Transfer and Decontamination of S. aureus in Transmission Routes Regarding Hands and Contact Surfaces.

Hand hygiene, cleaning and disinfection are pre-requirements for hygiene management in hospital settings and the food industry. In order to facilitate risk management, different contamination scenarios and interventions need to be evaluated. In the present study data on transfer rates and reductions of Staphylococcus aureus were provided in an experimental set-up using artificial skin. Using this methodology, test persons were not exposed with pathogenic bacteria. An exposure assessment model was developed and applied to evaluate different contamination routes and hygiene interventions. The transfer rates of S. aureus from inoculated VITRO-SKIN® to fomites were calculated from blotting series. The VITRO-SKIN® was more prone to spread bacteria than fomites. When different surfaces were cleaned, the reduction of S. aureus varied between <1 and 7 log CFU. It could not be concluded that a certain coupon material, cleaning agent, cleaning wipe, soiling or humidity consistently resulted in a high or low reduction of S. aureus. The reduction of S. aureus and E. coli during hand washing was evaluated on artificial skin, VITRO-SKIN®. The reduction of E. coli on VITRO-SKIN® was similar to the log reduction obtained when washing human hands. The S. aureus count on a human hand was both calculated in different scenarios describing different contamination routes starting from a contaminated hand using the exposure assessment model, and measured on an experimental setup using VITRO-SKIN® for validation. A linear relationship was obtained between the analysed level of S. aureus and the calculated level. However, the calculated levels of S. aureus on the VITRO-SKIN® in the scenarios were 1-1.5 log lower than the analysed level. One of the scenarios was used to study the effect of interventions like hand washing and cleaning of surfaces.

Extended staphylococcal enterotoxin D expression in ham products.

Staphylococcal enterotoxin D (SED) is one of the most frequently recovered enterotoxins in staphylococcal food poisoning (SFP) outbreaks. The expression and production of SED were investigated in three ham products, i.e. boiled ham, smoked ham and dry-cured Serrano ham incubated at room temperature for seven days. Staphylococcus aureus was also, as a reference, grown in cultivation broth during optimal growth conditions for seven days. In boiled and smoked ham, continuous sed expression was observed throughout the incubation period with a second increase in sed expression found after five days of incubation. In smoked ham, nine times less SED per colony-forming unit of S. aureus was detected than in boiled ham. In boiled ham, the SED levels unpredictably decreased after three days of incubation. In the Serrano ham, SED was detected after five days of incubation although S. aureus growth was poor and sed expression was too low to determine. After five days of incubation, all three products contained enough SED to cause SFP. These results show that the specific production levels of SED vary in the different ham products, and that toxin production was in part uncoupled from bacterial growth.

Prolonged expression and production of Staphylococcus aureus enterotoxin A in processed pork meat.

The bacteriophage-encoded staphylococcal enterotoxin A (SEA) is the toxin most frequently reported to be involved in staphylococcal food poisoning. In this study, sea expression and SEA formation were studied in four processed pork products: boiled ham, hot-smoked ham, Serrano ham (dry-cured Spanish ham) and black pepper salami. The products were selected because of their differences in intrinsic factors. As a reference, Staphylococcus aureus was cultivated under favorable planktonic growth conditions. Expression was mainly linked to bacterial growth for both meat products and broth cultures. In liquid broth, however, the relative level of sea mRNA peaked in the late exponential phase and then rapidly declined, while in the meat products allowing immediate growth, i.e. boiled and smoked ham, active sea expression occurred throughout the incubation period of seven days. Lower levels of sea mRNA and SEA were found in smoked ham compared to boiled ham, although viable counts of S. aureus on the two products were similar. Furthermore, the SEA concentration in the boiled ham reached a maximum after three days of incubation and then unpredictably decreased. In the Serrano ham, no increase in cell number was observed until day seven, and sea expression and extracellular SEA could only be detected on days five and seven. Finally, the black pepper salami with low pH and competing microbiota proved to be a difficult environment for the survival of S. aureus. The molecular mechanism behind the behaviour of S. aureus SEA expression is discussed in connection to the life-cycle of the SEA-encoding bacteriophage and the microbial communities in these pork products.

Use of the selective agar medium CREAD for monitoring the level of airborne spoilage moulds in cheese production.

