Scientific Publications - Work Done by Microbiology Reader Bioscreen C

Journal of Food Protection, Vol. 64, No. 1, 2001, Pages 51-57

Modeling the Growth Boundary of Staphylococcus aureus for Risk Assessment Purposes

Cynthia M. Stewart, Martin B. Cole, J. David Legan, Louise Slade, Mark H. Vandeven and Donald W. Schaffner
 

ABSTRACT

Knowing the precise boundary for growth of Staphylococcus aureus is critical for food safety risk assessment, especially in the formulation of safe, shelf-stable foods with intermediate relative humidity (RH) values. To date, most studies and resulting models have led to the presumption that S. aureus is osmotolerant. However, most studies and resulting models have focused on growth kinetics using NaCl as the humectant. In this study, glycerol was used to investigate the effects of a glassforming nonionic humectant to avoid speci? c metabolic aspects of membrane ion transport. The experiments were designed to produce a growth boundary model as a tool for risk assessment. The statistical effects and interactions of RH (84 to 95% adjusted by glycerol), initial pH (4.5 to 7.0 adjusted by HCl), and potassium sorbate (0, 500, or 1,000 ppm) or calcium propionate (0, 500, or 1,000 ppm) on the aerobic growth of a ? ve-strain S. aureus cocktail in brain heart infusion broth were explored. Inoculated broths were distributed into microtiter plates and incubated at 37°C over appropriate saturated salt slurries to maintain RH. Growth was monitored by turbidity during a 24-week period. Toxin production was explored by enterotoxin assay. The 1,280 generated data points were analyzed by SAS LIFEREG procedures, which showed all studied parameters significantly affected the growth responses of S. aureus with interactions between RH and pH. The resulting growth/no growth boundary is presented.

_______

INTRODUCTION

 Staphylococcus aureus is one of the leading causes of foodborne illness and is ranked as one of the most prevalent causes of gastroenteritis worldwide (12), even though the occurrences of this foodborne disease are grossly underreported. S. aureus was determined to be the etiological agent in 367 (19.6%) of 1,869 documented bacterial foodborne outbreaks in the United States (1). Approximately 25 major outbreaks of Staphylococcus food poisoning occur annually in the United States (10). S. aureus has been estimated by the Centers for Disease Control and Prevention to cause 185,060 illnesses, 1,753 hospitalizations, and 2 deaths per year in the United States, all of which are via consumption of contaminated foods (18). Because of its prevalence as a food poisoning organism, S. aureus has been extensively studied to determine the physical and chemical parameters that affect its growth and toxin formation (3, 4, 11, 12). S. aureus is ubiquitous to the mucous membranes and skin of warm-blooded animals. It is a poor competitor with other bacteria and is easily destroyed by cooking temperatures; however, its toxins can survive heat processing equivalent to that given to low-acid canned foods (11). Staphylococcal food poisoning frequently occurs when food is contaminated after cooking by a person carrying the organism, and subsequently the food is temperature abused for several hours, which allows for toxin production before consumption. It is the ingestion of the toxin that causes the foodborne disease.

