BAVARIAN RESEARCH ASSOCIATION PRIONS
LMU 19 a Occurrence and Stability of the BSE Agent in Foodstuff (primarily in Milk and Milk Products) and in the Environment
A scientific-based risk assessment on the occurrence of the BSE agent in and its potential secretion with milk was still missing at the beginning of the second funding period. Generation of topical data is of high political interest on behalf of consumer protection. After having established a rapid and sensitive ELISA during the first funding period and its validation on a broad organ-spectrum of the main domestic ruminant species cattle, sheep and goat, experiments performed during the second funding span focussed on mammary gland tissue and milk samples of respective species. On the one hand investigations aimed to detect expression of cellular prion protein in above mentioned samples by the means of ELISA, Western-Blot and Immunohistochemistry, on the other hand milk samples from experimentally infected BSE and Scrapie sheep were analysed. Results from clinically healthy animals revealed that PrPc expression in the mammary gland and the secretion in milk differed in a species-specific manner. While PrPc in the bovine mammary gland can only be found in actively secreting epithelial cells, protein expression in sheep is present independent from the stage of lactation. Goat mammary tissue revealed an intermediate expression and distribution. PrPc detection in bovine milk failed, whereas the protein is present in ovine and caprine milk fractions. Samples from BSE and Scrapie infected sheep have not been completely investigated up to the delivery date of this report. Nevertheless, three of the Scrapie and one BSE infected animals tested positively in colostrum when ELISA was applied. With exception of one sample results could be confirmed in Western-Blot although no bands of the PrPsc molecular weight were detectable. Visible band pattern resembled the cellular isoform of PrP. Nevertheless, results presented herein show for the first time the presence of PK-resistant non-typical prion protein isoform in milk. A final evaluation of potential infectivity on behalf of consumer protection could only be made in a follow-up project wherein bioassay experiments should be performed. The aim partner 2’s study was to investigate whether bacteria known to occur in foods may be able to produce proteases which can cleave PrPSc at pH 8 and at 30 C. The main focus was therefore laid on potentially proteolytic bacterial isolates applied in cheese manufacturing. Surface-ripened cheese (smear-ripened cheese) is characterized by development of complex microbial consortia on the surface, forming a biofilm-like association during ripening. Although these consortia are extremely heterogeneous, they mostly consist of yeasts and a wider variety of gram-positive bacteria, such as Arthrobacter, Corynebacterium, Brevibacterium, Brachybacterium, Microbacterium, Staphylococcus, and others . Some of these bacteria are known to be highly proteolytic. During ripening, proteases and lipases are secreted from the cells and subsequently diffuse from the surface of the cheese to its interior. Due to enzymatic degradation of protein and fat, they contribute to the desired taste and flavour of cheese. The number of bacteria present on a surface-area can exceed 109 cells/cm2. Identification of proteolytic enzymes able to degrade PrPSc under native (cheese-ripening) conditions would potentially permit development of novel procedures for PrPSc ‘‘decontamination’’. Mild conditions for enzymatic PrPsc degradation are of general interest for food production and processing procedures, and will possibly be useful to minimize the consumers’ risk, if the proteases remain active in different environments. In addition, the cleaning of machinery, instruments, and tools (e. g. surgical instruments) could be another field of application for such proteases [2, 3, 4]. In this study bacterial supernatants were screened to identify PrPSc degrading proteases. Initially, a broad-range screening was undertaken using a wide range of bacteria available from the Weihenstephan Culture Collection. Strains under study were as follows: Actinomycetales (Micrococcineae and Corynebacterineae) and Bacillales (Bacillaceae and Staphylococcaceae). Many of them were originally isolated from cheese surfaces. Additionally, strains of the genus Bacillus were included since bacilli are often present in food, are known protease producers, and ‘‘probiotic’’ Bacillus strains are frequently added to animal feeds . A total of 691 bacterial isolates belonging to 164 species, 12 genera, and 9 families and were included in the screening approach. The following genera were found to be protease secreting: Arthrobacter, Kocuria, Brevibacterium, Brachybacterium, Microbacterium, Curtobacterium, Corynebacterium, Dietzia, Rhodococcus, Bacillus, Staphylococcus and Serratia. Out of nearly 700 tested microbial cultures, six strains were found to secret adequate amounts of proteases sufficient to significantly degrade the prion protein in hamster brain homogenates. In some cases, PrPSc was degraded to a level undetectable by Western-Blot assay. These six strains are: Arthrobacter nicotianae, Bacillus licheniformis, Brachybacterium conglomeratum, Brachybacterium tyrofermentans and Staphylococcus sciuri and Serratia proteamaculans. The protease-producing Serratia identified in our study was originally isolated as a contaminant from a Bacillus culture included in preliminary Azocoll degradation tests. In a second part of the project we purified and characterised the proteases secreted from