Variation in Listeria monocytogenes dose responses in relation to subtypes encoding a full-length or truncated internalin A.

Internalin A (InlA; encoded by inlA) facilitates the crossing of the intestinal barrier by Listeria monocytogenes. Mutations leading to a premature stop codon (PMSC) in inlA and thus attenuated mammalian virulence have been reported. We recently characterized 502 L. monocytogenes food isolates from a retail survey and 507 human clinical isolates from multiple U.S. states with respect to the presence/absence of inlA mutations. The objective of this study was to investigate the hypothesis that dose responses for human listeriosis vary between L. monocytogenes strains with and those without a PMSC in inlA. Subtype-specific prevalence and concentration distributions in food, along with epidemiologic and consumption data, were input into established dose-response models to generate an r value (probability of a cell causing illness). Under the conservative assumption that L. monocytogenes levels at retail represent levels consumed, mean log(10) r values were -8.1 and -10.7 for L. monocytogenes subtypes with genes encoding a full-length and a truncated InlA, respectively. L. monocytogenes carrying a 5' frameshift mutation in a homopolymeric tract showed a mean log(10) r value of -12.1. Confidence intervals for the r values and their differences varied depending on subtypes. When the increase in concentration of L. monocytogenes subtypes between retail and consumption was considered, mean log(10) r values were reduced to -10.4, -13.8, and -12.8 for the subtypes with genes encoding a full-length InlA, for the subtypes carrying a PMSC in inlA, and for all L. monocytogenes isolates regardless of subtype, respectively. Our study provides further quantitative evidence that L. monocytogenes subtypes vary in abilities and relative likelihoods of causing human disease, which were mechanistically related to defined genetic markers.

Survey of Salmonella in raw tree nuts at retail in the United States.

The objective of this survey was to estimate the prevalence, contamination level, and genetic diversity of Salmonella in selected raw, shelled tree nuts (Brazil nuts, cashews, hazelnuts, macadamia nuts, pecans, pine nuts, pistachios, and walnuts) at retail markets in the United States. A total of 3,374 samples of eight tree nuts were collected from different types of retail stores and markets nationwide between September 2015 and March 2017. These samples (375 g) were analyzed using a modified FDA's BAM Salmonella culture method. Of the 3,374 samples, 15 (0.44%) (95% confidence interval [CI] [0.25, 0.73]) were culturally confirmed as containing Salmonella; 17 isolates were obtained. Among these isolates, there were 11 serotypes. Salmonella was not detected in Brazil nuts (296), hazelnuts (487), pecans (510), pine nuts (500), and walnuts (498). Salmonella prevalence estimates in cashews (510), macadamia (278), and pistachios (295) were 0.20% (95% CI [<0.01, 1.09]), 2.52% (95% CI [1.02, 5.12]), and 2.37% (95% CI [0.96, 4.83]), respectively. The rates of Salmonella isolation from major/big-chain supermarkets (1381), small-chain supermarkets (328), discount/variety/drug stores (1329), and online (336) were 0.29% (95% CI [0.08, 0.74]), 0.30% (95% CI [0.01, 1.69]), 0.45% (95% CI [0.17, 0.98]), and 1.19% (95% CI [0.33, 3.02]), respectively. Salmonella prevalence in organic (530) and conventional (2,844) nuts was not different statistically (P = 0.0601). Of the enumerated samples (15), 80% had Salmonella levels ≤0.0092 most probable number (MPN)/g. The highest contamination level observed was 0.75 MPN/g. The prevalence and contamination levels of Salmonella in the tree nuts analyzed were generally comparable to previous reports. Pulsed-field gel electrophoresis, serotype, and sequencing data all demonstrated that Salmonella population in nuts is very diverse genetically. PRACTICAL APPLICATION: The prevalence, contamination level, and genetic diversity of Salmonella in eight types of tree nuts (3,374 samples collected nationwide) revealed in this survey could help the development of mitigation strategies to reduce public health risks associated with consumption of these nuts.

