Fate of Staphylococcus aureus on Vacuum-Packaged Ready-to-Eat Meat Products Stored at 218C

The U.S. Department of Agriculture has established standards for the composition and shelf stability of various readyto-eat meat products. These standards may include product pH, moisture:protein ratio, and water activity (aw) values. It is unclear how closely these standards are based on the potential for pathogen growth or toxin production. Because the vacuum packaging used on most ready-to-eat meat products inhibits mold, Staphylococcus aureus is the pathogen most likely to grow on products with reduced aw and increased percentage of water-phase salt. In this study, 34 samples of various ready-to-eat meat products were inoculated with a three-strain mixture of S. aureus, vacuum packaged, and stored at 218C for 4 weeks. S. aureus numbers decreased by 1.1 to 5.6 log CFU on fermented products (pH # 5.1) with a wide range of salt concentrations and moisture content. Similarly, S. aureus numbers decreased by 3.2 to 4.5 log CFU on dried nonacidified jerky (aw # 0.82; moisture:protein ratio of #0.8). Products that were not fermented or dried clearly supported S. aureus growth and cannot be considered shelf stable. The product pH and moisture:protein ratio were the two compositional factors most highly correlated (R2 5 0.84) with S. aureus survival and growth for the types of products tested, but pH and aw or pH and percentage of water-phase salt also may provide useful predictive guidance (R2 5 0.81 and 0.77, respectively). Several federal standards exist for the composition of ready-to-eat (RTE) meat products. The standards are used to define both product characteristics and shelf stability. The compositional factors commonly used by regulators in establishing compositional and shelf-stability standards are moisture:protein ratio (MPR), water activity (aw), and pH. For example, nonrefrigerated semidry shelf-stable sausage must (i) have an MPR of #3.1 and a pH value of #5.0, (ii) have an MPR of #1.9 at any pH, or (iii) have a pH of #4.5 (or 4.6 with an aw of #0.91) and an internal brine concentration of $5% and must be intact (or vacuum packaged if sliced), cured, and smoked (19). With experience, processors can establish the relationship between MPR and product ‘‘shrink’’ or yield, which is relatively easy to determine. Similarly, a pH meter is relatively affordable for processors and can easily be used to determine product pH. Small-scale processors may be less likely to measure aw, however, because of the relatively high price of an aw meter. Food microbiologists, when evaluating the potential for pathogenic bacterial growth on meat products, commonly consider pH and either aw or percentage of water-phase salt (%WPS). Small-scale processors could build up a database relating product formulation and yield to %WPS, but this approach would require expenditures for analyses of water and salt percentages by a commercial laboratory. Because * Author for correspondence. Tel: 608-265-4801; Fax: 608-262-6872; E-mail [email protected]. of all these variables, different criteria may be used by regulators, processors, and scientists to evaluate shelf stability of RTE meat products. Currently, there are no readily available models that allow the interchanging of MPR, aw, and %WPS in determining whether an RTE meat product is shelf stable. Shelf stability can be defined as product characteristics that prevent the growth of pathogenic microorganisms under normal storage conditions. Thus, to experimentally determine shelf stability of RTE meat products that have reduced moisture and/or pH and added salt, a target pathogen should be identified. The bacterial pathogen commonly regarded as having the highest tolerance to reduced aw or increased salt concentration is Staphylococcus aureus (13). Mycotoxigenic or antibiotic-producing mold species can grow on meat products (1, 5, 14, 15, 17) and at lower aw values than are tolerated by S. aureus. Thus, these molds are a possible food safety hazard for RTE meat products. However, vacuum