Catherine W. Donnelly

Psychrotrophic Growth and Thermal Inactivation of Listeria monocytogenes as a Function of Milk Composition

Listeria monocytogenes strains 19111, 19113, 19115, F5027 and F5069 were grown in 11% nonfat milk solids, skim milk and whole milk at 4, 10, 22, and 37°C to determine the influence of temperature and milk composition on growth and thermal resistance. Milk composition affected cellular growth. The psychrotrophic growth of L. monocytogenes serotype 4b strains was enhanced in whole milk when compared to skim milk or 11% NFMS. This enhancement of psychrotrophic growth was not observed for serotype 1 or 3 strains. The stimulatory effect of whole milk on serotype 4b L. monocytogenes strains was most dramatic at 10°C where cells increased from 7.9 × 10° to 5.8 × 106 CFU/ml within 48 h. Milk composition did not affect the thermal resistance of L. monocytogenes . All strains used in this study had a D62.7°C value of 1.0 min or less, therefore, pasteurization as defined by current FDA guidelines should eliminate this organism from raw milk with a large margin of safety. Post-pasteurization contamination of dairy products with L. monocytogenes must be eliminated since the psychrotrophic nature of this organism ensures survival and proliferation during refrigerated storage.

Comparison of Heat Resistance of Listeria monocytogenes in Milk as Determined by Two Methods

The thermal resistance of 3 strains of Listeria monocytogenes was compared using test tube versus sealed tube methods of thermal inactivation. All L. monocytogenes strains were rapidly inactivated in milk when survival was measured using sealed tube thermal inactivation methods. Calculated D62°C values ranged between 0.1-0.4 min for the three strains tested. In contrast, total inactivation of L. monocytogenes populations using test tube methods of thermal inactivation could not be accomplished within 30 min at 62°C. Extensive tailing of survivor curves was consistently observed. When an initial population of 5 × 106 L. monocytogenes /ml was heated at 72, 82, or 92°C, consistent survival of a population of 102-103 L. monocytogenes /ml after 30 min was observed. The results prove that the test tube method for measuring thermal resistance of L. monocytogenes is inaccurate. Reports of extraordinary heat resistance based upon this method are correspondingly inaccurate. L. monocytogenes cells, dispersed freely in milk, will not survive pasteurization.

Low Incidence of Foodborne Pathogens of Concern in Raw Milk Utilized for Farmstead Cheese Production

Overall milk quality and prevalence of four target pathogens in raw milk destined for farmstead cheesemaking was examined. Raw milk samples were collected weekly from June to September 2006 from 11 farmstead cheese operations manufacturing raw milk cheese from cows, goats, and sheeps milk. Samples were screened for Listeria monocytogenes, Staphylococcus aureus, Salmonella, and Escherichia coli O157:H7 both quantitatively (direct plating) and qualitatively (PCR). Overall, 96.8% of samples had standard plate counts of < 100,000 CFU/ml, 42.7% of which were < 1,000 CFU/ml. Although no federal standards exist for coliforms in raw milk, 61% of samples tested conformed to pasteurized milk standards under the U.S. Pasteurized Milk Ordinance (PMO) at < 10 CFU/ml. All cow and sheep milk samples and 93.8% of goat milk samples were within the limits dictated by the PMO for somatic cell counts. Of the 11 farms, 8 (73%) produced samples that were positive for S. aureus, which was detected in 34.6% (46 of 133) of milk samples. L. monocytogenes was isolated from three milk samples (2.3%), two of which were from the same farm. E. coli O157:H7 was recovered from one sample of goats milk for an overall incidence of 0.75%. Salmonella was not recovered from any of the 133 samples. The findings of this study suggest that most raw milk intended for farmstead cheesemaking is of high microbiological quality with a low incidence of pathogens. These data will help inform risk assessments associated with the microbiological safety of farmstead cheeses, particularly those manufactured from raw milk.

Diversity of Listeria ribotypes recovered from dairy cattle, silage, and dairy processing environments

