The Influence of Heteroresistance on Minimum Inhibitory Concentration, Investigated Using Weak-Acid Stress in Food Spoilage Yeasts.

Populations of microbial cells may resist environmental stress by maintaining a high population-median resistance (IC50) or, potentially, a high variability in resistance between individual cells (heteroresistance); where heteroresistance would allow certain cells to resist high stress, provided the population was sufficiently large to include resistant cells. This study sets out to test the hypothesis that both IC50 and heteroresistance may contribute to conventional minimal inhibitory concentration (MIC) determinations, using the example of spoilage-yeast resistance to the preservative sorbic acid. Across a panel of 26 diverse yeast species, both heteroresistance and particularly IC50 were positively correlated with predicted MIC. A focused panel of 29 different isolates of a particular spoilage yeast was also examined (isolates previously recorded as Zygosaccharomyces bailii, but genome resequencing revealing that several were in fact hybrid species, Z. parabailii and Z. pseudobailii). Applying a novel high-throughput assay for heteroresistance, it was found that IC50 but not heteroresistance was positively correlated with predicted MIC when considered across all isolates of this panel, but the heteroresistance-MIC interaction differed for the individual Zygosaccharomyces subspecies. Z. pseudobailii exhibited higher heteroresistance than Z. parabailii whereas the reverse was true for IC50, suggesting possible alternative strategies for achieving high MIC between subspecies. This work highlights the limitations of conventional MIC measurements due to the effect of heteroresistance in certain organisms, as the measured resistance can vary markedly with population (inoculum) size. IMPORTANCE Food spoilage by fungi is a leading cause of food waste, with specialized food spoilage yeasts capable of growth at preservative concentrations above the legal limit, in part due to heteroresistance allowing small subpopulations of cells to exhibit extreme preservative resistance. Whereas heteroresistance has been characterized in numerous ecological contexts, measuring this phenotype systematically and assessing its importance are not encompassed by conventional assay methods. The development here of a high-throughput method for measuring heteroresistance, amenable to automation, addresses this issue and has enabled characterization of the contribution that heteroresistance may make to conventional MIC measurements. We used the example of sorbic acid heteroresistance in spoilage yeasts like Zygosaccharomyces spp., but the approach is relevant to other fungi and other inhibitors, including antifungals. The work shows how median resistance, heteroresistance, and inoculum size should all be considered when selecting appropriate inhibitor doses in real-world antimicrobial applications such as food preservation.

Variability and uncertainty analysis of the cross-contamination ratios of salmonella during pork cutting.

The transfer ratio of bacteria from one surface to another is often estimated from laboratory experiments and quantified by dividing the expected number of bacteria on the recipient surface by the expected number of bacteria on the donor surface. Yet, the expected number of bacteria on each surface is uncertain due to the limited number of colonies that are counted and/or samples that can be analyzed. The expected transfer ratio is, therefore, also uncertain and its estimate may exceed 1 if real transfer is close to 100%. In addition, the transferred fractions vary over experiments but it is unclear, using this approach, how to combine uncertainty and variability into one estimate for the transfer ratio. A Bayesian network model was proposed that allows the combination of uncertainty within one experiment and variability over multiple experiments and prevents inappropriate values for the transfer ratio. Model functionality was shown using data from a laboratory experiment in which the transfer of Salmonella was determined from contaminated pork meat to a butcher's knife, and vice versa. Recovery efficiency of bacteria from both surfaces was also determined and accounted for in the analysis. Transfer ratio probability distributions showed a large variability, with a mean value of 0.19 for the transfer of Salmonella from pork meat to the knife and 0.58 for the transfer of Salmonella from the knife to pork meat. The proposed Bayesian model can be used for analyzing data from similar study designs in which uncertainty should be combined with variability.

Critical comparison of statistical methods for quantifying variability and uncertainty of microbial responses from experimental data.

