Abstract
To optimize the sanitation treatment of ready-to-eat (RTE) intermediate-moisture foods (IMF), the behavior of Listeria monocytogenes Scott A (CIP 103575), L. innocua (NTC 11288), Salmonella enterica serovar Typhimurium (CECT 443), and Escherichia coli O157:H7 (CECT 4972) following treatment with electron-beam irradiation has been studied. As food matrixes, three RTE vacuum-packed products (Iberian dry-cured ham, dry beef [cecina], and smoked tuna) were used. Although an irradiation treatment is not necessary when the 102 colony-forming units/g microbiological criterion for L. monocytogenes is applied, a treatment of 1.5 kGy must be applied to IMFs to meet the food safety objective in the case of the “zero tolerance” criterion for the three strains. The IMF products presented negligible modifications of color (L*, a*, and b*), sensory (appearance, odor, and flavor), and rheology (hardness, springiness, adhesiveness, cohesiveness, gumminess, chewiness, and breaking strength) parameters at doses lower than 2 kGy. Therefore, the treatment of 1.5 kGy warrants safe IMF with sensory properties similar to those of the genuine products.
Introduction
T
An a w below 0.98 can have a marked effect on the microbiota of some foods. This is the case, for example, in production of dry fermented sausages where lactic acid bacteria and members of the family Micrococcaceae are dominant because of the reduction of the a w from 0.99 to 0.96 as a consequence of the addition of curing salts (Ordóñez et al., 1999). At a lower a w (less than 0.90–0.92) only the growth of Gram-positive cocci may be expected. No bacterial growth is observed at a w <0.85 (ICMSF, 1980).
Among pathogens, the bacteria of most concern in foods at a relative low a w are Staphylococcus aureus and Listeria monocytogenes. The former may grow even at an a w as low as 0.85 (Troller and Stinson, 1975), but no enterotoxin production has been detected at an a w lower than 0.93 (Smith et al., 1983). The European Commission No. 1441/2007 regulation (CEC, 2007) on microbiological criteria for food established the concentration of L. monocytogenes for RTE foods able to support the growth of this bacterium as an absence in a 25-g amount. However, the limit is a concentration below 100 colony-forming units (CFU)/g when L. monocytogenes is not able to multiply. Thus RTE products with an a w ≤0.92 are automatically included in this category.
Iberian dry-cured ham is made from the whole pork leg of the Iberian pig. Its sensory attributes, the manufacturing process, and other properties are reported elsewhere (López et al., 1992; Martín et al., 1999). In brief, hams are salted and dried (6–9 months). Afterward, they are ripened in cellars under ambient air for 18–24 months. The final product (commonly 6–8 kg) presents a pH of about 6.0, an a w of 0.85–0.88, and a moisture of around 45%. Cecina (Celtic “ciercina” means “cierzo,” the name of a Northern Iberian wind) is made with muscle pieces (7–12 kg) from the hind legs of beef. Cecina production is also described elsewhere (García et al., 1995). In brief, pieces are salted, dried, and cured (6–8 months) in a similar way to dry-cured ham, and they are usually smoked. The final pieces (3–6 kg) have a pH of close to 6.0, an a w <0.88, and a moisture of around 45%. Smoked tuna is made from separated muscle regions of the fish, which are then cut into strips or chunks (0.5–1.5 kg), salted, and cured for about a month. The final product presents an a w <0.88. Considering the a w of the three products, they may be included in the category of IMF. Therefore, these foods are microbiologically stable, and they are endowed with a very long shelf life (over 4–6 months). The only possible spoilage may be due to the growth of molds and fatty acid autooxidation when they are packed in a non-anoxic atmosphere. Likewise, they are of no concern from a public health perspective, and their consumption presents no hazard immediately after portions are cut from the whole piece.
