- Email: [email protected]
Survival of Escherichia coli O157:H7 and non-pathogenic E. coli on irradiated and non-irradiated beef surfaces
Meat Science 83 (2009) 468–473
Contents lists available at ScienceDirect
Meat Science journal homepage: www.elsevier.com/locate/meatsci
Survival of Escherichia coli O157:H7 and non-pathogenic E. coli on irradiated and non-irradiated beef surfaces D.M. Prendergast a,*, K.M. Crowley a, D.A. McDowell b, J.J. Sheridan a a b
Food Safety Department, Ashtown Food Research Centre, Teagasc, Ashtown, Dublin 15, Ireland Food Microbiology Research Unit, NICHE, The University of Ulster, Jordanstown, Newtownabbey, County Antrim, Northern Ireland
a r t i c l e
i n f o
Article history: Received 17 January 2009 Received in revised form 22 May 2009 Accepted 15 June 2009
Keywords: Escherichia coli Beef Sites Irradiated Non-irradiated
a b s t r a c t This study examined changes in numbers of pathogenic (PEC) and non-pathogenic (NPEC) Escherichia coli during storage at 10 °C on the surfaces of irradiated (IR) and non-irradiated (NIR) meat pieces excised from the neck, brisket and rump of beef carcasses and in Brain Heart Infusion Broth (BHI) and Maximum Recovery Diluent (MRD). On irradiated meat pieces, there were significant differences between mean PEC and NPEC counts at all sites. Differences in counts were also observed between IR and NIR surfaces and among the three meat sites for both E. coli types. These differences occurred only on IR samples, suggesting that the irradiation associated reductions in normal beef surface flora influenced survival of both E. coli types. PEC and NPEC counts increased during storage in BHI, but only NPEC counts increased in MRD. The results of this study highlight the impact of meat surface type and the presence/absence of the normal beef carcass surface flora on E. coli survival and/or growth during meat storage. Such previously unreported effects, and their precise mechanisms, have direct implications in the development and application of accurate models for the prediction of the safety and shelf life of stored meat. Ó 2009 Published by Elsevier Ltd.
1. Introduction Effective prediction of the bacteriological safety of chilled meats requires accurate information on the numbers/types of pathogenic bacteria present, and their interactions with the contaminated meat surfaces during chill storage. However, valid concerns about the experimental introduction of pathogens into food environments mean that it is rarely possible to directly observe such interactions under commercial conditions. Thus many studies have used indicator organisms (Prendergast, Rowe, & Sheridan, 2007), which can provide useful data, but may give less information of real pathogen food interactions. A number of broth based studies (Eblen, Annous, & Sapers, 2005; Salter, Ross, & McMeekin, 1998) have suggested that nonpathogenic E. coli strains can be used as indicators in modelling the growth of pathogenic E. coli O157:H7, and in evaluating the efficacy of interventions designed to reduce the persistence/numbers of this pathogen during meat processing and storage. In more general terms, the use of data from broth experiments to predict E. coli growth and survival on beef surfaces may be unreliable, because pathogens in broths face different challenges from those posed by meat surfaces. Food structure may have a profound effect on pathogen growth and this has been demonstrated for Listeria monocytogenes, where significant differences in growth were * Corresponding author. Tel.: +353 1 8059500; fax: +353 1 8059550. E-mail address: [email protected] (D.M. Prendergast). 0309-1740/$ - see front matter Ó 2009 Published by Elsevier Ltd. doi:10.1016/j.meatsci.2009.06.024
demonstrated between a broth and a gel system and a broth and an agar surface (Wilson et al., 2002; Koutsoumanis, Kendall, & Sofos, 2004). The influence of surface structure in relation to meat spoilage has also been recognised (Nychas, Skandamis, Tassou, & Koutsoumanis, 2008). For example, a number of recent studies have demonstrated that survival rates of (non-pathogenic) E. coli are different at different sites on beef carcasses (Prendergast and Sheridan, unpublished; Kinsella, 2008). As well as the above physical and environmental factors, pathogen survival and growth has been shown to be significantly influenced by the presence and nature of the natural flora of meat. Thus, Nissen, Maugesten, and Lea (2001) demonstrated more growth of E. coli O157:H7 on heat treated, lactic acid decontaminated beef pieces than on untreated control samples during aerobic storage at 10 °C, which is in agreement with Vold, Holck, Wasteson, and Nissen (2000), who noted that the presence of a beef microflora inhibited the growth of E. coli O157:H7 in beef mince aerobically stored at 12 °C. However, such interactions between pathogens and the wider beef microflora, may be influenced by other physical and/or environmental factors, as Berry and Koohmaraie (2001) reported no such effects on nonminced beef surfaces. A number of studies have examined survival and/or growth of E. coli O157:H7 and other pathogens on excised/cut surfaces (Kinsella et al., 2008; Logue, Sheridan, & Harrington, 2004; Nissen et al., 2001) but much less is known about the survival and/or growth of this pathogen on uncut surfaces, i.e. surfaces with intact
469
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
membranes. Studies of E. coli inoculated onto beef carcass surfaces have demonstrated that growth and survival rates are affected by the nature of the inoculated site (Kinsella, 2008; Prendergast and Sheridan, unpublished). Thus, during holding at a range of chill temperatures, some counts on the cut surface of the neck increased (while the rest remained unchanged) but on the brisket, some counts were reduced (while the rest remained unchanged). On the rump, all counts were reduced, frequently to almost zero. Overall, the results of such studies highlight the need for a more comprehensive understanding of the physical and biological factors modulating bacterial survival and growth on beef carcasses and derived products, in the development of accurate models of the growth of E. coli O157 (and other pathogens) in these complex food environments. The present study examined the growth/survival of E. coli O157:H7 and a potential index organism (generic E. coli) on excised irradiated and non-irradiated beef neck, brisket and rump surfaces, and compared the growth of these organisms on these surfaces, with their growth in broth systems.
2. Materials and methods 2.1. Slaughter and inoculation sites The study used an experimental abattoir at Ashtown Food Research Centre (AFRC), designed for low-throughput slaughter and processing of cattle, sheep and pigs. Six grass fed Hereford Fresian heifers (624 months) were processed under normal commercial slaughter conditions, i.e. lairage, stunning, exsanguination, dehiding, and evisceration, followed by carcass splitting, weighing and washing, to produce carcasses with a mean weight of 122.5 kg (range of 110.0–131.5 kg). Three sites on each carcass were identified as follows: (i) neck, (ii) brisket and (iii) rump. A fibrous membrane of connective tissue or fascia covers the latter two sites, while the selected neck site was the lean surface exposed during slaughter (i.e. not covered by fascia). Two animals were slaughtered per day, with each day representing a complete treatment. The treatment was repeated on three separate days. 2.2. Production of beef samples Immediately after carcass washing, surface samples (approximately 300 cm2, and 5 mm deep) were aseptically excised from each selected site, placed in sterile stomacher bags and transported to the on-site laboratory. From these surface samples cylindrical meat pieces (approximate diameter 2.5 cm2) were aseptically excised using sterile coring punches. On each occasion, 26 2.5 cm2 samples were excised from each of the three carcass sites. Samples were placed in separate 12.5 7.7 cm Cryovac BB4 bags (Sealed Air Cryovac, Cambridge, UK), vacuum packed using a Vac Star, model S220 (Sugiez, Switzerland) and chilled to 4 °C within 1 h. 2.3. Sample irradiation Vacuum packed samples (486) were transported under refrigeration (<10 °C) to the Agrifood Biosciences Institute, Northern Ireland, and irradiated at 4 °C (±1 °C) to a target dose of 5.0 kGy, using a cobalt 60 source (Gammabeam 650, Nordion International Inc., Kanata, Canada) at a dose rate of 5.33 kGy h 1. Irradiation dose was monitored and confirmed using four red perspex dosimeters (Type 4034AN, Harwell Dosimeters Ltd., Harwell, UK) placed throughout the packed meat pieces. Used dosimeters were spectrophotometrically examined (640 nm), and doses calculated using calibration graphs provided by the National Physical Laboratory
