An updated review of Listeria monocytogenes in the pork meat industry and its products

Authors


D. Thévenot, Unité de Microbiologie Alimentaire et Prévisionnelle, Ecole Nationale Vétérinaire de Lyon, BP 83, 69280 Marcy l’étoile, France. E-mail: d.thevenot@vet-lyon.fr

Abstract

Pork meat and processed pork products have been the sources of outbreaks of listeriosis in France and in other European countries during the last decade. The aim of this review is to understand how contamination, survival and growth of Listeria monocytogenes can occur in pork meat products. This study discusses the presence of L. monocytogenes in raw pork meat, in the processing environment and in finished products. The prevalence of L. monocytogenes generally increases from the farm to the manufacturing plants and this mainly due to cross-contamination. In many cases, this pathogen is present in raw pork meat at low or moderate levels, but foods involved in listeriosis outbreaks are those in which the organism has multiplied to reach levels significantly higher than 1000 CFU g−1. In such cases, L. monocytogenes has been able to survive and/or to grow despite the hurdles encountered during the manufacturing and conservation processes. Accordingly, attention must be paid to the design of food-processing equipment and to the effectiveness of the cleaning and disinfecting procedures in factories. Finally, the production of safe pork meat products is based on the implementation of general preventive measures such as Good Hygiene Practices, Good Manufacturing and the Hazard Analysis Critical Control Point.

Introduction

Listeria monocytogenes is a Gram-positive ubiquitous bacterium which is widely distributed in the environment. It has been isolated from a variety of sources including soil, vegetation, silage, faecal material, sewage and water (AFSSA 2000). This facultative intracellular bacterium can cause listeriosis, a severe invasive illness in humans, which may result in death. The risk of contracting listeriosis is high in immuno-compromised persons, the elderly, pregnant women and neonates. Thirteen serotypes of L. monocytogenes have been identified, but only three serotypes (1/2a, 1/2b and 4b) are associated with the majority of sporadic cases of listeriosis; serotype 4b is linked to almost all recent outbreaks (Rocourt and Bille 1997).

The transmission of this pathogen by contaminated food was first conclusively demonstrated by epidemiological and laboratory investigations in 1983 (Schlech et al. 1983). Listeria monocytogenes has been recovered from many different foods, and conversely, a variety of different food items such as raw and processed meats, soft cheese, raw milk, hot dogs, seafood and fresh vegetables have been linked to both sporadic cases and outbreaks of listeriosis (AFSSA 2000; FICT 2002). Pork meat and processed pork products, such as deli meats, have been implicated in listeria outbreaks in France (Jacquet et al. 1995; Goulet et al. 1998) and in other European countries (Jay 1996; Loncarevic et al. 1997) during the past decade. Listeria monocytogenes is of particular concern in raw, undercooked or ready-to-eat foodstuffs (AFSSA 2000). Because the organism is ubiquitous, food-processing industries are easily contaminated through raw food stuffs (AFSSA 2000). Listeria monocytogenes may then subsist in these industries because it grows at low temperatures, adheres to various food contact surfaces and certain strains have adapted to disinfectants (Salvat et al. 1995). In fact, Nesbakken et al. (1996) found L. monocytogenes at every stage of the fresh pork meat industry, with increasing prevalence from the slaughterhouse to the cutting room.

The human population responses to exposures to a foodborne pathogen are highly variable. Disease incidence is dependent on a variety of factors, including virulence of the pathogen, dose (the number of pathogens ingested), the general health and immune status of the host and the attributes of the food matrix that alter microbial or host status (Risk Assessment Drafting Group 2004). However, the incidence of L. monocytogenes in meat products is generally low, even if the pathogen is present at low or moderate levels (Encinas et al. 1999; AFSSA 2000; FICT 2002). Even if a single bacterial cell has the potential to cause disease, epidemiological data indicate that foods involved in listeriosis outbreaks are those in which the organism has multiplied and in general have reached levels significantly >1000 CFU g−1 (Ross et al. 2002; Risk Assessment Drafting Group 2004).