The problem arises when these foods are transformed into RTE products. This involves several operations, namely, either cutting or slicing followed by packing, which increase the contamination risk. Even when those operations are performed under very strict hygienic conditions, there is no guarantee that the RTE-IMF products will be consistently free of some pathogens because there is no bactericidal step during the process. The pathogens of most concern are those in which the zero tolerance criterion (that is, absence in a 25-g amount) is applied because they are usually the most harmful pathogens, and their detection implies a previous enrichment. Among the “tolerance zero” organisms the most interesting are, undoubtedly, Escherichia coli O157:H7, Salmonella spp., and L. monocytogenes. The first is because the infectious dose is very low (Meng et al., 2001), from less than 100 cells (AGA, 1995) to about one dozen (Tilden et al., 1996). The interest in Salmonella spp. and L. monocytogenes lies in that they are ubiquitous organisms, although the latter is only of concern in the case of the U.S. Department of Agriculture regulations. The slices of RTE-IMF products are usually packed in domestic portions. It is not possible to use traditional technologies for sanitizing the packed RTE products. Nevertheless, we have previously demonstrated that electron-beam (E-beam) irradiation is a very effective way to reduce pathogens numbers present in a variety of RTE foods to a safe level (Cabeza et al., 2007, 2009; Hoz et al., 2008; Cambero et al., 2011). Therefore, the present work centers on the sanitation of RTE-IMF products of animal origin. This work was focused to carefully adjust the irradiation doses to achieve an adequate level of microbial safety with minimum changes in sensory attribute.
Materials and Methods
Organisms
Within the species S. enterica, Salmonella Enteritidis and Salmonella Typhimurium are the serovars most frequently isolated from foods (Hendriksen et al., 2011). We have chosen Salmonella Typhimurium (strain CECT 443) for experiments because we have observed (Cabeza et al., 2009) that this serovar is more radioresistant than Salmonella Enteritidis. L. monocytogenes Scott A (CIP 103575), E. coli O157:H7 (CECT 4972), and, as a surrogate for L. monocytogenes, L. innocua (NTC 11288) were also used. The inoculum preparations of the Salmonella and Listeria strains have been previously described (Cabeza et al., 2007, 2009). E. coli O157:H7 was prepared in a similar way. The slices of IMF were contaminated by immersion in a beaker for a few seconds. In experiments, a large number of cells (about 108 CFU/mL) were used to precisely calculate the radioresistant parameters.
Sample preparation and irradiation treatment
Pieces (2–5 kg) of the IMF were purchased in a local market. Then, they were sliced (3 or 15 mm thick) using an electric machine, whose rotating blade was previously deeply cleaned to release any organic particle. The slices were packed and prepared as described for fermented sausages (Cabeza et al., 2009). To assess the rheological and sensory properties many bags (10–12) per IMF product with four to six non-inoculated slices each were used. Samples were treated with an industrial E-beam radiation source, which operates at 10 MeV, located in Tarancón, Cuenca, Spain. The radiation doses used were between 1 and 3 kGy. The dose absorbed by samples was verified by determining the absorbance of cellulose triacetate dosimeters (ASTM, 2000), simultaneously irradiated. Experiments were performed at room temperature (18–20°C) by triplicate. The temperature increase during treatment was less than 2°C. After irradiation, samples were stored at 4°C until use. Three irradiation treatments were carried out, at three different times, following the procedure described above.
Microbial analysis
Although the bacterial load of the IMF is low, selective media were used for survival counts to avoid the growth of natural microbiota. Salmonella sp. and E. coli sp. were enumerated in red violet glucose agar. To count survivors of L. innocua and L. monocytogenes, the selective medium Palcam was chosen. To count the survivors, the slices (about 10 g) were homogenized with 10 mL of sterile saline solution in a Stomacher® (Seward, Worthing, UK) bag. Counts were determined on the surface of plates by using a spiral plate system (Eddy Jet; IUL Instrument, Barcelona, Spain). Plates were incubated at 37°C for 24 h. Colonies were enumerated with an automatic counter (Countermat Flash; IUL Instrument).