S. aureus is highly salt tolerant and has been reported to grow at relative humidities (RHs) as low as 85% in NaCl concentrations up to 25% (wt/wt) (11). The RHs limiting growth are typically higher when humectants other than NaCl are used to control RH (11). Notermans and Heuvelman (19) reported that at a water activity of 0.98 to 0.93, sucrose was more favorable for growth than NaCl, but at a water activity of 0.87, NaCl was more favorable. Marshall et al. (17) reported that glycerol inhibition of the growth rate of S. aureus was about 10% greater than that caused by NaCl at water activity levels between 0.96 and 0.90. Empirical microbiological models can be designed to predict how microorganisms relate to the environment. Predictive models are typically broadly applied for the screening of products and/or processes before challenge or shelflife studies. However, it must be remembered that models are only tools to be used to help design products with food safety concerns addressed at the beginning of the development process and that there must be validation of the system through challenge or shelf-life studies with the real product. Both the U.S. Department of Agriculture Pathogen Modeling Program (PMP) and the Food MicroModel (FMM) developed in the United Kingdom through the Ministry of Agriculture, Fisheries and Food include S. aureus growth kinetic models. The PMP models the effects of temperature, initial pH, NaCl concentration, and sodium nitrite concentration in a broth system on aerobic and anaerobic growth kinetics of the organism (7). The FMM models the growth responses as affected by NaCl concentration, pH, and storage temperature in a model broth system (29). The PMP also includes a model for the survival of S. aureus in nongrowth conditions, which was developed from a model broth system that includes the effects and interactions of pH controlled by lactate buffer, lactic acid concentration, NaCl concentration, sodium nitrite concentration, and varying temperatures (31). For all of the above-mentionedmodels, NaCl was the humectant chosen to control the RH in the system, and the resulting models are kinetic in nature and, therefore, do not allow the boundary for growth/no growth to be estimated. In contrast to kinetic modeling, probability modeling focuses on determining if the microorganism of concern will or will not grow, in other words, determining the growth/no growth interface (21). This becomes increasingly important when pathogens are of concern, because the rate of growth may be less important than the fact that the organism is present and may be able to grow to an infectious dose or produce toxins (21). In this study, the statistical effects and interactions of RH controlled by glycerol, initial pH, and potassium sorbate or calcium propionate on the boundary for S. aureus growth were modeled

MATERIALS AND METHODS

 Organisms and media preparation.
The . ve S. aureus strains used to create the bacterial cocktail used in this study were S. aureus ATCC 13565 (American Type Culture Collection, Rockville, Md.), which produces staphylococcal enterotoxin A (SEA), ATCC 14458 (SEB), and ATCC 27664 (SEE); D-2 (SED) obtained from Toxin Technology, Inc. (Sarasota, Fla.); and A-100 (SEA) obtained from U.S. Army Natick Labs (Natick,Mass.). All strains showed typical growth on Baird-Parker agar plates and were coagulase positive. Their identi. cation as S. aureus was con- . rmed by use of a Riboprinter (Qualicon, Wilmington, Del.). Stock cultures were grown overnight at 37°C in brain heart infusion broth (BHI; Difco Laboratories, Detroit, Mich.), suspended in a 2:1 broth-glycerol solution, and stored at -80°C until needed. The BHI broth was reconstituted with the appropriate ratio of water to glycerol to achieve the desired RHs for the study (84, 88, 92, or 95% RH). Appropriate amounts of 10% stock solutions of potassium sorbate (Hoechst, Frankfurt, Germany) or calcium propionate (Aldrich Chemical Co., Milwaukee, Wis.) were added to achieve . nal concentrations of 500 or 1,000 ppm. The pH was adjusted using 0.1 or 1.0 N HCl (J. T. Baker, Inc., Phillipsburg, N.J.; pH 4.5, 5.0, 6.0, or 7.0). Final RH was determined using the Aqualab CX-2 water activity meter (Decagon Devices Inc., Pullman, Wash.). It was determined that the RH of the media did not change more than 60.003% when measured at ambient temperature versus 37°C. Each of the 80 broths was . lter sterilized and stored in screw-capped tubes at 4°C until needed.

Bioscreen microtiter plate preparation, incubation, and

measurements.
The . ve stock cultures were removed from -80°C storage, and one loopful was transferred into 9 ml of BHI broth, separately. The inoculated broths were incubated for 18 h at 37°C. The . ve cultures were individually adjusted to an optical density (OD) at 530 nm of 0.750 to 0.780 (Perkin-Elmer 35 spectrophotometer) with 0.1% sterile peptone to achieve a concentration of approximately 108 CFU/ml. Two milliliters of each diluted culture was transferred into one sterile test tube and vortexed to create the S. aureus cocktail used in the studies. Three 1:10 dilutions with sterile 0.1% peptone water were used to make a . nal working cocktail with a concentration of approximately 105 CFU/ ml. This working cocktail was kept on ice until used to inoculate the various media used in each experiment. A 1-ml sample of the working cocktail was taken, serially diluted, spread plated onto Baird-Parker agar plates, incubated at 37°C for 48 h, and counted to determine initial concentrations of cells.

