|Home | About | Journals | Submit | Contact Us | Français|
The presence of pathogens in dairy products is often associated with contamination via bacteria attached to food-processing equipment, especially from areas where cleaning/sanitation is difficult. In this study, the attachment of Listeria monocytogenes on stainless steel (SS), followed by detachment and growth in foods, was evaluated under conditions simulating a dairy processing environment. Initially, SS coupons were immersed in milk, vanilla custard, and yogurt inoculated with the pathogen (107 CFU/ml or CFU/g) and incubated at two temperatures (5 and 20°C) for 7 days. By the end of incubation, cells were mechanically detached from coupons and used to inoculate freshly pasteurized milk which was subsequently stored at 5°C for 20 days. The suspended cells in all three products in which SS coupons were immersed were also used to inoculate freshly pasteurized milk (5°C for 20 days). When SS coupons were immersed in milk, shorter lag phases were obtained for detached than for planktonically grown cells, regardless of the preincubation temperature (5 or 20°C). The opposite was observed when custard incubated at 20°C was used to prepare the two types of inocula. However, in this case, a significant increase in growth rate was also evident when the inoculum was derived from detached cells. In another parallel study, while L. monocytogenes was not detectable on SS coupons after 7 days of incubation (at 5°C) in inoculated yogurt, marked detachment and growth were observed when these coupons were subsequently transferred and incubated at 5°C in fresh milk or/and custard. Overall, the results obtained extend our knowledge on the risk related to contamination of dairy products with detached L. monocytogenes cells.
Listeria monocytogenes is ubiquitous in nature due to its inherent ability to survive and grow under a wide range of adverse environmental conditions, such as refrigeration temperatures, high acidity and salinity, and reduced water activity (16). This microorganism is a major concern for the food industry, since it is the causal agent of listeriosis, a severe disease with high hospitalization and case-fatality rates (approximately 91% and 30%, respectively) (25). According to the European Centre for Disease Control and Prevention, listeriosis was the fifth most common zoonotic infection in Europe in 2006 (14), while it accounts for approximately 28% of the deaths resulting from food-borne illnesses in the United States (34).
In the food industry, inadequately cleaned food-processing equipment (e.g., stainless steel [SS] surfaces) constitutes a potential source for L. monocytogenes, resulting in contamination of foods which come in contact with such equipment (36). Even though adherence to strict sanitation practices should minimize the risk of survivors on surfaces, existing evidence suggests that a considerable risk may occur in sites of processing plants which are not easily cleaned or sanitized, such as those that do not allow direct access of sanitation equipment for abrasion (e.g., edges, convex surfaces, etc.) (43, 45). Attachment to surfaces is believed to be important for the survival and persistence of this pathogen in such environments, with some strains able to remain on equipment surfaces for several years (32, 37). Thus, L. monocytogenes has been shown to adhere to and form biofilms on various food contact surfaces under laboratory conditions (3, 42, 44). Furthermore, attached L. monocytogenes cells are more difficult to mechanically remove from surfaces and are more resistant to sanitizers than their free-living counterparts (15, 40).
Dairy products have been implicated in outbreaks of listeriosis (10, 31). However, most of the in vitro studies of the growth and survival of L. monocytogenes in such products have used strains previously cultivated planktonically (41). Although the results obtained in these studies are of great value, such studies have not taken into consideration that cells contaminating a product in a food-processing environment are usually attached to surfaces enclosed in biofilms. Limited information is available on the kinetic behavior of L. monocytogenes in dairy products inoculated with detached cells, although preincubation conditions have been shown to influence subsequent growth and survival of L. monocytogenes in foods (7, 13, 17, 18). Given the major physiological differences between attached and planktonic cells (15, 27, 48), an effect on subsequent growth might be possible.
Considering the above, the main objective of the present study was to assess the influence of L. monocytogenes preincubation conditions with respect to mode of growth (either attached to SS or grown suspended in dairy products) on the subsequent growth of this pathogen in milk (at 5°C for 20 days). To prepare the two types of inocula, two different growth media (milk and vanilla custard) and temperatures (5 and 20°C) were studied. The unforced detachment of L. monocytogenes cells from SS coupons and growth in two dairy products (milk and custard) at 5°C for 20 days was also evaluated. In the latter case, previous attachment of cells to the coupons was done under especially adverse preincubation conditions (in yogurt at 5°C for 7 days).