The shelf life of irradiated samples was only studied in the smoked tuna because it was the IMF product with the highest a w (0.88). It was determined by periodically counting the total bacterial number in samples stored at 4°C. Non-irradiated samples were used as controls.
Physicochemical analyses, color measurement, texture profile analysis, tensile test, and sensory analysis
The performance of these analyses has been described previously (Cabeza et al., 2007, 2009; Herrero et al., 2007). In brief, color measurements of samples were performed using a tristimulus colorimeter (Minolta Chroma Meter CR300; Minolta Corp., Ramsey, NJ). The texture profile analysis and tensile test were carried out with a TA.XT2i SMS Texture Analyzer (Stable Micro Systems Ltd., Godalming, UK) using a P/25 cylindrical probe for texture profile analysis or a tensile grip (A/TGT) for the tensile test. The pH was determined in a homogenate of the sample with distilled water (1:10) (wt/vol), using a Crison Digit-501 pH meter (Crison Instruments Ltd., Barcelona). The a w was measured using a Decagon CX1 hygrometer (Decagon Devices Inc., Pullman, WA) at 25°C. The sensory analyses involved a panel of 20 tasters selected from the members of the Department of Nutrition, Food Science, and Food Technology of the Complutense University of Madrid (Spain). The panel members were previously trained in the sensory assessment of meat and smoked fish products. Triangular, rank order, and descriptive tests were performed (Benedito et al., 2011). The tests were carried in individual booths built according to the International Standards Organization DP 66.58 (ISO, 1981) criteria. White fluorescent (for appearance) and red (for odor and taste) light was used.
Statistical analysis
The regression coefficients (R 2) of curves were calculated by Excel (Microsoft, Redmond, WA). Data obtained from physicochemical determinations were statistically analyzed by one-way ANOVA using a Statgraphics Plus version 5.0 program (Statpoint Technologies, Warrenton, VA).
Results and Discussion
Estimation of food safety features
The reasoning used to assess the potential contamination of an RTE meat product during slicing and packaging has been previously described (Cabeza et al., 2007). It was concluded that the contamination of cooked ham with L. monocytogenes during its transformation into an RTE product can reach, in the worst of cases, 10 cells. This may also be applied to RTE-IMF because the same operations are performed for its conversion into an RTE product. Because salmonellae are ubiquitous organisms, a comparable degree of contamination may be supposed. Although the contamination of RTE products by E. coli O157:H7 is lower (about 1 cell/g) than that of Listeria or salmonellae (Bohaychuk et al., 2006), the same level of contamination may be considered because the risk for people may be even higher. Thus, in all cases, the initial level of hazard (H0) is 1 (log10 10).
Some health authorities (for example, the U.S. Department of Agriculture) recommend a “zero tolerance” policy for L. monocytogenes in all RTE products, which means a food safety objective (FSO) of 4 CFU/100 g (log10=−1.39). However, given that a stricter tolerance of “not detected in 25 g” does not provide a higher level of protection (Ross et al., 2000), both the ICMSF (2001) and SCVPH (2005) take the view that the criterion could be 100 CFU/g (log10=2) at the time of consumption, when the product is not able to support the growth of L. monocytogenes (SCVPH, 2005). The microbiological criterion for Salmonella spp. and E. coli O157:H7 is an “absence in 25 g” practically all over the world.
Because none of the three microorganisms can grow in the IMF product during the shelf life, the performance objective (PeO) [maximum concentration of a hazard in a food at a specific phase in the food chain before the time of consumption that provides an FSO (Gorris, 2005)] is zero (that is, there will be no hazard increase regardless of whether the product is going to market with either the 100 CFU/g or “zero tolerance” microbiological criterion).
Nevertheless, the performance criterion (PeC) [the effect in concentration of a hazard in a food that must be achieved by the application of one or more control measures to provide an FSO (Gorris, 2005)] is not the same for both markets. In the case of 100 CFU/g acceptance, no sanitation treatment is necessary because the assumed contamination (10 cells/g) is lower than the FSO (100 CFU/g). In the case of “zero tolerance,” the PeC may be calculated using the expression FSO=H0 – PeC + PeO (ICMSF, 2001) and the mentioned data (FSO=−1.39; H0=1; PeO=0) with a final result of PeC=2.39 D reduction of the original load of the three organisms.