(Teddington, UK). The sample preparation process, i.e. excision, vacuum packaging, dispatch, irradiation, and return was carried out under refrigeration conditions within 72 h. Returned, irradiated samples were stored at 0 °C for no more than 24 h. Immediately prior to inoculation, six irradiated samples were randomly selected from each sample type, placed in 30 ml BHI at 25 °C for 18 h, plated onto Oxoid Plate Count Agar (PCA) (Oxoid, Basingstoke, UK), incubated at 30 °C for 3 days, and examined to confirm sterility. 2.4. Inoculum preparation Six bovine isolates of E. coli O157:H7, each of which carried the hlyA and eaeA genes were obtained from the AFRC’s culture collection. Four of the isolates carried the vt1 and vt2 genes, while one each carried either the vt1 or vt2 gene only. All six isolates were made resistant to 50 lg ml 1 nalidixic acid (Sigma–Aldrich, St. Louis, US) and 1000 lg ml 1 streptomycin sulphate (Sigma–Aldrich) by the method of Park (1978). A single non-pathogenic bovine hide isolate of E. coli, was obtained from the AFRC culture collection. In a separate experiment, a six isolate cocktail of non-pathogenic E. coli was prepared to determine if there was a difference in the growth kinetics between the single isolate and a cocktail of strains. The pathogenic and non-pathogenic isolates were individually cultured in Oxoid Brain Heart Infusion Broth (BHI) at 37 °C for 16–18 h. Stationary phase cells were recovered using an Eppendorf model 5403 refrigerated centrifuge at 3000g for 10 min at 4 °C (Eppendorf, Hamburg, Germany), washed three times in Oxoid maximum recovery diluent (MRD) by centrifugation and resuspended/diluted in MRD to give a final inoculum of approximately 5 log10 CFU ml 1. 2.5. Meat surface inoculation Thirteen irradiated (IR) and thirteen non-irradiated control (NIRC) samples from each carcass site were inoculated with a 5 ll inoculum containing approximately 3 log10 CFU of the cocktail of pathogenic E. coli (PEC) or of the non-pathogenic E. coli (NPEC) and held at room temperature for 30 min (to facilitate cell adherence). Inoculated samples (13) and uninoculated controls (13) were transferred to Aqualab cups (Labcell, Basingstoke, UK), covered with caps and placed in refrigerated incubators at 10 °C. Inoculated and uninoculated samples from each site (neck, rump and brisket) were removed after 0, 6, 12, 18, 24, 30, 36, 42, 48, 54, 60, 66 and 72 h, and examined as described below. 2.6. Temperature monitoring Temperatures of the refrigerated incubators were monitored at three positions, top, middle and bottom every 30 min for 72 h, using Type T copper–constantan thermocouples and the data was recorded using a model 2040 Grant Squirrel data logger (Grant Instruments, Cambridge, UK). At the end of each experiment, data were downloaded to MicrosoftÒ Office Excel 2003. 2.7. Enumeration of pathogenic (PEC) and non-pathogenic E. coli (NPEC) Each recovered sample was pulsified for 30 s in 10 ml MRD in a sterile 100 150 mm stomacher bag (Seward, Norfold, UK) using a model PUL Pulsifier (Filtraflex, Ontario, Canada). Undamaged E. coli O157:H7 were enumerated by direct plating of 1 ml aliquots of the above homogenates onto Oxoid Sorbitol MacConkey Agar (SMAC), each supplemented with 50 lg ml 1 nalidixic acid and 1000 lg ml 1 streptomycin sulphate. TM
470
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
Undamaged non-pathogenic E. coli were enumerated by direct plating of 1 ml aliquots of the above homogenates onto MacConkey Agar, supplemented with 50 lg ml 1 nalidixic acid and 1000 lg ml 1 streptomycin sulphate. Sublethally damaged E. coli O157:H7 and non-pathogenic cell numbers were enumerated by a resuscitation procedure, i.e. samples were plated onto Oxoid Tryptone Soya agar (TSA), incubated for 2 h at 25 °C, over poured with SMAC or MacConkey (supplemented with 50 lg ml 1 nalidixic acid and 1000 lg ml 1 streptomycin sulphate), and further incubated at 37 °C for 24 h. 2.8. Total viable counts on meat samples At all sampling times, total viable counts (TVC’s) were enumerated from the inoculated and uninoculated IR and NIR samples. Samples were transferred into sterile stomacher bags containing 10 ml MRD, pulsified as described above, plated onto Oxoid plate count agar (PCA) and incubated at 30 °C for 3 days. 2.9. Growth of PEC and NPEC in MRD and BHI Ten ml volumes (12) of MRD in 30 ml polypropylene tubes with caps (Sarstedt, Numbrecht, Germany) were each inoculated with 5 ll of approximately 3 log10 cfu of PEC or NPEC, and held at 10 °C. Tubes were removed after 0, 6, 12, 18, 24, 30, 36, 42, 48, 54, 60, 66 and 72 h and PEC/NPEC numbers were estimated by direct plating onto SMAC and MacConkey agar supplemented with nalidixic acid and streptomycin sulphate as described above. Ten ml volumes (12) of BHI in 30 ml polypropylene tubes with caps (Sarstedt, Numbrecht, Germany) were each inoculated with 5 ll of approximately 3 log10 cfu of PEC or NPEC, and held at 10 °C. Tubes were removed after 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 11 h and PEC/NPEC numbers were estimated by direct plating onto SMAC and MacConkey agar supplemented with nalidixic acid and streptomycin sulphate as described above.
In almost all cases, there were no significant differences between direct and overlay counts of PEC or NPEC at each sample occasion on IR and NIR surfaces. Thus only overlay counts are presented (Table 1). The study compared mean (overlay) counts (1) between organisms, i.e. PEC and NPEC, (2) between IR and NIR surfaces, and (3) among meat sites (neck, brisket and rump.) (1) Non-pathogenic (NPEC) counts were significantly higher than PEC counts at all three sites on IR samples (P < 0.05), while on NIR samples, there were no significant differences between the counts for these organisms. (2) Counts on IR brisket surfaces were significantly higher than on NIR brisket surfaces, for both PEC and NPEC counts (P < 0.001). On the neck and on the rump, IR counts were not significantly different from NIR counts, except in the case of PEC IR counts on the rump, which were significantly lower than PEC NIR counts (P < 0.05). (3) Among meat sites, counts on the neck were significantly higher than on the brisket or rump for both PEC and NPEC counts, on almost all (IR and NIR) sites (P < 0.001). IR brisket counts were significantly higher than IR rump counts for both PEC and NPEC (P < 0.001), but NIR brisket counts were not significantly different from NIR rump counts. The study compared overlay counts during storage for up to 72 h. There were no significant differences among the time zero counts, between organisms (PEC and NPEC), between IR and NIR surfaces, and among meat sites (neck, brisket rump).
Table 1 Survival of pathogenic (PEC) and non-pathogenic E. coli (NPEC) (log10 CFU cm sterile (IR) and non-sterile (NIR) beef samples for 72 h at 10 °C.
) on
Beef site
Neck
Status
IR
NIR
IR
NIR
IR
NIR
Time after inoculation (h) 0 6 12 18 24 30 36 42 48 54 60 66 72 Mean
PEC 2.81 2.71 2.67 2.64 2.70 2.91 3.19 3.47 3.18 3.19 3.35 3.51 3.46 3.06
2.91 2.81 2.94 3.06 3.12 3.21 3.38 3.55 3.33 3.73 3.80 3.86 4.13 3.37
2.64 2.53 2.43 2.33 2.32 2.33 2.38 2.43 2.71 2.43 2.52 2.61 2.20 2.45
2.34 2.25 2.01 1.77 1.92 1.92 1.38 0.84 1.94 0.73 1.03 1.31 1.70 1.62
2.54 1.69 1.62 1.55 1.60 1.40 1.31 1.17 1.34 1.36 1.05 1.30 1.11 1.46
2.31 2.19 2.17 2.15 1.65 1.56 1.76 1.96 2.22 1.94 1.83 1.71 2.51 2.00
3. Results
0 6 12 18 24 30 36 42 48 54 60 66 72 Mean
NPEC 2.49 2.58 2.88 3.17 3.06 3.08 3.44 3.80 4.45 4.15 4.00 3.84 4.71 3.51
2.52 3.03 3.05 3.07 2.97 3.17 3.30 3.43 4.05 4.34 4.38 4.43 4.31 3.54
2.69 2.19 2.45 2.70 2.82 2.81 3.22 3.63 3.65 3.71 4.05 4.38 4.50 3.29
2.06 1.87 1.82 1.76 1.93 1.82 1.90 1.99 1.96 1.77 1.74 1.72 1.89 1.86
2.43 2.11 2.33 2.54 0.86 1.72 2.58 3.45 2.29 2.02 1.45 0.91 1.67 2.03
2.17 2.01 2.05 2.09 1.48 1.41 1.89 2.41 1.88 2.64 2.90 3.16 1.91 2.15
The mean temperature of the incubators during storage of beef and broth samples were 10.2 ± 0.5 °C.
Standard error of differences between means: columns: 0.47; rows: 0.43 Degrees of freedom: columns: 251; rows: 288 Standard error of differences for means: 0.21 Degrees of freedom for means: 22.
2.10. Statistical analysis All statistical tests examining bacterial counts were performed using log transformations. For all data, 0.998 was added to the original values prior to log transformation giving zero counts a value of 0.0009 on the log scale, while making no appreciable difference to the log values of all other (non-zero) counts (Prendergast, Sheridan, Daly, McDowell, & Blair, 2004). This allowed manipulation of all the data (including zero counts), with minimal significant effect on the overall results. The experiment was designed as a split-split-plot with temperature and three carcass sites on the main plots, counting method on the subplots and time on the sub-sub-plots. Analysis of variance (ANOVA) was carried out at the P < 0.05 level of significance using Genstat 5 (Rothamsted Experimental Station, Harpenden, United Kingdom). The growth kinetics of E. coli O157:H7 and non-pathogenic E. coli in BHI and MRD broths were determined using the Gompertz function, where parameters A, B, C and M were used to calculate the exponential growth rate (ERG; log10 CFU ml 1 d 1), generation time (GT; day) and lag phase duration (LPD; day). Linear regression analysis was used to compare the exponential growth rates (slopes) between experiments in MRD, using either a single nonpathogenic isolate or a six isolate cocktail.