Listeria strains isolated over the past 10 years from farms and dairy processing environments were subjected to strain-specific ribotyping using the automated Riboprinter microbial characterization system, alpha version (E. I. du Pont de Nemours & Co., Inc.). A total of 388 Listeria isolates from 20 different dairy processing facilities were examined along with 44 silage, 14 raw milk bulk tank, and 29 dairy cattle (26 udder quarter milk, 1 brain, 1 liver, and 1 aborted fetus) isolates. These 475 isolates included 93 L. monocytogenes , 362 L. innocua , 11 L. welshimeri , 6 L. seeligeri , 2 L. grayi , and 1 L. ivanovii strains. Thirty-seven different Listeria ribotypes (RTs) comprising 16 L. monocytogenes (including five known clinical RTs responsible for foodborne listeriosis), 12 L. innocua , 5 L. welshimeri , 2 L. seeligeri , 1 L. ivanovii , and 1 L. grayi were identified. Greatest diversity was seen among isolates from dairy processing facilities with 14 of 16 (87.5%) of the L. monocytogenes RTs (including five clinical RTs) and 19 of 21 (90.5%) of the non- L. monocytogenes RTs detected. Sixty-five of the 93 L. monocytogenes isolates belonged to a group of five clinical RTs. These five clinical RTs included one RT unique to dairy processing environments, two RTs common to dairy processing environments and silage, and one RT common to dairy processing environments, silage, and dairy cattle with the last RT appearing in dairy processing environments, silage, raw milk bulk tanks, and dairy cattle. These findings, which support the link between on-farm sources of Listeria contamination (dairy cattle, raw milk, silage) and subsequent contamination of dairy processing environments, stress the importance of farm-based HACCP programs for controling listeriae.

Environmental sources of Listeria and Yersinia in Vermont dairy plants

This survey was conducted to identify specific environmental sources of Listeria and Yersinia in Vermont dairy plants, and to further determine whether the type of plant and specific conditions existing within plants influenced the incidence of positive microbiological results. A total of 361 environmental samples, focusing on floors and other nonproduct contact surfaces, was taken from all of Vermonts 34 dairy processing plants. The incidence of Listeria monocytogenes (1.4%) was low compared to the incidence of Listeria innocua (16.1%). While only 2.5% yielded other Yersinia species, 10.5 % of the sites were positive for Yersinia enterocolitica . Sites positive for either Listeria or Yersinia were statistically more likely to produce a positive result for both (P<.05). Fluid plants had the highest incidence of both Listeria and Yersinia when compared to cheese plants or other types of dairy manufacturing plants. Areas associated with case washers in fluid plants had the highest incidence of microbial contamination. An additional area of concern for all types of plants was sanitizing floor mats and foot baths from which positive microbiological results were obtained. Contamination in wet areas was significantly greater than in dry areas of the plants (P<.05). Identification of the sources and conditions associated with these problematic bacterial pathogens is an important step in learning to control their incidence in dairy processing environments.

Effect of incubation temperature on repair of heat-injured Listeria in milk

Heat-injured Listeria species were examined for their ability to repair in pasteurized whole and 2% (fat) bovine milk. Listeria monocytogenes F5069 (serotype 4B) and F5027 (serotype 1/2a) and Listeria innocua CWD139 were heated at 55°C. After 20 min, 99% of the surviving population was injured as determined by their inability to grow in the presence of 4% NaCl. Bacterial cells were immediately suspended in sterile milk at a concentration of 102 to 103 per ml and incubated at 4, 10, 26 and 37°C. For all of the Listeria tested, repair at 4°C was initiated between days 8 and 10 and was complete between days 16 and 19; at 10°C, repair began immediately and was complete in 4 d; at 26 and 37°C, repair was complete by 13 and 9 h, respectively. The kinetics of repair were similar in whole and 2% (fat) milk. The relationship between the time required for repair and increasing temperature was nonlinear and indicated that repair of heat-injured Listeria in milk is highly sensitive to minor increases in temperature. Current Listeria detection techniques are not adequate for the detection of injured organisms. The public health consequences associated with failure to detect injured L. monocytogenes which subsequently repair in milk may be significant.

Behavior of Escherichia coli O157:H7 during the manufacture and aging of Gouda and stirred-curd Cheddar cheeses manufactured from raw milk.

This study was conducted to examine the fate of Escherichia coli O157:H7 during the manufacture and aging of Gouda and stirred-curd Cheddar cheeses made from raw milk. Cheeses were manufactured from unpasteurized milk experimentally contaminated with one of three strains of E. coli O157:H7 at an approximate population level of 20 CFU/ml. Samples of milk, whey, curd, and cheese were collected for enumeration of bacteria throughout the manufacturing and aging process. Overall, bacterial counts in both cheese types increased almost 10-fold from initial inoculation levels in milk to approximately 145 CFU/g found in cheeses on day 1. From this point, counts dropped significantly over 60 days to mean levels of 25 and 5 CFU/g in Cheddar and Gouda, respectively. Levels of E. coli O157:H7 fell and stayed below 5 CFU/g after an average of 94 and 108 days in Gouda and Cheddar, respectively, yet remained detectable after selective enrichment for more than 270 days in both cheese types. Changes in pathogen levels observed throughout manufacture and aging did not significantly differ by cheese type. In agreement with results of previous studies, our results suggest that the 60-day aging requirement alone is insufficient to completely eliminate levels of viable E. coli O157:H7 in Gouda or stirred-curd Cheddar cheese manufactured from raw milk contaminated with low levels of this pathogen.

