Improving and Comparing Probiotic Plate Count Methods by Analytical Procedure Lifecycle Management

Probiotics are live microorganisms that confer a health benefit to the host when administered in adequate amounts. This definition links probiotic efficacy to microbial viability. The current gold standard assay for probiotic potency is enumeration using classical microbiology plating-based procedures, yielding results in colony-forming units (CFU). One drawback to plating-based procedures is high variability due to intrinsic and extrinsic uncertainties. These uncertainties make comparison between analytical procedures challenging. In this article, we provide tools to reduce measurement uncertainty and strengthen the reliability of probiotic enumerations by using analytical procedure lifecycle management (APLM). APLM is a tool that uses a step-by-step process to define procedure performance based on the concept that the reportable value (final CFU result) must be fit for its intended use. Once the procedure performance is defined, the information gathered through APLM can be used to evaluate and compare procedures. Here, we discuss the theory behind applying APLM and give practical information about its application to CFU enumeration procedures for probiotics using a simulated example and data set. Data collected in a manufacturer’s development laboratory is included to support application of the concept. Implementation of APLM can lead to reduced variability by identifying specific factors (e.g., the dilution step) with significant impact on the variability and providing insights to procedural modifications that lead to process improvement. Understanding and control of the analytical procedure is improved by using these tools. The probiotics industry can confidently apply the information and analytical results generated to make decisions about processes and formulation, including overage requirements. One benefit of this approach is that companies can reduce overage costs. More reliable procedures for viable cell count determinations will improve the quality evaluation of probiotic products, and hence manufacturing procedures, while ensuring that products deliver clinically demonstrated beneficial doses.

Improving End-User Trust in the Quality of Commercial Probiotic Products

In a rapidly growing global probiotic market, end-users have difficulty distinguishing between high quality and poor quality products. This ambiguity threatens the trust consumers and healthcare providers have in probiotic products. To address this problem, we recommend that companies undergo third-party evaluations to certify probiotic quality and label accuracy. In order to communicate about product quality to end-users, indication of certification on product labels is helpful, although not all manufacturers choose to use this approach. Herein we discuss: third-party certification, the process of setting standards for identity, purity, and quantification of probiotics; some emerging methodologies useful for quality assessment; and some technical challenges unique to managing quality of live microbial products. This review provides insights of an Expert Panel engaged in this process and aims to update the reader on relevant current scientific methodologies. Establishing validated methodologies for all aspects of quality assessment is an essential component of this process and can be facilitated by established organizations, such as United States Pharmacopeia. Emerging methodologies including whole genome sequencing and flow cytometry are poised to play important roles in these processes.

Bacillus cereus food poisoning and its toxins

The genus Bacillus includes members that demonstrate a wide range of diversity from physiology and ecological niche to DNA sequence and gene regulation. The species of most interest tend to be known for their pathogenicity and are closely linked genetically. Bacillus anthracis causes anthrax, and Bacillus thuringiensis is widely used for its insecticidal properties but has also been associated with foodborne disease. Bacillus cereus causes two types of food poisoning, the emetic and diarrheal syndromes, and a variety of local and systemic infections. Although in this review we provide information on the genus and a variety of species, the primary focus is on the B. cereus strains and toxins that are involved in foodborne illness. B. cereus produces a large number of potential virulence factors, but for the majority of these factors their roles in specific infections have not been established. To date, only cereulide and the tripartite hemolysin BL have been identified specifically as emetic and diarrheal toxins, respectively. Nonhemolytic enterotoxin, a homolog of hemolysin BL, also has been associated with the diarrheal syndrome. Recent findings regarding these and other putative enterotoxins are discussed.

Heterogeneity observed in the components of hemolysin BL, an enterotoxin produced by Bacillus cereus

Hemolysin BL (HBL), a diarrheal enterotoxin originally isolated from Bacillus cereus strain F837/76, is composed of three antigenically distinct proteins designated B, L1, and L2. All three components are required for biological activity. Here, we report antigenic and physical variations in HBL components produced by other B. cereus isolates. Reactions of partial identity were observed in double gel immunodiffusion assays using antibodies to highly purified B, L1, and L2 components of F837/76 and culture supernatants of strains F837/76 and S1C. Western blot analysis showed that F837/76 produced one 38-kDa B protein, one 38-kDa L1, and one 43-kDa L1 protein. In strain S1C, two B (38 and 42 kDa), two L1 (38 and 41 kDa), and one L1 (43 kDa) proteins were detected. Further Western blot analysis of 127 B. cereus isolates showed that 90 produced one or more of the three HBL components. Approximately half of these 90 isolates (43/90; 48%) produced protein profiles which differed from that of F837/76. A total of four B, two L1, and three L2 component profiles with proteins of different sizes were observed. Individual strains produced various combinations of single or multiple bands of each component. In addition, some strains produced only one or two of the three HBL components. The public health significance of these strains is unknown, as all three components are required for biological activity. The data presented here demonstrates a high degree of heterogeneity in HBL and provide the basis for further studies to characterize the variations in HBL and to determine the role of the variant components in pathogenicity.