Brisket
2
Rump
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
PEC counts on IR surfaces did not change over time on the neck or brisket, but were significantly reduced on the rump (P < 0.001). NPEC counts on IR surfaces showed significant increases on the neck and brisket (P < 0.001), but were unchanged on the rump, during storage. PEC counts on NIR surfaces increased on the neck (P < 0.05), with no change on either the brisket or rump. Similar results were observed in relation to NPEC counts (P < 0.001). Significant increases, i.e. growth was observed on the neck (1PEC/2NPEC) and the brisket (1NPEC), but not on the rump. A comparison between irradiated and non-irradiated uninoculated controls revealed that irradiation had not completely sterilised all the samples. A small residual microflora remained after irradiation, which was significantly smaller than on the non-irradiated flora at all sites (P < 0.01). Mean TVC’s on inoculated IR meat surfaces were overwhelming composed of the inoculated E. coli cells. On NIR surfaces, counts on the neck were significantly higher on inoculated surfaces (P < 0.001), than on uninoculated surfaces (Table 2). Differences between these counts were related to the ability of PEC and NPEC to grow on the neck surface, in the presence of the normal meat surface microflora as indicated above (Table 1). Differences between the growth/survival of inoculated PEC and NPEC were also evident from the TVC counts on IR samples. On the three sites inoculated with PEC, the TVCs were always significantly lower (P < 0.001), than the TVC counts from NPEC inoculated surfaces. This effect was also noted on NIR samples, but only for the neck and brisket (P < 0.05). The growth kinetics of PEC and NPEC in BHI broth, were similar (Table 3). In MRD, their growth kinetics was different, i.e. NPEC grew within 3 days, but PEC failed to grow within this period. NPEC growth was much slower in MRD than PEC or NPEC in BHI. The shorter lag time for NPEC in MRD suggested that NPEC adapted to the minimal growth conditions of MRD more effectively than it did in BHI. As only a single isolate of NPEC was used in these experiments, as opposed to a cocktail of six isolates for PEC, a second growth experiment was undertaken in MRD using a cocktail of six
Table 2 Mean total viable counts (log10 CFU cm 2) on irradiated and non-rradiated beef sites, uninoculated and inoculated with pathogenic (PEC) or non-pathogenic (NPEC) E. coli. Sites
Inoculum
Irradiated Uninoculated
Neck Brisket Rump
PEC NPEC PEC NPEC PEC NPEC
Non-irradiated Inoculated
1.93 1.74 0.84 1.30 0.82 1.15
3.20 4.62 1.53 3.03 1.52 2.53
Uninoculated
Inoculated
3.40 2.74 2.35 2.93 1.57 2.26
4.17 4.79 2.42 3.04 2.14 2.61
Standard error of differences between means: 0.25 Degrees of freedom: 48.
Table 3 The influence of media on the growth kinetics of pathogenic E. coli (PEC) and nonpathogenic E. coli (NPEC) at 10 °C. Medium
BHI MRD
Organism
PEC NPEC PEC * NPEC 1 NPEC 2
Growth parameters LPD (h)
EGR (log10 CFU h
39.24 32.11 NG 18.04 19.68
0.112 0.098 NG 0.043 0.051
1
)
GT (h) 2.70 3.07 NG 7.00 6.00
LPD: lag phase duration; EGR: exponential growth rate; GT: generation time; BHI: brain heart infusion broth; MRD: maximum recovery diluent; NG: no growth; *NPEC 1: experiment using a single isolate; and NPEC 2: experiment using a six isolate cocktail.
471
non-pathogenic isolates. The data in Table 3 show that the growth kinetics were similar using a single isolate or a cocktail of nonpathogenic isolates. It was also demonstrated that the exponential growth rates (slopes) from the single isolate and the cocktail were not significantly different.
4. Discussion In this study the growth/survival at 10 °C of pathogenic and non-pathogenic E. coli isolates were compared over time on irradiated and non-irradiated beef surfaces and in broth. The study noted differences between E. coli numbers on irradiated and nonirradiated samples, with greater growth of both inocula on some irradiated samples. The study observed that non-pathogenic E. coli counts were higher than E. coli O157 counts on irradiated samples of all three meat surfaces. However, this was not the case on non-irradiated surfaces. Overall, growth was most frequently observed on the neck, with more growth of non-pathogenic E. coli, than of E. coli O157. This greater growth of non-pathogenic E. coli was also observed as part of the TVC’s from inoculated samples. The ability of non-pathogenic E. coli to outgrow E. coli O157 was similarly observed under the minimal nutrition conditions provided by MRD broth. The observation that non-pathogenic counts were higher than pathogenic counts on irradiated surfaces, but not on non-irradiated surfaces, suggests that irradiation may be affecting the chemical or physical composition of the meat surface. Alternatively it may be inactivating the natural microflora, favouring the growth of the non-pathogenic isolate. The possibility exists that irradiation renders the meat surface less favourable to the growth of the pathogen, by the production of radiolytic compounds. In the present study, samples were irradiated with a dose of 5.0 kGy, which gave a reduction of the order of 2–3 log. This result is in line with a previous study, which reported that a dose of 5.0 kGy led to a 2 log reduction in the microflora of fresh beef pieces (Aziz, Mahrous, & Youssef, 2002). In addition to the microbiological changes, such rates of irradiation have been reported to induce a range of chemical changes in irradiated stored beef. These changes arise from the oxidation of proteins and lipids as a result of the action of radiolytically induced free radicals (Grolichová, Dvorák, & Musilová, 2004; Rowe, Maddock, Londergan, & Lonergan, 2004). The impacts of such free radicals in relation to bacterial survival on meat surfaces are not known, and further investigations of such interactions may be necessary if meat irradiation is to be more widely used. For example, the contribution of inactivated bacterial cells, in the growth of any cells, which survive irradiation, may require further investigation. Alternatively, the inactivation of the meat microflora may render the meat surface less capable of supporting microbial growth. The ability of the surface microflora to provide a variety of metabolites or nutrients during growth is well established (Gill & Newton, 1977; Tsigarida & Nychas, 2001). The metabolites provided by one organism may produce a more favourable growth environment for another by a process known as metabiosis (Gram et al., 2002). Such an absolute requirement for a sufficiently favourable environment for the growth of PEC is confirmed by the observation that the pathogenic isolates could not grow in the minimal conditions provided by MRD, while the nonpathogenic isolate could grow in this medium. This study observed greater growth of both inocula on irradiated than non-irradiated brisket samples suggesting that the presence/absence of the normal beef microflora affects the growth of E. coli at this site. The influence of the beef microflora on pathogen growth and survival is not clear. Berry and Koohmaraie (2001)
472
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
substantially reduced the normal meat microflora on lean beef surfaces using a germicidal UV treatment. These surfaces were sequentially inoculated with different dilutions of a beef microflora and E. coli O157:H7. These authors reported that pathogen growth at 12 °C in the presence of a beef microflora of log 5.0–6.0 was similar to that on uninoculated samples. A study of the growth of E. coli O157:H7 in beef mince, in the presence or absence of four different aerobic background floras (log 5.0–6.0) showed that after storage at 12 °C, pathogen growth was inhibited in the presence of three of the background floras, but not in the fourth (Vold et al., 2000). When no background flora was added to the beef mince the pathogen grew on two of the four samples. The authors suggested that the presence of a background flora, whether naturally present or added to mince, resulted in inhibition of pathogen growth. Alternatively, these differences may relate to changes in beef mince composition. Such effects have been recorded in samples stored at 5.0–7.0 °C. Jay, Vilai, and Hughes (2003) noted that of four samples tested, one yielded 21 different bacterial genera, while two other samples yielded only seven. In the present study, one or more of the above processes may have been in operation, leading to the inhibition of pathogen growth. However, the authors consider it important to note that the current study found that the normal beef microflora, at naturally occurring levels on the brisket surface, induced significant reductions in E. coli, and particularly in E. coli O157, on this site. While this effect was not universally observed at the other sites, this is the first report of such an effect by the naturally occurring microflora, which is perhaps more significant than previous reports which noted inhibition at artificially increased levels of beef microflora. As well as the microflora, the different meat substrates may have significant roles in preventing or allowing the growth/survival of the E. coli types. In terms of the pathogen on irradiated surfaces, counts on neck samples were significantly different from those on the brisket and rump, and brisket counts were significantly different from those on the rump. The same pattern was observed among non-irradiated pathogen counts, except that brisket counts were not different from those on the rump. Non-pathogenic counts were the same as the pathogen counts, on irradiated and non-irradiated surfaces, with one exception where there was no difference between irradiated neck and brisket counts. Differences in survival of E. coli related to beef sites has been recorded previously on inoculated carcasses after storage at 10 °C for 72 h (unpublished work by Prendergast and Sheridan). In that study, there were significant differences in the survival of E. coli, with growth on the neck, no change in counts on the brisket and a decline in counts to almost zero on the rump. In contrast, in the present study there was growth on the neck, but no change in counts on the brisket or rump over time, for pathogenic and non-pathogenic E. coli. The data from these experiments demonstrated that while the neck site behaved in a similar manner with growth of E. coli, the other sites behaved differently. The difference between these two experiments may be related to the influence of changing carcass temperature and aw at the different sites, which occurs during chilling but is not present on meat pieces incubated at a specific temperature in a closed container where the relative humidity is uniformly high at all times. These parameters have been shown to have a major impact on the survival of E. coli inoculated on beef surfaces at different sites, both under commercial and experimental conditions. During storage, the meat microflora grew on non-irradiated samples on the three meat types, particularly on the neck, as did the two inoculated E. coli types. While the growth of non-pathogenic E. coli on the irradiated brisket could be related to the removal of the surface microflora, this was not the case for the neck samples. At this site, growth of both E. coli types occurred