Criteria or recommendations for tolerable levels of L. monocytogenes in processed foods have been established in some countries. For example, the USA practices ‘zero tolerance’ (no organisms found in 25 g of a food product, Shank et al. 1996) while Canada and France apply different norms according to the foodstuff. The probability of contracting listeriosis is thought to be very low when food contamination was below 100 CFU g−1. Therefore, a level of not more than 100 CFU g−1 is tolerated in certain foodstuffs, while zero tolerance is applied to foods which support L. monocytogenes growth and have extended shelf-lives (AFSSA 2000). There is a great deal of uncertainty concerning L. monocytogenes counts because the actual level in serving of food may vary considerably from one portion to another (Risk Assessment Drafting Group 2004). Several countries have therefore concluded that the complete absence of L. monocytogenes in certain foods is an unrealistic and unattainable requirement which does not have a significant positive impact on public health and may even divert resources and attention from other preventive measures. Despite these recommendations and drastic economic losses to industries due to recall products contaminated with L. monocytogenes, several processed food-related outbreaks of listeriosis have occurred during the last decade (Goulet et al. 1998).

The aim of this review is first, to update knowledge about the origin of L. monocytogenes contamination in pork meat, products and processing environment; and secondly, to understand how the organism is able to develop or survive the manufacturing processes, and then cause sporadic cases or outbreaks of listeriosis. This review details and discusses our current understanding of these issues. Special attention is drawn to the different serotypes and clonal lineages that are found in foods and are implicated in foodborne listeriosis.

Contamination of raw pork meat, meat processing environments and pork meat products

Raw pork meat

Listeria monocytogenes occurs frequently in raw pork meat (Norrung et al. 1999), although the origin of the contamination is unclear. Listeria monocytogenes has been occasionally isolated on farms from the faeces and skin of presumably healthy pigs (Skovgaard and Norrung 1989). The organism is thought to be harboured in the intestinal tract: the prevalence of L. monocytogenes in pig faecal samples ranges from 0% to 47%; the highest prevalences are reported in Eastern Europe (Felon et al. 1996). Husbandry practices that involve feeding pigs dry feed or silage, rearing pigs in closed houses, and maintaining specific pathogen-free herds may account for some of the reported variation in the incidence of L. monocytogenes in healthy pigs (Felon et al. 1996).

Carcasses are thought to be contaminated when the large intestine is ruptured during evisceration (Skovgaard and Norrung 1989). However, Kanuganti et al. (2002), detected L. monocytogenes in only 4% of pork carcasses sampled but not in the rectal contents of animals prior to slaughter. This may be due to major slaughter plants practice of evacuating the rectum prior to evisceration, thus minimizing rectal contents and thus subsequent carcass contamination. Authors have suggested not all L. monocytogenes detected in carcasses have a faecal origin. Bunčićet al. (1991) showed that pigs were more likely to harbour L. monocytogenes in their tonsils than to excrete the bacteria in their faeces.

Autio et al. (2000) also found that 14% of pig tongues and 12% of tonsils sampled in slaughterhouses contained L. monocytogenes. Interestingly, Kanuganti et al. (2002) showed that L. monocytogenes was detected more often in pig tonsil homogenates (7·1% of 252 samples) than in tonsil scrapings collected on the farm (3% of samples). Reported prevalence of L. monocytogenes in tonsils range from 0% to 61%; this range is probably due to differences in sampling techniques and/or farm management methods (Felon et al. 1996).

Autio et al. (2000) found that the occurrence of L. monocytogenes differed among slaughterhouses. Furthermore, the level of contamination of viscera (tongue, oesophagus, trachea, lungs, heart, diaphragm, kidneys and liver); was particularly high (64%). They hypothesized that L. monocytogenes spread through contact between the tonsils and tongue and the other viscera and carcass during the evisceration process. Moreover, Kanuganti et al. (2002) detected L. monocytogenes less often in the tissues of freshly slaughtered pigs (0·8–2·4% of samples) or of small intestines (8·3–9·3% of samples) than in ground pork (45–50·2% of samples). Furthermore, an extensive study of meat contamination levels in the meat processing industry indicated that chilling and cutting significantly increased the contamination of pork meat (Nesbakken et al. 1996) while Van der Elzen and SNIJders (1993) found that the environmental prevalence of the pathogen in chilling–cutting areas to be as high as 71–100%. These findings strongly suggest that postslaughter processing is a significant cause of meat contamination, and that contamination is amplified in the chilling and cutting room environment (Nesbakken et al. 1996). Pulsed field gel electrophoresis (PFGE) studies of Listeria strains collected both from raw meat and work surfaces during meat processing have demonstrated conclusively that raw meat can contaminate meat processing environments. Reciprocally, contaminated equipment can, in turn, contaminate meat products during processing (Giovannacci et al. 1999, Thevenot et al. 2004).