Enterotoxic activity of hemolysin BL from Bacillus cereus

Bacillus cereus causes exotoxin-mediated diarrheal food poisoning. Hemolysin BL (HBL) is a well-characterized B. cereus toxin composed of three components (B, L1, and L2) that together possess hemolytic, cytotoxic, dermonecrotic, and vascular permeability activities. Here, we show that HBL causes fluid accumulation in ligated rabbit ileal loops at a dose of 5 micrograms of each component per loop. Maximal fluid responses occurred for combinations of all three components at > or = 25 micrograms of each component per loop. Individual components and binary combinations did not cause significant fluid accumulation at 25 micrograms of each component. Specific antisera to HBL components inhibited the fluid accumulation response of crude culture supernatant from B. cereus F837/76. These antisera were tested against an antiserum to a partially characterized multicomponent diarrheal toxin described previously by Thompson et al. (N. E. Thompson, M. J. Ketterhagen, M. S. Bergdoll, and E. J. Shantz, Infect. Immun. 43:887-894, 1984). Immunoblot and immunoprecipitation analyses indicate that HBL and that toxin are identical. These results confirm previous speculation that HBL is a tripartite enterotoxin that, as for all of its other known activities, requires all three components for maximal activity.

Growth and penetration of Salmonella enteritidis, Salmonella heidelberg and Salmonella typhimurium in eggs

Eggs and egg dishes are important vehicles for Salmonella infections. Salmonella enteritidis, Salmonella typhimurium and Salmonella heidelberg, which can be isolated from chicken ovaries and feces, have been implicated in approximately 50% of the foodborne salmonellosis outbreaks in the United States. In this study, the growth of these three organisms, inoculated into yolks and albumen, was compared at 4, 10 and 25 degrees C. Regardless of whether 10(2) cfu/g or 10(4) cfu/g was inoculated into the yolk or albumen, populations of all strains increased 3 logs or more in number in one day when incubated at 25 degrees C. Maximum numbers of Salmonella ranged from 10(8) to 10(10) cfu/g. All strains grew at 10 degrees C, but peak numbers were lower and occurred later than those at 25 degrees C. Populations of the three Salmonella strains inoculated into eggs stored at 4 degrees C grew sporadically; in some test groups populations declined. The potential for Salmonella in contaminated feces to establish in the interior of eggs was examined by monitoring shell penetration. At 25 degrees C, all three Salmonella strains penetrated the shell in 3 days, but at 4 degrees C, only S. typhimurium was found in one membrane sample. When hatchery conditions were simulated by incubating eggs at 35 degrees C for 30 min followed by storage at 4 degrees C, penetration was enhanced. Penetration was observed by day 1-3 when eggs were exposed to 10(4) cfu Salmonella/g feces. Increasing the inoculum to 10(6) cfu/g feces resulted in 50-75% of the contents of eggs to be contaminated by day 1. All Salmonella-positive samples were detected by enrichment. Results of this study indicate that S. enteritidis, S. typhimurium, or S. heidelberg present in feces can penetrate to the interior of eggs and grow during storage.

Variable colonization of chickens perorally inoculated with Escherichia coli O157:H7 and subsequent contamination of eggs

Challenging 1-day-old White Leghorn chicks perorally with 2.6 x 10(1) to 2.6 x 10(5) Escherichia coli O157:H7 bacteria per chick resulted in cecal colonization at all levels. Two of six chicks inoculated with only 2.6 x 10(1) E. coli O157:H7 bacteria carried 10(3) to 10(4) E. coli O157:H7 bacteria per g of cecal tissue when sacrificed 3 months postinoculation. E. coli O157:H7 colonization persisted at least 10 to 11 months when chicks were administered 10(8) E. coli O157:H7 bacteria. Eggs from five hens that were fecal shedders of E. coli O157:H7 until the termination of the study (10 to 11 months) were assayed for E. coli O157:H7. The organism was isolated from the shells of 14 of 101 (13.9%) eggs but not from the yolks and whites. Considering that chicks can be readily colonized by small populations of E. coli O157:H7 and continue to be long-term shedders, it is possible that chickens and hen eggs can serve as vehicles of this human pathogen.