on irradiated and non-irradiated samples, indicating that the neck could support the growth of the natural microflora and the inoculated organisms, as shown by the higher counts for the TVC’s on inoculated samples. The ability of the neck site to support growth may be related to the increased availability of a specific nutrient, such as glucose, associated with the cut surface of the neck, compared to the brisket and rump where membranes cover the meat surface, possibly making nutrients less readily available. According to Gill and Newton (1978) however, the growth of bacteria on cut and intact surfaces will be similar as surface membranes will not be a barrier to diffusion of solutes to the meat surface. In addition to these factors, differences in meat structure per se must also be considered in relation to E. coli growth. The influence of surface structure on the survival of pathogens on meat surfaces has been recognised in relation to beef carcasses (Prendergast et al., 2007). During the development of rigor, structural changes occur on meat surfaces in the form of gaps due to shrinkage in muscle fibers and other structures, which aid bacterial survival due to cell penetration into these gaps (Frank, 2001; Thomas, O’Rourke, & McMeekin, 1987). In addition, rupture of capillary blood vessels occurs in the connective tissue, fat and muscle (Belk, Scanga, Smith, & Grandin, 2002). The blood contains low molecular weight compounds, such as glucose, glycerol, glycine betaine and carnitine (Smith, 1996) which are available to bacterial cells on surfaces and can aid in survival (Gill & Newton, 1980). Overall, the results of this study showed that NPEC grew better that PEC on all meat surfaces, suggesting that the non-pathogenic E. coli was better fitted to growth on the naturally contaminated beef surfaces than PEC. Such fitness may relate to differences in abilities to compete with the natural microflora as demonstrated in this study, or to the greater ability of the non-pathogen to grow under partially nutrient limited conditions. The second of these possibilities is confirmed by the broth results in the current study, where the NPEC was able to grow in the much more nutrient restricted environment posed by MRD. In conclusion this study demonstrated that there were differences in growth and survival between pathogenic and non-pathogenic E. coli isolates on meat surfaces and in broth. It was also apparent that the natural microflora of the meat surface had a major influence on the survival of pathogenic and non-pathogenic E. coli, as did the nature of the different meat surfaces. This study has demonstrated that at the levels of contamination frequently observed under commercial conditions, the normal meat microflora has a significant effect on the growth and survival of E. coli on some meat surfaces. Overall the results of this study demonstrate the challenges in using model organisms, and model liquid and/or surfaces, and environmental conditions in the development of accurate models for the prediction of the survival/growth of pathogens on meat surfaces under commercial conditions. Acknowledgements This research has been carried out with the financial support of the Commission of the European Communities, specific RTD program ‘‘Quality of Life and Management of Living Resources” Project No. QlK1-CT2002-02545. The authors wish to thank Dr. Eileen Stewart, Queens University Belfast for irradiation of the beef samples. Sincere thanks are also due to Dr. Dermot Harrington and to Ms. Paula Reid, Teagasc for their assistance with the statistical analysis. References Aziz, N. H., Mahrous, S. R., & Youssef, B. M. (2002). Effect of gamma-ray and microwaves treatment on the shelf-life of beef products stored at 5 C. Food Control, 13, 437–444.
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473 Belk, K. E., Scanga, J. A., Smith, G. C., & Grandin, T. (2002). The relationship between good handling/stunning and meat quality in beef, pork and lamb. In Proceedings of the American Meat Institute Foundation, animal handling and stunning conference, February 21–22.. Berry, E. D., & Koohmaraie, M. (2001). Effect of different levels of beef bacterial microflora on the growth and survival of Escherichia coli O157:H7 on beef carcass tissue. Journal of Food Protection, 64, 1138–1144. Eblen, D. R., Annous, B. A., & Sapers, G. A. (2005). Studies to select appropriate nonpathogenic surrogate Escherichia coli strains for potential use in place of Escherichia coli O157:H7 and Salmonella in pilot Plant Studies. Journal of food protection, 68, 282–291. Frank, J. F. (2001). Microbial attachment to food and food contact surfaces. Advances in Food and Nutrition Research, 43, 320–370. Gill, C. O., & Newton, K. G. (1977). The development of aerobic spoilage flora on meat stored at chill temperatures. Journal of Applied Bacteriology, 43, 189–195. Gill, C. O., & Newton, K. G. (1978). The ecology of bacterial spoilage of fresh meat at chill temperatures. Meat Science, 2, 207–217. Gill, C. O., & Newton, K. G. (1980). Development of bacterial spoilage at adipose tissue surfaces of fresh meat. Applied and Environmental Microbiology, 39, 1076–1077. Gram, L., Ravn, L., Rasch, M., Bruhn, J. B., Christensen, A. B., & Givskov, M. (2002). Food spoilage—interactions between food spoilage bacteria. International Journal of Food Microbiology, 78, 79–97. Grolichová, M., Dvorák, P., & Musilová, H. (2004). Employing ionizing radiation to enhance food safety – A review. ACTA Veterinaria Brunensis, 73, 143–149. Jay, J. M., Vilai, J. P., & Hughes, M. E. (2003). Profile and activity of the bacterial biota of ground beef held from freshness to spoilage at 5–7 °C. International Journal of Food Microbiology, 81, 105–111. Kinsella, K. J., Prendergast, D. M., McCann, M. S., Blair, I. S., McDowell, D. A., & Sheridan, J. J. (2008). The survival of Salmonella enterica serovar Typhimurium DT104 and Total Viable Counts on beef surfaces at different relative humidity’s and temperatures. Journal of Applied Microbiology, 102, 171–180. Kinsella, K. J. (2008). Defining the critical limits for the use of chilling as a CCP in a beef HACCP plan. PhD thesis. Chapter 6: The influence of commercial chilling on the survival of E. coli K12 surface inoculated at different sites on beef carcasses and the effect on the Total Viable Counts. University of Ulster, Jordanstown campus, Newtownabbey, Co. Antrim, pp. 110. Koutsoumanis, K. P., Kendall, P. A., & Sofos, J. N. (2004). A comparative study on growth limits of Listeria monocytogenes as affected by temperature, pH and aw when grown in suspension or on a solid surface. Food Microbiology, 21, 415–422.
473
Logue, C. M., Sheridan, J. J., & Harrington, D. (2004). Studies of steam decontamination of beef inoculated with Escherichia coli O157:H7 and its effects on subsequent storage. Journal of Applied Microbiology, 98, 741–751. Nissen, H., Maugesten, T., & Lea, P. (2001). Survival and growth of Escherichia coli O157:H7, Yersinia enterocolitica and Salmonella enteritidis on decontaminated and untreated meat. Meat Science, 57, 291–298. Nychas, G-J. E., Skandamis, P. N., Tassou, C. C., & Koutsoumanis, K. P. (2008). Meat spoilage during distribution. Meat Science, 78, 77–89. Park, R. W. A. (1978). The isolation and use of streptomycin resistant mutants for following development of bacteria in mixed populations. In D. Lovelock & R. Davies (Eds.), Techniques for the study of mixed populations (pp. 107–112). London: Academic Press. Prendergast, D. M., Sheridan, J. J., Daly, D. J., McDowell, D. A., & Blair, I. S. (2004). The use of a marked strain of Pseudomonas fluorescens to model the spread of brain tissue to the musculature of cattle after shooting with a captive bolt gun. Journal of Applied Microbiology, 96, 437–446. Prendergast, D. M., Rowe, T. A., & Sheridan, J. J. (2007). Survival of Listeria innocua on hot and cold beef carcass surfaces. Journal of Applied Microbiology, 103, 2721–2729. Rowe, L. J., Maddock, K. R., Londergan, S. M., & Lonergan, E. H. (2004). Influence of early postmortem protein oxidation on beef quality. Journal of Animal Science, 82, 785–793. Salter, M. A., Ross, T., & McMeekin, T. A. (1998). Applicability of a model for nonpathogenic Escherichia coli for predicting the growth of pathogenic Escherichia coli. Journal of Applied Microbiology, 85, 357–364. Smith, L. T. (1996). Role of osmolytes in adaptation of osmotically stressed and chillstressed Listeria monocytogenes grown in liquid media and on processed meat surfaces. Applied and Environmental Microbiology, 62, 3088–3093. Thomas, C. J., O’Rourke, R. D., & McMeekin, T. A. (1987). Bacterial penetration of chicken breast muscle. Food Microbiology, 4, 87–95. Tsigarida, E., & Nychas, G.-J. E. (2001). Ecophysiological attributes of a Lactobacillus sp. and a Pseudomonas sp. on sterile beef fillets in relation to storage temperature and film permeability. Journal of Applied Microbiology, 90, 696–705. Vold, L., Holck, A., Wasteson, Y., & Nissen, H. (2000). High levels of background flora inhibits growth of Escherichia coli O157:H7 in ground beef. International Journal of Food Microbiology, 56, 219–225. Wilson, P. D. G., Brocklehurst, T. F., Arino, S., Thuault, D., Jakobsen, M., Lange, M., et al. (2002). Modelling microbial growth in structured foods: towards a unified approach. International Journal of Food Microbiology, 73, 275–289.