It was investigated if a selective medium for common cheese spoiling moulds (CREAD) could give more relevant information than a general mould medium in hygienic air-sampling in cheese factories. A total of 126 air-samples were taken in six Nordic cheese factories using the general mould medium DG18 and CREAD. The level and genera of air-borne mould was determined. Identification to species-level was performed for a selection of samples. In five cheese factories the mycobiota was dominated by Penicillium spp. and in one cheese factory by Cladosporium spp. The concentration of air-borne moulds varied between the cheese factories ranging from 1 to 270 cfu/m3 on DG18 with a median value of 17. The number of mould colonies was in general lower at CREAD. Identification indicated that CREAD supported growth of common spoilage moulds for cheese, such as Penicillium palitans and P. commune. The mycobiota on DG18 also consisted of moulds not commonly associated with spoilage of cheese, such as Cladosporium spp., P. brevicompactum and P. chrysogenum. Contamination of cheese with mould is periodically a problem in production of semi-hard cheese and the level of air-borne mould is therefore routinely monitored in cheese factories. A clear correlation between the total number of moulds in air and mould growth on products is not always found. The conclusion from the investigation is that it is recommended to use a selective medium for cheese spoilage moulds, such as CREAD in hygienic monitoring.

The origin of off-odours in packaged rucola (Eruca sativa).

Rucola (Eruca sativa) was decontaminated and then reinoculated with selected microorganisms. The produce was then stored in three different atmospheres and at two temperatures. The accumulation of off-odours in the packaging headspace was analysed. A dozen compounds were detected by olfactometry but only dimethyl sulphide and dimethyl disulphide were considered to have a strong or moderate intensity. Thus, they were identified as the substances causing an unpleasant smell inside the bags. Inoculation with microorganisms resulted in higher production of off-odours. Samples inoculated with Pseudomonadaceae&Xanthamonadaceae were particularly potent in producing the two sulphides. The off-odour problem was much more prominent in samples that were kept in a packaging material that did not allow gas exchange resulting in oxygen levels below 1%. Higher levels of sulphides were detected at 8°C than at 4°C.

Influence of long-chain polyphosphate and heat treatment on Clostridium cochlearium and Clostridium sporogenes isolated from processed cheese spread.

The outgrowth of Clostridium spp. spores causes spoilage in processed cheese products due to gas and off-odor formation. The present study focuses on the response of spores of Clostridium sporogenes and Clostridium cochlearium at 25 degrees C to polyphosphate, both alone and in combination with heat treatment. The two strains used were isolated from spoiled cheese spread. The addition of 1.5% polyphosphate but not 0.75% polyphosphate totally inhibited the growth of C. sporogenes SIK4.3; in contrast, 0.75% polyphosphate was sufficient to totally inhibit C. cochlearium CCUG 45978. The highest polyphosphate concentration tested (1.5%) was sporicidal for C. sporogenes SIK4.3 but not for C. cochlearium CCUG 45978. When 0.75% polyphosphate Bekaplus FS was combined with a holding time of 5 min at 98 degrees C, no survival or growth of C. sporogenes SIK4.3 was detected; however, the same effect was not achieved through heating alone or through application of polyphosphate alone. C. cochlearium CCUG 45978 was more heat tolerant, as shown by higher D-values. In conclusion, the results strongly suggest that polyphosphate Bekaplus FS has the potential to restrict the growth of C. sporogenes and C. cochlearium in cheese spread stored at ambient storage temperature. Experiments with cheese are needed in order to verify this effect.

Characterization of Clostridium spp. isolated from spoiled processed cheese products.

Of 42 spoiled cheese spread products, 35 were found to harbor Clostridium spp. Typical signs of spoilage were gas production and off-odor. The identity was determined for about half of the isolates (n = 124) by Analytab Products (API), Biolog, the RiboPrinter System, 16S rDNA sequencing, cellular fatty acid analysis, or some combination of these. The majority of isolates were identified as Clostridium sporogenes (in 33% of products), but Clostridium cochlearium (in 12% of products) and Clostridium tyrobutyricum (in 2% of products) were also retrieved. Similarity analysis of the riboprint patterns for 21 isolates resulted in the identification of 10 ribogroups. A high degree of relatedness was observed between isolates of C. sporogenes originating from products produced 3 years apart, indicating a common and, over time, persistent source of infection. The spoilage potential of 11 well-characterized isolates and two culture collection strains was analyzed by inoculating shrimp cheese spread with single cultures and then storing them at 37 degrees C. Tubes inoculated with C. tyrobutyricum did not show any visible signs of growth (e.g., coagulation, discoloration, gas formation) in the cheese spread. After 2 weeks of incubation, tubes inoculated with C. cochlearium or C. sporogenes showed gas-holes, syneresis with separation of coagulated casein and liquid, and a change in color of the cheese. The amount of CO2 produced by C. cochlearium strains was approximately one-third that produced by the majority of C. sporogenes strains. To our knowledge, this is the first study to isolate and identify C. cochlearium as a spoilage organism in cheese spread.

