Behavior of spoilage bacteria and Salmonella enterica subspecies enterica O:4,5 in vacuum-packaged beef during refrigeration

In this study, the kinetic parameters of mesophilic, psychrotrophic and lactic acid bacteria in vacuum-packed beef at 1 °C and 4 °C were estimated from experimental growth curves produced by samples stored during 21 and 60 days, respectively. In a separate experiment, the survival of multidrug resistant (MDR) Salmonella enterica O:4,5 at 1°C was also characterized. The shelf-life of vacuum-packed beef stored at 4 °C was estimated at 16.1 days (95% CI: 14.8 – 17.3 days), whereas at 1 °C it was longer than 21 days because the mesophiles count estimated towards the end of the experiment was 12.5 ln CFU.g-1 (95% CI: 11.8 – 13.3 ln CFU.g-1) which is lower than the shelf-life reference value. At 1 °C, inoculated Salmonella was reduced in 6.61 ln CFU.g-1 (2.87 log CFU.g-1). These results demonstrated the importance of establishing in legislation, especially in Brazil, standard values of deteriorating microorganisms in beef for maintaining product quality.


INTRODUCTION
The meat deterioration process is a combination of complex biological and chemical events caused by different bacterial populations. The most frequent microorganisms in fresh meat are Acinetobacter, Pseudomonas, Brochothrix, Flavobacterium, Psychrobacter, Moraxella, Staphylococcus, Micrococcus, lactic acid bacteria (LAB) and different Enterobacteriace genera (DOULGERAKI et al., 2012). These microorganisms are known as specific spoilage organisms (SSO) and understanding the development of this microbial community is paramount to ensure meat quality. Meat deterioration is characterized by discoloration, strong odor and slime production. It is mainly dependent of the storage temperature and packaging conditions (EFSA, 2014).
The shelf-life of a product is the time that it can be stored under appropriate conditions without changing its sensory characteristics in order to be fit for consumption. Maintenance of the cold storage chain is indispensable for controlling of SSO. The meat must be stored, transported and marketed at a temperature not exceeding 7 °C (BRAZIL, 1996). Despite the temperature limit recommendations, vacuum-packed beef shelf-life varies between different countries, estimated at 24 to 26 weeks at -1 °C in Australia (MLA, 2016), 6 weeks at 0 °C in Europe (EFSA, 2014), 9 weeks at -2.2°C in the USA (NATIONAL Silva et al. CATTLEMEN 'S BEEF ASSOCIATION, 2009) and 18 weeks at -1.5°C in Canada (GILL et al., 2013).
Salmonella is a foodborne pathogen frequently recovered from beef, as reported by the Food Safety and Inspection Service (FSIS, 2016). Thus, studies on Salmonella behavior under the recommended refrigeration temperatures are required. Predictive models describing microbiological growth or inactivation influenced by specific environmental conditions such as temperature, pH and water activity may predict process improvements and ensure food safety (GONZALES-BARRON et al., 2014).
Thus, considering the variability between refrigerated and vacuum-packed beef shelf-lives, defined by different legislations around the world, and the necessity to predict Salmonella behavior under such conditions, this study aimed to estimate the shelf-life of vacuum-packaged beef stored at temperatures between 1 °C and 4 °C, and estimate the extent of Salmonella inactivation when it is vacuumpacked and stored at 1 °C.

MATERIALS AND METHODS
Sirloin samples (Longissimus dorsi) were purchased from an abattoir located in the city of Cuiabá, Mato Grosso state, Brazil. This establishment works under the Federal Inspection Service (SIF) and complying with Good Manufacturing Practices (GMP) conditions. Total mesophilic microorganisms, psychrotrophic and lactic acid bacteria counts were quantified to characterize initial loads.
This research was divided into two experiments: the first (shelf-life study) evaluating mesophilic, psychrotrophic and lactic acid bacteria counts in vacuum-packed beef stored at 1 °C ± 0.5 °C and 4 °C ± 0.5 °C; and the second study checking the behavior of Salmonella enterica O:4,5 multidrug resistant (MDR) inoculated in vacuum-packed beef stored at 1 ºC ± 0.5 °C. A total of 45 steaks (21 for analysis at 1 °C and 24 for evaluation at 4 °C) weighing 100g each were used in the first experiment. For the second experiment, a total of 15 steaks weighing 25g each were analyzed. Three steaks per analysis day were analyzed in both experiments. All steaks were individually vacuum-packaged in low-density, gastight polyethylene bags (150 cm 3 /24h.m 2 .bar to CO 2 , 35 cm 3 /24h.m 2 .bar to O 2 and 1.4 cm 3 /24h.m 2 .bar to N 2 at 22 ºC) with the aid of BD-420 vacuum sealer (R-Baião, Ubá, Minas Gerais, Brazil).