Prevalence of Salmonella in Cashews, Hazelnuts, Macadamia Nuts, Pecans, Pine Nuts, and Walnuts in the United States.

Nuts have been identified as a vector for salmonellosis. The objective of this project was to estimate the prevalence and contamination level of Salmonella in raw tree nuts (cashews, pecans, hazelnuts, macadamia nuts, pine nuts, and walnuts) at retail markets in the United States. A total of 3,656 samples of six types of tree nuts were collected from different types of retail stores and markets nationwide between October 2014 and October 2015. These samples were analyzed using a modified version of the Salmonella culture method from the U.S. Food and Drug Administration's Bacteriological Analytical Manual. Of the 3,656 samples collected and tested, 32 were culturally confirmed as containing Salmonella. These isolates represented 25 serotypes. Salmonella was not detected in pecans and in-shell hazelnuts. Salmonella prevalence estimates (and 95% confidence intervals) in cashews, shelled hazelnuts, pine nuts, walnuts, and macadamia nuts were 0.55% [0.15, 1.40], 0.35% [0.04, 1.20], 0.48% [0.10, 1.40], 1.20% [0.53, 2.40], and 4.20% [2.40, 6.90], respectively. The rates of Salmonella isolation from major or big chain supermarkets, small chain supermarkets, discount, variety, or drug stores, and online were 0.64% [0.38, 1.00], 1.60% [0.80, 2.90], 0.00% [0.00, 2.40], and 13.64% [2.90, 35.00], respectively (Cochran-Mantel-Haenszel test: P = 0.02). The rates of Salmonella isolation for conventional and organic nuts were not significantly different. Of the samples containing Salmonella, 60.7% had levels less than 0.003 most probable number (MPN)/g. The highest contamination level observed was 0.092 MPN/g. The prevalence and levels of Salmonella in these tree nut samples were comparable to those previously reported for similar foods.

Thermal resistance parameters for pathogens in white grape juice concentrate.

The heat resistance of Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes that were in stationary phase, had been exposed to high osmotic pressure, or were acid adapted was evaluated in white grape juice concentrate (58 degrees Brix, pH 3.3). The most heat-resistant cells of all three pathogens were those exposed to high osmotic pressure or in stationary phase. Unlike in single-strength juices, in concentrate the acid-adapted cells for all three pathogens were less heat resistant than were cells in the other physiological states. E. coli O157:H7 had the highest heat resistance for all temperatures tested (e.g., D62 degrees C = 1.8 +/- 0.3 min, with a z-value of 9.9 +/- 0.6 degrees C). L. monocytogenes exposed to high osmotic pressure had the highest z-value (12.3 +/- 1.2 degrees C), although its D-values for all temperatures tested were lower (e.g., D62 degrees C = 0.93 +/- 0.1 min) than those for E. coli O157:H7. Salmonella was the most sensitive of the pathogens under all conditions. Based on the results obtained in this study, one example of a heat treatment that will inactivate 5 log units of all three pathogens in white grape juice concentrate was calculated as 1.5 min at 71.1 degrees C (z = 10.3 degrees C). Validation studies confirmed the predicted D71 degrees C for E. coli O157:H7 exposed to high osmotic pressure.

Attributing risk to Listeria monocytogenes subgroups: dose response in relation to genetic lineages.

