Loss of viability of Listeria monocytogenes in contaminated processed cheese during storage at 4, 12 and 22 °C

Food Microbiology 27 (2010) 809e818

Contents lists available at ScienceDirect

Food Microbiology journal homepage: www.elsevier.com/locate/fm

Loss of viability of Listeria monocytogenes in contaminated processed cheese during storage at 4, 12 and 22
C Apostolos S. Angelidis*, Paraskevi Boutsiouki, Demetrios K. Papageorgiou Laboratory of Milk Hygiene and Technology, Department of Food Hygiene and Technology, School of Veterinary Medicine, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 January 2010 Received in revised form 28 April 2010 Accepted 29 April 2010 Available online 7 May 2010

The behaviour of Listeria monocytogenes in a processed cheese product was evaluated over time by inoculating the product with three different L. monocytogenes strains (Scott A, CA and a strain isolated from processed cheese) at three different inoculation levels (ca. 6  105, ca. 6  103 and 102 CFU/g of cheese or less) and after storage of the contaminated products at 4, 12 or 22
C. Growth of L. monocytogenes was not observed in any of the experimental trials (experiments involving different combinations of strain, inoculum level and storage temperature) throughout the storage period. L. monocytogenes populations decreased over time with a rate that was strain- and storage temperature-dependent. Nonetheless, for cheeses that had been inoculated with the higher inoculum and stored at 4
C viable populations of L. monocytogenes could be detected for up to nine months post-inoculation. The L. monocytogenes survival curves obtained from the different trials were characterised by a post-inoculation phase during which the populations remained essentially unchanged (lag phase) followed by a phase of logarithmic decline. The duration of the lag phase and the rate of inactivation of L. monocytogenes in the different trials were estimated based on data from the linear descending portions of the survival curves. In addition, a non-linear Weibull-type equation was fitted to the data from each survival curve with satisfactory results. The results of the present study emphasize that, according to the definition laid down in the European Union Regulation 1441/2007, the processed cheese product tested in this work should be considered and classified as one that does not support the growth of L. monocytogenes under reasonable foreseeable conditions of distribution and storage. However, postprocessing contamination of the product should be austerely avoided as the pathogen can survive in the product for extended periods of time, particularly under refrigerated storage (4
C). Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Listeria Processed cheese Survival Weibull model

1. Introduction Listeria monocytogenes is a food-borne pathogen that is widely distributed in nature and the causative agent of listeriosis, a serious and often life-threatening disease (Ryser and Marth, 2007). Owing to its elaborate physiological adaptation mechanisms, L. monocytogenes can survive and even proliferate in a variety of foods under adverse environmental conditions such as low pH, high salinity and low temperature (Angelidis et al., 1999; Hado and Yousef, 2007). Recent epidemiological data from eight European Union (E.U.) Member States have indicated that the incidence of listeriosis in humans has increased or remained relatively high since the year 2000, with the majority of cases concerning the elderly and those with predisposing medical conditions (Goulet et al., 2008).

* Corresponding author. Tel.: þ30 2310 999862; fax: þ30 2310 999803. E-mail address: [email protected] (A.S. Angelidis). 0740-0020/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.fm.2010.04.017

Beginning 2006, Commission Regulation (EC) 2073/2005 (amended by Commission Regulation (EC) 1441/2007) became effective for all E.U. Member States (European Commission, 2005, 2007). Compared to previously existing legislation, of particular interest are the legislative modifications regarding L. monocytogenes in ready-to-eat (RTE) foods. Thus, RTE foods are legislatively distinguished based on the intended target population (i.e., infants or people with special medical conditions versus all other human sub-populations). RTE foods for infants or for special medical purposes are still required to be free of L. monocytogenes (absence of the pathogen in 25 g of food in a 10-unit sampling plan). RTE foods other than those intended for infants or special medical purposes are then subdivided into those that are able to support the growth of L. monocytogenes and into those that are not. Products “with pH  4.4 or aw  0.92, products with pH  5.0 and aw  0.94 and products with a shelf-life of less than five days” are considered as RTE foods that are unable to support L. monocytogenes growth. The Regulation also states that “other categories of products can also belong to this category, subject to scientific justification”. Finally, the

810

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

L. monocytogenes criteria are adjusted according to their temporal stage in the food chain. Thus, for RTE foods that are able to support the growth of L. monocytogenes, the legislation compels the absence of the pathogen (in 25 g) “before the food has left the immediate control of the food business operator, who has produced it”, but allows for up to 100 CFU/g in “products placed on the market during their shelf-life”. The 100-CFU/g limit also applies throughout the shelf-life of marketed RTE foods that are unable to support L. monocytogenes growth. Under the new E.U. legislation, therefore, it is essential for manufacturers of RTE foods to be able to demonstrate to the competent authorities the L. monocytogenes-food category in which their products belong to. One way of providing such justification for products that do not meet the fore-mentioned physicochemical and shelf-life limits is through the conduct of challenge tests. The aim of the current work, therefore, was to characterise a grated processed cheese product with respect to its ability or not to support the growth of L. monocytogenes. Pasteurised processed cheeses are typically produced by blending natural cheeses of different ages and degrees of maturity in the presence of emulsifying salts and other dairy and non-dairy ingredients such as colours, flavourings, fats, spices, salts, acidulants and preservatives. Following the preparation of the desired formulation, the ingredients are mixed under heating. The melted homogeneous mass is then packaged, cooled and stored to yield a product with an extended shelf-life (Kapoor and Metzger, 2008). The minimum heat treatment required by the United States Code of Federal Regulations is 65.5
C for 30 sec (FDA, 2008). The formulation of the processed cheese product tested in this work includes a specific type of hard cheese, which is ripened for three months, milk casein, butter, sodium chloride (2.8%), sorbic acid (0.11%), phosphates (2.7%) as emulsifiers and water. The mixture is heated (80  1
C) under continuous mixing for 3e4 min in industrial 200-kg capacity, Stephan-type cookers in batch operation with direct steam injection. The melted homogeneous cheese mass is hot-filled in plastic package (film) in rectangular stainless steel hoops (12-kg block sizes) and then cooled in continuous flow water tunnels for 4e5 h until the cheese temperature reaches 15
C. The hoops containing cooled cheese are subsequently transferred to 4
C-cold rooms for at least 3 days for further cooling in order to achieve the desired firmness and texture. In the final step of manufacture, the cooled and solidified cheese is removed from the steel hoops and feeds the grating line where the cheese package is removed and the product is grated. The grated cheese is subsequently packaged under modified atmosphere (30% CO2 and 70% N2). Cheese packages are stored at 2e8
C during storage and distribution. The product, which is commercially sold as TandoÒ processed cheese, has a shelf-life of 12 months. According to the definitions laid down in Regulation 2073/2005, the product can not be a priori classified to the category of RTE foods that are unable to support L. monocytogenes growth based on its pH, aw and its intended shelf-life. L. monocytogenes is often isolated from the environment (e.g. floors, drains) of cheese companies even when good sanitation and hygiene protocols are in place (Pritchard et al., 1994; Kornacki and Gurtler, 2007). The present study was therefore undertaken to evaluate the behaviour of L. monocytogenes when the pathogen is introduced to the processed cheese as a post-processing contaminant and the product is subsequently stored under refrigeration (4
C) or temperature abuse (12 and 22
C) conditions. Such a contamination could, in theory, occur after the cooling of the melted cheese mass and before the final packaging, e.g. during the grating process. The influence of strain and level of contamination was also studied by using three strains of L. monocytogenes of different origin and serotype and three different levels of initial contamination.