Improvement and validation of RAPD in combination with PFGE analysis of Salmonella enterica ssp. enterica serovar Senftenberg strains isolated from feed mills.

In 1995 and 1996 a Swedish feed mill had problems due to a persistent contamination of Salmonella enterica spp. enterica serovar Senftenberg that was difficult to eliminate. Forty-eight strains isolated from the feed mill, together with unrelated strains included to evaluate the discriminatory power and reproducibility, were analysed by pulsed-field gel electrophoresis (PFGE). The source of contamination in the feed mill was identified and preventative measures were taken, that led to a resolution of the problem. A previously developed randomly amplified polymorphic DNA (RAPD) protocol was used, to evaluate a rapid and low-cost alternative to PFGE typing. The use of the alternative thermostable DNA polymerase Tth was shown to increase the reproducibility of the RAPD analysis. The reproducibility, in terms of Pearson's and Dice's similarity coefficients for duplicate runs, increased from 72.0 +/- 16.9% and 72.3 +/- 12.9% for Taq to 91.6 +/- 7.5% and 90.9 +/- 5.3% for the fingerprints obtained for the RAPD method employing Tth DNA polymerase. Simpson's index of diversity was calculated and found to be 0.580 for RAPD and 0.896 for PFGE. All of the seven RAPD types could be subdivided into one or more PFGE types, whereas none of the 22 PFGE types was divided into more than one RAPD type. RAPD provides a simple, rapid and powerful screening method that can be used to initially select isolates for further analysis by PFGE.

Genetic profiling of Campylobacter jejuni strains from humans infected in Sweden or in Thailand, and from healthy Swedish chickens, studied by pulsed-field gel electrophoresis (PFGE).

The major objective of the study was to explore the genomic diversity between Campylobacter jejuni (C.jejuni) from different sources as a tool for epidemiological considerations. Subtyping was performed using pulsed-field gel electrophoresis (PFGE) and the enzyme used for cleavage was SmaI. Isolates originated from humans infected in Sweden (n=49) and Thailand (n=32) and from healthy Swedish chickens (n=51). Eight PFGE groups were formed in a dendrogram and 48% of the isolates belonged to 1 of these groups. In 2 PFGE groups, strains from humans infected in both Sweden and Thailand were represented. Four of the PFGE groups comprised high frequencies of strains from domestic human infection, as well as from healthy chickens. The PFGE pattern was also compared with the antibiotic resistance pattern in all the above-mentioned isolates. In conclusion, C.jejuni was a diverse group based on PFGE genotyping; about 24% of the clones from Swedish patients and healthy Swedish chickens were similar; and there was no correlation between the antibiotic resistance pattern and the PFGE profiling among the studied strains. Our findings are also in accordance with our hypothesis that there may be similarities between Swedish and Thai strains, which might support a theory of globally occurring C.jejuni.

Comparison of genotyping methods by application to Salmonella livingstone strains associated with an outbreak of human salmonellosis.

During 2000 and 2001, an outbreak of human salmonellosis occurred in Sweden and Norway, caused by Salmonella livingstone. In this study, the genotypic differences between three strains obtained from food sources during the outbreak, two human strains and 27 more or less unrelated strains were analysed, using the three methods; automated ribotyping, pulsed field gel electrophoresis (PFGE) and randomly amplified polymorphic DNA (RAPD). Each method was evaluated regarding its discriminatory ability, reproducibility and typeability. Simpson's discriminatory index calculated for each method was 0.556 for automated ribotyping, 0.766 for PFGE and 0.236 for RAPD. The reproducibility, defined as the minimum similarity between individual replicates in a cluster analysis, was 96% for automated ribotyping and PFGE, and 90% for RAPD. All the strains were typeable with each method. When combining results for the three genotyping methods, it was found that RAPD did not increase the discriminatory index and was therefore excluded from further analysis. Using a combination of the results obtained from ribotyping and PFGE (D=0.855), two strains that had been isolated from feed factories during 1998 were shown to be identical to the outbreak strain, indicating a possible route of contamination due to a clone of Salmonella livingstone persisting in feed producing facilities. No connection to poultry was established.