packaging is commonly used to prevent mold growth on RTE meats, and there is no recent history of mycotoxin production associated with commercial RTE meats. Although the environmental pathogen Listeria monocytogenes also poses a risk for postprocessing contamination of RTE products, it has a higher minimum aw for growth than does S. aureus, 0.92 (20) versus 0.86 under aerobic conditions (13). Anaerobic conditions result in an even higher minimum aw (0.88) for S. aureus growth (11). Processing treatments reducing aw and preventing S. aureus J. Food Prot., Vol. 68, No. 9 1912 INGHAM ET AL. growth should also prevent L. monocytogenes growth. Therefore, the shelf stability of nonrefrigerated semidry sausage and perhaps other RTE meat products could be defined strictly in terms of whether S. aureus growth occurs. The objectives of the present study were to (i) experimentally determine the growth potential for S. aureus on a variety of RTE meat products with known MPR, aw, pH, and %WPS and (ii) determine critical compositional values associated with the likelihood of S. aureus growth. MATERIALS AND METHODS Inoculum preparation. A cocktail was prepared using three broadly representative strains of S. aureus. Strain FRI 1007 was obtained from the laboratory of Dr. Amy Wong (Food Research Institute, University of Wisconsin–Madison) and was originally isolated from a Genoa salami. Strain ATCC 12600 (American Type Culture Collection, Manassas, Va.) has been designated as a type strain since 1958 (3). This strain was originally isolated from pleural fluid, has typical infective characteristics, is coagulase positive and DNase positive, and is capable of growth in milk and potato. However, it reportedly does not produce enterotoxins A, B, C, or D (10). Strain ATCC 25923 is a clinical isolate used in antibiotic susceptibility testing and screening of plant-based compounds, e.g., spice essential oils (4, 6, 8, 9), and animal-based compounds (18) for antibacterial activity. Stock cultures were maintained at 2208C in brain heart infusion broth (BHIB; Difco, Becton Dickinson, Sparks, Md.) with 10% (wt/vol) added glycerol (Fisher Scientific, Itasca, Ill.). Working cultures were maintained at 48C on brain heart infusion agar (BHIA; Difco, Becton Dickinson) and were prepared monthly from frozen stock cultures. To obtain working cultures, each strain was cultured twice at 358C for 18 to 24 h in BHIB, streaked to BHIA, incubated 18 to 24 h at 358C, examined for purity, and then stored at 48C. Inoculation cultures were prepared for each strain by transferring a loopful of growth from the working culture plate to 9 ml of BHIB and incubating at 358C for 20 to 24 h, resulting in a concentration of about 8 log CFU/ml. To prepare the three-strain inoculum cocktail, the BHIB cultures were mixed by vortexing, combined into a sterile 50-ml conical centrifuge tube (Falcon brand, Fisher), and centrifuged for 12 min at 5,000 3 g. The supernatant was decanted, and the pellets were resuspended to the original volume in Butterfield’s phosphate diluent (BPD; Nelson Jameson, Marshfield, Wis.). The resulting cocktail was serially diluted in BPD and plated on Petrifilm Staph Express (PF-SE; 3M Microbiology, St. Paul, Minn.) to confirm that the initial cell concentration was about 8 log CFU/ml. The PF-SE plates were incubated at 358C for 24 h. Preparation of meat products. A total of 34 RTE meat products were obtained from a local grocery store over a 5-month period. All products were transported to the laboratory within 15 min and refrigerated (58C) until used. The products included those that are generally shelf stable, e.g., dried salamis, beef jerky, beef snack sticks, and pepperoni (Table 1), and others that required refrigerated storage (e.g., cooked salami and bologna; Table 2). The aw of a representative product piece (cross section) that included the outer surface was measured with a Decagon water activity meter (AquaLab Series 3 TE, Pullman, Wash.). The pH of a similar piece was measured after preparing a 1:5 slurry in distilled