L. monocytogenes Scott A (serotype 4b, epidemic strain, human isolate), kindly provided by Eddy Smid (Agrotechnological Research Institute ATO-DLO, Wageningen, The Netherlands), was used throughout the study. Although L. monocytogenes Scott A does not attach to surfaces as strongly as other strains, its attachment ability has been ranked in the middle of those of a list of multiple clinical and food isolates (12, 49). This strain was selected due to its clinical origin and the strong epidemiologic association of serotype 4b with human listeriosis (25, 29). Moreover, a single strain was used so that the findings on the comparative growth of planktonic and attached cells following detachment were not affected by the dominance of different strains under different growth conditions. Stock cultures were maintained in tryptic soy broth supplemented with yeast extract (TSBYE) (Biolife Italiana Srl, Milan, Italy) supplemented with 20% glycerol at −20°C and were regenerated by transferring 0.05 ml of the frozen culture into 10 ml of TSBYE and incubating at 30°C for 24 h. Aliquots (0.1 ml) of activated cultures were transferred to 10 ml of TSBYE, incubated at 30°C for 18 h, and then harvested by centrifugation (5,000 × g for 10 min at 4°C; Heraeus Instruments Megafuge 1.0 R). The cell pellet was washed and resuspended twice in 10 ml of Ringer's solution (Ringer's tablets; Merck, Darmstadt, Germany) before inoculation. Planktonic growth prior to attachment on abiotic surfaces aimed to simulate situations where cells from liquid food residues settle on food contact surfaces.
High-temperature-, short-time-pasteurized whole cows' milk (cartons of 1 liter), traditional vanilla custard (packages of 170 g), and yogurt (packages of 200 g) were purchased from a local market (within 18 to 24 h of their production) and transferred (at 4°C) to the laboratory for inoculation. Milk had a typical composition of 88% water, 3.5% fat, 3.2% protein, and 4.6% carbohydrate, with a pH ranging from 6.6 to 6.7. Vanilla custard, otherwise called “vanilla cream,” is a traditional milk-based dessert which is very popular to Mediterranean countries (41). It is made from cows' milk, sugar, modified starch of tapioca, wheat and corn starch, egg yolk, and vanillin (41). To ensure its microbiological safety, the custard is subjected to thermal treatment (82°C for 30 min) before packing. According to the manufacturer's specifications, the final product contains 4.6% fat, 2.9% protein, and 17.8% carbohydrate and has a pH ranging from 6.3 to 6.7. The yogurt was made of cows' milk and contained 4% fat, 4.5% protein, and 6.5% carbohydrate according to the manufacturer's specifications, with a pH ranging from 4 to 4.4.
SS coupons (2 by 5 cm, type AISI-304, no. 2b finish, 0.1 cm thick; Halyvourgiki Inc., Athens, Greece) were the abiotic substrates used for L. monocytogenes adhesion studies, since SS is a material commonly used for the manufacture of food-processing equipment (4). The coupons were initially soaked in acetone (overnight) to remove any debris and grease from the manufacturing process. Coupons were then washed in commercial detergent solution, rinsed thoroughly with distilled water, air dried, and finally sterilized by autoclaving at 121°C for 15 min before use.