Ten milliliters of each of the 80 broths as described above was aseptically transferred to sterile tubes, and 100 ml of the S. aureus cocktail was added to the broth to achieve an initial concentration of approximately 103 CFU/ml. The inoculated broth (400 ml) was aseptically transferred into sterile, 100-well microtiter plates in eight replicate wells per medium. The outer wells of the microtiter plates were . lled with uninoculated medium to act as a partial moisture loss barrier and as uninoculated controls. Each plate contained media at one RH level only. A piece of sterile Thermaseal . lm was placed on top of the open wells, and the lid was also in place during incubation. The plates were placed on a rack in a plastic sealable container, which had the bottom . lled with appropriate saturated salt slurries to maintain the environmental RH to ensure RH of the media in the plates remained as stable as possible. The saturated salt solutions used were as follows: ZnSO4 · H2O (83% RH @ 37°C), KNO (89% RH @ 37°C), KPO4 (93% RH @ 37°C), and K2Cr2O7 (96% RH @ 37°C). The containers were closed and placed into a 37°C incubator. The plates were periodically removed from incubation, and the OD of the 640 individual wells was measured for up to 6 months using the wideband . lter on the Bioscreen C system (Labsystems Oy, Helsinki, Finland). The experimentswere repeated on separate dates. A total of 1,280 wells were monitored.

Data collection.

The S. aureus population was considered to have shown growth when the Bioscreen wideband OD (ODwb) measurement increased from an initial reading of 0.200 to 0.220 to 0.350 or higher. Time to growth (TTG) was determined by calculating the geometric mean of the time of the last measurement that showed no growth (ODwb <0.350) and the . rst time point that showed growth (ODwb >0.350). In instances of no growth, TTG was censored at the . nal measurement time of 168 days.

OD measurements were made and transferred from the Bioscreen C ASCII . le to a DOS text editor and then imported into Statistica software (StatSoft, Tulsa, Okla.) for initial analysis. Data were then transferred to a Microsoft Excel spreadsheet to determine TTG and then to SAS for modeling using LIFEREG (SAS Institute Inc., Cary, N.C.).

Statistical analysis.
The TTG data along with a censoring indicator and the RH, pH, and potassium sorbate or calcium propionate levels were input data for construction of two separate models: one based on the potassiumsorbate data set and one based on the calcium propionate data set. These two data sets used a common block of data generated from the experimental conditions, which had no preservatives added. The SAS LIFEREG procedure was used to develop predictive models of ln TTG as a function of the factors RH, pH, and preservative level. By default, the procedure . ts a model to the log of the dependent variable. The resulting model can easily be transformed to a regular time scale. In growth modeling, there are often conditions where no growth occurs; therefore, TTG is sometimes censored. Under these conditions, ordinary least-squares regression is not applicable, and special procedures are required. The LIFEREG procedure can accommodate such censored data and uses maximum likelihood estimation methods to . nd regression coef. cients. When creating models with the LIFEREG procedure, there is a need to have many replicates of conditions equally spaced out about the matrix, with approximately 50% of the conditions allowing and 50% not allowing growth for best-. tting purposes. Therefore, the ranges of conditions studied (RH, pH, and preservative levels) were chosen to satisfy these requirements.

Toxin production.
Toxin production for experimental conditions close to the growth or no growth border was explored by enterotoxin assay using the VIDAS Staph Enterotoxin Assay. The VIDAS Staph Enterotoxin Assay is a qualitative enzyme-linked � uorescent immunoassay performed in the automated mini-VIDAS instrument (bioMerieux, Inc., Hazelwood, Mo.). After the . nal ODwb measurements were completed (24 weeks), the media from the microtiter plates were removed for assay. The broth from each of the eight replicate wells was removed via pipet and combined into two microcentrifuge tubes. The samples were spun down for 10 min, and the supernatant from the two tubes was combined. The pH of the supernatant was adjusted with 1 N NaOH to a pH of 6.0 to 8.0. A half a milliliter of the sample was placed into the appropriate well in a test strip, the strips were placed into the mini-VIDAS, and the assays were run. A positive and negative control were run with each set of test strips. The test is sensitive to 1 ng of toxin per ml of sample.