water. A representative sample of each product (ca. 50 g) was placed in a sample bag and sent to a commercial testing laboratory to be analyzed for water percentage (forced-air oven determination by method 950.46Bb, AOAC International, Gaithersburg, Md.) and salt percentage (potentiometric determination by method 980.25, AOAC). From these results, %WPS was calculated. Product compositional information is shown in Tables 1 and 2. Nine slices were then cut from each RTE meat product with a knife that was sanitized in 70% (vol/vol) ethanol. Each slice contained 3 by 5 cm of the outer surface of the meat product, was 1 cm thick, and weighed about 20 g. Each piece was placed in a biosafety hood on aluminum foil that had previously been sanitized with 70% (vol/vol) ethanol and UV light. Inoculating and plating procedure. The outer surface of each product piece was inoculated because S. aureus is considered a postprocessing contaminant, i.e., typical thermal processes used to make RTE meat products destroy it. To inoculate the RTE meat slices, 0.025 ml of the undiluted cocktail (containing about 6.5 log CFU) was transferred to the product outer surface and distributed as evenly as possible with a sterile bent plastic spreader (Daigger, Vernon Hills, Ill.). The product slices were then allowed to dry for at least 30 min. The small inoculum volume and use of a drying step were intended to minimize any changes in pH or aw (not actually measured). Any changes that occurred would have resulted in higher pH or aw and thus would have been more likely to encourage S. aureus survival than to reduce it. Six of the nine slices per sample were vacuum packaged (0.8 atm; Food Saver bags and packaging machine, Tilia, Inc., San Francisco, Calif.) and stored at 218C for sampling after 1 and 4 weeks. The three remaining product slices were analyzed immediately for the number of S. aureus cells per sample. Each entire product slice was aseptically added to a sample bag with 99 ml of BPD and stomached for 2 min at medium speed (Stomacher 400 lab blender, Fisher) to attain an initial dilution of about 1020.8 g/ml or 1020.8 cm2/ml. This initial dilution was conservatively denoted as 1021. Serial dilutions were made in BPD, plated on PF-SE, and incubated at 358C for 24 h. The PF-SE plates were then examined for the typical red to purple colonies of S. aureus. For each RTE meat product, one presumptive S. aureus colony was selected at each sampling time for confirmation testing. One typical colony was transferred to BHIA and incubated at 358C for 24 h. Resulting colonies were tested for Gram stain reaction, cellular morphology, and catalase activity. Throughout the study, all presumptive isolates were confirmed as S. aureus. Three pieces of each product that had been vacuum packaged and stored at 218C were analyzed after 1 and 4 weeks. At each sampling time, the log CFU per square centimeter was calculated for each piece, and the mean value was calculated for each product. A value of 0.9 log CFU/cm2 was assigned, based on a conservative assumption of 9 CFU/cm2 (12) when no colonies were present for the least dilute plating. Similarly, values of 2.0 and 3.0 log CFU/cm2 were assigned when no colonies were detected on 1022 and 1023 plates, respectively. Statistical analysis. To evaluate the pairwise relationship between pH and MPR, aw, or %WPS and the change in S. aureus populations (in log CFU), correlation coefficients (r values) were computed (16). To evaluate the overall relationship between change in S. aureus populations and pairs of pH and another compositional value, linear regression was performed (16). An R2 value, which indicates the proportion of variation in a response variable (i.e., change in S. aureus population) that is explained by linear regression on one or more independent variables, was used to assess the goodness of fit of each linear regression model. RESULTS AND DISCUSSION Most of the products studied met applicable U.S. Department of Agriculture (USDA) compositional standards J. Food Prot., Vol. 68, No. 9 