SS surfaces exposed to different types of dairy products (milk, vanilla custard, and yogurt) were used to simulate harborage sites within processing plants that cannot be easily cleaned and sanitized. Bacterial attachment was evaluated at 5 and 20°C to simulate the environmental conditions typically encountered in dairy processing environments. To investigate the hypotheses described above, three main studies were carried out. Initially, the adhesion of L. monocytogenes on SS coupons immersed in milk, custard, and yogurt for 7 days at 5 and 20°C was evaluated. The selection of 7 days of habituation was made in order to allow sufficient time for several cycles of dissociation events and subsequent regrowth of the biofilm to occur (47). This experiment aimed to determine the effect of the environment (i.e., growth medium and temperature) on the ability of L. monocytogenes to attach and to form biofilm on food-soiled SS coupons. At the next stage, two types of L. monocytogenes cells (i.e., detached from food-soiled surfaces or suspended in surrounding product) were used to inoculate milk in order to evaluate the effect of the previous environment on the subsequent growth kinetics of L. monocytogenes at 5°C. The experiment focused directly on the comparative growth of the two types of cells in case of different cross-contamination scenarios. Specifically, cells growing in suspension aimed to simulate cross-contamination of milk from product waste, purge, or residues, whereas attached cells represented contamination from soiled surfaces. Finally, the ability of L. monocytogenes cells to detach (unforced) from SS coupons soiled with yogurt and to disperse in noninoculated dairy products, such as milk and vanilla custard, at 5°C was evaluated. All three studies were performed at least twice, and in each replicate study the samples were analyzed in duplicate (n = 4).
Prior to inoculation, portions (40 ml) of freshly pasteurized milk were aseptically transferred into 50-ml sterile plastic centrifuge tubes. Inoculation of custard and yogurt was performed in their commercial packages (170 g for vanilla custard and 200 g for yogurt). Specifically, a 0.5-ml aliquot of appropriately diluted L. monocytogenes culture was added into the center of each package and/or tube in order to obtain an initial population density of approximately 107 CFU/g or CFU/ml as determined in preliminary trials. To ensure uniform distribution of the inoculum, milk samples were mixed by vortexing (30 s), whereas samples of custard and yogurt were thoroughly stirred with a sterile spatula. Following inoculation, individual sterile SS coupons were vertically immersed in the center of the inoculated packages and/or tubes, which were stored in high-precision (±0.5°C) incubation chambers (MIR-153; Sanyo Electric Co., Osaka, Japan) for 7 days at 5°C or 20°C. Uninoculated control samples were also held under the same conditions and analyzed frequently to confirm the absence of L. monocytogenes using the ISO 11290 enrichment protocol (23).
Detachment of attached cells from the SS coupons was performed by using the bead vortexing method (19), which has been established as the most suitable method for removal of attached bacteria (30). Briefly, after 7 days of incubation, each SS coupon was carefully removed from the dairy product, using sterile forceps, and was thoroughly rinsed with 25 ml of Ringer's solution in order to remove both food residue and loosely attached cells. The coupon was then transferred to a new 50-ml centrifuge tube containing 40 ml of Ringer's solution and 12 sterile glass beads (diameter, 5 mm) and was subsequently vortexed for 2 min at maximum intensity in order to detach the cells from the coupon. Quantification of attached L. monocytogenes cells and total microflora (TVC) was performed by surface plating 0.1-ml aliquots of appropriate 10-fold serial dilutions on duplicate plates of Palcam Listeria selective agar (Palcam; Biolife) and tryptic soy agar (Biolife) supplemented with 0.6% yeast extract (TSAYE), respectively. Formed colonies were counted after incubation of plates at 37°C for 48 h (Palcam) or at 30°C for 72 h (TSAYE).
The populations of suspended L. monocytogenes and TVC in the dairy products from which the SS coupons had been removed were also determined. Specifically, portions (1 ml for milk and 25 g for custard and yogurt) from the dairy samples were homogenized and serially diluted (1:10 dilution) in Ringer's solution. Aliquots (0.1 ml) of the appropriate dilution were surface plated in duplicate as described above. Moreover, typical L. monocytogenes colonies from Palcam and TSAYE plates were biochemically confirmed according to ISO 11290 (23).
Milk and custard incubated at 5°C and 20°C for 7 days were the environmental conditions used to harvest attached and suspended L. monocytogenes cells as described above. Aliquots (1 ml) of the appropriate dilution of each bacterial suspension (attached or suspended) were used to inoculate freshly pasteurized milk (100 ml), as described above, in order to obtain an initial population density of approximately 102 CFU/ml. All inoculated milk samples were statically incubated at 5°C for 20 days, and the suspended L. monocytogenes population was quantified every 24 h following the procedures described above.