Meat is mainly contaminated with L. monocytogenes serotypes 1/2a, 1/2b and 1/2c (Hof and Rocourt 1992). We also found that this is true for raw pork meat (Thevenot et al. 2005b). Most listeriose cases are sporadic and frequently involve serotypes 1/2a and 1/2b (Schuchat et al. 1991). The presence and predominance of serotype 1/2a and to a lesser extent 1/2b in meat processing plants might be a source of sporadic cases. We found serotype 4b strains in raw pork meat entering French deli meat factories (Thevenot et al. 2005b), although Giovannacci et al. (1999) did not found this strain in pork slaughtering and cutting plants. Nevertheless, this is of major public health importance as serotype 4b is responsible of the majority of listeria outbreaks in France (Goulet et al. 1998).

Pork meat processing environments

In general, the primary source of food contamination by L. monocytogenes before release to consumers appears to be the processing environment (Kathariou 2002).

Several studies have shown that L. monocytogenes strains isolated from meat processing environments are frequently of serotypes 1/2a, 1/2b and 1/2c (Jay 1996; Chasseignaux et al. 2002; Thevenot et al. 2005b). We also isolated serotypes 4b and 4e in French meat curing factories (Thevenot et al. 2005b).

Different molecular techniques (amplified fragment length polymorphism, PFGE, PCR and ribotyping) have shown that the genotypes of L. monocytogenes strains collected from pork meat processing environments are extremely diverse (Chasseignaux et al. 2001; Lundén et al. 2003a,b;Martinez et al. 2003; Thevenot et al. 2006). High genotype diversity among strains collected both from surfaces during processing and processed products may indicate either a continuous source of inoculation by raw material or the persistence of several strains in spite of the cleaning and disinfection operations. Indeed, both Lundén et al. (2003a, 2003b) and Thevenot et al. (2006) found both transient (sporadic) and resident (persistent) L. monocytogenes strains in meat processing environments.

Listeria monocytogenes may become established in the processing environment and survive for long period of time. Giovannacci et al. (1999) and Lundén et al. (2002) pointed out the persistence of L. monocytogenes strains over a year for two pork processing plants and for 3 years in a grinder respectively.

Listeria monocytogenes adheres to the inert surfaces encountered in the food-processing environments, although there are differences in both the extent and adsorption rate according to the type of surface, pretreatment, environmental conditions and bacterial serotypes (AFSSA 2000). Biofilms develop as a result of both the adsorption and adherence of free-floating cells, and the continued growth of cells in the biofilm matrix. Listeria monocytogenes can incorporate into a biofilm, although Chae and Schraft (2000) showed that not all L. monocytogenes strains had the same ability to evolve into a mature biofilm. These authors noted that the multiplication rate of free-floating L. monocytogenes cells vs biofilm cell forms was different and suggested that growth behaviour differed according to the support media. Kalmokoff et al. (2001) found little differences in the adsorption rates of different strains but significant differences in their adherence and ability to form a biofilm. The highly adherent strains produced fibrils which were generally absent in low-adherence strains (Vatanyoopaisatn et al. 2000). Norwood and Gilmour (2000) examined 111 L. monocytogenes strains and found that serotype 1/2c adhered significantly more to stainless steel than other serotypes over a 24-h period. Lundén et al. (2000) also noted differences in the short-term adsorption of resident and nonresident strains onto a stainless steel model surface, with strains within the 1/2c serotype demonstrating the highest degree of adsorption.

A study led by Autio et al. (2003) on L. monocytogenes strains recovered in 11 food-processing plants showed that certain genotypes were better able to cause resident (persistent) contamination of food facilities than other genotypes. These authors hypothesized that resident strains adapt to food-processing facilities via natural selection. However, Lundén et al. (2003a, 2003b) observed several L. monocytogenes strains with similar PFGE types which were classified as persistent in one factory but not another, emphasizing the complex nature of persistent and nonpersistent contamination. Lundén et al. (2003a, 2003b) did suggest that with additional samples been collected, some of the nonpersistent L. monocytogenes strains might have been recovered more frequently, and thus have been categorized as persistent.

Processed meat products

Processed meat products may be contaminated by L. monocytogenes at several stages: either the raw ingredients are contaminated and the manufacturing process is insufficient to sterilize the product or by contact with contaminated unprocessed raw materials, unclean surfaces or people (Samelis et al. 1998; Chasseignaux et al. 2001). The latter routes are well recognized (Reij and De Aantrekker 2004). Poor personnel hygiene, including simple procedures such as hand washing, has been identified as a causative mode of transmission of the pathogens (AFSSA 2000). Cross-contamination may occur at any stage between the meat processing plant to the final consumer in the home (Reij and De Aantrekker 2004). Contamination in the home often occurs when the same cooking utensils such as cutting boards are used on raw contaminated foods and then sterile food (either cooked or ready-to-eat processed) (AFSSA 2000).