Variability and uncertainty are important factors for quantitative microbiological risk assessment (QMRA). In this context, variability refers to inherent sources of variation, whereas uncertainty refers to imprecise knowledge or lack of it. In this work we compare three statistical methods to estimate variability in the kinetic parameters of microbial populations: mixed-effect models, multilevel Bayesian models, and a simplified algebraic method previously suggested. We use two case studies that analyse the influence of three levels of variability: (1) between-strain variability (different strains of the same species), (2) within-strain variability (biologically independent reproductions of the same strain) and, at the most nested level, (3) experimental variability (species independent technical lab variability resulting in uncertainty about the population characteristic of interest) on the growth and inactivation of Listeria monocytogenes. We demonstrate that the algebraic method, although relatively easy to use, overestimates the contribution of between-strain and within-strain variability due to the propagation of experimental variability in the nested experimental design. The magnitude of the bias is proportional to the variance of the lower levels and inversely proportional to the number of repetitions. This bias was very relevant in the case study related to growth, whereas for the case study on inactivation the resulting insights in variability were practically independent of the method used. The mixed-effects model and the multilevel Bayesian models calculate unbiased estimates for all levels of variability in all the cases tested. Consequently, we recommend using the algebraic method for initial screenings due to its simplicity. However, to obtain parameter estimates for QMRA, the more complex methods should generally be used to obtain unbiased estimates.

Validation of EN ISO 6579-1 - Microbiology of the food chain - Horizontal method for the detection, enumeration and serotyping of Salmonella - Part 1 detection of Salmonella spp.

The European and International Standard method for the detection of Salmonella spp. in samples from the primary production stage, EN ISO 6579:2002/Amd.1:2007, was validated by an interlaboratory study in the frame of Mandate M/381, ordered by the European Commission and accepted by the European Standardisation Organisation (CEN). In addition to this study, results from two interlaboratory studies organised earlier by the European Union Reference Laboratory (EURL) for Salmonella were used for determination of the performance characteristics. Parallel to the performance evaluation for the Mandate, the revision of EN ISO 6579:2002 started. Part of this revision was the incorporation of the standardised method for detection of Salmonella in samples from the primary production stage (EN ISO 6579:2002/Amd.1:2007) and its performance characteristics in the new part 1 of EN ISO 6579. The 2002 version of EN ISO 6579 already contained performance characteristics for the detection of Salmonella in food samples, but LOD50 values (contamination level at which 50% of the samples are found positive) were not yet included. To be in line with the performance characteristics determined for detection of Salmonella spp. in samples from the primary production stage, LOD50 values for detection of Salmonella in food samples were calculated from the raw data of the validation studies performed in 2000. In this paper, the performance characteristics of EN ISO 6579-1:2017 are determined based not only on the results of the interlaboratory study carried out in 2013 under the Mandate, but also on several other interlaboratory studies. These performance characteristics consist of specificity, sensitivity and LOD50.

Next generation microbiological risk assessment-Potential of omics data for hazard characterisation.

According to the World Health Organization estimates in 2015, 600 million people fall ill every year from contaminated food and 420,000 die. Microbial risk assessment (MRA) was developed as a tool to reduce and prevent risks presented by pathogens and/or their toxins. MRA is organized in four steps to analyse information and assist in both designing appropriate control options and implementation of regulatory decisions and programs. Among the four steps, hazard characterisation is performed to establish the probability and severity of a disease outcome, which is determined as function of the dose of toxin and/or pathogen ingested. This dose-response relationship is subject to both variability and uncertainty. The purpose of this review/opinion article is to discuss how Next Generation Omics can impact hazard characterisation and, more precisely, how it can improve our understanding of variability and limit the uncertainty in the dose-response relation. The expansion of omics tools (e.g. genomics, transcriptomics, proteomics and metabolomics) allows for a better understanding of pathogenicity mechanisms and virulence levels of bacterial strains. Detection and identification of virulence genes, comparative genomics, analyses of mRNA and protein levels and the development of biomarkers can help in building a mechanistic dose-response model to predict disease severity. In this respect, systems biology can help to identify critical system characteristics that confer virulence and explain variability between strains. Despite challenges in the integration of omics into risk assessment, some omics methods have already been used by regulatory agencies for hazard identification. Standardized methods, reproducibility and datasets obtained from realistic conditions remain a challenge, and are needed to improve accuracy of hazard characterisation. When these improvements are realized, they will allow the health authorities and government policy makers to prioritize hazards more accurately and thus refine surveillance programs with the collaboration of all stakeholders of the food chain.