SS coupons bearing attached L. monocytogenes cells habituated in yogurt for 7 days at 5°C as described above were transferred to uninoculated custard and milk. Specifically, by the end of habituation, SS coupons were aseptically removed, rinsed thoroughly with 25 ml of Ringer's solution, and finally transferred into either custard (170 g in its original package) or freshly pasteurized milk (50 ml in a plastic centrifuge tube). The population of coupon-attached cells before their transfer was also determined. Samples were statically incubated at 5°C for 20 days, while the L. monocytogenes population in the two dairy products was quantified every 48 h following the procedures described above.
Attached cells were also examined using epifluorescence microscopy (N-400F; Optika Microscopes, Italy) according to the protocol described by Pan et al. (40). Briefly, SS coupons habituated for 7 days in the different dairy products were aseptically removed from the products, rinsed thoroughly with 25 ml of Ringer's solution, and transferred to a small petri dish (diameter, 5 cm). Samples were stained with 0.01% acridine orange (Sigma-Aldrich Ltd., Greece) for 5 min at room temperature. Subsequently, coupons were rinsed three times with Ringer's solution to remove excess stain. Images of attached cells taken with a digital (charge-coupled device) camera (E-330; Olympus, Greece) were processed using Image-Pro Plus image analysis software (version 4.5; Media Cybernetics, Silver Spring, MD).
The data from plate counts (CFU per gram or per ml) of duplicate samples from two independent experiments (n = 4) were transformed to log10 values. The log10-transformed data for attached and suspended cells (from SS coupons) of L. monocytogenes were fitted to the Baranyi model with the in-house program DMFit version 2.1 (available at http://www.ifr.ac.uk/Safety/DMfit/default.html) in order to estimate the growth kinetics of the two populations. For each growth curve, the maximum specific growth rate (μmax; days−1), the lag time (tlag; days), the lower asymptote (yo; log CFU/ml), and the upper asymptote (yend; log CFU/ml) were determined. The growth data as well as the parameter estimates of the fitting then were analyzed by analysis of variance by the general linear model procedure of the SPSS statistical package (SPSS 10.0.1 for Windows; SPSS, Inc., Chicago, IL). Tukey's multiple-range test was used to compare means. All differences are reported at a significance level of alpha 0.05.
Bacterial attachment on SS coupons and proliferation in different food matrices were more favorable (P < 0.05) at 20°C than at 5°C (Fig. (Fig.1).1). At 20°C, significantly (P < 0.05) higher numbers of attached populations were recovered from coupons placed in custard (5.30 and 5.42 log10 CFU/cm2 for L. monocytogenes and TVC, respectively) than from those placed in milk (4.49 and 4.55 log10 CFU/cm2 for L. monocytogenes and TVC, respectively). It should be noted that at 20°C for both dairy products, similar levels of attached L. monocytogenes and TVC were observed, suggesting that the pathogen dominated TVC. At 5°C, the lowest level of attached cells was observed in yogurt, whereas milk and custard resulted in similar levels of attachment (approximately 3.7 log10 CFU/cm2 and 3.9 log10 CFU/cm2 for L. monocytogenes and TVC, respectively). Furthermore, incubation of milk and yogurt at 5°C resulted in significantly (P < 0.05) higher levels of attached TVC than of attached L. monocytogenes, indicating the existence of a mixed attached population on the SS coupons. Moreover, microscopic observation of the coupons placed in yogurt revealed the presence of attached cells (Fig. (Fig.2).2). Regardless of storage temperature, the bacterial populations growing in suspension were significantly greater (P < 0.05) than the attached populations (Fig. (Fig.1).1). Except for in yogurt, no major differences were seen between the suspended populations of L. monocytogenes and TVC (Fig. (Fig.11).