Inactivation of Escherichia coli, Listeria innocua and Saccharomyces cerevisiae in relation to membrane permeabilization and subsequent leakage of intracellular compounds due to pulsed electric field processing.

Membrane permeabilization, caused by pulsed electric field (PEF) processing of microbial cells, was investigated by measurement of propidium iodide (PI) uptake with flow cytometry. Inactivation of Escherichia coli, Listeria innocua and Saccharomyces cerevisiae was determined by viable counts, and leakage of intracellular compounds, such as ATP and UV-absorbing substances, was measured in the extracellular environment. Electrical field strength and pulse duration influenced membrane permeabilization of all three tested organisms of which S. cerevisiae was the most PEF sensitive, followed by E. coli and L. innocua. It was shown by viable counts, PI uptake and leakage of intracellular compounds that L. innocua was the most resistant. Increased inactivation corresponded to greater numbers of permeabilized cells, which were reflected by increased PI uptake and larger amounts of intracellular compounds leaking from cells. For E. coli and L. innocua, a linear relationship was observed between the number of inactivated cells (determined as CFU) and cells with permeated membranes (determined by PI uptake), with higher number of inactivated cells than permeated cells. Increased leakage of intracellular compounds with increasing treatment severity provided further evidence that cells were permeabilized. For S. cerevisiae, there was higher PI uptake after PEF treatments, although very little or no inactivation was observed. Results suggest that E. coli and L. innocua cells, which took up PI, lost their ability to multiply, whereas cells of S. cerevisiae, which also took up PI, were not necessarily lethally permeabilized.

Growth of pulsed electric field exposed Escherichia coli in relation to inactivation and environmental factors.

Pulsed electric fields (PEF) have been proven to inactivate microorganisms during nonthermal conditions and have the potential to replace thermal processing as a method for food preservation. However, there is a need to understand the recovery and growth of survivors and potentially injured microorganisms following PEF processing. The purpose of this investigation was to study the growth of Escherichia coli at 10 degrees C following exposure to electrical field strengths (15, 22.5 and 30 kV/cm) in relation to inactivation and the amount of potentially sublethally injured cells. One medium was used as both a treatment medium and an incubation medium, to study the influence of environmental factors on the inactivation and the growth of the surviving population. The pH (5.0, 6.0 and 7.0) and water activity (1.00, 0.985 and 0.97) of the medium was varied by adding HCl and glycerol, respectively. Growth was followed continuously by measuring the optical density. The time-to-detection (td) and the maximum specific growth rate (micromax) were calculated from these data. Results showed that the PEF process did not cause any obvious sublethal injury to the E. coli cells. The number of survivors was a consequence of the combination of electrical field strength and environmental factors, with pH being the most prominent. Interestingly, the micromax of subsequent growth was influenced by the applied electrical field strength during the process, with an increased micromax at more intense electrical field strengths. In addition, the micromax was also influenced by the pH and water activity. The td, which could theoretically be considered as an increase in shelf life, was found to depend on a complex correlation between electrical field strength, pH and water activity. That could be explained by the fact that the td is a combination of the number of survivors, the recovery of sublethal injured cells and the growth rate of the survivors.

Quantitative interaction effects of carbon dioxide, sodium chloride, and sodium nitrite on neurotoxin gene expression in nonproteolytic Clostridium botulinum type B.