Till now, there has been sporadic information on the occurrence of L. monocytogenes in processed pork meat products. First, because contamination may occur after manufacturing, in the home kitchen (AFSSA 2000). Secondly, many meat products are speciality items which are region or country specific.

Nevertheless, studies have documented the prevalence of L. monocytogenes processed raw pork meat products. In Japan, Inoue et al. (2000) observed that 20·6% of ready-to-cook minced pork samples were contaminated with essentially 1/2a and 1/2b serotypes. In the USA, 22·9% of the ground pork meat products and sausages sampled by Duffy et al. (2001) were contaminated with the pathogen. Additionally, L. monocytogenes was found in 10·6% and 10% of dried raw sausages in Chile (Cordano and Rocourt 2001) and France (Thevenot et al. 2005b) respectively.

A survey of cooked pork meat products by the French hygiene control services (DGCCRF 1996) between 1993 and 1996 showed that 16% of tongue, 9% of ‘pâté’, 11·7% of ‘rillettes’, 6·02% of salami and 13·1% of hams sampled were contaminated by L. monocytogenes; 1%, 1% 1·7%, 0·2% and 1·1% of the contaminated products, respectively, had more than 10 CFU g−1 of L. monocytogenes.

As with raw pork meat, serotypes 1/2a, 1/2c and 1/2b strains are most common in processed pork products (Jay 1996), while serotype 4b is rare (Greenwood et al. 1991; Hayes et al. 1991). However, in our recent study (Thevenot et al. 2006), we isolated several 4b strains in two different French pork meat curing factories that had the same PFGE types as the serotype 4b reference strain (CIP 78·38). An additional 45 strains (collected mostly from one factory) shared nearly identical PFGE profiles with a strain of L. monocytogenes isolated from a human clinical case of listeriosis (CIP 105550). The similitude between profiles suggests that these strains are closely related. Martinez et al. (2003) also found the same PFGE types between strains isolated from a human listeriosis case and a chicken processing company. The presence of 4b strains in the meat processing industry may result in sporadic cases or an outbreak of listeriosis, particularly if foodstuffs are consumed by persons in the at-risk population.

Meat processing environment and meat adapted clones

The presence of strains sharing identical pulsotypes in raw meat, meat processing environment and products is evidence that certain clones have adapted to meat processing plants and processed products (Thevenot et al. 2006). Chasseignaux et al. (2001) observed the same profiles in several plants which were from different geographical location. Autio et al. (2002) also observed that certain strains were recovered repeatedly in the finished pork-meat products of several unrelated factories, indicating that strains are not always plant specific. Certain strains may be more widely distributed in nature, and thus be more easily (re)introduced into processing plants via raw material. What is more, some clones may be better adapted both to raw meat and to meat processing environments and finished products.

The war between L. monocytogenes and the pork meat processing industry

Listeria monocytogenes has been isolated from an extensive range of pork meat products (AFSSA 2000; FICT 2002). The growth and survival of a micro-organism is dependent on its ability to overcome environmental hurdles encountered during the manufacturing and conservation process, such as temperature, pH, water activity, metal ions and nutrient availability. Most bacteria, including L. monocytogenes, are generally resistant to small changes in a specific environmental parameter, but severe or multiple changes stimulate complex stress responses that are generally directed to survival instead of growth (Booth 1998). The various environmental obstacles that are utilized by the food-processing industry are discussed below.

Food-mediated hurdles to L. monocytogenes growth

Physical and chemical hurdles

Temperature.  Temperature is the most important hurdle encountered by L. monocytogenes. It grows in a temperature range of +0·4 to 45°C, while the optimum growth temperature lies between 30 and 37°C (AFSSA 2000). Listeria monocytogenes grows in refrigerated foods (Augustin 1999), and this is of major importance for risk assessment of foodstuffs: even when initial contamination is low, the organism can multiply during refrigeration and reach levels up to 100 CFU g−1 (AFSSA 2000). Besides, the temperature home refrigerators is often closer to 9°C than 4°C (Sergelidis et al. 1997), which also favours L. monocytogenes growth.