Multivariate statistical analysis revealed that the preincubation conditions (storage temperature, growth medium, and mode of bacterial growth) significantly (P < 0.05) affected the subsequent growth kinetics of L. monocytogenes in milk at 5°C. Habituation at 20°C delayed (P < 0.05) subsequent growth of both attached and suspended cells of L. monocytogenes in milk, as seen by an increased lag phase, compared to habituation at 5°C (Table (Table1;1; Fig. Fig.3).3). At 5°C, no significant difference (P < 0.05) was observed between the two types of inocula in the subsequent growth kinetics (lag phase and growth rate) of L. monocytogenes in milk (Table (Table1).1). Regardless of temperature, habituation of L. monocytogenes in milk decreased the lag phase of cells previously attached to SS coupons compared to cells previously grown in suspension (Table (Table1;1; Fig. Fig.3).3). However, for populations habituated in milk, a slightly higher growth rate was observed for suspended cells than for attached cells (Table (Table1).1). In contrast, habituation in custard increased the lag phase of cells previously attached to SS coupons compared to cells previously grown in suspension (Table (Table1;1; Fig. Fig.3)3) In general, the highest maximum specific growth rates were observed when the habituation conditions were identical to the subsequent storage conditions (Table (Table1).1). Moreover, total bacterial counts reached approximately 6 to 7 log10 CFU/ml, regardless of preincubation conditions (data not shown).
Although L. monocytogenes was not detectable on SS coupons after 7 days of incubation in yogurt (Fig. (Fig.1)1) the transfer of whole coupons into uninoculated milk and custard resulted in the detachment and proliferation of the pathogen in both products (Fig. (Fig.4).4). More specifically, detectable growth of detached L. monocytogenes cells occurred after 9 and 10 days in custard and milk, respectively (Fig. (Fig.4).4). However, by the end of incubation (21 days), higher growth was obtained in custard (4.4 log10 CFU/g) than in milk (3.7 log10 CFU/ml) (Fig. (Fig.44).
The ability of bacteria, including L. monocytogenes, to adhere to and colonize surfaces is influenced by many environmental factors (26) that affect the physiological characteristics of bacteria (6, 11, 33) and/or alter the chemistry of the surface (26). The type and the composition of food residue on food-processing equipment have been suggested to influence both the population levels of attached cells and their resistance to disinfectants (6, 20, 24, 33, 44). Moreover, in accordance with previous studies, attachment of L. monocytogenes was found to increase with increasing temperature (6, 19, 33, 38). It has been suggested that these differences in attachment are independent of increases in cell density (33). Generally, enhanced attachment has been attributed to the increased production of flagella, which are highly related to cell attachment on surfaces under static conditions (9, 50). In contrast, Lemon et al. (28) showed that flagellum-mediated motility, and not flagella, is critical for both adhesion and biofilm formation of L. monocytogenes on abiotic surfaces. It has been speculated that the differences between studies are mainly due to different experimental approaches which in turn result in bacterial changes, including changes in pH, oxygen tension, and nutrient availability (47).
In contrast to temperature, low pH has been shown to stimulate initial adhesion of L. monocytogenes to SS, as negative groups on the cell surface become protonated at low pH (6). Thus, the limited attachment of L. monocytogenes in yogurt-soiled surfaces as found in the current study (Fig. (Fig.1)1) could be indicative of the inhibiting effect of low pH on the subsequent steps in the development of biofilm (i.e., microcolony formation, maturation, and regrowth of biofilm). Therefore, further studies are needed to elucidate the effect that such acidic environments may have on the different steps of biofilm development. Moreover, survival of the pathogen in the surrounding medium (Fig. (Fig.1)1) may be attributable to acid adaptation mechanisms induced by habituation in yogurt for 7 days (2).
The natural flora of milk and custard did not seem to inhibit attachment of L. monocytogenes, since the enumerated populations on surfaces immersed in these products ranged from 3.5 to 5.5 log CFU/cm2 (Fig. (Fig.1).1). Listeria commonly exists as a part of multispecies biofilms with other bacteria in food-processing facilities, and competitive microflora was shown to either enhance or inhibit L. monocytogenes attachment to surfaces (5, 8, 52). The presence of other bacterial species may also render the pathogen more resistant to stress conditions (5). On the other hand, Zhao et al. (52) demonstrated that some bacterial species may inhibit L. monocytogenes by producing antilisterial metabolites. Thus, the impact that the resident microorganisms may have on the biofilm-forming capacity of L. monocytogenes needs further investigation.