2. Materials and methods 2.1. L. monocytogenes strains Three L. monocytogenes strains were used in this study. The first strain, L. monocytogenes Scott A (clinical isolate, serotype 4b), was selected as one of the test strains because previous research has shown that strain Scott A is particularly resistant and survives for prolonged periods of time in contaminated dairy products (Papageorgiou and Marth, 1989a,b). The second strain was one of the L. monocytogenes strains that were isolated during the investigations of the soft-cheese listeriosis outbreak in Los Angeles in 1986. Previous research has shown that this strain, herein designated as L. monocytogenes CA, is less resistant in terms of its ability to survive for extended time periods in contaminated dairy products (Papageorgiou and Marth, 1989a,b). The third strain, designated as L. monocytogenes IS951, is a processed cheese isolate. The serotypic group of strains CA and IS951 was determined as described below. Bacterial strains were stored at 70
C in tryptic soy broth (TSB, Biolife Italiana S.r.l., Milano, Italy) as glycerol (20%) stocks. 2.2. Inoculum preparation Unless specified otherwise, all media and reagents were obtained from Biolife. After growth on tryptic soy agar (TSA) plates, each strain was grown separately in TSB at 30
C for 24 h and then sub-cultured in fresh TSB at 30
C for an additional 20 h. Fully grown (stationary phase) cultures in TSB were then centrifuged (10,000 RPM for 10 min) in an Eppendorf mini spin plus centrifuge (Eppendorf, AG, Hamburg, Germany) at room temperature. The supernatants were discarded and the cells were washed twice (resuspended and centrifuged as described above) in sterile ¼-strength Ringer’s solution (LAB M Limited, Lancashire, U.K.). The washed pellets were brought up and serially diluted in Ringer’s solution to yield the required inocula (“high”, “medium” or “low”) as described below. 2.3. Cheese inoculation, packaging and storage Cheese was aseptically removed from its commercial package (typically plastic bags containing 500 g or 1 kg of grated processed cheese) and 25-g portions were aseptically transferred in 400-mL sterile stomacher bags (BagLightÒ, Interscience, St. Nom, France). Forty mL of washed and appropriately diluted cell suspension were spread drop-wise, as uniformly as possible, over the cheese mass inside each bag to yield three types of initial inocula. The “high” inoculum was designed to yield ca. 6  105 CFU/g, the “medium” inoculum consisted of ca. 6  103 CFU/g and the “low” inoculum was intended to be 102 CFU/g or less. Control samples for each experiment were inoculated with 40 mL of sterile ¼-strength Ringer’s solution. Immediately upon their inoculation, the bags containing inoculated cheese were hand-massaged for 10 s to achieve a more uniform distribution of the inoculum in the grated cheese mass. Subsequently, the bags were heat-sealed in a controlled atmosphere environment (30% CO2/70% N2) using a YANG sealer (YANG Technologie Alimentari, Model No Atlantis 25, Vertemate, Italy) in order to mimic the product’s commercial atmospheric packaging conditions. Sealed bags were appropriately labelled and stored at 4, 12 or 22
C in high-precision controlled temperature incubators (Model MIR-253, Sanyo Electric Co., Ltd, Osaka, Japan) for up to one year. Thus, a total of 27 different experimental conditions (trials) were generated and tested over time in duplicate (3 strains  3 inoculum levels  3 storage temperatures  2 replications) resulting in 27 L. monocytogenes survival curves.

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

2.4. Cheese composition, microbiological and physicochemical characteristics Several measurements were conducted on samples of grated processed cheese at different time points. i) Prior to the onset of the experiments (inoculations) the pH and aw values of the grated cheese from 10 different production batches selected at random over a six-month period were evaluated in duplicate. ii) On the day of the experiments and prior to the inoculations, the pH, aw, the aerobic plate count (APC) as well as the moisture and fat content of the cheese used for the inoculations were measured (four determinations from the batch used for the experiments). The potential effect of the liquid inoculum on the aw value of the cheese was investigated by comparing the aw of uninoculated cheese with the aw value of cheese (25 g) that had been fortified and mixed with 40 mL of ¼-strength Ringer’s solution. iii) The pH and aw values of cheese samples were also evaluated in duplicate half-way as well as after the end of each experiment for trials of all three storage temperatures. In addition, two 25-g portions of the cheese used for the experiments were tested for the presence of L. monocytogenes using the ISO 11290-1 detection method (ISO, 1996). Standard methods were used for the determination of fat and moisture content and for the determination of APC (Wehr and Frank, 2004). The aw was measured at 25
C using an Aqualab water activity meter (Aqualab, model series 3, Decagon devices Inc., Pullman, WA), whereas the pH was measured after blending 25 g of cheese in quadruple volume (100 mL) of distilled and deionised water using a Consort pH meter (Consort, model C830, Turnhoot, Belgium), equipped with a Hanna FC-100 electrode (Hanna Instruments, Woonsocket, RI). 2.5. L. monocytogenes strain serotyping The serotypic group of strains CA and IS951 were determined by the multiplex PCR method developed by Doumith et al. (2004) with minor modifications. This multiplex PCR method distinguishes L. monocytogenes isolates in four major serovar groups. Four L. monocytogenes serotype reference strains, purchased from the strain collection of the Institute Pasteur (CIP) in France, were tested in parallel with strains Scott A, CA and IS951 as controls: L. monocytogenes strain CIP 104794 (serovar 1/2a), CIP 105449 (serovar 1/2b), CIP 105448 (serovar 1/2c) and CIP 78.38 (serovar 4b). The primers used, the PCR setup and the PCR amplification conditions were performed as described by Doumith et al. (2004) with the following modifications: i) The bacterial DNA from each strain was isolated from overnight TSB cultures (30
C) using a commercially available kit (Standard protocol for cultured cells, NucleoSpinÒ, Macherey-Nagel, GmbH & Co. KG). ii) The concentration of the primer sets lmo737 and ORF2819 was decreased to 0.5 mM in the reaction mixture and the concentration of the primer set lmo1118 was increased to 2.0 mM in the reaction mixture. iii) The final reaction volume per PCR tube was 50 mL and two units of DyNAzyme II DNA polymerase (Finnzymes Oy, Espoo, Finland) were used per reaction. Ten mL of each reaction mixture were mixed with 2 mL of a 10x loading buffer (Takara Bio Inc, Japan) and loaded into the wells of a 2% agarose gel containing 0.01% of the GelRed nucleic acid gel stain (Biotium, Hayward, CA). Five mL of a 100-bp DNA ladder solution (GenScript Corporation, Piscataway, NJ) were mixed with 1 mL of a 6 DNA loading buffer and the mixture was run on the same gel as a molecular weight marker. The gel electrophoresis was conducted at 100 V for 90 min in 1 TAE buffer (40 mM Triseacetate, 1 mM EDTA, pH 8.0). PCR products were visualized under UV light using a BioDoc-ItÒ Gel Documentation system (UVP, Cambridge, UK).