Contents lists available at ScienceDirect
Meat Science journal homepage: www.elsevier.com/locate/meatsci
Survival of Escherichia coli O157:H7 and non-pathogenic E. coli on irradiated and non-irradiated beef surfaces D.M. Prendergast a,*, K.M. Crowley a, D.A. McDowell b, J.J. Sheridan a a b
Food Safety Department, Ashtown Food Research Centre, Teagasc, Ashtown, Dublin 15, Ireland Food Microbiology Research Unit, NICHE, The University of Ulster, Jordanstown, Newtownabbey, County Antrim, Northern Ireland
a r t i c l e
i n f o
Article history: Received 17 January 2009 Received in revised form 22 May 2009 Accepted 15 June 2009
Keywords: Escherichia coli Beef Sites Irradiated Non-irradiated
a b s t r a c t This study examined changes in numbers of pathogenic (PEC) and non-pathogenic (NPEC) Escherichia coli during storage at 10 °C on the surfaces of irradiated (IR) and non-irradiated (NIR) meat pieces excised from the neck, brisket and rump of beef carcasses and in Brain Heart Infusion Broth (BHI) and Maximum Recovery Diluent (MRD). On irradiated meat pieces, there were significant differences between mean PEC and NPEC counts at all sites. Differences in counts were also observed between IR and NIR surfaces and among the three meat sites for both E. coli types. These differences occurred only on IR samples, suggesting that the irradiation associated reductions in normal beef surface flora influenced survival of both E. coli types. PEC and NPEC counts increased during storage in BHI, but only NPEC counts increased in MRD. The results of this study highlight the impact of meat surface type and the presence/absence of the normal beef carcass surface flora on E. coli survival and/or growth during meat storage. Such previously unreported effects, and their precise mechanisms, have direct implications in the development and application of accurate models for the prediction of the safety and shelf life of stored meat. Ó 2009 Published by Elsevier Ltd.
1. Introduction Effective prediction of the bacteriological safety of chilled meats requires accurate information on the numbers/types of pathogenic bacteria present, and their interactions with the contaminated meat surfaces during chill storage. However, valid concerns about the experimental introduction of pathogens into food environments mean that it is rarely possible to directly observe such interactions under commercial conditions. Thus many studies have used indicator organisms (Prendergast, Rowe, & Sheridan, 2007), which can provide useful data, but may give less information of real pathogen food interactions. A number of broth based studies (Eblen, Annous, & Sapers, 2005; Salter, Ross, & McMeekin, 1998) have suggested that nonpathogenic E. coli strains can be used as indicators in modelling the growth of pathogenic E. coli O157:H7, and in evaluating the efficacy of interventions designed to reduce the persistence/numbers of this pathogen during meat processing and storage. In more general terms, the use of data from broth experiments to predict E. coli growth and survival on beef surfaces may be unreliable, because pathogens in broths face different challenges from those posed by meat surfaces. Food structure may have a profound effect on pathogen growth and this has been demonstrated for Listeria monocytogenes, where significant differences in growth were * Corresponding author. Tel.: +353 1 8059500; fax: +353 1 8059550. E-mail address: [email protected] (D.M. Prendergast). 0309-1740/$ - see front matter Ó 2009 Published by Elsevier Ltd. doi:10.1016/j.meatsci.2009.06.024
demonstrated between a broth and a gel system and a broth and an agar surface (Wilson et al., 2002; Koutsoumanis, Kendall, & Sofos, 2004). The influence of surface structure in relation to meat spoilage has also been recognised (Nychas, Skandamis, Tassou, & Koutsoumanis, 2008). For example, a number of recent studies have demonstrated that survival rates of (non-pathogenic) E. coli are different at different sites on beef carcasses (Prendergast and Sheridan, unpublished; Kinsella, 2008). As well as the above physical and environmental factors, pathogen survival and growth has been shown to be significantly influenced by the presence and nature of the natural flora of meat. Thus, Nissen, Maugesten, and Lea (2001) demonstrated more growth of E. coli O157:H7 on heat treated, lactic acid decontaminated beef pieces than on untreated control samples during aerobic storage at 10 °C, which is in agreement with Vold, Holck, Wasteson, and Nissen (2000), who noted that the presence of a beef microflora inhibited the growth of E. coli O157:H7 in beef mince aerobically stored at 12 °C. However, such interactions between pathogens and the wider beef microflora, may be influenced by other physical and/or environmental factors, as Berry and Koohmaraie (2001) reported no such effects on nonminced beef surfaces. A number of studies have examined survival and/or growth of E. coli O157:H7 and other pathogens on excised/cut surfaces (Kinsella et al., 2008; Logue, Sheridan, & Harrington, 2004; Nissen et al., 2001) but much less is known about the survival and/or growth of this pathogen on uncut surfaces, i.e. surfaces with intact
469
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
membranes. Studies of E. coli inoculated onto beef carcass surfaces have demonstrated that growth and survival rates are affected by the nature of the inoculated site (Kinsella, 2008; Prendergast and Sheridan, unpublished). Thus, during holding at a range of chill temperatures, some counts on the cut surface of the neck increased (while the rest remained unchanged) but on the brisket, some counts were reduced (while the rest remained unchanged). On the rump, all counts were reduced, frequently to almost zero. Overall, the results of such studies highlight the need for a more comprehensive understanding of the physical and biological factors modulating bacterial survival and growth on beef carcasses and derived products, in the development of accurate models of the growth of E. coli O157 (and other pathogens) in these complex food environments. The present study examined the growth/survival of E. coli O157:H7 and a potential index organism (generic E. coli) on excised irradiated and non-irradiated beef neck, brisket and rump surfaces, and compared the growth of these organisms on these surfaces, with their growth in broth systems.
