilities was also important to ensure that a potentially synergistic combination was not
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
missed. For any one organism, if the concentration of either of the antimicrobial agents being tested was a little too high or low, the ability of the test method to identify synergy was compromised (unpublished data). Although this could be overcome by testing a greater number of intermediate concentrations of each compound, we feel that it was adequately compensated for by the widely differing susceptibility of the strains used in the test panel. In order to quantify our results in terms of the synergistic potential of the combination systems, we chose to measure fundamental parameters relevant to the organisms potential to cause spoilage or disease, i.e., how long was the lag phase, what cell numbers were reached after 24 h, and was the population viable? By summing the unweighted normalised scores of three critical growth parameters, we generated a single numerical value that allowed the relative effectiveness of different combinations of inhibitory compounds to be evaluated. Although some authors have been careful to define their usage of the term synergy and to distinguish between synergistic and additive responses (ter Steeg et al., 1999), there is no clear agreement in the literature with regard to quantifiable definitions for these terms. We chose to define the nature of each interaction as either: indifferent, additive, or synergistic in a manner similar to that proposed by Bonapace et al. (2000) for evaluation of the synergistic interaction of classical antibiotics against Acinetobacter baumannii. We chose to distinguish combinations of compounds acting in an additive manner (where the degree of inhibition was greater than could be accounted for by the action of compound 1, or the action of compound 2, acting independently) from those showing a synergistic interaction (where the degree of inhibition was greater than could be accounted for by the action of compound 1 plus the action of compound 2) because we believe it is the truly synergistic interactions that are of principal interest in food preservation applications. In the present study, the application of this method clearly identified the previously established synergistic interactions between nisin and monolaurin, nisin and LPS, and monolaurin and LPS. Mansour et al. (1999) and Mansour and Milliere (2001) showed that while nisin or monolaurin were not effective against Bacillus strains, the combination of the two was synergistic and resulted in complete inhibition of
257
spore outgrowth in milk. In the present study, nisin and lauricidink were shown to have synergistic or additive effect against both Gram-positive and Gramnegative bacteria. Combinations of nisin and LPS have been reported to give synergistic activity against L. monocytogenes, both when these agents were added simultaneously (Zapico et al., 1998) and when added in two steps (Zapico et al., 1998; Boussouel et al., 2000). In the present study, this combination was identified as yielding additive or synergistic interactions against a wide range of organisms, confirming its potential for use in a number of food preservative applications. McLay et al. (2002) have previously reported that combinations of monolaurin and LPS resulted in synergistic inhibitory activity against S. aureus and E. coli O157, a result confirmed in the present study and extended to include a number of food pathogens and food spoilage organisms not previously examined. Totarol, a diterpenoid isolated from the root bark of Podocarpus nagi (Ying and Kubo, 1991), has been reported to have bactericidal activity against a diverse range of bacterial species, including methicillin-resistant strains of S. aureus (MRSA) (Muroi and Kubo, 1996), but has not previously been considered for use in food preservative applications. It was included in the present study for comparison against the established synergistic combinations. Although not as potent as the other combinations tested, combined with either nisin or the LPS, it did show a modest enhancement of activity against a number of organisms, particularly the Gram-negative organisms. By contrast, totarol showed no enhanced activity in combination with lauricidink against any organism.
5. Conclusion The use of a microtiter-based assay and automated data recording and processing enabled us to rapidly and effectively test the inhibitory potential of combinations of antimicrobial agents against a panel of eight bacteria. The differentiation of response into categories labelled as indifferent, additive, or synergistic allowed the readily identification of combinations having potential for use in multihurdle food preservation processes. The method successfully identified
258
M. Dufour et al. / International Journal of Food Microbiology 85 (2003) 249–258
three previously established effective combinations and was successful in identifying two novel inhibitory combinations (nisin – totarol and LPS – totarol) that may be useful in food-based applications.
Acknowledgements This work was supported in part by Tatua Cooperative Dairy (Morrinsville, New Zealand).