Detached L. monocytogenes cells from soiled surfaces may contaminate foods and proliferate under refrigeration. Our findings also suggested that habituation of L. monocytogenes on surfaces at 20°C delayed subsequent growth of the detached cells in milk at 5°C compared to previous habituation at 5°C. Such findings are supported by studies demonstrating that exposure to temperature downshifts increased the lag time of planktonic L. monocytogenes cells (7, 13). In addition to temperature, the variability of food matrices for bacterial habituation and attachment markedly affected the growth behavior of suspended and detached cells. Geornaras et al. (17, 18) investigated the effect of inoculum origin on the survival and growth kinetics of L. monocytogenes during postprocess control of commercial frankfurters and smoked sausages. They found that the ecological background of the cells (i.e., attached to SS or grown in suspension) affected the growth behavior of the pathogen, including the lag time and the general response to various antimicrobial solutions (17, 18).
In the present study, the observed variations in growth of detached L. monocytogenes cells from surfaces soiled with different products likely depended on the temperature, the state (e.g., liquid or solid), and the composition of the food as well as the available nutrients. Specifically, detached cells from milk-soiled surfaces had better subsequent growth in milk than those from custard-soiled surfaces at both high and low temperatures. Presumably, detachment and dispersal of cells in a food ecosystem different from that where attachment took place increased the adaptive response. Adaptation to a new environment is analogous to the lag time (35), and studies with liquid laboratory media have shown that adaptation of planktonically grown L. monocytogenes cells increases with abrupt environmental shifts (i.e., pH and water activity) (35, 46). Cell adaptation is an important issue in challenge studies, since preparation of inocula under conditions different from those encountered in the targeted food could result in underestimation of bacterial growth and survival. In addition, the observation that detached cells from milk exhibited faster growth than their suspended counterparts underlines the importance of the physiological or metabolic history of cells for their subsequent growth (35, 46). It can be speculated that in the case of contamination, the risk of L. monocytogenes growth may be higher when bacteria are derived from biofilms rather than from residual liquids or product waste (e.g., whey or purge). Therefore, studies evaluating the effect of preservation treatments or modeling of microbial responses should also consider cells detached from biofilms.
Attached L. monocytogenes cells habituated in yogurt seemed to detach from the SS coupons and migrate to the favorable medium of milk or custard at 5°C (Fig. (Fig.4).4). This observation suggests that the bead vortexing procedure and/or the conventional plating technique used in this study could underestimate the attached microbial population on SS coupons and thus the actual risk of contaminated surfaces (1). It has also been suggested that in order to survive under adverse environmental conditions, many food-borne pathogens enter a viable but nonculturable state (i.e., a dormant state) (51) in which they maintain metabolic activity and pathogenicity but cannot be cultured by common laboratory methods (39). This conclusion is supported by a previous study demonstrating that cells that were metabolically active but not culturable by conventional plating techniques (detection limit, 1.3 log10 CFU/cm2) remained attached to the SS coupons after the bead vortexing (19). Furthermore, custard was more favorable than milk for detachment of L. monocytogenes from soiled surfaces and proliferation (Fig. (Fig.4).4). This may be associated with the structural heterogeneity and difference in nutrient composition that contributed to the faster detachment and proliferation of custard-derived cells. Moreover, the density, structure, and strength of the attached microbial population are among the factors influencing bacterial detachment (21, 22, 36).
In conclusion, the results of this work indicate that L. monocytogenes cells attached to equipment surfaces may constitute a significant risk due to possible detachment and multiplication in food products. Moreover, detached cells may exhibit a higher tolerance to stressful conditions than cells in suspension, which depends on the growth environment before and after their transfer. Detached cells, although old or injured, still pose a major public health threat to the food industry. Food soil, as ecological background, does affect both the growth and behavior of L. monocytogenes, and this effect differs between planktonic and biofilm states of growth. However, future research should evaluate similar contamination scenarios with multiple L. monocytogenes strains, also assessing genotypic changes due to attachment and detachment of the bacterium. This would improve understanding of the mechanisms that allow persistence of L. monocytogenes in dairy plants.
This work was funded by the European Union Integrated project “BIOTRACER: Improved bio-traceability of unintended microorganisms and their substances in food and feed chains” (proposal/contract no. 036272). E. D. Giaouris acknowledges the Greek State Scholarships Foundation (I.K.Y.) for providing a fellowship for postdoctoral studies.
Published ahead of print on 18 September 2009.