2. Materials and methods 2.1. Slaughter and inoculation sites The study used an experimental abattoir at Ashtown Food Research Centre (AFRC), designed for low-throughput slaughter and processing of cattle, sheep and pigs. Six grass fed Hereford Fresian heifers (624 months) were processed under normal commercial slaughter conditions, i.e. lairage, stunning, exsanguination, dehiding, and evisceration, followed by carcass splitting, weighing and washing, to produce carcasses with a mean weight of 122.5 kg (range of 110.0–131.5 kg). Three sites on each carcass were identified as follows: (i) neck, (ii) brisket and (iii) rump. A fibrous membrane of connective tissue or fascia covers the latter two sites, while the selected neck site was the lean surface exposed during slaughter (i.e. not covered by fascia). Two animals were slaughtered per day, with each day representing a complete treatment. The treatment was repeated on three separate days. 2.2. Production of beef samples Immediately after carcass washing, surface samples (approximately 300 cm2, and 5 mm deep) were aseptically excised from each selected site, placed in sterile stomacher bags and transported to the on-site laboratory. From these surface samples cylindrical meat pieces (approximate diameter 2.5 cm2) were aseptically excised using sterile coring punches. On each occasion, 26 2.5 cm2 samples were excised from each of the three carcass sites. Samples were placed in separate 12.5 7.7 cm Cryovac BB4 bags (Sealed Air Cryovac, Cambridge, UK), vacuum packed using a Vac Star, model S220 (Sugiez, Switzerland) and chilled to 4 °C within 1 h. 2.3. Sample irradiation Vacuum packed samples (486) were transported under refrigeration (<10 °C) to the Agrifood Biosciences Institute, Northern Ireland, and irradiated at 4 °C (±1 °C) to a target dose of 5.0 kGy, using a cobalt 60 source (Gammabeam 650, Nordion International Inc., Kanata, Canada) at a dose rate of 5.33 kGy h 1. Irradiation dose was monitored and confirmed using four red perspex dosimeters (Type 4034AN, Harwell Dosimeters Ltd., Harwell, UK) placed throughout the packed meat pieces. Used dosimeters were spectrophotometrically examined (640 nm), and doses calculated using calibration graphs provided by the National Physical Laboratory
(Teddington, UK). The sample preparation process, i.e. excision, vacuum packaging, dispatch, irradiation, and return was carried out under refrigeration conditions within 72 h. Returned, irradiated samples were stored at 0 °C for no more than 24 h. Immediately prior to inoculation, six irradiated samples were randomly selected from each sample type, placed in 30 ml BHI at 25 °C for 18 h, plated onto Oxoid Plate Count Agar (PCA) (Oxoid, Basingstoke, UK), incubated at 30 °C for 3 days, and examined to confirm sterility. 2.4. Inoculum preparation Six bovine isolates of E. coli O157:H7, each of which carried the hlyA and eaeA genes were obtained from the AFRC’s culture collection. Four of the isolates carried the vt1 and vt2 genes, while one each carried either the vt1 or vt2 gene only. All six isolates were made resistant to 50 lg ml 1 nalidixic acid (Sigma–Aldrich, St. Louis, US) and 1000 lg ml 1 streptomycin sulphate (Sigma–Aldrich) by the method of Park (1978). A single non-pathogenic bovine hide isolate of E. coli, was obtained from the AFRC culture collection. In a separate experiment, a six isolate cocktail of non-pathogenic E. coli was prepared to determine if there was a difference in the growth kinetics between the single isolate and a cocktail of strains. The pathogenic and non-pathogenic isolates were individually cultured in Oxoid Brain Heart Infusion Broth (BHI) at 37 °C for 16–18 h. Stationary phase cells were recovered using an Eppendorf model 5403 refrigerated centrifuge at 3000g for 10 min at 4 °C (Eppendorf, Hamburg, Germany), washed three times in Oxoid maximum recovery diluent (MRD) by centrifugation and resuspended/diluted in MRD to give a final inoculum of approximately 5 log10 CFU ml 1. 2.5. Meat surface inoculation Thirteen irradiated (IR) and thirteen non-irradiated control (NIRC) samples from each carcass site were inoculated with a 5 ll inoculum containing approximately 3 log10 CFU of the cocktail of pathogenic E. coli (PEC) or of the non-pathogenic E. coli (NPEC) and held at room temperature for 30 min (to facilitate cell adherence). Inoculated samples (13) and uninoculated controls (13) were transferred to Aqualab cups (Labcell, Basingstoke, UK), covered with caps and placed in refrigerated incubators at 10 °C. Inoculated and uninoculated samples from each site (neck, rump and brisket) were removed after 0, 6, 12, 18, 24, 30, 36, 42, 48, 54, 60, 66 and 72 h, and examined as described below. 2.6. Temperature monitoring Temperatures of the refrigerated incubators were monitored at three positions, top, middle and bottom every 30 min for 72 h, using Type T copper–constantan thermocouples and the data was recorded using a model 2040 Grant Squirrel data logger (Grant Instruments, Cambridge, UK). At the end of each experiment, data were downloaded to MicrosoftÒ Office Excel 2003. 2.7. Enumeration of pathogenic (PEC) and non-pathogenic E. coli (NPEC) Each recovered sample was pulsified for 30 s in 10 ml MRD in a sterile 100 150 mm stomacher bag (Seward, Norfold, UK) using a model PUL Pulsifier (Filtraflex, Ontario, Canada). Undamaged E. coli O157:H7 were enumerated by direct plating of 1 ml aliquots of the above homogenates onto Oxoid Sorbitol MacConkey Agar (SMAC), each supplemented with 50 lg ml 1 nalidixic acid and 1000 lg ml 1 streptomycin sulphate. TM
470
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
Undamaged non-pathogenic E. coli were enumerated by direct plating of 1 ml aliquots of the above homogenates onto MacConkey Agar, supplemented with 50 lg ml 1 nalidixic acid and 1000 lg ml 1 streptomycin sulphate. Sublethally damaged E. coli O157:H7 and non-pathogenic cell numbers were enumerated by a resuscitation procedure, i.e. samples were plated onto Oxoid Tryptone Soya agar (TSA), incubated for 2 h at 25 °C, over poured with SMAC or MacConkey (supplemented with 50 lg ml 1 nalidixic acid and 1000 lg ml 1 streptomycin sulphate), and further incubated at 37 °C for 24 h. 2.8. Total viable counts on meat samples At all sampling times, total viable counts (TVC’s) were enumerated from the inoculated and uninoculated IR and NIR samples. Samples were transferred into sterile stomacher bags containing 10 ml MRD, pulsified as described above, plated onto Oxoid plate count agar (PCA) and incubated at 30 °C for 3 days. 2.9. Growth of PEC and NPEC in MRD and BHI Ten ml volumes (12) of MRD in 30 ml polypropylene tubes with caps (Sarstedt, Numbrecht, Germany) were each inoculated with 5 ll of approximately 3 log10 cfu of PEC or NPEC, and held at 10 °C. Tubes were removed after 0, 6, 12, 18, 24, 30, 36, 42, 48, 54, 60, 66 and 72 h and PEC/NPEC numbers were estimated by direct plating onto SMAC and MacConkey agar supplemented with nalidixic acid and streptomycin sulphate as described above. Ten ml volumes (12) of BHI in 30 ml polypropylene tubes with caps (Sarstedt, Numbrecht, Germany) were each inoculated with 5 ll of approximately 3 log10 cfu of PEC or NPEC, and held at 10 °C. Tubes were removed after 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 11 h and PEC/NPEC numbers were estimated by direct plating onto SMAC and MacConkey agar supplemented with nalidixic acid and streptomycin sulphate as described above.
In almost all cases, there were no significant differences between direct and overlay counts of PEC or NPEC at each sample occasion on IR and NIR surfaces. Thus only overlay counts are presented (Table 1). The study compared mean (overlay) counts (1) between organisms, i.e. PEC and NPEC, (2) between IR and NIR surfaces, and (3) among meat sites (neck, brisket and rump.) (1) Non-pathogenic (NPEC) counts were significantly higher than PEC counts at all three sites on IR samples (P < 0.05), while on NIR samples, there were no significant differences between the counts for these organisms. (2) Counts on IR brisket surfaces were significantly higher than on NIR brisket surfaces, for both PEC and NPEC counts (P < 0.001). On the neck and on the rump, IR counts were not significantly different from NIR counts, except in the case of PEC IR counts on the rump, which were significantly lower than PEC NIR counts (P < 0.05). (3) Among meat sites, counts on the neck were significantly higher than on the brisket or rump for both PEC and NPEC counts, on almost all (IR and NIR) sites (P < 0.001). IR brisket counts were significantly higher than IR rump counts for both PEC and NPEC (P < 0.001), but NIR brisket counts were not significantly different from NIR rump counts. The study compared overlay counts during storage for up to 72 h. There were no significant differences among the time zero counts, between organisms (PEC and NPEC), between IR and NIR surfaces, and among meat sites (neck, brisket rump).
Table 1 Survival of pathogenic (PEC) and non-pathogenic E. coli (NPEC) (log10 CFU cm sterile (IR) and non-sterile (NIR) beef samples for 72 h at 10 °C.
) on
Beef site
Neck
Status
IR
NIR
IR
NIR
IR
NIR
Time after inoculation (h) 0 6 12 18 24 30 36 42 48 54 60 66 72 Mean
PEC 2.81 2.71 2.67 2.64 2.70 2.91 3.19 3.47 3.18 3.19 3.35 3.51 3.46 3.06
2.91 2.81 2.94 3.06 3.12 3.21 3.38 3.55 3.33 3.73 3.80 3.86 4.13 3.37
2.64 2.53 2.43 2.33 2.32 2.33 2.38 2.43 2.71 2.43 2.52 2.61 2.20 2.45
2.34 2.25 2.01 1.77 1.92 1.92 1.38 0.84 1.94 0.73 1.03 1.31 1.70 1.62
2.54 1.69 1.62 1.55 1.60 1.40 1.31 1.17 1.34 1.36 1.05 1.30 1.11 1.46
2.31 2.19 2.17 2.15 1.65 1.56 1.76 1.96 2.22 1.94 1.83 1.71 2.51 2.00
3. Results
0 6 12 18 24 30 36 42 48 54 60 66 72 Mean
NPEC 2.49 2.58 2.88 3.17 3.06 3.08 3.44 3.80 4.45 4.15 4.00 3.84 4.71 3.51
2.52 3.03 3.05 3.07 2.97 3.17 3.30 3.43 4.05 4.34 4.38 4.43 4.31 3.54
2.69 2.19 2.45 2.70 2.82 2.81 3.22 3.63 3.65 3.71 4.05 4.38 4.50 3.29
2.06 1.87 1.82 1.76 1.93 1.82 1.90 1.99 1.96 1.77 1.74 1.72 1.89 1.86
2.43 2.11 2.33 2.54 0.86 1.72 2.58 3.45 2.29 2.02 1.45 0.91 1.67 2.03
2.17 2.01 2.05 2.09 1.48 1.41 1.89 2.41 1.88 2.64 2.90 3.16 1.91 2.15
The mean temperature of the incubators during storage of beef and broth samples were 10.2 ± 0.5 °C.