Fleming et al. (1985) showed that the thermotolerance of L. monocytogenes is higher than many other nonspore-forming foodborne pathogens. Its thermotolerance may increase after exposure to a variety of environmental stress conditions including heating at sublethal temperatures or osmotic and acidic shocks. Linton et al. (1992) observed that the effect of a heat shock on L. monocytogenes was highly dependent on temperature and treatment duration: higher temperatures and longer treatments increased resistance. Factors that are intrinsic to the underlying foodstuff also affect influence thermotolerance (AFSSA 2000). Pork meat products which are cooked at very high temperatures are quite likely free of live L. monocytogenes cells. However, products such as ‘rillettes’, which are cooked at fairly low temperatures (50–60°C) may not be L. monocytogenes free even after if cooking times are long (FICT 2002).

At present, there is little evidence that certain strains of L. monocytogenes are more thermotolerant than others (De Jesus and Whiting 2003).

pH. Listeria monocytogenes is reported to exhibit viability at pH levels between 4·6 and 9·6 with an optimum at 7·1 (AFSSA 2000).

Listeria monocytogenes may be subject to low pH during processing of fermented specialties such as cured dried sausages (saucisson and chorizo) (Foegeding et al. 1992; Thevenot et al. 2005a). It also survives highly acidic pH characteristic of human stomachs and macrophage phagosomes (Cotter and Hill 2003). This property can be considered as a virulence factor as many other bacteria are not able to survive such conditions. The glutamate decarboxylases (GAD) system, which modulates intracellular pH in certain Gram-positive cells (Cotter and Hill 2003), is thought to be the key mechanism by which L. monocytogenes maintains pH homeostasis in acid environments. Strains which display reduced GAD activity are more sensitive to gastric fluid, and furthermore, GAD activity is significantly correlated with acid tolerance in this organism (Cotter and Hill 2003). The GAD system plays a role in the acid tolerance during both logarithmic and stationary growth phases and is required for the induction of an optimal acid tolerance response (Cotter and Hill 2003). However, Vialette et al. (2003) tested the pH sensitivity of L. monocytogenes strains which were isolated from meat products and food-processing environments, and found that bacteria were usually more sensitive in the exponential growth phase than in the stationary phase.

The usual pH of meats (5·0–5·5) can induce an acid tolerance response in L. monocytogenes, which in turn decreases the organisms sensitivity to even lower pH (Faleiro et al. 2003). Several authors have noted that acid adaptation can induce cross-protection against several other damaging factors including osmotic stress (O'Driscoll et al. 1996; Vasseur et al. 2001), and Faleiro et al. (2003) have indicated that such cross-stress protection is strain dependent. The molecular processes involved have not yet been identified. Finally, sublethal stresses other than acid, i.e. osmotic, heat and low temperatures, do not appear to affect the acid resistance of this bacterial species (Koutsoumanis et al. 2003).

NaCl. Listeria monocytogenes may also encounter high salt concentrations in dried sausages, raw dried ham or cooked meats (Foegeding et al. 1992; Dabin and Jussiaux 1994). This organism is able to grow at high salt concentrations (up to 10% NaCl), although salt tolerance varies with pH and temperature (AFSSA 2000). For example, L. monocytogenes survives in cheese brining systems (Faleiro et al. 2003). Strains of L. monocytogenes that have adapted to acid environment may become cross-protected against several damaging factors including osmotic stress (O'Driscoll et al. 1996; Vasseur et al. 1999). However, osmo-adaptation does not induce strong cross-protection against other stress factors (Lou and Yousef 1997).

Aw.  Water activity is strongly linked with the pH and salt content of the meat products. For example, during the drying stage of cured raw sausage products (e.g. saucisson and chorizo), the water holding capacity decreases, as does the pH, which approaches the isoelectric point (Tyopponen et al. 2003). The addition of salt to the sausage mix limits water activity, thereby inhibiting the growth of many spoilage and pathogenic bacteria including L. monocytogenes (Lücke 1985). After drying, the water activity is <0·90, which also inhibits bacterial growth (Tyopponen et al. 2003). Unfortunately, some manufacturers tend to reduce drying times in order to increase profitability. Insufficiently dried products may have water activity levels which permit the growth of L. monocytogenes (Aw = 0·92) (AFSSA 2000), and could be associated with a higher risk of listeriosis.