The objective of this study was to evaluate the hypothesis that the dose-response relationship for Listeria monocytogenes in humans varies with genotypic lineage or subtype. The linkages between molecular subtyping data and enumeration data for L. monocytogenes subtypes in foods consumed by the at-risk population were examined to test this hypothesis. We applied a conditional probability model to conduct a subtype-specific dose-response analysis, with the focus on invasive listeriosis. L. monocytogenes differed not only in the molecular subtype and lineage but also in the contamination level when isolates of the pathogen occurred in retail samples of ready-to-eat foods. Using the exponential model parameter r-value as a measure (essentially the probability of a single cell causing illness), we found that the virulence varied among L. monocytogenes lineages by several orders of magnitude. Under the assumptions made, for L. monocytogenes lineages I and II the consumption of a single cell would result in listeriosis with log average probabilities of -7.88 (equivalent to once in 10(7.78) times) and -10.3, respectively, as compared with -9.72 for L. monocytogenes independent of subtype. A greater difference in r-values was found for selected ribotypes. The uncertainty about the r-value estimates was small compared with the large differences in the virulence parameters themselves. Thus, for L. monocytogenes both subtype and the number of cells consumed matter, highlighting the usefulness of considering both exposure concentration and subtype prevalence in dose-response analysis. As advances are made in molecular subtyping and quantitative tools for dose-response analysis, further studies integrating genomic data into quantitative risk assessments will enable better attribution of disease risk to L. monocytogenes subtypes.

Survey of Listeria monocytogenes in ready-to-eat foods.

The purpose of this study was to develop data on the risk of listeriosis to support a science-based strategy for addressing Listeria monocytogenes in foods in the United States. Eight categories of ready-to-eat foods were collected over 14 to 23 months from retail markets at Maryland and northern California FoodNet sites. The product categories included luncheon meats, deli salads, fresh soft "Hispanic-style" cheeses, bagged salads, blue-veined and soft mold-ripened cheeses, smoked seafood, and seafood salads. The presence and levels of L. monocytogenes in the samples were determined by rapid DNA-based assays in combination with culture methods. Of 31,705 samples tested, 577 were positive. The overall prevalence was 1.82%. with prevalences ranging from 0.17 to 4.7% among the product categories. L. monocytogenes levels in the positive samples varied from <0.3 MPN (most probable number) per g to 1.5 x 10(5) CFU/g, with 402 samples having levels of <0.3 MPN/g, 21 samples having levels of >10(2) CFU/g, and the rest of the samples having intermediate levels. No obvious trends with respect to seasonality were observed. Significant differences (P < 0.05) between the sampling sites were found, with higher prevalences for threes categories in northern California and for two categories in Maryland. Significantly (P < 0.001) higher prevalences were found for in-store-packaged samples than for manufacturer-packaged samples of luncheon meats, deli salads, and seafood salads, while 16 of the 21 samples with higher counts were manufacturer packaged. The data collected in this study help to fill gaps in the knowledge about the occurrence of L. monocytogenes in foods, and this new information should be useful in the assessment of the risk posed by L. monocytogenes to consumers.

Listeria monocytogenes: low levels equal low risk.

Because of the public health significance of L. monocytogenes, U.S. regulatory agencies established a policy whereby ready-to-eat foods contaminated with the organism at a detectable level are deemed adulterated. This "zero tolerance" policy, however, makes no distinction between foods contaminated at high and low levels. We have reported elsewhere that a survey of over 31,000 ready-to-eat retail food samples, representing eight product categories, showed an overall prevalence rate of 1.82% for these foods. In this study, we used the food survey data in combination with concurrent data regarding illness in the population consuming the foods, together with other variable factors, to derive a dose-response model. The confidence interval for prevalence was 1.68 to 1.97%. L. monocytogenes levels, which ranged from -2 to 6 log CFU/g, were adequately described by the distribution beta (0.29, 2.68, -1.69, 6.1). An exponential dose-response model was obtained, with an R value (essentially the probability of a single cell causing illness) of 1.76 x 10(-10) for the population at the highest risk. A microbial risk assessment based on the model shows that an alternative to the zero tolerance strategy has a greater risk reduction potential and suggests that a management strategy focusing on the concentration of L. monocytogenes rather than its presence alone may have a greater impact on the improvement of public health by facilitating the development of control measures to limit the maximum levels of L. monocytogenes in foods.

Longitudinal studies on Listeria in smoked fish plants: impact of intervention strategies on contamination patterns.