811

2.6. Microbiological analyses Immediately upon inoculation as well as at multiple time points throughout the shelf-life of the product (1 year), L monocytogenes counts in the inoculated product were determined by testing duplicate samples (i.e. 2 bags) per inoculum level, L. monocytogenes strain and storage temperature. The frequency of examination (testing frequency) as well as the total number of samples that were examined per each strain, temperature and inoculum level was decided a priori based on preliminary experiments (data not shown), but also depended on the behaviour of the pathogen under each actual experimental condition (trial). In general, the testing frequency was higher for samples stored at higher temperatures (i.e. 22
C > 12
C > 4
C) because in these trials the L monocytogenes counts over time tended to change faster (see Results section). Hence, a typical testing frequency was approximately every 3 weeks for samples stored at 4
C, every 1 week for samples stored at 12
C and every 3e4 days for samples stored at 22
C. At each sampling point the sealed stomacher bags containing contaminated cheese were removed from their respective incubator (4, 12 or 22
C) and opened aseptically with the use of alcohol and flame-disinfected scissors. The enumeration of L. monocytogenes was conducted essentially according to the scheme laid down in the ISO 11290-2 method (ISO, 1998). Briefly, 225 mL of sterile Buffered Peptone Water (BPW) were added to each bag. The contents were stomached for 4 min (two min at speed setting “6”, followed by two min at speed setting “8”) using a Bagmixer 400 stomacher (Interscience). Bags were left at 20
C for 1 h and then their contents were serially diluted and plated onto ALOA agar medium. Plates were incubated at 37
C for 48 h and characteristic L. monocytogenes colonies were enumerated. The plates were then transferred back in the incubator for an additional incubation period of 24 h after which an additional count was performed. If the latter count was higher than the 48-h count (typically the case in experiments conducted at 4
C), the 72-h count results were recorded. At several time points throughout the experiments selected colonies from ALOA agar plates were biochemically confirmed as L. monocytogenes as described in the ISO method. The confirmation was conducted mainly on isolated colonies from ALOA plates when ALOA was used as the detection medium and less frequently on colonies from ALOA enumeration plates. The limit of detection of the enumeration method was 5 CFU/g (spread plating of 2 mL from the 101 suspension onto 8 predried, 90 mm-diameter, ALOA agar plates). At each time point the counts from the two bags were averaged and expressed as log CFU/g. For each of the 27 experimental conditions (trials), when the L. monocytogenes counts in the cheese reached a very low level (i.e. close to 10 CFU/g) the detection/enrichment protocol ISO 11290-1 (ISO, 1996) was used in parallel with the enumeration protocol. To enable this, the cheese samples were stomached as described above, with the exception that half-Frazer broth was used as the suspending medium. A portion of the suspension was used for enumeration purposes by spread plating on ALOA agar plates as described above, while the remaining suspension was incubated at 30
C for 24 h (primary enrichment). One hundred mL of the enriched suspension was used to inoculate tubes containing 10 mL of Fraser broth (FB), which were subsequently incubated at 37
C for 48 h (secondary enrichment). Following incubation, loopfuls from the FB tubes were streaked onto ALOA agar plates which were incubated for 48 h at 37
C to determine the presence or absence of viable L. monocytogenes cells in the cheese samples. For each experimental condition testing ceased when negative enrichment results were obtained in duplicate in four successive test dates. It should be noted that for purposes of graphical clarity different arbitrary numerical values were assigned to enrichment-positive results from cheese samples that had been inoculated with the

812

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

“high”, “medium” and “low” inocula. Thus, when bacterial counts in cheese samples decreased to below the level of detection via enumeration but, nonetheless, the presence of viable listeria could be verified via the application of the enrichment protocol, the value of log4 (0.6) was assigned to enrichment-positive samples that had been inoculated with the “high” inoculum. Likewise, enrichmentpositive results for samples that had been initially contaminated with the “medium” and the “low” inocula were assigned the values of log 3 (0.48) and log 2 (0.30), respectively. All enrichmentenegative results were assigned the value of log 1 (0). 2.7. Kinetics of L. monocytogenes inactivation The L. monocytogenes survival curves obtained from the different trials (see Results section) were characterised by an initial postinoculation phase of variable duration during which the L. monocytogenes populations remained essentially unchanged (apparent lag phase). Subsequent to the apparent lag phase was a phase of logarithmic decline which displayed a variable, depending on the experimental trial, albeit noticeable degree of downward curvature (concavity). Therefore, the quantitative data obtained from the “high” and “medium” inoculum trials (18 trials) were subjected to two different analyses for describing the inactivation kinetics of L. monocytogenes. The “low” inoculum trials yielded limited quantitative data (i.e. only relatively few data points were available from enumeration analyses as opposed to detection analyses) and were therefore not used for kinetic analyses. In both approaches and for all experimental trials (i.e. experiments involving different combinations of L. monocytogenes strain, initial inoculum level and storage temperature) the average values (log CFU/g) from duplicate determinations for each time point were used as the dependent variable. The first method for describing the inactivation of L. monocytogenes in the different trials was to adopt the classical first-order (linear) kinetic approach. In this approach the inactivation rate constant (m) and the duration of the apparent lag phase (l) of the L. monocytogenes populations in contaminated cheese were calculated from the log CFU/g vs. time graphs as follows: By visually observing the graphs, following inoculation (time 0) there was an initial apparent lag phase during which the L. monocytogenes populations remained essentially unchanged. The lag phase was followed by a linear descending portion signifying a period of bacterial population decline. Therefore, an equation of the form:

log N ¼ a þ mt

(1)

was fitted to the data points that were within the linear descending portion of the survival curve. In the above equation, N is the number of viable L. monocytogenes (CFU/g), a is the regression equation constant (log CFU/g), m is the slope (days1) and t is time (days). The slope of the above equation was taken as the rate of inactivation (inactivation constant). For each graph, a minimum of three (albeit in most cases more than four) data points in the linear descending portion of the inactivation curve that displayed a value for the coefficient of determination (R2) of 0.90 or higher were selected to provide an estimate and a 95% Confidence Interval (95% C.I.) for m via regression analysis. Analogous to the meaning of the growth rate constant in bacterial growth curves (number of decimal increases in the number of bacteria that occur per unit time in an exponentially growing culture), the inactivation or death rate constant signifies the number of decimal reductions per unit time in an exponentially dying culture. The term apparent lag phase herein denotes the time period in each survival curve from the day of product inoculation (time 0) to the onset of reduction in the L. monocytogenes populations. This period is often designated as the “shoulder” of a survival curve.

During the apparent lag phase, there is no change in the numbers of bacteria. Therefore, for t  l, log N ¼ log N0, where N0 is the level of initial inoculation (CFU/g). The duration of the lag phase (in days) was determined by substituting in equation (1) for t ¼ l, log N ¼ log N0 and solving for l:

l ¼ ðlog N0  aÞ=m

(2)

For these calculations only the average m values obtained from the corresponding regression equations were used and therefore single lag phase estimates were obtained. The second approach for describing the inactivation of L. monocytogenes during storage of the contaminated cheese relied on the utilization of all the quantitative data points from each semilogarithmic survival curve. Hence, the data from time zero (inoculation day) up to, but not including, the data points obtained with culture enrichments were fitted into the following equation:

logðNðtÞÞ ¼ logðN0 Þ  kt n

(3)

Equation (3) represents the cumulative form of the Weibull distribution presented as a semilogarithmic relationship (Peleg and Cole, 1998). In Equation (3), N(t) is the number of surviving L. monocytogenes cells (CFU/g) at time t, N0, the initial number of L. monocytogenes cells (initial inoculum, CFU/g), k the rate of inactivation (daysn) and n is a constant. In the above equation, when the constant n assumes values lower than 1, the survival curve is characterised by an upward concavity. In such situations it is hypothesized that the bacterial cells are getting adapted or more resistant to the “treatment” they are exposed to as time progresses (tailing). On the contrary, when n > 1 the shape of the survival curve is characterised by an upward concavity and in these situations it is proposed that the microbial cells are becoming increasingly damaged over time (Peleg and Cole, 1998; van Boekel, 2002). For each curve the parameters N0, k and n were simulated using the solver function of Microsoft Excel. 3. Results 3.1. Proximate analyses of cheese Proximate analyses for pH, moisture, fat and water activity are summarised in Table 1. The composition of the processed cheese blend is a function of the composition of the raw ingredients used in the formulation and in particular a function of the composition of the hard cheese. Hence, the physicochemical characteristics of processed cheeses may, in theory, vary somewhat from batch to batch. Therefore, prior to the onset of the experiments, the two important intrinsic factors affecting microbial growth and survival, namely the pH and the aw, were examined in TandoÒ grated processed cheese obtained from 10 different production batches selected at random within a 6-month period. The pH and aw values were found to be quite constant from batch-to-batch (pH ¼ 5.03  0.02 and aw ¼ 0.93  0.002). It appears, therefore, that with respect to these two physicochemical characteristics there was minimal variation among the products from different production batches over time. Also, the batch of cheese that was used for the experiments had very similar values in terms of its pH and aw Table 1 Proximate analysis and aerobic plate counts of the batch of processed cheese that was used for the experiments in this study.a Moisture (%)

Fat (%)

pH

aw

APCb [log (CFU/g)]

41.9  0.2

22  0.6

5.00  0.01

0.9273  0.0005

2.16  0.08

a b

Values are the means  standard deviation of four independent determinations. APC, Aerobic Plate Count.

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

with the other batches of the same product over time (Table 1). This constancy in the pH and aw values of the product overtime may be attributed to the fact that the manufacturing company uses the same type of hard cheese, obtained from the same manufacturer, as the major ingredient in the processed cheese blend and abides to a strict recipe for the formulation of the product. The pH and aw values of the cheese used for the experiments remained practically unchanged throughout the storage period (data not shown). APC results of uninoculated samples were consistently low (Table 1) indicating that the process of cheese grating as well as the handling of cheese during the preparation for inoculation did not result in substantial product contamination. The low APC in processed cheese can be attributed to the high heat treatment that the product undergoes in the cookers, a treatment that essentially allows only for the survival of bacterial spores. L. monocytogenes was not detected in the uninoculated cheese used for the experiments. Finally, the addition of the inoculum (cells suspended in 40 mL Ringers) in the cheese did not result in changes in the aw of the product as the aw values of cheese samples with or without 40 mL of added Ringers differed (without any particular direction) only in the third decimal unit of measurement (data not shown). 3.2. L. monocytogenes strain serotyping Fig. 1 shows the results of the multiplex PCR amplifications. Strain L. monocytogenes CA (Fig. 1, lane 2) yielded the same DNA amplification pattern as that of the serotype 1/2c reference strain CIP105448 (lane 8). The amplified genes were lmo1118 (906 bp), lmo737 (691 bp) and prs (370 bp). Therefore, L. monocytogenes CA is classified to the serotypic group 2, as defined by Doumith et al. (2004). Group 2 includes the L. monocytogenes serotypes 1/2c and 3c. The processed cheese isolate, strain L. monocytogenes IS951 (lane 4) yielded the same PCR amplicons with those of the serotype 1/2a reference strain CIP104794 (lane 6). The amplified genes were lmo737 and prs. Therefore, L. monocytogenes IS951 belongs to the serotypic group 1 as defined by Doumith et al. (2004), which includes the L. monocytogenes serotypes 1/2a and 3a. L. monocytogenes Scott A (lane 3), belonging to the serotype 4b, was correctly classified into group 4 with this multiplex PCR procedure as it yielded the same PCR amplicons as those of the serotype 4b reference strain CIP 78.38 (lane 9). Group 4 encompasses serovars 4b, 4d and 4e and the amplified genes are ORF2110 (597 bp), ORF2819 (471 bp) and prs (Doumith et al., 2004). 3.3. Behaviour of L. monocytogenes during storage of inoculated cheeses at 4, 12 and 22
C A simple, yet practical way of comparing the overall survival of L. monocytogenes in the different experimental trials is to report the