FATE OF S. AUREUS ON RTE MEAT PRODUCTS 1913 TABLE 1. Staphylococcus aureus on certain ready-to-eat meat products stored under vacuum at 218C Product category Processor Compositiona MPR aw %WPS pH S. aureus (log CFU/cm2) atb: 0 days 7 days 28 days Jerky A A B B 0.4 0.7 0.6 0.8 0.68 0.82 0.69 0.76 18.7 10.6 15.4 14.2 5.7 6.0 6.4 6.1 5.9 6 0.2 6.6 6 0.3 5.9 6 0 5.9 6 0 3.3 6 0.8 5.1 6 0.5 4.2 6 0.2 4.9 6 0.1 2.3 6 0.6 (2) 2.8 6 0.5 1.4 6 0.4 2.7 6 0.2 Beef snack stick C D 2.0 1.7 0.88 0.85 7.1 9.0 4.6 4.9 5.4 6 0.2 5.9 6 0.2 2.0 6 0 (3) 2.0 6 0 (3) 0.9 6 0 (3) 1.2 6 0.4 (1) Pepperoni E F 0.9 1.7 0.76 0.88 19.0 11.6 4.9 4.9 5.9 6 0.1 6.3 6 0.1 2.9 6 0.1 2.7 6 0.1 1.7 6 0.5 0.9 6 0 (3) G 1.5 0.86 13.1 4.6 6.5 (n 5 2) 2.0 6 0.1 (2) 0.9 6 0 (3) Salami, dried J E E J E Q 1.5 1.1 1.0 1.7 2.5 1.6 0.88 0.79 0.76 0.87 0.92 0.87 9.3 16.3 17.1 9.3 6.5 8.0 4.9 5.1 4.8 4.9 4.9 4.8 6.2 6 0.3 6.4 6 0.1 5.7 6 0.2 6.2 6 0.1 6.1 6 0.1 6.1 6 0.1 5.8 6 0 3.6 6 0.1 LEc LE LE LE 3.2 6 0.6 (2) 2.5 6 0.1 0.9 6 0 (3) 2.0 6 0.1 2.2 6 1.1 3.2 6 0.3 Summer sausage D H L C P 3.2 3.3 3.0 3.1 2.7 0.93 0.95 0.93 0.96 0.95 5.8 4.3 5.7 4.7 6.5 4.8 4.5 4.7 4.4 4.5 6.0 6 0.1 6.3 6 0.1 6.2 6 0.2 4.4 6 0.5 5.6 6 0.1 4.9 6 0.2 3.9 6 0.4 2.4 6 0.5 3.0 6 0 (1) 3.0 6 0 (3) 2.1 6 0.3 (1) 0.9 6 0 (3) 0.9 6 0 (3) 0.9 6 0 (3) 0.9 6 0.1 (2) H R 2.9 3.2 0.96 0.94 4.5 5.9 4.5 4.5 5.7 6 0.1 5.0 6 0.3 3.0 6 0.1 (2) 3.0 6 0 (3) 0.9 6 0 (3) 0.9 6 0 (3) E D D K K 2.4 3.0 3.3 3.2 2.9 0.94 0.95 0.93 0.94 0.95 6.5 5.2 5.9 5.1 5.0 4.9 4.9 4.9 4.8 4.8 5.8 6 0.1 5.6 6 0.1 5.8 6 0.2 6.0 6 0.1 5.5 6 0.3 5.1 6 0.3 4.7 6 0.5 5.3 6 0.1 5.3 6 0.1 5.1 6 0.3 1.4 6 0.9 (1) 3.1 6 0.8 4.7 6 1.0 4.5 6 0.1 4.3 6 0.5 a Compositional values are for a single representative sample. MPR, moisture:protein ratio; %WPS, % water-phase salt (brine concentration). b Each log CFU per square centimeter value is the mean 6 standard deviation of three samples unless otherwise indicated. The number in parentheses denotes the number of samples that resulted in no colonies on the least dilute plate. Values of 0.9, 2.0, and 3.0 were assigned when this plate was the 1021, 1022, and 1023 dilution, respectively. c LE, lab error (no data). TABLE 2. Staphylococcus aureus on ready-to-eat meat products during storage under vacuum at 218C Product category Processor Compositiona MPRa aw %WPS pH S. aureus (log CFU/cm2) atb: 0 days 7 days 28 days Bologna Salami, cooked H I M N 5.3 4.0 4.7 4.4 0.97 0.96 0.96 0.95 2.9 3.7 3.7 3.9 6.3 6.2 6.6 6.4 6.4 6 0 6.1 6 0.1 6.3 6 0.1 6.2 6 0 6.5 6 0 6.5 6 0 LEc LE 8.5 6 0.1 7.5 6 0 8.9 (n 5 2) 8.3 6 0.1 O O P 5.4 4.2 3.8 0.97 0.95 0.95 2.0 3.8 5.4 6.6 6.5 6.2 6.3 6 0.1 6.3 6 0.1 6.1 6 0.1 LE LE LE 8.5 (n 5 2) 8.1 (n 5 2) 8.7 6 0.1 a Compositional values are for a single representative sample. MPR, moisture:protein ratio; %WPS, % water-phase salt. b Each log CFU per square centimeter value is the mean 6 standard deviation of three samples unless otherwise indicated. c LE, lab error (no data). for either product identity or shelf stability (nonrefrigerated semidry sausage). The level of nonstandard products consistently found in the marketplace is not known. Of the four jerky products tested, three met the USDA MPR standard for jerky of #0.75. A fourth sample had an MPR value of 0.80. However, on all four jerky products S. aureus decreased in numbers by 1.0 to 2.6 log CFU after 1 week and by 3.2 to 4.5 log CFU after 4 weeks (Table 1). Of three pepperoni products studied, two met the USDA standard of an MPR of #1.6, and one had an MPR value of 1.7. S. aureus decreased in numbers on pepperoni products by 3.0 to 4.5 log CFU after 1 week and was undetectable on two J. Food Prot., Vol. 68, No. 9 1914 INGHAM ET AL. TABLE 3. Mathematical prediction of change in Staphylococcus aureus populations (Y) on vacuum-packaged ready-to-eat meat products during 4 weeks of storage at 218Ca Predictive variables Equation of regression line R2 pH and MPR pH and aw pH and %WPS Y 5 216.21 1 1.08(MPR) 1 2.05(pH) Y 5 232.03 1 16.25(aw) 1 2.83(pH) Y 5 214.01 2 0.25(%WPS) 1 2.56(pH) 0.84 0.81 0.77 