The steaks were stored in a Biochemical Oxygen Demand (B.O.D.) incubator (model BT 71,Biotech,Piracicaba,São Paulo,Brazil) at 1 °C ± 0.5 °C for 21 days, and 4 °C ± 0.5 °C for 60 days. Total plate colony counts were performed on days 0, 3, 7, 13, 16, 20, 21 and 0, 1, 4, 6, 10, 32, 46 and 60 for mesophilic aerobic bacteria, psychrotrophic bacteria and lactic acid bacteria according to the methods proposed by American Public Health Association (APHA, 2001). To this end, 10 g samples were separated and mixed with 90 mL of a peptone saline solution (0.85% NaCl and 0.1% peptone) and serial dilutions were prepared. For the mesophilic bacteria analyses, 0.1mL of the appropriate dilutions were added by the spread plate technique in plate count agar (PCA) (Kasvi, São José dos Pinhais, Paraná, Brazil) and incubated at 36 °C ± 1 °C for 48 hours. The same inoculation and culture medium were used for psychrotrophic counts, with incubations carried out at 7 °C ± 1 °C for 10 days. For the lactic acid bacteria counts, 1.0 mL inoculum of appropriate dilutions was placed on the plate and MRS agar (Kasvi, São José dos Pinhais, Paraná, Brazil) was added by the pour plate technique with overlays, followed by incubation under anaerobiosis at 30 °C ± 1 °C for 48 hours. Three steaks were analyzed at each sampling point, in duplicate per dilution, and the pH value was determined in triplicate for every analysis day for each steak, using a digital pH meter (Kasvi, São José dos Pinhais, Paraná, Brazil) after adding 10-grams sample in 90 mL of distilled water.
In the second experiment, the meat samples were previously evaluated to determine the absence of Salmonella spp. (ISO, 2002). The Salmonella enterica subspecies enterica O:4,5, a MDR organism previously characterized by CUNHA-NETO et al. (2018), was used in this study. The isolated strain was stored at -80 °C. A loop full of isolated strain was added to 9 mL of TSB broth (Kasvi, São José dos Pinhais, Paraná, Brazil) and incubated at 36°C ± 1°C for 24 hours. Serial dilutions (up to 10 -10 ) were prepared in peptone saline solution, and 0.1 mL of each dilution, in duplicate, were added, spread on XLD agar plate (Kasvi, São José dos Pinhais, Paraná, Brazil). These plates were incubated at 36 °C ± 1°C for 24 hours. The characteristic colonies were counted, and the Salmonella concentrations in TSB broth were determined. Subsequently, the inoculum concentration was adjusted up to 6 log CFU.g -1 . Aliquots (60 mL) of this inoculum were prepared and 25g steaks were immersed in this inoculum for 1 minute. After drying for 60s in a laminar flux chamber, the steaks were individually vacuum-packed and stored in a B.O.D. incubator (model BT 71,Biotech,Piracicaba,São Paulo,Brazil) at 1 °C ± 0.5 °C for 21 days.
The analyses were conducted on the initial day and on days 3, 6, 12 and 21 of storage.
Total plate count analyses for Salmonella colonies on XLD agar (Kasvi, São José dos Pinhais, Paraná, Brazil) were performed. On each day, three 25g steaks were removed from the B.O.D. incubator, unpacked and individually stomached (model MA 440/CF, Marconi, Piracicaba, São Paulo, Brazil). One steak sample was homogenized for 1 minute in 225 mL of peptone saline solution (0.85% NaCl and 0.1% peptone). Serial dilutions (up to 10 -4 ) were prepared, and 0.1 mL of the appropriate dilutions were spread onto XLD agar and incubated at 36 °C ± 1 °C for 24 hours. Counting was performed with the aid of a colony counter (Prolab, São Paulo, Brazil), and the results were expressed in log CFU.g -1 .
Both experiments were conducted in a completely randomized design with the data expressed as the means and standard error (SE). For the shelflife analysis, the log CFU.g -1 counts were converted to natural logarithm, and the reduced Huang model without lag phase (HUANG, 2008) was adjusted to each of the experimental growth curves of the spoilage indicators, using the statistical package R (R DEVELOPMENT CORE TEAM, 2018). The kinetic parameters, initial microbial load (Y 0 in ln CFU.g -1 ), maximum microbial load (Y max in ln CFU.g -1 ) and maximum growth rate (μ max in h -1 ) were estimated for each of the three microbial groups at both 1 ºC and 4 ºC.