Other factors. Listeria monocytogenes can be subjected to other potential stress in meat products. Microbial growth in certain meat products may consume all the available oxygen in the matrix. The growth of aerobic spoilage bacteria will be affected, but not that of L. monocytogenes because it is an aerobic/anaerobic organism. Encinas et al. (1999) found significant differences in numbers of Listeria between chorizos formulated with hot and mild paprika. Other spices, such as pepper, cardamom and garlic may also have antioxidative and antimicrobial properties (Tyopponen et al. 2003). Finally, smoke, which contains phenols, carbonyls and different organic acids, may inhibit different bacteria on the surfaces of cured pork products. According to Encinas et al. (1999), the manufacturing process and smoking significantly reduced Listeria counts in sausages and ham.

Microflora

Natural microflora.  Three groups of different bacteria are present in raw pork meat. The first regroups ‘useful’ organisms such as Micrococcus, Staphylococcus (levels ranging from 104 to 105 CFU g−1), lactic acid bacteria (LAB) (from 103 to 104 CFU g−1); and yeast (from 103 to 104 CFU g−1) (Samelis et al. 1998). LAB produce lactic acid, resulting in low pH values (Leistner 1995). This can be beneficial in raw meat products. For example, certain raw sausage specialties are obtained with natural LAB contamination (Dabin and Jussiaux 1994). Manufacturers may add LAB starters and glucose (which is an energy source for LAB) to the meat matrix enhance and control the drop in pH (Ross et al. 2002). Micrococcus, Staphylococcus and yeast produce aromatic compounds which contribute to the flavour of cured products (Montel et al. 1996).

The second group is composed of spoilage bacteria, such as hetero-fermentative LAB, pseudomonads and enterobacteria (Samelis et al. 1998). Their presence in raw meats can lead to spoilage during the manufacturing process (Dabin and Jussiaux 1994). The last group is represented by pathogenic bacteria such as Campylobacter, Clostridium botulinum and Clostridium perfringens, Salmonella spp. and L. monocytogenes (Dabin and Jussiaux 1994).

Microbial stability of cooked, fermented or raw meat products occurs when spoilage and pathogenic bacteria are reduced/removed and when useful micro-organisms are utilized properly (Leistner 2000; Ross et al. 2002).

Starters.  Modern food processing employs a range of technologies to ensure that the bacteriological quality of food is maintained from the time of manufacture to the time of consumption (Ross et al. 2002). Food fermentation involves the oxidation of carbohydrates into various metabolites (principally organic acids, alcohols and carbon dioxide) which limit the growth of spoilage and/or pathogenic flora. A number of desirable compounds are also synthesized, which affect the flavour of foods (diacetyl and acetaldehyde) or their nutritive value (vitamins, antioxidants and bioactive peptides) (Ross et al. 2002; Tyopponen et al. 2003). Food fermentation has been practiced for millennia, and as a result there is a tremendous variety of fermented meat specialties worldwide (Ross et al. 2002).

LAB are primarily responsible for many of the microbial transformations found in the more common fermented meat products (Ross et al. 2002). LAB produce high amounts of lactic acid that in turn lower the pH value (5·0) of the meat matrix, effectively eliminating or inhibiting spoilage and pathogen contaminants (Tyopponen et al. 2003).

LAB may also inhibit the growth of L. monocytogenes: Encinas et al. (1999) isolated L. monocytogenes from chorizo manufactured without lactic starters, but not in chorizos with starters. Kang and Fung (2000) confirmed the inhibitory effect of LAB starters. They counted L. monocytogenes in salami manufactured with and without LAB starters; the level was 2·2 log10 higher in the latter. Furthermore, meat processing lines are continuously contaminated through incoming raw products, yet lines which utilize the fermentation process do not remain persistently contaminating (Lundén et al. 2003a,b). This suggests that competing flora may prevent L. monocytogenes from becoming established on processing surfaces. Finally, Mahoney and Henriksson (2003) showed that the ability of L. monocytogenes to cause foodborne listeriosis is dependent on the nature of the food matrix; a traditional meat starter culture inhibits growth of the pathogen during passage through the gastrointestinal tract.

Many bacteria of the LAB family produce protein inhibitors referred to collectively as bacteriocins (Ross et al. 2002). In general, bacteriocins either depolarize the target cell membrane or inhibit cell wall synthesis. They can be divided into three main groups: the lantibiotic family [small bacteriocins composed of one or two peptides of c. 3 kDa (Nisin; Sigma-Aldrich, St Quentin Fallavier, France)], small nonmodified peptides (<30 kDa) and large heat-labile proteins (>30 kDa). De Martinis and Franco (1998) found that Lactobacillus sake produces a bacteriocin-like substance that can be used to inhibit the growth of L. monocytogenes in a Brazilian sausage speciality. Foegeding et al. (1992) further showed that pediocin increased food safety during both the fermentation and drying portions of sausage manufacturing.