Standard error of differences between means: columns: 0.47; rows: 0.43 Degrees of freedom: columns: 251; rows: 288 Standard error of differences for means: 0.21 Degrees of freedom for means: 22.
2.10. Statistical analysis All statistical tests examining bacterial counts were performed using log transformations. For all data, 0.998 was added to the original values prior to log transformation giving zero counts a value of 0.0009 on the log scale, while making no appreciable difference to the log values of all other (non-zero) counts (Prendergast, Sheridan, Daly, McDowell, & Blair, 2004). This allowed manipulation of all the data (including zero counts), with minimal significant effect on the overall results. The experiment was designed as a split-split-plot with temperature and three carcass sites on the main plots, counting method on the subplots and time on the sub-sub-plots. Analysis of variance (ANOVA) was carried out at the P < 0.05 level of significance using Genstat 5 (Rothamsted Experimental Station, Harpenden, United Kingdom). The growth kinetics of E. coli O157:H7 and non-pathogenic E. coli in BHI and MRD broths were determined using the Gompertz function, where parameters A, B, C and M were used to calculate the exponential growth rate (ERG; log10 CFU ml 1 d 1), generation time (GT; day) and lag phase duration (LPD; day). Linear regression analysis was used to compare the exponential growth rates (slopes) between experiments in MRD, using either a single nonpathogenic isolate or a six isolate cocktail.
Brisket
2
Rump
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
PEC counts on IR surfaces did not change over time on the neck or brisket, but were significantly reduced on the rump (P < 0.001). NPEC counts on IR surfaces showed significant increases on the neck and brisket (P < 0.001), but were unchanged on the rump, during storage. PEC counts on NIR surfaces increased on the neck (P < 0.05), with no change on either the brisket or rump. Similar results were observed in relation to NPEC counts (P < 0.001). Significant increases, i.e. growth was observed on the neck (1PEC/2NPEC) and the brisket (1NPEC), but not on the rump. A comparison between irradiated and non-irradiated uninoculated controls revealed that irradiation had not completely sterilised all the samples. A small residual microflora remained after irradiation, which was significantly smaller than on the non-irradiated flora at all sites (P < 0.01). Mean TVC’s on inoculated IR meat surfaces were overwhelming composed of the inoculated E. coli cells. On NIR surfaces, counts on the neck were significantly higher on inoculated surfaces (P < 0.001), than on uninoculated surfaces (Table 2). Differences between these counts were related to the ability of PEC and NPEC to grow on the neck surface, in the presence of the normal meat surface microflora as indicated above (Table 1). Differences between the growth/survival of inoculated PEC and NPEC were also evident from the TVC counts on IR samples. On the three sites inoculated with PEC, the TVCs were always significantly lower (P < 0.001), than the TVC counts from NPEC inoculated surfaces. This effect was also noted on NIR samples, but only for the neck and brisket (P < 0.05). The growth kinetics of PEC and NPEC in BHI broth, were similar (Table 3). In MRD, their growth kinetics was different, i.e. NPEC grew within 3 days, but PEC failed to grow within this period. NPEC growth was much slower in MRD than PEC or NPEC in BHI. The shorter lag time for NPEC in MRD suggested that NPEC adapted to the minimal growth conditions of MRD more effectively than it did in BHI. As only a single isolate of NPEC was used in these experiments, as opposed to a cocktail of six isolates for PEC, a second growth experiment was undertaken in MRD using a cocktail of six
Table 2 Mean total viable counts (log10 CFU cm 2) on irradiated and non-rradiated beef sites, uninoculated and inoculated with pathogenic (PEC) or non-pathogenic (NPEC) E. coli. Sites
Inoculum
Irradiated Uninoculated
Neck Brisket Rump
PEC NPEC PEC NPEC PEC NPEC
Non-irradiated Inoculated
1.93 1.74 0.84 1.30 0.82 1.15
3.20 4.62 1.53 3.03 1.52 2.53
Uninoculated
Inoculated
3.40 2.74 2.35 2.93 1.57 2.26
4.17 4.79 2.42 3.04 2.14 2.61
Standard error of differences between means: 0.25 Degrees of freedom: 48.
Table 3 The influence of media on the growth kinetics of pathogenic E. coli (PEC) and nonpathogenic E. coli (NPEC) at 10 °C. Medium
BHI MRD
Organism
PEC NPEC PEC * NPEC 1 NPEC 2
Growth parameters LPD (h)
EGR (log10 CFU h
39.24 32.11 NG 18.04 19.68
0.112 0.098 NG 0.043 0.051
1
)
GT (h) 2.70 3.07 NG 7.00 6.00
LPD: lag phase duration; EGR: exponential growth rate; GT: generation time; BHI: brain heart infusion broth; MRD: maximum recovery diluent; NG: no growth; *NPEC 1: experiment using a single isolate; and NPEC 2: experiment using a six isolate cocktail.
471
non-pathogenic isolates. The data in Table 3 show that the growth kinetics were similar using a single isolate or a cocktail of nonpathogenic isolates. It was also demonstrated that the exponential growth rates (slopes) from the single isolate and the cocktail were not significantly different.
4. Discussion In this study the growth/survival at 10 °C of pathogenic and non-pathogenic E. coli isolates were compared over time on irradiated and non-irradiated beef surfaces and in broth. The study noted differences between E. coli numbers on irradiated and nonirradiated samples, with greater growth of both inocula on some irradiated samples. The study observed that non-pathogenic E. coli counts were higher than E. coli O157 counts on irradiated samples of all three meat surfaces. However, this was not the case on non-irradiated surfaces. Overall, growth was most frequently observed on the neck, with more growth of non-pathogenic E. coli, than of E. coli O157. This greater growth of non-pathogenic E. coli was also observed as part of the TVC’s from inoculated samples. The ability of non-pathogenic E. coli to outgrow E. coli O157 was similarly observed under the minimal nutrition conditions provided by MRD broth. The observation that non-pathogenic counts were higher than pathogenic counts on irradiated surfaces, but not on non-irradiated surfaces, suggests that irradiation may be affecting the chemical or physical composition of the meat surface. Alternatively it may be inactivating the natural microflora, favouring the growth of the non-pathogenic isolate. The possibility exists that irradiation renders the meat surface less favourable to the growth of the pathogen, by the production of radiolytic compounds. In the present study, samples were irradiated with a dose of 5.0 kGy, which gave a reduction of the order of 2–3 log. This result is in line with a previous study, which reported that a dose of 5.0 kGy led to a 2 log reduction in the microflora of fresh beef pieces (Aziz, Mahrous, & Youssef, 2002). In addition to the microbiological changes, such rates of irradiation have been reported to induce a range of chemical changes in irradiated stored beef. These changes arise from the oxidation of proteins and lipids as a result of the action of radiolytically induced free radicals (Grolichová, Dvorák, & Musilová, 2004; Rowe, Maddock, Londergan, & Lonergan, 2004). The impacts of such free radicals in relation to bacterial survival on meat surfaces are not known, and further investigations of such interactions may be necessary if meat irradiation is to be more widely used. For example, the contribution of inactivated bacterial cells, in the growth of any cells, which survive irradiation, may require further investigation. Alternatively, the inactivation of the meat microflora may render the meat surface less capable of supporting microbial growth. The ability of the surface microflora to provide a variety of metabolites or nutrients during growth is well established (Gill & Newton, 1977; Tsigarida & Nychas, 2001). The metabolites provided by one organism may produce a more favourable growth environment for another by a process known as metabiosis (Gram et al., 2002). Such an absolute requirement for a sufficiently favourable environment for the growth of PEC is confirmed by the observation that the pathogenic isolates could not grow in the minimal conditions provided by MRD, while the nonpathogenic isolate could grow in this medium. This study observed greater growth of both inocula on irradiated than non-irradiated brisket samples suggesting that the presence/absence of the normal beef microflora affects the growth of E. coli at this site. The influence of the beef microflora on pathogen growth and survival is not clear. Berry and Koohmaraie (2001)
472
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473