813

“time to eradication” i.e. the post-inoculation time for each trial beyond which all of the trial’s examined samples tested negative to culture enrichment. This approach can be considered analogous to the “time-to-detection” approach that is used in experiments dealing with bacterial growth. It should be noted that only an approximation of the “time to eradication” for each different trial can be obtained, as its accuracy depends on the frequency and intensity of sample examination. For all experimental trials conducted herein, sample examination was ceased after four successive enrichment-negative results were obtained, on the assumption that the pathogen had been completely eliminated from the remaining samples. However, in these types of experiments, no matter how tiny, one can not rule out the possibility that some of the remaining samples in the pool of samples that were inoculated at day 0 may still contain viable L. monocytogenes. The behaviour of L. monocytogenes in artificially contaminated processed cheese stored at 4
C is shown in Fig. 2A, D and G. Growth of L. monocytogenes was not observed in any of the trials. Pathogen populations remained relatively unchanged for an estimated storage period of approximately 93 and 100 days, 96 and 98 days and 71 and 68 days in processed cheeses inoculated with the “high” and “medium” inocula of strains Scott A, CA and IS951, respectively (Figs. 2 and 3). After these periods, the L. monocytogenes populations started to decline with comparable rates (Figs. 2 and 4) that equaled to approximately 0.05 log CFU/g per storage day. As a result, the pathogen populations decreased to levels below the culture enrichment detection limit after approximately 300 and 171, 280 and 235, and 286 and 178 days of storage in cheese samples inoculated with the “high” and “medium” inocula of strains Scott A, CA and IS951, respectively. Analogous were the results of the “low” inoculum trials. In these trials the populations of L. monocytogenes declined with storage time and the last enrichment-positive results were recorded at post-inoculation days 115, 126 and 112 for cheese samples inoculated with strains Scott A, CA and IS951, respectively. The behaviour of L. monocytogenes in artificially contaminated processed cheese stored at 12
C is shown in Fig. 2B, E and H. Storage of cheeses at 12
C was selected to represent a scenario of severe temperature abuse, such as the possibility of storage of contaminated cheese at a malfunctioning consumer refrigerator (Azevedo et al., 2005; Xanthiakos et al., 2005). However, despite of the elevated storage temperature, the results were analogous to those obtained in the 4
C storage trials, i.e. no growth of L. monocytogenes was observed in any of the trials, regardless of strain and inoculation level. In all 12
C-trials, however, the estimated duration of the apparent lag phase was substantially shorter than the respective duration at 4
C. This reduction in lag phase was particularly pronounced for the Scott A and CA trials where the duration was shortened by approximately half and two-thirds,

Fig. 1. Agarose gel electrophoresis of DNA fragments generated by multiplex PCR. Lanes 1-10, GenScript Corporation 100-bp DNA ladder (from top to bottom 1500 bp, 1000 bp, 800 bp, 600 bp, 500 bp, 400 bp, and 300 bp), L. monocytogenes CA, L. monocytogenes Scott A, L. monocytogenes IS951, negative control (H20), L. monocytogenes reference strain CIP 104794 (serovars 1/2a), L. monocytogenes reference strain CIP 105449 (serovars 1/2b), L. monocytogenes reference strain CIP 105448 (serovars 1/2c), L. monocytogenes reference strain CIP 78.38 (serovars 4b), 100-bp DNA ladder. CIP, Collection of the Institute Pasteur.

814

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

A

B6

6 Low

Low

Medium

5

Medium

5

High

High

4

log(CFU/g)...

log(CFU/g)...

4

3

3

2

2

1

1

0

0 0

50

100

150

200

250

300

350

0

50

100 Time (d)

Time (d)

C

D

6

150

7 Low

Low

5

200

6

Medium

Medium High

High

5

log(CFU/g)...

log(CFU/g)...

4

3

4

3

2 2

1

1

0

0 0

20

40

0

60

50

100

E

150

200

250

300

F

7

400

7

Low

6

Low

6

Medium

Medium

High

High

5

log(CFU/g)...

5

log(CFU/g)...

350

Time (d)

Time (d)

4

3

4

3

2

2

1

1

0

0 0

50

100

150

200

Time (d)

0

20

40

60

80

Time (d)

Fig. 2. Fate of three strains of L. monocytogenes in experimentally contaminated processed cheese during storage at 4
C, 12
C or 22
C. The limit of detection of the enumeration method was 5 CFU/g (0.7 logCFU/g). Enrichment-positive results for samples that had been inoculated with the “High”, “Medium” and “Low” inocula have been arbitrarily assigned the values of log 4 (0.6), log 3 (0.48) and log 2 (0.30), respectively. All enrichmentenegative results are arbitrarily assigned the value of log 1 (0). Scott A e 4
C (A), Scott A e 12
C (B), Scott A e 22
C (C), CA e 4
C (D), CA e 12
C (E), CA e 22
C (F), IS951 e 4
C (G), IS951 e 12
C (H), IS951 e 22
C (I).

respectively (Fig. 3). In contrast to the reductions observed in lag phase duration between the 4- and 12
C-trials, when the differences in the rates of inactivation that were observed between the 4- and 12
C-trials are concerned, only in the IS951 12
C-trials was there an increase in the rate of inactivation (m) of L. monocytogenes compared to the corresponding trials conducted at 4
C (Fig. 4). Overall, the storage of contaminated cheese at 12
C was

substantially more detrimental to the survival of L. monocytogenes compared to its survival during storage at 4
C. A comparison of Fig. 2A with B, D with E and G with H shows that, irrespective of strain and inoculum level, the last enrichment-positive samples in trials conducted at 12
C were recorded considerably earlier compared to the last enrichment-positive samples recorded in the respective 4
C-trials.

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

G

H

6

815

6

Low

5

Low

5

Medium

Medium

High

High

4

log(CFU/g)...

log(CFU/g)...

4

3

3

2

2

1

1

0

0 0

50

100

150

200

250

300

350

0

400

50

100

I

150

200

Time (d)

Time (d) 6 Low

5

Medium High

log(CFU/g)...

4

3

2

1

0 0

20

40

60

Time (d)

Fig. 2. (continued).