The effects of carbon dioxide, sodium chloride, and sodium nitrite on type B botulinum neurotoxin (BoNT/B) gene (cntB) expression in nonproteolytic Clostridium botulinum were investigated in a tryptone-peptone-yeast extract (TPY) medium. Various concentrations of these selected food preservatives were studied by using a complete factorial design in order to quantitatively study interaction effects, as well as main effects, on the following responses: lag phase duration (LPD), growth rate, relative cntB expression, and extracellular BoNT/B production. Multiple linear regression was used to set up six statistical models to quantify and predict these responses. All combinations of NaCl and NaNO(2) in the growth medium resulted in a prolonged lag phase duration and in a reduction in the specific growth rate. In contrast, the relative BoNT/B gene expression was unchanged, as determined by the cntB-specific quantitative reverse transcription-PCR method. This was confirmed when we measured the extracellular BoNT/B concentration by an enzyme-linked immunosorbent assay. CO(2) was found to have a major effect on gene expression when the cntB mRNA levels were monitored in the mid-exponential, late exponential, and late stationary growth phases. The expression of cntB relative to the expression of the 16S rRNA gene was stimulated by an elevated CO(2) concentration; the cntB mRNA level was fivefold greater in a 70% CO(2) atmosphere than in a 10% CO(2) atmosphere. These findings were also confirmed when we analyzed the extracellular BoNT/B concentration; we found that the concentrations were 27 ng x ml(-1). unit of optical density(-1) in the 10% CO(2) atmosphere and 126 ng x ml(-1). unit of optical density(-1) in the 70% CO(2) atmosphere.

Relative neurotoxin gene expression in clostridium botulinum type B, determined using quantitative reverse transcription-PCR.

A quantitative reverse transcription-PCR (qRT-PCR) method was developed to monitor the relative expression of the type B botulinum neurotoxin (BoNT/B) gene (cntB) in Clostridium botulinum. The levels of cntB mRNA in five type B strains were accurately monitored by using primers specific for cntB and for the reference gene encoding the 16S rRNA. The patterns and relative expression of cntB were different in the different strains. Except for one of the strains investigated, an increase in cntB expression was observed when the bacteria entered the early stationary growth phase. In the proteolytic strain C. botulinum ATCC 7949, the level of cntB mRNA was four- to fivefold higher than the corresponding levels in the other strains. This was confirmed when we quantified the production of extracellular BoNT/B by an enzyme-linked immunosorbent assay and measured the toxicity of BoNT/B by a mouse bioassay. When the effect of exposure to air on cntB expression was investigated, no decline in the relative expression was observed in spite of an 83% reduction in the viable count based on the initial cell number. Instead, the level of cntB mRNA remained the same. When there was an increase in the sodium nitrite concentration, the bacteria needed a longer adjustment time in the medium before exponential growth occurred. In addition, there was a reduction in the expression of cntB compared to the expression of the 16S rRNA gene at higher sodium nitrite concentrations. This was most obvious in the late exponential growth phase, but at the highest sodium nitrite concentration investigated, 45 ppm, a one- to threefold decline in the cntB mRNA level was observed in all growth phases.

Interlaboratory random amplified polymorphic DNA typing of Yersinia enterocolitica and Y. enterocolitica-like bacteria.

A random amplified polymorphic DNA (RAPD) protocol was developed for interlaboratory use to discriminate food-borne Yersinia enterocolitica O:3 from other serogroups of Y. enterocolitica and from Y. enterocolitica-like species. Factors that were studied regarding the RAPD performance were choice of primers and concentration of PCR reagents (template DNA, MgCl(2), primer and Taq DNA polymerase). A factorial design experiment was performed to identify the optimal concentrations of the PCR reagents. The experiment showed that the concentration of the PCR reagents tested significantly affected the number of distinct RAPD products. The RAPD protocol developed was evaluated regarding its discrimination ability using 70 different Yersinia strains. Cluster analysis of the RAPD patterns obtained revealed three main groups representing (i) Y. pseudotuberculosis, (ii) Y. enterocolitica and (iii) Y. kristensenii, Y. frederiksenii, Y. intermedia and Y. ruckeri. Within the Y. enterocolitica group, the European serovar (O:3) and the North American serovar (O:8) could be clearly separated from each other. All Y. enterocolitica O:3 strains were found in one cluster which could be further divided into two subclusters, representing the geographical origin of the isolates. Thus, one of the subclusters contained Y. enterocolitica O:3 strains originating from Sweden, Finland and Norway, while Danish and English O:3 strains were found in another subcluster together with O:9 and O:5,27 strains. The repeatability (intralaboratory) and reproducibility (interlaboratory) of the RAPD protocol were tested using 15 Yersinia strains representing different RAPD patterns. The intralaboratory and the interlaboratory studies gave similarity coefficients of the same magnitude (generally >70%) for the individual strains. In the present study, it was shown that interreproducible RAPD results could be achieved by appropriate optimisation of the RAPD protocol. Furthermore, the study reflects the heterogeneous genetic diversity of the Y. enterocolitica species.