Comparison of Infectious Dose of Listeria monocytogenes F5817 as Determined for Normal Versus Compromised C57B1/6J Mice

The infectious dose of Listeria monocytogenes F5817, a serotype 4b human patient isolate, was determined following oral challenge in normal and compromised C57B1/6J mice. In an attempt to mimic human populations previously shown to be at risk to ingestion of L. monocytogenes , groups of mice used in this study consisted of the following: mice pretreated with hydrocortisone acetate or cimetidine; pregnant mice (12-14 d gestation); or beige mutants of C57B1/6J mice (deficient in lysosome production within monocytes and granulocytes). Mice were gavaged with varying levels of L. monocytogenes suspended in sterile 11% non-fat milk solids (NFMS). Upon expiration, the spleen, liver, lung, and brain were aseptically removed from mice. Organs were plated on LPM agar, and colonies were enumerated and biochemically confirmed as L. monocytogenes . Mice were considered infected if L. monocytogenes was recovered from at least one of the examined organs. In normal resistant C57B1/6J mice, the infectious dose 50 (ID50) ranged from 3.24-4.55 log10 CFU. In comparison, the ID50 for mice treated with 2.5 mg hydrocortisone acetate/day for 3d prior to infection decreased to 0.41 log10 CFU (range -1.91-2.74 log10 CFU). Administration of 0.25 mg hydrocortisone acetate/day for 3d prior to infection resulted in an ID50 similar to that calculated for normal mice. The ID50 calculated for pregnant mice was 2.48 log10 CFU, a value not significantly different from that of normal control mice. The response of beige mutants and cimetidine treated mice was comparable to that of normal controls, with ID50 values of 4.00 and 3.30 log10 CFU, respectively.

Increased detection of acid-injured Escherichia coli O157:H7 in autoclaved apple cider by using nonselective repair on trypticase soy agar

Three different acid-resistant strains of Escherichia coli O157:H7 were inoculated individually and as a cocktail into sterile apple cider (pH 3.2) at a level of approximately 105 cells per ml and incubated at 2°C. Samples were plated on Trypticase soy agar (TSA), violet red bile agar (VRBA), sorbitol MacConkey agar (SMA), and Petrifilm E. coli count plates (Petrifilm) at 24-h intervals. Repair of acid-injured cells was assessed by surface plating cider samples on TSA and allowing a 2-h room-temperature incubation period followed by overlaying with double-strength VRBA or SMA. Since SMA is a surface plate medium, the repair procedure was modified by overlaying SMA with Trypticase soy broth after 2 h of room-temperature incubation. Populations of all three strains and the cocktail of strains decreased rapidly in apple cider and approached undetectable levels within 72 h. At 24 and 48 h, 98.4% and >99% of the E. coli populations were injured, respectively. Repair procedures significantly (α = 0.05) increased detection of E. coli O157:H7. After 72 h E. coli O157:H7 was not detected by using SMA and Petrifilm; however, it was detected using repair procedures. Although detection levels were increased with resuscitation procedures, the levels detected were still lower than those obtained using nonselective TSA. This research confirms the need for special recovery steps when analyzing acidic food products suspected of containing E. coli O157:H7.

Incidence and seasonal variation of Listeria species in bulk tank goat's milk.

Four hundred and fifty raw goats milk samples obtained from the bulk tanks of 39 goat farms were analyzed for Listeria spp. over a 1-year period. Modified versions of the U.S. Department of Agriculture Food Safety and Inspection Service (USDA-FSIS) and Food and Drug Administration (FDA) protocols were used for recovery of Listeria. Overall, 35 (7.8%) samples yielded Listeria spp. with Listeria monocytogenes identified in 17 of the 35 (3.8%) Listeria-positive samples. Listeria innocua was detected in 26 (5.8%) samples. Eight milk samples contained both L. monocytogenes and L. innocua. Milk samples from 18 of the 39 (46.2%) farms were positive for Listeria at least once during this 1-year study. The modified USDA-FSIS method, which used Listeria repair broth rather than University of Vermont (UVM) broth for primary enrichment followed by a 4-h nonselective incubation period, yielded more Listeria-positive samples (77.1%) than the FDA method (51.4%). All L. monocytogenes isolates belonged to serotypes 1 (62.6%) or 4 (37.4%). Moreover, five different Listeria ribotypes were identified from 34 selected L. monocytogenes isolates, 2 of which were deemed to be of clinical importance. Listeria isolation rates were markedly higher during winter (14.3%) and spring (10.4%) as compared to autumn (5.3%) and summer (0.9%) with these trends similar to those previously reported for cows milk.