Cleaning and disinfecting products and procedures

Typical cleaning and disinfecting processes in food-processing environments involve a series of steps, each with a specific purpose (Tamplin 1980). A pre-rinse is performed using water to remove gross, loose material, followed by cleaning with suitable detergents (often alkaline) in order to remove residual substances. The detergent is then rinsed off, antimicrobial chemicals are applied in order to sanitize surfaces, and a final post-rinse with water removes these chemicals. The cleaning step is important because it removes organic matter from processing surfaces (Chasseignaux et al. 2001; Thevenot et al. 2005b). Besides harbouring bacterium, organic matter can reduce the efficiency of disinfectants [such as quaternary ammonium compounds (QACs)] or prevent disinfectants from reaching bacterium at the recommended concentrations (Gibson et al. 1999). Furthermore, Taormina and Beuchat (2002) showed that L. monocytogenes strains that were exposed to cleaning solutions that which did not affect cell viability were more sensitive to subsequent treatment with sanitizing chemicals.

Several studies have indicated that adherent micro-organisms may be much more resistant to sanitizing compounds than free-floating cells (Somers et al. 1994; Fatemi and Frank 1999). Furthermore, antimicrobial agents that are highly effective in liquid media may be less active in more complex foodstuffs (Norwood and Gilmour 2000). Multispecies biofilms, which combine the effects of bacterial shielding and the production of extracellular polymeric substances, also increase bacterial resistance to antimicrobial agents (Costerton et al. 1995). Finally, food and food-processing equipment, together with the highly buffered system typical of meats, offer micro-organisms a complex environment that suppresses the efficacy of antimicrobial agents (Korber et al. 2002).

During the last decade, numerous studies have focussed on sanitizing (disinfecting) products and their effects on L. monocytogenes. In a survey by Holah et al. (2002), QACs were markedly the most commonly used disinfectant in the food industry. This is primarily due to their biocide property (QACs cause generalized membrane damage and inactive cellular enzymes) combined with their excellent nontainting, nontoxic and noncorrosive nature (on human skin and hard surfaces), which makes them the ideal choice for manual application in the food industry. Still, a minimum inhibition concentration has increased the resistance of L. monocytogenes to QACs in strains isolated from a range of environmental, food, animal and clinical sources (Mereghetti et al. 2000). Mechanisms of such a resistance remain unclear. Lemaître et al. (1998) found extrachromosomal DNA in all resistant strains tested and concluded that this resistance was plasmid mediated. On the other hand, Mereghetti et al. (2000) observed resistant L. monocytogenes strains with and without plasmids. Romanova et al. (2002) confirmed the latter hypothesis: indeed all isolates tested in their study contained the mdrL gene, which encodes an efflux pump that confers resistance to QACs and is both chromosome- and plasmid borne. Finally, Holah et al. (2002) suggested that persistence was not due to resistance towards disinfectants but to physical adaptation mechanisms (surface attachment, biofilm formation, reduced growth rate and quiescence).

Several studies underlined the importance of utilizing proper procedures and concentrations during the sanitizing process. For example, the concentrations of sodium hypochlorite recommended by suppliers are more than sufficient to eliminate free-floating L. monocytogenes but are ineffective on bacteria in multispecies biofilms cultured on a stainless steel surface (Norwood and Gilmour 2000). There is also a significant difference in the effectiveness of sanitizing solutions on L. monocytogenes when applied to stainless steel or conveyor belt surfaces (Bremer et al. 2002). This study suggested that the decreased effectiveness of chlorine on the conveyor belt surface was due to limited penetration of the disinfectant: bacteria were relatively protected by the fabric weave. The authors concluded that during the sanitizing process, little consideration is given to the physical properties of the material being sanitized.

Authors have also tested different procedures to improve sanitizing processes. For example, Vasseur et al. (2001) tested combinations of chemicals that could decrease or eliminate L. monocytogenes surviving in food-processing plants. Chemicals that induced pH shocks were more efficient when used in combination with other parameters. An acid treatment (pH = 5·4) followed by an alkaline treatment (pH = 10·5) was not very effective against L. monocytogenes, whereas the opposite combination led to a three-log reduction of the bacterial population. Whatever the bacterial species, the most efficient treatments were combinations of alkaline, osmotic and biocide shocks (Vasseur et al. 2001).