Phenotypic Prediction: Linking in vitro Virulence to the Genomics of 59 Salmonella enterica Strains.

The increased availability of whole-genome-sequencing techniques generates a wealth of DNA data on numerous organisms, including foodborne pathogens such as Salmonella. However, how these data can be used to improve microbial risk assessment and understanding of Salmonella epidemiology remains a challenge. The aim of this study was to assess variability in in vitro virulence and genetic characteristics between and within different serovars. The phenotypic behavior of 59 strains of 32 different Salmonella enterica serovars from animal, human and food origin was assessed in an in vitro gastro-intestinal tract (GIT) system and they were analyzed for the presence of 233 putative virulence genes as markers for phenotypic prediction. The probability of in vitro infection, P(inf), defined as the fraction of infectious cells passing from inoculation to host cell invasion at the last stage of the GIT system, was interpreted as the in vitro virulence. Results showed that the (average) P(inf) of Salmonella serovars ranged from 5.3E-05 (S. Kedougou) to 5.2E-01 (S. Typhimurium). In general, a higher P(inf) on serovar level corresponded to higher reported human incidence from epidemiological reporting data. Of the 233 virulence genes investigated, only 101 showed variability in presence/absence among the strains. In vitro P(inf) was found to be positively associated with the presence of specific plasmid related virulence genes (mig-5, pef, rck, and spv). However, not all serovars with a relatively high P(inf), > 1E-02, could be linked with these specific genes. Moreover, some outbreak related strains (S. Heidelberg and S. Thompson) did not reveal this association with P(inf). No clear association with in vitro virulence P(inf) was identified when grouping serovars with the same virulence gene profile (virulence plasmid, Typhoid toxin, peg operon and stk operon). This study shows that the in vitro P(inf) variation among individual strains from the same serovar is larger than that found between serovars. Therefore, ranking P(inf) of S. enterica on serovar level alone, or in combination with a serovar specific virulence gene profile, cannot be recommended. The attribution of single biological phenomena to individual strains or serovars is not sufficient to improve the hazard characterization for S. enterica. Future microbial risk assessments, including virulence gene profiles, require a systematic approach linked to epidemiological studies rather than revealing differences in characteristics on serovar level alone.

Probabilities in norovirus outbreak diagnosis.

BACKGROUND: Noroviruses are recognized as the most common cause of outbreaks of acute gastroenteritis. Yet, diagnostic testing for norovirus is based mostly on RNA detection by RT-PCR, which is not widely available. While antigen detection tests (ELISAs) are easier to perform, they are in general less sensitive. OBJECTIVES: Our aim was to provide a scientific basis for declaring norovirus as the causative agent of an outbreak of acute gastroenteritis. STUDY DESIGN: Statistical analysis used binomial distribution to determine the minimal number of positive samples, and the probability of detecting the required number of positive samples, for different tests, required to assign norovirus as the causative agent of an outbreak of acute gastroenteritis. RESULTS: For either a standard RT-PCR or a commercially available ELISA, finding only 1 sample positive out of 2, 3 or 4 samples is sufficient to assign norovirus as the causative agent of an outbreak of acute gastroenteritis. However, when ELISA is used, the probability of detecting this required minimum number of positive samples is low when small numbers of samples are tested (57% when 2 samples are tested; 72% when 3 samples are tested). In order to reach a 90% probability of detecting a norovirus outbreak (false negativity at outbreak level <10%), at least 3 samples should be tested using RT-PCR, and 6 samples when using an ELISA. CONCLUSIONS: The sensitivity for NoV outbreak diagnosis will increase from 57% to 92%, or from 84% to 96%, for ELISA or RT-PCR respectively, when sample size increases from 2 to 6. Thus, using ELISA instead of RT-PCR for the detection of norovirus in stool samples will result in considerable numbers of false negative outbreaks unless a minimum of 6 samples are tested per outbreak.