The behaviour of L. monocytogenes in contaminated processed cheese stored at 22
C is shown in Fig. 2C, F and I. Storage of cheeses at 22
C was chosen in order to test the ability of the processed cheese to support L. monocytogenes growth under conditions of extreme temperature abuse (e.g. during room storage or in

situations where refrigeration is not available). The 22
C-trials also serve as tests for assessing the antimicrobial properties of the processed cheese product itself. Similar to the trials conducted at the other two storage temperatures no growth of L. monocytogenes was observed in any of the 22
C-trials, regardless of strain and inoculum

120

Lag phase, λ (d)

100 80 60 40 20

4

12 Sc ottA

22

4

12 CA

22

4

12

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

High

Medium

0

22

IS951

L. m onocytogenes strain, inoculum le v e l and storage te mpe rature (oC) Fig. 3. Estimated values for the apparent lag phase, l (in days), which indicates the time from product inoculation to the onset of inactivation of three strains of Listeria monocytogenes in processed cheese inoculated with two different inocula (“High”, “Medium”) and stored at 4, 12 or 22
C.

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

0 -0.05 -0.1 -0.15 -0.2 -0.25

4

12

22

4

Sc ottA

12

22

4

CA

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

High

Medium

-0.3 High

-1

Inactivation constant, μ (d )

816

12

22

IS951 o

L. m onocytogenes strain, inoculum le v e l and storage te mpe rature ( C) Fig. 4. Average calculated values for the inactivation (death) rate m (reductions in the viable populations of L. monocytogenes in log CFU/g per day of storage) of three strains of Listeria monocytogenes in processed cheese inoculated with two different inocula (“High”, “Medium”) and stored at 4, 12 or 22
C obtained via regression analysis assuming linear inactivation kinetics. Error bars indicate the 95% Confidence Intervals.

level. In fact, in all of the 22
C-trials the L. monocytogenes populations decreased to below the limit of enumeration (0.7 log CFU/g) after less than 2 months of storage. The last enrichment-positive result was recorded from a sample contaminated with the “high” inoculum of strain CA 53 days post-inoculation. 3.4. Description of L. monocytogenes inactivation The values of the inactivation rate m, which were calculated from the linear portion of the log CFU/g vs. time survival curves along with the corresponding 95% C.I.s are shown in Fig. 4. The values of l that were estimated as described in the Materials and Methods are presented in Fig. 3. The higher the storage temperature, the shorter was the duration of the observed lag phase, hence the faster was the onset of post-inoculation inactivation of L. monocytogenes. Also, the rates of inactivation of L. monocytogenes in the high storage temperature trials (22
C) were noticeably faster than those observed during storage of the contaminated cheese at the other two temperatures. The wide uncertainly around the point estimate of m observed in some of the “medium”-inoculum trials (e.g. in the ScottA-4
C and IS951-22
C trials, Fig. 4) may be attributed to the limited number of data points that were within a reasonably linear range and which were therefore used for estimating m. The second approach for describing the inactivation kinetics of L. monocytogenes relied on the utilization of all the quantitative data points from each semilogarithmic survival curve. This approach was based on the visual observation that the assumption of first-order kinetics may not necessarily apply for the entirety of the inactivation regions in the survival curves. As shown in Fig. 2, the semilogarithmic survival curves that were obtained from the different experimental trials are characterised by an initial “shoulder” of variable duration followed by a decline (inactivation region), which, in most cases, has a noticeable downward concavity. Under this approach all the available quantitative data obtained from each trial were fitted into a non-linear, Weibull-type inactivation equation (Peleg and Cole, 1998; van Boekel, 2002). The results showed that, for the majority of the experimental trials, the application of the equation to the data resulted in highly satisfactory fits as evidenced by the fact that in 14 out of the 18 trials the resulting R-square values were greater than 0.95 (Table 2). Also, the fitting of equation (1) to the data yielded n values which were substantially different (higher) than 1 (Table 2). Fig. 5A

through C presents typical examples of such applications where the appropriateness of equation (1) is illustrated. The applicability of the use of Weibull-type models to describe bacterial inactivation as a result of different imposed external treatments (e.g. heating, high pressure) has been previously documented and reviewed (Peleg and Cole, 1998; van Boekel, 2002). The current work showed that such an approach can also be applied to adequately describe the inactivation of a food-borne pathogen during the storage of a contaminated food. In this case, the antimicrobials present in the formulation of the cheese product itself create a “hostile environment” in which the L. monocytogenes cells are exposed to. In line with previous interpretations for survival curves displaying downward concavity (van Boekel, 2002), the downward concavity observed in the semilogarithmic survival curves obtained from the trials conducted in this work may be an indication that the remaining populations of L. monocytogenes in the contaminated processed cheese become increasingly damaged as storage time progresses. Table 2 Kinetic parameters of inactivation of three different strains of Listeria monocytogenes in processed cheese inoculated with two different levels and stored at three different temperatures. The parameters k and n were estimated by fitting equation log(N (t)) ¼ log(N0)  ktn to the survival data. R2

0.6314 0.0022 0.0405 0.0220 6.9891 0.2178

1.63 2.84 2.53 2.65 1.78 2.96

0.8958 0.8981 0.9903 0.9801 0.9787 0.9966

High Medium High Medium High Medium

1.7438 0.6697 0.3930 0.0129 4.2495 0.2160

1.46 1.64 2.14 2.96 1.89 2.97

0.9286 0.8503 0.9685 0.9893 0.9609 0.9937

High Medium High Medium High Medium

0.0483 0.0122 0.0103 0.0261 24.0763 0.0376

2.25 2.50 2.90 2.65 1.43 3.68

0.9703 0.9679 0.9702 0.9681 0.9681 0.9949

Storage temperature (
C)

Inoculum level

Scott A

4

High Medium High Medium High Medium

12 22 CA

4 12 22

IS951

4 12 22

k (103)

n

Strain

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

A

4

IS951/"Medium"/4 deg. C - Observed data (average) Model [log(N) = 3.69 - 0.000012t^2.5] R-sq. = 0.9679

log(N)

3

2

1

0 0

B

50

100 Time (d)

5

200

CA/"Medium"/ 12 deg. C - Observed data (average)

4 log(N)

150

Model [log(N) = 4.09 - 0.000013t^2.96] R-sq. = 0.9893

3 2 1 0 0

20

40

60

80

100

Time (d)

C

4

Scott A/"Medium"/ 22 deg. C Observed data (average)

log(N)

3

Model [log(N) = 3.75 - 0.00022t^2.96] R-sq. = 0.9966

2

1

0 0

10

20

30

40

50

Time (d) Fig. 5. Examples of typical semilogarithmic survival curves of L. monocytogenes in processed cheese fitted with equation log (N(t)) ¼ log(N0)  ktn as a model. Noticeable in the curves is the downward concavity indicating deviations from first-order inactivation kinetics.