Physicochemical composition
The chemical composition of the products is shown in Table 1. In the context of this work, the most important feature is the a w, which was lower than 0.90 in all cases. As mentioned above, this parameter is low enough both to stabilize the final product from a technological perspective and to avoid the growth of pathogens. Accordingly, these products present long shelf life at room temperature without their sensory properties being affected. The only spoilage agents that could act are lipid oxidation, mold growth, and excessive dehydration. However, these RTE-IMF are usually vacuum packaged, and therefore the above phenomena are effectively prevented.
Data are mean±SD values.
Safety aspects
The behavior of bacteria against radiation fits first-order inactivation kinetics yielding straight lines with high regression coefficients (R 2) (Table 2). Table 2 also shows the survival equations and the D-values obtained. L. monocytogenes presented D-values (0.42–0.49) in the range of the data reported in pork (Cabeza et al., 2007; Hoz et al., 2008) and other meat products (Patterson, 1989; Thayer and Boyd, 1995; Zhu et al., 2009), for which D-values of 0.36 and 0.58 kGy were reported. L. innocua presented higher D-values (0.46–0.57 kGy) than L. monocytogenes, which confirms previous results (Cabeza et al., 2007; Hoz et al., 2008).
y=log colony-forming units/g; x=dose (kGy).
D10, irradiation decimal reduction value.
The D-values (0.52–0.63) obtained here for Salmonella Typhimurium in the three matrixes (Table 2) are in total agreement with the results of other authors (Thayer et al., 1990; Grant and Petterson, 1992; Cabeza et al., 2009).
The data about the radioresistance of E. coli O157:H7 consistently offer lower D-values than those of L. monocytogenes and Salmonella Typhimurium. For instance, in several foods the recorded D-values were lower than 0.30 kGy (Thayer and Boyd, 1993; Clavero et al., 1994; Buchanan et al., 1998; Chirinos et al., 2002). The D-values shown in Table 2 (0.25–0.29 kGy) illustrate once more the radiosensitivity of E. coli O157:H7.
To calculate the E-beam dose needed to achieve the FSO (the process criterion [PrC]) in each RTE-IMF product, the highest D-value must be chosen, which then has to be multiplied by the PeC (that is, 2.39 D reductions) to yield the PrC. E. coli O157:H7 is not used because it presented the lowest D-values. In the Iberian ham and smoked tuna, Salmonella Typhimurium was the highest radioresistant strain. Therefore, the PrC will be 1.27 kGy for the Iberian ham and 1.51 kGy for the smoked tuna. L. innocua presented the highest D-value (0.57 kGy) in the case of the cecina. Then, the PrC will be 1.36 kGy. As the D-values of both species of Listeria and Salmonella Typhimurium are very close, a PrC of 1.5 kGy for IMF products could be generalized. This treatment would cause reductions of about 3 D for both organisms and reductions of about 5 D for E. coli O157:H7. The reduction of 5 D is close to the one proposed by ICMSF (2001) for cooking of beef patties with respect to E. coli O157:H7, which was stated to be a reduction of 4.4 D. From ICMSF (2001) data, in frankfurters, a reduction of 3.39 D was calculated to achieve the criterion of “absence in 25 g” of L. monocytogenes cells by a thermal process. In the case of salmonellae, the lowest infectious dose recorded has been 28 cells (Vought and Tatini, 1998). This means that assuming an initial contamination of this bacterium of 10 cells/g, 3 D reductions would lead to a final number of 1/100 g (fourfold lower than the microbiological criterion), a realistic level of safety assuming a serving of 125 g.