RESULTS

A threshold ODwb value of 0.350 was used to score wells with S. aureus growth (ODwb $ 0.350) versus wells with no growth (ODwb , 0.350). This threshold value was determined in two ways. First, ODwb measurements were compared with plate counts, and it was determined that an ODwb reading of 0.350 was approximately the . rst ODwb value where an increase in turbidity regularly correlated to an increase in plate counts, indicating the approximate ODwb where the instrument was sensitive enough to accurately determine an increase in cell numbers by a change in turbidity (data not shown). Second, the data were analyzed to determine TTG in each well using threshold ODwb values for growth of 0.300, 0.350, and 0.400, and models were created and compared (models using data from ODwb threshold values of 0.300 and 0.400 not shown). The models created with the threshold value set at 0.300 did not make biological sense. We suspect that this was because this ODwb corresponded to a level that is lower than the sensitivity of the instrument to measure an increase in cell numbers. When the threshold ODwb value was raised to 0.400, the models were similar to those created with the TTG data when the threshold ODwb of 0.350 was used; however, the TTG data when the threshold ODwb of 0.350 was used gave a slightly more conservative model. For these reasons, the models were developed with TTG data obtained by making the ODwb of 0.350 or more the lower limit for determining that growth had occurred in a particular well.

The models included all three main effects: RH (rh), pH (ph), and potassium sorbate (sorb) or calcium propionate (cal), their quadratic effects (RH·RH 5 rh2, pH·pH 5 ph2, and potassium sorbate·potassium sorbate 5 sorb2 or calcium propionate· calcium propionate 5 cal2), and the three two-factor interactions (rh ·ph, rh ·sorb, ph·sorb or rh ·ph, rh ·cal, ph· cal). These main effects, quadratic effects, and interactions are collectively referred to as the factors. LIFEREG outputs a table of regression coef. cient estimates and approximate chi-square distribution P values for each factor in the model. The relative importance of each factor can be judged by the P value: factors with small P values are most in� uential and predictive of the TTG. LIFEREG also allows the user to specify the error distribution to account for the variation in TTG not explained by the regression model. For model development purposes, Weibull, lognormal, and log-logistic distributions were considered. It was found that all three distributions gave very similar models; the regression coef. cients were similar in sign and magnitude. Since the three distributions are not all from the same class of distributions, it is not possible to formally test for goodness of . t using likelihood ratio tests. However, since all models produced similar regression equations, we selected the log-logistic distribution model, because it had the largest sample log likelihood.

The model development process is an iterative process. The . rst analysis of the data created models that were all inclusive; this allowed for the determination of those factors that had a signi. cant effect on the growth of S. aureus. All of the main effects, quadratic effects, and two-factor interactions were signi. cant for each model (P <0.005) except for the interaction between RH and potassium sorbate in the potassium sorbate model and the calcium propionate quadratic term in the calcium propionate model. For this reason, in each model the insigni. cant factor was dropped and the data were reanalyzed to develop the . nal models (Tables 1 and 2). The resulting model equation from the calcium propionate data was 2.250 to 3.703· rh - 1.488·ph + 0.418· cal + 1.572· rh² + 0.343·ph² + 0.814· rh·ph - 0.130·rh · cal - 0.221·ph·cal, and the resulting model equation from the potassium sorbate data was 2.624 to 3.938·rh - 2.421·ph + 0.907· sorb + 1.638· rh² + 0.887·ph² - 0.190· sorb² + 0.903·rh ·ph - 0.756·ph·sorb. From the model equations, the TTG contour plots for calcium propionate (Fig. 1) and for potassium sorbate (Fig. 2) were created. For diagnostic purposes, plots of ln predicted time to growth versus ln observed time to growth were examined (Fig. 3). Plots of residuals versus main effects were also examined, and no unusual patterns or anomalies were detected (data not shown).