In order to determine Salmonella kinetic parameters in vacuum-packed meat at 1 ºC, an empirical exponential decay model (Equation 1) without shoulder was fitted to the experimental survival curve (LEIKE, 2001), where Y(t) represents the microbial counts at time t, Y 0 the initial microbial counts (ln CFU.g -1 ), Y final the final asymptotic microbial concentration (ln CFU.g -1 ) and α the decay rate (day -1 ).

RESULTS AND DISCUSSION
Initial mesophile concentrations (Y 0 ) were determined as 4.33 and 5.71 ln CFU.g -1 respectively for 1 °C and 4 °C, while initial lactic acid bacteria concentrations were lower, at 3.53 ln CFU.g -1 at 1 °C and 3.17 ln CFU.g -1 at 4 °C. Psychrotrophic bacteria displayed a low initial concentration at 1 °C of 3.73 ln CFU.g -1 , and a higher concentration of 5.57 ln CFU.g -1 at 4 °C (Table 1).
For mesophiles at 1 °C, the end of the stationary phase was not evidenced after 21 days storage period by evaluating the adjusted growth curve ( Figure 1A). This indicated that the maximum concentration is likely to be greater than 12.50 ln CFU.g -1 determined at the end of the experiment, while at 4 ºC the end of the stationary phase occurred close to 21 days of storage, at a concentration of 20.19 ln CFU.g -1 ( Figure 1B).
Concerning the lactic acid bacteria growth curve at 1 ºC ( Figure 1C), microbial concentration was determined with 3.53 ln CFU.g -1 , without graphically displaying a relative exponential increase. At 4 ºC Table 1 -Means and standard error of the kinetic parameters for mesophilic, lactic acid and psychrotrophic bacteria in vacuum-packed beef stored at 1 °C and 4 °C during 21 and 60 days respectively, and means, standard error and p-values of the parameter estimates for Salmonella enterica subspecies enterica O:4,5 in vacuum-packed meat stored at 1 ºC for 21 days.
Silva et al.
( Figure 1D) the beginning of the stationary phase was observed shortly before approximately 19 days of storage, with a maximum concentration of 10.93 ln CFU.g -1 , and a 7.76 ln CFU.g -1 increase. In a study carried by REID et al. (2017), increased counts were observed from an initial concentration of 6.21 ln CFU/cm² to 13.57 ln CFU/ cm² for mesophilic counts, and 3.63 ln CFU/cm² to 11.34 ln CFU/cm² for LAB counts respectively in vacuum-packed meat stored at 2 °C for 6 weeks. BARROS-VELÁZQUEZ et al. (2003) analyzed mesophiles and lactic acid bacteria concentrations in vacuum-packed meat stored at 4 ºC for 70 days. They determined an initial amount of 10.88 ln CFU.g -1 for mesophiles and 9.29 ln CFU.g -1 for lactic acid bacteria. At 20 days of storage, the concentrations of both microorganism groups were greater than 18.4 ln CFU.g -1 , while at the end of 40 days, the concentrations of mesophiles and LAB were 25.3 and 21.96 ln CFU.g -1 , respectively. Concerning the psychrotrophic bacteria adjusted growth curves, the end of the exponential phase was reached approximately between 17 and 21 days of storage at 1 ºC ( Figure 1E), stabilizing at a maximum concentration of 14.7 ln CFU.g -1 , but for 4 ºC ( Figure 1F) psychrotrophic bacteria were in exponential phase at the end of 21 days, reaching 18.93 ln CFU.g -1 . It is important to note that all the microorganisms reached the stationary phase at 4 ºC.
No standards for mesophilic, psychrotrophic and lactic acid bacteria are established in Brazilian legislation (BRAZIL, 2001). However, the current revision of the law recommends as unacceptable aerobic mesophiles concentrations above of 6 log CFU.g -1 (13.82 ln CFU.g -1 ) in beef (BRAZIL, 2019). The European Community Commission recommends a maximum limit of 15.38 ln CFU.g -1 in cattle meat, while Chinese legislation considers values over 13.8 ln CFU.g -1 as unsatisfactory (REGULATION E.C., 2005;CENTRE FOR FOOD SAFETY, 2014). These recommendations are lower than the result observed for this microorganism at 4 ºC ( Table 1).
