INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Escherichia coli O157:H7 is commonly associated with foods of animal origin, and dairy cattle are a major reservoir of the organism (3, 14). Several outbreaks of hemorrhagic colitis and hemolytic-uremic syndrome have been attributed to consumption of undercooked ground beef (6, 8, 15, 19, 25), although it appears that the incidence of the pathogen in ground beef and the frequency of isolation of the pathogen from this food are low (8). It is now recognized that refrigeration is not sufficient to eliminate growth of a number of food-borne pathogens, including enterohemorrhagic E. coli. Growth and verotoxin production in broth have been reported to occur at temperatures as low as 7.9°C (16). Moreover, ground beef may be exposed to temperatures of 5 to 10°C during handling, conditions that may permit survival and growth of the pathogen.

Although ground beef has been a major food implicated in E. coli outbreaks, there have been limited studies on the fate of the organism in fresh beef held at low temperatures. The initial numbers of background microorganisms in market ground meats are high, typically 5 to 6 log₁₀ CFU per g, and they rapidly increase during storage at low temperatures. In contrast, the numbers of pathogen cells reported to cause illness after consumption of beef patties are very low. Levels of less than 5 CFU of E. coli O157:H7 per g were reported in ground beef that was implicated in a 1993 outbreak (4, 12). As a result, cultural methods may not detect the pathogen when it is present in foods at levels considerably lower than the levels of the background microorganisms. Several conventional methods used for detection of low numbers of the pathogen were found to be unsuitable and very time-consuming (13). In order to circumvent these problems, Palumbo and coworkers irradiated meat with 3 kGy to reduce the number of background microorganisms from 6 to 3 log₁₀ CFU per g before inoculation with a similar level of the pathogen (17). Viable E. coli O157:H7 cells were enumerated by surface plating on sorbitol MacConkey agar, and the background organisms were enumerated on tryptic soy agar plates. Ansay and coworkers treated ground beef with 5.4 to 7.5 kGy prior to inoculation with E. coli strains in their study of survival of the pathogen during frozen and refrigerated storage (2). Cell numbers were determined on sorbitol MacConkey agar.

The purpose of the present study was to monitor the fate of E. coli O157:H7 in ground beef stored from freshness to spoilage at refrigerator (2°C) and abuse (10°C) temperatures by using a strain expressing the Aequorea victoria green fluorescent protein (gfp) gene. The gfp gene was cloned (18), and a recombinant strain of the pathogen was constructed by Fratamico and coworkers (7). Sodium salts of lactate (SL), diacetate (SDA), and citrate have been reported to inhibit microorganisms in meats (1, 20-23). The effects of these compounds on E. coli O157:H7 and the background microorganisms were also examined in this study.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Test organism and inoculum preparation. E. coli O157:H7 strain SEA 13B88 (from the Odwalla cider outbreak; Food and Drug Administration) transformed with plasmid pGFPuv by Fratamico and coworkers at the Eastern Regional Research Center, USDA Agricultural Research Service, Wyndmoor, Pa., was used in this study (7). The 27-kDa green fluorescent protein encoded by the Aequorea gfp gene emits green light at 509 nm and is very stable in the presence of heat, detergents, and proteases (5, 24). For experiments, overnight cultures of the organism were grown in tryptic soy broth (Difco, Detroit, Mich.) at 35°C, and the cultures were diluted with sterile 0.1% peptone water (PW) before meat samples were inoculated.

Chemicals. The compounds tested were SL (PURAC America Inc., Lincolnshire, Ill.), SDA (Niacet, Niagara Falls, N.Y.), and a sodium salt of buffered citrate (SC), which was prepared by mixing sodium citrate (Allied Chemicals, New York, N.Y.) and citric acid (Fisher Scientific, Fair Lawn, N.J.) (15:1, by weight).

Meat samples. Fresh ground beef was obtained from local meat markets. For each trial the meat in the package was divided into five batches, including a control batch and batches treated with salts as follows (percentages were determined on a dry weight basis): 0.9 and 1.8% SL; 0.1 and 0.2% SDA; 1 and 2% SC; 0.9% SL plus 0.1% SDA; and 1.8% SL plus 0.2% SDA. All of the salt concentrations except 2% SC were equal to or less than the concentrations allowed for use with meat products (0.2% SDA, 2% SL, and 1.3% SC). Concentrated aqueous solutions of the salts were added to the meat samples at a level of 10 ml/kg. Water alone was added to the controls. After addition of the salts and thorough mixing, each batch was divided into two portions. One portion was used to study changes in the total-aerobe population in the meat during storage; the other was inoculated with the pGFP-bearing E. coli O157:H7 strain. The cultures (volume, 0.1 ml) were thoroughly mixed with 25-g meat samples in small plastic beakers covered with a double layer of aluminum foil. Meat samples containing E. coli O157:H7 at final concentrations of approximately 3 and 5 log₁₀ CFU/g were tested. All samples were stored aerobically at 2°C. Meat samples containing high levels of salts were also tested during storage at 10°C.