There was a slight difference in the curvature of the contour lines in the plots for the media with no preservatives data for each model (Figs. 1 and 2), despite the fact that the same data for the no preservative condition were used for developing each model. This difference occurs because two distinctly different mathematicalmodels were produced by the analysis, and when different models are used to predict TTG at a zero preservative level, they will produce slightly different predictions. It should be noted that the differences are small and mainly in the low pH areas of the plots. In practice, the target product formulation would be well away from the growth boundary to allow for inherent process variation.

FIGURE 1. Predicted time to growth of S. aureus based on the combined effects of RH (%), pH, and calcium propionate. The contours are scaled in intervals of 30 to 180 days. When moving down from the growth in less than 30-day regions, the . rst contour encountered is the 30-day line. When moving up from the no growth region, the . rst contour encountered is the 180-day line. The region above the 180-day contour line is not a plateau but a series of ever-increasing contours or longer TTG. (A) 0 ppm of calcium propionate; (B) 500 ppm of calcium propionate; (C) 1,000 ppm of calcium propionate.+, positive for enterotoxin;-, negative for enterotoxin.

FIGURE 2. Predicted time to growth of S. aureus based on the combined effects of RH (%), pH, and potassium sorbate. The contours are scaled in intervals of 30 to 180 days. When moving down from the growth in less than 30-day regions, the . rst contour encountered is the 30-day line. When moving up from the no growth region, the . rst contour encountered is the 180-day line. The region above the 180-day contour line is not a plateau but a series of ever-increasing contours or longer TTG. (A) 0 ppm of potassium sorbate; (B) 500 ppm of potassium sorbate; (C) 1,000 ppm of potassium sorbate. +, positive for enterotoxin; -, negative for enterotoxin.

FIGURE 3. Plots of ln predicted time to growth versus ln observed time to growth for S. aureus. (A) Calcium propionate model; (B) potassium sorbate model.

TABLE 1. Parameters of the model to predict log10 time to growth based on RH (rh), pH (ph), and calcium propionate (cal)

TABLE 2. Parameters of the model to predict log10 time to growth based on RH (rh), pH (ph), and potassium sorbate (sorb)

All of the models make good biological sense. As conditions become more and more unfavorable for growth, the contour lines are closer together, indicating conditions are approaching those that do not allow growth of the organism. As the RH or pH of the system decreases, a corresponding increase in TTG is seen, and the no growth area of the contour plot increases in size. The addition of potassium sorbate at low pH (pH 4.5 to 5.5) dramatically changes the contour plots with steep curvature in the low pH area of the plots. The no growth boundary with respect to RH is increased from approximately 89% to approximately 93% when 1,000 ppm of potassium sorbate is present at pH 4.5. This effect is not as evident when calcium propionate is present; the no growth boundary with respect to RH only increased approximately 1.5% from 88 to 89.5% when 1,000 ppm of calcium propionate is present at pH 4.5. There is little change in the boundary with respect to RH when the system is at pH 7.0 with the addition of either preservative studied.

Various regions on the contour plots generated by the models were explored with enterotoxin assays for all conditions where the . nal ODwb measurement was borderline. Also, assays were conducted for all samples with treatments at 84% RH, pH 7.0, and 84% RH, pH 4.5, although all ODwb measurements for treatments at 84% RH were less than 0.350. The results are indicated by a plus sign (toxin present) or minus sign (no toxin present) on the contour plots (Figs. 1 and 2). A total of 71 assays were run, and in every case where the data were scored as ‘‘no growth,’’ the toxin assay results were negative for toxin, and in every case where the data were scored as ‘‘growth,’’ the toxin assay results were positive for the presence of toxin. This further supports the choice of an ODwb of 0.350 as the correct cutoff value for time to growth in the individual wells.