Four ready-to-eat smoked fish plants were monitored for 2 years to study Listeria contamination patterns and the impact of plant-specific Listeria control strategies, including employee training and targeted sanitation procedures, on Listeria contamination patterns. Samples from the processing plant environment and from raw and finished product were collected monthly and tested for Listeria spp. and Listeria monocytogenes. Before implementation of intervention strategies, 19.2% of raw product samples (n = 276), 8.7% of finished product samples (n = 275), and 26.1% of environmental samples (n = 617) tested positive for Listeria spp. During and after implementation of Listeria control strategies, 19.0% of raw product samples (n = 242), 7.0% of finished product samples (n = 244), and 19.5% of environmental samples (n = 527) were positive for Listeria spp. In one of the four fish plants (plant 4), no environmental samples were positive for L. monocytogenes, and this plant was thus excluded from statistical analyses. Based on data pooled from plants 1, 2, and 3, environmental Listeria spp. prevalence was significantly lower (P < 0.05) for nonfood contact surfaces and the finished product area and for the overall core environmental samples after implementation of control strategies. Listeria prevalence for floor drains was similar before and after implementation of controls (49.6 and 54.2%, respectively). Regression analysis revealed a significant positive relationship (P < 0.05) between L. monocytogenes prevalence in the environment and in finished products before implementation of control strategies; however, this relationship was absolved by implementation of Listeria control strategies. Molecular subtyping (EcoRI ribotyping) revealed that specific L. monocytogenes ribotypes persisted in three processing plants over time. These persistent ribotypes were responsible for all six finished product contamination events detected in plant 1. Ribotype data also indicated that incoming raw material is only rarely a direct source of finished product contamination. While these data indicate that plant-specific Listeria control strategies can reduce cross-contamination and prevalence of Listeria spp. and L. monocytogenes in the plant environment, elimination of persistent L. monocytogenes strains remains a considerable challenge.

Tracking of Listeria monocytogenes in smoked fish processing plants.

Four smoked fish processing plants were used as a model system to characterize Listeria monocytogenes contamination patterns in ready-to-eat food production environments. Each of the four plants was sampled monthly for approximately 1 year. At each sampling, four to six raw fish and four to six finished product samples were collected from corresponding lots. In addition, 12 to 14 environmental sponge samples were collected several hours after the start of production at sites selected as being likely contamination sources. A total of 234 raw fish, 233 finished products, and 553 environmental samples were tested. Presumptive Listeria spp. were isolated from 16.7% of the raw fish samples, 9.0% of the finished product samples, and 27.3% of the environmental samples. L. monocytogenes was isolated from 3.8% of the raw fish samples (0 to 10%, depending on the plant), 1.3% of the finished product samples (0 to 3.3%), and 12.8% of the environmental samples (0 to 29.8%). Among the environmental samples, L. monocytogenes was found in 23.7% of the samples taken from drains, 4.8% of the samples taken from food contact surfaces, 10.4% of the samples taken from employee contact surfaces (aprons, hands, and door handles), and 12.3% of the samples taken from other nonfood contact surfaces. Listeria spp. were isolated from environmental samples in each of the four plants, whereas L. monocytogenes was not found in any of the environmental samples from one plant. Overall, the L. monocytogenes prevalence in the plant environment showed a statistically significant (P < 0.0001) positive relationship with the prevalence of this organism in finished product samples. Automated EcoRI ribotyping differentiated 15 ribotypes among the 83 L. monocytogenes isolates. For each of the three plants with L. monocytogenes-positive environmental samples, one or two ribotypes seemed to persist in the plant environment during the study period. In one plant, a specific L. monocytogenes ribotype represented 44% of the L. monocytogenes-positive environmental samples and was also responsible for one of the two finished product positives found in this plant. In another plant, a specific L. monocytogenes ribotype persisted in the raw fish handling area. However, this ribotype was never isolated from the finished product area in this plant, indicating that this operation has achieved effective separation of raw and finished product areas. Molecular subtyping methods can help identify plant-specific L. monocytogenes contamination routes and thus provide the knowledge needed to implement improved L. monocytogenes control strategies.