Prevalence of Clostridium botulinum types B, E and F in faecal samples from Swedish cattle.

Faeces were collected from 60 cows at three slaughterhouses situated in southern and central Sweden. The faecal samples were collected during two sampling periods over the year, summer and winter. All samples were analysed for the presence of Clostridium botulinum spores, according to a combined selection and enrichment PCR procedure. One PCR assay was specific for part of the type B neurotoxin gene, while the other assay was specific for both type E and F neurotoxin genes. The prevalence of C. botulinum in Swedish cattle was established to be 73% for non-proteolytic type B and less than 5% for types E and F. Twenty-eight (64%) of the positive faecal samples had a spore load of less than 4 spores/g. Statistical analysis (ANOVA) showed that seasonal variation (summer and winter) had a significant effect on the prevalence of C. botulinum type B in cattle, whereas the effect of geographical location of rearing of the cattle (southern and central Sweden) was less significant.

Evaluation of selective enrichment PCR procedures for Yersinia enterocolitica.

Four enrichment PCR protocols for detecting unlysed cells of pathogenic Yersinia enterocolitica were studied. First, the probability of detecting Y. enterocolitica cells of known concentrations by a multiplex PCR assay was determined, and it was found to follow a logistic regression model. From this model, the probability of detecting Y enterocolitica at a specific concentration could be estimated; for example, the detection probability of 10(4) CFU/ml was estimated to be 85.4%. The protocols were evaluated on enrichment cultures inoculated with 10(2) CFU/ml Y. enterocolitica and 10(2)-10(6) CFU/ml of a defined background flora. For each protocol, the time for sample withdrawal and the presence of background flora were studied with respect to PCR detection. The optimal point in time of sample withdrawal was found to be different for each protocol employed. Early detection was favoured by concentrating the target cells, and the most rapid PCR detection of Y. enterocolitica was achieved with enrichment in Yersinia-PCR-compatible-enrichment (YPCE) medium for 3 h at 25 degrees C, followed by a centrifugation prior to PCR analysis. For detection of Y. enterocolitica in the presence of high concentrations (10(6) CFU/ml) of background flora, a long incubation time followed by density centrifugation and a dilution step was most successful. The protocol that gave the most reliable PCR detection in the presence of 10(6) CFU/ml background flora included 24 h incubation in Yersinia-selective-enrichment (YSE) broth at 25 degrees C, followed by Percoll density centrifugation, and a 100 times dilution prior to PCR analysis.

Comparison of shelf life of vacuum-packed pork and beef.

The relevance of the intrinsic factors of meat to the sensorial shelf life of vacuum-packed, cold-stored minced pork and beef was investigated. The intrinsic factors studied were the pH and the concentrations of glycogen, glucose, glucose-6-phosphate, l-lactate and fat. The initial bacterial loading was the same on all the meat. High correlations were found between the initial values of pH, fat and l-lactate, respectively, and the rate of spoilage. Using partial least square regression, it was shown that changes in the pH and the concentrations of l-lactate and glucose-6-phosphate during storage were able to explain 68% of the variation observed in the rate of spoilage. No relationship was found between spoilage and the origin of the meat (pork or beef).

Antibacterial Effect of the Lactoperoxidase/Thiocyanate/Hydrogen Peroxide System Against Strains of Campylobacter Isolated from Poultry.

The antibacterial effect of the lactoperoxidase/thiocyanate/hydrogen peroxide system (lactoperoxidase system) was tested against strains of Campylobacter jejuni and Campylobacter coli isolated from poultry. The effect was studied at different pH-values and temperatures in UHT-milk (control); UHT-milk containing lactoperoxidase, sodiumthiocyanate and hydrogen peroxide (lactoperoxidase system); UHT-milk containing lactoperoxidase, sodiumthiocyanate, hydrogen peroxide and sodiumpyrosulfite (inactivated lactoperoxidase system). The lactoperoxidase system had a strong bactericidal effect against C. jejuni and C. coli . The bactericidal effect was more rapid at 37°C compared to 20°C. The effect of lactoperoxidase system decreased with decreasing pH-values. The fastest reduction in viable numbers was obtained at pH 6.6 and 37°C.