DISCUSSION

Comparison of the models developed in this study with other published work is dif. cult, because most previously published studies have used NaCl to control the RH of the system, whereas this study has used glycerol. The limits of S. aureus growth, or the growth of any microorganism, cannot be predicted by the RH of the system alone; instead, both the RH of the system and the physical properties of the humectant used should be considered. For example, NaCl is a non–glass-forming ionic humectant that has speci . c effects on membrane transport systems. Therefore, both the osmotic and ionic stresses placed on bacterial cells by NaCl could affect the ability of those cells to grow. Glycerol is a nonionic glass former that passively permeates the cell membrane. Therefore, glycerol would not have a direct effect on ion transport systems. Unlike NaCl, glycerol forms an aqueous glass. The glass transition temperature (Tg) is a speci. c characteristic for each glass-forming compound. The in� uence of a solute’s Tg on molecular mobility directly affects the amount of osmotic stress a speci. c solute places on cells. For these reasons, it should be expected that the limits of growth based on RH would be different when various humectants are used to control RH, and, therefore, comparing studies based on RH alone could lead to false conclusions (24).

The differences in modeling approaches also make comparisons dif. cult. Both the FMM and PMP have very few kinetic curves generated in which cells would be under stressed conditions, the most dif. cult conditions in which to obtain kinetic information. Also, there are few true replicates in these areas, with most of the experimental design similar to a central composite design in which most of the replicates are in the middle of the design. Kinetic models also model the mean value, which, due to natural biovariability in the bacterial population, can be meaningless under stressed conditions. The LIFEREG procedure used to create the boundary models handles this by . tting the distribution to the error term, hence allowing probabilities to be calculated. This is useful for risk assessment purposes and allows the model to be easily used in Monte Carlo simulations. One consequence of modeling this way is that there is a need to have many replicates equally spaced about the matrix, with approximately 50% of the conditions allowing and 50% not allowing growth for best-. tting purposes. The log time to growth is used to normalize the variance. It has generally been accepted that the limits for growth of S. aureus are 85% RH when NaCl is used as the humectant and 89% when glycerol is used (11). Our models, based on glycerol, show the limiting RH for growth to be approximately 86% (at pH 7.0), which disagrees with work published by Marshall et al. (17), who found the growth limits were 89 and 86% RH when glycerol and NaCl were used as the humectants, respectively. This disagreement could be due to strain variation or to the fact that only quarter-strength BHI broth was used in the work by Marshall et al. (17), whereas full-strength BHI broth was used in our model system. Shapero et al. (23) reported growth of S. aureus at 88% RH on tryptic soy agar plates with RH adjusted with glycerol but did not conduct experiments with the system at any lower RH values. Studies with sucrose (a glass former that is not cell membrane permeable and has its own Tg) used as the humectant showed growth at 90% RH but not at 87% RH (19), and Scott (22) reported growth at 88 but not at 86% RH. Broughall et al. (5, 6) developed models for growth of S. aureus in UHT milk, with RH adjusted by the addition of glucose (a glass former that is not cell membrane permeable and has a different Tg). They found growth at 88% RH but did not explore the RH boundary further, because the additional glucose caused the system to become saturated. There is published literature that describes the ability of S. aureus to grow in various food systems, mainly meats and cheeses, but most of these systems used NaCl as the RH depressant (2, 13, 15, 16, 20, 25, 30) and, therefore, cannot be used to validate the models created with glycerol. Often in experiments conducted with food systems, the pH of the product is not reported, nor is the list of ingredients published, which also leads to dif. culty when trying to compare data from various published studies.