Thermal resistance parameters for Shiga toxin-producing Escherichia coli in apple juice.

The purpose of the present study was to determine the heat resistance of six non-O157 Shiga toxin-producing Escherichia coli (STEC) serotypes in comparison to E. coli O157:H7 in single-strength apple juice without pulp. The thermal parameters for stationary-phase and acid-adapted cells of E. coli strains from serogroups O26, O45, O103, O111, O121, O145, and O157:H7 were determined by using an immersed coil apparatus. The most heat-sensitive serotype in the present study was O26. Stationary-phase cells for serotypes O145, O121, and O45 had the highest D(56°C)-value among the six non-O157 serotypes studied, although all were significantly lower (P < 0.05) than that of E. coli O157:H7. At 60°C E. coli O157:H7 and O103 demonstrated the highest D-values (1.37 ± 0.23 and 1.07 ± 0.03 min, respectively). The D(62°C) for the most heat-resistant strain belonging to the serotype O145 was similar (P > 0.05) to that for the most resistant O157:H7 strain (0.61 ± 0.17 and 0.60 ± 0.09 min, respectively). The heat resistance for stationary-phase cells was generally equal to or higher than that of acid-adapted counterparts. Although E. coli O157:H7 revealed D-values similar to or higher than the individual six non-O157 STEC serotypes in apple juice, the z-values for most non-O157 STEC tested strains were greater than those of E. coli O157:H7. When data were used to calculate heat resistance parameters at a temperature recommended in U.S. Food and Drug Administration guidance to industry, the D(71.1°C) for E. coli O157:H7 and non-O157 STEC serotypes were not significantly different (P > 0.05).

Relatedness of Listeria monocytogenes Isolates recovered from selected ready-to-eat foods and listeriosis patients in the United States.

Pulsed-field gel electrophoresis and serotyping were performed for 544 isolates of Listeria monocytogenes, including 502 isolates recovered from contaminated samples from 31,705 retail ready-to-eat (RTE) food products and 42 isolates recovered from human cases of listeriosis. The isolates were from Maryland (294 isolates) and California (250 isolates) and were collected in 2000 and 2001. The isolates were placed into 16 AscI pulsogroups (level of relatedness within each group, > or =66%), 139 AscI pulsotypes (levels of relatedness, > or =25% to 100%), and eight serotypes (serotypes 1/2a, 1/2b, 1/2c, 3a, 3b, 4b, 4c, and 4d). The most frequently found pulsotypes belonged to either pulsogroup A (150 food isolates plus 4 clinical isolates) or pulsogroup B (104 food isolates plus 5 clinical isolates). The majority of the 502 food isolates were either serotype 1/2a (298 isolates) or serotype 1/2b (133 isolates), whereas the majority of the 42 clinical isolates were either serotype 1/2a (19 isolates) or serotype 4b (15 isolates). Additionally, 13 clinical isolates displayed pulsotypes also found in food isolates, whereas the remaining 29 clinical isolates displayed 24 unique pulsotypes. These data indicate that most (86%) of the L. monocytogenes subtypes found in the RTE foods sampled belonged to only two serotypes and that 90% of the isolates displayed 73 pulsotypes, with 107 isolates displaying pulsotype 1. These data should help define the distribution and relatedness of isolates found in RTE foods in comparison with isolates that cause listeriosis.

Interaction of Factors to Control Microbial Spoilage of Refrigerated Foods 1.

A variety of factors can prevent growth of microorganisms. Combining inhibitory factors can result in considerable improvement in the microbial stability of foods. Suitable combinations of growth-limiting factors at subinhibitory levels can be devised so that certain microorganisms can no longer proliferate in the product (the hurdles concept). Knowledge of the effectiveness of a wide range of combinations of hurdles for a variety of microorganisms would be valuable in product development in allowing predictions of microbial stability and safety of new formulations. Data generated in the laboratory could be used to predict the effect that changing certain factors would have relative to other factors with regard to increasing or decreasing microbial stability. These types of predictions are particularly important with refrigerated foods since the storage temperature is frequently the primary hurdle, and temperature abuse is not uncommon.