Effect of trisodium phosphate on biofilm and planktonic cells of Campylobacter jejuni, Escherichia coli O157: H7, Listeria monocytogenes and Salmonella typhimurium

Trisodium phosphate (TSP) has been approved by the United States Department of Agriculture as a post-chill antimicrobial treatment for raw poultry. This study examines the effectiveness of TSP against planktonic (suspended) and biofilm (attached) cells of Campylobacter jejuni, Escherichia coli O157:H7, Listeria monocytogenes and Salmonella typhimurium at room temperature (RT) and 10 degrees C. At either temperature E. coli O157:H7 was the most sensitive to TSP treatments; 10(6) cfu/ml of planktonic or 10(5) cfu/cm2 of biofilm cells were eliminated by a 30 s treatment with 1% TSP. Campylobacter jejuni was slightly less sensitive. Listeria monocytogenes was the most resistant to the effect of TSP, requiring exposure to 8% TSP for 10 min (RT) or 20 min (10 degrees C) to reduce biofilm bacteria by at least one log. Biofilm cells of S. typhimurium and Listeria monocytogenes were more resistant than planktonic cells. Salmonella typhimurium was more sensitive to treatments using TSP at 10 degrees C than at RT. In contrast, L. monocytogenes was more resistant to TSP at 10 degrees C. Trisodium phosphate appears to be an effective treatment for reducing populations of C. jejuni, E. coli O157:H7 and S. typhimurium. This product has the potential to be used for reduction of bacterial counts on other food products besides raw poultry or on food and non-food contact surfaces.

Inhibition of Campylobacter jejuni colonization in chicks by defined competitive exclusion bacteria

Campylobacter enteritis in humans has been linked to consumption of chicken. Reducing Campylobacter jejuni colonization in chickens can potentially reduce Campylobacter infections in humans. In this study, the reduction of C. jejuni colonization in chicks by oral administration of defined competitive exclusion (CE) cultures, 2.5% dietary carbohydrates, or CE cultures and carbohydrates was examined. Prevention, elimination, or direct challenge of Campylobacter infection was simulated by administering treatments before, after, or at the same time as that of the Campylobacter inoculation. Additionally, the effect of maintaining high levels of protective bacteria was evaluated by administering CE cultures on days 1 and 4 (booster treatment). All treatments reduced C. jejuni colonization. Protection by aerobically grown CE cultures was not statistically different from that by anaerobically grown CE cultures. A combination of Citrobacter diversus 22, Klebsiella pneumoniae 23, and Escherichia coli 25 (CE 3) was the most effective CE treatment. Maintaining high numbers of CE isolates by administering CE boosters did not increase protection. The greatest reduction of Campylobacter colonization was observed in schemes to prevent or eliminate C. jejuni infection. C. jejuni was not detected in the ceca of birds receiving the prevention treatment, CE 3 with mannose, representing a 62% reduction in the colonization rate. The protection factor (PF), a value combining the colonization rate and the level of infection, for CE 3 with mannose was high (> 13.2). Fructo-oligosaccharides alone strongly prevented Campylobacter colonization. Only 8% of the chicks in this group were colonized, with a PF of > 14.3.(ABSTRACT TRUNCATED AT 250 WORDS)

Reduction of Campylobacter jejuni colonization of chicks by cecum-colonizing bacteria producing anti-C. jejuni metabolites

Cecum-colonizing bacteria were isolated from Campylobacter jejuni-free White Leghorn (Gallus domesticus) laying hens and screened for the ability to produce anti-C. jejuni metabolites. Nine isolates were obtained that possessed this characteristic. The peroral administration of the nine isolates as a mixture (ca. 10(9) per chick) to 1-day-old chicks was followed 1 week later by peroral inoculation of Campylobacter jejuni (ca. 10(9) per chick) to determine if the cecal isolates could protect chicks from colonization by campylobacters. The nine-strain mixture of cecal bacteria provided from 41 to 85% protection from C. jejuni colonization. The protective bacteria were reduced to a mixture of three strains on the basis of their ability to utilize mucin as a sole substrate for growth. These strains included Klebsiella pneumoniae 23, Citrobacter diversus 22, and Escherichia coli (O13:H-) 25. Four feeding trials with this three-strain mixture provided from 43 to 100% (average, 78%) protection from C. jejuni colonization. The dominant cecal bacterium of chicks treated with the three-strain mixture was consistently E. coli O13:H-. Similarly, three trials with only E. coli 25 used as the protective bacterium resulted in 49 to 72% (average, 59%) protection from C. jejuni colonization, with E. coli O13:H- being the dominant cecal bacterium in all cases. Although not completely effective, E. coli 25 substantially reduced the incidence of C. jejuni colonization of chicks. For all trials, fewer C. jejuni were present in the ceca of colonized chicks receiving the protective bacteria before exposure to C. jejuni than in chicks receiving only C. jejuni.(ABSTRACT TRUNCATED AT 250 WORDS)