The effectiveness of potassium sorbate to inhibit growth was considerably greater compared with calcium propionate, which was especially noticeable at low pH. Potassium sorbate and calcium propionate have some key similarities that may lead to the belief that they should act similarly as preservatives; potassium sorbate has a pKa value of 4.74 and a molecular weight of 150.2, whereas calcium propionate has a pKa value of 4.87 and molecular weight of 186.2. Both have been reported in the literature to act as weak acids; the molecules remain undissociated at low pH, diffuse through the bulk lipids of the cell membrane, then enter and dissociate in the cytoplasm, where the pH is close to neutral. It is the dissociated molecules that decrease the internal pH of the cell, which may prevent growth by perturbing metabolism (14, 28). Recent studies have shown that sorbic acid does have inhibitory effects in the undissociated form, suggesting not only that it may act as a weak acid but also that it may have a secondary mode of action associated with membrane-resident sorbic acid that acts as an uncoupler, which may disable active transport (8, 26, 27). It has also been reported that for S. aureus the minimum inhibitory concentration of propionic acid is much greater than sorbic acid (8, 9). This suggests that the levels of calcium propionate used in this study that are in line with the typical use levels in foods may have been too low to observe the inhibitory effect of calcium propionate on S. aureus growth.

In summary, two models describing the growth boundaries for S. aureus with respect to RH controlled by glycerol, pH, and preservative were developed. The growth boundaries were explored with toxin assays. For the . rst time, these models describe the growth/no growth boundaries for S. aureus with respect to RH controlled by glycerol and pH in addition to the preservatives potassium sorbate and calcium propionate. These models will allow product developers to visualize the ‘‘safe space’’ for the formulation of shelf-stable intermediate moisture foods using these preservation factors, will allow microbiologists to assess risks more effectively over a wide range of products, and will ultimately allow the consumer to have greater assurance of food safety.

REFERENCES

1. Bean, N. H., and P. M. Grif. n. 1990. Foodborne disease outbreaks in the United States, 1973–1987: pathogens, vehicles and trends. J. Food. Prot. 53:804–817.

2. Bennett, R. W., and W. T. Amos. 1983. Staphylococcus aureus growth and toxin production in imitation cheeses. J. Food Sci. 48: 1670–1673.

3. Bergdoll, M. S. 1989. Staphylococcus aureus, p. 463–513. In M. P. Doyle (ed.), Foodborne bacterial pathogens. Marcel Dekker, Inc., New York.

4. Bergdoll, M. S. 1992. Staphylococcal intoxication in mass feeding, p. 25–47. In A. T. Tu (ed.), Food poisoning: handbook of natural toxins, vol. 7. Marcel Dekker, Inc., New York.

5. Broughall, J. M., P. A. Anslow, and D. C. Kilsby. 1983. Hazard analysis applied to microbial growth in foods: development of mathematical models describing the effect of water activity. J. Appl. Bacteriol. 55:101–110.

6. Broughall, J. M., and C. Brown. 1984. Hazard analysis applied to microbial growth in foods: development and application of threedimensional models to predict bacterial growth. Food Microbiol. 1: 13–22.

7. Buchanan, R. L., J. L. Smith, C. McColgan, B. S. Marmer, M. Golden, and B. Dell. 1993. Response surface models for the effects of temperature, pH, sodium chloride, and sodium nitrite on the aerobic and anaerobic growth of Staphylococcus aureus 196E. J. Food Saf. 13:159–175.

8. Eklund, T. 1983. The antimicrobial effect of dissociated and undissociated sorbic acid at different pH levels. J. Appl. Bacteriol. 54: 383–389.

9. Eklund, T. 1985. Inhibition of microbial growth at different pH levels by benzoic and propionic acids and esters of p-hydroxybenzoic acid. Intl. J. Food Mirobiol. 2:159–167.

10. Holmberg, S. D., and P. A. Blake. 1984. Staphylococcal food poisoning in the United States. JAMA 251:487–489.

11. ICMSF. 1996. Staphylococcus aureus, p. 299–333. In T. A. Roberts, A. C. Baird-Parker, and R. B. Tompkin (ed.), Microorganisms in foods 5: characteristics of microbial pathogens. Blackie Academic & Professional, London.

12. Jablonski, L. M., and G. A. Bohach. 1997. Staphylococcus aureus, p. 353–375. In M. P. Doyle, L. R. Beuchat, and T. J. Montville (ed.), Food microbiology: fundamentals and frontiers. ASM Press, Washington, D.C.