Enumeration of microorganisms. Duplicate beakers were removed from storage every 1 to 3 days and used for testing. Meat samples (11 g) were transferred from these beakers to a stomacher bag (Seward Medical, London, United Kingdom), 1:10 (wt/wt) dilutions were prepared with PW, and the contents were blended for 2 min (Stomacher 400; Seward). To determine aerobic plate counts, serial dilutions were made in PW, and the appropriate dilutions were plated onto prepoured plate count agar. Appropriate dilutions prepared with meat samples stored at 2°C were also plated onto Pseudomonas isolation agar (PIA) (Difco) and violet red bile agar (Difco) plates. Colonies were counted after incubation at 25°C for 48 h (PIA) or at 35°C for 24 h (violet red bile agar). Numbers of E. coli O157:H7 cells were determined by plating appropriate dilutions onto plate count agar plates and incubating the plates at 35°C for 24 h. Fluorescent colonies were visualized and counted under a UV lamp (wavelength, 350 to 400 nm; Raytech Industries, Inc., Stafford Springs, Conn.).

Sensory characteristics and pH measurements. Changes in the odor and the consistency of meat samples were recorded immediately after the samples were prepared and every 2 or 3 days for samples stored at 2°C and daily for samples stored at 10°C by three researchers trained in assessing fresh meat quality. Meat pH values were determined by inserting a pH electrode (model 720A; Orion Research Inc., Boston, Mass.) directly into meat homogenates (1:10 dilution).

Statistical analysis. Three trials were performed with low salt concentrations at 2°C, and two trials were performed with high salt concentrations at 2 and 10°C. Duplicate measurements were obtained for all samples. Data were analyzed by using the analysis of variance procedure of the SAS statistical package (SAS Institute Inc., Cary, N.C.). Means for total-aerobe populations were compared by using a least significant difference procedure at the 5% level of significance for each treatment at 2 and 10°C.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

The initial levels of total aerobes in the meat samples at the time of purchase ranged from 5.2 to 6.5 log₁₀ CFU/g. Spoilage began in controls and in samples that received inocula consisting of 3 or 5 log₁₀ CFU of the pathogen per g after 4 to 8 days of incubation at 2°C. The meat developed off odors, the surface became slimy, and the level of total aerobes (predominantly gram-negative organisms) was 7.5 log₁₀ CFU/g or higher. At the beginning of the experiments the pH of untreated meat and the pH of meat after the salts were added ranged from 5.5 to 5.9, and the pH increased as spoilage proceeded.

Effect of salts on total aerobes. The effects of low and high concentrations of the salts on the levels of total aerobes in meat stored at 2°C are shown in Fig. 1 and 2, respectively, for samples that were not inoculated with the pathogen. The data for samples inoculated with the pathogen were not significantly different (P < 0.05). There was not a significant difference (P > 0.05) between the levels of total aerobes in untreated samples and the levels of total aerobes in samples treated with 1 or 2% SC during storage at 2°C (Fig. 1 and 2). There was also no difference in the time of onset of spoilage, as assessed by sensory changes, in untreated and SC-treated meat. SL significantly (P < 0.05) suppressed bacterial growth and delayed the onset of spoilage (Fig. 1 and 2). The shelf life was influenced by lactate concentration, and the level of total aerobes was less than 7 log₁₀ CFU/g in meat treated with 1.8% SL even after 14 days. Treatment with SDA significantly (P < 0.05) inhibited the growth of the total-aerobe population, and this treatment was the most effective treatment for extending the shelf life of the meat (Fig. 1 and 2). There was not a significant difference between the effect of SDA alone and the effect of SDA in combination with SL (P > 0.05). At the incipient spoilage stage the aerobes in meat samples treated with SC or SL were predominantly gram-negative organisms, as they were in untreated samples. The PIA plate counts for meat samples treated with SDA declined by 2 log₁₀ CFU/g after refrigeration for 8 days or longer at 2°C. The orders of effectiveness of the compounds at low and high concentrations were similar (Fig. 1 and 2). The pH values of untreated and SC- or SL-treated meat samples increased rapidly during storage to 6.2 to 6.5 or more, but the pH values of SDA-treated samples did not change.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Aerobic plate counts in ground beef stored at 2°C for 14 days. Symbols: , control; , 1% SC; , 0.9% SL; , 0.1% SDA; , combination of 0.9% SL and 0.1% SDA.