Rheological, color, and sensorial aspects
Table 3 shows the E-beam effect on selected parameters of RTE-IMF. Results from the treatment with 3 kGy have been omitted because optimal results occurred with doses lower than 2 kGy (namely, 1.5 kGy). In smoked tuna, a decrease (p<0.05) of redness was clearly observed during storage (Table 3). Nevertheless, the behavior of this parameter has to be related with the fixation of the muscle pigment (myoglobin) instead of with the irradiation because no significant differences (p>0.05) were found between treated and control samples. However, a higher (p<0.05) yellowness (b* value) was observed at 3 kGy (that is, 7.37±0.70 at 0 kGy vs. 9.03±0.58 at 3 kGy). In contrast, in Iberian ham, a higher red color intensity was reflected by the increase of the a* parameter when 3 kGy was applied (that is, 14.77±1.70 at 0 kGy vs. 18.46±0.65 at 3 kGy). Ahn et al. (2004) reported that in cured meat there was a reduction of nitrosoheme pigments due to a denitrosylation of nitrosomyoglobin immediately after the ionizing treatments. In the case of the cecina, there were no significant changes in color parameters during the shelf life.
Electron-beam treatment was at 0 (control), 1, and 2 kGy. Unless otherwise is stated no significant differences were found between control (untreated) and the treated samples.
The biceps femoris muscles dissected were used for the determinations.
Value determined after 1 day of storage once the treatment was applied.
Value determined after 15 (smoked tuna) and 24 (Iberian dry-cured ham and cecina) days of storage at 4°C once the treatment was applied.
Significant differences (p<0.05) were found. The actual values for 0, 1, and 2 kGy were 33.61±5.60B, 44.72±6.17A,B, and 45.08±5.07A N, respectively (A,B: values with different capital letters indicate significant differences [p<0.05]).
TPA, texture profile analysis.
In the texture features, a minor significant difference (p<0.05) was found only in the smoked tuna (Table 3). This affected the hardness, in that samples treated at 1 and 2 kGy were somewhat harder. Yoon (2003) has also reported an increase in hardness of treated cooked chicken breast at 2.9 kGy, due to shrinkage of the myofibrils. However, other authors have reported that the textures of meat products are not affected by E-beam treatments at doses ≤3 kGy (Lee and Ahn, 2005; Cabeza et al., 2007; 2009; Hoz et al., 2008).
In the sensory analysis, no significant differences (p≤0.05) were detected in the triangular analysis, except for the odor in the smoked tuna, just after the treatment at 2 kGy; this disappeared after storage at 4°C for 15 days. These results are in accordance with those obtained in the descriptive test, where a decrease in the smoke curing odor and a release of a slight stale odor were detected. Adverse sensory effects in several ionizing-treated muscle products have been reported (Jo and Ahn, 2000; Brewer, 2009; Medina et al., 2009), caused by reactions that occur. Off-odors due to ionizing treatment of products have been associated with amino acid degradations (Jo and Ahn, 2000) and the generation of sulfur compounds (Yan et al., 2006; Brewer, 2009). In the present work, the above changes were not detected. Higher doses of radiation (that is, >2 kGy) are probably necessary to observe them.
Shelf life of treated products
The RTE-IMF products have a very long storage period. However, the shelf life of smoked tuna was analyzed because, among the three, this product is the one with the highest a w (0.88). Table 4 shows the results. After 78 days of storage, the bacterial count was lower than 104 CFU/g, even in the control batch. From a sensory point of view, no differences were found.
Data represent the total viable count (log colony-forming units [CFU]/g) in plate count agar.
NA, not analyzed.
Conclusions
The E-beam treatment of RTE-IMF with doses of 1.5 kGy allows us to reach the FSO for L. monocytogenes, Salmonella Typhimurium, and E. coli O157H7 with practically no modifications of quality attributes.
Footnotes
Acknowledgments
The present work has been supported by the MICINN with Project CARNISENUSA (CSD2007-00016), included in the CONSOLIDER-INGENIO 2010 issue and AGL2010-19158.
Disclosure Statement
No competing financial interests exist.