13. Koenig, S., and E. H. Marth. 1982. Behavior of Staphylococcus aureus in cheddar cheese made with sodium chloride or a mixture of sodium chloride and potassium chloride. J. Food Prot. 45:996–1002.

14. Krebs, H. A., D. Wiggins, M. Stubbs, A. Sols, and F. Bedoya. 1983. Studies on the mechanism of the antifungal action of benzoate. Biochem. J. 214:657–663.

15. Kreisman, L. N., and T. P. Labuza. 1978. Storage stability of intermediate moisture food process cheese food products. J. Food Sci. 43:341–344.

16. Magrini, R. C., J. Chirife, and J. L. Parada. 1983. A study of Staphylococcus aureus growth in model systems and processed cheese. J. Food Sci. 48:882–885.

17. Marshall, B. J., D. F. Ohye, and J. H. B. Christian. 1971. Tolerance of bacteria to high concentrations of NaCl and glycerol in the growth medium. Appl. Microbiol. 21:363–364.

18. Mead, P. S., L. Slutsker, V. Dietz, L. F. McCraig, J. S. Bresee, C. Shapiro, P. M. Grif. n, and R. V. Tauxe. 1999. Food-related illness and death in the United States. Emerg. Infect. Dis. 5:607–625.

19. Notermans, S., and C. J. Heuvelman. 1983. Combined effect of water activity, pH and sub-optimal temperature on growth and enterotoxin production by Staphylococcus aureus. J. Food Sci. 48:1832–1840.

20. Rajkowski, K. T., F. Schultz, F. Negron, and A. Dicello. 1994. Effect of water activity on the growth of Staphylococcus aureus at meatcheese interfaces. J. Food Saf. 14:219–227.

21. Ratkowsky, D. A., and T. Ross. 1995. Modelling the bacterial growth/no growth interface. Lett. Appl. Microbiol. 20:29–33.

22. Scott, W. J. 1953. Water relations of Staphylococcus aureus at 30°C. Aust. J. Biol. Sci. 6:549–564.

23. Shapero, M., D. A. Nelson, and T. P. Labuza. 1978. Ethanol inhibition of Staphylococcus aureus at limited water activity. J. Food Sci. 43:1467–1469.

24. Slade, L., and H. Levine. 1988. Non-equilibrium behavior of small carbohydrate-water systems. Pure Appl. Chem. 60:1841–1864.

25. Smith, J. L., and S. A. Palumbo. 1980. Inhibition of aerobic and anaerobic growth of Staphylococcus aureus in a model sausage system. J. Food Saf. 4:221–233.

26. Stratford, M., and P. A. Anslow. 1996. Comparison of the inhibitory action on Saccharomyces cerevisiae on weak-acid preservatives, uncouplers, and medium-chain fatty acids. FEMS Microbiol. Lett. 142: 53–58.

27. Stratford, M., and P. A. Anslow. 1998. Evidence that sorbic acid does not inhibit yeast as a classic ‘‘weak acid preservative.‘‘ Lett. Appl. Microbiol. 27:203–206.

28. Stratford, M. A., and A. H. Rose. 1986. Transport of sulfur dioxide by Saccharomyces cerevisiae. J. Gen. Microbiol. 132:1–6.

29. Sutherland, J. P., A. J. Bayliss, and T. A. Roberts. 1994. Predictive modelling of growth of Staphylococcus aureus: the effects of temperature, pH and sodium chloride. Int. J. Food Microbiol. 21:217– 236.

30. Walls, I., V. N. Scott, and D. T. Bernard. 1996. Validation of predictive mathematical models describing growth of Staphylococcus aureus. J. Food Prot. 59:11–15.

31. Whiting, R. C., S. Sackitey, S. Calderone, K. Morely, and J. G. Phillips. 1996. Model for the survival of Staphylococcus aureus in nongrowth environments. Int. J. Food Microbiol. 31:231–243.

(Full Text online - PDF)

© 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