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Aerobic plate counts in ground beef stored at 2°C for 14 days. Symbols: , control; , 2% SC; , 1.8% SL; , 0.2% SDA; , combination of 1.8% SL and 0.2% SDA.

View larger version (18K): [in this window] [in a new window]   FIG. 3.   Aerobic plate counts in ground beef stored at 10°C for 7 days. Symbols: , control; , 2% SC; , 1.8% SL; , 0.2% SDA; , combination of 1.8% SL and 0.2% SDA.

Behavior of E. coli O157:H7 in beef. Since the pGFP-containing strain was originally isolated from cider, we performed growth studies in tryptic soy broth and sterile cooked meat with the transformed strain, the parent strain, and an E. coli O157:H7 beef isolate (EC 505B). The results showed that the growth kinetics of the three strains were similar. Fluorescence of the gfp recombinant strain was stable throughout the study. The colonies present after 24 h of incubation were large, and the fluorescence was intense even in the presence of high numbers of background microorganisms. The transformed strain was recovered from untreated meat samples and each of the salt-treated meat samples throughout storage, even after 18 days of incubation at 2°C and 10 days of incubation at 10°C, demonstrating that the organism was resistant to the salts and the combinations of salts at both levels (data not shown). For all of the samples (the control samples and the samples that received the two levels of salts and the combinations), there were significant decreases (P < 0.05) in the number of E. coli O157:H7 cells (1 log₁₀ CFU/g) during storage at 2°C when the meat samples were considered spoiled (data not shown). The small decreases observed at 10°C were not significant. While the results obtained when we used inocula containing 3 and 5 log₁₀ CFU/g were similar, using the large inoculum facilitated enumeration of the strain.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In previous studies performed to determine the fate of E. coli O157:H7 in raw meat, the researchers had a problem enumerating low numbers of the pathogen in the presence of increasing numbers of the background microorganisms. This problem was overcome by eradicating some or all of the interfering microorganisms by irradiation prior to inoculation with the pathogen (2, 17). However, the drawbacks of this treatment include the loss of amino acids, the development of oxidation products from lipids, especially during aerobic storage, and other changes in the meat components that may affect survival and growth of microorganisms and mask the effect, if any, of the background microflora on the pathogen. The organism used in this study, E. coli O157:H7 with a stable pGFP marker, required neither modification of the level of background microorganisms nor selective media. Retention of the Shiga toxin genes, attaching-and-effacing gene, and the 60-MDa plasmid in the strain used was confirmed by Fratamico (6a).

The microbial quality of the samples tested in this study, which resulted in aerobic plate counts of 5 to 6 log₁₀ CFU/g, is characteristic of fresh retail ground beef in our area. Lactate (1.8%) suppressed the growth of the total-aerobe population, while diacetate (0.2%) inhibited most of these organisms during storage at 2°C. However, none of the treatments tested influenced the behavior of the pathogen to a great extent, and the organism could be recovered from each of the treatment preparations incubated at both 2 and 10°C, even when the level of total aerobes exceeded 8 to 9 log₁₀ CFU/g. Ansay and coworkers (2) reported a reduction of 1.87 log₁₀ CFU/g in the number of E. coli O157:H7 cells in irradiated beef stored for 4 weeks at 2°C. The number of E. coli O157:H7 cells was also reduced in ground beef stored at 3 to 5°C, and E. coli cells were detectable when the level of total aerobes was 7 log₁₀ CFU/g but not when the level of total aerobes was 9 log₁₀ CFU/g (11). Survival of the pathogen in our study improved during storage at 10°C compared to storage at 2°C. Survival but not increases in the populations of enterohemorrhagic E. coli strains was also observed in ground beef stored at 5°C (17).

It has been suggested that survival of pathogens, such as Listeria monocytogenes and E. coli O157:H7, may be enhanced in the absence of competitive microorganisms and that microbial interference may provide protection against pathogens (9-11). Competition for nutrients and for attachment to adhesion sites and unfavorable changes in the substrate environment are some of the factors that may suppress pathogens (9). In the present study decreases of 1 log₁₀ CFU/g in the pathogen level were observed at 2°C only when the aerobic plate counts were greater than 7 log₁₀ CFU/g. The presence of high levels of total aerobes suggests that unfavorable changes in the environment produced by these microorganisms affected the survival of the pathogen. However, the persistence of E. coli O157:H7 in refrigerated ground beef and detection of surviving organisms even at the frank spoilage stage indicate that this food should be considered hazardous if it is contaminated with the pathogen irrespective of the number of background microorganisms. The recombinant E. coli strain used in this study facilitated studies of changes in populations in the presence of increasing numbers of background microorganisms and in the presence of antimicrobial agents. This organism was resistant to salts of the organic acids tested but should be useful in screening other potential inhibitors.