the six strains. In the screening part we tested the bacterial supernatant. Therefore protein purification was performed on concentrated supernatant from bacterial overnight cultures. In most cases of the protein purification the quantity of the protein present in the supernatant was the limitation for its identification. Proteins were separated according to their properties using several chromatographic columns. Subsequently, different fractions were loaded onto SDS-PAGE. Only the protein bands that were visible after a Coomassie staining could be cut out of the gels for sequencing. In addition there is only limited sequence information available from the database for bacteria like Brachybacterium conglomeratum and Brachybacterium tyrofermentans. We identified two zinc-metalloproteases able to degrade the PrPSc within a Western-Blot degradation assay. Both proteases have not yet been described. However, one protease isolated from the bacterial supernatant of Serratia proteamaculans has 92 % homology to a metalloprotease (Serralysin) of Serratia marcescens. The second metalloprotease was found in the supernatant and on the surface of Brachybacterium conglomeratum. In both cases purified enzymes are able to degrade the PrPSc to a certain degree under mild conditions. These findings seem to indicate that certain bacteria could produce specialised tools to degrade the PrPSc. Third part of the project includes studies on microbial degradation of PrPsc. The influence of a complex microflora residing in the gastrointestinal tract of animals on prion protein plays a crucial role with respect to early TSE pathogenesis and the potential infectivity of faeces resulting in environmental contamination. However, it is unknown whether infectious prion protein, considered to be very stable, is inactivated by microbial processes in the gastrointestinal tract of animals. Feedstuffs consumed by ruminants are initially exposed to microbial fermentation in the rumen prior to gastric and intestinal digestion. In particular polygastric digestion of ruminants represents an efficient system to degrade food proteins by microbial fermentation processes in rumen and colon. In this study, intestinal contents from healthy cattle, calves and pigs taken immediately after slaughter were applied to assess the ability of these microbial consortia to inactivate PrPSc. For that purpose, the consortia were incubated with brain homogenates of scrapie (strain 263K) infected hamsters and BSE infected cattle, respectively. The detection of PrPSc as a surrogate marker for prion infection was accomplished by Western-Blot. Biochemical analyses indicate the ability of complex ruminal and colonic microbiota of cattle to decrease scrapie associated prion protein up to immunochemically undetectable levels under physiological conditions. Similar results were obtained incubating scrapie brain homogenates together with intestinal microbiota of calves and pigs, respectively. Biochemical analyses by immunoblotting have given evidence of scrapie associated prion protein degradation by bovine intestinal microbiota. Comprehensive examinations in which PrPSc was visualized by Western-Blotting after immunolabeling with the monoclonal antibody 3F4 have previously shown a close quantitative correlation between PrPSc amounts and infectivity in the brains of hamsters infected with 263K scrapie agent (Beekes et al., 1996).Thus, in vivo hamster bioassays were performed with degraded samples of scrapie brain homogenates in order to prove concomitance of the loss of anti-prion antibody 3F4 immunoreactivity and inactivation of PrPSc. The results demonstrated significant prion infectivity after degradation of infected hamster brain by the gastrointestinal microflora of cattle. Thus, even in the absence of Western-Blot signals, infectivity is still present. This might be caused by PrPSc at levels below the threshold of immunochemical detection, or by a sub-fraction of infectious prion protein not detectable by immunochemical methods. Finally, the possibility of present infectious molecules or structures other than PrPSc must be considered. In contrast to the results of the scrapie associated prion protein degradation assay, incubation of the bovine gastrointestinal microbiota with BSE associated prion protein did not decrease the PrPSc signal in Western-Blot. This implicates a greater stability of BSE associated prion protein towards microbial degradation processes in the gastrointestinal tract, although that it was shown that BSE agents are less resistant to PK digestion than scrapie (263K) strains (Kuczius and Groschup, 1999). However, the results of the long-term incubations studies indicate that BSE associated prion protein is susceptible to degradation by the bovine gastrointestinal microbiota. Extended incubation time might further reduce PrPSc signal to undetectable levels. Though, the inactivation of BSE associated prion protein by the gastrointestinal microbiota of cattle has to be further validated independently in animal bioassays. In summary, these data suggest that it is impropriate to use PrPSc as surrogate marker for TSE infectivity in inactivation experiments. Conclusively, the use of Western-Blots or immunoassay formats may not provide an indication of the levels of prion inactivation. Therefore, the relationship between the loss of PrPSc signals based on immunodetection in vitro and the loss of prion infectivity in vivo has to be comparatively analysed for each inactivation experiment. Finally, it must be considered that the environment might be contaminated through cattle shedding infected faeces.