Quantification of Salmonella Survival and Infection in an In vitro Model of the Human Intestinal Tract as Proxy for Foodborne Pathogens.

Different techniques are available for assessing differences in virulence of bacterial foodborne pathogens. The use of animal models or human volunteers is not expedient for various reasons; the use of epidemiological data is often hampered by lack of crucial data. In this paper, we describe a static, sequential gastrointestinal tract (GIT) model system in which foodborne pathogens are exposed to simulated gastric and intestinal contents of the human digestive tract, including the interaction of pathogens with the intestinal epithelium. The system can be employed with any foodborne bacterial pathogens. Five strains of Salmonella Heidelberg and one strain of Salmonella Typhimurium were used to assess the robustness of the system. Four S. Heidelberg strains originated from an outbreak, the fifth S. Heidelberg strain and the S. Typhimurium strain originated from routine meat inspections. Data from plate counts, collected for determining the numbers of surviving bacteria in each stage, were used to quantify both the experimental uncertainty and biological variability of pathogen survival throughout the system. For this, a hierarchical Bayesian framework using Markov chain Monte Carlo (MCMC) was employed. The model system is able to distinguish serovars/strains for in vitro infectivity when accounting for within strain biological variability and experimental uncertainty.

Comparative Exposure Assessment of ESBL-Producing Escherichia coli through Meat Consumption.

The presence of extended-spectrum ß-lactamase (ESBL) and plasmidic AmpC (pAmpC) producing Escherichia coli (EEC) in food animals, especially broilers, has become a major public health concern. The aim of the present study was to quantify the EEC exposure of humans in The Netherlands through the consumption of meat from different food animals. Calculations were done with a simplified Quantitative Microbiological Risk Assessment (QMRA) model. The model took the effect of pre-retail processing, storage at the consumers home and preparation in the kitchen (cross-contamination and heating) on EEC numbers on/in the raw meat products into account. The contribution of beef products (78%) to the total EEC exposure of the Dutch population through the consumption of meat was much higher than for chicken (18%), pork (4.5%), veal (0.1%) and lamb (0%). After slaughter, chicken meat accounted for 97% of total EEC load on meat, but chicken meat experienced a relatively large effect of heating during food preparation. Exposure via consumption of filet americain (a minced beef product consumed raw) was predicted to be highest (61% of total EEC exposure), followed by chicken fillet (13%). It was estimated that only 18% of EEC exposure occurred via cross-contamination during preparation in the kitchen, which was the only route by which EEC survived for surface-contaminated products. Sensitivity analysis showed that model output is not sensitive for most parameters. However, EEC concentration on meat other than chicken meat was an important data gap. In conclusion, the model assessed that consumption of beef products led to a higher exposure to EEC than chicken products, although the prevalence of EEC on raw chicken meat was much higher than on beef. The (relative) risk of this exposure for public health is yet unknown given the lack of a modelling framework and of exposure studies for other potential transmission routes.

Emetic toxin-producing strains of Bacillus cereus show distinct characteristics within the Bacillus cereus group.