4. Discussion Although an older Canadian study revealed that a small percentage of commercially available processed cheeses may contain substantial numbers of aerobic and anaerobic sporeformers (Warburton et al., 1986), pasteurised processed cheeses maintain a good safety record in the United States with very few foodborne disease outbreaks having been attributed to contaminated processed cheese products. To date, the few reported outbreaks of foodborne illness associated with processed cheese products involved only Clostridium botulinum intoxications (Glass and Doyle, 2005). Consequently, the vast majority of research regarding the microbiological safety of processed cheeses has focused on the choice of appropriate formulations that are inhibitory to the outgrowth of C. botulinum spores and botulinal toxin production (Kautter et al., 1979; Tanaka et al., 1986; Eckner et al., 1994; Glass and Johnson, 2004). Research on the behaviour of L. monocytogenes in contaminated processed cheeses has been scarce. Essentially, other than the work of Glass et al. (1998) and Genigeorgis et al. (1991), no other study has

817

looked into the behaviour of this foodborne pathogen in processed cheeses. Glass et al. (1998) inoculated commercial processed cheese slices (aw ¼ 0.92e0.93; pH ¼ 5.61e5.84) with a three-strain L. monocytogenes cocktail (103 CFU/g) and monitored the populations of the pathogen over a 96-h storage period at 30
C. The authors reported a minor decline of approximately 0.6 log CFU/g in the viable populations of L. monocytogenes. Genigeorgis et al. (1991) tested the ability of 49 different market cheeses to support the aerobic growth of a five-strain L. monocytogenes cocktail as a post-processing contaminant (3.95 to 4.36 log CFU/g) under three storage temperatures (4, 8 and 30
C). American processed cheese (pH ¼ 5.7 þ sorbic and citric acid) was amongst the cheeses that did not support the growth of L. monocytogenes. The authors reported a maximum death of 2.09 log CFU/g after 7 days of storage at 30
C, 0.9 log CFU/g after 36 days at 8
C and 0.18 log CFU/g after 36 days of storage at 4
C. Monterey Jack processed cheese (pH ¼ 5.7 þ sorbic and citric acid) was also not amenable to L. monocytogenes growth displaying maximum reductions in the pathogen populations of more than 2.36 log CFU/g after 9 days of storage at 30
C, 2.36 log CFU/g after 36 days of storage at 8
C and 1.84 log CFU/g after 36 days of storage at 4
C. Similar pathogen population reduction rates were reported for Piedmont processed cheese (pH ¼ 6.4 þ sorbic and citric acid). According to Regulation 1441/2007, challenge testing is considered as one of the tools available for determining whether a RTE food is able or unable to support the growth of L. monocytogenes and, if it does, whether or not the 100 CFU/g limit throughout the shelf-life will be met or not. Because certain steps (e.g. grating and final packaging) during the manufacture of the grated processed cheese examined in this work may theoretically allow for contamination of the product with L. monocytogenes, this study evaluated the behaviour of the pathogen in artificially contaminated product during storage at three different temperatures. Three different inoculum levels of three L. monocytogenes strains of different serovars and origin were used. L. monocytogenes growth was not observed in any of the trials conducted irrespective of strain, initial contamination level or storage temperature. In general, the lower the storage temperature of the cheese, the more prolonged the survival of L. monocytogenes was, i.e. storage of the inoculated cheese at 12
C and even more so at 22
C was more detrimental to the survival of L. monocytogenes than storage at 4
C. The strong anti-listerial properties displayed by the cheese product tested herein obviously originate from the antimicrobial ingredients included in its formulation. In fact, the observed inactivation of L. monocytogenes may be attributed to the combined action of sodium chloride, sorbic acid and phosphates along with the relatively low pH of the product. The higher death rates observed in the L. monocytogenes populations in contaminated samples stored at 22
C compared to the other storage temperatures tested may be explained by the higher bacterial metabolism at the higher storage temperature and hence the increased lethal effect of the preservatives. These observations are in agreement with previously published results regarding the survival of L. monocytogenes in salted whey and skim milk where the storage at 4
C resulted in prolonged pathogen survival compared to storage at 22
C (Papageorgiou and Marth, 1989b). The physicochemical characteristics of the processed cheese product tested in this study are sufficient to inhibit growth of the pathogen, but insufficient to readily eliminate the pathogen particularly under refrigerated storage. The survival of L. monocytogenes for prolonged periods of time in refrigerated contaminated processed cheese may pose a safety threat particularly due to the risk of direct or indirect contamination of other RTE foods that may be amenable to L. monocytogenes growth. Hence, although the product displayed a clear anti-listerial activity at all temperatures tested, every effort should be made to avoid post-pasteurization contamination.

818

A.S. Angelidis et al. / Food Microbiology 27 (2010) 809e818

The use of a Weibull-type model to describe the inactivation kinetics of L. monocytogenes under the different experimental trials tested in this work was a more comprehensive alternative to the classical linear inactivation approach, as it took into account the entire range of the available quantitative data and on the same time resulted in very good fits. Future research should therefore further investigate and explore the applicability of different inactivation models for describing the inactivation or behaviour of food-borne pathogens during the storage of contaminated products at different temperatures. 5. Conclusions Challenge tests were conducted according to previously published guidelines (Scott et al., 2005) to evaluate the ability of a processed cheese product to support the growth of L. monocytogenes. No growth of L. monocytogenes was observed in 27 different trials involving three different L. monocytogenes strains, three different levels of initial inoculum and three different storage temperatures. In fact, in all trials, the initial post-inoculation periods during which the bacterial populations remained relatively unchanged were followed by periods of pathogen inactivation which was substantially more rapid during storage at 22
C. In addition, and with respect to Regulation 1441/2007, in none of the low inoculum trials with an initial inoculum below 100 CFU/g did the pathogen exceed the 100 CFU/g limit. Hence, according to the experimental results presented here, and under reasonable foreseeable conditions of distribution and storage, this processed cheese product belongs to the category of RTE foods that do not support the growth of L. monocytogenes. However, prior to onset of inactivation, viable L. monocytogenes may persist in the contaminated product for prolonged periods of time, particularly when the product is stored at low temperatures. The semilogarithmic survival curves of L. monocytogenes obtained from the different experimental trials displayed noticeable downward concavity of varying degree indicating deviations from first order microbial inactivation kinetics. A Weibulltype equation was fitted to the survival data of the different trials resulting in very promising fits and yielding n values which were substantially different (higher) than 1. Acknowledgements This research was supported/funded by VIOTROS SA via the Aristotle University of Thessaloniki Research Committee Grant No 82413. References Angelidis, A.S., Smith, L.T., Smith, G.M., 1999. Compatible solute accumulation by Listeria monocytogenes grown on milk whey. FASEB J. 13, A1533eA1535. Suppl. S. Azevedo, I., Regalo, M., Mena, C., Almeida, G., Carneiro, L., Teixeira, P., Hogg, T., Gibbs, P.A., 2005. Incidence of Listeria spp. in domestic refrigerators in Portugal. Food Control 16, 121e124. Doumith, M., Buchrieser, C., Glaser, P., Jacquet, C., Martin, P., 2004. Differentiation of the major Listeria monocytogenes serovars by multiplex PCR. J. Clin. Microbiol. 42, 3819e3822.