Rates of Thermal Inactivation of Listeria monocytogenes in Beef and Fermented Beaker Sausage

Rates of thermal inactivation of a five-strain mixture of Listeria monocytogenes were determined in ground beef roast and fermented beaker sausage. Studies were also done on ground beef contaminated with L. monocytogenes Scott A from an experimentally infected cow. D-values for the five-strain mixture at 54.4, 57.2, 60.0, and 62.8°C were 22.4, 15.7, 4.47, and 2.56 min, respectively, for ground beef roast. D-values for fermented beaker sausage at 48.9, 51.7, 54.4, and 60.0°C were 98.6, 44.4, 20.1, 11.2, and 9.13 min, respectively. D-values for the single strain of L. monocytogenes mixture in ground beef from the infected cow were about two to four times less at equivalent temperatures than those of the five-strain L. monocytogenes mixture in ground beef roast. Results from the five-strain mixture indicate that L. monocytogenes is about four times more heat resistant that Salmonella in ground beef roast.

Fate of Listeria monocytogenes in tissues of experimentally infected cattle and in hard salami

Muscle, organ, and lymphoid tissues of four Holstein cows experimentally inoculated (intravenously) with Listeria monocytogenes were examined 2, 6, or 54 days postinoculation for the presence of the organism by direct plating and cold enrichment procedures. L. monocytogenes was isolated from 66% of the tissues sampled; 38% of the isolations were attributed to the use of cold enrichment. Isolation of the organism from muscle tissue was possible only with animals inoculated 2 days before slaughter. The fate of L. monocytogenes during the manufacture and storage of fermented hard salami made from this meat also was determined. Three sausage treatments were evaluated: (i) uninoculated control sausage, (ii) "naturally" contaminated sausage (NC) made from meat of an experimentally inoculated cow, and (iii) sausage made from beef inoculated with a laboratory culture of L. monocytogenes (I). Initial Listeria levels in NC and I sausage were 10(3) CFU/g in trial 1 and 10(4) CFU/g in trial 2. Numbers of L. monocytogenes decreased by approximately 1 log10 CFU/g during fermentation and decreased further during drying and refrigerated storage. Small numbers (less than or equal to 20 CFU/g) of L. monocytogenes were present in I and NC sausage at the end of 12 weeks of refrigerated storage; recovery of these organisms generally depended on the use of an enrichment procedure. The results indicate that L. monocytogenes does not multiply during the fermentation and drying processes typical of hard salami manufacture but that survival may occur if the organism is initially present at greater than or equal to 10(3) CFU/g.

Isolation of Escherichia coli O157:H7 from retail fresh meats and poultry

A total of 896 samples of retail fresh meats and poultry was assayed for Escherichia coli serogroup O157:H7 by a hydrophobic grid membrane filter-immunoblot procedure developed specifically to isolate the organism from foods. The procedure involves several steps, including selective enrichment, filtration of enrichment culture through hydrophobic grid membrane filters, incubation of each filter on nitrocellulose paper on selective agar, preparation of an immunoblot (by using antiserum to E. coli O157:H7 culture filtrate) of each nitrocellulose paper, selection from the filters of colonies which corresponded to immunopositive sites on blots, screening of isolates by a Biken test for precipitin lines from metabolites and antiserum to E. coli O157:H7 culture filtrate, and confirmation of isolates as Vero cell cytotoxic E. coli O157:H7 by biochemical, serological, and Vero cell cytotoxicity tests. E. coli O157:H7 was isolated from 6 (3.7%) of 164 beef, 4 (1.5%) of 264 pork, 4 (1.5%) of 263 poultry, and 4 (2.0%) of 205 lamb samples. One of 14 pork samples and 5 of 17 beef samples contaminated with the organism were from Calgary, Alberta, Canada, grocery stores, whereas all other contaminated samples were from Madison, Wis., retail outlets. This is the first report of the isolation of E. coli O157:H7 from food other than ground beef, and results indicate that the organism is not a rare contaminant of fresh meats and poultry.