One hundred representative strains of Bacillus cereus were selected from a total collection of 372 B. cereus strains using two typing methods (RAPD and FT-IR) to investigate if emetic toxin-producing hazardous B. cereus strains possess characteristic growth and heat resistance profiles. The strains were classified into three groups: emetic toxin (cereulide)-producing strains (n=17), strains connected to diarrheal foodborne outbreaks (n=40) and food-environment strains (n=43), these latter not producing the emetic toxin. Our study revealed a shift in growth limits towards higher temperatures for the emetic strains, regardless of their origin. None of the emetic toxin-producing strains were able to grow below 10 degrees Celsius. In contrast, 11% (9 food-environment strains) out of the 83 non-emetic toxin-producing strains were able to grow at 4 degrees Celsius and 49% at 7 degrees Celsius (28 diarrheal and 13 food-environment strains). non-emetic toxin-producing strains. All emetic toxin-producing strains were able to grow at 48 degrees Celsius, but only 39% (16 diarrheal and 16 food-environment strains) of the non-emetic toxin-producing strains grew at this temperature. Spores from the emetic toxin-producing strains showed, on average, a higher heat resistance at 90 degrees Celsius and a lower germination, particularly at 7 degrees Celsius, than spores from the other strains. No difference between the three groups in their growth kinetics at 24 degrees Celsius, 37 degrees Celsius, and pH 5.0, 7.0, and 8.0 was observed. Our survey shows that emetic toxin-producing strains of B. cereus have distinct characteristics, which could have important implication for the risk assessment of the emetic type of B. cereus caused food poisoning. For instance, emetic strains still represent a special risk in heat-processed foods or preheated foods that are kept warm (in restaurants and cafeterias), but should not pose a risk in refrigerated foods.

First step in using molecular data for microbial food safety risk assessment; hazard identification of Escherichia coli O157:H7 by coupling genomic data with in vitro adherence to human epithelial cells.

The potential for using whole genome sequencing (WGS) data in microbiological risk assessment (MRA) has been discussed on several occasions since the beginning of this century. Still, the proposed heuristic approaches have never been applied in a practical framework. This is due to the non-trivial problem of mapping microbial information consisting of thousands of loci onto a probabilistic scale for risks. The paradigm change for MRA involves translation of multidimensional microbial genotypic information to much reduced (integrated) phenotypic information and onwards to a single measure of human risk (i.e. probability of illness). In this paper a first approach in methodology development is described for the application of WGS data in MRA; this is supported by a practical example. That is, combining genetic data (single nucleotide polymorphisms; SNPs) for Shiga toxin-producing Escherichia coli (STEC) O157 with phenotypic data (in vitro adherence to epithelial cells as a proxy for virulence) leads to hazard identification in a Genome Wide Association Study (GWAS). This application revealed practical implications when using SNP data for MRA. These can be summarized by considering the following main issues: optimum sample size for valid inference on population level, correction for population structure, quantification and calibration of results, reproducibility of the analysis, links with epidemiological data, anchoring and integration of results into a systems biology approach for the translation of molecular studies to human health risk. Future developments in genetic data analysis for MRA should aim at resolving the mapping problem of processing genetic sequences to come to a quantitative description of risk. The development of a clustering scheme focusing on biologically relevant information of the microbe involved would be a useful approach in molecular data reduction for risk assessment.

Microbiological risk from minimally processed packaged salads in the Dutch food chain.

The objective of this study was to evaluate the microbial hazard associated with the consumption of mixed salads produced under standard conditions. The presence of Salmonella, Campylobacter spp., and Escherichia coli O157 in the Dutch production chain of mixed salads was determined. Microbial prevalence and concentration data from a microbiological surveillance study were used as inputs for the quantitative microbial risk assessment. Chain logistics, production figures, and consumption patterns were combined with the survey data for the risk assessment chain approach. The results of the sample analysis were used to track events from contamination through human illness. Wide 95% confidence intervals around the mean were found for estimated annual numbers of illnesses resulting from the consumption of mixed salads contaminated with Salmonella Typhimurium DT104 (0 to 10,300 cases), Campylobacter spp. (0 to 92,000 cases), or E. coli (0 to 800 cases). The main sources of uncertainty are the lack of decontamination data (i.e., produce washing during processing) and an appropriate dose-response relationship.