a Predictive variables are pH and moisture:protein ratio (MPR), water activity (aw), or % water-phase salt (%WPS). The R2 value indicates the proportion of variation in the change in S. aureus populations that is explained by linear regression on pH and one other variable. The three other variables are listed in decreasing order of importance. pepperoni products after 4 weeks (Table 1). A total of six dried salami products were studied; five products met the USDA MPR standard of #1.9, and one had an MPR of 2.5. On dried salami, S. aureus did not survive well, with decreases of 2.9 to 4.8 log CFU after 4 weeks (Table 1). In terms of pathogenic bacterial growth, all of the jerky, pepperoni, and dried salami products tested could be considered shelf stable. At the time of purchase, however, only the jerky and some of the dried salami products were displayed at ambient temperatures. The USDA standards for a nonrefrigerated semidry sausage product are that it (i) have an MPR of #3.1 and a pH of #5.0, (ii) have an MPR of #1.9 at any pH, or (iii) have a pH of #4.5 (or 4.6 combined with an aw of ,0.91) and an internal %WPS of $5.0% and must be sold in an intact form (or sliced and vacuum packaged), cured with nitrite or nitrate, and smoked with wood (4). The two beef snack stick products tested met these criteria, and for these two products S. aureus numbers decreased 3.4 to 3.9 log CFU after 1 week (Table 1). These products were, however, displayed under refrigeration at the grocery store. Five summer sausage products of the 12 tested did not meet either of the first two options of the USDA shelf-stability standards. In each of the five products, the MPR exceeded 3.1. However, the product pH values were 4.5 to 4.9, and one product met the third option in the shelf-stability standard. For these five products, S. aureus numbers declined by 0.5 to 2.4 log CFU after 1 week and by 1.1 to 5.4 log CFU after 4 weeks (Table 1). For summer sausages not meeting any of the options in the shelf-stability standard, labels did not indicate that the packaged product should be refrigerated prior to opening the package. On the seven summer sausage products with MPR # 3.1, S. aureus decreased in numbers by 0.4 to 3.8 log CFU after 1 week and by 1.2 to 5.3 log CFU after 4 weeks (Table 1). Strict adherence to the shelf-stability standard did not seem to be related to S. aureus survival on the summer sausages tested. Summer sausage compositional factors other than MPR, aw, %WPS, and pH (e.g., spices or smokeborne compounds) may have affected S. aureus survival on these products. Dramatic differences in day 0 S. aureus concentrations for summer sausages made by processors C and R may reflect interproduct differences in spices or smoking procedures used. As expected, the cooked salami and bologna products did not meet USDA shelf-stability standards and supported S. aureus growth (1.4 to 2.6 log CFU increase after 4 weeks; Table 2). The results of these inoculation studies suggest that the USDA shelf-stability standard and the product standards for beef jerky, pepperoni, and dried salami are conservative in terms of preventing pathogenic bacterial growth. Statistical analysis revealed that pH could be used in combination with MPR, aw, or %WPS to predict the change in log CFU for S. aureus on vacuum-packaged RTE meat products during 4 weeks of 218C storage. Regression equations were developed to enable predictions; the R2 values were 0.77 when %WPS and pH were the predictive variables, 0.81 when aw and pH were the predictive variables, and 0.84 when MPR and pH were the predictive variables (Table 3). Higher R2 values may have been obtained if three predictive variables had been used. However, this threevariable analysis was not done because of the low likelihood of a processor obtaining data on all three compositional values for their products. Examination of the correlation coefficients (r values) associated with the Table 3 regression equations revealed that pH had the greatest impact on change in S. aureus numbers (r 5 