Eckner, K.F., Dustman, W.A., Rys-Rodriguez, A.A., 1994. Contribution of composition, physicochemical characteristics and polyphosphates to the microbial safety of pasteurised cheese spreads. J. Food Prot. 57, 295e300. European Commission, 2005. Commission Regulation (EC) No. 2073/2005 of 15 November 2005 on microbiological criteria for foodstuffs. Off. J. Eur. Union. L338, 1e26. European Commission, 2007. Commission Regulation (EC) No 1441/2007 of 5 December 2007 amending Regulation (EC) No 2073/2005 on microbiological criteria for foodstuffs. Off. J. Eur. Union L332, 12e29. Food and Drug Administration, 2008. 21 CFR, Part 133.169 to 133.180. Food and Drug Administration, Washington, DC. Genigeorgis, C., Carniciu, M., Dutulescu, D., Farver, T.B., 1991. Growth and survival of Listeria monocytogenes in market cheeses stored at 4 to 30
C. J. Food Prot. 54, 662e668. Glass, K.A., Kaufman, K.M., Johnson, E.A., 1998. Survival of bacterial pathogens in pasteurised process cheese slices stored at 30
C. J. Food Prot. 61, 290e294. Glass, K.A., Johnson, E.A., 2004. Factors that contribute to the botulinal safety of reduced-fat and fat-free process cheese products. J. Food Prot. 67, 1687e1693. Glass, K., Doyle, M.E., 2005. Safety of Processed Cheese. A Review of the Scientific Literature. Food Research Institute, Madison, WI. Goulet, V., Hedberg, C., Le Monnier, A., de Valk, H., 2008. Increasing incidence of listeriosis in France and other European countries. Emerg. Infect. Dis. 14, 734e740. Hado, B.H., Yousef, A.E., 2007. Characteristics of Listeria monocytogenes important to food processors. In: Ryser, E.T., Marth, E.H. (Eds.), Listeria, Listeriosis and Food Safety. CRC Press, Boca Raton, FL, pp. 157e213. International Organization for Standardization (ISO), 1996. ISO 11290-1:1996/Amd 1-2004. Microbiology of Food and Animal Feedingstuffs e Horizontal Method for the Detection and Enumeration of Listeria monocytogenes e Part 1: Detection Method. ISO, Geneva, Switzerland. International Organization for Standardization (ISO), 1998. ISO 11290-2:1998/Amd 1 2004. Microbiology of Food and Animal Feedingstuffs e Horizontal Method for the Detection and Enumeration of Listeria monocytogenes e Part 2: Enumeration Method. ISO, Geneva, Switzerland. Kapoor, R., Metzger, L.E., 2008. Process cheese: scientific and technological aspects e a review. Compr. Rev. Food Sci. Food Saf. 7, 194e214. Kautter, D.A., Lilly, T.J.R., Lynt, R.K., Solomon, H.M., 1979. Toxin production by Clostridium botulinum in shelf-stable pasteurised process cheese spreads. J. Food Prot. 42, 784e786. Kornacki, J.L., Gurtler, J.B., 2007. Incidence and control of Listeria in food processing facilities. In: Ryser, E.T., Marth, E.H. (Eds.), Listeria, Listeriosis and Food Safety. CRC Press, Boca Raton, FL, pp. 681e766. Papageorgiou, D.K., Marth, E.H., 1989a. Fate of Listeria monocytogenes during the manufacture, ripening and storage of Feta cheese. J. Food Prot. 52, 82e87. Papageorgiou, D.K., Marth, E.H., 1989b. Behaviour of Listeria monocytogenes at 4 and 22
C in whey and skim milk containing 6 or 12% sodium chloride. J. Food Prot. 52, 625e630. Peleg, M., Cole, M.B., 1998. Reinterpretation of microbial survival curves. Crit. Rev. Food Sci. 38, 353e380. Pritchard, T.J., Beliveau, C.M., Flandres, K.J., Donnelly, C.W., 1994. Increased incidence of Listeria species in dairy processing plants having adjacent farm facilities. J. Food Prot. 57, 770e775. Ryser, E.T., Marth, E.H., 2007. In: Listeria, Listeriosis and Food Safety. CRC Press, Boca Raton, FL. Scott, V.N., Swanson, K.M.J., Freier, T.A., Pruett, W.P.J.R., Sveum, W.H., Hall, P.A., Smoot, L.A., Brown, D.G., 2005. Guidelines for conducting Listeria monocytogenes challenge testing of foods. Food Prot. Trends 25, 818e825. Tanaka, N., Traisman, E., Plantinga, P., Finn, L., Flom, W., Meske, L., Guggisberg, J., 1986. Evaluation of factors involved in antibotulinal properties of pasteurised process cheese spreads. J. Food Prot. 49, 526e531. van Boekel, M.A.J.S., 2002. On the use of the Weibull model to describe thermal inactivation of microbial vegetative cells. Int. J. Food Microbiol. 74, 139e159. Warburton, D.W., Peterkin, P.I., Weiss, K.F., 1986. A survey of the microbiological quality of processed cheese products. J. Food Prot. 49, 229e230. 232. Wehr, H.M., Frank, J.F., 2004. Standard Methods for the Examination of Dairy Products, 17th ed. American Public Health Association, Washington, D.C. Xanthiakos, K., Angelidis, A.S., Koutsoumanis, K., 2005. Dynamic modelling of Listeria monocytogenes growth in pasteurised milk during storage in domestic refrigerators. In: Intrafood 2005, Innovations in Traditional Foods, 2005 EFFoST Annual Meeting (25e28 October 2005, Valencia, Spain). Proceedings, vol. I, pp. 81e84.