Comparison of Procedures for Isolating Listeria monocytogenes in Soft, Surface-Ripened Cheese

Ninety samples of soft, surface-ripened cheese from a lot previously identified to contain Listeria were assayed for Listeria monocytogenes by three procedures. These included: (a) cold enrichment, (b) the Food and Drug Administration enrichment procedure, and (c) the selective enrichment procedure of Doyle and Schoeni (Appl. Environ. Microbiol. 15:1127, 1986). L. monocytogenes was isolated from 41 of the 90 cheese samples. The organism was isolated from only 9 of the 41 L. monocytogenes -positive samples by more than one procedure. Most isolations (21) were made by the cold enrichment procedure, with 16 and 13 isolations made by the FDA and Doyle-Schoeni procedures, respectively. In most instances, the organism was isolated from a cheese sample by only one procedure.

Selective-enrichment procedure for isolation of Listeria monocytogenes from fecal and biologic specimens

A selective-enrichment procedure (SEP) was developed to isolate Listeria monocytogenes from fecal and biologic specimens. This procedure was compared with direct plating with McBride listeria agar and 2-, 4-, and 8-week cold-enrichment procedures in recovering L. monocytogenes from mouse fecal, liver, and brain specimens. Although the SEP occasionally did not isolate the organism from specimens proved positive by the other procedures, the SEP isolated L. monocytogenes from about two and five times as many specimens as the cold-enrichment and direct-plating procedures, respectively.

Isolation of Campylobacter jejuni from retail mushrooms

Campylobacter jejuni was isolated from 3 (1.5%) of 200 retail, polyvinyl chloride film-wrapped, fresh mushrooms. These results indicate that fresh mushrooms may indeed be a source of C. jejuni and support previously reported epidemiological data (Seattle-King County Department of Public Health, Surveillance of the Flow of Salmonella and Campylobacter in a Community, 1984) which revealed an an elevated relative risk of developing campylobacter enteritis in individuals who consume mushrooms.

Colonization of chicken cecae by Escherichia coli associated with hemorrhagic colitis

Bacterial enumeration, histologic examination, and immunoperoxidase staining demonstrated the ability of an Escherichia coli strain associated with hemorrhagic colitis (serotype O157:H7) to colonize chicken cecae for up to 90 days postinoculation after a peroral challenge at 1 day of age. The bacteria induced mild, transient, mucous membrane damage confined to the proximal cecae of healthy, normal-appearing chickens, principally at 14 to 28 days postinoculation. Attachment, effacement, and penetration of the cecal surface epithelium by E. coli O157:H7 were observed. With the exception of splenic, hepatic, and cecal tonsil immune-related changes and cecal damage and colonization, no other organ systems or portions of the gastrointestinal tract were affected by the bacteria. Bacterial counts indicated that E. coli O157:H7 was predominantly present in the cecae (often at levels greater than 10(6) CFU/g of tissue and contents) and to a lesser extent in the colon. Our results suggest that E. coli O157:H7 colonizes chicken cecae and is passed through the colon with fecal excrement. The ability of this organism to colonize chicken cecae indicates that chickens may serve as hosts and possibly as reservoirs for E. coli O157:H7.

Survival and growth characteristics of Escherichia coli associated with hemorrhagic colitis

Escherichia coli O157:H7 in ground beef was more sensitive to heat than salmonellae, but survived for 9 months at -20 degrees C with little change in number. The organisms grew well in Trypticase soy broth (BBL Microbiology Systems) between 30 and 42 degrees C, with 37 degrees C being optimal for growth. E. coli O157:H7 grew poorly in the temperature range (44 to 45.5 degrees C) generally used for recovery of E. coli from foods.

Fate of Salmonella typhimurium and Staphylococcus aureus in Meat Salads Prepared with Mayonnaise

Staphylococcus aureus and Salmonella typhimurium were tested for their ability to survive and to multiply in meat salads prepared with different concentrations of mayonnaise and held at 4, 22, and 32 C. When mayonnaise was added to meat salads in amounts recommended by recipes from a reputable cookbook, it inactivated a substantial portion of the initial population of both S. aureus (30-60%) and S. typhimurium (20-25%). Salads that were refrigerated at 4 C for 24 h evidenced very little growth of either organism whether mayonnaise was present or not. Storing salads at 22 or 32 C for 5 h resulted in <1.0 log₁₀ increase of either organism with the greatest increase occurring in salads containing no mayonnaise. Mayonnaise retarded but did not prevent the growth of S. aureus or S. typhimurium in salads stored at 22 or 32 C for 24 h. Increasing the concentration of mayonnaise in salads increased the degree to which growth of these organisms was delayed. Contrary to popular belief, the presence of mayonnaise in meat salads tends to retard rather than enhance growth of food-borne pathogens. However, addition of mayonnaise should not be considered a substitute for refrigeration for preserving meat salads from the growth of food-borne pathogens.