substantially reduced the normal meat microflora on lean beef surfaces using a germicidal UV treatment. These surfaces were sequentially inoculated with different dilutions of a beef microflora and E. coli O157:H7. These authors reported that pathogen growth at 12 °C in the presence of a beef microflora of log 5.0–6.0 was similar to that on uninoculated samples. A study of the growth of E. coli O157:H7 in beef mince, in the presence or absence of four different aerobic background floras (log 5.0–6.0) showed that after storage at 12 °C, pathogen growth was inhibited in the presence of three of the background floras, but not in the fourth (Vold et al., 2000). When no background flora was added to the beef mince the pathogen grew on two of the four samples. The authors suggested that the presence of a background flora, whether naturally present or added to mince, resulted in inhibition of pathogen growth. Alternatively, these differences may relate to changes in beef mince composition. Such effects have been recorded in samples stored at 5.0–7.0 °C. Jay, Vilai, and Hughes (2003) noted that of four samples tested, one yielded 21 different bacterial genera, while two other samples yielded only seven. In the present study, one or more of the above processes may have been in operation, leading to the inhibition of pathogen growth. However, the authors consider it important to note that the current study found that the normal beef microflora, at naturally occurring levels on the brisket surface, induced significant reductions in E. coli, and particularly in E. coli O157, on this site. While this effect was not universally observed at the other sites, this is the first report of such an effect by the naturally occurring microflora, which is perhaps more significant than previous reports which noted inhibition at artificially increased levels of beef microflora. As well as the microflora, the different meat substrates may have significant roles in preventing or allowing the growth/survival of the E. coli types. In terms of the pathogen on irradiated surfaces, counts on neck samples were significantly different from those on the brisket and rump, and brisket counts were significantly different from those on the rump. The same pattern was observed among non-irradiated pathogen counts, except that brisket counts were not different from those on the rump. Non-pathogenic counts were the same as the pathogen counts, on irradiated and non-irradiated surfaces, with one exception where there was no difference between irradiated neck and brisket counts. Differences in survival of E. coli related to beef sites has been recorded previously on inoculated carcasses after storage at 10 °C for 72 h (unpublished work by Prendergast and Sheridan). In that study, there were significant differences in the survival of E. coli, with growth on the neck, no change in counts on the brisket and a decline in counts to almost zero on the rump. In contrast, in the present study there was growth on the neck, but no change in counts on the brisket or rump over time, for pathogenic and non-pathogenic E. coli. The data from these experiments demonstrated that while the neck site behaved in a similar manner with growth of E. coli, the other sites behaved differently. The difference between these two experiments may be related to the influence of changing carcass temperature and aw at the different sites, which occurs during chilling but is not present on meat pieces incubated at a specific temperature in a closed container where the relative humidity is uniformly high at all times. These parameters have been shown to have a major impact on the survival of E. coli inoculated on beef surfaces at different sites, both under commercial and experimental conditions. During storage, the meat microflora grew on non-irradiated samples on the three meat types, particularly on the neck, as did the two inoculated E. coli types. While the growth of non-pathogenic E. coli on the irradiated brisket could be related to the removal of the surface microflora, this was not the case for the neck samples. At this site, growth of both E. coli types occurred
on irradiated and non-irradiated samples, indicating that the neck could support the growth of the natural microflora and the inoculated organisms, as shown by the higher counts for the TVC’s on inoculated samples. The ability of the neck site to support growth may be related to the increased availability of a specific nutrient, such as glucose, associated with the cut surface of the neck, compared to the brisket and rump where membranes cover the meat surface, possibly making nutrients less readily available. According to Gill and Newton (1978) however, the growth of bacteria on cut and intact surfaces will be similar as surface membranes will not be a barrier to diffusion of solutes to the meat surface. In addition to these factors, differences in meat structure per se must also be considered in relation to E. coli growth. The influence of surface structure on the survival of pathogens on meat surfaces has been recognised in relation to beef carcasses (Prendergast et al., 2007). During the development of rigor, structural changes occur on meat surfaces in the form of gaps due to shrinkage in muscle fibers and other structures, which aid bacterial survival due to cell penetration into these gaps (Frank, 2001; Thomas, O’Rourke, & McMeekin, 1987). In addition, rupture of capillary blood vessels occurs in the connective tissue, fat and muscle (Belk, Scanga, Smith, & Grandin, 2002). The blood contains low molecular weight compounds, such as glucose, glycerol, glycine betaine and carnitine (Smith, 1996) which are available to bacterial cells on surfaces and can aid in survival (Gill & Newton, 1980). Overall, the results of this study showed that NPEC grew better that PEC on all meat surfaces, suggesting that the non-pathogenic E. coli was better fitted to growth on the naturally contaminated beef surfaces than PEC. Such fitness may relate to differences in abilities to compete with the natural microflora as demonstrated in this study, or to the greater ability of the non-pathogen to grow under partially nutrient limited conditions. The second of these possibilities is confirmed by the broth results in the current study, where the NPEC was able to grow in the much more nutrient restricted environment posed by MRD. In conclusion this study demonstrated that there were differences in growth and survival between pathogenic and non-pathogenic E. coli isolates on meat surfaces and in broth. It was also apparent that the natural microflora of the meat surface had a major influence on the survival of pathogenic and non-pathogenic E. coli, as did the nature of the different meat surfaces. This study has demonstrated that at the levels of contamination frequently observed under commercial conditions, the normal meat microflora has a significant effect on the growth and survival of E. coli on some meat surfaces. Overall the results of this study demonstrate the challenges in using model organisms, and model liquid and/or surfaces, and environmental conditions in the development of accurate models for the prediction of the survival/growth of pathogens on meat surfaces under commercial conditions. Acknowledgements This research has been carried out with the financial support of the Commission of the European Communities, specific RTD program ‘‘Quality of Life and Management of Living Resources” Project No. QlK1-CT2002-02545. The authors wish to thank Dr. Eileen Stewart, Queens University Belfast for irradiation of the beef samples. Sincere thanks are also due to Dr. Dermot Harrington and to Ms. Paula Reid, Teagasc for their assistance with the statistical analysis. References Aziz, N. H., Mahrous, S. R., & Youssef, B. M. (2002). Effect of gamma-ray and microwaves treatment on the shelf-life of beef products stored at 5 C. Food Control, 13, 437–444.
D.M. Prendergast et al. / Meat Science 83 (2009) 468–473 Belk, K. E., Scanga, J. A., Smith, G. C., & Grandin, T. (2002). The relationship between good handling/stunning and meat quality in beef, pork and lamb. In Proceedings of the American Meat Institute Foundation, animal handling and stunning conference, February 21–22.
473
Logue, C. M., Sheridan, J. J., & Harrington, D. (2004). Studies of steam decontamination of beef inoculated with Escherichia coli O157:H7 and its effects on subsequent storage. Journal of Applied Microbiology, 98, 741–751. Nissen, H., Maugesten, T., & Lea, P. (2001). Survival and growth of Escherichia coli O157:H7, Yersinia enterocolitica and Salmonella enteritidis on decontaminated and untreated meat. Meat Science, 57, 291–298. Nychas, G-J. E., Skandamis, P. N., Tassou, C. C., & Koutsoumanis, K. P. (2008). Meat spoilage during distribution. Meat Science, 78, 77–89. Park, R. W. A. (1978). The isolation and use of streptomycin resistant mutants for following development of bacteria in mixed populations. In D. Lovelock & R. Davies (Eds.), Techniques for the study of mixed populations (pp. 107–112). London: Academic Press. Prendergast, D. M., Sheridan, J. J., Daly, D. J., McDowell, D. A., & Blair, I. S. (2004). The use of a marked strain of Pseudomonas fluorescens to model the spread of brain tissue to the musculature of cattle after shooting with a captive bolt gun. Journal of Applied Microbiology, 96, 437–446. Prendergast, D. M., Rowe, T. A., & Sheridan, J. J. (2007). Survival of Listeria innocua on hot and cold beef carcass surfaces. Journal of Applied Microbiology, 103, 2721–2729. Rowe, L. J., Maddock, K. R., Londergan, S. M., & Lonergan, E. H. (2004). Influence of early postmortem protein oxidation on beef quality. Journal of Animal Science, 82, 785–793. Salter, M. A., Ross, T., & McMeekin, T. A. (1998). Applicability of a model for nonpathogenic Escherichia coli for predicting the growth of pathogenic Escherichia coli. Journal of Applied Microbiology, 85, 357–364. Smith, L. T. (1996). Role of osmolytes in adaptation of osmotically stressed and chillstressed Listeria monocytogenes grown in liquid media and on processed meat surfaces. Applied and Environmental Microbiology, 62, 3088–3093. Thomas, C. J., O’Rourke, R. D., & McMeekin, T. A. (1987). Bacterial penetration of chicken breast muscle. Food Microbiology, 4, 87–95. Tsigarida, E., & Nychas, G.-J. E. (2001). Ecophysiological attributes of a Lactobacillus sp. and a Pseudomonas sp. on sterile beef fillets in relation to storage temperature and film permeability. Journal of Applied Microbiology, 90, 696–705. Vold, L., Holck, A., Wasteson, Y., & Nissen, H. (2000). High levels of background flora inhibits growth of Escherichia coli O157:H7 in ground beef. International Journal of Food Microbiology, 56, 219–225. Wilson, P. D. G., Brocklehurst, T. F., Arino, S., Thuault, D., Jakobsen, M., Lange, M., et al. (2002). Modelling microbial growth in structured foods: towards a unified approach. International Journal of Food Microbiology, 73, 275–289.