Prevalence and concentration of bacterial pathogens in raw produce and minimally processed packaged salads produced in and for the Netherlands.

Recent outbreaks with vegetable or fruits as vehicles have raised interest in the characterization of the public health risk due to microbial contamination of these commodities. Because qualitative and quantitative data regarding prevalence and concentration of various microbes are lacking, we conducted a survey to estimate the prevalence and contamination level of raw produce and the resulting minimally processed packaged salads as sold in The Netherlands. A dedicated sampling plan accounted for the amount of processed produce in relation to the amount of products, laboratory capacity, and seasonal influences. Over 1,800 samples of produce and over 1,900 samples of ready-to-eat mixed salads were investigated for Salmonella enterica serovars, Campylobacter spp., Escherichia coli O157, and Listeria monocytogenes. The overall prevalence in raw produce varied between 0.11% for E. coli O157 and L. monocytogenes and 0.38% for Salmonella. Prevalence point estimates for specific produce/pathogen combinations ranged for Salmonella from 0.53% in iceberg lettuce to 5.1% in cucumber. For Campylobacter, this ranged from 0.83% in endive to 2.7% in oak tree lettuce. These data will be used to determine the public health risk posed by the consumption of ready-to-eat mixed salads in The Netherlands.

A quantitative approach towards a better understanding of the dynamics of Salmonella spp. in a pork slaughter-line.

Pork contributes significantly to the public health disease burden caused by Salmonella infections. During the slaughter process pig carcasses can become contaminated with Salmonella. Contamination at the slaughter-line is initiated by pigs carrying Salmonella on their skin or in their faeces. Another contamination route could be resident flora present on the slaughter equipment. To unravel the contribution of these two potential sources of Salmonella a quantitative study was conducted. Process equipment (belly openers and carcass splitters), faeces and carcasses (skin and cutting surfaces) along the slaughter-line were sampled at 11 sampling days spanning a period of 4 months. Most samples taken directly after killing were positive for Salmonella. On 96.6% of the skin samples Salmonella was identified, whereas a lower number of animals tested positive in their rectum (62.5%). The prevalence of Salmonella clearly declined on the carcasses at the re-work station, either on the cut section or on the skin of the carcass or both (35.9%). Throughout the sampling period of the slaughter-line the total number of Salmonella per animal was almost 2 log lower at the re-work station in comparison to directly after slaughter. Seven different serovars were identified during the study with S. Derby (41%) and S. Typhimurium (29%) as the most prominent types. A recurring S. Rissen contamination of one of the carcass splitters indicated the presence of an endemic 'house flora' in the slaughterhouse studied. On many instances several serotypes per individual sample were found. The enumeration of Salmonella and the genotyping data gave unique insight in the dynamics of transmission of this pathogen in a slaughter-line. The data of the presented study support the hypothesis that resident flora on slaughter equipment was a relevant source for contamination of pork.

Circulation of group 2 coronaviruses in a bat species common to urban areas in Western Europe.

Fecal samples of 211 bats representing 13 different bat species from 31 locations in the Netherlands were analyzed for the presence of coronaviruses (CoV) using a genus-wide reverse transcription (RT)-polymerase chain reaction. CoVs are known for their high potential for interspecies transmission, including zoonotic transmission with bats as reservoir hosts. For the first time, a group 2 CoV was found in a bat, Pipistrellus pipistrellus, in Europe. This is of particular interest for public health as the reservoir host is a species that is common to urban areas in most of Europe and notorious for its close interactions with humans. Four verspertilionid bat species were found to excrete group 1 CoVs, viz. Myotis daubentonii, M. dasycneme, P. pipistrellus, and Nyctalus noctula. The last species is a newly identified reservoir. The overall prevalence was 16.9% and positive bats were found at multiple widespread locations. The circulating group 1 CoV lineages were rather species associated than location associated.