Validation of Predictive Mathematical Models Describing the Growth of Escherichia coli O157:H7 in Raw Ground Beef.

The growth of Escherichia coli O157:H7 in raw ground beef was investigated at 12°C, 20°C, and 35°C at pH 5.7 (unadjusted) and adjusted to pH 6.3 to 6.4. These growth data were fitted to the Gompertz equation and the resulting growth kinetics were compared with predictions from the U.S. Department of Agriculture Pathogen Modeling Program. Close agreement with the model was obtained at pH 5.7, but at pH 6.4, growth was more rapid than predicted. The U.S. Department of Agriculture Food Safety and Inspection Service has used this predictive model for developing proposed regulations on time-temperature requirements for carcass cooling. As there may be considerable differences in the microenvironment of raw ground beef and a beef carcass, the validity of using predictive models for estimating growth rates on a carcass should be determined by performing growth studies on carcass surfaces.

Validation of Predictive Mathematical Models Describing the Growth of Listeria monocytogenes.

Growth of Listeria monocytogenes and Listeria innocua in commercially available sterile homogeneous foods was investigated at different temperatures, pH values, and NaCl concentrations. Growth data were fitted to the Gompertz equation and the resulting growth kinetics were compared with predictions from the Pathogen Modeling Program and Food MicroModel. In general, good agreement was obtained when comparing growth rates and generation times for both models. Differences were observed when comparing lag phases, which ranged from 117 h shorter to 4.9 h longer than predicted for L. monocytogenes . Despite differences in lag phase, under most conditions, the models gave good predictions of microbial growth. Predictive modeling appears to be a useful tool in determining growth rates of Listeria in foods.

Heat Resistance of Spores of Non-Proteolytic Type B Clostridium botulinum.

The heat resistance of spores of non-proteolytic type B Clostridium botulinum was compared to that of type E and proteolytic type B spores. Spore suspensions were produced in a biphasic medium consisting of beef heart agar overlaid with a liquid phase containing trypticase, peptone, glucose, starch and cysteine. Thermal death time curves were established for seven strains heated in phosphate buffer. In general, spore suspensions of non-proteolytic type B strains had greater thermal resistance than type E strains. Decimal reduction times at 82.2°C, established by linear regression analyses of data, ranged from 1.49 to 32.3 min, but the higher heat resistances were not obtained consistently, even with different spore suspensions of the same strain. None of the spore suspensions of non-proteolytic, type B C. botulinum demonstrated heat resistance comparable to that of the proteolytic type B spores.

Heat Resistance of Talaromyces flavus and Neosartorya fischeri Isolated from Commercial Fruit Juices.

The heat resistance of two molds believed to have survived the thermal process applied to two commercial "shelf stable" fruit juices was studied. Neosartorya fischeri had a D-value of 1.4 min at 190°F (87.8°C) and a z-value of 10°F (5.6°C). Talaromyces flavus had a D-value of 2.2 min at 195°F (90.6°C) and a z-value of 9.5°F (5.2°C). Under certain conditions, both molds possess sufficient heat resistance to survive commercial thermal processes if ascospores are present in sufficient numbers.

Growth, Inhibition, and Survival of Listeria monocytogenes as Affected by Acidic Conditions.