The International Commission on Microbiological Specifications for Foods (ICMSF) advocates a count of 10 7 CFU.g -1 (16.1 ln CFU.g -1 ) of mesophiles as reference value in beef. The large number of microorganisms results in decreased meat glucose content, and the energy metabolism of deteriorating microorganisms degrades amino acids, producing metabolites such as ammonia, amines, sulfur compounds, aldehydes and ketones, which generate unpleasant odor, color and taste (ICMSF, 1986).
Based on the reference value of 16.1 ln CFU.g -1 of mesophiles counts for end of shelf-life (ICMSF, 1986), in the present study we can state that, at 1 °C, the shelf-life of vacuum-packed beef is longer than 21 days since the maximum value of mesophiles at the end of the experiment was 12.5 ln CFU.g -1 (95% CI: 13.3 -11.8 ln CFU.g -1 ) ( Table 1, and Figure 1 A). For storage at 4 °C, the shelf-life of vacuum-packed beef was 16.1 days (95% CI: 14.8 -17.3 days) ( Figure 1B), where the concentration of mesophiles was exceeded the value recommended by the ICMSF (1986), as mesophile growth rates at 4 °C were higher than at 1 °C.
Beef producing industries must pay attention to the shelf-life studies of the product, as a transition from 1°C to 4°C is sufficient to promote considerable increases in the growth rate of these spoilage microorganisms, and it is paramount to maintain the lowest possible temperatures in all processes.
Regarding pH values in the vacuum-packed meat, no significant change between the initial and final concentrations (P > 0.05) were observed for the meat samples stored at 1 ºC (5.497 ± 0.135 to 5.463 ± 0.069) or 4 ºC (5.696 ± 0.153 to 5.840 ± 0.129); although, the meat samples stored at 4 ºC presented greater pH variations.
In the study of Salmonella behavior in vacuum-packed beef during refrigeration, the initial Salmonella concentration was of 13.94 ln CFU.g -1 , and it declines in function of the time as can be graphically observed from the beginning of the experiment (Figure 1 G).
At the end of 21 days of storage, the final Salmonella concentration was of 7.324 ln CFU.g -1 (Table 1), demonstrating the action of low temperature and vacuum package in reducing the pathogen's population viability. In addition, this decline may also have occurred due to the accumulation of toxic metabolites by SSO growth during this period (Table  1, Figures 1A, 1C and 1E).
During the 21 days, a decline (P<.0001) in Salmonella concentrations from 6.616 ln CFU.g -1 , almost half of the initial concentration with decline rate of the 0.3 ln CFU.g -1 /day, occurred. However, it should be noted that from the initial concentration used, the microorganisms were not inactivated during 21 days under cooling.
Decreased Salmonella growth has been previously reported in vacuum-packed minced meat (DJORDJEVIĆ et al., 2018). Conversely, a study using a much lower storage temperature did not detect significant differences in the amounts of S. Typhimurium and S. Brandenberg inoculated in vacuum-packed meat stored at -1.5°C for 8 weeks (DYKES, 2001). Silva et al. The decay rate of Salmonella enterica O:4,5 in vacuum-packed beef meat estimated in this study (0.300 ± 0.083 ln CFU.g -1 .d -1 ; Table 1) is comparable to the inactivation rates (0.365 -0.578 ln CFU.g -1 .d -1 ) estimated in another study that assessed the evolution of Salmonella Typhimurium in vacuum-packed pork meat, also stored at 1 °C ( VAN LAACK et al., 1993). The extent of comparability is noteworthy, considering the differences in food matrix (beef versus pork), microbial quantification method (CFU versus most-probable-number) and Salmonella strain.
Maintaining Salmonella populations at low temperatures (cooling and/or freezing) during food storage represents a significant physiological challenge for the microorganism, promoting bacterial stress and metabolism alterations, such as cell membrane structure, DNA and transcriptional and translational response modifications. These changes render essential molecular processes, such as protein biosynthesis, ineffective; therefore, influencing bacterial growth rates (RICKE et al., 2018). Understanding Salmonella behavior in low temperatures can aid public health authorities, as well as all those involved in cattle meat marketing, on adequately storing meat and meat derivatives.

CONCLUSION
A shelf-life of over 21 days was estimated for vacuum-packed meat stored at 1 °C, while the shelf-life of meat stored at 4 ºC was 16.1 days. This could support regulation and decision-making within the meat industry, preventing the marketing of low nutritional value meat. Concerning the Salmonella enterica enterica serovar O:4,5, a growth rate decline was noted at 1 °C, demonstrating that lower conservation temperatures in conjunction with vacuum packaging is an effective strategy to reduce consumer exposure to Salmonella through the consumption of beef meat.