References Bonapace, C.R., White, R.L., Friedrich, L.V., Bosso, J.A., 2000. Evaluation of antibiotic synergy against Acinetobacter baumannii: a comparison with Etest, time-kill, and checkerboard methods. Diagn. Microbiol. Infect. Dis. 38, 43 – 50. Boussouel, N., Mathieu, F., Revol-Junelles, A.-M., Milliere, J.-B., 2000. Effects of combinations of lactoperoxidase system and nisin on the behaviour of Listeria monocytogenes ATCC 15313 in skim milk. Int. J. Food Microbiol. 61, 169 – 175. Cressy, H.K., Jerrett, A.R., Osborne, C.M., Bremer, P.J., 2003. A novel method for the reduction of cell numbers of Listeria monocytogenes by freezing in combination with an essential oil in bacteriological media. J. Food Prot. (in press). du Toit, E.A., Rautenbach, M., 2000. A sensitive standardised micro-gel well diffusion assay for the determination of antimicrobial activity. J. Microbiol. Methods 42, 159 – 165. Garcia-Gimeno, R.M., Hervas-Martinez, C., de Siloniz, M.I., 2002. Improving artificial neural networks with a pruning methodology and genetic algorithms for their application in microbial growth prediction in food. Int. J. Food Microbiol. 72, 19 – 30. Garcia-Graelis, C., Valckx, C., Michiels, C.W., 2000. Inactivation of Escherichia coli and Listeria innocua in milk by combined treatment of high hydrostatic pressure and the lactoperoxidase system. Appl. Environ. Microbiol. 66, 4173 – 4179. Giannuzzi, L., Contreras, E., Zaritzky, N., 1999. Modeling the aerobic growth and decline of Staphylococcus aureus as affected by pH and potassium sorbate concentration. J. Food Prot. 62, 356 – 362. Gould, G.W., 2000. Preservation: past, present and future. Br. Med. Bull. 56, 84 – 96. Gould, G.W., Abee, T., Granum, P.E., Jones, M.V., 1995. Physiology of food poisoning microorganisms and the major problems in food poisoning control. Int. J. Food Microbiol. 28, 121 – 128. Leistner, L., 2000. Basic aspects of food preservation by hurdle technology. Int. J. Food Microbiol. 55, 181 – 186. Lopez-Malo, A., Guerrero, S., Alzamora, S.M., 2000. Probabilistic modelling of Saccharomyces cerevisiae inhibition under the ef-
fects of water activity, pH, and potassium sorbate concentration. J. Food Prot. 63, 91 – 95. Mansour, M., Milliere, J.B., 2001. An inhibitory synergistic effect of a nisin – monolaurin combination on Bacillus sp. vegetative cells in milk. Food Microbiol. 18, 87 – 94. Mansour, M., Amri, D., Bouttefroy, A., Linder, M., Milliere, J.B., 1999. Inhibition of Bacillus licheniformis spore growth in milk by nisin, monolaurin, and pH combinations. J. Appl. Microbiol. 86, 311 – 324. Mazzotta, A.S., Montville, T.J., 1999. Characterisation of fatty acid composition, spore germination, and thermal resistance in a nisin-resistant mutant of Clostridium botulinum 169B and in the wild-type strain. Appl. Environ. Microbiol. 65, 659 – 664. McLay, J.C., Kennedy, M.J., O’Rourke, A.-L., Elliot, R.M., Simmonds, R.S., 2002. Inhibition of bacterial foodborne pathogens by the lactoperoxidase system in combination with monolaurin. Int. J. Food Microbiol. 73, 1 – 9. Muroi, H., Kubo, I., 1996. Antibacterial activity of anacardic acid and totarol, alone and in combination with methicillin, against methicillin-resistant Staphylococcus aureus. J. Appl. Bacteriol. 80, 387 – 394. Nikaido, H., Vaara, M., 1987. Outer membrane. In: Neidhardt, F.C. (Ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. American Society for Microbiology, Washington, DC, pp. 7 – 22. Oh, D.H., Marshall, D.L., 1993. Influence of temperature, pH, and glycerol monolaurate on growth and survival of Listeria monocytogenes. J. Food Prot. 56, 740 – 749. Roberts, J.A., 2000. Economic aspects of food-borne outbreaks and their control. Br. Med. Bull. 56, 133 – 141. Roberts, C.M., Hoover, D.G., 1996. Sensitivity of Bacillus coagulans spores to combinations of high hydrostatic pressure, heat, acidity and nisin. J. Appl. Bacteriol. 81, 363 – 368. Shimoni, E., Labuza, T.P., 2000. Modeling pathogen growth in meat products: future challenges. Trends Food Sci. Technol. 11, 394 – 402. Stevens, K.A., Sheldon, B.W., Klapes, N.A., Klaenhammer, T.R., 1991. Nisin treatment for inactivation of Salmonella species and other Gram-negative bacteria. Appl. Environ. Microbiol. 57, 3613 – 3615. Sutherland, J.P., Baylis, A.J., Roberts, T.A., 1994. Predictive modelling of growth of Staphylococcus aureus: the effects of temperature, pH and sodium chloride. Int. J. Food Microbiol. 21, 217 – 236. ter Steeg, P.F., Hellemons, J.C., Kok, A.E., 1999. Synergistic actions of nisin, sublethal ultrahigh pressure, and reduced temperature on bacteria and yeast. Appl. Environ. Microbiol. 65, 4148 – 4154. Wang, G.-H., 1992. Inhibition and inactivation of five species of foodborne pathogens by chitosan. J. Food Prot. 55, 916 – 919. Ying, B.-P., Kubo, I., 1991. Complete 1H and 13C NMR assignments of totarol and its derivatives. Phytochemistry 30, 1951 – 1955. Zapico, P., Medina, M., Gaya, P., Nunez, M., 1998. Synergistic effect of nisin and the lactoperoxidase system on Listeria monocytogenes in skim milk. Int. J. Food Microbiol. 40, 35 – 42.