0.75), followed by MPR (r 5 0.74), %WPS (r 5 20.52), and aw (r 5 0.44). The regression equations do not account for temperature, inoculum concentration, interstrain differences in survival, or inhibitory substances such as spices or smokeborne compounds that may affect S. aureus survival in other situations. Therefore, further experiments and statistical analyses are necessary prior to widespread application of regression-based predictions. None of the products tested were scored in the category of a change in S. aureus numbers from 21.1 to 1.3 log CFU after 4 weeks. Given this gap in the data set for which the regression equations were determined, the boundary for predicting shelf stability could be set as conditions resulting in at least a 1.1-log decrease in S. aureus numbers. Using this criterion, our experimental results, and the r values obtained, processors could consider vacuum-packaged summer sausage with a pH of #4.9 and an MPR of #3.3 to be shelf stable in terms of pathogenic bacterial growth. The minimum pH value for S. aureus toxigenesis has been reported as 5.3 in certain sausage products (2), 5.4 or 5.0 under anaerobic conditions in laboratory medium (7, 11), and 4.5 under aerobic conditions in laboratory medium (11). At pH 5.3, the predicted MPR, aw, and %WPS resulting in a decrease in S. aureus numbers of 1.1 log CFU after 4 weeks (calculated from Table 3 equations) were 3.9, 0.98, and 2.6, respectively. In isolation, each of these values hardly appears likely to restrict S. aureus growth. Thus, J. Food Prot., Vol. 68, No. 9 FATE OF S. AUREUS ON RTE MEAT PRODUCTS 1915 virtually any fermented (pH # 5.3) RTE meat product could be considered shelf stable when vacuum packaged and stored at room temperature. This conclusion is clearly less restrictive than that suggested by our experimental results. Further research is warranted to determine more precisely the limits of RTE meat product shelf stability. In our study, the nonfermented, nondried RTE products supported S. aureus growth. However, the question remains whether relatively high-salt, moderately dried RTE meat products actually would support S. aureus growth. According to predictions derived from Table 3 equations, the MPR or aw must be reduced to 2.6 or 0.86, respectively, or the %WPS must be at least 9.4 for a high-pH (6.0) product to cause a 1.1-log decrease in S. aureus numbers after 4 weeks. These predictions should be tested in future studies. The regression equations in Table 3 did not appear to be useful for products with a pH typical of postrigor meat. For pH 5.5, the MPR predicted for a 1.1-log decrease in S. aureus numbers was 3.5, a value typical of raw meat. The predicted aw and %WPS values for a 1.1-log decrease in S. aureus numbers (0.95 and 4.3, respectively) also appear unlikely to inhibit growth. Clearly, further studies are necessary to test and refine the predictive equations before such tools could be used by meat processors. The results presented here strongly suggest that the USDA standards for nonrefrigerated semidry shelf-stable sausage are conservative in terms of pathogenic bacterial growth during vacuum-packaged storage at 218C. Laboratory studies clearly indicated that S. aureus numbers decreased on fermented (pH # 5.1) products with a wide range of salt concentrations and moisture content. Similarly, S. aureus numbers decreased on dried (aw # 0.82 or MPR # 0.8) nonacidified jerky products. RTE meat products that were not fermented or dried clearly supported S. aureus growth and cannot be considered shelf stable. The product pH and MPR were the compositional factors best correlated with S. aureus survival and growth for the types of products tested, but pH and aw, or pH and %WPS also may provide useful predictive guidance for S. aureus growth. These compositional factors probably could not be used to predict shelf stability of unfermented products that had been subjected to mild drying and/or salting treatments. Further studies are needed to develop predictive models for the shelf stability of those products.