Mereghetti et al. (2000) suggests that the rotation of disinfectants will prevent the development of resistant strains. However, Lundén et al. (2003a, 2003b) demonstrated that rotation of disinfecting agent with differing mechanisms of action in food-processing plants may not have the desired effects because of disinfectant cross-adaptation. Rotation of disinfectants may nevertheless lengthen the adaptation periods.

Finally, Holah et al. (2002) observed an increasing tendency for disinfectants to be left on surfaces and not rinsed off prior to recommencing production. Manufacturers claim that residual disinfectants will prevent subsequent surface microbial development, although low, sublethal concentrations may in fact enhance resistant.

In most food-processing environments and more precisely in pork meat industries, the daily use of sanitizers at correct concentrations preceded by detergent-aided cleaning, will remove all adhering organisms or biofilms (Norwood and Gilmour 2000).

Food processing lines

Food processing equipment plays an important part in sustaining contamination and in contaminating processed foodstuffs (Lundén et al. 2002; Reij and De Aantrekker 2004). Factories with complex processing lines are more susceptible to persistent L. monocytogenes contamination (Thevenot et al. 2005b), and furthermore, it is difficult to eradicate L. monocytogenes from contaminated processing lines and equipment (Miettinen et al. 1999; Autio et al. 2003).

Mechanical cleaning is the most effective way to detach adhered cells and biofilms (Gibson et al. 1999). However, poorly designed equipment will have small spaces and narrow openings that are difficult or impossible to reach, resulting in poor mechanical cleaning. It is from these locations that the main danger of cross-contamination of pathogens from mature biofilm exists (Miettinen et al. 1999; Autio et al. 2003). Even sanitizers may not reach the site of contamination at effective concentrations when applied according to the manufacturer's recommendations. It may well be necessary to use sanitizer at much higher concentrations in these particular areas of the food-processing environment.

Attention must be paid to a correct design of equipment and helpful recommendations and guidelines for the validation of equipment cleanability have been published (European Hygienic Equipment Design Group (EHEDG) 1997; Reij and De Aantrekker 2004). Finally, compartmentalization of the processing line and disassembly processing machines should help control L. monocytogenes contamination in meat and poultry processing plants (Lundén et al. 2003a,b).

In summary, cleaning and disinfecting procedures for food-processing machines combine mechanical energy and chemical products. When these procedures are adequately applied, both persistent and nonpersistent L. monocytogenes strains can be eliminated (Reij and De Aantrekker 2004).

Conclusion

Listeria monocytogenes poses a microbiological risk in processed foods containing pork meats. The bacterium is able to adapt to hurdles encountered during the manufacturing and conservation process. In many European countries, a concentration of up to 100 CFU g−1 of food is tolerated in processed foodstuffs. The production of safe foods containing pork meat is contingent on proper hygiene at all levels of the food chain.

Suppliers of raw pork products must abide by strict microbiological standards, thus limiting initial contamination of processing lines and food products with L. monocytogenes. Even so the contamination of processing lines is unavoidable at present. Therefore, special care must be taken to ensure that cleaning and disinfecting procedures are applied correctly. General preventive measures such as Good Hygiene Practices and Good Manufacturing Practices will help limit the risk of spread food-borne pathogens, including L. monocytogenes. Hazard Analysis Critical Control Point techniques should be applied to each food-processing plant in order to identify and control sources of L. monocytogenes contamination and dissemination. End product control, challenge testing and L. monocytogenes predictive growth models can all be used to ensure that the end product is safe. An alternative approach could be to group food products according to intrinsic and extrinsic parameters, which limit microbial growth. In the case of L. monocytogenes, the main factors are pH, Aw, composition of the food product, natural antimicrobial components, biological structures, temperature, atmosphere, and competitive flora. It is a scientific sound and interesting option to group foods based on the set of hurdles encountered during manufacturing process in order to establish their stability towards L. monocytogenes and determine their status: safe or with potential danger for persons at risk. Finally, the consumers must scrupulously respect the product's shelf life and the cold chain.

Acknowledgements

This work was supported by funds from Rhône-Alpes and Massif Central regions, from Office National Interprofessionnel des Viandes de l'Elevage et de l'Aviculture (OFIVAL) and Association Nationale de la Recherche Technique (ANRT) and from manufacturers.

Ancillary