Growth and survival of four strains of Listeria monocytogenes under acidic conditions were investigated. Tryptic soy broth with yeast extract (TSBYE) was acidified with acetic, citric, hydrochloric, lactic, or propionic acid to pH 4.0-6.0, inoculated with L. monocytogenes and incubated at 30 or 4°C. The minimum test pH at which L. monocytogenes did not grow (inhibitory pH) was determined for each acid. In the pH range tested, this inhibitory pH was 5.0 for propionic acid, 4.5 for acetic and lactic acids, and 4.0 for citric and hydrochloric acids. All four strains gave similar results. Subsequent studies were conducted at 10 and 30°C to determine changes in cell populations in TSBYE adjusted to each inhibitory pH. Initial populations of viable cells (104 CFU/ml) were reduced to <10 CFU/ml within 1-3 weeks at 30°C, whereas at 10°C, L. monocytogenes survived for 11-12 weeks in acetic, citric, or propionic acid-adjusted media and for 6 weeks in media adjusted with hydrochloric or lactic acid. The concentration of undissociated lactic acid was 0.002 M at pH 4.5.

Validation of Predictive Mathematical Models Describing Growth of Staphylococcus aureus.

An investigation was performed on the growth of Staphylococcus aureus in a commercially available, sterile, homogeneous food at 12°C with 1.2 and 5.9% NaCl; at 25°C with 10.4% NaCl; and at 20 and 35°C with 1.2, 5.3, 12.5, and 15.8% NaCl; over a pH range of 5.5 to 7.5. Growth data were fitted to the Gompertz equation and the resulting growth kinetics were compared with predictions from the Pathogen Modeling Program (PMP) and Food MicroModel (FMM). For the PMP, predicted lag-phase durations varied from 0.5 to 130 h longer than the observed values. In general, close agreement with growth rates was obtained but there was a 10-fold difference in one case. For FMM, predicted lag-phase durations ranged from 27 h shorter to 47 h longer than the observed values. Again, close agreement with growth rates was obtained, but in one case a fivefold difference was observed. In general, for the sterile foods used under the growth conditions tested, the models underestimated the growth of S. aureus . This implies that while the models can be used as a guide to indicate growth rates in foods they should not be relied upon as the sole determinant of the product's safety.

A Comprehensive Approach to Reducing the Risk of Allergens in Foods.

The control of food allergens in a food-processing plant presents numerous challenges. An allergen prevention plan must determine potential sources of contaminating allergens and appropriate controls to prevent their introduction into food products. These controls may include scheduling production of allergen-containing products at the end of manufacturing runs, appropriate labeling and use of rework, equipment and system-design considerations, and thorough cleaning of lines after running allergen-containing food products. Proper labeling of food products, effective management of label inventories, control of ingredients from suppliers, and training of employees are key factors in allergen control.

Listeria monocytogenes isolates from foods and humans form distinct but overlapping populations.

A total of 502 Listeria monocytogenes isolates from food and 492 from humans were subtyped by EcoRI ribotyping and PCR-restriction fragment length polymorphism analysis of the virulence gene hly. Isolates were further classified into genetic lineages based on subtyping results. Food isolates were obtained through a survey of selected ready-to-eat food products in Maryland and California in 2000 and 2001. Human isolates comprised 42 isolates from invasive listeriosis cases reported in Maryland and California during 2000 and 2001 as well as an additional 450 isolates from cases that had occurred throughout the United States, predominantly from 1997 to 2001. Assignment of isolates to lineages and to the majority of L. monocytogenes subtypes was significantly associated with the isolate source (food or human), although most subtypes and lineages included both human and food isolates. Some subtypes were also significantly associated with isolation from specific food types. Tissue culture plaque assay characterization of the 42 human isolates from Maryland and California and of 91 representative food isolates revealed significantly higher average infectivity and cell-to-cell spread for the human isolates, further supporting the hypothesis that food and human isolates form distinct populations. Combined analysis of subtype and cytopathogenicity data showed that strains classified into specific ribotypes previously linked to multiple human listeriosis outbreaks, as well as those classified into lineage I, are more common among human cases and generate larger plaques than other subtypes, suggesting that these subtypes may represent particularly virulent clonal groups. These data will provide a framework for prediction of the public health risk associated with specific L. monocytogenes subtypes.